See the latest EHS federal and state regulatory updates due to COVID-19

EPA first proposed regulations for stationary internal combustion engines (ICE) on December 19, 2002. It affected a relatively small number of stationary engines: reciprocating ICE (RICE) with a power rating of greater than 500 hp that operated at major sources of hazardous air pollutants (HAPs). It took more than 18 months for EPA to finalize the regulation, which was codified in 40 CFR Part 63, Subpart ZZZZ. That regulation – commonly known as the RICE MACT – and the process it took to develop it was an omen of things to come.

EQ Fall 2014 - Engine Rules picIn the 12 years since the first proposal, EPA has published 25 Federal Register (FR) notices related to RICE MACT revisions/expansions, corrections, and reconsiderations and two other, separate regulations affecting stationary engines: 40 CFR Part 60, Subpart IIII and Subpart JJJJ, which regulate new ICE. In addition to the regulations themselves, EPA has published thousands of pages of regulatory preamble, guidance memorandums, training materials, compliance tools and templates, letters, and e-mails. This article discusses ten of the key elements of the rules and guidance, and how these elements apply to permitting and compliance projects. While there are many nuanced aspects of these rules that could be discussed, the following are the most common or potentially the most significant.

1. What is a Stationary Engine?

This question speaks to the fundamental applicability of the regulations. Each of the three regulations stipulates that only “stationary” ICE are affected. In contrast, “mobile” ICE, or, more specifically, “nonroad” ICE, are not subject to these regulations (rather, they are subject to mobile source regulations that principally affect the manufacturers of such engines). Thus, the definition of a nonroad ICE is critical to determining applicability.

From 40 CFR 1068.30, a nonroad ICE is, in summary, an ICE that is in or on a piece of equipment that is propelled while operating or is portable or transportable. It is the latter part of the definition that is of most importance in the decision of stationary vs. nonroad. To be portable or transportable means that the engine (1) was designed to be moved, e.g., it is on wheels or skids or it has handle(s) for carrying, and (2) is actually moved at least once per year (or shorter for seasonal sources) from one location (building, structure, facility, or installation) to another.

The definition of nonroad ICE has two critical implications for what may have otherwise been classified as stationary engines. First, an engine that is moved from location to location, even within the same plant site, may qualify as a nonroad engine. Example situations include a portable generator that is moved around within a mine to power various equipment or a trailer-mounted grinder that is used at a forest products facility to clear various, separate piles of wood waste. One of the key guidance documents pertaining to the first example situation is a December 5, 2008 letter from EPA Region 5 to the Minnesota Pollution Control Agency.

Second, the 12-month provision for portability found in the definition of nonroad provides for what is effectively a temporary-use exemption for stationary engines. The common situation where such an exemption may be considered is that of a new site for which grid power is not yet available and portable generators are brought in until grid power is supplied less than one year later. It is critically important to note that EPA and state/local regulatory agencies expect operators to make good-faith determinations of whether or not their engine(s) quality as “nonroad” based on solid forecasts of future use(s). For example, if the “installation” or purpose of an engine is to operate for more than one year then that engine should be treated from the start as a stationary ICE even if it is portable/transportable and even if it is replaced frequently (e.g., monthly) with other engines that accomplish the same purpose (in such a case, the original engine and all subsequent replacement engines should be considered stationary). Similarly, engines that are used in place of a stationary engine – while the stationary engine is being overhauled for instance – should be treated as stationary ICE. The expectations about an operator’s intentions for an engine, and a healthy discussion of the consequences of circumvention, is perhaps best presented in a September 2, 2009 Alaska DEC memorandum.

2. Certification Expirations

For many engines, the simplest method of complying with the two New Source Performance Standards (NSPS) regulations, 40 CFR 60, Subparts IIII and JJJJ, is to purchase engines for which the manufacturer has received a Certificate of Conformity from EPA. The only other substantive compliance requirement for operators of such engines is to properly operate and maintain the engine. However, these certificates have expiration dates and/or disclaimers that activate upon a certain number of hours of operation. Are the engines out of compliance upon certificate expiration or upon reaching the end of the warranty period?

The answer Based on the December 2007 rule preamble addressing the July 11, 2005 proposed rule, neither document expiration nor operation in excess of warranty periods results in noncompliance with the NSPS regulations.

3. Notifications

The RICE MACT and two NSPS regulations contain a myriad of notification requirements that are easily forgotten among all the other responsibilities of an environmental manager. Table 1 provides a summary of the most common notification deadlines for a stationary engine. Note that some or all of these notification requirements do not apply to many engine types.

