Relationships between Private Domestic Wells and Underground Storage Tanks: Evaluation of Mapping and Plume Transport Tool Implementations


EPA/600/R-17/282 | September 2017 | www.epa.gov/research
United States
Environmental Protection
Agency
oEPA
Relationships between
Private Domestic Wells
and Underground Storage
Tanks: Evaluation of Mapping
and Plume Transport Tool
Implementations
Office of Research and Development
National Risk Management Research Laboratory | Groundwater, Watershed, and Ecosystem Restoration Division

-------
Relationships between Private Domestic
Wells and Underground Storage Tanks:
Evaluation of Mapping and Plume
Transport Tool Implementations
James W. Weaver
United States Environmental Protection Agency
Office of Research and Development
Ada, OK 74820
Andrew R. Murray
Oak Ridge I nstitute for Science and Education
Cincinnati, OH 45268
Anish Khanal
Oak Ridge Institute for Science and Education
Ada, OK 74820
Fran V. Kremer
United States Environmental Protection Agency
Office of Research and Development
Cincinnati, OH 45268
l

-------
Abstract
Well water from private domestic we I Is (hereafter private wells) is often not tested as private owners
are exempt from sampling requirements of the Safe Drinking Water Act. Numerous incidents of
contamination of waterin private wells have been reported, however. Potential contaminant sources,
like underground storage tanks, are widespread acrossthe United States. This report describes a pilot
project using a geographic information systems (GIS) application that was developed to display locations
of underground storage tanks and indicate the likelihood that the re are private wells within several
selected distances. A few locations can be selected by the application user or, when data from an entire
state or region are available, a large area can be viewed at a glance. The pilot project was developed
for Oklahoma, because of the large amount of freely available data, but could be extended to other
locations where data are available.
2

-------
Table of Contents
Abstract	2
Table of Contents	3
List of Figures	5
Quality Assurance	6
Introduction	7
Private Well Locations	7
Underground Storage Tank Locations	8
Locations of Potential Impacts	8
Evaluation	10
Application Documentation	12
Introduction	12
Opening Screen	12
Application Tools	13
Legend	13
Operational Layer List	13
Charts	14
Select	16
Search	16
Base map	17
Attribute Table	17
Relational Tables	21
Single-Tank Analysis	22
Defining a Location	23
Setting Potential Impact Distance	23
Calculate Estimated Numberof Domestic Wells	25
Within the Potential Impact Distance	25
3

-------
References	27
Appendix A: Assessment of Plume Extents	29
Length of BTEX Plumes	29
Summary of Simulations	31
Conductivity	31
Dispersivity	32
Biological Degradation (Half-life)	33
Effect of Source Concentration on Plume Length	34
Retraction of Plumes	34
Result Summary	38
Appendix B: Implementation Details	39
Summary	39
Tools	40
4

-------
List of Figures
Figure 1. Example of selected tank site showing 1500 ft radius buffer distance, estimated number of
private wells (53), housing units (251) and population (520)	9
Figure 2. Locations of regulated tanks in the Oklahoma City area. Circles represent 1500 ft radius zone
aroundtanks which are colored bythe estimated numbers of private wells withinthe zone. Red isthe
highestdensity of wells inthe 1500 ft radiusand greenthe lowest. Areas with most red-colored circles
generally lack public water or are in areas of legacy historical private well use	10
Figure 3. Longitudinal dispersivity data from Gelharetal. (1992) plotted against "1/10" of length scale,
and the weighted regression formula of Xu and Eckstein, 1995. Most of the reliable data were at plume
scales of 10 to 100 m	30
Figure 4. Plume lengths at different conductivities	31
Figure 5. Plume lengths at different dispersivities	32
Figure 6. The plume length at different biological degradation rate	33
Figure 7. The effect of changing source concentration (Co) on plume length	34
Figure 8. Extent of contaminant plume as defined by 0.005 mg/L concentration showing plume
retraction in low conductivity aquifer material. (Solution C14of van Genuchten (1981), dispersivity of 10
m, and biodegradation half-life of 1.1 year)	35
Figure 9. Extent of contaminant plume as defined by 0.005 mg/L concentration showing plume
retraction in moderately conductive aquifermaterial. (Solution C14of van Genuchten (1981),
dispersivity of 10 m, and biodegradation half-life of 1.1 year)	36
Figure 10. Extent of contaminant plume as defined by 0.005 mg/L concentration showing plume
retraction in highly conductive aquifermaterial. (Solution C14of van Genuchten (1981) , dispersivity of
10 m, and biodegradation half-life of 1.1 year)	37
5

-------
Quality Assurance
This project was conducted underan approved U.S. EPA quality assurance project plan (ORD Project
QA ID #G-GWERD-0019367). Evaluation of the data and estimates methods used forthe software
tool described herein are presented inthe EPA report EPA 600/R-17/175, titled "Proximity of Private
DomesticWellsto Underground Storage Tanks: Oklahoma Pilot Study".

-------
Introduction
Ground water contamination carries the risk of impacting private domestic we I Is (hereafter simply
"private wells"). Because these wells are not covered by the Safe Drinking Water Act, testing of the
water supply is at the discretion of the owner (U.S. EPA 2004). In many cases well owners are unaware
that watershould be tested, what parameters should betested,orhowtogo about testing (see Ridpath
et al., 2016). As a consequence, private well water often goes without testing, and numerous incidents
where people have been exposed to contaminated water have been documented (e.g., Anderet al.
2016, DeSimone et al., 2009, Schaider et al. 2016, U.S. EPA 2002).
In the U.S., a common ground water contaminant source that potentially threatens well users is
underground storage tanks (USTs). Petroleum product releases have been reported from over 530,000
underground storage tanks, with almost 71,000 cleanups remainingto be completed (US EPA 2016).
One of the main potential pathwaysforexposureto petroleum hydrocarbonsoriginatingfrom leaking
underground storage tanks isthe consumption of waterfrom private domestic wells.
The purpose of this Geographic Information Systems (GIS) Application (APP) is to demonstrate a method
to showthe relationship between point sources of subsurface contaminantsand private wells. Because
of the large number of reported releases, underground storage tanks are taken as the potential source
of leaksforthis work.
Private Well Locations
Knowledgeof private well locations varies by state. Most of the data, however, are limited by
undercounting (record keeping may have only started recently, orall required recordsare not reported),
lack of easy access to data (e.g., data held as paperor PDF copies, cost to requesterfor release of data),
or legal restrictions (Weave retal., 2017). Because of these limitations we adopt the estimation
methods developed by Weaveretal. (2017) and Murray etal.,(2017) which use the 1990 census as a
baseline, and projects forward in time using census and/or state agency private well records. Originally
the work focused on a pilot study in Oklahoma (Weaveretal., 2017), but was extended by Murray et al.
(2017) to the entire U.S. Additionally, Murray increased the resolution by reducing the spatial data unit
from the census block group (average area of 42.12 km2) to the census block (average area of 0.826
km2).
7

