1. Mercury Monitoring and Reporting Requirements under the MATS Rule
EPRI Conference---Louisville, KY
May 2012
Matthew Boze & Robert Vollaro
USEPA, CAMD

2. Background
On February 16, 2012, EPA published the Mercury and Toxics Standards (MATS) rule (40 CFR 63, Subpart UUUUU), establishing national emissions limitations and work practice standards for certain hazardous air pollutants (HAP) emitted from coal-fired and oil-fired electric utility steam generating units.
Compliance with the rule is required:
For existing units, by April 16, 2015
For new or reconstructed units, by April 16, 2012, or upon startup (first-fire), whichever is later
Compliance must be demonstrated, by means of performance testing, no later than 180 days after the above dates
Mercury (Hg) is a regulated HAP under the rule.
For existing units, the rule specifies both heat input-based emission limits (lb/TBtu) and electrical output-based limits (lb/GWh) for Hg.

3. Background (cont’d)
Compliance with either Hg limit is acceptable for existing EGUs.
For new or reconstructed units, only electrical output-based emission limits are specified.
The magnitude of the applicable Hg emission limit depends on which “subcategory” the EGU is in. The general subcategories and applicable Hg limits (on a 30-day rolling average basis) are:
Coal-fired units designed to burn coal with GCV ≥ 8,300 Btu/lb
1.2 lb/TBtu or 1.3 x 10-2 lb/GWh for existing units
2.0 x 10-4 lb/GWh for new or reconstructed units
Coal-fired units designed to burn low-rank virgin coal (GCV < 8,300 Btu/lb)
4.0 lb/TBtu or 4.0 x 10-2 lb/GWh for existing units
4.0 x 10-2 lb/GWh for new or reconstructed units
IGCC units
2.5 lb/TBtu or 3.0 x 10-2 lb/GWh for existing units
3.0 x 10-3 lb/GWh for new or reconstructed units

4. Background (cont’d)
Liquid oil-fired units
2.0 x 10-1 lb/TBtu or 2.0 x 10-3 lb/GWh for existing units (continental)
1.0 x 10-4 lb/GWh for new or reconstructed units (continental)
4.0 x 10-2 lb/TBtu or 4.0 x 10-4 lb/GWh for existing units (non-continental)
4.0 x 10-4 lb/GWh for new or reconstructed units (non-continental)
Solid oil-derived fuel (i.e., petroleum coke)-fired units
2.0 x 10-1 lb/TBtu or 2.0 x lb/GWh for existing units
2.0 x 10-3 lb/GWh for new or reconstructed units
Limited-use units in the liquid oil-fired units subcategory (< 8% annual capacity factor, over a 2-year period) are not subject to emissions limits for mercury
Existing units in the same subcategory at the same facility may use emissions averaging for Hg.
If the units in an averaging group are designed to burn coal with GCV ≥ 8,300 Btu/lb, there are alternate Hg limits for the group (1.0 lb/TBtu or 1.1 x 10-2 lb/GWh on a 90-day rolling average).

5. Hg Compliance Options
Certain units can comply with the rule by qualifying as “Low-Emitting EGUs” (LEEs). This option may not be used for units with main stack and bypass stack configurations.
For Hg, a LEE is a unit that:
Emits at less than 10% of the applicable emissions limit; or
Has the potential to emit ≤ 29.0 lb of Hg per year
To qualify for LEE status, a demonstration test is required.
EPA Method 30B is used to measure Hg concentration for 30 unit operating days
Paired sorbent traps are used for the test.
Each pair of traps may be used for up to 10 operating days
Additional data (i.e., stack gas flow rate, CO2 or O2 concentration, moisture, electrical load, as applicable) are collected during the test. EPA reference methods or certified CEMS that meet Part 75 requirements are used to provide the flow rate, CO2, O2, or moisture data.
The additional data are used either to: (1) convert the measured Hg concentrations to the units of the emission limit; or (2) enable the EGU’s potential annual Hg mass emissions to be calculated.
The rule includes special provisions for applying the LEE methodology to units that share a common stack and units that have multiple stack (or duct) exhaust configurations.
The 30 operating day demonstration test must be repeated at least once every 12 months, to show that the unit continues to qualify as a LEE for Hg.

