Emission Rate Estimating and Data Quality

9. Emission rate estimating and data quality

Section 11 of the Regulation sets out the requirements for estimating and refining emission rates. The ESDM report must document both the facility operating conditions (that are relevant to estimating emission rates based on the averaging time of the standard or ministry POI limit) and the development of emission rate estimates for the significant air contaminants discharged from the facility. In general, section 12 sets out the requirements to “refine” emission rate estimates when the combined effect of sections 10 and 11 result in predictions of exceedences of ministry POI Limits. Also note that section 12 of the Regulation allows a person to proceed directly to abatement as an alternative to refinement. Under section 28 of the Regulation, there are notification requirements for exceedences of ministry POI Limits once refinement has been completed. However, if a facility determines that refinement will still likely show an exceedence of a ministry POI Limit, they may choose to focus their efforts on abatement (as opposed to refinement) - see subsection 12 (3) of the Regulation. Note that if this option is chosen the person must notify the ministry of an exceedence under section 28 and submit an abatement plan as per section 29. Please note that there are additional notification requirements under sections 25 and 30 of the Regulation that are also required before refinement is completed. For more information on notification requirements, see Chapter 2.3 Updates to ESDM Reports, chapter 12.3 Upper Risk Thresholds, and Chapter 13 Other Considerations That Require Follow Up Actions.

Refinement of emission rates may include source testing across a range of operating conditions or the ability to assess emissions in accordance with a plan approved by the director under paragraph 3 of subsection 11 (1) of the Regulation. These plans are typically referred to as a Combined Assessment of Modelled and Monitored Results (CAMM). In summary, sections 10, 11 and 12 of the Regulation work together to ensure that the assessment of maximum POI concentration are as accurate as possible and do not under-estimate actual concentrations.

9.1 Emission rate and estimation techniques

Subsection 11 (1) of the Regulation states:

Paragraph 1 of subsection11(1) of the Regulation states that an approved dispersion model must be used with an emission rate that is “at least as high as the maximum emission rate that the source of contaminant is reasonably capable of”. This section of the Regulation ensures that the emission rates estimates that are used by a facility in its initial assessment are conservative. If a facility can show it meets the air standards at the maximum POI using conservative emission rate estimates, then no further assessment is required for those contaminants.

If this initial assessment shows the standard (or ministry POI Limit) may be exceeded, the Regulation requires a facility either to “refine” its assessment or to take steps to come into compliance or meet the ministry POI Limit. Paragraphs 2 and 3 of subsection11 (1) of the Regulation describe ways to refine the assessment by using other less conservative but more accurate methods to assess emission rates including:

• An emission rate that, for the relevant averaging period, is derived from site- specific testing of the source of contaminant across a range of operating conditions (paragraph 2 of subsection 11 (1) of the Regulation); and/or
• An emission rate that, for the relevant averaging period, is derived from a combination of a method that complies with paragraph 1 or 2 of subsection 11 (1) and ambient monitoring, according to a plan approved by the Director as likely to provide an accurate reflection of emissions (paragraph 3 of subsection 11 (1) of the Regulation).

Paragraph 3 of subsection 11 (1) of the Regulation typically refers to the Combined Assessment of Modelled and Monitored Results (CAMM) as an emission rate refinement tool. For more information, please refer to the Technical Bulletin on CAMM as an Emission Rate Refinement Tool, as amended from time to time.

Emissions rates are directly linked to operating conditions at the facility. Subsection 10 (1) paragraph 1 of the Regulation states that a facility must use:

A scenario that, for the relevant averaging period, assumes operating conditions for the facility that would result in the highest concentration of the contaminant at a point of impingement that the facility is capable of.

In summary, the emission rate estimating must be either:

• “conservative”^{footnote 14[14]}, as represented by paragraph 1 of subsection 11 (1); or
• as accurate as possible, as represented by the methodologies set out in paragraphs 2 and 3 of subsection 11 (1).

The concepts of accuracy and conservatism in emission rate estimating work together. For example, the need for conservatism in the estimating technique should be inversely proportional to the degree of accuracy of the technique. The greater the accuracy, the less there is a need for conservatism in the emission rate estimating method.

In many cases, emission rate estimating is an iterative process where estimates start out conservative and are then refined to be more accurate and less conservative when earlier iterations result in an exceedence of a ministry POI Limit. Although the emission rate estimating methodologies described in paragraph 2 and 3 of subsection 11 (1) of the Regulation can be selected at any time, they also represent the end of the iterative or refinement process.

