Subject:	Meeting with Portland Cement Association and Affiliates  -  NESHAP Monitoring Issues

Location:	U.S. Environmental Protection Agency, Research Triangle Park, NC

Date: 		March 22, 2012

Participants:	

Portland Cement Association (PCA)
Andy O'Hare, PCA
Dan Nugent, Buzzi Unicem
Mike Pelan, Lafarge 
Satish Sheth, Cemex
Hector Ybanez, Holcim


U.S. Environmental Protection Agency (EPA)
Keith Barnett
Colin Boswell
Sharon Nizich
Robin Segall
Peter Westlin

RTI International
Matt Hakos
Mike Laney

A conference call was held with the PCA for the purpose of discussing monitoring issues identified by industry.  The list of issues identified by industry and submitted to EPA is attached.

                  PCA -- NESHAP Monitoring Issues (12-17-10)
                                   Parameter
                                     Issue
                                   Citation
                                  Requirement
                                     Issue
                                Recommendation
                                   Comments
PM (Kiln/Cooler)
Performance
§63.1348(a)(1); §63.1349(b)(1)
Use first 30 operating days with CEMS
There are many issues to discuss on PM CEMS that were not in the original proposal, so they were not in our original comments.
EPA needs to revisit PM CEMS as they apply to cement plants.
See Lafarge settlement terms for 30-day sum of PM Mass readings divided by  30 days of clinker production.

Additional question on the issue:  Throughout this section, the phrase "must be determined based on the first 30 operating days you operate" the CEM is used.  Does this mean that if a source is equipped with that monitor prior to the compliance date, it must use that data to determine initial compliance in the first 30 days of operation?
PM (Kiln/Cooler)
Performance
§63.1349(b)(1)
(iii) and (iv)
Discusses initial compliance based on first 30-days of CEMS; requires an emission rate computed for each run based on concentrations in gr/scf.  Includes equations for stack emissions and for stack+bypass emissions.
Is "each run" represented by each hour since the clinker data required is an hourly rate? Or each 15 minutes since §63.1350(b)(4) requires readings to be collected at least every 15 minutes.  Then hourly data to be summed to daily data and then averaged into a 30-day rolling average.
Clarify
PCA will provide suggested language
PM (Kiln/Cooler)
Compliance
§60.63(c)(1); §63.1348 (b)(2)(i); §63.1350(b)
PM CEMS per PS-11 and Procedure 2; Performance test method for PS-11 must be Method 5 or 5i; Relative Response Audits annually and Response Correlation Audits every 3 years;  §60.63 (f) also references PS-2
In order to range the PM CEMS for correlations the facility must operate short term above the PM limit by intentionally increasing PM emissions.  Such periods may include times prior to and after the test run to establish, verify and equilibrate the elevated emissions and to address testing issues or re-tests.  Also, PM correlation testing may require a minimum of 20 runs, which could require multiple weeks.
Expand §63.1357  · These periods should not be included in the compliance determination.  · EPA should explicitly state that it is acceptable to operate above the limit for this purpose.  · EPA should explicitly protect owners or operators against any enforcement actions, penalties. or citizens suits.
Partially covered in existing §63.1357 "Temporary, conditioned exemption from particulate matter and opacity standards"
PM (Kiln/Cooler)
Compliance
§60.63 (c)(3); §63.1350(b)(4)

NSPS  -  30-operating day rolling average =




NESHAP - Collect readings at least every 15 minutes.  Sum the hourly data to daily data and then into a 30-day rolling average. 



Is there any inconsistency here between NSPS and NESHAP?

How do you have a reading every 15 minutes if the clinker rate is calculated hourly?  Or do you take the mass of PM each 15 minutes and add together then divide by the hourly clinker production?  Or is it actually 30-days worth of summed PM mass readings divided by 30-days of clinker production?  This is the only place and only parameter for which the EPA mentions how to do a 30-day rolling average.
Need to address the averaging methodology for all parameters.  Should be consistent.  Should not require 15-minute values, which conflicts with 60.8(g)(3) requiring reduction of CEMS data to hourly values.  Task Force agreed for clinker based emissions limits that rolling 30 day average should be the summed 30 days of mass emissions over the summed 30 days of clinker production.  Also address issue of PS-2 vs. PS-6.  Also see reporting requirements near bottom of this table.
See Lafarge settlement terms for 30-day sum of PM Mass readings divided by  30 days of clinker production.  

