MP & Associates: GPS-X Slideshow

Welcome to the GPS-X Slideshow! This slideshow presents many of the powerful features of GPS-X, and demonstrates how a GPS-X layout is developed from start to finish. Please take a few moments to work through the slideshow; alternatively, you may wish to wait for all of the images to load and then print out a hard copy of the slideshow so that you may review it at your leisure.

This is GPS-X's main window. It contains GPS-X's main menu bar (the buttons labelled "FILE", "VIEW", "BUILD", etc., are drop-down menus which each contain several GPS-X commands); it also contains buttons which bring up other windows (the "LOCATOR" and "PROCESS TABLE" buttons) as well as buttons which modify the drawing mode (the "REPLACE", "DELETE", "MOVE", "REVERSE", and "BLOCK" buttons).

The large area below this row of buttons is GPS-X's "drawing board"; the layout of the treatment plant to be modelled is created here.

For this slideshow, we will be building a layout representing a treatment plant with both primary and secondary treatment; the secondary treatment will be carried out using the activated sludge process, using two parallel liquid trains.

The first step is to bring up the GPS-X Process Table; this is a window which contains all of the objects available in GPS-X (see Slide 3 for a labelled close-up of the Process Table). From the Process Table, objects can be selected and placed on the drawing board. In this slide, we have placed an influent object on the drawing board.

The "LOCATOR" window can be used to select an area of the drawing board for viewing. Here we have selected an area large enough to hold our completed layout.

The GPS-X Process Table contains allows easy access to all of the objects supported by GPS-X. To place one of these objects on the drawing board, the user simply selects the desired object with the mouse, and then clicks on the desired location of that object in the drawing board.

^*The acronyms used in this figure are defined below:

CSTR: Continuous Stirred Tank Reactor
PFT: Plug Flow Tank
SBC: Submerged Biological Contactor
RBC: Rotating Biological Contactor
DAF: Dissolved Air Flotation
SBR: Sequencing Batch Reactor

Several objects have been added to the layout. At the far left is the influent object, which represents the influent entering the plant. To the right of the influent object is a pump object; this will allow by-passing of the plant during very high influent flow conditions. To its right is a combiner object, followed by a primary clarifier. Two liquid trains have also been added, each containing a plug-flow reactor object (representing an aeration tank) and a secondary clarifier.

Several objects have been added to the layout. To the right of the influent object at the far left is a pump object; this will allow by-passing of the plant during very high influent flow conditions. To its right is a primary settler, followed by a splitter object. Two liquid trains have also been added, each containing a plug-flow reactor object (representing an aeration tank) and a secondary clarifier.

The layout of the plant has been completed: several objects have been added, and the connections between the objects have been defined -- all accomplished with just a few clicks of the mouse. The only new object type is the right-most object in the layout, which is an effluent object representing the effluent discharge point. The new objects are at the far right in the layout: an effluent object representing the effluent discharge point, and two combiner objects.

Once the layout has been completed, it is necessary to define which models are to be used for the objects in the layout. For example, the user can choose from one of three possible models for the influent object:

asm2: for influents characterized using COD and suspended solids as the main components
bodbased: used when COD data is not available
states: useful only if a full influent characterization has been performed and the state variables have been calculated manually

The model for the secondary clarifier in the first (topmost) liquid train is being defined. The user can choose from several settling models, including reactive and non-reactive models, as well as one-dimensional and two-dimensional models.

After models have been assigned to each of the objects, the next step is to define the necessary parameters. For the plug flow reactor object, the following types of parameters can be set:

the physical parameters, which include the volume and size of the tank, as well as the number of "cells" into which it is divided;
the operational parameters, which include parameters controlling the aeration of the reactor, aeration control (P/PI/PID), as well as the recycle between cells in the tank;
the stoichiometric parameters, which include ratios such as COD to suspended solids;
the kinetic parameters, which define rate constants for the various reactions.
the stoichiometric parameters, which include ratios such as COD to suspended solids;
the kinetic parameters, which define rate constants for the various reactions; and
the consistency parameters, which define alarm bounds for certain variables in each unit process object.

