&EPA
United States
Environmental Protection
Agency
Office of Water
(4802)
EPA-813-B-95-007
September 1995
Applicability of Wellhead
Protection Area Delineation
To Domestic Wells
A Case Study
-------
APPLICABILITY OF
WELLHEAD PROTECTION AREA DELINEATION
TO DOMESTIC WELLS:
A CASE STUDY
Prepared by
Marilyn Ginsberg
Ground Water Protection Division
Office of Ground Water and Drinking Water
U.S. Environmental Protection Agency
Washington, DC 20460
September 1995
-------
Contents
Introduction .
Protection of Private Wells .
1
1
13
Case Study: Tillery Area, North Carolina
1
Background
1
Hydrogeology
3
Welihead Protection Program Development
3
Delineation Approach
3
Method Selection
3
Parameter Values
5
Criterion and Thresholds
5
Water Discharge Rate
5
Domestic Wells
5
Non-domestic Wells
7
Model Runs
7
Confmed Aquifers
7
Surficial Aquifer
7
Results
8
Surficial Aquifer
8
Confmed Aquifers
10
Conclusions
Recommendations
10
12
References Cited
-------
List of Tables and Figures
Table 1. Hydrogeologic Units in the Study Area 4
Table 2. Long-Axis Lengths of Calculated Wellhead Protection Areas for Wells
in the Surficial Aquifer 8
Figure 1. Map Showing Water-table Elevation, Tillery area, North Carolina 2
Figure 2. Map Showing Subarea Boundaries, Tillery area, North Carolina 6
Figure 3. Calculated Wellhead Protection Areas 9
-------
APPLICABILITY OF WELLIIEAI) PROTECTION
AREA DELINEATION TO DOMESTIC WELLS:
A CASE STUDY
Introduction
The 1986 Amendments to the Safe Drinking Water Act (Act) established the Wellhead
Protection (WHP) Program to protect ground water that supplies drinking water to public water-supply
(PWS) wells and weilfields. Welihead protection for a community supplied by numerous private wells
requires a different approach than that for wellhead protection of PWS wells. The higher density of
private wells within a community may cause wellhead protection areas to overlap where hydrogeology
is not sufficiently known and long ground-water travel times are needed to meet protection goals.
Protection of Private Wells
Before developing a private-wellhead protection plan, a community should consider its
protection goals and determine if a wellhead protection program will help meet them. Several
communities have already expressed interest in developing a program. Tillery, NC is one such
community. The Wellhead Protection Area (WHPA) delineation approach developed for that
community and lessons learned from that activity are described below.
Case Study: Tifiery Area, North Carolina
Background
The greater Tillery, North Carolina, study area is an approximately 130 square mile area in
north-central Halifax County (Figure 1, insert). The area is rural; the population is approximately
3,000.
Over 95 percent of the communitys water supplies are obtained from about 1,700 small-
capacity domestic wells, one PWS school well and seven wells supplying stock and irrigation water.
Community residents, having become increasingly concerned about possible contamination of drinking-
water wells, decided to develop and implement a community-based, private-wellhead protection
program. The Environmental Protection Agency (EPA), interested in assessing the advisability of such
a protection program, offered technical assistance in the delineation portion of the program. All
management decisions would be made by the community.
-------
-.
I i t S S
N J - gs i -
,,--- - Ta I VN I I $ -
I J /
: : \ _ . J - -___) , I -
I % _k / - - v--, (
\ \\ - N . : _ - - I
I / S.
, S
1 S S 7 - - -
- - , 2 \
/ ,. , , , P
. , , ( / -_ 15
I i_i ,;; k: \ \ \ ? f
I , s. ,, &. , / - S
, II ? I --. . . ... -. . - 10
J 1 .- .j i / 7 c
,1-_ 30 - _; i .. _ - - j_ I :: - --
1 , i -
, iT ,/ . . \ -
,,,/_
-. :
; ____% -...- Dawson1 . % S I I
qL_ L_, Z 0 0 J S -_ . S 30 , , - 1.
\ _. ,) S. S
- .3. S . S .. . - I I I
__Y\. I
1 \ .. --.- rp S. ,_ / \
uP 5 25- -S \ S 7 S
S S S S - S
S - . , , \ 5.)
