AERA guide | Section 1: Facility characterization

As a first step, you need to characterize your facility. To do this, you must:

  • Identify air emission sources
  • Identify chemicals of potential interest
  • Estimate emissions

Provide all of the information identified in this Microsoft Office document icon AERA-05 Emission form, including:

  • A description of the process and effort expended in identifying chemicals of potential interest
  • Limits and/or assumptions
  • Identification of any changes made between submittals
  • Assumptions made for air toxic speciation and references

You must also attach an emissions calculation spreadsheet, including all pertinent information needed to review the specific air toxics emission rates, including, but not limited to:

  • Hourly emission rate calculations
  • Annual emission rate calculations
  • Emission factors with references 
  • Control efficiencies with references
  • Process throughputs
  • Any proposed permit limits that may impact emission rate calculations
  • Stack and fugitive source parameters (this may be found in the air dispersion modeling tables or forms)

An emissions unit is any piece of equipment or any process that emits pollutants into the air, including stacks, vents, and fugitive processes. The Microsoft Office document icon AERA-05 Emission form, together with the emissions calculation spreadsheet, are used to document which sources are present at the facility and which ones will be included in the AERA.

Here are some of the common emission sources:

  • Combustion stack/vent point sources
  • Non-combustion stack/vent point sources
  • Onsite mobile source tail pipe emissions
  • Idling vehicle tail pipe emissions
  • Onsite fugitive emission sources
    • paved roads
    • unpaved roads
    • storage/surge piles
    • material handling operations
    • valve, tanks, equipment leaks

Fugitive emissions sources

Operations without a specific stack or vent that exhaust into outdoor air through building ventilation or their emissions escape through doors or windows (e.g., parts washers), are called fugitive sources.

These sources are included in the AERA and can be modeled using the RASS at a screening level. Examples of emissions from fugitive sources include volatile organic compounds from outdoor leaking valves, hydrogen sulfide from uncovered wastewater treatment plants, and particulates blowing from outdoor stockpiles.

A note about natural gas combustion sources

Emissions due to combustion of natural gas as a fuel for boilers or other equipment need to be included in the RASS and Q/CHI Spreadsheet. If backup fuels are permitted, such as in the case of a natural gas curtailment, the worst-case emissions, by pollutant from each fuel source, needs to be assessed in the AERA. Although some natural gas combustion activities were once exempted from the AERA, these emissions now need to be quantified because of the general increase in natural gas combustion and the availability of natural gas emissions information.

Emission sources that do not have to be quantified

Certain types of emission sources do not need to be included in the quantitative risk estimates, as described below. These sources include:

  • Some “insignificant activity” emissions sources
  • Emergency generators (follow link for quantitative exclusion criteria)
  • Non-continuous Sources: associated with blasting, some start-up or shut-down situations, upset, and/or emergency situations
  • Sources emitting air toxics with total risks summing to less than risk-driver levels
  • Other (e.g., case-by-case determination on vehicle emissions)

Some "insignificant activity" emissions sources

Some “insignificant activities” as defined in Minn. R. 7007.1300 may emit substantial amounts of air toxics or small amounts of highly toxic chemicals that need to be included in the AERA process.

However, the emissions associated with an “insignificant activity” may be excluded from quantitative risk estimates if:

  • the activity emits an air toxic that does not have an IHB listed in the RASS, or
  • the activity emits an air toxic that is also emitted by sources/units already included in the emission inventory, and the contribution of the individual insignificant activity is less than 1% of the total emission inventory for an air toxic (hourly for acute and annual for chronic).

If an emissions source does not meet one of these two tests, then it must be included in the RASS. Documentation needs to be provided indicating why an activity is excluded.

Emergency generators

The emissions from an internal combustion engine associated with an “emergency generator” or fire pump are generally not quantified in the AERA. All other engines are included if not exempted as an insignificant activity emissions source. An “emergency generator” is only operated when unforeseen conditions result in disruption of electrical power to the stationary source. An emergency generator may not be omitted from quantification in the AERA if it is a part of a peak shaving contract, reduced use contract, or if it is used as a standby source during periods when power is available from the utility. The definition of emergency generators used for the purposes of an AERA comes from this EPA memo:


Maintaining emergency generators as a backup power source is recognized by the MPCA as being essential; however, they are sources of air toxics. Emergency generators require testing with a specific frequency and load. Testing regimens can be frequent (e.g., weekly) and if there are multiple generators, testing can be performed simultaneously or can be conducted within a period of several hours. This may result in very high localized concentrations of pollution which can represent significant risk. The MPCA therefore requests a project proposer to inventory, characterize, and certify emergency generators and fire pumps at the facility using this form:

