Hello all!
I have a question for all the more advanced WaterGems users, regarding demand allocation.
I have a whole city's water distribution network uploaded in watergems (something like a 650km of distribution pipes and 90km of transportation lines). The only consumption nodes I have are the ones that were created when importing the pipe shapefiles, so they are not representative of the actual population distribution or water consumption. The only data regarding consumption I have are the daily produced and fed water flows from the WTP: roughly 158.000 m3/day.
What should be the best way, in your opinion, to distribute demands through the whole network? The idea is to calibrate the model so that we can study the water inflows from reservoir to reservoir (the transportation lines between reservoirs also supply water along their way, so we need to estimate if they are enough for both distribution and transport or if we need to built new lines only for water transport from reservoir to reservoir, with no supply).
I have tried to erase the smaller diameter pipes (skelebration is also an option but I decided to erase directly because it allows me to choose my "backbone" and skelebrator always leaves many smaller diameter pipes, even after several skelebration steps), leave only the networks backbone and then, using Thiessen polygons created in WG, move the newly created orphaned nodes (as a result of the erased smaller pipes) demands to the remaining nodes in the system's "backbone". But this solution still has the initial problem: how to allocate demands throughout the whole network initially?
Thanks in advance for your reply!
Best regards
Rui
Rui,
Based on what you've listed that you have for data on the demand for the system it seems like you only have 1 option, but if you could get a little more information about your demands, service areas, and flow boundaries that may open up your possibilities. The one option that I can think of is to equally distribute your total demand over the entire system given that you don't seem to have any more data other the total consumption. The process that I would use to do this is to go into the demand control center (Tools > Demand Control Center), use the option to initialize the demands for all your elements (junctions and/or hydrants), right click the column header for the "Demand(Base)" column, choose "globally edit", and enter your average demand (the total demand divide by the number of junctions in your model). That will at least give you a place to start. If you don't know how to globally edit your demands and want to have a look further we have a wiki technote that describes the process:
communities.bentley.com/.../7557.how-to-edit-demands-globally.aspx
I also wanted to add that after you distributed your flow equally using the method above and you could then manually go in an make some adjustments if you got further information or knew that there was more demand in one location versus. another. I'm not sure how accurately you could do this because it would probably be estimating, but it might also prove useful.
If you had some additional information such billing meter data and known service areas that would be useful because then you could use a tool like Loadbuilder (Tools > Loadbuilder) and choose from one of the various options such as distribution by nearest node, nearest pipe, billing meter aggregation, etc...Loadbuilder is a tool that we have, which uses shapefiles that define things like your flow boundary area, service area, billing meter data, population by area (landuse), etc..., to allocate demands to your nodes automatically. There are 9 different methods and you can find more information about these methods in the help document that comes with the WaterGEMS software titled "Loadbuilder Wizard". This information will give you an idea of what type of data you would need to make each method work.
I hope that helps out.
Mark
Answer Verified By: Sushma Choure