Power plant Clean Air rules: Reducing your risk from PSD lookbacks

By Robynn Andracsek, PE, Burns & McDonnell

Congress last amended the Clean Air Act over 30 years ago. Since then, our nation has changed drastically: politically, technologically, and industrially. However, the air regulations remain riddled with undefined terms (“routine”), unfinalized guidance documents (the “puzzle book”), and contradictory court rulings (routine for industry/routine for a boiler).

Yet one policy has strengthened over time: the use of netting to reduce risk of Prevention of Significant Deterioration (PSD) lookbacks. A PSD permit is needed for the highest emitting facilities or for modifications at those facilities and can take 12-18 months for issuance. It requires installation of the best available controls, dispersion modeling to show compliance with air quality standards, and a public comment period. In a nutshell, this is the permit that utility plants either need or try to avoid.


When the PSD rules were written in the late 1960s, the electric utility industry argued that its oldest plants should be exempt from new, stringent limits under the logic that in 10-20 years those plants would soon be replaced by new plants built under PSD and subject to Best Available Control Technology (BACT). However, utilities proved adept at nursing along these older plants, avoiding the need to decommission or replace them, and were able to maintain the coal-fired boilers without major PSD- triggering upgrades. Thus, the grandfathered facilities, already considered old in 1970, grew older still. EPA then decided to reinterpret the exemption for “routine maintenance.” No definition exists in the PSD rules for what “routine” means.

40 CFR 52.21(b)(2)(i-iii) Major modification means any physical change in or change in the method of operation of a major stationary source that would result in: a significant emissions increase…of a regulated NSR pollutant…. A physical change or change in the method of operation shall not include… Routine maintenance, repair and replacement.

From 2000 to 2015, EPA pursued a “Coal-Fired Power Plant Enforcement” initiative.[1] EPA accused power plants of performing PSD modifications without receiving a PSD permit (and installing BACT). Utilities asserted that they were merely doing “routine” maintenance. The odds were not in the energy producers’ favor. As shown in Figure 1, in just 15 years, 113 plants paid or installed $21 billion dollars’ worth of fines and control devices due to this enforcement initiative. Many coal-fired boilers shut down or repowered with natural gas.

The Failure of NSR Reform

All parties agree that New Source Review (the parent program of PSD) is complicated. It inhibits technical improvements, economic developments, efficiency increases, pollution control projects, and innovation by using a patchwork of regulations, interpretation memos, and court opinions instead of clearly written rules. In 2002, EPA tried to “reform” NSR with a series of policies that would clarify the rule, streamline plant efficiency improvements, and define “routine” maintenance. After a series of adverse court rulings, very little exists of the “reform” other than a few rule text remnants:

“Routine maintenance, repair and replacement shall include, but not be limited to, any activity(s) that meets the requirements of the equipment replacement provisions…By court order on December 24, 2003, [this] sentence…is stayed indefinitely.

The WEPCO vs. Reilly court case serves as the best guideline of routineness. In 1988, WEPCO sued the EPA for failing to grant a PSD exemption under the RMRR provisions. The court’s decision (known as the “WEPCO rule”) described a test for RMRR using four factors: nature and extent, purpose, frequency, and cost (see Table 1). Under WEPCo analyses, projects such as retubing, overhauling economizers, and replacing superheater outlet pendant elements have been determined to not be routine. To date, the “WEPCo rule” remains the only detailed explanation of what is and is not routine maintenance. The regulation itself has not been changed. There are no definitive lists of possible modifications or direction as to under which classification each would fall. Every change must be analyzed, and all conclusions are open to debate. In short, utilities are left to guess their PSD applicability.

Table 1: The WEPCO Test for RMRR

Nature and Extent* What is the size, function, or importance of replacement parts?
* Does it require pre-approval of a state commission?
* Do internal documents cite the repair as “non-routine”?
* Can it be repaired while operating or during outage?
* Will an entire emissions unit be replaced?
*How much time will the modification take?
Purpose* Will it extend the useful life of the unit?
* Will it keep the unit operating in its present condition?
* Will it allow enhanced operation?
* Will it increase capacity, operating rate, utilization, or fuel adaptability?
Frequency*How often the change is performed in a typical unit’s life?
* How often is the change performed in industry?
Cost*What is the absolute cost and relative cost of replacing the unit?
*What is the amount of capital expenses?
*Will it be paid for out of the operating budget?

Certain phrases are “red flags” when relying on the WEPCo test to demonstrate a modification was “routine,” such as:

  • Restore original capacity
  • De-bottleneck
  • Increase capacity
  • Life-extension
  • Unprecedented
  • Capitalized expenditures
  • Decreasing down time
  • Change/Increase heat rate or MW
  • Increase reliability
  • Fuel consumption changes

When EPA has looked back as expenditure requests for “routine” projects, these descriptions have been used by the courts to determine that the modification was in fact not “routine.” Without a contemporaneous determination of baseline emissions and net increase, many projects were ruled to have violated PSD and been subject to large fines and costly retrofit projects.