A form such as this one allows the user to enter the model parameters. This form allows the kinetic parameters of the plug flow reactor to be set; additional parameters can be accessed using the "NEXT" button. The values shown here are the default values for the kinetic parameters of the plug flow reactor object.

The check marks in the squares to the left of the first two parameters indicate that we intend to manipulate these parameters during the course of a simulation. (The check marks are placed here as an example only; we do not actually want to manipulate these parameters for this example.)

In this example, we want to be able to control the influent flow rate as the simulation progresses. The required steps are:

Place a check mark beside the influent flow variable in the influent object;
Create a new control window, and call it "Influent Flow" (since this is the first control we have created, it has the ID of 1);
In the "CONTROL SETUP" window, assign the influent flow variable to the control window with an ID of 1, make it a slider control, and give the slider minimum and maximum values of 0 and 10000 m³/day, respectively.

In addition to slider controls, incremental up/down controls and on/off controls can be defined.

Once the controller has been defined, it can be displayed with the "SHOW" item of the "CONTROLS" menu.

This slider allows the user to manipulate the flow between 0 m³/day and 10,000 m³/day; this is done by dragging the handle of the slider with the mouse. The number to the left of the slider represents the current value of the slider.

This slide shows the different types of controls that can be defined. (Please note, however, that the controls shown here are not used in this example, with the exception of the influent flow slider).

On/off controls are often used for activating and deactivating automatic controllers; the on/off control shown in this slide controls the operation of the automatic dissolved oxygen controller.

The "controller type" control is a drop-down control. Available options for the "controller type" variable are P, PI, and PID; this control allows the user to select from these options.

The "DO Setpoint" control is an up/down control; the value of the dissolved oxygen setpoint can be incremented or decremented by clicking on the left and right arrows, respectively.

Each object has a number of "display variables" that can be placed on various types of graphs. Here, we wish to place the dissolved oxygen profile on a graph, so we need to display the "state variables" dialog (since dissolved oxygen is a state variable).

The steps required to complete our graph of the dissolved oxygen profile are:

Place a check mark beside the dissolved oxygen display variable in the "state variables" dialog of the plug flow reactor object.
In the "OUTPUTS" dialog, create a new graph window (using the "INSERT" button), assign it the name "DO Profile", and make it a bar graph. Since this is the first graph window we have created, it is assigned an ID of 1.
In the "OUTPUTS SETUP" window, assign the dissolved oxygen variable to the graph window with an ID of 1 (the graph window we have just created), and enter the minimum and maximum bounds for the y-axis of this variable.

By placing a check mark in the "SAVE" box in the "OUTPUTS SETUP" window, the simulation results for that variable will be saved to a file; these results can then be imported into any spreadsheet.

Also note that there are several types of graphs available to the user; these are:

X-Y: Standard line graphs with linear x and y axes.
X-Y SCROLLING: X-Y graphs with a "scrolling" x-axis -- the time span shown on the graph is fixed, but the upper and lower bounds increase as the simulation progresses.
BAR CHART: Bar charts showing the values of an array variable (for example, dissolved oxygen in a multi-tank plug flow reactor).
DIGITAL: Numeric values of variables.
3D BAR CHART: Three-dimensional bar charts showing the values of three-dimensional array variables.
GREYSCALE: Like the 3D BAR CHART, these graphs display the values of three-dimensional array variables; the values are represented as varying shades of grey.

GPS-X's DEFINE command can be used to define parameters which are not related to any particular object or model, but rather are "plant-wide" in scope. One such variable which is used frequently in wastewater engineering is the solids retention time (SRT), which is selected above.

