1. ## Question operating tanks

I have a pipe that connects a city pipe connection with a tank (attached figure). The city connection is represented by a reservoir with a PRV.
Total head PRV = 118 m
Total head tank = 91 m
Pipe connection: 300 mm diameter; 200 m reach

The only way to balance the system is to have a big flow about 41500 cmd. Indeed, the model provides the good result. However, that would mean a very hiigh water consumption (if we locate a meter for billing purposes, the water volume to be paid would be very high)

Is there a way to maintain a higher head in the PRV, so that I can reduce the flow. Or does the model necessarily have to be in balance; thus, I the only way to reduce the flow would be to reduce pipe diameter in order to increase head losses.

Another issue is that the tank received such big flow, although it would be full. This creates a problem with water age. Is it possible to apply a valve, so that the pipe gets closed when the tank is full (to control the water volume received by the tank)

Would there be a problem or a dangerous condition if we force the system to remain unbalanced? (to have higher pressure in the PRV)

photo (1).JPG

Here are a few comments and suggestions:

1) Most modelers use a FCV vs. a PRV for this application for the following reasons: Balancing demands is very difficult to do with a PRV b/c it is meant to adjust flow to meet a specific head and therefore is difficult to use to get a flow balance over a 24 hour or longer period. Using a FCV one can set the flow to equal the expected base demand which is typically what occurs at a Water Treatment plant. I would consider making this change as it would likely make your model much easier to have a flow balance.
.
2) If you do use a PRV, know that flow is governed by the head downstream of the valve. To increase flow you increase the pressure setting (and thus the head) and to reduce the flow you would decrease the pressure setting to reduce the head. Note that with a PRV your tank level would generally likely tend to want to match the HGL of the PRV unless the flows in the pipe in-between the PRV and the tank are experiencing high head loss. If high headloss is occurring, You may wish to increase the pipe size as this usually represents "fictitious pipes" and not real pipes.
.
3) Balancing supply and demand is generally critical in most EPS models and should generally always be done to avoid tanks running out of water or getting 100% full. When a tank reaches the minimum level or maximum level, EPANET closes all pipes connected to that tank "temporarily". This can cause a model to crash if the model cannot satisfy demands or has no place to send excess water supply. If your downstream tank fills or empties completely depending on any other water supply this may cause your model issues. So this is why most modelers shoot to have supply and demand balanced. Summing demands in the input tables is usually easy to do in order to know what a supply should be. Then the modeler would set the FCV flow based on the needed supply from that water source. This would be difficult to do with a PRV, but not impossible as long as the pipelines connected do not cause excessive headloss and thus restrict how high the water level in the tank can rise. You can have a model that is not in balance and still work for a time, but if you run the model long enough will generally have an issue at some point, so in general it is always preferred to have a balanced model especially if you do any water age runs which take long run times to complete.
.
4) Controls on the Valve. It is also a nice feature to add simple controls on the valve so that it will close and open as needed based on the level in the receiving tank. This is best accomplished using a simple control on the valve itself. Simple controls when you select the valve in the model explorer are found in the white hand on the blue pipe icon next to the initial status icon. The key to using controls on a valve is to make sure when you want the valve to run make sure to apply a setting vs. telling the valve to have an "open" status. To run as a control valve you must apply a setting as "open" status tells EPANET to make it act like an open pipe. Eg: a) Valve Z closed at tank level > "X1" (where X1 is the high level you want but make sure this is less than the tank max level) b) Valve Z Setting = Y at tank level < "X2" (where X2 is the low level you want but make sure this is greater than the tank min level). These allow greater flexibility in adjusting the valve setting so that you can have an imbalance in the valve setting compared to the demand by simply allowing the valve operation to cycle. However this will only work if the valve setting allows flows in excess of the demand (i.e. you have too much supply) and will not resolve when the setting is too low.
.
I believe this answers each of your individual questions below, but please respond here if you need any further clarification on any point.

Patrick Moore

I have a pipe that connects a city pipe connection with a tank (attached figure). The city connection is represented by a reservoir with a PRV.
Total head PRV = 118 m
Total head tank = 91 m
Pipe connection: 300 mm diameter; 200 m reach

The only way to balance the system is to have a big flow about 41500 cmd. Indeed, the model provides the good result. However, that would mean a very hiigh water consumption (if we locate a meter for billing purposes, the water volume to be paid would be very high)

Is there a way to maintain a higher head in the PRV, so that I can reduce the flow. Or does the model necessarily have to be in balance; thus, I the only way to reduce the flow would be to reduce pipe diameter in order to increase head losses.

Another issue is that the tank received such big flow, although it would be full. This creates a problem with water age. Is it possible to apply a valve, so that the pipe gets closed when the tank is full (to control the water volume received by the tank)

Would there be a problem or a dangerous condition if we force the system to remain unbalanced? (to have higher pressure in the PRV)

photo (1).JPG