United States Environmental Protection Agency

Office of Emergency Management
Consequence Management Advisory Team
Erlanger, Kentucky 41018

May 2013

Radiological Survey of
Coldwater Creek
North St. Louis County, Missouri

Airborne Spectral Photometric Environmental

Collection	Technology


-------
Coldwater Creek Survey

May 2013

Team Members

EPA Region 7

Matthew Jefferson, Superfund Remedial Project Manager

ASPECT

Mark Thomas, PhD - Scientist, Team Lead

John Cardarelli II, PhD, CHP, CIH, PE - Health Physicist, Rad Lead
Timothy Curry, MS, PE - Finance and Operations
Paul Kudarauskas, MLA - Environmental Scientist

Kalman Co. Inc., Contract Support:

Jeff Stapleton, MS - Principal Engineer
Robert Kroutil, PhD - Senior Project Engineer
Dave Miller, PE - Integration Engineer

Airborne ASPECT Inc., Contract Support:

Sam Fritcher, President

Beorn Leger, Pilot

Ken Whitehead, Pilot

Richard Rousseau, System Operator

Mike Scarborough, System Operator

ii


-------
Coldwater Creek Survey

May 2013

Table of Contents

Executive Summary	iv

Acronyms and Abbreviations	v

1.0 Introduction	1

2.0 Descriptions of the Sites and Survey Areas	2

3.0 Natural Sources of Background Radiation	5

4.0 Survey Equipment and Data Collection Procedures	7

4.1	Radiation Detectors	7

4.2	Flight Parameters	7

5.0 Data Analyses	8

5.1 Radiological	9

6.0 Results	13

6.1	Radiological Results	13

6.2	Electronic Data	16

Appendix I : Uranium Decay Chain	17

Appendix II	18

Discussion about radiological uncertainties associated with airborne systems	18

Background radiation	18

Secular Equilibrium Assumption	18

Atmospheric Temperature and Pressure	19

Soil moisture and Precipitation	19

Topography and vegetation cover	19

Spatial Considerations	19

Comparing ground samples and airborne measurements	20

Geo-Spatial Accuracy	21

References	22

iii


-------
Coldwater Creek Survey

May 2013

Executive Summary

The United States Environmental Protection Agency (EPA), Office of Emergency Management
(OEM), Chemical Biological Radiological and Nuclear (CBRN) Consequence Management
Advisory Team (CMAT) manages the Airborne Spectral Photometric Environmental Collection
Technology (ASPECT) Program. This program provides scientific and technical support
nationwide to characterize the environment using airborne technologies for environmental
assessments, homeland security events, and emergency responses.

In January 2013, the Agency for Toxic Substances and Disease Registry (ATSDR) asked EPA
Region 7 if it would be possible to collect additional data along Coldwater Creek due to several
health concerns received from the community. ATSDR believed this additional data would assist
in addressing those health concerns. EPA Region 7 requested that the ASPECT Program conduct
a radiological survey over the Coldwater Creek area in North St. Louis County, Missouri. The
survey was conducted on March 8, 2013 between 10:00 a.m. and 12:00 noon. Investigations by
the EPA, the United States Department of Energy (DOE) and the United States Army Corps of
Engineers (USACE) have attributed potential radiological contamination in Coldwater Creek to
runoff or windblown migration of prior storage of uranium-processing residues and wastes from
the North County portion of the St Louis Formerly Utilized Sites Remedial Action Program
(FUSRAP) sites. The St. Louis FUSRAP Downtown and North County sites were placed on the
Superfund National Priorities List (NPL) in 1989. The USACE has removed the North County
sources of these wastes, which came from ore-processing activities at the Downtown portion of
the St. Louis FUSRAP sites.

The purpose of the radiological survey was to identify areas of elevated gamma radiation in the
Coldwater Creek areas. The ASPECT results for Coldwater Creek showed surface gamma
emissions consistent with background levels throughout the Coldwater Creek survey area.

RADIOLOGICAL

About 2,200 gamma radiation measurements were collected and none indicated excess uranium
or uranium decay products. The ASPECT measures gamma radiation from Bismuth-214 which is
the ninth decay product in the Uranium-238 decay chain because Uranium-238 is not a strong
gamma emitter. In this survey, Bismuth-214 most likely indicates the presence of Radium-226
(the fifth decay product of Uranium-238) rather than Uranium-238 since the original uranium ore
was chemically separated from the rest of its decay products. The separation process invalidates
a key assumption in the algorithms used to estimate equivalent uranium concentrations from the
gamma radiation data; therefore, throughout this report "equivalent radium" will be reported
instead of equivalent uranium. No elevated gamma radiation measurements were detected
during the Coldwater Creek Survey.

