vปEPA
United States
Environmental Protection
Agency
OCCURRENCE OF
RELEASES WITH
THE POTENTIAL TO
IMPACT SOURCES
OF DRINKING
WATER
Office of Water (MC-140)
EPA 817-R-21-001
February 2021
-------
Office of Water (MC-140) EPA 817-R-21 -001 February 2021
-------
Disclaimer
The Water Security Division of the Office of Ground Water and Drinking Water has reviewed and
approved the report "Occurrence of Releases with the Potential to Impact Sources of Drinking Water"
for publication in February 2021. This document is intended for use by the drinking water sector to
better understand the risk of potential releases into sources of drinking water. It may provide
information useful for conducting Risk and Resilience Assessments, as required under America's Water
Infrastructure Act (AWIA) of 2018.
AWIA, Section 2013 requires community water systems to conduct Risk and Resilience Assessments,
which must consider important system assets, including source water. This report demonstrates that
releases to sources of drinking water occurred at an average rate of 393 releases per year over the 10-
year study period. Furthermore, the report demonstrates this risk is not equally distributed across the
water sector - some community water systems are at substantially greater risk of releases to their
source water. To address this risk, this report recommends that community water systems conduct an
inventory of facilities that could release a harmful substance into their source water as part of their
AWIA Risk and Resilience Assessments. An important resource for developing contamination threat
inventories is Tier II chemical inventory data collected under the Emergency Planning and Community
Right to Know Act (EPCRA). AWIA, Section 2018 amended EPCRA to provide community water systems
with access to Tier II chemical inventory data.
This report is new. It does not modify or replace any previous EPA guidance documents. This document
does not impose legally binding requirements on any party. The information in this document is
intended solely to recommend or suggest and does not imply any requirements. Neither the U.S.
Government nor any of its employees, contractors or their employees make any warranty, expressed or
implied, or assumes any legal liability or responsibility for any third party's use of any information,
product or process discussed in this document, or represents that its use by such party would not
infringe on privately owned rights. Mention of trade names or commercial products does not constitute
endorsement or recommendation for use.
Questions concerning this document should be addressed to WQ SRSffiepa.gov or the following
contact:
Steve Allgeier
USEPA Water Security Division
26 West Martin Luther King Drive
Mail Code 140
Cincinnati, OH 45268
(513)569-7131
Allgeier.Steve@epa.gov
-------
Acknowledgements
The document was developed by the U.S. EPA Water Security Division, with additional support provided
under U.S. EPA contract EP-C-15-022. The following individuals contributed to the development of this
document:
Emily Smith, Corona Environmental Consulting
Margaret Kearns, Corona Environmental Consulting
Adam McKeagney, Corona Environmental Consulting
Peer review of this document was provided by the following individuals:
Jennifer Heymann, American Water
Dawn Ison, U.S. EPA, Office of Ground Water and Drinking Water, Water Security Division
Kevin M. Morley, American Water Works Association, Federal Relations
Richard Stuck, Greater Cincinnati Water Works
David Travers, U.S. EPA, Office of Ground Water and Drinking Water, Water Security Division
Karen Wirth, U.S. EPA, Office of Ground Water and Drinking Water, Drinking Water Protection
Division
-------
Table of Contents
ACKNOWLEDGEMENTS II
LIST OF FIGURES IV
LIST OF TABLES V
ABBREVIATIONS VI
SECTION 1: INTRODUCTION 1
1.1 Background 1
1.2 Objectives 4
1.3 Scope 4
SECTION 2: METHODOLOGY 6
2.1 Data Sources 6
2.2 Data Processing 7
2.3 Defining Zones of Concern 9
2.4 Identifying Significant Releases 10
2.5 Record Review 11
2.6 Limitations of the Methodology 12
SECTION 3: RESULTS AND DISCUSSION 13
3.1 Occurrence of Significant Releases 13
3.2 Temporal Occurrence 15
3.3 Occurrence by Geographic Location 16
3.4 Materials Released 21
3.5 Cause of Incidents and Responsible Parties 29
3.6 Occurrence of Releases by Source Water Zone of Concern 32
SECTION 4: SUMMARY AND CONCLUSIONS 37
SECTION 5: RECOMMENDATIONS 39
REFERENCES 40
GLOSSARY 43
APPENDICES 44
Appendix A: Fields Extracted from the NRC Reports and Used in Analysis 45
Appendix B: Rules to Standardize Names of Bodies of Water 46
Appendix C: Unit Conversion Multipliers 47
Appendix D: Process for Preparing Spatial Data 48
Appendix E: Materials Excluded from Analysis 49
-------
List of Figures
Figure 1. Example of NHD Flowline and Waterbody Representations 6
Figure 2. Example of a Surface Water Zone of Concern 9
Figure 3. Example of Groundwater Zones of Concern 10
Figure 4. Annual Occurrence of Incidents Potentially Impacting Sources of Drinking Water 15
Figure 5. Geographic Distribution of Incidents Potentially Impacting Sources of Drinking Water between January 1,
2010 and December 31,2019 (Map shows the location of 1,111 incidents with precise lat/long coordinates and
2,820INCIDENTS WITH ESTIMATED LAT/LONG COORDINATES) 16
Figure 6. Total Number of Incidents and Cumulative Volume Released per State between January 1,2010 and December
31, 2019 17
Figure 7. Total Number of Incidents per HUC-2 Region between January 1,2010 and December 31,2019 18
Figure 8. Incident Counts and Total Volume Released into Named Bodies of Water (BOW) and Unknown or Unnamed
Bodies of Water 20
Figure 9. Thirty-one Waterbodies with the Greatest Number of Incidents or Largest Cumulative Volume Released to
Water across the U.S. (A volume less than 0.5 kgal is displayed as 0 in the figure: Neches River (0.13 kgal), Cape
Fear River (0.40 kgal), and Calcasieu River (0.45 kgal)) 21
Figure 10. Total Number of Incidents and Cumulative Volume Released for each Material Category^ volume less than
0.5 KGAL IS DISPLAYED AS 0 IN THE FIGURE: CYANIDE COMPOUNDS (0.21 KGAL) AND RADIOLOGICAL (NO VOLUMES REPORTED)) . 22
Figure 11. Geographic Distribution of Releases of Refined Oil that Occurred between January 1, 2010 and December 31,
2019 25
Figure 12. Geographic Distribution of Releases of Crude Petroleum that Occurred between January 1,2010 and
December 31,2019 26
Figure 13. Geographic Distribution of Releases of Wastewater that Occurred between January 1,2010 and December
31, 2019 27
Figure 14. Geographic Distribution of Releases of Drilling Fluids that Occurred between January 1,2010 and December
31, 2019 28
Figure 15. Geographic Distribution of Releases of Coal Combustion By-Products that Occurred between January 1,
2010 and December 31,2019 29
Figure 16. Total Number of Incidents and Cumulative Volume of Material Released for each Cause between January 1,
2010 and December 31,2019 30
Figure 17. Total Number of Incidents and Cumulative Volume of Material Released by each Responsible Party Category
between January 1,2010 and December 31,2019 31
Figure 18. Total Number of Incidents and Cumulative Volume of Material Released byType of Incident between January
1,2010 and December 31,2019 32
Figure 19. Frequency of Releases to Source Water ZOCs between January 1,2010 and December 31,2019 (total number
ofSWZOCs = 4,929, total number of GWZOCs = 106,816) 34
Figure 20. Geographic Distribution of Releases in SW and GWZOCs between January 1,2010 and December 31,2019 .36
IV
-------
List of Tables
Table 1. Examples of Significant Source Water Contamination Incidents (2010 to 2019) 1
Table 2. Material Categories 8
Table 3. Filters Applied to Identify Records for Inclusion in the Analysis 14
Table 4. Twenty-five Counties with the Greatest Number of Incidents across the U.S 19
Table 5. Count of Releases Involving a Volume within the Indicated Range for Each Material Category^ volume
RELEASED LESS THAN 0.5 KGAL IS DISPLAYED AS 0 IN THE TABLE: CYANIDE COMPOUNDS (0.21 KGAL) AND RADIOLOGICAL (NO
VOLUMES REPORTED)) 24
Table 6. Most Commonly Released Materials for each Responsible Party Category 32
Table 7. Surface Water Zones of Concern that Experienced the Most Releases 35
Table 8. Source Water Zones of Concern that Experienced more than One Release and which have the Greatest Number
of Releases per Square Mile 35
v
-------
Abbreviations
ATSDR
Agency for Toxic Substances and Disease Registry
CWA-HS
Clean Water Act Hazardous Substances
CWS
Community Water System
GW
Groundwater
HSEES
Hazardous Substance Emergency Events Surveillance
HUC
Hydrologic Unit Code
kgal
One thousand gallons
lat/long
Latitude and Longitude coordinates
NHD
National Hydrography Dataset
NRC
National Response Center
NTSIP
National Toxic Substances Incidents Program
ORSANCO
Ohio River Valley Water Sanitation Commission
PCBs
Polychlorinated biphenyls
PHMSA
Pipeline and Hazardous Materials Safety Administration
SDWIS
Safe Drinking Water Information System
SW
Surface water
TRI
Toxics Release Inventory
U.S. CG
United States Coast Guard
U.S. CSB
United States Chemical Safety and Hazard Investigation Board
U.S. EIA
United States Energy Information Administration
U.S. EPA
United States Environmental Protection Agency
ZOC
Zone of Concern
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
Section 1: Introduction
1.1 Background
Releases of harmful chemicals through accidents or unpermitted discharges into sources of drinking
water can cause significant problems for public water systems and the communities they serve.
Potential consequences can include adverse impacts on public health, interruptions in water service,
loss of public confidence, increased treatment costs, damage to water system infrastructure, and cost to
mitigate the impacts of the release. Congress recognized the importance of this risk to source water by
including Section 2018 in America's Water Infrastructure Act which authorizes community water
systems to access hazardous chemical inventory data and requires that these systems receive prompt
notification of spills contaminating their source water (U.S. Congress, 2018).
1.1.1 Notable Releases to Source Waters
Several source water contamination incidents have been reported in the media, and a few notable
contamination incidents are summarized in Table 1. The amount of material released in these examples
ranged from 10,000 gallons to more than 11 million gallons. Materials released include coal ash, mine
waste, wastewater, and uncommon industrial chemicals like crude methylcyclohexane methanol. The
causes of the releases shown in Table 1 include equipment failure, operator error, and natural disasters.
Regardless of their specific conditions, all of the releases share one thing in common - they significantly
degraded the quality of a source of drinking water.
Table 1. Examples of Significant Source Water Contamination Incidents (2010 to 2019)
Year
Waterbody
Description
2011
Mulberry Fork, AL
(NRC# 975693)
On May 8, 2011, approximately 1.6 million gallons of untreated wastewater was
released from American Proteins into the Mulberry Fork, a source of drinking water
for the City of Birmingham. The release resulted from tornado damage to a
wastewater treatment basin. Subsequent releases from this same facility include a
release of 80,000 gallons of wastewater in May/June 2015 and a release of 900
gallons of sulfuric acid in August 2016 (Sack, 2016).
2014
Elk River, WV
(NRC#: 1070627)
On January 9, 2014, approximately 10,000 gallons of a mixture containing
methylcyclohexane methanol (MCHM) was released to the Elk River, due to
corrosion in an above ground storage tank. The Elk River is the drinking water
source for Charleston, WV. The contamination incident resulted in a "do not use"
order for approximately 300,000 residents for 4 to 9 days (U.S. CSB, 2016; Rosen
et al, 2014).
2014
Dan River, NC
(NRC#: 1073040)
On February 2, 2014, approximately 39,000 tons (11 million gallons) of coal ash
(containing arsenic, cadmium, lead, mercury, and other metals) was released into
the Dan River from the Duke Energy Dan River Steam Station near Eden, NC. The
release resulted from failure of a stormwater pipe that allowed the contents of the
coal ash impoundment to leak into the river for several days. The release impacted
water quality at several drinking water intakes in North Carolina and Virginia.
Testing of treated water at the downstream community water systems indicated that
National Primary Drinking Water Standards were met (U.S. EPA, 2014). However,
elevated concentrations in the source water may have exceeded these standards,
and no information was available for contaminants not regulated under the Safe
Drinking Water Act.
2015
Yellowstone River,
MT
(NRC# 1105969)
On January 17, 2015, a ruptured oil pipeline leaked approximately 40,000 gallons of
crude oil into the Yellowstone River in Montana. The release impacted the drinking
water source for the nearby town ofGlendive (Beker, 2015). Testing of treated
water in Glendive showed no contamination, however, residents reported odors of
diesel fuel prompting the system to issue a "do not use" notice (National Park Trips
Media, 2017).
1
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
Year
Waterbody
Description
2015
Cement Creek &
Animas River, CO
(NRC#: 1124824)
On August 5, 2015, approximately 3 million gallons of mine waste (containing
arsenic, lead, and other metals) was released from the Gold King mine complex to
Cement Creek near Silverton, CO. The release was due to a breach in the
containment structure that occurred during an inspection. Cement Creek is a
tributary of the Animas River which flows into the San Juan River near Farmington,
NM. The contaminant plume reached Lake Powell on August 12. Five drinking water
systems draw water from the Animas River, and these systems closed their intakes
and/or issued "do not use" notices. Advisories were issued for private domestic
wells along the Animas River (U.S. EPA, 2015a).
