MODPUMP - MODFLOW-simulated pumping test model

MODPUMP Categories: flow models - saturated zone, pumping test programs, MODFLOW programs

MODPUMP Detailed Description


The MODFLOW SIMULATED PUMPING TEST model (MODPUMP) is a computer program that allows calculation of aquifer parameters for single or multiple (up to three) layer aquifer systems by simulating field pumping test data. The program allows calculation of hydraulic conductivity, storage coefficient, specific yield and vertical leakance for up to three-layer aquifer systems. The aquifer parameters are calculated by trial and error using graphical data points matching techniques where the observed drawdown at various time periods from a field pumping test are graphically compared to MODFLOW-simulated drawdown. MODFLOW refers to "A Modular Three-Dimensional Finite-Difference Ground-Water Flow Model," McDonald and Harbaugh, USGS, 1984. The MODPUMP program allows the user to input initial trial hydraulic parameters in one to two minutes in a user-friendly environment (total of four windows of automatically prompted input fields) to execute the MODFLOW model from within the program and to graphically compare the observed drawdown with the MODFLOW-calculated drawdown. In general, the MODPUMP program is a preprocessor and postprocessor for the MODFLOW model specifically designed to reduce field pumping test data and calculated calibrated values of the hydraulic parameters of multiple-layer aquifer systems. In multiple-layer aquifers, the layer from which the water is withdrawn and the layers of each observation point (up to four points are allowed) can be specified by the user. This option of vertical specification of pumping and observation wells allows a better definition of the hydraulic behavior of the aquifer system and a more accurate calibration of the groundwater flow model.

The most important aspect of this program is that it allows the user to simulate complete or partial field pumping test data (i.e., variable pumping rates during the test, interruptions of pumping, observation wells installed in layers other than the layer of water withdrawal, limited data points, etc.). Most classical methods of field pumping test data reduction assumes constant pumping rates, depth and screen intervals of all observation wells to be the same as the pumping well, sufficiently long periods of pumping to generate data for curve matching, etc., which are normally difficult to achieve in real-life pumping tests. The MODPUMP program can be used to simulate any field pumping test data since the test data is being simulated by a three-dimensional finite-difference groundwater flow model.

Once the field pumping test results have been simulated to an acceptable level by trial and error using the graphical matching technique and adjusting the hydraulic parameters, the user can produce a hard copy of the simulated hydraulic parameters and the resulting graph of matched data points. The hard copy can be produced on a LaserJet printer, Epson/Citizen printer or in ASCII format. The printout of input and output data will be in a report quality format, sufficient for insertion in a report, as supporting documentation of the pumping test data reduction.

Four data input screens are available for MODPUMP. They are as follows:

The Project Name/Number
prompt line requires specification of a name for the simulation. This is for identification purposes only. This name will appear on the tabular and graphical printouts and graphical view screens.

The Number of Active Aquifers prompt allows specification of the total number of aquifers that will be simulated for the specific field aquifer pumping test conditions. Prior to initiating the model set up, it is necessary to determine the number of aquifers (or layers) to be simulated by MODPUMP. Typically, this will be dictated by the actual subsurface conditions and/or by the depth of pumping well and various observation wells. For example, if the actual aquifer system consists of a uniform soil extending from the surface to a depth of 35 feet and the pumping test data was collected from a pumping well installed to a depth of 20 feet, one observation well should be installed to a depth of 20 feet and a second observation well to a depth of 35 feet (screen from 20 to 35 feet). The model set up should then be with two aquifers, the first extending from surface to 20 feet and the second from 20 to 35 feet in order to characterize the withdrawal and observation properly.

The Anisotropy Factor - Ratio of Kx/Ky prompt allows specification of the directional hydraulic conductivity of the aquifer. If the hydraulic conductivity is the same in the X and Y direction of the grid, specify 1.0. For hydraulic conductivity of 20% higher in the X direction (as compared to the Y direction), specify 1.2, etc. This factor will be applied to all aquifers (layers) simulated by this model.