-------
Underground Storage Tank Locations
Upon its creation by the Superfund Amendments and Reauthorization Act, the federal underground
storage tank program was delegated to the states. Each state manages their program according to state
priorities and the federal government col lects only selected data on each program. The locations of
regulated orleakingtanksare held bythe states. With data froman entire state, obviously,the
locations of tanks with the most potential for impacting private wells can be seen at a glance. Where
this information is not available, locations can be selected and the number of potentially-impacted wells
be determined.
Locations of Potential Impacts
In the following example regulated tanks are shown as circles of a specified radius. I nth is case 1500 ft
was chosen, although 1000 ft is also available, and for single tanks any distance can be chosen. This
distance correspondstoa maximum reasonable benzene, toluene, ethyl benzene and xylene (BTEX)
plume extent based on empirical data analysis and plume modeling (see Appendix A: Assessment of
Plume Extents). The distance roughly follows from plume studies and mode ling (see Appendix) which
show an upper bound of BTEX plume length on this order of magnitude. Site-specific investigation,
however, is the only way to determine the actual direction and extent of contaminant plumes. The
information presented inthis APP is intended asa screeningapproach forplanning:
•	If a tank (or planned tank location) is nearby, are there private wells within a specified potential
impactzone?
•	Should cleanupsorinspections be prioritized in areas with a high density of private well use?
In areas where datasetsgivingthe locationsof all underground storage tanks are available, a location
can be selected onthe mapand estimatesofthe numberof private wells, housingunitsand population
can be generated for use r-input potential impact zone areas (Figure 1). This usage of the tool addresses
the first question above. Details on the procedure for using these features are found in the section
"Single-Tank Analysis".
Areas of co-location of underground storage tanks and high reliance on private wells can thus be seen at
a glance (Figure 2). Details on the procedure for using these features are found in the section "Opening
Screen".
The approach of usinga potential impactzone follows from fixed-radius methodsof wellhead protection
(U.S. EPA, 1994). Absent a site-specific investigation, the direction of ground water flow is not known.
Because of preferred directions of ground water flow (in response to pressure gradients), however, it is
unlikelythattheentireareaofeach potential impactzone would be impacted.
8

-------
©


resultLayer
Sum Est_Wells divided by Area in Square Kilometers
80.2681943 - 80.2681943
others
bufferLayer


resultLayer
1500
4
0.656506029116601
0.656626474977382
52.6965535066006
520
251
Zoom to
Figure 1, Example of selected tank site showing 1500 ft radius buffer distance, estimated number of
private wells (53), housing units (251) and population (520).
9

-------
QUAIL RlDGE
John Kilpatrick Tpke
SPUING
I/ALLEY
ESTATES
tu^ef
fCHER
Lake
Hefner
The Village
Nicho ;¦ (r .
Aluma
,B£EN ^TURES
Vv'arr Acres
wEsrGArJ\
Oklahoma Gijy JQ-
wEsrs i
Smith Village A
DdW
WlNDfb&S? t
Tinker fr).
, Air VlfilOGEFIl
'Jorce
Vallfv firook
JHOGEWEW
ESTATES
Moore
iGE SOUTl
Piedmoht
turner
Jones
Woods
LIBERTY tigjf*
w Murray | © 2017 Microsoft Corporation and its c
Figure 2. Locations of regulated tanks in the Oklahoma City area. Circles represent 1500 ft radius zone
around tanks which are colored by the estimated numbers of private wells within the zone. Red is the
highestdensity of wells inthe 1500 ft radiusand greenthe lowest. Areas with most red-colored circles
generally lack public water or are in areas of legacy historical private well use.
Evaluation
Assessmentofthe spatial relationship between underground storage tanksitesand private wells
requires certain assumptions. These are necessary for two reasons. The first is because of data
limitationsthe locationsofall private wells are not known. Further, locations of known private wells in
the U.S. are not compiled at the national level. Existing state records are generally limited, because in
most states, reporting requirements have been imposed only recently. Thus in this work, private well
locationsare estimated bythe methods developed by Weaveretal. (2017).
The second reason is that the potential for contaminant impact on a well depends on factors influencing
movement and degradation from the source of contaminants to the well. The question is "does the
contaminant reach the well at a high enough concentration to create a health risk?" Answering this
question from mapping application is limited because contaminant transport depends on factors that
are not mapped in detail on a national basis. These factors include hydraulic conductivity, sorption and
heterogeneity of aquifer materials; the amounts of available electron acceptorsto drive biodegradation;
10

-------
and the local gradient of the watertable. The later can be estimated in a gross fash ion from surface
topography, but is limited by heterogeneity, well pumping rates, and impervious surface cover.
As noted in Appendix A: Assessment of Plume Extents, empirical data and modeling independently
provide estimates of the plume extent. Empirical data are limited bythe numberof sitesexamined and
the quality of site data; modeling is limited by assumptions built into the models, data limitations, and,
in this case, the lack of calibration to fie Id conditions. Plume extents thus calculated from these sources
can best be viewed as a guide for site-specific investigation, as it is site-specific investigation,
monitoring, and samplingthatestablishesactual plume extents.
Thus the results provided in this report are best viewed as guides to program planning and additional
investigation as map-derived data are limited and assumptions are used for determining both private
well locations and the extent of contamination from a presumed source.
11

-------
Application Documentation
Introduction
By combining publicly available data on Oklahoma's regulated underground storage tanks and estimates
of domestic ground water we 11 use at the Census Block level, the end user is able to use this application
to explore the varying potential vulnerability of contamination state wide for Oklahoma. Application was
made to Oklahoma fordemonstration purposes. Foruse in other states,tank location data could be
added to the EPA-de rived estimates of private we 11 usage. The following describes the capabilities of
the application and isa general guide on howto usethose capabilities.
Opening Screen
The application opens up to a map viewer with three basic components. The first, and most basic, are
the navigation controls in the top left of the window which allow you to zoom in and out, navigate to
the initial map extent (home button) or navigate to your location (currently limited to Oklahoma). The
second set of components are the application tools on the right-hand side. Here you will find tools
which will allow you to manipulate the map display, as well as to facilitate some simple statistical
analyses. The third component is the attribute table which will allow you to view specific information
relatingto facilities with USTsas well as information on specifictanks. Each of these will be discussed in
detail below.
5400096
*2706590
9000 S
PENNSYLVANIA
140022*
Moor*
Howufds Sorv>c» 729 NW 12TW
Onser
1400714
•97.51
16.000.00
Oklahoma UST Vulnerability