2017
Ohio River, KY /
OH
(NRC#: 1200030)
On December 19, 2017, an estimated 467,000 gallons of urea ammonium nitrate
was released when a barge suffered catastrophic failure. The release threatened
the drinking water supply for downstream utilities in Louisville, KY, Evansville, IN,
and Henderson, KY (ORSANCO, 2018). Staff from Louisville Water Co. and
ORSANCO monitored the river conditions daily and then hourly as the spill flowed to
Louisville. Strategic management of the intake rate helped the system avoid pulling
in water at peak contaminant concentration (Louisville Water Company, 2017).
NRC#: National Response Center report number
1.1.2 Previous Research
The examples listed in Table 1 were reported in widely distributed media, but such reporting is the
exception. Most releases into sources of drinking water receive scant attention outside of notification to
the responsible parties, responders, and ideally to affected community water systems. Because releases
to the environment are under-reported, there is no definitive assessment of the number or impact of
releases to water. However, several research efforts have attempted to characterize the occurrences of
releases that impacted sources of drinking water.
A research group at the University of Mississippi developed a database of releases into sources of
drinking water that occurred between 1990 and 2006. Two sources of information were used to
populate this database, the National Response Center (NRC) and Hazardous Substance Emergency
Events Surveillance (HSEES), both of which are described later in this section. A stated objective of the
project was to capture releases that impacted drinking water infrastructure; however, the project report
did not provide the methodology for making this determination (Zhu et al, 2009). The database is no
longer available from the project website.
During the development of the Clean Water Act Hazardous Substances Spill Prevention Proposed Rule,
U.S. EPA analyzed NRC records to identify releases involving Clean Water Act Hazardous Substances
(CWA-HS) over a 10-year period between 2007 and 2016. Over this period, a total of 285,867 incidents
were reported to the NRC, of which 9,416 (3.3%) involved the release of a CWA-HS with 3,140 (1.1%) of
these releases reaching water. This analysis reported that polychlorinated biphenyls (PCBs) were the
most commonly released CWA-HS, involved in 59% of CWA-HS releases that reached water. The next
four most frequently released CWA-HS were: sulfuric acid, sodium hydroxide, ammonia, and benzene
(U.S. EPA, 2018).
Several studies have focused on the release of oil and related materials to water. One such study
analyzed 6,622 spills from 21,300 unconventional oil and gas extraction wells in four states
(Pennsylvania, North Dakota, Colorado, and New Mexico) from 2005 to 2014. The U.S. Forest Service's
Forest to Faucets "index of importance as a source of drinking water" was used to evaluate the risk to
drinking water supplies. Releases were characterized with respect to location (state), material released,
and volume released. The most commonly released materials were: production waste streams (brine,
2
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
flowback, and produced water), crude oil, hydraulic fracturing solution, drilling waste (drilling mud,
cuttings, and drilling fluid), and production chemicals (hydrochloric acid, antifreeze, surfactants, and
glycol). The volume of material released ranged from 0.0001 to 991 kgal (Maloney et al, 2017).
U.S. EPA conducted a study to evaluate the potential impact of releases associated with hydraulic
fracturing on surface and ground waters. State databases were used to identify releases that occurred
between January 2006 and April 2012 in ten states with the most hydraulic fracturing activity reported
at the time of the study: Arkansas, Colorado, Louisiana, New Mexico, North Dakota, Oklahoma,
Pennsylvania, Texas, Utah, and Wyoming. Data sources were searched separately using a combination of
filters, keywords, and line-by-line reviews, with additional details provided by the state, service
company, and well operator where applicable. Of the approximately 36,000 release records identified,
24,000 (66%) were determined to be unrelated to hydraulic fracturing, and most of the remaining
12,000 (33%) records had insufficient data to make the determination. 457 release records could be
linked to hydraulic fracturing, of which 370 (81% of hydraulic fracturing-related spills) reported the
volume released, which ranged from fewer than 5 gallons to more than 1.3 million gallons. Fifty-six
percent of these 370 records involved a release volume less than 1,000 gallons, and accounted for only
3% of the total volume released by the 370 incidents. The majority (57%) of volume released came from
a single spill of 1.3 million gallons of flowback and produced water. Storage units were the most
common source of a release and failure of container integrity (e.g. holes, seal failures) were generally
associated with larger release volumes (U.S. EPA, 2015b).
A study evaluating oil releases over a 20-year period from 1980 to 2000 determined that crude oil
accounted for the greatest volume released while light fuels accounted for the greatest number of
releases (Etkin, 2004). In a later study, U.S. EPA Region 5 analyzed NRC records to characterize the
vulnerability of sub-watersheds in its states (Illinois, Indiana, Michigan, Minnesota, Ohio, and Wisconsin)
to releases of crude oil (Brody et al, 2012).
1.1.3 Release Tracking Databases
In addition to snapshot studies, some state and federal programs track source water spills over time.
The NRC, for example, is an emergency call center that fields initial reports of releases and forwards that
information to the appropriate federal or state agencies. The NRC posts release reports to their website
for every calendar year starting in 1990. These reports contain initial information about the incident,
and in most cases this information has not been validated or investigated by a response agency (U.S. CG,
2020). A number of studies have used the NRC as a primary source of information about releases,
including: Balasubramanian and Louvar, 2004; Etkin, 2004; Howard et al, 2008; Zhu et al, 2009; Brody et
al, 2012; and U.S. EPA, 2018.
The National Toxic Substances Incidents Program (NTSIP), managed by the Department of Health and
Human Services, Agency for Toxic Substances and Disease Registry (ATSDR) was active between 2010
and 2018. Release of toxic substances from seven states (Louisiana, New York, North Carolina, Oregon,
Tennessee, Wisconsin, and Utah) were captured in NTSIP. This program replaced a similar program,
Hazardous Substance Emergency Events Surveillance (HSEES), which was active from 1990 through 2009.
The HSEES monitored for incidents in 14 partner states (ATSDR, 2018).
The Toxic Release Inventory (TRI) is a program created under the Emergency Planning and Community
Right to Know Act. TRI requires industries that meet specific criteria to file an annual report
documenting releases of certain toxic chemicals to air, land, and water that may pose a threat to human
3
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
health and to the environment. TRI reporting is limited to a list of approximately 755 individual
chemicals and 33 chemical categories. TRI reporting limits change over time, complicating temporal
analysis of releases. Facilities must submit annual reporting forms for each chemical if they
manufacture, process, or otherwise use the chemical in amounts above established levels (U.S. EPA,
2020).
The Pipeline and Hazardous Materials Safety Administration (PHMSA) is managed by the U.S.
Department of Transportation and maintains a Hazmat Incident Database that contains information
from the Hazardous Materials Incident Report Form 5800.1. This database includes information on the
quantity of material released, mode of transportation, packaging information, and impacts of hazardous
materials released during transportation (PHMSA, 2020).
Collectively, these studies and data collection efforts provide valuable insight into the occurrence of
releases that impact sources of drinking water. However, the efforts described are either limited in
scope (i.e., a limited number of substances or a limited geographic region), or they do not differentiate
between releases to any media and releases specifically to sources of drinking water. A review of
published studies failed to yield a comprehensive, national study of releases into sources of drinking
water in the U.S.
1.2 Objectives
The objective of the study described in this report was to characterize the occurrence of releases into
sources of drinking water used by community water systems in the U.S.
Specifically, the study evaluated:
Temporal occurrence of releases between 2010 and 2019 (full calendar years)
Geographic occurrence of releases
Type and amount of material released
Responsible party and cause of releases
Distribution of the number of releases impacting individual community water systems
Furthermore, the results of this study are intended to inform Risk and Resilience Assessments, as
required under AWIA, Section 2013. One of the assets that must be considered in these assessments is
source water, and as shown by the results presented in this report, some community water systems face
a risk of spills and releases into their source of drinking water. Systems that have experienced source
water contamination incidents previously, or determine that they are at risk, should consider developing
an inventory of facilities that could release a harmful substance into their source water. An important
resource for developing contamination threat inventories is Tier II chemical inventory data collected
under the Emergency Planning and Community Right to Know Act (EPCRA). AWIA, Section 2018
amended EPCRA to provide community water systems with access to Tier II chemical inventory data.
1.3 Scope
The scope of this analysis was limited to the following:
Releases reported to the NRC. The NRC is a national call center that receives initial reports of
releases of any material into any medium. While the NRC is the most comprehensive source of
information about releases in the U.S., releases do occur that are not reported to the NRC.
4
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
Community water systems. This analysis was limited to releases into source water zones of
concern (described in Section 2.3) for community water systems, as defined in Section 1401(15)
of the Safe Drinking Water Act. Community water systems were considered in this analysis,
rather than all public water systems, because the former are required to conduct risk and
resilience assessments, and one asset that must be considered in these assessments is source
water.
5
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
Section 2: Methodology
2.1 Data Sources
Safe Drinking Water Information System
(SDWIS) was used to identify community water
systems from among the larger universe of
public water systems and provided information
such as the source water type (e.g., surface
water, ground water) and location of intakes
and wells for community water systems.
Media reports were used to investigate,
validate, and in some cases correct initial NRC
reports for significant releases (i.e., large
volume releases). Media reports from reputable
outlets were considered more reliable than the
NRC reports because the latter are preliminary
and often incomplete, while the former use
sources such as representatives from state and
federal response agencies and drinking water
systems to obtain details about the incident.
National Hydrography Dataset (NHD) served as
the primary source of information about
location of surface waterbodies. The high-
resolution NHD waterbody areas and flowlines
were used when available, otherwise medium-
resolution NHD flowlines and waterbody
boundaries were used. Figure 1 provides an
example of NHD waterbody and flowline
representations.
National Response Center (NRC) served as the
primary source of information about releases.
The NRC annual release reports for 2010
through 2019 were downloaded from the NRC
website (nrc.uscg.mil). A list of the fields
extracted from the NRC reports to support this
analysis is provided in Appendix A.
Figure 1. Example of NHD Flowline and Waterbody Representations
6
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
2.2 Data Processing
While the data fields in the NRC record are standardized, data entry errors occurred. Most errors involve
incorrect spelling and inconsistent naming of record attributes such as the names of waterbodies and
materials released. Additionally, details such as the name of the material released and the units for the
volume of material released are not standardized, which results in the use of synonyms or ambiguous
identifiers and several different volumetric or weight units. In some cases, important details about the
spill such as additional materials released and volumes of the materials released were included in the
free-text Incident Details field but not carried over into the appropriate, specific data fields. Finally, a
significant number of records were missing information important to the analysis, such as the precise
location of the release, the name of the material released, and volume of material released.
The following data processing was performed prior to analysis.
Populate missing information for the name and volume of material released by parsing and
searching the Incident Details field, which is a free-text field used to capture non-standard
information provided by the individual reporting the release.
Amount of material in water was assumed to be equal to the volume of material released if the
Incident Details field provided information to indicate the release likely occurred directly into a
waterbody and if the Amount of Material Released field was populated while the Amount in
Water field was empty. As an example, if a record reported that 1,000 gallons of diesel fuel was
released, and the Incident Details stated that the release was from a rail car that derailed into a
stream, it would be assumed that the Amount in Water was 1,000 gallons if this field was left
blank.
Names of bodies of water were corrected for spelling errors and standardized to a common
name for each waterbody. Rules used to standardize the names of bodies of water are provided
in Appendix B.
Names of materials released were corrected for spelling errors.
Material categories were developed to group similar materials together to support an aggregate
analysis for trends in the types of material released (see Table 2). Note that these material
categories are limited with respect to understanding potential consequences. Each specific
material has unique properties that will impact fate and transport, treatability, and public health
concerns.
Units for the amount of material released and amount in water were normalized to gallons
where possible. The conversion factors used are listed in Appendix C.
Location data was standardized by removing extraneous characters and applying algorithms to
identify a best address for the location of each release (when address information was provided
in the NRC record). The best address was then geocoded to support spatial analysis. The
methods used to prepare location data are described in Appendix D.