1350(b)(4)- This paragraph must be modified to more clearly indicate that the PM 30-day average is computed by "Sum the hourly data ... into a 30-day average."
PM (Kiln/Cooler)
Compliance
§60.64(b)(1)
Compliance based on §60.63(c) (i.e., CEMS 30-day average); PM emission rate is calculated on an equation for main stack emissions.
Discusses compliance based on a CEMS 30-day rolling average in §60.63(c), but §60.64(b)(1) requires an emission rate computed based on concentrations in gr/scf and with no reference to 30 days.
Clarify. PS 11 requires evaluating performance relative to PM concentration, but recording in units of the standard. Equation should be flexible to work with units as reported by CEMS.
PCA will provide suggested language
PM (Kiln/Cooler)
Other
PS-11
PM correlation for CEMS
The cement industry asserts that changes in particle size, refractive index, particle density, etc. may adversely affect the representativeness of the PS 11 PM correlation over time, and that EPA has not adequately investigated these issues at the variety of kiln systems and plant configurations where PM CEMS will be required.
When such changes are suspected, the plant may choose to perform new correlation testing or other investigations, and a new correlation may be established, and the results of previous monitoring may be adjusted or rejected as non-representative.
PCA will provide suggested language
Opacity (RM/FM)
Performance
§63.1349(b)(2)
3-hr Method 9
§63.1348(a)(2) for initial opacity compliance only addresses those sources in §63.1345, which does not include RM and FM, but the Performance testing requirements under §63.1349(b)(2) is not restricted as such
§63.1348(a)(2) should address RM and FM at major sources.
PCA will provide suggested language
Opacity (RM/FM)
Compliance
§63.1348(b) (3)(i) §63.1350(f)(2)
Daily 6-minute VEs by Method 22 on mill sweep and air separator dust collectors.  If VEs observed retest w/in 24 hours.  If still VEs do 30-minute Method 9 and implement corrective action w/in 1 hour if  >10%
§63.1348(b)(3) for continuous opacity compliance only addresses those sources in §63.1345, which does not include RM and FM, but under §63.1348(b)(3)(i) it references §63.1350(f)(2) which is for RM and FM and under §63.1348(b)(3)(ii) and (iii) it references COMS and BLD in lieu of daily VEs which are for RM and FM.
§63.1348(b)(3) should also apply to RM and FM at major sources.
There is no specified opacity standard for Raw Mills and Finish Mills in 1345, but there are in 1350(f).  PCA will provide suggested language.
Opacity (RM/FM)
Compliance
§63.1350(f)
Indicates monitoring per O&M developed per §63.1350(p) and opacity emissions monitoring plan per §63.1350(o)
§63.1350(o) has nothing to do with emission monitoring plans.  §63.1350(p) is not an O&M plan it is an emissions monitoring plan.
Reference opacity emissions monitoring plan per §63.1350(p)
Make recommended change.
Opacity (Other)
Compliance
§63.1350(f)
Indicates monitoring per O&M developed per §63.1347.
Why not per opacity emissions monitoring plan per §63.1350(p)?
Reference opacity emissions monitoring plan per §63.1350(p)
Make recommended change.
D/F
Performance
§63.1348(a)(3)(ii) and (iii); §63.1349(b)(3)(iv) and (vi)
References temperature and injection rate operating limits in §63.1344
§63.1344 deals with Affirmative Defense
Make corrections
Make recommended change.
D/F
Compliance
§63.1348(b)(4); §63.1350(g)
Continuous temperature monitors at inlets to dust collectors. 3-hour rolling average.
Currently the requirement is a 3-hour rolling average temperature based on 180 successive one-minute averages (i.e., each minute compute a new 3-hour average using the previous 179 valid data points) and new rule is a 3-hour rolling average each hour calculated using the previous three hours of available 1-minute data (could be less than 180 data points).  For in-line mills each 3-hour rolling average begins anew when mill goes up or down.
Keep the current methodology.  Facilities are already set up to monitor that way.  Otherwise EPA needs to explain why they changed it and how it provides better compliance.
Per the definitions in §63.1341, "Rolling Average" means the average of all one-minute averages over the averaging period. 