Attribute Tables
Lave?
Operational layers
Available Layers
12

-------
Application Tools
© «i»
The application toolbar contains eight distinct tools; the legend, operation layers, charts,
select, search, basemap selection, analysis, and draw.
Legend
Legend

f U * Q,
i
©
Vulnerability Based on Wells
Wells Within 1,500 ft.
||1 0-1
¦ 2-5
~ 6-10
1 11-25
26-135
Clicking on the legend reveals the symbology of all the currently
selected operation layers. Whenthe application isfirstopened,the
defaultoperational layernamed "Wells Within 1500 ft" will appearin
the map vie we rand the legend will appear as is shown in the figure to
the right. This layer portrays a 1500 ft potential impact around each
facility with an underground storage tank and is color coded based on the estimated number of
domestic wellsthat it contains.
Operational Layer List
The laye r I ist d isplays al I of the available layersthat can
be added to the map. It will initially look I ike this figure
to the right, but clicking on the left-hand arrows will
expandthe categoriesand reveal individual layers. The
available layersarethe initial 1500ft potential impact
zone layer, an alternative 1000ft potential impact
zone, and a layer showing estimated domestic we I Is at
the Census block level.
:= ^ jM
#3 Q, ¦
!¦ ®
Layer List

A X
Operational layers
~ I | Domestic Wells (Census Block)
~ ( | Vulnerability Based on Tanks
t Q Vulnerability Based on Wells
.v'
Tank Data
To account for preferred plumetransportdirections,
sectorsare drawn withinthe circlesto indicatethe
preferred direction based on the surface topography
and an assumed 60% variation inthe gradientdirection
(Haitjemaand Mitchell-Burker, 2005 ). This approach is based on the premise that ground water flows
in directions that mimic the surface topography. Because of a number of factors, the actual ground
water flow direction may differ from that predicted from surface topography. Key factors, especially
changing infiltration, pumpingfrom wells which can be highly variable, and subsurface heterogeneity,
require site-specific analysis and investigation. Thus the results from the transport tool are presented as
a firstapproximation which requires further site-specific data col lection, which is likely to require field
sampling. Again, site-specific investigation is the only way to determine the actual direction of
contaminant transport, butthere may be some information gained in estimatingthe majordirectionsof
13

-------
transport. Impacts should not be ruled out because ofthe locations of the sectors. Sectorsforthe 1000
ft and 1500 potential impactzonesare available.
Lastly, in the "Vulnerability Based on Tanks" category, the re is a layerthat shows the number of physical
tanks within a distance of 1500 ft from any point in the state. When using the Vulnerability Based on
Tanks category, turn offthe "Well's within 1500 ft" layer, to ensure only one layeris made visible ata
time to reduce the chance of confusing one or more layers with similar color schemes.
- Domestic Wells (Census Block)
... ¦r|v'[ Vulnerability Based on Wells
• ••
Est. Wells
^ >25 to 100
^ > 10 to 25
> 5 to 10
~ UST Facilities
Active Non-LUST
¦ Active LUST
¦ Inactive, Non-LUST
• ••
> 0 to 5
¦ Inactive, LUST

0 to 0
"| Possible 1,000 ft Plume
• ••
~ I Vulnerability Based on Tanks
• ••

~ |v| Tanks within 1,500 ft
~ ( [ Possible 1,500 ft Plume
• ••

¦ 1
"

2-5
1,000 ft. Facility Buffer
• ••
6-10
¦ 11 - 25
~

¦ >25
Wells Within 1,500 ft.
¦ 0-1
¦ 2-5
• ••
~ 6-10
B 11-25
I 26-135
Charts
The one available chart istitled "Estimated
Number of We I Is within Varying Distances", this
chart allows the end userto define a geographic
area and then to visualize the estimated wells
potentially vulnerable to LUST contamination
through a bar graphthat wiil display estimated
number of wellsforthe estimated 1000 and

u p :
¦ ©
-------
1500 ft sectorareas, and the 1000 and 1500 ft potential impactzones.
To create a chart, select "Estimated NumberofWells Within Varying Distances", then click "Use spatial
filterto limit features". Then you can either use the current map extent, or predefine an area which will
let you draw a specific area of interest, then click apply.
Choctaw Propane -
* Use spatial filterto limit features
Only features intersecting the current map area
• Only features intersecting a user-defined area
The resulting chart pops up in the window. A
magnifying glass is visible in the top right corner of the
results which will popoutthe graphto make it more
readable. Redflagsappearinthe map viewerwhich
denotethefacilitiesthatwere included inthegraph.
The x-axis re presents estimated number of we I Is that
are vulnerable. Blue denotesthe 1000ft sector, orange
denotes the 1500 ft sector, grey re presents the 1000 ft
potential impact zone, and yellow re presents the 1500
ft potential impact zone.
9 \ rJ / ~
m m m * *
< Options Chart Results
Estimated Number of Wells within Varvina Distances
Rockys Place ¦
Goddard Ready Mixed Corcrete ¦
Choctaw Propane Ire
15
-------
Select
The select tool allows the use rto draw a rectangle, circle, or polygon to select specific areas of interest
for further investigation using the attribute table1. When you select specific sites, you can then run
simple statistics on those areas of interest within the attribute table (more on that below).
SL Oklahoma UST Vulnerability
Well* Within 1.500
BJ Uptioni »
FAC.NUM
- fry men V -.nuo gB U
Numb*' of Tinici 'it me
«r Mtacoari K*freif>
Wo'tii Gti

de Total Cipeety n Wellj >n i .CN
G»!lon, fW.
30ft WaU»t»1j00ft Wein.n 1.000'
Fume 8u««r
5501076
1 Godoa'C Reaoy
Mixed Concrete
16665 NE23STR Cnotnw
35.50
¦912*
4,OOOjOO 0-27
2.02 456
5502041
4 Hamngton
Service Ce*«»r
*4003 ME 23RD Ooca»
3549
-97.28
28.000.00 0.97
2-22 520
16 features 0 selected

¦ ¦
¦ ¦
cf
%
111
& H

The search tool simply allows the use rto input a location to relocate the map extent. The user could
searchfora city, zipcode,address,etc...
1 The chart tool contains a separate selection tool,
16
-------
Basemap
The basemaptool is usedtoselectthe backgroundforthe BasemapGallery
map extent. The default isaerial imagery with labels, but
many other options exist such as road maps, topographic
maps, and various others, asshown inthe figure tothe
right.
Light Gray National Oceans
Canvas Geographic
OpenStreetMap Streets Terrain with
Labels
Topographic USA Topo Maps USGSNatonal
Map
Attribute Table
The attribute table allows the use rto dive deeper into the information on specific facilities and
individual tanks. When the application is opened, the attribute initially appears at the bottom of the
screen. It can be hidden or expanded by clicking the arrow at the top of the table. There are two tables
that are associated with operational layers; "UST Facilities" and "Wells within 1500 ft" A third table is
the individual tank data which is a relational database connected to both of the othertwo operation
l ayers. Esse ntial ly the tables forfacil ities and wells within 1500 ft are the same and are boththere for
convenience when selecting from different layers. The select tool is ideal to use in conjunction with the
attribute table and will allow a closer look at specific areas of interest.
Imagery
Imagery with
Canvas
Labels
17
-------
Example:
As an end userifyou spot a clusterof
highly vulnerable areas (red) to the
east of Oklahoma City (pictured to the
right), the select tool enablesyouto
draw a polygon around a specific
area to conduct a more in-depth
analysison a specificarea of interest.

