HEALTH CONSULTATION

LAWRENCE LIVERMORE NATIONAL LABORATORY
BIG TREES PARK 1998 SAMPLING
LIVERMORE, CALIFORNIA

DESCRIPTIVE STATISTICS

Descriptive Statistics

Descriptive statistics are a tool to characterize a sample. It computes the mean, variance, median, mode, standard deviation, kurtosis, skewness, range, minimum, and maximum of a distribution. The definitions listed here were copied from the spreadsheet help screens used by ATSDR in the data analysis.

The mean is another term for the average of a list of numbers.

The standard error is the sample standard deviation.

The median is the middle value in a range of values in a data set arranged in ascending or descending order. The median gives a more robust estimation of the central value in a distribution than the mean or average.

The standard deviation is the population standard deviation.

The kurtosis indicates how close the data are to a normal distribution. The closer to zero, the more normal the distribution.

The skewness of the data characterizes the degree of asymmetry of a distribution around its mean value. A positive result means that the distribution is skewed to the right (the median is less than the mean). A negative result means that the distribution is skewed to the left (the median is greater than the mean). A skewness value of 0 indicates the distribution is symmetrical around its mean.

The confidence level (0.95) is the probability corresponding to the normal distribution and must be greater than zero (0) but less than 1.

Currently used computer spreadsheets calculate the descriptive statistics of numbers when the data sets are specified. Because any number of variations of the data set can have the descriptive statistics calculated, we are only including a small representative sample of the data evaluated by ATSDR. An electronic spreadsheet is available for distribution. This spreadsheet is approximately 6 megabytes in size.

On the following pages, we have included the descriptive statistics concerning the concentrations of plutonium 239/240 included the data that did not meet ATSDR criteria for analysis. The data sets include: 1) the grid samples at 5 centimeters; 2) all park locations at a depth of 5 centimeters (not including the grid); 3) the tree wells at a depth of 0 to 45 centimeters; 4) outside the tree wells at a depth of 0 to 45 centimeters; and 5) t-test of grid at 5 centimeters as compared to the tree wells at 0 to 45 centimeters. Also included in this analysis is data in which the hot spots have been removed from the grid values.

RESPONSE TO PUBLIC COMMENTS

LLNL Comments/Pu Health Consult from the Environmental Restoration Division

Page 2, paragraph 2
The "Big Trees Park extension" was generally referred to as the "Big Trees Park eastern extension" in the sampling plan.

ATSDR Response: We have corrected the terminology by adding "eastern" and verified the correct term throughout the remainder of the document.

Page 2, paragraph 3
We suggest that the text "The first publicly reported sampling" be replaced with "The first report of sampling" or something similar. As it stands, the text could be misinterpreted to imply that there was previous unreported sampling; this is not the case.

ATSDR Response: The text has been modified to read "The first sampling ..."

We suggest that the clause "and the two additional parks in Livermore" be deleted. The suggestion that the two additional parks might be above background is based on a comparison of samples analyzed with very different data quality objectives: 1 gram aliquots for the parks versus 300 to 1000 gram aliquots for the background data. Such comparisons are not valid.

ATSDR Response: Although we agree that data quality objectives between the various sampling events were different as well as the methods of analysis, the statement that the additional parks were sampled is factually accurate and the initial 1994 report did suggest elevated levels of plutonium. Therefore, we have not made any additional changes.

Page 2, last paragraph, continuing onto page 3
We agree that additional discussion of the previous (1995) data sets is not necessary. In particular, the criticism of this report voiced by the Tri-Valley CAREs representative at the September 16, 1999 Site Team meeting concerning lack of discussion of the 1995 results was incorrect because the 1998 results and 1995 results are entirely consistent with respect to both risk and pathway assessments. Therefore, there is no loss from omitting discussion of the 1995 results from this health consultation.

ATSDR Response: Thank you for your comment.

Page 3, section titled "Statistical considerations for radiological data analysis"
We would like to encourage the use of the term "uncertainty" rather than "error" throughout. Although the term "error" is common among frequent users of radiological data, it may be somewhat misleading to the public, because it refers to unavoidable variation due to the nature of radioactive decay, not mistakes.

ATSDR Response: We have reviewed the NCRP commentary concerning the use of the terms error and uncertainty. To clarify, we have changed the document where necessary.

Page 6, third paragraph
We suggest that some text be added to this paragraph to make it clear that the background values cited (0.0026 pCi/g, 0.0032 pCi/g, 0.0054 pCi/g) are central values, not upper limits. In particular, it should be expected that approximately half of all background measurements will be above these values.

For example, in the last sentence the phrase "... a value twice the reported background..." could be changed to "...a value twice the reported background geometric mean...."

ATSDR Response: The footnote on that page indicates the geometric mean is a central tendency. We have added a statement in that note to indicate that some samples will be below and others above the central values. We have also modified the last sentence as appropriate.

It could also be mentioned that the required reporting level, 0.005 pCi/g, is below the upper range of background, which is an additional reason why the reporting level is valid for the purposes of the sampling plan.

ATSDR Response: Thank you for your comment.

Page 7, starting with the section titled "Pathway Analyses - Radiological Considerations" and continuing through Table II on page 9.

Some of the discussions and interpretations regarding the air pathway are difficult to understand. In particular, inferences about the air pathway based on comparisons between different parts of the park are difficult to understand.

We suggest that you may want to focus the discussion on two arguments:

The first of these is found in the middle of the second complete paragraph on page 8, in these sentences: "Those samples at the laboratory boundary, MET and MESQ, have Pu concentrations lower than soil samples collected in the park. This would not be expected if air dispersion were the pathway-Pu is a dense material and would deposit at higher concentrations closer to the release point."

ATSDR Response: As stated at the September 16, 1999, site team meeting, we were concerned about this section as well. ATSDR has re-written the section and we believe the text is clearer and more explanatory.

The second argument is found in the first full paragraph on page 9, "The GIS spatial analysis indicated that the Pu is not uniformly distributed and these values did not fit a continuous distribution as would be expected from an atmospheric dispersion pattern." In fact, even allowing for local variations in wind patterns due to trees, the spatial distribution is still not uniform enough to be caused by an air pathway.

ATSDR Response: We agree. Thank you for your comment.

