Pump is producing flow rates way below design point at design pressures

[rick81122]

Hey,

I have a problem with two split case centrifugal potable water booster pumps that are not pumping water at any where near the design point of 600 gpm at 220 feet. The pumps are sister pumps installed to operate in a three pump parallel configuration with another old pump, with suction and discharge manifolds. The suction piping seems to be fairly well designed with approx. 3 feet of straight 4" diameter suction piping into each pump. The suction side is out of a large storage tank with approx. 18 feet of static suction head on the pumps. The discharge side is fairly simple with a check valve and an old school air and oil pump control valve. Both pumps are driven by 60 HP 460 volt three phase motors. The motors are designed to spin at 3500 rmp at 60 HZ.

I have installed a pressure gauge right at the discharge of both pumps, to assure myself of the system curve operating point. Pump 1 (pumping alone) is operating against 98 psi (226 feet), and produces 380 gpm. Pump 2 (pumping alone) is operates against 93 psi (214 feet) and produces 210 gpm. Both of these operating points make no sense according to the pump curves I have obtained from the manufacturer. Pump 1 should be making approx. 500 to 550 gpm when seeing 226 feet of head, and pump 2 should be making 620 to 640 gpm when pushing against 214 feet of head. It seems like the pumps are running off the curves, which is impossible.

I figured it had to be power related. I have trouble shot the power supply and am seeing seeing consistent voltages between the legs of 490, 486, and 490 when the pumps are running. The full load amps rating of the motors is 68 amps. I am seeing 52, 54, and 50 amps per leg with pump 1 running, and 44, 49, and 47 amps per lag with pump 2 running. The lower current draw corresponds with the reduced flow rate, but the power and motors seem to be ok.

I am wondering if the impellers have been installed backwards, or the motors have been wired to the power supply incorrectly. My next move is to open the cases and look at the impellers, and to make sure the leads on the motor are connected according to the diagram I just received from the motor manufacturer as per a 480 volt power supply.

My question is what else should I look for to be causing these pumps to operate well below the rated flow rate. I am assuming that by placing gauges at the discharge and reading this direct pressure I am getting an accurate indication of the TDH the pumps are working at. I probably need to see if I can install gauges on the suction side and measure the difference in pressure across the pumps as a better indication of TDH?

Another clue is that when all three pumps operate at the same time including the old pump that is operating well at its design point of around 380 gpm and 220 feet of head, Pump 2 gets starved and gets hot after a few minutes.

Any ideas?

Thanks,

Rick

 

[JJPellin]

Power is certainly not the problem. An induction motor cannot be slowed down much below rated speed without tripping off. Options I would consider are these:

Motor rotation. - No need to wait for shutdown and check motor leads. Touch the spinning shaft with a soda straw and verify rotation.

Impeller orientation - If impeller is double suction, it could be installed backwards.

Suction restriction - A plugged suction strainer or a valve with a dropped gate. Verify suction pressure right at the pump flange.

Incorrect impeller - Wrong pattern or diameter.

Johnny Pellin

 

[DubMac]

Most likely you are "pumping back through" one of the pumps, or at least one of the pumps (when two are running) is producing more head and restricting the flow of the second pump.

In parallel pumping the curves should be fairly steep; if they are not, one of the pumps can "wander" up and down the perf. curve as it tries to equalize the pressure seen from the other pump. And with a check valve, it is possible for the "stronger" pump to pump the check valve shut on the "weaker" pump; that is most likely what has happened when you say "pump2 gets starved and heats up".

If you can post a copy of the curves of the new pumps, and also the old one if you have it; most any of the experts on this site could unravel your problem. It would also be advisable for you to draw up a "total" performance curve. That is, you should make one curve that represents the two new pumps running. Also another curve which shows all 3 pumps running.

 

[rmw]

I'm highly suspicious of the check valve on the "old" pump. Actually, I'm highly suspicious of any check valve.

rmw

 

[Artisi]

Are these 2 pumps a new installation or have they run been installed and operating previously?

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[bimr]

Your piping is horribly undersized. I am surprised that you can 380 gpm out of a pump.

The 4-Inch suction should have been 8-Inch if you were planning to operate with a single 600 gpm pump, and larger if you desired to operate multiple pumps. The discharge should be at least 6-Inch, preferably 8-Inch.

The pumps are probably not cavitating because the pumps are throttle back by the undersized piping. The pumps are actually internally recirculating, that is why the pumps are producing heat.

I would suggest that you shut off the pumps immediately before you burn up the pumps.

Management also looks down on double posting.

 

[bimr]

600 GPM through a 4-Inch pipe is about 15 ft/sec. What was the designer of this system thinking?

Reasonable pipe velocities depend on the application. There is no correct velocity for all applications. Here is a general guideline.

