Examples of Delineation Methods
Return to: TABLE OF CONTENTS, Water Quality Home Page, DEQ Home Page
SAFE YIELD CALCULATION
CALCULATED FIXED RADIUS EXAMPLE
CONCEPTUAL MODEL DEVELOPMENT
POROUS MEDIA ASSUMPTION
SAFE YIELD CALCULATION Top of Appendix A
The safe yield calculation is designed to determine the maximum pump rate that can be sustained without lowering the water level in the well bore to the screened or perforated interval (assuming the pump intake is not placed at a shallower level).
The first step in the process is to determine the specific capacity of the well. This calculation yields the relationship between pump rate and drawdown. To determine specific capacity, the well is pumped at a constant rate at a discharge approximately 50 percent of pump capacity. Pumping continues until the drawdown has stabilized. From a practical perspective, drawdown can be considered stabilized if the amount of change in pumping level between measurements (1 hour interval) is less than 5 percent of the total drawdown to that point. The minimum duration of the pumping interval should be 4 hours; specific capacity can best be determined in conjunction with an aquifer test (see Section Aquifer Tests).
After drawdown has stabilized, the specific capacity (SC)
can be determined as follows:
SC = Q/s
Where "Q" is the pump rate in gallons per minute
and "s" is the drawdown in feet. SC then has units of gpm/ft. As an example,
assume that a pump rate of 250 gpm produces a drawdown of 18 feet. The specific capacity
then equals 250 gpm/18 feet = 13.9 gpm/ft drawdown. This implies that for every 13.9 gpm
of pump rate, the drawdown produced will be one foot.
In the diagram below, a well is illustrated showing the
static water level (SWL), screened or perforated interval, or water-bearing zone in an
uncased well (SI), and depth to the top of the screens (SD).
In calculating the safe yield, we first determine the
available water (AW) in the well as follows:
AW = SD - SWL
The available water is the distance that the water level can be lowered
without encountering the screened- or water-bearing interval. Note that in many areas, the
SWL varies seasonally as a function of recharge to the aquifer and demand. Water systems
should base their safe yield on the lowest SWL of the year and continue to monitor that
level to make certain that the resource is not in a state of decline. In addition, systems
may want to build in a safety factor by multiplying AW by some value less than one, e.g.,
0.90, to prevent accidentally lowering the pumping level below the screens.
As an example, if the seasonally lowest SWL for the
system was 23 feet and the depth to the water-bearing zone or the screened interval was 51
feet, the AW would be calculated as follows, using a 90 percent safety factor:
AW = (51-23) ´ 0.9 = 25.2 feet
Figure A-1: Example Well Characteristics
The safe yield (SY) for the well is obtained by multiplying the
available water (AW) times the specific capacity (SC). Figure A-1 illustrates this
example calculation of the safe yield.
SY = AW ´ SC
This calculation indicates that the well could be pumped at 350 gpm without lowering the pumping level to the screens. It should be noted that depending upon other use in the area, the aquifer may not be able to sustain that pump rate. Static water levels in the aquifer should be monitored to ensure that long-term decline of water levels in the aquifer are not indicated.
Top of Appendix A
CALCULATED FIXED RADIUS
EXAMPLE Top of Appendix A
The Calculated Fixed Radius (CFR) method for wellhead delineation determines the volume of the aquifer that will be required to supply water to the well for a period of time equivalent to the time-of-travel (TOT) criteria established by the state. In Oregon, the TOT applied to the CFR technique is 15 years. The CFR technique assumes a uniform aquifer and a flat water table, i.e., negligible groundwater flow. As a result, the volume calculated is a cylinder and the wellhead protection area is a circle (see Figure 3-3, Section 3.3). The radius (R) of the circle depends on the TOT, the adjusted pump rate (Qa; see text), the effective porosity (ne) and the screened/perforated or water-bearing interval (I). The appropriate equation is as follows:
r=The radius of circle (actually the cylinder);
Qa=Cubic feet/year (= gpm ´ 70267);
ne=A dimensionless number between 0 and 1.0 (see Table 3-1, Section 3.3).
