That jug is not a green accessory. It is a natural gas product in a plastic bottle.

Diesel exhaust fluid is 32.5 percent urea and 67.5 percent deionized water by mass (ISO 22241). The water is the easy part. The urea is ammonia that already met carbon dioxide. The ammonia is hydrogen that already met nitrogen. The hydrogen came from methane. The methane came out of a well.

Subscribe

That is the whole chain. The rest of this post is the plant floor version.

Methane becomes hydrogen

Almost every U.S. ammonia plant starts with steam methane reforming. Natural gas is the feedstock and the fuel. EPA’s greenhouse gas inventory is blunt about it. All but one domestic ammonia facility uses natural gas as feedstock. The exception is a Kansas plant that runs petroleum coke (EPA, Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990–2023).

Steam and methane go into a reformer at high heat. You get hydrogen and carbon oxides. A shift converter turns more of the carbon monoxide into carbon dioxide and still more hydrogen. Air separation, or a process air compressor on the front of the train, supplies the nitrogen.

USGS puts the geography where the gas is. In 2025, ammonia was produced by 18 companies at 38 plants in 19 states. About 57 percent of U.S. capacity sat in Louisiana, Oklahoma, and Texas because those states hold the gas that feeds the reformers. Active plants ran at about 80 percent of rated capacity. Estimated 2025 production was 14.0 million metric tons on a nitrogen-content basis (USGS, Mineral Commodity Summaries 2026).

Subscribe

This is not a side stream. IEA’s Ammonia Technology Roadmap still describes the world fleet the same way. Just over 70 percent of global ammonia production runs on natural-gas steam reforming. Most of the rest is coal gasification, concentrated in China. That gas-based fleet pulled about 170 billion cubic meters of natural gas, roughly 20 percent of industrial gas demand, when IEA published the roadmap (IEA, Ammonia Technology Roadmap).

If someone tells you the blue jug is “just water and fertilizer,” they skipped the reformer.

Hydrogen becomes ammonia

Haber-Bosch is the unit that made cheap nitrogen possible. Nitrogen from air. Hydrogen from the reformer. Iron catalyst. High pressure. High temperature. Ammonia out the back.

IEA’s latest hydrogen ledger shows how little of this chemistry has changed clothes. Global hydrogen demand grew almost 3 percent in 2025 and surpassed 100 million metric tons. Almost all of that demand stayed in the old jobs, refining and industry, which includes ammonia. Low-emissions hydrogen reached almost 1 million metric tons in 2025. IEA expects it to clear 1 percent of global production for the first time in 2026 (IEA, Global Hydrogen Review 2026).

Read that again. After a decade of hydrogen keynotes, more than 98 percent of the world’s hydrogen is still the fossil kind. The molecule in your DEF started in that pile.

Subscribe

USGS adds the use split on this side of the water. About 88 percent of domestic ammonia goes to fertilizer. The rest becomes explosives, plastics, fibers, resins, and a long list of other compounds. DEF lives in that “other” column, but it is the same molecule leaving the same converters (USGS, Mineral Commodity Summaries 2026).

I wrote the food version of this chemistry in The Fossil Fuels on Your Plate. The plate and the tailpipe share a plant.

Ammonia and CO₂ become urea

Urea is not mined. It is manufactured.

Two ammonia molecules and one carbon dioxide molecule make ammonium carbamate, then urea and water. Plants that already stripped CO₂ off the hydrogen train use that same CO₂ as a feedstock. The carbon does not disappear. It gets parked in the urea molecule until the SCR catalyst breaks the fluid back apart in the exhaust pipe.

EPA tracks the carbon that does not get parked. Net CO₂ from U.S. ammonia production was 12.2 million metric tons CO₂ equivalent in 2023, or 12,211 thousand metric tons of CO₂. That is the industrial-process number after the inventory accounts for CO₂ consumed to make urea. Gross U.S. greenhouse gas emissions that year were 6,197.3 million metric tons CO₂ equivalent. Ammonia production is a real source. It is not the national inventory (EPA, Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990–2023).

The urea that goes into DEF has to be cleaner than farm-grade product. ISO 22241 caps biuret, aldehydes, phosphates, and a short list of metals because those contaminants poison an SCR catalyst. Automotive-grade urea is still urea. The specification is tighter. The feedstock is not different.

Urea water meets diesel NOx

EPA’s 2010 heavy-duty on-highway standard cut NOx to 0.20 grams per brake horsepower-hour. Most of the industry met it with urea-based selective catalytic reduction. A doser sprays DEF into hot exhaust. Heat breaks the urea back into ammonia. On the catalyst, ammonia reduces NOx to nitrogen and water.

Subscribe

That is the same ammonia you just followed out of the reformer, temporarily stored as urea so a truck can carry it without a pressure vessel.

The air-quality result is in EPA’s own trend reports, not in a press kit. Combined emissions of six key pollutants fell 78 percent from 1970 to 2023 while the economy grew more than 320 percent. Annual nitrogen dioxide concentrations fell 62 percent since 1990 (EPA, Our Nation’s Air: Trends Through 2023).

SCR did not do that work alone. Ultra-low-sulfur diesel, particulate filters, and a lot of unglamorous maintenance did their share. But the 2010 NOx limit is why the blue jug exists. You cannot run a modern diesel freight fleet to that standard on good intentions.

I walked through the barrel’s other products in What a Barrel of Oil Actually Makes. Diesel moves the freight. Natural gas makes the fluid that lets the diesel meet the rule.

The emissions ledger, without the costume

Two books get mixed on purpose.

Book one is criteria pollutants. NOx. NO₂. Ozone precursors. That book got better because industry built hardware, including this hardware.

Book two is greenhouse gases. Ammonia plants emit CO₂ because steam reforming of methane makes hydrogen and CO₂. EPA counted 12.2 million metric tons CO₂ equivalent from that source in 2023. IEA counted more than 99 million metric tons of conventional hydrogen against less than 1 million tons of the low-emissions kind in 2025.

Subscribe

Neither book says the blue jug is magic. Both books say the same physical fact. The country cleaned up diesel NOx by installing a chemical plant on the side of every late-model truck, then filling that plant with a natural-gas derivative.

The civilization pattern is the one I laid out in US Electricity Powered 2,500% GDP Growth. CO₂ Emissions Rose Just 6%. Output went up. The dirty part of the exhaust went down. That happened because operators, refiners, and chemical plants built systems, not because someone renamed the feedstock.

Bottom line

American heavy industry is why the freight moves and why the tailpipe got cleaner.

Petroleum is not only fuel. Natural gas is not only heat. Methane becomes hydrogen. Hydrogen becomes ammonia. Ammonia and CO₂ become urea. Urea and water become the blue jug. The jug becomes nitrogen and water in the exhaust.

Wind turbines, solar panels, and battery packs have the same kind of buried feedstock chain. I have already written the oil-in-the-machine version. DEF is the gas-in-the-muffler version. Verify before you believe the sticker on the shelf.

If a policy erases the gas well, it also erases the hydrogen, the ammonia, the urea, the fertilizer, the explosives, and the fluid that keeps a 2010-and-later diesel legal. That is not a slogan. That is the process flow.

The industrial sector in America is why this chemistry runs at scale, with the safest and most reliable production we have. Keep it in the pipe long enough to finish the reaction.

-Nate

Subscribe

Subscribe to Hammer and Anvil. Paid subscribers get more in depth insights.

SOURCES

IEA, Global Hydrogen Review 2026. https://www.iea.org/reports/global-hydrogen-review-2026
IEA, Ammonia Technology Roadmap. https://www.iea.org/reports/ammonia-technology-roadmap
EPA, Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990–2023. https://www.epa.gov/ghgemissions/inventory-us-greenhouse-gas-emissions-and-sinks
EPA, Our Nation’s Air: Trends Through 2023. https://www.epa.gov/air-trends
USGS, Mineral Commodity Summaries 2026, Nitrogen (Fixed)-Ammonia. https://pubs.usgs.gov/periodicals/mcs2026/mcs2026-nitrogen.pdf
ISO 22241-1, Diesel engines - NOx reduction agent AUS 32 - Quality requirements (urea content 32.5% mass fraction).
EPA heavy-duty on-highway NOx standard, 0.20 g/bhp-hr, 2010 and later.

Built and Operating

Google’s Pryor Creek campus in Mayes County is the state’s primary hyperscale site. It has operated since 2011, spans multiple buildings, and continues to expand as part of Google’s cumulative Oklahoma investment of more than $15 billion. Power comes mainly through the Grand River Dam Authority.

Subscribe

Smaller operational facilities include TierPoint sites in Oklahoma City and Tulsa, a Cerebras Systems AI facility in Oklahoma City that uses closed-loop liquid cooling, Core Scientific’s initial Muskogee capacity, and various enterprise and carrier hotels such as those operated by Lumen, DXC at the Cherokee Industrial Park, and others. When these smaller sites are counted, trackers typically list a few dozen facilities statewide. Publicly reported operating capacity reaches into the hundreds of megawatts once the Google campus is included.

Under Construction

Google’s Stillwater campus in Payne County covers roughly 400 acres near Perkins and Richmond roads. It is planned for up to six buildings of about 300,000 square feet each and forms part of Google’s $9 billion 2025 Oklahoma package. Construction is underway with power supplied by Oklahoma Gas and Electric.

Share

Project Clydesdale sits on 506 acres in unincorporated Tulsa County near Owasso and the Cherokee Industrial Park. Beale Infrastructure broke ground in October 2025. On August 19, 2026, Google confirmed it is the end user and operator. The project includes a new on-site Public Service Company of Oklahoma substation funded by the developer and about 1.5 miles of water-line upgrades for a rural water district. Initial investment estimates are in the $1 billion range, with phases of roughly 200,000 square feet each.

Meta’s Project Anthem in east Tulsa at Fair Oaks Innovation Park broke ground on April 21, 2026. It is Meta’s first Oklahoma data center and is designed for AI workloads across more than 2 million square feet on roughly 340 acres. Investment exceeds $1 billion. Peak construction is expected to involve more than 1,000 workers, with about 100 permanent jobs once operations begin around 2028. Meta has committed to covering full water and wastewater costs and funding local infrastructure upgrades. A second-phase application was withdrawn earlier in 2026.

Subscribe

Core Scientific’s Muskogee campus is advancing high-performance computing capacity. An initial facility of roughly 70–100 MW critical load leased to CoreWeave is in final commissioning for 2026 delivery. A second unleased building is under construction with delivery targeted later. Broader expansion plans, including the Polaris acquisition, point toward a campus potential of roughly 1.5 GW gross power over time.

Google is also advancing campuses in Muskogee County near Summit and Council Hill as part of the same $9 billion package. Early development and construction activity have been reported, with power arrangements through OG&E.