EQ Fall 2014 - Table 1

4. What Really Is an Emergency?

The regulatory definition of and operational restrictions for emergency ICE are straightforward: operate only during emergencies except for up to a total of 100 hours per year in specified non-emergency situations. (Note that (a) more than 100 hours per year is allowed in certain limited situation and (b) some states restrict non-emergency operation to less than allowed by the federal regulations). Several areas of clarification are needed regarding the emergency-use provisions.

What happens if you exceed the non-emergency hours of operation limits? The answer depends on the guidance document. EPA’s April 2, 2013 Q&A document, the most recent, comprehensive extra-regulatory guidance document published by EPA about stationary engines, indicates that operation of an emergency engine in non-emergency situations beyond the allowable limitations by any amount would recategorize the engine from emergency to non-emergency. However, in the preamble to the 2013 RICE MACT amendments, EPA stated that such situations call for case-by-case decisions. Presumably, the situation could be handled like any other deviation.

Can you use an emergency engine to prevent an emergency? The adage, “an ounce of prevention is worth a pound of cure”, does not hold true for stationary emergency engines. On numerous occasions, EPA has been approached about situations where engines are used to prevent forthcoming disasters. The most common example is turning an engine on when a storm is approaching that is sure to flood areas and/or interrupt power. In each case EPA has said that such use is of a non-emergency nature and therefore is subject to the hours of operation limitations.

Is it always worth the hassle? Compliance with RICE MACT emergency-use provisions is not necessarily easier than compliance with non-emergency provisions. In fact, for the five types of engines listed below, requirements are generally less stringent for non-emergency varieties. In both cases, the primary compliance requirement involves maintenance of the ICE (e.g., oil changes, inspections of belts, hoses, etc.); emergency ICE has the added hours of operation limitations and the recordkeeping and reporting requirements that come along with them.

  • Existing RICE < 100 hp at major sources
  • Existing compression ignition (CI) RICE < 300 hp at area sources
  • Existing four-stroke (4S) RICE < 500 hp at area sources
  • Existing two-stroke (2S) RICE at area sources
  • Existing 4S RICE > 500 hp at remote area sources
5. The “Gap”

There is a regulatory gap (loophole) for many stationary engines. In a nutshell, if an area source engine is constructed (installed on site or contracted to be installed on site) after 6/12/06 and is manufactured prior to whichever of the five NSPS manufacture dates applies, then the engine is not subject to any compliance requirements (regardless of when the engine was ordered). Note that a gap engine is still technically an “affected source” under the RICE MACT. The figure on page 9 graphically depicts the fundamental applicability requirements for the three stationary engine rules – the five NSPS manufacture dates are listed on the right side of the NSPS arrows.

EQ Fall 2014 - Arrow Graphic

6. Common Emission Calculation Mistakes

Engine air emissions calculations are straightforward – most often taking the form of an activity rate (either power output or fuel heat input) multiplied by an emission factor wherein the “trick” is to simply cancel the units. Despite the simplistic form, there are several common engine emission calculation mistakes.

Improperly de-rating the horsepower value. First, the two sets of regulations – NSPS vs. MACT – use different definitions of engine power. The RICE MACT uses “site-rated hp”, which allows for adjusting the maximum horsepower rating based on site conditions such as elevation (rule of thumb: three percent per 1,000 foot), but NSPS regulations refer to “maximum engine power”, which does not allow for de-rating for any reason. For potential-to-emit (PTE) calculations, using the maximum power rating for the engine is generally recommended. Also, it is common for engines to be coupled to pumps, compressors, or other equipment that effectively limits the power output of the engine. These constraints should not be considered when calculating PTE or when determining applicability of the three stationary engine rules.

Not accounting for engine efficiency. Engines are not very efficient – 30 to 40 percent in most cases with some CI (diesel) engines exceeding 40 percent. That is, every unit of mechanical power output requires about three time as much power input in the form of fuel heating value. In AP-42, EPA suggests the use of 7,000 MMBtu/hp-hr as a conversion factor. This is a misnomer; it is not a conversion factor. The straight conversion factor is 2,544 MMBtu/hp-hr, but it does not account for efficiency. With the 7,000 MMBtu/hp-hr factor, EPA is assuming a 36.6 percent efficiency in addition to the conversion factor, meaning that it is what manufacturers refer to as a fuel consumption rate or a heat rate. Ultimately, when using emission factors from multiple references, it is critical to match up the emission factors to the activity rate in terms of input or output and to make sure efficiency is appropriately considered.

Using an inappropriate fuel heating value. Fuel heating value is a critical variable when completing certain conversions necessary for emissions calculations. When referring to the heating value of a fuel, engine manufacturers use lower heating value (LHV) whereas EPA uses higher heating value (HHV). The difference between LHV and HHV is the amount of heat it takes to vaporize water created by combustion, and it can be significant, e.g., it is approximately 10 percent for natural gas / methane. As with output vs. input, it is important to confirm that emission factors and activity rates are on the same basis in terms of heat value.