Suggestions for minimizing emissions and impacts from emergency generators as well as detailed information on emergency generator use are provided in this form, and in this document:

Screening out sources

Emissions from all units or sources at the facility needs to be evaluated in an AERA, either within the quantitative risk estimation or qualitative descriptions, with the exception of those screened out using methods described in the section Screening out pollutants and sources using the RASS

Once all relevant emission sources are identified, an inventory of air toxics emitted or potentially emitted needs to be provided. These air toxics are called Chemicals of Potential Interest (COPI), and include those air toxics with inhalation health benchmarks (IHBs), some criteria pollutants with health benchmarks, and pollutants for which an IHB is not available.

COPI quantitatively assessed for risk—Chemicals with IHBs

Air toxics with readily available inhalation health benchmarks (IHBs) (i.e., those listed in the ToxValue tab of the RASS), which are emitted, or potentially emitted are quantified in the AERA for risk estimates. If emissions data are not available for an air toxic with an IHB, and if the project proposer and MPCA staff agree that it is unreasonable to make a comparison to a similar type of air emissions source, a project proposer does not have to provide emissions data for the air toxic. However, a project proposer must describe attempts made to identify emission factors (e.g., list databases consulted, literature reviewed, internet searches, industry databases, personal interviews with experts).

COPI qualitatively assessed—Chemicals without IHBs

If there is no toxicity value in the hierarchy, nor is there sufficient toxicity information for the MDH to develop an IHB (see MDH Review of Air Toxics without IHBs), a pollutant-specific quantitative analysis is not necessary. However, some project proposers have provided qualitative information comparing modeled results to occupational health IHBs.

Air toxics emitted or potentially emitted without IHBs will be screened by MPCA staff for USEPA-designated hazardous air pollutants (HAPs), respiratory sensitizers, persistent, bioaccumulative toxicants, and emerging pollutants of concern.

Criteria pollutants

Criteria pollutants are those pollutants with federal or state ambient air quality standards that include PM2.5, PM10, TSP, hydrogen sulfide, nitrogen dioxide, sulfur dioxide, carbon monoxide, lead, and ozone, for various time averaging scenarios. Air toxics are air pollutants that are known or suspected to cause cancer or other adverse health effects, and may include criteria pollutants that have an inhalation health benchmark in addition to the air quality standard. The MPCA defines air toxics as any pollutant that has a health benchmark from the AERA hierarchy of toxicity information.

Usually projects that conduct an AERA also perform separate, criteria pollutant modeling. If this is the case, a qualitative summary of and reference to the separate criteria pollutant analysis is sufficient.

Natural gas boiler air toxics

All air toxics emitted from natural gas combustion sources, with AP-42 emissions values need to be evaluated quantitatively with the following exceptions, which have E rated emission factors based on detection limits. These air toxics do not need to be included in quantitative emission estimates from natural gas fired boilers because of the associated uncertainty. Instead, these air toxics need to be discussed qualitatively.

  • 56-49-5 3-Methylchloranthrene
  • 57-97-6 7,12-Dimethylbenz(a)anthracene
  • 83-32-9 Acenaphthene
  • 203-96-8 Acenaphthylene
  • 120-12-7 Anthracene
  • 56-55-3 Benz(a)anthracene
  • 50-32-8 Benzo(a)pyrene
  • 205-99-2 Benzo(b)fluoranthene
  • 191-24-2 Benzo(g,h,i)perylene
  • 205-82-3 Benzo(k)fluoranthene
  • 218-01-9 Chrysene
  • 53-70-3 Dibenzo(a,h)anthracene
  • 193-39-5 Indeno(1,2,3-cd)pyrene
  • 7440-41-7 Beryllium
  • 7782-49-2 Selenium

Screening out pollutants

All emitted pollutants need to go through the quantitative AERA process with the exception of those screened out using the methods described in the “Screening out pollutants and sources using the RASS” section, or those specifically addressed in this guidance.

The foundation of all qualitative and quantitative information in an AERA is a comprehensive list of air toxics and their hourly and annual emission rates. As such, it is crucial that a project proposer provide high quality information about the air toxics emission rates for a complete list of COPI. Proposed emission rates that are below an emission source’s unrestricted emission rate or below a permit limit may become the basis of new enforceable permit limits. Project proposers need to submit current facility emissions based upon current permitted limits and total facility emissions after the proposed modification.

Emission factors

Process-related emission factors must be found or developed for each air toxic before emission rates can be estimated. An emission factor is a representative value that relates the quantity of an air toxic released to the atmosphere with an associated activity.

