7 replies to this topic

[bikrom]

Dear all,

 

I have some basic doubts arising due to a recent event. I have attached a sketch of closed loop cooling water circuit for our compressors in the excel file. The only parameters which can be monitored are pressure and temperature downstream of the cooler.

 

Only one of the two cooling water pumps is i/c at any time. Normally, cooling water is sent to only one of the compressors, whichever is in operation. Valve in the make-up line is kept very slightly open so as to allow a little overflow from the overflow line. There is no controller in the discharge line of the pumps.

 

Here are my doubts,

 

a ) If water is sent to both compressors simultaneously (without changing make-up flow), we observe a fall in the disch. Pressure. (Generally from 3.5 kg/cm2g to 2.5 kg/cm2g). Does this necessarily mean that the flow rate through the pump has increased putting extra load on the pump? Or does this simply mean that NPSHA has reduced and level of water in the tank has fallen? Or is it a bit of both?

 

b ) In above case, to maintain the disch. Pressure at 3.5 Kg/cm2g, we now increase the make-up flow slightly. Will the flow rate increase now? Will it put any extra load on the motor?

 

c ) Finally, will the maximum NPSHA be (maximum upstream make-up pressure + height of the tank)? (if make-up valve is completely open)

 

What I basically am confused about is:

 

1) Effect of sending water to both compressors, (without increasing make-up flow) on flow and head of the pump.

 

2) Effect of increasing the make-up flow(and hence overflow) on NPSHA, flow and head of the pump.

 

What actually happened follows. Compressor-1 and cooling water pump -1 were running. Compressor-2 had to be started for a routine change-over. Cooling water valve to compressor-2 was hence opened without changing make-up flow. This caused cooling water pump disc. Pressure (P) to fall from 3.5 to 2.5 kg/cm2g. After a short interval, make-up flow to tank was increased until disc. Pressure (P) rose back to 3.5 Kg/cm2g.

 

Due to certain complications, change-over which normally is over in 30 minutes, could not be completed for the next 3 hrs. During this whole time, cooling water was going to both compressors. After three hours, cooling water pump -1 tripped due to overload and cwp-2 started automatically. 45 minutes later, cwp-2 also tripped due to overload. The reason behind this overload is what I want to understand.

 

Other problems in the cooling water circuit have not been ruled out though. (higher than expected pressure drop, underperforming motor and such).

 

Any help is highly appreciated.

 

Thankfully

Bikrom

Attached Files

•  hydraulics.xlsx   14.01KB   40 downloads

Edited by bikrom, 31 July 2013 - 09:15 AM.

[meetyourmaker]

here are my answers

I have assumed that the tank is open to atmosphere( its pressure is quite low) and makeup water is not being used to pressurize the tank.

a) when you send the cooling water to both compressors the flow through the pump will INCREASE. the pump operating point will shift to the right on its curve and amps drawn by the motor will increase. So since the flow has increased the differential head developed by the pump will decrease . You said the system is a closed loop system so under steady state conditions the level will not fall since whatever extra the pump draws from the tank it is putting it back. there is no accumulation!. innitially when you lined up the second compressor valve the level in tank may have fallen if the line was empty, resulting in sudden decrease in pump pressure(from 3.5kg to 2.5kg/cm2) and reduction in tank level.

)  increasing makeup flow will have no effect on pump load/flow.

c) assuming tank pressure is not equal to make up stream pressure, the NPSHA will be= tank pressure(seems to be atmospheric in this case)+liquid head as measured from pump impeller eye-liquid vapor pressure at pumping temp-line frictional losses.

If a you keep the level in tank constant i.e it is always overflowing with constant makeup, the NPSH remains the same as calculated from formula above

here is my take on your problem

The pumps tripped beacuse the motor high amp setting was reached, this occured because the flow and hence the current drawn by the pump motor increased when you started sending water to both compressors.

1) so when sending water to both compressors,  the pump flow will increase and head will decrease

2) Like I said there should be no effect of the makeup on NPSHA flow and head

Edited by meetyourmaker, 31 July 2013 - 02:46 PM.

[bikrom]

Thanks a lot for your answer. Please bear with me some more.

 

If the tank were open to atmosphere (that is no roof), all you have said would be perfectly correct. However the tank in question happens to have a roof. The overflow line is on this roof (so the tank is always full). Overflow line is a short tube of 6" which is then connected to a hose draining at grade level. I should have mentioned this earlier. Sorry for the trouble.

 

Pressure in the make-up line upstream of the valve is 6.0 Kg/cm2g.(attached excel file in first post).

 

now

 

a ) I guess your answer still holds true for part a) i.e., as long make-up flow valve is not opened, level and hence NPSHA will remain same.(and the load and flow will increase and head will decrease)

 

b ) Will this closed roof cause the NPSHA to increase and the load to decrease on opening the make-up flow valve (and have no significant effect on flow and head)? 