Approved or Planned

Project Spring in Sand Springs covers 827 acres. White Rose Partners is the developer with Google as the prospective end user. The city council approved rezoning in early 2026. Construction is targeted for 2027–2029. Litigation over annexation and a conservation easement continues.

Subscribe

Beale’s Project Mustang at the Claremore Industrial Park in Rogers County received city approval in May 2026. It is described as air-cooled with multi-phase potential and associated tax incentives.

IREN’s Project Emerald in Pittsburg County near Kiowa involves roughly 2,000 acres and a planned load in the 1.2–1.6 GW range. It remains in development. Tax-incentive discussions were tabled and sent back for renegotiation after public pushback in mid-2026. Power ramp is projected around 2028.

Other merchant proposals, including Cloverleaf Infrastructure’s Project Open Sky near Piedmont and Beltline Energy sites in the Yukon-Luther-Piedmont corridor, sit in various stages of planning. Several face local delays or contests.

Canceled or Withdrawn

Beale’s Project Atlas in Coweta was the clearest case of a project stopped by local opposition. The proposal covered about 270 acres with multi-phase potential. The Coweta Planning Commission recommended denial of rezoning 4–1 in January 2026 after packed meetings focused on water demand, power load, noise, proximity to homes, and the use of nondisclosure agreements. Beale withdrew the application on March 30, 2026. Without rezoning the project could not proceed.

Subscribe

An unnamed prospect in Broken Arrow saw its letter of intent expire and effectively ended. A few other early-stage proposals have been delayed or stalled by recalls, resignations, and political pressure rather than formal permanent bans.

Temporary Moratoriums and Policy Changes

Multiple municipalities have enacted temporary moratoriums through the end of 2026 (or into 2027 in some cases) to study zoning, water, and power impacts. These include Oklahoma City, Tulsa (with exemptions for already-advanced projects such as Anthem), Edmond, Luther, Norman, Pawhuska, and Broken Arrow.

The Cherokee Nation implemented a policy in early August 2026 prohibiting new hyperscale data centers on tribally owned and trust lands. A citizen survey of 1,593 respondents found 64 percent opposition. The Nation also expects early consultation for projects on non-tribal land within the reservation. Projects such as Clydesdale and Mustang sit on private land and are not directly halted by the tribal policy.

At the state level, the Data Center Customer Ratepayer Protection Act of 2026 requires customers of 75 MW or more to cover their own infrastructure costs. Water reforms have increased oversight, though facilities that buy from municipal systems still leave usage less visible. Some projects emphasize closed-loop or air cooling; others rely on municipal supplies.

Power and Water Context

Large loads connect primarily through the Southwest Power Pool via utilities including OG&E, PSO, and GRDA. Utilities have filed for new generation and transmission to meet demand. Developers routinely highlight construction jobs in the hundreds to more than 1,000 at peak, permanent operational headcounts often in the 50–100 range per large campus, infrastructure upgrades they fund, and payments in lieu of taxes. Critics focus on cumulative grid strain, water intensity where evaporative cooling is used, noise, limited long-term local hiring relative to scale, and early secrecy around end users.

Subscribe

The densest activity sits in the northeastern corridor from Pryor through Owasso, Claremore, and Tulsa, with growing scale in Muskogee County. Central Oklahoma has seen heavier political resistance relative to actual construction under way. Statuses continue to change with permits, court rulings, and utility dockets. Local government agendas, Oklahoma Department of Environmental Quality air permits for backup generation, utility filings, and company announcements remain the most reliable public sources for site-specific updates.

Share

The knowledge does not disappear. It just stops flowing.

This is how normalization of deviation starts. A small difference gets accepted because it is not big enough to force a shutdown or slow the unit. The next shift sees the same reading and treats it as normal. Over time the gap grows. By the time the difference is large enough to matter, the system has already trained people to treat the early signal as background noise. The first person who noticed it learned that speaking up cost more than staying quiet.

Subscribe

Most plants do not set out to punish the person who flags a problem. The system does it anyway. Flagging creates more paperwork. It can delay production or force an uncomfortable conversation with a supervisor who is already under pressure. Sometimes the flag is treated like an accusation. Other times nothing happens at all. When the last three write-ups went nowhere, the fourth one starts to feel pointless. Experienced operators and mechanics learn this faster than anyone. The newer people still write everything up. The veterans have already figured out the real rules of the place.

I have watched this pattern play out across different plants and different crews. The conversations do not stop. Operators and mechanics still talk about the issues among themselves on the unit. Those grumpy side conversations continue. What stops is the formal signal. The knowledge stays trapped at the level where it cannot drive action, cannot enter a work order, and cannot protect the process.

Subscribe

This is an incentives problem. If your metrics mainly celebrate clean audits and zero reportable incidents, you are measuring the absence of failure after the fact. You are not measuring the presence of learning before the fact. The plants that run reliably treat an early flag as valuable information instead of a disruption. They reward the person who catches the problem while it is still small. They close the loop so the person who spoke up sees that the flag mattered. Over time that changes what people are willing to raise.

Ownership changes the dynamic more than almost any other factor. When operators and maintenance both have clear decision rights over the asset, the conversation shifts. Fixing the gauge early becomes something either group can initiate without waiting for layers of approval. It is no longer one group creating work for the other. It becomes mutual benefit. Operations gets a more reliable process and fewer surprises. Maintenance gets fewer emergency calls and less reactive work at inconvenient hours. Both sides protect the product that stays in the pipe. Both sides keep people safer.

Bottom-up systems work when the people closest to the equipment have both the authority and the reason to act on what they see. Clear decision rights give them the authority. Aligned incentives give them the reason. Without both, even good people eventually stop raising the small things. The system quietly teaches them that silence is safer than signal.

The knowledge is already on the floor. It lives in the people who stand next to the equipment every shift. The real question is whether the system makes it safer and smarter to speak up than to stay quiet.

What is one place in your system where you could reduce bureaucracy and give operators or maintenance clearer decision rights so the small issues get acted on instead of walked past?

-Nate

Subscribe

Operators and maintenance live with the equipment every shift. They catch the change in sound, the vibration that has not hit the analyzer yet, the seal that is starting to weep. The board can look fine while the person standing next to the asset already knows something is off. That knowledge is real. It is also fragile.

Too often it stays locked in grumpy talks between ops and maintenance. A mechanic mentions something to an operator on the way out of the unit. The operator nods. They both know. Then the conversation ends. Nothing gets written down. No work order gets generated. Leadership never hears it. The issue sits there until it grows into something louder, more expensive, or harder to fix.

I have always lived by a simple rule: when we keep it in the pipe, we keep people safe and we protect the value-add products we sell for good profit. That only works when the people closest to the process can get what they know into the system. When the knowledge stays trapped between two hard hats on the unit, the pipe is already starting to leak value.

In eighteen years across plants, the pattern is consistent. The gap between what gets documented and what actually happens on the floor is where risk hides. It is also where a lot of the reliability and environmental performance problems begin. Small things that an experienced operator or mechanic would flag early become larger problems because the system never received the signal.

Frontline feedback is some of the most important input an environmental or process safety person can get. Systems alone will not pull it out. Listening will. But listening requires more than an open-door policy. It requires giving the people who see the problem clear ownership and the authority to act on it, and making sure the incentives reward speaking up instead of staying quiet.

Most management systems were designed for auditability and traceability. They are good at proving what was checked and when. They are far weaker at capturing the kind of knowledge that lives in pattern recognition, muscle memory, and years of standing next to the same equipment under different conditions. That is not a software failure. It is a design reality. The system can only process what someone decides is worth entering.

The plants that close this gap treat the people closest to the process as a primary source of truth, not an afterthought. They make it easy and safe to raise the small things. They close the loop so the person who spoke up sees that it mattered. Over time, that changes the culture. Grumpy side conversations start turning into useful signals instead of dead ends.

Bottom-up knowledge does not replace engineered controls, procedures, or formal risk assessments. It makes them sharper. It gives them the early warning data they were missing. When that knowledge flows, safety improves, reliability improves, and the product stays in the pipe longer.

Safe, reliable, compliant operations depend on more than what the system can trend. They depend on what the people on the floor already know, and whether the organization is structured to hear it.

What have you seen stay locked in those grumpy talks instead of making it into the system?

6CO2 + 6H2O + light energy -> C6H12O6 + 6O2

As long as that forest stands, that carbon is securely locked away. It is a natural, God-given, self-sustaining system.

The Green Paradox and Destroying the Sink

Current green energy policies incentivize clearcutting standing forests, grinding them into pellets, shipping them across oceans on diesel carriers, and burning them for electricity. This practice actively worsens the CO2 problem in two ways: instant release and destroying the defense.

• Trees that took a century to absorb carbon are incinerated in seconds, instantly vaporizing stored carbon back into the atmosphere.

• Removing the forest eliminates the natural filter that would otherwise continue absorbing future emissions.

The Accounting Loophole

How does burning trees get classified as 100 percent renewable and net-zero? It relies on a paperwork trick. The UN Intergovernmental Panel on Climate Change (IPCC) designed rules to avoid double-counting emissions, once during harvest and once during burning. They decided emissions from burning wood should be recorded as zero in the energy sector, assuming the harvesting country would log the carbon loss in its land-use ledger.

However, political frameworks took this math trick literally. They codified biomass as carbon-neutral, allowing utilities to completely erase their smokestack emissions on paper. A rule meant to prevent double-counting became a loophole where no one counts the carbon at all.

Taxpayer-Funded Deforestation

Because international regulations treat wood smoke as zero-emission, governments use taxpayer money to heavily subsidize this destruction. For example, the United Kingdom's Drax Power Station, a retrofitted coal facility, collected a record £999 million (nearly $1.3 billion USD) in public green subsidies in 2025 alone.

These subsidies are funded directly by surcharges on British household electricity bills. Citizens are paying a premium to burn American forests just so politicians can cross out a number on a spreadsheet and declare they are meeting renewable targets.

The Physics: Wood vs. Coal

Atmospheric physics ignores political ledgers. Wood is a highly inefficient fuel source with low energy density and high moisture. Generating one megawatt-hour (MWh) of electricity requires burning a vastly larger physical mass of wood pellets than coal.

Official power plant data from the U.S. Environmental Protection Agency (EPA) analyzed by the Partnership for Policy Integrity shows that a standard utility biomass boiler emits roughly 3,000 pounds of CO2 per MWh. By comparison, a conventional coal facility emits about 2,180 pounds. Switching to wood pellets increases smokestack CO2 emissions by 40 to 50 percent, creating a carbon debt that takes 50 to 100 years of forest regrowth to repay.