5, S j - , __i S S
-. z - 1 Scotiar
I. - Neck . S
S / __. -_____%5__ S ,
WATER-TABLE Water-Table Contour in
MAP Meters Above Sea Level
Study Area Location Study Area Boundary
Tiflery Area, North Carolina
Case Study - -- Stream
Raleih Swamp
N Village
A _________ ( ) HogFarm
2 Km. ( j Prison Farm
Figure 1. Map Showing Water-table Elevation, Tillery Area, North Carolina
2
-------
Hydrogeology
Only one useful well log is available for the study area. However, based on information for
the surrounding area, the following generalized sequence of hydrogeologic units (Winner and Coble,
1989, and Giese et al., 1991) is described (Table 1).
Wellhead Protection Program Development
All known domestic wells are of very-low yield and withdraw from the surficial aquifer.
Older formations provide small to moderate supplies to the PWS well and to agricultural wells in the
study area.
Developing the delineation component of a community-based, private-welihead protection
program requires a somewhat different approach than that for a State-level program focusing on
protection of PWS wells. The differences between these types of programs are related to the
feasibility of accurately describing the hydrogeology at each well site, and to the level of ability of the
implementors (typically community residents) to transfer calculated WHPA boundaries to each well.
Community needs and availability of funds for technical expertise are critical issues to
consider during development of the WHPA-delineation approach. The Tillery-area community is an
economically disadvantaged one lacking funds for technical assistance or collection of additional data.
Tillery residents also lack the technical training to measure water levels or accurately locate wells on
topographic maps.
The communitys WHP goal is to protect wells until public water supplies can be purchased
from outside the County in approximately 15-20 years. The focus of the protection is to help
residents manage their property to control potential sources of contamination to their drinking-water
wells.
Because community members will be implementing the program, the issue arose regarding the
number of hydrogeologically discrete subareas into which the study area should be divided. (More
subareas allow greater accuracy in WHPA calculation, but make transfer of calculated boundaries to
individual wells more prone to error by the implementor.)
Delineation Approach
Method Selection
After consideration of data availability, costs and desired level of accuracy, EPA chose to use
the WHPA 2.2 GPTRAC computer model to calculate WHPA boundaries. The code is a steady-state,
two-dimensional, user-friendly, semi-analytical flow model developed for EPA that calculates,
contours and displays ground-water time-of-travel isochrons around a pumping well (Blanford and
Huyakom, 1993).
3
-------
Table 1. Hydrogeologic Units in the Study Area.
Unit I
Description J
Comments
Surficial aquifer (water-table
aquifer)
beds of sand, silty sand, and
clay, possibly with isolated sand
and gravels;
yields small supplies;
source for all known
domestic wells in the study
area
Yorktown confining unit (may
include some post-Yorktown
clays) 2
clay and sandy clay with some
local beds of fine sand or shell 3
Yorktown Aquifer
composed largely of fine sand,
silty and clayey sand, and clay
and is characterized by shells
and shell beds throughout 4
may yield small to
moderate supplies; no
known wells are supplied
by this unit in the study
area
Upper Cape Fear confining
unit
nearly continuous clay, silty
clay and sandy clay beds 5
Upper Cape Fear Aquifer
layers of sand, some silty or
clayey, alternating with clay
layers
may yield moderate to large
supplies; supplies two
known wells in the study
area
Lower Cape Fear confining
unit
clay and sandy clay beds 6
Lower Cape Fear Aquifer
layers of sand, some silty or
clayey, alternating with clay
layers
may yield moderate to large
supplies; supplies six
known wells in the study
area
Crystalline bedrock
varies among metavolcanics
schists, gneisses, granites
no known wells developed
in these rock types in the
study area
after Winner and Coble, 1989.
2 GieseetaL, 1991.
Winner and Coble, 1989.
Ibid.
Ibid.
6 Ibid.
4
-------
l arameter Values
The WHPA 2.2 model requires values for several parameters: recharge to the aquifer,
transmissivity of the aquifer (and confining layer, if any), porosity, hydraulic gradient, aquifer (and
confining layer, if any) thickness, and well-discharge rate. If these variables are not sufficiently
constant across an area, division into subareas is required.
