Anyone with WQ Calibrator experience?

[MikeK]

Hello.
Just joined the forum today, although I have been using Innovyze's products for approximately the last 10 years. Better late, etc.
Trying to use InfoWater's WQ Calibrator to help our client identify the best location(s) for chlorine booster stations in their distribution system. Tessie's looking at my model now, because it won't run and I probably didn't set it up correctly.
If anyone has experience with the WQ Calibrator and can relay your experience, ins-and-outs, etc., I would be grateful.

[Patrick Moore]

MikeK,

As it unfortunately looks like there has not been any responses to your thread as of yet, I thought I would chime in.

From my experience there has not been a high number of utilities active in modeling water quality of constituents in the last five or so years, although they are being done. I think part of that is the amount of effort it takes to gather the field data needed to complete the WQ calibration which , in my own experience, has been often more than a utility has been able to invest time and resources to given the information they have been able to glean from simply running water age simulations. Water age simulations have been often used as a surrogate to identify areas where water quality problems could occur, and requires simply a reasonably calibrated hydraulic model to use. It has become viewed that many water quality concerns such as a low chlorine residual or growth of disinfection byproducts are related to water age and thus if operations can eliminate high water age, they have a good chance of eliminating the water quality issue. It's not perfect, but this simplistic approach has become widespread in the industry and may be why so few have chimed in with experience using the tool.

However, for utilities that are investing in a full constituent water quality model such as for tracking the chlorine residual in a system, the most challenging part of creating a WQ model is that the pipe decay coefficients must be developed using field test data and are fit to the system so that the field decal and the model decay produces results within the calibration tolerance. Thus, in WQ calibrating users adjust the pipe wall coefficients to best match the field and model predicted data. That one step can often feel more like an art that a science as there are not a lot of guidelines to go by in how best to complete it. This is where the WQ Calibration tool can be very helpful as it allows a user to create groupings of pipes to change the wall coefficients of pipes in that group together to get a best fit to the field data using a more automated process.

There is still a lot of engineering judgment that is involved in calibrating any model, especially WQ models. It is always recommended that any calibration use as many different calibration points as possible, as this allows for the best fit when using the WQ Calibrator tool. You may or may not be aware that the Innovyze User Guide for the WQ tool was installed during the software installation and that it contains a very good tutorial in how to use the feature in the software. These User guides are usually located in the C:\Program Files (x86)\InfoWater\Help folder on a windows machine. The file is called "InfoWater WQ Calibrator UG.pdf". Section 3 is the quick start tutorial and uses an example file already installed on your machine to help the user walk through how it works. These user guide tutorials are often very similar to the basic feature training tools used in the in person training classes and are a good way to get familiar with the tools.

This User Guide recommends that for pipe groups that pipes of similar age, material, and diameter be grouped together. The model will assume all pipes in a group will have identical wall coefficients. Those groupings are what seem to be the most important factor in the calibration and the software can utilize up to 50 different pipe groupings. The use of genetic algorithms in the calibration software allow a wide variety of values to be used to help find the best values by use of brute force to examine lots of different options. This would be difficult to do by hand. I have always thought that selection of the groupings is key to making the WQ calibration tool work the best.

Its good to hear Tessie at Innovyze support has been helping you. Please feel free to contact Innovyze support if you need further assistance.

Patrick Moore
Innovyze Support.

[ESYSU]

I want to learn more about modeling water quality as well. I am particularly interested in modeling: water age, chlorine, and trihalomethanes (THMs).

Just to make sure I am understanding everything correctly so far, could you clarify the following:

1. Are the Section 3 steps suppose to be done with different bounds, concentrations, and thresholds for each (water age, chlorine, and THMs)?
Meaning there are multiple sets of pipe wall coefficients depending on what is being modeled? Or is there only one set of pipe wall coefficients that works to model all?
2. Does the WQ Calibrator provide the pipe wall coefficients only? Or does it also account for the bulk coefficient?

Sorry if the questions aren't clear. I am still trying to wrap my head around it.

[Patrick Moore]

Elena,

Thank you for your questions.

For some Background WQ information and how the model works:

The best description of how the Water Quality functions work (i.e a focus on the theory) is often found in the InfoWater User Guide document in Section 4 See section 4.4 on page 4-29), Section 2 describes the capabilities of the software (see section 2.5 to 2.7 on pages 2-8 to 2-20 for Water quality, and Section 3 is a quick start tutorial (The WQ model section starts on page 3-26 and extended Multi species tutorial #4 on page 3-82).