Reasonable Velocities for the Flow of Water through Pipe:

Boiler Feed.............8 to 15 ft/sec
Pump Suction ............4 to 7 ft/sec
General Service.........4 to 10 ft/sec
City.......................to 7 ft/sec
Transmission Pipelines...3 to 5 ft/sec

 

[Artisi]

Agreed, 4" pipe for 600gpm, what were they thinking, unless the system has been upgraded from a much lower flowrate. However, with only 3ft of inlet pipe the losses are tolerable but the high velocity may well be causing some inlet flow problems effecting the pump performance. A sketch / photo of the inlet pipe / pumps might be of help.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[bimr]

From the post, it would appear that smaller pumps were replaced with larger pumps. The existing pipework was not upsized.

Centrifugal pumps will operate on the manufacturer's curve if the pumps are installed correctly.

 

[DubMac]

With the undersized piping, it would be a good idea to get a system curve sketched as well.

 

[khalid01]

When I first read your post, I immediately jumped to the new impellers size. I believe it worth double checking the 'old' pump impeller size and compare it with the other new pumps. As you know, a slight change in the impeller diameter can make a big difference in pump performance.

I tend to agree with the previous posts. Sounds that the design of the suction piping was not done properly. You cant pass 600 gpm through a 4" pipe (but I am not sure if THAT is the root cause of the problem). Also symmetry of suction piping for pumps in parallel operation is very important, so maybe a sketch of the piping system will clarify this.

Are the new pumps purchased from the same manufacturer of the old one? Same model?

Have you noticed any 'abnormal' sound coming from the suction piping when you run the new pumps? Any sign of excessive vibration? Have you taken any vibration readings when you tested the pumps?

 

[123engr]

rick81122

I agree with the posts above. I would advise your team to painstakingly check, redesign and validate the whole pumping system. This includes redesign of both the suction and discharge piping, pumps design, etc.



I am what I am by His grace

 

[QualityTime]

"....I am wondering if the impellers have been installed backwards, or the motors have been wired to the power supply incorrectly. My next move is to open the cases and look at the impellers, and to make sure the leads on the motor are connected according to the diagram I just received from the motor manufacturer as per a 480 volt power supply.

My question is what else should I look for to be causing these pumps to operate well below the rated flow rate. I am assuming that by placing gauges at the discharge and reading this direct pressure I am getting an accurate indication of the TDH the pumps are working at. I probably need to see if I can install gauges on the suction side and measure the difference in pressure across the pumps as a better indication of TDH?..."


A pressure gauge and a suction gauge is required is required to determine TDH. If it is under a flooded suction conditon subtract the suction reading from the discharge reading and then add the difference in velocity head between the suction and discharge if the suction and discharge flange sizes are different. If it is under a suction lift condition add the two gauge readings and then add the difference in velocity head between the suction and discharge if the suction and discharge flange sizes are different. Make sure you make the correction for the gauage readings to the pump centerline datum.

Do a shut in head test. If it does not match the curve that MIGHT be and indication it is running backwards.


Throttle the discharge valve to develop readings and plot it agains the pump curve and tell us what you observe.

 

[QualityTime]

I want to make a clarification to my last post:

"...A pressure gauge and a suction gauge is required is required to determine TDH. If it is under a flooded suction conditon subtract the suction reading from the discharge reading and then add the difference in velocity head between the suction and discharge if the suction and discharge flange sizes are different. If it is under a suction lift condition A GAGE THAT CAN READ NEGAVTIVE PRESSURE IS REQUIRED. Add the two gauge readings and then add the difference in velocity head between the suction and discharge if the suction and discharge flange sizes are different. Make sure you make the correction for the gauage readings to the pump centerline datum...."

 

[QualityTime]

One more time, !!!I want to make a clarification to my last post:

"...A PUMP DISCHARGE pressure gauge and a PUMP suction gauge is required is required to determine TDH. If it is under a flooded suction conditon subtract the suction reading from the discharge reading and then add the difference in velocity head between the suction and discharge if the suction and discharge flange sizes are different. If it is under a suction lift condition A GAGE THAT CAN READ NEGAVTIVE PRESSURE IS REQUIRED. Add the two gauge readings and then add the difference in velocity head between the suction and discharge if the suction and discharge flange sizes are different. Make sure you make the correction for the gauage readings to the pump centerline datum...."

 

[brandonbw]

I am curious what size the dry well is? We had a system designed that was less GPM and about half the TDH, and that called for 6" pipe. Something we have also been seeing is undersized dry wells and I was curious if this was one of those.

B+W Engineering and Design
Los Angeles Civil Engineer and Structural Engineer

http://bwengr.com

|

http://bwstructuralengineer.com

|

http://bwcivilengineer.com

 

[EnergyProfessional]

"I have installed a pressure gauge right at the discharge of both pumps, to assure myself of the system curve operating point." do you only read discharge pressure? How are you actually determining the flowrate? I hope you are not just using the pressure gain over the pump....

I mean you say you don't get enough flow, how do you know?

and regarding pipe sizes, ASHRAE recommends 1-4 ftH2O/100 ft. You actually have to calculate that for the specific fluid and pipe size and flow. ft/s is just a rule of thumb. suction and discharge size variance imho only matters in open systems or where you have hydraulic separation (i.e. secondary/primary loop). In single loops it doesn't matter because what the pump pushes out, comes pushed back in.

I'm not sure your background, but you should hire an engineer to size that properly (don't hire the guy who originally designed this!)