Example 1 - The Calculation
A community of 400 has an average daily water use of 140,000 gallons. The water is derived from two separate wells both of which produce equally1. The average daily use of each well, therefore, is:
140,000 gpd per 2 Wells = 70,000 gpd per Well
70,000 gpd per 1,440 min/day = 49 gpm per Well
Qa = 49 gpm ´ 1.25 = 61 gpm =
61 gpm ´ 70,267 = 4,286,287 ft3 / year
From the well report:
I=22 feet (-33 to -55 feet);
ne=0.2 (Table A-1 - for sand and gravel);
TOT=15 years (pop. <500, Figure 3-1, Section 3.3)
1 If data were available that indicated differing pump rates
for the two wells, the total usage would be divided between the wells in a manner
consistent with those pump rates.
Substituting into the above equation yields:
r = 1,761 feet.
The wellhead protection area for this well, therefore, is constructed by
drawing a circle of radius 1,761 feet around the wellhead. Internal delineations can be
constructed by substituting different values of TOT in the equation. For example, the
radius corresponding to a 6-month TOT would be 322 feet.
If the second well, with the same production rate and in the same
aquifer, was perforated through 15 feet instead of 22 feet, a value of 15 would be used
for "I" in the equation. The resulting value of "r" for the second
well using a 15-year TOT would be 2,133 feet. Because a smaller aquifer thickness is being
used by the second well (15 versus 22 feet), a larger radius is required to supply the
same amount of water.
Example 2 - Overlapping Wells
The city in the example above noted that after drawing the appropriate circles around wells 1 and 2 that the areas overlapped. The city correctly reasoned that because the wells were drawing from the same interval, the wells could not each draw the same water from the overlapping area. To compensate for the overlapping area, the city chose to represent the two smaller wells as a single well whose production rate was equal to the sum of the two individual wells (140,000 gpd). The hypothetical well was located exactly half way along a line between the two active wells. If the production rate of the two wells differed, then the placement of the hypothetical well would be closer to the higher producing well.
Figure A-2: Determination of Location ofSingle Hypothetical Well
NOTE: Determination of location of single hypothetical well to replace actual wells in the case where individual CFR circles overlap. One circle is drawn with hypothetical well as center point having discharge equal to sum of all individual wells. (See text for explanation.)
Figure A-2a illustrates this case. Two wells are separated by
1,000 feet. Well 1 (point A) pumps at 300 gallons per minute, while well 2 (point B) pumps
at 100 gallons per minute. The location of the hypothetical well (producing 400 gallons
per minute) would clearly be closer to "A" (well 1) than "B" (well 2).
Well A pumps at 0.75 of the total discharge (300 gpm/ - 300 gpm + 100 gpm); therefore, the
location of the hypothetical well would be 75 percent (i.e., 750 feet) of the way from
"B" to "A" (point C in Figure A-2b). The radius of the
hypothetical well would be calculated based on the 400 gpm figure and drawn from point C.
If three wells, all with overlapping circles, had been involved in this
case, the hypothetical well, with a discharge equal to the sum of the three individual
wells, would be located along the line segment "C-D" in Figure A-2c. If
the pump rate of well 3 (point D) was 200 gpm, than the hypothetical well would be placed
0.33 (i.e., 200 gpm/(200 gpm + 400 gpm)) of the distance towards "D" along
segment "C-D". In calculating the radius for the three wells, the discharge
would be 600 gpm and the radius would be drawn from point E in Figure A-2d.
If the wells have different intervals, use an average thickness in the calculation, not the sum.
This process of drawing line segments can be continued to include
additional wells as necessary. Further, the final placement of the hypothetical well is
independent of the order in which the segments are drawn as long as the appropriate pump
rates are used in the calculation. For example (Figure A-2d), the point E would
have been the same if we would have started with segment "B-D" and then went to
segment "A-B". This method should not be construed as a necessarily accurate way
of compensating for the overlap. Rather, it represents an approximate solution to the
If proximal wells are drawing from separate aquifers, it is permissible
for the circles to overlap at the surface.
