cn\<\
M E M O R A N D U M
DATE: October 29, 1992
TO: Designated PRP Group
West Lake Landfill Site
FROM: McLaren/Hart
Ray Forrester
Brent Finley
Jack Buddenbaum
SUBJECT: NRC SUMMARY REPORT - ANALYSIS
•1> 07 996^7

3. o

40249243

SUPERFUND RECORDS

Several key issues arise from the manner in which the radiological data have been

interpreted in the U.S. Nuclear Regulatory Commission (NRC) Summary Report concerning

the West Lake Landfill Site. This Memorandum briefly reviews the data that have been

collected, examines how that data have been interpreted, and identifies certain overestimates

on the part of the NRC concerning the nature and extent of radioactive contamination at

the Site.

NRC’s Beliefs Concerning Current and Future Health Risks

The NRC documents reviewed state that, under current conditions, the Site does not pose

a risk, nor has it in the past; however, those documents state that, in the future, levels of

radioactivity in soils located at the Site will eventually reach levels that pose a risk and,

therefore, remediation is required to protect against future risks. Indeed, the documents

conclude that levels of radioactivity will reach such magnitudes that on-site cleanup options

may not be feasible and that a costly off-site disposal option may be the only alternative.

These conclusions are based primarily on soil data that was collected over 10 years ago. A

careful review of the surficial and subsurficial boring data, and NRC’s interpretation of

those data, suggests that the radioactive inventory at the Site may have been exaggerated

to the point where the feasibility of less costly options, including no action, may have been

prematurely dismissed and may very well be acceptable for much of the material at the Site.

Data Collection and Interpretation

Briefly, the NRC contractor measured gamma emissions in 18 subsurface borings placed in

areas that were believed to be the most contaminated. The estimates of average levels of

Ra, Th and U and the total radioactive inventory were derived from the information

obtained from these subsurface borings and from approximately 10 surface samples. It is

interesting to note that the subsurface borings data showed that 7 of the 18 borings did not

Westlake Landfill

October 29, 1992

Page 2

have a single sample greater than background concentrations (>2.5 pCi/g Ra), and the

other 5 borings had levels above background only in the top 3 feet. Only the remaining 6

borings had levels of radioactivity that could be considered to be above background

throughout the boring. Apparently, the Site contains localized confined “hot spots”; it is not

contaminated in a homogeneous fashion.

The NRCs interpretation of the data collected, and its assumptions can be summarized as

follows:

Assumption #1: The average specific activity of Ra in all samples from all

of the borings is 90 pCi/g.

Assumption #2: Radiochemical analyses of 10 surface samples showed that

all Th:Bi ratios ranged from 4:1-40:1; other samples taken at Latty Avenue

reportedly had ratios ranging up to 300:1. As a conservative measure, it was

assumed by NRC that the average Th:Bi ratio in the landfill is 100:1. Since

Bi and Ra are assumed to be present in equivalent concentrations, the Th:Ra

ratios are also assumed to be 100:1. Therefore, the average Th concentration

in the landfill is 9,000 pCi/g.

Assumption #3: By ‘Visually integrating” the total volume of soil that contains

>5 pCi/g Ra and assuming that the average Ra concentration in the material

is 90 pCi/g, the total Ra inventory of the site is 14 Ci Ra. Therefore, the

total Th inventory of the site is 1400 Ci Th.

Assumption #4: Because of the large Th:Ra ratio, future decay of Th to Ra

will cause Ra activity to increase 5-fold over the next 100 years, 9-fold over

the next 200 years, and 35-fold over the next 1,000 years. Therefore, Ra

concentrations will eventually exceed soil criteria for most on-site disposal

options and off-site disposal may be required.

In summary, NRC has assumed, based on indirect measurement methods and conservative

assumptions, that Ra and Th levels are 90 pCi/g and 9,000 pCi/g, respectively, and that the

levels of Ra will eventually exceed acceptable concentrations because Th decay to Ra is not

in equilibrium.

However, a few critical assumptions made by NRC have resulted in gross overestimates of

the amount of Ra and Th present in the landfill. These assumptions may have a significant

impact on the disposal options. These are outlined below.

ciaren

•/aft”

Westlake Landfill

October 29, 1992

Page 3

Issue #1 – Ra Concentrations: A review of the data indicates that the actual

average activity of Ra in all subsurface borings is far less than 90 pCi/g.

Indeed, it is clear that well over half of the samples are at background and

approximately 90% of the samples are far less than 90 pCi/g. The NRC

apparently mistakenly took the arithmetic means of data to derive an average

concentration. However, arithmetic means are appropriate only for data that

are normally distributed. The Ra data at the Site are lognormally distributed,

and therefore, the geometric mean should have been calculated. Lognormal

distributions occur when the bulk of the data are grouped around a certain

value and a small percent of the samples have much higher concentrations

than the rest (outliers). This almost always occurs with “hot spot” data.

Several EPA guidance documents indicate that it is inappropriate to take the

arithmetic mean of data that are not normally distributed.

The impact of taking the arithmetic mean of these data is that the average Ra

concentration throughout the landfill has been overestimated. The geometric

mean, which is a more accurate measure of the central tendency of the data,

is less than 10 pCi/g. Hence, the NRC has overestimated the Ra

concentrations throughout the landfill by almost at least an order of

magnitude.

