View Full Version : InfoWater demand numbers

April 18, 2016, 06:40 AM
Where can I go in InfoWater to verify the results of the demand allocator without having to rerun the allocator?

Patrick Moore
April 18, 2016, 12:10 PM

Generally best practice methods I used for years were as follows for verifying demands:

1) Know the total demand expected in the demand allocator (total demand from meters, or expected demand from land use worked out manually) You can then use this to compare that all of the demand expected to be allocated was allocated.
2) After allocation make a screen capture of the demand summary report window created by the allocator. If you close this window you generally can't get this back without rerunning the allocator. So make sure you copy it before you close it. If this number is different than the expected values in number 1 above, there generally is something you will need to resolve in your allocation. Could be a spatial projection issue or meters outside Thiessen polygon boundaries, but something is causing demands to not get allocated properly.
3) As a final check I always recommend checking the Demand in the input DB tables after allocation to verify the values are as expected. This can verify that demand was not allocated to inactive nodes unexpectedly, that demands were not applied multiple times (can be an issue depending on the settings you have in the allocator and demands not purged before rerunning). This final test is the most important. If demand total are not as expected at this step these need to be resolved as the totals shown in the DB tables are the values the model uses to run. Be sure to know that demand can get allocated to inactive elements in certain circumstances (Note: Inactive elements don't generally show up in the DB tables unless you set the DB editor to open inactive elements.)

So essentially the best verification method is to first compare the allocated totals match the expected allocation values and then the verify that all of the demand made it to active nodes in the scenario.

Please let me know if you have further questions on this methodology.

Patrick Moore