Other variables which can be defined with this command are:

DAVG: defines a variable which represents the daily average of the specified GPS-X variable.
MAVG: defines a variable which represents the moving average of the specified GPS-X variable.
TOTAL: defines a variable which represents the integral over time of the specified GPS-X variable.
ADF: defines a variable which represents the filtered value of the specified GPS-X variable. An advanced custom filtering method, called the Adaptive Data Filter, is used to filter the data.
MFLOW: defines a variable which represents the mass flow rate of the specified GPS-X variable.
F/M: defines a variable which represents the food-to-mass ratio for the specified objects.

Once the layout has been completed and the desired controls and displays have been defined, it is necessary to compile the model in order to simulate the plant. Normally, this can be done in one step by clicking on the "BUILD" button; however, in certain special situations the code and the executable files must be generated separately, as shown above.

Normally, for models of moderate size, the compilation step takes only one or two minutes.

Once the model has been compiled, simulations can be run from the "SIMULATION CONTROL" window. "Scenarios" can also be defined from this window. In GPS-X, a scenario is simply a list of commands that are recorded when you make changes to the object data entry forms. Saving a scenario stores these commands for later retrieval; running the scenario involves the playback of these instructions.

Here, we have defined a "Demo Run" scenario, in which we have modified the influent flow rate and the wastage flow rate in the plug flow reactors.

Often, it is desired to run the model to steady-state before performing dynamic simulations. In GPS-X, this is done by checking the "STEADY STATE" box and then starting the simulation. Here, the steady-state simulation has achieved 100% convergence, as indicated by the "CONVERGENCE" bar in the lower-right corner of the "SIMULATION CONTROL" window.

This slide shows a one-day simulation in progress. The graph in the upper-right corner shows the influent flow rate (in red) as well as the effluent suspended solids (in blue). The histogram in the lower-right corner shows the instantaneous dissolved oxygen profile in one of the plug flow reactors.

The influent flow rate has been modified from its initial value by manually adjusting the position of the slider in the upper-left corner.

Suppose that you had both influent flow rate data and effluent suspended solids data for a real plant. In order to calibrate the model, you would need to be able to compare the simulation's output variables of interest (here, effluent suspended solids) for a simulation which uses the same time-varying inputs as your real plant (here, influent flow rate).

In GPS-X, this is easily accomplished: the influent flow rate controller can be defined as a "file-input" type controller, meaning that it takes data from an external data file. Effluent suspended solids data can be displayed automatically alongside the simulation results; the data is plotted as plusses ("+") while the simulation results are continuous lines.

The case in this slide is obviously synthetic; it has been created from the saved results of a previous simulation, which was then re-run with slight modifications to the secondary clarifier's settling parameters. However, it demonstrates the ease with which GPS-X can be used to develop calibrated models.

Often, it is necessary to determine the sensitivity of the model's results to fluctuations in certain parameters -- for example, to obtain a measure of the error in the model's predictions. This can be easily accomplished with the GPS-X's ANALYZE Tool. Here, we will set up a steady-state sensitivity analysis of the mixed liquor suspended solids to the wastage flow rate from the plug flow reactor.

This slide shows the results of the completed sensitivity analysis. Graphs of this type can be used to determine the amount of wastage necessary to achieve a desired level of MLSS.

The GPS-X OPTIMIZER can be used to automate many of the tedious tasks required to calibrate a model. In this slide we are simultaneously optimizing the value of two parameters which control the kinetics of the biological reactions in a batch reactor. The OPTIMIZER will find the values of the parameters which results in the best possible fit betwen the simulated soluble substrate values and those supplied in an external data file.

This is the end of the GPS-X slideshow. While we have covered many of GPS-X's features, GPS-X has many more powerful features that make it the best wastewater treatment plant simulator available. Examples of features not covered in this slideshow include:

Model customization
Report generation
On-line capability
Dynamic parameter estimation
Automated respirogram evaluation
Maximum likelihood estimation
MATLAB^� support
Multivariable control design with GMI^TM
More...