Understanding emission requirements for standby gensets
To obtain an air emissions permit for facilities that have stationary diesel emergency standby generators, it is necessary to comply with U.S. Environmental Protection Agency (EPA) and local regulatory requirements. Ensuring that the system design takes into account these regulatory requirements can have a significant impact on overall facility layout and cost.
The EPA’s regulations are relatively complex and have been in a state of flux. This article provides an overview of the EPA regulatory framework with a concentration only on those requirements for stationary diesel emergency standby generators that are greater than 500 hp. This size range is commonly encountered in larger data centers, hospitals, and municipal infrastructure. This article also focuses on new installations only; it does not offer insight into the rules that govern existing retrofit installations.
This article also attempts to identify some of the key EPA terminology, such as National Ambient Air Quality Standards (NAAQS), New Source Performance Standards (NSPS), Reciprocating Internal Combustion Engine (RICE) National Emission Standards for Hazardous Air Pollutants (NESHAP), and Tiers 2, 3, and 4, and put them into a framework that allows critical power engineers to understand the current regulations and how those regulations influence the design of new facilities.
A diesel engine generates certain emissions that the EPA considers to be criteria pollutants. Criteria pollutants are deemed to be serious health risks and are measured by the EPA throughout the U.S. in geographic entities called areas. The key criteria pollutants associated with diesel engines are: nitrogen dioxide (NO2), particulate matter (PM), and carbon monoxide (CO) (see Figure 1).
Explaining the regulatory environment
The Clean Air Act forms the regulatory basis for all air compliance activity (see Figure 2). It was originally established in the early 1970s. The most important recent major amendments to the Act occurred in 1990. These amendments recognized the need to consider the available technology as a component in determining achievable standards. The EPA terminology for this is maximum achievable control technology (MACT). Cost-effective technology advances in MACT have created the platform for the EPA to look at new emission requirements for diesel engines.
As part of its risk assessment, the EPA allows emergency engines to meet somewhat lower standards than nonemergency units because of the lower annual operating hours. The definition of “emergency” can be relatively complex. Clearly, a utility outage is an emergency condition. There are specific definitions for voltage and frequency variations for electrical reliability that can also constitute an emergency situation. In general, a total of 100 hr/yr is allocated to emergency generators for maintenance and testing. Of these 100 hr, the EPA currently allows up to 50 hr to be used for demand response programs in some jurisdictions. However, this aspect is currently under review and may be removed. There are no restrictions on the number of run hours for the engine when it is being used under emergency conditions.
EPA regulatory framework
The EPA regulatory framework has several components based on the details of implementation. These components include NAAQS; RICE NESHAP; NSPS; and Tiers 2, 3, and 4. They are covered in detail in the following paragraphs.
NAAQS: Maintaining a NAAQS is a fundamental concept of the Clean Air Act. A NAAQS is based on limits that are designed to ensure healthy air quality for all citizens regardless of where in the U.S. they live. As part of a NAAQS, the EPA defines six criteria pollutants. The modern lean-burn diesel engine has improved dramatically in recent years, but can still contribute significantly to three of the criteria pollutants: NO2, PM, and CO.
NO2 is one of the constituents of NOX, the formation of which is largely a function of combustion temperature. Typically, a higher combustion temperature results in a higher level of NOX formation. PM is also a function of combustion temperature. Typically, a higher combustion temperature results in less PM formation. As a result, undesirable NOX and PM formation act in opposing directions when engine designers are investigating combustion temperature. Engine efficiency also typically improves at higher combustion temperatures, which is another important consideration for engine designers. CO is often a reflection of incomplete oxidation of fuel in the combustion chamber. Most major diesel engine manufacturers have optimized their combustion processes to such an extent that often CO regulatory requirements are not an issue.
The required targets and the timetable for NAAQS implementation are always changing and apply to each of the criteria pollutants. The EPA goes through a public consultation process to establish the required NAAQS levels for each criteria pollutant. The U.S. is divided into a set of areas, and the EPA performs measurements of the criteria pollutants in each area. Areas that do not meet the NAAQS targets for criteria pollutants are deemed nonattainment areas. For each nonattainment area, the affected state is required to prepare a state implementation plan (SIP) to resolve the issue and achieve attainment. The U.S. Northeast (Maine to Northern Virginia) is a special case. Because this air shed is highly populated, it has more stringent air quality standards. The EPA calls this area the Ozone Transport Region.
