Data Usability and Reporting
for Emission Rate Calculations

In dealing with a waste matrix analysis (e.g., waste feed testing), Sample Quantitation Limits (SQLs) are the most appropriate detection limit to use for risk assessment purposes as already documented in EPA's Guidance for Data Usability in Risk Assessment (Part A) Final(EPA 9285.7-09A, April 1992), also known as DUGS. For risk assessment purposes, DUGS defines SQLs as "the method detection limit adjusted to reflect sample-specification such as dilution or use of a smaller sample aliquot for analysis due to matrix effects or the high concentration of some analytes."

EPA has recognized that data quality objectives differ somewhat from the initial characterization of an unknown, potentially gross contaminated medium to the verification testing of a treated or "cleaned" medium. In soil for example, SQLs may not be low enough to verify that clean up methods have met the required clean up standards. In this case, special analytical techniques may be applied to provide lower detection limits or detection limits other than the SQLs are evaluated. In this respect, stack gas sampling is comparable to the verification sampling which follows treatment or removal actions in other environmental programs.

In dealing with stack gas sampling data, lower detection limits are achievable due to less interferences associated with the type of matrix and the lower level of contaminants potentially present. Since lower detection limits are essential to assessing the absence, or minute amount of a constituent verifiably present in the stack gas and ensuring permit limit compliance, EPA's Human Health Risk Assessment Protocol for Hazardous Waste Combustion Facilities, Peer Review Draft (EPA530-D-98-001 A, B, and C; July 1999), commonly referred to as the HHRAP, specifies alternative detection limits for emissions data reporting. Please refer to EPA's HHRAP Sections 2.4.1 and 2.4.2 for the discussion on data reported as non-detected in emissions testing. In order to supplement the HHRAP, EPA Region 6 has developed the following guidelines in order for a facility to effectively meet data quality objectives for combustion risk assessments:

I. Waste Feed Data Reporting: For any data reported as non-detected during waste feed testing (metals or organics), please have the laboratory report SQLs as defined in DUGS, above.

II. Emissions Data Reporting: For any data reported as non-detected during emissions testing (metals or organics), please have the laboratory report the following type of detection limits: SQLs as defined in DUGS and Method Detection Limits (MDLs) or Estimated Detection Levels (EDLs) as defined in the HHRAP. The following discussion explains the need for each set of detection limits and how the laboratory can report results for inclusion into the TB/RB Report.

A. EPA Emission Rate Spreadsheets use MDLs for non-dioxin/furan organics and inorganics to calculate Reliable Detection Levels (RDL=2.623 X MDL) in accordance with EPA's HHRAP.

1. MDLs are specific to each laboratory, and must be explicitly requested from the lab for all chemicals analyzed and subsequently reported in the trial burn report. The procedure to determine MDLs is defined in 40 CFR Part 136, Appendix B. Although 40 CFR Part 136, Appendix A provides MDLs for individual methods based upon ideal conditions, MDLs are laboratory specific and should be obtained directly from the laboratory conducting the analyses. The laboratory will conduct the testing necessary to calculate MDLs on at least an annual basis. Please direct the laboratory to provide MDL summary tables for each type of analysis (per matrix) that is pertinent to the specific trial burn/risk burn testing program for your facility.
   

   For example, MDLs for all metal analytes are based upon the method used for each. For the volatile organic sampling train (VOST), two separate MDLs should be provided, one for the mass measured from the tenax or tenax charcoal samples and one for the concentration measured in the condensate samples. It should be noted that the MDLs used to calculate the RDL for the condensate are multiplied by the 25 ml volume required for the MDL determination. Actual detections in the condensate (or use of SQLs, see items 2 and 3 below) require use of the actual sample volume analyzed for the VOST condensate to calculate the correct mass of contaminant quantified in the sample matrix. EPA spreadsheets sum the RDLs or detected values from the tenax, tenax charcoal, and condensate fractions of the VOST in order to calculate the total train mass.

2. Based upon EPA's experience with other facilities that have completed testing during 1997 and 1998, not all compounds quantifiable by the method and ultimately reported by the laboratory for a specific analysis have had "MDLs" predetermined for reporting purposes. If compounds are quantified that do not have MDLs determined, then EPA will use SQLs for risk assessment purposes as defined above per DUGS. This may increase risk estimates depending on the SQL obtained for each particular analyte. Therefore, the facility should clarify for the laboratory prior to testing that MDLs should be reported for the full list of chemicals typically quantified by the method.
   

3. The laboratory may report both "detection limits" and/or even "sample quantification limits" for chemicals quantified if no direction was specified prior to analysis. Different laboratories may define these terms differently and may not define these terms consistent with EPA terminology for risk assessment purposes. Therefore, please provide the laboratory with EPA's definition of SQLs for risk assessment purposes (SQLs defined above per DUGS). The laboratory should report SQLs on the individual raw data forms. This information may be used if for any reason the MDL for a specific analyte is undetermined.
   

B. EPA Emission Rate Spreadsheets use EDLs to calculate emission rates for dioxin/furan organic compounds which are not detected in accordance with EPA's HHRAP.

1. The HHRAP specifies use of EDLs which are defined within the method. The SW-846 definition does not include the use of empirical factors or other mathematical manipulations specific to the laboratory. If possible, please have the lab report EDLs as defined by SW-846, without any adjustments made for the type of column used.
   

2. Since laboratories typically report an adjusted EDL based upon the type of column used (they multiply the EDL by an "F factor"), EPA spreadsheets allow adjustment of the lab reported EDL back to the standard EDL defined in SW-846 (lab reported EDL ÷ lab reported "F factor"). The lab reports the F factor either in the case narrative or directly on raw data sheets, or both.
   

The above guidelines have been developed to help facilities understand those specific issues which need to be communicated to the laboratory conducting analytical work and the contractor responsible for compiling and reporting data for inclusion in the TB/RB report. However, the type of testing conducted for combustion risk assessments must also be comprehensive in order to utilize facility specific information most effectively.

In general, all waste streams should be sampled for a full scan analysis (i.e., analyses for full, commercially available analyte lists based upon testing for volatile organics, semi-volatile organics, metals, and PCBs if applicable). For stack sampling, please see EPA's HHRAP Volume 2, Appendix A-3 Attachment, Tables Of Compound-Specific Parameter Values, pages A-3-51 through A-3-60 for a complete list of possible compounds which can be quantified during emissions testing. Each facility may have different compounds of potential concern and/or interest and should communicate their specific needs to the laboratory doing the analytical work. The facility should also direct the lab to evaluate and report the top 30 Tentatively Identified Compounds, if possible. EPA's Guidance on Collection of Emissions Data to Support Site-Specific Risk Assessments at Hazardous Waste Combustion Facilities, Peer Review Draft (EPA530-D-98-002; August 1998), which is the companion document to EPA's HHRAP, is very useful for checking correct sampling procedure with correct analytical procedure depending upon the specific data quality objectives.