1350(g)(4) - Did EPA really mean to include the words "Each hour"?  It should be "each minute" to preserve the current methodology for D/F.
THC
Performance
§63.1348(a)(4)(i); §63.1349(b)(4)(i)
§63.1349(b)(4)(i)(A) states that compliance is demonstrated with a RATA when CEMS accuracy is w/in 20% or RATA is w/in 10% of the standard.  THC span (as propane) is 50 ppmvd.  §63.1349(b)(4)(i)(B)  states that initial compliance is based on the first 30 operating days with CEMS
Which is used for demonstrating compliance?
Make corrections
This is just an error by EPA, it means that the RATA is SUCESSFUL when the criteria are met, compliance with emission standard is based on the monitoring results
THC
Compliance
§63.1348(b)(6); §63.1350(i)(1)
THC CEMS per PS-8 and Procedure 1
PS-8 requirement is to be w/in 2.5% of the span of 50 ppmvd, which is only 1.25 ppmvd for the CD, where as previously acceptable PS-8A allowed 3 ppmvd for the CD.  The proposed NESHAP was for PS-8A, therefore we did not get an opportunity to comment on PS-8.
Allow previously acceptable PS-8A.  
PS8A CD limits are +/- 3 ppm or +/13% of span, Span is 50 to 100 ppm.  Always used PS8A and 100 ppm span for cement plants burning hazardous waste.
THC
Compliance
§63.1350(i)(2)
For alkali bypass stack may use initial and subsequent performance tests.
Not mentioned for performance/initial compliance
 
PCA will provide suggested language
Alt. THC/ OHAPs
Performance
§63.1348(a)(4)(ii through v); §63.1349(b)(4)(ii)
Method 320 or ASTM D6348-03.  FTIR, three 1-hr runs mill-on & mill-off. Use average HAPs concentrations for initial compliance and weighted (mill-on and mill-off) average THC CEMS concentration to set site-specific THC limit.  Repeat every 30 months.
Method 320 quantitation limits > 9 ppm.  §63.1352 allows the alternate use of Method 18 "solely for use in applicability determinations" per §63.1340 (in which they never discuss applicability determinations).    Note:  the HAPs is based on the average value for mill-on/mill-off but the alternate THC CEMs value is based on a weighted average.  Also need to define "weighted average".
Allow alternate methods for compliance.  The HAPs concentrations should be for the weighted average for mill-on and mill-off values.  Define how "weighted average" is calculated.
What exactly is a weighted THC average for Mill on/Mill off?
Seek EPA insight and PCA will provide suggested language.
Alt. THC/ OHAPs
Performance
§63.1341 (Total Organic HAP definition)
Only the measured concentration of the listed analytes that are present at concentrations exceeding one-half the quantitation limit of the analytical method are to be used in the sum.  If any of the analytes are not detected or are detected at concentrations less than one-half the quantitation limit of the analytical method, the concentration of those analytes will be assumed to be zero for the purposes of calculating the total organic HAP for this subpart.
Method 320 quantitation limits > 9 ppm.
Allow alternate methods for compliance.
Requires definitions of "quantification limit" and "detection limit".  Encourages use of bad testers and poor test equipment to game the system.  Such data will not withstand any legal scrutiny.  Contrary to reasonable scientific principles and practices.  Method 18 must be allowed in rule or by response to petitions.
Alt. THC/ OHAPs
Compliance
§63.1348(b)(6); §63.1350(j)
THC CEMS per PS-8 and Procedure 1 (same as above) or PS-15 and Procedure 1.
Is the facility out of compliance until the alternate limit is set?  Must alternate limit be set prior to the compliance date?  The proposed NESHAP was for PS-8A, therefore we did not get an opportunity to comment on PS-8.
Provide sufficient time after the compliance date to set the alternate TCH limit without the facility being considered to be out of compliance.
The PS8 RATA which compares FID to FID is arbitrary.  If both the installed and test FIDs disagree when both are calibrated to known standards, which is correct?  What is the appropriate corrective action for a failed RATA?
Seek EPA insight and PCA will provide suggested language.
Hg
Performance
§63.1348(a)(5); §63.1349(b)(5)
Use first 30 operating days with CEMS, or 30 daily(?) sorbent traps
Are sorbent traps daily for the initial compliance?
 
