As shown above, the average railway congestion rate of the busiest sections of rail in Metropolitan Tokyo, Osaka and Nagoya in the 7th year of the Reiwa Era (2025) reached 143%, 117% and 126% respectively. This represents an increase of a few percentage points over the last several years. What do these numbers all mean? Does 139% congestion look like this:

The shocking 200% congestion rates reported in Japan and South Korea do not necessarily indicate more crowding than a 150% rate for the notoriously congested Guangzhou Metro Line 3; they are simply measured against a much emptier definition of a 'full' train. We need to understand what a “congestion rate” actually measures and why the calculations behind it leaves so much room for interpretation.

How to Calculate Congestion Rates

To start, we have to ask: What is congestion? According to the Cambridge Dictionary, congestion is defined as:

a situation in which a place is too blocked or crowded, causing difficulties

擁擠，阻塞

Extending to urban railways, congestion rates are a measure of how blocked or crowded train cars are, attempting to quantify the difficulties it is causing to passengers. Train car congestion can be rationalized as when passenger demand exceeds the train car’s capacity. I have briefly noted in my review of the new 2024 metro openings for the Beijing Subway on how to calculate line capacity below:

But to recap, the theoretical capacity of a train line is defined by:

The capacity of a line is compared with the actual observed demand to create a ratio between theoretical capacity and actual usage (observed demand):

In Japan, this capacity ratio is known as 混雑率 (congestion rate) while Mainland China uses 满载率 (full load rate). Operators in East Asia usually present train congestion rates as a percentage of rated capacity while North American agencies typically express it as a ratio, called a volume/capacity ratio (V/C ratio) with 1 being “at capacity”, anything over 1 is overcapacity and anything under 1 is below capacity.

The above Beijing Subway article did not cover how planners determine the assumptions about the capacity of each train car (the divisor) that feed into the theoretical capacity equation. In the capacity ratio calculation, the assumptions driving capacity numbers of each train car can greatly change the magnitude of the divisor even if the dividend of observed passenger volumes remain the same. Now that we understand the basic formula for congestion rates, let's examine the variable that causes the most international confusion: how each country defines the “capacity” part of the equation.

Variations of the Capacity Equation Divisor

While I expect there is not much debate about how to count passenger volumes, what the world largely does not agree on is how to define the passenger capacity of a train car. Ask a North American, and a congested train car could be a train that has all seats taken with a bunch of standees. While someone from South Asia would think unless the train has people hanging out the side; it is crowded, not congested. The train might have the same design and size and experience the same demand, but will have different calculated rates of congestion if wildly different assumptions about rated capacities (i.e., what people consider a full train) are used.

The capacity of a train car is the number of people you can fit inside it. How you arrange the passengers inside is determined by the design of the train and assumptions on the passengers comfort expectations. The inside of said train car has two official sources of capacity: seats and floor space for standees.1 One of the biggest factors driving the capacity of a railcar is the ratio of seats and standees. This ratio is quite intuitive and luckily countries around the world largely agree/converge on the suite of methods used to calculate trains with different seating arrangements.

Start with Seating

On one end of the train capacity calculation spectrum are North American commuter operations and global intercity train applications. Here train capacity is usually defined as the number of seats on the train and the train cars are designed to maximize seating. If there are standees in the train, that train is overcapacity. This means in these cases capacity is just number of seats + zero standees. Additionally, seated capacity is also defined to be very discrete; each seat is assumed to hold exactly one person. I know in practice this may not be true in some edge cases. For example, you could have situations where children are doubling up with adults on seats but this is a rounding error compared to how interpretations of other aspects on how capacity is calculated have a greater effect on the result. Again, train car capacity calculations are about providing a sense of train capacity.

Some North American commuter or global intercity train operations may define capacity with an additional amount of standees. The capacity for extra standees can be arbitrarily defined as an additional factor, say 20% of seated capacity on a high speed train in China (which sells standing tickets) like below:

Or some operators use an absolute number (10 additional people per train car) with the assumption that standees will hang out in the vestibule of said intercity or commuter train and that is within design capacity in their eyes.

Add More Standees

In urban/suburban and regional railway contexts, capacity is always the sum of seated capacity and standing capacity. Standing room is the primary source of capacity. However, unlike seated capacity, standing capacity is not a discrete measurement and relies on an assumption of standee density. Standee density is defined as the number of passengers divided by train floor area available for standing. Operators typically express standee density as passengers per square meter. The amount of physical standing area in a train car is affected by the train’s seating arrangements, number of doors and the size of the train. For the same size of train, the more seats you have, the less train floor area you can devote to packing in standees.

You can see that an urban train with longitudinal seating has a lot more train floor area devoted to standees. In that configuration, standees would make up the bulk of rated capacity. To drive the point home, below is the plan layout of different seating configurations for the refurbished JR-E E129 series train with a “liner” variant that can switch between full longitudinal or fully transverse seating. While the regular “commuter” train with mixed seating configuration:

At the extreme end of the capacity spectrum lies a train car design with no seats at all. The so-called “cattle car” configuration.

Bonus: Remove all the Seats

The other end of the extreme from intercity and North American commuter operations is if there are no seats also known as the cattle car. Cattle cars were famously used during Japan’s boom era, deployed on the JR-E inner suburban lines like the Yamanote or Chuo-Sobu Local services. Such seatless designs are very high capacity and full or partial cattle cars have reappeared in Beijing, Guangzhou, Shanghai and Seoul in the face of increasing congestion, in contrast to Osaka and Tokyo where they have largely disappeared. Capacity for such a vehicle is simply the floor area of the train multiplied by the assumed standee density as there is zero seated capacity.

A useful mental experiment is to imagine your favorite metro train car with all its seats removed. You will find that removing all seats leads to a very high per-car passenger capacity. By extension different assumed standee densities will affect the assessment of what is deemed overcapacity for the same train car design. This mirrors how people culturally (North American vs. East Asian crowding) and contextually (urban vs. intercity) have different views on what constitutes a full train. The single most influential number in a train car’s capacity calculation is the assumed standee density. This is where global metrics diverge most sharply.

Standee Density, the Key Determinant of Rated Capacity

Broadly speaking congestion is subjective, as such the benchmarks for it will vary wildly around the world. This subjectivity is reflected in how across the world standee density is assumed with very different density thresholds even if all applications being compared operate in the urban rapid transit context. The high standee ratio usually found on urban and suburban trains leaves capacity calculations depend heavily on what planners assume to be an “acceptable” standee density. Again you can’t assume more than one person can use a seat but you certainly can assume more people can squeeze into one square meter of train floorspace for standees. This changes how capacity numbers are interpreted.

As a result, different standee density standards will lead to wildly different capacity figures for the exact same seating layout and train size. For example, Line 1 Yonge University in Toronto operated at capacity in 2019 with a demand of 28,000-30,000 passengers/hour/direction (pphpd). The Toronto Transit Commission (TTC) assumes the TR series trains used on Line 1 have a per-train capacity of 1,080 passengers assuming a standee density of 4 passengers/m². In fact, if you assume a standee density of 6 passengers/m², as the manufacturer (Bombardier now Alstom) does, the per-train capacity of the TTC TR series holds 1,458 pax.

As an example of how cattle cars vastly increase capacity, using the datasheet for the TR series from the manufacturer’s website above, a “cabless” intermediate B or C car carries 68 seated passengers and 182 standees. So each intermediate TR series car with a mixed transverse and longitudinal seating layout holds 250 passengers assuming 6 passengers/m² of standee density.

If one intermediate “cabless” TR series car were a cattle car, the capacity is just the internal floor area multiplied by an assumed standee density. Assuming 6 passengers/m² of standee density such a cattle car would hold over 400 standees. Even assuming a looser 4 passengers/m² such a hypothetical car would still hold over 270 standees. You can see:

1. The power of cattle cars but you are sacrificing comfort for capacity.

2. How the standee density assumption greatly changes the rated capacity of a train car even if the design doesn’t change.

So what is a Full Train?

This subjectivity of how you define a train as full fuels this misconception floating around about East Asian train congestion statistics. For example, there have been situations where South Korea and Japan report rush hour trains running at close to 200% capacity, captivating minds in Western popular culture. The vibes are generally not incorrect, those Oshiya were really required at major congested Japanese train lines in boom era Japan, particularly in greater Osaka and Tokyo. Seoul’s infamous Line 9 “Hell Line” (지옥철; ji-ok-cheol) as described by S.Y. Lee on their S(ubstack)-Bahn substack was still extremely congested based on my personal experience in 2016.

As noted above throughout history South Korean and Japanese urban railways reporting +200% crowding numbers fueled their reputation for overcrowded trains. China is also known for congested metro trains but really only report crowding statistics ranging from 100% to 150%.2 Korea may have a hell line but China has the notoriously congested “Hell West Road” Station (地狱西路; Di-yu Xi-lu; Dei-yok Sai-lo) or Tiyu Xilu Station. This Guangzhou Metro station is serving the most congested metro line on Earth, Line 3. However, Line 3 is not officially operating at 200% capacity:

This raises an important question: some context is needed behind the meaning of these numbers from Japan and South Korea. Both countries use surprising assumptions on what a full train standee density is. This is in contrast with China which assumes crowding in a very different way. Below are examples of Korean and Japanese metro system crowding metrics used today:

What is interesting is how qualitative the metrics are for both Korea and Japan. This is likely due to how crowding data is commonly collected, via manual visual inspection on site.

Korean Definition

Starting with Korea, the metrics above note a South Korean train deemed at 100% capacity still has quite a bit of room to move around. 100% is defined in S Korea with three rows of standees between the longitudinal seats and doors, 2 standees around each door. Seoul’s metro trains are quite wide by international standards being over 3.1m wide so with my eyeball math, three rows of standees between the two longitudinal seats works out to be about 3 passengers/m² of standee density. Each standard Seoul Metro train car has four 1.3 meter wide double stream doors per side of each car and uses all longitudinal seating. With these assumptions in mind, each standard 20m long and 3.1m wide Seoul Metro car assumes about 160 passengers as 100% “at capacity”.

According to the Seoul Metro peak loadings on Line 4 and 7 reaches 193% and 164% capacity respectively. Traditionally, the congested line in Seoul are Line 9 on the south side of the Han River which operated over 200% but recent improvements on Line 9 I suspect has reduced crowding.

Japanese Definition

Shifting focus to Japan, the MLIT metrics are even more qualitative in nature, determining crowding based on visual descriptions like:

100% — Commuters have enough personal space and are able to take a seat or stand while holding onto the straps or hand rails.

150% — Commuters have enough personal space to read a newspaper.

180% — Commuters must fold newspapers to read.

200% — Commuters are pressed against each other in each compartment but can still read small magazines.

250% — Commuters are pressed against each other, unable to move.

An older expanded MLIT train crowding standard from over 20 years ago included a bonus 300% reference where “[Commuters] are near their limits and there is physical risk.”:

This is a holdover threshold from Japanese Bubble Economy Era congestion where some trains in the Kansai and Kanto area did reach these extreme levels of congestion, particularly parts of the Kanto JNR regional system:

Declining ridership caused by the Lost Decades of the Japanese economy in addition to gradual rail service improvements combined with changing consumer tastes (No one reads a newspaper on the train these days) prompted MLIT to revise their 20 year old train qualitative crowding metrics in August 2025. The new standard removes references to paper reading materials and redefines the higher levels of crowding.

Which are as follows:

100% — Train at capacity (all seats, straphangers, and handholds near the door taken)

150% — Standee’s shoulders don’t touch, crowding around the doors

180% — Standee’s shoulders touch and there’s slight pressure. People near the door feel cramped and have difficulty turning around

200% — Bodies are touching and there is a great sense of pressure. People near the door are unable to move.

Note that MLIT has removed descriptions for 250% crowding as such extreme crowding does not exist in Japan anymore in the Reiwa Era (more on this in a future substack so Like, Share and Subscribe!). Such crowding lives on in the Guangzhou Metro today. It also further reduced how crowded a “full train” is. 100% capacity is now “all seats, straphangers, and handholds near the door taken” which still is a very empty train as there is a lot of standing space not near the doors. I guess this is a more pragmatic human centered approach that the Japanese have now adopted to measure crowding. It also reflects how train crowding has been changing in Japan (read: decreasing). Japan knows that everyone has experienced seeing a train with everyone packed around doors but you can see lots of space to stand in the area between the doors. However you would still think:

As a side note, in addition to qualitative measures, Japan, like many other countries also has a more quantitative method to quickly calculate capacity. Assuming a certain seating configuration, a general factor averaging out a set standee density, with the average amount of space one seat takes up in a train is developed. This general factor would be used in an equation below:

The MLIT's Annual Report of Railway Statistics (鉄道統計年報) and Urban Railway Congestion Survey normalize crowding using this method, with a general factor of 0.35 meters of train floor space per person. Train floorspace is different from standee space and is just the internal floor area of the traincar. In addition, the factor is counting both seated and standing passengers. Essentially in this method, the general factor assumes the seating layout of the train car to be using fully longitudinal seats and a door density of four dual stream doors per side of a 20m Japanese or three doors per side for shorter 17-18 m cars. The method then works out an average allotted space per passenger regardless if they are sitting or standing. Notice the inverse of this general factor is 2.9 passengers/m² of train floor area. Rounding this inverse to account for the fact that a person in a seat takes up more space, you get an implied standee density in this method for a “full” train to be just over 3.25 passengers/m².

As you can see regardless of methods used for calculation, Japan and S Korea both in their own way essentially assume 100% capacity is an urban/suburban train with all seats taken and standees standing at just above 3 passengers/m² of density. My experience with both countries confirms this. Korean trains are increasingly more crowded than Japanese trains but you can still walk around a train defined as at 100% capacity using either Japanese or Korean crowding metrics.

Crowding Definitions around the World

Outside of Japan and Korea, the classic standard for planning was 6 passengers/m² of standee density is 100% capacity. Widely used around the world in the 1970s. Today outside of East Asia the maximum capacity standard is converging on 4 passengers/m² of standee density as shown by the Toronto TR Series example above.

This organic shift is logical, I reviewed the New York’s Hub Bound Report data on peak crowding into Manhattan in 2016-2017. Based on my knowledge of the North American subway commuter zeitgeist, sections of the the New York Subway’s Lexington, Broadway/7th Ave and L Train services were always well known to be the network’s most congested during peak periods. The report data supports my impressions. In 2016, before the opening of the 2nd Ave Subway Phase 1, the Lexington, Broadway/7th Ave Express services and the 14th Street Tunnel (L Train) are the only services providing on average less than 3.8 square feet of train floor space per commuter during the peak hour.

In non-freedom units, 3.8 square feet of train floor space/passenger also matches Japan’s general factor of 0.35 meters of train floor space per person. The inverse of that yields an average density of over 3 passengers/m² of train floor area. Which when adjusted for seated passengers the standee density approaches 4 passengers/m². Many of MTA’s recent high priority capital projects in the 2010s are geared to relieving crowding on these lines, which regularly exceeded 4 passengers/m². Such as:

• L Train signalling and electrical upgrades for more frequent service.

• 2nd Ave Subway Phase 1 and 2 which would provide relief to both of the North Manhattan A Division corridors.

These projects were consistently worked on and supported despite MTA’s project delivery incompetence and high costs. This shows that when crowding approaches the threshold of 4 passengers/m² of standee density, the public starts putting pressure on pols. This creates political capital to prioritize and sustain improvement projects to fix these crowding issues.

In Toronto, where the Yonge Line between Eglinton and Bloor Stations was commonly noted to be congested but political football delayed the “Downtown Relief Line” project designed to remedy this. Since the late 2000s demand has slightly exceeded the scheduled capacity of the Yonge Line. As noted above, the TTC assumes the capacity of the Yonge Line’s fleet of TR series’ using a maximum standee density of 4 passengers/m². As approaches crowding 4 passengers/m², passengers are already not entering trains at some sections due to congestion, instead opting to wait for a less crowded train.

This congestion finally reached a point where a relief line project received enough support from the general public and the powers that be, was funded, and broke ground in 2019. Known as the Ontario Line; the half underground downscaled automated metro has had delivery problems to become one of the most expensive metro projects on Earth to construct. These two examples confirm that at least in North America, 4 passengers/m² is the maximum standee density that engineers should design for.

Hopping overseas, this shift to 4 passengers/m² maximum capacity is epitomized in Hong Kong. Hong Kong today has a two tiered threshold straddling old and new. With a suggested 4 passengers/m² as “full” and the older 6 passengers/m² as an ultimate but non-sustainable crush load. This was done after observations by the MTR Corporation noted how unrealistic 6 passengers/m² is to achieve in practice these days. Passengers now simply crowd platforms as they hope for a less crowded next train. People, particularly in the developed world, are not as motivated to squeeze themselves onto a train these days. As the MTR Corporation put it to the HKSAR Government in April 2016:

I’m sure parallels of this mentality can now be found in commuters in all other parts of the world leading to this convergence. While North America, Europe, and Hong Kong are converging on 4 passengers/m² as the practical maximum, China officially holds onto a much denser standard.

The China Syndrome

In contrast, China takes a different approach, as I noted in one of my first Substack posts, China’s metro design standards, GB 50157-2013 “Code for Design of Metro” (地铁设计规范), assumes a standee density of 6 passengers/m² as “at capacity”. While a crush loaded train is assumed to have a standee density of 9 passengers/m². See footnote at the bottom of table I posted below:

China to this day officially designs to this older standee density standard that is out of step with Chinese (and broader global) commuter preference. Some cities in China have independently planned and designed lines assuming a lower 5 passengers/m² for “passenger comfort”, after observing underbuilt disasters in Guangzhou and Beijing from afar. Those disasters made planners realize that 6 passengers/m² of standee density is not really a good target from a service experience perspective and you should not optimize train size to such a high level of crowding. This is why some smaller less dense Tier 2 cities like Shijiazhuang, Urumqi, Taiyuan or Zhengzhou have designed and built lines with higher capacity Type A trains.

Putting it All Together

Right out of the gate you can see how these standee density assumptions affect the calculated capacity and crowding rates, and then ultimately shape train design and transportation policy decisions. We have now surveyed the major global standards: 3 ppsm (Japan/Korea), 4 ppsm (emerging Western norm), and 6 ppsm (China’s official baseline). Let’s put these numbers side by side to see their real-world impact.

If measured using Japanese standards, an officially 100% full Chinese metro train would rate at just under 200% capacity in Japan or South Korea (which assume 3 passengers/m²) or 150% in Europe and North America (converging on a 4 passengers/m² assumption). A crush loaded Chinese train with 9 passengers/m² of standee density is rated to be 150% capacity in China but would be classed as the mythical 300% capacity in Japan. In China the closest their metro trains reached 300% capacity (if using the Japanese methodology of measuring crowding) is Guangzhou Metro Line 3 and 5 as well as Beijing Line 13 in the early 2010s:

I have been presenting many abstract concepts in this post. To understand what these abstract percentages mean for your morning commute, let's look at the literal floor space beneath your feet. To help you visualize, I marked out a 1 m² space in my flat for reference. I then put in some instant noodles and an 8 pack of regular 330mL coke cans I have on hand for scale:

3 passengers/m²: The Japanese and Korean Baseline

I place three pairs of shoes varying from US size 7 to 9 in my 1 m² box. If you are ever standing in a train and want to know what is the crowding level look down and compare with the series of images below. If you see feet with about this spacing during your standing train ride, this is 3 passengers/m² of standee density. Japan and South Korea would consider a train like this with all seats taken as operating close to 100% capacity.

Before you judge, look at the image above again. Remember that a human's torso, chest, and arms are much wider than their literal footprint. In addition people here could be wearing coats, carrying bags, not taking their backpacks off, watching shorts on their phone etc. all of this needs to fit inside the marked out box. Just imagine you are the person wearing the orange and black shoes at the bottom of the image. The door is at the top, and you are trying to exit the train. This can help you conceptualize what it feels like moving around in a train deemed by the Japanese and South Koreans to be just under 100% capacity feels like. This density would be considered ~75% capacity in North American Subways and ~50% capacity in China. Regardless, at this level, you can still move quite freely.

4 passengers/m²: The Emerging International Standard

I add one more “passenger” and below is now 4 passengers/m² of standee density. This is the average density the busiest section of TTC Subway Yonge Line was operating at during the AM Peak in 2018 and the crowding you should expect when taking any Hong Kong MTR line during rush hour today. It is also the level of crowding that starts making the news in North America as a problem to fix.

Again, visualize that the hypothetical people filling those shoes are carrying bags and not removing their backpacks. You, at the bottom, need to exit through the train doors at the top of the image. Imagine moving around the train with this level of crowding. In this situation I think you would most likely require a lot of side stepping and “‘scuse me’s”. This is the converging international standard for standee density for an urban/suburban train being at 100% capacity, used in North America and Europe. In Japan and South Korea this would be about 120-135% capacity and China would deem it to be under 70% capacity.

6 passengers/m²: The Chinese Standard

I add two more “passengers”. This is the standee density for the old definition of 100% capacity used around the world and still officially used in China today, 6 passengers/m² of standee density. Hong Kong and some train manufacturers may call it “crush loaded” density as a higher tier of capacity on top of the maximum capacity threshold where the more modern 4 passengers/m² of standee density is used.

You will have physical contact with all adjacent passengers at all times. If you need to exit the train in the above environment, you will be squeezing between passengers and maybe doing a little pushing. This would be 125% capacity in North America or a bit more crowded than the busiest sections of the Lexington Line in the New York City Subway during the AM peak before the opening of the 2nd Ave Subway Phase 1. By Japanese and Korean crowding standards this is close to 200% capacity.

This is the level of crowding that prompted Seoul Metro to consider putting some cattle cars on Seoul Metro Line 4. This is also the average level of crowding you see on the most crowded sections of train lines in Japan before the ridership collapse following the COVID-19 pandemic. Before that period 2017-2018 eleven segments on the Tokyo area during the AM peak regularly approached 200% capacity according to the MLIT. Including greatest hits like the Tokyo Metro Tozai Line, Tokyu Den-en Toshi Line, JR-E Chuo Rapid, Tokaido, Chuo-Sobu Local and Keihin-Tohoku Line.

Following the analysis I presented in the beginning of the post, 180%-200% crowding reported by the MLIT equals about 6-6.25 passengers/m² of standee density.

9 passengers/m²: Crush Load

I add three more “passengers”, making it 9 passengers/m². This is the old definition of “crush loaded” and basically is not thought about anymore in the Western World. You need really motivated commuters and/or pushers a la Japan to consistently pack trains to this average.

I personally experienced this on Guangzhou Metro Line 3 before relief lines opened in the 2020s, and on Seoul Metro Line 9 before service adjustments increased capacity. It was a deeply dehumanizing experience; you can not move. Your arms are pinned to your sides. Every time your chest expands to inhale, your chest pushes out and presses against the six other people around you. When the six people surrounding your body inhale, their chests push against you at different times because total strangers packed in a metro car are obviously not synchronizing their breathing with each other. When you want to leave the train and you are not right at the door there will be a lot of pushing. Sometimes there will be shouting. You will be stepping on people’s feet but people will step on your feet with greater frequency. Would not recommend.

Interpreting Congestion Rates in Practice

Going back to the start of article where Japan’s MLIT has published national train crowding data for 2025. When you read that Tokyo’s average train congestion rate is 143%. Just remember, despite all the technical jargon surrounding this metric, that figure is not an objective physical measurement but a bureaucratic baseline. The real story lies in what the local transit agency decides to call a ‘full’ train. The next time you see a congestion rate, remember that the number tells you as much about the country's standards as it does about your commute. The real question is not “how crowded is the train?” but “what does your country consider acceptable?”.

Now that you have that perspective, let’s compare the MLIT report’s crowding numbers with other places around the world in the next post coming soon so…

Thanks for reading this global analysis of train crowding. Please like, share, and subscribe for more news and analysis of global railway operations.

Or drop me a Ko-fi if you like what you read!

The area is famous for their lavish and bombastic Chinese New Year celebrations with the hero’s dance parade; Yingge (Mandarin), Ying Goh (Cantonese) or Eng-ko (Teochew Min-nan) and Parade of Deities: Youshen (Mandarin), Jau San (Cantonese) or Iu-Sing (Teochew Min-nan), where each village carries their deity around in sedan chairs for worship. Food-wise, Chaoshan is known for its fragrant, aromatic beef, particularly made into beef balls and served in hot pot. The area is also known for a specific series of seafood, offal, soybean-based and goose cold dishes served with a side of thin congee. In Hong Kong, this style of cuisine is known as da laang (打冷) and is commonly eaten as a late-night meal or snack, called siu ye (宵夜).

A fascinating observation one of my relatives made was that when Chaoshan immigrants arrived in Hong Kong to escape the turmoil of Mainland China, unlike the Shanghainese migrants which arrived to Hong Kong with some affluence, Chaoshan immigrants’ cuisine and bland, thin congee was viewed as low-end food in Hong Kong. One can reductively rationalize that thin, watery congee is a response to a history of poverty. Over time, thanks to Li Ka-shing’s iconic life story and other very successful Hong Kong tycoons who came from Chaoshan, Chaoshan cuisine in Hong Kong now has a high-end edge. Prestigious versions of Chaozhou congee and cold dishes are now commonly served in HK banquet halls, with a more flavorful broth, living proof of the “hierarchy of taste” theory in action.

The local Teochew dialect of the Min-nan variety of Chinese is mostly mutually unintelligible with Yue (Cantonese) and Putonghua/Guoyu (Mandarin) varieties and somewhat intelligible with Taiwanese “Taigi” and traditional Min prestige dialects spoken around Xiamen, Quanzhou and Fuzhou. But enough about the linguistic and culinary landscape, you are here for the trains. So are there any? Sadly, the Chaoshan area has been truly forgotten by Guangdong Province, which pours investment into the Cantonese/Hakka-dominated regions of the Greater Bay Area.

The Shantou-Chaozhou-Jieyang area has a population exceeding 10 million people, a rich history, but no rail transit system. It is, with nearby Quanzhou, one of the most populated urban areas within China not served by any urban rail. Part of the reason is the relatively dispersed nature of the metropolitan area. The population is split between three prefectural cities, which hide the true size of the urban area in the eyes of a central government organ who would be approving a metro proposal. The functional, contiguously built-up urban areas of the Shantou-Chaozhou-Jieyang area IMHO consist of:

• The Shantou core, located on the lower estuary of the Shantou River consisting of the central urban areas of Shantou Prefectural City such as Longhu, Jinping, Haojiang and Chenghai Districts and parts of what was formerly Fengxi District of neighbouring Chaozhou Prefectural City. Shantou’s urban area serves as one of the two cultural centers as well as the economic heart of the region. This urban area has just over 2.8 million people. Denoted below in orange.

• The Chaozhou core, the other cultural hub, has 1.5 million people in a built-up area on the Hanjiang River system which consists of Xiangqiao and Chao’an Districts of Chaozhou Prefectural City. Denoted below in yellow.

• The Jieyang core has 1.5 million people in the built-up urban core of Jieyang Prefectural City which consists of Rongcheng and Jiedong Districts on the upper sections of the Shantou River. Denoted below in fuchsia.

• The Ling River (Jiang) corridor consisting of dispersed but dense peri-urban developments of Shantou Prefectural City Chaonan and Chaoyang Districts and the adjacent Puning, a sub-prefecture “satellite city”, of Jieyang Prefectural City. The total population of this contiguously built up corridor is almost 5 million people but lacks a well defined center and is loosely connected. Puning is a center but (to me) it seems its social-economic connections with the Shantou areas in the south are quite tenuous even when quite physically contiguous and proximal. Denoted below in cyan.

In other words, the total population spread across these three distinct cores is comparable to the combined populations of Vancouver, Calgary, Edmonton and Montreal, all packed into three adjacent valley deltas surrounded by mountains. A Chinese Rhine-Ruhr with no local rail transit.

The Past: First Contact

The first time I arrived in Shantou was to support my relatives participating in the 2023 Shantou Marathon. Naturally we all took HSR to Shantou as there was a direct service that took us from Hong Kong city center (Kowloon West) to Shantou’s city center railway station in just over 3 hours.

Our train from Hong Kong took the XRL to Shenzhen North and did a reversing movement onto the Shenzhen-Xiamen Railway, a Class I 250kph mixed freight and passenger HSL (High Speed Line) that is part of the broader coastal HSR corridor up to Ningbo. This line is one of the busiest and most congested HSLs in China. Passenger services have become so dense that almost all freight traffic has been pushed out.

The Shenzhen-Xiamen Railway is one of the few lines to have its speed restored to design speeds after the 2011 HSR crash in Wenzhou led to a performative nationwide downgrade in speeds. 350/380 kph lines were downgraded to 300 kph operations, 250 kph lines to 200 kph and 200 kph upgraded lines reduced to 160 kph. Due to congestion and strong financial performance, the Shenzhen-Xiamen Railway was restored to its designed 250 kph service speeds on 10 April 2021, one of the first HSLs in China to do so.

A medium term remedy to congestion on this corridor is a 350kph rated HSL being constructed to parallel the Shenzhen-Xiamen Railway. The section of interest being constructed is the 350kph Guangzhou-Shantou HSR that bypasses the existing congested 250kph mixed traffic Shenzhen-Xiamen Railway. This will hopefully one day allow the Shenzhen-Xiamen Railway to go back to hosting some freight services. The branch lines of the Guangzhou-Shantou HSR will allow trains to run directly between the city centers of Guangzhou or Shenzhen or Hong Kong to Shantou and beyond to Fujian Province.

The Shenzhen-Xiamen Railway’s alignment prioritizes overall speed and directly serves neither Chaozhou nor Shantou city centers. Instead the line runs straight through the Chaoshan region on what was at the time a strip of countryside between the two cities. An ex-urban compromise station was built roughly between the two cities and was the traditional way for both cities to access the HSR services on the Shenzhen-Xiamen Line.

In August 2018, new connection tracks were built to allow services from the east west Shenzhen-Xiamen HSR to connect to the legacy north south single track Guangzhou-Meizhou-Shantou Railway. This allowed HSR trains to swoop down south and directly reach Shantou’s “legacy” city center station.

Using these connection tracks, our HSR service switched off the Shenzhen-Xiamen HSL and onto the said conventional Guangzhou-Meizhou-Shantou Railway, which had been modernized on an elevated, double-tracked alignment. This resulted in the last leg of my HSR journey into Shantou passing densely packed urban villages running up against the Guangzhou-Meizhou-Shantou Railway viaduct I’m on, giving a very Japanese feeling.

After weaving between packed mid-rises as if we were taking the Shinkansen between Shinagawa and Ueno, we entered and terminated at Shantou’s legacy stub-ended station, the current terminus of the Guangzhou-Meizhou-Shantou Railway. While most people would think that a legacy station should be very close to the city center, it is important to note that China’s railway system is relatively new. Even before the high speed era, many cities in China did not have any railway access for most of their history.

Not that Shantou was not trying, throughout the 19th century British Hong Kong based Jardine Matheson (now known as Jardines) and Butterfield & Swire (now known as Swire Group) was interested in building railway to support Shantou port but was never able to get the approval from the Qing Government. The Qing was skeptical of railways and the ulterior motives of foreign involvement in such projects. This skepticism was not entirely unfounded. Shantou eventually did build a railway late 19th century with the completion of the ChaoChow–Swatow Railway built with funding and support from SE Asian Overseas Chinese that trace their ancestry from Chaozhou. The project was spearheaded by a Hakka SE Asian Businessman Tjong A Fie and his brother who both grew up in Chaozhou.

The railway was dismantled by retreating Kuomintang forces in the 1930s to prevent use by the invading Imperial Japanese Army during the Second World War. Decades passed until Shantou finally regained conventional railway service with the completion of the Guangzhou-Meizhou-Shantou Railway and Shantou Station…. in 1995. Despite the 50 years of turmoil China underwent after the demolition of the ChaoChow–Swatow Railway, Shantou’s urban area grew. This new “legacy” station was not as close to the city center as one might expect.

Shantou Railway Station is a stub ended station where passenger services need to terminate or reverse in and out of. Hence most HSR services on the Shenzhen-Xiamen HSL that run through the Chaoshan Region just stop at the compromise Chaoshan Station and continue blasting through the Chaoshan Region to the Pearl River Delta or Fujian Province. Guangzhou-Meizhou-Shantou Railway does continue south as a short spur to Shantou’s container port.

When we finally arrived at the Shantou Station, we got off and saw a massive under construction structure looming behind the station platforms. This structure is Shantou’s new revitalized station, built to accept the new Shantou-Guangzhou HSL, which detours away from the Shenzhen-Xiamen Railway to directly serve the city center Shantou Station via a coastal alignment and an undersea tunnel.

Unlike the existing through services of the Shenzhen-Xiamen Railway that streaks across the north of Shantou or detour off via Guangzhou-Meizhou-Shantou Railway to directly serve Shantou Station, the Guangzhou-Shantou HSL will instead come from the south via a new tunnel under the Shantou Estuary. This will bring 350kph HSR service directly to Shantou’s city center. In the future these faster 350kph trains will also not need to dead end and terminate to reach Shantou station. The new HSL will instead allow trains to run through onto another under construction high speed line, the Shantou-Xiamen HSL, which is a continuation of a 350kph class HSL paralleling the older 250kph mixed traffic line. The completion of the Shantou-Xiamen HSL will connect the Guangzhou-Shantou HSL with the 350kph Xiamen-Fuzhou HSL, providing full corridor wide relief and redundancy to the older 250kph Shenzhen-Xiamen-Fuzhou Line that also allows all through running trains to serve Shantou city center.

We proceeded to exit the packed station as passengers from our terminating 16-car (two 8-car sets coupled) train queued behind another crowd that disembarked from a different 8+8 car train that terminated a few minutes before ours. As I exited the station, I found out that the taxi pickup location was a 10 minute walk away in a parking lot. Most likely a complication from the station’s current renovation activities.

We took a taxi to our hotel, dropped off our luggage, and then took a bus into Shantou city center. As we got stuck in traffic, we passed by the highlight of my trip: station construction sites for the future regional railway system serving the Shantou-Chaozhou-Jieyang area. Shantou, the largest city in the area, has not been able to secure approval for construction of a metro/light metro from the central government. However the prefectural cities of Shantou, Chaozhou and Jieyang as well as the Guangdong provincial government did get started on an urban-regional rail system with city-centre tunnels built to national railway standards with all-stop and express trains reaching up to 160 kph. When this system opens, instead of being stuck in traffic in a taxi or crowded bus we would be on a regional train running underground from Shantou Railway station to the city center.

And boy… Shantou has a lot of traffic. This is what happens when a 10 million population urban area has no higher order transit. However, this regional railway system, called the East Guangdong Regional Railway, is the first step to improving railway access in the Shantou-Chaozhou-Jieyang area. The first phase of the system started construction in 2021 and is 140km long with 30 stations. The system runs mostly elevated between the three major centers of the Shantou-Chaozhou-Jieyang area in a "ring+line" formation but is tunnelled under each of the cities’ centers like some Chinese S-bahn. The system also will connect the region's airport and various high speed railway stations.

In Shantou city center the core of the regional railway system runs underground right under the busiest parts of the city along Zhongshan Road. The next day I walked along Zhongshan Ave to see the construction progress. I saw lots of cut and cover sites constructing the station boxes.

This core city center tunnel serves the dense areas of Shantou which are characterized by mid-rise commieblocks and urban villages. The map below shows the typical built form of Shantou’s city center. The area is characterized by typical 70s era mid-rise commieblocks (purple and yellow) with urban villages stuffed in between. The urban villages consist of mid rise handshake buildings (green, orange and blue) packed wall to wall on all four sides:

Below is the street level of the various urban villages in Shantou city center, packed with 4 to 10 story “handshake buildings” named after their extremely dense urban form where “you can reach out the window of your flat and shake the hand of your neighbour in the flat of the adjacent building”.

A lot of old Deng-era commieblocks in Shantou are beginning to be retrofitted with external elevators and had their lower floors re‑purposed from residential flats into commercial and retail spaces.

Shantou is a city of rich history, as expected given the vast overseas population that calls it their ancestral home. The railway station area is in the west but the old historical area of Shantou is in the east, filled with Chinese Republican-era Tonglau buildings now renovated to be a tourist zone. The regional railway tunnel connects both areas as it runs lengthwise along the southern coastal areas of urban Shantou.

During my stay of a few days in Shantou, I was notified that a friend also happened to be visiting the area but only in Chaozhou. Since I was only visiting Shantou, the timing and logistics didn’t really work out. She arrived in Chaozhou the morning of the day I was leaving. I didn’t want to just take a taxi over to Chaozhou, spend two hours, and then rush back to Shantou Station to catch my train to Guangzhou. Also I discovered the last intercity coach for the day from Shantou to Chaozhou departed at 2PM which is insane. If only there were a frequent regional railway that allowed me to easily get between the two cities and any railway station in the area.

Instead, that day I decided to arrive to Shantou Station a few hours earlier than my train’s departure and then used the time to explore the mix of vertical factories, urban villages and residential compounds surrounding the station. The vertical factories are interesting as they have a very Hong Kong/Taiwanese design aesthetic to them with some colour added to the buildings, reflecting the traditional sources of investment in the area.

I headed into the station, where the old structure’s dated design was showing. As a pedestrian I had to access the station using some well-maintained but dreary passages that looked shoehorned into the vast car infrastructure supporting the old station. I entered a well maintained concrete maze that was admittedly well signed and barriered, but the wayfinding to guide me (as a pedestrian) was only in Chinese. It was jarring to be walking through a series of parking structures and road tunnels with luggage. If not for the signage, I felt like I should not belong in these areas.

So I left Shantou for Guangzhou via the best way to travel, Chinese HSR. I got some last minute pictures of old Shantou station for posterity. It and the associated concrete passages I used will be completely demolished next year to make way for the second phase of the new station. The new modern station will have better pedestrian connections to the surrounding areas via elevated plazas as shown by a poster in the old station hall proudly showing a better future is in the works.

I arrived on the platform to board my train where once again the construction site loomed over the old one.

As I headed north on the Guangzhou–Meizhou–Shantou railway to get back to the Shenzhen-Xiamen HSL, I caught glimpses of the under construction at-grade sections of the East Guangdong Regional Railway tracks/ road bed which parallels sections of the Guangzhou-Meizhou-Shantou Railway as the eastern ring of the regional railway system makes its way to Chaozhou.

When leaving for Guangzhou I was hoping my train would use some sections of the Guangzhou-Shantou HSL that parallels the Shenzhen-Xiamen HSL, which I was expecting would open in late 2023. This never happened. So instead we took the traditional route of getting to Guangzhou from Shantou via Shenzhen on the slower Shenzhen-Xiamen railway. As a result I just got views of worktrains preparing for the opening of the Guangzhou-Shantou HSL at stations like Shanwei where the services of the two parallel lines can interline with each other.

The next day I found out I was literally a couple hours too early; parts of the Guangzhou-Shantou HSL would open the next day while I was in Guangzhou.

In Guangzhou I reviewed their year end 2023 metro openings and the opening of a new railway station.

The Present: Redux Feat. Chaozhou

I went with a different relative to the Chaoshan Area in early 2026, so slightly more than two years later. This time going to Chaozhou was firmly planned into the itinerary. Once again I traveled by high speed train from Hong Kong to get to the Chaoshan area. This time with the opening of the new Guangzhou-Shantou Line travel times between Hong Kong and Shantou have been reduced from 3 hours to just over 2 hours and the capacity killing reversing movement at Shenzhen North was removed. Instead with the opening of the new line, we ran through Shenzhen North Station and switched to the Shenzhen-Nanchang HSL up to Huizhou then switched to the newly completed Guangzhou-Shantou HSL via some connection tracks.

I now blasted at 350kph across coastal Guangdong on the Guangzhou-Shantou HSL. Sometimes I saw the older 250kph Shenzhen-Xiamen line (with associated regional services running on it) at key stations as the two lines dance and weave around each other flying over the fish ponds and rice paddies between Shenzhen and Shantou.

I arrived at Huilai Railway Station, a new station on the Guangzhou-Shantou HSL, and saw the platform and branch line of the under construction Jiehui Railway. This line is not part of the “official” East Guangdong Regional Railway program and is effectively a single track freight railway spur to a petrochemical plant from regular railways in Jieyang City. However, now the line is being built as an electrified single track mixed passenger freight railway with passenger stations, becoming an informal extension of the East Guangdong Regional System.

This is an interesting and uncommon setup in China and allows the East Guangdong Regional service to be extended from Jieyang to Puning and Huilai HSR stations. I hope to see positive results, as light mixed passenger and freight lines are quite rare in China but hold huge promise in allowing marginal “shortline” or “branch line” railway projects to be viable.

In the shot below you can see my HST fly over the Jiehui Railway. The passenger branch that will terminate at Huilai Railway Station curves to the left, while the freight branch to the petrochemical plant curves to the right.

Just before I plunged under the Donggou Mountains and crossed the Shantou Estuary via a new undersea tunnel, I saw the under construction freight railway branch that connects with a new relocated Shantou container port which will replace the old city center one I saw last year.

This single track freight railway is a continuation of the Guangzhou–Meizhou–Shantou railway to connect with the new container port that has been relocated across the estuary, away from the city center. This southern freight spur extension shares the new undersea tunnel that the Guangzhou-Shantou HSL runs in, creating a three-track corridor.

Emerging from the undersea tunnel I saw the tunnel approaches for the East Guangdong Regional Railway tunnels that run east west under Shantou’s city center that curve north to join the Guangzhou–Meizhou–Shantou railway freight spur and my train on the Guangzhou-Shantou HSL. The addition of this pair of regional railway tracks creates a five track corridor that heads north up to the new Shantou Railway Station. Below is the road bed for the three tracks: one for freight services and two for the East Guangdong Regional railway system, in addition to the two HSL tracks my train was riding on.

Upon getting off my train, I saw that the old station building I was in years ago had been completely demolished. Below are the satellite images of Shantou Station between my first visit in 2023 (left) and 2025 (right).

The last two images below are taken looking towards the west side of the station where the old Shantou Station would have stood. I could not see the old station platforms and structures from the new station anymore as they are completely demolished.

I headed down to the departure hall of the new station. I again looked towards the west side and saw a large blank wall which hides the second phase of the station.

The new Shantou station has the classic new Chinese station style but now following the new trend of a greyer paint scheme in contrast to the all-white used in the 2010s.

I left the station and passed through the new ride-hailing pick-up and drop-off area. The area has EV charging infrastructure and a ride hailing prompt system showing where your rideshare is waiting for you. It is pretty neat, but such overbuilt infrastructure would not have been needed if there were a regional/urban railway system serving the Chaoshan area. Being JRUrbane, I reject unnecessary ride hailing. Instead I headed to the bus terminal which was in a nice sunken plaza with an LCD screen cycling through next bus info and a terminal map proclaiming “SHANTOU WELCOMES YOU”. Sorry no pictures of the terminal as I saw my bus was about to depart and we ran to board but the bus terminal was really nice and clean.

In Shantou I took my relative to a few tourist sights, some of them the same as my last visit. This visit I didn’t have time to revisit the construction sites of the regional railway tunnel that runs under Zhongshan Road. I’m sure the project has been making tons of progress since 2023.

Just like 2023, I found Shantou city center rustic.

After seeing the joys of Shantou, we made our way via taxi to Chaozhou. On the way we stopped by Cihong Former Residence, an old residential compound of a prominent Chinese-Thai family now turned into a tourist site.

Along the way to Chaozhou we got further glimpses of the under construction East Guangdong Regional Railway. This time the elevated sections that run between the city centers. In a hypothetical future when this regional railway system is complete we would have used it to get between Shantou and Chaozhou. The system even has a station serving the Cihong Former Residence so we could have made that trip via regional rail too once the system is operational.

When I arrived in Chaozhou and got settled in, I went to explore the city, which resembled Shantou in many ways, with plenty of Tonglau buildings. Chaozhou’s old center has a famous street with rows and rows of paifang gates. I also found that Chaozhou’s old town was much more vibrant than Shantou’s. We also learned that Da Laang is a Hong Kong modified Chaozhou thing. The exact assortment of cold platter food found in Da Laang does not exist in authentic Chaozhou cuisine.

After spending a day looking at the historical sights of Chaozhou including the old city wall, stone pagoda and a floating bridge, I headed to the commercial center of Chaozhou to see the under construction underground sections of the East Guangdong Regional Railway. Like Shantou, the East Guangdong Regional Railway runs underground through Chaozhou to directly serve its dense urban core.

One of the stations is in Chaozhou's People's Square where there seemed to be sets of slurry processing facilities that were pressing dirt out and dumping the cakes of what I assume to be the filtered leftovers of the TBM slurry from the tunneling works of Chaozhou’s regional rail city center tunnel.

Like Shantou many older apartment buildings in Chaozhou are getting retrofitted with external elevators.

The area’s lack of a regional railway system was once again apparent as we had to get a taxi from the commercial center of Chaozhou to the nearest HSR station, Chaoshan Station, Chaoshan area’s original compromise high speed railway station between the centers of Chaozhou and Shantou on the older Shenzhen-Xiamen Line.