Example 3 - Delineation of A Wellfield
In cases where nonoverlapping CFRs are drawn around individual wells in a wellfield, it is reasonable to delineate a wellfield protection area comprising all of the individual CFRs. For example, Figure A-3 shows a case where the WHPAs for four wells within a wellfield have been delineated using the CFR technique. As can be seen in the figure, the individual circles are separated from one another by small areas that according to the CFR technique, overlie portions of the aquifer that do not contribute to the well.
Given the uncertainty of the CFR technique, however, it would be prudent to include those inner areas within the wellfield protection area. Accordingly, it is recommended that in such situations, all of the CFRs be enclosed in a smooth curve (dashed line in Figure A-3) to provide a conservative approach to protection.
Figure A-3: Calculated Fixed Radius Method Applied toFour Distinct Wells
NOTE: Because of uncertainty associated with the calculated fixed radius method, a singleline (dashed in figure) is drawn to encompass all the individual circles. The area enclosed by the dashed line should be considered as the wellhead protection area for the wellfield. (Shown individually as "x".)
Example 4 - Well Near A Boundary
1. Stream Boundary - In those cases where the CFR of a well intersects a perennial stream, the CFR can simply be truncated at the stream (Figure A-4a).
2. No-Flow Boundary - If the CFR intersects a no-flow hydrogeologic boundary, e.g., impermeable
bedrock valley wall, adjustments to the position of the circle should be made (Figure
A-4b). This can be accomplished by shifting the CFR relative to the boundary. The
shift should be perpendicular to the general trend of the boundary. The CFR should be
moved until the circle is tangential to the boundary. The area between the circle and the
boundary should be included in the delineated WHPA (Figure A-4c).
The adjustments above should not be regarded as technically defensible
methods of correcting for the presence of the boundary. Rather they reflect an attempt to
improve the estimate of the area involved in supplying water to the well.
Figure A-4: Calculated Fixed Radius (CFR) Method in Proximity to A Hydrogeologic Boundary
(a) For wells where the circle delineated by the CFR method intersects a perennial stream, the circle should be truncated at the stream.
(b) CFR method applied to a well adjacent to a no-flow boundary, except valley alluvial sediments against the bedrock valley wall.
(c) The circle drawn in (b) should be moved directly away (perpendicular) from the boundary (see arrow) until the circle is tangential to the boundary. The area within the circle and the area between the circle and the boundary should be included in the WHPA.
Top of Appendix A
CONCEPTUAL MODEL DEVELOPMENT Top of Appendix A
As indicated in Figure 3-3 (Step 3 - Flowchart), public water systems having a population >500 must develop a conceptual model as a framework for the delineation of the WHPA. If the population is between 500 and 3,300 the conceptual model can be constructed from regional data. For larger populations site-specific data must be incorporated.
The various components of the conceptual model are listed below. The
requirements below represent what the State views as the minimum data needed for a
framework to construct the conceptual model. They may, of course, be modified based on
best professional judgment.
Components of The Regional Conceptual Model (Population 500-3,300)
1. Compilation of Regional Studies - A literature search should be conducted to locate any existing geological/hydrogeological publications or reports that have included the area of concern in their study. This literature search can be greatly facilitated by contacting local and state agencies that deal with water issues: Watermaster's Office, Water Resources Department, Regional office of the Department of Environmental Quality, Groundwater Section of the DEQ, U. S. Geological Survey, Drinking Water Program of the Oregon Health Division, the Department of Geology and Mineral Industries (DOGAMI), Geology Departments at area colleges and universities.
2. Development of A Well-Log Inventory
- Experience from the Demonstration Projects indicates that this
task can be most efficiently accomplished by the water system because of its familiarity
with the local area. Well reports for wells within the area should be obtained from the
local watermaster's office or the Salem office of the Water Resources Department. It is
recommended that the well reports from the section within which the well of concern occurs
and the eight (8) surrounding sections should be obtained. Wells should be located on
appropriate 7.5 minute quadrangle maps. Depending on available information and the pump
rate, however, well reports from fewer sections may be sufficient. Other features of
interest, e.g., springs, surface waterbodies, high-production wells, etc., should also be
noted at this time.
3. Fence Diagram - The well-log inventory should be utilized to identify the hydrologic units
within the area. Data allowing, spatial variation of the units should be identified
through the use of a fence diagram, constructed from a minimum of two mutually
perpendicular sets of three cross sections. Ideally, these should be oriented so that one
set is approximately parallel to the direction of groundwater flow.
