Getting Started

To open the PFAS_SAT, do the following steps:

1- Open the conda command prompt.

2- Activate the environment:

conda activate PFAS_SAT

3- Open python to run PFAS_SAT:

python

4- Run PFAS_SAT in python:

import PFAS_SAT as ps
ps.PFAS_SAT()

PFAS SAT

Here is the PFAS_SAT start screen (Fig. 1). The user interface includes the following tabs:

1. Start

2. Waste Material Properties tab: Shows the default data for waste materials, and you can edit, import, or export the data through this tab.

3. Process Models tab: Shows the process models and type of waste materials that each can treat.

4. Process Models Input Data tab: Shows the default input data for process models, and you can edit, import, or export the data through this tab.

5. Define System tab: In this tab, you can create a scenario and define which processes are used to treat the waste material.

6. Flow Analysis tab: This tab shows the Sankey diagram and data for PFAS flows in the treatment system.

7. Monte Carlo Simulation tab: In this tab, you can define/change uncertainty distributions for the input data and perform a Monte Carlo simulation.

8. Sensitivity Analysis tab: In this tab, you can perform sensitivity analysis on your system and study the effect of input parameters on the PFAS fate.

You can create a new project by clicking on the Start New Project button [9].

Fig. 1 Start tab

Create new project

Waste Materials Tab

The first step of creating a new project is to check the default data for the waste material that you want to simulate. If you have the data for waste material (e.g., PFAS concentration), you can update the default data in the Waste Material Properties tab. Fig. 2 shows how you can edit, import, or export data for waste materials. If you have data in csv format, you can import it by using the Browse [1] and Import [2] Buttons.

Note

If you are importing data, the data should be in the csv format and have the same column names as the default data files. We suggest to copy our data files and edit them to keep the structure.

Warning

Some of the waste materials like Contaminated Water, Contaminated Soil, etc. are too case dependent and the user should only create a scenario with them if he/she has the data for PFAS concentration.

You can change the waste material via drop-down list [3] and its help button [4] displays more information about the sources of waste material and potential treatment options. You can see the uncertainty distributions for data by checking the checkbox [5]. The Uncertainty Distribution Help button [6] shows you how to read the distributions. You can change the data in the table [7] and don’t forget to click the Update button [9] otherwise your changes will be lost. You can remove all the uncertainty distribution by the Clear Uncertainty button [8]. Export button [9] lets you export the data in ‘csv’ format and then you can import your data file next time that you want to use PFAS_SAT. When you are done with data for waste materials, you can go to the next tab by clicking on the Define Process Models button [11].

Fig. 2 Waste Materials tab

Process Models Tab

The Fig. 3 shows the Process Models tab. You can select the process model by the drop-down list [1] and see the waste materials that the selected process can treat/accept. The help button [2] displays more information about how we modeled the process. You can import your input data (csv file) for the process models by the Browse button [3] and don’t forget to check the User Defined radio button if you are importing data. You can change the types of waste that each process model accept and click the Update button [6] to save them. You can go to the next tab by clicking on the Check Process Models Input Data button [7].

Fig. 3 Process Models tab

Process Models Input Data Tab

You should check the input data for the process models, especially the operational input data that are case dependent (e.g., Windrow height or width in the composting process). This tab is similar to Waste Materials Tab (Fig. 2) and you can edit, import, or export the data.

Warning

If you are changing the input data for process models, check the minimum and maximum range. The model doesn’t check that your data is valid. Wrong data can results in wrong results or produce an error.

Warning

Don’t change the name or unit of the parameters unless you are also updating the code.

Note

You can develop new process models for the technologies that are included in the current version of PFAS_SAT or revise current models. See Develop New Process Models.

Fig. 4 Process Models Input Data tab

Define System Tab

Fig. 5 shows the Define System tab. In this tab, you should select the starting waste material [1] and it’s mass flow [2] that you want to simulate and click the Setup Scenario button. Treatment Processes frame [5] will show all the potential processes that can be used to handle the starting waste material and the residuals/products from the downstream processes. If some of these processes don’t exist in your region/state, deselect them so they won’t be used in the treatment network.

Warning

The treatment network should include at least one process for the staring material and each of the products/residuals produced in the downstream processes. The pop-up warnings will help you add the required processes.

After selecting the treatment options [5], click on the Create Network button [6] to create the Treatment Network Parameters table [7]. In the table, you should allocate the starting materials and other products/residuals to the treatment processes and then click on the Update Network button [8]. After updating the network parameters, you will see the network graph [9].

Now you have a complete scenario, and you can do further analysis include Flow Analysis, Monte Carlo simulation, or Sensitivity analysis on your scenario.

Warning

You should check and define the allocation of waste materials to treatment processes in Treatment Network Parameters. The sum of the fractions (allocations) for each waste/product should be 1 otherwise, you will get the following pop-up error that tells you which of the allocations are wrong.