Leaving Chaoshan via the slower Shenzhen-Xiamen Line I caught glimpses of the under construction East Guangdong regional railway system which would serve Chaoshan Station. This section of the regional system then heads to serve Chaoshan’s Airport and beyond to another urban city center tunnel cutting under Jieyang’s city center.

I arrive to Puning Station and again saw the Jiehui Railway. This single track mixed passenger and freight traffic railway will allow limited East Guangdong Regional services to peripherally serve Puning city and its HSR station. Unlike the last time I saw this railway, I was on the parallel Guangzhou-Shantou HSL where the Jiehui Railway connected (terminated) at Huilai Station. This time this rail spur was crossing the Shenzhen-Xiamen Railway and connected with Puning Station, allowing for interchange between regional and high speed services.

This is a departure from how the East Guangdong Regional system serves the other urban cores in the Shantou-Chaozhou-Jieyang area which uses tunnels cutting across the city center.

Going back to Hong Kong, it cannot be overstated how intensely used the Shenzhen/Guangzhou-Xiamen corridor is. You will see services from both Xiamen-Shenzhen and Guangzhou-Shantou lines weave into each other at key junction stations where both parallel lines meet such as Shanwei.

Services can interline between the faster and slower lines using connection tracks. Unlike the last time I took a Shenzhen-Xiamen HSL service in 2023, my train actually switched to the faster and now open Guangzhou-Shantou HSL at this station.

The Future: No longer China’s Largest Urban Area with no Urban Transit?

Apart from the under construction regional railway which is more of a provincial endeavor, will there be an urban railway system serving the Shantou-Chaozhou-Jieyang area? Can’t say that they didn’t give it the ol’ try. Shantou started construction of a system in 2016 under the approval and auspices of the General Office of the State Council 2003 Document Number 81. Document Number 81 outlines the minimum socio-economic criteria for a city attempting to build a rail transit system, whether it be a “light rail” or a metro “rapid transit” system. At the time it was as follows:

Shantou Prefectural City leading up to 2017, had a general budget revenue of 15 billion yuan and a GDP of 237 billion yuan. Shantou had a total resident population of 5.5 million with an urbanization rate of 69.2% meaning an urban population of 3.8 million people in 2017. At face value, in theory, Shantou could have opted to build a metro system at the time.

I suspect there was some squishiness in the interpretation of the population criteria and perhaps planners opted to evaluate based on the urban populations of the central urban area only (Jinping, Longhu and Haojiang districts). Remember there is a bit of Shantou Prefectural City (Chaoyang and Chaonan Districts) that are located in the neighbouring Ling River (Jiang) valley and are quite urbanized but not really contagious to the central “core” of Shantou.

Following this more conservative thinking, the urban population of Shantou’s central urban areas works out to about 1.8 million people. In addition, Shantou’s dispersed urban nature most likely didn’t allow for a concentration passenger traffic flows. There is no massive office district with wall to wall skyscrapers in Shantou in 2017, today or planned in the future so I doubt that planners were able to submit a report claiming that any part of a proposed Shantou Metro would reach 30,000 people per hour per direction in the future.

It does not help that some of the continuous built up areas of central Shantou is in neighbouring Chaozhou Prefectural City which no doubt hampers the socioeconomic numbers in determining macro metro system planning scale. Current regulations do not allow for combining populations, government revenue or GDP of adjacent prefectural cities to meet approval criteria. At a micro level, some promising prospective metro projects running in the contiguous urban area would require coordination across prefectural boundaries. Those would be off the table or more difficult to get approval.

As a result of these consideration, what actually happened was Shantou opted to build a monorail system based on BYD’s skyrail system, which was built in Yinchuan. Under Chinese parlance, a “light rail” system includes elevated light metros, monorail or AGT technologies. As shown above even with a conservative interpretation of population, 2017 Shantou did meet the criteria for building a light rail system.

However, one year later the central government revised the minimum socio-economic criteria for constructing urban rail transit systems in a new decree under the General Office of the State Council 2018 Document Number 52. Document Number 52 raised the economic and planning eligibility of Chinese cities to construct a rail transit system significantly, as well as adding more fiscal requirements.

My understanding is Shantou still mostly met the new requirements for a light rail system but Shantou’s annual government general budget revenue in 2018 on the eve of circulation of Document Number 52 fell to only 13.2 billion yuan. This drop may have tripped the suspension. Over the next few years a more rigorous formalized process for rail transit evaluations and approvals was implemented, enforced and tightened by the central government and Shantou’s little light rail project was continuously found to be non-compliant and stalled.

Fast forward to today, in 2025, Shantou’s annual government general budget revenue reached 17 billion yuan and GDP was 302 billion yuan which actually was a contraction compared to 2024. Based on the current regulations above, by 2025 Shantou could jumpstart and continue to build a light rail system. However, other (IMHO arbitrary) additional administrative moratoriums have been put on to metro construction approvals at the national level in recent years, barring Shantou for building anything in the 2020s into today.

This is ridiculous, based on my observations, the Shantou urban area does need a small urban transit system. Shantou and the broader area has the population, density and travel demand. The core Shantou urban cluster alone almost has the population of Montreal and is more densely populated. The domestic and foreigner tourist boom into China (though for the Chaoshan Area it more domestic tourists) has placed more pressure on the transport system sparking some complaints on the poor transport arrangements for big events, like the well known lavish Chinese New Year celebrations the region is known for. Every time I come to the area, I am confronted by the congestion around key commercial areas and the lack of convenience in getting anywhere. If you think everyone can just get around using ride-hailing, I got news for you.

That alone should be enough to support and justify building an urban rail system. The under development regional railway is just that, a regional system with wide stop spacings for travel between urban areas. It will be a great system but it does not properly provide local access to the entirety to Shantou’s, Chaozhou’s and Jieyang’s urban cores. The Shantou-Chaozhou corridor alone is urban enough to support one line running along the Chaoshan Road (S233)/ Xin’an Avenue (X088) corridor.

My proposed corridor was not even the corridor Shantou Metro first tried to develop. The initial Shantou Metro monorail line was planned to run east west north of the regional railway tunnel serving dense urban villages like below:

My prospective line would run into approval issues as it crosses into and mostly serves Chaozhou Prefectural City in the north. You wouldn’t know because the central Shantou built up area spills over the perfectural city border. Also note my proposal roughly follows the alignment of the old ChaoChow–Swatow Railway. It’s almost like the 1920s Late Qing and Chinese Republican era development surrounding the old railway is pre-car and still conducive to public transit many decades later. I hope in the future, that the macro policies and environment improves so that Shantou can finally get around to building some urban rail transit to complement its new regional railway system and high speed railway links.

So thanks for reading my experiences and transit thoughts with an often forgotten corner of Guangdong Province. To continue seeing the evolution of what is the largest Chinese urban area without an urban railway system: like, share and subscribe.

Or drop me a Ko-fi if you like what you read!

Line 13 is a key north south trunk line connecting the dense residential suburbs of Guangming District in the north with the technology hubs of Nanshan District in the south by crossing a large natural barrier in the Shiyan area, home to the Yangtai Mountains. It currently terminates at the Shenzhen Bay Port, a gateway to Hong Kong.

Line 13 is notable as it runs through the various technology clusters in Nanshan District, a series of dense skyscraper districts dotted with the global headquarters or major R&D centers of world leading technology companies such as, but not limited to, Qualcomm, Microsoft, DJI, TP-Link, Tencent, Game Science, ZTE, Hytera, Xiaomi, Mindray, Kingdee, Lenovo, VIVO, TCL, Han's Laser, Alibaba and Baidu. As such, Line 13 was well known during detailed design to be a flagship line of the Shenzhen Metro. I guess that is also why Shenzhen Metro Line 13 is colored "Substack Orange™”.

The opening of Line 13 Phase II North last week fulfills the vision of this extremely high capacity north south metro line that crosses technology filled Nanshan District. Yes, I’m ignoring the Line 13 Phase II South Project which is not all that consequential; it parallels a less congested section of Line 2 and mostly serves Shekou, a suburban area filled with expats and entertainment.

A Minor Line Through Fishing Villages

The initial basic concept of Line 13 was envisioned as a peripheral tangential line connecting the proposed Qianhai CBD (Houhai was considered part of Qianhai back then) to the village of Shangwu in the 2008 Shenzhen Rail Network Planning Scheme masterplan. This is essentially what Line 13 Phase I is today.

It was always surprising looking back how Line 13 was viewed as a peripheral line but it was important to note that the areas Line 13 was proposed to go to in the mid 2000s were still very undeveloped and some sections were fish ponds. Many of Shenzhen’s world-class companies were still minor ODM manufacturers (like BBK Electronics), were just beginning their global expansion (like TCL and ZTE), or were humble startups that nobody had ever heard of (like DJI and Game Science). These companies did gradually locate or remain near Line 13 years later.

As this Line 13 concept was refined in the 2010s, adjustments were made since publication of the metro master plan. Shenzhen, at the time the fastest growing city on Earth, did what it did best; it rapidly grew. The Houhai CBD plan was formalized and a thicket of supertalls and skyscrapers was constructed, the Shenzhen Bay Port to Hong Kong was opened, vertical factories assembling electronics at the Shenzhen High-tech Industrial Park gave way to an ocean of office, research and “talent housing” towers as Shenzhen became the global Silicon Valley of hardware. The China Resources Dachong mega-redevelopment added copious amounts of residential and office towers north of Shenzhen High-tech Industrial Park. Meanwhile, the San Francisco Bay Area continues to cosplay as a small town suburban office park.

By the mid-2010s, Shenzhen began working on a number of additional developments along the proposed Line 13. Of note is the Liuxiandong CBD, another technology and research skyscraper district north of Shenzhen High-tech Industrial Park. Later on, the Shenzhen Railway Hub Master Plan (2016-2030) was released, announcing that China Railways proposed a new flagship high speed railway station for Shenzhen, called Shenzhen Xili Station. Xili Station would be larger than the existing monster-sized Shenzhen North. This new monster station would be constructed at the Xili Freight Station between Liuxiandong CBD and Shenzhen High-tech Industrial Park. The new railway station would be partly underground and connected by a series of underground tunnels to a number of surrounding high speed and conventional railways, a Stuttgart 21 on Chinese steroids.

These latter two developments shifted the planning of Line 13, originally planned to deviate east to interchange with Line 5, Shenzhen’s busy middle orbital line, at Xili Station. This Xili Station is not to be confused with the under-construction CR Shenzhen Xili Railway Station; they are not the same. This is the reason why the Line 7 platforms at Xili Station, which opened in 2016, are extremely wide; they were reservations for Line 13 to run down the middle and provide a cross platform interchange (CPI) with Line 7 and have a direct platform to platform interchange with Line 5 that ran under.

But do not fret, the reservation for CPI at Xili Station will now be used by the under construction Line 27 Phase I.

The redesigned section of Line 13 saw no eastern deviation to Xili Station. It now ran straight through the Xili area to directly serve the then proposed Liuxiandong CBD and properly connect with the proposed metro reservations under Shenzhen Xili Railway Station. As such, the interchange between Line 13 and Line 5 was shifted one station west on Line 5 to Liuxiandong Station, which better anchors the proposed Liuxiandong CBD. This sudden shift of Line 13 to interchange with Line 5 at another station will lead to complications upon opening later on; stay tuned.

With all these developments in the 2010s, anticipated demand from all the technology/ knowledge workers was high. In addition, Line 13 was one of the few north south lines in the west side of Shenzhen. With these in mind, planners reimagined and redesigned Shenzhen Line 13 from a peripheral line to a high capacity trunk line. The line was designed to support 8 car Type A trains running up to 100kph.

The design for Line 13 Phase I from Shenzhen Bay Port to Shangwu was approved as part of Shenzhen Urban Rail Transit Phase IV Construction Plan (2017-2022). Line 13 Phase I is 23 km long and has 16 stations. The project started construction in 2018, one year after national approval. That same year, Shenzhen submitted and got approval for a revision to its Phase IV Construction Plan which added more metro projects. Of note here, Line 13 got a Phase II which extended the line on both ends, adding 18.8 km and 11 stations in the north and 4.47 km and 3 stations in the south. So just as Line 13 Phase I started construction, Phase II extensions on both ends of the Phase I project were approved; this is what Chinese netizens called Shenzhen Speed. Both north and south extensions of Phase II started construction in 2020.

The whole line will be complete in 2027 with the opening of Line 13 Phase II South. Line 13 has been proposed, designed, constructed, commissioned and is now shuttling around Shenzhen’s world class technology workforce while “experts” in San Francisco argue which layer of the Earth’s mantle BART needs to tunnel through to get to San Jose.

Today, with Phase I and most of Phase II open, Line 13 is now a monster 45km long line stretching from the Shenzhen border with Dongguan in the north to Shenzhen Bay and the Pacific Ocean in the south. In the middle are a dense medley of office towers housing the world’s largest technology firms. Line 13’s entire Phase I and II North, totaling over 45 km of fully underground 8-car Type A, cost about 300 million INT$/km to construct or about 15 billion INT$. This budget is half the cost of the mostly elevated Ontario Line light metro being constructed in Toronto. Yet Shenzhen Metro Line 13 has a vastly more extravagant scope, using some of the highest capacity metro trains on the planet capable of conveying up to 75,000 people per hour per direction and running completely underground under office and residential districts that are vastly more dense than most of urban North America.

Although Line 13 was given an express line designation and designed to use 8 car Type A trains that operate up to 100kph, the stop spacings are very tight. Line 13’s macro level stops are spaced every 2 km but this is obscured by the long stationless segments going through mountains (see line map in beginning of this post). The segments running through the urban areas of Shenzhen have quite tight stop spacings. This is a very Hong Kong Ex-KCR configuration with tight spacing in urban areas then long segments under unpopulated mountain areas (See West Rail Line or East Rail Line). It might be why MTR had interest in being involved with Line 13.

Opening of Line 13 Phase I

The first section of Line 13, Line 13 Phase I South, opened on 28 December 2024 but I was busy with huge openings in Guangzhou. I am only one man who likes trains and China opens train lines faster than I can visit all of them. I did get time to go over to Shenzhen and review it a few days after it opened and wrote about it below:

After those openings, I seriously dropped the ball in covering Shenzhen. On 28 September 2025, a short 4.9km, 3 station extension of Line 6 Branch Phase 2 and a 9.5 km, 8 station Line 16 Phase II opened. During that time I was in Yunnan Province and Vietnam where I also missed some openings in Guangzhou. However, the openings in Guangzhou were quite interesting so I did go back to see them as soon as I had time. As for Shenzhen’s openings… :x

I did take a look at Line 6 Phase 2 South later on when I was in Dongguan for the opening day of its second metro line, as it terminated frustratingly close to the Shenzhen Metro but did not connect to it. The Line 6 Branch does now connect with Line 13 (spoilers). As I neared the end of 2025, Line 13 Phase I North opened exactly one year later on 28 December 2025. Again I was distracted by openings in Guangzhou. Given that Shenzhen Metro Line 13 touches a lot of what I missed, I plan to bundle my coverage of the Line 13 Phase II North 2026 opening day ride with other interesting features of Shenzhen Metro Line 13 openings I missed in 2025 in this longform post.

Opening Day Experience of Phase II North

It was early morning in Hong Kong and I debated on how to get to Shenzhen. I could take a 9XX series cross harbour bus and transfer (free) to a BX series cross border bus to Shenzhen Bay Port but I opted to go via the ERL and cross at Lok Ma Chau/Futian Port. After entering Shenzhen, I did an obligatory transfer between Line 10 and Line 11 at the now iconic Gangxia North Station, which never gets old for me.

I got a new Line 11 train and took Line 11, an express east west line that blasts across Shenzhen and stops only at her various CBDs, to Houhai Station, named after the CBD it serves, and transferred to Line 13. While transferring I saw a long queue at Houhai Station’s elevator. I’d like to think they are eager people going to participate in the opening of Line 13’s Phase II North Extension.

Once at Line 13’s platforms, I saw the system maps and (MTR-esque) Line 13 Strip maps had been updated with the new extension.

The official first train ceremony happened at the northern terminus of the new extension, Lisonglang Station, where the first SB train would depart. The ceremony was complete with a Hong Kong style Day 1 opening rush. These opening day railfan rushes were starting to gain traction recently in Guangdong among younger railfans as observed with the opening of Guangzhou Metro Line 12 last year, most likely due to influence from Hong Kong’s railfan scene.

In contrast, I did the less ceremonious first NB train into the new extension at Shangwu Station. A yellow heavy rain warning was issued and if I were to participate in the day one official rush, I would have to wait outside Lisonglang Station in the rain. So upon arriving at Shangwu Station on the currently operating Line 13 Phase I, my NB train made a terminating movement by switching over to the operational SB platform. This termination movement (and associated Mandarin, Cantonese and English termination announcements) will no longer be made as frequently in an hour as trains would soon be able to run north through Shangwu Station into the newly opened Line 13 Phase II North.

I got off the terminating train at the soon to be former terminal of Line 13. Initially there were some crowds when I arrived one hour early but about 30 minutes before the first NB train was scheduled to depart (11:18 AM sharp) the crowds of eager fans thickened considerably; most likely families had finished brunch in the urban villages surrounding Shangwu Station.

Line 13 trains are very CRRC copy-paste feeling but also have a noticeably higher trim level than the other Shenzhen CRRC metro trains most likely due to MTR’s differing visual design standards. This is similar to MTR’s other Shenzhen Metro line, Line 4, which had some fairly customized trains that are unlike typical Chinese trains, built by China Southern Rolling Stock Corp (one of the predecessors of CRRC) in collaboration with ABB. Line 13’s trains are strangely built in CRRC’s Jiangmen plant, not in CRRC’s flagship EMU plants in Changchun (which built the MTR C-Train), Nanjing, Qingdao (which built the MTR Q-Train) or Zhuzhou. Below is one of those MTR influenced Line 13 trains serving as the first NB train to enter Phase II North extension. Also note the individual LCD screens over each door, a new passenger information standard being rolled out across new metro lines in China.

We rode the first NB train up into the new extension, oh yeah, the massive 8 car 3.08 meters wide and over 185 meter long trains are fully automated.

I arrived at the new northern terminus of Line 13, Lisonglang Station, the northern Line 13 terminus where the official first train ceremony was launched. I hung around the station and soaked in the new station smell.

The Lisonglang Station concourse was transformed into a little carnival for the opening after the official “first train” SB train departed from Lisonglang.

At street level, the terminus is situated way out in the Shenzhen suburbs bordering Dongguan, and the station’s exits were surrounded by vertical factories and new TOD residential developments.

I headed back down and took a Line 13 train south to explore the rest of the extension. I headed to Fenghuang Town Station, which was a boldly colored interchange station with an elevated section of Line 6. The underground Line 13 and the elevated Line 6 platforms were connected with a long interchange corridor.

I then headed to see Guangmingcheng Station, which has a cross-platform interchange (CPI) with Line 6 Branch.

When I was at the day-one opening of Dongguan Metro Line 1, I took Dongguan Metro Line 1 to the stop closest to the Shenzhen border. I then walked across the border into Shenzhen. When in Shenzhen, I went to Shenzhen Metro Line 6 Branch and took a detour to see Line 6 Branch Phase 2. When I visited one of the stations of this project, Guangmingcheng, the Line 13 CPI was not open yet and Line 13 platforms were boarded up. Below are the platforms in late 2025 before the opening of Line 13 Phase II North with Line 13 platforms all covered up:

You can find out more about my experience with this day one opening in Dongguan and my subsequent metro ride back to Hong Kong below:

The fun thing about a CPI between a monster 8 car Type A line and a much shorter 6 car Type B line is that you can see the difference in required platform length between the two train configurations.

After touring the new Line 13 Phase II extension, my sights were set south to continue a tour of an extension I missed in December 2025.

A Retroactive Review of Line 13 Phase I North

I continued further south on Line 13 for a retroactive review of Line 13 Phase I North. Line 13 Phase I North opened on 28 December 2025 but I was busy in Guangzhou that day. This extension’s flagship station, Shigu, lies at the center of the Liuxiandong CBD and features bold lights and astral motifs on the platform level.

The concourse of Shigu Station is a literal cathedral to metro travel. I will reiterate: I can’t help but notice the temporariness of MTR’s touches in the design, with logos as stickers in contrast with the earlier Line 4 which suggests that MTR had a stronger hand to play in design and was a very blatant MTR feel and design. This is due to the procurement model. MTR is more of an operator for Line 13 compared to the complete rail-plus-property Build-Operate-Transfer (BOT) model Shenzhen used with MTR on Line 4.

More amenities are coming to Shigu Station as the CBD on top gets built out:

I exited Shigu Station and was surrounded by the office and “Talent Housing” towers of the Liuxiandong CBD, home to TP-Link, ZTE, Transsion, DJI and UBTECH among others; including several rare earth refining and advanced materials research labs.

The next interesting station in Line 13 Phase I North is Shenzhen University Station, an interchange station with Line 1. The Line 13 section has bold swirling architectural features.

Other stations of Line 13 Phase I North are not so interesting. However, I think the adjacent sibling stations of High Tech Central and High Tech North, both serving the Shenzhen High-Tech Industrial Park, are fascinating. Both have knockout panels in their platforms as reservations for interchange passages with perpendicular lines. The structure of both stations is almost exactly the same.

Opening Day Woes

Line 13 Phase I North opened in late December 2025 and complaints of congestion from commuters immediately came in. When I visited Line 13 when only Phase I South was operating, trains were quite empty. Line 13 Phase I was complete by late 2025, with both North and South sections opened and demand rapidly grew. Netizens called Line 13 the shortest but busiest metro line in China. MTR does know how to pick winners. The pain point for commuters is that Line 13 trains, while large and spacious, were crowded. Yet the service ran infrequently (every 6 minutes) due to a lack of trains ordered for Phase I.

A Humble Station, Thrust into the Limelight

The epicenter of this controversy is the congestion at Liuxiandong, the station that suddenly was decided to be an interchange between Line 13 and Line 5, as well as thrust into the role as an anchor for the namesake CBD but was never designed to be. Liuxiandong Station was built during an era where stations not planned to be an interchange/flagship station were a bit more cramped/austere compared to the cathedrals of today such as Shigu Station. As the Liuxiandong CBD was being built out, Liuxiandong was the only metro station in the area. The station bore the brunt of all the walk-in demand from the numerous new large office developments in the area and developed a reputation for being a little too congested during peak periods.

The opening of Line 13 Phase I North allowed some Liuxiandong CBD tech workers to shift away from Liuxiandong Station and use Shigu Station to access the metro instead. However, the opening of Line 13 Phase I North replaced congestion at Liuxiandong Station from walk-in traffic to transfer traffic between Line 5 and Line 13. Transfer traffic was consistently overwhelming Line 5 platforms at Liuxiandong mere days after the opening of Line 13 Phase I North. The video below was taken on 29 December 2025, the first workday after the opening of Phase I North, transforming Line 5 Liuxiandong Station into an interchange station with Line 13:

Making things worse, Liuxiandong Station, a station never designed to be an interchange station let alone one with a trunk line, used a “transfer via concourse interchange design”. In this design, passengers transferring between (for example) Line 13 to Line 5 need to take an escalator from the Line 13 platforms up to the common Liuxiandong Station concourse and take another escalator down. This design puts a lot of pressure on the vertical circulation heading to and from the platforms.

At Liuxiandong Station a bottleneck was created from the limited escalators and stairs heading to and from the Line 5 platforms. These bottlenecks are causing backups in the platform and concourse as large crowds of passengers try to both leave and enter Line 5, which runs as frequently as every 2 minutes during peak periods.

Line 13 has a slightly different issue; being a newer portion of the station designed with interchange flows in mind, there is sufficient vertical circulation between the platforms to and from the concourse. However, the limited frequency of Line 13 is causing platform congestion from queuing passengers attempting to board. Social media after the opening of Line 13 Phase I North showed constant posts complaining of the transfer experience at Liuxiandong Station with oppressive crowds and large amounts of flow control.

The only consolation is that the opening of Line 13 Phase I North allowed Line 13 to reach up to Line 6 in the north at Shangwu Station, strengthening the network of the Shenzhen Metro. Some media observers noted that this Line 13 extension might have relieved congestion at Shenzhen North Railway Station, where Line 6 and Line 5 meet, by providing an alternative route. The Line 5 platforms of Shenzhen North Railway Station had numerous severe crowding incidents throughout 2025, at times nearly shutting down the station.

Not Shenzhen’s First Rodeo

This is not the first time Shenzhen Metro had congestion from under designed stations. Line 1’s High-tech Park Station is one station away from Line 1’s interchange with Line 13 at Shenzhen University Station. High-tech Park Station was built during the same 2010s era as Liuxiandong Station on Line 5. Like Liuxiandong Station with the Liuxiandong CBD, High-tech Park Station was, for the longest period, the only station to serve the Shenzhen High-tech Industrial Park. As Shenzhen became the hardware technology capital of the world, Shenzhen High-tech Industrial Park rapidly grew into a forest of tech offices and research towers. The area is now home to TCL, Mindray, Skyworth and Game Science among others.

In the late 2010s, High-tech Park Station was congested from all the new technology companies opening in the Shenzhen High-tech Industrial Park. In response, Shenzhen Metro had to retrofit the station in 2020 with two new sets of stairs using the existing open to below spaces in the concourse. Shenzhen Metro had already extended Line 9 to the Shenzhen High-Tech Industrial Park to relieve Line 1; the Line 9 Phase II West extension opened in late 2019. This is why Line 9 has a huge jog north after Hongshuwan Station.

On opening day of Line 13 Phase II North, I did go to High-tech Park Station to whet my appetite for the progenitor of what congestion relief solutions were implemented in Liuxiandong Station. I arrived at the station and saw the two sets of stairs whose styling looked a little out of place compared to the cohesive look that fixtures in Chinese metro stations typically take.

The “aftermarket” looking nature is very apparent on the new stairs at High-tech Park Station but hey, I heard one great man who had huge influence in Shenzhen say it doesn’t matter if the cat is black or white…

In addition, you can see the initial station circulation of the cramped High-tech Park Station was completely overhauled to accommodate the massive growth of commuters. New faregates were finagled in wherever they fit and positioned to be as close to each of the new stairs as possible, giving a very haphazard look.

In the future, Line 20, an under construction parallel express line for Line 1 (the second relief line for Line 1 after Line 11) will interchange at High-tech Park Station. Hopefully the station can be massively expanded as part of the interchange station enabling works. In addition, Line 27, the line that took over Line 13’s interchange reservations at Xili Station, is proposed to have its Phase II extended to interchange at High-tech Park Station.

Second Rodeo

Two weeks after the opening of Line 13 Phase I North and constantly receiving complaints about congestion at Liuxiandong Station. Shenzhen Metro in a wild development, rapidly retrofitted additional stairs between Liuxiandong Station’s Line 5 platforms and concourse. The retrofit done at High-tech Park Station was repeated for Liuxiandong Station on an accelerated scale. By the week of 16 January 2026, Shenzhen Metro started making holes in the tiles for the structural frame supporting the new stairs to mate with the platform structure and removing ceiling architectural finishes as well as marking out demolition locations.

Shenzhen Metro then prepared the surfaces to punch a big hole in the ceiling of platform level to the concourse. All this happened as the station operated normally (read: congested).

The hole was punched on a weekend requiring a full station closure with Line 5 trains bypassing the station all weekend. I saw notices of weekend work when I was in Huizhou making my way back to Hong Kong via the Shenzhen Metro Line 14 that week.

By 31 January 2026 the steel structure and stairs were substantially complete. This was constructed in a span of just two weeks:

My Experience

I arrived at the revamped Liuxiandong Station via Line 13 as a continuation of my opening day tour to see its transfer operations and how the new stairs were holding up. In the Line 13 concourse on my way to Line 5, I was confronted with a lot of special peak period directions, signage (only in Chinese) and passenger control equipment, the scars of a congested station.

I can see how Liuxiandong Station was congested leading to the massive number of complaints after the opening of Line 13 Phase I North. The station concourse originally just had two sets of vertical circulation elements to the Line 5 platform below, with two sets consisting of stairs and an escalator each. The new additional retrofitted stairs adopted a scissor configuration to maximize capacity and add another vertical circulation element on top of the two originals. These stairs are even more “aftermarket” looking than those at High-tech Park Station.

At the platform level, the quick build nature of the two new sets of interlocking stairs is readily apparent. It is as described above: punch a hole in the station concourse floor slab, reinforce the edges with I-beams and finesse two sets of stairs into the space.

The hype about Liuxiandong Station’s heavy use and crowding problems was real. During my visit of the station, it was quite busy with people transferring between Lines 5 and 13 despite it being a Sunday early afternoon.

After my tour of Liuxiandong CBD, I continued south on a Line 13 train to see Shigu Station. Then I reached another oddity of Line 13, the unopened Xili Station that will serve the aforementioned CR Xili Railway Station. The railway station and metro are not yet open. Line 13 trains are bypassing the nearly completed metro station.

Operations after the near completion of Line 13

Opening day and my tour of Line 13 after the opening of Phase II North were on a Sunday, the next day was a workday. If you thought the netizen complaints blasting Line 13 in early 2026 after the full opening of Phase I North were bad, my scan of social media in the subsequent workdays this week after the opening of Line 13 Phase II North showed an even more congested Line 13. The sticking point is now the low frequency. MTR was running the entire line with trains every 5-6 minutes during my Sunday visit. On Monday, the first workday with the new extension, the frequency was not much higher and led to complaints from commuters after a congested morning rush hour with the new major extension. Line 13’s daily ridership rapidly grew to 450,010 passengers that Monday, 29 June 2026. Seven months ago when only Line 13 Phase I South was open, ridership hovered at around 90,000 trips per day.

After complaints from Monday morning, MTR announced that morning rush hour frequency would be increased the next day (Tuesday) using short turn service patterns as well as pressing additional trains into service. The section between Deya Road and Houhai Stations would run every 3 minutes during peak periods while the other sections would be served every 6 minutes.

These changes in train operations led to some improvements on Tuesday though complaints were still being made. Line 13’s daily ridership reached 450,900 on Tuesday, 30 June 2026, breaking the single day Line 13 ridership record despite that day being a typical day. The busiest section is estimated to already convey over 35,000 people per hour per direction during every peak period, more than any rapid transit line in North America. Ridership again grew the next day, reaching 463,900 on Wednesday, 1 July 2026, though it was a holiday in neighbouring Hong Kong so I would expect a bump in ridership as Line 13 does connect with the Shenzhen Bay boundary crossing. Regardless, the daily growth trajectory of Line 13 is obvious and ridership has more than quintupled (increased 5-fold) in the span of seven months. Line 13 will no doubt be another +1 million per day ridership trunk line in Shenzhen Metro’s rapidly growing pantheon of high performance lines.

Netizens observed that the “struggles” with providing adequate service on Line 13 mirror the experience with MTR’s management of Shenzhen Metro Line 4 during the 2010s. Line 4 was running 4 car trains for years after the Line 4 Phase II North extension greatly extended the line north into the suburbs, leading to severe congestion. It didn’t help that MTR/Shenzhen rapidly built large numbers of rail-plus-property developments around each Line 4 station. MTR was heavily criticized for being very slow to shift from 4-car Type A trains to the ultimate 6-car configuration that Line 4 had been designed for, as well as for increasing frequencies; much to the chagrin of local media and the general population of Longhua District. Line 13 on the other hand was not designed by the MTR Corporation and is running even higher capacity full fat 8 car Type A trains but at a low frequency. From the public’s perspective, MTR, as the operator, is to blame.

The Future of Shenzhen’s Premier Tech Line

Congestion on Line 13 after the opening of Phase II shows that more trains need to be deployed and maybe even purchased to achieve a higher frequency. This is especially important as Shenzhen has been clamping down on illegal electric scooters which may drive more overall demand for metro travel. In addition, Xili Railway Station will open in the coming years. If Shenzhen North Railway Station is any gauge of demand, which drives that metro station to be consistently the top five busiest in the system, I expect a huge amount of demand to arrive at Line 13. In the mid-future the next solution is a relief line, Line 29 (in bold red), whose Phase I is under construction. It parallels Line 13 and can be an accessory to its development.

Regarding Line 13, Dongguan is now a popular place to relocate R&D bases due to Shenzhen’s increasing land costs. An extension of Line 13 north across the border into Dongguan would connect to Huawei’s massive Oxhorn R&D base, which is only 7 km away from Line 13’s current northern terminus behind a set of mountains. It will also allow Shenzhen Metro Line 13 to connect with Dongguan Metro Line 3. So hopefully over time MTR can ramp up service on Line 13. The line is positioning itself to be an indispensable addition to the Greater Bay Area transportation network and quietly supporting China’s technology sector.

So thanks for reading my experiences with another first day metro opening in Shenzhen. To continue seeing the evolution of what will be the world’s largest urban railway system: like, share and subscribe for more long form adventures and technical analysis in East Asia.

Or drop me a Ko-fi if you like what you read!

Hong Kong’s Long and Rich Aviation History

HK’s modern jet-age aviation history was anchored by the iconic Kai Tak Airport in Kowloon Bay. Kai Tak was a single runway airport jutting into Victoria Harbour surrounded by mountains and urban density, home to the world famous ‘Hong Kong Turn’ landing approach, in which pilots first used the Instrument Landing System (ILS) to fly low and close to Checkerboard Hill in Kowloon. After confirming their position and approach, pilots had to manually make a dramatic, sharp 47‑degree visual right turn, then quickly adjust to land on Kai Tak’s runway. During this segment, the plane flew so low over Kowloon that it felt like the landing gear was scraping the TV antennas on the tenement buildings, and the port-side wing tip skimmed the Kowloon Walled City.

In 1973, the British colonial government of Hong Kong was already considering a location to relocate Kai Tak’s operations. 13 locations across the territory were studied; six were shortlisted, but the government sat on the proposal for a few more years.

By the late 1980s Kai Tak International Airport was saturated with demand, giving urgency to the British colonial government’s re-examination of the new airport proposal. HK Kai Tak airport was forecast to reach absolute capacity in the mid‑1990s. By 1996, it was exceeding the 24 million passengers per year capacity, handling 29.5 million international passengers and 1.56 million tonnes of international cargo. These numbers made it the third busiest airport in the world for passengers and the busiest airfreight hub in the world.

The six shortlisted locations were narrowed down to two and a decision was made for the final site. In 1989, the HK government announced the decision to relocate the airport to a new reclaimed island north of Lantau Island and to connect it with a completely new transport backbone dotted with new HK-style dense suburban developments. The plan was called the Airport Core Programme, an ambitious package of projects that included not only a new two runway airport built on a reclaimed island but also:

• A new expressway system (North Lantau Expressway, Lantau Link, Route 3 Kwai Chung to Tsing Yi, West Kowloon Expressway, Western Harbour Crossing) and a new MTR Airport Express service that linked the airport to the rest of Hong Kong; it included what was then the world’s longest combined road/rail bridge and a third cross-harbour tunnel.

• Several rail+property developments, including a new town (Tung Chung) as well as newly reclaimed land hosting Hong Kong’s two tallest buildings: IFC on Central Reclamation Phase I and ICC on West Kowloon Reclamation.

• All these developments are connected by a new MTR local line, the Tung Chung Line.

The new airport was called Hong Kong Chek Lap Kok International Airport and it opened in 1998, one year after the Handover of Hong Kong and one year before the Handover of Macau. The Airport Core Programme’s Rail+Property developments were phased in over time in the 2000s, well timed with Hong Kong’s post reunification real estate boom.

Since the completion of the Airport Core Programme, Hong Kong Chek Lap Kok International Airport operated a two-runway system as Hong Kong was governed under the One Country, Two Systems framework with China. Usage grew rapidly, eating up all the raw capacity of the new airport. HKIA became the world’s busiest cargo airport for 16 consecutive years (since 2010), far surpassing famous Western airfreight hubs such as the FedEx “Superhub” at Memphis International Airport, the UPS “Worldport” at Louisville Muhammad Ali International Airport and “CargoCity” at Frankfurt Airport. At the same time during the 2010s before the COVID-19 era, HKIA was consistently among the top ten busiest airports in the world.

By the 2010s, HKIA had reached its practical maximum capacity of 68 aircraft movements per hour. Forecasts showed future demand would far outstrip this, threatening Hong Kong’s status as a leading global aviation hub. Since the handover, Mainland China’s economy has grown rapidly. Neighboring Pearl River Delta megacities of Guangzhou and Shenzhen opened and expanded massive world class airports in Baiyun and Bao’an respectively. Another mega world class airport in Foshan is being built at the time of writing. Building a third runway at HKIA was identified as the only long-term solution, hence the introduction of the Three-Runway System (3RS) expansion plan.

In 2015, the Hong Kong SAR Executive Council gave the green light for the 3RS expansion plan and funding was approved. To expand the two-runway system to a three-runway system with a new terminal and additional supporting facilities, the airport island was expanded north. With the full 3RS operational, HKIA’s capacity will rise to handle 120 million passengers and 10 million tonnes of cargo annually, roughly a 50% increase over the old two-runway limits. Environmental impact assessments were completed after significant legal challenges. Full-scale land reclamation started in mid-2016.

The Hong Kong-Zhuhai-Macau Bridge (HZMB) opened in 2018 with HK side boundary crossing facilities built on a reclaimed island beside the airport. Bonded bus service over the HZMB commenced in December 2023. This service allows HKIA international flight passengers coming to and from Macau and Zhuhai to check in and drop their luggage on the Macau/Zhuhai side of the HZMB boundary crossing and board a special “bonded” bus that directly enters the HKIA restricted airside zone. Passengers can then proceed directly to their gate and board their flight. This avoids the need for Macau and Zhuhai passengers arriving at or departing from HKIA to clear Hong Kong immigration.

Since COVID, HKIA has had a delayed recovery in its annual passenger numbers, while still maintaining its world topping airfreight volumes, processing over 5 million tonnes of airfreight in 2025 using large multiple level vertical logistics centers. However, it has been the world’s second fastest growing passenger airport in 2024 and 2025. It grew 34.3% in 2024, reaching 53 million passengers in its first full year of normal operations since COVID. In 2025, it served over 60 million annual passengers, a 14.9% annual growth. Passenger growth is expected to exceed 10% in 2026, on track to reenter the world’s top 10 busiest airports by passenger volume in the next few years.

The 3RS master plan required T2 to be rebuilt into a full‑featured terminal capable of serving the vastly expanded Hong Kong Airport with a new northern runway and satellite terminal. This is not the first time Hong Kong International Airport has had a T2. The original T2 opened in 2007. It was not a standalone terminal with aircraft gates; instead it was a large departure only hall with shopping, entertainment and check-in facilities known as SkyCity. With its emphasis on shopping, T2 back then was more of a restaurant mall with two rows of check-in desks in the middle (expandable to four). Below are historical 2007 images of the old Terminal 2 from GAKEI.com that include a shot of the to-scale model for HKIA Terminal 2. Note the huge amount of shopping and minimal check-in infrastructure.

At its peak, it handled around 30 airlines and served as an overflow check-in facility for Terminal 1 (T1). However, all passengers still boarded flights from T1’s piers and satellites via an APM that transported passengers who had just checked in at Terminal 2. All arrivals were still processed at Terminal 1.

The old T2 was closed in 2019 for renovations as part of the 3RS master plan. The old terminal was torn down and a new one was constructed. The floor area was doubled and new arrivals facilities that the old T2 never had were added such as immigration, baggage claim and customs. SkyPark Mall was relocated to a larger structure on the east in collaboration with the HK developer New World Development as part of their K11 brand now called K11 SKIES.

Note how the airport wraps around the elevated MTR Airport Express Station, which was designed from the get-go to be flanked by two terminals. The elevated station structure built in the late 1990s was designed to support Spanish solution platforms.

The reconstruction of the Hong Kong Airport Terminal 2 has been something I have been following for years now and is a constant background of my life when I arrive at HKIA on my many travels. It is always cool to see the sweeping roof being assembled as I leave Hong Kong for a vacation. Ultimately, a second APM system will connect the T2 hall with the brand‑new satellite concourse south of the new north Runway.

In the future, a system like the ULTRA PRT used in Heathrow Airport will connect the HKIA T2 hall/K11 SKIES to the HZMB port, allowing direct coach services to Toisan or Zhaoqing.

Glimpses of the Future

The first section of the renovated Terminal 2 actually opened on 23 September 2025. This section is the massive new intercity coach terminal for airport buses bound for Mainland China (and Macau). I toured it a few days later after sending off some visiting friends at the airport (I commonly do this for some reason). Upon leaving, I noticed a light behind the Airport Express train/platform screen doors on the Terminal 1 side of the platform. Upon closer inspection, I saw they were getting ready to implement a Spanish solution layout at the MTR Airport Station with HK Airport Terminal 1 on one side and Terminal 2 on the other side.

I left the MTR Airport Express station and headed down and under the station to get to the other side of the Airport Express tracks as the Terminal 2 platforms were not open yet. I entered the newly opened intercity bus terminal which had that big new building smell. When I looked up, I could see the not‑yet‑opened Terminal 2 arrival and departure floors and skywalks above the bus terminal with cool LED mood lighting.

When I looked up on the other side, I could see the skywalks leading from the still under construction HK Airport Terminal 2 floors with the corresponding new unopened Airport Express platforms. The light I saw behind the train/PSDs was from there.

The Check-in Hall Opening

After the opening of the intercity bus terminal, what does Hong Kong International Airport’s newest addition look like? I headed to HKIA via Airport Express (the world’s best Airport Link service) with my friend, whom I was seeing off. On the way I saw that the under-construction infill station, Tung Chung East, was nearly complete.

Upon arrival, we saw the first feature of the newly opened Terminal 2, a Spanish Solution setup (which no one expects) at the HKIA MTR Airport Express Station.

My friend’s Cathay Pacific (remember the best airline) flight required check-in in Terminal 1 the old fashioned way. However, at the departure board you can now see the revival of the two color terminal ID sorting system with blue for Terminal 1 and golden yellow for Terminal 2.

After bidding farewell to my friend, I headed to explore Terminal 2 using the wayfinding that had been updated with Terminal 2’s opening. I crossed over to Terminal 2 on the arrivals level.

However, since the Airport Express is now a Spanish solution with Terminal 1 on one side and Terminal 2 on the other, intrusive thoughts took over: “I can just wait for an Airport Express train and when both doors are open I can just punch through the train from Terminal 1 over to Terminal 2”, which I performed to great success:

I explored the Terminal 2 side of the Airport Express station lower departure platforms where passengers board Airport Express trains bound for Hong Kong, which is just as airy and bright as the original Terminal 1 side.

The MTR Airport Station is a double deck stacked station to accommodate the Spanish Solution for both directions.

Then I headed up to the HKIA arrival platforms of the MTR Airport Express where arriving passengers from Hong Kong alight and head to check in to their flights. The design of this level echoes that of the legacy Terminal 1 side but with a warmer tone.

Legacy late 1990s era Terminal 1 departures level platforms for reference:

Let’s look at the experience of heading to Terminal 2 for a departing flight via Airport Express. Beyond the Airport MTR Station, you cross the highest skybridge, above the sweeping hall of the long distance coach terminal that I reviewed in the section above. Here passengers can connect with coaches to Mainland China.

Walking beyond the coach terminal hall, you enter Terminal 2 proper, which has a circulation design mirroring Terminal 1’s:

• Passengers arriving by Airport Express walk up a gently sloping ramp that crosses over a dramatic moat where you can see the arrivals below. That ramp takes arriving Airport Express passengers to the check-in hall at the departures level.

• Conversely, passengers arriving by airport bus walk down a gently sloping ramp that also crosses over that same dramatic moat where you can see the arrivals below. That ramp takes arriving airport bus passengers down to the same check-in hall at the departures level.

Coming from the Airport Express station, I walked up the ramp and finally arrived at Terminal 2’s check-in hall. I looked up to see the new spectacular swooshing roof, high tech futuristic check-in counters and colorful but hard seating.

The area is plastered with lots of LCD screens:

Unlike the old Terminal 2 which was dominated by retail uses, the new Terminal 2 has all the retail options shoved to the back of the terminal. Behind that retail is a skybridge to the new K11 SKIES luxury mall complex, Hong Kong’s answer to Singapore’s Jewel Changi Airport entertainment hub. You can see the purple K11 SKIES complex in the background.

After checking in passengers then cross airside control, enter the airport restricted section and take the escalator down to the basement APM platforms surrounded by LED screen walls. The basement of Terminal 2 has APM trains heading to Terminal 1 where passengers can board their flights. As noted above in the history section, in the future these basement platforms will have a separate APM service going to Terminal 2.

I left the check-in hall and headed up the ramp to the Terminal 2 airport drop off area with nice newly paved roads and sidewalks. The new traffic signals still even have construction dust on them. Like Terminal 1 this area is where Hong Kong’s excellent Airport Bus services drop off passengers for departing flights. Passengers arriving at Terminal 2 need to head under the Airport Express Station to get to the original airport bus pick-up terminal.

Like Terminal 1, Terminal 2 is served by a parade of airport buses that reach all over Hong Kong.

Overall, a great plane/bus spotting location!