There are accurate and representative data to derive emission factors for the sources contained in Minn. R. 7005.0100, Subp. 10a.

There are fewer emissions measurements available for air toxics than criteria pollutants. This means that there is usually a higher uncertainty and a greater variability associated with air toxics emissions estimates. As in other steps of an AERA, and risk assessment in general, conservative assumptions are made when there is uncertainty in data. Many emission factors are arithmetic averages of a data set, and need not automatically be assumed to represent maximum emissions for a source type on either a short or long term basis.

Emission factors that are arithmetic averages

Emission factors, such as those found in AP-42, are often arithmetic averages of the available data set. If used without modifying to account for being an arithmetic average, the MPCA will describe the resulting risk estimate as potentially underestimated (see discussion of the development of emission factors from stack test data below). Describe any mitigating factors if AP-42 factors are believed to be upper-bound estimates in the Microsoft Office document icon AERA-05 Emission form.

Sources of emission factors

Emission factors can be either source-specific or generic. Source-specific emission factors are derived from source-specific emission testing, mass balance or chemical analysis and are preferred in comparison to generic emission factors. The MPCA has worked with certain industrial sectors such as metal mining, iron ore processing, electric services and coal burning facilities to develop source-specific emission factors. Generic emission factors are usually derived from actual measurements of the emissions from representative sources/processes, and are assumed to be the long-term averages for all facilities in the source category.

Internet searches can identify similar facilities, especially those that have been permitted in other states. Other states’ analyses often can reveal additional monitoring data from emissions that could be used to estimate source specific emission factors.

USEPA’s AP-42 is the most common source for non-facility specific emission factors. However, before using these factors directly, consider their derivation and applicability. AP-42 factors can be assumed to represent long-term emissions for a source, but care needs to be exercised in estimating hourly emissions. AP-42 guidance directs users to review the literature and technology and to be aware of circumstances that might cause specific sources to have emissions characteristics that are different from generic sources.

Few data sources will contain emissions from emerging or novel air emission processes. Furthermore, emerging or novel processes are generally not included in emission factor databases. Reasonable effort is expected to identify emissions associated with these types of sources by examining professional literature or interviewing expert authorities. Characterizing emissions from similar sources (e.g., by fuel type, process) might be appropriate in these instances, while recognizing the attendant limitations.

Table 1 below describes common sources of emissions information.

 

Table 1 Types of sources of data for air toxics assessments Comments on data quality
 
URL Link
 
Air emission test data from a project proposer’s own facility or similar facilities elsewhere
 
Test data are very useful for developing the list of chemicals emitted from a facility, along with an emissions rate. Cannot be used to exclude chemicals not tested.
 
Facility or source specific
 
U.S. Environmental Protection Agency’s (USEPA) AP-42 
 
Emissions data published by USEPA which is categorized according to data quality. Provides criteria pollutant emission factors and for most emission sources, toxics emissions factors.  Also will often include emissions summaries for source types for which a MACT standard has been developed.  AP-42 factors can be assumed to represent long-term emissions for a source, but care needs to be exercised when using them. AP-42 guidance directs users to review the literature and technology to be aware of circumstances that might cause sources to have emissions characteristics that are different from other typical existing sources.
 
http://www.epa.gov/ttn/chief/ap42/c00s00.pdf

http://www.epa.gov/ttn/chief/ap42/index.html

 
USEPA’s Factor Information Retrieval (FIRE) Data System
 
USEPA’s “staging area” for air toxics emissions factors. Somewhat complete information for combustion sources, but incomplete for emissions from manufacturing units.
 
http://cfpub.epa.gov/webfire/
 
Material Safety Data Sheets
 
Very reliable source of air toxics content information for painting and other coating or evaporative uses. MSDS sheets may not be reliable data sources for estimating  emissions where chemical reactions are involved.
 
http://www.ilpi.com/msds/#Internet
 
Chemical analyses of feedstocks and products
 
They are very useful for developing the list of chemicals emitted from a facility, along with an emissions rate. Cannot be used to exclude chemicals not tested.
 