6. Hg Compliance Options (cont’d)
What are the compliance requirements for EGUs that do not qualify as LEEs ?
A liquid oil-fired unit that does not qualify as a LEE for total HAP metals (including Hg), individual HAP metals (including Hg), or filterable PM must demonstrate compliance in accordance with § (c)(2)(ii), either by:
Conducting an initial performance test and subsequent quarterly stack tests to show that the filterable PM emissions limit is met;
Using a certified PM CEMS to demonstrate initial and on-going compliance with the filterable PM emission limit; or
Conducting an initial performance test and subsequent annual stack tests to show that the filterable PM emissions limit is met, and using a PM CPMS to monitor continuous performance (with respect to an operating limit) in-between the annual stack tests.
A coal-fired unit, IGCC , or solid oil-derived fuel (i.e., petroleum coke)-fired unit that does not qualify as a LEE for Hg must continuously monitor Hg concentration using Hg CEMS or sorbent trap monitoring systems, in accordance with Appendix A of Subpart UUUUU.

7. Hg CEMS Principle: Typical Hg CEMS Components:
Representative samples of flue gas are continuously extracted from the stack or duct
Particulate matter is removed (i.e., filtered) from the gas samples
Hg analyzer measures vapor phase Hg0 and Hg2+ (simultaneously or separately)
Typical Hg CEMS Components:
Probe
PM filter
Sample conditioning
Sample transport (sample line and pump)
Gas analyzer (with or without converters)
Calibration gas system

8. Typical Hg CEMS

9. Hg CEMS (cont’d) Types of systems:
Dilution extractive – wet basis
Extractive – dry basis
Development of a quality assurance/quality control (QA/QC) program is required.

10. Sorbent Trap Monitoring Systems
Principle:
Known volumes of flue gas are continuously extracted from the stack or duct through paired, in-stack sorbent traps. Vapor phase Hg is collected on the sorbent medium.
Typical sorbent medium is halogenated carbon
Paired sorbent traps are used for quality-assurance purposes and to ensure measurement precision
A pair of sorbent traps is typically used for 24 to 168 hours (but may be used for up to 14 operating days when the Hg concentration is very low), before being removed and analyzed.
Components:
Dual trains w/sampling probes and paired sorbent traps. The sorbent traps have three sections: (1) main sample collection section; (2) breakthrough section; and (3) section spiked with a known mass of Hg0 , for QA/QC .
Moisture removal system
Vacuum pump
Sample gas flow meter

11. Typical Sorbent Trap Monitoring System---PS 12B

12. Sorbent Trap Monitoring Systems (cont’d)
Probe
3-section
sorbent traps

13. Sorbent Trap Monitoring Systems (cont’d)
The sorbent trap system must be operated and maintained according to Performance Specification 12B in 40 CFR Part 60, Appendix B.
Leak checks must be performed (before and after sampling)
Paired sorbent trap consistency/agreement is required
A minimum % recovery of the spiked Hg0 in the third section is required
Sample recovery techniques include acid leaching, digestion, thermal desorption (see next slide).
Analytical methods include UV atomic fluorescence, UV atomic absorption, and X-ray fluorescence
Analyzer calibration is required
Development of a QA/QC program is also required.

14. Thermal Desorption System
Thermal desorption system allows rapid, on-site measurement of mercury in sorbent traps
Alternative to traditional wet chemical analysis methods

15. Monitoring System Certification
Each Hg CEMS or sorbent trap monitoring system must be initially certified.
For Hg CEMS, the following certification tests are required:
7-day calibration error test, using elemental or oxidized Hg standards
Linearity check, using elemental Hg standards
3-level system integrity check, using oxidized Hg standards
Cycle time test, using elemental or oxidized Hg standards. This test is not required for integrated batch sampling-type CEMS
Relative accuracy test audit (RATA). Note that a bias test is not required. Allowable reference methods for the RATA include EPA Methods 29, 30A, 30B, and ASTM D (the Ontario Hydro Method)
All Hg calibration gas standards must be traceable to the National Institute of Standards and Technology (NIST). EPA , in cooperation with industry, NIST, and outside consultants, has developed interim traceability protocols. These are cited in Subpart UUUUU, Appendix A, sections and

16. Monitoring System Certification (cont’d)
For sorbent trap monitoring systems, only a RATA is required.
During the RATA, the monitoring system must be operated according to Performance Specification 12B
The sorbent material must be the same type that is used for daily operation of the system
The sorbent traps used for the RATA may be smaller than those used for daily operation
The allowable breakthrough in the second section of the traps depends on the Hg concentration level in the stack gas (see Footnote 1 in Table 3, below).
The required tests and performance specifications that must be met for initial certification of Hg CEMS are summarized in Table 1.
The RATA specifications for the initial certification of a sorbent trap monitoring system are the same as those shown in Table 1.