When seeking an air permit for a new diesel emergency generator, if there is a NAAQS issue, it will most likely relate to NO2. In 2010, the EPA proposed limits based on an hourly worst-case scenario of 100 parts per billion. It is not uncommon, during certain times, for background concentrations in nonattainment areas to be high enough that very little NO2 needs to be added to make an installation exceed the limit. Prior to 2010, the NO2 limit was based on a yearly average.
By mid-2013, each state was to have submitted a SIP for its nonattainment areas with respect to NO2. When a major data center, hospital, or other installation installs significant capacity of new diesel standby generators, the typical hourly worst-case scenario occurs during the full load test of the units. Modeling is done of the site, typically using the EPA’s atmospheric dispersion modeling (AERMOD) system. AERMOD is a mathematical simulation of how pollutants will disperse into the atmosphere. The modeling takes into account the topography of the site, its major emissions sources, prevailing wind conditions, and other factors that could lead to worst-case conditions.
RICE NESHAP and NSPS: The RICE NESHAP requirements from the EPA have received a lot of attention in the last few years, largely because of the impact these requirements have on existing nonemergency diesel and natural gas generators. These requirements have meant that many existing nonemergency diesel generators have had to add oxidation catalysts and other equipment to their engines. In keeping with the overall focus of this article on new emergency diesel generators, we will review RICE NESHAP and NSPS from this standpoint.
A facility is deemed by the EPA to be an area source if it has the potential to emit less than 10 tons/yr of any single hazardous air pollutant or less than 25 tons/yr of any combination of hazardous air pollutants. A major source has emissions greater than the area source levels. Typically, major sources have more stringent requirements.
The EPA has classified more than 70 area source categories. Examples include stationary reciprocating internal combustion engines (RICE) and boilers. Each of these categories has special NESHAP requirements and an associated timeline.
While NESHAP can impact new and existing RICE, NSPS applies to only new installations. As with RICE NESHAP, NSPS typically specifies performance standards that are defined within the EPA Tier levels discussed later in this article.
For the critical power engineer, RICE NESHAP and NSPS are typically not major issues for new emergency diesel gensets greater than 500 hp. Since 2008, all major manufacturers have produced engines that meet RICE NESHAP and NSPS requirements for new emergency diesel engines. To meet these requirements for a new diesel emergency engine, the engine must be certified to at least Tier 3; if it is greater than 752 hp, it must be certified to at least Tier 2. Most of the resulting obligations from RICE NESHAP apply to the facility operators, not the critical power engineer designing the facility. For example, site operators should use ultra-low sulfur diesel (ULSD) fuel. This is not a big constraint because ULSD has been in wide use since 2007. The facility operator must also record emergency operation with reference to a nonresettable hour meter and make this information available to the EPA if requested. There are other relatively straightforward record keeping and maintenance obligations for facility operators to maintain compliance with RICE NESHAP.
Tiers 2, 3, and 4: There has been a lot of press coverage on Tier 4 and its subsets Tier 4i (interim) and Tier 4f (final). The Tier 4 standards have had a huge impact on engine manufacturers because significant emissions reductions have been required to meet these standards. It is not uncommon for a large T4 stationary engine to cost 40% more than a similar power Tier 2 or Tier 3 engine because of the extensive emissions aftertreatment equipment that may be required. In addition, large stationary T4 gensets often require significantly more space allocation than Tier 2 or Tier 3 units.
The concept of EPA Tiers started in the early 1990s. The current level for new stationary nonemergency diesel engines exceeding 560 hp is Tier 4i, and by January 2015, Tier 4f will be in place for large stationary gensets. Under Tier 4, a large engine is considered to be one that exceeds 752 hp, whereas under RICE NESHAP, it is 500 hp. In general, EPA T4 standards target on-highway, off-road mobile sources and stationary nonemergency engine-driven generators. EPA T4 is not required for emergency gensets, but some engine vendors are advocating use of T4 engines to ensure there are no operating restrictions beyond the current 100 hr maintenance and testing limit currently in place. If a new engine is not T4, it must have a permanent label indicating that it is for emergency use only. It is important to note that, in addition to significant extra cost and space requirements, there can be some significant disadvantages to using T4-certified engines for emergency applications. For example, under current EPA rules, a certified T4 emergency engine used in a data center must shut down if the urea is unavailable. This is not a desirable situation for an emergency generator running during a long utility outage.
