Pumping Test in Leaky and Unconfined Aquifers
I. Leaky aquifers
Theis solution assumes that all the water pumped is from storage within
the aquifer. When this assumption is not met, the observed response to
pumping deviates from the Theis solution, as illustrated below:
3
s
5
1 10 100 1000 10000
Observation data
t
Tests Affected by Leaky Confining Beds
(Heath, 1983)
Theis:
no leakage
Leakage from
storage in
confining bed
Leakage from
other aquifers
Theis solution K’, S’ b’
K, S b
Q r S 2
s= w( u ) where u=
4πT 4Tt
Semi-confined (Hantush & Jacob)
(leakage from elastic storage in confining bed)
Q
s= H (u, β )
4πT
1/ 2
r
where β = (S ' S )1/ 2 B = Tb'
4B K'
Leaky (some water from other aquifer)
s=
Q
4πT
w u, r B( )
r K'
where =r
B Tb'
II. Water Table Aquifer
Approach #1 (Approximation)
The change in transmissivity due to dewatering is assumed to be negligible,
that is, T ≈ Kb , where b is the initial aquifer saturated thickness.
Then the Theis equation can be applied to an unconfined aquifer with storage
coefficient for confined aquifer replaced by specific yield, Sy,
Q r 2 Sy
s= w(u), u=
4πT 4Tt
This approach is reasonable only when the drawdown is
s ≤ bthat is,
significantly smaller than the saturated thickness,
Approach #2: (Concept of delayed yield)
Stage 1: Similar to response in confined aquifer. No physical dewatering yet.
Water pumped is from storage. The value of S estimated from aquifer test
represents confined storage coefficient.
Stage 2: Intermediate, similar to response in leaky aquifer. Water pumped is
partly from storage and partly from gravity drainage.
Stage 3: Water pumped is drawn mostly from gravity drainage. Drawdown
response conforms to some Theis-type curve again. The value of S estimated
from aquifer test represents Sy (specific yield) for unconfined aquifer.
1 2 3
Neuman solution
Q
s= w (u A , u B , Γ )
4πT
r2S
where uA = for early time-drawdown data
4Tt
r 2 Sy
uB = for later time-drawdown data
4Tt
r2Kz b is initial aquifer thickness
Γ= 2
b Kh
Effect of partial penetrating wells
The effect of partial penetrating wells is negligible if
(1) The length of pumping well screen is close to
aquifer saturated thickness
(2) The observation well is fully penetrating
(3) The observation well is located sufficiently Partially Partially Fully
penetrating penetrating penetrating
far away, i.e.,
pumping observation observation
r > 1.5b K h K v well well well
r
b
Well Loss
There is usually a difference between the aquifer head and the water level inside
the well bore. This difference is referred to as ‘well loss’. The magnitude of well
loss is controlled by pumping rate, well diameter, degree of partial penetration, and
aquifer properties.
Slug Test
(also known as piezometer test, falling/rising head test, or response test)
A known volume of water
is quickly taken from (or
added to) a well. The rate
at which the water level
falls or rises is measured.
• Inexpensive alternative
to pumping test
• Estimated parameters
represents much smaller
area than pumping test
Three major methods to analyze slug test data:
• Papadopulos, Bredehoeft & Cooper
(confined aquifer, fully penetrated well)
• Hvorslev
(confined aquifer, partially penetrated well)
• Bouwer & Rice
(unconfined aquifer, fully or partially penetrated well)
r 2 ln (Le R )
K=
2 Let37
t37: time when the
h(t)/h0 water level
reaches 37% of
initial change
Kansas Geological Survey Guidelines for Slug Tests
¾Three or more slug tests should be performed at a given well. If the well has not been properly developed prior to slug testing,
the slug test itself might cause additional development. This can be detected by a shift in the calculated transmissivity during the
repeated slug tests. Also the slug test can mobilize fine material, which can result in a decrease in formation transsmissivity.
¾Two or more different initial displacements should be used during testing of a given well. A series of tests with initial head
displacement, Ho, which varies by a factor of at least 2, should be employed; however, the first and last tests should have the same
Ho so that a dynamic skin effect can be detected.
¾The slug should be introduced in a near-instantaneous fashion and a good estimate of the initial displacement should be
obtained. This is a basic assumption of the slug-test methodology. While it is easy to accomplish in a low permeability system, it
can be difficult to do in a system that has a very rapid response.
¾Appropriate data-acquisition equipment should be used. Manual methods of measurement might be satisfactory for a well in a
low permeability sediment that responds over a period of many minutes to hours. For more permeable formations where the total
period of response can be less than one minute, a data logger with a pressure transducer is mandatory. The pressure transducer
should have the proper sensitivity for the planned head displacement.
¾An observation well should be employed for estimation of the storage parameter. A single well is fairly insensitive to the
value of the storage parameter for several reasons. If possible, a nearby monitoring well screened across the same interval should be
used to determine the storage coefficient.
¾Method chosen for data analysis should be appropriate for site conditions. Ensure that the chosen method is correct: confined
or unconfined aquifer, fully penetrating or partially penetrating well.
¾Use of pre- and post analysis plots should be an integral component of the analysis. Even though care was taken to select the
proper method of analysis based on the field conditions (point 6), it is still possible that either the method which was selected, or
some assumption that went into the analysis, is wrong. Careful examination of the data plot versus the theoretical model of the test
data should be made. While this seems to be an obvious point, in many cases the field data are analyzed with an automated curve-
fitting program. A human should still review the final plot.
¾Appropriate well construction parameters should be employed. The effective screen length and screen diameter are critical
parameters. If the well is gravel packed, the length and diameter of the gravel pack should be used if the gravel pack is significantly
more permeable than the formation. However, if well development is incomplete, the screen length may be more appropriate than
the length of the gravel pack.
.
Key Reference: Butler, J.J., 1997, The Design, Performance, and Analysis of Slug Tests, Lewis Publishers, FL.