The Total Number of Simulation Periods During and After Pumping prompt allows specification of the number of points in time for which the model will calculate drawdown at various observation wells for comparison with the observed data. In general, the number of points should be three or more during pumping and three or more after pumping. An excessive number of points will result in long calculation times by the MODFLOW model and the final accuracy of the results is not expected to increase accordingly. It is recommended to use a total of 6 to 10 simulation periods.

The Diameter of Pumping Well Borehole prompt allows selection of the actual size (diameter) of the borehole (not the well casing) into which the pumping well was installed. This is used by the model to calculate the size and proportions of the finite-difference grid system and it is important to accurately specify this value.

The Total Number of Observation Wells prompt allows specification of the total number of observation wells for which drawdown data is available and will be used for model simulation. Wells that exist but are not specifically used for simulation/calibration need not be specified. All observation wells in all aquifer (layers) must be specified herein. The maximum number of observation wells allowed is four.

The Distance from Center of Pumping Well to Observation Well # 1, 2, 3 and 4 prompt allows specification of the precise distance (measured) from the center of the pumping well to the center of the corresponding observation well. The user shall be aware of the orientation of various observation wells in relation to the pumping well when anisotropic conditions are being modeled (see preceding paragraph).

The input prompts of the second screen include all the hydraulic parameters that will need adjustment for the trial and error simulation of the pumping test data with the MODFLOW model. In general, the hydraulic parameters of this screen prompt are self-explanatory except for the specific variations used with the MODFLOW model. In the MODPUMP program, all aquifers are identified as TYPE 3 in the MODFLOW model which means "Confined/Unconfined" aquifers. These types of aquifers are treated as leaky confined when the ground-water level is above the top of aquifer and as unconfined when the ground water recedes below the top of aquifer. Therefore, by specifying the top and bottom elevation for all aquifers, it is possible to simulate all types of field conditions (i.e., unconfined, confined and confined/unconfined). To specify an unconfined aquifer for the top layer, simply specify the top of the aquifer at a sufficiently high elevation (i.e., ground surface or higher). Thin semi-confining layers can be represented by the leakance value and need not be modeled as separate aquifers or layers if the horizontal flow component is negligible (see Leakance Section below). The following is a brief presentation and discussion of each input parameter.

The Average Hydraulic Conductivity is the actual hydraulic conductivity (or permeability) of the corresponding aquifer (or soil layer) which is representative of the average value between the bottom and top of aquifer elevations. For confined aquifers with a given transmissivity, assume an approximate thickness of the aquifer and divide the transmissivity by the assumed thickness to calculate the equivalent hydraulic conductivity.

The Storage Coefficient of Aquifer prompt allows specification of the actual storage coefficient when the aquifer is confined (groundwater level above top of aquifer). The MODFLOW model incorporates conversion of variable transmissivity and specific yield to constant transmissivity and storage coefficients when the aquifer converts from unconfined to confined and vice-versa. Therefore, for this prompt, specify a storage coefficient when the aquifer is confined (i.e., groundwater level or potentiometric level is above the top of aquifer elevation). If the aquifer is unconfined, then specify any value since it will not be utilized by the model.

The Specific Yield or Fillable Porosity prompt allows specification of the actual specific yield, or the fillable porosity. This value is used when the aquifer is unconfined (i.e., the groundwater or potentiometric level below the top of the aquifer elevation). If the aquifer will be confined throughout the pumping test simulation, then specify any value since it will not be used by the model.

The Average Elevation of Top of Aquifer, the Starting Elevation of Groundwater Level, and the Average Elevation of Bottom of Aquifer are self-explanatory except that the top of an aquifer can not be specified at a higher elevation than the bottom of the aquifer above it. If a semi-confining layer is being represented only by the leakance value, the top of the aquifer below can be specified at an elevation lower than the bottom of the aquifer above it. Otherwise, the top of an underlying aquifer should always coincide with the bottom of the aquifer above it.

