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

Partially Closed Valves in water supply system

Hello,

I'm modeling a high pressured water supply system full of partially closed valves (which people use as some kind of "prv's" to reduce pressure because they don't have enough money to buy it).
The information they gave me is D(mm) of a valve, type of the valve, pressure before and after the valve (M h2o), and how much flow goes through it after partially closing it (l/s). 

for instance,   pressure before and after the valve is 75m, valve D=150mm, after partially closing it becomes after the valve 20m and 5 l/s goes through it.


I used hc=kv2/2g  equation to find headloss coefficient setting

75 - 20 = k(4Q/3.14D2)2/2g

k=13750


but how precise is it? how precise would it be in eps modeling?

How to model this kind of system in watergems? 

Parents
  • Hello Levon, 

    Which valve element you are using to model partially closed valves? Is it TCV? 

    Are you changing the closure % of valve time to time during EPS or is it constant? 

    Please read more about calculating headloss coefficient and modeling valve closure using link below. 

    Configuring TCV relative closure patterns for EPS 

    How are headlosses determined for TCV's with the different initial status settings? 

    If this is not what you are looking for then please elaborate your query. 

    Regards,

    Sushma Choure

    Bentley Technical Suppport

  • Hi Sushma,

    I'm using TCV, steady-state condition.

    I don't know how to represent a valve that is partially closed.

    (for instance, in real life pressure before and after the valve is 75m, valve D=150mm. After partially closing TCV pressure becomes after the valve 20m and 5 l/s flow goes through it.). how to simulate this?


    What would you do?

    Would it be precise to put PRV+FCV to simulate partially closed TCV? or there are other options?

    in my city model, I have more than 100 such kinds of situations. 

  • To answer this question, I would need to know the type of valves you are using and how they are operated. Are these gate valves, butterfly valves, globe valves,...? Are they adjusted hourly, daily, annually, never?

    The simplest way to model this is to develop a curve of percent open vs. minor loss K and knowing the percent open, you can either use a TCV or simply put a minor loss on the pipe where the valve is located. Knowing the percent open of each valve for 100 valves can be difficult, especially if they are frequently adjusted by a range of human operators who may not communicate the status to the modeler.

    I have a problem understanding why anyone would operate their system like this. Manually throttling a valve to control pressure results in the valve being closed significantly during low flow periods. However,  when a high flow is needed, these valves dramatically reduce system capacity. Remember that head loss is a function of flow squared. It's like paying for a 200 mm pipe but converting it into a 50 mm pipe.

    In general, this is not a good way to operate a system. It is better to define pressure zones and feed them through PRVs. A PRV can maintain a good pressure during low demand times but when demand increases, can open wider to enable a large flow to pass. There is no need to put an FCV in series with a PRV unless you are actually installing such a configuration.

    Answer Verified By: levon batoyan 

  • Thanks for all of your responses,

    the system is very old with high NRW, we inherit it from the soviet union. Operators are doing this to reduce NRW until directors find the money for prv's.

    In this case, it is a globe valve, which is adjusted annually.

Reply Children
No Data