Oklahoma C ty
19 &S SANTA
Q 6 Johnson
Manufacturing. 'nc
?000 S SUNNYLANE P O
BOX 95129
Ntoore
9200 S POLE ROAD
In the figure above, we see that we have selected 100 features from the "Wells within 1500 ft" layer,
which are now colored light blue. When we click on the wells within 1500 ft tab in the attribute table
we seethe number of selected features in the bottom left corner. To show only the selected records in
the attribute table, click the options drop down and then "Show selected records."
18
-------
;;; Options ~ [ Fiher by Map Extent | Q Zoom to (X] dear Selection 0 Refresh
@ Show Selected Records
ame
Address
City
Latitude
Longitude
Total Capacity in Gallons
Estimated Welts in 1,000
Estimated We
)H Show Related Records

ft Piume
ft Plume
T Filter

onoco #2706590
9000 S PENNSYLVANIA
Oklahoma City
35.3766
-97.5473
71,000.00
0.00
0.00
O Show/Hide Columns
Export Selected to CSV

Eleven #46
600 SW 4th St
Moore
35.3339
-97.4948
68,900.00
0.00
0.00

4
Mart 27
714NW27TH
Moore
35.3630
-97.5033
36,000.00
0.01
0.02

2
Coot Simpson
Construction Co.
5301 WEST INDIAN HILL
ROAD
Norman
35.2912
-97.5163
20,000.00
0.40
0.97

1
Masters Moving &
Storage Co.
3303 BROCE COURT
Norman
35.2538
-97.4900
8,000.00
0.00
0.00

4
Howards Service Center
729 NW 12TH
Moore
35.3488
-97.5010
20,150.00
0.00
0.00

2
Midco, Inc.
4230 S W 134TH
Moore
35.3333
-97.5954
13,500.00
0.30
0.60
The selected recordsthen appear in the attribute table as highlighted rows (see below). Each row
represents a single facility with eitherl or more USTs on site. You can now run simple statistics on
these facilities which re present our defined area of interest by clicking on the title of any column and
cl icki ng "statistics".
1,500 ft Tank Data
3p Extent | Q Zoom to (3 Clear Selection 0 Refresh
Name
Address
City
Latitude
Longitude
Total Capacity in Gallons
Estimated Welis
ft Plume
in 1.000 Estimated Wells
ft Plume
in 1,500 Estimated We

Kerr Mcgee #6152
OnCue #106
3300 SE 15THST
8917SE29THST
Del City
Midwest City
35 4497
35.4357
-97.4551
-97.3712
16,500.00
52,300.00
0.00
0.03
0.01
0.03
Sort Descending
Statistics

Kerr Mcgee #8488

Oklahoma City

0.26

All Day Marts LLC
5529 SE 15 STREET
Del City
35.4497
-97.4243
62,000.00
0.25
0.96
0.74

Total #4425
4301 SE 15THST
Del City
35.4498
.97.4414
54,000.00
0.00
0.00
0.10

Us Pollution Control inc
4725 SE 59TH
Oklahoma City
35.4062
-97.4296
10,000.00
0.01
0.02
2.87

Buddy's Produce
8716 SE 15th
Midwest City
35.4496
-97.3732
14,900.00
Q.OO
0.28
0.86

C St L Foods
4308 SE 29TH STR
DelCitv
35.4349
-97.4414
28.000.00
0.03
0.37
0.20

This wiil bring up the statistics window forthe specified
column (in this case total estimated wells within 1500 ft of a
facility, see below). This shows that the re are an estimated
4,019 domesticwellscloserthan 1500 ft to a UST facility
within our area of interest. There are facilities that have no
wells within 1500 ft, but on average (forthisarea)there are
an estimated forty wellswithin 1500 ft ofeach facility. The
maximum estimated wellswithin 1500 ft of a single facility
inthisarea is 135. Runningstatisticsagainonthecolumn
labeled "Total Capacity in Gallons"you seethatthisarea
has a total storage capacity of 1.74 million gal Ions with an
average tank capacity of 17,450 gallons.
Statistics
Fed : Estimated We
Buffer
Number of Value
Sum ofValues
Minimum
Maximum
Average
X
n 1,500 ft
100
4,019.36
0.02
135.04
40.19
Standard Dev aton 25.56
Totake it one step further, you can usethefilteroption
(also underthe optionstab) and create an expression to refinethe results so we only see active
facilities:
19
-------
O Add an expression O Add a set
Facility Status (Strinc * is " Active » O
Value Field • Unique
OK I Cancel
Click OK, and we see that only 44 of our previously selected lOOfacilitiesare still active. We can create
a second filterto refine itfurthertoonlythose sitesthat have been confirmed to have hada leaking
tank:
Fa ci 1 ity Statu s (Strl n c ~
1
is
Active

Value
Field
• Unique
LUST Facility (String' ~
i
IS
YES
C Value (-'Field '¦•'Unique
We now have 17 active facil ities that have been classified as LUST sites. Notethat when you use
the filter option in the attribute table, the map extent will be updated to only show those sitesthat
satisfy your filter expression. As you add filter expressions, tank sites that do not satisfy the expressions
are removed from the visible map. Likewise, the selected records in the attribute table will also be
removed so that the only viewable records are those that you originally selected that also satisfy the
filterexpressions.
20
-------
£ Oklahoma UST Vulnerability
Oklahoma UST Vulnerability
OZoofnto 8 O—l HlKMJ" C "•'¦Hit-.
i 0 Zoom to E) Ciw Mitaon O &•»•««>•
hunibw of Tankt
Numb«f o* T»nk» N»m«
Relational Tables
Information on specific facilities is available from the attribute table. With any number of rows selected
in eitherthe "UST Facilities" or "We I Is within 1500 ft" table, click options (top left of attribute table) and
then click "Show Related Records." You will notice the attribute table tab switch to "Tank Data" and the
records shown re present the individual tanks at each of the sites from your previous selection. You will
then be able to see how many active and inactive tanks each facility has and what their capacities are as
well as what type of fuel they contain "Gasoline / Diesel/E-85 etc...).
21
-------
Quickly using the map to find information
Simply clicking on the map leads quickly to
information from eitherthe "UST Facilities" layer
or the "Wells within 1500 ft" layer. This brings up
a pop-up window showing relevant information
on the selected facility. The image to the right
showsthisspecificfacility has3 Underground
Storage Tanks, is an active facility and is a LUST
site.
Clickon "Tank Data" underthe RelatedTables
heading. This brings up a list of the three tanks.
Clickthe symbol in the bottom right and then
select "View in attribute table." This shows the
information for each of the three tanks in the
attribute table. We find that the three tanks at
this gas station all hold gasoline. Two tanks have
a capacity of 12,000 gallonsandthe otherhas a
capacity of 10,000 gallons. All three tanks are
classified as currently in use and having leaked at
some point. This particulargas station has
roughly 30 domestic we I Is estimated to be within
1500 feet of the facility.
Single-Tank Analysis
Wherethere isa lack of publicly available information givingthe locations of all underground storage
tanks, the APP user may define a single tank location. The process involves three steps; (1) defining a
location, (2) settingthe potential impactdistance, and (3) calculatingthe estimated numberof wells
within the defined potential impact distance.
[CityArmyj
Area of Vulnerability (1,500 ft): Circle K
#2704000
Circle K
#2704000
Related Tables:
22
-------
Defining a Location
Defining a specific location is
accomplished by usingthe draw
tool with the 'point'draw mode
and clickinga pointon the map. It
is possible to select more than one
location at a time, however, the
time the analysis takes to run will
increase with each additional
location. It is recommended to
use one location at a time.

