The Western States Petroleum Association (WSPA), composed of the major oil refining and transportation companies operating in the western states, has recently released their report, "The California Groundwater Purging Study for Petroleum Hydrocarbons." This report was unsigned, but prepared by SECOR. WSPA's cover letter states, "The study's findings show that with few exceptions, purging does not have a significant affect on petroleum hydrocarbon constituent concentrations in groundwater samples. Therefore, purging should not be necessary to ensure that groundwater samples are representative of formation water." This conclusion is contrary to current practices. Most guidance documents recommend that purging, or even low flow purging, is necessary to get a representative sample from wells. This is especially true when volatiles are being sampled. How, then, did WSPA and SECOR come to their contrary conclusion?

Michael Barcelona, Ph.D., formerly with the Illinois Water Survey and now Director of the "National Center for Integrated Bioremediation Research and Development" at the University of Michigan, Ann Arbor, was asked to peer review the study. Dr. Barcelona reviewed the initial workplan and the final report. His general comments on the final report included, "The study seems to have been well organized and documented although it is clear that they adopted their original study design with its flaws. I had communicated my concerns on the design to you last year. It is somewhat disheartening to read the final product which suffers from the serious bias I had identified at that time. Most of the problems with the work arise from the highly variable, poorly controlled purging methods which were expected to cause bias in sample results. Comparison of the biased unpurged bailer sample with another biased (bailed) sample after purging is fruitless. These methods problems are significant and evidenced adequately by the high (relative) standard deviation of the relative percent difference between field duplicates (Table 5-13)." In essence, Dr. Barcelona is saying the bailer is a poor sampling device and that it is meaningless to design a study that is based on a device that biases the sample.

WSPA and SECOR were made aware of these concerns prior to the study being initiated. WSPA responded to Dr Barcelona's initial concerns with the following comments, "While we believe that your concerns are valid, WSPA's study is not intended to evaluate the relative merits of alternative purging methods. Rather, it is intended to investigate differences in groundwater samples collected before and after purging, based on current practices and requirements. The success of this effort is contingent on acceptance of study results by regulators under the current system. Therefore, the protocol must simulate current field practices." (Emphasis not added.)

Some wells were even purged with a vacuum truck. Dr. Barcelona raises a poignant point. He stated, "The greatest systematic difference [25%-vacuum (truck) pump] shows this as a biased method. Analytical and sampling errors can easily be controlled to <1% of natural variability. Why would anyone accept the use of such a biased purging method?"

In 1994, GRA held a short course on groundwater sampling. Robert Puls, Ph.D., from the United States Environmental Protection Agency's Kerr lab in Ada, Oklahoma gave a demonstration on sampling devices. The short course indicated that bailers are poor sampling devices. Dr. Puls went on to say that he could bias the analytical results by using bailers.

One effect of the WSPA study is to question why we allow the current poor sample practices. It is unfortunate that it is easier to let the current, but poor, practices continue when we know they are questionable. WSPA may have a lot to gain by not purging. They state that, "based on a survey of six major oil companies, annual purging costs are about $2,000 per site. Therefore, with approximately 12,000 leaking underground storage tank (LUST) sites, the saving could be on the order of $24 million" by not purging.

It seems appropriate to question current practices. The tools for sampling have improved. Perhaps it is appropriate for the GRA Technical Committee and other members of GRA to work with regulatory agencies, counties, and the regulated community to select sampling devices that minimize purge water and control the natural variability to <1%.

If you would like a copy of the study and Dr. Barcelona's comments, please send $5.00 to GRA to cover reproduction and mailing costs.

Floyd Flood has been in the water business for most of his life. Currently, he is editor of HydroVisions.