17. Monitoring System Certification (cont’d)
Table 1: Certification Requirements for Hg CEMS
1 “R” is the reference gas value; “A” is the analyzer reading; “Aavg” is the average analyzer reading; “RMavg” is the mean reference method value; and “Cavg” is the mean CEMS value
2 Using NIST-traceable elemental or oxidized Hg standards. The test may be performed on 7 consecutive operating days, if necessary.
3 Using NIST-traceable elemental Hg standards
4 Using NIST-traceable oxidized Hg standards
5 Using NIST-traceable elemental or oxidized Hg standards. This test is not required for integrated batch sampling-type CEMS.
For this required certification test.....
The main performance specification1 is.....
The alternate performance specification1 is.....
And the conditions of the alternate specification are.....
7-day calibration error test2
|R - A| ≤ 5.0% of span value, for both the zero and upscale gases, on each of the 7 days
|R - A| ≤1.0 µg/scm
The alternate specification may be used on any day of the test.
Linearity check3
|R - Aavg | ≤ 10.0% of the reference gas concentration at each calibration gas level
|R - Aavg | ≤ 0.8 µg/scm
The alternate specification may be used at any gas level
3-level system integrity check4
|R - Aavg | ≤ 10.0% of the reference gas concentration at each calibration gas level
RATA
20.0% RA
|RMavg - Cavg | ≤ 1.0 µg/scm
RMavg < 5.0 µg/scm
Cycle time test5
15 minutes

18. On-going Quality-Assurance
To ensure that certified Hg CEMS and sorbent trap monitoring systems continue to provide accurate data, the systems are subject to on-going quality assurance and quality control (QA/QC) requirements.
For Hg CEMS, the following QA tests are required:
Daily calibration error tests , using elemental or oxidized Hg standards
Weekly single-point system integrity checks , using oxidized Hg standards
Quarterly linearity checks, using elemental Hg standards or 3-level system integrity checks using oxidized Hg standards
Annual RATAs
For sorbent trap monitoring systems:
The system must be operated and maintained according to PS 12B
Annual RATAs are required
Quarterly calibration of the gas flow meter and auxiliary equipment is required
The on-going QA/QC requirements and acceptance criteria for Hg CEMS and sorbent trap monitoring systems are summarized in Tables 2 and 3, respectively.

19. On-going Quality Assurance (cont’d)
Table 2: On-going QA /QC Requirements for Hg CEMS
1 Weekly means once every 7 operating days
2 NIST-traceable Hg standards are required
Perform this type of QA test....
At this frequency....
With these qualifications and exceptions....
Acceptance criteria….
Calibration error test
Daily
Use either a mid- or high- level gas
Use either elemental or oxidized Hg2
Calibrations are not required when the unit is not in operation.
|R - A| ≤ 5.0% of span value or
|R - A| ≤ 1.0µg/scm
Single-level system integrity check
Weekly1
Use oxidized Hg ---either mid- or high-level
Not required if daily calibrations are done with a NIST-traceable source of oxidized Hg |
R - Aavg| ≤ 10.0% of the reference gas value
|R - Aavg| ≤ 0.8 µg/scm

20. On-going Quality Assurance (cont’d)
Table 2: On-going QA /QC Requirements for Hg CEMS (cont’d)
3 A “QA operating quarter” is a calendar quarter with at least 168 unit or stack operating hours
4 NIST-traceable Hg standards are required
5 “Quarterly” means once every QA operating quarter
6 Annual” means once every four QA operating quarters
Perform this QA test....
At this frequency…
With these qualifications and exceptions....
Acceptance criteria….
Linearity check or
3-level system integrity check
Quarterly5
Required in each “QA operating quarter”3 ---and no less than once every 4 calendar quarters
168 operating hour grace period available
Use elemental Hg for linearity check4
Use oxidized Hg for system integrity check4
|R - Aavg | ≤ 10.0% of the reference gas value, at each calibration gas level
|R - Aavg| ≤ 0.8 µg/scm
RATA
Annual6
Test deadline may be extended for “non-QA operating quarters”, up to a maximum of 8 quarters from the quarter of the previous test.
720 operating hour grace period available
20.0% RA
|RMavg - Cavg | ≤ 1.0 µg/scm, if
RMavg < 5.0 µg/scm