Throughout this section, the phrase "must be determined based on the first 30 operating days you operate" the CEM is used.  Does this mean that if a source is equipped with that monitor prior to the compliance date, it must use that data to determine initial compliance in the first 30 days of operation?
Seek EPA insight and PCA will provide suggested language.
Hg
                                       
                                 63.1348(b)(7)
Demonstrate continuous compliance by a 30-day rolling average Hg emission rate (normal operation) and 7-day rolling average Hg concentration (startup/shutdown).  
How does the facility utilize a sorbent trap system and demonstrate compliance with two different emissions standards over two different operating modes/periods?
Clarify
Once the facility switches from "startup mode" to "normal operation," the emissions standards change.  However, the sorbent traps do not capture the distinction between startup and normal operation.  Do the traps need to be replaced every time there is a change between startup mode and normal operation?  Should there be dual sorbent trap systems such that one is used for startup/shutdown operation and one is used for normal operation?  
Seek EPA insight and PCA will provide suggested language.
Hg
                                       
                                 63.1348(b)(7)
Demonstrate continuous compliance by a 30-day rolling average Hg emission rate (normal operation) and 7-day rolling average Hg concentration (startup/shutdown).  
Compliance with the Hg standard after that fact.
If this method provides acceptable data (it mimics the reference method and from my understanding has demonstrated average data availability of 95%), EPA should lengthen averaging periods to greater than 7 days for startup/shutdown (e.g., 30 days) and greater than 30-days for normal operation (e.g., 90 days).
Assuming a sorbent trap can operate for a period of 7 days (during normal operation), between 20-25% of the 30-day compliance period will be determined after the fact.  This leaves the facility with less opportunity to "adjust" the system in response to adverse compliance data.  A 90-day compliance period leaves 7-8% uncertainty during the "last" 7-day sorbent trap period.  With this approach, the facility has more opportunity to adjust the system.
HCl
Performance
§63.1348(a)(6)(i); §63.1349(b)(6)(i)
Wet scrubber.  Method 321 and establish operating parameters.  Initial compliance is the arithmetic average of all runs.  Repeat every 30 months.
 EPA has specified a "typical lower quantification range" of 0.1  -  1 ppm for Method 321, but has not provided a definition of how to calculate or procedure to determine the quantification range, particularly in the presence of water vapor and as effected by the sampling train/PM filter.
EPA should delete "typical lower quantification range, should revise Method 321 and include DL and PQL determination procedures, performance based
 
HCl
Performance
§63.1348(a)(6)(ii); §63.1349(b)(6)(ii)
No wet scrubber.  Use first 30 operating days with CEMS
HCl CEMS are not currently available that can meet the specified requirements or reliably continuously monitor HCl emissions at the levels required to demonstrate compliance with the standards.

No NIST Traceable cal gas.

PS-15 is not adequate, EPA acknowledges in Response to Comments that an HCl-specific PS is needed.
 EPA should provide documentation on HCl CEMS capable of meeting these requirements.





Need NIST traceable cal gas.

EPA should stay this requirement until HCl-specific PS is developed.
Note in the Proposed CISWI MACT the EPA stated:  "Although final performance specifications are not yet available for HCl CEMS and multi-metals CEMS, EPA is considering development of performance specifications."  EPA has proposed PS-13 (Specifications and Test Procedures for Hydrochloric Acid Continuous Monitoring Systems in Stationary Sources) of appendix B of 40 CFR part 60.

There is currently no known HCl CEM available that operates in the range specified by the rule. Thermo agrees with PCA assessment.
HCl
Compliance
§63.1348(b)(8)(i); §63.1350(l)(1)
No wet scrubber.  HCl CEMS per PS-15 and Procedure 1, except RATA requirements per section 11.1.1 and 12.0 of PS-15.
 
Allow other analytical techniques.
Argue against EPA's assertion that such CEMs are available.  Get statements from reputable CEMS manufacturers.  Note in the Proposed CISWI MACT the EPA stated:  "Although final performance specifications are not yet available for HCl CEMS and multi-metals CEMS, EPA is considering development of performance specifications."  EPA has proposed PS-13 (Specifications and Test Procedures for Hydrochloric Acid Continuous Monitoring Systems in Stationary Sources) of appendix B of 40 CFR part 60.

There is currently no known HCl CEM available that operates in the range specified by the rule. Thermo agrees with PCA assessment.
HCl
Other
§63.1352(a)
Allows the alternate use of Method 320 "solely for use in applicability determinations" per §63.1340
Method 320 is not appropriate for measurement of HCl especially at low concentrations, does not address the necessary considerations contained in Method 321
Method 320 for HCl should not be allowed.
 
Parameter Monitoring & Continuous Emission Rate Monitoring
 
§63.1350(n)(3)
CERM flow rate monitor 5% accuracy
Per RATA this should be 10%
Accuracy needs to be per RATA requirements.
 
