November 18, 2001

Jessica -

Sorry, but I can't get my fax working this weekend. So, hoping that regular mail will get through in a timely manner. I have included a portion of the Tetra Tech document that uses a linear relationship between flow and PCB content. The document contains much data generated from flow measurements. I have included an example. One additional weakness of results from this method is the poor correlation that occurs. Notice that the resulting concentrations are very close to 1.5 ug/L. Yet, if one looks at some actual examples of concentrations in the charts shows concentrations around storm events are far in excess of 1.5 ug/L. So, it is difficult to figure what database was used to establish the correlation. The method is likely incorrect and the data are probably inappropriate. This is not to even question the possibility of data distortion by analyzing only the water and ignoring the sediment. I brought this up with you in the past and sent my critique of data quality issues. The problem was not isolated. Please take a look at the July and August, 2001 Illinois Central/Quarry Springs data. Even though they are citing 8082 which requires 1,000 mL., and the CoC says they sent in 1,000 mL., somehow 100 to 200 mL. is not analyzed. How could that be?

I am also including a set of questions posed on the design and installation of the IC Spring Treatment Plant. Much of that is still pertinent. In the same reports noted above I would also recommend that you review the location of the Quarry Spring samples and the relative lower PCB results. I cannot tell whether it is the discrepancy in volume/sediment analysis, or whether it is location of sampling. By sampling downstream from the spring exit point, sparging of the water occurs due to stream turbulence. PCBs are volatilized and water concentrations decrease, but because the contaminant is now in the air. Each storm event mobilizes great quantities of PCBs and what doesn't get caught in the sediment, has great potential for dispersal into a community and regional contamination source.

Please also review the 303d list of streams in Indiana contaminated with PCBs and mercury. Most are receiving their burden by airborne transport. Bloomington's sites are a source for more than just the problems in Richland and Clear Creeks. I am very interested in seeing the methodology of the air sampling and dye trace studies. Very little of the actual site contamination was removed in the "cleanups" and the real routes of contamination via streams and air are being successfully addressed. Please also take a look at some of the COPA postings on PCB volatilization to help you understand the magnitude of PCB release through the air. They are quite volatile.

I know that you have many other sites and a limited number of hours. As I said, I would be willing to dig through the IDEM records for the information, but I feel that what I need has not been sent to you. These are field records gathered by Viacom and Tetra Tech. Copies are not in the admin record at the library. I hope that you have strong influence in the management of information and can request copies for IDEM/s own record repository. Thanks again for your help.