7
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
Table 2. Material Categories
Category
Materials Most Commonly Released that Fall within the Category
Acid
Sulfuric Acid, Hydrochloric Acid, Phosphoric Acid
Alcohol
Denatured Alcohol, Bourbon
Antifreeze
Ethylene Glycol, Propylene Glycol
Caustic Material
Sodium Hydroxide, Caustic Soda Solution
Chlorine
Sodium Hypochlorite, Chlorine
Coal Combustion By-Products
Coal Ash, Creosote, Fly Ash
Crude Petroleum
Crude Oil
Cyanide Compounds
Sodium Cyanide Solution
Drilling Fluid
Drilling Brine, Produced Water, Drilling Mud
Fertilizer
Fertilizer, Anhydrous Ammonia, Urea, Ammonium Nitrate Urea Solution
Firefighting Foam
Fire Fighting Foam, Fire Fighting Water, AFFF (Aqueous Film Forming
Foam)
Food Products
Milk, Vegetable Oil, Palm Oil
Metals and Metalloids
Arsenic, Lead
Mine Waste
Mine Waste, Mine Water
Organic Solvents
Toluene, Ethyl Alcohol, Ethanol, Acetone
Paint
Paint, Oil Based Paint
Pesticides/Herbicides
Dieldrin, Paraquat Dichloride, Diphenylamine, Insecticide
Radiological Materials
Radioactive Material, Uranium, Radium
Refined Oil
Automotive Gasoline, Fuel Oil, Diesel Oil, Hydraulic Oil
Salt Water
Saltwater
Transformer Oil
Transformer Oil, Mineral Oil, Polychlorinated Biphenyls
Unknown Material
Unknown Chemicals, Unknown Material
Wastewater
Sewage, Wastewater
8
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
2.3 Defining Zones of Concern
Zones of concern (ZOC) were developed using the locations
of surface water intakes and groundwater wells. Releases
with adequate location data were analyzed to determine
whether they occurred within a ZOC for a community water
system. The criteria for establishing surface water (SW)
ZOCs are shown in the callout box and are consistent with
criteria for establishing source water area delineations for
conducting a source water contamination threat inventory
(U.S. EPA, 2006).
The confidential version of SDWIS was used to identify 5,119 surface water intakes for community water
systems serving a population greater than 1,000 customers. SW ZOCs were delineated for 4,899 of these
surface water intakes, while ZOCs could not be delineated for 220 intakes because intakes were located
too far away from the NHD flowlines, on the shores of lakes or reservoirs, or in areas with highly
complex NHD flowlines. Three of the 220 surface water intakes for which a ZOC was not delineated
belong to community water systems serving more than 500,000 people. The intakes for these three
large systems are located in highly protected source water areas that contain no industry, chemical
storage (other than that maintained by the community water system), transportation routes, or
pipelines, and thus are unlikely to have experienced a release. The 220 SW ZOCs that did not delineate
properly were reviewed and 30 were selected for manual processing: 26 were selected because the
associated community water system has a population served greater than or equal to 100,000 or a
number of service connections greater than or equal to 30,000. An additional four SW ZOCs were
selected because the intakes are located in a watershed that experienced releases in other SW ZOCs.
Applying these adjustments, the total number of SW ZOCs considered in this analysis is 4,929 (4,899 that
automatically delineated plus 30 that were manually delineated).
Surface Water (SW) ZOCs: The ZOC for
each surface water intake extends 50
miles upstream, V* mile downstream,
includes all major tributaries, and
includes a Va mile buffer inland from
the waterbody area boundary (see
Figure 2 for an example).
20 miles
A PWS Intake Location
NHD Flowline
Surface Water Zone of Concern
Figure 2. Example of a Surface Water Zone of Concern
9
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
Guidance from U.S. EPA states that an "arbitrary fixed
radius" can be used as a first approximation for a source
water area delineation (U.S. EPA, 2006). Using this method,
106,816 GW ZOCs were delineated.
The SW ZOCs and GW ZOCs used in this study were defined to be conservative, meaning that they cover
a large area in order to capture most releases that could have impacted water quality at the intake or
wellhead. However, whether a specific release would impact water quality at the point of withdrawal
depends on several factors, such as volume of material released, characteristics of the material released,
size and flow of waterbody, etc. There have been large releases that have impacted utilities more than
50 miles downstream of the point of release. Conversely, small releases of certain contaminants may
have no appreciable impact on water quality within one mile from the point of release.
Groundwater (GW) ZOCs: The ZOC for
each groundwater well is defined by a
Zi mile radius around the well location
(see Figure 3 for an example).
Ground Water Zone of Concern
Wellhead Location
Figure 3. Example of Groundwater Zones of Concern
2.4 Identifying Significant Releases
In calendar years 2010 through 2019, a total of 281,141 releases were reported to the NRC, The
following criteria were used to identify releases with the potential to significantly contaminate a source
of drinking water.
Releases in which fewer than 100 gallons of material were released were excluded from the
analysis, with the exception of highly toxic materials as described in the third bullet.
Records for which the volume of material released was not reported were evaluated for
indicators that they had the potential to release a volume greater than 100 gallons. Specifically,
releases with records that did not report a volume, but which met any of the following criteria,
were included in this analysis:
o Release resulting in contamination of the water supply
o Release from a storage tank with a capacity greater than or equal to 1,000 gallons
10
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
o Release involving a freight train or freight car
o Release involving a barge or a tanker
o Release resulting in a closed rail track or waterway
o Release resulting in property damages
Releases of a highly toxic material were included in the analysis regardless of the amount of
material released (including records for which the volume released was not reported). The
materials identified in the NRC record were screened based on relative acute toxicity, and those
with the lowest toxicity threshold were categorized as highly toxic materials for this analysis.
The four classes of highly toxic materials identified in the NRC record include:
o Arsenic compounds
o Cyanide compounds
o Pesticides, insecticides, and herbicides
o Radionuclides
Releases with location information sufficient to geocode the record were analyzed to determine
if they fell within a ZOC for a community water system. Releases that occurred outside of a ZOC
were excluded from the analysis, with the exception of 14 unique incidents that were reported
to have impacted drinking water systems further than 50 miles downstream of the release
location.
Releases without location information were retained in the analysis if the volume released
exceeded 100 gallons, had the potential to release more than 100 gallons for records in which
the volume released was not recorded, or released a highly toxic material. However, other filters
were still applied to these records, as described in the following bullet.
Remaining records were screened to remove those with the following attributes:
o Reports generated during drills
o Releases that occurred outside of the U.S.
o Releases to air
o Releases that occurred offshore or in specific bodies of water that were verified to be
unconnected to a source of drinking water
o Releases involving materials unlikely to change water quality (e.g., sand, aggregate,
steel). A complete list of materials excluded can be found in Appendix E.
o Duplicate records
One release not captured in the NRC record, but reported in the media, was added to the
dataset for this analysis: a release of 794 kgal of oil and produced water in McKittrick, CA in
2019.
2.5 Record Review
Following the data processing steps and removal of records of releases unlikely to significantly impact a
source of drinking water, the quality of the remaining records was further assessed. Due to the volume
of records it was infeasible to review them all. Instead, the following methods were used to screen and
identify records for further review:
Records for releases of 2,000 gallons of material or more were reviewed. While most of these
releases were confirmed, a small number were not credible and thus removed from the analysis.
Reports deemed not credible generally involved volumes released greater than 1 million gallons
of material that could not be verified through another source, or releases of large volumes of
material rarely stored in large quantities (e.g., release of 100,000 gallons of radioactive waste).
Records for releases in which the Incident Description contained words such as "neighbor" or
"parked car" were reviewed. In most cases, records that contained these words or phrases
11
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
pertained to releases that were unlikely to significantly impact sources of drinking water (e.g., a
"neighbor" dumping used motor oil in a sewer or a "parked car" leaking fluids).
Records that updated information about an incident reported under another record identifier
were reviewed. If multiple NRC records pertaining to the same incident were identified,
information from the multiple records was consolidated into a single NRC record.
If information in an NRC record was deemed questionable or incomplete, research was conducted to
investigate the details of the incident. Information from the NRC record such as location, date, material
released, and responsible party was used to search media reports for corroborating information and
additional details. When available, reports from established media outlets were used to populate
missing fields or correct information provided in the initial NRC report. The details of more than 200
records were corrected using information identified in the Incident Description field or obtained from
media reports.
2.6 Limitations of the Methodology
The analysis considers only releases reported to the NRC, with the exception of one release that
was identified through media reports, as noted in Section 2.4.
NRC reports are preliminary and in most cases the information is not reviewed or corrected by
response agencies.
The NRC record likely underrepresents the total number of releases that occur. Anecdotal
information suggests that releases that are first reported to a 911 call center may not be
reported to the NRC. Additionally, less obvious releases, such as combined sewer overflows,
may not be reliably reported to the NRC.
NRC reports are often incomplete and missing important information such as the location of the
release, the material released, or the volume of material released.
Assumptions were made to identify releases that had the potential to significantly impact a
source of drinking water. Notably, NRC records for releases in which the reported amount of
material released was less than 100 gallons were removed from the analysis, except for releases
of highly toxic chemicals. In cases where the reported amount underestimated the actual
amount of material released, a significant release could have been incorrectly removed from the
analysis.
The criteria used to develop ZOCs were by necessity generic. It is possible that releases
significantly impacting a source of drinking water occurred outside a zone of concern.
Conversely, it is also possible that releases within a zone of concern did not significantly impact
the source water.
12
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
Section 3: Results and Discussion
Results from the analysis of releases that potentially impacted a source of drinking water are presented
in the following subsections:
3.1
Provides a summary of the number and volume of releases potentially impacting source
water
3.2
3.3
3.4
3.5
3.6
Presents the temporal occurrence of releases to water over the 10-year study period
Presents the geographic occurrence of releases to water
Presents the occurrence of releases to water involving different material categories
Presents the reported causes and parties responsible for releases to water
Presents distribution of release occurrences within SW and GW ZOCs
3.1 Occurrence of Significant Releases
The total number of releases reported to the NRC between January 1, 2010 and December 31, 2019,
plus one release identified outside of the NRC record, was 281,142. After the records were processed
according to the methodology described in Section 2, there were 5,806 records remaining in this
analysis. The impact of the various filters in reducing the number of records is shown in Table 3. The
filter resulting in the greatest reduction in the number of records was the "significant release" filter,
which consisted of the following criteria:
Records involving an unknown amount of material released during an incident that did not
involve large volume transportation (e.g., barge or rail transport) nor result in significant impacts
(e.g., closure of a waterway, contamination of the water supply). 144,497 records were removed
based on this criterion.
Records involving a known amount of material released but below the threshold of 100 gallons.
95,354 records were removed based on this criterion.
Records reporting the release of a highly toxic material, as described in Section 2.4, were
retained regardless of the amount of material released. There were 880 records involving the
release of a highly toxic material that were retained even though the amount released was less
than 100 gallons.
13
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
As shown in Table 3, the net number of records removed under the significant release criteria was
238,971 (85%).
Table 3. Filters Applied to Identify Records for Inclusion in the Analysis
Filter Criterion
# Records Removed
Records that did not meet criteria for a significant release or involve the release
of a highly toxic material
238,971 (85%)
Records of releases into bodies of water not used for drinking water
20,552 (7.3%)
Records of releases that occurred offshore
11,056 (3.9%)
Duplicate records
2,845 (1.0%)
Records that were generated during drills
993 (0.4%)
Records of releases involving materials that would not degrade water quality
549 (0.2%)
Records of releases that occurred outside of the U.S.
377 (0.1%)
Throughout this report, the term incident refers to a specific event that resulted in the release of at
least one material, while the term release refers to the release of a specific material. Each NRC record
relates to a unique release of a specific material. A single incident can result in multiple releases and
thus generate multiple NRC records. For example, an accident involving a tanker truck that spilled fuel
oil from its cargo tank, diesel fuel from its saddle tanks, and coolant from its radiator would generate
three NRC records relating to this single incident. The 5,806 NRC records identified in this analysis
correspond to 3,931 unique incidents, many of which resulted in multiple releases.
Of these 3,931 incidents:
1,111 reported precise location information
3,907 reported the name(s) of material(s) released
3,114 reported the volume(s) of material(s) released, with a
total of 38,940,397 gallons
1,884 reported the volume of material that reached water, with
a total of 36,010,550 gallons
3,860 reported water as the medium affected by the incident, 4
reported ballast, 31 reported land, 22 reported other, 10 reported soil, 1 reported subsurface,
and 3 reported unknown
All 3,931 incidents reported a nearby body of water that was affected
Over the 10-year study
period 3,931 incidents
were identified with
the potential to impact
a source of drinking
water.
14
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
3.2 Temporal Occurrence
Figure 4 shows the occurrence of releases with the potential to impact sources of drinking water during
the 10-year timeframe considered in this study. The average number of incidents per year over this
period was 393 with a standard deviation of 81. The number of incidents per year ranged from 481 in
2011 to 229 in 2019 and shows a decreasing trend over the 10-year period. This same trend was
observed in the complete NRC dataset (i.e., before applying the filters described in Section 2).
Specifically, for the complete NRC dataset, the number of releases decreased from 27,809 in 2010 to
23,587 in 2019, and the number of releases from 2015 through 2019 was always less than 25,000.
14,000
12,000
10,000 a
8,000
CD
_OI
6,000
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
3.3 Occurrence by Geographic Location
Of the 3,931 unique incidents considered in this analysis, 1,111
(28%) included precise latitude and longitude (lat/long) coordinates.
