TM 5-815-5/AFM 88-5, Chap 6/NAVFAC P-418
correlation between D10 and kh shown in figure C-4
charge should include the exact time that the observa-
was based on laboratory sieve analyses and on such
tion was made.
well-flow tests.
(3) As changes in barometric pressure may cause
the water level in test wells to fluctuate, the baromet-
d. General test procedures.
ric pressure should be recorded during the test.
(1) Before a pump test is started, the test well
(4) When a pumping test is started, changes in
should be pumped for a brief period to ensure that the
water levels occur rapidly, and readings should be
pumping equipment and measuring devices are func-
taken as often as practicable for certain selected pie-
tioning properly and to determine the approximate
zometers (e.g., t = 2, 5, 8, 10, 15, 20, 30, 45, and 60
valve and power settings for the test. The water level
minutes) after which the period between observations
in the well and all observation piezometers should be
may be increased. Sufficient readings should be taken
observed for at least 24 hours prior to the test to deter-
to define accurately a curve of water level or draw-
mine the initial groundwater table. If the groundwater
down versus (log) elapsed pumping time. After pump-
prior to the test is not stable, observations should be
ing has stopped, the rate of groundwater-level recov-
continued until the rate of change is clearly estab-
ery should be observed. Frequently, such data are im-
lished; these data should be used to adjust the actual
portant in evaluating the performance and charac-
test drawdown data to an approximate equilibrium
teristics of an aquifer.
condition for analysis. Pumping of any wells in the
vicinity of the test well, which may influence the test
C-3. Equilibrium pumping test.
results, should be regulated to discharge at a constant,
uninterrupted rate prior to and during the complete
test.
is pumped at a constant rate until the drawdown in the
(2) Drawdown observations in the test well itself
well and piezometers becomes stable.
are generally less reliable than those in the piezom-
b. A typical time-drawdown curve for a piezometer
eters because of pump vibrations and momentary
near a test well is plotted to an arithmetical scale in
variations in the pumping rate that cause fluctuations
figure C-5 and to a semilog scale in figure C-6. (The
in the water surface within the well. A sounding tube
computations in fig C-6 are discussed subsequently.)
with small perforations installed inside the well screen
Generally, a time-drawdown curve plotted to a semilog
can be used to dampen the fluctuation in the water
scale becomes straight after the first few minutes of
level and improve the accuracy of well soundings. All
pumping. If true equilibrium conditions are estab-
observations of the groundwater level and pump dis-
lished, the drawdown curve will become horizontal.
(Courtesy of UOP Johnson Division)
Figure C-5. Drawdown in an observation well versus pumping time (arithmetical scale).
C-5
Previous Page
