Pump Efficiency

[impeller1]

Folks,

We have 6 vertical turbine pumps (3 stage) which are rated at delivering 4400 gpm. From commissioning, which was approximately 9 years ago, the pumps delivered around 4300 - 4400 gpm. Fine. Recently, we have seen the performance drop to 3900 - 4000 gpm, around a 10% drop in performance. We have serviced these pumps, checked all wear ring clearances, runouts, bearing bushing clearances and checked for any casing leaks. We did everything I think could possibly be done, yet the pump isnt delivering much more, around the same 4000gpm. A Process consultant on site is adamant that the pump is still the problem no matter how hard I try to explain to him the pump is governed by the system it's in and that will dictate where on the curve it operates. The pump pumps into RO membranes which can become, clogged, fouled or damaged. I would like to get your opinions, do you think it is a system problem as I think or am I not addressing some issue with the pump. The only proof I have that it's the system is that I know for a fact that I've checked everything regarding the pump and I dont see anything else that can be done - pump was fitted with new shafts, sleeves, bushings and clearances are in spec.

 

[ione]

Is meanwhile changed anything in the pipeline (i.e. supplemental fittings or everything which could have increased the pressure drop and so changed the response of the pump to the new system’s conditions)? Can you check the ampere absorption of the pump’s motor and compare it with a value measured when the flow delivered was 4400 gpm?

If the pump has been completely overhauled and it is ok, I would be inclined to think of a system’s problem.

 

[ione]

Have you checked and overhauled just the pump or also the motor? If not check for correct voltage and speed: reduced speed means reduced flow rate (being all the other conditions consistent).

 

[BigInch]

Is the pump's total discharge head equal to the corresponding value for your new flowrate as shown on the pump curve. Remember to include any suction pressure changes too. If all pumps are operating on point, then you should most likely looking for problems elsewhere.

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[impeller1]

BigInch, Well of the pumps that were refurbished, they're not 100% on point with the design curve, maybe a 5% difference to what it's supposed to be. I will do further testing and post the results with the design curve.

 

[rmw]

You may have a stack up of small problems. Not one big problem but a combination of slight wear ring wear, impeller wear, etc.

Are the flows you gave determined pump by pump individually or taking the flow of all pumps in parallel and dividing by six?

Can you post the original TDH and the current TDH.

Any build up in the pit near the pump suction(s)? Any change in the pit level?

rmw

 

[CH5OH]

by turbine pumps you mean steam turbine driven? or?
if yes, rpm still the same?
if no, you might have internal leakage (between stages or from discharge to suction)
if the proces is to blame, you should see a difference in discharge pressure (unless the pump curve is really flat)
less flow means higher discharge pressure

 

[Artisi]

Sounds like a normal / time reduction in capacity due to slight wear and increased clearances, fouling of the pipe work (increasing head).
Is there any changes in the inlet supply level, have the RO elements been changed or is there any likely head change after the RO unit/s

 

[JJPellin]

Some additional information would be valuable. In the service you describe, I would expect the pump to be installed in a can, not in an open pit. Is there a strainer in the suction line? We have experienced a number of instances where "start-up" strainers were accidentally left in place and eventually became restricted. Are you sure you are setting the lift correctly. We have pumps of this type with radial wear rings that also have a lateral ring off the end of the impeller eye. Depending on the configuration, the correct lift setting could be 0.030" or it could be 1/4". Verify the proper lift for you pump from the manufacturer. Pumping into an RO I would expect some grade of clean water, perhaps deminerized water. Demin water can be very corrosive. If the pump wetted surfaces have become rough from general corrosion, the pumping efficiency and capacity could be reduced. There are commercially available coatings that promise to restore efficiency by reducing flow losses and turbulence.

If the water is not demineralized, it could be prone to lime buildup in the piping. I have seen 10 inch pipe with only about a 4 inch hole left down the center in very hard water services.

Are you sure that the flow measurement is correct. If it measuring a pressure drop across an orifice plate, the plate could be washed out resulting in an artificially low reading. Has the flow meter been calibrated? Has the orifice plate been inspected?


Johnny Pellin

 

[HEC]

Based on measurements being within 5% of curve, I would suggest that the pump is infact operating on the curve and the issue is with the pipework. Look for fouling or barnicles!


Mark Hutton

 

[MikeCool]

If as you say, today's pump after refurbishment is the same as the new pump, the problem is not with the pump. You describe a 10% drop in flow, but you don't discuss the system head? Has it gone up? During the 9 years of operation, it's certain that your piping is somewhat fouled. This adds to the resistance of flow. Unless something else has changed (motor rpm?), an increase in system head is the culprit. Fire your consultant.

Mike Cool
Mechanical Engineer
American Stainless Pumps, Inc.
Los Angeles California USA

http://www.aspumps.com

 

[Artisi]

As a "pump guy" I must remind everyone - initial analysis of a sight problem always indicates it's the pump at fault, irrespective of the problem - flow, head, vibration, power consumption or noise, it's the pump.

 

[Artisi]

read site problem

 

[ccfowler]

impeller1,

I agree with the earlier posts that the suggest that the problem is probably within the connected system and not in the pumps, but you didn't mention anything about the pump motors or the electrical supply. If there have been any electrical supply changes (other greater loads on the same supply, different supply transformer, etc.), or the motors have been rebuilt, re-wound, or replaced, then it is possible that the pumps may actually be operating at slightly lower shaft speeds resulting in slightly reduced performance without any significant physical changes or degradation within the pumps themselves.

You didn't mention any sort of VFD, so I'm assuming that the nightmare of VFD's and multiple pumps is not a part of your problem. If VFD's are involved, then there is a really big can of worms to deal with.

 

[Slagathor]

impeller1,

BigInch and MikeCool pointed out the issue. You are only looking at 1/2 of the info....FLOW. A pump curve is defined by 2D points. Flow AND head.

When that pump was brand new....you could have run the system...choked down a valve...adding friction...and pushed the pump back left to a lower flow. Would THAT have been a pump problem?

No. Operations personnel...and more than a fair share of engineers make this mistake. To judge a pumps performance...you have to look at flow head points...not just flow and heads alone.

Run a real pump test. Measure flow and head at each point. This requires good pressure gauges on suction and discharge. Correct for gauge elev...and don't forget to calc velocity head (though that should not matter much in a high head application).

RO pumps are high head...so your pump curve could be flat. That means a small change in friction (membrane 'gunk') could push you left quite a bit. I would be that you are requiring the pumps to work against more pressure...pushing your point left.