Check Valve on Minimum Flow Line 1

[ghamsa]

We have 4 running centrifugal pumps. Each pump is equipped with 8” minimum flow recycle line. Each recycle line has flow control valve and check valve (downstream of the control valve. All 4 pumps discharge into one common header, 24”.

There will be additional 5th pump identical to the other 4 pumps and connected to the same 24” discharge header. Unfortunately, this new pump came without check valve downstream of the minimum recycle valve.
We need to start up the new 5th pump but in case this pump shutdowns or trips the minimum flow recycle control valve will fully open. My question is: will there be major pressure disturbance in the 24” disparage header? Or the other running pumps will work to maintain stable pressure in the discharge header?


Note: we need to start up the new pump but we are planning to add check valve after sometime from pump start up.

 

[jeap]

Hi ghamsa,

First of all, a question: Assuming that each of your pumps has a check valve and a shut off valve (as most of the systems have). The recycle line starts upstream or downstream of this check valve? If it starts downstream of the pump check valve I consider it is not necessary to install an additional check valve downstream or the control valve.

If the recycle line starts upstream the pump check valve you need to install an additional check valve downstream the control valve in order to avoid the recicling fluid to run through the pump when it is stopped. This assumption is considering that you have a minimum flow control that opens the control valve if it is sensed a low flow at the pump discharge.

An additional option is to program an interlock to close the control valve when the pump stops. It could be done by venting the air feed line to the control valve with a three way solenoid valve, or course if your control valve is Fail to Close.

Regards,
jeap

 

[dcasto]

jeap, do have those backwards?

 

[ghamsa]

How can I predict the pressure drop in the 30" header (570 psig) in the first 5 mints. once the 8" minimum flow control valve opens fully? the minimum flow control valve is connected the pump suction header (16" with pressure of 270 psig)

Thank you all

 

[BigInch]

Why do you need a check in the recycle line? I assume your recycle line runs from pump discharge to suction or back to a tank. Since those are (almost or always) at lower pressure than the pump discharge, there should be no need for a check to prevent flow from suction or tank to the pump discharge. Are you sure you don't mean a bypass line? A bypass line would normally have a check valve.

The best way to make start up predictions and be sure you have a smooth path to stable flow operations is to use a pipeline or pump system simulation program that has very accurate transient analysis with control equipment modeling capabilities. Let me know if you need additional info.

http://virtualpipeline.spaces.msn.com

 

[jeap]

dcasto,

please look the diagram:

http://i3.uploadem.com/10042007/pagina.jpg

It is possible under the conditions I show you...

Need to know a scheme of the ghamsa case.

 

[BigInch]

Right. The diagram helped a lot.

First question is why is the recycle line open when the pump is stopped? It should be closed (and interlocked if remote controlled) until the pump begins spinning up. Once the pump begins spinning up, pressure builds very rapidly and any flow from any other recycling pump into that segment would normally be very small and quickly reverse direction as the starting pump builds pressure. If it trips on the start-up, close the valve.

I don't see any other "probable" reason that flow would come from the recycle lines into the first pump's discharge. You would not normally have a pressure drop in that segment of pipe to drive any flow in towards the pump discharge, provided that the tank is not a pressure tank with a pressue higher than the discharge pressure of your pumps (Doubtful) or unless the recycle line became plugged at the tank. Basically all flow should take the path of least resistance, which would be straight to the tank. If pump #1 was off, the chance for backflow to its discharge from pump #2 would be greater (but still doubtful unless the recycle line was plugged again), but supposedly pump #1 discharge valve would be closed in that situation. If that valve was open, you might backflow into pump #1, but again only if the recycle line was plugged at the tank AND the valve was open. Normally it is not required to design for two simultaneous failures, if the probability of failure (plugging at the tank) is very low and there is little chance for a dangerous situation to develop. Assuming your recycle line is designed for full pump shut-off pressure, no problem. If pump #1 was on and recycling, all flow would always go to the tank from that pump too, unless again a plug at the tank.

If the valve was open AND the tank inlet was plugged AND another pump was recycling, you might get some backflow through pump #1. Is that a bad probability, given two failures would have to had occured and another pump would have to be recycling at the same time? Is the suction piping and suction flange on pump #1 designed for the pump discharge pressure? Maybe if its a low pressure pump, maybe not if its a high pressure pump. If not, you may want to have a check valve in the recycle line.

I'd say, no you don't need a check in the recycle line, but they're not too expensive, if you feel better with one.

http://virtualpipeline.spaces.msn.com

 

[dcasto]

I can't see the diagram, but you are in good hands with BigInch.