~
p
p
A Q>

A X
Click an analysis tool to execute
Create Plume Buffers

biB a i
v£>
Setting Potential Impact Distance
Estimate Wells in Plume Buffers
vl)
After having chosen a distance for delineation of the potential impact zone according to considerations
given in "AppendixA: Assessment of Plume Extents" or by agency policy (i.e., 1000 ft),the selected
distance isentered. With a location defined bythe selected point,the 'Create Plume Buffers"tool
(within the Analysis tool set) is used to create a potential vulnerability area forthe tank.
23
-------
By default, the selection for(l) "Choose layer
containing features to buffer" should be set to "Points"
which isa reference to the userdefined location inthe
previous step. (2) "Enter buffersize" is setto define the
radius of potential impact zone (3) Finally'Result layer
name' is used to designate the resulting layer (4) Select
'Run Analysis'.
:E $ Li
i * o> ::
! ©
Analysis

^ X

Create Plume Buffers
1 Choose layer containing features to buffer
Points
©
©
2 Enter buffer size
0
Distance
Enter buffer size
©
Field
1500 Feet
To create multiple
buffers, enter distances
separated by spaces (2 3
5).
Options
3 Result layer name
Potential Area of Vulnerability
©
©
Back I Run Analysis
24
-------
Calculate Estimated Number of Domestic Wells
Within the Potential Impact Distance
The 'Estimate Wells in Plume Buffers'tool allowsthe
userto calculate the estimated numberof wells within
the area that was created from the previous step
(Setting Potential Impact Distance). 'bufferLayer' will
bethe inputfor(l), regardless of whatyou namedthe
outputfromthe 'Create Plume Buffer" tool. The layer
to summarize is'Wells', (3), allowsthe userto
calculate statistics forthe defined area of vulnerability.
In thisexample we includetotal estimated domestic
wells(Est_Wells), estimated population, and
estimated housingunits. Leave (4) asthe default. Give
your output a unique name (5). Run the analysis.
¦—
A. i
|B Q,
¦¦
¦¦
©
•
Analysis

^ X
Estimate Wells in Plume Buffers
1 Choose an area layer to summarize other
features within its boundaries
bufferLayer
2 Choose a layer to summarize
Wells
3 Add statistics from the layer to summarize
.y Sum Area in Square Kilomet...
©
©
©
Est Wells
Sum
Population Sum
Housing_... - Sum
Field Statistic
4 Choose field to group by (optional)
Field
Add minority, majority
Add percentages
5 Result layer name
Summary Results
Curr«r,t mao «XT«n?
©
Back I Run Analysis
25
-------
The result will be a circulararea drawn aroundthe selected point. The name ofthe layerwill, by
default, appear as'resultLayer' and can be seen in the legend. Clicking on the map in the defined area
will display a pop-up box with the results. You may need to click the right-facing arrow in the top right
corner ofthe pop-up box to view the correct layer data. The pop-up box will then display the calculated
area ofthe potential impactarea,the estimated wells, housingunitsand population. Inthiscase, we
estimatethereto be 53 wells, 251 housingunitsand a population of 520.