The comparison with the 1980 CDHS sampling results should probably be limited to the 1980 sampling locations closer to the laboratory than Big Trees Park. However, this would have very little affect on the results. Please verify the value 0.068 pCi/g for the geometric mean of the 1980 CDHS sampling results; we suspect this is a typographical error. Our calculation shows 0.0069 pCi/g, which supports the first of the two arguments above.

ATSDR Response: We have re-evaluated that section of the text as well and modified the discussion. The value we calculated was in error as we mis-read the column heading as pCi/g when in actuality the heading was pCi/100 grams. ATSDR also now reports the concentration ranges in the downwind direction as reported by CDHS.

The statement in the last paragraph on page 8 "In those areas that are downwind of the laboratory, the concentrations of Pu in soils are higher than those concentrations found in the park" is incorrect. As stated on page 6, the downwind geometric mean concentration is about 0.0054 pCi/g, which is smaller than the 0.017 pCi/g geometric mean concentration cited for the park later in the same paragraph. The correct value for the 1980 CDHS downwind study, 0.0069 (or 0.0068) pCi/g is also less than in samples from the park. More than 20 of the grid samples have concentrations larger than the largest downwind measurement.

ATSDR Response: See our response to your previous comment. The text has been corrected to reinforce our statements that the plutonium concentrations in the park are higher than those concentrations found in the downwind directions.

We suggest that the discussion of the t-tests in Table II include some mention of the variability or range of the data. For example, in the second full paragraph on page 8, the sentence beginning with "In essence, the soils in the top 5 cm from the grid locations are not significantly different from the soils collected from the other 5 cm samples within the park..." is correct with respect to the average concentrations of the samples that meet the ATSDR criteria defined on page 4. However, the ranges of these two data sets are significantly different (as shown by the F-test results in Table II), so it is not quite correct to say the soils are not significantly different. The grid locations include several samples with much higher concentrations than the other areas (but well below EPA's risk based guideline). Taking this into account in the comparison provides additional support for the GIS spatial analysis.

ATSDR Response: We have modified the text as appropriate by adding the ranges of the values used in our evaluation of the samples collected at the 5 cm depth.

To make it easier for the reader, we suggest that "X" in Table II be replaced with "NS" for "not significant."

ATSDR Response: The appropriate change has been made in the table.

Page 10, Table III and preceding paragraph

A more detailed examination of the ratios reveals that the ratio 0.53 for samples meeting ATSDR's criteria in Big Trees Park, as shown in Table III, comes from two types of samples. The first, consisting of the samples with the highest 239+240Pu concentrations, has an average ratio of about 0.10. Since 0.10 exceeds the fallout ratio, it is clear that the highest concentrations are not due to fallout. The second consists of samples with low 239+240Pu values and unusually high 238Pu values. These samples also depart from the usual fallout pattern, but since the levels are low, the departure is probably due to sample heterogeneity.

ATSDR Response: Thank you for your comment.

Page 13, section titled "Pathway Analyses - Metals Considerations"
Although we agree with the conclusions, we have another view of how the data support the conclusion. We understand ATSDR's perspective that only ratios greater than 2.0 are considered significantly different. However, we suggest that ratios between 1.0 and 2.0 are significant and relevant to the pathway question if there are enough of them. The fact that for every metal, nine of the 10 trees had a ratio greater than one in the 0-45 cm interval indicates that even though the ratios are less than two, the differences are real and due to the presence of an additional metal-bearing material in the soil next to the tree roots.

ATSDR Response: Thank you for your comment.

Comments from community group, regional, and national organizations.

(1) The data presented in ATSDR's report do not disprove air and water pathways.

In general, the evidence presented does not support ATSDR's definitive conclusions regarding the implausibility of an air or water pathway. ATSDR's analysis of the air pathway includes contradictory statements. For example the report states on page 8, "If the Pu reached the park via the air pathway, the Pu concentrations in the associated grid locations would be significantly different from the Pu at other surface areas in the park. This would not necessarily rule out the air dispersion pathway as other locations within the park could have been impacted by the air releases". On page 8, you also state that "If the Pu in the top 5 cm of soils is significantly different from that in the samples collected 10 and 20 cm, then either air dispersion or other methods of distribution are possible".

ATSDR Response: As with any technical document or report of this nature, there is never incontrovertible proof that a pathway is feasible or can be ruled out. Our discussions of the air and water pathways are such discussions. One has to evaluate the data using the best available data, procedures of data analysis, interpretation and use of knowledge physical data associated with environmental dispersion modeling, and the physiochemical nature of plutonium. The process ATSDR used to accomplish this difficult task has been discussed at a site team meeting and is outlined in the consultation. We have also discussed what the data do not supply. Therefore, we believe our analysis to be correct based on the available information. We believe the data collected during the 1998 sampling of Big Trees Park strongly suggests that the most plausible method whereby plutonium reached the park is via the sewage sludge pathway. The probability that plutonium was deposited by air or other pathway is remote and this determination, we believe, is supported by the data.

It appears from Table II on page 9, that for the t-test results, the grid samples are statistically different from all ten comparisons except three: "other park locations", "tree well", and "grid (10 cm)"; for the F test, the grid samples appear to be significantly different from all ten comparisons. Since the plutonium found in the top 5 cm of soil appears to be, in fact, present in significantly greater concentrations than is the plutonium found at 10 and 20 cm -- these results seem inconsistent with your conclusion that an air pathway is not viable.

ATSDR Response: On the one hand, a t-test evaluates the averages or mean of two sets of samples to determine if they are similar. On the other, the F-test is a procedure to determine if the variations within (or ranges of) the two sets of sample measurements are significantly different. Therefore, they report different statistical parameters and should not be used to compare a t-test result to an F-test result. ATSDR used both analyses to 1) determine if the average value of plutonium was different from the other averages and then used the F-test to determine if the means where identical, then how did the variations compare. We have also clarified the discussion of the air pathway and given the ranges of the plutonium found in those areas of the park. Plutonium, in the form of plutonium oxide, the form most commonly found in the environment, is not a highly mobile radionuclide and will not migrate in soils. Only in those conditions where soils have been disturbed will one find distributions of plutonium at depth. If the plutonium in the park was applied as a surface dressing and no additional deep soil disturbing activities occurred, then the concentration of plutonium would be higher at the surface.