Parameter values for this study were obtained from the literature; no additional data were
collected. Because water-level data from wells were lacking, surface-water features on topographic
maps (U. S. Geological Survey, 1985, 1974a, 1974b, 1962, 1961 and 1960) were used to develop a
water-table map (Figure 1) from which hydraulic gradients could be obtained. Porosity of all aquifers
was estimated to be 0.2 (20 percent). U. S. Geological Survey reports (Giese et al., 1991; Winner
and Coble, 1989) supporting the Coastal Plain Aquifer-System Analysis provided the information
from which all other aquifer-parameter values were obtained or estimated for use in the WHPA 2.2
model.
The absence of reliable, site-specific geologic information for the study area, made moot the
issue of subdividing the area based on hydrogeologic variability. However, the surficial aquifer was
divided into six subareas based on hydraulic gradient (Figure 2).
Criterion and Thresholds
EPA selected the ground-water time of travel (TOT) criterion and the 2-, 5-, 10- and 20-year
thresholds. The community would choose the level-of-protection threshold that best met community
needs.
Water Discharge Rate
EPA assumed that all wells pump continuously year round at a constant rate, rather than
periodically. The rate assumed for each well was that rate that would provide the total yearly well
discharge estimated for the well. This approach was necessitated by the steady-state nature of the
WHPA 2.2 model, the lack of availability of pumping schedules and the large number of wells in the
study area. Where more than one well withdrawing from the same formation serves the same facility,
the total demand at the facility was considered to be met by one well only. Total daily discharges
were determined as described below.
Domestic Wells
The community was unable to provide an estimate of daily well yield. The number of
residents per home ranges from one to eight, more homes having four residents than any other
number (Grant, 1995). EPA considered four to be a high estimate; however, because a high estimate
yields conservative (that is, protective) WHPAs, four residents per well was assumed for WFIPA
calculations. National water-use statistics show that average, domestic, per capita use in the
southeastern U.S. is 100 gallons per day (gpd) (Solley et al., 1993). EPA chose 400 gpdlwell as an
estimate of local household demand.
5
-------
I
4
21
sE
J 1 ____
S
fS.4lp
HYDRAULIC-GRADIENT
SUBAREAS MAP
Study Area Location
Tillery Area, North Carolina ___________
Case Study , _/
N
0 2Km.
Figure 2. Map Showing Subarea Boundaries, Tifiery Area, North Carolina
6
A
0 2MiIes
)
s .
Ne
Hydraulic-Gradient
Subarea Boundary
Study Area Boundary
Stream
Swamp
R Village
-
I
I
I
-I
1
I
I
E j - -
J
1
I
/,1
I
I
A
p
---- -- - - E
I
I!
c
A
/
)
E
B
I
I
E
I
I
S ) I
I
-S
S..
-------
Non-domestic Wells
Two agricultural wells are located at the prison farm and are screened in the Lower Cape
Fear Aquifer. EPA estimated from the purposes of the wells that the equivalent continuous discharge
for the facility was 5.1 gallons per minute (gpm). The PWS well is a school well that withdraws
from the Lower Cape Fear Aquifer. A school officials estimate of water use was 3,600 gpd
(2.5 gpm).
Two additional wells screened in the Lower Cape Fear Aquifer supply two hog farms for
agricultural and human-consumption purposes. Demand at each hog farm was estimated to be
86.8 gpm.
A third hog farm is supplied by three wells withdrawing from the Upper Cape Fear Aquifer.
These wells are believed to serve agricultural and human consumption needs. The total average
facility demand was estimated to be 14.6 gpm.
Model Rims
Confined Aquifers
The WHPA 2.2 model provides the user with confined-, semi-confined-, and unconflned-
aquifer options. In the study area, all aquifers below the water-table aquifer are confined to some
extent. To test the tightness of the confining layers, two 20-year scenarios were each run for the
Upper Cape Fear Aquifer and for the Lower Cape Fear Aquifer; ground-water flow in each aquifer
was modeled individually. In each scenario, the well representing the greatest facility discharge in the
formation was modeled. That is, for each scenario, the discharge of the Lower Cape Fear well was
86.8 gpm; the discharge for the Upper Cape Fear well was 14.6 gpm.