PSD Netting Reduces Risk

PSD is triggered through a two-part test: 1) Is there a PSD modification? and 2) Is there a significant emissions increase? Using the WEPCo test for Part 1, most projects are PSD modifications. Whether or not there is an emissions increase after the project becomes of utmost importance. Thus, triggering PSD rests on Part 2: the potential emissions increase obtained through PSD netting. There is no such thing as a “like kind” replacement.

Proactive utilities conduct PSD netting calculations before each maintenance outage to document their baseline emissions and determine if future operation should be managed below calculated netting emissions rates. When conducted under attorney-client privilege, these calculations provide risk mitigation.

Before you pick up a pencil, review these netting considerations:

  • Calculations are always done on a tons per year (tpy) basis; thus a decrease in short term emissions with an increase in capacity factor can increase the tpy and trip PSD.
  • The baseline or past actual emissions are the average tpy from a consecutive 24-month period in the five years from start of construction of the new project at a utility plant (10 years if the facility is not an electric utility). This is NOT the permitted or allowed emissions; it is the actual emissions.
  • The baseline actual emissions 24-month period may differ for each regulated pollutant. If there are multiple emissions units emitting the pollutant, the baseline actual emissions 24-month period must be the same for that pollutant.
  • The baseline must be established before the modification project or outage.
  • Exclude any projects which were already included in a PSD permit
  • All other non-PSD projects in this 5 or 10 year past actual period must be included at their potential to emit (aka contemporaneous increases).
  • Include contemporaneous decreases but only actual emissions can be used (such as for a removed auxiliary boiler). Use the consecutive 24-month period to calculate the negative tpy. Since this only helps decrease the net emissions, this is optional.
  • Actual emissions that exceed a limit cannot be used in the past actuals.

The basic PSD netting formula is:

The advanced PSD netting formula allows a few extra steps that can be of great help to utilities:

  • Projected future emissions can be used instead of potential future emissions. For example, use the projected capacity factor accounting for predicted demand and downtime instead of 100% capacity factor.
  • This is key. Utilities are one of the few industries that have detailed records of “could have been accommodated” emissions. When a utility bids into the Regional Transmission Organization, records of generated of how many kilowatt-hours  could have been produced. When compared to the actual kilowatt-hours operated, the “could have been accommodated” kilowatt-hours and the associated emissions can be calculated. This has the effect of drastically increasing the baseline emissions for a peaking plant or any other plant that had more availability than called upon by the Regional Transmission Organization.
  • “Increased utilization due to product demand growth” is another, less useful, exemption to the netting procedure. If future operation increases due to demand growth which was unknown at the time of the project, that demand growth can be excluded from the netting. This is rarely useful.

The advanced PSD netting formula is:

Compare the net emission increase to the PSD significant emission rates, shown in Table 2 (only the most common pollutants are shown). Please note, greenhouse gases  cannot trigger PSD unless another pollutant also triggers PSD.

Project Emissions Accounting Change

An October 21, 2020, EPA modification to the PSD rules now allows netting to take place in two steps. Assuming a single piece of equipment is being replaced, the revised netting allows the first step to look at the full scope of the project emissions, both increases and decreases. For example, a boiler replacement could examine the future projected actuals minus the baseline actuals in Step 1. Only if that calculation exceeded the SER would contemporaneous increases and decreases from other projects at the source have to be included in Step 2.

This results in an abbreviated Step 1 equation:

This modification has not been adapted by all states and is of limited use but does avoid the requirement to track down the contemporaneous increases at a site that has undergone constant small project changes.

Reasonable Possibility

EPA established “reasonable possibility” record keeping and reporting regulations in 40 CFR 52.21(r)(6)(vi) as follows. However, state rules might supersede these requirements.

  • The 50% threshold varies depending on if “could have accommodated” emissions are used.
  • Establish baseline emissions in writing.
  • Keep post project records of actual emissions.
  • May need to report annually to agency for 5 or 10 years after the modification to document that actual emissions did not exceed the netting calculation.

What’s In It for You?

Before your modifications, use attorney-client privilege to conduct a PSD netting study to document baseline emissions, take credit for “could have been accommodated” emissions, and determine if records or reports are needed under reasonable possibility or state regulations. Since the Clean Air Act is not expected to be amended in the foreseeable future, PSD netting calculations provide the best risk mitigation against PSD lookbacks and huge fines, but only if done in advance of the modification project or outage.

[1] https://www.epa.gov/enforcement/coal-fired-power-plant-enforcement.

About the author: Robynn Andracsek is a regular contributing editor and P.E., Burns & McDonnell.

No posts to display