 

c ) Will the maximum possible NPSHA (when make-up valve is fully open) be = 6.0 + liquid head as measured from pump impeller eye-liquid vapor pressure at pumping temp-line frictional losses.

 

Two conclusions I need to arrive at are:

 

1) It seems that the pump is not designed to supply water to both the compressors at a time.

 

2) Does opening up the make-up flow valve help in any tiny way to reduce the load on the motor? (keeping in mind that pump has to discharge at the top of the same tank from bottom of which it is taking suction)

 

Thanks in advance.

 

With regards

Bikrom

Edited by bikrom, 31 July 2013 - 08:23 PM.

[breizh]

Bikrom,

Is this happening suddenly ?  Do you have a way to measure the flow rate ? Can you use a non intrusive flowmeter (ultrasonic type) for example ?  Nothing wrong with the suction line (obstruction ) or with the equipment (pump) or discharge line (Check valve)?

What about Fouling ?

 

Hope this helps

 

Breizh

Edited by breizh, 31 July 2013 - 11:12 PM.

[meetyourmaker]

 

a ) I guess your answer still holds true for part a) i.e., as long make-up flow valve is not opened, level and hence NPSHA will remain same.(and the load and flow will increase and head will decrease)

 

b ) Will this closed roof cause the NPSHA to increase and the load to decrease on opening the make-up flow valve (and have no significant effect on flow and head)? 

 

c ) Will the maximum possible NPSHA (when make-up valve is fully open) be = 6.0 + liquid head as measured from pump impeller eye-liquid vapor pressure at pumping temp-line frictional losses.

 

Two conclusions I need to arrive at are:

 

1) It seems that the pump is not designed to supply water to both the compressors at a time.

 

2) Does opening up the make-up flow valve help in any tiny way to reduce the load on the motor? (keeping in mind that pump has to discharge at the top of the same tank from bottom of which it is taking suction)

 

First of all let me state that I still think this tank is atmospheric and the pressure of the water makeup stream has nothing to do with the tank operating pressure.

a, b>. On opening the makeup valve there is no change in the level of the tank, whatever extra water you are feeding into the tank is getting drained immediately so since level is constant and pressure in the tank remains atmospheric there is no change in npsha.

 

c)NO as ive stated above I believe tank pressure will be atmospheric and not makeup water pressure.

1) try to find out where the pump is operating on its curve and determine whetherthe pump or the motor is the bottleneck. It is not recommended to operate the pump at flows significantly greater than BEP flow.

2)This is a closed loop pumping system, the pump does not have to overcome any static head it only has to overcome the frictional resistance in the piping system so increasing tank pressure/ NPSH etc has little effect on the pump operating parameters as long as NPSHA>NPSHR. Only opening closing valves in the discharge piping will affect pump operating point.

 

[bikrom]

Thank you all. I will go through the pump curves and  other information. We will run both the pumps for the time being whenever required. breizh, this is happening suddenly. We intend to clean the whole system for fouling at next opportunity. we also intend to install a local ammeter.

 

meetyourmaker, thanks again. 

 

just one last question please. why are we getting a higher discharge pressure on increasing the make-up flow? (it happens every time even though no other change anywhere in the system is done) 

[meetyourmaker]

Do you increase the make up flow as soon as you line up the pump discharge to both compressors? how long does it take for the pump discharge pressure to normalize? Have you tried this without increasing makeup flow. Is the cooling water line to the second compressor filled up with water when you start sending cooling water to it?

[bikrom]

We tried to recreate the situation today. After lining up the flow to both compressors, we waited for 15 mins. Disc. pr. as usual had dropped from 3.7 to 2.6 kg/cm2g. We measured the amperage which had increased from initial value of 7.1 to 7.9 Amps.  (expected behaviour) .Disc. pr. didnot show any trend of increasing during these 15 mins.

 

Then the make-up flow was increased(by approx. 2000 Kg/hr) till we achieved a disc. pressure of 3.4 Kg/cm2g. kept it in this condition for next 15 minutes. No significant change in amperage was found. Make-up flow was further increased until disc. pressure rose to 3.7 kg/cm2g. Still no significant change in amperage found. Disc. pr. did not show any trend of decreasing during this period.

 

Yes, cooling water line of the second compressor is filled up always.

 

Disc. pr. doesn't change at all in either direction unless we open or close make-up flow valve.

 

Full load current is 10.6 Amps.(as per data sheet). Pressure drop from pump disc. to compressor circuit inlet is 0.5 kg/cm2 approx. We always maintain a visible overflow from the hose draining at grade level. Water temp. after cooler is in the range of 35 to 38 deg.C