4. Field Visit -
A field visit to the area provides a valuable perspective in developing the conceptual
model. It should be used to ground check the well log inventory and to determine if the
fence diagram is consistent with surface exposures.
5. Recharge to The Aquifer - The GPTRAC semi-analytical model allows for input of the amount of recharge to
the aquifer. In Oregon, the most common sources recharge to the aquifer are rainfall and
irrigation. Rainfall amounts can be derived from local records or from the Oregon
Precipitation Map (Oregon Climate Service, Oregon State University). Irrigation
application can be arrived at by local records or from estimates based on crop
requirements (Consumptive Use Requirements, Soil Conservation Service; see Hydraulic
Surplus Determination, OHD's Guidance Document for Monitoring Reduction through Use, and
Susceptibility Waivers). Infiltration coefficients depend highly on the nature of the
surface (e.g., residential, industrial, agricultural, etc.), the extent of the vegetation,
character of the rainfall, and, in the case of irrigation, the irrigation method (e.g.,
flood, sprinkler, drip, etc). Published recharge rate estimates are available for some
areas. Where they are not available, an estimate of one-third to one-half of the
precipitation rate is probably reasonable for natural recharge.
6. Other Wells -
Well reports from the area, supplemented by a local area review, should be evaluated to
identify any large production wells (e.g., irrigation or industrial) that might exist in
the area. Wells used specifically for recharge should also be located. These include dry
wells utilized for disposal of storm water or other wastewaters. 7. Hydrogeologic
Characteris-tics - Parameters utilized directly in the delineation of the wellhead
protection area include the hydraulic gradient, hydraulic characteristics of the
hydrogeologic units and the presence of hydrogeologic boundaries.
Table A-1: Approximate Values of Porosity and Hydraulic Conductivity for Various Aquifer Materials
|86 - 8,600||0.24 - 0.38|
|0.25 - 1,725||0.31 - 0.46|
|0.26 - 50||0.26 - 0.53|
|2.6 ´ 10-4 - 17||0.34 - 0.61|
|2.6 ´ 10-6 - 4.0 ´ 10-3||0.34 - 0.60|
|1.3 ´ 10-2 - 0.5||0.30 - 0.50|
|8.6 ´ 10-5 - 1.7||0.05 - 0.40|
|2.6 ´ 10-6 - 4.3 ´ 10-3||0.20 - 0.40|
|2.6 ´ 10-8 - 5.2 ´ 10-4||-|
|3 ´ 10-5 - 70||0.05 - 0.50|
|5.0 ´ 10-6 - 0.1||<0.01 - 0.05|
|0.1 - 5,180||0.05 - 0.35|
Unfractured Igneous/Metamorphic Rocks
|1.0 ´ 10-8 - 5.2 ´ 10-5||<0.01 - 0.05|
Fractured Igneous/Metamorphic Rocks
|2.1 ´ 10-3 - 85||<0.05 - 0.50|
* Hydraulic Conductivity:
Increased by - Presence of sand or gravel, increase in sorting, stratification, unconsolidated character, and high secondary porosity.
Decreased by - Presence of clay, poor sorting, unstratified character,
cementation, or compaction.
Increased by - Increase in sorting, rounded grains, small particle size, unconsolidated character, and high secondary porosity.
Decreased by - Poor sorting, irregular shaped particles, unstratified, large particle size, cementation/compaction.
Source: Domenico and Schwarz, 1990; EPA, 1994
Components of the Site-Specific Conceptual Model (Population 3,300 and greater)
1-6. Components 1 through 6 - Conduct components 1 through 6 of the Regional Conceptual Model as described above.
7. Hydrogeologic Characteristics - Parameters utilized directly in the delineation of the wellhead protection area
include the hydraulic gradient, hydraulic characteristics of the hydrogeologic units and
the presence of hydrogeologic boundaries.
Hydrologic Boundaries: Streams, lithologic boundaries, groundwater divides, etc., must be identified based on hydrogeologic mapping (well reports, conceptual model and field mapping).