The perspective of being dropped off at Terminal 2 from an airport bus service is similar to Terminal 1. After alighting from your bus and entering the terminal, you are welcomed with sweeping views of the check in hall as you walk down the ramp. A show stopping send off for tourists from Hong Kong or a confident “See you again back home!” for Hongkongers going on vacation.

After walking down the ramp to the check-in area, you can turn around to see a parade of buses dropping off passengers at the drop off area through a wide airy window. Just like Terminal 1.

I head down a few floors to the new arrival area that the old Terminal 2 didn’t have. It’s not as nice and iconic as the one in Terminal 1 where you might chase Asian celebrities, but it does the job. It is big but it’s a bit barren.

From there you can see the Terminal 2 arrival experience. You have just passed immigration, collected your bags, and passed customs. After exiting the airport restricted zone you have a straight, level, step-free path directly to the MTR Airport Express platform. Great if you have lots of bags and are unfamiliar with the airport. You just walk straight onto the train. This is actually an improvement compared to Terminal 1, which has some minor level changes and an arrival hall split into Halls A and B, lacking a straight line of sight from the arrival hall exit to the MTR Airport Express Station.

While walking you pass by renderings of the Hong Kong International Airport 3RS and Hong Kong‑Zhuhai‑Macau Bridge master plan, as well as advertisements for an expansion of the airport exhibition centre.

Then you cross over some skybridges over the long distance coach terminal hall.

There are also side passages that connect the Terminal 2 arrivals hall with the MTR Airport Express Station, crossing over the intercity bus terminal via separate skybridges. If you take the north side passage, you can see the covered PSDs on the section of platform reserved for the extension of Airport Express trainsets. The Airport Express system is designed to run 10 car trains. The Airport Express service is currently running 6 car + 1 baggage car sets, so the PSDs for two cars are covered.

On the Terminal 1 side the PSDs for the extra reserved platform length are not covered.

As I waited for my Airport Express train on my Terminal 2 platform I looked at the people from Terminal 1 waiting one track away. Also note how the PSD model on the Airport Express boarding/HKIA arrivals level is different from the PSDs used in the Airport Express alighting/HKIA departures level.

I boarded my Airport Express train for my free same day ride back into Hong Kong. You can see the Spanish solution in action. All Airport Express trains have information on the Spanish solution setup in every vestibule on the train.

So thanks for reading this aviation post. Please like, share and subscribe for more on the ground transport openings and analysis.

Or drop me a Ko-fi if you like what you read.

Guangzhou Metro is a comprehensive system for a massive city, but its wide coverage paired with its low capacity creates a strange combination that produces strange results. The outcome of this metro network that is thinly spread out across the vast and populous Guangzhou/Foshan area is overwhelming congestion. The network before the recent implementation of the various relief lines was also highly radial and funnels all traffic into a small handful of city center trunk lines, some built to light metro standards. These handful of city center metro lines are in itself thinly spread out across Guangzhou’s massive and dense urban core.

In short, Guangzhou is a Tokyo-scale city with Hong Kong density and culture, leading to similar urban rail passenger throughput achieved on the Guangzhou Metro compared to systems found in Hong Kong or Tokyo (over 50,000 people per hour per direction). Yet, Guangzhou was trying to meet this demand using small, European-sized metro trains (or even light metro trains) on a metro network density and capacity of a metro system serving a mid-sized European city, with disastrous results.

Other Chinese megacities, such as Beijing and Shanghai, brought congestion relatively under control by rapidly expanding their networks in lockstep with population and ridership growth. Relief lines did eventually come to Beijing Line 1 or Shanghai Line 2, with more anticipated to come online in the coming years. Shenzhen built out core metro capacity first in urban districts to only then start construction of expansions into the suburbs after the mid-2010s.

Wuhan, Hangzhou, Chongqing and Chengdu quickly committed to higher capacity designs after finishing their initial lines. Guangzhou’s more contentious local land ownership politics and land rights constrained city center metro expansion and delayed relief line projects. Influence from the Hong Kong MTR down the Pearl River meant Guangzhou Metro was incentivised to expand rapidly into the suburbs with radial lines serving copious amounts of greenfield Rail+Property TOD developments. Meanwhile, the city's general austerity (read: being a cheapskate) slowed the adoption of high-capacity metro designs until Guangzhou started building lines numbering in the double digits.

In these past four years, over 350 kilometers or 220 miles of new metro lines have been added to the Guangzhou/Foshan area. As I have covered their openings and history in a multipart series of series for Late 2023, Late 2024, Middle 2025, Sept 2025 and Late 2025. Instead of further expanding into the suburbs, most of these new additional Guangzhou Metro lines were added in the core of the system (Lines 10, 12, 13, 18 etc) or were orbital distributor lines (Line 7 and 11). These new city center metro lines are bearing results, starting in 2026, popular media noted that conditions on key congested Guangzhou Metro lines like Line 3 have improved with these numerous recent metro openings diverting traffic away. So after a series of visits to Guangzhou in 2026, I decided to compile a review on my personal experiences with the conditions on Line 3 and 5 after these recent relief projects and measures have been implemented.

The State of Play

During my review of Line 11’s opening a little over a year ago, I noted two of Line 11’s stations connecting to Guangzhou’s national railway stations that did not open with the rest of Line 11. One of these Line 11 stations, Guangzhou East Railway Station, opened on 29 September 2025 along with the other extensions I covered in a previous article. Guangzhou East Railway station will be come a massive 4 line metro interchange under a completely rebuilt double deck Guangzhou East Railway Station.

After seeing some Malaysian friends on New Year’s Eve and staying overnight at Guangzhou, I woke up to 2026 and decided to take a look at this newly opened platform for Line 11 at Guangzhou Railway station.

This new station is key to further unlocking the utility of Line 11 city center loop by allowing Line 3 and Line 1 passengers to transfer to Line 11. So on this New Years Day visit, I decided to look at these new Line 11 platforms for the first time, a few weeks after it officially opened. I arrived to Guangzhou East Railway station via Line 1 and make my way to the Line 11 platforms.

I arrived at the station and being New Year’s Day which is a high demand day for the Guangzhou Metro. I can see how the opening of this new station is shifting passenger flows in the Guangzhou Metro after a few weeks of operation as well as note the overall transfer experience. Arriving at the new platforms via Line 11, I headed up to the concourse where there was a lot of queueing from the massive Line 11 station into a narrow connecting corridor to Line 3. I actually think the opening of this Line 11 station might have made Line 3 worse.

We were corralled into a series of 100m looping queues to exit the Line 11 concourse into the legacy narrow passages heading Line 3, passing a great sticker poster of the metro trains that serve Guangzhou East Railway Station the A1 series for Line 1, the B1 series for Line 3 and the A9 series for Line 11.

Making my way into one of Line 3’s older station halls from Line 11’s concourse, there was a lot of passenger flow control. It took more than 10 minutes to pass through this section.

It is clear that Guangzhou Metro is putting in some tight passenger control to prevent Line 3’s platforms from being swamped. We were squeezed down one set of escalators out of three, to further throttle the flow of passengers heading to Line 3’s platforms. Again, I think Line 11 might have made Line 3 worse in some respects.

Once I made it to Line 3’s platforms, things seemed pretty smooth.

Line 3’s strange branch system limits frequencies in some core sections. You will see gaps of up to 4 minutes, as trains usually arrive every 2 minutes but some Line 3 services are diverted to the Tianhe Coach Terminal Branch.

Seeing the congestion on Line 3 at Guangzhou East Station I instead went to take the parallel Line 1 around to Tiyu Xilu Station, the busiest metro station in Mainland China. I arrived to a pretty orderly scene:

Heading up to the Tiyu Xilu Station concourse, a lot of passenger flow control was implemented. Passengers are herded into going around in loops before they enter either Line 3 or Line 1 platforms. As the old man behind me said in Cantonese “Can’t cross, it is barrier’ed here, ya come back.”

Zhujiang New Town Station, a key interchange between Lines 3 and Line 5.

I headed down to the Line 5 platforms which should be slightly relieved with the opening of Line 10 and Line 13 these past years.

Empty eastbound Line 5 trains were deployed to clear platforms at Zhujiang New Town. The initial Line 13 Phase 2 segment with Line 11 may not have sufficient diversion effect to improve the conditions on Line 5. We will have to wait for the opening of the rest of Line 13 Phase II to see the full effects of relief to Line 5.

Line 3 is still congested as of 2026 in the sense that more extreme passenger control measures still exist. Such as temporarily blocking passengers transferring from Line 5 for several minutes from entering Line 3 to allow for Line 3 platforms and trains to clear.

At the same time Guangzhou Metro staff are also blocking passengers from the concourse from entering Line 3. It is always amusing to see the torrent of passengers cascading down the escalators (or up the stairs) into the Line 3 platforms the moment the 'floodgates' open.

Congestion and crowd control measures still remain in effect on Sunday night at Guangzhou East Railway Station in the transfer corridor between Lines 3 and 11 when I visited Guangzhou in March 2026. I returned to Guangzhou again in April 2026 to visit a friend to similar conditions. Note both times I have no Line 3 pictures, I still have a reflex in my Guangzhou pathfinding cognition to instinctively avoid taking Line 3 whenever possible. Both times I arrived at Guangzhou East via Line 11 and take regional trains back to Home Kong.

Contrast above with my experience on New Year’s Day:

But overall the opening of Guangzhou East Railway Station connection with Line 3 and Line 11 has created a new fairly smooth but busy interchange. Luckily the queues don’t spill into the Line 11 concourse but we will have to see when the final section of Line 18 opens at this station in the future. Providing further relief to Line 3 but also creating a four line interchange thus putting pressure on Guangzhou East Railway Station itself. Below was a regular Sunday night in March 2026:

The Future of Line 3 and Guangzhou as a Whole

These recent new expansions have shown in data the level of relief, overall Guangzhou Metro ridership grew 5% in 2025 yet key congested urban lines saw ridership decreases across the board. Line 3’s ridership fell from 1.8 million passengers per day in 2024 to just over 1.7 million in 2025. Demonstrating that the “vibes” are correct, Line 3 has been less congested compared to my experiences in the late 2010s where at times Line 3 alone exceeded moving 2.7 million people per day. The incomplete Line 18, the world’s first high speed metro and first relief line to Line 3, saw ridership grow from 114,000 per day in 2024 to 141,300 in 2025, and was observed facing some congestion during AM rush hour. Meaning Line 18, even in it’s half finished form, is having a diversionary effect. Line 11’s first year average daily ridership exceeds 500,000 passengers per day and this is without the recent Guangzhou East and Guangzhou Railway Station mega interchange openings. When I was visiting Guangzhou on December 31, 2025, Guangzhou Metro broke a single day ridership record exceeding 14 million rides, with the Line 11 city center ring alone carrying 953,000 passengers that day.

Further relief to the core of Guangzhou Metro is on the way. With Line 12 central segment and the rest of Line 13 Phase II slated to open this year, relieving Lines 2, 5 and 6. While Line 18 North, Line 22 North, Line 24 are well underway in addition to Line 28, Guangzhou’s new East West high speed metro which recently broke ground.

The Guangzhou Foshan urban area continues to grow, with Guangzhou City alone adding over 400,000 new residents in the last 5 years. Coupled with the recent crackdown on electric scooters, whose rise as an alternative was an indirect response to Guangzhou Metro’s (years in the making) reputation for notorious congestion and chaos. Metro relief and expansion are still badly needed in the Ram City.

Currently in the Year of Our Lord 2026, Guangzhou is still mulling its Phase IV Metro Expansion plan which would see the construction of projects such as Line 19 and Line 26, Line 3’s second and third relief line. The urgency is real, as Shi Haiou (史海欧) Senior Chief Engineer of the Guangzhou Metro Design & Research Institute Co, Ltd., noted in late 2025 at the Guangzhou Greater Bay Area International Rail Transit Industry Exhibition that Guangzhou Phase IV Metro Expansion Plan should be submitted for national approval as soon as possible. In previous articles I wrote about the Guangzhou Metro in 2025, I noted regulatory procedure in China is that the next metro expansion plan can not be submitted unless 70% of proposed investment in the current metro expansion plan is implemented. With the mass opening of several Guangzhou Metro projects in 2025, Guangzhou has now exceeded that 70% threshold of the current Phase III Metro Expansion Plan. Moves are being made in local government organs. The 2026 Guangzhou Municipal Government Work Report noted that submission and Central Government approval of the Guangzhou Metro Phase IV Metro Expansion proposal is one of the key goals for 2026. I wish them the best in securing approval for even more metro expansion, Guangzhou still, in my observation, needs it.

With these thoughts in mind, I then took a crowded Guangzhou-Shenzhen Regional train back to Hong Kong in late April 2026. I have to go to Guangzhou again next week upon publishing this post and most likely again after a few weeks.

So thanks for reading this analysis of Guangzhou. Please like, share, and subscribe for more news and on the ground coverage of how the world’s most congested metro lines are no longer congested.

Or drop me a Ko-fi if you like what you read!

Nestled in the heart of Northwest China, Ningxia or Ningxia Hui Autonomous Region was historically the homeland of China’s Hui Muslim minority group and is becoming a modern logistics, chemical and renewable energy hub. The Hui people make up over one third of Ningxia’s population and maintain a distinct Islamic Chinese culture visible in their dress. This is reflected by the the thousands of mosques scattered across its arid/semi-arid landscape, punctuated by Ningxia’s striking terrain shaped by a tapestry of mountains, deserts and the Yellow River.

Ningxia historically plays an important role as a commercial stopover of the Silk Road’s northern route. Today it is a key strategic node in the Belt and Road Initiative. "Single Bill" rail-ship and China-EU intermodal container trains export Ningxia’s chemical and agricultural outputs to the rest of China and beyond to East/Southeast Asia and Europe/Central Asia respectively. Yinchuan is the bustling provincial capital and largest city in Ningxia with vibrant halal eateries and Hui culture.

Ningxia Through Cantonese Eyes

Ningxia is an interesting place, I personally don’t hear much about it down in Hong Kong but I did know it exists before deciding to travel there (you will see why later on). However, among us Cantonese Chinese, one of the things Ningxia today is famous for, is it’s 菜心 (Choi Sum) or 油菜 (You Cai) and Goji/Wolf Berries. But I think I’ll focus on Choi Sum which has a more interesting history. The modern synonymy between Ningxia and Choi Sum in the Hong Kong psyche can not be understated. Ningxia Choi Sum is famous to the point that when mentioning to my relatives in a cha chaan teng the night before I went on my trip, one of them piped up: “帶兩斤寧夏菜心返落來啦!” (Bring two catties of Ningxia Choi Sum back down here!).

Modern Ningxia, a northwestern Muslim region of China with no strong historical connections to southern Chinese cuisine, today is an important source of this staple vegetable widely used in South Chinese cooking despite being over 2,000 kilometers away. Choi Sum is so important to Southern Chinese cooking that when I lived in Canada, where the Chinese immigrant population historically skews Cantonese, non-Chinese people are quite familiar with Choi Sum along with Bok Choi. Yet, Choi Sum unlike Goji Berries isn’t even native to Ningxia.

Choi Sum needs relatively dry and cool (10-25 degrees) conditions to grow. These are the exact conditions South China experiences during the “winter”/dry season. So historically in South China, where farmers can have multiple harvests per year in the same field, Choi Sum was rotated in to be grown during the “winter”/dry season. While higher value crops are grown during the “summer”/wet season. As summer in South China is too hot and wet to grow Choi Shum. My older relatives noted how when they were young in Hong Kong just after winter they would eat a lot of Choi Sum as it was “in season”, widely available and cheap. Then in the early summer, Choi Sum stock in the Hong Kong wet markets would give way to other vegetables such as 豆苗 (Dau Miu) or be really expensive. This was long before national cold chain logistics were implemented and you ate what farmers nearby grew.

So how does Ningxia come in? The story goes that the various local governments of Ningxia, a poorer corner of China with limited (at the time) economic opportunities, were looking for ideas on “poverty alleviation”, particularly in rural areas. Given the land of Ningxia experiences a mix of **searches online** hemiboreal humid continental with cool arid/semi-arid desert climates, not much can grow there. Goji Berries, while VERY important for Chinese cooking and health, only has so big of a market; There are a lot of land and farmers in Ningxia. Officials identified that the agricultural industry in Ningxia can be further developed by growing Choi Sum in the summer with some irrigation. Essentially meaning Guangdong’s winter is Ningxia’s summer. With the national policy of toll free passage for trucks carrying fresh agricultural products, farmers and farmland in Ningxia can viably grow Choi Sum and have more markets to serve. This also frees up more productive farmland in South China to grow other things that the North can’t grow.

After some experiments introducing Choi Shum to various parts of Ningxia, several areas were found to be very suitable for its growth, then new local industry has taken hold. By 2024, Ningxia exports almost 80 semi-truckloads of Choi Sum to both the Yangtze and Pearl River Delta Megalopoli per day, feeding over 150 million people. I always wondered with the slow introduction and improvement of cold chain capacity of China Railways these past years if it is better to ship all this Choi Sum to the Yangtze and Pearl River Deltas via China Railways.

South of urban Yinchuan in Yongning County is a national hotspot of Choi Shum production. What is funny about Ningxia is despite the high volume of production in the area, local people still don’t normally eat Choi Shum. I guess it is because it is still viewed as Southern food. My Yinchuan hotel had Choi Shum as part of its breakfast and labelled it as Guangdong Choi Shum. However, I highly doubt it is from Guangdong instead of say, Yongning County 25 km south. Maybe they wanted to emphasize it is a Southern Chinese dish?

After the success of Choi Shum, local authorities are exploring if the microclimate around Ningxia can be used to develop a wine industry. Time will tell if this can be achieved. I had some Chinese red wine in Beijing back in 2017 and… I was not impressed. The terroir of that variety of wine needs work or something I don’t know. You can see a electronic billboard on the pedestrian overpass promoting Yinchuan’s wine industry on the top right below:

All this development was happening in tandem to large coal liquefaction plants being developed southeast of Yinchuan in Lingwu sub-prefectural city. One of the results is the flagship Ningdong Energy and Chemical Industry Base, containing the largest coal liquification/gasification petrochemical plant in the world. These plants, while more expensive to operate compared to normal petrochemical plants using crude oil/natural gas as a feedstock are also why China has been so resilient to the 2026 Strait of Hormuz crisis, by turning coal into plastics and fertilizer. So what did Yinchuan do with all this new agricultural and chemical wealth? Build a BRT of course.

Arrival to Yinchuan

I pick up my social media journey by taking a high speed train from Lanzhou. I take the Lanzhou Metro to Lanzhou West station to catch my train. My train was a CRH5G which is a special high/low temperature desert sand and alpine snow hardened version of the CRH5.

My train departs, leaving Lanzhou to use the Lanzhou-Xinjiang Railway passing China-EU freight trains along the way.

They really do need the CRH5G with all that hardening (IDK IP67?) there was a lot of blowing snow encountered by my trip.

Below is the countryside of Gansu Province. Gansu neighbors Ningxia in the south and is China's poorest province, with a GDP per capita of just under 7,500 USD. About the GDP per capita of Thailand or Azerbaijan. Gansu is trying to be one of China's literal "powerhouse" provinces with massive expansion of wind and solar generation capacity.

Note the amount of transmission lines below. All this power is sent to the rest of the country via China’s massive network of UHV (Ultra High Voltage), the high speed rail of electricity transmission, and also regular HV transmission lines. Like China’s high speed railway network, the scale of China’s UHV network has no other equal on Earth.

I enter Yinchuan City:

Yinchuan is the capital of Ningxia so the station is built with Islamic architecture and motifs. When I was there the Baotou–Yinchuan High-Speed Railway was still under construction but it opened a couple of months later in 23 December 2025. When I was there most trains came from the south via the Yinchuan–Lanzhou HSL and terminating at Yinchuan. I could imagine Yinchuan Railway Station is much busier today with through trains to Baotou, Beijing and beyond.

After walking across a large station square and park. I enter a xiaoqu where my hotel is located.

Below is the area around the Yinchuan Railway station. You can see all the “unnecessary” parks that surround this rail hub. Yinchuan Stations is located a little bit off from the center of Yinchuan’s urban area as Yinchuan successfully grew out west towards and beyond the station and its sideways T shaped urban area now largely surrounds the railway station. There is a farmer’s market on one of the empty lots near the Railway Station waiting for development. There you can buy 30 eggs for 15 RMB (2.25 USD).

After getting settled in I head out into the nearby BRT, which also serves the railway station. This BRT is my personal reason for knowing Yinchuan exists:

Yinchuan BRT

The Yinchuan BRT opened on 10 September 2012 and the “corridor proper” consists of a single 21 km east west line with 32 stations, operated by Yinchuan Public Transport Company. Construction of the BRT corridor took only four months. The BRT line zig zags along Yinchuan’s supergrid road network west to southeast. The system carries about 100,000 passengers daily, with a flat fare of 1 yuan and free one-hour transfers to 21 regular bus routes. The network of buses popping in an out of the BRT corridor forms a spine for Yinchuan. Yinchuan is shaped like a sideways T and Yinchuan BRT’s collection of lines do form the basic skeleton serving west and southern Yinchuan as shown below:

Today, the BRT network is operated with a fleet of mostly articulated electric buses and is an “open system” with regular non-articulated bus routes entering and exiting the corridor as needed. The BRT corridor proper has buses coming every 2-5 minutes most of the day. As such in 2025, public transportation modal share in Yinchuan accounts for 37% of motorized trips.1

For those true Chinese BRT heads, you might notice how the stations looked like Changzhou BRT stations in Jiangsu Province. This is no coincidence; Yinchuan’s BRT has had extensive technical assistance from the designers of the system in Changzhou. That also means that like Changzhou’s BRT, the design does not allow for an easy way to implement express stopping patterns.

The stations have platform screen doors and can berth 2-3 articulated buses, with sizing of each station depending on projected demand.

Access to BRT stations are generally fine for a median design but with the caveat that you need to cross a wide Chinese stroad in some places. At the busy Shangcheng (商城) or “Mall” BRT station the BRT station exit just dumps everyone onto a lightly marked road crossing that vehicles need to yield for pedestrians just from sheer volume of people crossing the street after alighting from a BRT bus.

South of the “Mall” the BRT station is named after, is a massive complex of indoor and outdoor lighting wholesalers.

I saw a model of outdoor light for sale only to see the exact same model in use at a park while walking back to my hotel the same day (after taking the BRT of course).

The actual “Mall” in question has a bizarre architectural scheme with tons of green painted exposed metal and small panel glass windows. Definitely a nod to Victorian Industrial Architecture.

Inside the mall there is some great food but mostly not local Ningxia food. There was even a mainland HK style Egg waffle chain. Of course I had to try it. Staff asked me were I’m from and I replied “Hong Kong”, she smiled and started making my order. She clearly knows the stakes. A wayward Southerner from the birthplace of the egg waffle has stumbled into this obscure part of China to test her egg waffle mettle.

“Let me know how it tastes.”, she said passing the hot lumps of cooked egg flour batter. It was good and that was communicated to her.

Anyways… The BRT line uses median lanes east Yinchuan Railway Station…

and but has curbside lanes west of Yinchuan Railway Station.

Despite the Yinchuan BRT operating for over 10 years there is still confusion of drivers entering the BRT lanes, realizing their mistake and slowly backing out.

In fact two days ago, a car has crashed into a Yinchuan BRT station. Luckily there was no casualties.

When the Yinchuan BRT first opened, the system used a fleet of (literally) green articulated Youngman JNP6181GC diesel buses. Also, note below how before the BRT lanes are hermetically sealed from general traffic lanes by traffic fencing. This is no longer the case. Today, fencing is reduced to a single row down the median of the road separating the two travel directions. Enforcement is done electronically to what seems to be great success, the BRT lanes are sanitized from private vehicles.

Today the Yinchuan BRT uses a fleet of fully electric 12m buses from Xiamen based Kinglong (the same model which also graces the excellent Xiamen BRT) and fully electric 18m articulated buses from Zhuhai based Guangtong.

Buses have electronic cameras with a live feed showing the sides of the bus beside the side mirror to further reduce blind spots, most likely in response to the rapid proliferation of electric scooters. They also assist with platform docking at BRT Stations.

Ⓑ Ⓡ Ⓣ

My favorite quirk of the Yinchuan BRT is the use of horizontal signal heads with three lamps spelling out Ⓑ Ⓡ Ⓣ . The lights are painted yellow too which made them look very North American. The three lamps are multicolor LEDs so all three letters in Ⓑ Ⓡ Ⓣ just change between red, yellow and green like any other traffic light to direct buses.

This is definitely not MUTCD (Manual on Uniform Traffic Control Devices) or Vienna Convention on Road Signs and Signals compliant.

The BYD Monorail

Yinchuan also used all the Choi Sum money to build the first BYD “Skyrail” monorail in a park, dubbed the very mass transit sounding “Yinchuan Skyrail Line 1”. This 5.7 km unidirectional loop route with eight stations circles around the Yinchuan Flower Expo Garden. It was the first public deployment of the BYD “Skyrail” monorail technology, now deployed at scale in São Paulo for their Metrô Line 17 (gold). As you can imagine a small loop line in a park is not useful. When I arrived, the system is currently shut down with everything blocked off and quiet.

Yes, this is a big white elephant project, Yinchuan should have tried to get BYD to build a real transit line.

My Thoughts on Yinchuan’s Transit?

This leads me to Yinchuan’s transit future, which lacks ambition. Yinchuan has about 2 million people and is a bit of a “linear city” despite being, as I mentioned before, a sideways T shape. The transit system is mediocre but the buses are quite frequent overall and are somewhat arranged in a grid that does successfully get you around. It certainly did for me most of the time. However, I found it difficult to get to key tourist sites in Yinchuan such as the Western Xia Imperial Tombs or Zhenbeibu Western Film Studio which are served by limited bus service.

I took a bus back from Zhenbeibu Western Film Studio but it required an interchange at the western terminus/depot on the edge of the city of one of the BRT’s feeder routes. These tourist sites are a little bit outside of Yinchuan City but could have had a bus every 20-15 mins that through runs into the BRT corridor like what is commonly done in South Korea.

The UNESCO Western Xia Imperial Tombs has rammed earth structures that date from the European High Middle Ages and is poorly served by Yinchuan City Transit. Again there should be a bus every 20-15 mins that through runs into the BRT corridor.

After the opening of BRT Line 1, Yinchuan was planning on building three more corridors but I could not find a map showing what lines were planned. More importantly, these three additional corridors were never built and it has been over 10 years. One BRT corridor is not enough to cover a Chinese city with 2 million population. Large parts of Yinchuan are still not served by the BRT or any other higher order transit.

While the BRT is an open system with buses outside the corridor entering BRT lanes with free transfers to other bus routes, there are some quite dense (or aspirational) areas in Yinchuan that do not have BRT service or any higher order transit for that matter. A new CBD in the north of the city centered on the twin 301m supertall Yinchuan Greenland Center complex doesn’t really have any compelling higher order transit. What is there currently can be basically summed up to a couple of bus routes that come every 10 mins or so. Yet the Yinchuan Greenland Center complex that anchors the development just towers over everything and you can see it everywhere.

Another area only partly served by the BRT is the current commercial area around People’s Square which is a few kilometers south of the Yinchuan Greenland Center. I would know because I took one of the regular bus routes between those two areas via Route 13, which comes every 12 minutes. The actual BRT corridor just misses the People’s Square area, running a superblock south as it zigzags to the southeast of Yinchuan. Below was me on B1 (the trunk only BRT route) passing just south of the People’s Square area:

People’s Square is surrounded by a bunch of cultural institutions which are then surrounded by office and high end residential and hotel buildings. It is a shame that the actual BRT corridor (with a ton of service) just misses this area. However, some bus routes that directly operate into the BRT corridor in the west do serve the area.

Here I passed by their main library, being curious I went inside and found the jackpot where I can learn Chinese railway engineering from first principles...

As noted above, this area is served by a frequent regular east west BRT feeder bus running along Beijing Middle Road. This bus, Route 101, runs down Beijing Middle Road until the BRT corridor zig zags up to also run on Beijing Middle Road. There Route 101 enters the BRT corridor. The bus is quite frequent and seems to come every 10 minutes or better. You can see how frequent the buses on this corridor are because the buses need to pull into the auxiliary road which doubles as a bike lane. This maneuver has very visibly polished a path on the road surface with their tires.

My Route 101 bus was pretty crowded too.

However, it would be nice if the BRT corridors were expanded to these aforementioned places. Maybe a sort of North South corridor to compliment the east west one that connects the new CBD in the north with the commercial and cultural area around People’s Square.

Part of the reason for the delay in expanding the BRT was that Yinchuan wanted to build a metro. The ever increasing social-economic requirements for a city to construct a metro system imposed by national authorities these past few years culminated in Yinchuan’s metro proposal to be formally rejected by the central government in 2024. What does that mean for Yinchuan’s transit future?

The rise of cheap ridesharing, EVs and electric scooters seems to be impacting transport planning at a local level these days. More annoyingly these modes are causing frustrating 3rd world-esque traffic problems that, based on my experience in China, didn’t exist 5 years ago. For example, the busy Huaiyuan Night Market is served by the curbside section of the BRT, yet is clearly a huge generator of non transit traffic problems with idling vehicles everywhere.

This is reflected in the some of my “perceived” mode choices people are observed using to access the Huaiyuan Night Market commercial area. The area is served by a large and well used BRT station but there are also tons of illegal chaotic curbside electric scooter parking, queueing of taxis, idling rideshare vehicles because people can’t be bothered to show up for their ride on time. The widespread demand for these new modes didn’t exist 5 years ago and no doubt Chinese local urban management systems have not kept up with their growth. If transit offerings were better along with much more aggressive enforcement no doubt these “big city urban diseases” would be mitigated or even eliminated.

The rise of these new modes does seem to dampen or at least confuse the direction of transportation planning for medium sized cities in China like Yinchuan. Is there sufficient demand for a metro today? Light metro? Should we double down and build more BRT? Call back BYD to build a monorail now that they are too busy burying the world in cheap, well made, full feature EVs? There have been some populist pushback against more BRT lines by motorists. However, for Yinchuan I think there is a glimmer of hope. Transit ridership remains high, most likely due to what I perceive as pretty tight enforcement of electric scooter behaviour and licenses. This is unlike the scooter transport anarchy chaos that happens in the South these days. Most scooters in Yinchuan seem to have licence plates for tracking violations and people still cycle around on their own bikes.

This means there are still some barrier/process for someone to make a decision and invest to get around Yinchuan via a scooter. Unlike South China where everyone and their dog can just buy an electric scooter with no plates/license, illegally modify with another motor and battery, immediately use it to run five red lights at 60kph then park anywhere for free. This led to a large population in the South to dump their bikes and transit passes for electric scooters these past 5 years.2 Due to this, there is a noticeable erosion of motivation to improve public transit that is not a metro, which an electric scooter, even breaking all the traffic rules, can not compete with, in these areas. Yinchuan seems to have somewhat avoided this scooter dependency induced death spiral as the buses are still generally frequent and well used. Yinchuan’s local authorities wisely should build on that small transportation policy success.

So thanks for reading this post of my partial Journey to the West. Like, share and subscribe for more transit oriented stories of my travels in East Asia.

Or drop me a Ko-fi if you like what you read.

I will be using some pictures of that trip in this article to show the conditions before and after the opening of Line 22 Phase I North on 29 December 2025. I was not able to reach Guangzhou for the opening day due to some work obligations, but I was going to meet a good Malaysian friend in Guangzhou on New Year’s Eve a few days later. So on New Year’s Eve, hours before seeing them, I went up to Guangzhou to take a look at this new late 2025 Guangzhou Metro addition.

Like last time, I arrived in Guangzhou from Hong Kong in the morning of 31 December 2025 via high speed rail. As I do on every visit for a new metro opening, I took a picture of the same Guangzhou Metro map that greets me each time at Guangzhou South Railway Station. This map had been replaced for the fourth time in just over a year due to the frequency of major metro line extension openings. This time the map was updated with the star of the show, the Line 22 Phase I North Extension:

Line 22, the Express Relief Line for Line 2

No story of a Guangzhou Metro line is complete without a story of the planning failures and the resulting congested system. Line 2 is a north south metro line that carries over a 1 million trips per day. Back in the early 1990s, when SYSTRA/MVA assisted in the design and masterplanning of Guangzhou Metro Lines 1 and 2, much more intelligent leadership was in charge at the time. Guangzhou Metro Lines 1 and 2 (parts of Line 2 are now Line 8) were semi-reasonably designed to use fairly high capacity six car Type A trains.

When the SYSTRA/MVA designed Line 1 and 2 masterplan was completed, ridership in the early years on both lines was “underperforming” with footfall being half of what was projected. Pundits criticized the Guangzhou Municipal Government for overbuilding and waste. As a result, later key trunk urban Guangzhou Metro lines such as 3, 4, 5, 6 and 7 all were assumed to not need high capacity and “optimized (derogatory)” to be built as 3-4 car light metros. This was done on the premise of the deeply flawed 小编组 高密度 or “small group (trains), high density (frequency)” metro design paradigm taking over China at the time. This paradigm gave rise to notorious examples, such as the disastrous Shanghai Metro Lines 5 and 6.

Courted by the Institute for Transportation and Development Policy (IDTP), “less important” Guangzhou transportation corridors started dabbling in building cheaper surface Bus Rapid Transit (BRT) corridors on Zhongshan Avenue (BRT built), Huangpu Avenue and Dongfeng Road (both to be served by Line 13).1 This austere era of light metro + BRT planning, to put things in technical terms, royally screwed Guangzhou with massive congestion upon the opening of each light metro or BRT line. These undersized light metro and BRT lines struggled to serve what would become one of the largest and densest cities in the world.

In hindsight, a massive high performance metro system would have been more suitable. While I think the South Korean duplicate BRT + Metro corridor model can be good, I believe what Guangzhou needs today is an all-hands on deck approach to expanding the metro and regional rail. The issue is, in 2026, with the rise of electric scooters, Chinese dominance of EVs and higher affluence of the general Chinese population, advocates no longer have the political tools to make an ambitious BRT program happen.

Luckily, due to early “overbuilding” (or, more accurately, proper vision by leadership and SYSTRA/MVA) Line 2’s higher capacity full sized metro trains, while congested, are not inhumanly so (see below). The same cannot be said about the walking disasters of Line 3, 4, 5 and 6 that did get built as light metros. Today, Guangzhou continues to grow and has already become one of the most populated cities in the world, so the over 20 year old Line 2 will need some help. Just as Line 18 is the relief line to the horribly congested Line 3, and Line 13 is the relief line to the slightly less horribly congested Line 5, so the north south high speed metro Line 22 can come in, as the express relief line to Line 2.

Line 2 in Mid-December before the opening of Line 22 Phase I North

A major notable role of Line 2 is it connects Guangzhou/Foshan’s main high speed railway station, Guangzhou South, with the old core of Guangzhou. However, being a regular 80kph 1-2km stop spacing metro line it takes 40 minutes to get from Guangzhou South to the old core of Guangzhou. The remote location of Guangzhou South HSR Station, like many of China’s suburban HSR stations, didn’t really stop everyone from taking high‑speed rail, but the +40 minute metro ride to the city center is not ideal. This is where Line 22 comes in, during planning known as the 南站快线 or South Station Express Line. Line 22 is an express line that will reduce the travel time from the city center of Guangzhou to Guangzhou South Station. The opening of Line 22 Phase I North to the edge of the city center provides a minor glimpse of the power of an express line. Despite Line 22 currently not being complete, not penetrating into the city center and the Phase I North section having some design issues (more on that later); Traveling the 15 km from Guangzhou South to Line 22’s current terminus, Fangcun, takes 19 minutes. Traveling on Line 2 the same distance (15 km), from Guangzhou South to The 2nd Workers’ Cultural Palace Station, on the other hand, takes 26 minutes.

Regardless, upon arriving in Guangzhou South HSR Station on 31 December 2025, I swooped down to Line 2’s boarding platforms and (unscientifically) observed the extent of the diversionary relief effect the days old Line 22 was having on Line 2; it now looks better!

Before the opening of Line 22 Phase I North, you might need to wait for 1-2 Line 2 trains to pass as shown by my video taken two weeks earlier:

However, given that there are literally over 1,000 high speed trains departing Guangzhou South Railway Station everyday, these days the metro platforms of Guangzhou South Station are still quite congested with people pouring in from Line 7 on the platform face across from the boarding platform of Line 2:

I headed to the Line 22 side of the station, and the signs of diversionary relief were apparent; there was now an endless stream of people coming out from the regional railway side of the station complex which now hosts Guangzhou high speed metro Line 22 (which in case you have not noticed was extended north a few days ago) and platforms for Guangzhou’s regional railway system:

On the northbound Line 22 platform it is nice to see such heavy use days after opening. Every person in this video would have crowded themselves onto Line 2 a few days ago:

Riding the new Extension

I boarded a northbound Line 22 train and took in the sights of the days old new extension. The entire Guangzhou Metro high speed metro Line 22 Phase 1 was constructed for a purchasing power parity (PPP) adjusted cost of about 220 million INT$/km. Line 22’s unit cost is less than the unit cost of construction for Toronto’s Finch West surface road median suburban tramway.

One of the issues I observed with Line 22 North Extension is the rather tight stop spacings for a high speed metro, possibly due to local pressure to get a metro station near their homes. This leads to the strange turns and curves all over Line 22 to “connect the dots”. This is not without precedent, as similar examples exist on the same line south of this new Line 22 extension. The former northern terminal of Line 22, Chentougang Station, is only ~3km away from the adjacent Guangzhou South Station. This is also an unusually close stop spacing for a high speed metro. However, Chentougang Station has some justifications for its existence:

• possessing overtaking tracks for express services

• is located beside Line 22’s Chentougang depot, enabling the value capture topside developments to be served.

• is planned to interchange with the under construction Foshan Metro Line 4.

The interchange with Foshan Metro Line 4 is key. Today, only Phase I of Foshan Metro Line 4 is under construction. This 55.2km 33 station Foshan-only section of Line 4 runs right up to the Guangzhou/Foshan city border, leaving a small 3km gap from the Line 4 Phase I terminal at the border to Chentougang Station of Line 22.

However, the opening of this new Line 22 extension continues this “tight” stop spacing that is not characteristic of a high speed metro. There are two more stations between Chentougang and Xilang, the next major interchange station on Line 22, which are Nanpu West and Nanjiao Stations. The existence of these stations is keeping the stop spacing in this section of Line 22 to around 3 km. In addition, there are numerous curves on this section of Line 22. Looking at a to scale map of the new extension, I realized why: when designing the alignment of Line 22, Guangzhou Metro needed to avoid Foshan City.

Nanpu West and Nanjiao Stations are the two new non-interchange stations added as part of Line 22 Phase I North. Both are quite dubious. Nanpu West is serving a bunch of residential areas on the west side of Nanpu Island while Line 2 serves the east. So Nanpu West kind of makes sense, since you might as well serve that part of the isolated island while Line 22 jogs around Foshan city.

The inclusion of Nanjiao Station in the area is extremely questionable. Huawei Station on the recently opened Line 10 is only a couple hundred meters north. Huawei Station’s raison d'être is to serve as a meeting point between Line 10 and Line 10 branch into Foshan City as well as be an interchange point for Foshan Metro Line 11. Also directly north of Huawei Station is Line 10’s depot with topside development. Foshan Metro Line 11 is a radial suburban metro line similar to Guangzhou Metro Lines 14 and 21. Like Guangzhou Metro Line 14 and 21, Foshan Metro Line 11 starts from the monster Guangzhou Metro Line 11 loop line and radiates out to the suburbs, in this case Shunde, Foshan in the south.

Huawei and Nanjiao Stations are not close enough to be interchanged with by a single Foshan Metro Line 11 station. At the same time, they are too close to be served by two distinct Line 11 stations. Foshan Metro Line 11’s suburban metro nature requires wide stop spacings to allow the GoA4 fully automated trains that reach up to 100 kph to stretch their wings. 500m is not an acceptable stop spacing for this type of line. Therefore, the approved north south Foshan Metro Line 11 needed to prioritize interchanging with either station. Due to historical planning quirks, Foshan Metro Line 11’s current design assumes an interchange with Line 10/ Line 10 Branch. Following this plan, Line 10’s Huawei Station was built with structural provisions for Foshan Metro Line 11 to interchange with it.

The result is Line 22’s Nanjiao Station has no planned lines to interchange with. Combined with the closeness to Huawei Station, all Nanjiao Station does is slow down Line 22 services. 100% the Line 22 express services will need to skip this station. Heck close the station if you have to. I would be more sympathetic if Nanjiao Station was designed to run close enough to interchange with Line 10/Line 10 Branch and Foshan Metro Line 11. More connectivity to the express Line 22 is always welcome, but that is not the case in the current plan. This station also does not have any overtaking loops for express trains to pass all stop trains, making Nanjiao Station today extremely useless even if the surrounding factory sites are redeveloped.

All these factors created a design for this new section of high speed metro Line 22 with tight stop spacings and numerous curves, all of which are somewhat unfit for a high speed metro. During my ride I observed speeds hovering around 120kph, topping out at maybe 130kph in sections. Hopefully, when they introduce express services to bypass stations, average speeds on Line 22 will increase. I also observed that all four stations in this new extension do not have any overtaking loops, which is disconcerting.

The extraneous Nanjiao Station is nothing too special with the bronze theme. Again, this station should have been dropped during planning or moved closer to interchange with Huawei Station or built with overtaking loops for express trains; Completely useless.

After looking at that disappointment of a station, I then arrived at Xilang Station whose platforms surprised me with the striking orange accents everywhere:

The concourse then transitions from a single color scheme to a two color design, with the addition of yellow accents:

Xilang Station is a four line interchange station between Lines 1, 10, 22 and the Guangfo Line (Foshan Metro Line 1). I reviewed Line 10’s opening a few months ago, whose Xilang station box is on the east side of the legacy at-grade Xilang Station Line 1 platforms. As noted in a previous article, Line 22’s days old Xilang Station box (which I was standing in) was on the other side of the Xilang Station Line 1 platforms (west). The underground station boxes of Line 10 and Line 22 sandwich the at-grade Xilang Station Line 1 platforms. To transfer between Lines 10 and 22, several tunnels have been mined under the legacy at-grade Line 1 platforms:

As noted in my previous article on Guangzhou Metro extensions that opened near National Day, I investigated the state of publicly accessible areas of interest for the Line 22 project. This time, about two weeks later, I returned to those same areas to see the changes post-opening. One of those areas was the transfer concourse of Line 10, where before the opening of Line 22, the transfer tunnels were boarded up. So I walked through the set of now completed tunnels mined under the Line 1 platforms to get from Line 22’s concourse to Line 10’s concourse:

I hopped back on the Line 22 and headed up to Line 22’s temporary terminal of Fangcun, another four line interchange station, this time with Lines 1, 11, 22 and the under construction future Line 28:

Note that Line 1 was mentioned again for this interchange station. That is because this section of Line 22 operates as an express bypass for this section of Line 1.

Like the Seoul GTX, Guangzhou’s high speed metro lines run very deep under the city center. So at Fangcun Station I headed up a long escalator into a dramatic transfer concourse filled with uncles and aunties taking pictures of the days old structures:

The columns on the south side of the Line 22 station concourse are quite special with an array of square indentations pressed into the concrete and painted with dark colors to really give a contrasting 3D effect. What is super fascinating is how the indentations fade out in size and depth the higher up the column your eyes go, drawing your gaze to the ceiling above. Throw in the cove notch chamfers on each corner of every column and you have a very striking effect.

It is pretty common in Guangzhou and Shenzhen for major interchange stations to have two sets of elevators from concourse to platform for redundancy and capacity.

The Line 22 transfer concourse is actually two levels with a second floor supported by a steel truss bolted to the ceiling of the station box to make for a columnless structure. This second floor is the network of unpaid passages connecting all the exits, alot of them not yet completed or awaiting connection with new Bai’etan CBD office towers above the station.

During my time here, I saw a lot of transfer traffic between Line 22 with Line 1 and Line 11. Again, every person in these pictures is no longer taking Line 2 which is a win for everyone. What is not shown in the videos and pictures is that the RAZER RGB LED highlights throughout the ceiling of the station very slowly shift into different colors. As you can tell from the media I presented here, I’m in the blue phase.

I headed to the surface and saw the under construction Bai’etan CBD and construction sites that are no doubt now preparing to build Fangcun’s Line 28 platforms. I then headed back down into Fangcun Station.

I explored the rest of Fangcun Station, which upon closer inspection revealed a cornucopia of unfinished passages. Some of these passages lead to adjacent under construction TOD developments, while others lead to unfinished new exits or to the reserved interchange station structures for the up and coming east west high speed metro Line 28.

View from the unpaid zone on the second floor of the Line 22 Fangcun Station concourse I mentioned earlier:

I headed up to the Line 1 & Line 11 concourse where it was clear the newer Line 11 and Line 22 concourses were sandwiching the legacy 90s era Line 1 station box, with access passages punched through Line 1’s station box on both sides.

After looking at Line 22 Phase I North, I headed on to another series of crowded trains to another location… Where? Stay tuned.