Moreover, among ATSDR's evidence disproving the air pathway is the statement, "The grid [top 5 cm?] samples are virtually identical [statistically] across the park and in the tree wells which apparently received the sewage sludge; this type of distribution would not be expected if the Pu had been deposited via the air pathway" (p.8) You seem to be saying here that one would not expect a random distribution [i.e., statistically identical] of grid samples with an air pathway. In apparent contrast, among the data used by LLNL to rule out an air pathway is that, according to LLNL's analysis, "The grid radials show a definite increase above background at the first grid radial. The pattern of distribution illustrated in Figure 4 is not consistent with aerial distribution because aerial distribution would show randomly distributed levels of plutonium above background throughout the grid" (Lawrence Livermore National Laboratory. Livermore Big Trees Park: 1998 Summary Results. Gallegos G., MacQueen D., Surano K. August 13, 1999 p. 18 of 24).

ATSDR Response: Although the complete data sets were available to all organizations, LLNL and ATSDR selected different number sets to perform the analysis. LLNL used a wider range of numbers than ATSDR. In essence, their analysis included those numbers that did not pass the ATSDR screening procedure as outlined in the public health consultation. The analysis performed by the laboratory and their numerical results, therefore, would be somewhat different than the ATSDR analysis. In addition, the LLNL discussion is pertaining to the initial sampling points in the grid area; whereas, the ATSDR evaluation was using the trees and associated tree wells and then comparing these values with the grid values.

ATSDR's assumption that because plutonium is a dense material it "would deposit at higher concentrations closer to the release point" may not be valid under all circumstances. This assumption does not account for the other factors that would also be expected to influence plutonium deposition, such as the height of the stack, the loft (e.g., whether there was a heat source at the time of release), the particle size distribution of the air emissions, the ambient temperature, prevailing winds, humidity and other weather conditions, etc.

ATSDR Response: Following an air release and where the plume touches the ground, heavy particles will settle out before lighter particles, only the distance from the release point changes; that is, heavy particles released to the atmosphere will settle out closer to the source than lighter particles. The atmospheric models, using an x, y, and z coordinate system, show this quite clearly with each of the atmospheric stability classes used in the models to calculate what is commonly called "chi over q" distributions, the concentration expected at ground surface as it relates to distance from the source. Since the "chi over q" calculations include release point, heat rise, wind speed, atmospheric turbulence, and distance from the source, we believe our statements are correct. Furthermore, there are a sufficient number of buildings between the park and the plutonium facilities at the lab that would affect the dispersion patterns; this is called the "wake effect." One result of the "wake effect" is that materials have a tendency to settle out sooner than they would if no buildings were present in the wind direction.

Looking at ATSDR's evidence disproving the air pathway takes us back to the previously quoted statement, "The grid samples are virtually identical across the park and in the tree wells which apparently received the sewage sludge." In this regard, we note that there is no direct evidence that the tree wells received sludge, rather, the City agency responsible for the park has consistently denied that plutonium-laced sewer sludge (or any other sewer sludge) was ever used in the park. Although sewer sludge may have played a role in contaminating Big Trees Park, it is difficult to interpret ATSDR's sewer sludge analysis as presented.

ATSDR Response: We agree no direct evidence exists that the park received sewage sludge. However, the data strongly suggest that plutonium contaminated materials were used during the establishment of the trees in the park. The only plutonium contaminated materials that may have been available was sewage sludge. The city has stated that it did not used sewage sludge during park construction; however, this does not rule out other civic organizations who could have landscaped the park.

The data presented also do not disprove a water pathway (i.e., the Arroyo Seco sampling). Elevated levels of plutonium were found in one sample at each of the two locations downstream of the concrete channel. Because the duplicate samples did not validate these results, and because the other Arroyo Seco samples were negative, you rule out a water pathway. However, although you assume that these two results are "false positives", isn't it is equally plausible that the duplicate samples were false negative results?

ATSDR Response: Uncertainties exist with all environmental samples. In order for duplicate samples run in the same lab at the same time using the identical reagents, equipment, etc., to be valid, identical if not very similar results must be obtained. If one were to disregard the duplicate samples when the data exists, then one is skewing the data inappropriately. We believe the laboratory should have reanalyzed the samples when the internal quality control (the duplicate) did not give a similar result. Therefore, the data are not reliable and we believe the statement is considered correct. The reasoning for this is three-fold: (1) During the laboratory quality control procedures (See Figure 1), one of the participating laboratories observed a similar result during their analysis. That is, one sample showed elevated levels of plutonium but not the duplicate. This particular laboratory reanalzed the sample 2 additional times and could not repeat the elevated reading. Therefore, the elevated level was considered suspect or that a "hot particle" in the sample was measured. The laboratory was not able to determine the cause of the elevated reading; (2) If the channel had been the source of contamination in the park, the channel soils could have been used to develop the park. ATSDR believes the distribution of contamination in the park sampling locations would have been different. In discussions with the laboratory and reviewing documents received by both the California Department of Health Services and ATSDR, the sewer lines from the laboratory and the surface water drainage characteristics do not indicate that a viable pathway exists from either the southwest or southeast corners of the laboratory. Modifications performed in those areas occurred in the early to mid 1970s and these changes channeled surface drainage to the center of the laboratory. The sewer lines running under the facilities in that quadrant of the laboratory and toward the arroyo were not connected to any facility that used plutonium; and (3) If the rechannelization of the arroyo had occurred after any releases to the channel, the former arroyo would have elevated levels of plutonium as well. This was not the case. We believe that any plutonium found in the arroyo most likely was derived from the deposition of sludge that may have inadvertently been placed in the channel.

Moreover, ATSDR appears to be overstating the interpretability of the negative results in the other Arroyo Seco samples. These samples may not accurately reflect historical levels of plutonium in the channel, as significant disturbances to the channel are known to have occurred (e.g., scoured by storms, and earth moving equipment) over the years. Further, the Arroyo Seco sampling results do not obviate the need for an investigation of the potential environmental impact of documented sewer line breaks from LLNL.

ATSDR Response: See our previous comment. Based on our investigation, when documented sewer lines broke, those lines did not connect any facilities using plutonium.

We recommend that ATSDR clearly state its decision logic and assumptions for air and water pathways, and resolve the inconsistencies and contradictory statements in the air pathway analysis.