Verify Before You Believe

The biomass boom proves why citizens must verify green energy claims. Celebrated emission drops are often entirely artificial, based on spreadsheet tricks that ignore the smoke billowing into the sky.

The industry claims harvested forests will eventually grow back, but the atmosphere reacts to emissions hitting the sky today, not a promise of re-absorption a century from now. Fossil fuels like coal and natural gas represent geologically compressed ancient biomass. Fossil fuels are organic. Utilizing these dense energy sources allows modern standing forests to remain intact, protecting the frontline ecosystems that actively filter our air.

SOURCES

Ember Carbon Accounting & Subsidy Report: https://ember-energy.org/latest-updates/drax-received-record-1-billion-in-biomass-subsidies-in-2025/

Partnership for Policy Integrity (PFPI) Carbon Data: https://www.pfpi.net/biomass-basics/

U.S. Environmental Protection Agency (EPA) Biomass Policy Regulations: https://www.epa.gov/climatechange/climate-change-regulatory-actions-and-initiatives

IPCC Greenhouse Gas Inventory Guidelines: https://www.ipcc-nggip.iges.or.jp/public/2006gl/

Once operational, the facility will run 24 hours a day, seven days a week, processing an initial capacity of 30,000 barrels per day (BPD) with up to one million barrels of storage. The project will bring 75 to 80 permanent, full-time operational jobs to Stephens County.

While traditional refining infrastructure focuses on standard, centralized complexes, the Duncan project represents a shift toward a decentralized, modular model. The technical design details clarify why this facility is one of the most modern refineries currently undergoing construction.

Modular Pre-Fabrication vs. Standard Infrastructure

Standard refining infrastructure requires significant upfront capital and construction timelines spanning multiple years. Because traditional plants are built-in-place on expansive footprints, engineering and engineering management are prolonged processes.

GFO's facility utilizes a Net-Zero Emissions Technology (NZET) framework, which relies on a modular layout.

• The processing units are prefabricated and scaled to fit a compact physical footprint, which compresses the construction timeline down to approximately 12 months.

• Lower upfront capital expenditures allow the plant to be built directly adjacent to existing regional oil production, reducing midstream transportation requirements.

Precision Separation vs. Thermal Cracking

To compare the operational mechanics of this facility to standard modern refineries, it helps to examine the physics of how crude oil is separated into commercial products like gasoline, diesel, jet fuel, and asphalt.

• Traditional refining relies heavily on thermal cracking, heating crude oil to high temperatures using external fuel sources. The NZET system utilizes a patented molecular separation process that operates at lower temperatures and lower pressures, reducing the baseline energy required to run the facility. 

• The Duncan facility utilizes a closed-loop system that captures and recycles its internal process gases to help power its own operations, preventing product loss and eliminating the need for routine open-air safety flaring.

• The facility is designed to integrate an advanced gas-absorption technology called Hot Potassium Carbonate (HPC). This system captures over 95% of internally generated carbon dioxide emissions from flue gases, which can then be combined with hydrogen feedstock to synthesize high-value products like Sustainable Aviation Fuel (SAF). 

Deploying smaller, localized footprints directly inside oil-producing regions like Oklahoma allows the energy industry to add processing capacity with lower overhead and an altered emissions profile compared to older industrial complexes.

Sources

Duncan Area Economic Development Foundation Announcement: https://ok-duncan.com/news/article/duncan-area-economic-development-foundation-and-stephens-county-commissioners-host-groundbreaking-ceremony-for-stephens-county-refinery-redevelopment

Clean Refineries Inc. / Green Fuels Operating Site Overview: https://www.cleanrefineries.com/green-fuels-ops

Clean Refineries Inc. Carbon Capture Partnership Release: https://www.cleanrefineries.com/post/clean-refineries-inc-subsidiary-gfo-partners-with-eden-carbon-solutions-to-transform-co2emissions-i

But if you operate an industrial facility, you know a secret that Washington politicians often ignore: The federal EPA is only the baseline.

In response to my last post, a reader asked how these federal shifts translate to everyday operations. The answer lies in a massive, distributed infrastructure of state regulators that completely dwarfs Washington's footprint. While the federal EPA employs roughly 12,500 people, the day-to-day execution of environmental law happens at the state level.

According to data from the Environmental Council of the States (ECOS), state agencies manage over 90% of federal environmental programs through delegation agreements and execute the vast majority of the nation's routine compliance enforcement. Collectively, these state agencies command a massive footprint, with total annual budgets scaling past $23 billion nationwide.

If you are dealing with a routine compliance inspection, a permit renewal, or a surprise audit, you usually aren't dealing with Washington. You are dealing with your local state capital. True regulatory simplification is a mirage because it requires untangling 51 different bureaucracies, not just one.

The Floor vs. The Ceiling: The Reality of "Primacy"

Under major federal statutes like the Clean Air Act, the Clean Water Act, and the Resource Conservation and Recovery Act (RCRA), the federal EPA sets minimum national standards. However, the system is designed for primacy, which means Congress intended for states to take over the day-to-day implementation and enforcement of these programs once their local regulations are approved as "at least as stringent as" federal rules.

But "at least as stringent" is a floor, not a ceiling. Many states explicitly choose to build a ceiling much higher than the federal floor. This manifests in three major ways:

1. More Stringent Limits and Expanded Scopes

States frequently regulate pollutants or industries that the federal government completely ignores, or they enforce much tighter thresholds.

California's Air Resources Board “famously” sets vehicle emissions and industrial standards that dictate national manufacturing because companies cannot afford to lose the California market.

The Texas Commission on Environmental Quality manages a highly industrialized coast where rules regarding Highly Reactive Volatile Organic Compounds and air permitting can be incredibly granular, featuring localized emissions cap-and-trade programs that go far beyond federal baselines.

Regulators in the Northeast and Midwest regularly step ahead to enact strict, enforceable regulations on emerging contaminants like PFAS in drinking water and soil long before federal EPA limits are finalized and litigated.

2. Redundant and Accelerated Reporting Timelines

Even if a state aligns with federal pollution limits, they rarely align on the paperwork. State Departments of Environmental Quality regularly require additional reporting layers, completely different software systems, and accelerated notification timelines.

If you have an accidental release or a permit deviation, federal law might give you a broader window to report. Your state agency, however, might demand an initial electronic notification within two hours, followed by a 14-day written engineering report.

A multi-state operator cannot simply create one standard compliance dashboard. A plant in Oklahoma, a plant in Ohio, and a plant in Pennsylvania will look entirely different from a reporting schedule perspective.

3. The Local Political Pendulum

Federal rollbacks take years to litigate and implement. State rules can change much faster based on local gubernatorial elections or regional environmental crises. When the federal government pulls back on enforcement, states with strict environmental mandates often ramp up their own enforcement to compensate, filling the vacuum. Conversely, underfunded state agencies might lack the resources to process paperwork efficiently, leading to costly project delays rather than "deregulatory" relief.

The Emotions of De-Regulation

When a federal administration promises to roll back a 1,000-page EPA rule, it makes for great headlines. But for the environmental health and safety manager on the plant floor, it rarely results in an immediate lighter workload.

If the federal EPA rescinds a rule, your state-level permit conditions do not automatically vanish. You are still legally bound by the terms of your state-issued Title V Air Permit or NPDES Water Permit until the state agency chooses to modify it, a process that can take years of bureaucratic back-and-forth.

Simplifying environmental compliance isn't a matter of rewriting code in Washington D.C. It requires addressing a decentralized, 51-headed regulatory hydra. Until we look at the state-level apparatus executing 90% of the nation's programs, any talk of true regulatory relief is just empty rhetoric.

This court-mandated action, following the SSM Litigation Group v. EPA decision, simply restores a long-standing legal framework. By reinstating sections 40 C.F.R. 70.6(g) and 71.6(g), the EPA brought back a common-sense defense for operators. It recognizes that even well-maintained facilities experience unavoidable emergencies. If a plant acts in good faith and minimizes emissions during a malfunction, it is protected from unjust civil penalties.

The EPA did not dismantle environmental oversight. It adjusted a single brick in a massive regulatory fortress.

The Physical Scale of Title 40

To understand how heavily regulated American industry remains, look at the physical size of the rulebook. Federal environmental regulations live inside Title 40 of the Code of Federal Regulations (CFR).

• A physical copy of Title 40 spans more than 30 thick volumes, split into roughly 63 distinct parts. It encompasses tens of thousands of pages of dense, multi-column technical text.

• The EPA’s jurisdiction is divided into over 1,000 separate programs, numbered from Part 1 to Part 1099.

• Part 63, which regulates hazardous air pollutants, is so comprehensive that it requires multiple printed books just to contain that single section.

The Math of "Command Words"

Policy analysts measure true regulatory burden by tracking binding restrictions. These are explicit command words carrying the force of law, specifically: shall, must, required, prohibited, and may not.

The EPA enforces well over 100,000 individual, legally binding commands. Out of the entire United States federal government, the EPA consistently ranks as one of the top two heaviest regulatory bodies.

For industrial operations, the EPA is the dominant compliance authority. Data from the Office of Management and Budget (OMB) shows that nearly 45 percent of all federal regulations impacting a manufacturing facility originate from the EPA alone.

The Bottom Line

Withdrawing the 2023 rule created zero new requirements and erased zero existing emission standards. It adjusted a legal sub-paragraph to match a federal court ruling.

The 100,000 command words are still active. The tens of thousands of pages of Title 40 remain in effect. Facilities are still legally required to invest heavily in maintenance, continuous emissions monitoring, and rapid-response containment systems.

Ignore the sensational headlines. The 30-volume reality proves that American industry remains meticulously regulated.

But the green illusion hides an even darker, deeply human cost. To harvest clean wind in the West, we rely on a supply chain that leaves a trail of toxic heavy metals, chronic illness, and neurological damage across developing industrial hubs, primarily in China.

When we think of wind turbines, we tend to picture elegant, weightless blades. In reality, they are massive consumers of base metals. A single megawatt of onshore wind capacity demands roughly 100 to 150 metric tons of steel. It also requires 2 to 3 metric tons of copper. Move those operations offshore, and the copper requirement jumps to 5 metric tons per megawatt because of the massive undersea cabling and heavy generators.

The United States cannot fulfill this demand on its own. Instead, we rely on global supply chains where one nation holds a functional monopoly. According to the International Energy Agency, China commands an average 70% global market share in processing strategic energy minerals. Even when raw ore is mined elsewhere, it is routinely shipped to Chinese smelting hubs for refinement. The situation is acute enough that the U.S. Geological Survey officially added copper to the U.S. Critical Minerals List after domestic production dropped significantly, leaving us heavily reliant on this overseas pipeline.