Paragraph 7 of subsection 26 (1) of the Regulation requires that an ESDM report contain the following information with respect to emission rates:

Chapters 9.1 Emission Rate and Estimation Techniques of this Procedure Document provides information on four basic methods to estimate air emission rates and provide guidance on developing emission rates that are, in most cases, consistent with paragraph 1 of subsection 11 (1) of the Regulation.

See Chapter 9.2 Data Quality for guidance on the requirement to assess the accuracy of the emission rate estimating methodologies. Also see Chapter 9.3 Refinement of Emission Rates for additional guidance on the requirements.

9.1.1 Emission factors

Emission factors are typically constants (usually expressed in mass emission per unit of production or mass of raw material input) which are applied to a process parameter or production rate to generate an emission rate. The most commonly used emission factors are those published by the United States Environmental Protection Agency (US EPA). The US EPA and other regulatory agencies use data from past source testing campaigns to develop emission factors for a variety of industrial processes. When well-documented emission factors (e.g., such as the factors presented in US EPA guidance) are applied to a source that they were meant for; and applied in a manner that will result in a conservative estimate of emission rates then the use of emission factors is, in most cases, consistent with paragraph 1 of subsection 11 (1) of the Regulation.

9.1.2 Mass balance calculations

A mass balance is an accounting of the material that enters and leaves from a process or reaction. Emissions resulting from many processes, such as painting or surface-coating, are a direct result (i.e., no chemical transformation through the process) of the raw material inputs to the processes. In these cases, emission rates can be estimated from the through-put of raw materials (e.g., paint and solvent usage).

Typically, mass balance calculations that assume 100% of material used is emitted to the air are considered to be conservative. However, a follow-up “field verification” of material usage may be required by the ministry depending on:

• the type of contaminants emitted;
• the magnitude of the POI concentrations relative to the relevant ministry POI Limit; and
• the presence of other evidence (e.g., odour complaints) of adverse effect.

In most cases, the use of a mass balance approach is consistent with paragraph 1 of subsection 11 (1) of the Regulation as long as the mass balance calculations are: (i) well documented (e.g., calculations are provided); (ii) used for emissions that are not chemically transformed through the process; and (iii) account for all discharges/flows of the contaminant used in the process.

9.1.3 Source tests

Source tests may be conducted at the facility to measure contaminant emission rates, required to prepare the ESDM report. Source testing is to be conducted following the requirements outlined in the Ontario Source Testing Code. Emission rates may also be based on source tests conducted on another similar process provided that the similarity of the process can be established. Emission rate estimates that are based on validated source tests are another acceptable approach to estimating emission rates. In order for the ministry to accept source test results, without further analysis, source tests should be validated by the ministry. A ministry validated source test means:

• source testing methods were submitted to the ministry in a pre-test plan;
• the ministry accepted the pre-test plan;
• the ministry had an opportunity to witness the test; and
• the final report was submitted to the ministry. (Note: if a ministry review of the report indicates that the source test results may not be accurate, the data should not be used)

Source test results that have had similar over-sight from other regulatory agencies may also be considered validated. Proof of this validation must be documented in the ESDM report, and shall be comparable to the validation typically undertaken by the ministry.

Results from un-validated source testing may be used to estimate emission rates but the quality of the estimates may be considered to be classified as uncertain. The quality of un-validated source testing can be improved through a review and comparison to other validated sources of information. For example, an executive summary that includes a summary of the sampling and analytical methodology; the name of the persons responsible for the testing; the process operating conditions that the tests were conducted under; the dates of testing; and a full test report that is available for review upon request. This may demonstrate that the un-validated data is in the range of expected emission rates for the facility. It is also possible to demonstrate this through partial post-testing validation. See Chapter 9.2 Data Quality for further information.

In summary, the use of source testing to estimate emission rates is, in most cases, acceptable under paragraph 1 (not paragraph 2) of subsection 11(1) of the Regulation when:

• source tests are validated;
• the source testing represents one specific operating condition; or
• sources tests are not validated but the results are demonstrated to be comparable to other validated emission rate information and are well documented.

When estimating emission rates according to paragraph 2 of subsection 11(1) of the Regulation, source testing is required to be conducted comprehensively across a full range of operating conditions. The range of operating conditions must be approved by the ministry before source testing is conducted. Testing must be conducted in accordance with a plan approved by the Director. The Director must also be given at least 15 days written notice of the source testing and an opportunity to witness the testing, and the results of the testing must be approved by the Director. For more information, see Chapter 9.2.1 “Highest Data Quality” Emission rate estimating Techniques.

