September 6, 2002

Dear Mr. Alcamo:

Tetra Tech EM Inc. (Tetra Tech) is submitting analytical results for eight surface water samples collected by Tetra Tech's subcontractor, Earth Tech, Inc., at the Detmer 2 Spring location near the above- referenced site during a storm event on July 22 and 23, 2002. The samples were collected at 1-hour intervals during the storm event using an autosampler. Tetra Tech's summary of field sampling activities dated August 7, 2002, discusses the autosampler's programming and setup at the location where the Detmer 2 Spring emerges into the West Fork of Stout's Creek near the ABB site. Tetra Tech's subcontractor laboratory, TriMatrix Laboratories, Inc. (TriMatrix), of Grand Rapids, Michigan, analyzed the surface water samples for (1) polychlorinated biphenyls (PCB) using U.S. Environmental Protection Agency (EPA) SW-846 Method 8082 and (2) total suspended solids using EPA Method 160.2.

Tetra Tech reviewed TriMatrix's analytical data to evaluate whether the data meet the data quality objectives specified in the quality assurance project plan for the above-referenced work assignment. Tetra Tech's review was conducted in general accordance with procedures discussed in EPA's "Contract Laboratory Program National Functional Guidelines for Organic Data Review" dated October 1999. The review procedures were adjusted as necessary.

The enclosure to this letter summarizes TriMatrix's sample analytical results and includes necessary data qualifiers applied by Tetra Tech. TriMatrix's quality control (QC) results included the results for initial and continuing calibrations, blanks, and laboratory control samples. The opening continuing calibration results for Aroclor 1016 on one column exceeded the QC limit. Specifically, the average continuing calibration result for this Aroclor exceeded the 15 percent difference QC criterion. Because the accompanying continuing calibration result for Aroclor 1016 on the other column was acceptable and because all other calibration results (including the closing continuing calibration results) were acceptable, no qualifications are warranted for this minor irregularity.

The PCB mixtures detected in seven of the eight samples were similar to both Aroclor 1242 and Aroclor 1248. TriMatrix identified the mixtures as Aroclor 1248. If the PCB mixture in the samples had been identified as Aroclor 1242, the quantitative results would have been about three-quarters of those presented in the enclosure. Also, for the seven samples with positive results, more than a three-fold quantitative difference was observed among the results for the five peaks on the quantitation column chromatogram. These differences are probably due to differences between the mixtures of PCBs in the surface water samples and the unweathered standards. The Aroclor 1248 results for the seven samples are therefore considered to be estimated.

The storm event surface water sample analytical results are usable for all purposes as qualified. If you have any questions regarding Tetra Tech's data review, please call me at (312) 946-6491.