March 30, 1999
Mr. Andy S. Counts
American Furniture Manufacturers Association
223 South Wrenn Street
Post Office Box HP-7
High Point, NC 27261
Dear Mr. Counts:
I am responding to your letter of February 17, 1999, our meeting of March 17, 1999, and your subsequent letter of March 19, 1999. Let me reassure you that we are pleased to have actual test data for this category of boilers. We are highly disappointed however, that we were not allowed to review the protocols in advance or to observe the tests performed in North Carolina as we had agreed at our meeting in August 1998.
The test reports you submitted raise some concerns. With the information submitted, we cannot determine whether the tests are representative of the boiler population for this industry. In addition, we cannot fully reconstruct firing rates and other aspects of operation during the tests to determine if they were typical and appropriate. Our Technical Services Section staff is attempting to review the reports, but we need additional information before DAQ can properly judge the adequacy and applicability of these tests. Preliminary review indicates that some serious errors in the calculations may exist. A list of data and clarifications needed before we can complete our review is enclosed. These items need resolution before additional testing begins, as well.
We continue to agree that the current AP-42 factor (on an energy basis) probably overestimates carbon monoxide (CO) and underestimates NOx. We expect the AP-42 update process will lead to lower values for CO and higher values for NOx. AFMA members and DAQ must make consistent and reasonable judgements on what is the "best available information" for estimating emissions from individual facilities and for reviewing the inventories submitted for the CY'98 inventory process. We seem to have a choice. We can put our efforts into working with the EPA and AFMA to assure that the final EPA factors are the best that they can be, or we can duplicate and possibly shortchange those results by developing an interim factor for NC.
Given the short amount of time before the CY'98 inventories are due, and the questions we have on the test reports submitted by AFMA, we do not believe either process can be accomplished in time for use in the current inventory. Furthermore, any interim factor would necessarily have a short life (immediately superceded when the EPA process is complete), and would probably be somewhat conservative. There is no fee impact for CO emissions in the CY'98 inventory, so this is not a significant consideration. In addition, any facilities that have pending permit classification actions or other such potential impacts (a limited number at most) would likely warrant testing anyway. Therefore, we do not believe it would be appropriate to make changes at this time.
Owners of "dry wood-fired" boilers in NC remain responsible, on an individual basis, for the best possible emission estimates. Facility officials certify their accuracy when they sign the inventory or permit application. We will continue to evaluate the submittals by using the energy-based factors in our spreadsheets, site-specific test results (with our advance review of protocols and on-site observation during testing) and/or other information that is appropriate for the particular facility. Thus, we will continue to approve emission estimates on a case-by-case basis, using the best available information.
We look forward to receipt of the additional information requested and will keep you informed of any decisions from our review of these tests. We would also appreciate your continuing plans to test an additional boiler with our involvement so that we can fold those anticipated results into this review and the EPA factor development process. We appreciate your continued assistance and participation in the AP-42 update process so that we can reach a collective understanding of the best available information to estimate emissions.
B. Keith Overcash, P.E.
Alan McConnell, Kilpatrick Stockton LLP
Alan W. Klimek
DAQ Regional Supervisors
DAQ Section Chiefs
Members, Wood/Furniture Emission Factor Task Force
Attachment: AFMA Test Report Data Needs Request Summary:
Michael Pjetraj - 3/19/99
For the two test reports, encompassing four boilers, there was little to no data on the operating conditions of the boilers. Documentation of the following points is necessary for a thorough review of the test data.
- The maximum firing rate for each boiler
- The firing rate at the time of testing.
- Post combustion control devices
- Percentage of control for each device (to determine pre-control emissions for comparison).
- Control of combustion process & parameters monitored.
- Percentage of excess air during the testing.
- Type of fuel (We are assuming 100% process waste, however, please indicate if any other fuel was present as well as, any bark, or any treated (varnished etc) wood).
- Type and age of the boilers.
There are some specific questions regarding the testing that need to be addressed. These are the same type of questions that would be asked of any test results.
- CO spikes: During the test, there are several CO spikes that exceed the calibrated span as indicated by the DAS printouts. The methodology for continuous analyzers requires that the test be voided and repeated if there is a single value above the calibrated span. How do the spikes affect data quality?
- Similar to the general questions: % excess air, heat input while testing, controls, fuels.
- Where did the O2/CO2 data for PM runs come from?
- Why was run #1 for the particulate matter aborted? Why was the particulate catch analyzed and reported anyway (lab data sheets)?
- The calculation of heat input does not follow any standard calculation that the DAQ follows. How were the heat inputs calculated?
- Why was the final lb./mm Btu calculated from a constant heat input of 14 mm Btu/hr rather than the Method 19 calculations? Using the calculated heat input or the Method 19 calculations, there is an increase in CO and NOx emissions 30+% greater than those reported. The PM data is increased approximately 44%.
- The PM test results indicated a highly fluctuating air draft. The excess air percentage increased from 131% on run 1 to 380% and 462% on runs 2 and 3, respectively. How does this variation affect the test results? Is there any reason for this type of variation?
- Was condensible PM measured along with filterable? If so, please provide.
Last Modified: Fri April 04 16:39:39 2014