Lat/long coordinates were estimated for an additional 2,820
incidents, using the centroid of the smallest region identified in the
NRC report, most often a city or town. The precise or estimated
lat/long coordinates for these incidents are mapped in Figure 5,
which shows geographic clustering around urban areas, industrial
hubs, resource extraction hubs, and transportation corridors.
r
r
28%
L 72%
LI
J
Incident with Locational Data
Incident without Locational Data
Density of Releases
Releases
Figure 5. Geographic Distribution of Incidents Potentially Impacting Sources of Drinking Water
between January 1, 2010 and December 31, 2019 (Map shows the location of 1,111 incidents with
precise lat/long coordinates and 2,820 incidents with estimated lat/long coordinates)
16
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
Figure 6 shows the number of incidents (black circles) and cumulative volume released in kgal (blue
shading and number within state boundary) to water in each state over the 10-year study period. States
with the greatest number of incidents include: Texas (303), California (244), Oklahoma (236), Louisiana
(205), and Pennsylvania (169).
ฎ * ฎ
214 * 251
119
41 ' Hf r 'A- 'J*
%
1.011
ฉ ฎ- ~
tfRs 810 105 Cra 106
ฎ ,n
ฉ ฉ @
ฉ 10B 27 U*
ฉ
3.353
^ 108 u* t
126
226
@ 103
ฉ ฎ 1,530 ฎ
437 50 - 156
% ฎ
ฎ I i
1.960
Volume Released (KGal)
S
173
10,784
3
100
200
303
ฉ
* l
Figure 6. Total Number of Incidents and Cumulative Volume Released per State between January
1, 2010 and December 31, 2019
Figure 6 also shows the cumulative volume released in each state over the 10-year period considered in
this analysis. Four states experienced a total volume released greater than 2,000 kgal: North Carolina
(10,784 kgal), Georgia (7,781 kgal), Colorado (3,353 kgal), and California (2,268 kgal). The large total
volumes released in these four states were driven by a single, large-volume incident in each state:
In North Carolina, a single incident involving the release of 10,491 kgal of coal ash into the Dan
River accounts for 97% of the total volume released during the study period.
In Georgia, a single incident involving the release of 7,593 kgal of wastewater into Sugar Creek
and the Withlacoochee River accounts for 98% of the total volume released during the study
period.
in Colorado, a single incident involving the release of 3,000 kgal of mine waste into Cement
Creek and the Animas River accounts for 89% of the total volume released during the study
period.
17
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
In California, a single incident involving the release of 794 kgal of oil and produced water into an
unnamed body of water accounts for 35% of the total volume released during the study period.
Figure 7 shows the number of incidents released in each hydrologic unit code region (HUC-2) over the
10-year study period. A numeric ID is shown for each HUC-2 region in the white hexagon, and the
number of incidents occurring in each HUC-2 region is shown in the black circle. HUC codes with the
greatest number of incidents include: [3] South Atlantic-Gulf Region (542), [5] Ohio Region (531), and [2]
Mid-Atlantic Region (480). As shown in Figure 7, eight of the HUC-2 regions extend into neighboring
countries ([17], [10], [9], [4], [1], [18], [15], and [13]), however, the analysis considered only releases
that occurred within the borders of the U.S. (i.e., releases occurring in Canada and Mexico were not
considered in this analysis).
09
.04:
02 s
'16]
'05
06
08'
#o
20 HUC Number
Incidents
Major Rivers
ฉ Number of Incidents within HUC
Figure 7. Total Number of Incidents per HUC-2 Region between January 1, 2010 and December 31,
2019
Table 4 shows the 25 counties across the U.S with the greatest number of incidents. The states with the
most counties on this list are Louisiana with 4 counties, Texas with 3 counties, and Oklahoma with 2
counties. The four Louisiana counties are located close together in the southern metropolitan region of
the state, which contains oil and gas extraction operations. The Mississippi River flows through each
county and two of these counties border the Gulf of Mexico. The three Texas counties are in eastern
Texas, with the adjacent counties of Tarrant and Dallas in the northeastern region and Harris County in
the southeast region. These counties contain large urban areas with extensive resource extraction and
industrial development. The Oklahoma counties are Osage and Carter. Osage County in northern
18
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
Oklahoma, is bounded by the Arkansas River, and contains resource extraction mines and oil and gas
wells. Carter County is located in southern Oklahoma and contains extensive oil, gas, and mineral
extraction operations.
Table 4. Twenty-five Counties with the Greatest Number of Incidents across the U.S.
(Top Material Categories Released includes counts for any unique material released during the incident.)
County
State
No. of
Incidents
County Characteristics
Top Material Categories
Released (Count)
Osage
OK
117
Population 49,000 (21/mi2)
Resource Extraction - Oil & Gas/Mines
Borders the Arkansas River
Drilling Fluid (55)
Crude Petroleum (40)
Salt Water (40)
Los Angeles
CA
91
Population 9,818,000 (2,100/mi2)
Resource Extraction - Oil & Gas/Mines
Bisected by the Los Angeles River
Refined Oil (40)
Wastewater (28)
Harris
TX
37
Population 4,713,000 (2,730/mi2)
Resource Extraction - Oil & Gas/Mines
San Jacinto & Buffalo Bayou Rivers
Refined Oil (25)
Cook
IL
36
Population 5,150,000 (5,450/mi2)
Des Plaines & Calumet, Rivers
Borders Lake Michigan
Refined Oil (24)
Queens
NY
30
Population 2,254,000 (20,900/mi2)
JFK International Airport
Borders the East River
Refined Oil (14)
Anti-Freeze (6)
Natrona
WY
28
Population 80,000 (14/mi2)
Resource Extraction - Oil & Gas/Mines
North Platte River
Drilling Fluid (24)
Duval
FL
24
Population 958,000 (1,200/mi2)
St. Johns River
Refined Oil (14)
Weld
CO
20
Population 324,000 (76/mi2)
Resource Extraction - Oil & Gas/Mines
South Platte River
Drilling Fluid (11)
Refined Oil (5)
Crude Petroleum (5)
King
WA
20
Population 2,253,000 (1,000/mi2)
Resource Extraction - Mines
SEA-TAC International Airport
Duwamish River, Green River
Borders the Puget Sound
Refined Oil (14)
Orleans
LA
19
Population 390,000 (2,000/mi2)
Resource Extraction - Oil & Gas
Borders the Mississippi River, Lake
Pontchartrain, and Lake Borgne
Refined Oil (15)
Carter
OK
19
Population 48,000 (58/mi2)
Resource Extraction - Oil & Gas/ Mines
Washita River & Caddo Creek
Crude Petroleum (13)
Drilling Fluid (11)
Mobile
AL
18
Population 413,000 (337/mi2)
Resource Extraction - Mines
Mobile River & Gulf of Mexico
Refined Oil (11)
Plaquemines
LA
18
Population 23,000 (9/mi2)
Resource Extraction - Oil & Gas
Mississippi River & Gulf of Mexico
Refined Oil (13)
Wayne
Ml
18
Population 1,749,000 (3,000/mi2)
Borders Detroit River & Lake St. Clair
Refined Oil (14)
Dallas
TX
18
Population 2,636,000 (3,000/mi2)
Resource Extraction - Oil & Gas/ Mines
Trinity River
Refined Oil (11)
Wastewater (4)
New Haven
CT
17
Population 855,000 (1,000/mi2)
Resource Extraction - Mines
Quinnipiac & Housatonic Rivers
Refined Oil (12)
19
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
County
State
No. of
Incidents
County Characteristics
Top Material Categories
Released (Count)
St. James
LA
17
Population 21,000 (86/mi2)
Resource Extraction - Oil & Gas
Mississippi River
Refined Oil (10)
Crude Petroleum (4)
Westchester
NY
17
Population 968,000 (2,000/mi2)
Borders the Hudson River and Long
Island Sound
Refined Oil (10)
Transformer Oil (4)
Allegheny
PA
17
Population 1,216,000 (1,700/mi2)
Major Industrial Hub
Commerce transport along Ohio,
Allegheny, and Monongahela Rivers
Refined Oil (13)
Tarrant
TX
17
Population 2,103,000 (2,100/mi2)
Resource Extraction - Oil & Gas
W. Fork Trinity River & Clear Fork
Trinity River
Refined Oil (11)
St. Louis
MO
16
Population 994,000 (1,900/mi2)
Resource Extraction - Mines
St. Louis Lambert International Airport
Borders the Missouri, Mississippi, &
Meramec Rivers
Refined Oil (13)
Cuyahoga
OH
16
Population 1,235,000 (2,800/mi2)
Resource Extraction - Oil & Gas
Cuyahoga & Rocky Rivers
Borders Lake Erie
Refined Oil (13)
Shelby
TN
16
Population 937,000 (1,200/mi2)
Resource Extraction - Mines
Mississippi, Wolf, and Loosahatchie
Rivers
Refined Oil (14)
E. Baton
Rouge
LA
15
Population 440,000 (940/mi2)
Resource Extraction - Oil & Gas
Mississippi River
Refined Oil (8)
Harrison
WV
15
Population 68,000 (170/mi2)
Resource Extraction - Oil & Gas
West Fork River
Refined Oil (5)
Drilling Fluid (4)
Figure 8 shows the number of incidents and total volume released into "unknown or unnamed" bodies
of water or into "named" bodies of water. The majority (61%) of incidents occur into an unknown or
unnamed waterbody, however, most of the volume released (85%) occurs into named bodies of water.
This discrepancy may be due to more complete reporting of releases involving large volumes (i.e., more
information, including the name of the body of water, is reported for large releases).
Incident Counts
r
r 39%
1
I /
61% J
iy
A
Unknown or
Unnamed
Named BOW
Total Volume Released
Unknown or
Unnamed
Named BOW
Figure 8. Incident Counts and Total Volume Released into Named Bodies of Water (BOW) and
Unknown or Unnamed Bodies of Water
20
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
The thirty-one named waterbodies with the greatest number of incidents or largest total volume
released are shown in Figure 9. The waterbodies with the
greatest number of incidents include the Mississippi River (139
incidents), Ohio River (77 incidents), and Hudson River (29
incidents). Figure 8 also shows the waterbodies into which the
largest cumulative volumes were released: the Dan River (10,491
kgal), Withlacoochee River (7,593 kgal), and Cement
Creek/Animas River (3,002 kgal). Notably, these three bodies of
water were impacted by a small number of incidents: one, one,
and two, respectively. Similar to the analysis of release by state,
this analysis shows that a relatively small number of very large
releases dominate the distribution of cumulative volumes
released across waterbodies.
Regardless of the geographic
boundaries used in the
analysis, three characteristics
are common to regions with
a high occurrence of
releases: (1) urban areas, (2)
transportation corridors, or
(3) resource extraction
activity.
160
140
120
I/)
2 100
0
_c
"Z 80
2
1 1
<
1 I
O
x
O
< 5
<
u
<
12,000
10,000
8,000
6,000
4,000
1,600
^ 1,418 2,000
~ ~
1 1
Figure 9. Thirty-one Waterbodies with the Greatest Number of Incidents or Largest Cumulative
Volume Released to Water across the U.S. (A volume less than 0.5 kgal is displayed as 0 in the figure:
Neches River (0.13 kgal), Cape Fear River (0.40 kgal), and Calcasieu River (0.45 kgal))
3.4 Materials Released
Of the 3,931 unique incidents considered in this analysis, 3,907 (99.4%)
incidents included the name of the material(s) released. These incidents
involved 840 different materials, which were grouped into the categories
listed in Table 2. Incidents involving the release of multiple materials are
counted under multiple material categories resulting in a total of 4,226
unique material releases, of which 3,250 reported the volume released and
1,954 reported the amount in water.
The frequency of releases and total volume released involving each of
these material categories is shown in Figure 10. Materials in the Refined Oil
category are the most frequently released materials by a significant margin.
0.6%
Material Spilled with Name of
Material Released
Material Spilled without Name of
Material Released
21
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
There were 2,401 (56.8%) releases of Refined Oil over the 10-year study period. Within the Refined Oil
category, the most commonly released material is diesel, accounting for 45.7% of releases of Refined
Oil, followed by unknown oil (10.5%), and gasoline (8.4%). Releases of gasoline accounted for the largest
total volume released, 588 kgal (29.7%), within the Refined Oil category.
The next most frequently released material category was Crude Petroleum, with 398 (9.4%) releases,
followed by Wastewater with 285 (6.7%) releases, Drilling Fluid 231 (5.5%), Transformer Oil 164 (3.9%),
and Other Materials 116 (2.7%). The Other Materials category includes a wide range of materials, most
of which are poorly characterized and were involved in three or fewer incidents. The largest volume of a
named material captured under the Other Materials category was a release of 44 kgal of stearic acid
into an unnamed body of water in Massachusetts in 2011.
3000
2500
2000
1500
E 1000
500
2,401
1,979
(0.8)
14,014
10,769
ฆ Count
Volume
(Average Volume Released
(kg=l))
398
2,051 3-503
(5.2) 285 164
It49-2),^ 2) 103
I I i
116
113
cc
Q
=J
co
3
1
3
o
<
LU
<
ฃ
LU
1
jx!
cc
cc
O
cc
1
u
2
cc
LU
<
LU
l/)
O
Q_
<
CC
ฃ
cE
co
-z.
