valve closure

Hello Jafar,

Based on the Valve Type and headloss from the Flow Setting (Initial), the calculated initial relative closure of your FCV is 99.0% (see the "results" section of the FCV properties). However, your operating rule pattern starts at 0% (Starting Relative closure) which means it calls for the valve to be fully open at the start of the simulation. So, in the first transient timestep, the valve is opening from 99% closed, to 0% closed (and then starts to close back up over 3 seconds).

The 99% closed initial position is due to the headloss of 137.75 m in order to reduce the flow to the desired setting of 1200 L/min (again, see the results section of the FCV).

First, make sure the flow setting (and units) are correct, along with other boundary conditions in the model, such as the downstream reservoir (is its elevation set too low?)

Next, you'll need to set the "starting relative closure" in your operating rule pattern to match the calculated relative closure from the initial conditions (see "results" section of the FCV properties). Remember that this initial relative closure may change if you change other things in the model, since the FCV will dynamically change its headloss in the initial conditions (not during the transient though) to maintain the flow setting. I recommend holding that initial position for a few seconds at the start of the simulation, so you can confirm that the model is stable (with no "initial surge")

You can read more about this and other related valve closure concepts in this article from our Wiki: Modeling Reference - Valve Closure

The following article talks about how a dynamically-operating valve such as a FCV can be converted to a TCV, which has a bit more flexibility and control over the relative closure: Converting a PRV PSV or FCV to a Throttle Control Valve (TCV) in HAMMER

Hi Jess,

Thank you so much for your great instruction and description, it was really helpful for me.

I'm going to read the articles you mentioned above, but a quick question, is there any possibility to set the valve initial position in valve operational pattern equal to calculated one for each valve? I have lots of valve in my case and each valve has own initial condition, I should define lots of operational pattern for the valves. 

Cheers

Jafar

Hello Jafar,

Jafar Dashtbani said:

but a quick question, is there any possibility to set the valve initial position in valve operational pattern equal to calculated one for each valve?

Yes you can. But as Jesse mentioned above you have to be careful in using the "Starting Relative Closure". It should match with the relative closure calculated in the initial conditions. Since you have a bunch of valves you can checkout the result field "Relative Closure (Calculated)%" for all of them. Then you can use this as the starting relative closure in your operating curve for each of these valves. This would lead to different operational patterns for different valves.

If your valves are going to follow the same operational pattern but their initial relative closures are different you can simply duplicate the operational pattern for each valve and enter the calculated relative closure for each of them.

If you have a lot of data you can consider developing these operational patterns in Excel and then importing and assigning them via ModelBuilder.

Here is an article which shows how to import demands and associated patterns via ModelBuilder. You can similarly try for valve operational patterns.

Importing Demands From A Spreadsheet

Hope this helps.

If you have a large number of FCVs, then you may want to consider a different approach using a different valve type such as a TCV, where you can enter a specific initial starting relative closure (which you could set the same for all valves and therefor only use one pattern). What do these valves represent exactly?

Hi Jesse,

Hi Yashodhan,

thanks both for your reply,

my case is large fueling network with lots of hydrants point for fueling. each point has own FCV that operator inters the set point in start of fueling. so both flow and inlet pressure in variable considering the location of point in the network and number of fueling point in service. I should try with TCV. 

Cheers