EPA regulatory framework summary
As mentioned previously, this article looks mainly at large diesel engines used in emergency standby applications. Figure 3 shows the EPA regulatory impacts for this type of application. If the critical power facility is large (has engines that exceed 500 hp), is located in a nonattainment area, and does full load testing, it may require some form of NOX or NO2 mitigation. Table 1 summarizes the EPA regulations associated with large stationary diesel engines used in emergency applications.
Technology to deal with air emissions from diesel engines
For large stationary diesel engines up to and including Tier 3, engine manufacturers have adopted many innovative technologies that typically focus on in-cylinder optimizations. Looking beyond Tier 3, much of the focus has been on exhaust aftertreatment technologies. For diesel engines, the most common aftertreatment emission control technologies are:
- Oxidation catalyst to deal with CO and unburned hydrocarbons
- Diesel particulate filter to meet PM requirements
- Selective catalytic reduction (SCR) to meet NOX requirements.
As mentioned previously, often NOX becomes the constraining pollutant from a NAAQS standpoint. All diesel engines will also require some level of exhaust silencing. As a result, a common configuration for large critical power facilities in nonattainment areas is to use Tier 2 for engines exceeding 752 hp and Tier 3 for engines less than 752 hp in combination with an SCR and silencing.
Oxidation catalysts and PM filters
For diesel engines, oxidation catalysts are often combined with particulate filters. This can be done by applying the catalysts, which are usually platinum-group metals, to a particulate filter. Another common approach is to have separate oxidation catalysts upstream of the particulate filters. The oxidation catalyst creates heat by oxidizing unburned hydrocarbons and shifts NO, creating a favorable environment for the particulate filters to regenerate.
SCR works by injecting a reductant, usually a 32.5% concentration of urea into the exhaust stream. The urea is converted into ammonia (NH3) in the hot exhaust stream. In the presence of a catalyst, the NH3 combines with the NOX in the exhaust to produce harmless water vapor and nitrogen. Many SCR systems can achieve NOX reductions of 95% or more. Some exhaust aftertreatment vendors offer multifunction systems that combine SCR, silencing, and slots that can be filled, if required, with oxidation catalysts and PM filters. This gives the critical power engineer a lot of flexibility, allowing him or her to add catalysts and filters late in the project cycle without impacting the size of the emissions unit and the surrounding piping should it be required for the air permit. Figure 4 shows an SCR system that combines silencing and other emissions functions in a single cube mounted on an enclosure that houses a large standby diesel genset.
The critical power engineer faces significant air compliance challenges because of the regulatory environment. These challenges are compounded if the site location is not fully finalized when the initial design is done. A change in air shed location could lead to a significant change in the results of the AERMOD simulation. A change in emissions mitigation requirements could then have a significant impact on the physical space required for various aftertreatment devices.
Until recently, aftertreatment was done using separate devices for each emissions function. Figure 5 shows an illustration of a separate silencer and SCR system in the exhaust stream of a large generator used in a data center. The physical space required for the devices and the complex piping and expansion joints required between them makes this arrangement a large and overly complex system.
Some vendors now offer exhaust aftertreatment systems that combine all required functions in a single cube. These multifunction systems can contain any combination of SCR, silencing, oxidation catalyst, and PM filters in the same cube. This makes installation much easier and allows critical power engineers to design systems that meet the regulatory requirements of any air shed in the U.S. The cube is typically installed above the engine. As a result, it does not take up much more space than a conventional silencer (see Figure 6).
The regulatory requirements for obtaining an air permit for large scale critical power facilities using stationary diesel engines is continuing to become more complex. It is important for critical power engineers to understand the overall regulatory framework and build enough flexibility into their design to ensure that the requirements for an air permit can be met.
Bob Stelzer is the chief technical officer for Safety Power Inc., Mississauga, Ont. He leads the engineering team that developed the company’s ecoCUBE family of products, which has been configured for more than 40 engine types from most of the world’s major engine manufacturers. He is a mechanical engineer with a master’s degree in engineering.