Minor Loss at PRV


I was modeling a scenario with a PRV and realized that the PRV is not working in the field but the operational team is not using the Bypass. It turns out that the supplier of this PRV indicates a Headloss of around 4 or 5 mH2O. 
Can I enter this head loss using the Minor Headloss field? Which coefficient should I use?

In this image I obtained field data with the model and I am concluding that this Headloss of 5 mH2O occurs basically in all time Steps.

  • To model a fixed, specific headloss, you can use the Pressure Breaker Valve (PBV). See more here: Modeling a Constant Headloss

    If the headloss varies with flow, you could calculate an equivalent minor loss coefficient K (Headloss = K*V^2/2g) and place that on a pipe instead.


    Jesse Dringoli
    Technical Support Manager, OpenFlows
    Bentley Communities Site Administrator
    Bentley Systems, Inc.

    Answer Verified By: Fabio Lobo Araujo 

  • Fabio, I suspect the PRV manufacturer is referencing the pressure loss that happens on PRVs fitted with a conventional 3 port Pilot with the bonnet venting its pressure to the downstream port.   An alternative arrangement that some designs use is to vent the bonnet to atmosphere with a solenoid valve.  If the PRV is the second type, this will follow a standard Minor Loss formula of H = k . v^2 / 2g

    If the PRV is a 3 port Pilot venting the bonnet to the downstream port , the the Head Loss vs Flow Curves look a bit like this.  It will change slightly manufacturer to manufacturer, in this case this is for the Singer Model 106.   There is probably a bit of miscommunication from the PRV supplier about how the PRV performs with the Pilot Control fully wound open to vent the bonnet as much as possible........they don't so much have a "minimum pressure loss".  Yes there is a smaller minimum pressure differential to overcome the opposing bonnet forces to get the valve to initially crack open, but more relevant in the more common design cases is that they instead have a "minimum pressure loss before they full open".

    Note that it isn't a constant head loss coefficient, it is a 2 part compound curve with a transition in between.  There is an initial high relative minor loss coefficient in the low flow range that occurs as the flow is trying to lift the valve plate up off its seal, but is fighting the against the bonnet that has both a spring, and the the downstream water pressure trying to keep the valve closed, with the flow rate momentum not quite able to push the valve opening very far open.  As the flow gets high enough, the PRV finally reaches the second, flatter curve when enough flow rate cause enough momentum change/force upwards on the valve plate to be in the "full open" position, and in this higher flow position the PRV has a more conventional k value minor loss coefficient.

    It isn't particularly easy to model in network analysis packages.  What we did was decide which part of the flow range we were interested, and use only 1 of the below slopes to figure out the PRV minor loss coefficient value and use this.  Doesn't 100% model the PRV accurately, but it does OK for the most common flow design scenarios we compute.

    Thinking about it, you could model a PRV that has its Pilot wound all the way open reasonably accurately with a WaterCAD GPV  .  GPVs allow the user to put in any headloss curve they wish and so the below manufacturer curves can be manually input this way.

    Answer Verified By: Fabio Lobo Araujo