In the first scenario, the vertical hydraulic conductivity estimated from Giese et a!., 1991, and
confining-unit thickness estimated from Winner and Coble, 1989, were input. In the second scenario,
for purposes of comparison, the aquifer was modeled as totally confined.
For each formation, both scenarios produced virtually identical WHPAs. That is, for a well
discharge equal to, or less than, that input to the model, 1) hydraulic interference with a well in a
different aquifer will not occur, and 2) the tightness of the overlying unit allows the aquifer to be
treated as completely confined.
That these aquifers may be treated as highly confined has importance in the WHP Program,
because confining layers offer some degree of protection for a well. The Tillery community might
choose to use a smaller-than-calculated WHPA for the seven private wells discharging from the
Lower Cape Fear and the Upper Cape Fear aquifers.
Surficial Aquifer
The largest value of the water-table gradient was selected in each subarea, because this
condition results in the longest (most protective) contaminant-flow distances. In each subarea, a
7
-------
model scenario was run to test for well interference among wells located 500 feet apart (the most
closely spaced conditions in the aquifer in the study area). This scenario showed that no well
interference occurred unless wells were aligned directly downgrailient from each other. In that case,
there was no interference parallel to the prepumping direction of ground-water flow and only slight
interference perpendicular to that direction. For this reason, well interference was considered
negligible and model scenarios for each subarea were run for 2, 5, 10 and 20 years (Figure 3) for one
well only.
Results
Because the axis of a WHPA is oriented parallel to ground-water flow (assuming aquifer
isotropy and homogeneity), errors in depicting the position or curvature of water-table contours result
in misorientation of the WHPA. Such errors are particularly significant in the case of narrow
WHPAs, resulting from wells with small discharge relative to the regional ground-water flow, like
those WHPAs calculated for wells in the surficial aquifer in the study area. In these cases, a minor
error may cause the calculated WHPA to lie outside the true area of ground-water contribution.
Compounding these errors are the errors associated with failure of the aquifer to meet the assumptions
of isotropy and homogeneity.
Surficial Aquifer
The unlikelihood that the calculated narrow WHPA boundaries could be precisely oriented,
led EPA to provide the community with circular WHPAs for wells in the surficial aquifer. These
WHPAs were constructed with radii equal in length to the long axes of the calculated WHPAs, as
measured from the well to the upgradient boundary (see Table 2).
Table 2. Long-Axis Lengths of Calculated Wellhead Protection Areas for Wells
in the Surficial Aquifer.
Subarea
2-year WHFA
(in fed)
5-year WHPA
(in feet)
10-year WIIPA
cm feet)
20-year WIIPA
(in feet)
A
200
450
950
1,800
B
200
500
1,000
1,950
C
350
800
1,700
3,300
D
450
1,000
2,000
3,950
E
550
1,400
2,750
5,600
F
1,050
2,600
5,250
10,500
8
-------
Figure 3. Calculated Wellhead Protection Areas
Surf icial Aquifer Subarea A
1100
0
1100
0
Surf icial Aquifer Subarea B
- 1100
0
1100
0
1100
20 years
lOyears
S years
2years
1100
0
1100
0
1100
0 -
1100
0
1100
20 years
l oyears-
5years
2years
>00
-9000 000 -3000
-12000
1100
0
1100
0
1100
0
1100
0
1100
-1200
1100-
0
I
-3000
0
I
30
C
20 years
10 years
Syears
2years
I
-3000
o
3000
)0
>0 -12
1100
0
1100
0
1100
0
1100
0
1100
Surf icial Aquifer Subarea E
0-
1100-
0-
- 1100
0
1100
1100
Surf icial Aquifer Subarea 0
20 years
loyears
5years
2years
-9000 -6000
-3000
0
300
Surf icial Aquifer Subarea F
20 years
10 years
5years
-
2years
-
0
C
20 years
10 years
5years
2years
-12000
1100
0
1100
0
1100
0
1100
0
1100
>0 -1>
-12000 -9000 -6000 -3000 o - 30
5500-
Upper Cape Fear Aquifer
3300
1100
0 c:::: :::) 2oyearS
-1100-
-3300-
I I I
-9000 -6000 -3000 0 3000
fee?