<
Q
3
o
X
1
cc
u
1-
O
88
369
<
ฃ
64
138
O
i
<
61
284
~6ir~
231
3
O
o
o
o
53
574
51
95
-30-
132
28
18
24
60
21
15
8
1,428
(178.6)
3,000
<
w
LU
N
1
<
~
LU
2
CC
LU
CC
CC
o
y
CC
I
O
<
j
<
~r
(J
u
-Z.
u
<
z
1
i
CO
3
o
X
<
U
cc
3
Q
O
O
X
O
u
<
3
o
Q
Z
<
16,000
14,000
12,000
10,000
8,000
6,000
4,000
2,000
0
<
o
Material Category
Figure 10. Total Number of Incidents and Cumulative Volume Released for each Material Category
(A volume less than 0.5 kgal is displayed as 0 in the figure: Cyanide Compounds (0.21 kgal) and
Radiological (no volumes reported))
There were 164 (3.9%) releases of transformer oil over the 10-year period. Transformer oil is most often
released from pole mounted transformers that can contain 50 to 100 gallons of transformer oil;
however, transformers can be much larger, requiring significantly larger volumes of oil (Power Partners
Inc., 2009). Of the 164 releases involving transformer oil, twenty-two released volumes between 1 kgal
and 1.4 kgal. These releases included discharges from vehicles carrying transformer oil, large
transformers damaged during flooding, sub-station transformer discharges, and releases from storage
tanks. The NRC records for releases of transformer oil were not always clear whether the oil contained
PCBs. Through the 1970s, transformer oil often contained PCBs. However, the Toxic Substances Control
Act banned the production of new PCBs in 1979, thus it is hypothesized that very few releases of
transformer oil during the study period involved oils containing PCBs.
Figure 10 also shows the cumulative volume released for each material category over the study period.
Wastewater releases accounted for 14,014 kgal (36%) of the total volume released over the 10-year
22
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
study period (38,940 kgal), an average volume per incident of 49.2 kgal. There were 3 releases of
wastewater of 1,000 kgal or more. The largest release occurred in 2019 where 7,592 kgal of wastewater
were released into the Withlacoochee River in Georgia.
The second largest release of wastewater occurred in
2011 when 1,600 kgal were released into Mulberry Fork
in Alabama.
Coal Combustion By-Products, which include fly ash,
bottom ash, and boiler slag among other material, had the second highest cumulative volume released.
Coal Combustion By-Products accounted for 10,769 kgal (27.7%) of the total volume released, an
average volume per incident of 512.8 kgal. This total volume is dominated by a single release of 10,491
kgal of coal ash (specifically, fly ash) into the Dan River in North Carolina in 2014. Coal combustion by-
products can present a serious threat to water quality since they contain toxic metals and metalloids
including arsenic, cadmium, lead, and mercury.
Drilling Fluid had the third highest cumulative volume released at 3,503 kgal (9%), an average volume
per incident of 15.2 kgal. Drilling fluids often contain brine that can increase the bromide concentration
in the source water. Some community water systems have attributed increases in the concentration of
brominated disinfection by-products to contamination from drilling fluids (States et al, 2013).
Mine Waste had the fourth highest total volume released at 3,000 kgal (7.7%). Because there was only
one incident involving the release of mine wasted in this analysis, the average volume per incident is
also 3,000 kgal. Similar to coal-combustion by-products, mine waste may contain high concentrations of
toxic metals.
Table 5 shows the number of releases that occurred within a specified range of volumes for each of the
material categories. Four volume ranges were considered: less than 1 kgal; between 1 and 10 kgal;
between 10 and 100 kgal; and greater than 100 kgal. With only one exception, Alcohol, most releases for
each material category were less than 1 kgal, and the number of releases in each volume range
decreases as volume increases. This table also illustrates that occurrence of large volume releases varies
across material categories. The following material categories were involved in releases larger than 10
kgal, but less than 100 kgal (shaded yellow in the table): Caustic Material, Fire Fighting Foam, Metals and
Metalloids, Other Materials, Refined Oil, Salt Water, and Transformer Oil. The following material
categories were involved in releases larger than 100 kgal (shaded orange in the table): Alcohol, Coal
Combustion By-products, Crude Petroleum, Drilling Fluid, Fertilizer, Food Products, Mine Waste, Organic
Solvents, Unknown Material, and Wastewater.
While refined oils are involved in
the largest number of releases,
wastewater accounts for the largest
cumulative volume released.
23
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
Table 5. Count of Releases Involving a Volume within the Indicated Range for Each Material
Category (A volume released less than 0.5 kgal is displayed as 0 in the table: Cyanide Compounds (0.21
kgal) and Radiological (no volumes reported))
Material
Category
No. of
Releases
Total
Volume
Released
(kgal)
Number of Releases in Volume Range
Largest
Volume
Released
(kgal)
Not
Reported
< 1
(kgal)
1 -10
(kgal)
10 -
100
(kgal)
> 100
(kgal)
TOTAL
4,226
38,940
976
2,301
786
134
29
N/A
Acid
57
41
12
33
12
0
0
7
Alcohol
8
1,428
4
0
3
0
1
1,418
Antifreeze
36
16
14
17
5
0
0
4
Caustic
Material
51
95
8
26
15
2
0
24
Chlorine
28
18
7
13
8
0
0
5
Coal
Combustion
By-Products
21
10,769
7
6
4
2
2
10,491
Crude
Petroleum
398
1,257
36
260
91
9
2
794
Cyanide
Compounds
1
0
0
1
0
0
0
0.21
Drilling Fluid
231
3,503
6
85
106
28
6
977
Fertilizer
53
574
13
24
13
2
1
467
Fire Fighting
Foam
24
60
3
13
5
3
0
20
Food
Products
61
231
13
22
22
3
1
100
Metals and
Metalloids
64
138
44
13
4
3
0
80
Mine Waste
1
3,000
0
0
0
0
1
3,000
Organic
Solvents
61
284
28
11
19
2
1
178
Other
Materials
116
113
44
53
18
1
0
44
Paint
21
15
7
9
5
0
0
4
Pesticides/
Herbicides
20
6
8
11
1
0
0
4
Radiological
6
0
6
0
0
0
0
N/A
Refined Oil
2,401
2,773
605
1,461
308
27
0
88
Salt Water
88
369
5
31
41
11
0
33
Transformer
Oil
164
103
13
129
20
2
0
14
Unknown
Material
30
132
23
4
1
1
1
100
Wastewater
285
14,014
70
79
85
38
13
7,592
The results in Table 5 show that the following material categories account for 79% of the number of
releases and 83% of the total volume released: Refined Oil, Crude Petroleum, Wastewater, Drilling
Fluids, and Coal Combustion By-Products. To investigate the geographic distribution of releases in these
categories, they were mapped in Figures 11 -15. These maps show the precise and estimated lat/long
coordinates as points, while the shading shows the density of release occurrence, with darker shading
indicating a higher occurrence of releases.
Figure 11 shows the geographic distribution of releases of Refined Oil, indicating a widespread
distribution of releases in urban areas. Regions with a high density of Petroleum Product releases
24
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
include Western New Jersey/Eastern Pennsylvania; Western Pennsylvania/West Virginia/Eastern Ohio;
Southeastern Louisiana, and Central Oklahoma/Northcentral Texas. The states with the most releases of
Refined Oil include: Texas with 146 (6%), Ohio with 122 (5%), Pennsylvania with 120 (5%), and Louisiana
with 120 (5%). The majority of releases in Texas occurred into drainage and collection areas (69),
followed by unnamed creeks, streams, or tributaries (16). The remaining releases of Refined Oil in Texas
occurred into various named creeks or rivers, with the Neches River experiencing the most (6). The
majority of releases in Ohio also occurred into drainage and collection areas (48), followed by the Ohio
River (22), and unnamed creeks, streams, or tributaries (16). The bodies of water impacted by the most
releases of Refined Oil include: drainage and collection (932), unnamed creeks, streams, or tributaries
(256), the Mississippi River (102), and the Ohio River (73). Releases occurring on the Mississippi River
and Ohio River can impact adjacent and downstream states. Releases of Refined Oil into the Mississippi
River occurred in Louisiana (67), Missouri (10), Illinois (5), Minnesota (5), Mississippi (5), Kentucky (3),
Tennessee (3), Iowa (2), Arizona (1), and Wisconsin (1). Releases of Refined Oil into the Ohio River
occurred in Ohio (22), Kentucky (16), West Virginia (14), Indiana (10), Pennsylvania (6), and Illinois (5).
i*
*%v.
C *! A
Refined Oil Release
Density of Releases
ฆ Shale Plays (USEIA)
Figure 11. Geographic Distribution of Releases of Refined Oil that Occurred between January 1,
2010 and December 31, 2019
25
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
Figure 12 shows the geographic distribution of releases of Crude Petroleum with hot spots in Central
Oklahoma/Southeast Kansas, Southeast Texas/Southern Louisiana, and West Virginia/Eastern Kentucky.
There is significant overlap between these hot spots and the major shale plays that are sources of oil
and natural gas. The states with the most releases of Crude Petroleum include: Oklahoma with 98 (25%),
Texas with 62 (16%), Louisiana with 41 (10%), and Kansas with 28 (7%). The majority of releases in
Oklahoma occurred into unnamed creeks, streams, and tributaries (33), followed by drainage and
collection areas (18). The remaining releases of Crude Petroleum in Oklahoma occurred into various
named creeks or rivers, with the Little Chief Creek experiencing the most (4). The majority of releases in
Texas occurred into unnamed creeks, streams, or tributaries (13), followed by drainage and collection
areas (6). The remaining releases of Crude Petroleum in Texas occurred into various named creeks or
rivers, with Neches River and Sabine River experiencing the most (4 each). The bodies of water impacted
by the most releases of Crude Petroleum include: unnamed creeks, streams, or tributaries (90), drainage
and collection areas (69), unnamed reservoirs, lakes, or ponds (30), unnamed wetland area (18), and the
Mississippi River (10). Releases of Crude Petroleum into the Mississippi River occurred in Louisiana (7),
Illinois (2), and Mississippi (1).
Crude Petroleum Release
Density of Releases
ฆ Shale Plays (USEIA)
Figure 12. Geographic Distribution of Releases of Crude Petroleum that Occurred between
January 1, 2010 and December 31, 2019
26
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
Figure 13 shows the geographic distribution of releases of Wastewater in urban areas across the U.S.,
similar to the distribution of releases of Refined Oil. This result is expected given that refined oils and
wastewater are prevalent in urban areas. Regions with a high density of Wastewater releases include
Southern California and Southwestern Pennsylvania. The states with the most releases of Wastewater
include: California with 54 (19%), Virginia with 25 (9%), and Texas with 21 (7%). The majority of releases
in California occurred into drainage and collection areas (36), followed by the Los Angeles River (4), the
San Joaquin River (2), and Hutchinson Creek (2). The remaining releases of Wastewater in California
occurred in various named creeks or rivers. The majority of releases in Virginia occurred into drainage
and collection areas (14), followed by drinking water (3), Skiffes Creek (3), and the Elizabeth River (2).
The bodies of water impacted by the most releases of Wastewater releases include: drainage and
collection areas (101), unnamed creeks, streams, or tributaries (22), drinking water (20), and unnamed
reservoirs, lakes, or ponds (8).
#
2010 Census-designated Urban Area
Density of Releases
Wastewater Release
Figure 13. Geographic Distribution of Releases of Wastewater that Occurred between January 1,
2010 and December 31, 2019
27
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
Figure 14 shows the geographic distribution of releases of Drilling Fluid, with regions experiencing a
large number of releases in Northern Oklahoma/Southern Kansas, Northeastern Texas/Southcentral
Oklahoma, Central Wyoming, Northeastern North Dakota, and Northern West Virginia/Eastern
Ohio/Southwestern Pennsylvania. All of these regions contain large shale plays and associated oil and
gas extraction operations that use large quantities of drilling fluids. The states with the most releases of
Drilling Fluid include: Oklahoma with 81 (35%), followed by Wyoming with 31 (13%), and Texas with 24
(10%). The majority of releases in Oklahoma occurred into unnamed creeks, streams, or tributaries (23),
followed by a drainage or collection area (22), and unnamed reservoirs, lakes, or ponds (12). The
remaining releases of Drilling Fluid in Oklahoma occurred into various named creeks or rivers, with Hay
Creek experiencing the most (3). The majority of releases in Wyoming also occurred into drainage and
collection areas (15), followed by Castle Creek (4). The bodies of water impacted by the most releases of
Drilling Fluid include: drainage and collection areas (61), unnamed creeks, streams, or tributaries (51),
unnamed reservoirs, lakes, or ponds (18), and South Platte River in Colorado (6).