Refining millions of tons of steel and copper on a coal-dominant grid comes with a steep price tag, and the people living near these processing hubs are paying it with their health.

Subscribe

The refining facilities concentrated in China discharge a dangerous mix of heavy metals into local air, topsoil, and water tables. Because these industrial zones overlap with dense agricultural regions, the toxicological toll on local communities is severe:

• Cadmium: Released during copper and zinc smelting, this carcinogen migrates easily into local wheat and rice crops. Chronic ingestion triggers kidney failure and attacks bone metabolism, causing extreme brittleness and bone decay.

• Arsenic: A byproduct of copper refining that hitches a ride on airborne PM2.5 dust. Inhalation or ingestion damages cellular DNA, causing sharp spikes in skin, bladder, and lung cancers.

• Lead: Thermal processing of base metals blankets nearby towns in neurotoxic dust. Children living near these smelters consistently exhibit elevated blood lead levels, resulting in permanent cognitive impairment.

• Hexavalent Chromium: Used to forge the specialized steel alloys inside turbine nacelles and gearboxes, this compound inflicts severe oxidative DNA damage, leading to high rates of occupational lung cancer.

• Mercury: Vaporized by coal-fired steel plants and smelters, mercury settles into aquatic ecosystems and builds up in the food chain. It targets the central nervous system, permanently damaging motor skills and cognitive faculties, particularly in unborn children.

The ultimate irony of modern wind energy is a story of spatial and moral displacement. To erect a turbine that boasts “zero emissions” in an American field, vast amounts of coal are burned and devastating poisons are dumped into the food and water supplies of communities half a world away.

We are not just importing turbine components. We are outsourcing the carcinogenic and neurological costs of this energy source.

Sources and Further Reading

International Energy Agency (IEA) Critical Minerals Report and Tracker. IEA Critical Minerals

U.S. Geological Survey (USGS) Mineral Commodity Summaries. USGS Publications Warehouse

U.S. Department of Energy (DOE) Wind Energy Technologies Office. Energy.gov Wind Office

Agency for Toxic Substances and Disease Registry (ATSDR). CDC ATSDR Toxic Substances

First-generation wind installations are maturing. As they hit their 20- to 25-year operational lifespans, we are discovering that while we rushed to build these giants, we have absolutely no viable plan for their massive, messy retirement. The result is the emergence of a new kind of American landscape, the high-density, open-air turbine graveyard.

From 2.5 to 400 Gigawatts: The Looming Waste Tidal Wave

The scale of this challenge is not static; it is compounding at an exponential rate. To understand why this disposal bottleneck is turning into a crisis, look at the sheer speed of wind energy's expansion across the United States.

Subscribe

In the year 2000, U.S. installed wind capacity sat at a modest 2.5 gigawatts. By the end of 2023, that number exploded to roughly 148 to 150 gigawatts, a staggering 59-fold increase in just over two decades. Looking forward, projections from the Department of Energy indicate that capacity could approach 400 gigawatts by 2050 as grids push for higher renewable shares.

Because these massive blades have a strict service life of 20 to 25 years, this historical growth guarantees a steep, inescapable rise in decommissioned material. National Renewable Energy Laboratory (NREL) analyses project that cumulative U.S. blade waste will reach approximately 2.2 million tons by 2050. Annual retirements are set to accelerate markedly through the 2030s and 2040s as the massive fleets installed in recent decades simultaneously reach end-of-life and undergo "repowering" overhauls. We are building a mountain of waste faster than we can dig the holes to bury it.

The "Hard-Boiled Egg" Failure

The industry likes to highlight that roughly 90% of a turbine (including the steel tower, concrete foundation, and internal metallic gears) is economically recycled. They conveniently ignore the remaining portion: the highly concentrated, non-recyclable composite waste stream that makes up the blades.

The technical reality is an insurmountable chemical barrier. To create blades light enough to spin but strong enough to withstand extreme environmental loading, manufacturers mix fiberglass with cured thermoset resins like epoxy. Curing triggers an irreversible chemical reaction that permanently locks the polymer chains into a rigid, three-dimensional network. Once cured, the material cannot be melted, reshaped, or dissolved.

It is physically comparable to a hard-boiled egg. Once the chemical transition occurs, you cannot un-boil the egg back into its raw, liquid state. This irreversible cross-linking prevents the plastic matrix from being separated from the reinforcing fiberglass, rendering these massive structures exceptionally resistant to conventional recycling and standard municipal waste management methods.

A Trail of Commercial Failures

Wind energy advocates frequently point to "emerging technologies" to reassure the public that a solution is just around the corner. But official data tells a completely different story. Reports from NREL and the Association of State and Territorial Solid Waste Management Officials (ASTSWMO) reveal that true, closed-loop blade recycling has failed to achieve commercial-scale success.

Subscribe

Every highly publicized alternative has quietly stalled:

• Mechanical recycling via shredding and grinding degrades the fiberglass pieces so severely that the output can only be downcycled as low-value filler or aggregate. It fails to recover usable resins or fibers at competitive economics, leaving it non-viable for true recycling.

• Pyrolysis and thermal processes involve baking the blades at extreme temperatures to burn off resins. This requires immense energy inputs, destroys the structural strength of the glass fibers, and carries capital costs too high to survive market economics.

• Chemical recycling methods like solvolysis and hydrolysis can break down resins in a lab setting, but they have failed to advance to a commercial scale in the U.S. due to extreme process complexity and massive infrastructure gaps.

• Repurposing initiatives have repeatedly turned out to be illusions, such as a widely praised 2020 Wyoming law allowing old blades to be used as backfill material for coal mine reclamation. Years after the legislation was enacted, official sources note that not a single blade has actually been used for this purpose.

Even heavily promoted commercial partnerships, like burning blades for cement co-processing, are fundamentally misleading. Incinerating blades in cement kilns is classified as downcycling rather than material recovery. It simply utilizes the silica as a cheap sand substitute while burning the plastic, functioning as a high-temperature method of making the waste disappear rather than recycling it.

Why Landfills Can’t Stomach Them

Because real-world recycling options remain severely limited, wind developers routinely choose landfill burial as the most cost-effective and legally permissible option. However, the physical scale of modern blades presents a nightmare for landfill operators:

• Standard landfill machinery cannot compact these structures because they are engineered specifically to resist extreme physical crushing.

• Landfill operators are forced to waste significant on-site resources using specialized heavy equipment to cut the blades into shorter segments.

• Even when sectioned, the curved pieces exhibit material memory and refuse to stack efficiently, resulting in a massive, permanent loss of premium landfill airspace.

Because of these operational difficulties, many small municipal waste facilities are outright refusing to accept turbine blades. This forces operators to transport them long distances to regional mega-landfills or let them pile up in plain sight.

Mapping the Crisis

The concentration of these massive, non-biodegradable structures transfers severe spatial, visual, and ecological burdens onto local communities. It is easy to read statistics about acreage, but it is another thing entirely to see it from the road.

Subscribe

This is a street-level view taken directly from a public highway south of Sweetwater, Texas. The scale is staggering. Row upon row of hollow, white blades disappear over the horizon, stacked on metal racks directly adjacent to local fields. You can barely comprehend the total area covered from this perspective, and this is just a single roadside viewpoint in one town. This isn't a temporary storage yard; it is an open-air graveyard for a failed lifecycle plan.

These storage sites represent the physical manifestation of the industry’s disposal bottleneck. Looking at Google Earth data for just four documented regional stockpiles across the country confirms the overwhelming physical footprint of this waste:

The Central Facility in Sweetwater, Texas consumes 35.6 acres of land with a 1.2-mile perimeter, acting as a massive regional hub for blade accumulation and preparatory cutting.

The South Stockpile in Sweetwater, Texas covers an additional 9.96 acres, creating a separate high-density open-air storage zone just down the road.

The Newton, Iowa stockpile yard utilizes 8.88 acres along industrial corridors for tight, linear blade staging.

The Casper, Wyoming disposal site at 1886 Station Road claims 5.00 acres of open terrain for high-density stockpiling prior to landfill burial.

Combined, just these four locations account for nearly 60 acres of land consumed by unrecyclable blade waste sitting above ground, a footprint that will inevitably grow as the larger fleets installed in recent decades hit retirement.

Environmental Liability and Financial Fallout

The problem extends beyond spatial consumption. When turbine blades undergo mechanical cutting or experience structural degradation, they generate fine fiberglass particles and composite dust. Rainwater falling on these open-air graveyards washes this particulate debris directly into adjacent drainage ditches, agricultural swales, and local waterways. The physical hazards of this composite fragmentation were explicitly demonstrated during the recent Vineyard Wind structural failure, where thousands of sharp, needle-like fiberglass shards washed ashore along New England coastlines.

Furthermore, federal regulations have created a massive safety vacuum. Because cured blades are chemically stable, the EPA does not classify them as hazardous waste under the Resource Conservation and Recovery Act. With no federal laws mandating recycling or financial recovery, end-of-life decisions are left entirely to market forces that favor dumping.

Subscribe

This creates an intense financial risk for local communities. Wind developers frequently secure long-term leases with private landowners using shell companies or Special Purpose Vehicles. If a wind farm operator faces bankruptcy or corporate restructuring at the end of a project's life, the primary decommissioning liability can fall directly onto the private landowner or local taxpayers. Without heavy industrial cranes, specialized transport vessels, and millions of dollars, a private citizen or small town cannot safely dismantle and remove these towering structures.

The current model of wind energy is fundamentally unsustainable. If we cannot manage the physical waste we generate today, we are simply creating permanent, plastic monuments to our rush for clean power, and the mess is only going to get worse.

Sources

Association of State and Territorial Solid Waste Management Officials (ASTSWMO) Fact Sheet on Wind Turbine Blades: https://astswmo.org/files/Resources/Hazardous_Waste/2022-11-Wind-Turbine-Blades-Fact-Sheet.pdf

National Renewable Energy Laboratory (NREL) Clean Energy Materials and Waste Strategy: https://www.nrel.gov

U.S. Department of Energy (DOE) Wind Energy Technologies Office: https://www.energy.gov/eere/wind/wind-energy-technologies-office

But if you peel back the layers of that turbine, down past the white paint, into the fiberglass blades, inside the gearbox, and deep into the concrete foundation buried in the earth, you find reality.

A wind turbine is not a replacement for oil and gas. It is a product of them.

Let’s look at the hammer-hard facts of what it actually takes to build a “zero-emission” machine from the ground up, and why the wind industry is entirely dependent on the fossil fuel industry for its very existence.

Part 1: The Blades Are Literally Made of Petroleum

When you think of a wind turbine blade, don't think of metal or wood. Think of advanced plastics. A modern utility-scale turbine blade is a massive, hollow shell of high-tech chemistry, and every ounce of that chemistry begins at a petrochemical refinery.