21. On-going Quality Assurance (cont’d)
Table 3: Normal Operation and QA/QC for Sorbent Trap Systems (PS 12B)
1 For the RATA, Section 2 breakthrough depends on CHg stack gas Hg concentration. Allowable breakthrough is ≤ 10% of Section 1 mass if CHg is > 1 µg/m3; ≤ 20% of Section 1 mass if CHg is > 0.5 and ≤ 1 µg/m3; ≤ 50% of Section 1 mass if CHg is > 0.1 and ≤ 0.5 µg/m3. There is no breakthrough criterion if CHg is < 0.1 µg/m3 .
QA/QC test or specification
Acceptance criteria
Frequency
Consequences if not met
Pre-monitoring leak check
≤ 4% of target sampling rate
Prior to monitoring
Monitoring must not commence until the leak check is passed.
Post-monitoring leak check
≤ 4% of average sampling rate
After monitoring
Invalidate the data from the paired traps or, if certain conditions are met, report adjusted data from a single trap (PS 12B, Section ).
Ratio of stack gas flow rate to sample flow rate
Hourly ratio may not deviate from the reference ratio by more than ± 25%.
Every hour throughout monitoring period
Sorbent trap section 2 breakthrough
≤ 5% of Section 1 Hg mass
(for daily operation)1
Every sample

22. On-going Quality Assurance (cont’d)
Table 3: Normal Operation and QA/QC for Sorbent Trap Systems (PS 12B) (cont’d)
QA/QC test or specification
Acceptance criteria
Frequency
Consequences if not met
Paired sorbent trap agreement
≤ 10% Relative Deviation (RD) if the average concentration is > 1.0 µg/m3
Every sample
Either invalidate the data from the paired traps or report the results from the trap with the higher Hg concentration.
≤ 20% RD if the average concentration is ≤ 1.0 µg/m3
Results also acceptable if difference between paired traps is ≤ 0.03 µg/m3
Spike Recovery Study
Average recovery between 85% and 115% for each of the 3 spike concentration levels
Prior to analyzing field samples and prior to use of new sorbent media
Field samples must not be analyzed until the percent recovery criteria has been met.
Multipoint analyzer calibration
Each analyzer reading within ± 10% of true value and r2 ≥ 0.99
On the day of analysis, before analyzing any samples
Recalibrate until successful
Analysis of independent calibration standard.
Within ± 10% of true value
Following daily calibration, prior to analyzing field samples
Recalibrate and repeat independent standard analysis until successful.
Spike recovery from section 3 of both sorbent traps
75–125% of spike amount
Invalidate the data from the paired traps or, if certain conditions are met, report adjusted data from a single trap (PS 12B, Section ).

23. On-going Quality Assurance (cont’d)
Table 3: Normal Operation and QA/QC for Sorbent Trap Systems (PS 12B) (cont’d) 2 “Annually” means once every four QA operating quarters. Limited extensions of the RATA deadline are allowed for “non-QA” quarters---the maximum allowable extension is 8 calendar quarters from the previous RATA. A 720 operating hour grace period is available.
QA/QC test or specification
Acceptance criteria
Frequency
Consequences if not met
Relative Accuracy
RA ≤ 20.0% of RM mean value; or if RM mean value ≤ 5.0 µg/scm, absolute difference between RM and sorbent trap monitoring system mean values ≤ 1.0 µg/scm
RA specification must be met for initial certification and annually2 thereafter
Data from the system are invalid until a RATA is passed.
Gas flow meter calibration
An initial calibration factor (Y) has been determined at 3 settings; for mass flow meters, initial calibration with stack gas has been performed. For subsequent calibrations, Y within ± 5% of average value from the most recent 3-point calibration
At 3 settings prior to initial use and at least quarterly at one setting thereafter
Recalibrate meter at 3 settings to determine a new value of Y.
Temperature sensor calibration
Absolute temperature measured by sensor within ± 1.5% of a reference sensor
Prior to initial use and at least quarterly thereafter
Recalibrate; sensor may not be used until specification is met.
Barometer calibration
Absolute pressure measured by instrument within ± 10 mm Hg of reading with a NIST-traceable barometer
Recalibrate; instrument may not be used until specification is met.

24. Hg Emissions Reporting
For units that qualify as LEEs for Hg, the results of all 30-day demonstration tests must be submitted to EPA’s WebFIRE database, using the Electronic Reporting Tool (ERT). The results must be submitted within 60 days after completing each test.
For units that continuously monitor Hg emissions:
A monitoring plan must be developed and maintained.
The Hg CEMS or sorbent trap monitoring system(s) and any required additional monitoring systems must be represented in the monitoring plan.
Monitoring plan data must be submitted to EPA electronically and updated when necessary, using the Emissions Collection and Monitoring Plan System (ECMPS) Client Tool.
Hourly Hg concentration data must be reported electronically using the ECMPS Client Tool.
Data from Hg CEMS are reduced to hourly averages
For sorbent trap monitoring systems, the Hg concentration obtained from the analysis of a pair of traps is reported for each hour of the data collection period.