, J/# ^
I I Density of Releases
Drilling Fluid Release
ฆ Shale Plays (USEIA)
Figure 14. Geographic Distribution of Releases of Drilling Fluids that Occurred between January 1,
2010 and December 31, 2019
28
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
Figure 15 shows the geographic distribution of releases of Coal Combustion By-Products, with a
concentration of releases occurring in Western Pennsylvania/Eastern Ohio/Northern West Virginia. The
states with the most releases of Coal Combustion By-Products include: Ohio with 3 (14%) and North
Carolina with 2 (10%). The high number of releases in these regions is likely a result of the large number
of active and closed coal-fired power plants in these same regions. The three releases in Ohio occurred
into Block House Hollow, the Ohio River, and Riddles Run while the two releases in North Carolina
occurred into the Dan River and the Neuse River. The bodies of water impacted by the most releases of
Coal Combustion By-Products include: drainage and collection areas (2), drinking water (2), Hudson
River (2) which only impacted New York, and the Dan River (1) which impacted only North Carolina.
Coal Combustion By-Product Release
Coal Power Plant Density (US KIA)
H Density of Releases
Figure 15. Geographic Distribution of Releases of Coal Combustion By-Products that Occurred
between January 1, 2010 and December 31, 2019
3.5 Cause of Incidents and Responsible Parties
The NRC record includes standardized fields for recording the reported cause and responsible party of a
release. The causes of releases reported to the NRC and considered in this analysis include:
Equipment Failure: including ruptures in storage vessels and pipelines, over pressurization, and
explosions
Transportation Accident: including accidents occurring on roadways, railways, and waterbodies
that involve a vehicle or vessel
Operator Error: including failure to follow procedures for transport and transfer of materials
Natural Phenomena: including hurricanes, flooding, and tornados
29
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
Dumping: including unpermitted discharges as well as illegal dumping by private companies and
citizens
Other: cause known but other than the standard causes listed above.
Unknown: cause of the release unknown or not reported
All 3,931 incidents included in this analysis have information about the cause. The number of incidents
and cumulative volume of releases attributed to each cause are shown in Figure 16. Equipment Failure
was reported as the leading cause of an incident 27% (1,078), was responsible for the 49% (19,260 kgal)
of the cumulative volume released, and had an average volume released per incident of 17.9 kgal.
Transportation Accidents caused 13% (505) of the incidents but only 2% (820 kgal) of the cumulative
volume released, and thus had a much lower average volume per incident of 1.6 kgal. Natural
Phenomena caused only 5% (212) of the incidents but accounted for 5% (1,807 kgal) of the cumulative
volume released, with an average volume released of 8.5 kgal.
1200
1000
a 800
600
-Q
E
400
200
1.078
19,260
(17.9 )
878
11,506
(13.1)
767
4,654
(6.1)
I Count
1 Volume
(Average Volume Released (kgal))
505
820
(1.6)
365
665
(1.8)
212
1,807
126
228
1.8)
EQUIPMENT UNKNOWN
FAILURE
OTHER
TRANSPORT
ACCIDENT
Cause of Incident
OPERATOR NATURAL DUMPING
ERROR PHENOMENON
25,000
20,000
15,000
ce
10,000 E
5,000
Figure 16. Total Number of Incidents and Cumulative Volume of Material Released for each Cause
between January 1, 2010 and December 31, 2019
Because wastewater accounts for most of the cumulative volume released, the causes of wastewater
releases were investigated. The majority of wastewater releases were caused by equipment failure
(100), other causes (57), unknown causes (52), and illegal dumping of materials (29). Pipeline breaks and
pump failures were the two most common types of equipment failures leading to wastewater releases.
All 3,931 incidents included in this analysis listed the responsible party using the following designations:
Private Enterprise
Private Citizen
Federal Government
30
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
State Government
Local Government / Public Utility (two unique NRC designations that were combined for this
analysis)
Military
Tribe
Other
Unknown
Figure 17 shows the number of incidents and total volume released
attributed to the above categories of responsible party. Private
Enterprises was listed as the responsible party for 69% (2,712) of the total
number of incidents, 87% (34,308 kgal) of the total volume released, and
an average volume per incident of 12.7 kgal. Within this category
Chaparral Energy, an oil and natural gas producing company in Oklahoma,
was responsible for the largest number of incidents (53), which involved
petroleum products and drilling fluids. Local Government / Public Utilities
were responsible for 6% (236) of the incidents, 8% (3,329 kgal) of the total volume released, and an
average volume per incident of 14.1 kgal. Most incidents caused by Local Government / Public Utility
involved the release of wastewater from municipal wastewater systems. Private Citizens were
responsible for 4% (140) of the total number of incidents, 0.15% (59 kgal) of the total volume released,
and an average volume per incident of 0.4 kgal.
2,712
3000 34,308
(12.7)
2500
ฃ 2000
QJ
l
S 1500
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
Table 6. Most Commonly Released Materials for each Responsible Party Category
Responsible Party Category
Most Commonly Released Materials
Private Enterprise
Diesel Oil, Crude Oil, Unleaded Automotive Gasoline
Unknown
Diesel Oil, Unknown Oil, Fuel Oil, Unleaded Automotive Gasoline
Local Government / Public Utility
Wastewater, Hydraulic Oil, Transformer Oil
Military
Wastewater, Diesel Oil, Jet Fuel
Other
Diesel Oil, Crude Oil, Other Oil
Federal Government
Hydraulic Oil, Diesel Oil, Gasoline
State Government
Diesel Oil, Hydraulic Oil, Fuel Oil
Tribe
Wastewater
Figure 18 shows the number of incidents and total volume released for various types of incidents. The
most common type of incident involved Other Transportation, not including Watercraft and Railroads,
which are captured under separate categories. While Other Transportation accounted for a large
number of incidents, 1,257 (32%), they only accounted for 1,475 kgal (4%) of the cumulative volume
released, and thus had an average volume released per incident of only 1.2 kgal. The second most
common type of incident were those at Fixed Facilities, which were involved in 1,081 (27%) incidents,
accounted for 29,272 kgal (75%) of the cumulative volume released, and had the largest average volume
released per incident of 27.1 kgal. While Pipeline breaks account for only 354 (9%) incidents, they
account for the second largest total volume released 4,500 kgal (12%). The average size of a release
from a pipeline break is 12.7 kgal. The NRC allows releases to be assigned to both Fixed Facilities and
Storage Tanks (which are also fixed facilities). It is unclear how these two categories are differentiated,
and it's likely that there is some overlap - specifically, some releases from Fixed Facilities likely came
from storage tanks at those facilities.
1400
1200
1000
800
o
fc 600
400
200
1,257
1,081
1,475
29,272
ฆ Count
Volume
(Average Volume Released
(kgal))
519
2,814
(5.4)
354
4,500
(12.7)
Other
Transportation
Fixed Facility
Watercraft Storage Tank Pipeline
Type of Incident
136
163
(1.2)
Railroad
48
13
(0.3)
Unknown
35,000
30,000
25,000
20,000 J!
15,000
10,000 งฆ
5,000
Figure 18. Total Number of Incidents and Cumulative Volume of Material Released by Type of
Incident between January 1, 2010 and December 31, 2019
3.6 Occurrence of Releases by Source Water Zone of Concern
In Sections 3.1 through 3.5, the analysis considered the occurrence and attributes of releases, treating
each incident as a unique data point. This section considers the occurrence of unique incidents within a
32
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
source water ZOC for all community water system intakes and wells included in the analysis. In cases
where ZOCs for multiple intakes or wells overlap, the same incident can impact multiple ZOCs.
Furthermore, this analysis includes only the 1,111 incidents for which there is precise location data (i.e.,
lat/Long coordinates). Of these, 1,097 fell within a ZOC, as defined in Section 2.3. While the remaining
14 incidents did not fall within a ZOC, it was confirmed that the contaminant plume from the incident
reached one or more downstream intakes that were more than 50 miles from the release location and
prompted a response (e.g., closing an intake, deploying booms, collecting and analyzing samples). These
incidents were manually reviewed and assigned to SW and/or GW ZOCs by tracing the closest NHD
flowline and identifying the closest downstream ZOC(s) that may have been affected by the incident.
Figure 19 shows the number of community water systems whose source water ZOC experienced at least
"n" incidents, where "n" is the number indicated on the y-axis. The figure shows a significant difference
in the occurrence of incidents for SW ZOCs and GW ZOCs. Of the 4,929 SW intakes considered in this
analysis, 85% (4,204) did not have an incident occur within their ZOC under the assumptions of this
analysis. The remaining 725 SW intakes experienced at least one
incident that potentially impacted their source of drinking water. While
the distribution of incidents in SW ZOCs drops off rapidly, there were a
small number of ZOCs that experienced a significantly larger number of
incidents. Twenty-five SW ZOCs experienced at least 10 incidents (on
average, one release per year) and three SW ZOCs experienced 22
incidents over the 10-year study period.
Of the 106,816 GW wells considered in this analysis, 0.7% (700) experienced at least one incident within
its ZOC under the assumptions of this analysis. The distribution of incidents in GW ZOCs drops off much
more quickly compared with the distribution for SW ZOCs. The maximum number of incidents in a GW
ZOC was four, experienced within only four GW ZOCs.
The data in Figure 19 suggest that SW ZOCs are more likely to experience a release than are GW ZOCs:
15% for SW ZOCs vs. 0.7% for GW ZOCs. Also, SW ZOCs are more likely to experience a greater number
of releases compared with GW ZOCs. However, this analysis does not consider the prevalence of long-
term or chronic contamination of ground water aquifers.
SW ZOCs are more
likely to experience a
release than are GW
ZOCs: 15% for SW ZOCs
vs. 0.7% for GW ZOCs.
33
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
1600
725
700
359
48
233
9
1154
95
| ฆ 4ฐ " 2o9 25 20 15 12
I ฆ ฆ - 2. 2. ฐ ฃ
ISW Intake
IGW Well
25
20
15
12
6
5
5
4
4
3
3
3
3
ฃ
ฃ
0
0
0
0
0
0
0
0
0
10
11
12
13
14
15
16
17
18
19
20
21
22
Number of Incidents (n)
Figure 19. Frequency of Releases to Source Water ZOCs between January 1, 2010 and December
31, 2019 (total number of SW ZOCs = 4,929, total number of GW ZOCs = 106,816)
Table 7 presents attributes of the six SW ZOCs with the greatest number of releases. The three ZOCs
with the greatest number of incidents are in Osage County, Oklahoma and overlap almost completely
and thus experienced the same 22 incidents. Osage county has a significant oil, gas, and mineral
extraction industry. The Pennsylvania ZOC with the second largest number of incidents (18) contains a
large urban area with major river commerce routes. The majority of incidents in this ZOC involved
Refined Oil products.
The area of the ZOCs varied by more than three orders of magnitude, which could be a factor in the
number of incidents experienced in a ZOC. The areas of the SW ZOCs ranged from 0.42 square miles to
1,518 square miles, with an average area of 345 square miles. GW ZOCs used a standard 0.5-mile radius,
and thus had a uniform area of 0.78 square miles. To account for the impact of area on incident
occurrence, the incident count was normalized by the area of the ZOC. The normalized incident count
ranged from 0.0008 to 2.36 incidents per square mile. However, large normalized incident counts can
result from ZOCs with small areas. Because GW ZOCs have a small uniform area of 0.78 square miles,
they have an artificially large normalized count. GW ZOCs with one incident have a normalized count of
1.28 per square mile and those with four releases have a normalized count of 5.13 per square mile.
34
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
Table 7. Surface Water Zones of Concern that Experienced the Most Releases
ID
State
Counties
within ZOC
Zone Area
(Sq. Miles)
Number of
Incidents
Incidents per
Sq. Mile
Top Material Categories
Spilled (Count)
A
OK
Osage
303
22
0.072
Drilling Brine (16)
B
OK
Osage
303
22
0.072
Drilling Brine (16)
C
OK
Osage
303
22
0.072
Drilling Brine (16)
D
PA
Allegheny
Armstrong
Washington
1,066
18
0.017
Refined Oil (8)
Transformer Oil (4)
E
PA
Allegheny
Washington
935
16
0.017
Refined Oil (8)
F
CT /
NY
Fairfield
New Haven
New London
Westchester
193
14
0.073
Refined Oil (7)
Table 8 presents the five SW ZOCs that experienced more than one incident, and which had the largest
number of incidents per square mile. Two ZOCs from Louisiana (I and J) that experienced a large number
of incidents (12 and 13, respectively) also have a large number of releases per square mile (0.40 and
0.32 incidents per square mile, respectively). However, the other three zones experienced a modest
number of releases (4 or 5). A commonality among the five ZOCs with the largest number of incidents
per square mile is that they all have relatively small areas (3 to 41 square miles) - much smaller than the
average zonal area of 345 square miles. Thus, in all five zones, it is the small zonal area that is largely
responsible for the high normalized count.
