Slug Test Analysis
The program 'SLUGT2' is an updated version of program 'SLUGT' and computes hydraulic conductivity values based on the analysis of slug-test data.
commonly used for slug-test analysis included in SLUGT2 are:
(1) Method of Cooper, Bredehoeft and Papadopulos;
(2) Method of Bouwer and Rice; and
(3) Modified Hvorslev Method.
In addition, effects due to air entrapped in the completion region of a piezometer or well are estimated using the method of Keller and van der Kamp as a computation option. The program is intended to be run in batch mode, but there is an option to enter the input data from the keyboard as prompted one by one on the monitor. The batch mode involves the preparation of a data file that may include data for a series of slug tests. Data input and output may be in metric (SI) units (centimeter-meter) or inch-feet units. SLUGT2 executes in a Microsoft Runtime Windows Environment (MRWE) with a windows-style interface.
(1) a detailed results file including input and output values;
(2) an Excel-compatible table summarizing the results for all the tests analyzed in the run, and
(3) an Excelcompatible output file containing displacement (H) vs. time data for each case to facilitate preparation of graphs of log(H) vs. time.
The user may find it valuable to plot the reduced data to determine if certain data points should be excluded in a possible second run. In many cases the observation points all fall close to a best-fit straight line and the results of the slug-test analysis are ready for use. In other cases, the early-time points may fall on one straight line, and the later-time points on another straight line, the latter often having a flatter slope. In that case, the user will have to decide whether or not to rerun the program using only the earlier-time points or the later-time points for computation of K and T.
Distribution of the public domain SLUGT2 program includes source code, executable image, and example data sets. The user's manual includes installation instructions and program documentation.
Modifications to SLUGT in Developing SLUGT2
1. General Changes in Code
a. SLUGT2 now accommodates the input of data either in metric units (meters and centimeters) or the English system (feet and inches). In the case of the metric option, output is in cm/sec for K values and cm2/sec for T values, while for the English system K values are in ft/day and T in ft2/day.
b. SLUGT2 now creates a new output file ('Slugt2_summary.out') that gives a concise summary of the permeability results from all the tests analyzed in a given run, as well as the H0 used or calculated for each case. This output can conveniently be imported into Excel to readily produce a summary of results table. The name of this file can be suitably renamed by the user if desired.
c. Improved output in the main results file so that each method is clearly
labeled and the results are readily comparable.
d. Eliminated unnecessary space in output to save paper, etc.
e. The code itself is laid out in a more organized way with a separate
subroutine for each of the three methods for calculating the T and K
f. SLUGT2 can accommodate data either as depth-to-water values or as actual displacement (H) values. In the former case, the program computes the H values based on the value of the depth to static water level input by the user.
2. Changes to Code in Calculating T using Cooper, Bredehoeft and Papadopulos (CBP) Method
a. Modified criteria for calculating H0. Previous method on occasion resulted
in an unrealistically high value for H0. Briefly, H0 is first estimated by
extrapolating to time zero a linear-regression line based on the earliest
points. However, if this value is found to be greater than the H0
computed from the slug volume, then H0 is set equal to the value
computed from the slug volume. Finally, if any of the displacement (H)
values from the data points is found to be higher than the current H0,
then H0 is set equal to 1.1 times the highest H value in the data.
b. Added new input parameter—H0, to permit user to explicitly specify what
H0 he or she may actually want. This is important if second or third runs
are necessary in which data points deviating from the best-fit log(H) vs
time line might be eliminated. In such a case one can input as H0 that
which was calculated/used from the first run. If H0 is input, then this
overrides the calculation described above.
c. Eliminated Trange/Tbar as a criterion for determining the best-fit curve in the CBP method. This was because developer believes this is not as reliable a method as using the minimum root mean square (RMS) approach. Now the program uses only the RMS method to select the alpha, and hence the T, that best fits the data.
3. Changes to Code in Calculating K using Bouwer and Rice (B&R) and Hvorslev Methods
a. Applied corrected calculation for rc for case when the water level is within screen, and applied it to the Hvorslev method as well as to B&R Method. Whatever rc is used is indicated in the main results output file. Established the criterion that whenever more than half of the rise (fall) of water level during tests is within the screen, then the effective rc is computed, as suggested by Bouwer, as: rw (corrected) = sqrt(rw**2. + por*(rh**2. - rw**2.)), where rw is radius of the screened or open interval of the well, rh is the radius of the drilled hole, and por is the porosity of the filter pack.
b. Improved method for calculating effective screen length (required for the B&R and Hvorslev Methods) when the change in the head is all or partially within screen. When the head change occurs within the screen, an average of the wetted portion of the screen during the test is used as the effective screen length.
c. Provided additional case for the NAVFAC method (as a sub-set under the Hvorslev Method) so that all three cases related to differing degrees of partial penetration are now included.
d. Set rw equal to radius of drilled hole, rh, in B&R and Hvorslev's method, per Butler (1997). This was not applied to the CBP method as rw in that case is only used in calculating the storativity S.
e. Changed method of computing the values of A, B and C in B&R method. In program SLUGT, the values were interpolated from a set of data points obtained by visually reading off the values for these parameters in the original paper by Bouwer and Rice (1967). Now in SLUGT2 the values for A, B and C are computed using analytical best-fit equations given by D. Van Rooy (1988) as given in Butler (1997).
f. As used in SLUGT2, the slope of the log(H) vs. time plot for the B&R and Hvorslev methods is computed from the linear regression on all the data points, without successively eliminating later data points (as was done in SLUGT). As mentioned above, the user should plot log(H) vs time for each analysis to select what points to eliminate, if any, in a second run.
CLICK HERE TO DOWNLOAD
Author: A.C. Mills (AMEC E&I, Inc.)