• The blade gets its strength from layers of fiberglass or carbon fiber, but those fibers are completely useless without a glue to hold them rigid. Manufacturers use thousands of pounds of epoxy resins to bind the blades together. Epoxy is synthesized from chemical feedstocks like bisphenol-A, which are cracked directly from crude oil.

• To keep the blades lightweight enough to spin in a light breeze, the insides are stuffed with structural foam cores. These are made from Polyvinyl Chloride (PVC) foam or Polyethylene Terephthalate (PET) foam, which are 100% synthetic plastics birthed from natural gas and petroleum processing.

• The tips of these blades spin at speeds exceeding 150 miles per hour, meaning raindrops, hail, and dust hit the blade like sandpaper. To keep the fiberglass from tearing apart, the blades are coated in thick polyurethane weatherproofing, which is a crude oil derivative.

If you removed every petroleum-based molecule from a wind turbine blade, you would be left with a pile of loose, floppy glass fibers.

Subscribe

Part 2: The Drivetrain Needs Oil to Move

Even if the wind is free, the mechanical parts turning that wind into electricity require heavy-duty fossil fuels to avoid grinding themselves to pieces and failing completely.

Inside the bus-sized box at the top of the tower sits a massive gearbox and a series of heavy bearings. A standard onshore turbine requires 53 to 211 gallons of high-performance gear oil. The massive offshore turbines operating in the ocean can require up to 2,113 gallons of specialized synthetic oil.

These are not standard lubricants. They are highly engineered synthetic polyalphaolefins (PAOs) or mineral oils. They utilize highly refined petroleum base stocks mixed with chemical additives derived from fossil fuel processing. Furthermore, the hydraulic systems that twist the blades to catch the wind run on petroleum-based hydraulic fluids kept under immense pressure. Without these petroleum products, the internal components would seize up and fail in a matter of hours.

Subscribe

Part 3: The Industrial Heat of Metals and Concrete

To understand the full footprint, we have to look back before the turbine is even assembled, all the way to the raw manufacturing of steel and concrete. This is where natural gas and coal do the heavy lifting that electricity simply cannot do.

• A single wind turbine requires hundreds of tons of structural steel. Smelting iron ore into steel requires temperatures exceeding 1,500°C (2,700°F). To reach those temperatures and chemically strip the oxygen out of iron ore, the global supply chain relies entirely on coking coal and natural gas.

• To keep a 300-ton tower from blowing over, it must be anchored into a massive underground block of concrete. Making cement, the glue in concrete, requires baking limestone in industrial kilns at 1,400°C. These kilns are overwhelmingly fired by coal, petroleum coke, or natural gas because the electric grid cannot deliver that level of massive, volumetric heat.

• Concrete alone has no tensile strength; it will snap under the intense leverage of the wind without a dense skeleton of steel rebar inside it. A single utility-scale turbine foundation requires anywhere from 50 to over 100 metric tons of steel rebar. Just like the tower itself, every pound of this rebar is born from high-heat blast furnaces fueled by coking coal and natural gas.

Part 4: The Heavy Diesel Muscle of Construction

A wind turbine cannot walk itself to a ridge line or plant itself in the ocean. The entire process of moving, excavating, and erecting these massive structures is a high-octane logistics operation powered entirely by heavy petroleum distillates.

• Transporting a single turbine requires up to a dozen specialized, multi-axle semi-trucks to haul the massive tower segments, the nacelle, and the 200-foot blades from manufacturing plants to the job site. These fleets burn thousands of gallons of diesel fuel just to get the components on-site.

• Preparing the ground requires a fleet of heavy earthmovers, bulldozers, and excavators to clear land and dig out the massive foundation pits. These machines run exclusively on diesel fuel.

• Erecting the turbine requires some of the largest cranes in the world. A main lift crane capable of hoisting a 100-ton nacelle 400 feet into the air requires its own separate convoy of transport trucks just to assemble it, and burns through hundreds of gallons of diesel fuel per day while operating.

The Bottom Line

Acknowledging this reality changes the entire conversation around the energy transition. Renewable energy infrastructure is not separate from the fossil fuel industry; it is entirely enabled by it.

The fundamental truth is that oil, gas, and coal provide the foundational materials, the mechanical survival, the transport logistics, and the high-density industrial heat required to build modern wind power. Every wind turbine spinning in America today is, from its very conception, anchored firmly to the global oil and gas supply chain. You cannot build the “green” grid of tomorrow without using the fossil fuels of today.

Sources

Coronado, D., & Wenske, J. (2018). Monitoring the oil of wind-turbine gearboxes: Main degradation indicators and detection methods. Machines, 6(2), 25.

González-Reyes, G. A., Bayo-Besteiro, S., Vich Llobet, J., & Añel, J. A. (2020). Environmental and economic constraints on the use of lubricant oils for wind and hydropower generation: The case of NATURGY. Sustainability, 12(10), 4242.

Mishnaevsky, L., Branner, Kim., Petersen, H., Beauson, J., McGugan, M., & Sørensen, B. (2017). Materials for wind turbine blades: An overview. Materials, 10(11), 1285.

The energy electricity sector has been demonized in America. It is time to look at the facts so you can make your own decision.

A useful metric gets little attention. It compares annual carbon dioxide emissions against electricity production, population, and gross domestic product.

The numbers tell a clear story.

The United States has grown dramatically across every major measure since 1970 and carbon dioxide emissions have barely moved.

The table below displays official data for every five years since 1970. That was the year the Environmental Protection Agency (EPA) was formed. Population figures come from the U.S. Census Bureau. Electricity net generation is the total annual domestic output from the U.S. Energy Information Administration, measured in terawatt-hours. Carbon dioxide emissions are energy-related totals from the same agency, in million metric tons. Gross domestic product is nominal figures from the Bureau of Economic Analysis, in billions of current dollars. The 2025 entries use the latest available estimates.

From 1970 to today:

• Population grew +65%.

• Gross domestic product grew +2,500%.

• Electricity generation grew +175%.

All this growth and the energy-related carbon dioxide emissions grew only 6%. That is not a typo. The same sector blamed for environmental destruction powered the largest economic expansion in American history and increased its total emissions by roughly one-sixteenth of what the economy grew.

Emissions did not rise in a straight line. They peaked near 6,000 million metric tons around 2005 and have fallen since. Current levels are close to where they started in 1970. That decline happened while the economy and population continued to expand.

The Environmental Protection Agency opened on December 2, 1970. That same year, Congress passed the Clean Air Act. It set national air quality standards. It also required new power plants to meet technology-based limits on sulfur dioxide, nitrogen oxides, and particulate matter.

The 1977 amendments strengthened rules for new and modified facilities. They required the best available control technology. Plants responded by installing electrostatic precipitators to capture particulates. They added early flue-gas desulfurization systems to remove sulfur dioxide. Low-nitrogen-oxide burners followed.

Subscribe

The 1990 Clean Air Act amendments created the Acid Rain Program. This cap-and-trade system imposed declining national limits on sulfur dioxide and nitrogen oxide emissions from coal-fired power plants. Utilities installed flue-gas desulfurization scrubbers on most coal units. Selective catalytic reduction systems for nitrogen oxides came next. These controls became standard across the fleet by the mid-2000s. They sharply cut those pollutants.

The drop in carbon dioxide emissions after 2005 is notable. It happened while electricity output and economic activity kept rising. The main contributor was the spread of natural-gas combined-cycle plants. These units convert fuel to electricity more efficiently than older coal plants. They produce roughly half the carbon dioxide per kilowatt-hour generated. The broader shift toward natural gas in the generation mix reinforced this effect.

https://substack.com/@nathanhammer/note/p-197295243?r=4ojm62&utm_medium=ios&utm_source=notes-share-action

The Supreme Court ruled in 2007 that the Environmental Protection Agency could regulate greenhouse gases under the Clean Air Act. The agency issued its first carbon dioxide performance standards for new power plants in 2015. Standards for existing plants followed but were later revised. Carbon capture and storage technology has been recognized as available for further control. Large-scale deployment remained limited through the period shown.

The record is straightforward. Population up 65%. GDP up 2,500%. Electricity up 175%. CO2 up 6%. The electricity sector powered a growing nation and nearly held emissions flat across more than five decades. These are the measurable facts.​​​​​​​​​​​​​​​​

I put together a wireframe rendering of a tank battery specifically to walk through this. Every pipe, valve, containment wall, and alarm system you see in that image exists for a reason. Most of it is required by law.

All of it matters.

So let’s go layer by layer and close the gap between what people imagine and what’s actually operating on these sites.

What You’re Actually Looking At

A tank battery is a cluster of aboveground storage tanks where crude oil is held temporarily after production. They sit at well sites, gathering facilities, and terminals across the country. They are not unregulated. They are not unmonitored. And they are not left to chance.

Here’s what’s required to operate one responsibly and what’s actually there.

Subscribe

Layer 1: The Permits (Before a Drop of Oil Goes In)

Before this facility ever receives product, the operator has already navigated a stack of regulatory requirements:

A Spill Prevention, Control, and Countermeasure (SPCC) plan is federally mandated for any aboveground oil storage above applicable thresholds. It outlines exactly how spills are prevented, controlled, and responded to, in writing, stamped by a licensed engineer.

On top of that, air quality permits under the Clean Air Act govern the vapor control systems. State-issued oil operations permits cover tank inventory, drainage controls, and site compliance. And stormwater management permits ensure that any runoff interacting with the facility is captured and handled properly.

That’s four separate regulatory frameworks, minimum, before operations begin.

Layer 2: The Vapor Controls (The Pipes Aren’t Just Pipes)

See that overhead pipe manifold on top of the tanks? That’s part of a closed vent system and it’s one of the most misunderstood pieces of equipment on a tank battery.

Tanks breathe. As product moves in and out, and as temperatures change throughout the day, vapors are generated. Left uncontrolled, those vapors, primarily volatile organic compounds (VOCs), would vent to atmosphere.

The closed vent system doesn’t let that happen. It routes those vapors to a vapor recovery unit or enclosed combustor, where they’re either recovered as usable product or destroyed. The control efficiency on these systems runs 95 to 98 percent or higher, meaning nearly all of the potential atmospheric emissions are captured or eliminated. That’s not a best-case number. That’s what EPA standards require.

Subscribe

Layer 3: Secondary Containment (The Wall Around Everything)

The lined structure surrounding the tanks, what’s called the dike or berm, is not decorative. It is an engineered containment system, sized by a qualified professional to hold the entire volume of the largest single tank in the battery.