So thanks for reading this late post. Like, share and subscribe for more transport news and analysis about the world’s fastest growing metro system, which still grows faster than I can type.

Or drop me a Ko-fi if you like what you read.

After the Expo, JNR retained the service and named it Shin-Kaisoku 新快速 (literally “New Rapid”). The accompanying English translation is “Special Rapid”, again for historical reasons. The impact of the Special Rapid service on the Tokaido/San’yo Main Line during Japan’s boom‑era urbanization of the ’70s and ’80s cannot be overstated. The regional economic integration of Kyoto, Osaka and Kobe or Keihanshin1 was greatly assisted by the JNR Special Rapid Service on top of the major Kansai private railways.

Special Rapid services were gradually extended to operate west and east beyond Kobe and Kyoto, switching to run as an all-stop local and serve outer suburban or even ex-urban areas of Kansai such as the western Hyogo Prefecture and northern Shiga Prefecture. The core of this express regional service ran on what today would be dubbed the JR-West Kyoto Line and Kobe Line.

Special Rapid trains run on a 15 minutes clockface most of the day2 and new rapidly takes you from Osaka to Kyoto or Kobe in 30 minutes or 20 minutes respectively on a pair of express tracks of the JR-West Kyoto and Kobe Lines. On the parallel slow tracks, all‑stop local services make the same journeys away from Osaka in 40 minutes to Kyoto and 35 minutes to Kobe, respectively. Today it is an old friend of mine that greets me every time I come to Kansai.

While Japan was reaping the benefits of its investment, across the Pacific, Toronto was dreaming of similar modernisation.

Across the (Bigger) Pond

During this same era, Toronto was proposing its own improvements to the un-electrified Government of Ontario Transit or GO Transit network. Throughout the 1970s, during Toronto’s (at the time) single lost decade of transit expansion, there were other repeated and failed attempts to build a better regional transit system (GO-Urban). Of note is the SuperGO plan, which called for the electrification and implementation of frequent locomotive hauled bi-level electric mainline services as frequently as every 4 minutes (or less!). Under this plan, a pair of dedicated electrified tracks for GO Transit services paralleling the existing pair of Canadian National (CN) tracks would be built in an expanded quadruple track corridor of what would today be the Kosei-sen and Koto-sen… I mean Lakeshore East Line and Lakeshore West Line.

This track expansion would allow for outer suburban “mid-distance” express trains to provide near non-stop transport from Oakville to Pickering and beyond every 10-15 min. While “short-run” trains would provide parallel local services to the inner suburbs. As this practical conventional SuperGO plan was limping along, another distraction was brewing in the background.

小编组，高密度 Small Groups, High Density

In the early 1980s, the provincial government of Ontario announced the “Government of Ontario Advanced Light Rail Transit” (GO-ALRT) plan, a regional light metro system competing with the SuperGO concept. Instead of SuperGO’s large electric locomotive hauled trains, GO-ALRT proposed using small frequent electric trains propelled by linear induction motor (LIM) traction. Very Guangzhou Metro Line 4 coded. The technology was advocated by the government owned Urban Transportation Development Corporation (UTDC). The GO-ALRT system as proposed in 1982 was estimated to be fully built out 20 years later, meaning a 2002 completion date. As an aside GO-ALRT’s name must have been the forerunner of the contemporary East Asian practice of calling light metros, light rail.

The UTDC was redesigning the system for use in the GO‑ALRT plan in tandem with the ongoing Vancouver SkyTrain and Scarborough Rapid Transit (SRT) projects. All three projects used modified designs initially developed for the failed GO-Urban plan. The SkyTrain and SRT used the Intermediate Capacity Transit System (ICTS) Mark I light metro vehicle. The initial design for the GO‑ALRT proposal, unlike the ICTS Mark I, called for light three car units. Each GO‑ALRT unit is 36 meters long and just over 2.5 meters wide. Units can be connected together into 2, 3 or 4 unit consists (6, 9 and 12 cars) to form a larger train that is between 72 meters to 144 meters long.

Over time, as the GO-ALRT project evolved, the physical dimensions of each train car were revised to be increasingly larger. I suspect this was done to guarantee to reach a required peak hour sectional throughput capacity. Eventually, each car was proposed to be roughly 23 m long and 2.8 m wide, organized into two‑car married pairs. Coupling two or three pairs would form a 4‑ to 6‑car train, 92 m to 137 m in length. Instead of linear motors, propulsion would now be provided by conventional rotary motors in each car, an Electric Multiple Unit (EMU) if you may. A GO-ALRT line was planned to operate at high top speeds, up to 120kph, to take advantage of wide stop spacings as expected for a regional (no longer light) metro system. Signaling was proposed to use the original SELTrac system developed in Canada to allow EMU trains to arrive every 2 minutes.

By the end of the 1980s, both SuperGO and GO-ALRT plans were abandoned due to lack of interest, prefacing the continued lost decades of Toronto transit expansion. The SELTrac technology was sold and resold to several European firms, eventually evolving into the modern communications‑based train control (CBTC) solution now deployed across thousands of kilometres of urban railways worldwide. ICTS technology was used to build a few systems in Beijing and Kuala Lumpur and ended up in the hands of Alstom.

The Start of a Better GO Transit

It was the late 90s, Japan had entered the Heisei Era and its asset price bubble burst, beginning their own Lost Decade. At the time there was uncertainty over the need to pluralize the term. More importantly in 2000, in case anyone had not noticed: GO-ALRT was never implemented in Toronto. The Lost Decade of Toronto transit expansion had become the Lost Decades of Toronto transit expansion. However, glimmers of hope began to appear after 2002. The GO Transit Rail Improvement Program (GO TRIP) was announced by the provincial government and construction quietly began a few years later. GO TRIP was what was then a massive commuter railway improvement programme to increase the capacity and reliability of the GO commuter rail network as well as providing some limited off peak and counter peak service. The program involved:

• adding more tracks, such as a third track on the Lakeshore East and West Lines;

• grade separating rail junctions in several places where GO Trains and freight trains cross each other such as the Snider Diamond on the then called Bradford Line, West Toronto Diamond on the then called Georgetown Line or the Hagerman Diamond on the Stouffville Line;

• later on, outright purchasing ownership of tracks from private freight companies (mostly Canadian National Railway) like parts of the Weston, Oakville, Newmarket, Bala and Kingston Subdivisions.

The actual improvements to GO Transit commuter rail service directly enabled by GO TRIP were modest but the infrastructure (sometimes literally) laid the groundwork for potentially greater things such as….

Réseau Express Régional

GO Regional Express Rail (GO RER), the follow‑up, more ambitious regional rail program, was announced in 2008 as part of the “Big Move”. The “Big Move” was a sweeping regional public transportation capital expansion programme for the Greater Toronto and Hamilton Area that attempted to end the lost decades of Toronto transit expansion with a collection of subway, LRT, BRT and regional railway projects. With the Big Move, a new regional transportation agency for the Greater Toronto and Hamilton Area was formed to manage and plan everything, called Metrolinx. Metrolinx, the Big Move and GO RER revived the conversation on electric regional trains for Toronto.

By the early 2010s, Torontonians were scoffing at the high cost of subway expansion as they tore each other apart in the Subway vs LRT debates, prolonging the lost decades of Toronto transit expansion. In hindsight, the costs for building subways quoted at the time, even after adjusting for inflation, were still incredibly cheap. This stands in stark contrast to the world record breaking numbers in 2026, when Toronto was infected with a severe case of late stage North American transit construction cost disease. However, the GO RER plan used surface running electric trains on existing at-grade corridors which have much more modest construction costs but can still offer extremely high capacity…. see the JR-W Kansai urban network. This made the GO RER program a very logical choice to prioritize from a technocratic perspective.

GO RER proposed for most of GO Transit’s regional train lines to operate all day in both directions and at frequencies of 15 mins or better, creating an “urban network” if you will. The plan was a big departure from the existing typical North American peak period, peak direction, diesel commuter rail operations GO Transit is currently running but enhanced by the GO TRIP programme. To achieve this intense regional railway service, GO Trains were proposed to be electrified and Metrolinx even teased the use of EMUs, not unlike systems in Europe and East Asia.

Sadly, as “logical” as the GO RER programme was, the hype surrounding it dissipated as quickly as it appeared. The initial electrification report published for the GO RER programme lacked ambition and downplayed recommendations for transitioning to an EMU fleet. By 2011, the big plans for electric regional express rail fell by the wayside of the mainline as the Subway vs LRT debates continued to roar past in parallel. GO RER however, was not cancelled and still clinging to life as major upgrading projects were still underway that felt like GO TRIP on steroids. Projects such as the ambitious Georgetown South Project, which expanded a three to four track surface Kitchener Line corridor into a six to eight track grade separated corridor in an urban setting that would not look out of place in Japan. However, there was minimal progress on other supporting infrastructure such as electrification and re-signaling needed to achieve a service worthy of the RER name as trouble was brewing on the horizon.

But a Momentary Distraction…

The 2014 Toronto mayoral election reached another fever pitch of transportation expansion debates in a city scarred by such discourse. John Tory was a major candidate and sold the public on an upgraded electric mainline service on the Stouffville and Kitchener Lines dubbed SmartTrack. This transit plan, like all Toronto transit plans, was highly controversial in concept, design, feasibility and funding model. The plan’s hope was that getting Torontonian Bloor-Danforth Subway (red line below, I don’t know why it is red it should be green) commuters to ride a cheaper to build alternative surface electric trains would provide relief to the meagre subway system and avoid the need to build a new downtown subway line. Regardless, apart from advocating for more urban infill stations in the City of Toronto and calling for (maybe) trains every 5 mins during rush hour, it was not clear how SmartTrack was any different from what was being proposed by the existing half dead GO RER program everyone had forgotten about at the time.

For the cynical, SmartTrack did its job: John Tory was successfully elected as Mayor in 2014 and the UK-based consultant that quickly developed the SmartTrack plan was handsomely compensated. During his first tenure as Mayor, Mayor Tory basically did nothing to move forward his landmark transit plan on which he had staked his election campaign upon. John Tory was reelected twice in 2018 and 2022. In each subsequent term SmartTrack was further de-scoped and also fell by the wayside. The collateral damage? Another momentary distraction from actually expanding Toronto’s non-expanding transit system. Today, SmartTrack lives on as a corpse consisting of only three new surface infill railway stations costing over 1.7 billion CAD in total. For those counting at home, Wuhan built a new over 10km fully elevated 10 station city center light metro line for the same price:

GO RER 2: Just called GO Expansion now

In 2017, the drama around SmartTrack was winding down and the reinstatement of the first few Transit City LRT lines cooled the Subway vs LRT debate. A second revival of the GO RER program was announced, now named GO Expansion, in a bittersweet moment for Toronto transit watchers as the quite nice GO RER name was dropped. This also raised doubts about the level of ambition on the project. If I were to be cynical, this revival was conveniently announced in the lead up to the 2018 Ontario Provincial General Election. Regardless, GO Expansion was now placed as a higher priority as votes matter more than actual transit, with early off corridor (Off-Cor) enabling works confirmed to start at the end of 2017. At the same time, Metrolinx started the procurement process to find a consortium to handle the large task of operating the existing services and upgrading the infrastructure to enable GO Expansion’s overarching but still vague vision of an electric regional railway network. At around the same time as the announcement, I joked with friends at work at an engineering consulting firm, roughly saying:

In early 2022, Metrolinx selected ONxpress Transportation Partners to implement GO Expansion. ONxpress is a consortium of Spanish and Canadian constructors plus Alstom and most famously Deutsche Bahn, the German national railway company. My guess that GO Expansion would look like the Nürnberg S-Bahn is looking quite realistic. By late 2022, Metrolinx announced the vision ONxpress Transportation Partners proposed for GO Expansion. It was an ambitious vision far exceeding the initial promise of “two way service at least every 15 min”. Key electrified GO lines would each have frequencies of 3 to 8 minutes (20 to 7.5 trains per hour) during peak periods at stations served by express and local services. This vision exceeds what was originally promised and is drifting towards the Berlin S-Bahn in scale. Though interestingly the vision also made no mention of the use of EMUs. If implemented, this ONxpress vision would give Toronto one of the best regional railway systems in the Americas and be on par with a large electric suburban system in Europe. Things were starting to look good for Toronto electric regional rail, and quickly too.

Fast forward to the mid-2020s, and the GO Expansion Program seems to be experiencing a death by a thousand descopes and is on the verge of collapse. As shown in Reece Martin’s blog posts on Next Metro and video, along with a constellation of newspaper articles published by Jack Hauen on the Trillium and Andy Takagi on The Star. The contract with the consortium tasked with operations and delivery, ONxpress, was terminated. Also terminated with it was the big vision for world class electric regional rail. Scope was drastically decreased, now with Metrolinx internally discussing a Minimum Viable Product using an apt, if unintentionally ironic, acronym of MVP. The proposed MVP would only partially electrify the Lakeshore Lines by… 2038. Proposed quadruple tracking work on remaining three track sections of the Lakeshore and Kitchener Corridors was postponed. Ditto for when any improvement works on the other lines will be done, as they have been now relegated to a mystical Phase 2.

Remember, this current incarnation of improving Toronto’s existing commuter rail system was announced as GO TRIP in 2002. The first Chengdu Metro project was approved in 2004, the first line started construction in 2005. Today, the Chengdu Metro has grown to become (as of writing) the fourth longest metro system on the planet with 664 km of route and is expected to reach over 1,000 km of subway lines by 2038. Only then will Toronto hopefully be able to partly upgrade one GO Transit line into (maybe?) a 3 track electric traction line.

It now increasingly seems that GO Expansion’s current ambition is to turn GO Transit into a slightly better NJ Transit. I think part of the reason for the demise of the GO RER/GO Expansion project is the lack of a clear vision. As much kudos as I will give to ONxpress for its ambitious vision, it was still quite vague on what exactly the service would look like. By the time that vision was almost fully articulated, no-one wanted to institute the change to accommodate it.

What Now?

It seems to me that the GO Expansion Program has devolved into a new sort of tabula rasa. This provides me with an opportunity to wax poetic about an alternative vision of an electric regional railway network for the GTHA or the even wider Greater Golden Horseshoe. I am not the first to look into this. Dr. Jonathan English has extensively written over the years on the need to leverage the GO Transit network into a regional surface rapid transit system. The latest culmination of his vision is the Trillium Regional Railway Network Plan developed during his tenure as the Director of Transportation Policy at the Toronto Board of Trade (TBOT). A local transportation planner, Ontario Traffic Man, also sketched out an ultimate vision of what they think GO Expansion should look like. My meagre contribution to the “yet another transit plan for Toronto” discourse is to imagine something in between a “Nürnberg S-Bahn with Berlin’s electric locomotive hauled Regional-Express network thrown on top and a full blown JR-East Greater Tokyo network.” in scale for Toronto. That something looks like the JR-West Kansai regional network.

While everyone fawns over JR-East’s massive Kanto (Tokyo) regional network, I think most cities (not only Toronto) should be inspired by what is done in JR-West. Why? Because JR-West’s various urban networks are realistic for a large non‑Japanese city to implement, with a bit of ambition. JR-West still has some impressive infrastructure like a long quadruple track corridor (Sanyo/Tokaido Mainline), a cross city urban tunnel (JR-W Tozai Line) but also some great operational design to make the best use of what they have (see operations on the JR-W Hanwa and Yamatoji networks) or passenger/freight interactions (Sanyo/Tokaido Mainline fast tracks). The great branding and organization of each “city/urban network” in Kansai, Okayama and Hiroshima is the cherry on top. The communication of how individual lines and services work together in corridors is much better executed in Kansai than in Kanto (Tokyo).

Contrast this with JR-East’s Kanto regional network in Tokyo. JR-East clearly had the luxury of throwing money at any problem to create vast multi-track corridors segregated for each fast, slow and freight service. Tokyo today has massive surface multitrack railway corridors sorted by stopping patterns that reach a combined 150,000-200,000 people per hour per direction. All these lines, even mid-distance ones like the Ueno-Tokyo or Sobu Rapid/ Yokosuka services, use massive 300m long trains coming around 25 times per hour and essentially operating as an overgrown metro service.

Trains, even the mid-distance regional ones, have a high density of doors and largely metro style longitudinal seating. Tokyo’s regional railway offerings, from a passenger experience perspective, on the Ueno-Tokyo, Joban/Narita Rapid, Chuo Rapid, Sobu Rapid/Yokosuka and Yamanote Freight Line services are spartan and force a lot of standees. At the same time other countries (and Japanese cities) offer many more passenger amenities, seating and comfort for such long outer suburban commuter services. Not to mention JR-E’s habit of deemphasizing intercity limited express runs during rush hours for more commuter capacity (see Chuo Rapid) which may not be a luxury that other cities have.

Almost no city in the world needs this capacity or requires such sacrifices to passenger comfort as well as reductions in freight and intercity operations in a regional rail system to serve urban and suburban roles. It can be very hard for cities that do not have existing surface railway right of ways to leverage or simply the urban mass to even envision a grand regional network of Tokyo’s scale and feel. This makes it hard for advocates to even sell such a vision to the layman or even political leaders. Perhaps the surface railways of São Paulo, Mumbai, Delhi, Kolkata or Jakarta could really take a look at the specifics of what Tokyo did. You know, cities with populations well exceeding 20 million people that have five or more existing multi track radial surface rail corridors ripe for upgrading in a five directions of commuting operations strategy kind of way like Tokyo. All they need is ambition.

Other massive +25 million population cities with limited legacy railway corridors like Cairo, Lagos, Shanghai, Guangzhou and Karachi are better off building a giant overgrown, greenfield, master planned metro system with express high speed metro lines. In other words the opportunities to build something like Tokyo’s JR-E network for most cities in this day and age seem farfetched. This no doubt drives a lot of skepticism and impedes buy-in if Tokyo’s regional railway system is brought up as an example in most cities.

Toronto, Canada’s largest city, is neither as big nor as dense as those aforementioned juggernauts. However, Toronto does have numerous broad legacy multi-track surface railway corridors. Yet I feel that the regional railway system of the Japanese region most similar in size to Greater Toronto, the Chukyo (Greater Nagoya) metropolitan area, is not ambitious or inspiring enough. Greater Nagoya’s regional rail system basically can be described as Meitetsu with the table scraps of the respective JR Tokai and Kintetsu networks. When you dive into the service offerings and infrastructure of Greater Nagoya, you see no integrated quadruple track operations with designated fast and slow tracks and surprisingly low frequencies in some areas. When you look at the “table scraps” of the respective JR Tokai and Kintetsu networks that serve Nagoya, there is some quite intensive service but nothing really inspiring.

Yet, I believe Greater Nagoya’s regional system is not aspirational enough for the Greater Toronto Area and Hamilton area. Yes, I’m discounting the Meitetsu network as Toronto does not really have any infrastructure, operating environment and/or service architecture that is analogous to it. Meitetsu is basically an isolated Munich S-Bahn consisting of many branch lines feeding into a single common through city track pair… on Japanese drugs. Even the stations on the common dual track trunk such as Meitetsu Nagoya have Spanish Solution platforms to maximize throughput of the trunk like the Munich S-Bahn. Not many parallels can be found in Toronto’s existing infrastructure.

The populations of both Greater Nagoya and Toronto urban areas are similar, but Toronto’s existing mainline rail infrastructure is much more similar to JR-W Kansai Urban(e) Network. Toronto already has or had proposed a few cool integrated quadruple track corridors as part of GO Expansion. Nagoya on the other hand does not really have any such examples. Again to put the JR-West Kansai network’s scale and infrastructure in perspective, the network essentially operates as one long through running quadruple track corridor (Lines A/B/C), an underground city center tunnel (Line G/H sometimes Line A) and a loop line with a ton of branches feeding into it (Lines O/P/Q/R/S/T).

Most of the JR-W’s Kansai network infrastructure (everything but the core of Line A) consists of two track lines that require careful planning and operation on a mix of freight, local, express and regional passenger services. It is common for different service types (not stopping patterns) in the busy parts of the JR-W network to interact and overtake each other with overtaking loops at a few stations. Some parts of the network even can be described as fairly intensive urban single track operations. In contrast, Tokyo’s JR regional system mostly has the luxury of separating and double tracking everything everytime lines get busy. Therefore if I were being strategic but still trying to push the envelope for better; the scale, operations and expected offerings of the JR-West Kansai network can be a better pitch for a truly visionary Toronto mainline urban/suburban/regional passenger service than Nagoya. Yet it is still realistic enough to pitch for buy-in, implement and sustain compared to Tokyo. It is not that crazy.

The JR Urban(e) Network Vision

So begins the second part of the article where I attempt to sketch out a semi-realistic vision for what an electric frequent GO Expansion plan should look like using a lot of examples from JR-W’s Kansai Network. I try to envision an ambitious service plan making the best use of existing, under construction and already proposed (meaning feasible) upgrades. Keeping in mind to minimize “out of the blue” changes on infrastructure that go beyond what is already planned or under construction while still having some ambitious vision. However, Canada typically lacks grand vision these days so I would not be surprised if this plan falls upon deaf ears.

Regardless, why I picked JR-W for inspiration is that they historically needed to make the best of their existing infrastructure since JNR dumped all their money on Tokyo’s five regional railway corridors and the first Shinkansen. As such JR-W’s Kansai regional network has a lot more parallels and design pressures with Toronto infrastructure-wise, making it a good model for an ambitious start of an electric regional railway plan for five of the seven GO Transit lines.

Station Spacing

Stop spacing on the GO Transit system has been a contentious issue. Metrolinx is notorious for analysis that is biased against infill stations. Metrolinx’s bias comes from economic calculations that heavily weigh travel time impacts for existing upstream passengers and the operating assumption of having an all local service blend for most lines. If an additional station is proposed, it will be systematically penalized as trains stopping and serving the new proposed station are calculated to be “delaying” existing passengers. Japan gets around this by just having various express and local stopping patterns and using high performance EMUs. Regardless, a quick review of key lines in the JR-W Kansai Network shows a station every 2 to 3 kilometers. This vision of GO Expansion will target a similar station spacing.

The Service

I will outline the vision for what an electric frequent GO system would look like with heavy inspiration from JR West. I am assuming a lot of planned and under construction infrastructure will be in place and will only discuss any infrastructure additions that are not planned as part of the original GO Expansion plan. I will be assuming the lines being discussed have my recommended ~2km station spacings by constructing a ton of new infill stations as well as full electric traction.

Dual track line: Barrie Line (BR)

The GO Transit Barrie line (BR) should be like the JR-W Gakkentoshi Line (also known as Line H or Katamachi Line), a mostly dual track line but the quieter end in Nara is single tracked. The Gakkentoshi Line’s service planning uses a zonal system stopping pattern operated as three basic zones with certain trains skipping stops as they enter urban Osaka. Speeding up travel times for commuters from the suburbs.

On the JR-W Gakkentoshi Line, Regional Rapid trains skip four stations in Osaka’s urban zone and run local in the inner and outer suburban zones. Rapid trains skip seven stations in the urban and inner suburban zones then run local. Local trains run all stop but terminate at the edge of the urban and inner suburban zones. Local trains rarely run the length of the line fully end to end.

Below is the schedule of a major urban station on the Gakkentoshi Line heading inbound. Regional Rapid departures are in green text, Rapid departures are in yellow text and local departures are in black text.

Frequency fades the further away from the city center as local (and some regional rapid) trains short turn, as shown by the Gakkentoshi Line schedule below at major stations where the service transitions between the respective urban, inner suburban and outer suburban zones. You see a lot of regional rapid (green text) and rapid trains (yellow text) interspersed with the local trains (black text). As you reach the outer suburban zones all trains are some type of rapid train but service goes down to hourly/half-hourly off peak.

Why do I bring the Gakkentoshi Line as a comparison for the future of the Barrie Line? You see the JR-West schedules, stopping patterns and service intensity I presented above and ask: What is the level of infrastructure required to run this service? The answer is that the Gakkentoshi Line is an electric double track line with stations every 2-3km and 3 to 4 stations on the double track urban and inner suburban portion of the line are equipped with overtaking tracks. The outer suburban section is single track with passing loops at most stations. See below a track map of the Gakkentoshi/ Katamachi Line (Line H) created by 配線略図.net where from left to right and looking down, you are moving east to west:

If you compare the trackmap above with the infrastructure of the Barrie Line today, you can see the Barrie Line will be more than halfway there with double tracking work committed from Union all the way up to Aurora Station. What is missing is electrification, more infill stations (every 2-3km) and some overtaking loops for the various zonal express stopping patterns to overtake each other. Yes, there are still differences between the Barrie Line and JR‑West’s Gakkentoshi Line; such as some minor freight operations in a short section of the Barrie Line to serve a small oil depot in North York. Ultimately, I believe GO Transit could achieve a service intensity a few trains per hour less than the service provided by JR-W on the Gakkentoshi Line on the Barrie Line.

As such, the ultimate service plan for the Barrie Line could run similarly to the Gakkentoshi Line. I believe the Barrie Line will have a more commuter oriented demand profile where there is less demand for direct Toronto-Barrie trips but more “as the train heads closer to the city as it fills up” like the Gakkentoshi Line. So a system of zonal express patterns is preferable. The city zone starts at Union and heads up to Maple Station in Vaughan; the inner suburban zone will be from Maple to Newmarket Stations and the outer suburban zone reaches up to Allendale Waterfront Station in Barrie. There will be different tiers of zonal express service patterns running all stop to serve a specific suburban zone but running express and stopping only at major stations in other zones. For this vision of an electric GO Transit system they will be called Sectional Express services.

The ultimate configuration of a full electric traction GO Regional Express Rail Barrie Line could operate like a slightly less frequent JR-W Gakkentoshi Line. The list below could be the ultimate service configuration of a full electric traction GO Regional Express Rail Barrie Line during the peak periods in each direction:

• 3-4 trains per hour of Local trains serving the city and inner suburban zone

• 3-4 trains per hour of Sectional Express3 (JR-W calls them Regional Rapid on Gakkentoshi Line) trains, stopping only at major stations in the city section (Union to Maple). Then running local in the inner suburban zone and some trains continuing out to Allendale Waterfront (Barrie) in the outer suburban zone running local.

• 2-3 trains per hour of Express trains (JR-W calls them Rapid on Gakkentoshi Line), stopping only at major stations in the city and inner suburban zone (Union to Newmarket) then running local to Allendale Waterfront (Barrie).

See the schedules below for the Gakkentoshi Line heading inbound for what the level of service a typical station in each zone of the Barrie Line would see in this vision for GO Expansion. The urban section of the Gakkentoshi Line is the busiest (leftmost schedule), served by alternating Rapid, Regional Rapid and Local trains. The Barrie Line can operate a similar service plan with alternating Rapid Express, Regional Rapid Sectional Express and Local trains. Services gradually are added or cut back as the Barrie Line moves towards/away from the city as they do in the Gakkentoshi Line. Of course if there are concerns with creating Akazu no Fumikiri (開かずの踏切)4 situations on the Barrie Line, the frequency of the local trains can be cut back at Caledonia Station just south of the first at-grade crossing (Castlefield Ave) on the Barrie Line.

I could imagine in an alternative service pattern, an electrified more urban(e) Barrie Line which removes all the commuter zones. That schedule would compress into a simpler two stopping pattern local/express combo with more Rapid trains per hour that skip maybe two thirds of the stops on the Barrie Line. In this vision in each direction during rush hour the Barrie Line could see maybe five or six local trains per hour, some of which short turn at Maple Station south of the inner suburban zone. On top of that, four or five Express trains per hour, gradually cutting back service as you move away from the city until maybe 2-3 trains per hour reach Barrie.

Regardless, the Barrie Line itself is already mostly isolated from freight and intercity movements and is entirely owned by GO Transit/Metrolinx. As shown by what JR-W has achieved on the Gakkentoshi Line, the services I proposed here are within the realm of possibility infrastructure wise for a two track line with some line end single tracking.

Quadruple Track Corridor: Kitchener (KI)/Stouffville (ST)/Lakeshore Lines (LW-LE)

Kitchener and Lakeshore Lines are quadruple track corridors and should operate like what is done on the JR-W San’yo/Tokaido Mainlines (also known as JR-W Line A), namely the services branded as the Kobe/Kyoto/Biwako Line sections. On the long quadruple track sections of the San’yo/ Tokaido Mainline, different rail services are sorted by track, with a pair of slow tracks and a pair of fast tracks. As shown in the schedules of the San’yo/Tokaido corridor below, JR-W ran frequent Local (black text) and Rapid (yellow text) services on the slow pair of tracks. While intercity limited express trains (red text in next set of schedules), freight trains (not shown) and regional trains such as the aforementioned Special Rapid service (blue text) ran on the fast tracks.

As noted in the schedules above, the San’yo/ Tokaido Mainline corridor’s fast tracks also host a large volume of limited express intercity trains/ Airport Express Haruka trains and even Sleeper Limited Express trains. The Tokaido Mainline between Osaka and Kyoto (also known as the Kyoto Line) hosts 4-6 limited express trains per hour during the AM peak on the fast tracks. At the same time, these limited express intercity trains are sharing the fast tracks with the Special Rapid regional service and freight trains serving JR-Freight Kyoto, Suita, Osaka, (formerly) Umeda, Kobe and Himeji freight terminals along the way.

Uniquely here, Rapid services (yellow text) which effectively are zonal express services GO Transit is familiar with share the slow tracks with Local trains using smart scheduling and strategic overtaking loops accessible from the slow tracks to allow both services to co-exist. This arrangement is partly due to historical reasons from the JNR era but it also effectively sorts services in a way that fully utilizes the capacity of a quadruple track corridor. It would be a waste of capacity to dedicate a full track pair just to Local services and try to shove all the Rapid (Sectional/Zonal Express), Special Rapid, Limited Express and freight services into the fast tracks and hope they coexist.

This is why the JR-W model of running a quadruple track corridor, born from a quirk of history, is worth emulating. In contrast to models in North America where strict separation in planning and operations leads to a quadruple track corridor defaulting into operating as simply two isolated parallel railways. From the SuperGO plan to how WMATA and MARC/VRE in Greater Washington Area run beside each other in “joint corridors”, N American multitrack corridors don’t interact in an integrated way. JR-W sees Rapid (Sectional/Zonal Express), Special Rapid (Regional), Limited Express (Intercity) and freight services as all important. How do you allocate the two track pairs in the corridor for each service? Who can mix/synergize? What services can coexist together? The North American model ends up throwing out all the different services and hyperfocusing on two extreme ones for each track pair. For example, a corridor where WMATA’s Metrorail lines run adjacent to other mainline railways only provides:

• a commuter oriented frequent passenger Metrorail service with wide (+3km) stop spacing by Metrorail, basically a suburban commuter express service

• Freight service on mainline rail

• A terrible, infrequent, mostly peak direction and mostly peak period regional service by VRE and MARC.

JR-W would take a quadruple track corridor like the San’yo/Tokaido and deliver:

• a local service stopping at stations every 1–2 km.

• a sectional express commuter oriented stopping pattern (what JR‑W calls Rapid on Line A/B and Regional Rapid on other lines)

• an express regional service (what JR‑W calls Special Rapid on Line A/B and Rapid on other lines)

• at the same time still accommodating substantial intercity (Limited Express) and freight services.

All services described in the above list are frequent, two way, all day and run on just four tracks with some brains. No passenger market segment is left unserved for decades; all are integrated into a seamless system between the various passenger services JR-W had on offer.

Quadruple track corridors in Toronto for GO RER can generally be arranged and operated in a similar way with the caveat that the fast tracks most likely have to be on the outside of the corridor in a fast-slow-slow-fast arrangement. This is done in some parts of the quadruple track JR-W Kobe Line, where 4 door Local and 3 door Rapid services use the center two tracks:

Whilst Special Rapid, Limited Express and freight trains use the outer tracks at the same station, not stopping because JR-W determined this will be a station that only serves services on the slow tracks:

This final ultimate quadruple track arrangement is proposed to be emulated on the GO Transit network, as the fast tracks would host any of the corridor’s freight movements. The GO network has a number of adjacent industrial railway freight customers that still ship by rail via spur/siding access. Putting the fast tracks on the outside of the corridor will help reduce overall movement conflicts serving these industrial railway freight customers to trains running on the slow tracks. However, GO Transit currently is building out their railway corridors on the assumption of a slow-fast-fast-slow arrangement, as Ontario Traffic Man has sketched out in their trackmap of the Lakeshore East corridor in the City of Toronto:

Because GO is building to a slow‑fast‑fast‑slow arrangement, adjustments to existing station platform arrangements on all of GO’s proposed quadruple track corridors will be required.

Lakeshore Corridor

The Lakeshore Corridor is Toronto’s premier corridor with the most ridership potential and was proposed to have significant sections quadruple tracked in GO Expansion. However, this ambition is currently in doubt. It is in so many ways similar to JR-W’s own Kansai premier quadruple track corridor, the aforementioned San’yo/ Tokaido Mainline (Line A). As such the Lakeshore Corridor or Lakeshore East and West Lines should bring back that ambition and adopt a similar quadruple track operational design JR-W runs today on the San’yo/ Tokaido Mainline. Such as using a fast-slow-slow-fast track arrangement:

This plan will deemphasize existing zonal express trains operating on the corridor.5 These zonal express services exist historically on the Lakeshore East/West and Kitchener Lines, known simply as “express” trains. These services historically operated as a few runs during the peak period, running all stop in peak direction of the line’s outer section then running non-stop into/out of the city center. For example, “express” services on the Lakeshore East Line stop at all stations between Pickering and Oshawa Stations, then run non-stop to Union Station in Toronto’s downtown. This express service stopped running in 2021 and for some reason in the Year of Our Lord 2026 there still seems to be no good reason for the suspension. While on the Lakeshore West Line “express” services stop at all stations between Aldershot and Clarkson Stations and then run non-stop to Union Station. In this vision for the future GO regional rail system, these services will not be completely removed from the Lakeshore E-W and Kitchener Lines and operate all day both ways. They will be renamed as Sectional Express, are analogous to JR-W’s Rapid services on Line A/B or Regional Rapid services on other JR-W Kansai lines and operate like the zonal express services proposed for the Barrie Line.

A more Japanese style, limited-stop express service that runs all day both ways and stops only at major stations, like JR-W’s flagship Special Rapid service (blue line below) on the San’yo/ Tokaido Mainline 東海道本線・山陽本線 (Line A), would be prioritized in service planning. This is in contrast with a service design using layers of zonal express services found on the Gakkentoshi Line (as Regional Rapid or Rapid services) or proposed in this article for the Barrie Line. Zonal express style services still appear on the San’yo/ Tokaido Mainline (Line A) as Rapid services (orange line below) and operate on the slow tracks sharing them with the Local trains (grey line below):

The corridor’s design will encourage passengers to transfer between Express services with Local services at major interchange stations. This is done to better leverage the quadruple track nature of the Lakeshore and Kitchener corridors. In addition, the Lakeshore corridor’s urban geographic environment is similar to the JR-W Sanyo/ Tokaido Mainline in that there are major nodes anchoring the length of the corridor such as Hamilton, Oakville Midtown, Port Credit, Humber Bay Shores, Exhibition, Union (Downtown Toronto), West Harbour, Scarborough GO (transfer to Stouffville Locals) and Ritson Road (Downtown Oshawa). This is not unlike the nodes of Himeji, Kobe, Osaka (Umeda), Shin-Osaka, Kyoto and Otsu that the Special Rapid serves non-stop on the Sanyo/ Tokaido Mainline. With these considerations in mind I will explain the vision for the Lakeshore corridor in two halves:

Stouffville/Lakeshore East Lines

The Lakeshore East corridor was planned to be a quadruple track corridor between Union and Oshawa Station as part of the original GO Expansion plan. For the Lakeshore East Line, the inner and outer suburban zones will be the dual track sections beyond Pickering Station, with Oshawa Station serving as the boundary between inner and outer suburban zones. While the inner/outer suburban zone of the Stouffville Line is single track north of Unionville Station.

Right out of the gate, one thing to note is I doubt the corridor between Scarborough GO and Union Stations can be expanded into a six track corridor. The SuperGO plan originally proposed five tracks between Scarborough GO and Union Stations. This means the right of way could support having more than quadruple tracks but having an odd number of tracks is not optimal.

Therefore, the Stouffville Line services are constrained by sharing with Lakeshore East Line services on the Lakeshore Corridor between Scarborough and East Harbour. The Lakeshore East Line was ultimately proposed in the original ONxpress GO Expansion plan to be expanded into a four track corridor. However, a track imbalance still exists. That is why I think the Lakeshore East and Stouffville Lines have to be planned together. If we are to take inspiration from JR-W on the (also) eastern end of the Kansai network, Kosei Line services need to merge and interact with Biwako line services using the quadruple track west of Kyoto Station. Then Kosei and Biwako line services all squeeze into the quadruple track Kyoto Line. This leads to, like at Scarborough GO Station, an imbalance of track pairs as shown by a track map of the Biwako/ Tokaido Main Line (Line A) created by 配線略図.net where from left to right and looking down, you are moving east to west:

Most services on the Kyoto Line JR京都線 (Line A) head into the Biwako Line 琵琶湖線 (Line A) which is still quadruple track east of Kyoto. These include 3-4 special rapid trains per hour from the fast tracks of the Kyoto Line, which continue onto the fast tracks of the Biwako Line. The dual track Kosei Line 湖西線 (Line B) branches off the Biwako Line and gets one special rapid train per hour all day through operating to and from the Kyoto Line fast tracks. While the rest of the Kosei Line services are local trains that terminate at Kyoto (denoted below with a 京), freeing up the Kyoto Line slow tracks for more inbound Local and Rapid services from the Biwako Line and resolving the track imbalance heading into the city center of Osaka (right side of map below denoted by 大阪).

The ultimate service plan for the Stouffville/ Lakeshore East Lines in this vision would be run similarly to the JR-W Kyoto/ Biwako (Line A) sections of the Tokaido Mainline and Kosei Line (Line B) as described above. In this plan the city zone is up to Pickering Station on the Lakeshore Line and Unionville on the Stouffville Line. The Stouffville Line still can operate like the Kosei Line with express trains (blue text) every hour. Since the Stouffville Line has no intercity trains (limited express red text) we can assume it can support more express trains. Maybe even extend the hourly Stouffville/Lakeshore express trains to serve and terminate at Uxbridge. The dual track section of the Stouffville Line between Scarborough GO and Unionville will need some overtaking loops to allow express trains to pass the locals. The Kosei Line schedule can be achieved and even accommodate a few freight services using the infrastructure as shown by a track map of the Kosei Line (Line B) created by 配線略図.net where from left to right and looking down, you are moving north to south:

What is different about Stouffville/ Lakeshore East system compared with JR-W Biwako/ Kosei (Line A/ B) system is after the two “branches” diverge at Scarborough GO Station the Lakeshore Line, even if it was widened to four tracks, will still have a bottleneck further east when it reaches the junction with the busy CN York Subdivision at Pickering Station. The York Subdivision is Toronto’s primary freight by-pass line, the Musashino Line or Guangshi Railway of Toronto. This line handles large amounts of North American style >kilometer long “land barge” freight trains. I do not even know if we can get any GO Transit express trains onto those tracks. As such, most likely two of the Lakeshore Line tracks east of Pickering into Durham Region will have to remain fully dedicated to freight traffic with VIA intercity trains.

The most realistic result is that the corridor beyond this junction will allocate one pair of tracks for freight trains and VIA intercity services while the other pair currently owned by GO Transit will only have the capacity to host the Lakeshore Line Sectional Express and Express trains. I propose the GO Lakeshore East Line to be widened to quadruple track up to Pickering. To understand how train schedules can accommodate this transition from a quadruple track GO Transit corridor to GO Transit having access to only two tracks east of Pickering Station, we can again look to JR-W’s Kansai Network:

JR-W has a similar dilemma on the Sanyo Mainline (Line A) on the western side of their Kansai Network. The quadruple track Sanyo Mainline becomes dual track west of Nishi-Akashi Station. As shown below by a track map of the Sanyo Mainline (Line A) created by 配線略図.net where from left to right and looking down, you are moving east to west:

JR-W’s schedules (see below) on the San’yo Mainline react to accommodate this change in infrastructure. Most San’yo Mainline Local trains (black text in schedule, gray line on the above map) on the slow tracks terminate at Nishi-Akashi 西明石 (denoted in schedule below with 西) or even earlier at Suma 須磨 (denoted in schedule below with 須). While the 4-5 Rapid trains per hour (yellow text in schedule, orange line on above map) skip stops but still use the slow tracks, all continuing into the dual track section. However, as denoted in the map above in light orange Rapid trains now run local, stopping at all stations beyond Nishi-Akashi Station up to Kakogawa 加古川 (denoted in schedule below with 加), with some continuing beyond Himeji City to Aboshi 網干 (denoted in schedule below with 網).

Finally, the 3 to 6 Special Rapid trains per hour (blue text in schedule, blue line on above map) from the fast tracks also all continue into the dual track section beyond Nishi-Akashi but still remain express skipping stops up to Himeji or beyond. Once in the dual track section, Special Rapid trains use overtaking loops to pass the Rapid trains that are now running local. Once at Himeji Station 姫路, most Special Rapid trains terminate. Some Special Rapid trains continue beyond Himeji as all stop local trains that eventually terminate at Kamigori 上郡 on the Sanyo Mainline or bifurcate and through operate into the Ako Line 赤穂線.

Following what JR-W’s example above at Nishi-Akashi Station, I propose some Lakeshore East Local trains turn off and through run into the Stouffville Line at Scarborough GO Station and the rest of the Lakeshore Line local trains terminate at Pickering Station. The Sectional Express and Express trains will run into the dedicated GO Transit dual track section beyond Pickering Station. Sectional Express trains will now run local, mirroring old historical service pattern of the Lakeshore East “Express” train. Express trains continue to stop only at major stations up to Oshawa. To handle the interactions between the Sectional Express and Express stopping patterns from sharing the one pair of GO Transit tracks, Express trains will use timed overtakes at stations with overtaking loops to pass the Sectional Express trains. In summary, east of Pickering Station, the following arrangements can be made:

• VIA intercity trains can switch off the Lakeshore East fast tracks to the pair of CN freight tracks.

• Express trains from the Lakeshore East fast tracks head over to the pair of GO Transit tracks. The Sectional Express trains will continue to run local beyond this station and Express trains will only stop at major stations until Ritson Road (Downtown Oshawa) then run local to Newcastle.

• Terminate most Local trains coming in on the slow tracks at Pickering Station as well as divert some Local trains into the Stouffville Line.

Putting everything together and taking inspiration from what was achieved by JR-W Kyoto/Biwako and Kosei Lines as well as the Sanyo Mainline. The below list could be the ultimate service configuration of a full electric traction GO Regional Express Rail Lakeshore East/Stouffville Line during the peak periods each direction:

• ~6 trains per hour of local services using the Lakeshore Line slow tracks serving the urban areas up to Pickering Station. Some (2-3 of the ~6 trains per hour) through operating as locals into the Stouffville Line instead of going to Pickering. In addition:

  • 3 trains per hour of shuttle local services on the Stouffville Line will terminate at Scarborough GO Station and do not enter the Lakeshore East corridor. Maybe transfers are timed with the 3-4 express trains per hour that run on the Lakeshore East corridor. This shuttle/local pattern mirrors the approach JR‑W uses to resolve track imbalance on the Kosei and Biwako Lines as they converge into the Kyoto Line. However, the Stouffville Line north of Unionville is single tracked with minimal space for passing loops. Meaning some local services will need to be cut back at Unionville Station to accommodate this constraint.

• 3-4 trains per hour of Sectional Express trains on Lakeshore Line slow tracks only. Stopping only at major stations in the urban section. Then Sectional Express trains run local in the dual track inner suburban zone beyond Pickering Station up to maybe Oshawa/Ritson Road (Downtown Oshawa).

• 3-4 trains per hour of express trains using the Lakeshore Line fast tracks stopping only at major stations in the city and inner suburban zone (Union to Oshawa) then running local in the outer suburban zone beyond Oshawa to Newcastle. In addition:

  • 1-2 trains per hour of express trains using the Lakeshore Line fast tracks. Stopping only at major stations in the city and inner suburban zone of the Lakeshore East Line from Union and Scarborough GO Stations. After which these trains diverge and through operate into the Stouffville Line, running express stopping only at major stations up to Unionville. Then run all stop local to Uxbridge.

Again as achieved by JR-W, 4 to 6 express trains per hour on the fast tracks of the Quadruple track corridor between Pickering and Union Station leave room for ~4 intercity VIA trains (or even Alto HSR trains) and a few short freight trains per hour (or a few long ones every day). Just like what is done on the fast tracks of the JR-West Kobe/Kyoto (Line A) corridor.