>0
00 -9000 -6000 -3000 0 30
Lower Cape Fear Aquifer
3300
1100
-3300
-5500
-12000
-12000
- I I
-9000 -6000 -3000 0
feet
9
-------
EPA considered the error associated with overly large WHPAs to be far less of a problem
than the likely error that misoriented WHPAs would lie outside the true area of ground-water
contribution to the well. That is, EPA considered the error of over protection to be less of a problem
than the failure to protect.
EPA suggested to the community that where WHPAs overlapped, for the sake of simplicity,
the protection area should be drawn along the common boundary. That is, well interference should
be considered negligible. Where a WHPA included part of more than one subarea, the community
would decide whether or not to modify the WHPA in some way to reflect the effect of the adjacent
subarea(s) on WHPA size.
Twenty-year circular boundaries were not provided to the community, because 20-year
WHPAs would provide protection similar to the 10-year boundaries; that is, 10- and 20- year
boundaries would encompass virtually the entire study area. Each set of 2-, 5-, and 10-year
boundaries was coded by letter to the appropriate subarea and was provided on transparent film at the
1:24,000 scale.
Confined Aquifers
The potentiometric surfaces (Giese et al., 1991) of the confined aquifers are smoother than the
water table in the surficial aquifer, because the potentiometric surfaces are not surface-water
controlled. As a result, WIIPAs are somewhat easier to orient. EPA provided the Tillery community
with calculated 20-year WHPAs for the maximum confined-aquifer discharges noted above and
considered those boundaries to be quite protective. As stated above, the community may wish to
designate smaller WHPAs.
North Carolinas Wellhead Protection Program suggests a sethack of approximately 300 feet
for a PWS well pumping 2.5 gpm. Although non-community transient wells are not included in the
State Program, EPA suggested that the community consider using the 300 foot setback for the
Dawson Crossroad school well. (The community might also want to consider using the 1000-foot
sethack necessary to enter the North Carolina Monitoring Waiver Program.)
Conclusions
The Safe Drinking Water Act, through the WHP Program provides a mechanism for
differential protection of ground water that supplies PWS wells and wellfields. These wells and
wellfields are, in most cases, sufficiently separated that their WHPAs are distinct and differentially
protectable. However, application of the WHP Program to areas supplied by numerous individual
wells is problematic, because the goal of isolating clearly defined areas for priority protection may not
be achievable.
The Tillery community hopes to define WHPAs with large enough TOTs to allow natural
remediation of contaminants until an alternative water supply becomes available in 15-20 years.
Residents hope to preferentially protect and manage contaminant sources within WHPAs and, where
10
-------
possible, site such sources external to the WHPAs. Because WHPAs are so large (virtually the entire
study area is incorporated into 10- and 20- year circular WHPAs and much of subareas C, D, E and
F is incorporated into 5-year circular WHPAs), larger TOTs cause individual wells to become part of
one large welifield. That is, by moving sources of contamination away from one well, they are
frequently moved closer to another in a common protection area. When multiple wells constitute one
large welifield, the entire aquifer within the weilfields WHPA must be managed. Thus, for large
areas with many wells, WHP becomes synonymous with aquifer protection.
The community has three management options. The first two options require that the
community be able to approximately locate their wells on topographic maps and determine the very
general direction of ground-water flow. Two types of uncertainty are present in these first two
options, the uncertainty associated with well location and the hydrogeologic uncertainty that causes
imprecision in WHPA calculation and in the magnitude and direction of the water-table gradient.
The first option is that residents identify the general direction of ground-water flow where
they live and, wherever possible, site contaminant sources downgradient from their wells. Sources
can be located fairly close (about 25 feet) to a well; as can be seen from Figure 3, the WHPA
boundary in the downgradient direction is near the well. Although these sources will be upgradient
from other wells, TOT to neighboring wells can be maximized in this fashion. The ground-water
TOT between a well and its downgradient neighbor can be estimated by comparing the distance
between the wells to the lengths of WHPA radii.