ATSDR Response: We have clarified the document where appropriate.

(2) ATSDR is ignoring the public health implications of its own findings.

ATSDR has determined that elevated plutonium in Big Trees Park is from local LLNL activities, and not from atmospheric testing of nuclear weapons designed by LLNL. The results of the 1998 soil sampling at Big Trees Park demonstrate plutonium at levels up to 310 times higher than background concentration. In this report, ATSDR defines background levels of plutonium as 0.0025 pCi/g (p.6). The results in Table 1. (p.7) show that 25% of all samples from Big Trees Park (94 of 379 samples) had plutonium at levels greater than double the background concentration. Because all data less than 0.005 pCi/g have been excluded from the analysis, it is not possible to determine how many samples were above background levels of plutonium. The results presented in LLNL's Big Trees Park 1998 Summary Results (Table 3. P. 9) show that 43% of all the grid samples (81 of 188 samples) have plutonium at levels above background concentration. Plutonium was previously detected in Big Trees Park in 1994 and 1995. Plutonium was also found by the US EPA in 1994 in Sycamore Grove and Sunflower Street parks. ATSDR reports that in 1973, the soil in the home yards of three LLNL employees who had used the sludge had plutonium at levels more than 700 times background concentration. It is not disputed that plutonium-contaminated sludge was distributed widely throughout the community over many years.

Taken together, these results indicate that LLNL activities have resulted in plutonium at levels at least several hundred times background concentration at various locations throughout the Livermore community.

As a public health agency, ATSDR is well aware that there are potentially significant public health impacts of exposing large numbers of people, including children and other vulnerable populations, to a cancer-causing substance, over long periods of time. As indicated by its recommendation that "no further actions are needed," ATSDR is ignoring the potential public health implications of finding a long-lived, cancer-causing substance distributed widely throughout a well-populated, and growing, community.

ATSDR Response: ATSDR agrees that the plutonium in the park is the result of laboratory activities and ATSDR agrees that the plutonium is above levels considered background. Furthermore, all radioactive materials have been classified as cancer causing agents. However, the scientific community and groups have differing opinions as to what level cancer is caused by agents such as plutonium. Plutonium is an alpha emitter and there are at least 32 alpha emitting radionuclides that occur in nature with decay energies similar to plutonium and may be present in the human body. Only their deposition location in the bone may vary (bone surfaces or bone volume). Human data exists that clearly show the workers from the Manhattan Project and other nuclear related activities who have plutonium in their bodies at concentrations of 1000 times higher than at any soil concentration found in Livermore have led normal lives and life spans. Those from the Manhattan Project are approaching their 70s and 80s and their death rates are similar to those individuals with no plutonium in their system. For sources of information for this statement, we recommend you review the reports of the United States Transuranium and Uranium Registry at www.tricity.wsu.edu/~ustur/ where many of these studies are documented.

We find ATSDR's unwillingness to acknowledge and make recommendations that would address the public health concerns associated with plutonium contamination throughout the Livermore environment to be in direct opposition to its stated mission, "to prevent harm to human health and diminished quality of life from exposure to hazardous substances found at waste sites, in unplanned releases, and in other sources of pollution present in the environment".

ATSDR Response: At those sites where ATSDR has found levels of contaminants above health concerns, the agency has made significant recommendations for actions, including the issuing of health advisories. ATSDR has identified 5 sites easily accessible to the public and contaminated with radioactive materials above levels of health concern and issued public health advisories. This is about 20% of all sites for which ATSDR has issued these health advisories. The contamination at Big Trees Park and the city of Livermore are below these levels of either exposure or radiological dose.

Although ATSDR concludes that the plutonium in Big Trees Park is from sludge, and although it acknowledges that contaminated sludge was widely distributed throughout Livermore, ATSDR does not recommend additional sampling. ATSDR concludes that no further sampling is needed because disturbances to the soil over the years would make sampling locations difficult to identify, and, that if the three households tested in 1973 are an indication of the plutonium levels at other households, the levels would not be of health concern.

ATSDR Response: As mentioned in a previous health consultation approved by the site team, the levels of plutonium in Livermore soils is below levels of public health concern. The radiological dose is less than 5 millirem and in actuality, less than 2 millirem. There is a theoretical risk associated with this increased exposure but that risk is on the order of about 1 excess death per year per 10 million exposed. The Environmental Protection Agency Region IX has established a clean up level of 2.5 pCi/g in soil for Pu 239. Based on the current knowledge of radiation induced adverse health effects, there is no basis for ATSDR to recommend additional actions regarding plutonium in the city of Livermore.

As to the difficulty in identifying households for sampling, ATSDR should recommend a program to gather voluntary information from interested households and to provide sampling on demand for Livermore Valley residents. ATSDR should also support implementation of a toll-free call-in number to gather such information and to answer questions about plutonium contamination. This suggestion was endorsed by US EPA Region IX in 1998.

As to whether the three homes sampled in 1973 are representative of other homes in the area, ATSDR presents no rationale as to why this should be the case. No data are offered to suggest that the sludge pile at the Livermore Water Reclamation Plant was a homogenous mixture of sludge and plutonium. The 1967 LLNL memo that describes the plutonium release reveals that the amount of plutonium discharged was actually not known, but rather was determined by inference. ATSDR/CDHS have shown that the amount of plutonium in the effluent at the Livermore Water Reclamation Plant was likely to have been underestimated. Therefore, the amount of plutonium in a homeowner's sludge cannot be determined from existing data. ATSDR's conclusion as to potential health impacts of the levels of plutonium at the three households is addressed below.

ATSDR Response: ATSDR also supported the issue of a telephone service to allow residents of Livermore to request assistance as discussed at a meeting at LLNL in November 1997, and attended by state, federal, and laboratory personnel. Perhaps a major problem in the sampling of yards is that over time, home designs change, people change residences, and soils are moved about the yard. Because plutonium is very difficult to detect other than by direct soil samples, the question is "how would we know where to sample?" There is no way to determine if the sludge used at these employee homes and yards is representative without the log book. The three yards that were sampled are the only available data. The state and ATSDR have received more recent data of sludge sampled in later years, 1970s, where the plutonium was measured by the more accurate procedure of alpha spectroscopy. The distributions of plutonium in the drying beds is dependent on the procedure the water reclamation plant used to dry the sewage and distribute the sludge in their drying bed. We discussed these concentrations in the drying beds at the September 16, 1999, site team meeting.