Issue #2 – Th:Ra Ratios: The 10 surface soil samples contained Th:Ra ratios

ranging from 4:1 to 40:1. However, only 2 ratios were greater than 30:1 and

the mean ratio was approximately 18:1. As with the Ra data, the Th:Ra ratios

are lognormally distributed because there are 1 to 2 “outliers”. Regardless,

NRC essentially ignored these ratios and instead referred to samples taken

off-site at Latty Avenue that were approximately 300:1. NRC arbitrarily

settled on a ratio of 100:1. As noted above, this is 5 times greater than the

true mean of the ratios measured on-site. Hence, NRC’s estimate that

average Th concentrations are 9,000 pCi/g may be at least 50-fold too high.

Issue #3 – Validity of Radiochemical Analyses: An apparent inconsistency

arises in the U:Bi ratios measured in the radiochemical analyses of the surface

soils and those measured by gamma analyses in the subsurface borings. As

is pointed out several times in the Summary Report, U:Ra ratios (and,

therefore, U:Bi ratios) in the subsurface borings are within a very narrow

range, approximately 1:2 to 1:10. However, in the radiochemical analyses of

the surface borings, the U:Bi ratios (and, therefore, the U:Ra ratios) range

c/argn

•/art

Westlake Landfill

October 29, 1992

Page 4

from 2:1 to 1:4,000. This may be indicative of poor data quality in the

radiochemical analyses (e.g., perhaps the instrument was not properly

calibrated). If so, then the thorium data, on which much of NRC’s

assumptions hinge, may not be entirely credible.

Issue #4 – Radiochemical Inventory: The radiochemical inventory for Ra was

estimated by taking the “mean” Ra concentration of 90 pCi/g and multiplying

it by the reported volume of soil that contains > 5 pCi/g Ra. This was done

by “visual integration”. The accuracy of the “visual integration” technique is

unknown. However, if one examines only the Ra data that are >5 pCi/g, the

values are once again lognormally distributed, with a geometric mean of

approximately 10 pCi/g. Hence, the total inventory of Ra at the Site is likely

to be less than 2 Ci, and not 14 Ci as stated in the NRC document. Similarly,

the total Th inventory may very well be 50-fold less than 1400 Ci

(approximately 28 Ci).

Issue #5 – Future Concentrations of Ra in the Landfill: The Summary Report

suggests that Ra concentrations will increase by 5-fold over the next 100 years,

9-fold over the next 200 years, and 35-fold over the next 1,000 years. This

would result in future Ra concentrations that exceed the NRC Branch

Technical Position (BTP) Option 4 of 200 pCi/g. If one assumes a 10 pCi/g

mean Ra concentration and a mean Th:Ra ratio of 18:1, then future Ra

concentrations are: 18 pCi/g in 100 years, 25 pCi/g in 200 years, 72 pCi/g in

1000 years, 161 pCi/g in 5000 years, and a maximum concentration of 186

pCi/g. These values are within Option 4 disposal limits. Accordingly, it is

entirely possible that soil concentrations of Th and Ra in much of the landfill

site will remain within the cleanup targets; only a few isolated “hot spots” may

require specific treatment.

In summary,

• The average Ra concentration and total radioactive inventory has been overestimated

by an order of magnitude.

• The average Th concentration and total radioactive inventory has been overestimated

by 50-fold.

• The radiochemical analyses upon which NRC has developed their estimates of

thorium concentrations may be suspect.

ciaren -im—

Westlake Landfill

October 29, 1992

Page 5

It is entirely possible that a vast majority of the fill can be demonstrated to be within

guidelines now and for hundreds of years into the future.

Thorium Analysis vs. Th:Ra Ratios

The degree of remediation will be driven almost entirely by the established amount of

thorium in the landfill. Accordingly, during future site investigation activities, actual

radiochemical analysis for thorium should be performed, rather than to continue with the

indirect method of estimating thorium levels from radium levels. The elevated Ra currently

in the landfill has not been derived from the elevated levels of Th. The half life of Th is

in the thousands of years, hence, the elevated Ra is primarily due to the enrichment process

or portions of existing Ra being passed along in the ore refining process. In other words

any “ratio” of Th:Ra is strictly an artifact and has no bearing on any real relationship to Thto-

Ra decay segment. Therefore, the focus should be on Th analysis in future

measurements. Once accurate estimates of the Th inventory are established, the future

ingrowth of Ra and simultaneously decay-correct for existing Ra levels can be calculated to

yield a grand total for future Ra concentrations.

Baseline Risk Assessment

EPA will eventually require a baseline risk assessment. ChemRisk has successfully

negotiated with EPA to allow ChemRisk to prepare the baseline assessment on other sites

with full acceptability to EPA. ChemRisk’s experience with radiochemical risk assessments

should give it high credibility with Region VII. Also, few, if any, EPA contractors can match

ChemRisk’s experience in assessing health risks associated with exposure to radioactive

material. At an appropriate stage, ChemRisk is willing, at the request of the PRP’s, to

arrange a meeting with Region VII staff to introduce them to ChemRisk and to demonstrate

ChemRisk’s experience on the issues that must be addressed in performing a radiochemical

risk assessment.

ciaren

~