The second option is that quarter-circle WHPAs be used. These WIIPAs would provide a
buffer on the sides of the narrow, calculated WHPAs. If the community uses quarter-circle
WHPAs with radii equal to those of circular WHPAs, much of the study area would still be
incorporated into the 10-year WHPAs, but 5-year WHPAs would be differentially protectable in most
of the study area (that is, in subareas A, C, D and much of E, and in B except near Dawson
Crossroads where well density is high). However, with this option, some areas of actual ground-
water contribution to wells likely would fail to be protected. Most of these failures would occur in
areas where the direction of the hydraulic gradient radically changes, and also near the top of hills
where zones of contribution are, because the hydraulic gradient is 0, circular.
The third option is that the community selects the 2-year TOT (differentially protectable
WHPAs in all subareas except F) and vigorously manages those WHPAs. Because 2 years is so short
a period, management of WHPAs should be coupled with contaminant-reduction measures across the
entire study area. Although this approach includes WHP techniques (e. g., defining TOT and
direction of travel in order to delineate the WHPA), the approach has a large component of aquifer
protection. (Regardless of the option chosen, a high-priority activity should be to reduce the
likelihood of drinking-water contamination resulting from the poor well construction prevalent in the
area. Although replacement of wells is not economically realistic, less costly improvements should be
considered, such as placing concrete pads, andlor berming the soil, around wells.)
The larger the TOT selected, the more individual well owners will need to rely on each other
in order to manage contaminant sources that have the potential to affect their own and each others
wells. Cooperative source management by neighbors may also be needed where smaller WHPAs
intersect property boundaries. In this, as in any other aquifer-protection activity, cooperation among
neighbors is a key to protection.
11
-------
Recommendations
A WHP Program will not provide distinct management areas for numerous low-capacity wells
where the hydrogeology is not sufficiently known and long travel times are needed to meet protection
goals. Before embarking on a costly WHP effort, a community should consider its management goal,
hydrogeologic setting and population distribution.
The location of private wells close together causes their WHPAs to overlap, necessitating
protection of large community areas as opposed to discrete areas where differential management can
be applied. Drinking-water protection may not be achievable by relocation of contaminant sources,
but rather by contanhimint reduction or vigorous contanlinant-source management.
12
-------
References Cited
Blanford, T. N., and Huyakorn, P. S., 1993, WHPA: A Modular Semi-Analytical Model for the
Delineation of Welihead Protection Areas, Documentation and Users Guide, Version 2.2.
International Ground Water Modeling Center, Colorado School of Mines, Golden, Colorado.
Giese, G. L., Eimers, J. L., and Coble, R. W., 1991, Simulation of Ground-Water Flow in the
Coastal Plain Aquifer System of North Carolina. U. S. Geological Survey, Open-File Report 90-372,
178 pages.
Grant, Gary, 1995, Concerned Citizens of Tillery, Oral Communication.
Solley, W. B., Pierce, R. R., and Penman, H. A., 1993, Estimated Use of Water in the United States
in 1990. U. S. Geological Survey Circular 1081.
U. S. Environmental Protection Agency, 1987, Guidelines for Delineation of Welihead Protection
Areas. EPA-440/6-87-010.
U. S. Geological Survey, 1985, Roanoke Rapids, North Carolina, 1:100,000-scale Metric
Topographic Map, 36077-A l-TM-100.
U. S. Geological Survey, 1974a, Boones Crossing Quadrangle, North Carolina, 7.5 Minute Series
(Topographic) Map, Scale 1:24,000.
U. S. Geological Survey, 1974b, Halifax Quadrangle, North Carolina, 7.5 Minute Series
(Topographic) Map, Scale 1:24,000.
U. S. Geological Survey, 1962, Scotland Neck Quadrangle, North Carolina, 7.5 Minute Series
(Topographic) Map, Scale 1:24,000.
U. S. Geological Survey, 1961, Enfield Quadrangle, North Carolina, 7.5 Minute Series
(Topographic) Map, Scale 1:24,000.
U. S. Geological Survey, 1960, Dawson Crossroads Quadrangle, North Carolina, 7.5 Minute Series
(Topographic) Map, Scale 1:24,000.
Winner, M. D., Jr., and Coble, R. W., 1989, Hydrogeologic Framework of the North Carolina
Coastal Plain Aquifer System. U.S. Geological Survey, Open-File Report 87-690, 155 pages.
13
------- |