 

[ghamsa]

Thank you all for participations. I will try to describe my case again. Sorry, I could not make a clear picture.
We have four centrifugal pumps each is 130,000 bbl per day in capacity. Normally, three pumps are running and one is on standby. The pumping system is consisting of the following components:
? Suction header: size is 16”with operating pressure of 270 psig (all four pumps take suction from this common 16”header)
? Discharge header: size 30”with operating pressure of 570 psig (all four pumps discharge into this common header)
? Pump pressure at flange discharge is 590 psig
? Pump discharge line is 14” size
? Each pump discharge is equipped with a check valve followed by a block valve.
? Each pump is served with 8”minimum recycle control valve that is branched from each pump 14”discharge line and has check valve upstream of the minimum recycle control valve.
? Each 8”minium recycle line is connected into the 16”suction header
? The intent of the adding a check valve upstream of the 8”minium recycle valve is to avoid release of discharged flow to 16”suction header during pump shutdown , trips or minimum control valve air signal failure
There is a new 5th pump being added to these four pumps. Unfortunately, the 5th pump came without a check valve upstream of the 8”minium recycle line. With addition of this new pump, we will have 3 running pumps and 2 standby pumps.
My question: if the 5th pump trips, the control logic will fully opens up the 8”minium recycle control valve back to the 16”suction header, which is at 270 psig. How this sudden opening of the 8”minium recycle control valve will disturb the operating pressure in the 30”dishcarage pressure, which is at 570 psig, in the first 5 minutes?? Will this overload the other 2 running pumps, or would the 30”discharage header pressure, 570 psig, drop to close the suction pressure??
I hope this description is clear and I thank you all for your help

 

[ghamsa]

Note: Please consider the (?) in the above thread as bullets.

 

[BigInch]

I think the control logic is backwards.

When the pump trips, CLOSE the valve. I believe that the valve should be open ONLY when the pump has a RUNNING status and when RPM status is LOW (probably < 20% of BEP revs).

When you push the start button, the valve needs to,

1.) OPEN, and stay open until,

A.) The pump reaches something around 20% operating speed, at which time you need to be closing that valve to build sufficient pressure to open the pump discharge check and get into the header. It is doubtful you will build sufficient pressure, if a recycle valve is open.

Or,

B.) The pump does not get a RUNNING status immediately after starting, then CLOSE.

When the pump trips, the recycle valve should be closing.

So, to make the logic very simple,

IF 0 <= PUMP_RPM AND PUMP_RPM <= 20% BEP_RPM 
  OPEN RECYCLE VALVE
ELSE
  CLOSE RECYCLE VALVE
END IF

You have said nothing to indicate to me that the recycle valve should be OPEN when the pump is tripping or is stopped. I believe it should be closed.

P.S. There are NO VSDs on those pumps, right? If there are VSDs, you probably don't need any recycle line, if you have the control logic correct, so in that case just try it with the recycle valve CLOSED all the time.

http://virtualpipeline.spaces.msn.com

 

[ghamsa]

There are NO VSDs on these pumps. We close minimum recycle control valve during pump start up in order to build pressure.
I agree our logic is backward. The valve should close when pump shutdowns or trips.
Just for the sake of answering my thread, can you please answer the question in the last paragraph on the thread?”
My question: if the 5th pump trips, the control logic will fully opens up the 8”minium recycle control valve back to the 16”suction header, which is at 270 psig. How this sudden opening of the 8”minium recycle control valve will disturb the operating pressure in the 30”dishcarage pressure, which is at 570 psig, in the first 5 minutes?? Will this overload the other 2 running pumps, or would the 30”discharage header pressure, 570 psig, drop to close the suction pressure??

Many thanks

 

[BigInch]

OK. But first, as I said before, if the pump trips the recycle valve should close. No need for recycling on a pump with power off.

Its a bit hard to say exactly how the header will respond, without putting it in the simulator, but I'll try to offer some ideas.

If the other 4 pumps are running, you will lose 1/5 of the flow into the header, so the pressure will reduce until the system responds somehow. The system will respond with a combination of 2 things happening.

1) the 4 running pumps will tend to increase flowrate, and
2) the downstream header and piping will pick up that flow and the resistance to flow of that piping system will increase. You will eventually balance out at the new intersection point of the pump curves superimposed on your pipe system curve. The time it takes to do that will depend on how fast the downstream pipeline can fill and back up pressure, so that's about its length and diameter. If you didn't have control valves in the discharge, it should be at its new increased pressure approximately about the same time it takes for a pressure wave to travel down the pipeline and return back to the pumps. Since you have control valves in the pump discharge, they will also affect how long the downstream pipeline fills as they move to try to maintain whatever set pressure you have. Since you might have those actuators set fast or slow or for holding a constant pressure on the header or a constant pressure on the pipeline, I've no idea what effect they will have. Another effect comes from the motor torque curve and the pump moments of intertia. This is more important to include, if they are large pumps.

If this is a critical application, the only way to handle it is to do a full pump, valve, pipeline, and recycle valve simulation with various operational configurations for the pump group.

http://virtualpipeline.spaces.msn.com

 

[ghamsa]

BigInch BigThankyou

Many thanks