The Lakeshore East Line was proposed to be quadruple track all the way to Oshawa, but that seems to be descoped in the MVP for GO Expansion. This needs to be brought back. Apart from that, trackwise everything is almost there on the Lakeshore East and Stouffville Lines to achieve this proposed service plan except for: electrification, infill stations and a bunch of overtaking loops all around. The Stouffville Line will be double tracked from Scarborough GO to Unionville Stations and single track the rest of the way to Old Elm (or even Uxbridge). What is really missing in terms of major infrastructure for the Stouffville Line is a flyover at Scarborough GO Station and the design should be like the Yamashina Station trackmap below drawn by 配線略図.net. The fast tracks are on the outside of the quadruple track corridor:

There are still a bunch of crossings in eastern Scarborough between Pickering and Guildwood Stations that may need to be addressed if Akazu no Fumikiri (開かずの踏切) situations are a concern.

Lakeshore West Line

The other side of the Lakeshore corridor will follow a similar pattern:

• The corridor will deemphasize existing zonal express trains. Instead, a more Japanese style limited-stop express service that runs all day in both directions and stops only at major stations will be prioritized.

• Local and Sectional Express trains will run on the slow pair of tracks. While Express, VIA intercity and the occasional freight services will run on the fast pair of tracks.

However, this time the quadruple track corridor will run up to Aldershot Station then immediately split into two branches. One going to Hamilton GO Center and the other heading to East Harbour/ Stoney Creek and beyond to Niagara via the CN Grimsby Subdivision. The Hamilton GO Center branch will mostly be single track as it is constrained by freight movements and the single track Hunter Street tunnel.

The CN Grimsby Subdivision running through the West Harbour area, Lower Hamilton and terminating in Niagara Falls Station will most likely remain owned by CN Railway. As such the level of service offered in this “Niagara Falls” Branch may be limited.

We can again take inspiration from JR-W. This time back to the east at Kusatsu Station where the quadruple track corridor of the Tokaido Main Line ends but continues as a dual track line with a single track branch line, the Kusatsu Line (Line C). As shown by a track map of the Biwako/ Tokaido Main Line (Line A) created by 配線略図.net where from left to right and looking down, you are moving east to west:

The single track nature of the Kusatsu Line limits frequency to around 3 trains per hour (left schedule below). Many trains through run from the Kusatsu Line into the Biwako Line to Kyoto (denoted below left schedule with a 京). There is even one run as a Sanyo/Tokaido Mainline Rapid to Himeji, almost 150km away (denoted with a ▼ in left schedule below). At the same time if not running into the Kusatsu Line, remaining outbound Biwako services must squeeze into the remaining two track section of the Biwako Line beyond the Kusatsu Line to Ritto. Conversely, throughout the day a Special Rapid train every hour terminates at Kusatsu Station (denoted with a 草 in right schedule below) or about 1-2 local trains per hour through operate into the Kusatsu Line in the evening (denoted with a 柘 in right schedule below).

Putting everything together: taking inspiration from what was achieved on the JR-W Biwako/Kusatsu Lines at Kusatsu Station, the operational arrangements for a quadruple track corridor and service plans outlined for Lakeshore East. The list below could be the ultimate service configuration of a full electric traction GO Regional Express Rail Lakeshore West Line during peak periods in each direction:

• ~6 trains per hour of local services using the Lakeshore West Line slow tracks serving all stops in the city and inner suburbs. Some trains (1-2 of the ~6 trains per hour) head into the CN Grimsby Subdivision, maybe terminating at Stoney Creek Station to provide service for Lower Hamilton but most will terminate at Aldershot Station.

• 3-4 trains per hour of Sectional Express trains on Lakeshore West Line slow tracks, stopping only at major stations in the city section between Union and Port Credit Stations. Then Sectional Express trains run local in the dual track inner suburban zone beyond Port Credit Station. This service will closely but not exactly follow the historical “express” service provided on the Lakeshore West Line. Half the trains head to Hamilton GO Center and the other half terminates at Aldershot Station.

• 4-6 trains per hour of express trains using the Lakeshore Line fast tracks stopping only at major stations in the city and inner suburban zone. Of the 4-6 express trains, half (2-3 trains per hour) will enter the branch to Hamilton GO Center. This is a completely express service stopping only at major stations between Hamilton and Union. The other 2-3 trains per hour will continue to run express into the CN Grimsby Subdivision up to Stoney Creek Station using overtaking tracks to pass the handful of Lakeshore West Local trains and then run local in the outer suburban zone up to Niagara Station.

Like the Lakeshore East Line, the Lakeshore West Line was proposed to be quadruple track all the way to Aldershot Station, but that seems to be descoped in the MVP for GO Expansion. This needs to be brought back to achieve what was outlined in the service plan listed above. Apart from that, trackwise everything is almost there on the Lakeshore West to achieve this proposed service plan except for: electrification, infill stations and a bunch of overtaking loops all around. A flyover will be needed at Aldershot Station just like what is shown at Kusatsu Station. Lastly there are still a bunch of crossings that still need to be removed, particularly in Mississauga, if Akazu no Fumikiri (開かずの踏切) are a concern.

Kitchener Line/UP Express/Bolton Line

Union-Pearson (UP)/Kitchener outer suburban express services can operate like Hanwa Rapid or the quadruple track corridor JR-W originally envisioned the Hanwa Line to be (but that is another story). A key feature of the Hanwa Line Rapids is the Airport and Kishuji Rapid services that run as one train and split at Hineno station. One set goes to Kansai International Airport via a branch line while the other continues south on the Hanwa Line to Wakayama.

This is shown by JR-W’s schedule below where Kansai Airport/Kishuji Rapid trains are denoted by two sets of different color Kanji 関快・紀快 (blue・yellow text) north of Hineno Station (below left). The Kishuji Rapid is just denoted as 紀快 (yellow text) south of Hineno Station on the Hanwa Line, as the airport set has split off into the Kansai Airport branch line and all Local trains have terminated at Hineno Station (below right). The Kitchener Line Express/UP Express services can follow a similar service concept.

Another thing to note is JR-W manages to run 4 trains per hour of each stopping pattern: Rapid (blue text), Regional Rapid (green text), and Local (black text) on the two track line Hanwa Line using overtaking loops. Not to mention the 1-2 trains per hour of near non-stop limited express Haruka and Kuroshio services that run on top of the regular service on the Hanwa Line. This again shows the subtle difference in historical planning ethos of JR-W compared to the more infrastructure heavy approach used by JR-E.

By splitting trains it will be possible to reuse the existing UP airport spur viaduct, whatever that can be saved from its poorly built design. Configurations can be 9+3 or 6+3 car consists with the 3 car unit being for the UP Express Service. By having trains split at Woodbine Station, UP Express trains can still be short 3 car units and not run into congestion issues while serving the rest of the Kitchener Line corridor as an express service for non-airport use.

This solves the conundrum Metrolinx always struggled to grapple with the design/market position of the UP Express within the GO Transit Network ever since harmonization of UP Express and GO Transit fares at intermediate stops. Implying that the UP Express also has a non-airport express role. Tensions boiled over when Metrolinx once proposed that UP Express runs alternate between a non-stop service that skips two intermediate urban stations and an all stop service. This effectively was a service cut for non-airport commuters boarding the UP Express at the two intermediate urban stations to get downtown. UP Express trains only come every 15 minutes and there was no service increase for this proposal, much to the immediate chagrin of the general public upon announcement leading to prompt cancellation. Metrolinx most likely wanted to reduce crowding “caused by” Kitchener Line commuters who are starved of any decent GO Transit service and who use the UP Express as an alternative. This was impacting the “Airport Express” experience of UP Express services. However, my proposal solves this by coupling UP Express trains with Kitchener Express trains at Woodbine Station, adding capacity exactly when the combined train serves the two (now major) intermediate urban stations on the Kitchener corridor.

This vision proposes full quadruple tracking of Kitchener Line corridor from Union to at least Mount Pleasant Station. Assuming similar track infrastructure and arrangements on the Kitchener Line as presented about the quadruple track Lakeshore East and West Corridors, services are sorted into each pair of fast and slow tracks. The slow tracks will host Local and Sectional Express trains while the fast tracks host the Express trains to Kitchener/Airport, VIA Intercity and freight trains. West of Bramalea Station, two tracks will need to be allocated for CN to run heavy freight services and then turn off the Kitchener Line into York Subdivision, Toronto’s railway freight bypass line. As such, the section of the Kitchener Line running through Brampton City is proposed to have a similar track allocation as on the Lakeshore East Line in Durham Region east of Pickering Station: A joint quadruple track corridor with a pair of CN heavy freight tracks and a pair of dedicated GO Transit tracks hosting different GO RER inner and outer suburban services. The different stopping patterns can co-exist on the two GO Transit tracks using overtaking loops.

Track constraints in this section of the Kitchener Line are similar to those described for the Lakeshore East Line east of Pickering Station. Similarly, The proposed service arrangements should be modelled on and inspired by the JR-W San’yo Mainline around Nishi-Akashi Station. Below is the level service JR-W achieved on the dual and quadruple track sections of the San’yo Mainline around Nishi-Akashi Station. The GO Transit Kitchener Line should aspire to the service levels of Local (black), Section Express (Yellow) and Express (Blue) services heading inbound into Toronto shown by the JR-W schedule below. Imagine this level of service provided by GO Transit on the dual track section at Brampton Station (left) and quadruple track section at Dundas West Station (right):

If we where to translate the above to a more frequent electric Kitchener Line. Urban local services will terminate at Bramalea Station as the GO RER accessible tracks go from a quadruple track corridor with two separate fast/slow pairs of tracks east of Bramalea Station to a single pair in the west continuing through Brampton. While remaining inner and outer suburban Sectional Express and Express services will use the two dedicated GO Transit tracks running through Brampton. VIA Intercity and residual Kitchener Line freight services using the fast tracks with the Kitchener Express trains will shift over to the pair of CN owned heavy freight tracks running through Brampton adjacent to the two dedicated GO Transit tracks.

Putting everything together and taking inspiration from what was already achieved by JR-W, the list below could be the ultimate service configuration of a full electric traction GO Regional Express Rail Kitchener/Bolton Line during peak periods in each direction:

• ~6 trains per hour of Local services using the Kitchener Line slow tracks, 6 trains per hour serving the city stopping at every station up to Bramalea Station while 2 of those ~6 trains per hour can head up to the proposed Bolton Line.

• 3-4 trains per hour of Sectional Express services using the Kitchener Line slow tracks that run local between Mount Pleasant and Bramalea Station then express to Union stopping only at major stations.

• 4-6 trains per hour Express/Airport Express services using the Kitchener Line fast tracks. Each train will split at Woodbine Station with one set heading to Pearson Airport while the other longer set runs express to Kitchener.

Like the Lakeshore East and West Line, the Kitchener Line was proposed to be quadruple track all the way to Bramalea Station with parts of the Kitchener Line already widened to four tracks as part of the Georgetown South Project. However, there remain significant sections between Bramalea and Union Station that are not quadruple track and seem to be descoped in the MVP for GO Expansion. This needs to be brought back to achieve the service levels listed above. In addition, the Kitchener Line corridor running through the City of Brampton is currently only three tracks, in a tight right of way and completely owned by CN Railway.

Expanding this section of the Kitchener Line running through the City of Brampton to four tracks with a pair of tracks dedicated to freight/VIA intercity and another strictly to GO RER services remains feasible and should also be pursued to allow for the service levels proposed in the list above. Operations in this arrangement will be similar to what already exists on the Lakeshore East Line going through Durham Region. Lastly, flyovers will be needed at Halton Hills (see map above) to allow the CN freight tracks to not interfere with the GO tracks when they head south as well as a flyover when the Airport spur diverges from the Kitchener Line west of Woodbine Station. As for grade crossings, the Georgetown South Project removed most if not all of them. (やったー!) Finally, the normal suspects: electrification, infill stations and a bunch of overtaking loops all around.

Union Station Imbalance

Regional rail best practice generally suggests as much service through run across the city center as possible. This will help preserve precious capacity in already constrained city center corridors. Ontario Traffic Man articulated in their vision for GO RER to have as many Lakeshore East Line trains through operate into the Lakeshore West and Kitchener/Airport Lines as possible. This means that not a single Lakeshore East Line train shall terminate in Toronto’s city center. In addition, Dr. English noted in their TBOT Trillium Regional Railway Network Plan that there is an imbalance of tracks and resulting service on both sides of Union Station.

Major JR-W hubs such as Tennoji or Osaka Stations operate on similar principles, mostly through running stations with minimal services terminating. Infrastructure is provided to terminate some services at Tennoji or Osaka Stations to balance out the capacity mismatch on each side of the station. Those services that do terminate in these central stations are Local services, reserving any capacity on the constraining side for through running as many express/regional trains as possible.

The west side of Union Station can support 6 pairs of tracks:

• Lakeshore West quadruple track

• Kitchener/UP Express quadruple track

• one pair for the Milton Line and another for the Barrie Line

At the same time the east side of Union only has two pairs (Lakeshore East quadruple track) or maybe three pairs if we do something about the unelectrified single track Richmond Hill Line. Dr. English proposed just that in his excellent report proposing the Trillium Regional Rail Plan, using the lower section of Richmond Hill Line as a short electric spur line with terminal station infrastructure to allow for the abundance of GO RER services from the west to run through Union Station and the Union Station Rail Corridor (URSC) then turn around on the Richmond Hill Line. Termination infrastructure at new “Shoulder Stations” on Spadina and Cherry Streets will additionally help correct the imbalance of services on each side of Union.

This is similar to how JR-W handles the huge track imbalance with the Hanwa (Line R) and Yamatoji Lines (Kansai Main Line or Line Q). Where you have services from three track pairs trying to squeeze into the single pair on the Osaka Loop Line. To mitigate this, many Hanwa and Yamatoji Line services (mostly the Local services some Regional Rapids) terminate at Tennoji or the short historical Minatomachi Station branch that is now JR-Namba Station. This reduces the number of trains using the remaining limited capacity of the two track Osaka Loop Line (Line O) that is not taken up by the local services of Osaka Loop proper. As such most services from the Hanwa and Yamatoji Line that make it into the Osaka Loop Line are Rapid or Regional Rapid services.

Short Spur: Richmond Hill Line

A short electrified spur line that used the lower section of the Richmond Hill Line to peel off and turn around trains away from the URSC as proposed in the Trillium Regional Railway Network Plan. This is not unlike how JR-W uses the JR-Namba Station to turn around Yamatoji Line trains that were not able to squeeze onto the Osaka Loop Line. Below is the trackmap of the lower side of the Osaka Loop Line by 配線略図.net looking left to right is east to west. Yamatoji and Hanwa Line trains have numerous opportunities to terminate at Tennoji, Shin-Imamiya and JR-Namba Stations.

This means a short section of the Richmond Hill Line will need to be electrified and double tracked to allow space for some local trains that can not fit into the Lakeshore East Line’s schedule to terminate off corridor in a similar way to how JR-W utilizes JR-Namba terminus. If you are really ambitious, perhaps electric service can be extended up to Thorncliffe Park via the Don Branch Subdivision. To reduce the need for flyovers, the services using the northern-most tracks of the URSC (Barrie Line/Kitchener Line) should be directed to terminate in the lower sections of the Richmond Hill Line.

Stub Terminal: East Harbour

East Harbour Station can be like JR-W Tennoji station with a stub-end termination complex where most local Hanwa Line Local trains terminate while the Rapids through operate into the Osaka Loop Line on adjacent through running platforms. These termination platforms would most likely need to be between where the current Ontario Line and Lakeshore Line through platforms are proposed. This was done to avoid the need for flyovers and allow for the most Lakeshore services to through operate with each other. New through tracks and platforms for the Lakeshore East Line could be built on the south side.

Given the location of East Harbour is east of the Richmond Hill Line where most Barrie services peel off to terminate, the next northern-most tracks of the URSC are used by Kitchener Line services. East Harbour’s stub‑terminal complex would therefore be an ideal place to turn back Kitchener Line Locals and any Kitchener/Airport Express trains that cannot through operate onto the Lakeshore East Line as Express services.

Shoulder Stations: Spadina and Cherry Streets

Given the more ambitious service plans I proposed in this article, “Shoulder Stations” advocated by Reece Martin of Next Metro fame can further help spread out the remaining termination infrastructure requirements by leveraging the entire length of the URSC like what was done in Osaka around Tennoji. This is on top of running electric trains to terminate somewhere up the Richmond Hill Line as proposed by Dr. English and the stub-terminal platforms at East Harbour. By spreading out where services can reverse among four different locations, the termination infrastructure at each individual station/location will be minimized. Thereby reducing the required width of the infill shoulder station on the narrow URSC ROW and/or allowing for wider platforms overall.

What should this minimal termination infrastructure at shoulder stations look like? On the Gakkentoshi, Tozai, and Takarazuka services (Lines H/G) some westbound services quickly terminate at Tsukaguchi Station (denoted by 塚 in schedule below) on the Takarazuka Line during the AM peak after crossing Osaka city center.

In addition to Tsukaguchi Station (denoted by 塚) many AM rush hour outbound trains on the Takarazuka Line terminate at Takarazuka Station (denoted by 宝) and Shin-Sanda Station (denoted by 新三). Below is a trackmap of the turnaround infrastructure required to achieve this. Some configurations like those found in Shin-Sanda Station even simultaneously enable overtaking of Regional Rapid and Rapid trains while also allowing trains to reverse. According to the Fukuchiyama Line schedule above, Shin-Sanda Station can reverse around 6 trains per hour and Tsukaguchi Station reverses 3 trains per hour while 3-5 trains per hour are reversed at Takarazuka Station.

Similar performance can be expected from such minimal 1-2 track reversing platforms at Spadina and Cherry Streets. The location of these reversing platforms inside the station will be determined by which service is expected to mostly reverse at the station in question.

Future Rolling Stock

Last but not least, no Regional Railway vision is complete without a discussion on the rolling stock. The highest performing regional railway systems in the world, like the JR-W Kansai Network, use EMUs. Metrolinx was already resistant to the introduction of EMUs even after the announcement of GO RER. Metrolinx’s landmark electrification report published in 2010 concluded that an electric fleet would be hauled by locomotives and make use of Metrolinx’s vast existing fleet of unpowered double deck Bombardier Bi-Level cars. A report performing a post analysis of Metrolinx’s 2010 electrification report by Transport Action Canada noted that Metrolinx’s preference for an electric locomotive hauled fleet was driven by concerns that an EMU fleet threatened Metrolinx’s current (at the time) 10 year fleet strategy assuming a large locomotive hauled Bi-Level fleet.

Since the publication of Metrolinx’s electrification report in 2010, Metrolinx has ordered 349 Bi‑Level series IX coaches (now Alstom). That total is enough to form nearly 30 twelve‑coach trains or 58 six‑coach trains. All while existing older 181 units of Bi-Level series VIII cars are being refurbished. This is ever sinking cost fallacy as Metrolinx continuously hoards ever more unpowered Bombardier Bi-level cars to be hauled by locomotives.

In 2026, as leaked via some excellent investigative journalism by the Trillium, Metrolinx is working on a new, updated draft fleet strategy. The article found that Metrolinx is still suggesting electric locomotives will remain the core of its fleet strategy. More than a decade after GO RER was announced, Metrolinx still attempts to protect the locomotive‑hauled paradigm. I assume it is because they now have so many Bi-level cars, many recently bought from the previous 2010s fleet strategy. Also many of GO Transit’s aging Bi-levels are in poor condition today. Yet the previous 2010s fleet strategy was developed with the same line of thinking (we have so many bilevel cars we must keep them!) over ten years ago. You can never improve if you keep kicking the ballast down the tracks like that.

When will the EMU transition happen? Time will tell… No, the time is now. A clear vision must be articulated and reflected in (ideally) this or (definitely) the next 10 year fleet strategy for a transition to an EMU based operation, like the highest performing regional railway systems in the world, like the one run by JR-W around the Kansai region:

Back to Reality

JR-W runs a mixed fleet of various EMU rolling stock to reflect the tension of being a regional railway system and a substitute for a metro system. Just like what an ambitious system in Toronto would entail. Local trains have the same seating configurations as local metro systems in Kansai. The exception is the Osaka Loop Line local services which use 8 car JR-W 323 series sets with a full rapid transit style longitudinal seating layout but 3 doors per car. The JR-W 207 and 321 series have 4 doors per 20m car with full rapid transit style longitudinal seating. These trains are 7-8 cars long and made with 3, 4 and 7 car sets used on the very urban San’yo/Tokaido Line local and Tozai Line services.

JR-W Kansai Rapid/Regional Rapid services exclusively use 3 door cars but with different variations of transverse seating. 1+2 seating for more standing room is found on the 223-0, 223-2500, 225-5000 and 225-5100 series used on the Hanwa Line services. While other models such as the 223-1000, 223-2000, 225-0, 225-100, and 225-6000 series have 2+2 seating and are found on San’yo/Tokaido Line Special Rapid or Yamatoji/Higashi-Osaka Line Rapid services as well as general use on more peripheral suburban/peri-urban lines in the JR-W Kansai urban network such as the Sagano or Nara Lines. Some formations of the 223 and 225 series have bathrooms on board. I will also note that the JR-W 223 and 225 series basically have the same functional/dimensional train design as the final iteration of rolling design proposed in the GO-ALRT plan. The EMUs of GO Transit can mirror this configuration and they don’t even have to be Japanese.

Local Service EMUs

EMUs for more local urban services should be like JR-W’s with more doors per car compared to EMUs used for express/regional services and feel like subway trains. Translated to GO Transit environment, a local urban EMU could use 85ft cars with 4 doors on each side and be 6-8 car formations. Inside each car is a mix of longitudinal and transverse seating like the TTC subway trains. This translates into 150-200m long trains with mostly longitudinal seating. This vision echoes the fleet plans of the 1970s: both SuperGO and GO‑ALRT proposed vehicles of a similar scale when Greater Toronto was half its current size and population.

Sectional Express/Express Service EMUs

EMUs for sectional express or express services should have fewer doors and more seats. These trains will have a mixed “medium distance” as well as urban commuter character and should be designed to reflect this mixed role. These EMUs can use the same 85ft long cars as local trains but with 3 doors and more comfortable seats in either transverse 2+2 or 1+2 seating configurations. Sectional express and the Airport Express services should prefer trains with more transverse 1+2 seating. JR-W does this to better accommodate airport traffic luggage while also being a good way to boost standee capacity for non-airport traffic use and keeping a more comfortable “medium distance” regional transverse seating arrangement:

While the regional express services might fully use transverse 2+2 seating with a bathroom and luggage racks on board. In addition all of these express EMUs should have folding seats near the doors that lock up during rush hours for more capacity:

On top of that, those transverse seats can be set back from the doors like those found in the 323 series. This will create large open areas for standees/leanees around each door to serve rush hour crushes while keeping the more regional transverse seating configuration. Again these trains will have a mixed “medium distance” and commuter role.

The EMUs used in express/regional services are less frequent and will experience rush hour crushes so will need to be formed into longer consists. Looking at what was deployed on JR-W’s flagship Line A Special Rapid, GO Transit should run 10 to 12 car long 3 door trains using 85ft cars for its express services. That is a 250m to 300m long train. Trains used on Sectional Express services should be somewhere in between the designs and train lengths used on Local and Express services.

Where to use them?

The fleet allocation again should follow what is done on JR-W. The high capacity 4 door subway-style EMUs should be deployed on local services for the Lakeshore and Kitchener/Bolton Stouffville and Barrie lines. While 3 door regional EMUs with 1+2 transverse seating would be deployed on the UP Express (Airport), Barrie and Stouffville Sectional Express services. Finally 3 door trains with 2+2 transverse seating should be used on Lakeshore/ Stouffville/ Kitchener/ Barrie Express services. Imagine you are at Hamilton GO Center and a semi-frequent 3-door 12 car limited stop Express EMU with comfy seating pulls up and takes you to Downtown Toronto in less than 40 min, all day both ways:

What now? (Again)

Over the years, every time I visit Kansai and ride JR-W’s Kansai Regional Network, I can not help but think about how GO Transit’s expansion programme can also achieve similar results. This post is just the tip of the Shibata-type coupler. There are so many other minutiae that need to be reviewed and investigated: Perhaps services to Stratford and Niagara can be handled by diesel locomotives. Bolton Line may need to use Battery EMUs. What about a future Phase 2? Frequent two way electric service on the Richmond Hill and/or Milton Lines? A new line on the CP Midtown corridor? So many thoughts of what a vision for a regional frequent electric regional system for the Greater Toronto and Hamilton Area can be. I’ll end off with another of my Japanese train videos, I think Union Station looking like below circa 2035 already makes GO RER with JR-W characteristics a very compelling vision worth pursuing:

Anyway, thanks for reading this really long post that I get reminded to write everytime I visit Kansai. Like, Share and Subscribe for more videos and thoughts about trains from East Asia and beyond.

Or drop me a Ko-fi if you like what you read.

I am late in drafting this but I was also in Kunming, Dali, Lijiang and later Vietnam when Guangzhou Metro opened these lines. I expected these lines to open at the end of the year, as is typical of many Chinese metro systems. By the time I arrived back in Hong Kong, it was already mid-November. So I visited these new lines in December 2025 to see what was up and maybe look at what the conditions look like for lines that still have not opened such as the Line 22 and the Guangzhou-Qingyuan Regional Line.1 These lines were to open at the end of 2025, and I am writing this passage in February 2026. I then had to go to Japan (so stay tuned). Again, Guangzhou opens transit lines faster than I can craft a comprehensive narrative of the experience and history for them. So in December 2025 I hopped on a high speed train from Hong Kong and arrived at Guangzhou South less than 1 hour later:

I headed to the basement of Guangzhou South and was confronted with a metro map showing the new lines that opened in Guangzhou before National Day. This is the third time they had to change the maps in less than 12 months. They will have to change them once again in late December 2025.

Conditions Before Line 22’s End of Year Opening

Before I explored the newly opened metro lines that opened a few weeks ago, I checked out the state of other lines that were to open in a few weeks. Namely, high speed Metro Line 22 North Extension. This project would provide relief to the congested Line 2 and allow for quick access from Guangzhou city center to Guangzhou South high speed railway station. Line 2 was still busy with high speed railway passengers even at noon on a Tuesday in December sans-Line 22 North:

I made my way to check out the existing southern portion of Line 22 which for a few years was a simple high speed shuttle between Panyu (with Line 18) and the South Railway Station. Traffic was always light on Line 22… for now.

However, Line 22 had already put in the new maps for the northern extension opening in a few days.

I would also like to point out how pretty the passenger info on Line 22 and the Guangzhou regional services platform with high resolution LCD and LED screens.

As noted above Line 22 North was not open during my visit and would open in a few weeks. Therefore I had to slog it on Line 2 the old fashioned way. While riding Line 2, I decided to go to Line 22’s future interchange stations to take a look, such as the massive future four line interchange at Xilang Station. This sudden change in plan was made after I passed by the two lines that would take me from Line 2 to Xilang Station, the recently opened Line 10 and the Guangfo Line, where I needed to interchange at either Dongxiao South or Nanzhou Stations respectively. So instead I got off Line 2 at Changgang Station to change to Line 8. At Changgang Station, I caught a Guangzhou Metro employee changing the system map in preparation for Line 22 North opening in the coming days.

I asked them when Line 22 North would open, but unlike last time with the Line 5 East extension, Guangzhou Metro staff did not give me a straight answer. :(

I took Line 8 to the Guangfo Line and arrived at Xilang Station, where all the entrances and connecting passages to Line 22 were still boarded up. This is particularly true in the vast and newly opened Line 10 transfer hall, which has numerous connections that run under Line 1’s structure to the Line 22 platforms on the other side of Xilang Station. I described the structure of Xilang Station in another piece when I toured the opening of Line 10 a few months ago in June.

Line 14 Phase II North

After looking around Xilang Station I headed to my first stop of the new line, which was Line 14 Phase 2 North. The Line 14 Phase 2 North segment runs 8.8 km long from Lejia Road to Jiahewanggang, with 7 stations and is fully underground. Line 14 was supposed to be an express radial suburban line to Conghua, and its design is reflected as such with trains operating up to 120kph and passing loops at stations for express services to overtake local trains.

Line 14 Phase I opened on 28 December 2017 as a long suburban line starting at Jiahewanggang Station and heading northeast up to the neighboring town of Conghua (though Conghua is still in Guangzhou Prefectural City), serving the northern suburbs of Guangzhou on the way. Jiahewanggang Station is a notoriously congested metro hub where Line 14 terminates and connects to Line 2 and Line 3 Airport Branch. After the opening of Line 11 partly relieving Tiyu Xilu Station, Jiahewanggang Station regularly trades blows with Tiyu Xilu for the title of Guangzhou Metro’s busiest metro station. This arrangement meant the huge volumes of commuters from northern Guangzhou and Conghua need to pile onto Line 14, then get dumped onto the congested Line 2 and the legendarily congested Line 3. The Phase II extension would take Line 14 further south, deep into Guangzhou city center, and terminate at Guangzhou Railway Station. There, Line 14 could connect with Guangzhou’s monster city center ring line, Line 11. This allows Line 14 passengers to bypass Jiahewanggang and all of Line 2/Line 3, decongesting the broader Guangzhou Metro network.

During planning, Line 14 was mandated to fulfill a role as an express suburban metro line. This was achieved not only with passing loops (as noted above in the largely elevated Phase I) but also with the line designed for trains to reach 120kph. For underground sections, equipping stations with passing loops is expensive. Therefore, high average speeds in those sections were achieved by wide station spacings instead, with minimal variation in skip stop patterns.

Those considerations noted above meant that in this section, Line 14 express services were planned to stop at all stations like the local services. This led to the entirely underground Line 14 Phase II South Extension segment being designed and built with no passing loops at any stations. Big Tokyu Shin-Tamagawa Line (underground segment of the Den-en-toshi Line) energy. As such, the original Line 14 Phase II plan, being fully underground with no express overtaking infrastructure, opted for very wide station spacings to maintain Line 14’s role as an express line. The initial Line 14 Phase II plan between Jiahewanggang and Guangzhou Railway stations proposed only three intermediate stations. Despite the Line 14 Phase II project having a length of almost 12km and running under the densely populated inner city of Guangzhou, over time, at the request of residents, the planning of Line 14 Phase II included more and more intermediate stations.

These developments added today’s Pengbian (彭边), Hebian (鹤边), Mawu (马务) and Lejia Road (乐嘉路) stations to Line 14’s Phase II plan. With the addition of four more intermediate stations the average stop spacing of the Line 14 Phase II South Extension was halved compared to the original proposal. The final result of the Line 14 Phase II South Extension is shown below:

I assume the planning authorities also relented over time and added more stations to this section of Line 14 as with the simultaneous development of a new concept taking over the role of providing express regional services in Guangzhou: the high speed metro. The impact this concept of high speed metro had on metro planning cannot be understated: the northern extension of Line 14 Phase II (see EIA map above, blue line to Liangkou or 良口 station) was cancelled and replaced with a high speed metro proposal. Rapid connections to Guangzhou city center from Line 14 would be achieved by transferring to high speed metro Line 18 or 22. Therefore, there was less of a need to travel along Line 14 end to end when commuters could just transfer to an even faster high speed metro line. Combined with Line 14 Phase II running under the dense inner suburban Guangzhou neighbourhoods:

I arrived via Line 12 West (future Substack) to Xinshixu Station, which very obviously looks like two stations bolted together in an L shape with different wall panel colors, despite the two parts of the station being built at almost the exact same time. Xinshixu station and Line 12 West opened in 29 June 2025 and the Line 14 North Segment opened in 29 September 2025 just before National Day. So in the span of a few months, the urban villages surrounding the newly built Xinshixu Station went from having no metro station to an interchange between two major lines.

Of the additional stations added to Line 14 Phase II, Lejia Road Station is one of the most fascinating. Lejia Road Station was added to the planning of Line 14 as a normal infill through station. Later on, it became clear that the Line 14 segment to Guangzhou Railway Station would not be completed on time due to complications with Guangzhou Railway Station’s ongoing redevelopment. As Line 14’s actual terminal platforms at Guangzhou Railway Station remained under construction, Line 14 trains needed somewhere to terminate. So Lejia Road Station, instead of being a standard infill underground island platform station, was completely redesigned at the last minute to be Line 14’s new temporary terminal station. Today, it is a side platform station with crossovers and tail tracks to allow for Line 14 trains to terminate and reverse.

Heading up to the concourse of Lejia Road, I saw a lot of displays and found out the other side platform was closed, meaning Line 14 trains were reversing using only a single track and platform at Lejia Road.

The rest of the Lejia Road concourse was covered in promotional displays:

The opening of only the northern segment of Line 14 Phase II, combined with the delays of Line 12’s central segment, leaves a strange tangle of lines north of Guangzhou that does not connect to the Line 11 ring line. As noted in my review of Line 11 when it opened a year earlier, Line 11’s platforms at Guangzhou Railway station were not open. Similar to Line 14 Phase II South, Guangzhou Railway Station redevelopment plans delayed the construction of Line 11’s Guangzhou Railway Station platforms. Both Lines 11 and 14 (yellow and cyan boxes, respectively) share the same eastern part of the sprawling five line Guangzhou Metro station complex in front of Guangzhou Railway Station, which is still under heavy construction:

As such, it was not possible to interchange between Line 11 and Line 2, a line that was also carrying all the people from the dead ended existing segment of Lines 12 and 14. This forced people on Line 14 (and some on Line 12) to pile onto Line 2. Yet those Line 2 passengers could not transfer to Line 11, limiting the utility of Line 11 as an urban city center circulator.

Hopefully, the opening of the Line 12 central segment to Guangzhou University of Chinese Medicine Station on Line 11 would help divert all the traffic (particularly from Line 14) being funneled into Line 2 to instead head into the high capacity Line 11. This should avoid the impact and uncertainty of any delays from Line 14 and Line 11’s connections to Guangzhou Railway Station and reduce a lot of unnecessary transfers like at Xinshixu Station between Lines 12 and 14:

This could also be seen at the interchanges of Line 12 at Baiyun Cultural Square Stations, where I saw large volumes of people transferring between all Lines 2, 12 and 14. All were attempting to navigate the numerous frustrating temporary dead ends that existed in that area of the Guangzhou Metro network described above. When the network is fully built out, many of these inefficient transfer movements should disappear.

Line 13 Phase 2 East

After seeing Line 14, I took the metro to look at another metro project that opened during National Day: the Line 13 Phase II East Segment. Line 13 is envisioned to be a major east west line designed to decongest the Guangzhou Metro network. Currently, east west travel in central Guangzhou is dependent on two very congested light metro lines: Line 5 and Line 6. Line 5 in particular has to be the world’s hardest working light metro line. Serving well over 1 million passengers a day on 6 car Type L trains that are slightly over 100m in length, Line 5 is extremely congested. In 2018, as reported by China Association of Metros the (CAMET), the busiest section of Line 5 was moving 54,000 people per hour per direction (pphpd) during the AM peak hours.

This figure for Guangzhou Metro Line 5 is close to the passenger demands experienced by the JR East Yamanote Line in around the same year, but on short 100m long 6 car trains. The Yamanote Line uses 220m long trains with 11 cars.

To further put things into perspective, the busiest section of the Yurakucho Line has a similar sectional demand to Guangzhou Metro Line 5 (56,000 pphpd). However, the Yurakucho Line uses massive 10 car Tokyo Metro trains while Guangzhou Metro Line 5’s six car Type L trains have slightly less capacity than an 8 car Toei Oedo Line train.

In response to ever increasing congestion on the underbuilt portions of the Guangzhou Metro system (so most of it, really), planners designed the next generation of Guangzhou Metro lines with massive capacity in mind. Lines such as Line 11, Line 18, Line 22, Line 26 and Line 28, which are planned to use large high capacity 8 car Type A or Type D trains. One such line for our story is Line 13, which is envisioned to be a high capacity crosstown city center relief line for Line 5, a highly anticipated addition to the Guangzhou Metro.

History has not been kind to the commuters of Line 5, which faced ever increasing traffic demands. In 2017, Line 13 Phase I opened. Its only connection to the rest of the Guangzhou Metro network was with Line 5. This meant that, from a network perspective, Line 13 Phase I behaved like a branch line that dumped commuters onto Line 5. Ironically, the opening of the first phase of the Line 5 relief line made Line 5’s congestion worse. This development was also comical, as Line 13 uses massive eight car Type A trains but terminates at Line 5 and funnels everyone to transfer onto Line 5’s undersized 6 car Type L light metro trains. A single Line 13 eight car drum cross section Type A train has the same capacity as two Line 5 six car Type L trains.

Later on, the opening of Line 21 western extension in 2019 included using Line 11 tracks to temporarily allow Line 21 services to connect and terminate at Line 5’s Yuancun Station. This connection with Line 5 was done to avoid Line 21 from becoming disconnected from the rest of the Guangzhou Metro network before Line 11 could open. This also created some interesting engineering at Tianhe Park Station to adapt infrastructure designed for Line 11’s massive eight car Type A trains to serve Line 21’s smaller six car Type B trains. However, it also meant that without Line 11 (and the rest of Line 13), Line 5 was one of the few connections Line 21 had to the rest of the Guangzhou Metro network. Between 2019 and 2024, this arrangement forced Line 21 to dump huge numbers of suburban commuters onto Line 5. It cannot be stressed how both the Line 13 and Line 21 openings used Line 5 as their only connection to Guangzhou city center, creating a massive bottleneck on Line 5 and in eastern Guangzhou.

Line 13 Phase II was supposed to extend Line 13 across Guangzhou, creating a high capacity city center crosstown line and finally providing an alternative route to Line 5. When building a massive metro line under a densely populated city, construction delays are inevitable. As such, on National Day 2025, Line 13 Phase II was only able to open an eastern 8.6km long four station segment that extended Line 13 west up to Tianhe Park Station. The good thing was that the opening of this first Line 13 Phase II segment transformed Line 13 from a branch line dumping everyone into Line 5 to something of a relief line that at least took passengers up to the eastern side of the Line 11 loop, bypassing the need to use Line 5.

So how did this new Line 13 extension look? I arrived at Tianhe Park Station, a massive interchange station between Lines 11, 13 and 21. The station is designed for massive transfer capacity, with Line 11 and 21 having same direction cross platform interchange and Line 13 running underneath with Spanish solution platforms. This station and the adjacent Yuancun Station have a strange history involving the temporary use of Line 11 platforms and tunnels for Line 21 while the rest of Line 11 was under construction. This was all done to allow Line 21 to access the rest of the Guangzhou Metro network while Line 11 was under construction: By dumping everyone onto the congested Line 5 instead. The evolution of Tianhe Park Station is all described in an article I wrote below:

When I was visiting Tianhe Park Station after touring the Line 5 East extension back in 2023, only the concourse of the Line 11 and Line 21 platforms was open. The connections to the Line 13 station box were all boarded up with nice, glossy, permanent looking pale green panels, seen in the background of the video I took in 2023 below:

The video below is one taken during my December 2025 visit. I was now standing behind the huge circular LED display shown in the earlier 2023 video above. The massive concourse of Tianhe Park Station was now used to handle the huge anticipated volumes of traffic from Line 11 and Line 21 to/from Line 13 to head into the city center (and vice versa) but, more importantly, to avoid using Line 5. You can see in the background of the 2025 video below that the pale green panels from the 2023 video have been completely removed to reveal a series of “arches” that lead to the newly opened Line 13 concourse.

Line 13’s station box plunges into the station boxes of Lines 11 and 21 at a 45 degree angle in plan view. The station was built cut and cover in staggered stages, with the station box of Lines 11 and 21 built first and Line 13’s starting a few years later.

The top left picture below shows that the “arches” are the interface between the cross platform station box of Line 11 and 21 and the Line 13 station box. The row of arches is actually a series of access points punched right into the side of the Lines 11 and 21 station box to reach the separate but also wide Line 13 station box, all to allow for maximum permeability for transfers between the three lines. You can see based on the orientation of the ceiling panels, that Line 13’s station box is oriented at a skewed angle vis-à-vis Line 11/21 station box. The other pictures below show Line 13’s wide concourse, which hosts Spanish solution platforms below.

A few months ago, every person crossing here in the video below would need to squeeze onto Line 5. Now they are using Line 11 and interchanging at Tianhe Park to Line 13 instead, completely avoiding Line 5.

When you reach Line 13’s platforms, no one expects the Spanish solution. However, only the eastbound track is used; the other “westbound” track is used for storing trains, as shown by the lack of wayfinding on the PSDs serving that track (bottom left picture).

I head to the west end of the Line 13 alighting platforms to look at the new very clean westbound “tail track” of Line 13, begging to host trains heading into Guangzhou city center. These tracks would be put to use when the rest of Line 13 Phase II opened.

I took the next departing Line 13 massive eight car Type A eastbound train to explore the other stations. I sat in the last train car with a Guangzhou Metro employee carrying around a digital track gauge and total station (with mount to do top of rail referencing). This brought back some good track engineer memories.

The rest of the train was quite full leaving eastbound from Tianhe Park Station; again, everyone pictured here on my massive eight car Type A Line 13 train would have had to squeeze onto Line 5 trains with half the capacity a few months earlier.

I took Line 13 east to another interesting station that opened with the Line 13 Phase II east section, Tianhe Zhucun Station. Due to complicated geology in the area and the need to reduce disruption to an overpass, a BRT line and an urban village above, the station was built in two parts. The southern side of the station hosts Line 13’s eastbound platforms, which are built cut and cover. The cut and cover structure allows for wide platforms and a full length concourse on top. The cut and cover structure is capped with arches to support the soil cover on top and allow for a columnless, airy design in the concourse.

The westbound platform is a mined tube on the northern side of the station and is connected to the eastbound cut & cover section of the station using a series of north south cross passages. The westbound platform does not have a concourse, so all station access is done via crossing over to the concourse above the eastbound platform.

On top of Tianhe Zhucun Station, the Guangzhou BRT runs undisturbed, only disturbed by road traffic breaking the rules of the reserved BRT lanes.

Then I realized it was really late, and I rushed back down to take the metro to Guangzhou East Railway Station and catch a regional train back down to Hong Kong.

So thanks for reading this late post. Like, share and subscribe for more transport news and analysis about the world’s fastest growing metro system, which grows faster than I can type.

Or drop me a Ko-fi if you like what you read.

Arrival

I had been to Hanoi in 2018 and Ho Chi Minh City in 2024, but this time I arrived by sleeper train from Yunnan Province. As I outlined in a Twitter thread. Overall the trip was nice and cheap; the refurbished trains are clean and comfortable.

However, it is a huge indictment of Vietnam Railways that it takes 8 hours to get from Lào Cai to Hà Nội, a distance of just under 300km, or the distance from Toronto to Detroit or Tokyo to Nagoya. This is why sleeper services are available for such a short distance in the first place. The unelectrified colonial meter gauge single track railway is really showing its age and is not fit for purpose in the new Vietnamese economy. The journey consists of a lot of slow rolling through villages and thick jungles, occasionally stopping to wait for trains to pass. Đường sắt Việt Nam needs actual Đổi Mới not a train service named Đổi Mới.

Of course, since my visit and the posting of this article, Vietnam broke ground on a new standard gauge electric double track mainline that will connect Hải Phòng - Hà Nội - Lào Cai with the China Railways network into Yunnan Province. This will be a boon for trade. There is a lot of trade between Vietnam and China and some of that trade is already going by rail.

The Joy of Hanoi

So I arrive in Hà Nội and eat lots of Xôi, Bún chả and Bánh mì Cột Đèn in the bustling city of 6 million and growing people. Upon leaving ̶G̶a̶r̶e̶ I mean Ga Hà Nội, I already see signs of Hanoi Metro development with a large construction site for the eastern segment of Line 3. Hopefully Vietnam Railways improves their services so the Hanoi Railway Station becomes the major destination for the Metro Station. Right now Hanoi Railway Station is large and pretty clean but quite quiet, with minimal intercity service consisting of a few departures per day, not unlike a major Canadian intercity railway station.

Back when I visited in 2018, the Vietnam Military History Museum was near the Hanoi Flagtower. The congested museum was moved to a less congested site out in the suburbs. The exhibition is now housed in a massive new building fit for the confidence of new Vietnam. My Vietnamese friends raved about the new museum and highly recommended it to me. So one day I headed out to southwest Hanoi to see the new Museum. I took a xe buýt from the Hanoi Old Quarter out to the suburbs. In Hanoi, Vinfast has been manufacturing battery electric buses and they are deployed in E-xx series routes. The buses are nice, clean, feel a bit Chinese, have charging ports and are a joy to ride in as we cruised through inner city Hanoi. Some routes (I think it was E-11) are even free to ride!

When riding the bus we even pass under a section of the Hanoi Metro (Line 2A) before take the CT03 expressway out to suburban Hanoi, which looks like suburban China. Our bus trip was cut short and we were kicked off the bus because of flooding on the CT03 expressway frontage road ahead and the bus needed to detour away. We ended up walking for over 30 minutes on that expressway frontage road to get to the Vietnam Military History Museum.

My Vietnamese friends are correct, the new Vietnam Military History Museum was cool and it was worth the long bus ride and sudden hike to see.

I took a taxi to the nearest metro station which in this case is Line 2A. Line 2A is a 13.1km long, 12 station fully elevated line built with Chinese assistance. The taxi driver spoke a little bit of Chinese and called me 地鐵先生 (Mr. Metro) after I explained to him that I just need to get to the closest metro station. There were a lot of flooded streets in the area and he wanted direction on a preferred detour route so I told him my deal: “get me to the nearest metro station, I don’t care which one”. Which in hindsight is a strange request but whatever, I need to experience rail transit.

