The contaminant treatment system (CTS) package is contained in the \mf6cts\mf6cts.py:Mf6Cts class. Several examples usages can be found in the autotest\cts_mf6_test.py script - these tests are "self-contained", in that the precompiled MODFLOW 6 binary library files, as well as the required python dependencies are all present in this repo, so one should be able to "just run" these tests ("should")...
There is a simple example notebook in examples/. Also available on binder:
The Mf6Cts class is designed to work with separate flow-then-transport
modeling scheme. As such, the flow model and transport models should be
in separate directories. Additionally, the FMI process is expected to
be used, and, if the MAW package is used in the flow model, then the
MWT package is expected in the transport model. Users are expected to
make the flow-model binary output files needed by the FMI available.
The CTS input file follows the standard MODFLOW 6 input format style. Through the use of period blocks, a CTS system can be (re-)defined for each stress period in the model. Each period block must identify the CTS system that is being defined. In this way, numerous different CTS systems can be (re-)defined at the stress period level. An optional efficiency can be specified; efficiency ranges from 0.0 (no dissolved-phase mass is removed) to 1.0 (all dissolved-phase mass is removed). If not specified, efficiency is assumed 1.0---the CTS instance removes all mass.
Within each period block, information related to the components of the CTS must be defined. These include the component package type (e.g. "wel", "maw", etc), the component instance of the package type (as defined in the flow model nam file, e.g. "wel_1", "maw_cts", etc), whether the component is an extractor ("out") or an injector ("in"), and the indexing information for the component. For non-MAW type components, if the model is a structured grid, then the indexing information is the layer, row and column location of the component; if the model is unstructured, then the indexing information should be node-number of the component. For MAW-type components, the indexing information is the "wellno" value in the MAW package.
Below is an example CTS input file:
begin period 2 cts 1 efficiency 0.361
wel wel_0 out 3 21 21
wel wel_0 out 3 21 23
maw maw_0 out 2
wel wel_0 in 1 1 1
maw maw_0 in 1
end period 2 cts 1
begin period 2 cts 2 efficiency 0.909
wel wel_0 out 3 20 3
wel wel_0 in 1 33 1
wel wel_0 in 1 33 33
wel wel_0 in 1 1 33
end period 2 cts 2
begin period 5 cts 1 efficiency 0.925
maw maw_0 out 2
maw maw_0 in 1
end period 3 cts 1
begin period 7 cts 2 efficiency 0.222
wel wel_0 out 3 20 3
wel wel_0 in 1 33 1
wel wel_0 in 1 33 33
wel wel_0 in 1 1 33
end period 3 cts 2
In this example file, we see that for stress period 2, CTS instance "1" has three extraction wells and one injection well, while CTS instance "2" has one extraction well and three injection wells. Using the MODFLOW 6 convention, CTS instance "1" continues from stress period 2 to stress period 5 as it is defined in stress period 2. Similarly, CTS instance 2 continues from stress period 2 to stress period 7 as it was defined in stress period 2. At stress period 5, CTS instance "1" is reconfigured to be a single injector and single extractor system of just MAW-type wells. At stress period 7, CTS instance "2" changes its efficiency value but contains the same injection and extraction wells.
The Mf6Cts class can also be driven from the command line:
python mf6cts.py <config_file.py>
where <config_file.py> is the name of a configuration file that is a
simple python source file listing the required arguments:
-
cts_filename: the name of the cts file -
lib_name: the name of the MF6 shared library file -
transport_dir: the directory holding the transport model files -
flow_dir: the directory holding the flow model files -
is_structured: a boolean flag indicating if the model is a structured grid (a value of1indicates a structured grid) -
flow_output_files: a python list of the flow model output binary files that theFMIpackage is expecting
An example configuration file:
cts_filename=’model.cts’
lib_name=’libmf6.so’
transport_dir=’fivespot_maw_t_api’
flow_dir=’fivespot_maw_api’
is_structured=True
flow_output_files=[’gwf.hds’,’gwf.bud’,’gwf.maw.bud’]
The MODFLOW 6 CTS package write several comma-separated-value (CSV)
files summarizing the performance of the CTS system for the both the
flow and transport models and from both the CTS node and CTS system
perspective. The flow model summary CSV files are written in the flow
model directory (e.g. flow_dir) and are named
"gwf_cts_flow_node_summary.csv" and "gwf_cts_flow_system_summary.csv";
these files summarize the flow-model extraction and injection aspects of
CTS instance performance and include information about the requested and
actual rates and cumulative volumes extracted and/or injected for each
CTS system and its nodes across all flow solution stress periods and
time steps.
The transport-model summary CSV files are written in the transport model
directory (e.g. transport_dir) and are named
"gwt_cts_node_summary.csv" and "gwt_cts_system_summary.csv"; these files
summarize the transport-model extraction and injection aspects of CTS
instance performance and include information about the extracted and
injected dissolved-phase mass as both a rate and cumulative mass for
each stress period and time step, as well as cumulative mass across the
entire solution period. These files also contain information about the
blended injected concentration if an efficiency less than 1.0 is used.