Carelessness about emission factors. Emission factors can come from a variety of sources. The most common source is engine vendor/manufacturer information. When using such data, it is important to note any exclusions. For example, vendors will often exclude formaldehyde from their published VOC emission factor. This can be acceptable for regulatory applicability purposes, but it is not appropriate for PTE calculation purposes since formaldehyde is a VOC. The next most common source of emission factors is AP-42, which, for engines, should be considered suspect as the data is old and not necessarily categorized in the most appropriate manner. Additionally, such emission factors should always be used with caution because they are averages – meaning that roughly half of all tested sources emitted at levels higher than the published factor.

When calculating emissions, it is always critical to compare to any applicable standards, and in many cases it is advisable to use the emission standards to establish PTE. However, for engines, using the NSPS-applicable Tier/mobile source standards is not always recommended for a variety of reasons. First, the Tier/mobile source standards are established for a family of engines based on weighted averages, not necessarily for any specific engine within the family nor maximum engine power. So some engines may actually emit at higher rates than the applicable standard. Many engines also emit at rates much less than the Tier standards – especially for carbon monoxide (CO) – and using the standard as a PTE emission factor can result in very high estimates (exceeding major source thresholds for example).

Not considering the full effects of controls. The most common add-on control technologies for engines are oxidation catalysts (OxCat or two-way catalyst), which are used to control lean burn engines, and non-selective catalytic reduction (NSCR or three-way catalyst), which are used to control rich burn engines. Both technologies control emissions of CO and hydrocarbons by further oxidizing the pollutants (NSCR also reduces NOX). These control techniques increase the amount of carbon dioxide (CO2) emitted, which may need to be accounted for in emissions calculations. Also, OxCat substantially changes the NOX chemistry – further oxidizing NO to NO2. This potentially creates an ambient air concern that may have been avoided without the use of OxCat since the national ambient air quality standard (NAAQS) is specific to NO2, not total NOX. Lastly, it is important to bear in mind any unique fuel constituents when considering catalytic controls as many metals and especially sulfur can damage the catalyst rendering it partially or completely ineffective.

7. Performance Testing

Exhaust gas volumetric flow rate is a critical parameter for calculating mass emission rates with performance testing. Method 2 is most commonly used to measure velocity, which is in turn used to calculate volumetric flow rate. However, it is not necessarily the best (most accurate) method available. Method 2 typically has a 10 percent margin of error in good conditions, and that error margin can increase substantially in the exhausts of reciprocating engines with variable flow rates/velocities. Calculation of emission rates following Method 19 procedures should be considered as an alternate to Method 2 based calculations. Method 19 calculations are based on fuel flow and excess air (O2 and/or CO2) measurements.

Another test method concern involves the various options available for measuring hydrocarbon emissions. Although Method 25A is most commonly used, alternative methods – Method 18 and Method 320 – can be more accurate than Method 25A, but they are also more expensive and more susceptible to interferences. Furthermore, while Method 25A does not identify any specific organic compounds in the VOC mix, Method 18 and Method 320 do. Sources may wish to consider whether they want to identify specific organic compounds when there is no regulatory requirement to do so, especially in an atmosphere where such information can be misinterpreted and misused.

8. Site-Specific Monitoring Plans

With respect to site-specific monitoring plans (SSMPs), the most important issue is to simply to have one available for review that satisfies RICE MACT provisions. See Trinity’s Winter 2014 EQ for an article about preparing SSMPs.

The second concern about SSMPs, or really about the monitoring provisions of the RICE MACT regulation itself, is that the rule is virtually silent about required pressure drop measurements. It says nothing about data averaging, quality control, calibrations – only that monthly measurements should not deviate more than two inches of water (in-H2O) from the measurement taken during the initial performance test.

9. Permitting Advice

For engines, unlike many emissions units, mutually agreeable permit conditions are typically the result of education of, not negotiation with, agency permit writers. Between the complexity of these rules and the necessity of knowing many other rules as well, it is simply not reasonable to expect most permit writers to be experts on the stationary engine regulations. As such, permit applications should be explicit about which provisions are applicable and which, non-standard compliance options, if any, are being used.

10. Embracing Reconstruction

For most projects involving large expenditures, a key goal is to avoid qualifying as reconstruction. However, because of how the stationary engine rules relate to one another, reconstructing an engine can sometimes be beneficial. In one case an existing 4S rich burn engine, which would be subject to continuous monitoring requirements under the RICE MACT, was intentionally reconstructed so that the only applicable RICE MACT requirement would be to comply with NSPS JJJJ, which requires performance testing and a maintenance plan. Over the life of the engine, recordkeeping will be substantially reduced. One caution regarding this strategy is that the cost aggregation method utilized will set a precedent for future evaluations.