The math behind it accounts for the displacement of tanks, piping, and equipment inside the dike. It also incorporates freeboard for a worst-case rain event, specifically the volume of a 25-year, 24-hour storm for that geographic location. Engineered to hold a quarter century rainfall event at a minimum.

The practical meaning: if the largest tank released its entire contents and it was simultaneously raining harder than almost any storm that location has seen in a quarter century, everything stays inside the containment. It doesn’t reach the soil. It doesn’t reach groundwater. It stays on site in containment until it is cleaned up.

Layer 4: Transfer Controls (The Drip Pans Nobody Talks About)

Loading and unloading a tank battery is where incidental spills most commonly occur at hose connections, couplings, and transfer points. The industry knows this, which is why containment drip pans are required beneath connection points during all transfers.

Any drips or minor releases are captured immediately and directed into the secondary containment system. It’s a simple control, and it works.

Layer 5: Mechanical Integrity (Inspections Before Failure, Not After)

Tanks don’t last forever, and regulators know it. That’s why aboveground storage tanks are subject to periodic inspection under API 653, the industry standard for in-service tank assessment.

These inspections use visual examination, ultrasonic testing, and other nondestructive methods to measure wall thickness, corrosion rates, and weld integrity. The goal is to find problems before they become failures. Issues identified early get repaired. Tanks that don’t pass don’t stay in service.

Basically an X-ray for the tanks.

Subscribe

Layer 6: Overfill Prevention (Multiple Alarms, Not One)

Federal regulations require high-level alarms integrated into the overfill prevention system. Level indicators continuously monitor tank volumes. When levels approach or reach a predetermined threshold, the system activates audible and visual alarms and sends operator notifications via text alerts or control room displays.

This isn’t a single alarm with no backup. It’s a system designed to give operators time to respond before anything becomes a problem.

Layer 7: Real-Time Monitoring (SCADA, Always Watching)

Beyond the alarms, modern tank batteries run automated sensors and SCADA systems (Supervisory Control and Data Acquisition) that provide continuous, real-time data on tank volumes, flow rates, and operational conditions.

Any sudden change that could indicate a leak or anomaly shows up immediately. Where automated control valves are installed, operators can make remote adjustments without boots on the ground. This is 24/7 surveillance of a facility that most people drive past without a second thought.

Layer 8: Physical Security (It’s Locked Down)

The perimeter has a chain link fence topped with barbed wire, with lighting for visibility and a locked gate controlling site access. During non-operational periods, loading and unloading pipes are secured with locks.

This keeps out vandalism, theft, and accidental interference, each of which could otherwise affect tank integrity in ways the sensors and alarms would catch, but that nobody wants to deal with in the first place.

Subscribe

The Point

Eight layers of protection. Four regulatory frameworks. Engineered containment. Real-time monitoring. Mandatory inspections. Emission controls running at 95-plus percent efficiency.

That’s what a compliant tank battery is. Not a rusty barrel sitting in a field. Not a ticking environmental time bomb. A highly regulated, multi-barrier system designed and legally required to prevent, detect, and respond to anything that might go wrong.

The industry isn’t perfect. No industry is. But the gap between public perception and operational reality on something as basic as oil storage is wide enough to drive a tanker truck through.

Now you know what’s actually inside the fence.

Nathan Hammer has 18 years of field experience across 29 states in regulated industrial environments. He writes about compliance, operations, and the real mechanics of the energy industry at nathanhammer.substack.com.

Most people think of oil and gas as fuel. You pump it, burn it, move somewhere. That’s the story that gets told.

Here’s the story that doesn’t: the same industry pulling hydrocarbons out of the ground is also the reason the grocery store shelves aren’t empty - and why the air around power plants is cleaner than it used to be.

Let’s walk through it.

Subscribe

It Starts With Natural Gas

The dominant industrial process for feeding the world begins with methane - the primary component of natural gas. Through a process called steam reforming, methane is broken apart using high heat and a nickel catalyst to produce hydrogen. That hydrogen is then combined with nitrogen pulled from the air and forced together under enormous pressure - up to 300 atmospheres - in the presence of an iron catalyst.

The result is ammonia.

That’s the Haber-Bosch process, developed in the early 1900s, and it is arguably the single most important industrial achievement in human history. Some estimates suggest half the nitrogen atoms in your body passed through this process at some point. Without it, the earth could not feed its current population.

Ammonia is the starting point for nearly everything that follows.

Subscribe

From Ammonia to What’s on Your Plate

Roughly 80 percent of global ammonia production goes directly into fertilizer. But the chain doesn’t stop at raw ammonia. It gets converted into:

Urea - the most widely used nitrogen fertilizer in the world. Ammonia reacts with carbon dioxide (often captured as a byproduct of the same production process) to form urea. That urea gets spread on crop fields globally.

Ammonium nitrate - used in both fertilizers and, in blended form, explosives for mining and construction.

Ammonium phosphate (MAP and DAP) - combines nitrogen with phosphorus for balanced crop nutrition.

Ammonium sulfate - 21 percent nitrogen, 24 percent sulfur. Especially useful in sulfur-deficient soils, which are more common than most people realize.

Every bag of fertilizer you’ve ever seen hauled down a farm road traces back to a natural gas well.

Now Add Sulfur - and That Comes From Oil Too

Here’s the parallel track most people don’t know exists.

When crude oil is refined and sour natural gas is processed, the hydrogen sulfide that gets extracted - which would otherwise be a toxic waste problem - gets run through something called the Claus process. That process converts it into elemental sulfur at 95 to 98 percent recovery rates.

This sulfur isn’t waste. It’s a feedstock.

Subscribe

It gets converted into sulfuric acid, which then treats phosphate rock to produce phosphoric acid, which then gets combined with ammonia (from natural gas, remember) to create the ammonium phosphate fertilizers mentioned above. The two supply chains - natural gas and petroleum refining - converge right there in the fertilizer bag.

Sulfur also goes into slow-release elemental sulfur fertilizers and potassium sulfate, which supports crops that need both potassium and sulfur in balance.

Ammonia Also Cleans the Air at Power Plants

This one surprises most people. The same ammonia synthesized from natural gas is injected directly into the exhaust streams of coal, natural gas, and oil-fired power plants to cut nitrogen oxide emissions - the pollutants that contribute to smog and acid rain.

The technology is called selective catalytic reduction, or SCR. Ammonia gets injected into the hot flue gas upstream of a catalyst bed, where it reacts with nitrogen oxides and converts them into ordinary nitrogen and water vapor. About 80 percent of operating SCR systems at electric utilities run on ammonia rather than any alternative. Commercial systems routinely achieve 70 to 90 percent NOx removal, with well-tuned installations hitting 95 percent or higher.

This is distinct from scrubbers, which target sulfur dioxide using limestone. SCR is specifically a NOx control - and ammonia is the reason it works.

So the next time a power plant is held up as an example of fossil fuel harm, the chemistry running its emission controls is also a product of the fossil fuel supply chain.

One More: The Truck in the Field Runs Cleaner Because of This

Diesel exhaust fluid - DEF, or AdBlue depending on your market - is a 32.5 percent urea solution in deionized water. Every modern diesel truck with a selective catalytic reduction system uses it. When injected into the hot exhaust stream, the urea breaks down into ammonia, which converts nitrogen oxides into harmless nitrogen and water.

That fluid in the blue jug at the fuel island? It starts at a natural gas well too.

The Point

The argument against fossil fuels is almost always framed around combustion - burning fuel, releasing carbon. That’s a fair debate to have.

But the full picture includes ammonia. It includes urea. It includes sulfuric acid, DEF, the fertilizers that support billions of acres of crop production worldwide, and the emission controls running on power plants right now. These aren’t combustion products. They’re chemical feedstocks derived from the same hydrocarbons the energy debate focuses almost entirely on burning.

You can hold a complicated view on energy policy and still recognize what the petrochemical supply chain actually delivers: food security, agricultural productivity, cleaner air, and industrial capacity at a scale nothing else currently replicates.

That’s worth understanding before forming an opinion either way.

Nathan Hammer writes about safety, industry, and the things the field actually knows that the office forgets. Subscribe for free below.

Subscribe

Share

Pull up to a pump, fill your tank, drive away. That’s the mental model most people have for what an oil refinery does. It turns crude oil into fuel. End of story.

That’s about 80 percent of the story. The other 20 percent is quietly running your entire life.

When refineries process crude oil, they don’t just squeeze out gasoline and diesel and throw the rest away. What’s left over, the heavier residues, the chemical streams, the recovered compounds, gets routed into downstream plants that manufacture things you use before you ever get to your car. Things you touched before you left the house this morning.

Here’s what that actually looks like.

Subscribe

The plastic bag your groceries came in. The bottle your shampoo lives in. The food container keeping your leftovers sealed in the fridge. Those come from polyethylene, which starts as naphtha, a lighter stream pulled off during refining, before it gets converted into plastic in a downstream petrochemical plant.

The carpet under your feet. The shirt on your back. The upholstery in your car. Synthetic fibers like polyester and nylon come from aromatic compounds called benzene, toluene, and xylene that are extracted from refinery streams. Cheap, durable, easy-care textiles that most of the world now depends on for clothing and home goods.

The road you drove to work on. Asphalt comes directly from the heaviest residues left after distillation. Every highway, parking lot, and roof in America is largely built from what’s left at the bottom of the barrel after everything useful has been pulled off the top.

The aspirin you took this morning. The syringe the nurse used at your last appointment. The plastic tubing in medical equipment. Benzene derivatives from refinery streams are precursors to everyday medications. The disposable plastics that keep healthcare sanitary come from the same petrochemical chain.

The fertilizer that grew your food. Refineries remove sulfur from crude oil as part of the refining process, a byproduct of cleaning up the fuel. That recovered sulfur gets converted into sulfuric acid and used in fertilizer production. The global food supply is partly sustained by something refineries pull out as a byproduct.

Subscribe

The oil in your engine. Lubricating oils and base stocks come from mid-range distillates. Without them, engines seize, industrial machinery fails, and household appliances wear out faster. Refinery-derived lubricants are why mechanical things last as long as they do.

The full list runs to ten major categories and thousands of individual products: detergents, paints, adhesives, medical supplies, industrial cleaners, aluminum anodes. Almost none of it gets talked about when people talk about oil.

The reason is simple. Fuels are visible. You buy them directly. The other stuff is embedded three or four steps deep in supply chains before it becomes something you recognize.

But that embedded 10 to 20 percent of every barrel is what makes packaging affordable, medicine accessible, food abundant, roads durable, and clothing cheap enough for everyone to own more than one set of clothes.

Refineries aren’t just gas stations at industrial scale. They’re the base layer of the physical economy, the part that almost nobody sees, running almost everything people use.

