Lab Exercise #4

Well Construction
Steady-state and Transient Flow to Wells
GEO 476/391C Hydrogeology

What is a Pump Test?

A pump test (A.K.A. an aquifer test) is used to estimate the hydraulic properties of an aquifer or
confining bed by pumping water out of a well and observing the drawndown in observation wells.
Pump tests can predict the ability of the formation to yield water; and to infer hydraulic boundary
conditions. From a pump test, we can determine transmissivity (T), and storativity (S).

Specific Storage (Ss): the volume of water released from a unit volume of aquifer for a unit drop
in head. Units [1/L]

Storativity (S): the volume of water released from a unit area of aquifer for unit drop in head. Units
[dimensionless]
S = Ss ⋅ b

Specific Storage: Calculating Ss from compressibility of the aquifer and fluid.

When we pump water out of an aquifer we are releasing water from storage and decreasing head.
We increase the effective stress when we pump and we decrease the fluid pressure.
PUMPING: σe p

(1) Increasing σe will cause compaction in the aquifer

dVw = −dVT = αVTdσ e
A unit total volume of 1 and unit decline in hydraulic head will give us:
dVw = αρg

(2) Decreasing p (fluid pressure) will cause the fluid to expand

The water will expand due to the decrease in fluid pressure dVw = −β Vwdp

For a Vt =1, then the change in fluid pressure is dp = ρgdh and Vw = ΦVt
And for a unit decline in hydraulic head (dh = -1)
dVw = β Φ ρg

Specific Storage (Ss) is equal the sum of dVw caused from the aquifer compressibility (α) and
dVw caused from the fluid compressisbility (β)

Ss = ρg( α + βΦ )

1 Physical Hydrogeology 476K and 191C

Purpose
The tests discussed below use response of a well to pumping primarily to determine the
hydraulic conductivity (K) and specific storage (Ss) or their two-dimensional, horizontal
equivalents, transmissivity and storativity (T and S

Common assumptions
The tests use a number of simplifying assumptions to make inferences about the media
being tested. The common assumptions are:

1. Flow is only horizontal to or from the well;

2. Flow is uniform with depth;
3. Non-darcian (i.e., nonlaminar) flow near the well can be ignored, or is nonexistent;
4. Well is fully penetrating;
5. Well is infinitely small in diameter;
6. Well discharge is constant during the test;
7. Pre-pumping water level is horizontal;
8. Water is released instantaneously from storage upon head drop; and
9. The aquifer is infinite (i.e., no hydraulic boundaries)

Many of these assumptions can be relaxed, but the resulting analyses are more complex.
Below we synopsize steady-state formulae, the Theis equation; the Cooper-Jacob simplifications;
leaky aquifers; unconfined aquifers; and corrections to some of the assumptions listed above.

STEADY-STATE FORMULAE

Confined aquifer (Thiem equation)

Q ln( r2 r1 )
h 2 − h1 = s1 − s 2 = (1)
2πT

where h1,2 and s1,2 are the heads and drawdowns,

respectively, at r1,2, all in [L]; Q is the pumping
rate [L3 t-1]; and T is transmissivity [L2 t-1].
Figure 1. Confined aquifer.

2 Physical Hydrogeology 476K and 191C

Unconfined aquifer

2 2 Q ln( r2 r1 )
h 2 − h1 = (2)
πK

Figure 2. Unconfined aquifer (left) and diagram of distances r1,2 from the pumping well (right).

In the unconfined case, the lowering of the

water table decreases the area available for flow,
as per Darcys law:

Q = K∇hA (3.1)
dh
=K 2πr1 h 1 (3.2)
dr
dh
=K 2πr2 h 2 (3.3)
dr
Note that r1h 1 < r2 h 2 .
Figure 3.

3 Physical Hydrogeology 476K and 191C

 THEIS EQUATION

Theis (1935) derived the following relationship, now called the Theis equation, based
upon an analogy to heat transfer:
Q
s= W( u ) (4.1)
4πT
where
∞ i
i +1 u
W( u ) = −0.577216 − ln( u ) + ∑ ( − 1) (4.2)
i =1 i ⋅ i!

where t is time since pumping commenced [t] at a constant rate of pumping Q [L3 t-1] and
r 2S
u= (4.3)
4Tt

The steps in a Theis curve analysis are as follows:

Step #l. - Plot log drawdown (s) versus log time (t) on the same scale of matching curves;

Step#2 -Overlay plot and match master curve - be sure to keep axes parallel;

Step #3 - Pick a convenient point and record values of: s, t, u, and W(u);

Step #4 - Solve equations:

Q Q
s= W( u ) → T = s = W( u ) (4.4)
4πT 4 πs

r 2S 4uTt
u= →S= 2 (4.5)
4Tt r

Recall that T == Kb, and S = Ss⋅b, where b = aquifer thickness.

