HI, I am trying to run a distribution model in EPS in which the source is a over head tank on the hillock which serves a colony in which every building consists of a service overhead tank. But when I run this model flow is getting multiplied and not showing the actual hydraulics. Can you help me in thisTWS all Tank.wtg I attached the model also for your reference.
Thanks in advance,
Eswar
Hello Eswar
You have uploaded only .wtg file, please upload .wtg.sqlite file as well without which the model cannot be opened.
How are you modeling building service overhead tanks, you could put a junction for this with total demand of the building.
Regards,
Sushma Choure
Bentley Technical Suppport
Hi Sushma
Sorry for that,Herewith I am attaching the Whole model All Tanks.zip
I need the tank level for each passing hour of each building over head tank, that's why I modeled it as a Tank.
Thanks
Try experimenting with the Calculation Options. I tried the following settings:
Hydraulic Timestep: 0.05
Convergence Check Frequency: 5
Convergence Check Cutoff: 500
Damping Limit: 0.01
Trials: 10,000
Accuracy: 0.0001
With these settings, the model ran successfully with no errors and only a few warnings. The results are that almost all of the tanks fill up completely and then stay essentially full to the end. The PRVs are just opening and closing to let demand through, and almost none of the zones have any tanks that are providing actually buffering to establish a consistent hydraulic grade.
The flow through the pipes is very erratic, but within reasonable bounds, as all the PRVs are constantly trying to open and close to provide exactly the right headloss to meet the demand.
In other words, with the full tanks and the numerous PRVs and lack of hydraulic grade buffering, this is a very tricky model to get accurate results for, so it requires very small timesteps and high number of trials.
Tank Levels:
Flows out of the Reservoir (They get erratic once demand starts at 4:00):
It looks as if these tanks may be so tiny that they fill up in a fraction of a time step. If this is the case, then you would need to take very small time steps if yo really wanted to model them.
This is analogous to trying to model the filling and draining of toilet tanks. You would need to take time steps on the order of a few seconds.
What is the problem you are trying to solve? If it is design of the piping network, as long as you provide sufficient flow to the junction nodes, that is usually adequate.
Hi Stephen,
I appreciate your effort. Can you send me the model you worked out, that will be helpful for me.
TWS all Tank - edit.wtg.zip
Here's the updated model. I'm not sure if you'll be able to open it, as I think I'm on a newer version of WaterGEMS. When I saved the file it updated the schema to a newer version.