X
(2 of 2)
resultLayer
0.656506029116601
0.656626474977382
52.6965535066006
520
251
resultLayer
Sum Est_Wells divided by Area in Square Kilometers
80.2681943-80.2681943
others
bufferLayer
b
Legend
26
-------
References
Ander, E.L., Watts, M.J., Smedley, P.L., Hamilton, E.M., Close, R.,Crabbe, H., Fletcher, T., Rime 11, A.,
Studden. M., Leornardi, G. 2016, Variability inthe chemistry of private drinkingwatersuppliesandthe
impact of domestic treatment systems on water quality, Environmental Geochemistry and Health,
38:1313-1332.
API. 1998. Characteristics of Dissolved Petroleum Hydrocarbon Plumes: Results from FourStudies.
American Petroleum Institute.
http://www.api.0rg/~/media/Files/EHS/Clean_Water/Bulletins/O8_Bull.pdf
Batu, V., van Genuchten, M.T. 1990. First-and Third-Type Boundary Conditions in Two-Dimensional
Solute Transport Modeling, Water Resources Research, 26(2), 339-350.
Con nor J. A., Kamath, R., Walker, K.L., and McHugh, T.E. 2015. Review of quantitative surveys of the
length and stability of MTBE, TBA, and benzene plumes in ground water at UST sites. Groundwater,
53(2), 195-206.
DeSimone, L.A., P.A., Hamilton, and R.J. Gilliom. 2009. The quality of our nation's waters—Quality of
waterfrom domestic wells in principalaquifersofthe United States, 1991-2004—Overview of major
findings. Circularl332. U.S. Geological Survey, 48 p.
Gelhar, L. W., C. Weltyand K. R. Rehfeldt, 1992. A critical reviewofdataon field-scaledispersion in
aquifers, Water Resources Research, 28(7): 1955-1974.
Haitjema, H.M., Mitchell-Burker, S. 2005. Are watertablesa subdued replica of the topography?,
Ground Water, 43(6), 781-786.
Murray, A. R., Weaver J. W., Kremer, F.V. 2017. Estimating Domestic Ground Water We 11 Use inthe
United States. For publication in Journal of the American Water Resources Association.
Ridpath, A., Taylor, E., Greenstreet, C., Martens, M., Wicke H., Martin, C. 2016, Description of
calls from private well owners to a national well water hotline, 2013, Science of the Total
Environment, 544, 601-605.
Schaider, L.A., Ackerman,J.M., Rude I, R.A. 2016, Septic systems as sources of organic wastewater
compounds in domestic drinking water we I Is in a shallow sand and gravel aquifer. Science of the Total
Environment, 547,470-481.
U.S. EPA. 1994. Handbook: Ground Water and Wellhead Protection, EPA/625/R-94/001. U.S.
Environmental Protection Agency.
27
-------
U.S. EPA. 2002. Drinking Waterfrom Household Wells, EPA816-K-02-003. Washington, D.C: United
States Environmental Protection Agency.
U.S. EPA. 2004. Understandingthe Safe DrinkingWaterAct. EPA816-F-04-030. Washington, DC: U.S.
Environmental Protection Agency, Office of Water.
Weaver,J.W. 2004. On-lineToolsforAssessingPetroleum Releases, EPA600/R-04/101, Research
Triangle Park, NC.
Weaver, J.W., Murray, A.R., Kremer, F.V. 2017. Estimation ofthe Proximity of Private DomesticWellsto
Underground Storage Tanks: Oklahoma Pilot Study. Accepted for publication in Science ofthe Total
Environment.
Xu, M., Eckstein, Y. 1995. Use of Weighted Least-Squares Methods in Evaluation ofthe Relationship
Between Dispersivity and Field Scale, Groundwater, DOI: 10.1111/j.l745-6584.1995.tb00035.x.
28
-------
Appendix A: Assessment of Plume Extents
Length of BTEX Plumes
Studies of the length of contaminant plumes indicate the expected extent of contamination from leaking
underground storagetanksites(API, 1998, Connoretal., 2015). Although based on limited data,these
studies indicatethatthe maximum observed extent of contaminant plumes isonthe orderof 1500 ft
(500m).
For confirmation of these distances, analytical models of ground watertransportwere constructed in
Java, and consisted of one-and two-dimensional analytical solutions (van Genuchten, 1981, Batu and
van Genuchten, 1990). To account for transport of benzene, a carcinogen, an initial concentration of 5.0
mg/L was selected as a baseline case. For comparison and exploration of the effect of source
concentration on plume length, simulations were also madeforsource concentration of 0.5,1.0, 25.0
mg/L. The plume extent wastaken asthe drinkingwatermaximumcontaminant level (MCL) of 0.005
mg/L. Thus the plume extentconsistsofallthe watercontaminatedatlevelsabovethe MCL. The initial
concentration, 5 mg/L, was selected as it is in the range of concentrations associated with fresh gasoline
(Weaver, 2004).
Several scenarios were simulated with parameters typical of leaking underground storage tank sites
(Table 1). Most important of these were varying conductivity of subsurface mate rials, dispersivity, and
biodegradation rates. The material types range from tight clays to highly conductive sands and gravel.
In each case the gradient was assumed to be 1/1000. The dispersivities were chosen to correspond to
the data tabulation of Gehlaretal. (1992) and the weighted regression to those data developed byXu
and Eckstein (1995). At a scale of about 3,000 ft (1000 m),the Xu and Eckstein (1995) regression result
isabout 10 m and higherestimatesof 100 m (one-tenththe plume length) were used (Figure 3). Note
that the data at this scale were judged to be of low re liability, and that conservative tracer experiments
were usedtogeneratethe results. The plume estimatesfromthe analytical solutionsare constrained by
the chosen end point of 0.005 mg/L. Actually these plumes, even considering biodegradation, would be
longeras concentrations exist belowthe threshold concentration of concern. The biodegradation rates
were determined from 365 and 730 day half-lives. The source width wastaken as 100 m forthe two-
dimensional model. The effects of conductivity, half-life and dispersivity on plume length are
summarized in the following sections.
29
-------
Table 1 Parameter values used in simulation.
Conductivity (m/d) Dispersivity(m) Half Life (d)
86.4
0.864
0.00864
Sand, Gravel
10
100
200
Silt, Silty Sand
10
100
200
Clay, Glacial Till
10
100
200
365
730
365
730
365
730
365
730
365
730
365
730
365
730
365
730
365
730
Gelhar, Welty and Rehfeldt [1992] Dispersivity Data
0
- 0
Rilliitllty:
— ™ ¦ Xj ird EcKitiln
* High
A I iTtwm td Im
~ Low
~ cr
~ ~
5 1 0
Scale [m]
Figure 3. Longitudinal dispersivity data from Gelharetal. (1992) plotted against "1/10" of length scale,
and the weighted regression formula of Xu and Eckstein, 1995. Most of the reliable data were at plume
scales of 10 to 100 m.
30
-------
Summary of Simulations
Conductivity
In one dimension, plume length increased with increasing conductivity and was proportional to the
order of magnitude of conductivity increase (Figure 4). The plume in the most conductive material (sand
and gravel) expanded forabout 2700 days and peaked at a distance of about 1100 ft (335 m).The plume
in least conductive soil (clay and glacial till) also expanded for 2700 days but peaked at distance at about
2 ft (0.6 m). Only small differences we re found between the one-and two-dimensional results,
presumably dueto biodegradation andthe width ofthe source.
10000
1000
% 100
Ol
4—
(U
D = 10 m2/day, X = 0.693/365 day"1

— —

—•
U
C
A3
4-"
(/)
5 10

0:-* 500 1000 1500 2000 2500 3000 3500 4000
0.1

Time (day)