Subscribe

Sources: U.S. Energy Information Administration refinery production data and established industry process descriptions.

Table of Contents

1. It Starts in the Ground

2. Turning Ore into Metal Takes Enormous Heat

3. Petroleum Is Also a Raw Material, Not Just a Fuel

4. Semiconductor Manufacturing Is Extremely Chemical-Intensive

5. Getting Parts and Phones Around the World

6. The Phone in Your Pocket

7. What This All Means

8. Changing The Narrative

1. It Starts in the Ground

Making an iPhone requires roughly 70 different elements from the periodic table. Copper, aluminum, gold, silver, cobalt, lithium, nickel, tin, tungsten, rare earth elements, silicon. Each of those has to be dug out of the earth somewhere, often on a different continent.

That digging runs on diesel. The excavators, haul trucks, drilling rigs, processing equipment, and support vehicles at every mine site burn fuel constantly. A single large haul truck can hold 1,000 gallons of diesel and burn through hundreds of gallons per shift. When you multiply that across every mine feeding materials into the global electronics supply chain, the fuel consumption is staggering. None of this stops if you remove petroleum from the equation. The mining simply does not happen.

2. Turning Ore into Metal Takes Enormous Heat

Raw ore is not usable. A chunk of rock pulled out of a lithium mine or a copper deposit has to be crushed, chemically treated, smelted, and refined into the pure material that can actually go into a battery or a circuit board. That process requires heat, and a lot of it.

Smelting aluminum requires temperatures above 1,700 degrees Fahrenheit. Copper refining involves acids and high-heat processing. Purifying the silicon used in chips is one of the most precise and energy-demanding operations in modern manufacturing. All of that heat and energy comes largely from fossil fuels, including natural gas, which is a workhorse for high-temperature industrial processes worldwide.

This is where a huge share of the petroleum burden in electronics actually lives, not in the final product, but in the transformation of raw materials into usable ones.

Subscribe

3. Petroleum Is Also a Raw Material, Not Just a Fuel

Here is the part most people miss entirely. Oil and gas are not only burned to power the process. They are also physically converted into the materials that go into the phone and everything around it.

Plastics in the phone’s internal components, cables, and connector housings come from petrochemical feedstocks. The adhesives that hold the screen, battery, and housing together are petroleum-derived. The solvents used to clean components during manufacturing come from petrochemical production. The coatings that protect circuit boards, the insulation on internal wiring, the protective films applied during fabrication, the inks used in packaging, and the binding agents in the box the phone ships in all trace back to oil and gas as their origin.

Apple has reduced visible plastic in its packaging over the years, and that is worth acknowledging. But the supply chain behind the phone has not been detached from petrochemical inputs. They are just less visible to the consumer.

4. Semiconductor Manufacturing Is Extremely Chemical-Intensive

The chips inside an iPhone are built in facilities called fabs. These are some of the most precisely controlled environments ever constructed, and they consume an extraordinary quantity of industrial chemicals.

Making a modern chip requires photoresist chemicals that react to light to etch circuit patterns at nanometer scale. It requires acids to strip and clean layers. It requires specialty gases to deposit materials atom by atom. It requires solvents to remove residues. The wafers move through hundreds of individual process steps, and many of those steps depend on chemicals that originate in the petrochemical supply chain. A single leading-edge chip fab can consume millions of gallons of ultrapure water per day — totaling hundreds of millions to billions of gallons per year — along with comparable volumes of chemical inputs.

This is not a small part of what makes your phone work. The chip is the phone, in many ways. Without semiconductor-grade chemistry, there is no A-series processor, no neural engine, no camera system.

5. Getting Parts and Phones Around the World

An iPhone is assembled in China, but its components come from Japan, South Korea, Taiwan, the United States, Germany, and dozens of other countries. Before any of those parts reach the assembly line, they travel by ship, truck, rail, and air. After assembly, the finished phone travels again.

That global movement runs almost entirely on petroleum. Marine shipping, which carries the bulk of international trade by volume, runs on bunker fuel. Air freight, which carries high-value electronics, runs on jet fuel. Trucking runs on diesel. There is no meaningful alternative infrastructure in place for any of these modes at scale.

The supply chain for a single iPhone touches thousands of miles of ocean freight and a web of ground transport before it ever reaches a retail shelf or a UPS driver’s truck.

6. The Phone in Your Pocket

Once you have the phone, it needs electricity to charge and operate. In the United States, roughly 40 percent of electricity generation comes from natural gas. In many other parts of the world where iPhones are used, coal and gas make up an even larger share of the grid. The use phase is smaller than manufacturing in overall impact, but it is still powered significantly by fossil fuels in most places where people actually live.

Subscribe

7. What This All Means

The iPhone is a remarkable object. It really does contain the work of thousands of engineers, scientists, and skilled tradespeople across dozens of countries. But it is also, in a very real sense, a product of the petroleum economy.

Oil and gas show up as diesel fuel in the mines. They show up as the heat source for smelters and refineries. They show up as the raw material for plastics, adhesives, coatings, solvents, and packaging. They show up as the chemical inputs for chip and battery manufacturing. They show up as the fuel that moves parts and finished devices across oceans and continents.

The next time someone says that oil and gas are old industries with no connection to the modern digital world, hand them your phone. Then ask them to explain exactly how it got there.

8. Changing the Narrative

The oil and gas industry has received a bad rap in recent years. It is time to look at the truth. The dedicated work of this sector powers every step of the supply chain that brings an iPhone — and countless other modern technologies — into our hands. From mining to manufacturing, transportation to energy, petroleum and natural gas make our digital lives possible. Acknowledging this reality does not diminish the need for responsible practices. It simply recognizes the foundational role these industries play in enabling the world we live in today.

Nathan Hammer is a marketing consultant and personal brand strategist specializing in HSE, process safety, and pipeline safety. He has 18 years of industrial experience and is co-founder of Trunkline.com.

The event took place in the Amine Regeneration Unit during scheduled maintenance. Two Repcon, Inc., boilermakers were removing isolation blinds from ARU6 piping that had been depressurized and prepared for work as part of a broader turnaround. The adjacent Sulfur Recovery Unit (SRU) was also in turnaround status.

The workers intended to remove Blind 407 on the prepared ARU6 line. Instead, they opened a flange on the nearby ARU7 acid-gas line, located about five feet away. The two piping segments were visually identical and lacked clear identification at the work location. An identification tag for Blind 407 had been placed on a railing above the correct flange, outside the workers’ normal line of sight. The boilermakers relied on a flange-locking device present on the ARU7 line, which resembled devices used throughout the de-inventoried SRU.

The ARU7 line contained approximately 27,000 pounds of pressurized acid gas, roughly 90 percent hydrogen sulfide. The workers were wearing supplied-air respirators consistent with the permit requirements but did not carry personal hydrogen sulfide monitors during this task. When the flange was opened, the release produced concentrations well above immediately dangerous levels.

Jose Perez, Jr., one of the Repcon boilermakers, was fatally injured near the release point. The second Repcon worker activated his escape respirator and reached safety. A light breeze of approximately 2 miles per hour carried the vapor cloud into the adjacent SRU. There, Jason Scott, an ISC Constructors employee working in the turnaround area without respiratory protection, was fatally exposed.

The release continued for nearly one hour until emergency responders isolated the line by reinstalling the flange.

The work permit issued that day covered multiple blind removals with differing hazards. It included a note stating “Ops present for each break,” but this requirement was not enforced as a clear hold point. The boilermakers and foreman understood verbal permission allowed them to proceed without an operator present for the final blinds. No dedicated SIMOPs review addressed the specific hazard of opening piping in the operating Amine Unit upwind of the active turnaround zone.

The U.S. Chemical Safety and Hazard Investigation Board determined that the incident resulted from the failure to positively identify the correct equipment. Contributing factors included inadequate labeling practices, a broad-scope work permit, gaps in hazard communication between operations and contractors, and the absence of effective controls for simultaneous operations.

These findings invite reflection through the five pillars of moral safety.

Integrity: Did everyone involved, operations, contractor supervision, and the workers themselves, have a true and shared understanding of what was live, what was isolated, and where the boundaries were? Or did each party assume someone else had confirmed the details?

Respect: Did we respect the contractors enough to give them unambiguous information at the point of work? And did the contractors respect the process enough to stop and verify when something did not feel right, even under turnaround pressure?

Communication: Did operations clearly communicate the hazards of working adjacent to live systems? Did contractor supervision ask the right questions before sending their crew in? When the permit said “Ops present for each break,” did anyone on either side treat that as a real requirement?

Stewardship: Operations owns the process and its hazards. Contractors own their crew’s safety in the field. Both failed to ask the same question that day: what happens to everyone around us if we get this wrong?

Compliance: The permit had a hold point. The crew had a foreman. Operations had a presence requirement. None of those layers caught this. Compliance is not a signature on a form. It is the moment someone actually speaks up and/or stops the job.

Subscribe now

These questions remind us that safety systems are only as strong as the human decisions and attention behind them. When breaking lines or removing blinds, confirm the exact equipment through visible, location-specific labels, operator walkthroughs where required, and clear assessment of nearby hazards. A simple five-foot difference in piping can have significant consequences if identification is not reliable.

Before any line break or flange opening, verify the equipment identity and consider who else could be affected by an error. Clear labeling at eye level, unambiguous permits, and active communication between teams remain essential safeguards.

If you have seen effective equipment identification practices or strong line-breaking procedures that reinforced these principles, feel free to share general lessons or approaches in the comments. These discussions help strengthen safety performance across our industry.

A big thanks to the U.S. Chemical Safety and Hazard Investigation Board for their dedication to conducting thorough investigations and for providing the detailed reports that help our industry learn from these events and continuously improve process safety.

Walk into a modern data center and you’ll find rows of servers stretching as far as you can see, all humming with activity. The heat they produce is immense, and keeping them cool requires industrial-scale solutions. The most common method used by the majority of large data centers today is something called evaporative cooling. Think of it as a giant version of those swamp coolers people use in dry climates. Here’s how it works: hot outdoor air gets pulled through wet pads. As the water on those pads evaporates, it cools the air down, which is then blown across the servers. It’s remarkably energy-efficient compared to traditional air conditioning, which is why it’s so popular. But there’s a catch. This method is constantly drinking water and breathing it out as vapor.

Some data centers use air-cooled chillers instead, which work more like the AC unit in your home, using refrigerant to chill the air. These systems often include cooling towers that also rely on water to release heat outside, though they’re sometimes chosen for locations where water is harder to come by. Then there’s the cutting edge: liquid cooling. As computers become more powerful to handle AI processing, air alone can’t keep up. Instead, special fluids are pumped directly to computer chips or servers are actually submerged in cooling liquid. The good news? These systems are often closed-loop, meaning the same liquid gets recycled over and over, using far less water than the evaporative methods.

