Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations SSS148 on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Centrifugal Pump Relief Valve Sizing

Status
Not open for further replies.

ChemEng33

Chemical
Jul 2, 2017
5
A new centrifugal pump has been installed which is capable of exceeding the pipework's design pressure when the discharge is blocked. The client would like a relief valve to be installed on the discharge of the pump which relieves to the suction side of the pump. I am trying to determine what the required relieving rate would be and would be very grateful for your thoughts.

From my understanding, as the discharge valve is closed the pressure will increase along the pump curve until the relief valve starts to lift (relief valve set pressure - pump suction pressure). The further the discharge valve is closed the more flow is bypassed across the relief valve. As the discharge valve is fully closed all the flow is bypassed across the relief valve and the pump suction pressure becomes the discharge pressure - frictional losses. The pressure will fall along the pump curve and flow will increase.

My questions therefore are firstly, is my understanding correct and how do you determine the point at which maximum flow through the relief valve would occur?

If I assume that the only pressure losses are through the orifice of the relief valve, would you calculate the pressure drop across the relief valve with incremental increases in flow until a point on the pump curve was found?

Thanks in advance
 
Replies continue below

Recommended for you

In your system as described, the PSV is designed to discharge the full flow at the pipe MAWP with 10% overpressure. The pressure drop of the PSV based on the inlet pressure (i.e. pump discharge pipe MAWP) and the outlets pressure (i.e. the pump suction pressure).

IMO, some issues of this pump system need to be addressed when blocked the pump outlet valve in operation, such as:
- overpressure the discharge pipe
- overheat the pump
- inadequate pump minimum flow requirement for a stable hydraulic flow

If to add a PSV to protect the discharge pipe from overpressure, the set pressure of the PSV could be based on the MAWP of the discharge pipe. The discharge flow could be between the full flow and the minimum pump flow. The required full flow is to protect the pipe, and the minimum flow is to protect the pump. And, one should to verify these conditions with pump's H-Q curve.
 
Your worst case is total blocked outlet.

The issue to determine first is what is the inlet pressure at no flow? Might / should be higher than at flowing condition.

Then subtract this pressure from your pipework design pressure. This should then be the maximum differential pressure the pump can supply before you go into overpressure.

So find this differential pressure on your pump curve.

This should then give you a pressure difference and flow rate for your relief valve.

mk3223 says the same thing and provides good guidance on other issues you need to address.

Are there intended to be high pressure trips set below the relief valve setting?
how long does the blocked in condition last?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.
 
Thank you very much for both of your responses.

I have already carried out the steps outlined above but I then started to think, what happens when flow begins as the relief valve begins to lift. The pump is now in a closed loop and the static head from the tanker is no longer relevant. The pump suction pressure becomes the discharge pressure minus the pressure drop across the relief valve. The differential pressure across the pump is now less and therefore the flow increases.

With regards to the other issues, the pump motor has load monitoring and high load trip set, the pump 'can' temperature is monitored and a high temp trip set, the pump discharge pipework has a pressure transmitter immediately downstream of the pump with a high pressure trip that is set below the new relief valve set pressure.
 
Thank you very much for both of your responses.

I have already carried out the steps that you have advised in your responses. But after doing so I began to think, what happens as the relief valve begins to lift and flow begins. The pump is now in a closed loop and the pump suction pressure = the pump discharge pressure minus the pressure drop across the relief valve. The differential pressure will decrease and the flow increases (as we move down the pump curve). As the flow through the relief valve increases the pressure drop increases. So my question is how do I determine the required full flow when the loop is closed and equilibrium is reached? Or is equilibrium ever reached and the relief valve will just 'chatter' away?

With regards to the issues raised, the pump motor has load monitoring and a high load trip, the pump can temperature is monitored and a high temp trip set and there is a pressure transmitter installed immediately downstream of the pump with a pressure trip set below the intended relief valve set pressure.
 
Apologies for the two posts.

I thought the first hadn't posted but must have been a delay.
 
It's a good thing that the shutdown mechanisms are in place to protect the pump and piping.

Yes, the PSV "close loop" around the pump is an issue when the PSV discharge is too "close" to the pump suction inlet. As the pump suction pressure goes up, it will be balanced out with the tank which is worked as a "buffer tank". The issue is that the short close loop is to raise the liquid temperature which may damage pump in a short period of time. To avoid it, the PSV discharge may be routed back to the tank, instead of pump suction.
 
Thanks mk3223,

Unfortunately, the source vessel is a tanker which is why I intended on routing the PSV discharge to the pump suction, rather than back to the tanker barrel.

Please can you explain what you mean by "balanced out"? Wouldn't the pressure at the pump suction be the same in this scenario regardless of whether the tanker bottom outlet was closed or not?
 
Yes, it's to be a totally close loop if the tanker outlet was closed. If the connection still open, some discharge liquid will push back to the tanker.
IMO, the PSV discharge can be routed back to the tanker if a nozzle was equipped on the top of the tanker, or you can request it.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor