Issue in quality simulation for a long time simulation

[HUGHYDRO]

Hi, I try to lauch a "age" quality simulation on 300 or 500 hours. I observed that QGISRed didn't succeed to finish simulation if the pattern is too small (pattern time step and/or quality time step), more annoying when I compare the results of quality simulation in QGIS RED or in EPANET 2.2 I don't have the same results (really not ...). An other point is that QGIS Red don't manage to simulate and give results in statistic (Max, Min, Average) as EPANET succeed.

Maybe I don't use it in the good way so If you can give me some informations or confirm my difficulties .

Thanks

[neslerel]

Hi @HUGHYDRO,

You are right, currently QGISRed doesn't support a big simulation (i.e., with a high number of report steps). There are two main reason: 1) we store the results in shapefiles with a column (in attribute table) for each report step and the number of columns is limited. 2) the way we have to recover Epanet results is very inneficient and it provoques that, in this type of models, the simualtions take a long time or it doesn't finish.

We have solutions for each one, but at the moment we want to increase the features of QGISRed that there are not included in Epanet. We hope in future version we can improve this behaivour.

About statistics, we are not generating a report with this data, but I think that with results layers and QGis it is possible to obtain this information. We also noted it to include in future versions.

Thanks for your comment!

[fmartine]

Hi @HUGHYDRO,

As Nestor said we are focused now on improving the capabilities of QGISRed to build models. Particularly we are working on refining the nodal demands allocation, well directly or through service connections, and also we have started to declare sensors. Improving the capabilities of visualizing the results will come in future versions, but in the meanwhile, you can export the model to EPANET.

However, I'm more worried about the issue that the QGISRed results differ from that produced by EPANET for an age simulation. If you can provide more details on this case we will go into deep to solve the problem. You can send your data confidentially to my e-mail at fmartine@hma.upv.es

[neslerel]

Hi @HUGHYDRO, could you send us a model (INP or QGISRed project) where the quality results don't match with Epanet results? If there is a problem we want to fix it.

Thanks!

[neslerel]

Hi @HUGHYDRO,

It seems there is difference in quality results depending on Epanet toolkit version. QGISRed uses the latest (2.2) and it gets the same results as Epanet 2.2 (GUI). Could you confirm us with what version are you comparing results?

Thanks!