(3) ATSDR selectively excludes relevant data from its analysis.

ATSDR's report does not incorporate the results of 1994 sampling at Big Trees, Sunflower Street, and Sycamore Grove parks, the 1995 results of sampling at Big Trees Park, and the Department of Health Services Radiological Health Branch's monitoring of LLNL plutonium discharges to the sanitary sewer. All of these data suggest the potential for significant plutonium contamination throughout Livermore. At the September 16, 1999 Site Team meeting, in response to a question as to why the 1995 Big Trees Park sampling data were excluded, Dr. Charp replied, that to be included, data had to have "data quality objectives." As ATSDR concludes that no additional sampling in the Livermore Valley is needed, in part because of the results of the 1973 sampling at three LLNL employee households, ATSDR should provide the full report, including the original purpose and data quality objectives of these 1973 sampling results.

ATSDR Response: The 1998 sampling event was not to reevaluate the levels of plutonium in Sunflower Park or Sycamore Grove Park but to determine by what method plutonium reached Big Trees Park and then perhaps apply this pathway to other areas of Livermore. The "data quality objectives" for the previous sampling events and the 1998 sampling event are not comparable. Among the issues associated with these data include the different sample size collected, the procedures used to evaluate the samples varied, and the statistical precision of the analyses different. This is the major problem comparing sampling results collected from similar locatons over time as procedures change and methodologies improve. However, one can look at trends in the data. The data trend in 1998 sampling is similar to the other sampling events. That is, the levels of plutonium are above background and below levels of recognized health concern. Based on the recent results for Big Trees Park, the possibility exists that plutonium in other areas of Livermore were the results of sewage sludge being used as a soil additive. The health consultation released by the California Department of Health Services, Environmental Health Investigations Branch states that the levels of plutonium in Big Trees Park was not at levels of health concern. Since the plutonium levels in the other parks are similar, these levels are not considered a level of health concern. Requests for the data quality objectives and related information for the 1973 sampling of the employee households should be directed toward LLNL as they are the originating organization and thus the custodians of this information.

Although the health impacts of exposure to "low-level" radiation are controversial, ATSDR's conclusions regarding the potential health impacts of plutonium contamination in Livermore are based on a comparison of the measured levels with the US Environmental Protection Agency (EPA) Preliminary Remediation Goal (PRG) and National Council on Radiation Protection and Measurements (NCRP) screening level recommendations.

ATSDR Response: Currently, ATSDR, a non-regulatory agency, has no agency derived screening levels for radioactive materials in soils. The agency has established a Minimal Risk Level (MRL) for ionizing radiation and that level is 100 millirem per year (1 millisievert per year). EPA Region IX established its PRG in accordance with Superfund methodology at a lower bound of 1 in a million risk. Another EPA directive establishes a clean up criteria of 15 millirem per year from all pathways (Office of Solid Waste and Emergency Response Directive OSWER No. 9200.4-18). ATSDR dose calculations included in a previous public health consultation estimate the dose at perhaps 2 millirem, a value of about 14% of the OSWER directive and well below the ATSDR MRL. The concentrations of plutonium in Big Trees Park are at levels below the regulatory limit, below the ATSDR MRL, and below documented levels that have been shown to cause adverse health outcomes.

ATSDR should present the underlying assumptions in these reference levels in a clear and complete fashion. These reference levels do not incorporate the breadth of current scientific information on the health impacts of "low-level" exposure to radiation. One fatal cancer per 100,000 person rem was expected in 1943, one per 10,000 was expected in 1985, and one per 2000 was expected in 1998 (Morgan K. and Peterson KM. The Angry Genie. One Man's Walk through the Nuclear Age. University of Oklahoma Press 1999 p.118). It suggests a significant scientific bias on the part of ATSDR to dismiss the potential for public health impacts of a community-wide, long-term exposure to plutonium using only EPA and NCRP reference levels. As to the validity of the NCRP reference level, Karl Morgan pointed out that "a cursory glance at NCRP, which set radiation protection standards in the United States, sheds some light on whose hand fed those who set levels of permissible exposure. ... Past sources of income for the NCRP included the DOE, Defense Nuclear Agency, Nuclear Regulatory Commission, U.S. Navy, American College of Radiology, Electric Power Institute, Institute of Nuclear Power Operations, NASA, and the Radiological Society of North America... the NCRP relies upon the nuclear-industrial complex for most of its funding other than income from publication sales (p.116).

ATSDR Response: The excess cancer rates you mention are excesses in the expected rates of cancer in the population. Currently, information from the American Cancer Society places the incidence of cancer at about 1 in 3 and the cancer death rate at 1 in 4. The cancer rates estimated from the exposure to "low-level" are hypothetical and can or cannot be disproven. This is especially true for exposures in the range of environmental exposure. There are areas on the planet where people are exposed to naturally occurring levels of radiation much higher that those in Livermore or other parts of the United States and these areas appear to have lower cancer rates than the United States. Thus, the controversy on the linear, no-threshold hypothesis, low-level (the term low never defined) exposures, and cancer relationship are cloudy at best. The NCRP, as a US Congress chartered organization is not a regulatory agency and cannot set standards but does make recommendations to the Federal government as well as the nuclear industry. ATSDR uses their recommendations when no regulations exist.

Moreover, these reference levels only address cancer. ATSDR's assessment excludes the risks of other potential, non-cancer, health impacts of exposure to radiation. In light of the uncertainty that surrounds our current understanding of environmental releases and human exposures, exclusive reliance on the use of these reference values also excludes current approaches to addressing scientific uncertainties, such as the Precautionary Principle. The Precautionary Principle can be summarized as, "When an activity raises threats to harm the environment or human health, precautionary measures should be taken even if some cause and effect relationships are not fully established scientifically".