iv


-------
Coldwater Creek Survey

May 2013

Acronyms and Abbreviations

AEC	Atomic Energy Commission

AGL	above ground level

ASPECT	Airborne Spectral Photometric Environmental Collection Technology

ATSDR	Agency for Toxic Substances and Disease Registry

Bi	bismuth

CBRN	Chemical Biological Radiological Nuclear

CERCLA	Comprehensive Environmental Response, Compensation, and Liability Act

CMAT	Consequence Management Advisory Team

cps	counts per second

DOE	Department of Energy

ENVI	Environment for Visualizing Images

EPA	Environmental Protection Agency

214

eRa	Equivalent Radium based on Bi region of interest

eTh	Equivalent Thorium based on 208T1 region of interest

214

eU	Equivalent Uranium based on Bi region of interest

FOV	Field of view

ft	feet

FUSRAP	Formerly Utilized Sites Remedial Action Program

GPS	Global Positioning System

Hz	hertz

IAEA	International Atomic Energy Agency

K	potassium

MeV	Mega electron volts

NCP	National Oil and Hazardous Substances Pollution Contingency Plan

Nal(Tl)	sodium iodide thallium drifted detector

NPL	National Priorities List

NORM	Naturally Occurring Radioactive Material

pCi/g	picocuries per gram

Ra	radium

ROD	Record of Decision

ROI	Region-of-Interest

Rn	radon

SLAPS	St. Louis Airport Sites

Th	thorium

T1	thallium

U	uranium

|jR/hr	microRoentgen per hour

USACE	United States Army Corps of Engineers


-------
Coldwater Creek Survey

March 2013

1.0 Introduction

The EPA initiated the Airborne Spectral Photometric Environmental Collection
Technology (ASPECT) Program shortly after 9/11. Its primary focus was the detection of
chemicals using an infrared line scanner coupled with a Fourier transform infrared
spectrometer mounted within an Aero Commander 680 twin-engine airplane. In 2008,
ASPECT significantly upgraded the radiological detector system to improve its airborne
gamma-screening and mapping capabilities. In 2012, a neutron detection system was
installed. Currently, ASPECT is the only program in the United States with a 24/7/365
operational platform that conducts remote sensing for hazardous chemicals,
gamma/neutron emitters, and aerial imaging. It has deployed to more than 130 incidents
involving emergency responses, homeland security events, and environmental
characterizations.

Up to a four member crew, two pilots and two technicians, operate the airplane. A
scientific support staff provides additional assessment and product development
commensurate with the site specific needs.

In January 2013, EPA Region 7 requested that the ASPECT Program conduct a
radiological survey over the Coldwater Creek area located in North St. Louis County,
Missouri. The survey was conducted on March 8, 2013.

The purpose of the radiological survey was to identify areas of elevated radiation
contamination as compared to normal background concentrations.* ASPECT uses
multiple algorithms to produce a variety of products for decision makers. One algorithm
requires measurements to be collected over an unaffected area to establish a local
background. This area was located near Cora Island, northeast of the survey area. These
measurements were used to determine the statistical significance for any excess eRa and
the results are represented in a product called a "sigma plot." One sigma represents one
standard deviation from expected background levels. While subsurface concentrations of
gamma-emitting isotopes can be detected by the instrumentation, self- shielding of the
ground limits its effective detection to a depth of about 30 centimeters or 12 inches
(Bristol, 1983).

* A "normal background" area was selected by the ASPECT subject matter experts to be an area northeast
of the site where no known contaminants exist.

Page 1 of22


-------
Coldwater Creek Survey

May 2013

2.0 Description of the Coldwater Creek Survey Area

Figure 1: Coldwater Creek area consists of an irregular shape area covering over 5,000 acres (8
square miles). It was separated into three sections designated A, B, and C for convenience.

The St. Louis FUSRAP sites are comprised of the "Downtown" site located at the
Mallinckrodt Chemical Plant in downtown St. Louis and the "North County" sites located
near the Lambert St. Louis International Airport in St. Louis, Missouri. The North
County sites consist of three areas previously used for storing radioactive and other
wastes from uranium processing operations conducted by the Atomic Energy
Commission (AEC) and its successor, DOE. None of the three areas is now owned by the
Federal Government.

The St. Louis Airport Site (SLAPS) area covers 21.7 acres immediately north of Lambert
St. Louis International Airport, approximately 15 miles northwest of downtown St. Louis.
It is bounded by a railroad track, Coldwater Creek (Figure 1), and McDonnell Boulevard.
Radioactive metal scrap and drums of waste were stored in the SLAPS area in uncovered
piles from 1947 to the mid-1960s, when they were transferred 0.5 mile northeast to the
Hazelwood Interim Storage Site (HISS) area. Buildings in the airport area were razed,
buried, and covered with clean fill after 1967. In 1969, the land was conveyed to the
Lambert St. Louis Airport Authority.

HISS and the Futura Coatings Co. plant cover 11 acres adjacent to Latty Avenue,
Coldwater Creek, and Hanley Avenue. In 1966, Continental Mining and Milling Co.
acquired the property and recovered uranium from wastes purchased from AEC's St.

Louis operations. In 1967, the company sold the property and by 1973, most processing
residues had been removed. LJnder the direction of the Nuclear Regulatory Commission
(NRC), the present owner excavated contaminated soil and stored it in two large piles in
the eastern portion of the 11 acres. Since the 1970s, Futura Coatings, a manufacturer of
plastic coatings, has leased the western portion.

Page 2 of21


-------
Coldwater Creek Survey

May 2013

Congress transferred responsibility for FUSRAP site characterization and remediation to
the USACE in October 1997 as part of the Energy and Water Development
Appropriations Act of 1998. USACE is remediating the remaining sites within the
framework of the Comprehensive Environmental Response, Compensation, and Liability
Act (CERCLA) and the National Oil and Hazardous Substances Pollution Contingency
Plan (NCP). While the DOE retains responsibility for FUSRAP, USACE implements the
program under a USACE/DOE Memorandum of Understanding.

