Validating the need for automatic calibration

00:03

A prerequisite for network model calibration is that overall flow balances—

00:08

—demands, leakage, and transfers—

00:12

—are in reasonable agreement for both the model and the real-world network for the specified time period.

00:19

To determine the performance of the current network,

00:21

and to validate that a calibration exercise needs to take place,

00:26

you need to compare the simulated pressure and flows with observed values.

00:31

The simulated values are from the results of a normal hydraulic simulation,

00:36

and the observed values are provided by a live data configuration.

00:41

You can then compare the live and predicted results using either graphs or grid reports.

00:48

Observed vs Predicted Reports allow you to compare summary flow, pressure,

00:54

or reservoir depth for all network objects of an appropriate type in the model that are linked to live data.

01:01

Simulation Graphs allow you to produce graphs of the comparison between observed

01:06

and predicted data for all the selected objects of appropriate type that are linked to live data.

01:12

For this exercise, you will produce an observed versus predicted report.

01:17

Begin by running a normal hydraulic simulation.

01:21

Create a new Run Group object.

01:23

In the Model Group, right-click the Calibration Model model group and select New > Run Group.

01:32

Then, right-click the new run group and select New > Run from the flyout.

01:39

The Schedule Hydraulic Run dialog opens.

01:43

In the Title field, type a name for the run.

01:46

In this example, “Normal Run” is entered.

01:50

Check the box next to Experimental.

01:53

Then, from the Model Group, drag and drop the Calibration network into the Network panel in the dialog.

02:01

The Control and Demand Diagram panels populate as well.

02:05

Click Save, and then Run.

02:08

Next, to produce the observed versus simulated grid,

02:12

double click on the Cal Control simulation results, then expand the Results menu and select Grid reports,

02:21

and then from the flyout, select Observed vs predicted.

02:26

The Grid Report (Observed vs Predicted) dialog box appears.

02:31

From the Model Group, drag and drop the Normal Run simulation into the Sim panel,

02:38

and then drag and drop the live data configuration into the Live Data Configuration panel.

02:45

Under Options, enable Pressure, and then click Produce Grid.

02:51

A table appears, providing some values and a verdict for each point in the network that has a live data pressure feed associated with it.

03:00

In this example, you can see that three points have a fair verdict;

03:05

therefore, running the automatic calibration module may improve the performance.

Video transcript

00:03

A prerequisite for network model calibration is that overall flow balances—

00:08

—demands, leakage, and transfers—

00:12

—are in reasonable agreement for both the model and the real-world network for the specified time period.

00:19

To determine the performance of the current network,

00:21

and to validate that a calibration exercise needs to take place,

00:26

you need to compare the simulated pressure and flows with observed values.

00:31

The simulated values are from the results of a normal hydraulic simulation,

00:36

and the observed values are provided by a live data configuration.

00:41

You can then compare the live and predicted results using either graphs or grid reports.

00:48

Observed vs Predicted Reports allow you to compare summary flow, pressure,

00:54

or reservoir depth for all network objects of an appropriate type in the model that are linked to live data.

01:01

Simulation Graphs allow you to produce graphs of the comparison between observed

01:06

and predicted data for all the selected objects of appropriate type that are linked to live data.

01:12

For this exercise, you will produce an observed versus predicted report.

01:17

Begin by running a normal hydraulic simulation.

01:21

Create a new Run Group object.

01:23

In the Model Group, right-click the Calibration Model model group and select New > Run Group.

01:32

Then, right-click the new run group and select New > Run from the flyout.

01:39

The Schedule Hydraulic Run dialog opens.

01:43

In the Title field, type a name for the run.

01:46

In this example, “Normal Run” is entered.

01:50

Check the box next to Experimental.

01:53

Then, from the Model Group, drag and drop the Calibration network into the Network panel in the dialog.

02:01

The Control and Demand Diagram panels populate as well.

02:05

Click Save, and then Run.

02:08

Next, to produce the observed versus simulated grid,

02:12

double click on the Cal Control simulation results, then expand the Results menu and select Grid reports,

02:21

and then from the flyout, select Observed vs predicted.

02:26

The Grid Report (Observed vs Predicted) dialog box appears.

02:31

From the Model Group, drag and drop the Normal Run simulation into the Sim panel,

02:38

and then drag and drop the live data configuration into the Live Data Configuration panel.

02:45

Under Options, enable Pressure, and then click Produce Grid.

02:51

A table appears, providing some values and a verdict for each point in the network that has a live data pressure feed associated with it.

03:00

In this example, you can see that three points have a fair verdict;

03:05

therefore, running the automatic calibration module may improve the performance.

Video quiz

Which of the following is NOT necessary when determining whether you can run an automatic calibration simulation or not?

(Select one)
Select an answer

1/1 questions left unanswered

Step-by-step guide

A prerequisite for network model calibration is that overall flow balances—demands, leakage, and transfers—are in reasonable agreement for both the model and the real-world network for the specified time period.

To determine the performance of the current network, and to validate that a calibration exercise needs to take place, compare the simulated pressure and flows with observed values:

  • Simulated values are from the results of a normal hydraulic simulation.
  • Observed values are provided by a live data configuration.

Compare the live and predicted results using either graphs or grid reports:

  • Observed vs Predicted Reports allow comparison of summary flow, pressure, or reservoir depth for all network objects of an appropriate type in the model that are linked to live data.
  • Simulation Graphs allow comparison between observed and predicted data for all the selected objects of appropriate type that are linked to live data.

To produce an observed versus predicted report:

  1. Create a new Run Group object.
  2. In the Model Group, right-click the Calibration Model model group.
  3. Select New > Run Group.

In the Model Group, Calibration Model model group shortcut, New > Run Group is selected.

  1. Right-click the new run group and select New > Run.
  2. The Schedule Hydraulic Run dialog box opens.
  3. In the Title field, type, “Normal Run”.
  4. Check the box next to Experimental.
  5. From the Model Group, drag and drop the Calibration network into the Network panel in the dialog box.

Note: The Control and Demand Diagram panels populate as well.

  1. Click Save.
  2. Click Run.

The Schedule Hydraulic Run dialog box displays with parameters set.

To produce the observed versus simulated grid:

  1. Double-click Cal Control simulation results.
  2. Expand the Results menu.
  3. Select Grid report… > Observed vs predicted….

From the Results menu, Grid Reports is selected and Observed vs predicted… is highlighted.

  1. The Grid Report (Observed vs Predicted) dialog box appears.
  2. From the Model Group, drag and drop the Normal Run simulation into the Sim panel.
  3. Drag and drop the live data configuration into the Live Data Configuration panel.
  4. Under Options, enable Pressure.
  5. Click Produce Grid.

The Observed v Predicted dialog box displays with the parameters set.

A table appears. Two points have a Fair verdict:

The Grid Report: Observed vs Predicted – Pressure showing the results, with the Verdict column highlighted and these live data objects showing as Fair or Good.

Therefore, running the automatic calibration module may improve the performance.

Was this information helpful?