This document is found on your computer in this directory: C:\Program Files (x86)\InfoWater\Help and is called the InfoWater Users Guide.pdf.

For answering your specific questions:

1. Are the Section 3 steps suppose to be done with different bounds, concentrations, and thresholds for each (water age, chlorine, and THMs)?
Innovyze Response - > The tutorial covers specific steps for multi species in Extended tutorial 4 on page 3-82
For Any specific WQ model you would need to develop initial conditions, pipe reactions for each constituent and bulk decay for each constituent for each water source. Pipe reaction rates are typically "fit" to a model using field test data as they are not measurable in the field.

I will have to research this further to confirm, but if you are doing Multi Species modeling you can create complex bulk relationships and complex pipe coefficient relationships that can be used. But the typical Multi species model is modeling constituents that interact with each other and thus should be modeled together. It sounds like you need to define bulk and pipe parameters for each constituent but in Multi-Species modeling they are defined slightly different than single constituents are using the Multi Species link parameter editor discussed in Step 6 below.

From the help file search for "Multi-Species Water Quality Modeling" for the overview in the help file on this topic. Here is what we describe in the help file for the basic methodology for a Multi Species Model:

Methodology

The outline below defines the basic process for
building and solving a Multi-Species Water Quality model.

1. Use Tools->Multi-Species Model command from the
   main menu to start the Multi-Species Model Dialog Box
   with which to create and edit multi-species water quality models. There are five
   system example models that are already built into the software and these models
   are named CHLORAMINE, DBP, INACTIVATION, PSM1 and PSM2. The details of these
   models are described in Section 4 of the User's Guide.
2. To start, define the species. Use the Multi-species
   editor - Species Tab to define the various species to be modeled. The parameters
   required for each specie include, the Units to be used, the Type of reaction
   (Bulk or Wall), the reaction expression
   type for pipes and tanks (P-Type and/or
   T-Type) and the term name of the reaction
   expression to be used for the pipes and tanks (P-Term and/or T-Term). The mathematical expressions that the
   term names represents are input in step 4.
3. Define the model coefficients using the
   Multi-species Editor - Parameter Tab and the Constant Tab to define the
   coefficient names and values.
4. Input the mathematical expressions of the
   differential-algebraic equation set that defines the reaction dynamics. Use the
   Multi-species Editor - Term Tab and the included Expression Builder to define the mathematical
   expression terms.
5. Set the model options using the Multi-Species Editor - General Tab.
6. For the elements involved, specify the source and initial water quality of
   each specie by using the and the model
   parameters by using the Multi-Species Link Parameter
   editor . All of the above
   toolbar buttons are available from the Model
   Explorer - Attribute Tab toolbar.
7. Select the Multi-Species model of interest in the Quality Tab of
   the Simulation Options and adjust the Thread Allocation setting if required from
   the Advance
   Tab of the Simulation Options.
8. Use the Run Manager
   to start the simulation.
9. When the Multi-Species model is finished, a report for each of the species
   defined by the Species Tab will be included in the usual hydraulic water quality
   results report for the model data elements. See the Output
   Report/Graph Manager for additional details.

Meaning there are multiple sets of pipe wall coefficients depending on what is being modeled? Or is there only one set of pipe wall coefficients that works to model all?
Innovyze Response - see above


2. Does the WQ Calibrator provide the pipe wall coefficients only? Or does it also account for the bulk coefficient?

Innovyze Response - typically since bulk decay coefficients are developed using lab tests they don't have to be "fit" like pipe decay values typically are. That is why the WQ calibrator is set up to help make the best "fit" for pipe decay rather than bulk decay. If you have multiple sources you can assign bulk decay for the entire system based on source tracing as the combined bulk decay is a function of the proportion from each source.

I want to learn more about modeling water quality as well. I am particularly interested in modeling: water age, chlorine, and trihalomethanes (THMs).

Just to make sure I am understanding everything correctly so far, could you clarify the following:

1. Are the Section 3 steps suppose to be done with different bounds, concentrations, and thresholds for each (water age, chlorine, and THMs)?
Meaning there are multiple sets of pipe wall coefficients depending on what is being modeled? Or is there only one set of pipe wall coefficients that works to model all?
2. Does the WQ Calibrator provide the pipe wall coefficients only? Or does it also account for the bulk coefficient?

Sorry if the questions aren't clear. I am still trying to wrap my head around it.