Pumps in series in a Pipeline system

I don't understand the notations "xxx+yyy km", nor where the block valves are in relation to the pumps. A sketch would help.

Because of the relatively flat pressure-versus-volume (head-versus-flow) curves of most centrifugal pumps, pressure-based controls are not usually very precise.

It appears that there is some sort of time delay between the detection of excess pressure and shutdown of the corresponding pump. If the time delays on pumps 2 and 3 are longer than the others, pumps 4 and 1 will shut down first.

This is only one idea, which may or may not be helpful, because I don't yet fully understand your system.

17 seconds is "sudden"? Maybe; how big are the pumps, valves, and pipes?

Total pipe is 873 km

PS-01 located at beginning (00km +00m)

PS-02 located at 228km +600m

PS-03 located at 448km +18m

PS-04 666km + 991m

Are you using some sort of software for this analysis and you are trying to understand the output?

What are the trip pressures at each pump? If they are all the same, then pump 1 is closer to its trip pressure than any of the rest.

If the software is a true dynamic simulation then it may be that the pump locations for 2 and 3 are such that they don't see the full pressure effect where as pump 1 may be at a point where the pressure effect is magnified by the acoustic response of the pipeline. This would be related to the speed of sound in the fluid being pumped.

At ps-02, (and others) is it 228km distance and 600 (m) elevation?

As was explained to me by the pipeline mechanics the high pressure trip on the outlet could mean the fluid being pumped was more than the system could handle, or valves could be not opened fully, or excessive differential pressure on a filter vessel / strainer or some unknown reason for excessive pressure.

What is your inlet pressure? It could be your outlet valve is pinching down to keep the inlet pressure high enough to prevent cavitation.

If PS-3 is not tripping when pressure is excessive or PS-4 is shut down, check your flow sensor, the pump should shut down if there is: excessive outlet pressure, insufficient inlet pressure or a "no flow" condition". Check the pressure and flow sensors for PS-3.

Why is BV-24 closing? I am not familiar with a surge analysis, as far as I know (mostly operator experience on pipeline / storage systems) you don't close block valves on a flowing pipeline unless you absolutely have to (something about burst pipes or filter vessels!?). An engineer should be able to mathmatically calculate the forces involved with slamming shut a block valve with out you risking your pipeline with an actual event.

Please provide more information.

Thanks,

Drew

I think he is doing simulations Drew. Not a real problem, but what if scenarios in the case of a valve closing.

What are the horsepower of the pumps?

and do you have the pump curves, as well as th impeller sizes of each pump.

BV-42 I take it as a buttefly valve, which can be pretty nasty closing., Does this butterfly valve have to seal off completely, t avoidstressto te lie, we have put ports in it, and used an extra valve frpositiv shut off. (it was a situaton).

curious on your line size on the inlet and outlet of the pumps as well as the size and wall thickness of the line itself.

p911

Greetings from EGYPT to all,

This is the modest contribution of PUMPMAN to this forum:

First we need to state the followings:

1- the flat curve is an advantage of a centrifugal pump as it means that the rate of Change of head is small or "minimum" with respect to the rate of change of flow rate.

2- we shall consider that the 4 pumps in series are of same flow rate ans same impeller design parameters as it should be for a multi pump systems in series.

3- providing that calculations are as close as possible to the system conditions in reality and the system components especially the sensors and controlling elements are sound and free of any defects, the closing of a valve and the shut down of a pump will not occur unless the pressure value per each has been reached.

Thus, we conclude:

A- The shut down of PS1 has occurred when its pressure has been reached due to the additional pressure created due to the the closing of the said valve. when the pressure built up to its preadjusted value is reached as its the farest from the valve

B- The PS3 did not shut down as its S/D pressure value meant for it has not been reached and it may have some way to releaf itself say through a branching or venting or something else

C= We got no comments on pump PS2, what is hapening to it?

The main reason for such a porblem may be that the pumps are not identical to each other. the pumps in series should be twins or as much similer to each other as possible to make sure that they shall act to outer insertins with same sensitivity

All coments are most welcomed

PUMPMAN

The spacing of the pumps is very nearly uniform, so each of them should be facing nearly identical pipe friction losses, if the pipeline itself is of uniform size. On the face of it, this all looks good. I am still inclined to suspect variations in the pressure settings or time delays of the shutdown controls, but it is still a bit of a guess. We don't know the line size or velocity, so it is hard to estimate the effect of decelerating this liquid column in 17 seconds. (Or have I missed something?)

The spacing of the pumps is very nearly uniform, so each of them should be facing nearly identical pipe friction losses, if the pipeline itself is of uniform size.

Does not necessary mean that all of the pumps are experiencing the same conditions.

The centrifugal pump always pumps the difference between the suction and discharge heads.

Plus what are the conditions of where the pumps are installed, is there sufficient straight length prior to the inlet?

what about NPSH. too much/little this creates all kinds o problems. Too much negative can begin the product to start boiling.

We don't know the line size or velocity, so it is hard to estimate the effect of decelerating this liquid column in 17 seconds. (Or have I missed something?)

maybe, lack of info there can be a number of issues creating one problem. And the issues are not where the problem shows its self.

p911

Hi Tornado, don't think you've missed anything, I also wondered if the control equipment is calibrated and checked for operation. With Centrifugal pumps if the flow is reduced, the running current drops, I wonder if there is underload detection, if so is it set correctly.

As a rule of thumb in the oil patch we set underload for ESP's to 80% of the running current once the WHP pressure/flow rate is constant.

However with horizontial pumping systems it can be lower as we do not have to worry about motor cooling (as with ESP's), but underload/low flow means the pump is in downthrust and pump life is reduced as it could be operating outside the OR.

With all due respect, the OP is referring to doing a pipeline surge analysis, which is typically a design process for a new system, and usually is not necessary for an existing system. The simulation is done to determine if in an emergency, say a truck runs into the line and breaks it open, then what is the dynamic response of the system due to the sudden closure of an emergency block valve to limit the amount of spillage. If not properly designed, and if the valve is not properly located, the pressure surge fed back into the pipeline due to the sudden valve closure can tear the pipeline apart and make the emergency worse.

I think some of the posters may not be familiar with this process. The following might help.

http://www.pipingdesign.com/articles/solutions_to_pressure_surge_in_piping_systems.pdf

From the OP

I am doing surge analysis of pipeline system (873 km) due to blockage of block valve.

Steve S, You say it enough times sometimes I can pick up on it, thanks for the link

Wish the OP would have answered

p911