Subscribe

But here’s the part that doesn’t show up in the glossy tech company brochures: most of the water used to cool data centers simply disappears. When a data center uses evaporative cooling or cooling towers, the water doesn’t just circulate and return to where it came from. Instead, somewhere between 70% and 80% of it literally evaporates into thin air. It transforms into water vapor that drifts away as steam. You can sometimes see it: that cloud rising from a data center facility. That’s water leaving the local water system forever. It doesn’t flow back into rivers or reservoirs. It doesn’t return to the municipal water supply. It just vanishes into the atmosphere. The remaining 20-30% doesn’t fare much better. It becomes what’s called “blowdown.” This is water so concentrated with minerals and salts from the cooling process that it can’t be used anymore. This gets flushed into the sewer system as wastewater.

Stay up to date with the industrial world - subscribe for free.

Subscribe

Now, you might be thinking: surely these industrial facilities use industrial water, right? Some recycled or lower-grade water that’s not meant for human consumption? Not quite. The surprising truth is that data centers often use the exact same drinking water that comes out of your kitchen faucet. Why? Because treated municipal water is clean and pure, which means it won’t clog up the sensitive cooling equipment with sediment or algae. Recent industry estimates suggest that 75% to 85% of data centers primarily rely on municipal potable water. Google, for instance, reported that about 78% of its data center water withdrawals in 2024 were potable water. Think about that for a moment: the same water you brush your teeth with is being evaporated by the thousands of gallons to keep servers cool so you can watch cat videos and check your email.

To be fair, the tech industry is beginning to wake up to this issue. Some companies are experimenting with recycled wastewater. This is the kind that flows through purple pipes in water-conscious communities. Others are investing in those closed-loop liquid cooling systems that dramatically cut water consumption. But the reality is that most data centers operating today still follow the old playbook: pull drinking water from the public supply and send most of it skyward as steam. As our appetite for cloud storage, streaming, and AI continues to grow, so does the invisible cost: the water that keeps the internet cool, one evaporated gallon at a time.​​​​​​​​​​​​​​​​

As of this week, there are over 18 new major projects planned or under construction. But what are they doing? And more importantly, how will they affect your water, your electric bill, and your neighborhood?

Here is the plain English guide to what is happening in our backyard.

What Are They Building?

Think of a data center as a "hotel for the internet." Inside these buildings are thousands of powerful computers stacked like bookshelves.

• Hyperscale Centers: These are the giants (Google, Meta/Facebook). They cover hundreds of acres and use massive amounts of power to run Artificial Intelligence (AI) and store your photos, emails, and documents.

• Colocation Centers: These are smaller facilities, like a self-storage unit for companies to rent space for their servers.

Why Oklahoma?

Tech companies love us for two reasons:

1. Cheap Power: We have affordable natural gas and wind energy.

2. Space: We have the land they need to build massive campuses.

The "Big Three" Concerns

While these projects bring construction jobs and tax revenue for schools, many locals are asking tough questions.

1. Will my electric bill go up?

This is the number one battle statewide. These facilities use enormous amounts of electricity. A single campus can use as much power as 50,000 homes. Utility companies (like OG&E and PSO) need to build new power lines and substations to support them.

• The Conflict: Consumer groups like AARP are fighting to make sure you do not pay for those upgrades. There is fear that residential rates could rise 10 to 15% to cover the costs.

• New Law Alert: Just days ago, Rep. Brad Boles filed House Bill 2992. If passed, this law would legally force data centers to pay for their own infrastructure upgrades so the cost does not fall on families.

2. What about our water?

Computers get hot. To keep them from melting, data centers use cooling systems that can evaporate millions of gallons of water a day.

• The Fix: Most new projects (like Google in Stillwater and Project Clydesdale in Owasso) are switching to "reclaimed water." This is treated wastewater (sewage that has been cleaned but is not for drinking) or river water, so they are not draining the municipal drinking supply.

• The Worry: In places like Coweta, residents are worried about drought years. If the Arkansas River gets low, they want to know if the data center gets priority over farmers.

3. The Noise ("The Hum")

In Luther and Sand Springs, residents are pushing back against the noise. Data centers use thousands of cooling fans. Up close, it can sound like a continuous leaf blower or a low drone. Residents are demanding rules that force the buildings to be far away from homes to dampen the sound.

Project Tracker: What Is Happening Near You?

Currently Under Construction (The "Done Deals")

• Google (Pryor & Stillwater): Google is the veteran here. They are expanding in Pryor and have started building a massive $9 billion campus in Stillwater.

• Owasso (Project Clydesdale): Located north of town, this massive project broke ground late last year. Update: Neighbors recently received letters about "pre-blast inspections," signaling that heavy construction (and potentially blasting) is about to start.

• Tulsa (Project Anthem): A huge site in East Tulsa linked to a major tech firm (likely Meta/Facebook). Construction has begun near the Fair Oaks Industrial Park.

Planned but Contested (The "Hot Spots")

• Luther (Beltline Energy): A proposed site near the Redbud power plant. Status: Heavy opposition. A local group called "Citizens for Responsible Development" is fighting the rezoning, citing noise and property value concerns.

• Coweta (Project Atlas): A cloud computing center planning to draw water from the river. Status: Zoning hearings are delayed. Residents have packed city hall to demand answers on water usage.

• Yukon (Gamma Resources): A proposed site near I-40. Status: Currently stalled. Zoning hearings have been deferred twice due to local outcry.

• Claremore (Project Mustang): A planned site in the Industrial Park. Status: In early talks. City leaders promise it will not raise utility rates, but residents are asking for more transparency regarding the "mystery" tenant.

Canceled

• Lawton (Boomtown): If you heard about a project here, it is gone. The lease was canceled in December 2025 after state delays.

The Bottom Line

Oklahoma is pivoting to become an AI capital. For some, this means better funding for schools and new tech jobs. For others, it means construction traffic and worries about resources.

What you can do: If a project is proposed in your town, look for the "Zoning Hearing" dates. That is your only real chance to speak up about where these buildings go and what rules they must follow.

The operator met the driver at the unloading area, a spot where five different fill lines clustered together. These lines were the gateways to various bulk storage tanks containing chemicals like caustic, acetic anhydride, and sodium hypochlorite. The operator reached for the sulfuric acid line, unlocked the padlock, and placed it on a piece of angle iron above the pipe. He later noted that he pointed toward the correct line to guide the driver. Satisfied that the “safe” line was now the only one accessible, the operator turned and walked back toward the Mod B building to complete his end-of-shift paperwork.

Left alone at the back of his trailer, the driver began his setup. He looked at the cluster of pipes. He saw a fill line with its dust cap unlocked... a clear signal in his mind that it was the intended destination. What he couldn’t see was that eighteen inches away, the sodium hypochlorite line was also unsecured. A small split ring on that pipe’s locking mechanism had corroded and snapped some time ago, making it impossible to secure the padlock. To the driver, the unlocked cap wasn’t a maintenance failure though... it was an invitation.

He pulled the cap, connected his heavy four-inch hose, and moved to the valves. Under the facility’s established procedures, the operator was required to remain present and be the one to physically open the fill line valve. But with the operator already back inside the building, the driver, wishing to be helpful and move the process along, took it upon himself to open the valves and start the transfer.

The Turning Point and the Toll

Within minutes, the routine was gone. Inside the control room, the operators didn’t hear an alarm at first; they smelled one. A sharp, suffocating scent filled the air. Outside, the reaction was visible. As the sulfuric acid pumped into the sodium hypochlorite tank, it triggered a massive exothermic reaction. A thick, greenish-yellow cloud of chlorine gas began to vent violently from the top of the 6,500-gallon bulk tank.

The driver, sitting in his cab, saw the world behind him disappear into a toxic fog. He attempted to reach the valves to shut down the flow, but the gas was too thick, stinging his eyes and lungs. He was forced to abandon his rig, fleeing for his life as the cloud began to drift over the plant fence line and toward the town of Atchison.

The consequences rippled far beyond the plant gates. The city of 11,000 people entered a state of emergency. Schools were evacuated, and thousands of residents were told to seal their windows and doors. By the time the cloud dissipated, over 140 people, including plant staff, the driver, and local citizens, required medical treatment. Six people were hospitalized with significant respiratory injuries. A simple connection to the wrong pipe had transformed a quiet morning into a public health crisis.

Reflection: The Five Pillars of a Safe Operation

When we analyze an event like the Atchison release, it’s easy to focus on the chemistry. But the real failure occurred long before the first drop of acid hit the wrong tank. It occurred in the quiet space between our intentions and our actions. This event provides a clear look at how our moral obligations... the pillars of our profession... uphold or undermine our safety.

Integrity and Stewardship are the foundations of maintenance. The broken split ring on the sodium hypochlorite line was a minor mechanical issue, but it created a major “error trap.” Stewardship means caring for the small things... labels, locks, and rings... because we know someone else is relying on them to stay safe. When a pipe labeled “Hydrochloric Acid” actually contains “Sulfuric Acid,” as was the case here, we have failed the test of integrity. We are providing false information to a colleague who is trusting us with their life.

Communication and Respect are found in the handoff. When the operator walked away before the hose was connected, he ended the conversation prematurely. Safety is a dialogue, not a monologue. By not staying to verify the connection, the operator missed the chance to show respect for the driver’s safety. Respect in our industry means never leaving a teammate alone in a high-hazard situation, regardless of how “routine” the task feels.

Finally, there is Compliance, but not as a matter of following rules for the sake of a clipboard. The requirement for the operator to open the valve was a critical safeguard. This rule existed because two sets of eyes are better than one. When we bypass a procedure to save five minutes on paperwork, we aren’t just being efficient; we are discarding the hard-earned lessons of those who came before us. Compliance is, at its heart, a way of honoring those lessons to ensure no one else has to learn them the hard way.

Actionable Insight

Procedural “shortcuts” often feel like efficiency until the moment they turn into a catastrophe. The requirement to have an operator open the valve was designed to prevent exactly what happened in Atchison.

A question for your next shift: Is there a task you perform where you’ve started “helping” by doing a teammate’s step, or where you’ve walked away before the “handshake” was truly finished?

We can’t change the past, but we can change the standards we hold ourselves to today. This is why Trunkline provides every vendor on our platform with our 5 Pillars for Moral-Based Safety. We believe that when the front lines lead from the bottom up, we can build a future where every worker comes home.

This account is based on the findings of the U.S. Chemical Safety and Hazard Investigation Board (CSB) final report. For the full report and details visit https://www.csb.gov. This is not legal advice and shall not be construed as such.