Fig. 5 Define System tab

Flow Analysis

All the data for PFAS flows in the treatment network are added to the PFAS Inventory which can be seen in the Flow Analysis tab (Fig. 6 [2]). The currently included endpoints are air, water, soil, landfill storage, reactivated activated carbon, destroyed/mineralized PFAS, and injection well. The row indexes for the Inventory are:

1. Flow name: Name of the flow in the process models.

2. Source: The process that produces the flow.

3. Target: The process that treats/receives the flow.

4. Unit: Unit for the PFAS flows.

5. PFAS: In the current version, 10 types of PFAS are tracked through the system.

Note

When you are doing the flow analysis, PFAS_SAT will show a pop-up window that displays the error in the PFAS balance. As some of the flows are loop (e.g., Sending the landfill leachate to WWT produces WWT solids that will be dumped in the landfill and result in landfill leachate generation, which was the starting material), PFAS_SAT is using the Cut-off approach to break the loop when the PFAS flows are less than Cut-off. So small errors in PFAS mass balance (< 5%) are acceptable. You can change the Cut-off from the menu/tools/Options.

Note

The Sankey is interactive, and you can change the view, export png, or see the labels for flow. Check the toolbar for Sankey Diagram. The source file of the Sankey diagram is also saved in html format in the directory that you opened the PFAS_SAT.

Fig. 6 Flow Analysis tab

Monte Carlo Simulation

You can define the uncertainty distributions for the input data and then run the Monte Carlo simulation to study the effect of uncertainty/variability in the input data. Fig. 7 [2] shows the Monte Carlo Simulation tab. Changing and updating the data is similar to the Waste Materials Tab and Process Models Input Data Tab. You should enter the number of simulations in the spin box [8] and click the Run button. The progress bar [9] will show you the progress and when the simulation is done, you can save the results in csv format or analyze the results to find the distributions or correlations by clicking on the Show Results button [10].

Fig. 7 Monte Carlo Simulation tab

Monte Carlo Simulation Results

Results from the Monte Carlo simulation will be shown in a new window when the simulation is done. You can view the results from the Monte Carlo simulation in the Data tab (Fig. 8), plot the results based on the inputs (Fig. 9), plot the distributions (Fig. 10) and calculate the correlations between the results and input data (Fig. 11) to find the most important parameters.

Fig. 8 Monte Carlo Simulation Results Table

Select the parameter for the x-axis [1] and y-axis [2] from the drop-down lists and then click the Update button to see the plot [4]. You can plot scatter and hexbin [3]. Use the toolbar [5] to change the view, export, or edit the plot.

Fig. 9 Plot Monte Carlo Simulation Results

Select the output from the drop-down list [1] and then click the Update button [3] to see the distribution plot [4]. You can plot a histogram, box plot, or density plot [2]. Use the toolbar [5] to change the view, export, or edit the plot.

Fig. 10 Plot Distributions of Monte Carlo Simulation Results

Select the endpoint from the drop-down list [1] and then click the Update button to see the correlation plot [2]. Use the toolbar [3] to change the view, export or edit the plot.

Fig. 11 Monte Carlo Simulation Results (Correlation data tab)

Sensitivity Analysis

You can perform parametric sensitivity analysis in the Sensitivity Analysis tab (Fig. 12) and export the results or plot them (Fig. 13). ‘Model’, ‘Category’ and ‘Parameter’ drop-down lists [1-3] will help you to choose a parameter for your analysis and then you can see more info in the Parameter Information screen[4] which will guide you to choose a realistic range for your analysis. Increase the number of steps, if your parameter has a non-linear effect. You can track the sum of the all PFAS types include in SAT or select only one type (e.g., PFOA) [6].

Fig. 12 Sensitivity Analysis tab

Select the parameter for the x-axis [1] and y-axis [2] from the drop-down lists and then click the Update button to see the plot [3]. Use the toolbar [4] to change the view, export, or edit the plot.

Fig. 13 Plot Sensitivity Analysis Results

Uncertainty Distribution

Tha stats_arrays package is used to define uncertain input parameters for the process models and waste materials. The table below shows the main uncertainty distributions that are currently used

Name	uncertainty_type	loc	scale	shape	minimum	maximum
Undefined	0	static value
No uncertainty	1	static value
Lognoraml	2	\(\boldsymbol{\mu}\)	\(\boldsymbol{\sigma}\)		Lower bound	Upper bound
Normal	3	\(\boldsymbol{\mu}\)	\(\boldsymbol{\sigma}\)		Lower bound	Upper bound
Uniform	4				Minimum	Maximum
Triangular	5	mode			Minimum	Maximum
Discrete Uniform	7	mode			Minimum	upper bound

Guideline to define uncertainty

1. Normal distributions (ID = 3): When there is sufficient published data.

2. Triangular distribution (ID = 5): When values are based on expert opinions with a reasonable value for the mode.

3. Uniform Distribution (ID=4): When only the range is known without preference for mode.

4. Lognormal distributions (ID=2):  When only one value is available or there is significant data and the value must be non-negative.

5. Discrete Uniform (ID=7): For True/False (0,1) parameters.(min=0,max=2).

Note

In Normal distribution, if the mean is too close to lower or upper bound (mostly for parameters that are fractions), use the triangular distribution.

Note

In Lognormal distribution, if the parameter is related to the emission factors, sigma should be in the range of 0.04 to 0.09 based on the quality of the data.

See also

For more information about distributions check stats_arrays website.

Develop New Process Models