Top of Appendix A
POROUS MEDIA ASSUMPTION Top of Appendix A
Of concern here is the rfelation between groundwater flow direction and
the gradient. In a homogeneous isotropic aquifer, groundwater will flow perpendicular to
head contours. It is, therefore, predictable through the use of simple analytical
expressions and groundwater velocity can be determined through Darcy's Law. If the aquifer
is fractured, groundwater flow direction may be significantly different than the gradient
direction (Bradbury et al., 1991; Bradbury and Muldoon, 1994). Further, Darcy's Law, which
assumes laminar flow, cannot be applied to groundwater flow through discrete fractures
(EPA, 1994). Heterogeneities within the aquifer, manifest as large hydraulic conductivity
contrasts, may also result in groundwater flow that departs from the general down-gradient
direction (EPA, 1994).
Recognizing Fractured Aquifers
The first step in determining whether the aquifer that is supplying the well of concern is fractured or porous is to evaluate the geologic province that the aquifer falls within. Wells producing from aquifers within alluvial basins are unlikely to be producing from a fractured aquifer, whereas those producing from any bedrock area outside or below the alluvial basins may be. A review of the well logs may provide useful information. If the producing zone is described as a sediment, e.g., sand, gravel or silt, it is justifiable to assume that the aquifer is porous. If on the other hand, the aquifer is described as bedrock, e.g., basalt, granite, sandstone, etc., further evaluation is necessary. Geologic maps and/or aerial photographs may be useful in a preliminary evaluation. Geologic maps will show the location and orientation of fractures (joints and faults) within an area. Aerial photographs may indicate the presence of fractures by linear features on the image.
Evaluation of The Porous Media Assumption
The EPA recommends the following "subjective criteria" for determining whether a fractured-rock aquifer can be regarded as a porous medium for purpose of application of analytical models discussed below (Bradbury et al., 1991).
1. Pumping Test Responses - A porous medium should respond
Note that it is the drawdown that is contoured, not the hydraulic head.
A symmetrical drawdown imposed on a regional gradient may yield an elliptical head pattern
around the well.
2. Fracture Scale to WHPA Scale - The observed fractures
should be numerous and small (size and spacing) relative to the delineated area. The EPA
offers as a general rule that if the WHPA dimensions are at least 100 times that of the
fracture spacing, the porous media assumption may be justified. The scale of fractures may
be estimated from geologic maps, from borehole geophysical studies, including down-hole
camera, and field investigations.
3. Distribution of Hydraulic Conductivity (K) - In an
aquifer that conforms to the porous media assumption, the distribution of "K"
should be relatively uniform or conform to a log-normal distribution. In a fractured
terrain, the distribution of "K" may yield a bimodal distribution, showing two
dominant "K" values, one corresponding to the conductivity along the fractures
and the other to the conductivity normal to the fractures.
4. Variations in Groundwater Chemistry - The EPA (1991)
provides several criteria related to groundwater composition that may indicate flow along
fracture surfaces. These include: (1) significant variations in composition in space and
time, and (2) low total dissolved solids and mineral saturation indices. It should be
noted that there are alternative explanations for the above observations and groundwater
chemistry is most useful if the data from a number of similarly constructed wells are
Even though a given fractured aquifer meets all criteria to be
considered as behaving as a porous medium, there is still a consideration to be made in
the application of traditional analytical tools. Bradbury and Muldoon (1994) Note that
even under such situations, the actual capture zone of the well is invariably larger than
that calculated through the use of the analytical methods.
Heterogeneities Within The Aquifer
If large hydraulic conductivity contrasts occur within the aquifer, either as a result of depositional or deformational histories, groundwater flow direction will change as groundwater crosses from one material into another. In general, the groundwater flow paths will be "refracted" into the medium having the higher value of "K" and will exhibit preferential flow within that medium even though that path may not be perpendicular to the regional groundwater gradient. A method of transforming the direction of flow to accommodate aquifer heterogeneities is given in EPA (1994) and references cited therein.
Top of Appendix A
AQUIFER TESTS Top of Appendix A
An aquifer test consists of a carefully planned interval of pumping and water level monitoring. We use the term "aquifer test" instead of "pump test" for two reasons. First of all, the objective is not just to test the pump, rather it is to better characterize the aquifer. Secondly, an aquifer test is a multi-phase effort that includes a number of steps in addition to just pumping the well.