9.1.4 Engineering calculations

Emission rates can also be developed from fundamental scientific principles and measurements. In this case, an engineering estimate can be based on operating conditions, data from the literature, thermodynamic and physical properties. Included in this category of emission rate estimates are calculations based on direct source measurements that are neither considered validated source tests nor comparable to validated results, as noted above.

The use of derived formulae (sometimes in combination with accepted empirical data) is also an acceptable emission rate estimating method as long as the approach is demonstrated through documentation and references to be based upon sound scientific and engineering principles. When engineering calculations are based upon fundamental scientific principles or based upon a recognized empirical relationship (e.g., the ministry Evaporation Equation in Appendix C); are well documented (e.g., calculations are provided); then the use of engineering calculations is, in most cases, consistent with paragraph 1 of subsection 11 (1) of the Regulation.

See Appendix C for further guidance and reference material to assist with estimating emission rates.

9.2 Data quality

Every emission rate estimate must include some quantification or qualification of the uncertainty of the estimation. For the purposes of this Procedure Document, the data qualification is referred to as Data Quality (i.e., the higher the quality of the data, the higher the accuracy and certainty of the emission rates and therefore assessment of POI concentrations).

This assessment of accuracy is required by subparagraph 7 iii of subsection 26 (1) of the Regulation and is intended to avoid any underestimations that may result in a false prediction of compliance.

A “Data Quality” classification system, that provides an indication of the accuracy of the emission rate estimating methods used in the ESDM report, is set out in Chapter 9.2 Data Quality of this Procedure Document. The basic concepts of the Data Quality classification system are: (i) the accuracy of the emission rate estimate is directly proportional to the level of Data Quality and (ii) the “conservatism” of the emission rate estimate generally decreases with increasing Data Quality.

This Data Quality classification system may generally be used to assess the accuracy and conservatism of the emission rate estimating method, consistent with paragraph 7 iii of subsection 26 (1) of the Regulation. However, on a case-by-case basis, the ministry may request a more comprehensive assessment of the accuracy and the likelihood of an underestimated emission rate. An assessment of the significance of the error in the estimate may also be required by the ministry. In addition, when there is sufficient data to quantify the range of expected emission rates (e.g., in the case of data from a continuous emission monitoring system) then it should be included in an ESDM report.

Example to illustrate concept of data quality (for emission rate estimates that satisfy para. 1 of s.11 (1) of the Regulation)

In one situation, there are two different emission rate estimates (one higher than the other) where both emission rate estimates have the same data quality. Only the higher of the two emission rate estimates in this example can satisfy the requirements of paragraph 1 of subsection 11(1) of the Regulation.

In another situation, there are two different emission rate estimates where the lower emission rate estimate has a higher data quality than the greater emission rate estimate. In this case, the lower emission rate estimate would satisfy paragraph 1 of subsection 11(1) of the Regulation since it has higher data quality.

9.2.1 “Highest Data Quality” emission rate estimating techniques

• Combined Assessment of Modelled and Monitored Results (CAMM): The highest quality data is provided by site-specific emission rate estimating and air dispersion modelling in combination with ambient monitoring of contaminants using an approach that has been validated by the ministry. This type of approach is typically used for fugitive emissions. The ministry Technical Bulletin: “Combined Assessment of Modelled and Monitored Results (CAMM) as an Emission Rate Refinement Tool” provides further guidance with respect to an acceptable plan under paragraph 3 of subsection 11 (1) of the Regulation. The guidance provided in the Technical Bulletin functions to ensure that the emission rates used in an approved dispersion model are “as accurate as possible”. A form is available on the ministry website to request approval of the modelling-monitoring plan (PIBS # 6323e) (as amended). (See also Chapter 9.3 on Refinement of Emission Rates).

• Comprehensive and Validated Source Testing Across a Full Range of Operating Conditions:The highest quality estimates of emission rates for point sources are derived from comprehensive equipment-specific testing that have been validated by the ministry or a regulatory agency acceptable to ministry. This approach involves source tests conducted to measure contaminant emission rates over a range of operating conditions where process data was available and recorded; and results in a statistically significant data set. The use of validated continuous emission monitoring systems also represents “Highest Data Quality”. This category is consistent with subsection 11 (1) paragraph 2 of the Regulation. (See also Chapter 9.3 on Refinement of Emission Rates).

