LAWRENCE LIVERMORE NATIONAL LABORATORY
BIG TREES PARK 1998 SAMPLING
In April 1999, the Environmental Health Investigations Branch of the California Department of Health Services, under a cooperative agreement with the Agency for Toxic Substances and Disease Registry (ATSDR), released a health consultation entitled "Lawrence Livermore National Laboratory Plutonium Contamination in Big Trees Park" . That public health consultation reported the results of an analysis of the potential health effects of plutonium-associated radiological doses that might be received by individuals visiting the park. The analysis indicated that health effects were unlikely and that the potential for plutonium-associated radiological doses in the park was not a matter of public health concern.
The question of how the plutonium (Pu1) reached the park was not completely addressed in that health consultation. ATSDR and California Department of Health Services (CDHS) made three specific recommendations regarding the need for additional sampling in Big Trees Park. These recommendations as published were as follows:
- We recommend further sampling of Big Trees Park and the Arroyo Seco Creek sediments to determine the vertical and horizontal extent of the Pu 239 contamination.
- We recommend further evaluation of the distribution of contaminated sludge throughout the Livermore Valley, and other areas. This would include assessing the different avenues for gathering information on where sludge may have been distributed, and if locations are identified that may have received contaminated sludge, assessing the feasibility of various approaches to characterize the potential plutonium in those areas.
- We recommend that potential contaminant releases from sewer line ruptures be investigated in future health consultations focusing on sediments in the Livermore Valley or LLNL discharges to the sanitary sewer system.
ATSDR is preparing this public health consultation to address the following issues: (1) overall
quality of the data collected during the 1998 Big Trees Park sampling event; (2) evaluation of the
possible pathways by which plutonium reached the park; (3) evaluation of metals contaminants
in Big Trees Park; and (4) if additional consideration should be given for an area-wide sampling
for plutonium in residential soils.
ATSDR held a series of public meetings in the city of Livermore in 1997 and 1998. One outcome of these meetings was that the regulatory agencies composed of the U.S. Environmental Protection Agency (EPA), the U.S. Department of Energy (DOE), and the California Department of Health Services Radiological Health Branch agreed that additional sampling in Big Trees Park was warranted. DOE and the Lawrence Livermore National Laboratory (LLNL), with regulatory oversight, developed a new sampling plan. This sampling plan was released for public comment and was finalized in October 1998 . This plan included a split sampling plan in which approximately 10% of the samples collected would be supplied to ATSDR for independent analysis .
As stated in the sampling plan, the purpose of the 1998 round of sampling was to describe methodologies, analytical procedures, and related tasks to determine how the plutonium identified in the previous public health consultation could have reached the park. Environmental samples were to be collected for pathway analysis. Soil samples were to be collected at various locations and depths in Big Trees Park. Major sampling locations included (1) samples selected by use of a grid to determine the possibility of an air dispersion pathway; (2) the current Arroyo Seco channel and the old channel before the realignment of the creek to determine the possibility of a water pathway resulting from releases of plutonium from the LLNL via sewer line breaks or by overland flow from the southeast quadrant of the laboratory; and (3) sampling in and areas around ornamental trees to determine the potential for a soil pathway resulting from the use of contaminated sewage sludge as a soil amendment. Other sampling areas included the Big Trees Park eastern extension; the playground of the Arroyo Seco Elementary School; and a disked area chosen on the basis of a request by a community member. The disk area was sampled to determine whether the dust and dirt in the area presented the possibility of a possible inhalation and ingestion pathway. While a detailed discussion of the rationale associated with these locations is beyond the scope of this consultation, the information is available in the 1998 Sampling Plan .
Previous Sampling Events
The results of the first sampling of Big Trees Park were released in 1994 when the EPA National Air and Radiation Environmental Laboratory was doing confirmatory surveys of plutonium contamination in the southeast quadrant of LLNL . Big Trees Park and two additional city parks in Livermore were chosen to represent areas considered background. Sampling results indicated that plutonium levels might be elevated in the surface soils of these parks. In 1995, in response to the discovery of plutonium in the soil samples from Big Trees Park, LLNL resampled the soil at Big Trees Park for verification purposes . Details of these sampling events can be found in the previous public health consultation  and two other documents [4, 5].