Table 8. Source Water Zones of Concern that Experienced more than One Release and which have
the Greatest Number of Releases per Square Mile
ID
State
Counties
Zone Area
Number of
Incidents/sq. mile
Top Material
Impacted
(Sq. Miles)
Incidents
Categories Spilled
(Count)
G
PA/NJ
Warren
3
4
1.34
Refined Oil (4)
H
CA
San Mateo
12
5
0.41
Refined Oil (5)
I
LA
Jefferson
Orleans
Plaquemines
St. Bernard
30
12
0.40
Refined Oil (7)
J
LA
Jefferson
Orleans
Plaquemines
St. Bernard
41
13
0.32
Refined Oil (7)
K
IL
Cook
14
4
0.28
Refined Oil (2)
Figure 20 shows the geographic distribution of all incidents that impacted a ZOC. Similar to Figure 5, the
distribution of releases within ZOCs shows a concentration of releases near major urban areas or regions
with significant oil and gas extraction operations. Two areas that experienced a large number of releases
are Southern Louisiana and Western Pennsylvania/Eastern Ohio. The states with the most releases
impacting a ZOC are: Ohio with 207 (8%), Texas with 204 (8%), Pennsylvania with 193 (7%), and
Louisiana with 162 (6%). Notably, this is a different ranking from the broader analysis of incidents by
state presented in Figure 6 in which the states with the most releases were: Texas (303), California
(244), Oklahoma (236), Louisiana (205), and Pennsylvania (169). The different trends depicted in Figures
35
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
6 and 20 indicates that there may not always be a correlation between the occurrences of releases to
water and the occurrence of releases in a ZOC.
I Density of Releases
2010 Census-designated Urban Area
Releases
Surface and Groundwater Zones
Figure 20. Geographic Distribution of Releases in SW and GW ZOCs between January 1, 2010 and
December 31, 2019
36
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
Section 4: Summary and Conclusions
2.1% of releases reported
to the NRC over a 10-year
period had the potential
to impact source water.
An analysis of releases reported to the NRC from January 1, 2010
through December 31, 2019 showed that 5,806 out of 281,142
records, or 2.1%, had the potential to impact sources of drinking
water. The criteria used to identify these releases include: volume
released greater than 100 gallons (or an indicator of a high volume
release when the volume was not recorded), releases of highly toxic
materials, and releases with location data and which fell within a ZOC or affected a nearby ZOC. There
were also criteria for excluding records, such as those that were generated during drills, occurred
outside of the U.S., were released to air, occurred offshore, or included materials that were unlikely to
change the water quality (e.g., sand, steel, aggregate). A downward trend in the number of releases
potentially impacting source water was observed between 2010 and 2019.
The 5,806 records were related to 3,931 unique incidents. Of the 3,931 unique incidents, 3,114 reported
the volume released and 1,884 reported the volume released to water. The total volume released over
this 10-year period was over 38,940 kgal. The volume released varied significantly from year to year,
from a low of 899 kgal in 2013 to a high of 11,477 kgal in 2014. Notably the year 2019 had both the
lowest number of unique incidents and the second largest cumulative volume released, which was
driven by two large releases: (1) a release of 7,593 kgal of wastewater into Sugar Creek and the
Withlacoochee River in Georgia and (2) a release of 1,418 kgal of
bourbon into Glenns Creek and the Kentucky River in Kentucky. This
reflects the general trend that most releases are relatively small;
54.4% were less than 1 kgal, while a small number of very large
releases are responsible for the majority of material released (0.7%
released more than 100 kgal).
54.4% of releases were
smaller than 1 kgal,
however, 0.7% released
more than 100 kgal.
Releases occurred in all 50 states and D.C. with the highest density of releases near urban areas,
industrial hubs, resource extraction hubs, and transportation corridors (including rivers used for cargo
transport). States with the greatest number of unique incidents include: Texas (303), California (244),
Oklahoma (236), Louisiana (205), and Pennsylvania (169). There was a large range in the total volume
released in each state, ranging from < 1 kgal in Rhode Island and Nevada to 10,784 kgal in North
Carolina. Four states experienced a total volume released to water greater than 2,000 kgal: North
Carolina (10,784 kgal), Georgia (7,781 kgal), Colorado (3,353 kgal), and California (2,268 kgal). The large
total volumes released to water in these five states were the result of one or two significant incidents in
each state. The Mississippi and Ohio Rivers were the two named bodies of water impacted by the
greatest number of releases.
There were 840 different materials released over the study period,
but the most commonly released materials by a significant margin
were Refined Oil products, which were involved in 56.8% (2,402) of
releases. However, the material categories responsible for the
largest total volume released were: Wastewater with 14,014 kgal
(36%), Coal Combustion By-products with 10,769 kgal (27.7%),
Drilling Fluid with 3,503 kgal (9%), and Mine Waste with 3,000 kgal
(7.7%).
The highest occurrence of
releases occurs near
urban areas, industrial
hubs, resource extraction
hubs, and transportation
corridors.
37
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
Equipment Failure was reported as the leading cause of incidents, 27% (1,078), and was responsible for
49% (19,260 kgal) of the total volume released. Transportation Accidents were another significant cause
of incidents accounting for 13% (505) of the incidents but only 2% (820 kgal) of the total volume
released - an average release volume of 1.62 kgal. By comparison, Natural Phenomena caused only 5%
(212) of the releases but accounted for 5% (1,807 kgal) of the total volume released - a significantly
larger average release volume of 8.52 kgal.
The number and volumes released within individual SW or GW ZOCs varied widely. Most SW ZOCs (85%)
and GW ZOCs (99.3%) considered in this study did not experience any incidents over the 10-year study
period. There were 725 SW ZOCs and 700 GW ZOCs that experienced at least one incident. Of the ZOCs
that experienced more than one incident, twenty-five SW ZOCs experienced at least 10 incidents, four
SW ZOCs experienced 22 unique incidents, and four GW ZOCs experienced four incidents.
The results of this analysis indicate regional trends in the risk of source water contamination. However,
the absence of a release in a particular area or ZOC should not be inferred to mean that the ZOC is free
from risk. Ultimately what matters is where the next significant release into water occurs and whether it
impacts a water system's source water. The following section provides recommendations to help water
systems understand and prepare for their unique risk profile.
38
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
Section 5: Recommendations
The findings from this study demonstrate there is a significant risk of releases into sources of drinking
water at a national scale. However, the risk to a community water system will depend on their unique
circumstances. To understand the relative risk of source water contamination to a community water
system, it is recommended that releases into source water be considered in an all-hazards risk
assessment, such as that required under America's Water Infrastructure Act (U.S. EPA, 2019a). Factors
to consider when assessing the risk of releases to source water include:
S History of releases into a community water system's source water protection
area (or a smaller zone of concern near the intake or wellhead)
S Land use and population in the source water protection area
S The number and capacity of resource extraction and refining operations (oil,
gas, minerals, etc.) in the source water protection area
S Volume and type of industrial activity in the source water protection area
S Use of waterbody for commodity transport
S Prevalence of hazmat transport on roads, rail lines, and waterways in the source water
protection area
S Prevalence of wastewater and stormwater outfalls in the source water protection area
S Prevalence of power generation facilities in the source water protection area
If the results of the all-hazards risk assessment indicate that the risk of releases to source water is
significant, consider actions to prepare for and mitigate that risk, such as:
S Conduct a thorough inventory of source water contamination threats and
prioritize those threats to focus attention on those that present the greatest
risk to a community water system
o Take advantage of the source water provisions of Section 2018 of
America's Water Infrastructure Act, which give community water
systems access to Tier II chemical inventory data collected under the
Emergency Planning and Community Right to Know Act (U.S. EPA, 2019b) for facilities in
a source water protection area
S Share contact information with facility owners and arrange for direct notification in the event of
a release from that facility
o Ensure that notifications of releases that are reported under the Emergency Planning
and Community Right to Know Act are promptly reported to community water systems
that could be impacted (U.S. EPA, 2019b)
S Develop relationships with Local Emergency Planning Committees and local first responders and
arrange for timely notification of releases, particularly those resulting from transportation
accidents, that could impact a source water
S Identify the materials stored, used, or discharged at facilities that pose the greatest acute risk to
source water
S Identify methods and laboratories that can analyze for these materials
S Evaluate the ability of treatment processes to remove or neutralize these materials
S Update emergency response plans to include procedures to respond to releases to source water
S Plan for controls such as booms or curtains that could protect a water intake from a
contaminant plume
Assessing
the risk
Actions to
prepare
39
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
References
ATSDR, 2018. National Toxic Substances Incidents Program. Retrieved from
https://www.atsdr.cdc.gov/ntsip/index.html
ATSDR, 2009. Hazardous Substance Emergency Events Surveillance. Retrieved from
https://www.atsdr.cdc.gov/hs/hsees/index.html
Balasubramanian, S.G. and Louvar. J.F., 2004. Study of major accidents and lessons learned. AlChE
Process Safety Progress.
Becker, R. A., 2015. Massive Oil Spill in Yellowstone River Contaminates Drinking Water. PBS. Retrieved
from https://www.pbs.org/wgbh/nova/article/massive-oil-spill-yellowstone-river-contaminates-
drinking-water/
Biesecker, M and Suderman, A, 2018. Dam breach at N.C. power plant causes coal ash spill into Cape
Fear River. Insurance Journal. Retrieved from
https://www.insurancejournal.com/news/southeast/2018/09/24/502096.htm
Brody, T.M., Bianca, P.D., and Krysa, J., 2012. Analysis of inland crude oil spill threats, vulnerabilities, and
emergency response in the Midwest United States. Risk Analysis, Vol. 32, No. 10.
Etkin D.S., 2004. Twenty-year trend analysis of oil spills in EPA oil jurisdiction. Freshwater Spills
Symposia. Retrieved from
https://archive.epa.gov/emergencies/content/fss/web/pdf/etkin_04.pdf
Grinberg, Emanuella, 2019. The Kentucky River is Brimming with Dead Fish After a Fire at a Bourbon
Warehouse. CNN. Retrieved from https://www.cnn.com/2019/07/07/us/jim-beam-warehouse-
fire-kentucky-environmental-impact-trnd/index.html
Howard, M., Morin, I., and Watts, K., 2008. Review of oil spill incidents reported to the National
Response Center from onshore oil production facilities. International Oil Spill Conference
Proceedings. Vol. 2008, No. 1, pp. 561-569. Retrieved from https://doi.org/10.7901/2169-3358-
2008-1-561
Louisville Water Company, 2017. Update on Ohio River Spill. Retrieved from
https://www.louisvillewater.com/newsroom/update-ohio-river-spill
Maloney, K.O., Baruch-Mordo, S., Patterson, L.A., Nicot, J.P., Entrekin, S.A., Fargione, J.E. Kiesecker, J.M.,
Konschnik, K.E., Ryan, J.N., Trainor, A.M., Saiers, J.E., Wiseman H.J., 2017. Unconventional oil
and gas spills: materials, volumes, and risks to surface waters in four states of the U.S. Science of
the Total Environment, 581-582 (2017) 369-377.
Murawski, J., 2018. Duke Energy Says Coal Ash Isn't Contaminating Cape Fear River; State Awaits Its Own
Tests. The News & Observer. Retrieved from
https://www.newsobserver.com/news/local/article218893790.html
40
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
National Park Trips Media, 2017. Pipeline Spills Oil into Yellowstone River Again. Retrieved from
https://www.yellowstonepark.com/news/pipeline-oil-spill-yellowstone-
river#:~:text=January%2017%2C%202015%2C%20an%20oil,River%20on%20near%20Glendive%
2C%20Montana.&text=The%20Bridger%20Pipeline%20Company%20reported,of%2042%2C000
%20gallons%20of%20oil.
ORSANCO, 2018. Ohio River Valley Water Sanitation Commission Annual Report 2018. Retrieved from
http://www.orsanco.org/wp-content/uploads/2018/10/2018_Annual_Report.pdf
PHMSA, 2020. U.S. Department of Transportation Pipeline and Hazardous Materials Safety
Administration. Retrieved from https://www.phmsa.dot.gov/hazmat-program-management-data-
and-statistics/data-operations/incident-statistics
Power Partners Inc., 2009. Single Phase Overhead Distribution Transformers, PDL 46-200, Athens, GA.
Retrieved from http://www.powerpartners-usa.com/wp-content/uploads/2016/01/single-
phase_product_spec_sheet.pdf?x30412
Rosen, J.S., Whelton, A.J., McGuire, M.J., Clancy, J.L., Bartrand, T., Eaton, A, Patterson, J., Dourson, M.,
Nance, P., and Adams, C., 2014. The crude MCHM chemical spill in Charleston, W.Va. JAWWA,
106:9
Sack, W, 2016. American Proteins to pay $50,000 fine for violations including Mulberry Fork acid spill.
The Cullman Tribune. Retrieved from https://www.cullmantribune.com/2016/12/01/updated-
american-proteins-to-pay-50000-fine-for-violations-including-mulberry-fork-acid-spill/
Sangul, H., Santella, N., Steinberg, L.J., and Cruz. A.M., 2012. Analysis of hazardous material releases due
to natural hazards in the United States. Disasters. Retrieved from
https://doi.Org/10.llll/j.1467-7717.2012.01272.x
States, S., Cyprych, G., Stoner, M., Wydra, F., Kuchta, J., Monnel, J. and Casson, L., 2013. Marcellus shale
drilling and brominated THMs in Pittsburgh, PA Drinking Water. JAWWA, 105:5
U.S. Congress, 2018. America's Water Infrastructure Act. Retrieved from
https://www.congress.gov/bill/115th-congress/senate-bill/3021/text
U.S. CSB, 2016. Chemical Spill Contaminates Public Water Supply in Charleston, West Virginia, Report
No. 2014-01-l-WV.