Facility or source specific
 
Reasonable attempts to find information not available else where

·      Peer-Reviewed technical literature

·      Conference proceedings

·      Trade organizations that maintain emissions databases or information

·      Industry publications

·      Trade group reports

 
Best engineering judgment and consideration of the following factors needs to be  used when developing emission factors:

(a) the precision and accuracy of the data

(b) the design and operational similarity between the emission units

(c) the size of the data set

(d) the availability of data of equal or greater quality

(e) operating conditions of the emissions unit when data was collected

(f) the data analysis procedures

 
Facility or source specific
 
Document for the Electric Generating Unit (EGU) National Emissions Inventory (NEI)
 
Data are obtained using the Energy Information Administration (EIA) – 767 electric power survey. These are USEPA derived emission factors.
 
https://www.epa.gov/air-emissions-inventories

 
·       California Air Toxics Emission Factors Search

·       California’s Best Available Control Technology

 
California database of emission factors. May provide emission factors for chemicals not available through  other sources.
 
http://www.arb.ca.gov/app/emsinv/catef_form.html

http://www.arb.ca.gov/bact/bact.htm

 
North Carolina DENR Division of Air Quality
 
  http://daq.state.nc.us/monitor/eminv/
 
USEPA tool: PM Calculator
 
  https://www.epa.gov/air-emissions-inventories/emission-inventory-tools
 
NCASI Technical Bulletins
 
Forest products industry group developed emission factors. (forest products—mainly paper but some wood products also).
 
http://www.ncasi.org/Programs/Reports-and-Articles/Technical-Bulletins-and-Special-Reports/Technical-Bulletins/Index.aspx
Novel Air Emission Sources
 
Few data sources will contain emissions from emerging or novel air emission processes (i.e. processes that are not common or have not had stack testing)
 
Reasonable effort is expected to identify toxic air emissions associated with these types of sources: professional literature, or interviewing expert authorities, Characterizing emissions from similar sources (e.g. by fuel type, process)
 
USEPA, Health Assessment Document for Diesel Engine Exhaust. 2002. USEPA/600/8-90/057F. Dioxin/furan combustion of diesel emissions, large variability, contact MPCA if you propose to use this source. https://cfpub.epa.gov/si/si_public_record_report.cfm?dirEntryId=29060&simpleSearch=1&searchAll=diesel

 

Note: Compilation of Air Pollutant Emission Factors (AP-42), Factor Information REtrieval (FIRE) as well as the California Air Toxics Emission Factors (CATEF) database are preferred information sources. In considering multiple sources of emissions information, USEPA’s emissions information presented in AP-42 needs to be used where available. When an emission factor for a pollutant is not available from AP-42; FIRE or CATEF may be used. If you determine that either FIRE, CATEF, or another published emission factor has been developed using more robust data than that used for AP-42, that alternate factor may be used. When using the proposed facility alternative emission factors, describe the number of tests used to generate the factor, and the similarity of the emission unit, operating conditions, and control equipment to the proposed facility.

AERA emissions

Emission estimates calculated for an AERA need to be the most accurate estimate of permitted emissions over the appropriate timeframe, with a reasonable certainty that air toxics emission rates are not underestimated. The availability of an acute or a chronic health benchmark will determine the “appropriate time frame” for each chemical. The MPCA permitting webpage includes source-specific and some pollutant-specific emissions estimation guidance. This needs to be the first stop for estimating emissions. The information required for AERA emission rates is the same as the information required for permit applications, with a few exceptions:

  • In addition to listed HAPs, the AERA needs to include air toxics that have toxicity values listed in the RASS.
  • The AERA reviewer will need to follow emission rate calculations from the process capacity and emission factor to the emission rate used in refined AERMOD modeling (grams/second) or entered into the RASS (pounds/hour or tons/year).

Emission rates used in AERAs

Project proposers need to assess the proposed “potential to emit” (PTE) scenario as defined by permit conditions. These permit-allowable emission rates take into account existing and proposed permit limits, rule-based limits, and any other enforceable limits.

Types of limits within a permit include numeric emission limits, operating limits, such as hours per year or use of control equipment, throughput limits, allowable fuels, and allowable materials. A permit may describe multiple or alternative operating scenarios. If emission rates reflecting a lower-emitting operating scenario are used in an AERA, they may become the basis for state-only enforceable limits within the facility’s permit.

Generating emission estimates

The methodology for calculating air toxics emissions needs to be the same at existing facilities for pre- and post- modification scenarios. If not, documentation and justification need to be provided.

Emission rates need to be estimated for the subset of chemicals on the COPI list with readily available IHBs (i.e., those listed in the ToxValue tab of the RASS). The RASS is designed to assess inhalation risks from long-term exposure to the average ambient air concentration during a year (chronic), short-term exposure to the maximum potential hourly ambient air concentration (acute), and mid-term exposure over a one-month period (sub-chronic). Maximum annual and the maximum hourly emission rate for each chemical must therefore be determined to conduct the AERA (monthly air concentrations for air toxics are estimated in the RASS based on annual emission rates).

Since AERA risk estimates generally rely on RASS, AERMOD, or HHRAP-based software, the emission rates are required to be in the appropriate units for the software used.

