This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

Supply demand higher than required demand in the system

Hi.
I tried to model an existing system using WaterGEMS. A simplified version of the model is attached as below.
CASE 1
I modelled a reservoir to represent the existing water treatment plant. The existing water treatment plant can only produce a maximum flow of 75MLD.
The reservoir supplies water to 4 tanks. In between, there are also 11 direct tapping points from the transmission pipe to their supply areas. The outflow demand for Tank 1, 2,3 and 4 are set to be 22MLD, 8MLD, 5MLD and 5MLD respectively. Whereas, the total demand for the 11 tapping point is 32MLD. Therefore, the total required demand for this system is 72MLD.
I tried to model the system without the introduction of FCV-21 in my previous model. When I checked the flow of the downstream pipe after the pump station, the flow is found to be 130MLD, which is not in the case of the existing system.
Therefore, I put a FCV which was set to 75MLD behind the reservoir, so that it will limit the supplied flow to be 75MLD.
May I know if this modelling approach is correct? If not, how can I model the system so that the supply flow will be at a maximum of 75MLD?
CASE 2
The same system as mentioned above is used, but FCV-21 is set to "inactive"  in this case so that the flow is not controlled at a maximum flow of 75MLD.
The reservoir supplies water to 4 tanks. In between, there are also 11 direct tapping points from the transmission pipe to their supply areas. The outflow demand for Tank 1, 2,3 and 4 are set to be 22MLD, 8MLD, 5MLD and 5MLD respectively. Whereas, the total demand for the 11 tapping point is 32MLD. Therefore, the total required demand for this system is 72MLD. The hydraulic pattern is also set to be fixed, so that there is no variation in the demand over time. The initial setting of all the 4 tanks are set to be empty at start.
After computing, I tried to check the WTP outflow again, which is circled as shown in the picture attached. The flow is supposed to be 72MLD since the required demand in the system is only 72 MLD. However, the flow in that pipe is found to be 104MLD. 
May I know why the supply demand is higher than the demand required? How can I rectify this issue if the system is modelled wrongly? Furthermore, is there any way to model the system so that the supply flow will be at a maximum of 75MLD in this case?
Thank you.
  • Hello,

    Check to see if there is flow into the tanks from the pumps in your model. If so, that is likely why you are seeing more flow in the pipe in question. If you have demands on the tank, the demand is drawn directly from the tank. However, there can still be flow into the tanks from the pump stations if the head at the pumps is higher hydraulic grade at the tanks. 

    If you look at the calculation summary, you can see a the flow demanded. This should match the total system demand at a given time step. The calculation summary also includes the flow supplied (from a reservoir) and the flow stored (the net flow at tanks in the system. 

    Regards,

    Scott

  • Hi.

    Thank you for your explanation. Since this is an existing system, which is only supply a maximum flow of 75MLD. Is there any way to model in such a case so that the flow from the pump station will be at 75MLD?

    Thank you.

  • First, you should model the pumps you have and see which combination can provide 75 MLD with the largest unit out of service. If they can't reach that flow, look into replacing some pumps or adding pumps. For a station that size, you'll probably want 4 or 5 pumps to give you good flexibility and reliability.

    But 364 days of the year will have demand less than 75 MLD. You need to look at which combination of pumps will work most efficiently at a variety of flows that you'll encounter during the year. Different combinations will work more efficiently at different flow rates.

    WaterGEMS pump combination analysis for pump station gets you that kind of information.

  • Thank you.

    Actually, the pump station can provide around 110MLD at most. However, the existing water treatment plant can only provide a maximum flow of 75MLD treated water to the supply zones. Therefore, the maximum supplied flow of 75MLD from the treatment plant should also be reflected on the model for analysis purpose. The problem is that the reservoir can produce up to 104MLD in my model, which does not reflect the on site situation.

    Therefore, I would like to get some advice on how to model the reservoir so that maximum 75MLD will be supplied.

    Thank you.

  • Hello Shinkean,

    Looking at your original post, I am recommending some options for you to try;

    1. As Scott mentioned, demands applied directly to tanks are drawn directly from the tank. This means that based on the HGL at the tanks the demands applied to it would be fulfilled and it does not necessarily mean that the tanks are first filled by the pumps. If you are running Steady State simulation with fixed demands then the Initial Elevation / Level of the tank will be considered as the tank HGL for the demands there, so it could be incorrect to assume that tanks are empty here. However, if you are running Extended Period Simulation (EPS) and allowing the tanks to fill then your system can be realistic.

    In this case rather than having demands directly assigned to the tank try using an junction downstream of each of the 4 tanks and assign the demand to that junction.

    2. Rather than putting an FCV right after the Reservoir (WTP), use it after the pumps for a more realistic estimate. As you say the pumps are providing around 110 MLD so an FCV set at 75 MLD should control the flow downstream. However, this is just so you get a flow of 75 MLD going towards the system. If in the existing system there is no such provision of an FCV, then you need to check how the pumps are functioning in your system with accurate pump curves.

    Let me know if this helps.


    Regards,

    Yashodhan Joshi