The St. Louis FUSRAP cleanup began as follows:

•	March 1974 - AEC established FUSRAP

•	October 4, 1989 - The EPA listed St. Louis FUSRAP Downtown and North
County sites on the NPL

•	June 29, 1990 - DOE and the EPA signed a Federal Facilities Agreement
committing DOE to clean up low-level radioactive-contaminated soils at the
Downtown and North County sites

•	August 27, 1998 - St. Louis Corps District issued a Record of Decision (ROD)
for the Downtown sites

•	September 5, 2005 - St. Louis Corps District issued a ROD for the North County
sites

The remedy for the St. Louis FUSRAP RODs involves USACE contractors excavating
radioactively-contaminated soils from numerous private and municipally owned
properties and shipping these soils by rail car to disposal facilities in Idaho or Utah. Soil
excavation work is ongoing in several locations in the Downtown and North County
properties. USACE excavated 177,000 cubic yards of soil from the Downtown sites and
852,000 cubic yards of soil around the North County sites through September 2012.

Investigations by EPA, DOE and USACE have attributed potential radiological
contamination in Coldwater Creek to runoff or windblown migration of the prior storage
of uranium-processing residues and wastes from the North County portion of the St.
Louis FUSRAP sites. USACE has removed the North County sources of these wastes,
which came from ore-processing activities at the Downtown portion of the St. Louis
FUSRAP sites. The USACE conducts bi-annual sediment and water sampling at six
different locations in Coldwater Creek as part of its environmental monitoring program.
USACE reviews and evaluates data in its annual environmental monitoring reports.

Although USACE has sampled sediment and water along Coldwater Creek since 1998,
some data gaps exist. As part of the plan to work from upstream to downstream, the
USACE sampled Coldwater Creek from McDonnell Boulevard to Frost Avenue in
October and November 2012. The results of the sampling will be summarized in a report
expected at the end of 2013. In addition, USACE is developing a sampling plan for the
portion of Coldwater Creek from Frost Avenue to St. Denis Bridge. Once the sampling
plan has been issued, the USACE will begin sampling this stretch of the creek. The
results will determine the density of sampling required throughout the remainder of the
creek to the mouth of the Missouri River. The purpose of the final round of sampling will

Page 3 of21


-------
Coldwater Creek Survey	May 2013

be to confirm the creek meets the North County ROD's cleanup requirements or to
identify and quantify any material requiring removal to meet these requirements.

Page 4 of 21


-------
Coldwater Creek Survey

May 2013

3.0 Natural Sources of Background Radiation

Naturally occurring radioactivity originates from cosmic radiation, cosmogenic
radioactivity, and primordial radioactive elements that were created at the beginning of
the earth about 4.5 billion years ago. Cosmic radiation consists of very high-energy
particles from extraterrestrial sources such as the sun (mainly alpha particles and protons)
and galactic radiation (mainly electrons and protons) and contributes to the total radiation
exposure on earth. The intensity of cosmic radiation increases with altitude, doubling
about every 6,000 ft, and with increasing latitude north and south of the equator. The
cosmic radiation level at sea level is about 3.2 |iR/h and nearly twice this level in
locations such as Denver, CO. (Grasty, et al., 1984).

Cosmogenic radioactivity results from cosmic radiation interacting with the earth's upper
atmosphere. Since this is an ongoing process, a steady state has been established
whereby cosmogenic radionuclides (e.g., 3H and 14C) are decaying at the same rate as
they are produced. These sources of radioactivity were not a focus of this survey and
were not included in the processing algorithms.

Primordial radioactive elements found in significant concentrations in the crustal material
of the earth are potassium, uranium and thorium. Potassium is one of the most abundant
elements in the Earth's crust (2.4% by mass). One out of every 10,000 potassium atoms is
radioactive potassium-40 (40K) with a half-life (the time it takes to decay to one half the
original amount) of 1.3 billion years. For every 100 40K atoms that decay, 11 become
Argon-40 (40A) and emit a 1.46 MeV gamma-ray.

Uranium is ubiquitous in the natural environment and is found in soil at various
concentrations with an average of about 1.2 pCi/g. Natural uranium consists of three
isotopes with about 99.3% being uranium-238 ( U), about 0.7% being uranium-235

235	234

( U), and a trace amount being uranium-234 ( U). Thorium-230 and Radium-226, as
decay products of Uranium-238 would be expected to have the same activity
concentrations as background Uranium-238 except that in some instances, changes in
soil chemistry may cause one species to migrate with the groundwater and disrupt the
local equilibrium so that the concentrations of Ra-226 and Th-230 may differ slightly
from the U-238 concentration. The ninth decay product of Uranium-238 is Bismuth-214
which is used to estimate the uranium present since it is relatively easy to detect.
Bismuth-214 has a very short half-life relative to Ra-226, Th-230 or U-238, therefore it
can be used to infer the presence of Ra-226, Th-230, and U-238 for airborne applications.
When it is used to estimate these isotopes, the precursor designator "e" (which means
equivalent) is used to identify that a decay product was used to estimate the Ra-226, Th-
230, or U-238 levels and is reported as eRa, eTh, and eU accordingly. See Appendix 1 for
the Uranium decay chain.