4 Physical Hydrogeology 476K and 191C

Questions (100 pts.):

1. (10 pts.) Sketch a cross section showing a cone of depression in a confined aquifer.
Derive the Theim equation for steady-state radial flow to a well in a confined aquifer.

2. (10 pts.) Sketch a cross section showing a cone of depression in an unconfined aquifer.
Derive the Theim equation for steady-state radial flow to a well in an unconfined aquifer.

3. (20 pts.) If a confined aquifer and an unconfined aquifer are pumped at the same rate, in
which aquifer will drawdown be greatest? (Assume that the aquifers are made of the
same materials and have the same basic properties.) Explain why in 100 words or less.

4. (20 pts.) Consider the two heterogeneous systems below. Each system has two separate
zones, each zone has a significantly different permeability than the other zone, and there
is an equivalent permeability for the overall system that is basically an “average” of the
two permeabilities. Howerver, the equivalent permeability for the system aligned parallel
to flow (A) is no the same as the equivalent permeability for the system aligned in series
(B). Derive an equation for each system that describes the overall permeability (Kequiv) of
that system. (Note: in one system, Kequiv will be an arithmetic mean of K1 and K2, and in
the other it will be a harmonic mean of K1 and K2.)

Some hints to get you started:

- assume ‘b’ (thickness) is equal in both zones

- for A: assume there is a discharge in zone one (Q1) and a different discharge in
zone 2 (Q2), and they add up to equal the total discharge of the system
(Q1 + Q2 = Qtotal)

- for B: assume that Q1 = Q2 = Qtotal Also note that the head drop across zone 1 (h2
– h1 = dh1) plus the head drop across zone 2 (h3 – h2 = dh2) equals the total head
drop (dh2 + dh1 = dhtotal)

- substitute in Darcy’s law and manipulate to get the different Kequiv equations.

A. B.

b K1 K2
K1 dh/dl
dh/dl
K2 b b b

h1 h2 h3

5 Physical Hydrogeology 476K and 191C

 5. (20 pts.) For each of the following two data sets, calculate T and S using the Theis
method. Use the type curve you were given in lab. Use two different match point to
show the non-dimensionality of this method.

A B
R = 100 ft R = 30 m
Q = 4 x 105 ft 3/day Q = 0.131 m3/s

s (ft) t (min) s (m) t (min)

3.8 1 0.14 1
5.2 2 0.22 2
6.2 3 0.28 3
7.0 4 0.32 4
7.6 6 0.37 6
8.3 7 0.39 7
8.8 10 0.44 10
10.0 20 0.55 20
12.2 40 0.66 40
15.7 80 0.75 80
16.5 100 0.81 100
19.0 300 0.99 300
20.3 500 1.06 500
21.6 1000 1.10 1000

6. (20 pts.) Using the T and S calculated in problem 5 (A) above, calculate drawdowns for
the following conditions:

r Time Pump Discharge

1 100 ft 10 min 40,000 ft 3/day
2 100 ft 1 day 40,000 ft 3/day
3 100 ft 1 month 40,000 ft 3/day
4 100 ft 1 year 40,000 ft 3/day
5 2 ft 1 day 40,000 ft 3/day
6 2 ft 1 day 1,000 ft 3/day
7 1 km 1 year 1,000 ft 3/day
8 1 km 1 year 10,000 ft 3/day
9 1 km 1 year 100,000 ft 3/day
10 1 km 10 year 100,000 ft 3/day

6 Physical Hydrogeology 476K and 191C

 Pump Test 2: The Cooper-Jacob Simplification
GEO 476/391C Hydrogeology

COOPER-JACOB SIMPLIFICATION
Cooper and Jacob (19) observed that as the variable decreases with time, the terms in W(u)
become increasingly small ( u i i ⋅ i! → 0 ) so that the well function can be approximated by
W(u) ≈ -0.5772 – ln(u) (1)
The Theis equation then becomes:
Q  r 2S  Q  4Tt 
s=  − 0.5772 − ln  =  ln 2 − 0.5772  (2)
4πT  4Tt  4πT  r S 
Therefore, in terms of log10 (ln X = 2.303 log10 X) so that equation (6) can be expressed as:

2.3Q  2.25Tt 
s= log 10  2  (3)
4 πT  r S 

There are 3 basic solutions using the Cooper-Jacob simplification: the time drawdown
test, the distance drawdown test, and the recovery test.

