water distribution model

I am now updating the 2017 model to include the future water mains and checking if I am meeting future my max day demand and fire flow needs. The city is expanding and will have new commercial, residential and industrial subdivisions. 

Currently there is one storage tank, which is supplying water just to the city and its new watermains (not neighboring city).

The water source filling the tank is just  from two connection points from the neighboring city. I am not entirely sure if it is gravity, all I know is that  the connection can supply a max of 500 gpm.

This is the way I modelled it but I am not sure it is correct:

I drew a reservoir and pump at each connection point. I have hydrant data near the connection point and I used that to develop my three point pump curve. I  then set the elevation of the reservoir so that I am able to get 500 gpm as flow supplied? I then distributed a max day demand of 400 gpm throughout the system to different junctions. Is that the correct way to do this?

Attached is my model.

Lawrence Model 2021 - MST.wtg

First, you only included the .wtg file, We also need to SQLite file. It's best to use File >Save to Package and upload them that way.

A lot depends on the relative importance of the upstream system. Are they just a backup or do they supply 100% of your water. If it's the latter, then you need to spend some effort understanding their system.

I'll assume you've done a good job modeling the downstream system. The success of the model depends on how you model the upstream. The best solution is to have them give you their model and you can merge them but that seldom happens. The next best thing is to model their system from their tank that supplies you, just including the most important pipes.

If the upstream system won't share their maps with you, then you need to look at the instrumentation at the connection points. Ideally you would have flow and pressure at the connection sent to your SCADA System. First try to model average days. 

Then conduct hydrant flow test just across the connection point. If you only have a single feed, there is an approach that you can use.

https://communities.bentley.com/products/hydraulics___hydrology/w/hydraulics_and_hydrology__wiki/9293/modeling-a-connection-to-an-existing-system

If you have multiple connections, this approach isn't very accurate.

While someone may say you can get say 500 gpm, at which locations and what pressure? If the upstream system has a tank at the connection point or it ties into a 48 in. pipe, you don't need to worry about the head loss in the upstream system. However, inmost cases, head loss in the upstream system is an issue.

Thanks for all the help! Really appreciate it. The connection points consists of two pump houses that draws from the lake and provides 100% of the  supply to the town.  The first connection is now providing 100% of the supply to the town. The second connection is eventually going to be also operational proving supply to the town as well.  I have hydrant flow test just across one of the connection points. So I guess I can use that.

So how should this be modelled when I have more than one connection point?

You mentioned hydrant flow tests but then you also appear to say that the "connections" are real pumps/water sources? Can you clarify?

As seen in the article, it is not valid to have two connections to an existing system (where the actual water source is further upstream) if they both go back to the same source, because the variation of inflow (for example from demands) from one source could impact the capacity of the other source, and cannot be reflected in the pump+reservoir approximation approach. So, if the actual water source is much further upstream and you have two connection points to it modeled with the pump+reservoir approach using hydrant flow tests, it would not be valid and you'd need to model a skeletonized version of the system back to the actual source, for more accurate results. Perhaps a sensitivity analysis trying both approaches would be helpful to give you more insight into the behavior of the connection points.

You mentioned hydrant flow tests but then you also appear to say that the "connections" are real pumps/water sources? Can you clarify?

As seen in the article, it is not valid to have two connections to an existing system (where the actual water source is further upstream) if they both go back to the same source, because the variation of inflow (for example from demands) from one source could impact the capacity of the other source, and cannot be reflected in the pump+reservoir approximation approach. So, if the actual water source is much further upstream and you have two connection points to it modeled with the pump+reservoir approach using hydrant flow tests, it would not be valid and you'd need to model a skeletonized version of the system back to the actual source, for more accurate results. Perhaps a sensitivity analysis trying both approaches would be helpful to give you more insight into the behavior of the connection points.