—•— K = 0.00864 m/day ID - • K =0.00864 m/day 2D -
K = 0.864 rrt/day 2D —*—K = 86.4 m/day ID
K = 0.864 m/day ID
HN— K =86.4 m/day 2D
Figure 4. Plume lengths at different conductivities
31
-------
Dispersivity
The plume length also increased with the dispersivity (Figure 5). The plume with disperisivity of 10
m2/day extended todistance of 1160 ft (350 m) and plume with dispersitivityof 200 m2/day extended to
distance of 2,300 ft (700 m). At lower longitudinal dispersivity, small differences are evident between
the one-and two-dimensional results. In the two-dimensional model, mass transport occurs lateral to
flow, and less mass istransported longitudinally;thusthe 2D model result hasa lowerplume extent
than the ID result.
K = 86.4 rn/day, A = 0.693/365 day1
2500
0
0 500 1000 1500 2000 2500 3000 3500 4000
time {day)
— h#-— D = 10 m2/day ID » D = 10 m2/day 2D — D =200 m2/day ID —D = 200 m2/day 2D
Figure 5. Plume lengths at different dispersivities.
32
-------
Biological Degradation (Half-life)
Longerhalf-lifecorrespondsto lowerbiological degradation rate, sothe plume lengthsare greaterwhen
longer half-lives are selected (Figure 6). A plume extended upto 1200 ft(350 m) at half-life of 365 days
and the plume extended to 1600 ft (490 m) at half-life of 730 days, when all other parameters were the
same. Similar results were obtained in either one-ortwo-dimensional simulations.
K = 86.4 rn/day, D = 10 rrr/day
1800
1600
1400
1200
1000
ro 800
600
400
200
0
500 1000 1500 2000 2500
Time (day)
3000
3500
4000
half-life = 365 day II: ¦
alf-life = 730 day ID ¦
• hlaf-iife = 730 day 2D
¦ half-life = 730 day 2D
Figure 6. The plume length at different biological degradation rate.
33
-------
Effectof Source Concentration on Plume Length
Althoughthe plume length dependson source concentration,the dependence is mild. Forsimulations
with source concentrations of 0.5 to 25 mg/L, the plume length increased by a factor of 2.6. Notably the
increase in source concentration was a factor of 50, indicatingthat changes in source concentration are
dampened bythe model whenthe resultsare expressed as plume length.
K = 86.4 m/day, A = 0.693/365 day"1, D = 200 m2/day
2500
2000
"S 1500
OJ
>¦*—
OJ
U
c
ra
^ 1000
500
0
0 1000 2000 3000 4000 5000 6000 7000 S000
Time (day)
—•^C0=0.5 mg/l —•—CG=lmg/l —#^C0=5mg/l —#-€0=25 mi/I
Figure 7. The effectofchangingsource concentration (Co) on plume length.