(1) Time drawdown test: 1 pumping well

In this test we measure drawdowns at a given point in the aquifer at different times. In
the rest of this section, ln = loge and log = log10.

2.3Q  2.25Tt  2.3Q  2.25T∆t 

s= log 2  → T = log  (4)
4 πT  r S  4 π∆ s 2
 r S 
2.25Tt 0
and S= (5)
r2

If we choose one = I log cycle of time, then

2.3Q 2.25Tt 0
T= and S = (6)
4 π∆ s r2

where t0 is the time intercept where the drawdown line intercepts the abscissa at s = 0.

Figure 4.

7 Physical Hydrogeology 476K and 191C

(2) Distance drawdown test: 1 pumping well and 1 or more observation wells

In this test, we measure the drawdown in two (or more) wells at the same time. The
procedure is the same as for time drawdown except
that we now subtract drawdowns at different
distances, not different times. If we choose a one log
cycle difference in radii (i.e., r2 = 10⋅r1), then

2.3Q
T= (7)
2π∆s
and
2.25Tt
S= 2 (8)
r0
Figure 5.

Where r0 is the intersection of the straight-line portion of the graph with the abscissa at s = 0.

(3) Recovery test: 1 pumping well

In this test a well is pumped for a certain time (t*),

then pumping stops and data is taken as the head rises (i.e.,
the well recovers). The formula is:

s' =
2.3Q
4 πT
log t ( )
t' (9)

where s' = residual drawdown and t’ = t = t*.

Note: Storativity (S) cannot be directly determined by this Figure 6.
method.

Limitations of the Cooper-Jacob and Theis methods

These equations are derived for confined aquifers. In unconfined aquifers or in the
presence of hydraulic boundaries, this method is not as accurate.

8 Physical Hydrogeology 476K and 191C

 PUMPING TESTS IN LEAKY AQUIFERS

Here we consider leakage down through an

overlying confining layer. The upper sand
maintains a constant head. K' is the
vertical hydraulic conductivity of the
confining layer. In addition we must
assume that: 1) flow in the upper confining
(leaky) layer of thickness b’ is vertical,
while the flow in the confined aquifer is
horizontal;2) the water table remains at a
constant level; and3) there is no loss of
water from storage in the confining layer. Figure 7.
The equation is:
s=
Q
4πT
(
W u, r
B
) (10)

where B = T ( K ' b')

PUMPING TESTS IN UNCONFINED AQUIFERS

In these tests, the

consideration of water produced
by the lowering of the water table
must be considered. As shown in
Figure 8, this response is
subdivided into 3 parts: the early
time, the dropping of the water
table, and very long times. Note
that there are the well function W
is a function of uA, uB, and a third
term, which is defined below.
Q
T= W( u A , u B , Γ ) (11)
4 πS
Figure 8. Unconfined aquifer response to pumping.
where
r 2S r 2 Sy r 2K z
uA = , uB = , Γ= 2 (12, 13, 14)
4Tt 4Tt b Kh
and T=Kh⋅b. The above formulation assumes that s<<b (i.e., the drawdown is small compared to
total saturated thickness of the unconfined aquifer). If this assumption is not valid, we use a
“corrected” drawdown (s’) in the calculations, where
s2
s' = s − (15)
2b
and s is the measured drawdown.

9 Physical Hydrogeology 476K and 191C

 CORRECTING FOR OTHER VARIATIONS IN AQUIFERS

Figure 9 on the following page depicts a number of variations from the ideal Theis curve,
here plotted as drawdown (s) as a function of t/r2.

Leaky or unconfined aquifers

Addresses above.

Figure 9. Graphs of variations from ideal conditions in a single monitoring well or piezometer.
The curves represent: (1) a thin phreatic or water-table or unconfined aquifer; (2) a
thick phreatic aquifer (compare with Figure 8); (3) a partially penetrating well; (4) a
barrier boundary; (5) a leaking aquifer, an unconfined aquifer (compare with Figure
8), or a recharge boundary condition.

Partially penetrating wells

Often, we ignore this effect where the observation well is fully penetrating or at some
distance away from the pumping well. Time drawdown curves approach more horizontal flow
with time. There are also some empirical corrections that can be made. Walton (1970) discusses
these.

Anisotropic media
If the medium is homogeneous, a flow net transformation can be made. The ellipse is
transformed into a circle for radial coordinates; the equations are solved; and the analyses are
back transformed to the elliptical coordinates.

