Project Mustang’s Air-Pollution Question: What Could a 1-Gigawatt Gas Campus Emit?

TAYLOR, Texas — Project Mustang’s proposed power system is large enough that its air emissions cannot be understood by looking only at the data-center buildings.

The current concept discussed before Taylor officials includes approximately 1,040 megawatts of on-site generation, potentially using 84 to 112 natural-gas generating units, along with diesel backup generators and black-start equipment.

That puts a second question behind the one examined in Part 1 of this series: once the generators and nearby homes are placed on the same map, what would all of that equipment actually put into the air?

There is an important line to draw at the outset. Natural-gas turbines, reciprocating engines and diesel generators are known sources of air pollutants. Those pollutants have been studied for decades, and many have well-established health effects at sufficient concentrations.

What WLP cannot yet determine is whether Project Mustang would expose any nearby resident to concentrations high enough to create a measurable health risk. That requires project-specific emissions data and air-dispersion modeling that have not yet been made public.

The hazard is known. The exposure is not.

The first pollutant to watch: nitrogen oxides

Combustion at high temperatures produces nitrogen oxides, or NOx. The group includes nitric oxide and nitrogen dioxide. Nitrogen dioxide, or NO₂, is the component most directly associated with respiratory effects.

EPA says elevated NO₂ can irritate airways and aggravate respiratory disease, particularly asthma. That does not mean every gas-fired generator creates an unsafe concentration beyond its property line. Equipment design, pollution controls, load, exhaust conditions, stack height, weather and distance all affect what reaches a particular receptor.

But for a project contemplating dozens of gas-fired units, NOx should be one of the first pollutants quantified.

It matters for another reason. NOx can react with volatile organic compounds in sunlight and contribute to ground-level ozone. The analysis therefore cannot stop with what leaves the stack. It also has to consider what those emissions can help form after they enter the atmosphere.

Fine particles are more complicated than visible smoke

Another pollutant that deserves attention is PM2.5 — fine particulate matter small enough to penetrate deep into the lungs.

EPA’s scientific reviews associate PM2.5 exposure with cardiovascular and respiratory effects. Natural-gas combustion generally produces much less direct particulate matter than coal, but that does not make particulate analysis irrelevant.

Diesel generators can emit particles directly. NOx can also contribute to secondary particulate matter that forms in the atmosphere after emissions leave the source.

That distinction matters because a modern gas-fired facility does not have to produce a visible plume of smoke to affect air-quality calculations.

Carbon monoxide and air toxics

Natural-gas engines and turbines can also emit carbon monoxide, or CO, which is produced when fuel does not burn completely. The amount emitted can vary significantly with the technology, operating load and pollution controls.

The less familiar part of the discussion involves hazardous air pollutants, commonly called air toxics.

EPA identifies stationary combustion turbines as sources of hazardous air pollutants including formaldehyde, benzene, acetaldehyde and toluene. Stationary reciprocating engines can also emit compounds including formaldehyde, acetaldehyde, acrolein, methanol and benzene.

Those compounds do not all behave the same way. Benzene is a known human carcinogen. Formaldehyde and acrolein can irritate the respiratory system at sufficient exposure.

Those toxicological facts are important, but they still do not answer the Project Mustang question. The relevant issue is how much Mustang would emit after controls are applied, how those emissions would disperse, and what concentrations would be predicted at surrounding homes.

Diesel generators need their own accounting

Project Mustang representatives have disclosed diesel backup generation and black-start capability. Those sources should not disappear into a footnote simply because they may run only during testing, emergencies or specific operating events.

The public still does not know the complete diesel inventory. A meaningful review would need the number of generators, their size and manufacturer, testing schedules, whether multiple units may run simultaneously, and the assumptions used for extended emergency operation.

A diesel unit operating a few hours for testing is very different from a gas engine operating continuously. But on a campus of this scale, even standby equipment deserves a source-by-source accounting.