AERA submittals need to include emission calculations spreadsheets including all of the information described on the MPCA emissions estimates for permitting webpage.

Special considerations for hourly emissions

If a permit has a ton per year limit on an air toxic, it is not appropriate to divide that by 8760 hours per year to obtain an hourly emission rate. The maximum hourly emission rate needs to be based on the hourly capacity of the unit and the appropriate emission factor. For short-term emissions, it is very likely that the estimated actual emissions are the same as the potential hourly emissions since many emission units can and do operate at full capacity for short-term periods.

Certain parameters should be adjusted for short-term emissions estimates. For example, when using a material balance method, if a range of toluene content is provided on the MSDS, the highest number of the range should be used to estimate the potential hourly toluene emissions. In the case of batch processes that last more than an hour, where the air toxic emission rate may fluctuate throughout the batch, the hourly air toxic emission rate needs to be based on the hour with the highest emission factor.

Using an emission factor developed from stack emission tests

For use in an AERA, the stack tests need to be conducted at maximum permitted operation. The values are generally collected in sets of three measurements. Many times, there may be two sets of three measurements. The preferred summary of these measurements for use in developing an emissions rate is a 95 percent upper confidence limit (95 UCL-AM) of an arithmetic mean. If the 95 UCL-AM exceeds the maximum measurement, the project proposer needs to use the maximum measurement to develop the emission rate. At times it may be advisable to use a mean plus 1 or more standard deviations, dependent on the distribution of the data and the number of values.

The 95-UCL is defined as a value that, when calculated repeatedly for randomly drawn subsets of site data, equals or exceeds the true mean 95% of the time. The 95 UCL-AM should not be confused with the 95th percentile of the measured data. Five or more total values with 4 of those values being detected measurements, are required to obtain 95 UCL-AM values from ProUCL (http://www.epa.gov/nerlesd1/tsc/software.htm).

All small data sets below 8-10 values will result in warnings from the ProUCL software. If there are less than five detected values in the data set, the highest measured stack test value needs to be used to develop annual and hourly emission rates. In some cases, no values are detected. In this case, the emission rate should be developed using the detection limit for that specific stack test. If any other value is proposed, MPCA review will be required.

It should be emphasized that the stack test data need to reflect the emissions unit under consideration or a similar emission unit(s). If the emission unit has been modified, previously generated test data may not be appropriate.

Emission rate calculation

An emission rate is calculated as follows:

[Emission rate = Process rate * Emission factor * (1— control efficiency)]

Note: This is a general expression, and therefore other parameters may be used.

Refining emissions estimates

If the total estimated risks generated by the AERA are greater than facility risk guidelines, a project proposer may choose to refine the emission rates for emission sources that contribute to the largest fraction of estimated risk. For example, the composition estimate of a coating might be specified more narrowly, or a stack test might be performed to reflect control efficiency or variability. In this way, an iterative method of estimating emissions based on the AERA results is used.

When emissions data are not available

Reasonable effort needs to be expended to identify sources of emissions data. If no data are available for an air toxic, and it is unreasonable to make a comparison to a similar type of air emissions source, the project proposer will not be asked to provide emissions data. However, all attempts made to identify emission factors (e.g., list database consulted, literature reviewed, internet searches, industry databases, personal interviews with experts) must be described.

Special air toxics considerations

The MPCA has found that several air toxics require careful treatment in estimating emissions. Guidance for many of these air toxics is provided here.

Diesel exhaust particulates

If available, particulate emission rates of PM2.5 need to be included in the RASS as “diesel exhaust particulate” for diesel-fired combustion sources; if not available, PM10 data needs to be used. This is an analysis of particulate exposure that is separate from analyses to predict impacts for comparison to the NAAQS, and is used to estimate the potential non-cancer health effects from long term diesel exposure.

Individual chemical constituents of diesel exhaust emissions are estimated from the emissions information sources cited in Table 1. These are calculated in order to estimate the potential carcinogenic health effects from long term diesel exposure.

Mercury

Form Microsoft Office document icon HG-01: Mercury Releases to Ambient Air is an editable spreadsheet and used to report mercury emissions and their calculations to the MPCA. Total mercury emissions need to be speciated into particulate bound, reactive gaseous and elemental mercury. The speciation of mercury is process specific and needs to follow this hierarchy of data sources:

  1. Facility or industry stack testing using the most current U.S. EPA method
  2. Methodology described in technical support document of the National Air Toxics Assessment or the U.S. EPA National Emissions Inventory.
  3. Default speciation of 20%-elemental, 60% particulate, and 20% oxidized (USEPA HHRAP, 2005).