Each of the previous sampling efforts had different data quality objectives. Because of this and several other reasons, the data between these sets should not have been analyzed as a combined group. ATSDR's reasons for this belief include, but are not limited to, a lack of available interlaboratory quality control and assurance information from independent laboratories; variations in the methodologies used for the analyses; issues concerning lower limits of detection; and the precision and accuracy for the intercomparison of data sets. These factors have been discussed at several public meetings, and the California Department of Health Services, ATSDR, and the regulatory agencies involved have also discussed these issues through a series of correspondence. Additional discussion of the previous data sets is not warranted and is beyond the scope of this consultation.
Data Quality, Quality Control, and Quality Assurance Activities
For the 1998 sampling of Big Trees Park, the issues of data quality were addressed through a more rigorous oversight by EPA and the California Department of Health Services; public participation via a comment period; a split sampling plan among LLNL, EPA, and ATSDR; and regulatory analysis of the split sampling data by EPA. ATSDR was involved in this process in the following three ways: (1) ATSDR served as a central collection point for public comments; (2) ATSDR arranged for an independent laboratory analysis of approximately 10% of the collected samples; and (3) ATSDR participated in the data quality review.
The data quality team was established with representatives from the Lawrence Livermore National Laboratory, the Radiological Health Branch of the California Department of Health Services, and from DOE, EPA, and ATSDR. Members of the team were selected on the basis of their knowledge of radiation chemistry, radiological measurement protocols, and familiarity with radiation measurement instrumentation. Identical data packages including results, quality control samples, protocols and analysis procedures were supplied and reviewed by this team. Discussions on the methods for the team's review were discussed in the DOE offices in Oakland, California, before the samples were collected. Present for these technical discussions were representatives from ATSDR, DOE, EPA, LLNL, and both the Radiological Health Branch and the Environmental Health Investigations Branch of the California Department of Health Services. Additional technical deliberations and discussions of data validation issues were conducted through conference telephone calls or by electronic mail.
The EPA Quality Assurance office in San Francisco also requested a review of the quality control and quality assurance issues by its Office of Radiation and Indoor Air in Montgomery, Alabama, and the Indoor Environments National Laboratory in Las Vegas, Nevada. The latter organization also reviewed the analytical methods and the quality control procedures of the participating laboratories, and evaluated the results with respect to the samples collected according to the split sampling plan. This evaluation concluded that, overall, the data among the laboratories were in close agreement and stated that "most of the data is useable without qualification" .
Statistical considerations for radiological data analysis
Typically, results from the analysis of radioactive samples are expressed as a 3-number set that includes the measured result, an indication of the uncertainty in the measurement, and the minimum detectable activity. Therefore, many statistical considerations must be taken into account during the evaluation of these samples, especially those samples at or near values considered background values. These considerations include method of sample collection and preservation, the methods of analysis, and laboratory quality control and assurance that the samples are handled identically. These issues all contribute to the general term "error," which is best defined as uncertainty in the measurements. Furthermore, the instruments used to measure the radioactivity also contribute to the uncertainty, with the uncertainty of the measurement denoted as the 2 sigma error (2). The 2 sigma error represents a 95% confidence interval that the reported value is correct with only a 5% chance that the "true" number is outside the range in the given interval. For example, 10 plus or minus 3 (10 ± 3) means that the measured value is 10 with an uncertainty of 3. This is interpreted as the possibility of the true number being between 7 and 13. Occasionally, the 2 sigma error is larger than the measured value, for example, 10 ± 15. With the uncertainty factor this great, the sample should be considered a nondetect or zero.
When analyzing radiological samples with concentrations that approach either the lowest level that can be detected by the instrumentation or the levels associated with background, the 2 sigma error increases dramatically. Therefore, finding 2 sigma values approaching 50% or more of the measured value is common in these situations and these data could be considered not significantly different from background; that is, a nondetection.
For this consultation, ATSDR analyzed the data using the following guidelines which were established by ATSDR before the start of the evaluation:
- Does the laboratory-derived value exceed the minimum detectable activity (MDA)? If the value is less than the MDA, then the sample measurement is not significant and no additional evaluation is needed.
- Does the laboratory-derived value exceed the reporting limit as outlined in the sampling plan? If the sample is above the reporting limit, continue with the evaluation.
- Is the measured laboratory-derived measurement greater than the MDA but less than the reporting level? If so, continue with the evaluation.
- Is the 2 interval less than the laboratory-derived measurement and what percentage does it represent? If the 2 interval is less than 25% of the laboratory reported value, consider the measurement a valid measurement. If the 2 interval is more than 25%, attempt to evaluate using other methods--such as duplicate samples.