• Note: If a source testing plan and notification to witness a source test is being submitted under paragraph 2 of subsection 11(1), this must be clearly identified in the submission. Facilities that are using paragraph 2 of subsection 11(1) of the Regulation for refinement of emissions must contact the ministry prior to source testing in order to confirm the range of operating conditions to be source tested for the purposes of “refinement”.

• Mass Balance: A mass balance technique may be considered to provide a Highest Data Quality estimate if:

  • 100% of the material balance is accounted for (e.g., 100% emitted to air);
  • it is reasonable to assume that the contaminants will not undergo a chemical transformation through the source/process;
  • the usage averaging period is similar to the averaging period for the air quality standard; and
  • the material usage information has been validated (e.g., through purchase records that are provided to the ministry upon request).

9.2.2 “Above-Average Data Quality” emission rate estimating techniques

• Validated Source Testing at One Specific Operating Condition: Emission rate estimates that are derived from validated source testing at one specific operating condition are anticipated to provide above-average quality of estimate of emission rates and POI concentrations.
• Mass Balance: A mass balance technique can be considered to provide emission rate estimates of above-average quality if:
  • 100% of the material balance is accounted for (e.g., 100% emitted to air);
  • it is reasonable to assume that the contaminants will not undergo a chemical transformation through the source/process; and
  • the usage averaging period is similar to the averaging period for the air quality standard.
• Emission Factors: Emission rate estimates that are developed from tests on a moderate to large number of sources where the source category population is sufficiently specific to minimize variability (e.g., US EPA, AP-42, emission factor quality rating of A or B) are anticipated to provide above-average quality of emission rate estimates.
• Engineering Calculations/Judgement: Emission rate estimates derived from fundamental scientific and engineering principles; and/or relevant empirical data can be considered above-average quality estimates if it is clear (e.g., the approach is recommended through ministry documentation) that the estimating technique will result in relatively conservative predictions.

9.2.3 “Average Data Quality” emission rate estimating techniques

• Emission Factors: Emission rate estimates that are developed from tests on a reasonable number of facilities where the source category population is sufficiently specific to minimize variability (e.g., US EPA, AP-42, emission factor quality rating of C) are anticipated to provide average quality emission rate estimates.
• Engineering Calculations/Judgement: Emission rate estimates derived from fundamental scientific and engineering principles; and/or relevant empirical data can be considered average data quality estimates.
• Partially Validated Source Testing at One Specific Operating Condition: emission rate estimates that are from source testing where the testing has only been partially validated (e.g., pre-test plan approval or post-testing assessment and documentation, by a regulatory agency, of the quality of the final source testing report) at a specific operating condition. These tests are anticipated to provide an average data quality estimate of emission rates.

9.2.4 “Marginal” or “Uncertain Data Quality” emission rate estimating techniques

• Un-Validated Source Testing at One Specific Operating Condition: emission rate estimates that are from un-validated source testing are considered to be of Uncertain Data Quality. The quality of un-validated source testing results may be up-graded to “Average Data Quality” through post-testing assessment (if enough information is available) by a regulatory agency or by comparison of the test results to alternative sources of validated information (e.g., validated source testing on similar equipment and/or comparison to emission factors of at least average data quality).
• Emission Factors: Emission rate estimates that are developed from tests on only a small number of facilities where there is evidence of variability within the source category population (e.g., US EPA, AP-42, emission factor quality rating of D or E) and/or the emission factor rating is uncertain are considered to have uncertain data quality.
• Calculations/Judgement: Emission rate estimates derived from calculations where the scientific/technical integrity of the approach is uncertain are considered to have uncertain data quality.

In many cases, the use of emission rate estimating methodologies that are classified as Marginal or Uncertain Data Quality may be the only method available to the facility. Where the maximum POI concentration from the facility is less than 10% of the ministry POI Limit, emission rate estimates of Marginal or Uncertain Data Quality, may be adequate. In most cases, where POI concentrations are more significant, emission rate estimates based on Marginal or Uncertain Data Quality may also be considered acceptable provided these emission rate estimates have been altered to be sufficiently conservative.

Depending on the nature of the contaminant and the magnitude of the POI concentrations, emission rate estimates that are identified as having Marginal or Uncertain Data Quality may warrant development of a range of estimated emission rates and POI concentrations and/or follow-up development of source-specific emission studies. In some cases, implementation of air pollution abatement may also be a reasonable requirement where Data Quality is Marginal or Uncertain and the likelihood and extent of exceedence of an air quality standard is high.

In the event that initial estimates of maximum POI concentration are above the ministry POI Limits then the above-noted Data Quality classification system also informs “refinement” to a higher level of accuracy in the predictions, as required by section 12 of the Regulation and as explained in Chapter 9.3 Refinement of Emission Rates.