He dropped us off at a metro station and we walked to the station entrance dodging tons of motorbikes rat running on the sidewalk to avoid rush hour traffic.

Tour of Hanoi’s higher order public transport offerings

I enter the metro station by going up some comically long stairs to escape the traffic chaos of Vietnamese afternoon rush hour below:

The station is quite nice; the concourse has this very “aspirational” system map of the Hanoi Metro masterplan which calls for a massive 12 line network. Despite only parts of the 2 lines are operating when I visited. Quite a Chinese move. I am not sure when this will be complete, but Line 2 and 5 are under construction on top of the already operating Line 2A and Line 3, both of which I will be taking a look at today. However, the issue I have with this map is that it does not show what is currently operating vs what is planned/under construction. My family member who was travelling with me looked at the map and said: “JRUrbane we were in Hanoi for a few days and never took the metro anywhere; why did you hide this massive metro system away from me?”

I’m sorry, 90% of the lines on the map do not exist! Comparing the map with what I know about Line 3, the solid lines I assume are elevated sections while dotted lines are underground sections. All this info is great but it does not help the average commuter (or tourist). Throw on top that, the map was showing a lot of proposed lines that are decades from realization, making it a very confusing map. Regardless, I buy single journey tickets from very Chinese looking ticket vending machines. Which gave me very Chinese style magnetic tickets.

Line 2A uses 4 car type B trains manufactured by the obscure Beijing Subway Rolling Stock Equipment Corp. a boutique Chinese metro car builder that is not part of CRRC Group. The station platforms look to only support 4 car trains.

When I enter the train it feels like I’m taking the metro in 2010s Tianjin or something, complete with dot LED line maps above each door. Which makes sense as the trains and the metro project started in 2011 as Vietnam’s second metro project (after Hanoi Line 3) and initially scheduled to be completed by 2016. Four years to build a new metro line is a very typical Chinese metro construction timeline. However, the project underwent huge issues: construction was only finished in 2018, the trains were delivered from China in 2017 and spent most of that time sitting around as issues with testing dogged the project. Line 2A only opened in 2021, a delay of 5 years.

Like older Beijing Subway rolling stock you can look into the cab:

As shown in the video above, I take the train north to the terminus Cát Linh. In a strange design choice this station is an elevated stub ended station with Spanish solution platforms (which no one expects) and seems to be the only station in the Hanoi Metro network to have some form of platform edge barriers. The platform barriers look to be able to have actual doors installed in the future. There I also learned that the Hanoi Metro accepts Visa.

It is very strange that this terminal is designed to be stub ended so any further northern extension will be difficult to implement. Across the street from the blind end of the stub terminal is a large block of tube houses with narrow streets. I doubt that Hanoi can easily extend Line 2A anywhere further north, a strange design to say the least.

According to the Hanoi Masterplan Line 2A is supposed to interface with Line 2 which is routed through the Old Quarter in a reverse branch shooting off a few stations south of Cát Linh. Luckily there will be an interchange here with the east west Line 3 but the underground section of Line 3 is still under construction. So as of 2025, the Hanoi Metro is a network of two disconnected lines.

If it was not clear before, the guideway of Line 2A is very Chinese in design, using non-segmented prefabricated prestressed concrete box girders.

In addition to the future under construction Line 3, the Hanoi BRT also passes through the area of Cát Linh. The World Bank was involved in this project and the resulting design was (rightfully) lampooned by the ITDP. The line is a “closed system” using dedicated rigid body buses equipped only with “wrong side” high level doors to interface with the median BRT platforms. Being a closed system BRT buses can not serve regular stops and regular buses do not serve BRT stops. This is of course not best practice and the usage shows: Ridership looks “meh” for a dense Asian city starved of higher order public transit. Hanoi should opt for something closer to the Nanning or Guangzhou BRT, an “open system” using regular buses with passing loops for express services as advocated by the ITDP.

In addition, the stations are only double bays and BRT buses do bunch as there are sections of the BRT line, like north of Cát Linh, that BRT buses need to run in mixed traffic. This leads to drivers having some hesitation when entering the station as they are not sure how many buses are ahead of them and what BRT station bays are filled up.

The buses are relatively frequent (every 3-5 mins during PM peak) but there was bunching. Again, for a key corridor in such a dense city I expect usage to shoot through the roof but really all I see is motorcycles and cars using the bus lane to dodge traffic and/or make a U-Turns.

As noted above the underground segment of Line 3 is still under construction so I had to ride a congested minibus from Line 2A’s Cát Linh terminal to the nearest Line 3 station. At times the minibus was paralleling the BRT.

Arriving at the nearest Line 3 station, Cầu Giấy, you can see from the design that the project consortium is from a completely different country than Line 2A. Line 3 is built using European (French) expertise. The Line 3 project started construction in 2010 and the elevated portion opened in 2024. Construction of the underground section of Line 3 continues to this day and will hopefully open in 2027, completing the 12.5km long, 12 station phase 1 of the Line 3 project. This makes the Hanoi Metro Line 3 almost as fast to build as the recently opened Phase 1 of the Eglinton Crosstown LRT project in Toronto.

Unlike Line 2A they do not have an aspirational but more down-to-earth map consisting of the two active lines. However, they still do show the unopened under construction underground Line 3 segment. Hence why Line 2A and 3 connect on the map when, based on my experience above, that is clearly not the case. When arriving at Cầu Giấy all fare machines were out of commission so I had to line up in a long line at a manned booth to buy tickets and the station staff on duty didn’t speak English… or Chinese (not saying they had to). With some pointing and gesturing I got my single ride tickets to the next station and paid using credit card. Ticket-wise unlike Line 2A the line uses RFID tokens. When Line 2A and Line 3 connect I’m not sure how the ticketing situation will work with both lines using different fare media. I guess being that Line 3 is underground and Line 2A is elevated at Cát Linh, some sort of out-of-station transfer will be used? Regardless, I excitedly get my tokens and head up to the platform to board a waiting train.

The trains used on Hanoi Metro Line 3 are 4 cars long and manufactured by Alstom, and they look very European. In terms of passenger information there are just two high resolution LCD screens on each end of the car with a sticker line map over each door. A bit austere if you ask me given the train seems designed later than Line 2A trains. Route information is also shown on the train’s onboard LCD TV screens, which are also advertising the Alstom train you are riding on.

The next station (Chùa Hà) you can see the standard station canopy used for a bunch of Line 3 stations which are quite light and airy with a surprisingly high level of intricacy. Line 3’s platforms look to be able to handle 6 car long trains in the long term.

I head out of Chùa Hà to take the bus back to the Old Quarter. Below is the street presence of the elevated Line 3 whose running lines use prefabricated U-shape tub girders.

Nation Building

So what do I think of all this? Being JRUrbane I can’t help but think about costs and state capacity. I went to Ho Chi Minh City in 2024 and caught a glimpse of their up and coming metro Line 1 that opened a few months later:

When planning the trip I was hoping that I could arrive when the line opens. However, all throughout 2024 there was serious “will it, will it not open” situationship vibes. Similar to the Eglinton Crosstown LRT in Toronto, for months Ho Chi Minh City Metro Line 1 looked very complete but did not announce an opening date for months. Alas, I arrived in April 2024, had a lot of fun in HCMC, but the HCMC Metro Line 1 actually opened months later in December 2024. The line is mostly elevated, 19.7km long with 14 stations and built with Japanese assistance.

So at the time of writing, I have personally interacted with all of Vietnam’s current active metro lines. I come to the inspiration of drafting this article. I talk a lot about state capacity and reducing transit construction costs. A common comment I get when I introduce projects in China, Spain, France or S Korea is from countries that can not for the life of them build transit projects cost effectively (Canada, US, UK etc). The common refrain from these commenters is to hire China, Spain, France or S Korea to build projects in Canada, US, UK etc. The issue I have here is that hiring a contractor/consortium from your preferred country to “bring expertise” may not be sufficient in providing the outcomes you desire. Vietnam is a prime example of this.

So here we have Vietnam entering its first foray into metro construction in the 2010s. Whether it was intentional or not the first three projects Hanoi Metro Line 3, Hanoi Metro Line 2A and HCMC Line 1 all went to different consortiums backed by European, Chinese and Japanese expertise and financing respectively. Was the Vietnamese government emulating China’s model of building state capacity and institutional learning during the 1990s? China with their, at the time, nascent high speed railway program essentially invited and collaborated with as many foreign countries identified with a track record of success as possible. Then, in a strategic and structured approach, China exposed it’s own engineers and planners to as many different ideas, standards and techniques from overseas. Iterating and developing on what was learned to create something that can work in China, and perhaps in the future surpass what was found overseas. Is this what was happening here in Vietnam? Regardless of whether this was a deliberate strategy by Vietnam, I wanted to see what was the final outcome of this three way “battle”. The result is shown below in international dollars (INT$) in an attempt to adjust for local wages and costs:

The reported data shows remarkable internal consistency: more stations with longer platforms and underground segments are more expensive. This is logical. That is the issue, it could mean that no matter who the Vietnamese hires to do the work: the Chinese, the Japanese, the French or even the Martians; the result will still be “Vietnamese Costs” which sadly are on the high side globally. The Vancouver Canada Line, which has more underground running but smaller trains was built cheaper. Wuhan Metro Line 1 phases 1 to 4, which has the same specs as Hanoi Metro Line 2A, such as being fully elevated supporting 4 car Type B trains (and built using Chinese standards), was built for just over half the cost at 95 million INT$/km. On top of cost issues there are the timeline issues: all three projects took well over 10 years to build.

This means in Vietnam’s case, they can change the nationality of the contractor/ consortium all they want but there is clearly something that is holding all costs and timelines to be consistently on the high side. The only thing these projects all have in common is they are pretty typical, mostly elevated metro lines in Vietnam. Meaning the consistent costs could be related to the regulatory environment, client-contractor culture, typical approaches to conflict resolution etc. of Vietnam. Whoever contracted to build the project will only have a minor effect on costs.

This leads to the nebulous concept of state capacity; I view institutional learning as a core competency of state capacity, a form of perpetual Đổi Mới if you may. In theory, such an environment would tease out the costs of the consortium’s home country as you are learning and adopting their practices, assuming scope and technical difficulty remains the same. You can hire the greatest practitioners from the greats in public transport planning and engineering: Spain, Turkey, France, Japan, S Korea, Singapore or China. But if your procurement strategy does not clearly encourage capacity building or there is a lack of administrative capacity to retain that new talent, that knowledge will flow through and out like a leaky No. 3 sieve. These three Vietnamese metro projects all started in the early 2010s, yet they all ended up at similar high costs regardless of which country assisted. Looking forward, the question now is: will the learnings from these three projects be applied in the next wave of Vietnamese projects to create positive cost outcomes?

Next Steps

Vietnam is currently working on a number of new rail projects with ground recently broken on Hanoi Metro Line 2 and 5, built with Japanese and Chinese technical assistance respectively. Ho Chi Minh Metro Line 2 is being developed with more in-house domestic expertise and financing. The projected costs are noted below, adjusted for local wages and prices, and are disappointingly high:

Of note, there is more underground running in this next wave of projects which obviously drives up costs. However based on the data above, Vietnam is building mostly underground urban lines in relatively poor alluvial soil conditions at higher costs compared to Bangkok’s MRT projects like the Orange Line or Purple Line North and South Segments. Granted Bangkok isn’t really the paragon of cost effective metro construction either. However, Bangkok’s geological and hydrological conditions are similar if not worse than either Hanoi or Ho Chi Minh Cities yet Bangkok’s lines are built for larger, higher capacity trains.

Most disappointing of all is Hanoi Metro Line 5. Line 5 is mostly running at grade in the median of the CT03 Thăng Long Avenue/Expressway with wide near 2km station spacings. Which is also the same expressway I got dumped off onto the side of when taking the bus to the new Vietnam Military History Museum.

The video below shows the flooding on CT03’s frontage road that was preventing my bus (and me) from reaching the Vietnam Military History Museum. The expressway mainlines of CT03 (where Line 5 will run in the median) are behind the trees on the left:

Expressway CT03 already has a wide reservation in the median for Line 5 to run at-grade down. I do not know why this line is so expensive. At over 200 million INT$/km it is more expensive than the Montréal Réseau Express Métropolitain project. Below you can see the massive green space reserved (the grove of palm trees) for Line 5 as I turn on to Expressway CT03.

Hanoi Metro Line 5 will run between and parallel to the pair of bridges shown below. There is no reason for this line to reach 200 million INT$/km in cost.

Of course on top of metro projects there is also the complicated saga of Vietnam’s flagship high speed railway project, the North South Express Railway, running the length of the country from Hanoi to Ho Chi Minh. The HSR project originally favored the use of Japanese Shinkansen technology; stalled due to sticker shock of the high price and later tried courting Chinese involvement; secretly tried to outsource development to Vinspeed of Vingroup; got caught from a leak and officially confirmed Vingroup’s interest in involvement; after which Vingroup then signed an MoU with Siemens for HSR development assistance only for Vingroup to get pushed out of the project weeks later. Now everyone is patiently waiting for any new developments on this HSR project that I view as one of the most promising HSLs in the world.

Vietnam needs to step it up. I and many tourists are in awe at the sheer volume of motorcycles running down Vietnam’s streets. However there was a (naïve) implicit understanding that this is transient. It is assumed that Vietnam’s large cities, like the large cities of developed East and South East Asia, will eventually converge to rely on a large efficient public transport network as the country continued to develop. With the recent rise of Vinfast as well as Vietnam’s rapidly growing economy, the Vietnamese are quickly switching to cheap Vinfast cars which will lead to legendary growth in congestion of Vietnam’s cities. Public transit has been hollowed out, with less than 5% of trips in Ho Chi Minh City made on public transit. When I visited in 2018 I noticed Hanoi’s streets essentially were an ocean of motorcycles. While Hanoi in 2025 had a lot more cars on the streets, many of them Vinfast EVs, is this the future of mobility that Vietnam aspires to?

So cảm ơn for reading this post, which is another change in geography. Like, share and subscribe for more transport news and analysis from around the world.

Or drop me a Ko-fi if you like what you read so I can enjoy a cà phê.

In this Part 2 article, I will be covering my experience and thoughts on Line 3, a new high capacity metro line that also has a massive new high speed railway station, Beijing Chaoyang Railway Station attached to it. I start my day strong with a cheap beef noodle and lamb skewer, the Breakfast of the Champions:

I head out and take a bus to a secret location which was under a lot of renovation upon my arrival for a big announcement. This might be a future Substack post… (subscribe for more)

Getting to and experiencing Line 3

But more importantly (depends on your perspective I guess), after touring the future Substack post I again take the bus to the nearest Beijing Subway station which is Tuofangying on Line 12.

I then take Line 12 Eastbound to the terminus of Line 12:

The terminus of Line 12 is Dongbabei station. Dongbabei provides same direction cross platform interchange with Line 3, another of Beijing’s “half consist” capable lines and the focus of this Part 2 article. Like Line 12, Line 3 uses 4 car Type A trains but all platforms are fully built out to support 8 car Type A trains.

Currently, both Line 12 and 3 both terminate at Dongbabei Station.

But in classic Chinese fashion Lines 3 and 12 both have projects that will extend both lines further east. Line 12 possibly being extended to Tongzhou and Line 3 to Caogezhuang where it can interchange and temporarily through operate with Line 20 Phase 1 also known in planning documents as Beijing Subway Line R4. Line 3’s eastern extension started construction in late December 2025. I head upstairs to look at Dongbabei’s concourse which is semi-underground as the entire cross platform station complex was built cut and cover.

The History of Line 3

I think it might be beneficial to stop here and go over the history of Beijing Subway Line 3, as sharp-eyed followers of the Beijing Subway will know that for its entire 50+ year existence from 1971, the Beijing Subway did not have a Line 3 until late 2024. In fact the first time China ever opened a Metro Line 3 was Shanghai in 1997 more than 25 years after opening of China’s first subway. Shanghai also opened China’s first Line 4 in 2001, Line 5 in 2003, Line 6 in 2007, Line 7 in 2009, Line 8 with Line 9 in 2007. It is only until 2008 when Beijing gets to have China’s first Line 10, then China’s first Line 11 goes back to Shanghai in 2009, Line 12 in 2013; back to Beijing for China’s first Line 13 which opened in 2002 followed by Beijing having China’s first Line 14 in 2013 and Line 15 in 2010; back to Shanghai for Line 16 in 2013, Line 17 in 2017 and Line 18 in 2020 and finally back to Beijing for China’s first Line 19 in 2021. After that, things shifted south with Shenzhen having China’s first Line 20 in 2021 and Guangzhou with China’s first Line 21 and Line 22 in 2018 and 2022 respectively.1 The following section I draw heavily on information and graphics from a blog post by Hat600.

From the Beginning

In the 1950s in the dawn of the People’s Republic of China, Beijing had planned a metro system using Soviet experts. The resulting plan was clearly a hybrid of Soviet and Chinese principles. A forced Soviet triangle with a Chinese style grid+ring line network. Line 3 was one of the North-South grids starting from what looks like today’s Liudaokou Station in the north heading down to just south of where today’s Jijiamiao Station is.

Going it Alone, in Chaos

The outbreak of the Korean War as well as internal political turmoil stalled China’s first subway project for years. Beijing’s subway planning was again paused during the disastrous Great Leap Forward campaign and resulting famines. As interest in the subway project picked up again, Moscow started to withdraw all of their subway technical experts after the Sino-Soviet split in 1961. Planning resumed after the Great Leap Forward campaign was officially terminated in 1962, using Chinese “experts” presumably armed with technical notes leftover from the Soviets and maybe some pointers from the East Germans. The result was a more modest plan, departing from the Sino-Soviet plan, was published in 1965 as shown below:

This 1965 system plan did not use line numbers but divided the system into Phase I, Phase II and Phase III projects (top right). Not to be confused with the modern retroactive Beijing Subway Phase I, Phase II and Phase III expansion plans created when China formalized its metro planning procedures in mid 2010s.

It seems from a network planning perspective no liberties were taken from the previous teachers in this 1965 plan. The new plan did not look anything like what was proposed with Soviet help. There are common themes such as a line from Xizhimen to the Summer Palace area as well as a long straight line from Fuxingmen to the Shijingshan area. These areas traditionally had huge military importance, showing the military-civil fusion part of Soviet metro planning remained in Beijing. However lines generally ran in straight lines under wide stroads (or where Beijing’s city walls slated for demolition for a wide expressway) unlike the Soviet plan which lines seemed to be drawn with minimal consideration of what is on the surface. This is due to the change in preferred metro construction/doctrine from classic Moscow Metro deep tunnels and stations for military-civil defense to shallow cut and cover running lines and stations that are for military-civil emergency mobility/evacuation.

Phase I, China’s first subway, finally started construction in 1965 just in time for the Cultural Revolution. Phase I mostly ran on western Chang’an Avenue but swings south at Fuxingmen to avoid the section of Chang’an Avenue running through the old city in front of the Forbidden City and Tian’anmen Square due to technical and (maybe) security concerns. Still, most of what was labelled as Phase I ended up being part of today’s Line 1, all of Phase II ended up being part of today’s Line 2. I guess this is just borne out of convenience. However most of Phase III ended up being part of… today’s Line 4. Hence the beginning of Line 3’s meandering planning history.

The Subway Line that Beijing Forgot

With the winding down of the Cultural Revolution and some other quite messy early 1970s Chinese political events, it was time to divert that revolutionary fervor into subway planning again. An updated Beijing Subway masterplan was published in 1974. At the time Phase I from the 1965 system plan was in operation with construction grinding all through the chaos of the Cultural Revolution and Phase II was under construction.

In this updated 1974 plan, the yet to start Phase III proposed in the 1965 plan was reformed into a proposed Line 3. This was the first official reference of Line 3 after the Sino-Soviet Split. The scale of Line 3 was massively expanded compared to the “1965’s Phase III Project” now extending west from the Fragrant Hills (today served by the Xijiao LRT) to Beijing Capital International Airport, creating a U-shaped crosstown line.

There was so much faith in the certainty of the 1974 plan that the, at the time under construction, 1965 plan Phase II (Today’s Line 2) built extra reserved platforms at Dongzhimen and Dongsi Shitiao Stations in anticipation for the implied next metro project, Line 3, as proposed in the 1974 plan. Phase II opened in 1984 from Jiangguomen to Fuxingmen as a line isolated from Phase I.

First Setback for Line 3

After the opening of Phase II, limited national resources delayed and constrained any further subway expansion in the 1980s just as China’s economy was beginning to take off in the advent of the Reform and Opening up Era spearheaded by Deng Xiaoping. Line 3, as proposed in the 1974 Plan, was assumed to start construction after the opening of Phase II but was consistently delayed as officials were preoccupied with quickly adjusting the newly opened Phase II into the loop line of Line 2 today after complaints from commuters.2

Since adjusting Phase II into a proper loop line that connected with Line 1 at Fuxingmen, Beijing Subway’s only interchange, congestion at said station from transfer traffic between Lines 1 and 2 as well as traffic coming from buses heading east on Chang’an Avenue motivated officials to again divert attention away from starting work on Line 3 in the late 1980s. This newfound motivation to extend Line 1 through the old city and ignore Line 3 was also bolstered by officials and engineers now being confident that cut and cover construction of Line 1 under Chang’an Avenue east threading between the Forbidden City and Tian’anmen Square was technically feasible. It also helps that the Japanese were willing to finance (maybe provide technical assistance too?) this Line 1 extension project through ODA (Official Development Assistance) to China as opposed to Line 3.

As an aside, China also got some Japanese rolling stock from the Tokyu Car Corporation as Japan got involved with the Line 1 extension project. The BSSA-3000 series (externally known as the M series or simply Beijing Subway 3000 series) are one of the few wholly foreign imported subway trains in Chinese history. These trains are rumored to be based on the Odakyu 9000 series with the look of typical mid 1980s Tokyo commuter rolling stock like the later batches of the Tobu 8000 and 10000 series or the Odakyu 1000 and 8000 series.3

With the Japanese on board, officials delayed Line 3 and focused resources on extending Line 1 also known as the Fuxingmen-Bawangfen (Fuba Line) project. Bawangfen was another name for the Sihui area (today’s Line 1 Sihui Xi and Dong stations) which was the terminus of Line 1 until recently. First, the line was extended one station east to Xidan known by the Japanese as the Beijing Subway Construction Project. This was followed up by the rest of the Fuba Line known by the Japanese as Beijing Subway 2nd Phase Construction Project. The project was mildly ambitious in scope with all Fuba line stations built to support 8 car trains instead of the 6 car standard used in the 1965 Phase I, Phase II and Phase III subway plans.

Of these two projects the section from Jiangguomen to Thermal Power Plant Stations (today’s Dawanglu Station) was proposed as a Phase III project way back in the 1965 plan.

While attention was diverted to the Fuba Line Project instead of Line 3, planning work for Line 3 did not stop, another larger expansion plan for the Beijing Subway was unveiled in 1993:

In this plan the conversation for Line 3 reached its zenith in ambition as the eastern proposed alignment was altered in the 1993 expansion plan to swing north serving Wangjing before going to the airport becoming a monster 48km long east west crosstown line.

Wangjing (at the time) was proposed to be a satellite CBD of Beijing and today has become an interesting place architecturally but still a bit dull. This north south alignment of Line 3 through Wangjing proposed in 1993 was is approximately replaced with Line 14 today.

In addition to changing planning priorities of Beijing shifting Line 3’s alignment in the 1993 plan, Line 3 was proposed to use a prototype of what would be a Type A car (22.6m long by 3.0m wide cars) in 6 car sets using 3rd rail traction power.

Unsurprisingly, officials balked at the perceived high cost of construction for the quite ambitious Line 3 proposal even if it was not built with Type A standards. Type A railcars would not appear in the Beijing Subway until the opening of Line 14 more than 20 years later. Those 1990s Beijing apparatchiks need to come to capitalist North America today to see what actual high cost of rapid transit construction is. This reluctance to build Line 3 was reflected in an updated Beijing Subway plan unveiled in the late 1990s:

Line 3’s scope to the airport and Fragrant Hills was cut back and replaced by a branch of Line 13 and a northern segment of Line 9 respectively. No doubt Line 3’s Type A design was dropped back down to Type B. This left Line 3 as proposed in the late 1990s as a “stubby” 30km long line from Wangjing to what looks like Sijiqing Qiao Station on Line 12 today.

Another Setback for Line 3

It is no wonder that after completion and adjustments of the Phase II Project (Line 2) as well as extending Line 1 as the Fuba Line, the proposition of the next project, Line 3, as proposed was not viewed as a very high priority. In addition, changes in Beijing’s urban population distribution and lack of funds have again shifted the planning priorities of Beijing Subway in the late 1990s. Beijing’s growth was now happening in the north and east of the city. The still expensive urban underground east-west stubby Line 3 was again deprioritized for cheap surface “city rail” extensions serving the Northern and Eastern suburbs.

Sharp eyed readers will notice such “city rail” lines are already showing up in the late 1990s plan posted above. These “city rail” extensions give us Line 13 and a further extension of Line 1/Fuba Line east with a new Bawangfen-Tongzhou Line (Batong Line) in the early 2000s. Some JRUrbane lore, the first time they visited Beijing was in 2003 after these city rail lines opened. Trains and stations were not air conditioned, there was no PSDs, you manually bought subway tickets from a booth that are manually hand stamped at the fareline and the Beijing Subway map looked like this:

There young JRUrbane did wonder why Line 3 did not exist when there was a Line 13.

One World, One Dream… without Line 3

On the verge of opening the aforementioned city rail lines, Beijing in theory should be again resuming detailed planning work on Line 3 as the next subway expansion project… Beijing won the bid to host the 2008 Summer Olympics in 2001. North-south connections are prioritized for the Olympics. This led to construction starting on an Airport Express Line and Line 4 (both which took over parts of Line 3 in the new late 1990s plan) as well as Line 5, Line 8 and Line 10 as work wrapped up on the “city rail” lines.

Line 4’s takeover shifted the planned alignment for the east west Line 3 further south in later plans but more importantly for this story, this -say it with me- further delayed the implementation of Line 3.

Okay. So here's how a synthetic CDO works. Let's say you don’t build Line 3…

After the opening of at least Line 5, 8 and 10 in time for the 2008 Olympics (Line 4 opened a year late) and the ongoing success of the Olympic Games in China over the summer, the bourgeoisie class of the United States of America were using most of society's wealth to play around trading high-risk subprime mortgages in mortgage-backed securities and collateralized debt obligations (CDOs). Which I was told, and I quote:

This reckless trading was the US bourgeoisie class’ attempt of preserving the value of their private property or at the very least maintain their hold on capital. Done to perpetuate of their economic dominance of American society after a collapse in US real estate asset values led to the “Great Recession” economic downturn in 2007. This gamble ultimately failed and led to the 2008 Global Financial Crisis.

As a response, the Chinese Government implemented a massive 4 trillion RMB stimulus plan in 2008-2009 in a Keynesian attempt to shore up its economy. This record breaking stimulus plan included building thousands of kilometers of new subway lines across China. The Beijing Subway planning went into overdrive, armed with subway expansion masterplans that are updated every year. Most of the stimulus plan focused on fast tracking shovel ready elevated radial suburban metro line projects that are cheaper to build and easier to design.

These new prioritized suburban lines such as the Daxing, Fangshan, Changping, Line 15 and Xijiao Lines will massively expand the Beijing Subway to suburban districts. In addition, a major east west high capacity urban crosstown trunk line was also fast tracked. Will it be Line 3?

The Mandate of Heaven against Line 3

The latest late 2000s Beijing Subway expansion masterplan delineates two east west crosstown lines running through the city center of Beijing: Line 3 and 6. However, unresolved issues with how to route these two key city center crosstown lines through the old city continued to delay both projects. The more southern Line 6 would run between the back of the historic Forbidden City and the front of Jingshan Park. With stimulus money coming in planners had to recommend one project for prioritization, since the two lines are parallel and very close to each other. As with the history of the Line 3 section in this really long post, I will draw heavily on information and graphics from a blog post by Hat600.

Line 3 maintained a proposal of a more northerly and “stubby” route in the 2000s. On top of its short length, this proposal of Line 3 was routed far from many major trip generators making it an overall unattractive line utility wise. Its only saving grace was the western end which connects with Line S1, a project with strong political support for quick implementation and wanted a straightline connection to key Beijing employment nodes. This is in contrast to Line 6 which touched the northern bits of the Guomao and Financial Street CBDs and was logically a higher value line but did not directly connect with Line S1. Building both lines at the same time was not an option.

To get a project approved, planners had to address and the contradictions and fold all considerations described above into a single project for prioritization. To prioritize the connection with Line S1 while also connecting to the Guomao and Financial Street CBDs it was proposed to swap Line 3 and Line 6 on each side of the Forbidden City and choose one line to fast track with the stimulus funding.

At the same time cultural protection concerns around the Forbidden City have increased as planning for Lines 3 and 6 were underway. Increasingly, officials have minimal appetite to accept risk to cultural relics in the area being impacted by subway construction. The powerful National Cultural Heritage Administration refused to budge on their ever increasing standards for risk mitigation to cultural relics. To manage these risks, subway lines increasingly got routed away from the Forbidden City. This is the “heart” of China after all, what is a couple-decade-old western underground train technology compared to the splendor of over 500 years of Chinese imperial history. As a response, designers proposed that Line 3 and 6 swapped their east and west portions with an interchange at Di’anmen (地安门). This swung both lines and stations away from the Forbidden City and reduced the risk of damage to cultural relics.

Despite Di’anmen being further away from the Forbidden City the new proposed location was still surrounded by key historical sites. Planners found it difficult to build a massive interchange station in the area which also required connecting to the north south Line 8. Line 8, at the time, was still considering to run under the entire central axis of the Forbidden City. A more out there proposal assumes it is feasible to tunnel Line 6 close to the northern part of the Forbidden City and Line 3 and 6 will instead swap places at Ganjiakou Shopping Center (甘家口商场) in the west to reduce the complexity of building such a large interchange station at Di’anmen.

Ganjiakou Mall Station will host an opposite direction cross platform interchange between the Lines 3 and 6 that will allow for strong same direction correspondence between the two lines reducing the travel time impacts of swapping corridors.

Ultimately to get something fast tracked for construction, a swapped version of Line 6 that swung north and took over the western portions of Line 3 was selected to move forward as a Phase I project. The problem of serving the northern side of the Forbidden City was postponed with the hope that when Line 3 starts construction, technological improvements can be found to allow for subway lines to run close to the Forbidden City with minimal impact. These events also meant the planning/ implementation of urban Line 3 fell to the sidelines once again.

Line 3 Phase 2 has to work around the ever increasing web of considerations involving the set in stone Line 6 Phase 1 and cultural protection concerns from the government and the public. The central section of Line 8 also faced similar planning delays as the aforementioned running under the entire Forbidden City plan became a non-starter. As planning for Line 3 continued, a proposal included a crazy alignment where Line 3 weaved and interchanged between Lines 5, 6 and 8 while still trying to run just north of the Forbidden City.

Planning constraints continued to pile on and engineers were unable to convince the National Cultural Heritage Administration that Line 3 would not impact the Forbidden City. As a result Line 3 was re-routed to run parallel to Line 6 between Ping’anli to Nanluogu Xiang to avoid running so close to the Forbidden City. This was the final configuration of Line 3 submitted for approval.

Return of Line 3

The rapid expansion of suburban subway coverage in the 2010s allowed suburban commuters to overwhelm the core urban sections of the Beijing Subway system even with Line 6 Phase I opening and providing core subway capacity. At the same time surface road congestion in Beijing rapidly increased as the Chinese economy continued to boom. This environment led to the legendary viral video of the crush of Changping Line passengers transferring to Line 13. As such, Beijing Subway planning shifted to increasing core subway capacity in the city center in their Phase 2 expansion program approved in 2015.

Line 3 was finally back in the conversation as a priority line to be built as part of Beijing Subway’s Phase 2 expansion. Construction and approvals for Line 3 Phase 1 were fast tracked as part of a second generation of high capacity crosstown express lines proposed for the Beijing Subway, along with Lines 12, 16, 17, 19 and 20. All these lines are planned to use high capacity 8 car Type A trains not unlike those found in the Hong Kong MTR system as I described in Part 1.

Line 3 Phase I West started construction in 2016 and opened together with Line 12 in December 15, 2024. Line 3 Phase I only covered the eastern segment of the line, the rest of Line 3, as approved, has yet to start construction as a Phase II. As noted in Part 1, I checked out Line 12 and Line 3 Phase 1 in December 19th and 20th respectively.

Exploring History in the Making

With all that background out of the way I enjoy the few days old Line 3 Phase I West, the first step manifested from decades of planning history. I board a waiting Line 3 train which are quite attractive SFM86 series trains by CRRC Sifang.

As I take Line 3 to Chaoyang Railway Station I make my way to the front and see a crowd of people peeking over the semi permanent barrier behind the driver at the ends of the train, similar ones are put up on Line 12:

Everyone wants to see this:

Chaoyang Railway Station

I arrive at Chaoyang Railway Station which has extra wide platforms for massive volumes of people changing to/from national railway trains, and in the future, Line 20.

I head up to the concourse of Line 3 Chaoyang Railway Station which still has props celebrating the opening of Line 3. There is evidence of some interest by the general public on opening of these Subway lines as I was the 3040th person to push the button.

The opening of Beijing Subway Line 3 finally brings Beijing Subway access to the massive Beijing Chaoyang Railway Station, which has been operating since 2021. A vast station with tons of trains heading to the Dongbei Region.

The exterior of the station is even more dramatic (the good Beijing weather helps):

Dongsi Shitiao Station

After soaking in the views of the massive railway station, I take Line 3 to its temporary terminus of Dongsi Shitiao Station, an interchange Station with Line 2 (Phase II of the Beijing Subway 1965 plan). As noted above in my history of Line 3, Dongsi Shitiao and Xizhimen opened in 1984 as a Line 2 stations both with platforms below reserved for a future Line 3 that didn’t arrive for decades. When Line 4 took over the planned Line 3 route in the west, Xizhimen’s platforms reserved for Line 3 were used by Line 4.

The eastern side of Line 3, where it would connect at Dongsi Shitiao Station, was not so lucky. By the time Line 3 Phase I West opened in December 2024 and I arrived to visit, it had been 40 years since those extra Line 3 platforms were completed.

The reserved platforms were originally designed for six car Type C trains, but Line 3 today was redesigned to be a heavy haul crosstown line using 8 car Type A trains. Therefore, the original reserved platforms (picture below, dark structure on top of the blue 三号线老站台 text) were refurbished to be a transfer level between Line 2 and 3. New larger platforms for Line 3 are mined on each side of the original reserved Line 3 station box (light grey structures on each side of the blue 三号线老站台 text). Both platforms are connected by cross passages that have escalators up to the transfer level (structure behind the blue 三号线老站台 text). All this was built while Line 2 trains continuously run above in a 40 year old structure (upper dark structure with red 二号线站台 text):

So when Line 3 Phase I West started construction in 2016 the original 1980s reserved platforms were rebuilt into a transfer level. The track pits on each side was filled in and a bunch of holes were punched in the platform floor for tons of escalators between Line 2 and 3 and various station exits.

So when I arrived at Dongsi Shitiao Station on my Line 3 train, I ran up to check out the transfer level. Below is what the same space in the above picture looks like today but I took the picture on the top of the stairs in the opposite direction. You can see in the center of the picture new escalators down to Line 3 were put in a new hole punched into the platform. New hexagonal LED lightboxes with diffusers evoke the old replaced 1980s light fixtures that used florescent tubes. You can also see the columns are not as slender, most likely due to some strengthening work. Note from here on there will be some comparison pictures before renovations taken by 李睿 from trainnets.com that just so happen to (roughly) match the angles I took.

Both track beds of this center platform station has been filled in with no trace of them ever existing (missed opportunity?). The original design would have tracks hugging each white wall on each side of the “room” with the two rows of columns running down a center platform like many early Beijing Subway Stations. The ceiling was lowered, most likely to accommodate the massive amount of HVAC and other E&M stations today has.

Standing where Line 3 trains would be running in the original station design:

Many Phase I and Phase II Beijing Subway stations generally did not have full length mezzanines but with an intermediate level where ticketing and station exit circulation is needed. Wherever such intermediate space is not needed, the station box has a double height open to below into the platform hall. This layout was generally maintained in the renovated Line 3 transfer level. However the iconic early Beijing Subway style large sweeping stairs that lead up to the intermediate level ticket hall and platform have now been demolished and ticketing areas extended into the platform hall as any access to the ticket hall from Line 3 can be directly made with a direct escalator (left of my image) without using the transfer level:

In the middle of the transfer layer, the ceiling gets really low because that is where Line 2’s perpendicular station structure flies over Line 3’s platform hall. It also seems the foundations holding Line 2 up have been massively strengthened. Compared to the pictures of the original design, the area has become quite claustrophobic.

This is also where you can see the old + shaped 1980s staircase originally designed for direct center platform to center platform transfers between Line 2 and 3. They were maintained to their original design and finishes and used for transfer traffic.

When you go up these start you enter the Line 2 platforms which still retain their original 1980s China vibe.

For decades the stairs to Line 3 were blocked with simple wire barriers, cheekily teasing you to look down into the ghost platforms of Line 3. I visited Beijing in 2003, 2016 and 2018 and passed by the station a few times over the years, aware of the history the stairs in the middle of the station bear. It is almost a spiritual experience to finally be able to walk up and down those same stairs.

As noted above early Beijing Subway stations, lack full length mezzanines and maintain a double height ceiling where an intermediate level is not needed. Line 2 just looks a lot messier.

I head back down to the lowest level to check out the new Line 3 platforms that where mined under all of the station structures I just toured. Below are the connecting passages between the two split platforms that directly run under the old Line 3 platforms.

Line 3 can get crowded:

High Speed Sleeper Home

After experiencing exploring around a bit, I hang around Beijing for a few more days with my relatives (and ride even more metro lines including a maglev). At the end of our trip we all take the high speed sleeper train down to Hong Kong. It was a great experience. I was given toiletries and snacks, the blankets are comfortable and warm.

Racing into the night:

One night you are in Beijing with freezing winter weather and the next morning you are in Shenzhen/Hong Kong with sunny late summer temperatures.

So thanks for reading this delayed post. Like, share and subscribe for more transport news and analysis from around the world.

Or drop me a Ko-fi if you like what you read!

In the 1960s and 1970s Toronto’s transit system was world renowned, “New York but run by the Swiss”. However, since the mid 1990s Toronto’s transport planning was characterized by political grandstanding, infighting and indecision. For 30 years (1995 to 2025) the Toronto Subway network only grew 9.7km, averaging ~300 meters of new rapid transit per year.1 All while the population and economy of Toronto steadily grew during these decades. The city that worked ground to a halt, fractured along geographic and ideological lines.

It was not all doom and gloom during this period (just mostly). Taras documents his interactions with Dr. Paul Mees, a famous Australian transportation scholar who studied how to provide public transportation to suburban auto oriented areas, in shaping his transit exploration itinerary in Toronto. Dr. Mees and later Toronto-native Dr. Jonathan English popularized the concept of the “Toronto Model”.

The Toronto Model

Conventional urban planning wisdom states that any successful public transport route can only exist with supportive land use. The “Toronto Model”, challenges this, demonstrating that providing frequent and available transit services can create a viable transit network despite a low density auto oriented built form that is, in theory, not conducive to generating any public transport ridership. The Toronto Model is essentially a counterfactual to “Density is Destiny”. Yes, higher urban densities help with the viability of transit and really push systems to reach Hong Kong or Seoul levels of modeshare and usage. Yet Toronto’s transit success emerges from a suburban fabric similar to the US. The City of Toronto’s public transport provider (not Greater Toronto think like Tokyo City proper’s Toei) the Toronto Transit Commission (TTC) has created a network of frequent suburban bus corridors with a broad daily service span within the City of Toronto. The result is that Toronto’s transit ridership along each of these individual suburban bus corridors is surprisingly high compared to counterparts in the US (and Australia). These bus corridors are not on the scale of Beijing’s Third Ring Road loop corridor, which Route 300/300快 during its heyday carried over 300,000 passengers per day. However, many of these suburban TTC bus corridors still carry tens of thousands of trips each weekday:

Arguably as a result of the Toronto Model, public transit modeshares in Greater Toronto is just over 20%, a modeshare higher than all metro areas in US and Canada apart from Greater New York. This is in contrast to Greater Toronto (population 7.5 million) having a small 70km subway rapid transit system that has struggled to expand for decades. To put that meager size in perspective, Oslo has a slightly larger subway system serving a metro area of 1.5 million people. Urban Qingdao has a similar population to the Greater Toronto area but a metro system that is 342.3 km (212.7 mi) long with over 150km (93 mi) of additional lines under construction. Beijing’s Third Ring Road loop bus corridor carries far fewer daily passengers today because Beijing built one of the world’s longest loop metro lines to relieve it, the 57km long 45 station Line 10 covers the Third Ring Road and now carries over 1.7 million passengers per day.

Toronto needed a good bus network to make the best of its limited metro network. Running a lot of buses in auto-oriented suburbs all day to feed a modest subway system can produce some respectable ridership and modeshare results; Hence, the Toronto Model. However Toronto’s buses, while frequent, suffer from poor average speeds, long travel times and notorious unreliability; relying on the brute force of frequency to keep the resulting transit experience as passable. The point is that eventually, like Beijing, the ridership built up by all this great bus service will be served by an upgraded higher order rapid transit line like a metro. Toronto’s city center streetcars are even worse, being some of, if not the, the slowest tram services in the world. The internet is littered with a litany of complaints about the unreliability and slow speeds of the TTC’s bus and streetcar operations. Showing that the Toronto Model great as it is, has its limits.

The ridership created by the Toronto Model has created a large enough political powerbase of transit users to advocate for transit improvements in Toronto. Interest is consistently high within Toronto’s local government and advocates for possible infrastructural improvements to address long standing grievances with TTC’s bus services and expanding on the successes built by the Toronto Model. The issue is that this motivation to improve meets the severe dysfunction and disagreement on the planning and actual implementation of transit improvements that has dogged Toronto for decades. Which is no doubt frustrating to Toronto’s transit advocates. Toronto wants it but can not do it. This leads the discussion from the Toronto Model to the Toronto Tragedy.

The Lost Decades (失われた30年)

While suburban Torontonians rode frequent but unreliable and slow buses, rapid transit expansion entered into tumultuous political debates in the 1980s and devolved into paralyzation by the 1990s. As noted in the lede of this article Toronto essentially had zero transit expansion during this period.

While the city still grew, buses were “holding down the fort” the best they can. At the same time, the constant political jockeying on how to improve Toronto’s public transport reached its zenith with the controversial years long battle over the Transit City plan, a controversial “ambitious” transportation plan proposed in 2007 by then Toronto Mayor David Miller. The plan calls for a 120 km (74.6 mi) web of semi underground and surface road median running “light rail transit” (LRT) lines to create a vast higher order transit network that would knit Toronto’s suburbs together.

The sell was that building these LRT lines will cost less, could be built within a generation and bring mass transportation to transport the masses. Largely surface running lines like the Finch West, Scarborough Malvern or Don Mills LRTs were estimated to cost about 70-90 million CAD per km to build, one third the cost of building an underground TTC subway line at the time. Some lines like the Eglinton Crosstown or Jane LRTs would require some sections to use significant tunneling to access denser areas with narrower streets but the argument goes that these projects are still cheaper to build than a fully underground subway line.

Transit City was extremely controversial, it was THE election topic driving the campaigns of the 2010 and 2014 Toronto municipal elections. Supporters called Transit City a cost effective technocratic solution of getting higher order transit to all corners of Toronto ASAP at a low cost. Detractors called it pejoratively as “just a streetcar” (not wrong), “Transit Shitty”, “Transfer City”, the “Gravy Train” (which needed to be stopped) and later used Mayor Rob Ford’s “Subway, Subways, Subways” as a battlecry. Gallons of ink has been spilled, the non stop clattering of keyboards rang across the six boroughs of Toronto, debating online and offline via newspaper op-eds and online forums. To put into perspective how much the Transit City debate was in the early 2010s Toronto zeitgeist, after the election of Rob Ford for Mayor in 2010, Mayor Ford cancelled Transit City on his first day in office.

Despite this, the first and most substantial line proposed in the Transit City plan, the central segment of the Eglinton Crosstown LRT (ECLRT) started construction on its underground grade separated portion in 2011. Mayor Ford constantly attempted to modify the now under construction ECLRT throughout his term as Mayor. Chiefly, half the length of the ECLRT first phase are surface street running sections east of the central segment and had not started construction at the time. Mayor Ford wanted to put these surface street running sections underground. This will transform the ECLRT into a Frankenstein fully grade separated subway using low floor LRVs (like Vienna U-Bahn U6 or Changchun Metro Line 4/8).