ATSDR Response: Current medical literature does not recognize radiation causing any physiological disease other than cancer although results from the atomic bomb survivors indicate other adverse health outcomes such as mental retardation and perhaps changes in cholesterol levels. However, these effects are caused by exposures and doses well above the levels observed as the result of environmental exposures. Cancer is the only endpoint associated with radiation that has been conclusively shown. The "Precautionary Principle" apparently evolved out of the 1984 conference concerning protection of the North Sea. In substance, the application of the principle is usually derived for chemicals whose effects are potentially toxic, persistent or bioaccumulative or where certain combinations or concentrations of chemicals could alter the physical and chemical state of soil or water. Therefore, some individuals suggest it should be applied to radiation exposures as well. Although there are many versions of this principle, the underlying theme involves making decisions based on limited information and limited funding.

Without the log book indicating where the sludge was possibly distributed and used, there is no way to determine appropriate sampling locations. The California Department of Health Services has agreed to continue investing the issue of sewage sludge distribution from the reclamation plant. Unless additional information s obtained indicating specific locations where sludge was used, further sampling efforts would be inconclusive. Neither the levels of radionuclides in Big Trees Park or in the yards of the workers that reportedly used the sludge are at levels of public health concern.

ATSDR's assessment excludes analysis of the total health risk from exposure to multiple contaminants. Dr. Owen Hoffman, of SENES, Oak Ridge, TN, who served as a technical consultant on behalf of the affected community to the Big Trees Park Sampling plan, recommended that an estimate of the total health risk from exposure to multiple contaminants be made explicitly, rather than comparing a measured soil concentration to a derived concentration limit (i.e., PRG). Dr. Hoffman recommended that measurements of contamination in soil should be extended to include an estimate of the tissue dose (in the case of radionuclides) and an overall estimate of the risk of adverse health outcomes (Draft Final Livermore Big Trees Park 1998 Soil Sampling Plan, July 31, 1998, p.113 of 157).

ATSDR Response: ATSDR discussed this issue with Dr. Hoffman who has agreed with our method of analysis. We offer the following as additional information: 1) The previous health consultation discussed the health issues associated with Big Trees Park and the result was that no public health concern existed at that time. Dr. Hoffman is comfortable with the approach (conservative approach) ATSDR used in that analysis; 2) The current health consultation expands on the previous document and discusses pathways only. Because the plutonium detected in the 1998 sampling is similar in values to the previous sampling events (the 1998 event is more precise than previous events), the impact on public health has not changed; 3) Dr. Hoffman, as a member of the NCRP and an author of NCRP 129, in essence supports the conclusions of the NCRP; that is, the screening values ATSDR applied to Big Trees Park are appropriate and; 4) The NCRP committee, of which Dr. Hoffman was the chair, responsible for developing a commentary on uncertainty analysis, recommends no additional analysis of uncertainty when the environmental levels found are in this order of magnitude.

ATSDR's dismissal of an air pathway does not consider the recent findings in a report of the Defense Nuclear Safety Board, "HEPA Filters Used in the Department of Energy's Hazardous Facilities (DNFSB/TECH-23, May 1999). This report documents DOE's failing infrastructure supporting its HEPA filter program. Data on the historical and current performance of LLNL's HEPA filter program, including documentation of the quality assurance and control measures would seem to be essential to making ATSDR's sweeping judgment about the possibility of air-borne releases of plutonium from LLNL.

ATSDR Response: The agency has been aware of the issues associated with safe use and misuse of HEPA filters at Department of Energy facilities. This includes improperly maintained filters, problems associated with that activity, and failure rates of HEPA filters. Nonetheless, environmental sampling is designed to detect releases to the environment, whether, air, soil, or water. Failure of HEPA filters is therefore indirectly monitored by environmental sampling and directly monitored by sampling stations connected to the release points of the plutonium facilities - stack and vent monitors. The results of these sampling events show no elevated levels of plutonium released to the atmosphere except to the east of the site.

(4) As was done on ATSDR's previous Health Consultation on Big Trees Park, please provide all the comments submitted to ATSDR on this Health Consultation and ATSDR's response to the comments as an appendix to the final report.

ATSDR Response: The comments are incorporated in this appendix.

(5) In general, the data are not presented in the report in a clear and meaningful way. For example, the report lacks simple descriptive statistics as to the nature of the distribution of plutonium in soil overall, and for each of the sampling areas (i.e., the grid, Arroyo Seco samples, etc.) It would be helpful to present the distribution of plutonium levels (i.e., X% were non- detects, X% were less than 0.001 pCi/g, X% were between 0.001-0.002 pCi/g etc.), the range, geometric mean and geometric standard deviation, and other relevant descriptions of the data. When statistical analysis is performed, it would be helpful to provide sufficient, non-technical, and clearly written, information as to what is being analyzed, what are the underlying assumptions of such an analysis, and how changing the assumptions might alter the result. Please include any visual displays such as the GIS results for the grid samples mentioned on page 8. The entire data set has been censored to exclude values less than 0.005 pCi/g plutonium. Please explain the effect that altering the data set has in the results.

ATSDR Response: ATSDR considered adding descriptive statistics in the original document but decided that, based on a comment at a previous public meeting, to keep the discussion of statistics at a minimum. We have added an appendix following the main body of the health consultation. This appendix contains limited descriptive statistics as well as a t-test and F-test for one set of analysis. ATSDR did not censor or exclude any data in which the measured value exceeded the detection limit. This includes those samples that were less than 0.005 pCi/g. As stated in the health consultation, the only samples ATSDR excluded were those that did not pass the requirements of our data selection. We did indeed use values both above and below the reporting limit of 0.005 pCi/g for our analysis. As an example, please refer to the text on pages 8, 10, 12, and several entries in Table V. All assumptions used for the data selection were listed in the health consultation on page 4. Standard techniques were used for the other aspects of the statistical analysis. In addition, the document prepared by LLNL did not exclude any data points and the result of that analysis is similar to the ATSDR findings. All data available to the agencies is also available to the public through the internet as discussed in the health consultation. As a reminder, the data used by ATSDR is available to the public through the web site listed in the health consultation.

The Environmental Health Investigations Branch (EHIB) of the California Department of Health Services (CDHS) is providing comments (Exhibit 1) on the health consultation entitled Lawrence Livermore National Laboratory Big Trees Park 1998 Sampling Livermore, California, written by the Agency for Toxic Substance and Disease Registry (ATSDR).

CDHS requests that ATSDR include the attached comments, and all others comments submitted, in the form which they have been written, and not in a summarized format. We believe this is important to the Lawrence Livermore National Laboratory (LLNL) site team process, which is committed to public transparency throughout the ATSDR public health assessment activities.