9.3 Refinement of emission rates

Development of emission rates and resulting POI concentrations can, in many cases, be an iterative process. Section 12 of the Regulation will determine the need for further refinement of emission rates. This section is entitled “Combined effect of assumptions for operating conditions and emission rates” and states:

If the results of an approved dispersion model indicate that discharges of a contaminant may result in an exceedence of an air quality standard or an adverse effect^{footnote 15[15]} then the modelling exercise is not complete. In order to run an approved dispersion model in accordance with the Regulation, section 12 must be complied with. Section 12 generally states that where an exceedence or adverse effect is indicated, one of the following requirements must be automatically completed:

1. Refinement: the approved dispersion model shall be used in conjunction with actual operating data in accordance with a scenario described in paragraph 2 of subsection 10 (1) and emission rate estimating methods set out in paragraphs 3 in subsection 11(1) of the Regulation (i.e., highest data quality emission rate estimates) (unless the Director accepts a submission under paragraphs 2 in subsection 11(1) of the Regulation), or
2. Abatement: submit an abatement plan (see s.12 (3) and s.29 of the Regulation) that sets out the steps for achieving compliance with the POI Limits to the ministry.

In summary, a conservative^{footnote 16[16]} emission rate estimating technique may be used if the ESDM report shows compliance with the ministry POI Limit. If, however, an ESDM report identifies an exceedence of a ministry POI Limit then the facility must either: (i) develop and implement an abatement plan that will achieve compliance with the Limit; and/or (ii) refine the ESDM report in accordance with section 12^{footnote 17[17] footnote 18[18]} of the Regulation. Note that the end-point for ESDM report refinement is the highest data quality obtained from a modelling-monitoring analysis. If refinement shows non-compliance, then abatement action will be required by the ministry.

The ministry expects that a facility date the results of modelling runs and that re-modelling using more accurate inputs be completed within approximately six months of a more conservative modelling run that indicates an exceedence. Where ambient monitoring or comprehensive source testing is required to refine the modelling, it is acknowledged that more than six months may be required. Where there is a more serious risk to human health or the re-modelling exercise is straightforward, it is expected that the refinement be completed in less than six months and as soon as possible.

Table 9-1: Overview of the section 26 ESDM report Requirements to Document Operating Conditions, Emission Rate Estimates and Data Quality

Emission rates and point of impingement concentrations at the maximum operating condition

According to paragraph 6 of subsection 26 (1) of the Regulation, each significant contaminant must have a description of the operating conditions that result in the maximum POI concentration (either that the facility is capable of or based on the actual operating conditions for the time period described in paragraph 2 of section 10 (1) of the Regulation), according to the averaging period for the relevant ministry POI Limit. This description of the operating condition for each significant contaminant should:

• identify each significant source (or group of sources) of the contaminant (including the specific modelled source name);
• describe the operating conditions of the significant sources that result in the maximum POI concentration for the contaminant, according to the averaging period for the ministry POI Limit; and
• be based upon information contained in the Facility Description section of the ESDM report with consideration for simultaneous versus sequential operations and emissions; design capacities; technical and operational limitations on production and the terms and conditions of ECAs issued to the facility.

Explanation of the emission rate estimating methodology and sample calculation (see section 11 of the Regulation)

According to subparagraph 7 i of subsection 26 (1) of the Regulation, an explanation of the method used to estimate emission rates must be included for each significant contaminant. The description should include a brief section for each significant source or source category. Any appropriate references should also be provided (e.g., emission factor document title and relevant information to assist a reviewer in obtaining a copy). If source testing is used to develop emission rate estimates then copies of an executive summary from the source testing report should be provided with full copies of the report available upon request. As per subparagraph 7 ii of subsection 26 (1), a sample calculation that illustrates each emission rate estimating method used should also be provided.

Assessment of accuracy/data quality

According to subparagraph 7 iii of subsection 26 (1) of the Regulation, an assessment must be provided for each significant contaminant, on the accuracy of each emission rate estimating method including an assessment of whether the method is more likely to overestimate or underestimate and an assessment of how significant this error might be. In most cases, the assessments required by subparagraph 7 iii of subsection 26 (1) of the Regulation can be satisfied by use of the Data Quality classification system described in Chapters 9.2.1 through 9.2.4 of this Procedure Document. This Data Quality classification system may generally be used to assess the accuracy and conservatism of the emission rate estimating method.