If MAW-type boundary conditions are included in CTS instance, then a MAW-specific summary file is written in the transport model directory named "gwt_maw_node_summary.csv". This file summarizes the dissolved-phase transport aspects of individual nodes that comprise each MAW boundary condition, including flow rate, concentration, cumulative volume and cumulative mass for each node in each MAW boundary condition that is included in a CTS instance.
A very simple test is included in the code repository to verify the functionality of the Mf6Cts package; the files needed to run this test are included in the "/verification" directory. The test consists of a cross-section model with 1 layer, 1 row, and 5 columns. The first column contains a GHB-type boundary with a specified concentration of 1.0. The last column contains a GHB-type boundary with a specified concentration of 0.0. In column 2, an extraction captures dissolved phase that enters the model domain. Column 3 in active. Column 4 contains an injection well that re-introduces extracted water after the CTS instance has "treated" it. The model has 3 stress periods. Stress period one has a CTS treatment efficiency of 0.0 (no mass removal), stress period 2 has a treatment efficiency of 0.5 (half mass removal), and stress period 3 removes all mass (efficiency equals 1.0). Additionally, the WEL package uses the "auto-flow-reduce" functionality such that the requested extraction rate cannot be satisfied by the model.
To execute the verification test, first copy the "./mf6cts/mf6cts.py" source file into the ./verification directory. Then copy the MODFLOW-6 shared library file from the appropriate "/bin" directory for the users operating system into the both the "./verification/fr1_test_api" and "./verification/fr1_test_t_api" directories and change the value of lib_name in "./verification/config_file.py" to correspond to the name of the shared library file. Finally at the terminal within the "./verification" directory, execute "python mf6cts.py config_file.py"
After running the verification test, users can check that the Mf6Cts functioned as expected in the following ways:
- Checking the locations of the extraction and injection wells: In the file "./verification/fr1_test_api/gwf_cts_flow_node_summary.csv" and "./verification/fr1_test_t_api/gwt_cts_node_summary.csv", the layer-row-column information for each CTS-associated WEL-type boundary is listed. These should correspond with the layer-row-column information in the CTS input file ("./verification/fr1_test_t_api/model.cts")
- Checking the flow-model extraction/injection rate balancing: Inspecting the flow-model listing file ("./verification/fr1_test_api/gwf.lst"), users can check that the cumulative and incremental values for WEL "IN" and "OUT" for each stress period are within 0.0001 units. Inspecting the CTS flow node summary, users can see the
requested_rateandactual_ratefor the extraction and injection wells. Theactual_rateshould be less than therequested_rateand the absolute value for theactual_ratefor extraction well and injection well should be within 0.0001 units. - Checking the transport-model treatment efficiency: Inspecting the transport-model listing file ("./verification/fr1_test_t_api/gwt.lst"), users should see that the simulated concentration for node 4 (the injection well location) should equal the extraction well concentration (node 2) with a value of 1.0; the simulated WEL "IN" mass should equal the simulated WEL "OUT" mass in the stress period 1 mass budget report. For stress period 2, the simulated concentration for node 4 should equal 0.5 (half of the extraction concentration) and the WEL "IN" MASS rate should be half of the "OUT" mass rate in the mass budget report for stress period 2. For stress period 3, the simulated concentration for node 4 should be near numerical zero and the WEL "IN" rate for stress period 3 should 0.0 in the mass budget report.
- read the CTS input file and instantiate the MODFLOW-6 flow model through the API
- for each stress period
- for each time step
- for each solution iteration except the first
- for each CTS instance
- get the requested extraction and injection rates for each WEL/MAW entry in this CTS. The entries for each CTS is stored in the CTS input file, while the rate information is stored in the corresponding MODFLOW-6 input file
- get the corresponding last iteration simulated extraction and injection rates through the API as an array stored in memory
- if the total simulated extraction rate is less than the total requested extraction rate, scale the requested injection rate for all injectors proportionally such that the total requested injection rate is equal to the total simulated extraction rate.
- reset the requested injection rate array pointer in memory through the API so that the next iteration will use these rates
- for each CTS instance
- once solution has converged or maximum number of iterations is reached, record CTS results to the node and system summary CSV files
- for each solution iteration except the first
- for each time step
- read the CTS input file and instantiate the MODFLOW-6 transport model through the API
- for each stress period
- for each time step
- for each solution iteration
- for each CTS instance
- accumulate the mass removed by each extractor by summing the concentration times flow rate product for the entries in this CTS. The CTS entries are stored in the CTS input file; the simulated concentration array and flow-rate array are accessed through the API.
- apply the treatment efficiency by multiplying the requested treatment efficiency times the total mass removed by extractors to yield the mass removed by the treatment system and the mass to be injected. The treatment efficiency for each CTS instance for each stress period is stored in the CTS input file.
- sum the total injection volume as flow rate times time-step length for all injectors.
- calculate the injection concentration as injection mass divided by total injection volume
- reset the injection well concentration array pointer in memory through the API so that the next iteration will use the calculated injection concentration resulting from the application of the treatment efficiency
- for each CTS instance
- Once the solution has converged or the maximum number of iterations is reached, record the CTS results in the node and system summary CSV files
- for each solution iteration
- for each time step