The determination of aquifer properties in the site-specific conceptual
model requires an aquifer test. This test should be at a constant rate for a minimum of 24
hours for a confined aquifer and 72 hours for an unconfined aquifer. Recovery should be
allowed to occur over the same time period as drawdown. Pump rate should be consistent
with normal production levels. In some cases, the well yields may be too small to allow
for prolonged pumping at normal production levels. In those cases, alternate methods,
e.g., slug tests (see below) may have to be used. Ideally, at least two observation wells
should be monitored during the test. If the aquifer has confined characteristics, a third
shallow observation well placed to explore leakance should be utilized.
The drop in the water levels (i.e., the drawdown, in the production well
and in nearby monitoring wells as a function of time) are monitored throughout the test.
The drawdown-time data is critical to the proper determination of aquifer properties.
Therefore, the monitoring is done according to a rigorous schedule in order to ensure that
the data obtained is useful.
How Are Aquifer Test Results Used?
There is a direct relationship between the aquifer's characteristics and the amount of drawdown relative to both the time since pumping began and the distance a monitoring well is from the production well. For example, if we know the aquifer's transmissivity (equal to the hydraulic conductivity multiplied by the aquifer's thickness) and storativity (the amount of water the aquifer releases), we can predict what the drawdown will be for a given pump rate after a certain time and at a certain distance from the well.
It follows, then, that we can use those same equations in reverse to
calculate transmissivity and storativity if we know the drawdown, time and distance.
Generally, the solutions are graphical; i.e., we plot the data and compare the graph to
theoretical solutions. There are many different types of solutions, depending on the
Designing The Aquifer Test
Aquifer tests can be expensive and time consuming. Therefore, it is important that the data be collected in a manner that will yield useful results in terms of aquifer characteristics. We recommend that you consult with a professional early on in the planning of the aquifer test so that the data collected will meet your needs and expectations. The Drinking OHD Water Program (503-731-4010) will provide technical assistance to help get you started.
Importantly, there is no "off-the-shelf" aquifer test plan
that is available. Critical questions of what pump rate, what test duration, what
additional wells should be involved and at what distance, what hydrogeologic boundaries
may affect the results, and what corrections need to be applied to the drawdown data, all
need to be addressed on a site by site basis. As discussed above, there are several
solution methods that are applied to aquifer test results depending on the hydrogeologic
setting and well construction characteristics. For the solutions to be valid, certain
criteria must be met with regard to how the test is performed. Clearly these criteria have
to be identified prior to running the test itself.
A conceptual model of the hydrogeologic setting based on well reports
and other data available should be prepared prior to the test. From this information, a
simulation of the aquifer test can be run prior to the actual test in order to identify
potential problems and critical data to be collected. This allows the test to be designed
to fit your specific setting and helps to ensure that the data collected will in fact
represent your part of the aquifer and groundwater flow system.
1. Inadequate Planning - The most common mistake is not planning the test adequately beforehand. As a result the data is only marginally useful and any use of it is open to some question.
2. Too Short of Test - Pump test data of limited duration (i.e., 1 to 4 hours) may be useful to evaluate the specific capacity of the well or to monitor the groundwater resource regionally, but is generally inadequate to define aquifer characteristics. The purpose of the test is to obtain representative values for those parameters that influence groundwater movement in the aquifer. The longer the duration of the test, the larger volume of aquifer involved; therefore, the more representative is the data of the aquifer.
In some instances, the early water derived from the well may be supplied
wholely by thin highly permeable beds. With further pumping, other less permeable zones
may be involved. If the pump test had been of limited duration, the resulting aquifer
characteristics might have reflected the thin beds only. Applying their characteristics to
the entire aquifer during delineation modeling would have led to erroneous results.
In unconfined aquifers, a test of limited duration may not capture the
delayed yield of the aquifer. In such cases, the results might indicate an artifically low
transmissivity or be mistaken for the impact of a nearby boundary.