U.S. EPA, 2006. How-to Manual: Update and Enhance Your Local Source Water Protection Assessment.
EPA 816-K-06-004. Office of Water, Washington, D.C. Retrieved from
https://www.epa.gov/sourcewaterprotection/how-manual-how-update-and-enhance-your-local-
source-water-protection
U.S. EPA, 2014. Duke Energy Coal Ash Spill in Eden, NC. Washington, D.C. Retrieved from
https://www.epa.gov/dukeenergy-coalash/history-and-response-timeline
41
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
U.S. EPA, 2015a. Emergency Response to August 2015 Release from Gold King Mine. Washington, D.C.
Retrieved from https://www.epa.gov/goldkingmine
U.S. EPA, 2015b. Review of State and Industry Spill Data: Characteristics of Hydraulic Fracturing-Related
Spills, EPA/601/R-14/001. Retrieved from https://www.epa.gov/sites/production/files/2015-
05/documents/hf_spills_report_final_5-12-15_508_km_sb.pdf
U.S. EPA, 2018. Clean Water Act Hazardous Substances Spill Prevention Proposed Rule. Retrieved from
https://www.epa.gov/rulemaking-preventing-hazardous-substance-spills/proposed-rulemaking-
clean-water-act-hazardous
U.S. EPA, 2019a. Overview of New Risk Assessment and Emergency Response Plan Requirements for
Community Water Systems. Retrieved from https://www.epa.gov/waterresilience/overview-new-
risk-assessment-and-emergency-response-plan-requirements-community
U.S. EPA, 2019b. America's Water Infrastructure Act, Section 2018: Amendments to the Emergency
Planning and Community Right to Know Act. Retrieved from
https://www.epa.gov/sites/production/files/2019-
05/documents/awia_sec_2018_factsheet_for_water_sector_final.pdf
U.S. EPA, 2020. Toxic Release Inventory. Retrieved from https://www.epa.gov/toxics-release-inventory-
tri-program
WTXL, 2019. More than 7 million Gallons of Sewage May Have Contaminated Withlacoochee River.
WTXL ABC 27. Retrieved from https://www.wtxl.com/news/local-news/sewage-spill-in-valdosta-
results-in-over-7-000-gallons-of-contamination-in-withlacoochee-river
Zhu, T., Jia, Y., Chao, X., Frihi, M., and Hammouri, M., 2009. Development of NCCHE chemical spill
incident database. Published by the National Center for Computational Hydroscience and
Engineering, School of Engineering, University of Mississippi. Retrieved from
https://www.ncche.olemiss.edu/
42
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
Glossary
Clean Water Act Hazardous Substances. A list of substances defined under authorities of Section
311(b)(2) of the Clean Water Act (Title 40 of the CFR, Part 116).
Community Water System. A system that provides water for human consumption through pipes or
other constructed conveyances and has at least fifteen service connections or regularly serves at least
twenty-five individuals, and which serves the same population year-round (as defined in SDWA section
1401(15)).
Hazardous Substance Emergency Events Surveillance. An environmental surveillance system
established by ATSDR to collect and analyze information about acute releases of hazardous substances
and threatened releases that result in a public health action such as an evacuation. The system was
active from 1990 through 2009.
National Hydrography Dataset. A dataset maintained by the USGS that represents the water drainage
network of the United States with features such as rivers, streams, canals, lakes, ponds, coastline, dams,
and stream gages.
National Response Center. The designated federal point of contact for reporting all oil, chemical,
radiological, biological and etiological discharges into the environment, anywhere in the United States
and its territories. The National Response Center is part of the federally established National Response
System and staffed 24 hours a day by the U.S. Coast Guard.
National Toxic Substances Incidents Program. A surveillance program established by ATSDR to track the
release of toxic substances, which was active between 2010 and 2018.
Toxic Release Inventory. A program created under the Emergency Planning and Community Right to
Know Act that requires certain industries to file an annual report documenting releases of certain toxic
chemicals that may pose a threat to human health and to the environment. Reporting is limited to a list
of approximately 755 individual chemicals and 33 chemical categories.
Zone of Concern. For surface water intakes, an area that extends 50 miles upstream, / mile
downstream, includes all major tributaries, and includes a % mile buffer inland from the waterbody area
boundary. For ground water wells, an area defined by a Vz mile radius around the well location. This
definition was used solely for the purposes of the study presented in this report.
43
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
Appendices
44
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
Appendix A: Fields Extracted from the NRC Reports and Used in
Analysis
CALLS - Contains Report receipt and Suspected Responsible Party information
FIELD NAME
FIELD DESCRIPTION
NRC Report Number
Unique Identifier assigned to each report (known as SEQNOS)
Call Type
Cateqorized Type of Call: INC = Incident; DRL = Drill
Responsible Company
Name of Suspected Responsible Company
Responsible Orq Type
Orqanization Type of the Suspected Responsible party
INCIDENT COMMONS - Contains general information which is common to all types of Incident Reports
such as location, type of incident, cause, etc.
FIELD NAME
FIELD DESCRIPTION
Description of Incident
Detailed explanation of the incident
Type of Incident
Specific type of incident beinq reported
Incident Cause
Cause of the incident
Incident Date Time
Date and time incident occurred, was discovered or planned
Incident Location
Additional details about the location of the spill (e.q., river mile marker)
Location Address
Street address or intersection nearest to the incident location
Location Nearest City
City or town nearest to the incident location
Location State
State where incident occurred
Location County
County where incident occurred
Location Zip
Postal zip code where incident occurred
Lat (Deq, Min, Sec, Quad)
Deqrees of Latitude for incident location
Lonq (Deq, Min, Sec, Quad)
Deqrees of Lonqitude for incident location
INCIDENT DETAILS - Contains information which further describes the situation and impact of the Incident.
FIELD NAME
FIELD DESCRIPTION
Damaqes
Indicates if there were any damaqes that occurred durinq the release
Waterway Corridor Closed
Indicates if any waterway traffic corridors were closed
Track Closed
Indicates if any rail tracks were closed
Medium Description
Medium affected as a result of the incident
Body of Water
Immediate body of water impacted by the incident if applicable
Water Supply Contaminated
Indicates if a drinkinq water source was contaminated by the release
Nearest River Mile Marker
The nearest river mile marker of the incident location if applicable
Offshore
Indicates if the incident location is offshore
INCIDENTS - Contains information specific to each type of Incident
FIELD NAME
FIELD DESCRIPTION
Type of Fixed Object
Cateqorized object type
Capacity of Tank
Capacity of the storaqe tank
Capacity of Tank Units
Unit of measure for the storaqe tank capacity
MATERIAL INVOLVED - Contains information specific to each material spilled
FIELD NAME
FIELD DESCRIPTION
CHRIS Code
Chemical Hazards Response Information System Code
CAS Number
Chemical Abstracts Service Reqistry Number
Amount Of Material
Amount of material released
Unit Of Measure
Unit of measure for amount released
Name of Material
Name of material released
Amount In Water
Amount of material that reached water
Unit of Measure Reach Water
Unit of measure for amount in water
TRAINS DETAIL - Contains information about the train(s) involved
FIELD NAME
FIELD DESCRIPTION
Train Type
Type of train involved
VESSELS DETAIL - Contains information about sea going vessels involved with an incident
FIELD NAME
FIELD DESCRIPTION
Vessel Type
Type of vessel involved
MOBILE DETAILS - Contains information about mobile vehicles involved with an incident
FIELD NAME
FIELD DESCRIPTION
Hazmat Carrier
Indicates if a transport for hazardous materials
45
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
Appendix B: Rules to Standardize Names of Bodies of Water
Standardized Name
If the Body of Water name contains the following:
Mississippi River
"Mississ"
Unnamed Creek, Stream, or Tributary
"Unnamed Creek," "Unknown Creek," "Local creek," or
"Unknown Tributary"
Genesee River
"Genes"
Monongahela River
"Monog," "Monon," "Mong," or "Monoag"
Schuylkill River
"Schuykill," "Schyukill," or"Schulkill"
Vermilion River
"Vermi"
Willamette River
"Wilam" or "Willam"
Passaic River
"Passaic"or "Passiac"
Tennessee River
"TN River" or "Tenn"
Allegheny River
"Alleg"
Androscoggin River
"Scog"
Delaware River
"Delaware"
Colorado River
"Colorado River"
Cuyahoga River
"Cuyahoga"
Elizabeth River
"Elizabeth River"
Acushnet River
"Acu"
Anacostia River
"Anac"
Anclote River
"And"
Arkansas River
"Arka"
Brazos River
"Braz"
Withlacoochee River
"Withl"
Arthur Kill
"Arthur Kill"
Kill Van Kull
"Kill V"
Canal
"Canal"
Unnamed Groundwater Source
"Water Well", "Well Water", "Well", "Groundwater", or
"Aquifer"
Unnamed Reservoir, Lake, or Pond
"Unknown Pond", "Reservoir", or "Unknown Lake"
Drainage and Collection
"Storm," "Drain," "Sewer," "Ditch", "Basin", "Runoff", or
"Culvert"
Drinking Water
"City Water", "Tap Water", "Drinking Water", "Drinking
Water Wells", or "Potable Water"
Unnamed Wetland Area
"Wetland," "Marsh," or "Swamp"
46
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
Appendix C: Unit Conversion Multipliers
For each unit of measure below, multiply the value reported in those units by the multiplier to convert
to gallons (U.S., liquid).
Barrels (petroleum) = 42.0
Barrels (not petroleum) = 31.5
Cubic Meters = 264.172
Cubic Yards = 201.974
Cups = 0.0625
Drops = 0.0000132086
Gallons = 1.0
Liters = 0.264
Ounces = 0.0078125
Pints = 0.125
Pounds = 0.12 (assuming a density of 8.345 pounds/gallon, which is the density of water)
Quarts = 0.25
Tablespoons = 0.00390625
Tons = 269.0 (assuming a density of 8.345 pounds/gallon, which is the density of water)
47
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
Appendix D: Process for Preparing Spatial Data
Generating a "best address" for records lacking latitude/longitude information:
o Removing leading and trailing white space
o Correcting format of address record by applying a script that matches one or more
numbers, followed by a space, and followed by one or more characters
o Reviewing the NRC fields: "Location Address", "Location Street 1", and "Location Street
2" for valid addresses
o Removing any value after a decimal in zip codes and adding a leading 0 to the zip code in
states whose zip codes are verified to contain leading zeros
o Addresses were run through the 2012-2013 version of the Max Rice/Juice Analytics
Excel geocoding tool and latitude & longitudes for each record with a validated "best
address" were provided
Geocoding records with latitude/longitude degrees, minutes, and seconds:
o Deleting duplicate rows
o Records with any missing information in the latitude/longitude degrees, minutes, or
seconds were omitted from the geocoding process
o Records with latitude degrees <18 or >72 were omitted from the geocoding process
(outside of the U.S.)
o Records with longitude degrees <66.5 or >180 were omitted from the geocoding process
(outside of the U.S.)
o Records with latitude/longitude minutes or seconds outside the range of 0-60 range
were omitted from the geocoding process
o Formulas were applied & latitude/longitude generated:
ฆ Latitude: degrees + (minutes/60) + (seconds/3600)
ฆ Longitude: (degrees + (minutes/60) + (seconds/3600))*-l
Latitudes & longitudes for each record were added to QGIS and clipped to the appropriate
delineated zone
48
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
Appendix E: Materials Excluded from Analysis
Releases of the following materials were excluded from the analysis, however, not all examples and
variations of the contaminant name are listed below.
AMMUNITION/EXPLOSIVES
BATTERIES AND RELATED PRODUCTS
BENTONITE/CLAY
COAL
CONSTRUCTION MATERIALS
o ROOFING SHINGLES
o STEEL COIL
o CEMENT/CONCRETE
o FLORESCENT LIGHTS
o WOOD STEPS
DYES
EARTH/MINERAL PRODUCTS
o GRAVEL
o GYPSUM
o LIMESTONE
o LEAVES
o DIRT
o ROCKS
FOAM PRODUCTS
o BROWN FOAM
o FOAM BLANKET
o BLUE DYE- FOAM MARKER
o STYRAFOAM
METAL AND SCRAP
o METAL SCRAPS
o METAL SHAVINGS
o JUNK AND SCRAP METAL
o OLD TIRES
o RUST AND METALLIC SUBSTANCE
MISCELLANEOUS WASTE
o BAG OF TRASH
o BURNING TIRES
o GARBAGE/DEBRIS
o TOILETS
VOLATILE CHEMICALS
o BUTANE
o METHANE
o CNG
o LNG
o NATURAL GAS
WATER
o CHILL WATER
49
-------
Occurrence of Releases with the Potential to Impact Sources of Drinking Water
o CHLORINATED WATER
o DRINKING WATER
o CITY WATER
o GROUNDWATER
o FIREWATER
o RAINWATER
50
------- |