Parameter Monitoring & Continuous Emission Rate Monitoring
 
§63.1350(o)(4&5)
Alternate Monitoring  -  Administrator will notify you of approval or denial within 90 days.....Neither submittal of the application or the Administrator's failure to approve or disapprove the application relieves you of the responsibility to comply with any provision...
Even if you submit information in a timely manner if the Administrator does not take action, the facility must still be in compliance.
If Administrator does not take action within the specified time-frame, the alternate method should be approved.
PCA will suggest language.
Parameter Monitoring & Continuous Emission Rate Monitoring
 
§63.1343(a)
Ensure appropriate corrections for moisture are made when measuring flowrates used to calculate particulate matter (PM) and mercury emissions. 
Why only applicable to Hg and PM?  Also confirm that default moisture from annual RATA is acceptable.

Why is the moisture correction on PM and HCl only?  Concentration based standards.  Does this REQUIRE continuous moisture monitoring?
If moisture corrections are required, the default moisture from annual RATA is acceptable.
This is what EPA is proposing for GHG Reporting rule.

Seek insight from EPA and PCA will provide suggested language.
Emissions Monitoring Plan & other General Monitoring Requirements
 
§60.63 (f) (i)(1); §63.1350 (p)  §60.64(b)(4) cross references to §63.1350 (p)  also
Develop emissions monitoring plan, and submit upon request.
As referenced per other  NESHAP sections -  PM CEMS, Opacity and Clinker refer to  §63.1350(o), which has to do with alternate monitoring not emission monitoring plans
"(o)" should be "(p)"
Opacity emissions monitoring plan required  -  if using alternative monitoring.  Does this apply to prior AMRs?
Emissions Monitoring Plan & other General Monitoring Requirements
 
§60.63 (f) (i)(1)
Submit this site-specific monitoring plan, if requested, at least 60 days before the initial performance evaluation of your CMS.
Notice to regulators is usually 30 days before the test and that is the point at which the regulator might request the plan.
Change to 30 days.  If regulator requests plan less than 30 days prior to the tests, then the test does not need to be delayed.
 
Emissions Monitoring Plan & other General Monitoring Requirements
 
§63.1348 (b)(1)(i)
"...site specific monitoring plan required by §63.1350 (o)."
"(o)" is for alternate monitoring
"(o)" should be "(p)"
 
Emissions Monitoring Plan & other General Monitoring Requirements
 
§63.1350 (p)(2)
"...must also address (o)(2)(i) through (iii) of this section."
These sections do not exist.  
"(o)" should be "(p)"
 
O&M Plan
 
§63.1347(a)
Develop O&M plan and submit for approval with part 70 permit application
1347(a)  - Most, if not all plants already have a Title V permit.  If the Title V permit has already been issued or a renewal application has already been deemed complete, must the OMP still be submitted to the Administrator for review and approval?  If approval is required, what does the Title V have to do with it and by what date must the plan be submitted?
Clarify
PCA will suggest language.
Performance Test Reporting & Other Reporting
 
§63.1349(d)
All reports must be signed by the "facility's manager"
How is "facility manager" defined?
Define "facility manager"
PCA suggests the same language as that under Part 75 for "designated representative"
Performance Test Reporting & Other Reporting
 
§63.1354(b)(9)(vi)
Semi-annual reports - (vi) Monthly rolling average mercury, THC, PM, and HCl (if applicable) emissions levels in the units of the applicable emissions limit for each kiln, clinker cooler, and raw material dryer.
Is this an average of the 30-day rolling averages calculated each day for that month, or is this an average of the daily values for the 30 (or 31) calendar days for the month?

Changes to SSM Semi-Annual Report formats effective 11/8/10?
Address this with issue of rolling averages for all parameters.
PCA will suggest language.
Operating Days
 
§63.1341; §63.1343
For 30-day rolling averages, operating days include only days of normal operation and do not include periods of operation during startup or shutdown.  The 30-day period means 30 consecutive operating days.
§63.1343(a)  -  the period of "consecutive" operating days applies Hg, PM and THC, but for HCl it does not reference "consecutive" operating days.
HCl should also be subject to "consecutive" operating days.
Address this with issue of rolling averages for all parameters.

The HCl discussion in 1343(a) makes no reference to "consecutive".  Why should HCl be computed differently than mercury, THC and/or PM?
Operating Days
 
§63.1341; §63.1343
For 7-day rolling averages, operating days include only days of operation during startup and shutdown and do not include periods of normal operation.  The 7-day period means 7 consecutive operating days.
§63.1343(a)  -  the period of "consecutive" operating days applies Hg, PM and THC, but for HCl it does not reference "consecutive" operating days.
HCl should also be subject to "consecutive" operating days.
Address this with issue of rolling averages for all parameters.