ATSDR Response: We agree with your comment.

1. The health consultation is particularly hard to read and understand the basis of the conclusions and recommendations because it does not adequately review the data or provide any of the actual data used in the statistical analyses. It would be helpful to the reader if ATSDR would provide, at the very least, descriptive statistics (i.e. means and range of values) for the various key areas, such as the inside of tree wells, outside of tree wells, the grid locations, and the former area of elevated value (Location 1) etc. Also, the rationale for any values that were excluded in the analyses should be described. Our understanding of ATSDR documents is that they are supposed to be written for the lay person but also be scientifically defensible. The lack of information severely limits the scientific value of the document.

ATSDR Response: We have clarified specific areas of the document to ease its readability. As for the inclusion of statistics, initially ATSDR decided not to include descriptive statistics. However, we have now added an appendix following the main document with limited descriptive statistics. If requested, we will be glad to furnish more information.

2. ATSDR concludes that the elevated plutonium in the park is not a result of air deposition. However, this conclusion does not seem to be supported by the rationale presented in the ATSDR health consultation, and at the ATSDR site team meeting on September 16, 1999. For example, the health consultation states, "Conclusion: Based on analyses of the grid samples and comparing them to areas where air deposition is known to have occurred, ATSDR does not believe that the presence of plutonium, in Big Trees Park is the result of air deposition" (page 10, paragraph 2). ATSDR seems to be referring to the samples designated MET and MESQ, which were collected from the western boundary of the laboratory, as the basis for this conclusion. ATSDR states, "those samples at the laboratory boundary, MET and MESQ, have Pu concentrations lower than soil samples collected in the park. This would not be expected if air dispersion were the pathway-Pu is a dense material and would deposit at higher concentrations closer to the release point" (page 8, paragraph 3 ).

   First, the document does not provide the actual values of the MET and MESQ sampling as a means of comparison. It is implied that these areas (MET and MESQ) are elevated above background (see above). Next, it is unclear whether areas at the western boundary of LLNL would indeed have higher concentrations of Pu than the park; there are no maps provided that show the location from where the samples were collected. One would expect samples collected from the northern end of the western boundary to show less impact from aerial dispersion originating at the southern end the site (plutonium facility), than a sample collected at the southern end of the western boundary. In fact, the aerial dispersion distances to the north end of the western boundary of the LLNL property are similar to the distance to Big Trees Park. Also, on-site changes such as construction activities may have occurred at some time, which could account for the differences in the values. In the absence of the actual values, sample locations, and a scientifically defensible rationale, it is not possible to draw definitive conclusions ruling out the air pathway.

   Additionally, during the discussion of air pathway at the site team meeting and according to the overheads provided by ATSDR which state, "grid locations did not differ from other park areas, and concentrations not different from: downwind areas (N,ENE,NE), fence line monitoring stations, and off site locations in the City of Livermore". Therefore, it can be concluded that there is no difference between downwind values, and concentrations in the park; and downwind values are collected from areas where air deposition is known to have occurred. Thus, based on the limited information presented in the health consultation, air deposition cannot be ruled out as a potential source of the elevated Pu in the park, with the exception of the 'hot spot" area (former Location 1).

ATSDR Response: We have clarified the discussion of the air pathway by including information more relevant to the downwind directions that compare to the distance Big Trees Park is from the laboratory. We have also included values for the north, northeast, and east northeast directions at the fenceline. The data from the California Department of Health Services show that as the distance from the fence increases, the concentrations of plutonium decrease until at the point downwind at the same distance as the park, then there is no discernable difference. We believe that the majority of the plutonium in the downwind direction was from two areas on the laboratory property during dry weather: 1) The holding pond prior to reengineering; and the former taxi way.

3. ATSDR concludes that based on the sampling results from the Arroyo Seco, " there is no indication that Pu ever entered the Arroyo Seco from LLNL outfall in the southwest corner of the site, from overland transport..." (page 11, paragraph 1). ATSDR does acknowledge that 2 samples, collected in the channel suggest the presence of Pu, but the duplicate did not validate these levels. ATSDR suggests that these elevated levels could have been a result of "deposition of Pu from other areas currently in the park, a random error associated with the samples collected in the Arroyo, or possibly an error in the laboratory analysis" (page 11, paragraph 1).

   CDHS believes the detection of Pu in the samples collected downstream of Big Trees Park is an indication that Pu may have entered the Arroyo at some time in the past. It does not seem that "deposition of Pu from other areas currently in the park" is a likely explanation based on the distance at which the downstream samples were collected from the park. CDHS does agree that "random error associated with the samples collected or an error in the laboratory analysis" may have occurred. However, these same reasons could also explain why the duplicate sample did not validate the original results. The fact that the samples collected upstream of Big Trees Park do not show elevated Pu may be a factor of time (when a potential release or migration occurred) relative to disturbances in the Arroyo such as weathering, scouring, sedimentation and development. Therefore, it does not seem that a definitive conclusion can be reached based on these sampling results.

ATSDR Response: Please see our response to a similar question posed by the community groups above.

4. ATSDR concludes that "sewage sludge containing Pu radionuclides was most likely applied to these ornamental trees as the area outside the tree wells is not contaminated" (page 11, paragraph 4).

   ATSDR seems to base this conclusion on samples collected from a depth of 0-45cm, which is essentially a composite of surface soil down to a depth of 45cm. Since none of the samples indicated the presence of elevated Pu at depths greater than 45cm, the depth at which a soil amendment would have been applied, it seems reasonable the elevated Pu found in 0-45cm sample may have been a result of Pu present in surface soil, which is consistent with several other areas in the park.

ATSDR Response: Typical methods of planting recommend a well depth of approximately the same depth as the root ball and 2 to 3 times as wide as the ball to allow for root growth. In some cases, recommendations include a depth twice as deep as the root ball then back fill to allow for softer soils under the root ball. Variation exists in this procedure, especially if bare root stock is used. With many trees, for example those in what the industry refers to as 5 gallon pots, the root ball is about 45 cm (about 18 inches) in height. Thus the composite sample is appropriate for this evaluation. If the sludge, as the data suggest, was used, then the plutonium would most likely be found in the top 45 cm.