During Mayor Rob Ford’s shenanigans with the ECLRT, Transit City supporters tried to revive the remaining first phase of the Transit City plan (Finch West and Sheppard East LRTs as well as the Scarborough light metro rebuild into an LRT). Additionally, in another breakthrough during the Rob Ford administration, the Finch West LRT was approved by council in 2012. All while the LRT vs Subway debate got so toxic for the Sheppard East corridor that the decision was kicked down the road to this day. This ultimately set the stage for today, with two Transit City LRT lines, the Finch West and Eglinton Crosstown LRTs making it into construction.

Toronto continued to debate on other transit expansion projects until Doug Ford, older brother of Rob Ford became premier of Ontario in 2018 and imposed his Subway plan on top of the two already under construction LRT lines. Steamrolling his subway projects over the toxic perennial debates on BRT, LRT or subway for stalled projects like: the replacement of the SRT, Yonge North Subway Extension/VIVA BRT South and Sheppard East Subway Extension/LRT. Even replacing the traditional North American-style Relief Line underground subway proposal with his Ontario Line semi-elevated automated light metro plan. This maneuver was successful, much to the displeasure and skepticism of the urban progressives and/or Transit City supporters, widely painted as a heavy handed provincial intervention.

Transit Wars: Episode IV - A New Hope

It is important to pause at this point to look at how Toronto got here politically. A clash of conflicting visions and values on the future of transit in Toronto. Toronto is a city with a dense vibrant core served by streetcars surrounded by oceans of post war auto oriented sprawl but with great suburban bus services all linked by a meager subway rapid transit network. When I lived and worked in Toronto it was commonly described by planning circles as: “Vienna surrounded by Phoenix”. The city center of Toronto is progressive, urban(e) served and shaped by streetcars with a culture that loved them despite their big flaws. Streetcars gave Toronto some European transit credibility (you know the Vienna part) in contrast to the pejorative stigma North American built form generally has among urbanists. To supporters, Transit City was an opportunity to catalyze urban transformation into the suburbs, making Toronto more of a “Vienna with bits of Phoenix around it”.

The suburbs were comparatively less enthusiastic about this urbanism. Despite the successes identified by Dr. Mees and Dr. English, most in the suburbs still drive the car and opponents to Transit City are concerned that the new LRTs being proposed or “imposed” depending on your political flavor would impact traffic. To Toronto’s (also broadly Canada’s) credit and in no part thanks to the relative successes in Canadian suburban transit (again the “Toronto Model”), pro-car Canadian suburban politics are not as violently against improving transit compared to those south of the 49th. As long as “the transit” is out of sight and has minimal impact, there will be a baseline level of support even among Canadian pro-car suburban politicos and their constituents. Heck, many of them might even use it from time to time. Hence the unwavering emphasis on suburban underground subway expansion advocated by that bloc. Canadian pro-car suburban politicians like Rob and Doug Ford advocating for developing subway rapid transit are in stark contrast to their pro-car US counterparts who have a proclivity to actively sabotage any mass transit service. Regardless of which side of the transit debate Torontonians were on (Transit City or Subway, Subways, Subways) what was missing in both camps was a vision for transit to be both built cost effectively but also provide a high quality of service.

During all drama of Transit City vs Subway, Subway, Subways; Toronto was developing into a “tall and sprawl” urban form. This may have exacerbated political views, with residents now living in two extremes of housing (40+ story towers OR single family homes) with nothing in between. The city center of Toronto gradually redeveloped to look more like Hong Kong, with all its problems (high costs, small living spaces) while the supposed benefits (hyper convenience, insane vibrancy, excellent public transportation) remained elusive. However, this evolution of Toronto’s urban form also produced little pockets of skyscrapers all over the suburbs. This emerging urban form is only served by the aforementioned frequent buses that are frequently unreliable and the city center increasingly looks like Shenzhen but with large sections only served by the slowest tram operations in the world.

My observations are the urban changes in Toronto and political fatigue between LRT vs Subway debates allowed for the emergence of a new “Third Front” in the Toronto transit wars: elevated automated rapid transit lines and surface regional mainline rail, cost effective but fast transport. Not a vision of Phoenix or Vienna but of Tokyo for Toronto. Taras’ in his book was very bullish on Transit City, proclaiming the plan could have turned Toronto into “the Strasbourg of North America” and was quite critical of Mayor Rob Ford and the suburban councilors machinations against Transit City for subways. This third front (me included) would not be sure why Toronto, the center of an urban area with 7.5 million people; now with pockets of areas looking like Seoul or Shanghai needs to be inspired by transit solutions in Strasbourg, France, a European urban area with less than 1 million people but that is neither here nor there I guess.

Regardless, my read is Transit City supporters prioritized aesthetics and potemkin transit coverage while Subway supporters prioritized impact reduction (to themselves mostly) above all else but both groups did not care about what mattered most: Providing high quality rapid transit to everyone at a reasonable cost. At the surface level, the modus operandi of this “third front” of elevated rapid transit advocates essentially is to maximize access to actual rapid transit and build it quickly at the lowest cost.

Canaries on the Tram Line

Looking back it was clear by the end of the 2010s and the “lost decades of Toronto transit expansion” it is arguable that Transit City supporters were quite successful. Two LRT lines from the first phase of the Transit City plan are well underway by the 2020s. As these projects dragged along huge issues began to be identified with the two surviving Transit City projects on the onset of construction and unveiling of preliminary alignment design plates from the consultants. It really felt that the vision guiding these light rail lines is style over substance. The LRT program looks to simply be replacing Toronto’s busiest bus corridors with a tram running down the middle of the road.

Reece Martin of Next Metro, who I would consider part of the “third front” and encouraged me to write this post, has extensively written and discussed about the design issues with the under construction Transit City lines and Toronto’s general history of sloppy tramway operations leading up to their opening. The alarm was clear, if Toronto’s track record of city center streetcar operations, which remember are some of the slowest in the world even with dedicated tram lanes, are any indication, these LRT lines are going to be a walking disaster. This is exacerbated by the insistence from Transit City supporters to only consider “imagined Europe” LRT designs that only run down the middle of streets for urban realm enhancing purposes but will lead to large numbers of traffic conflicts with suburban drivers. So it was no surprise that as these two under construction Transit City LRT Lines that survived the Mayor Rob Ford-era cuts dragged on, it became increasingly clear that the promises of improved speed and operations by the switch from buses to LRT will not materialize.

The Asian Canary on the Tram Line

I have raised this issue of the questionable travel time benefits the Finch West LRT can bring in early 2024. My own digging is based on my experience with riding Route 36 Finch West while living in Toronto years ago and working on surface transportation operations consulting work for the TTC before the Finch West LRT project. This whole personal investigation started because some Transit City fans were celebrating the baseless claim that the Finch West LRT will save 16 minutes of travel time between Humber College and Finch West Subway Station.

The assumption made here is a series of personal data points and vibes for travel time on Route 36 between Humber College and Finch West Subway Station (52 minutes). The 16 min time saving is derived from a propaganda video on the Finch West LRT. Metrolinx, the project owner and provincial government agency, put out this video advertising the expected 36 minutes end to end travel times for the Finch West LRT project. 52 subtracted by 36 is 16. So good, Finch West LRT will save time as the fan claimed. Not so fast… (pun so intended)

At the time, when I read that proclamation I thought something was fishy as this was not the data result I remember working with. The existing Route 36 bus travel time (+50 minutes) is way too high. So you have an unscientific discrete personal anecdote on travel time, while when I was consulting for the TTC on this route I had teams of subcontractors riding up and down Route 36 taking in field data, recording delay observations and holding GPS plotter data. Which I then validated with GPS CAD-AVL bus tracking data provided to me by the TTC. So I dove deeper on this gut feeling.

The TTC uses their past experience on bus operations for each route to inform adjustments to the route’s scheduled average speed and run time every couple of months. Broadly they can not assume too fast of an average speed divorced from the route’s actual conditions or each bus run will systemically end up classed as late. This would look bad and is not advised. As with everything in life, we should promise what we can reasonably offer (Metrolinx take note). TTC service planners can assume a conservative slightly slower average speed to account for unforeseen congestion and provide drivers with some space in the schedule to breathe and recover, called schedule padding. However it is also important to balance between aggressively pushing for speedy operations and schedule padding for certainty in schedules. A highly padded schedule will lead to buses driving slower if it just so happens the delays didn’t materialize that hour. Good supporting operations and service design should properly balance between the two. All this information is sent to route managers to direct drivers and publicly published by the TTC in service plans, large documents of planned service for every route in the TTC network.

To begin the deeper dive, I looked at TTC’s historic service plans for January 7, 2018 to February 17, 2018. Below, I can see the planned service for TTC route 36A, a branch that runs between Finch West and Humber College is scheduled during AM rush hour assuming a bus average speed of 18.1kph. This branch runs on the exact same segment that is being replaced by the Finch West LRT, with a some deviations at the end of the route before terminating at Humber College. These deviations add more than 1 km to its length compared to the then under construction Finch West LRT. This is why I will be focusing on comparing average speeds and not too fixated on end to end travel times. TTC route 36B ran the rest of the way between Finch West and Finch Subway Stations.

The average speed of 18.1kph is in line with average speed assumptions before Route 36 was split at Finch West Subway Station into A and B branches in 2018. Before Route 36A buses ran the full gamut from Humber College to Finch Subway Station. As shown by a similar 18.3kph assumed AM rush hour speed in an older service plan (November 26, 2017 to December 16, 2017) where Route 36A buses ran all the way to Finch Station via the at the time newly opened Finch West Station. So the travel time for Route 36 assumed by the planners at the TTC has been consistent.

For a little bit of background, Route 36 was operationally split at Finch West Station in 2018, most likely to prepare for the construction of the Finch West LRT and to take advantage of the newly opened Finch West subway station from the Toronto–York Spadina subway extension (TYSSE). By operationally splitting the bus route in the middle, delays can be controlled and prevented from cascading down the long route. Similar things were done to other routes like Route 25 Don Mills for Eglinton Crosstown LRT work. It also helps that when the Finch West LRT opens the Route 36A section can just disappear and the Route 36B segment between Finch West and Finch Stations can just remain as is.

Based on the information above, the 18.1kph planned average speed of Route 36A and using MATH; the travel time of the 36A Finch West local bus should take 38 minutes to travel 11.5km (I used cycling and waypoints to get Google Maps to tell me the true distance of the bus route see below) from Finch West Station to Humber College. Which is almost the same as the advertised Finch West LRT travel time of 36 minute covering the same distance.

The above has huge implications. The Finch West LRT is shorter than Route 36A by taking a shortcut down Hwy 27. Using the Metrolinx's advertised Finch West LRT end to end travel time of 36 minutes and the 10.5km total length, it works out that the average speed of the Finch West LRT is 17.5kph. This is all using Metrolinx’s provided numbers. At the time, when confronted with this result, I semi-refused to believe it. This means that the Finch West LRT is actually on average slower than the 36 Finch West local bus it will be replacing. (Stay tuned to the end it gets worse)

For a quick sanity check, local bus speeds should be similar to cycling. If you have cycled in outer suburban Toronto you would know your average speed is slightly less than a bus. When cycling with the same Toronto arterial bus route (This is not advised to do this BTW, cycle quieter streets please), you tend to leapfrog with the exact same bus. Passing the bus as it boards and alights passengers but that same bus passes you when going between stops. This is confirmed by the map above as cycling the same route and distance as the bus is reported to be 40 minutes end to end, close to the 38 minutes I calculated for Route 36A.

Regardless how did the Finch West Route 36A bus become reportedly so slow (+50 minutes end to end) after I went back to Hong Kong? I dug deeper and found since 2019 the TTC planners (to their credit) added running time and padded the 36 Finch West local bus schedules BECAUSE of construction congestion on Finch for the Finch West LRT, which has torn up the road along the entire 36A bus route. The assumed average speed of the bus went from 18.1kph to 14.8kph after construction started.

Notice back in the Metrolinx video (I saved below) they DO NOT note any travel time comparisons nor savings with the 36 Finch West local bus. Why? Simply because there are none to be found when using a proper apples to apples comparison with the 36 Finch West local bus without interference from construction of the LRT line. So to properly compare time savings you have to use a bus travel time before the LRT started construction like I did. There I found that end to end average speeds for the TTC local bus are scheduled to be faster than the advertised Finch West LRT scheduled speeds. Since scheduled speeds are generally a reflection of reality it is suffice for me to say back in early 2024 that the Finch West LRT would be similar in speed to the bus it replaced.

With all the above in mind, I concluded in early 2024 that the Finch West LRT will be about the same speed as the 36 Finch West local bus it replaced based on Route 36 travel times before the LRT started construction. A bombshell allegation at the time, much to the displeasure of some Transit City supporters. One local advocate wrote a rebuttal using data they collected but at the time I remained unconvinced as the graphs basically show that the Finch West LRT will only be slightly faster than the route 36 bus it will replace. Which at this point is splitting hairs when billions are being spent on this tramway project. I also might have been in China riding actual rapid transit at the time and could not care less. In addition, the rebuttal argues that streetcars in dedicated lanes will make the service more reliable. Again if you have ever taken a TTC streetcar service, even one of those “streetcar rapid transit” services that have their own lanes, is laughable.

The proof is in the pudding and hindsight is 20/20 (see later on in this post). I stand by my assessment: Toronto, as reported at the time, wasted well over 1 billion dollars for a poorly designed median suburban tramway on Finch West Avenue with minimal to no travel time improvement. But JRUrbane that is just the false promise of speed, what about LRT’s being more cost effective to build than Subways? As you will see in the next paragraph, building the Finch West LRT did not simply cost 1 billion dollars, it was closer to 3 billion as the project prodded along. I’m not sure the case for Transit City was there too.

The Cost Disease Python…

Apart from the disappointing anticipated speeds both LRT projects have been infected with huge North American cost inflation disease. The Crosstown LRT being half underground should be, as Transit City advocates argued, quick and cheap the build. It was not. Costs gradually increased and timelines severely lengthened as the two LRT projects dragged along. The Eglinton Crosstown LRT now costs 500 million INT$/km and the Finch West tramway is projected to cost 220 million INT$/km to build. The latter is a similar per km cost to building the underground Sheppard Subway back in the early 2000s (accounting for inflation) or the cost of a typical underground metro line in China (accounting somewhat for wages and inflation).

Toronto’s cost inflation disease, like a python wrapping around its meal, overwhelms and snuffs out any benefits a project can bring. It is cold comfort knowing Toronto blew billions just to replace a bus route with, at best, a slightly faster tramway. This is assuming the tramway even delivered what was promised if everything worked out (but it didn't). In addition, the overall dysfunction of North American transit construction timelines is another spit in the face for Transit City’s supporters. Apart from costing tens of billions of dollars, I would hardly call taking over 15 years (since 2011) to build the Eglinton Crosstown LRT as expedient construction. The Greater Hangzhou Metro grew from no metro in 2012 to almost 625km of it in 2025:

Too Little, Too Late

Regardless of the evidence put forward by the remnants of Subways, Subways, Subways and the newly created Third Front of light metro supporters (plus yours truly), Transit City supporters refused to believe the warning signs. Insisting these “tramways” are rapid transit and doing cute retcons like calling LRT “Light Rapid Transit”. Any industry professional worth their salt knows a tramway running down the middle of a street is not rapid transit. Light Rapid Transit as a concept exists in East Asia for light metro lines (like the Busan-Gimhae LRT I rode in 2022 below) which actually provide the service quality of a metro line. However, such technology is antithetical to what Transit City supporters want in a transit project aesthetically.

The only Transit City project to even be close to the concept of rapid transit was the original +30km Eglinton Crosstown LRT plan from Pearson Airport to Kennedy Station. Phase 1 (solid line) was the initial segment approved and started construction back in 2011. A western Phase 2 extension (dotted line) would start construction at a much later date and was originally planned to be median streetrunning.

In summary, the original Transit City plan for the ECLRT consists of a 10km long central underground tramway section with each end of the line envisioned as a 10km tramway running down the median of Eglinton Avenue. The ECLRT as described above was the most rapid transit-esque proposal in the entire Transit City plan, all other “Light Rapid Transit” lines of Transit City had minimal grade separation and were essentially tramways. Which shows how tenuous the Transit City program was with the concept of rapid transit, treating Toronto as some midsize German city.

This obfuscation is not new from Toronto’s Transit City supporting pro-streetcar lobby. Starting in 2009 20052 TTC’s system wide transit map “Rideguide” started labelling downtown tramways with median dedicated lanes (509/510 and 512) using an oxymoronic title of “streetcar rapid transit” up until 2014 when the system wide transit maps were redesigned. Despite these aforementioned tram corridors being some of the slowest tramway operations in the world with no Transit Signal Priority (TSP) and in dire need of operational fixes.

This concept of “streetcar rapid transit” introduced in 2005 was arguably a marketing ploy by pro-streetcar factions inside the TTC and Toronto City Council to hype up an ongoing project to put the 512 St. Clair streetcar on its own median lanes, which started construction in… 2005. The St. Clair streetcar rebuild is its own saga of political infighting, lawsuits and poor transit design. In summary, The City of Toronto spent 106 million CAD (136 million CAD in 2025 money) to redo some poles, put in curbs, renew stop shelters and make 7 kilometers of the streetscape pretty with lanes dedicated for a tramway down the middle. The result was minimal travel time savings on the 512 St. Clair streetcar when the rebuild project was completed in 2010. These travel time savings all but evaporated by the 2020s due to poor line operating practices. This will sound familiar later on.

Some unapologetically Toronto-centric reporters such as Stephen Wickens wrote an article that lampooned the sudden contrived TTC re-designation of these median lane tramway routes as rapid transit. Stephen also criticized the specific concept of “surface rapid transit” (median lane trams with many intersections and tight stop spacings) pushed by Toronto’s streetcar/LRT supporters. He received “threatening and unpleasant phone calls in the weeks that followed the publishing of his article” from streetcar/LRT fans, TTC and City Officials blinded by their zealotry for European style tramways and that someone had the gall to speak out against the worship of their Golden Calf.

Toronto did not learn its lesson, this deception came back with a vengeance in the early 2020s. The under construction Eglinton Crosstown and Finch West LRTs got subway numbering (TTC Rapid Transit Lines 5 and 6 respectively) and branding, thus elevating (not literally) them to rapid transit status comparable to an actual subway line. They even show up on the subway map like Eglinton Crosstown and Finch West LRTs are somehow real rapid transit now. Well, regardless of how you are going to brand these LRT lines, if you are going to blow more money and time building a tramway that France, Italy, Spain, S Korea or China builds an actual underground rapid transit line it better be good. However when “Toronto Rapid Transit Line 6” opened…

TRAMinator 2: Judgment Day

After all the trials and tribulations Line 6 opened on December 7, 2025. The opening day of Line 6 Finch West was a depressing disaster. Billions of dollars and years of work only to build a tramway that is slower, less frequent and less reliable than the TTC route 36 bus it replaced, a worse case scenario for its supporters. A typical end to end trip on the 10km long tramway ranges from 45 mins to over 1 hour, slower than the jogging speed of an average person. Since opening, every day had a significant service disruption. Supporters of Transit City for years aggressively advocating and ignoring warning signs 而家冚辦爛坐嗮喺度O曬嘴噉樣. The top brass at TTC and City of Toronto seemed to be caught flatfooted and unaware about how equitable transit oriented traffic operations can be achieved. To the point that a local advocate with technical experience on TSP design and implementation had to write a post explaining the basic first principles they got wrong.

Upon opening, Line 6 was plagued with archaic operational regulations mirroring or even exceeding the overbearing practices found with the TTC downtown streetcar network. TSP was conditional, only activating if the tram was late but trams are never late when it turns out their schedules are padded to be extremely slow. The median lane tramway had an alignment design speed of 60kph which is extremely low for a new rapid transit line. In a strange design choice the tramway installed an advanced train protection system that aggressively enforces low speeds in stations and at intersections in the name of “safety” (which I was told never stops… except a Finch West tram always seems to stop). This creates a tram service that crawls as cars and even TTC’s own buses blast past it.

It is rich that the Finch West Tramway and another project I recently covered, Dongguan Metro Line 1 both opened around the same time. Dongguan Metro Line 1 opened on 28 November 2025. I rode this new line on opening day and published my coverage of it just as the disaster of Line 6 in Toronto unfolded:

Both Dongguan Metro Line 1 and the Finch West Tramway even started construction at the same time (2019). Except Dongguan Metro Line 1 is a fully automated 120kph 6 car underground city center metro line while the Finch West Tramway struggles to keep up with the speed of a local bus. Also Dongguan Metro Line 1 was built for much less money than the Finch W Tramway.

The tons of bad press prompted the municipal government, now headed by Olivia Chow, to look into improving TSP to be more proactive and make other operational adjustments to “Toronto Rapid Transit Line 6” and even the existing neglected city center streetcars. We can remain optimistic that there can potentially be some improvements. However it is important to note, as Marco Chitti of Italian (urban) Letters puts it, TSP is not some magical panacea:

You can never get subway average speeds from a tramway running at-grade down a street. So I suspect the adjustments to TSP will improve speeds on the Finch West Tramway but it will be like squeezing blood from a stone. Even if the touted reforms are successful; Line 6 gets a bit faster and operates closer to initially promised speeds (slated for next year). Blowing over 200 million INT$/km on surface road median tramway that is only slightly faster than the local bus it replaced is a poor value proposition and does not vindicate the cost argument Transit City supporters bandy about. The fact these lines also took so long to build is another slap in the face to Toronto voters and commuters.

All this controversy around Line 6 is in the background of the looming opening of the much delayed Line 5 Eglinton Crosstown LRT. “TTC Rapid Transit Line 5” is rumored to be only slightly faster than Line 6 due to half of it being built and behaving like an actual rapid transit subway. Line 5 also costs more per km to build than some of the world’s most cutting edge, high capacity and cracked grade separated urban railways built under the city centers of France (RER E), S Korea (GTX) or China (Guangzhou Metro Line 11, 13, 18 and 22 or Shanghai Metro Line 14).

Transit City in Name Only

Taras’ called the chaotic political atmosphere surrounding the lack of clear acceptance on the Transit City plan the “Toronto Tragedy”. I think looking more broadly the bigger tragedy is that a group of supporters for Transit City, so fixated on a certain way transit needs to be expanded in Toronto, highjacked and gaslit everyone into spending almost 20 billion CAD building two mediocre LRT lines totaling 30km (with 10km underground).

Finch West Tramway and the upcoming travel time and cost effectiveness disaster known as the Eglinton Crosstown LRT will undoubtedly cause huge reputational damage to the concept of light rail in Toronto, or at least discredit the way light rail is designed as envisioned by Transit City’s supporters. I highly doubt that residents in Scarborough are now chomping at the bit to build the Sheppard East and Eglinton East as LRTs. We must make an informed decision as to whether our goal is trying to blow billions of dollars building a transportation system cosplaying as a European town or one which will best serve people. If we are building a transportation system to cosplay as a European town, the Finch West and Crosstown LRTs would be a good place to start. But if we are building a transportation system to best serve people, the Finch West and Crosstown LRTs as they are designed today is a good place to stop.

Why? because as I noted above the "Toronto Model has successfully produced a public transit modeshare in Greater Toronto of just over 20%. However at the same time private car modeshare is still over 70%, a clear majority. Hong Kong’s public transport modeshare is well over 70%, at the same time private car modeshare in Hong Kong is under 15%, yet no one is calling Hong Kong the car city with those ratios. Toronto’s transit discourse and advocacy needs to drastically improve if it wants a large scale modeshift to transit and be any sort of actual “first tier”, “world class”, “Transit City” like Paris, Seoul, Tokyo, Zurich or Hong Kong.

This post is a bit of a change in geography. Like, share and subscribe for more transport news and analysis from around the world.

Or drop me a Ko-fi if you like what you read!

Setting the Stage

Apart from being overshadowed by Hong Kong, Shenzhen and Guangzhou, the modesty of the Dongguan Metro is partly due to the very dispersed and industrial nature of Dongguan’s urban form and governance. Dongguan’s urban form is defined by hundreds of villages and towns dispersed over a large area with no obvious city center.

Dongguan Prefectural City’s local governance is also uniquely set up to reflect and reinforce this urban form. Dongguan is an amalgamation of 4 subdistricts and 28 towns all with quite equal (dispersed) political and economic power. In theory, the 4 subdistricts (colored yellow below in the first map) is the center of Dongguan but in practice their pull is limited. The towns are also loosely grouped into “subareas” (片区). Truly the Holy Roman Empire of Chinese municipal governance. This is in contrast with most Chinese prefectural cities having clear distinctions on satellite cities, urban/suburban districts and rural counties, which provide clarity on what the center of the urban area is and by extension where a metro line should be built. For example Shenyang (third map below) clearly labels non-core areas as cities and counties as well as districts (names with no county or city label) are informally split into urban and suburban.

The above environment leads to two considerations: Firstly, it makes planning/ justifying a metro system for such a city type not as apparent. Secondly, planning and prioritization of where a metro line should go is not as clear. Despite these hurtles, Dongguan was one of the first cities in China with this urban form to try to plan and build a metro.

The First Line… Line 2

Dongguan Metro opened its first metro line as Line 2 Phase 1 & 2 in 2016. It essentially operates as a 15 station rapid shuttle between the largest old town (the 4 subdistricts), new town and Dongguan’s two major railway stations. Dongguan Metro knew their lines have to be more regional in nature to match the dispersed, decentralized but dense urban form. Line 2 is almost 40km long but only has 15 stations. The trains are 6 car Type B trains but with less doors (3 per side per car instead of the standard 4 doors used across China) and more seats to fulfill its more regional role. In a rare move during the early 2010s Chinese Metro scene, Line 2 was designed to operate up to 120kph:

Despite Line 2’s design pioneering of many features, Dongguan is famous in Chinese Metrofan circles for being one of the slowest expanding networks in China. This is not surprising, with China’s traditional metro model (despite Dongguan already making pioneering adaptations) being a poor fit to Dongguan’s aforementioned industry heavy, highly dispersed urban form. It also does not help that Dongguan is sandwiched and overshadowed by the massive juggernauts of Guangzhou, Shenzhen and Hong Kong. After the opening of Line 2 Phase 1 & 2… Dongguan did not open anything for almost 10 years, much to netizens’ chagrin:

After opening Line 2 in 2016, Dongguan Metro did start detailed design and planning for Line 1 a few months later. Again, this is strange in booming China where one city would have many parallel metro projects in various stages of planning and construction done together in tandem at any given time. During this time, ridership on Line 2 was mediocre, as lone Line 2’s limited reach and wide stop spacings depressed ridership, just serving a daily average of over 130,000 trips per day in 2024.

Dongguan is at the forefront for understanding how to design a metro to serve the “highly dispersed but still dense Chinese industrial city” archetype, being one of the first cities of this type to build and open a metro. China’s existing legacy metro model (~2km stop spacing, 80-100kph, very high capacity) is highly effective in providing a baseline service for a large city with multiple clearly defined centers. However it is less clear how to provide a “basic metro service” suitable to serve a city like Dongguan.

So in the past few years, Dongguan has been struggling with how to design a metro to serve its unusual urban form. The results for Dongguan’s metro development are unsurprising: low ridership of Line 2 as well as the slow planning of later follow up lines. While the rest of China debates should these types of cities even build a metro. Similar cities in China (mainly in the south) like Quanzhou, Shantou-Chaoshan, Linyi, Huizhou, Zhongshan-Jiangmen are grappling with this problem in planning/justifying their networks. I understand Dongguan Metro’s vision: attempting a regional metro network with wide 2-4km stop spacings that link up the hundreds of villages and industrial (and now R&D parks) with metro services. However this vision only works with a true network, not a single line.

The Second Line… Line 1

Dongguan Metro Line 1 Phase 1 started construction in 2019 but interestingly its final design was only approved by the National Development and Reform Commission as part of a “Adjusted 2nd Phase Construction Plan for Dongguan’s Urban Rail Transit (2022-2030)” (东莞市城市轨道交通第二期建设规划调整（2022-2030年）) in March 2022, all while the project was well under construction:

As noted above, the final adjusted Line 1 Phase 1 metro project was only approved in March 2022 at an estimated to cost 165 million Purchasing Power Parity (PPP) adjusted International Dollars (INT$) per km or about 9.5 billion INT$ in total to construct. The reason for the delayed approval? A last minute adjustment of the alignments for Line 1 and 3 Phase 1. For the Line 1 Phase 1 project, 17.2km of elevated alignment in the south was switched from elevated to underground running. Official reports stated the total cost increased from 34.64 billion RMB to 37.98 billion RMB, a difference of 3.34 billion RMB. Of the total cost, “direct engineering costs” (I’m assuming it is hard construction costs) increased from 16.75 billion RMB to 23.92 billion RMB a change of 7.17 billion RMB. Why such a difference between the two deltas of total cost and hard costs? How can total cost only slightly increase but hard costs significantly increase? My suspicion is that running elevated incurs significant land acquisition costs which are usually included in reported total costs of a metro project in China. Switching to underground avoids land acquisition costs but underground running by definition increases hard construction costs.

The final approved Line 1 Phase 1 is a 57.5km long line with 25 stations. The line is now mostly underground but has a short 8km elevated section in the north (blue solid line). The dotted lines on both ends are short extensions that will connect Dongguan Metro Line 1 to Guangzhou/Foshan and Shenzhen/Hong Kong systems, creating a long chain of metro systems from Hong Kong to Guangzhou.

As I alluded to in the start of this post, Line 1 Phase 1 is the first step in linking the metro systems of the Pearl River Delta together. However, both ends of Line 1 today still stops short of the Guangzhou and Shenzhen borders, meaning the Pearl River Delta’s metro networks are still not unified. So as of 2025, it is not possible for me to go from South Horizons, Hong Kong to Huadu, Guangzhou a straight line distance of about 170km, strictly by metro.

Going to see the Metro

I knew I could not make it to the Dongguan Metro Line 1’s early morning opening ceremony from Hong Kong. Guangzhou typically opens their lines at noon but Dongguan Metro Line 1 had a two day limited public trial on November 26 and 27 just before the official opening. By then the last minute kinks should be ironed out and Dongguan Metro can start service (and have the opening ceremony) in the morning. So I planned to ride Dongguan Metro Line 1 on opening day but in the afternoon. So after doing some morning errands and paying some respects, I make my way to Lo Wu to cross over to Shenzhen:

Because I ended up in Lo Wu I needed to take the Guangzhou-Shenzhen Railway regional services to Dongguan instead using high speed rail.

The pair of tracks serving the regional rail services (tracks 3 and 4) of the quadruple track Guangzhou-Shenzhen Railway just got upgraded to 200kph operations on 5 November 2025, an increase from the traditional 180kph speed limit. Now non-stop “direct express” C8xxx series trains travelling from Shenzhen city center to Guangzhou city center takes less than 1 hour and depart every 30 minutes.

This upgrade is part of a suite of improvements to regional services on the Guangshen corridor. Which underwent a signal upgrade in 2023 that allowed daily regional services to increase from 118 to 146 trips per day. The improvement of speed in November 2025 was also accompanied by another increase in regional services to 178 trips per day. Today, one Guangshen regional train departs every 13 minutes on average throughout the day. During peak periods a Guangshen regional train leaves every 5 minutes.

I wait to enter my express train which, unlike the non-stop direct express, stops at a number of semi-major stops (like Dongguan or Zhangmutou). I’m mixed in with the tail end of the crowd taking the 13:00 direct express train that leaves 5 mins earlier on another platform:

Some people had to run to catch the 13:00 direct express train:

I head down to my platform with some Central Asians and Russians:

While another express train from Guangzhou pulls in as the 13:00 direct express to Guangzhou leaves. I quickly then hop on to my express train behind me:

My seat was the one strange “booth” face to face seats with a table found on every car of a CRH1A. The table had a copy of today’s “People’s Railroad”. It is wild that, sure the People’s Railroad is one “broadsheet” but, there are enough Chinese mainline railway news every day to fill a broadsheet. Look at this HSR tunnel, look at this bridge, look at this massive station that is getting ready to open etc.

I arrive at Dongguan Station and a CRH6A direct express train bound for Shenzhen passes mine:

I make my way from the China Railway mainline platforms to the Dongguan Metro platforms. On the way, I do indeed confirm that Line 1 (the blue line) is on the map and open today. I have to take the older Line 2 to Citizen Center Station to get to this newly opened Line 1.

New Metro Line Who This?

I arrive at Citizen Center Station, which recently was renamed (it was formerly Hongfu Road Station) and see the new stairwells heading down from Line 2’s platforms to the few hours young Line 1. Upon entering Line 1, I was hit with a strong “new construction” smell.

Citizen Center Station was built a decade ago with reservations for a Line 1 interchange. Before Line 1 opened, knockout panels for the stairwells directly connecting Line 2 with Line 1 can be seen on Line 1’s platforms.

Citizen Center Station is also planned to host an interchange with Guangzhou Metro Line 28, Guangzhou’s East West high speed express metro line that connects Foshan Guangzhou and Dongguan cities together. These three cities have a combined population of 40 million people but have yet to be fully conurbated. This will change with the arrival of Line 28. However, due to the more recent nature of the Line 28 project, I believe there have not been any reservations at Citizen Center Station to directly accommodate Line 28.

At Citizen Center Station Line 1 platforms, no one expects the Spanish Solution:

Line 1’s platforms, unusually, have a set of platform screen door control panels in the middle of the platform. I guess it is because of the Spanish solution setup. However, the arrow decals on the ground do not reflect the true operation of the Spanish solution (third image below). A true Spanish solution requires the separate platforms to be designated for boarding and alighting which the station’s passenger circulation plan to be designed accordingly. Citizen Center Station’s Line 1 platforms are designed for separate designated boarding and alighting roles as shown in the video above. Passengers board from the center platforms and the one way escalators are all set to convey people away from the side platforms (last image below) designated for aligning.

All systems go:

The trains have all the expected accoutrement: fully automated but with 2-4 people standing in front practicing shisa kanko and LCD screens with very Shenzhen Metro UI. Interestingly, while Line 1 uses 6 car Type B trains capable of running up to 120kph like Line 2, the trains use the standard 4 doors per side configuration in contrast with Line 2. Perhaps Dongguan envisions this line to be more urban(e) in nature.

A take Line 1 to its northern terminus near Guangzhou, flying through suburban Dongguan:

One Line 1’s depots, set up for topside development naturally:

Paralleling the Guangzhou–Huizhou Regional Railway:

I arrived at the terminus of Line 1, the elevated Dongguan West Railway station:

The terminus is a short distance away from the Guangzhou border. The tail tracks are just begging to be extended to meet Guangzhou Metro Line 5 which was recently extended to the Guangzhou Dongguan border.

Regional Railway Connections

The elevated Line 1 terminus station is connected to the nearby Dongguan West railway station via a footbridge, there are even express escalators that allow passengers to transfer between metro and intercity railway levels, skipping the intermediate levels of the tall metro station.

The area is a maze of railway viaducts. You can see why the Dongguan Metro station needed to be so tall. Line 1 like the parallel Guangzhou–Huizhou regional railway viaduct needs to clear the tall regional railway viaduct of another Guangzhou-Shenzhen regional line (this is a different regional line not the Guangzhou-Shenzhen Railway regional services I took to get to Dongguan) running perpendicular to Dongguan Metro Line 1:

The adjacent Dongguan West regional railway station is a two level interchange station with passing tracks connecting the two Guangdong-Hong Kong-Macao Greater Bay Area Regional Railway lines: Guangzhou-Shenzhen and Guangzhou–Huizhou Regional Lines. The latter is part of the Guangzhou regional system operated by Guangdong Intercity Railway Operation Corp, a subsidiary Guangzhou Metro Group. Due to the two lines having different operators, the Guangzhou–Huizhou Regional Railway does not even show the Guangzhou-Shenzhen Regional Line on their maps.

Below you can see the contrast in regional service between the Guangzhou–Huizhou Regional Line (below left) which is operated by Guangdong Intercity Railway Operation Corp with the lightly used Guangzhou-Shenzhen regional line (below right) which is operated by China Railways. Guangzhou–Huizhou Regional Line is more frequent and has multiple branded stopping patterns, some trains are designated using the rarely used S series numbering. However, we shall see how service changes on the Guangzhou-Shenzhen regional line when the line is extended south from Shenzhen Bao’an Airport to Qianhai then to the Hong Kong border and extended north via the Xinbaiguang Regional Railway to Guangzhou Baiyun Airport. The latter section is expected to be operated by Guangdong Intercity Railway Operation Corp.

I get back on to Line 1. Inspired by Dongguan West Railway station, I was curious about how Dongguan Metro integrates with other regional railway lines at other stations. Another feature of Dongguan’s public transport landscape is its reliance on regional railways. Dongguan is between Shenzhen and Guangzhou, so if you look at Dongguan’s rail transit planning over the years, Dongguan opportunistically assumes regional railways radiating from its two neighboring megacities will also serve Dongguan.

Regional railways very clearly make up a big component of Dongguan’s plans, as shown by the numerous lines, such as the aforementioned Guangzhou-Shenzhen and Guangzhou–Huizhou Regional Lines, crisscrossing Dongguan’s rail transit planning map as shown below. The map also shows the existing Metro Line 2 Phase 1 & 2 (purple), newly opened Line 1 Phase 1 (orange) as well as the under construction Line 2 Phase 3 (cyan). Metro Line 3 Phase 1 (dark blue) is approved and awaiting construction while Line 1 Phase 2 & 3 (dashed lines) are officially proposed as part of Phase 3 Dongguan Metro expansion that will connect to the Guangzhou and Shenzhen Metros.

Making my way downtown… as in, back to Dongguan’s new town center around Citizen Center station.

I arrive at Xiping Station on Line 2 is which is another interchange station between the Dongguan Metro and a regional railway line introduced earlier, the Guangzhou–Huizhou Regional Line. This regional railway runs underground through central Dongguan. Currently the situation of fare integration between Metro and regional railways in Dongguan is… very Japanese. The two operators barely acknowledge each other fare wise and the stations don’t have the exact same names. To transfer between Metro and regional rail at this station you head up to the concourse and exit the metro faregates, walk to the regional railway part of the concourse, enter using a different set of Guangdong Intercity Operation Corp faregates then head down. The regional railway section of the station still has decorations celebrating its extension to Guangzhou South Railway Station and simultaneous whole line take over by Guangdong Intercity Operation Corp last year. The regional railway station also is advertising Guangzhou Metro’s TOD opportunities.

There was clearly motivation for better metro regional railway integration. I observe the Dongguan Metro Line 2 platforms having provisions for a direct platform to platform interchange between the underground Dongguan Metro and regional railway stations. Conversely, pictures on Wikipedia by Tim Wu also showed a walled off connection to the Metro on the regional railway platforms:

Looking around the station I see there is a “Rail Transit Building” at one of the exits.

Intrigued, I decided to go see this “Rail Transit Building”. I follow the sign to the appropriate exit which leads to a confusing, still half under construction underground/sunken mall system.

I go outside and admire this “Rail Transit Building” which is on top of the confusing underground mall. The internet told me this tower is newly opened, 250m tall with 55 floors.

At the base of the “Rail Transit Building” a sign pointing you to the station complex, labeled the regional rail service properly instead of calling them “intercity”.

I explore and eat in the surrounding areas then proceed to reenter the metro.

I head back to the (still) few hours old Line 1 to see yet another regional railway connection at Dongchengnan Railway Station. Here, again, is another interchange between the Dongguan Metro and the Guangzhou–Huizhou Regional Line. It is another “exit faregate in concourse walk to other concourse then enter faregate” type of affair. In contrast to what I described with Xiping Station, which was planned or built a decade earlier, I do not think there are provisions for direct platform to platform transfer in this station. The station map does not even show the regional railway portion of the station. Is the optimistic Halcyon Days for regional rail/ metro integration that built Xiping Station in Dongguan dead?

The Long March

After finishing the tour of Dongguan’s metro and regional railway integration, I think about how to go back home to Hong Kong. Being that I’m on the southern side of the still less than day old Line 1 and the southern terminus of Line 1 almost reaches Shenzhen, I decide to take the rest of Line 1 to the end. Then I will figure out how to get into Shenzhen and take the Shenzhen Metro/HK MTR back home.

Taking the Line 1 south closer to Shenzhen, I notice a huge amount of stations with cross platform interchange provisions. Many station platforms have large blank walls but are not side platform stations. Station maps do show unused tracks in the station structure. I observed that provisions were made for a paired cross platform interchange, like those found in Mong Kok and Prince Edward Stations of Hong Kong MTR fame, at Songshan Lake (top row of pictures below) and Guangdong Medical University (bottom row of pictures below) Stations with the approved Dongguan Metro Line 3:

Going further south on Line 1, provisions are provided for a cross platform interchange with the proposed Line 1 branch service to Zhangmutou Town at Fengshan Station:

Approaching the Shenzhen border, I reach Huangjiang Station which is a half built station hosting provisions for a same direction cross platform interchange with the proposed Line 4. You can see from the station plan, layout and the strange bare artless walls hugging closely with standalone elevators and stairwells on the northbound platform. The station clearly is meant to be expanded to add another Line 4 platform on the Exit C and D side, completing the set of platforms for each line.

Line 4 Phase 1 is proposed to terminate at this station and head east to the Tangxia, Fenggang and Qingxi Towns, southern Dongguan towns that are northern extensions of Shenzhen’s suburbs. I suppose planners believe Line 4 has a hypothetical Phase 2 that will extend Line 4 the other direction west of Huangjiang Station and wanted to future proof for that without building the full cross platform interchange. However, this is entirely conjecture on my part.

I arrive to Line 1’s southern terminus, Meitang Station. It has a technology theme, I guess because it is as close to the Shenzhen border you can get on the Dongguan Metro today.

The tail tracks at Meitang Station are begging to be extended south into Shenzhen.

Currently there is a 4km gap between the termini of Dongguan Metro Line 1 and Shenzhen Metro Line 6 Branch. Both lines are planned to thru operate with a connection line (with one intermediate station) proposed to be constructed as part of Dongguan Metro Line 1 Phase 3 expansion project (green dotted line below).

I suspect why the connection line is deferred to a different phase is that Shenzhen Metro and Dongguan Metro still are not sure how they want to through operate together. From my perspective they could do a number of service patterns which affect the fleet planning of Dongguan Metro Line 1 and Shenzhen Metro Line 6B:

1. A forced transfer at the border of Shenzhen and Dongguan meaning an interchange between Line 1 and 6B at SUAT (or Shenzhen Institute of Advanced Technology Station) or 深理工站

2. A forced transfer at the proposed intermediate station in Dongguan Metro Line 1 Phase 3 expansion meaning an interchange between Line 1 and 6B at Huangjiang South Station or 黃江南

3. Shenzhen Metro Line 6B to take over of Dongguan Metro Line 1 up to Fengshan Station all Dongguan Metro Line 1 trains operate into the future Line 1 branch to Zhangmutou.

4. Full through operations: various Line 1 and Line 6B services run into partial or full lengths of each others’ lines.

From what I can gather with the Dongguan Metro Line 1 Phase 3 project. The intermediate station (Huangjiang South) was ultimately selected to be the temporary physical interchange and terminus for Line 1 and Line 6B services until the compatibilities between the two lines can be sorted out. Being that Huangjiang South was not planned or built during the shifting histories and planning of Lines 1 and 6B, planners have a tabula rasa to design the station with extra platforms to serve as the temporary interchange. This is not the first time such things happened in China. Suzhou Line 3 and S1 had similar interoperability uncertainties and were adjusted to through operate months after Line S1 opened. So the timelines for resolving through operation issues here looks to be short.

So how do I get to Shenzhen? I looked at the station map and saw that I can follow Gongming-Changping Road (公常路) south to Shenzhen. I knew I was close to Shenzhen and I also knew the nearest Shenzhen Metro Station ends at the border between Dongguan and Shenzhen. So decided to walk to there… only to see a branded shuttle between the Dongguan Metro Line 1 and Shenzhen Metro Line 6 that I could have taken. It even came around every 10 minutes. The metro exit also has a Shenzhen Metro -esque underground alignment map for construction locates.

But good thing I did walk the gap between the Dongguan Metro Line 1 and Shenzhen Metro Line 6 had many interesting things to look at like urban villages:

Shops filled with laser cutters, 3D printers, CNC machines, and other more traditional machine shop tools.

And vertical factories

I keep walking and suddenly see the characters Shenzhen coming out the ground and the sidewalks were suddenly nicer. I knew I entered Shenzhen.

Walked a bit further and see Dongguan Metro Line 1-Shenzhen Metro Line 6 shuttle bus dropping off passengers transferring between the two lines.

Reviewing the Shenzhen Metro underground alignment map for construction locates just outside of the metro station, you can see the tail tracks of Shenzhen Metro Line 6B run right up to the Shenzhen Dongguan border. The box (most likely TBM access shaft) at the end of the tail tracks on the map (top right) is right beside the Shenzhen character sculpture coming out the ground I passed by as I entered into Shenzhen. Shenzhen really wants the line to enter Dongguan.

I then enter SUAT (Shenzhen Institute of Advanced Technology) station and its namesake university. In other words: I made it.