The HCl discussion in 1343(a) makes no reference to "consecutive".  Why should HCl be computed differently than mercury, THC and/or PM?
NOx - NSPS
Compliance
§60.64 (c)
Calculate and record the 30-operating day rolling emission rate of NOx as the total of all hourly emissions data for a cement kiln in the preceding 30 days, divided by the total tons of clinker produced in that kiln during the same 30-operating day period using Equation 3.
Refers to a NOx concentration in gr/scf.  Should be "...preceding 30-operating days...."
Clarify.  Equation should be flexible to work with units as reported by CEMS.
Gas concentrations should all be ppm not mg/m3 except Hg which is ug/m3.  When fixed it has no impact.
SO2 - NSPS
Compliance
§60.64 (c)
Calculate and record the 30-operating day rolling emission rate of SO2 as the total of all hourly emissions data for a cement kiln in the preceding 30 days, divided by the total tons of clinker produced in that kiln during the same 30-operating day period using Equation 3.
Refers to a SO2 concentration in gr/scf.  Should be "...preceding 30-operating days...."
Clarify.  Equation should be flexible to work with units as reported by CEMS.
Gas concentrations should all be ppm not mg/m3 except Hg which is ug/m3.  When fixed it has no impact.
Opacity (Other)
Other
§60.64(b)(4)
Any sources other than kilns (including associated alkali bypass and cooler) subject to the 10 percent opacity limit must follow the appropriate monitoring procedures in § 63.1350(f), (m)(1) through (4), (m)(10) through (11), (o), and (p) of this chapter.
Therefore, any miscellaneous source per §60.62(c), which would include those located at Area Sources constructed, reconstructed or modified after Aug. 17, 1971, are subject to the requirements in § 63.1350(f), (m)(1) through (4), (m)(10) through (11), (o), and (p).  Also, when would the complaince date be for existing facilities? 
Should only apply to major sources.
 
 
 
§63.1340(b)(6)
§63.1340(b)(7)
§63.1340(b)(8)
Affected sources noted in these subsections can be at ANY portland cement plant.
Other sections (e.g., 63.1345) exclude these sources at area source facilities
Clarify in 63.1340 that these sources are only subject at Major source facilities
PCA will provide suggested language
Operating Days
 
§60.62(a); §63.1341; §63.1343
Operating day means any daily 24-hour period during which the kiln operates. Data attributed to an operating day includes all valid data obtained during the daily 24-hour period and excludes any measurements made when the kiln was not operating.
Poorly defined.  Does one-minute of operation constitute an operating day, or does a period that results in a valid CEMS data value constitute an operating day?  How is "operate" defined; when fuel is burning?  Startup and shutdown are in preamble but not defined in the rule.
Redefine
PCA will provide suggested language
 
 
§63.1343(a)
1343(a) defines the meaning of 30-day and 7-day rolling average as being consecutive for PM, mercury and THC only.  The section also indicates that an operating day includes all valid data obtained during the daily 24-hour period and excludes any measurements made when the kiln was not operating.
If data are excluded for a period that may comprise a day or more, how can it be consecutive if we must jump from one valid data set to the next?
Addition detail must be provided in this section to account for prolonged kiln down time.
PCA will provide suggested language
 
 
§63.1343(b)
Provides emissions limits and emission calculation methods for kiln and bypass stacks
Preamble notes "If a coal mill is heated with kiln exhaust, it is subject to all the emissions limits, and all associated monitoring, reporting, and recordkeeping requirements of this subpart." No provision for separately vented in-line coal mill in regulations.  Facilities with in-line coal mills vented separately from the main stack need to know what, if any, of the continuous monitoring requirements apply to the coal mill stack
Clarify if separately vented inline coal mill emissions need to be accounted.  If so, main stack monitoring should suffice, with the same concentrations as the kiln exhaust applied to the in-line coal mill percentage of total flow.
Reference Ash Grove alternative monitoring request
PM (Kiln/Cooler)
 