Thorium-232 is the parent radionuclide of one of the four primordial decay chains. It is
about four times more abundant in nature than uranium and also decays through a series
of decay products to a stable form of lead. Thorium-232 is not part of the Uranium decay
chain. The thorium content of rocks ranges between 0.9 pCi/g and 3.6 pCi/g with an
average concentration of about 1.3 pCi/g (Eisenbud, 1987). The ninth decay product,

Page 5 of21


-------
Coldwater Creek Survey

May 2013

208

thallium-208 ( Tl), is used to estimate the presence of thorium by its 2.61 MeV gamma-
ray emission.

All these primordial radionuclides are present in varied concentrations in building
materials which make-up part our naturally occurring radioactive background (Table 1)
(NCRP, 1987). Other radiation sources that contribute to our external radiation include
nuclear fallout and man-made radiation such as medical and industrial uses of radiation
or radioactive sources.

Table 1: Average concentrations of uranium and thorium in some building materials
Material	Uranium-238	Thorium-232

(pCi/g)	(pCi/g)

Granite

1.7

0.22

Sandstone

0.2

0.19

Cement

1.2

0.57

Limestone concrete

0.8

0.23

Sandstone concrete

0.3

0.23

Wallboard

0.4

0.32

By-product gypsum

5.0

1.78

Natural gypsum

0.4

0.2

Wood

-

-

Clay brick

3

1.2

Page 6 of 21


-------
Coldwater Creek Survey

May 2013

4.0	Survey Equipment and Data Collection Procedures

4.1	Radiation Detectors

The radiological detection technology consisted of two RSX-4
Units (Radiation Solutions, Inc.. 386 Watline Avenue,

Mississauga, Ontario, Canada) (Figure 2). Each unit was
equipped with four 2"x4"xl6" thallium-activated sodium
iodide (NaI[Tl]) scintillation crystals.

The Radiation Solutions RSX-4 unit was used during this
survey for airborne detection and measurement of low-level
gamma radiation from both naturally occurring and man-made
sources. It can also be used for ground-based measurements.

These units use advanced digital signal processing and
software techniques to produce spectral data equivalent to
laboratory quality. The unit is a fully integrated system that includes an individual high
resolution (1,024 channel) advanced digital spectrometer for each detector. A high level
of self diagnostics and performance verification routines such as auto gain stabilization
are implemented with an automatic error notification capability, assuring that the
resulting maps and products are of high quality and accuracy.

4.2	Flight Parameters

The ASPECT airplane used the following flight procedures for data collection on March
8,2013:

Altitude above ground level (AGL):	500 feet

Target Speed:	110 knots (125 mph)

Line Spacing:	500 feet

Data collection frequency:	1 Hz for radiological survey

Figure 2: USX4 unit
showing four detector
locations.

Page 7 of21


-------
Coldwater Creek Survey

Figure 3: Flight lines for the radiological survey over the Coldwater Creek site.

For environmental radiation surveys using a fixed-wing airplane, flying height above
ground level has been more or less standardized at 400 feet (IAEA 1991, 2003) and 5).
ASPECT target height for this survey was 500 feet to permit safer flying conditions.
Aerial and ground-based surveys collected over phosphate mines in central Florida
provided evidence that the increased altitude flight parameters have no significant effect
on the airplane sensitivity or resolution for environmental surveys (Cardarelli et al.,
2011a, 2011b).

5.0 Data Analyses

A unique feature of the ASPECT chemical and radiological technologies includes the
ability to process spectral data automatically in the airplane with a full reach back link to
the program QA/QC program. While data are generated in the airplane using automated
algorithms, a support data package is extracted by the reach back team and independently
reviewed for scientific validity and confirmation. The following sections detail the
analyses completed for this survey.

Page 8 of21


-------
Coldwater Creek Survey

May 2013

5.1 Radiological

Aerial gamma spectroscopy analyses have several distinctive considerations that must be
addressed in order to obtain accurate and meaningful products. Due to the unique
interactions of gamma rays with matter, special techniques are used to process the data.
For a uranium/radium survey, care must be taken to account for the background levels of

uranium/radium. This process was described in Section 3.
The ASPECT measures gamma radiation from Bismuth-
214 which is the ninth decay product in the Uranium-238
decay chain because Uranium-238 is not a strong gamma
emitter. In this survey, Bismuth-214 most likely indicates
the presence of Radium-226 (the fifth decay product of
Uranium-238) rather than Uranium-238 since the original
uranium ore was chemically separated from the rest of its
decay products. The separation process invalidates a key
assumption in the algorithms used to estimate equivalent
uranium concentrations; therefore, throughout this report
"equivalent radium" will be reported instead of equivalent
uranium.

Several environmental factors, such as moisture, may
significantly affect the detector response. Specifically,
precipitation disturbs the equilibrium of the uranium decay
chain and soil moisture actually shields some of the
gamma rays and prevents them from reaching the
detectors. There are several similar considerations that are
discussed in Appendix II.

In the days leading up to the survey, the St. Louis area had
received significant snowfall. During the survey, the
snowfall had melted, but the ground was likely fairly
saturated. This additional moisture in the ground would
serve as a partial shield and reduce the intensity of
radiation reaching the detectors. A 10 percent increase in
soil moisture would decrease the total count rate by about
10 percent. The higher than average energy from
Bismuth-214 would be slightly less affected, because soil
moisture affects the detection of lower energy gamma rays
more than higher energy gamma rays.