5. ATSDR further suggests that sewage sludge was added as a soil amendment based on the ratio of metal concentrations inside the tree well compared to outside the tree well in 1 of the 10 trees sampled (page 14, paragraph 1). This conclusion was reached based on 1% of the results that showed a difference in metal concentrations inside the tree well compared to outside the tree well.

   CDHS agrees that the metal ratios detected inside the tree well of Tree 3 indicate that some type of soil amendment may have been added. However, the concentration of metals detected in the tree well are consistent with many fertilizers and soil amendments, not just sewage sludge. Based on previous information provided by the City of Livermore Parks and Recreation Department, who have stated that sewage sludge was never used as a soil amendment in Big Trees Park, it seems more likely the concentration of metals found in Tree 3 may be a result of a soil amendment or fertilizer other than sewage sludge.

ATSDR Response: Although we presented the data from one tree, 9 of 10 trees for which samples were collected showed the similar trend. That is, the metals and plutonium inside the tree well was higher than outside the tree well. The data we showed was the most dramatic difference. In all cases, the ratios were at least 2 indicating that the metals were at least twice as high inside the wells as compared to outside the wells. Typical fertilizers used in gardens contain nitrogen, phosphorus, and potassium with various amounts of other metals such as those found in Big Trees Park with sewage sludge generally having higher concentrations of these metals than typical fertilizers. In some cases for specific metals, the application of typical fertilizers can double the average concentration of these metals in 45 years or less (Environmental Protection Agency document "Background report on fertilizer use, contaminants, and regulations" EPA 747-R-98-003, January 1999). For example, to double the soil concentration of cadmium using typical fertilizer at a typical rate would require 16 years while using the same fertilizer, the time to double the soil concentration of nickel would require 283 years. Because the nickel concentration is more than double the typical soil concentrations, the application of a typical fertilizer does not appear to be a viable explanation for the heavy metals. Although the concentrations of the metals found in the tree wells are well within the typical ranges of heavy metals found in fertilizer, ATSDR believes that typical fertilizers also would have been used in other areas of the park such as grassy areas. Because no elevated levels of heavy metals were found outside the tree wells, the sludge as a soil amendment (fertilizer) is the most viable source for the plutonium as well as the heavy metals.

6. ATSDR uses sampling results from yards of three LLNL employees who received sewage sludge in 1973 as a basis to conclude there is no need for additional sampling in the Livermore Valley (page 15, paragraph 1). The highest value of Pu-239 detected in the yards was measured at 1.84 13 pCi/g, which correlates relatively well with the concentration detected in the drying beds (2.6 pCi/g) at the Livermore Water Reclamation Plant (LWRP) during that same year. In other words, these samples seemed to reflect the sewage sludge measurements from the same time period (1973). This indicates the possibility that the amount of dilution that occurs in the drying beds may be much less than originally expected.

ATSDR Response: We assume that your statement of 1.84 13 pCi/g is in actuality 1.84 pCi/g  13% which is how the data were expressed. We agree with your supposition that the samples reflect the sewage sludge measurements.

CDHS agrees that this data is of value, because it further supports the need to investigate the distribution of sewage sludge, which occurred during the 1960s when alpha activity was measured at levels up to 297 pCi/g, in dried digester sludge.

ATSDR Response: Thank you for your comment. As previously discussed, there are technical issues drawing conclusions from measurements of gross alpha levels. Although it is a strong possibility that the 267 pCi/g level is indicative of the presence of radioactive material, no firm conclusions should be drawn from that number about the concentrations of plutonium at the treatment plant.

7. ATSDR concludes that "the most credible pathway whereby Pu radioisotopes were introduced into Big Trees Park was the application of sewage sludge. This is based on a comparison of concentration of both Pu and heavy metals collected from within tree wells to the concentrations present outside the tree wells" (page 16, paragraph 2).

   CDHS does not agree with this conclusion for several reasons. The first is based on the data and rationale presented in the health consultation (refer to comments 4 and 5). The second contradiction is that the sewage sludge pathway argument does not explain the main "hot spot" (Location 1). Elevated plutonium levels in this area were detected in the first sampling (USEPA,1994), confirmed in the next (LLNL,1995), and again found in this latest round of sampling (LLNL, 1998). Location 1 is not in an area that would have received a soil amendment, because there has not been any vegetation added to this area.

ATSDR Response: We disagree based on our detailed review of the laboratory data and their quality control and quality assurance documentation. In those instances where the laboratories performed internal sample duplicates, several times the elevated plutonium initially detected was not detected in the duplicate samples or even in a triplicate sample. This is indicative that a microscopic particle of plutonium may have been the source of the elevated readings in those particular samples. This is suggested by the data in Table V. This "particle" is consistent with metals composition of sludge in that aggregation of metals can occur, similar to crystalization. If air deposition was the most probably pathway, then the distribution would be more uniform across the sampling locations. This was not seen in the GIS analysis as the concentrations at the topmost measurements were not uniform and in fact showed several orders of magnitude variations. This would not be expected if air dispersion were the pathway. Measurements of plutonium in the samples collected from the current arroyo as well as the former arroyo do not consistently show elevated levels of plutonium. If the arroyo was the source, then the data should support that theory but the data do not indicate otherwise.

1. Elements have been assigned either one or two letter abbreviations. The abbreviation for plutonium is Pu. When a number is associated with the name or the abbreviation, the number indicates the mass of the particular element. For example, Pu 239 represents plutonium with a mass of 239. Two numbers are given if typical radiochemistry laboratory analyses cannot differentiate between them. For example, Pu 239/340 indicates that typical radiochemistry laboratory analyses cannot differentiate between Pu 239 and Pu 240 because the energies of their decay are almost identical and cannot be differentiated by the counting equipment.

2. The identification of nongovernment-affiliated private laboratories in this document is neither an endorsement nor a recommendation of their capabilities. They are mentioned for informational purposes only. ATSDR has an Interagency Agreement with the Division of Federal Occupational Health (DFOH), Health Resources Services Administration, Department of Health and Human Services. The laboratory services of Georgia Tech were obtained through a subcontract administered by DFOH.

3. The geometric mean is used to measure the central tendency of a data set, especially if the data set is not evenly distributed. Because of this, one would find that a number of samples will be above the reported geometric mean and others will be lower than this central value.

Table of Contents