§63.1343(b)(2)
Alternate PM Limit
Only allows adjustment for flow from kiln and cooler  -  what about exhaust from coal mill and/or by-pass if they go to main stack.  An issue that came up in our internal discussions today.  For PM on a combined stack, when and how often is the alternate PM limit determined?  Do you use the average exhaust flows for the same first 30-day CEMS operating performance period for the initial compliance?  Is it conducted and changed annually based on the average flows measured for the RATA?
Allow adjustment for flow from all sources to the main stack.
PCA will work on developing proposed alternative language.
Current language establishes a different standard for each kiln based on site specific clinker factor and air flow.
HCl
Other
§63.1343 Table 1 Footnote 4
If the kiln does not have a HCl CEM, the emissions limit is zero. 
This implies that if you have a wet scrubber and therefore do not need an HCl CEMS that your limit is zero  -  but you still have to prove compliance every 30-months by Method 321, and there will be some HCl in the emissions. Also note that the footnote only applies to #4  -  existing kiln startup and shutdown, and #6  -  new kiln normal operation.
Cannot have an emission limit of "zero".  Limit should also be 3 ppmvd.  Also applying to S/S at an existing kiln but normal operation at a new kiln is arbitrary.
The Footnote of 1343 (b) Table 1 states that emissions limit for HCl from major source kilns is ZERO in some cases, but not all.   Existing kilns under normal operations and new kilns during S/S are not footnoted.
 
 
§63.1348(a)(3)
§63.1346(c), (d) & (e)
(iii) If activated carbon injection is used and you are subject to a D/F emission limitation under §63.1343(b), you must demonstrate initial compliance with the activated carbon injection rate operating limits specified in §63.1344 by using the performance test methods and procedures in §63.1349(b)(3)(v). The average of the run injection rates will determine the applicable injection rate limit.
Although the definition "sorbent" broadly applies to ACI used to control any HAPs, 1348(a)(3) only indicates operating limits for ACI usage only apply if ACI is used for D/F compliance.  Need to clarify if ACI injection monitoring and limits only apply to D/F control strategies versus if ACI were used for THC or Hg compliance.
Clarify if ACI monitoring and injection rates only apply if ACI is necessary for D/F control.
Verify EPA intent and consider language changes that would help clarify that intent.
 
 
§63.1346(f)
Fly ash cannot be used in a cement kiln if it was from a process that used activated carbon to control Hg emissions, unless it can be shown that its use will not increase Hg emissions from the kiln. Once the kiln must comply with a mercury limit specified in § 63.1343, this paragraph no longer applies.
The last sentence of this section implies that this paragraph becomes effective on 11/8/10 and is no longer applicable after 9/9/13.
Need to confirm compliance dates and for portions of this citation.
This should be in technical correction of compliance dates.
 
 
§63.1348(a)(6)
HCl demonstration of initial compliance
Section only considers compliance using wet scrubbers and tray towers.  No consideration for compliance via dry sorbent injection.
Include initial compliance measures for alternative control techniques.
Was mentioned in the reconsideration supplemental list.   PCA will develop proposed alternative language.
 
 
§63.1348(a)(6)(ii)
If a source does not have a wet scrubber or tray tower, initial HCl compliance must be demonstrated using CEMS
Would using a dry scrubber also exempt the source from the HCl CEM requirement?
If so, rule should be appropriately revised to specify this.
Was mentioned in the reconsideration supplemental list.   PCA will develop proposed alternative language.
 
 
§63.1348(b)(1)(ii)
(ii) Except for periods of monitoring system malfunctions, repairs associated with monitoring system malfunctions, and required monitoring system quality assurance or quality control activities (including, as applicable, calibration checks and required zero and span adjustments), you must operate the monitoring system and collect data at all required intervals at all times the affected source is operating. Any period for which data collection is required and the operation of the CEMS is not otherwise exempt and for which the monitoring system is out-of-control and data are not available for required calculations constitutes a deviation from the monitoring requirements.
Any quantity of missing data is an automatic deviation under the Clean Air Act. The term deviation is too stringent for minor occurrences of monitoring missing data.
 
PCA will develop proposed alternative language.
 
 
§63.1349(d)
(d) Clinker Production Monitoring Requirements. If you are subject to an emissions limitation on particulate matter, mercury, NOx, or SO2 emissions (lb/ton of clinker), you must:
It does not seem appropriate for a MACT requirement to be triggered by NOx and SO2 limitations.
Request applicability for clinker production monitoring to not be triggered by NOx and SO2 limits.
PCA will propose removal of NOx and Sox reference
 
 
EPA's Policy Memo; May 16, 1995
facilities may switch to area source status at anytime until the "first compliance date" of the standard.
It is not clear whether the existing major sources can opt in as area sources of if once-in, always-in applies to the kilns.
Request clarification for once-in, always-in
Seek insight from EPA and PCA will provide suggested language.
Clinker Production Monitoring Requirements
 