Radiological spectral data are collected every second along
with GPS coordinates and other data reference
information. These data are subject to quality checks
within the Radiation Solutions internal processing
algorithms (e.g. gain stabilization) to ensure a good signal.
If any errors are encountered with a specific crystal during
the collection process, an error message is generated and

I

Upload data to
ASPECT servers for
postprocessing

Page 9 of 21


-------
Coldwater Creek Survey

May 2013

the data associated with that crystal are removed from further analyses.

Prior to the survey, the RSX-4 units go through a series of internal checks. When
powered up, the crystals go through an automated gain stabilization process. The process
uses naturally occurring radioelements of potassium, uranium, and thorium to ensure
proper spectral data collection. If no problems are detected, a green indicator light
notifies the user that all systems are good. A yellow light indicates a gain stabilization
issue with a particular crystal. This can be fixed by waiting for another automatic gain
stabilization process to occur or the user can disable the particular crystal via the
RadAssist Software application. A red light indicates another problem and would delay
the survey until it can be resolved.

The "background data" in this context includes radiation contributions from radon,
cosmic, and airplane sources. These are unwanted contributions to the radiation
measurements and must be subtracted from the raw measurements to properly estimate
radiation contributions from terrestrial sources only. Ideally, these data are collected over
water at the survey altitude but when a large body of water does not exist, research has
shown that an acceptable alternative is to collect data 3,000 ft above the ground (AGL)
(Bristow, 1983). At this altitude, atmospheric attenuation reduces the terrestrial radiation
to a negligible level but is still low enough that cosmic radiation is not significant.

A "test line" in this context is flown at survey altitude near the survey area. The line is
not expected to contain any known elevated concentrations of naturally occurring
radioactive material (NORM) or man-made radionuclides. For this survey, an area near
Cora Island, west of the site, was used for this purpose. Hence, this test line serves as the
natural background area (after the radon, cosmic, and airplane sources are subtracted)
from which the survey data is compared to determine if any statistical anomalies occur
within the survey area.

The calibration coefficients were determined based on methodology published by the
International Atomic Energy Agency (IAEA, 2003).

One of the possible software programs available to the ASPECT team for processing
radiological data is the Environment for Visualizing Images (ENVI) code. For this
survey, ENVI® Version 5.0; ASPECT Version 9.1.1.2, Build 1302282009 (Exelis Visual
Information Solutions, Boulder, CO) was used to produce excess eRa sigma point plots

214

showing locations where Bi was out of balance with the surrounding environment.
The process is depicted below.

Page 10 of 21


-------
Coldwater Creek Survey

May 2013

ENVI ASPECT Method

I

Live time correction

I

Subtract cosmic and airplane background
contribution (3,000 ft AGL)

I

"Test line" (determines "normal")
Height correction (|i=0.0018 m-1)
Calculate 214Bi ROI K-value (median)

I

Subtract radon contribution (test lines)

I

214

Determine net count rate for Bi and
standard deviation (sigma value; o)

I

Determine Sigma Values
(<-6o, -6 to -4; -4 to -2; -2 to 2; 2 to 4, 4
to 6, >6o)

I

Create excess eRadium sigma plots

The excess eRa sigma plots are used to help determine whether the detected radiation
associated with the Bi-214 is consistent with areas known not to contain any elevated
radiation signatures, e.g. a background area. Because the uranium/radium concentration
will vary slightly from point to point, a statistical analysis is used to help make this
determination. The first step of this process is to determine the background variation.
This is done by measuring an area that is close to the site but not contaminated by the site
or containing any similar contaminants from other sources. All of the site measurements
are then compared to this to make sure the variation is within the variation of the
background data. Points that are noticeably different from the background points are
likely to be of man-made origin. Excess eRa sigma points were determined using an
algorithm based on the assumption that natural background radioisotope contributions are
stable over large geographical areas. This will result in a spectral shape that remains
essentially constant over large count rate variations.

ASPECT used the ENVI code analysis wherein a background "test line" is flown with
similar characteristics in an area physically close to the survey location but not affected
by the contamination. This background is used to compare the readings by statistical
methods. For this survey, the area was near Cora Island just northeast of the site.

Page 11 of 21


-------
Coldwater Creek Survey

May 2013

214

To determine excess radium count rate, the region-of-interest (ROI) around Bi (1659
keV to 1860 keV) is compared to the ROI represented by nearly the entire spectrum,
called the Total Count ROI (36 keV to 3,027 keV). The count rate ratio between these
windows (e.g., Uranium ROI / Total Count Rate ROI) is relatively constant and is
referred to as the "K" value. A K-value was determined from the "test line" data
collected before and after each survey. The median K-value (e.g., most common Re-
value) was used in the algorithm to determine excess eRa.

K-value = 	Count rate in tarset region-of-interest

Count rate in "Total Count" region-of-interest

Excess activity can be estimated using the following formula:

Excess eRa activity = Measured eRa activity - Estimated eRa activity

Where:

Measured eRa activity = the measured count rate within the eRa ROI during the survey

Estimated eRa activity = K-value * measured count rate in Total Count ROI during the
survey

The equation for excess activity becomes:

EXCESS eRa = Measured eRa ROI - (K * Measured Total Counts ROI)

The most likely value of net "excess eRa" should be zero, and since radiological
disintegrations are randomly occurring events, the second-by-second "excess eRa" results
are statistically distributed about the mean in a normal Gaussian distribution (Figure 5).