The turbine-versus-engine question matters

One of the biggest remaining uncertainties is what the main natural-gas system will actually use.

Project Mustang discussions have referenced both gas turbines and reciprocating engines. That distinction is not cosmetic. Different machines have different combustion characteristics, exhaust temperatures, stack flows and emissions profiles. They can also use different pollution-control systems.

For NOx, controls can include lean-premix combustion or selective catalytic reduction. For carbon monoxide and certain organic pollutants, oxidation catalysts may be used.

The Mustang concept also references urea storage. Urea can be used with selective catalytic reduction systems to control NOx, making that detail potentially significant. But it remains an engineering inference — not confirmation of Mustang’s final emissions-control design.

Until the manufacturer, model, unit count and pollution controls are disclosed, precise emissions estimates would be speculation.

“Tier 4” does not cover the whole power plant

Taylor’s public discussion has repeatedly included the phrase “EPA Tier 4.” Without context, that can sound like a high emissions standard applying across the entire project.

The Aug. 18 hearing clarified an important distinction: Tier 4 applies to emergency or standby diesel equipment, not the main natural-gas generation plant.

That does not make Tier 4 meaningless. It means it should not be used as shorthand for the emissions performance of Mustang’s proposed 1,040-MW gas-generation system.

The main gas equipment will have its own applicable state and federal requirements, depending on the final technology, size, emissions and regulatory classification.

The air permit will tell us far more than a presentation

The most important regulatory record may ultimately be Project Mustang’s TCEQ air-permit application.

A detailed permitting package should identify individual emission sources, equipment types, fuels, operating rates, emission factors, pollution controls, hourly and annual emissions, stack parameters and operating limits. It should also show which state and federal programs apply.

Whether Mustang triggers federal programs such as Prevention of Significant Deterioration or Title V cannot responsibly be determined from the 1,040-MW figure alone. Applicability depends on potential emissions, controls and how the sources are classified.

Those records are unlikely to arrive immediately. But when they do, they should provide the first serious opportunity to compare the project’s public description with source-by-source emissions calculations.

Emissions are not the same as exposure

This is the most important distinction in the series.

Suppose a future emissions inventory lists tons per year of NOx, CO or formaldehyde. Those totals still do not tell a nearby resident what they would breathe.

After pollutants leave a stack, concentrations at ground level depend on source location, stack height, exhaust temperature and velocity, wind, atmospheric stability, terrain, nearby buildings and distance.

That is where AERMOD, EPA’s preferred regulatory dispersion model, becomes important. AERMOD can estimate concentrations at receptor locations around a facility using source characteristics and meteorological data.

For Mustang, the useful receptors should include the actual homes surrounding the project. Depending on the pollutant and applicable standard, the meaningful results could include predicted one-hour, 24-hour and annual concentrations at those residences.

The question Taylor should be asking now

Taylor does not need to decide whether natural-gas combustion can produce air pollution. Decades of emissions testing and regulation have already answered that.

The unanswered question is narrower: What will this particular collection of generators emit, and what concentrations will reach the people living around it?

That is measurable. But it cannot be answered from a PowerPoint presentation, a megawatt figure or a general statement that the project will comply with environmental rules.

Part 1 established why source-to-home location matters. Part 2 establishes why an emissions inventory by itself is still not enough. The two have to be combined.

Until the equipment list, emissions inventory, air-permit record and dispersion modeling are available, claims that Project Mustang either will harm nearby residents or cannot harm them go beyond the project-specific evidence.

Part 3 will examine how those emissions should be modeled at surrounding homes — including stack height, weather, distance, building downwash and what an AERMOD analysis can, and cannot, tell the public.

Sources

• U.S. EPA — Stationary Combustion Turbines / hazardous air pollutants
• U.S. EPA — Stationary Engines
• U.S. EPA — Nitrogen Dioxide
• U.S. EPA — Particulate Matter
• U.S. EPA — Ground-Level Ozone