I enter the station and take the Shenzhen Metro home but not before I detoured to Futian Metro Station Line 11 and check up on Shenzhen’s metro freight experiment. (That might be another Substack)

So thanks for reading my adventures in Dongguan. So to continue seeing the evolution of what will be the world’s largest urban railway system: like, share and subscribe for more long form adventures in East Asia.

Or drop me a Ko-fi if you like what you read!

This is a part 3 of a multi-part series where I explore the Guangzhou Metro lines that newly opened in June 2025. You can read the second part which provides more context below:

The Line 10 South Section Experience

After exploring old Xilang Station’s Line 1 platforms, I head down to Xilang Station’s new few hours young Line 10 station.

As I head down there are props celebrating the opening of Line 10 and 12 in the transfer concourse of Line 10 Xilang Station, similar to the ones set up when I was visiting the opening of Line 5 Phase 2 and Line 7 Phase 2 back in December 2023. At that time I observed interest in those openings from the general public was extremely limited. After a few hours of opening, I was only the 65th person to be interested enough to press the button. It seems the other passengers are not there to see the extensions (Line 5 Phase 2 and Line 7 Phase 2 East) but are passing through on their new daily routine of using the new metro extension.

This time, it has been several hours since the opening of Lines 10 and 12 and I am exactly the 4,300th person interested in the event at Xilang Station.1 It seems the enthusiasm is greater for extensions this time, perhaps due to the more centralized and consequential nature of these partial openings.

I head on down to the platform level and board a Line 10 train:

My Line 10 train leaves the Xilang Terminal and I enter…

Huawei Station (not that Huawei Station that is in Shenzhen and is named after the Huawei you are thinking about) which is secretly a massive station reserved for interchange with the Line 10 branch and Foshan Metro Line 11.

Planning documents show Huawei Station will have a cross platform interchange with a proposed branch of Line 10 to Foshan City. Line 10 trains will use the outer tracks of the station and Line 10 Branch trains will use the inner two tracks.

The Line 10 Branch is a long western extension of Guangzhou Metro Line 10 that heads deep into Foshan City, providing relief to the congested Guangfo Line (Foshan Metro Line 1), a light metro line that serves over 500,000 people per day.

When I arrived at the station, I could see that the two inner tracks were blocked off with a semi permanent white wall. These barriers are waiting for the day to give way and host platform screen doors with Line 10 Branch trains behind them, shuttling people to and from Guangzhou and Foshan.

Note that the station has extra wide platforms, as this station is also a future interchange station with Foshan Metro Line 11. Beneath the two sets of Line 10 platforms are reservations for the future perpendicular Foshan Metro Line 11. The west end of the platform has publicly accessible knock out panels enclosed with glass railings that are reserved spots for the vertical accesses connecting Line 10 to the future Foshan Metro Line 11 platforms.

Planning documents below show how Foshan Metro Line 11 will connect with Guangzhou Metro Line 10 at the west end of Huawei Station, hence why the knock out panels (and other reserved structures) are on the west side of both Line 10 platforms.

A String of Missed Connections

The next station of note I passed through on my tour of Line 10 South is Gongye Avenue South station, which frustratingly has a missed connection with Guangfo Metro Shixi Station. The original plan for this station called for a 200m long interchange passage to connect the two station boxes. However, the plan for an interchange was cancelled and Line 10’s Shixi Station box was designated as a separate station named Gongye Avenue South station.

The logic behind this cancellation is the perceived low demand for transfers between Line 10 and Guangfo Line at Shixi. Line 10 already connects with the Guangfo Line at the terminal station where I started, Xilang Station. Some netizens suspect there were land take issues with the adjacent Jinbi Garden residential estate impacting the viability of building this interchange corridor. However, I have found no evidence of this to be the case. I suspect it could be the Guangzhou Traffic Department (one of the most carbrained ones in China) not wanting traffic lanes to be closed to build the transfer corridor and is not being cooperative.

I continue on Line 10 and pass through an inactivated station called Sun Yat-sen University South Gate. The station has not opened with the rest of Line 10 as some surface ancillary station structures critical to the operation of the station (fire escape, ventilation etc) have still not been built due to land take issues.

I did not have time to whip out my camera to film as we passed by the unopened station, whose platforms are largely complete. Below is the POV from the train passing through the University South Gate station from Wilson_Z:

Similar to Gongye South Avenue and Shixi Stations, Sun Yat-sen University South Gate Station is a missed connection with Line 8’s Sun Yat-sen University Station. Originally Line 10 was planned to run underground through the central axis of Sun Yat-sen University and connect with Line 8. The University was skeptical of the plan from the onset on the grounds that vibrations would impact research and construction would damage the delicate university buildings dating from the Chinese Republican Era on each side of the axis.

Ultimately, Guangzhou Metro modified the original Line 10 alignment to divert away from the university grounds after getting feedback from students and the university administration. This meant that a Line 10 station that would have directly served the university and have a (somewhat long) fare paid interchange with Line 8 could not be built.

Universities opposing transit projects on their campus grounds are not new in China or anywhere else in the world. For example, Tsinghua University in Beijing has stopped Beijing Subway Line 9 and 15 extension proposals from entering the campus grounds. Even Guangzhou Metro Line 8’s current alignment today, which hugs the southern edge of Sun Yat-sen University and has an off-center Sun Yat-sen University station west of the university’s south gate, was also the result of pushback from the university on Line 8’s initial proposal to put a station directly serving campus grounds.

The above map reveals the culmination of Sun Yat-sen University’s decades long track record of opposing metro projects. Today it results in a missed connection between two major metro lines of Line 8 and 10, with the respective lines’ stations being over 500m apart and both stations shoved far away from the university grounds.

After passing the unopened Sun Yat-sen University South Gate Station, my Line 10 train heads to Binjiang East Road Station. This station was designed to be an interchange with Line 28, Guangzhou’s approved east-west high speed metro line. The station has reservations for Line 28 platforms to run under the now open Line 10 platforms.

Below is the Binjiang East Road Station plan from 广州市规划和自然资源局关于轨道交通十号线工程滨江东路站总平面规划方案的复函 showing the reservations for Line 28 platforms to run skewed under Line 10’s platforms.

Planning adjustments made to Line 28 in 2024 cancelled a few stations on the line including the station at Binjiang East Road, meaning Line 28 will run under Line 10 without an interchange. This change led to some controversy with netizens noting the increasingly low connectivity of Line 28 with the additional missed connection. However, given that Line 28 is an express high speed metro line, wide stop spacings are expected. It is difficult to provide a transfer for every Guangzhou Metro line Line 28 crosses while maintaining such wide stop spacing.

After passing through Binjiang East Road, I end my trip on Line 10 and end up back at Wuyangcun Station the 2nd to last station of the currently open Line 10 South segment.

As noted in Part 1 below, Line 10 will extend further northeast, taking over a branch of Line 3 and hopefully providing relief to Line 3’s operations.

I ran out of time to look at Line 12 West so this might have to happen in a part 3+ next time I am in Guangzhou, most likely when Guangzhou opens more metro lines at the end of the year which is in a few months. In the meantime I close out with these thoughts:

The Near Future

Edit: I was expecting all these lines listed below to be operational during China’s well known end of year opening bonanza in December. However, while gallivanting with family around Yunnan Province, I found out that Guangzhou quietly opened a lot of the lines I mentioned below just before National Day (October 1) instead of the usual last week of December. Of all the metro lines below, only the final section of Line 22 Phase 1 has not opened as of publishing this post.

With the opening of Line 10 and Line 12 East/West, I go back to what I posited at the beginning of Part 1 of this post: With this batch of new metro lines now open, when will the next openings be? For future Substack posts on the experience, history, and impacts of these Guangzhou Metro openings, please like, share and subscribe!

Based on online chatter, Guangzhou still has gas in the tank for more openings lined up near the end of this year. With the 3 metro segments I toured in June (Line 10 Phase 1 South and Line 12 East and West), there are still three more metro segments slated to open at the end of 2025. These sections are speculated to include the east section of Line 13 Phase 2, first section of Line 14 Phase 2 and final section of Line 22 Phase 1 as well as extensions to four regional railways as I noted in the tweet thread below:

When walking around the new Line 10 and 12 extensions, I see advertisements from Guangzhou Metro offering station retail spaces for lease at Line 14 Phase 2 and regional railway lines, confirming (in my eyes) that they are slated to open at the end of the year. Update: They opened these lines just before National Day (October 1)

In the meantime, it seems the new lines that opened these past few months (Line 10, Line 11 and Line 12) are making good progress on providing relief for Guangzhou’s notoriously congested metro system.

The number of stations with regular passenger flow restrictions (these are not fun; it sometimes takes up to 30 minutes to enter a station with passenger flow restrictions) has been steadily increasing every year. This year, the number of stations with regular passenger flow restrictions has been steadily decreasing despite ridership growth. With the few month old Line 11 reaching an average daily ridership of 550,000 trips per day in June 2025 growing by 19% since opening a few months ago in December 2024, I noticed that Line 5 has not been too congested during my Sunday tour, so Line 10 seems to be providing some relief.

The Future

Netizens noted that the Guangzhou Municipal Government may have reached the investment expenditure threshold of the currently approved phase of metro expansion (Phase III) with the opening of these new line segments. This means Guangzhou can submit the next phase of metro expansion (Phase IV) for national approval. It is currently not officially submitted to any higher government bodies for review but rumor has it that a number of projects are proposed in the new expansion program, totaling over 175 km of new metro lines. Projects include two new parallel relief lines for Line 3, Line 19 and Line 26, as well as a Phase II eastern extension of Line 10.

In addition, it lends credence to the rumor that Guangzhou now has the capacity to start construction on the approved and now finalized Line 28 (as noted above, without a transfer to Line 10) this year. Line 28 is Guangzhou’s first east west high speed regional metro.

I will be watching Guangzhou’s next moves on metro expansion closely. So to continue seeing the evolution of the world’s largest metro expansion program: like, share and subscribe for more of my personal experiences and thoughts on metro openings in China.

Or drop me a Ko-fi if you like what you read!

Apart from mandating the absorption of large amounts of private railway infrastructure, the Railway Nationalization Act also restricted development of new private railways. Japanese private railway capital was diverted to further develop Japan’s American-style electric interurbans, many at the time were legally classed as tramways. These later became the major/semi-major private railways we know today. During the legislative process of the Railway Nationalization Act, a group of investors inspired by “high speed” +100kph electric train developments overseas, used a different legal framework in an attempt to gain national approval for constructing a “high speed” non-regional (eg intercity) railway from Tokyo to Osaka. The prospective company to handle this new line was proposed to be called the “Japan Electric Railway Company”.

To save on cost of land acquisition, the Tokyo terminus of this proposed line is an outlying station at Hiro-o near Shibuya Village just outside of Tokyo’s built up area. The Osaka terminal will be at the actual Osaka Station. The line will take inspiration from high performance railways overseas at the turn of the century, using 1435mm standard gauge and use 3rd rail electric traction. Trains would be leaving every 30 mins and reach 120kph, later on perhaps 160kph or even 200kph. Assuming 120kph operations trains are planned to complete the trip from Tokyo to Osaka in less than 6 hours. This proposal was rejected just after the Railway Nationalization Act was promulgated and enforced in Year 40 of the Meiji Era (1907).

After absorbing and monopolizing intercity railways, the now massive Japanese Government Railways then began to look at other global powers to further industrialize and reform as the Showa Era began. During Year 14 of the Showa Era (1939), the Imperial Japanese Government, inspired by higher performance railways overseas (and I suspect still remembered the failed Japan Electric Railway proposal), unveiled a plan for a 1435mm “broad gauge” trunk line. The Japanese press opted for a more playful name for the project: The “Bullet Train” Plan.

The published Report on Japanese National Railway Trunkline (日本国有鉄道幹線調査会の答申について) from the National Archives of Japan describes the line as a railway built to higher standards than the Japanese Government Railway’s existing national Cape gauge network. The report argued that the existing Cape gauge network is reaching passenger and freight capacity but Japan’s limited road and waterway infrastructure can not match the speed, cost and scale of railway transport. Japan in the early Showa Era did not have any limited access expressways and most intercity roads are still unpaved. This was a direct result of the Meiji Government for decades prioritizing funding for ports and railways instead of road transport.

The project was proposed to use high powered steam and/or electric locomotives hauling passenger and freight trains running up to 150kph and, perhaps later on, up to 200kph. The first section of line to open would be from Tokyo via Nagoya/Osaka and run north of the Inland Sea to Shimonoseki, on the western end of Honshu just north of Kita-kyushu. In the future this line could be extended beyond the Japanese home islands to the Korean Peninsula, Manchuria and China via an undersea tunnel under the Korean Strait. It will allow sleeper Bullet Trains using this line to reach Seoul, Shenyang and Beijing from Tokyo in just over 24 hours, maybe less if 200kph speeds are realized. Even grander plans envisioned broad gauge trunk line trains to reach Western Europe (and connect with their “broad gauge” system) and Southeast Asia.

This line was proposed to be partly electrified (from Tokyo to Osaka/Kansai) with sections west of Kansai to be initially non-electrified upon opening. Meaning some Bullet Trains would need to use steam traction. The plan would use hypothetical Type HC51, HD53 and HD60 steam locomotives, upscaled from a variety of existing designs ranging from South Manchuria Railway’s Pashina-type locomotives hauling the Asia Express service and German Reich Railway’s experimental Class 05 locomotives. Electric traction would be provided by HEH50 locomotives envisioned to reach 210kph.

None of the proposed bullet train locomotives ever got built. The Saitama Railway Museum has some blueprints and a model of what a HC51 steam passenger bullet train locomotive would look like.

A blog by 新之介, recreated what a HD53 steam passenger bullet train locomotive would look like based on rough historical renderings/blueprints and the existing South Manchuria Railway Asia Express which the HD53 was purportedly going to be based on. The recreation seen below looks like a sleeker Mallard but Japanese. The HD53 was planned to be the high speed 200kph revenue service steam locomotive for superexpress passenger services.

The HD60 is the freight locomotive but visual or even technical information on that design is extremely sparse.

The cost of the entire project was estimated to be 172.5 billion yen which includes 10 billion yen for rolling stock but does not include financing costs. The project is expected to allow for trains to go from Tokyo to Osaka in about 4.5 hours. Construction of the broad gauge line from Tokyo to Shimonoseki, officially started on Year 15 of the Showa Era. Land was acquired all over Honshu Island, south and west of Tokyo, for the project’s right of way, sometimes using quite unsavory means as typical of Imperial Japan. Officials and surveyors would show up and drive stakes to mark land for the new bullet train line, sometimes without notice and/or permission of the original private owners of the land. Physical construction of the project was marked by the initiation of works on the Tanna (called Shin-Tanna), Nihonzaka and Osakayama Tunnels on the Tokaido section.

The project was halted due to Japan’s deteriorating situation during the Second Global War. However, the Nihonzaka and Osakayama Tunnels continued construction and was eventually opened and used by the zairaisen Tokaido Mainline when complete. During the project’s hiatus and war, Japanese Government Railways did take advantage of acquiring more land for the bullet train right of way entering into central urban areas at zero to rock bottom prices due to constant American firebombing.

After Japan lost the war and started rebuilding, congestion on the Tokaido Mainline prompted the revival of the “Broad Gauge Trunk Line/Bullet Train” project at the newly reformed Japan National Railways (JNR). At the time only about 20% of Japan’s first-class arterial roads were paved. National Route 1, one of the most important arterial roads in Japan connecting Tokyo with Osaka, was only two-thirds paved. There were still no limited access expressways in Japan. The alignment of this Bullet Train was modified slightly from the original Bullet Train plan and made maximal use of the land acquired from the Imperial-era. Compared with the older JGR Bullet Train plan, the first section of the JNR plan to open was cut back from Shimonoseki to Osaka. Leaving the section west of Shin-Osaka to Shimonoseki to be completed at a later date as the JNR San’yo Shinkansen. The land (aggressively) acquired for the San’yo Section during the war was disposed back to private owners.

Construction on the revived project was kicked off with a ceremony at the half finished Shin-Tanna Tunnel for the now renamed “Tokaido Shinkansen” or Tokaido new trunk line project. During construction of the Shinkansen, the completed Nihonzaka tunnel from the bullet train plan used by the Tokaido Mainline (today’s JR-Tokai Tokaido Mainline) was transferred for Tokaido Shinkansen use. In addition, the completed tunnel built beside the Osakayama Tunnel in Kansai intended to be used for the bullet train plan was to remain used for the express tracks of the quadruple track Tokaido Mainline (today a segment of JR-W’s Biwako Line) and the Tokaido Shinkansen plan rerouted the alignment to another separate new tunnel called the Higashiyama Tunnel. EMU operations where adopted for the Shinkansen project to the point where the vestigial surviving freight plans proposed using EMU-like trains not too dissimilar to the existing current day narrow gauge JR-Freight M250 series. However, later on, all plans for allowances for freight traffic on the Shinkansen was dropped.

The old Bullet Train plan called for Shin-Osaka Station to be located on the JR-W Kyoto Line between Higashi-Yodogawa Station and the Kanzaki River. Shin-Osaka Station from the Bullet Train project was shifted further south closer to Osaka City for Shinkansen project by using the right of way of a freight bypass line called the Northern Freight Line (北方貨物線). This version of Shin-Osaka is what was ultimately built today and the original right of way for the bullet train project became a major road called the Shonai-Shimo Line (庄内新庄線).

The first sections of the Shinkansen project completed was actually a section in Kansai between Osaka and Kyoto around Shimamoto. Here the Tokaido Shinkansen was planned to run parallel with Hankyu’s at-grade running interurban Kyoto Mainline. JNR and Hankyu decided to collaborate and upgrade the corridor to a full quad track elevated structure. The Shinkansen viaduct was built beside the Kyoto Mainline. When the Shinkansen viaduct was completed, for 8 months on Year 38 of the Showa Era (1963), Hankyu trains used the Shinkansen tracks while the Kyoto Mainline was demolished and rebuilt as a viaduct assisted by the common gauge used in both systems.

Hankyu ran trains on the Shinkansen tracks complete with temporary stations and using 1,500V DC traction but on modern Shinkansen auto-tensioning overhead catenary. After the Hankyu portion of the viaduct was complete, Hankyu trains switched back and the Shinkansen viaduct transitioned its substation equipment to output 25kV 60 Hz AC traction.

Below is my video of the corridor in 2024 when I was taking a Hankyu train to Kyoto.

The Tokaido Shinkansen opened on Year 39 of the Showa Era in time for the 1964 Tokyo Olympics, almost 25 years since the start of land acquisition and initial construction of the Bullet Train plan. Running up to 200kph, the Shinkansen became the fastest train service in the world. The Tokaido Shinkansen started with 30 pairs of passenger services per day, alternating with a hourly skip stop “Hikari” service with an hourly all stop “Kodama” service creating a combined 30 minute headway:

It is important to note by Year 40 of the Showa Era (1965) when the Shinkansen has been operating for one year, Japan’s entire limited access expressway network consists of a few short urban sections in Osaka and Tokyo and a section of the intercity Meishin Expressway from Osaka to Shiga Prefecture.

In a span of 50 years, this humble “new trunk line” grew to be one of the busiest high speed railway lines in the world. Growing from 60 services per day (~30 pairs) on opening, to up to 383 (~190 pairs) per day in Year 7 of the Reiwa Era. Chinese leader Deng Xiaoping had a short candid interview while riding the Shinkansen during his visit to Japan in October 1978 where he opined:

就感觉到快，有催人跑的意思。我们现在正合适坐这样的车。

It feels fast, [the train] gives you the desire to run. This train’s form is suitable for our use.

A Chinese movie retelling Deng’s visit to Japan also showed him pondering on the Shinkansen:

The opening of the Shinkansen allowed for better regional/local services on the old Zairaisen Tokaido Line. At the start of the Heisei Era, a faster “Nozomi” service stopping only at Nagoya and Kyoto was introduced, becoming the dominant service pattern by the 2000s. At the same time, speeds where gradually raised from 200kph to 285kph as fast non-stop megacity to megacity connections become increasingly important while non-megacity Japan become more automobile dependent with competition from the increasingly dense Japanese expressway network.

The Shinkansen was originally envisioned to simply be a higher performance broad gauge general railway for Japan or in other words, a literal “new trunk line”. Today, the original line basically operates as a passenger only metro-like shuttle between Kanto, Chubu and Kansai Regions with 16 trains coming every hour between 7 to 9AM and 5 to 8PM every day and 12 to 15 trains per hour at all other times of the day between 6AM to 11PM. Most trains are Nozomi services that only stop in stations in Greater Tokyo as well as the main station in Nagoya, Kyoto and (Shin-) Osaka completing the trip in 2.5 hours and moving almost half a million people per day.

Across the sea unlike Meiji Era Japan, Qing Dynasty China was traditionally very ambivalent or even outright hostile to the concept of railways. Europeans have been attempting to sell railway technology to Qing Dynasty China since the late Edo period just before the Meiji Restoration, only to be ignored. Still, the Qing Government managed to slowly accept the invention and built China’s first railways around Hebei and Taiwan Provinces before the end of the 1800s.

Fast forward over 100 years, modern China became very interested in industrializing and looked overseas for technology and ideas on development. China began to look at other global powers to further modernize as Deng Xiaoping’s Reform and Opening Up Era began. The impression of the Shinkansen on China has never faded after Deng’s trip to Japan. In the beginning of the 1990s China’s sprawling Ministry of Railways were inspired by higher performance railways overseas after seeing the success of other HSR systems in Japan and Europe. The justification at the time was that China’s existing conventional railway network is reaching passenger and freight capacity but China’s road and waterway infrastructure can not match the speed, cost and scale of railway transport.

Unlike Japan, planning for the Chinese HSR system started in the era of rapid motorization and mass aviation. China had a competing national expressway development program, airport network development masterplan and is motorizing as the economy rapidly grew. China’s first expressway opened in 1988 in Shanghai. 20 years later, China’s expressway network grew to almost 54,000 km by 2008. Only then did the first +300kph HSR line started operating.

Initial projects were “speed up campaigns” that rebuilt and slightly realigned existing conventional lines for higher speeds, sometimes pushing existing Chinese Republican Era lines to support up to 200kph operations. At times the upgrades are done at the expense of accommodation for freight service. China’s first true greenfield high speed line, the Qinhuangdao-Shenyang Passenger Dedicated Line opened in 2003. The line was designed to allow trains to reach 250kph but for years the line was running conventional SS9 electric locomotives hauling passenger cars at speeds up to 160kph. The first true HSR service in China using world class trains and infrastructure was opened with a similar backdrop as Japan; In time for the Olympics. This time it was the Beijing 2008 Olympics, where the Beijing-Tianjin Intercity HSL opened with trains running up to 350kph; becoming the fastest train service in the world and thrusting China into the high-speed railway era. The rest is history.

China’s cities in the late 90s were not bombed out by the United States nor did China have a 25 year head start in planning and land acquisition for a high speed railway system as China rapidly urbanized. Contrary to popular Western belief, urban and rural land can not simply be confiscated in 2000s China like it was in Imperial Japan. As such it was generally cheaper to locate HSR stations in the outskirts of growing early 2000s Chinese cities. There was also a -somewhat misguided- hope of developing a new district around all these new HSR stations like Xin-Guangzhou. Some of these outlying rural/small city new HSR stations remain surrounded by minimal development like “beet field” stations of the French TGV to this day. However, rapid Chinese urbanization put these new outlying HSR stations of many major Chinese cities in what is now considered the inner suburbs. Additionally, unlike Nihon, China’s HSR services stated to use existing tracks to enter legacy urban stations à la République Française as the entire Zhongguo railway network, high speed or otherwise, uses Nippon’s “broad gauge”.

15 years later, from the opening of the initial Beijing-Tianjin Intercity HSL, China constructed the largest, busiest and fastest high speed railway system on the planet. By 2024, the network was reaching 47,000km of lines, serving almost 10 million people per day with trains operating up to 350kph. Some lines are designed to host revenue operations of up to 380kph and construction recently started on a line between Chengdu and Chongqing that will allow under development trains to reach a revenue speed of 400kph. Increasingly, the goal of the Chinese HSR system has been to connect China’s numerous megacities with other megacities as quickly as possible with rural/small city new HSR stations becoming more and more peripheral to HSR ridership.

Ridership exploded as China’s HSR network was being built out with older HSLs getting saturated with more and more services. Key trunk HSR lines are approaching daily service volumes of the Tokaido Shinkansen, shown below in daily high speed railway service train pair maps made by 吡啶 et al.

The comparison of Chinese and Japanese service levels was only done above in 2024, an updated Chinese service level map for mid-2025 is shown below. Note in the span of one year, the massive growth in provided daily services; particularly on the Xi’an-Zhengzhou, Zhengzhou-Changsha, Changsha-Nanchang, Nanchang-Shenzhen, Shenzhen-Xiamen, Wuhan-Yichang Lines. Japan’s service levels have not changed much between 2024-2025.

At the top end of Chinese HSR service levels you have the Guangzhou-Shenzhen HSL averaging 171.5 pairs of trains per day, approaching the ~190 train pairs per day on the Tokaido Shinkansen. Other busy corridors include large sections of the Beijing-Guangzhou and Beijing-Shanghai HSL that exceed an average 150 pairs of trains per day. This is somewhere between the service levels of the San’yo and Tokaido Shinkansen west of Nagoya. Below map shows that almost half (47%) of the Mainland Chinese HSR system operates with the same or greater daily service intensity as Taiwan’s HSL or the quieter section of the San’yo Shinkansen west of Hiroshima. That same section of the San’yo Shinkansen has been proposed to be part of the initial “bullet train” plan almost 100 years ago.

It is important to note China’s key lines are close to the Tokaido Shinkansen’s service numbers because China’s HSR system has a longer operational span, with “red-eye” services that run well past 12PM into the night on top of China’s much vaunted sleeper high speed railway services. In contrast, Shinkansen services all stop running just after 11PM. Taiwan’s HSL starts a bit later in the day but ends at midnight. This extended operational span in Mainland China can add 10-20 additional pairs of trains per day. Regardless, like the Shinkansen over its +60 year history, the service levels of the 15 year old Chinese HSR system has been gradually increasing as ridership matures and grows. This is not to mention the Chinese HSR system is beginning to carry freight on high speed EMU trains, which are not counted in the maps:

China’s high speed railway system’s sheer size makes it undefinable in relation to other HSR systems. It increasingly operates like an amalgamation of high speed railway concepts from around the world. Many Chinese “high speed railway lines” are Class I 200-250kph lines that also serve conventional locomotive hauled passenger and freight services that reach 120-160kph, similar what Japan originally envisioned for its Bullet Train project. In addition, the CRH system increasingly looks more in line with the French TGV (very fast direct lines with outlying stations but central stations can be accessed with normal speed lines) with notes of German ICE (mixing of high speed 300kph services with regular train services and designing greenfield HSLs for slower heavier freight services).

As ridership for the Chinese HSR system grows and matures, services on new lines get ramped up to levels similar to the Shinkansen. Many newly opened lines may start out with 20-30 pairs of trains per day (sound familiar?) but grow to over 160 pairs per day a decade later, which is approaching the service intensity of mature core Shinkansen corridors like San’yo, Tohoku and Tokaido.

As noted before, half of the Mainland Chinese HSR system is running at service densities similar to the successful Taiwan HSL. However, it is important to note the “more lightly used” other half of the Mainland Chinese HSR system consists of newly opened lines still building up their ridership or Class I lines that also host conventional passenger and/or freight trains which are not counted. The map below shows the network when adjusted to count conventional locomotive hauled passenger train services but still does not account for any freight services (high speed EMU freight or regular freight). Even the meme HSL to Xinjiang looks pretty good with 30-70 pairs of passenger trains per day depending on the section.

Yes the HSL to Xinjiang can run regular passenger trains. Video below of a conventional passenger train running on the Lanzhou-Xinjiang HSL by 不会飞的CRH5G:

The map above weights each longer locomotive conventional train pair as 1.5 pairs. With this weighting, the quietest sections of the Lanzhou–Xinjiang HSL still sees the equivalent of about 30-40 pairs of passenger trains a day. To put that minimal level of service into perspective, 30-40 pairs of passenger trains a day is the same level of intercity service provided by Amtrak on the busiest section of the Northeast Corridor combined with Keystone services (south of New York). There are less than 10 pairs of VIA intercity trains a day on “The Corridor” between Toronto and Montreal.

The above map still does not show daily freight trains on mixed lines, which can reach 40-50 pairs on some corridors. For example, freight congestion of the older Lanzhou–Xinjiang slow line due to the growth of China-EU freight trains has forced some (I assumed lighter) freight trains to use the Lanzhou–Xinjiang HSL as a relief line. Regardless, if you want to discuss the wonderful nuances of Chinese HSR you can continue by reading my other long post below:

Regardless, China and Japan are juggernauts in the railway world by pushing railway technology to the limit in an attempt solve different problems in each respective country. Such as using massive scale (China) or high intensity (Japan). Strangely they also had similar approaches and opening rational when building their first HSR lines.

Like, share and subscribe for more wanderings of the mind on trains.

Or drop me a Ko-fi if you like what you read!

Pekingnology

China massively overbuilt high-speed rail, says leading economic geographer

What you are about to read is a lengthy, comprehensive, sharp — and at times even indignant — criticism of China’s high-speed rail development by one of the country’s leading official economic geographers recently…

Read more

a year ago · 53 likes · 12 comments · Yuxuan JIA, Xintian Li, Zhao Huiyi, and Zichen Wang

Or you can read the unadulterated Hanzi version straight from Mr. Land Transportation himself.

The contributing authors of the translated piece noted that it was:

lengthy, comprehensive, sharp — and at times even indignant

The author in this piece was lengthy, sharp — (*gasp* emdash) and ironically described as “at times even indignant”. Ironic because I don’t think Mr. Lu, hereinto referred to as the Author, was fair to China’s HSR system in his critique.

House cleaning

I will state for the record a subscriber (so like, share and subscribe) alerted me of this translated article hours after it was published, followed by Reece Martin of Next Metro fame asking for my thoughts on the translated text a few days later. During the long process of writing this a few additional subscribers reached out asking for my comment. I think it is important to set the record straight here not because China and/or its massive high speed railway system is infallible or something (it isn’t).

After translation the Author’s negative coverage has become used in two ways overseas:

• “Experts” that use the critique China’s high speed railway system as a talking point to discredit the entire concept of high speed rail in their countries. As such the Author’s article has been a commonly used as a citation in their protestations against high speed rail since the translated version was published weeks ago.

• Others still, strictly motivated by ideology, would use the Author’s article as a general dig against China’s model of development.

Uniquely some use the Author’s article to disprove of both China and high speed rail at the same time (they are special). Due to the contrived nature of the above two perspectives upholding the Author’s article for validation, hopefully the Author’s article is factually robust enough to support whatever high speed rail development they are arguing against in their own countries. Sadly this does not seem to be the case, as upon closer inspection, the Author’s article is riddled with questionable opinions, wild conjecture and most damningly, technical errors. Anyone using the Author’s article to argue against high speed rail should not be proud of themselves because it shows they did not do any research and are highly susceptible to very willingly take things at face value when it conveniently fits their narrative. They are not experts and their options derived from support of the Author’s article should be discarded.

In this initial post, I will focus on the more technical aspects of the article, the Author goes deep into finance and economics in many sections. My observation is many of these finance and economics opinions the Author gives tends to have a very narrow focus and is overall quite dismissive of broader benefits HSR can bring, a common skeptical slant the Author has. Maybe a further discussion on this can be done in a later part 2. Regardless, I will focus on the technical side in my review in this part 1. Why? Well, my view is that if you are going to (rightfully) criticize something, I expect you to get the facts at hand correct. The goes for overseas commentary citing this Author’s article to discredit the CRH system or oppose HSR schemes in their own country. Again, if you uncritically praised this Author’s article and/or used it a proof for your own arguments against other high speed railway schemes (whatever they may be) perhaps some lateral reading on the topic is in order for you. Doubly true if you claim to be a “transportation expert” using what Mr. Lu has produced here to support your argument. As broadly speaking, the Author makes many incorrect assumptions about how railways work or makes mistakes on basic Chinese geography. The article seems to be a long distillation of passages the Author has previously posted, as such it is long but I will review a few sections in this part 1, starting at the top.

For the Glory of High Speed Rail

The Author begins with a short general overview of recent general Chinese transportation and economic history. Then quickly dives to “specifically address the issues arising in China’s HSR planning and construction from the perspectives of rationality and economy.”, where he further elaborates in more detail on China’s high speed railway development, touching on initial transportation planning rationales. Reading Chinese high speed railway history from his perspective, I get the feeling that the Author’s views on China’s entire high speed railway program can be summarized as: “problematic waves and waves of irrational expansion, punctuated by eye watering statistics” which they use to drive their high speed railway overbuilding thesis with statements such as “70,000 km of high speed railway in China by 2035”, mirroring sensationalist Western media headlines, which also get China’s high speed railway development wrong.

He ends the section promising to “tackle all these issues, albeit from a peripheral angle, and can only expose a fraction of the full picture.”. I will also do the same in multiple parts. With that out of the way, the Author continues with a section on…

What is HSR?

This section is a bit of a contrived mess. The Author basically spends several paragraphs explaining that faster trains mean more expensive tracks. TL;DR, to the Author, high speed railways are just faster more expensive conventional railways. Yes, but that does not answer the question. What is the definition of HSR we will be using? What is the definition other people using? Is high speed rail a railway designed for speeds of +200kph? +250kph? +300kph? What if it is designed to support 300kph operations but currently runs passenger trains at 200kph and reserved for freight operations? Can passenger and freight trains both run on it? The Author also does not properly define what conventional railways are, noting a conventional railway is designed for max speed of 120kph in one paragraph then redefining it as a railway with a top speed of 160kph in another paragraph when it is convenient to his argument. (The latter is correct as China sees it)

Buried somewhere in the middle of this section is what I believe to be the Author’s actual definition of high speed rail: a passenger dedicated railway line with an ultimate design speed of at least 250kph. Which roughly lines up with China’s current official definition. The problem is so much of China’s network deemed “high speed rail” does not properly map to the Author’s definition and the official definition of HSR has shifted around over the years. An example would be the latter inclusion of distinctions such as “express rail” (快速铁路) in the Chinese technical lexicon. Overseas this is better known as higher speed rail and is now in contrast with passenger dedicated +250kph railways officially designated as high speed railways. Before the inclusion of the express rail concept, many older sub-250kph mixed traffic lines were classed as “high speed railways”. Overseas commenters (and the Author) still assume many of these express railway lines have the utility (or limited utility of being specialized only for +250kph passenger services) and design standards of conventional high speed lines.

This is why setting up definitions matter. Just stating “China is overbuilding HSR with 70,000 km of HSR planned by 2035” without context is clickbait when no one knows what is the definition of HSR and just assumes all the HSR lines in question are like the Shinkansen. This is how sensationalist stories of overbuilding sprout up. Is it 70,000km of the fastest top end passenger only French LGV-style lines by 2035 or 70,000km of HSR by 2035 but half of the lines are 200-250kph and some are designed to handle some pretty serious freight and regional services. One is by nature should be optimized for the busiest and most time sensitive travel markets meaning a very narrow utility. The other has a much broader utility and can not strictly be judged only for fast long distance passenger travel.

If you look at the Openrailwaymap above off the coast of Fujian north of Xiamen, are there two completely parallel high speed railway lines (labeled red) above? Is this overbuilding? Or is one of the lines a 200-250kph railway capable of running regional and freight services (Fuzhou-Xiamen Railway formerly the Fuzhou-Xiamen High Speed Railway but now designated as an express railway), while the newer one is optimized only for 350kph long haul passenger HSR services (current Fuzhou-Xiamen High Speed Railway). Both statements can led to two different reactions: “Ahhh! overbuilding” vs. “what are the ratio of 250kph to 350kph lines”. Knowing this we can ask about how both these “high speed lines” fit into the network:

• Where can we free up capacity to run new freight services?

• Is there network wide relief value?

• How can regional and long distance traffic be divided up between the two “high speed lines”?

• What are potential regional express railway services we can now set up from freed capacity?

• Does the extra capacity allow both long distance and regional express railway services be more frequent?

These nuances of railway design and roles are never discussed by the Author in this section. If the Author explored the wonderful flavors and variations of high speed rail that can and does exist in China, I think this section would be an enlightening value add to the conversation. Suggesting China’s high speed rail system is overbuilt and quoting “70,000km of high speed rail by 2035” in the same breath without properly defining what someone can consider high speed rail is misleading on the Author’s end at best, an obfuscation at worse. Again for your information, the answer for what the 70,000 km of HSR by 2035 in China is composed of, the latter is the more correct answer. This has huge implications on the critique of high speed rail network planning that critics seem to miss. But alas the takeaway the Author wants to note in this entire section is: high speed rail is so much more expensive to build and maintain than conventional rail. Which of course dovetails into his whole thesis.

On the planning and construction of China’s HSR

The second section is huge and broken into multiple subsections. The first sub-section “Fundamental principles to be followed in designing China’s HSR network” the Author outlines his principles for HSR planning that are, at face value, pretty reasonable and logical. However he somewhat derails in this section and ends off with a critique on China’s megalopolis rail planning, which in his eyes goes against his principles of high speed rail planning. In his view, there is a risk that government policymakers have been too heavily chasing regional travel time requirements and unnecessarily overusing high speed rail as a mode for middle to short distance regional travel instead of using other modes to fill this role. Again this goes back to the Author’s impreciseness with high speed rail’s definitions and being alarmist about overbuilding. I do not see large amounts of +250kph HSLs in China strictly built to serve only regional traffic. Again there is a definition problem as regular regional “ICR” or express railway or mixed passenger freight “Class I” lines may have some 250kph capable sections and get swept up into China’s high speed rail mileage, making the high speed rail network look overbuilt.

The Author’s issue seems to be with lines like the Guangzhou-Zhuhai or Chengdu-Dujiangyan “intercity” railways or ICRs, which are all regional lines that at times technically fit the broader definition of a high speed railway. Many ICRs have become heavily used for longer distance services instead of their intended regional role. These “successful regional HSRs” clearly are envisioned to be more like a Tsukuba Express on steroids and can be further operated as such. I don’t see how they can be overbuilt. Having a network of frequent semi-high speed regional trains that connect you to a nearby megacity in an hour is not a bad thing.

In fact, for some of these ICRs the solution is to build a parallel higher standard high speed lines to divert long distance trains away from the ICRs to allow for more regional services as those ICRs have become saturated with longer distance HSR trains. These ICRs, if run properly, should serve the mid to short distance market with frequent stopping services as originally intended. However, I’m sure the Author would just say that is overbuilding as there are now two parallel high speed railways in their eyes.

On top of that, many middle to short distance markets in China are being served by high speed metros, not high speed rail, something the Author supports building more of in his Article. However, the translators failed to localize and calls them regional “suburban light rail”. Regardless, with this in mind, the Author should be advocating for China to find ways to run its ICRs and express railways more like regional railways in Europe or Japan not, saying China is building too much HSR for short to medium distance travel, mistaking that many ICR lines are included in the tally.

Basic requirements for HSR planning and construction

This second sub-section, the Author starts with a fairly reasonable (but overused in transport planning, trust me) analogy on network planning using the human anatomy:

“It is said that the total length of blood vessels in the human body is about 96,000 km, while the abdominal and thoracic aorta measures only around 20 centimetres. Though the analogy may not be exact, it serves to make the point.”

It’s clear what is driving the Author’s alarmist stance is his insistence, at a macro level, that China’s overall high speed railway scale is too large and high speed railways should only be used to connect megacities and heavily populated areas. Which I would note, China is blanketed with such conditions (megacities and heavily populated areas) so China should have a large high speed railway network. China’s urban structure also makes his human anatomy analogy not work. Such an analogy only works if there is one “heart” that the “Aorta” feeds into. China clearly is a country with a lot of “hearts” that needs a large network of “Aortas”. Again, the Author seems to be narrowly assuming China’s high speed railway network is entirely made of +250kph passenger only lines which is incorrect.

“Although HSR entails significant capital investment and ongoing maintenance costs, it should only accounts for a small share of China’s total passenger turnover. The majority of land-based passenger transport is undertaken by conventional railway systems, highway networks, and urban public transportation infrastructure.”

These are all obfuscations by the Author. China currently already possesses a large conventional railway system, highway network and urban public transportation infrastructure, all which entails significant capital investment and ongoing maintenance costs. The country is hardly neglecting it’s vast and rapidly expanding conventional railway, roadway, and urban public transportation networks. China today has a:

• 162,000 km long railway network with 48,000 km of it being high speed railway (How much is express railway?)2.

• 5.4904 million km highway network (not the same as controlled-access expressway) with 190,700 km of it being controlled-access expressway.

• 12,161 km of urban rail transit with 9,306 km being metro lines. Nine of the top ten longest metro systems in the world are now in China.3

High speed rail is not (and was never envisioned to be) the bulk of China’s overall transport network (even after all the expansion plans) but a massive integral part of it. The new high speed railway lines provide capacity relief to existing conventional railways that allow the overall network to move more freight or frees up express passenger railways for more regional services. In fact, the Author should instead be actively arguing that China’s expressway expansion be curtailed and funds diverted to his pet projects of more conventional railways and “suburban light rail” not complaining about expanding high speed railways. A lot of urbanists would argue China’s expressway network is overbuilt.

Despite outlining “Fundamental principles to be followed in designing China’s HSR network”. The Author’s personal threshold on what should a viable high speed railway floats unrealistically well above that. Mr. Lu was on the record in 2006 implying the Beijing-Shanghai HSL, today one of the world’s busiest and most profitable HSLs, should have it’s construction delayed for an opening after 2030. In 2006. He further doubled down on that opinion in an interview in 2011:

This is a completely ridiculous track record of conservativism shown by the Author. By this standard, the Tokaido Shinkansen should be the only high speed railway in the world to be built. Ever. The Author’s hardline skepticism on the whole concept of high speed rail remains to this day so a scenario where there is essentially no high speed rail anywhere may not bother him. As shown by a passage later on:

“Why insist on HSR coverage? Numerous countries in Europe and North America have no HSR systems, yet are widely regarded as modern and well-connected.”

This is a silly opinion, and shows how flippant the Author is on transportation planning for high speed rail. These comments are why I think a lot of non-Chinese overseas “transport experts” latch on to this article as “proof” against any high speed railway schemes in their countries. The US or some small European country not having high speed rail is irrelevant and the Author's statement spits in the faces of advocates in those countries (rightfully) arguing for the initiation of high speed rail projects.

“Particularly among relevant authorities and some local officials at the prefectural and county levels, there is a widespread tendency to dismiss conventional railways, some not even bothering to consider them. HSR is proposed even where it is unnecessary; where 250 km/h would suffice, a demand is made for 350 km/h instead. In some cases, parallel high-speed lines are put forward based on dubious justifications. Can these authorities offer a clear and concrete rationale? I believe not. It’s like a child’s game of make-believe—devoid of reason.”

If the Author actually looked at examples of “parallel high speed lines” like the Fuzhou-Xiamen corridor example above or the Wuhan-Yichang (and beyond to Nanjing) corridor examples in the later sections of my post. You can see the nuance that can be had in railway planning. The fact that the Author is light on actual concrete examples shows he has nothing and does not understand the different roles and flavors of high speed railways can take in China. Many new mixed freight passenger express railways that the author claims are “creating” parallel high speed lines are simply just higher performance conventional railways and have a similar utility.

“Policy pronouncements calling for “HSR access for all prefecture-level cities” [There are 293 prefecture-level cities in China, Translator’s note] and “expressway access for every county” lack a sound scientific basis.”

This is a fair critique, not everywhere needs to have high speed railway or expressway access. There is a risk of overbuilding of any infrastructure when such performance metrics are blindly followed. But high speed rail is not the only transport mode that is being pushed by these unscientific KPIs, expressway projects also are appraised in a similar manner, arguably in an even more irrational way. Yet the Author’s body of work (including this article) remains quite silent on that topic compared to his substantial commentary on high speed rail over the years.

“The construction of heavy-haul rail corridors such as the Datong-Qinhuangdao and Shuozhou-Huanghua lines—dedicated to coal transport—effectively resolved long-standing logistics bottlenecks, particularly for Shanxi’s coal exports and north–to-south coal distribution.”

The Author is incorrect in two ways. Firstly, the push for more freight traffic (such as container, chemicals and automobile) to be handled by railways by national transportation outlines this past decade has lead to more demand and congestion on railways. Secondly, the author ignores that the building out of China’s HSR network allowed passenger demands for conventional raailways to fall as existing /new demand used the new EMU services running on the parallel newly built high speed and express railways. The result is the old conventional lines are freed up to carry much more freight. Allowing China to become the busiest freight railway system in the world. high speed rail was what “resolved long-standing logistics bottlenecks” not a few heavy haul coal railways as the Author believes (but those are cool too).

Meme Weeding Sidenote: Where is the Ridership?

At the end of the “Basic requirements for HSR planning and construction” sub-section. The Author notes:

“express services on conventional rail lines remain the most popular, the most heavily used, and the most revenue-generating segment of the network.”