10 Physical Hydrogeology 476K and 191C

Hydraulic boundaries: Image Well Theory

Image well theory using the Principle of Superposition is applied. This is depicted in
Figure 10 Here the fault zone brings low permeability material in contact with the aquifer. If
we can approximate this by an impermeable material, we superpose another pumping well across
the boundary and add drawdowns. For a stream or body of water that is not lowered by
pumping, but which recharges the aquifer, we use a constant head boundary, and add a recharge
well to maintain a zero drawdown condition at the stream. For simplicity, we assume that all
image wells have the same discharge as the pumping well.

Figure 10. Image wells for a no-flow boundary (upper) and a constant head boundary (lower).

For Discharging Image Well: impermeable boundary Q Q

s = s real + s image (16)

s=
Q
4 πT 4 πT
[
[ W( u real ) ] + Q W( u image ) ] (17) Q
Q

s=
Q
4 πT
[ ]
W( u real ) + W ( u image ) (18)
2 2
r S r S (19)
u real = r u image = i
4Tt 4Tt x x

For Recharging Image Well: recharging boundary

11 Physical Hydrogeology 476K and 191C

x x
 s = s real − s image (20)

s=
Q
4 πT 4πT
[
[ W( u real ) ] − Q W( u image ) ] (21)

s=
Q
4πT
[
W ( u real ) − W ( u image ) ] (22)
2 2
r S r S (23)
u real = r u image = i
4Tt 4Tt

Other Well Parameter Definitions

Radius of Influence: R is the radius of the cone of depression for a given time and a measured
drawdown at the pumping well for that time.
Tt
R = 1.5 (10)
S
scalculated s theoretical
Well Efficiency: = or =
sactual s measured

Well Loss: S w = CQ 2 Q = pumping rate sw = well loss C = well loss constant

Q
Total Drawdown: S total = S + Sw = W(u ) + CQ 2
4πT
Measured Drawdown: S measured = S la min ar + S w sw = due to turbulent flow

REFERENCES
Cooper, H. H., and Jacob, C. E., 1946, A generalized graphical method for evaluating formation
constants and summarizing well field history: Trans., American Geophys. Union, v.27,
p. 526-534.

Driscoll, F. G., 1986, Groundwater and Wells: Johnson Division, St. Paul, MN, 1089p. An
extensive, practical treatise.

Theis, C. V., 1935, The relation between the lowering of the piezometric surface and the rate and
duration of discharge of a well using groundwater storage: Trans., American Geophys. Union,
v.2, p. 519-524.

Walton, W. C., 1970, Groundwater Resource Evaluation: McGraw-Hill, New York, 664p.

Questions:

12 Physical Hydrogeology 476K and 191C

1. (20pts) Time-drawdown data collected at a distance of 100 ft. from a well pumping at a rate
of 192,000 cubic ft/day are as follows:

t (min) s (ft)
1 3.8
2 5.2
3 6.2
4 7.0
5 7.6
6 8.3
7 8.8
10 10.0
20 12.2
40 14.0
80 15.8
100 16.4
300 19.0
500 20.2
1000 21.6

Calculate transmissivity (T) and storativity (S) in an excel spreadsheet. Plot a graph showing
time versus drawdown. Refer back to distance drawdown test.

2. (15 pts) A 24-inch diameter well is located in an aquifer with a transmissivity of 1340 ft2/day
and a storativity of 0.05. A fully penetrating stream is located 1000 ft from the well. The well is
pumped at a rate of 96,000 ft3/day. Calculate the drawdown (1) at the stream and at the midpoint
between the well and the stream after one year of pumping.

3. (15 pts) The figure below gives some pumping test results as measured in an observation
well. The inflection in the drawdown curve is the result of a boundary some unknown distance
from the observation well. Given the following information:

• Transmissivity = 10,000 ft2/day

• Storativity = 0.0001
• Distance from observation well to the pumping well (r1) = 1000 ft.
• The time (t1) at which s1 is measured is 0.1 day
• The time (t2) at which s2 is measured is 10 days
• s2 = 2*s1 (the drawdown at t2 is twice the drawdown at t1)

Find the distance from the observation well to the image well that is responsible for this
behavior.

BONUS Questions:

13 Physical Hydrogeology 476K and 191C

4. (2 pts) Consider the following graph of drawdown versus time. What possible reasons could
cause the graph to deflect at time tx.

5. (2 pts) Assuming the pumping rate is continuous and constant, what might cause the
deflection numbers a, b, and c?

14 Physical Hydrogeology 476K and 191C