Page 12 of 21


-------
Coldwater Creek Survey

May 2013

Normal Gaussian Linear Distribution

/I

h-

P(x)

-	50% (PE) -

-	68.27% (ox) -
. 90 % _

-	95.45 %(2a) -

-	99.7 % (3o) -

Standard deviation (g, sigrna)
represents the spread of the
data about the mean. In this
survey, the mean value (net
"eRa") was zero.

1	g = 68.27% of the data

2	g = 95.45% of the data

3	g = 99.73% of the data

4	g = 99.99366% of the data

5	o = 99.99994% of the data

6	o = 99.999999% of the data

Figure 4: Normal Gaussian Distribution and associated confidence intervals.

Every measurement was scored according to its "sigma" value and color coded according
to the ranges in Figure 5. The color code and range were arbitrarily selected to limit the
risk of false positives to 1 in about 15,800,000 samples (greater than or less than 6
sigma).

Sigma Values (Excess Bismuth-214)

Less than-6.0

^^-2.0 to +2.0 Greater than +6.0

^ -6.0 to-4.0

<^+2.0 to +4.0

^ -4.0 to-2.0

((§)) +4.0 to +6.0

Figure 5: Standard Deviation Legend for Excess eRadium

6.0	Results

This survey on March 8, 2013 covered over 8 square miles of land and consisted of about
2,200 radiological data points.

6.1	Radiological Results

The radiological product consisted of an eRa sigma plot, which represents the number of
standard deviations from a normal background.

Page 13 of 21


-------
Coldwater Creek Survey

May 2013

6.1.1 eRa Sigma Plots

Since uranium (and radium) is a naturally occurring radionuclide and is ubiquitous in
nature, an analysis was conducted to determine the statistical significance of any
deviation from naturally occurring background levels. The analysis is referred to as a
sigma plot and is discussed in Section 5. Areas on a sigma plot with values greater than 4
sigma are very likely to contain uranium or its decay products in concentrations greater
than background, while values greater than 6 sigma almost certainly indicate above
background levels for uranium and its decay products. Of the nearly 2,200 data points
collected in this survey, none was within 4 to 6 sigma (standard deviations) from the
mean value and none was greater than 6 sigma from the mean.

Table 2 summarizes the sigma plot results for excess eRa for the entire survey Coldwater
Creek area. Approximately 94 percent of the area surveyed was below the 2 sigma
threshold. Less than 6 percent of the surveyed area fell between 2 and 4 sigma,
accounting for all of the data taken. No data points indicated variation from background
above the 4 sigma level. All areas were consistent with natural background.

Table 2: Statistical data of eRa results for each survey area.

Fit.













Block

Area

# Data

< 2 Sigma

> 2 Sigma

>4 Sigma

>6 Sigma

1

Coldwater Creek A

891

834

57

0

0

2

Coldwater Creek B

462

441

21

0

0

3

Coldwater Creek C

822

777

45

0

0



Totals

2,175

2,052

123

0

0



94.3%

5.7%

0%

0%

Page 14 of 21


-------
Coldwater Creek Survey

Figure 6: Excess eRadium Sigina Plot
Coldwater Creek Survey
March 8, 2013

May 2013

^Florissant

'©'Spanish Lak^

© 2013 Google

-0 Berkeley

'.lit 38 769515 Ion -90'?58646'

Sigma Values (Excess Bismuth-214)

Less than-6.0

(<§J> -2.0 to +2.0 Greater than+6.0

^ -6.0to-4.0

(@) +2.0 to +4.0

^ -4.0 to -2.0

+4.0 to +6.0

Flight Parameters

500 ft altitude
500 ft line spacing
110 knots

1 second acquisition time

All areas were consistent with natural background.

This image should not be used independently to assess potential health risks.
Additional information is necessary to make appropriate health-related decisions.

Page 15 of 21


-------
Coldwater Creek Survey

May 2013

6.2 Electronic Data

Access to the electronic data can be provided by contacting:

Matthew Jefferson

Superfund Remedial Project Manager for St. Louis FUSRAP/Coldwater
Creek

EPA Region 7

Jefferson.Matthew@epa. gov

Page 16 of 21


-------
Coldwater Creek Survey

May 2013

Appendix I : Uranium Decay Chain

Page 17 of 21


-------
Coldwater Creek Survey

May 2013

Appendix II

Discussion about radiological uncertainties associated with airborne
systems.

Ideally the airborne radiation measurements would be proportional to the average surface
concentrations of radioactive materials (mainly NORM). However, there are several
factors that can interfere with this relationship causing the results to be over- or under-
estimated, as described below. Additionally, two other sections in this Appendix discuss
how airborne data should be interpreted and compared to ground-based surface
measurements.

Background radiation

Airborne gamma-spectroscopy systems measure radiation originating from terrestrial,
radon, airplane, and cosmic sources. To obtain only the terrestrial contribution, all other
sources need to be accounted for (subtracted from the total counts), especially for this
survey where small differences are important. Radon gas is mobile and can escape from
rocks and soil and accumulate in the lower atmosphere. Radon concentrations vary from
day to day, with time of day, with weather conditions (e.g., inversions and stability class),
and with altitude. It is the largest contributor among background radiation and its decay

214

product, Bi, is used to estimate radium and uranium concentration in the soil. Radon is
normally accounted for in the processing algorithm by flying specific test lines before
and after each survey and comparing the results. Cosmic and airplane radiation (e.g.,
instrument panels and metals containing small amounts of NORM) also provide a small
contribution to the total counts. These are accounted for in the processing algorithm by
flying a "high-altitude" or "water" test line and subtracting these contributions for the
survey data.