The potential impact of hydrogeologic boundaries is an important issue
to recognize in aquifer tests. These boundaries (e.g., streams, geologic contacts,
groundwater divides, etc.) may significantly affect groundwater flow in the area. There
presence can be recognized in a longer test, but may be completely missed in a test of
3. Inadequate Recovery - As has already been discussed,
the important data that is collected is the amount of drawdown as a function of time in
the given well. If the pump has been on just prior to the test and the water level in the
well has not fully recovered (i.e., returned to its pre-pumping level), the drawdown
recorded subsequently will not accurately reflect the pumping conditions during the test.
As a result, the aquifer characteristics determined will be in error. The well should be
idle for a minimum of 16 hours prior to the aquifer test.
4. Inadequate Corrections to Drawdown - A number of
factors other than pumping can influence the water level in the well during the test.
These include long-term changes in the aquifer due to regional pumping or recharge
effects, changes in barometric pressure (especially for confined aquifers), changes in
surface water stage (especially for unconfined aquifers) and interference from nearby
pumping wells. If these features are identified and monitored before and during the test,
corrections can be made to the data.
5. Poor Monitoring Practices - We have seen aquifer test
data in which the water levels have been measured too infrequently or too imprecisely.
Careful monitoring is critical to the utility of the data. We provide recommendations for
frequency of monitoring water levels below. With respect to field measurements, it is
recommended that the pump rate be monitored on a 2-hour basis and the rate be maintained
within 10 percent of its starting value. Significant variations in the pump rate pose
large problems in interpreting the data.
Water level measurements should be determined to the nearest 0.01 feet.
Tapes marked in tenths/hundredths of feet should be used as opposed to inches/feet. Time
determinations should be made to the nearest minute, and if more than one observer is
involved, the measurements should be synchronous to within 1 percent of the time since
6. Improper Conveyance of Pumped Water - The water brought
to the surface during the pump test must be piped sufficiently far away from the
production and monitoring wells so that it will not seep back into the ground and
artificially recharge the aquifer in the vicinity of the well. This is particularly
important for aquifer tests involving unconfined aquifers.
7. Well Interference - During the pre-pumping, pumping,
and recovery phases, the presence of any other pumping wells within 1500 feet should be
noted. The pre-test simulation may help to better define the distance from the well
another well can be before interfering with the pumping level in the production or
Below are recommendations regarding the collection of data during the aquifer test's pre-pumping, pumping, and recovery phases.
NOTE: These are general recommendations only; modifications may
be necessary as dictated by the conceptual model and simulation results.
1. Duration - The pumping phase should be at a constant
rate for a minimum of 24 hours for a confined aquifer and 72 hours for an unconfined
2. Pump Rate - The pump rate should be at normal operating
levels, but care must be taken to avoid the possibility of excessive drawdown, i.e.,
lowering the water level to the perforations or screens, during the test. It may be
necessary to calculate the safe yield of the well and set the constant rate at 75 percent
of that value.
3. Observation Wells - If other wells, e.g., domestic or
irrigation (open to the same aquifer as the test well), are available in the vicinity
(e.g., within 1000 feet), they should be identified as possible observation (monitoring)
wells. The use of observation wells greatly enhances the ability to obtain representative
data during the test. The conceptual model and simulation will provide information as to
which wells can be used as a function of their depth and distance. If these observation
wells are screened over different portions of the aquifer, corrections to the drawdown
will probably be necessary. If the aquifer being evaluated is confined, it may be useful
to select an observation well completed within the overlying unconfined aquifer to
determine if there is any leakage from the overlying aquifer into the confined system.
4. Stream Stage - If there is a stream near the well being
tested, and the conceptual model or simulation suggests a potential connection, it may be
useful to periodically monitor the stage (depth and width) of that stream during the test.
In areas near the coast, tidal fluctuations should be considered
5. Pre-Pumping Phase - The well to be tested should remain
idle for at least 16 hours prior to the test. During that time, water level measurements
should be made at 16, 12, 3, 2 and 1 hours prior to initiating pumping. Within the hour
immediately proceeding pumping, water level measurements should be taken at 20 minute
intervals. The purpose of this exercise is to establish any long term trends in water
level changes that may be occurring. Barometric measurements of atmospheric pressure
(inches of mercury) should be made as well. Confined aquifers may show significant
responses, e.g., 0.5 to 1 foot, to large changes (e.g., 1 inch of mercury) in atmospheric
pressure. These measurements will allow the determination of the barometric efficiency of
the aquifer so that corrections can be applied to the drawdown data.