§60.63 (b)(2);  §63.1350(d)(2)
Determine, record and maintain an accuracy of the system for clinker or kiln feed within 30 days of the effective date of this rule (for existing sources).  During each quarter determine, record and maintain ongoing accuracy.
Why must this be done within 30 days of the effective date of the rule if we do not have to comply with a clinker based emission limit for 3 years?
Should not be applicable until 3 years.
1350(d) -  Is raw feed monitoring really due by 12/8/10?  Or, is it really just that we only have to "Determine, record, and maintain a record of the accuracy of the system of measuring hourly clinker production " as literally written?  This would not impose the +/-5% accuracy provisions of the previous paragraph until 9/9/2013.
 
 
§63.1345
Emissions limits for affected sources other than kilns; in-line kiln/raw mills; clinker coolers; new and reconstructed raw material dryers; and raw and finish mills, and open clinker piles.
The owner or operator of each new or existing raw material, clinker, or finished product storage bin; conveying system transfer point; bagging system; and bulk loading or unloading system; and each existing raw material dryer, at a facility which is a major source subject to the provisions of this subpart must not cause to be discharged any gases from these affected sources which exhibit opacity in excess of ten percent.
Crushers are obviously not subject to the Subpart per 1340(c).  However, it is not obvious whether other sources associated with raw material handling and storage including screen, material transfer points, etc. are included or not.
Clarify when OOO ends and LLL begins
PCA will develop proposed alternative language.
 
 
63.1348(a)(1), (4), (5) and (6)
Initial compliance with the respective standard must be demonstrated based on "the first 30 operating days" of CEM operation
1) The requirement implies that the initial compliance demonstration can preceed the Complaince Date if the facility installs and certifies the CEM prior to that date..    2) The requirement does not make provisions for the occurrence of SU/SD events within the first 30 days of CEM operation.
The regulations should be revised to state that initial compliance is not required to be demonstrated before the Compliance date.  The regulation shouild be revised to reflect that the initial compliance demonstration should exclude SU/SD operating days 
PCA will develop proposed alternative language.
 
 
63.1343(b)(2?)
Aternate PM limit for combined kiln/cooler emissions.
1) The prescribed formula relies on flow rate weighting per ton of "raw feed", but fails to define raw feed  How is this to be determined?   2)The basis for flow determination is also not defined.  It is unclear as to how the facility-specific PM limit is to be applied (i.e.. on a flow capacity basis or on an actual flow basis).  
Clarify the term "raw feed" and clairfy that the limit is to be detyermined on a flow capacity basis.
PCA will develop proposed alternative language.
 
 
63.1350(d)(3)
"Record the daily clinker production rates and kiln feed rates."
Facilities that choose to monitor clinker production directly do not need to, and should not be required to, monitor and record kiln feed rates.
Clarify that moniytoring/recording of kiln feed rates is not required if clinker production is directly measured.
PCA will develop proposed alternative language.
 
 
63.1350(k)(1)
The span value of the Hg CEM "must include the intended upper limit" of the mercury concentration measurement range.
The term "intended limit" is so vague as to be impossible to determine what is needed to comply.
The regulation should be revised with a more definitive requirement for the upper limit and one that the regulated entity can plan for in acquiring commercially available equipment.
PCA will develop proposed alternative language.
 
 
63.1350(l)
Must comply with (m)(1) through (m)(4)
The regulation at (m)(3) requires the determination (and presumably recording) of 3-hour block averages.  Since the limit is on a 30-day RA basis, why are 3-hour blocks needed?
Eliminate the requirement for 3-hr block averages.
 
 
 
63.1349(b)(6)
The HCl PT requirements include collecting stack gas volumetric flow rate data.
The HCl standard is a ppm-based standard and does not rely on stack flow.  The requirement to collect flow data is superfluous.
Elimiate the requirement for flow data.
 
 
 
63.1350(g)
Each hour, calculate "the three-hour" average temperature
This revised requirement changesw the monitoring basis of the APCD temperature from a 180-minute RA to a 3-hour hourly RA.  It is unclear whether this revision will require a new "initial" performance test.
Revert back to the use of 180-minute RA or clarify that this change does not require a new inital PT.
 
 
 
63.1357
Temporary exemption from PM and opacity standards for purposes of PM-CEM correlation testing.
EPA failed to extend this exemption to clinker coolers.  Since these will now also require PM-CEMS, they need to be provided a similar exemption.
Revise the requirement to accomodate coolers.
PCA will develop proposed alternative language.

