GCI TECH NOTES ©
Members of the panel were:Larry Denyer, USEPA OSW, designated HWC contact; John Chapman, Clean Air Engineering, Vice President; Jim Wright, Clean Air Engineering, Technical Director; Dave Gossman, Gossman Consulting, Inc., President; Laura Green, Cambridge Environmental, Inc., President; Steve Wisker, SICK Optic-Electronic, Inc.
The moderator for the panel discussion was Jeff Bossong of Clean Air Engineering. The first panel member to speak was Steve Wisker with SICK Optic Electronics. Steve spoke of the European experience with PM monitoring and referenced a standard of 10 mg/m3 normalized at 0o C as a half hour average. He also talked about the poor repeatability of Method 5, which he presented as a 50% variation in the testing he was familiar with. He pointed out that there would be modified Method 5 sampling in the next round of testing to account for longer sampling times.
In essence, the poor performance of the current Method 5 will cause the Method 5/PM CEM correlation that falls below about 20 mg/dscm to be meaningless. That is to say the correlation value used by the PM CEM whenever the concentration drops below about 20 mg/dscm could be grossly higher or lower than the actual particulate emissions. This would make a ten-minute or two-hour HRA run very high or very low. Also, the next time the Method 5/PM CEM correlation was run, the ten minute and two hour HRAs could be grossly different when there may be no difference in actual particulate emissions. This could lead to a situation where a facility has made no changes in any operating parameters since they performed their first Method 5/PM CEM correlation, and subsequently operated without an "unusual" number of PM exceedance AWFCOs. However, upon performing the required re-correlation of the Method 5/PM CEM, the facility may find that because of this new Method 5/PM CEM instrument output correlation, the resulting HRAs are skewed upward and the facility has numerous AWFCOs due to "high" particulate emission values. The facility would then be subjected to "administrative action" purely due to the poor repeatability of a test method required by the regulations. Or the opposite could happen, neither of which would be technically accurate. He touched briefly on the scattered light technology currently favored by EPA but did not appear overly optimistic that this would work
Attendee comment: There are weasel words in the proposed HWC which allow a source to get out of PM CEM testing. Many other attendees did not agree that a source would be able to get out of PM CEM testing. This was one proposed requirement that seemed pretty certain to get into the final rule.
The next panel presenter was Laura Green who addressed risk assessment and the HWC. She pointed out that sources will likely end up performing a risk assessment as part of the new proposed HWC regulations. She also pointed out that there are quantitative decisions associated with a risk assessment. She then briefly discussed the key elements of a risk assessment; location, location, location; emission rates; land use aspects; and how/who should do a risk assessment.
Following Laura Green was Dave Gossman of Gossman Consulting, Inc. who pointed out that EPA contractors whom are planning to perform mercury CEM testing, at the Holnam Holly Hill South Carolina facility have commented that the EPA proposed certification & calibration program provided for in the proposed regulation won't work. In short, it appears questionable as to whether the mercury CEM technology is where EPA believes it to be. Apparently, the sampling portion of the technology is a problem. On top of that, the amount of moisture in the stack elicited a near panic response from the equipment manufacturers. He pointed out that EPA has requested information/ comments in the proposed HWC rules over 200 times. He also spoke of his concern that USEPA was in effect requiring the use of SW-846 analytical methodologies even though some of those methods are not adequate and other approved methods such as ASTM methods are available and more appropriate. In addition, it was pointed out that under the proposed HWC rule, a malfunctioning continuous monitoring system would result in a waste fuel cut off. Dave also noted that there is a new law, recently passed by Congress, requiring the use of consensus based standards such as those offered by ASTM.
The next panel member to make a presentation was Jim Wright of Clean Air Engineering. He began his talk with the question, "Is everybody eager to perform testing every 18 months?" He discussed the testing requirements in general and pointed out that what scared him the most is the certification of the PM monitors. He also voiced concern over how the testing requirements under the new rule will affect or be affected by the EPA Guidance on Trial Burns which was issued in 1994 (i.e. will the old guidance become obsolete).
Attendee Comment: It is worth noting that the HWC preamble very plainly and soundly explains why 3 hours of PCDD/PCDF testing makes sense but the rule calls for six hours of testing in order to count non-detects as zero.
John Chapman of Clean Air Engineering was the next speaker. John talked about the double analytical costs that would be associated with the proposed method 23A changes. He pointed out that the proposed HWC regulations supposedly compare with the municipal waste combustor regulations however, MWCs only have a single extraction. As part of the detect/non-detect issue, he raised the hypothetical question of when is a particle not a particle?
Larry Denyer of USEPA was the final panel speaker of the evening. He informed the group that the proposed HWC comment deadline had been extended to August 19 even though there were only 12 requests. He also spoke of a number of meetings, including the Center for Responsible Waste Incineration (CRWI) and the Cement Kiln Recycling Coalition (CKRC). He explained that he was the official contact on the HWC rules because EPA develops regulations through the work group process and he had been the working chair for the group. He confirmed that the rule does request lots of comments and that he was looking forward to getting those comments. He pointed out that he had already been made aware of the three hours versus six hours contradiction for PCDD/PCDF testing and that had been a mistake. He had already referred it to the appropriate group for resolution. While he admittedly couldn't recall which part had been in error, a follow-up phone conversation with Larry indicates that the six hour requirement in the regulation portion of the proposed regulations is in error. The reasoning for three hours of dioxin testing explained in the preamble is correct and should have been reflected in the regulatory portion as well. He also spoke briefly about CEM certifications.
Attendee Comment: Stack testing methods have always had different interpretations however the proposed HWC rule doesn't seem to allow for that type of normal variation in certifying CEMs with manual stack sampling. In addition, by analyzing an increased number of impinger/sampling train portions, already existing problems are just made worse. Modified methods are supposedly going to result in lower detection levels through more analysis however Method 29 has 20-30% inherent variation which in effect takes up all of your margin of error for relative accuracy.
Larry Denyer acknowledged the questions/input and then went on to state that he had been made aware of the fact that including Be as one of the metals of concern was holding up the idea of multi-metals CEMs and that this was being reconsidered by the EPA. He also addressed the issue of moisture in the stack of cement kilns as being a problem in PM monitoring. Larry pointed out that he had notes, etc. concerning discussions about moisture in cement kiln stacks and at least one contractor had specifically said that this would not be a problem.
Attendee Comment: Perhaps the contractors had little or no experience with cement kilns stacks and were not speaking from a knowledge basis.
Much of the post-panel presentations discussions centered around risk assessment. Laura Green was asked about how different half-life considerations of PCDD/PCDFs might affect a risk assessment. She responded by stating that the only thing that really matters is the half-life on a blade of grass before a cow eats that blade of grass. If it involved a 15-20-30 day portion of the risk assessment equation then a different half-life would make a difference, otherwise it would not. While there is photodegredation on plants, PM bound PCDD/PCDF is not considered to be photodegradeable.
This panel discussion was a valuable exercise in that it brought incinerators, cement kilns and regulators together in one forum.