Secular Equilibrium Assumption

Secular equilibrium is assumed in order to estimate thorium or uranium concentrations

208	214

from one of its decay products, T1 or Bi respectively. Secular equilibrium exists
when the activity of a decay product equals that of its parent radionuclide. This can only
occur if the half-life of the decay product is much shorter than its parent and the decay
product stays with its parent in the environment. In this case, the measurement of 214Bi
gamma emission is used to estimate the concentration of its parent radionuclide, uranium,

222

if one assumes all the intermediate radionuclides stay with each other. However, Rn is
a noble gas with a half-life of 3.8 days and may de-gas from soils and rocks fissures due
to changes in weather conditions. Due to the relatively long half-life (relative to 214Bi)
and the combined effect of radon gas mobility and environmental "chemical" migration,
it is not certain whether the secular equilibrium assumption is reasonable. In addition,
human intervention in this natural chain of events may have caused an increased
uncertainty in uranium concentration estimates. This becomes more complex with
uranium ore waste materials, where the uranium has been extracted and the resulting
waste materials contain mostly uranium decay products, e.g. radium. In this situation, the

Page 18 of 21


-------
Coldwater Creek Survey

May 2013

eRa concentration would be a better estimate for radium concentration rather than
uranium concentrations, as is the case in this survey.

Atmospheric Temperature and Pressure

The density of air is a function of atmospheric temperature and pressure. Density
increases with cooler temperatures and higher pressures, causing a reduction in detection
of gamma-rays. This reduction in gamma-ray detection is called attenuation and it is also
a function of the gamma-ray energy. Higher energy gamma-rays are more likely to reach

214

the detectors than lower energy gamma-rays. For example, 50% of the Bi 1.76 MeV
gamma-rays will reach the detector at an altitude of 300 ft whereas only 44% of the 40K
1.46 MeV gamma-rays will reach the detector. Temperature and pressure changes
contribute little to the overall uncertainties associated with airborne detection systems as
compared to other factors. Despite the nominal correction, the ASPECT program
accounts for temperature and pressure effects.

Soil moisture and Precipitation

Soil moisture can be a significant source of error in gamma ray surveying. A 10%
increase in soil moisture will decrease the total count rate by about the same amount due
to absorption of the gamma rays by the water. Snow cover will cause an overall
reduction in the total count rate because it also attenuates (shields) the gamma rays from
reaching the detector. About 4 inches of fresh snow is equivalent to about 33 feet of air.
There was no significant precipitation during this survey; however, the ground was likely
saturated from recent snow melt.

Topography and vegetation cover

Topographic effect can be severe for both airborne and ground surveying. Both airborne
and ground-based detection systems are calibrated for an infinite plane source which is
referred to as 2% geometry (or flat a surface). If the surface has mesas, cliffs, valleys, and
large height fluctuations, then the calibration assumptions are not met and care must be
exercised in the interpretation of the data. Vegetation can affect the radiation detected
from an airborne platform in two ways: (1) the biomass can absorb and scatter the
radiation in the same way as snow leading to a reduced signal, or (2) it can increase the
signal if the biomass concentrated radionuclides found in the soil nutrients are present in
the leaves or surfaces of the vegetation.

Spatial Considerations

Ground-based environmental measurements are usually taken 3 ft above the ground with
a field of view of about 30 ft2. The ASPECT collected data at about 500 ft above the
ground with an effective field of view of about 10 acres. These aerial measurements
provide an average surface activity over the effective field of view. If the ground
activity varies significantly over the field of view, then the results from ground- and
aerial-based systems may not agree. It is not unusual to have differences as much as
several orders of magnitude depending on the survey altitude and the size and intensity of
the source material. For example, in the figures below, if the "A" circle represents the

* Attenuation coefficients of 0.0077m"1 for 1.76 MeV and 0.0064m"1 for 1.46 MeV.

Page 19 of 21


-------
Coldwater Creek Survey

May 2013

detector field of view and the surrounding area had no significant differences in surface
activity, a 500 ft aerial measured could correlate to a ground-based exposure-rate of 3.5
|iR/h, However, if all the activity was contained in a small area such as a single small
structure containing uranium waste materials (represented by the blue dot within the field
of view of "B"), a 500 ft aerial measurement may still provide the same exposure-rate
measurement but the actual ground-based measurements could be as high as 3,150 |iR/h.

Detector Field of View

Concentration A = Concentration B

Aerial measurement is	Aerial measurement

a good indicator of	will not capture

average ground	differences in smaller

activity.	areas of intense

activity.