The Leakance to Aquifer Below prompt allows specification of the leakance value between the center of the aquifer in which it is specified to the center of the aquifer below. This is the MODFLOW convention and it must be clearly understood and properly applied. Although in most applications of this model the leakance value will be calculated by trial and error, its understanding is essential for proper interpretation of calibration trends and efforts of parameter adjustment.

This data input screen prompt allows specification of the aquifer number (or layer number) from which the pumping test water is being extracted or pumped in and where the various observation wells exist. The screen prompt above indicates that the pumping well is located in aquifer 1 of a two-aquifer system. One observation well was installed in aquifer 1 while the second observation well was installed in aquifer 2. Note that the pumping well and the observation wells can exist in any of the active aquifers of the model setup.

The following is the fourth and the last screen prompt of the DATABASE option which allows specification of the length of the stress periods, the background recharge rate, the pumping rate and the observed drawdown at the observation wells.

This window prompt will automatically be repeated for each simulation period. For the first input field "Length of Simulation Period," enter the incremental time in minutes for the corresponding simulation period. Note that all times for simulation periods are incremental times and not cumulative. The length of simulation must be selected from the field test data. Typically, the following rules of thumb can be used to select the total number of simulations and the length of simulations:

1. Use as few simulation periods as possible to minimize computational time while still providing sufficient data points for the graphical curve-matching technique to calculate the hydraulic parameters.

2. Select time increments for which significant drawdown changes have occurred. Do not use repetitive time increments when minimal or no drawdown changes have been observed.

3. Select at least three simulation periods during pumping for better definition of the drawdown curve.

4. Avoid using initial short-term drawdown periods (i.e., initial 20 to 40 minutes). It is more accurate to match the drawdown data in the latter portion of the pumping test.

5. Select simulation periods to match with any change of pumping rates during the test.

The Average Background Recharge Rate is intended to simulate rainfall or artificial recharges to the aquifer being tested during the pumping test periods. Typically, there will not be any recharge and this input field shall be zero. However, on special occasions where significant rainfall recharge or artificial recharge (i.e., irrigation) is present, the rate in inches per day can be specified. The Pumping Rate prompt allows specification of the actual pumping rate from the test well (field measured) in gallons per minute. For pumping tests where water is being injected (pumped in), the recharge value must be entered as a negative value. The next four prompt lines allow specification of the actual observed drawdown at each of the active observation wells. It is important to remember the location of each observation well and its corresponding drawdown data since the program will compare these observed data with the MODFLOW-calculated drawdown extracted from the location and aquifer number specified for each observation well.

Selecting the VIEW option, a box prompt will appear requesting the number of observation well for which the comparison graph is to be generated for viewing. For this prompt, type in the number of the observation well (i.e., 1, 2, 3 or 4), and the program will read the MODFLOW drawdown results and the previously-specified observed drawdown values and generate a graph with two curves. One curve will be for the observed drawdown and the other for the MODFLOW-calculated drawdown. The drawdown values will be plotted against the specified time for visual review and evaluation. If the two curves closely match each other, the hydraulic parameters selected are accurate; if not, the hydraulic parameters need to be modified/adjusted (in DATABASE option) and the program must be rerun (in RUN option). Once the observed- and model-predicted drawdown curves are matching to an acceptable level (Note: exact matching may not be possible, since field conditions are rarely sufficiently uniform to produce a perfect match - an average match of all selected points would be sufficient for most cases), the graphical model results can be printed. When the graphic is displayed on the screen, a menu bar appears at the bottom for printing options. The menu bar allows the selection of three types of printing devices, namely, Epson FX850/Citizen Printers or equivalent, HP LaserJet Printers (must have capacity to print letter type Univers) and Print Screen. The Print Screen option is intended to allow printing with printers that are not compatible with the optional printers provided.

Scientific Software Group    P.O. Box 708188   Sandy, Utah 84070
Phone (801) 208-3011   Toll Free (U.S.) 1-866-620-9214    Fax (801) 302-1160    E-mail
Copyright 1998   Scientific Software Group    All rights reserved