Large pumping system 2

Hi friends:

I am designing a large cooling water system consisting of centrifugal pumps,cooling towers and shell and tube heat exchangers.I have been told that large centrifugal pumps should be equipped with a discharge pressure control valve.This valve will be partially open while pump is started and gradually opens to deliver at operating pressure.This is done to reduce water hammer and protect the pipework from large fluctations.

I am dealing with a pump capacity of about 16000 gpm and discharge pressure of 150 psi.Do I need such a pressure control system?What type of valves are used for this purpose? Whatis the time to be allowed for flow stabilization( before the valve can open say 100%)?

I would like to hear more about this from guys with experience on such large pumping systems.

Vic23

 

[TD2K]

Wow, that's a higher pressure for CW than I've typically seen.

The last place I worked had pumps in the 12,000 to 15,000 gpm range, turbine and electric drive, about 40,000 gpm total circulation. They were equipped with dampened check valves on the discharge to avoid having the check valves slammed shut on shut-down but no start-up control valve.

I never watched Operations start the pumps up but I wouldn't be surprised if the manual discharge valve was pinched closed, that's a pretty common practise.

 

[ChasBean1]

Vic, 16,000 GPM? If the drive is electric, the starting current for this pump must be tremendous. If it is a steam turbine pump, there should be a startup procedure in place so as to slowly admit steam to the turbine. If this monster is electric, you should use a variable speed drive (VSD) because, a) the Pump could start at 15 Hz (i.e., 25% capacity) and slowly ramp to its design setting, and b) incorporating VSD control will allow system flexibility and better system pressure control. The first cost of the VSD will probably pale in comparison to the operating cost of the pump.

Starting such a pump against a shutoff head (by using the valve control you mention) could be detrimental to the pump and might not significantly dampen the initial shock from the pump starting.

 

[MJCronin]

Vic23

This sounds a lot like a "Circ water" or "Cooling Tower water" system in a powerplant or a Chemical facility..... is it?

I am very suspicious of the 150 psi.....

Most typically, these systems circulate cooling water at 6-10 ft/second thru piping to many heat exchangers that are arranged in parallel service. The head on these pumps in a system arranged in this manner is commonly 75 to 125 FEET TDH....... the systems typically are all on the same general elevation. The 150 psi is typically the ANSI rating of the flanges in the piping system.

I am also suspicious of single pumps in the 16,000 gpm range....

Typically, a system of this type would have three or four pumps ( each at 50 or 33% respectively) with one spare. the pumps are typically vertical turbine or double suction design.

Tell us more about your system.... is cooling tower water ( aerated water) being circulated to the heat exchangers ? Are all of the heat exchangers in parallel configuration ?

How did you get the 150 psi as a pump discharge pressure ?

MJC

 

[sow]

I would be careful of starting the pumps closed valve head when your operating pressure is so high in the first place.

I had a similair problem with a higher flow/ lower head system. We were conscerned about hammer upon start up / shut down we decided it would be usefull to use either VSD's or soft starts ( be careful soft starts dont always have less starting current than star deltas). However this does not cater for the pumps being tripped, power failiure, accidents etc. Hence if you are likely to have a surge problem on start up you are going to get one when the power fails. You need to consider the risks of a surge event due to these events (liklihood and severity) before you decide wether you want additional surge alleviation measures.
I wouldnt hesitate to investigate the surge in more depth for a system of this expense/pressure.
A south african company called Cla-Val do the tpye of valves that you are talking about as well as othjer "air inlet" surge protection valves.

 

[cstron]

The design gets more commplex when the flow ducts are of more length. These create operational difficulties. Flow distribution in pipes effects Cooling tower performance. Seeing such a sytem operating at a high pressure i can visualise the difficulties. But still things can be operated by employing careful and lengthy operating procedures. please go through the deciding factor of discharge pressure once again.

 

[petern1]

We have large pumping systems for tailings slurry with 5 pumps in series each with a 2500HP motor pumping approximately 25,000 usgpm. The motors for all five pumps are fed from the same VFD, and as ChasBean1 says we start at reduced frequency and ramp up the speed. The discharge is usually filled with water from a suction line flush valve first.

Also as mentioned above, surge on shut down has proven to be a major problem, with severe water hammer. In a slurry system check valves are a no-no, so we incorporate a dump valve on the discharge. This dump valve is a fast-acting ball valve that opens immediately on a pump trip and dumps the pressure wave to a waste pond. If it doesn't work (i.e. if the instrument air supplying it is frozen) then look out.

In other slurry pump installations where we have no need to isolate on the suction side as we do above, we just shut the line down and let the line reverse flow through the pumps, elimianting the need for a check valve.

Our hot water pumps have 2000 HP motors with no VFD or soft start. These when started without pinching back the discharge valve cavitate and hammer severely. Starting them with the discharge valve pinched back about 25% open seems to work well.