Illustration of aerial measurement capabilities and interpretation of the results

Comparing ground samples and airborne measurements

Aerial measurements are correlated to ground concentrations through a set of calibration
coefficients. The ASPECT calibration coefficients for exposure-rate, potassium,
uranium, and thorium concentrations were derived from a well characterized
"calibration" strip of land near Las Vegas, Nevada. In-situ gamma spectroscopy and
pressurized ionization chambers measurements were used to characterize the area. One
must exercise caution when using a laboratory to analyze soil samples to verify or
validate aerial measurements because differences will occur. In addition to local
variations in radionuclide concentrations, which are likely to be the most significant
issue, differences may arise due to laboratory processing. Laboratory processing
typically includes drying, sieving and milling. These processes remove soil moisture,
rocks and vegetation, and will disrupt the equilibrium state of the decay chains due to
liberation of the noble gas radon. Thus reliance on 208T1 and 214Bi as indicators of 232Th

238

and U (as is assumed for aerial surveying) is made more complex. In addition, aerial
surveys cannot remove the effects of vegetation on gamma flux. Intercomparisons must
minimize these differences and recognize the effects of differences that cannot be
eliminated.

Page 20 of 21


-------
Coldwater Creek Survey

May 2013

Geo-Spatial Accuracy

All aerial measurements collected by the ASPECT airplane are geo-coded using latitude
and longitude. The position of the airplane at any time is established by interpolating
between positional data points of a non-differential global positioning system and
referencing the relevant position to the time that the measurement was made. Time of
observation is derived from the airplane computer network which is synchronized from a
master GPS receiver and has a maximum error of one second . Timing events based on
the network running the Windows-based operating system and the sensor timing triggers
have a time resolution of 50 milliseconds, so the controlling error in timing is the network
time. If this maximum timing error is coupled to the typical ground velocity of 55
meter/sec of the airplane, an instantaneous error of 55 meters is possible due to timing.
In addition, geo-positional accuracy is dependent on the instantaneous precision of the
non-differential GPS system which is typically better than 30 meters for any given
observation. This results in an absolute maximum instantaneous error of about 80 meters
in the direction of travel.

For measurements dependent on airplane attitude (photographs, IR images), three
additional errors are relevant and include the error of the inertial navigation unit (INU),
the systemic errors associated with sensor to INU mounting, and altitude errors above
ground. Angular errors associated with the INU are less than 0.5 degrees of arc.

Mounting error is minimized using detailed bore alignment of all sensors on the airplane
base plate and is less than 0.5 degrees of arc. If the maximum error is assumed, then an
error of 1.0 degree of arc will result. At an altitude of 150 meters (about 500 feet) this
error translates to about 10 meters. Altitude above ground is derived from the difference
in the height above the geoid (taken from the GPS) from the ground elevation derived
from a 30 meter digital elevation model. If an error of the model is assumed to be 10
meters and the GPS shows a typical maximum error of 10 meters, this results in an
altitude maximum error of 20 meters in altitude error. If this error is combined with
attitude and the instantaneous GPS positional error (assuming no internal receiver
compensation due to forward motion), then an error of about 50 meters will result. The
maximum forecasted error that should result from the airplane flying straight and level is
+/- 130 meters in the direction of travel and +/- 50 meters perpendicular to the direction
of travel. Statistical evaluation of collected ASPECT data has shown that typical errors
of +/- 22 meters in both the direction of and perpendicular to travel are typical.

Maximum errors of +/- 98 meters have been observed during high turbulence conditions.

* The ASPECT network is synchronized to the master GPS time at system start-up. If the observed
network/GPS time difference exceeds 1 sec, at any time after synchronization, the network clock is reset.

Page 21 of 21


-------
Coldwater Creek Survey

May 2013

References

BRISTOW, Q. (1983). Airborne y-ray spectrometry in Uranium Exploration.
Principles and Current Practice. International Journal of Applied Radiation and
Isotopes 34(1), 199-229.

CARDARELLI, J., THOMAS, M., and CURRY, T. (2011). Environmental
Protection Agency airborne detection capabilities. Health Physics Society Midyear
Meeting: Radiation Measurements. Charleston, South Carolina, February 8, 2011.
(page 27) Abstract available at

http://www.hps.org/documents/2011 midyear final program.pdf Accessed on 10
April 2013.

CARDARELLI, J., THOMAS, M., CURRY, T., KUDARAUSKAS, P., and
KAPPELMAN, D. (2011). Aerial and Ground Radiological Surveys: Phosphate
Mines in January 2011. US EPA. Available at

http://epa.gov/region4/superfund/images/nplmedia/pdfs/coroiflrt2011.pdf Accessed
on 12 March 2013.

EISENBUD, M., (1987). Environmental Radioactivity: From Natural Industrial and
Military Sources. 3rd Edition. Academic Press, Inc., New York, NY.

GRASTY, R.L., CARDSON, J.M. CHARBONNEAU, B.W., HOLMAN, P.B.,
(1984). Natural Background Radiation in Canada, Geol. Surv. Can. Bull. 360.

IAEA (1991). International Atomic Energy Agency. Airborne Gamma Ray
Spectrometer Surveying. Technical Report Series No. 323. (International Atomic
Energy Agency, Vienna).

IAEA (2003). International Atomic Energy Agency. Guidelines for radioelement
mapping using gamma ray spectrometry data. Technical Document 1363. IAEA,
Vienna. Available at http://www-

pub.iaea.org/mtcd/publications/pdf/te 1363 web.pdf. Accessed on 12 March 2013.

NCRP (1987). National Council on Radiation Protection and Measurements.
Exposure of the Population in the United States and Canada from Natural
Background Radiation. NCRP Report 94 (National Council on Radiation Protection
and Measurements, Bethesda, Maryland).

Page 22 of 21


-------