25. Hg Emissions Reporting (cont’d)
Quarterly electronic reporting of Hg emissions is required, no later than 30 days after the end of each calendar quarter.
Hourly records of any additional data needed to convert Hg concentration to units of the emission standard (e.g., stack gas flow rate and/or moisture content, CO2 or O2 concentration, electrical load) must be included in the electronic quarterly reports.
The additional monitoring system(s) used to convert Hg concentration to units of the standard must meet the requirements of 40 CFR Part 75.
For sorbent trap monitoring systems, certain other hourly data must be reported (i.e., sample flow rate, gas flow meter reading, and the ratio of stack gas flow rate to sample flow rate).
The data acquisition and handling system (DAHS) must be programmed to provide an hourly data stream in the units of the applicable Hg emission standard (lb/TBtu or lb/GWh).
Hg emission rates are calculated only for operating hours in which valid data are obtained for all parameters in the applicable equation(s). For the purposes of Subpart UUUUU, Part 75 substitute data are not considered to be valid data.

26. Hg Emissions Reporting (cont’d)
Operating hours in which a valid Hg emission rate is not obtained are counted as hours of monitoring system downtime.
Only unadjusted monitoring data are used to calculate Hg emission rates. Bias adjustment factors (BAFs) are not applied to any of the parameters.
The results of all required certification and QA tests of Hg CEMS, sorbent trap monitoring systems, and the additional monitoring system(s) used to convert Hg concentration to units of the standard must be reported electronically, using the ECMPS Client Tool.
For Hg CEMS and sorbent trap monitoring systems, RATA results and test details must also be submitted to the WebFIRE database via ERT, within 60 days after completing each test.
Compliance with the 30 (or in some cases, 90) operating day rolling average Hg emission rates is assessed in the semiannual reports required under §

27. Future Changes to ECMPS for Hg
EPA is reviewing the Hg reporting schema and reporting instructions (March 5, 2008) drafted for CAMR. This will be our starting point for integrating continuous Hg data into ECMPS.
Structurally, that schema is compatible with the MATS Hg continuous monitoring and reporting requirements
New method of determination codes (MODC) will be needed to handle the differences between the CAMR rule and the MATS rule.
Exclusion of Startup and Shutdown hours
Handling of invalid data hours – No Substitute Data
No Bias Adjustments
Etc.

28. Future Changes to ECMPS for Hg CEMS
Monitoring Plan will have to reactivate the <CalibrationStandard> record for reporting the basis of daily calibrations (i.e., Hg0 or Hg2+ )
Add 3-level System Integrity Check to the QA schema
Add the <WeeklySystemIntegrityData> record back to the Emissions schema. Program to check for this on an operating day basis when Hg0 is used in daily calibrations.
Hourly Hg concentration measurements to be recorded in MHV records
Hourly Hg emission rate data to be reported in DHV records

29. Future Changes to ECMPS for Hg Sorbent Trap Monitoring Systems
Add the QA record structure for reporting the required Gas Flow Meter Calibrations, GFM temperature sensor barometer calibration checks
Create special emissions data schema to support sorbent trap monitoring
Hourly GFM Data record
Sampling Train Data record
Sample Train Data would contain all the information necessary to determine the quality assurance status of each sampling period.
Spike Recovery
Breakthrough
Trap Agreement
Stack Flow to Sample Rate Ratio Check
Leak Check

30. Future Changes to ECMPS for Hg Sorbent Trap Monitoring Systems (continued)
Concentration determined from a sampling period would be entered into the appropriate hourly MHV records
This allows the data to be used to determine hourly emission rates and treated like Hg CEMS data for compliance averaging.
Also allows for a sorbent trap systems to be used as a backup to Hg CEMS, (if desired to keep data loss at a minimum).
Hourly Hg emission rate data to be reported in DHV records

31. ECMPS Process Forward
EPA is committed to ensuring that the continuous monitoring data for MATS is integrated into ECMPS in a manner that is consistent with the process currently used for Part 75 reporting.
Also, we will be working on more specific draft documents on the specific plans as they relate to MATS – ECMPS monitoring requirements
Reporting Schema
Reporting Instructions
Check Specifications
Stakeholder meetings
Goal is to have ECMPS and your DAHS programming updated, tested and ready to go by 1Q2015