t
#
~
~
#
—#

0-* ' *

Retraction of Plumes
Because of biodegradation and dispersion, the maximum extent of plumes may retract. This is
demonstrated both through mode ling and through the empirical data analysis of Connoretal. (2015).
Simulations were conducted with an exponentially-decayingflux source (solution C14of van Genuchten,
34
-------
1981) used rates of source depletion, expressed as half-lives of lyear, 2 year, 3 yearand 5 years2. As
before, a gradient of 1/1000 was assumed, and a dispersivity of 10m2/day was used. The biodegradation
half-life of l.lyr was usedforthe contaminant in ground water. The plumes contracted soonest for
lowest source half-lives, which are the least persistent sources (Figure 8to Figure 10).
K = 0.00864 m/day
2.5
0
0 1000 2000 3000 4000 5000 6000 7000 8000
Time (days)
9 Source Half-life = 2 year > Source half-life = 3 year > Source half-life = 5 yrs
Figure 8. Extent of contaminant plume asdefined by 0.005 mg/L concentration showing plume
retraction in low conductivity aquifer material. (Solution C14ofvan Genuchten (1981), dispersivity of 10
m, and biodegradation half-life of l.lyear).
2 Simulations with low source half-life values (lyear) and low conductivities (0.00864 m/day, 0.864
m/day) did not converge to a solution, presumably indicating that no contaminant plume forms under
these conditions.
35
-------
K = 0,864 m/day
45
5
0
0 1000 2000 3000 4000 5000 6000 7000 8000
Time (days)
—•—Source Half-life = 2 year —•—Source half-life = 3 year —•—Source half-life = 5 yrs
Figure 9. Extent of contaminant plume as defined by 0.005 mg/L concentration showing plume
retraction in moderately conductive aquifermaterial. (Solution C14of van Genuchten (1981),
dispersivity of 10 m, and biodegradation half-life of l.lyear).
36
-------
K = 86.4 m/day
1400
1200
1000
S 800
4—
cu
c
ft}
600
b
400
200
1000
2000
3000
4000
Time (days)
5000
6000
7000
8000
Source Half-life = 1 year
Source half-life = 3 yrs
Source half-life = 2 yrs
Source half-life = 5 yrs
Figure 10. Extent of contaminant plume as defined by 0.005 mg/L concentration showing plume
retraction in highly conductive aquifer material. (Solution C14of van Genuchten (1981) , dispersivity of
10 m, and biodegradation half-lifeof l.lyear).
37
-------
Result Summary
As shown above, the model solutions were highly dependent on the combination of parameters. The
plume length extended up to 3,810 ft (1160 m) underthe most favorable conditions (most conductive
soil, dispersivity of 200 m2/day, and half-lifeof 730 days). Conversely, plume lengths of only a few feet
were observed in case of the least conductive soils.
The study by Connoret al. (2015) compiled data from over 1,300 plumes of Benzene, 500 plumes of
methyl tert-butyl ether (MTBE) and 108 Plumes of tert-butyI Alcohol (TBA). They concluded that most of
the Benzene and MTBE plume lengths stabilized between 425 and 530 feet at concentration of 5 ng/L.
TheTBA plumeswere comparable in lengthtothe benzene and MTBE plumes. They observed that
plumes with lengths in excess of 1000 feet were extremely rare. The authors acknowledge, however,
that at many leakingunderground storage tanksitesground water plumes are incompletely
characterized because ofvarious limitationson sampling.
The results of the mode ling study performed forth is paper show that longer plumes are theoretically
possible ifthe hydraulic parametersofthe porous media are favorable, and biodegradation ratesare
low. However, these are not supported byfieldevidence,andthe plume length is highly dependenton
the choice of parameters. Dispersion, in particular, has orders of magnitude variability for conservative
tracers, and was picked arbitrarily forthe mode ling study. He nee, for planning purposes a maximum
plume length of 1000 to 1500 ft has been found reasonable by several states as itencompassesthe
known distribution of plume lengths.
38
-------
Appendix B: Implementation Details
Summary
The tool box forthe APP consists of three separate tools which will allow an end use rto input a point
layerrepresentingthe locations of LUST sitesand a flowdirection raster(easily created in mostGIS
software packages). Thefollowingdescribesthe varioustools.
Thisset oftoolstakesthe locations of LUST sitesand determinesthe generalized ground waterflow
direction awayfromthetank site based on a flowdirection raster, which iscreated separately priorto
usingthistoolset. Thisapproach is based on the premisethatground waterflow mimicsthe surface
topography. Because of a numberof factors, the actual ground waterflowdirection may differfrom
that predicted from surface topography. These factors (primarily pumping from we I Is and subsurface
heterogeneity, impervious surface cover re directing runoff) require site-specific analysis and
investigation.Thus,the resultsfromthetransporttool are presented asa firstapproximation which
requiresfurthersite-specificdata collection, which is likely to require field sampling.
In addition tothe physical effects,the outcomes of thistoolset are heavily dependentonthe input flow
direction raster. Generally, flowdirection rasters are created by defining the direction of steepest
dropoff from an origin cell tothe neighboring8cells (termed "queen connectivity"). The simplestform
of this iscalled an 8 directional (D8) flow rasterwhich will always return one of eight possibledirectional
values (increments of 45 degrees). This is the type of flow direction rasterthat iscreated by using the
flow direction tool in ArcGIS. The other, higher resolution option iscalled an infinite direction (D-
lnfinity)flow raster. A D-infinity flowdirection raster will determine the steepest dropoff from the
origin cell by determining the neighboring pair of eel Is that re present the steepest slope and then
calculating an angle between those two cells based on their proportional difference. One such example
of a D-infinity flow direction tool is freely and publicly available from the hydrology lab of David
Tarboton at Utah State University (http://hydrology.usu.edu/taudem/taudem5/index.html), and is
described as:
"Assigns a flow direction based on the D-infinity flow method using the steepest slope of a
triangular facet (Tarboton, 1997, "A New Method forthe Determination of Flow Directions and
Contributing Areas in Grid Digital Elevation Models," Water Resources Research, 33(2): 309-
319). Flow direction is defined as steepest downward slope on planar triangular facets on a block
centered grid. Flow direction is encoded as an angle in radians counter-clockwise from east as a
continuous (floating point) quantity between 0 and 2 pi. The flow direction angle is determined
as the direction of the steepest downward slope on the eight triangular facets formed in a 3 x 3
grid cell window centered on the grid cell of interest. The resulting flow in a grid is then usually
interpreted as being proportioned between the two neighboring cells that define the triangular
facet with the steepest downward slope."
39
-------
The LUST Plume Locations toolbox has three separate steps. The two important steps are
"Assign Plume Variables" which takes point locations and creates the necessary data that will be
used to create plumes based on desired plume distance and angular flow variables, and "Points
to Plumes" which createsthe actual plume areas based on the input variables from the previous
step. These tools are provided separately to increase efficiency and decrease computational
time. The third tool, "ESRI Flow Direction to Radians", converts an esri flow direction raster into
radians from east which will enable the "Assign Plume Variables" tool to accept it as an input.
Tools
El % JKl
A <¦*r*iam,mis Oil iPI til I*?1
gN*** rlUrnc i 9rfcHlNe&
f.*p[ rVi, f, ,« ,, |(
Tennis to Pliurwes
1. ESRI Flow Directionto Radians:
This tool is only necessary if the flow direction raster you intend to use is the output
fromthe "Flow Direction"tool in ArcMAP. It is encouraged that a D-infinityflow
direction raster is used in orderto maximize accuracy. As noted above, a tool that is
ready to use in ArcGISis available from Tau DEM at Utah State University. QGIS, a well-
known open source GIS software package alsooffersa D-infinityflowdirectiontool.
Input ESRI Flow Direction Raster
Output Folder
£3
; ESRI Flow Direction to
Radians
I Converts the output of an ESRI
;i flow direction raster into Radians
i i fro m east}
2. Assign Plume Variables:
Thistool preparesthe point layerto be used inthe 'Pointsto Plumes'tool.
40
-------
jj*3 Assign Plume Variables
~ X
* Flow Direction Raster (Radians)
* UST Point Layer
* Output Folder
Degrees from Flow
30
Linear Distance
500
X Field
LONGITUDE v
Y Field
| LATITUDE vl
The two files needed as inputs are the flow direction raster (in radians from east) and a point
layerrepresentingthe locationsof underground storage tanks. The point layerfortanks must
also have a unique id fie Id, as well as XY coordinate fie Ids, which must be labeled in the
appropriate boxes. The maximum distance and angle of flowforthe plume must be
determined. An annotated graphic of the tool is available on the following page. The steps this
tool cyclesthrough are as follows (A-E):
A. Extract the values from the flow direction rastertothe points re presenting the locationsof
the leakingunderground storage tanks. Every point location nowhasa direction (in
radians) of steepest path.
B. Add fieldstothe point layerrepresentingflowdirection (in degrees), left bounded angle(in
degrees),and right bounded angle(in degrees). These fields wi II be populated inthe next
steps.
C. Calculate the angle of steepest flow (in degrees) by converting radians to degrees, this will
populate one of the fields just created.
D. Calculate the remainingtwoemptyfields(leftand right bounded angles). Thisstepisa bit
tricky since, if you are calculating degrees difference from the steepest path, you could end
up with degree values <0or >360. Therefore, this step calculatesthe left and right
bounding angles by adding or subtracting the plume angle in put from the value of the
direction of steepest path. Itthen isolatesany values lessthan Oorgreaterthan 360 and
corrects them. Ex:-15 becomes345, and 380 becomes20. Atthis point, each tank point
now has fields representingthe angle of steepest flow, andthe direction ofthe leftand
right bounding lines representingthe specified angle ofthe plume.
Assign Plume Variables
This tool prepares a point layer representingthe source
of leaking tanks to be processed by tbe Points to
Plumes' tool
41
-------
E. The final step isto write the output pointfile sothat it may be input intothe 'pointsto
plumes'tool andto write an emptyfeature classthat will holdthe resultsofthe plume to
pointstool. When the pointsto plumestool runs, it will iterate through each tanksite
separately and append each result to this empty dataset.
Pointsto Plumes
The pointsto plumestooltakesthe prepared point layerand iterates through itto build
polygonsthat representthe estimated maximum plume areaforleakingtanks based on
given directions and angles. The steps this tool uses are described be low (A to F.)
A. Iterate through feature selection. Thistool runsan iteration foreach individualtank
site. This is necessary to avoid the clipping of overlapping polygons when the plume
area isdefined foreach site.
B. The 'Bearingdistance to line'tool is called to create defined boundsforthe plume. The
tool callsthe max distance, and angle fields from each feature which were calculated in
the previoustool. Usingtrigonometry,the
length ofthe sidesofthetriangle are
calculatedto run thistool. The outputsare
two separate (leftand right) boundinglines.
The output can beseentothe right, where
purple representsthe left bound and blue is
the right bound.
C. Once the bounding lines are in place, the
start and end points ofthe bounding lines
are converted into separate points, which
results in a separate pointfeature withthree
points (the origin and two end points)
D. The vertices are then converted into a polygon which representsthe initial plume area,
but because ofthetriangle shape ofthe plume, there are nowareasthatare withinthe
plume area but fartherthanthe defined maximumdistance.
When the bearing distance to line function was called, an in put of
500 meters maximumdistance ata 30 degree angle would
actually yield atriangle side of 577 meters.
E. The initial plume area is clipped based on a separate bufferthat
was created based onthe inputvaluesof maximum plume
distance. This will yield the cone shape so that all of the polygon
area is within the designated area (i.e. 500 meters).
42
-------
Sslif
O
F. Finally,aspatialjoin isrunto reattach the original site attributes
to the newly created polygon, allowing the end userto maintain site
specific information (owner/LUSTstatus/licensing/inspection etc...)
43
-------
SEPA
United States
Environmental Protection
Agency
Office of Research and Development (8101R)
1200 Pennsylvania Ave. NW
Washington, DC 20460
epa.gov/research
EPA/600/R-17/282
September 2017
-------