6. Pumping Phase - After initiation of the pumping, drawdown measurements in the production and observation wells should be made according to Table A-2. The most critical period of measurements are within the first 100 minutes, when the water levels are changing rapidly.
Table A-2: Drawdown Measurement
|Time Intervals||Time After Pumping Started|
|0 - 1 Minute||As Frequent as Practical|
|1 - 10 Minutes||1 Minute|
|10 - 100 Minutes||10 Minutes|
|100 - 300 Minutes||30 Minutes|
|300 - 1,000 Minutes||1 Hour|
|1,000 - 5,000 Minutes||4 Hours|
|5,000 - End||1 Day|
7. Recovery Phase - Water level measurements made during
the recovery of those water levels after the pump has been shut down should be taken at
the same frequency as the drawdown measurements during the pumping phase. As in the
drawdown phase, the most important information is obtained during the first 100 minutes.
Measurements should continue for the same duration as in the pumping phase, or until the
water levels have reached 95 percent recovery.
8. Measurement Devices - Water level and flow rate
measurement methods should be in accordance to Water Resource Department requirements (see
"Pump Test Requirements for Ground Water Right Holders" distributed by the
What About Low Yield Wells?
Instances occur for low demand wells where the well's yield will not permit long-term pumping. In other words, the aquifer cannot supply water to the well bore at the same rate as the pump extracts it. As a result, the pumping level drops below the perforations or pump intake and no water is produced. This is most common in areas where the aquifer has a low hydraulic conductivity, e.g., has a high proportion of fines or is characterized by low-density fractures.
In these instances, the aquifer test may be designed using a lower pump
rate, or may involve the use of slug tests. A slug test involves introducing an object or
volume of water into the well and recording how long it takes the water to return to its
initial level. This should not be necessary for a water supply well equipped with a pump.
As an example, a cylinder of solid PVC or aluminum, or a capped,
sand-filled PVC pipe, is lowered into the well. Its volume displaces the water to a higher
level. Over time, that water level will return to the initial level. Water levels versus
time are recorded during this "falling head" portion of the test. After the
water levels have returned to the original level, the cylinder can be withdrawn. This will
result in a dropping of the water level in the well. The time versus water level data
collected during this "rising head" portion of the test will yield a second
independent estimate of the hydraulic conductivity from the falling head portion.
There are a number of disadvantages to using slug test results for
estimating hydraulic conductivity. Perhaps the most limiting is the fact that the volume
of aquifer involved in the test is very small and therefore the results are not very
representative. A second problem is the fact that there is always borehole damage that
occurs when the well is drilled that changes the hydraulic conductivity of the aquifer at
the bore hole-water interface.
Finally, if the casing is not perforated, the water must
enter and leave from the bottom of the casing. This results in the data reflecting the
vertical hydraulic conductivity rather than the horizontal conductivity.
Who Can Perform Aquifer Tests?
Aquifer tests should be performed by qualified individuals. The conceptual model/simulation phase should be performed by registered geologists, engineering geologists or professional engineers, providing they have hydrogeological experience. The actual test itself can be conducted by experienced individuals in the above professional groups as well as licensed well drillers and certified water rights examiners.
Top of Appendix A
1. Bradbury, K.R., Muldoon, M.A., Zaporozec, A. and Levy, J.
1991. Delineation of Wellhead Protection Areas in Fractured Rocks. U.S.
Environmental Protection Agency, EPA 570/9-91-009.
2. Bradbury, K.R., Muldoon, M.A. 1994. Effects of Fracture
Density and Anisotropy on Delineation of Wellhead Protection Areas in Fracture-Rock
Aquifers. Applied Hydrogeology, 2: 17-23.
3. Domenico, P.A. and Schwarz, F.W. 1990. Physical and
Chemical Hydrogeology. John Wiley & Sons, New York, 824 p.
Top of Appendix A
Return to: TABLE OF CONTENTS, Water Quality Home Page, DEQ Home Page