positive displacement pump versus centrifugal pump 3

[Yobbo]

Hi,

I would like to start a discussion about the differences and the working range of centrifugal pumps and positive displacement pumps in combination with low viscosity medium like water. The reason for this is that I wonder whether the considerations to choose one or the other might partly still be based on aged arguments. I am afraid I cannot speak out of long term experience about the positive displacement pumps. But I do know that there has been a time, where there was a transitional phase where the centrifugal pump got more and more application as a replacement for the positive displacement pump.
Just to give a first description of obvious differences I learned I would like to name the following ones:

PDP : positive displacement pump
CP : centrifugal pump

disadvantages PDP versus CP
- principally a source of pressure fluctuations, although
there are some that produce far less fluctuations than
others.
cure: pressurized vessels for damping fluctuations
- in general more gland problems
- dynamic problems with check valves when applicable
- in general higher maintenance costs
- need for safety relieve valve to protect the piping
against exceeding the design pressure of the system
- construction is in general more complex

advantages PDP versus CP
- higher efficiency with lower flows
- higher head achievable with relatively simple construction
- controllability of flow is better

Reviewing my list of (dis)advantages I am inclined to conclude, that applying a PDP is less favourable. But maybe I forgot some aspects that are important too. Also I wonder whether the weighing of these factors is still correct. They were made at a time, where the required power was still cheap and therefore the efficiency were less important. Qualitively spoken the difference in maintenance costs may still be in favour of the centrifugal pump, but hasn't the development of PDP's (especially the rotary ones) made this difference less? My drive to get the differences in application area more clear is the fact that the PDP's have the potential of saving power. All of course just in situations where both types of pump may be applied.
For example the rotary displacement pump to me looks a candidate with the least disadvantages of a PDP, that could operate advantageously compared to a CP. Especially in the area of low flow condensate recovery to a high pressure condensate system.
I wonder if anyone has ever performed a study on this subject. I am looking forward to your reactions.

With best regards,





Karel Postulart, The Netherlands
Nuon Power Generation

 

[zdas04]

The dynamic vs. positive displacement discussion has been going on for nearly 100 years, and it still has more emotion than science.

I've never heard the claim that PD pumps used less energy than dynamic pumps before. In my experience the opposite is true. If I run a centrifugal pump in its sweet spot the power consumption is significantly less than a PD pump in the same conditions. The problem I see with dynamic pumps is that the sweet spot is so small. I recently did an analysis of an electrical submersible pump (ESP, multi-stage centrifugal pump) vs. a progressing cavity pump (PCP, rotary PD pump). The ESP had a turndown ratio of 1.6:1 (i.e., the maximum flow was 1.6 times the minimum flow) and the PCP had a turndown of 11.1:1--almost 10 times the operating range. If I had been certain beyond all doubt that I could control the flow from the reservoir in the ESP sweet spot then the power requirements were 2/3 the power requirements of the PCP in that same set of conditions (and half the motor I eventually specified for the full range).

My rule of thumb has always been that if I can control the suction and discharge pressure I want a designed-for-service dynamic pump (for the energy savings). If the variability is outside of my control then I want a PD pump.

And to think that energy efficiency wasn't as big a deal 50 years ago as it is today is to ignore the facts. There has never been a time that engineers did not consider operating costs in their analysis. Energy cost becomes more or less of a driver with the assumptions for future cost of energy, but it is never incidental.

David Simpson, PE
MuleShoe Engineering

"Belief" is the acceptance of an hypotheses in the absence of data.
"Prejudice" is having an opinion not supported by the preponderance of the data.
"Knowledge" is only found through the accumulation and analysis of data.

 

[micalbrch]

If you reduce your comparison to clean water applications, you will not find many applications where a PDP has advantages over a CP. High pressure cleaning or well injection are the only ones that come to my mind and "high pressure" is even not a very precise definition. But there are many other applications, especially not for water where PDP have advantages over CP and that concerns maintenance and spare parts costs, too. On the other hand the world's pump market is a 90 % CP market. So one can say that PDP cover more the niches.

Just some comments to your PDP disadvantages. An air driven diaphragm pump is a PDP but does not need a pressure relief valve and disphragm pumps in general (as well as peristaltic pumps) are glandless.

The world of PDP pumps is so diversified that it is a bit unfair to put all in one definition and compare them with CP. There are progressive cavity pumps, gear pumps, lobe pumps, diaphragm pumps (air driven, mechanical driven and hydraulically actuated), plunger pumps and single and double acting piston pumps.

 

[stanier]

PD pumps cost 3 to 10 times that of centrifugal pumps.

When a PD pump stops it stops dead and the fluid cannot pass through the pump. Unlike the dynamic response of a centrifugal pump where the pump can even reverse flow can pass forward or backward. (refer Four quadrant data). hence waterhammer is more of a concern and expensive mitigation measures may be required for PD pumps.

“The beautiful thing about learning is that no one can take it away from you.”
---B.B. King

http://waterhammer.hopout.com.au/

 

[stanier]

Suggest you go to

http://www.pump-zone.com

http://www.mcnallyinstitute.com

http://www.pumpsystemsmatter.org

http://www.pumpfundamentals.com/pump_glossary.htm#gl441

and read.

Cnetirfugal pump's capacity can be varied by change to the impeller diameter and associated drive.

Generally look at the industry that you are in and what others are doing. Rest assured that even 50 years ago they made decisions on logical engineering and commercial criteria. Because an industry has been doing it for a long time does not mean it is wrong.

“The beautiful thing about learning is that no one can take it away from you.”
---B.B. King

http://waterhammer.hopout.com.au/

 

[moltenmetal]

stanier: be careful- if upstream pressure is higher than downstream pressure, significant forward flow can occur through a PD pump when the pump is stopped, depending on the type of PD pump. That's why I like the PIP P&ID symbols used for pumps, rather than the generic circle with the arrows and the little base on it that people often use on PFDs. The PIP symbols give the user an idea of the flow path that the pump body itself represents as part of the piping system, especially if you go the extra step to draw in the check valves which are part of the pump (if the PD pump design you're using has any).

 

[stanier]

Hi moltenmetal,

Could you please elaborate on what PD pumps allow flow through when they are stopped. In my experience piston, piston diapragm, diaphragm, progressive cavity, gear, rotary vane and peristaltic pumps would not allow any flow except perhaps leakage past seals. this is not considered 'significant".

I was not referring to small leakages but the quite large flows that can occur with a centrifugal pump when it is tripped.



“The beautiful thing about learning is that no one can take it away from you.”
---B.B. King

http://waterhammer.hopout.com.au/

 

[Compositepro]

Many reciprocating pumps have spring loaded valves. Some valves aren't even spring loaded but close when the flow reverses, and drags the check valve closed. So flow can still continue through these pumps even if the pump is not running.

 

[CaracasEC]

The characteristic of a positive displacement pump includes:it is a constant volume, while centrifugal is variable volume, pdp has variable differential head while cent pump has constant head. PDP is relatively insensitive to liquid properties while cent pump is sensitive to liquid properties. PDP is relatively insensitive to system changes while cent pump is sensitive to system changes.PDP is not self limiting while cent pump is self limiting..

 

[stanier]

Suction NPSH needs to consider the acceration of the fluid for a PDP.

“The beautiful thing about learning is that no one can take it away from you.”
---B.B. King

http://waterhammer.hopout.com.au/

 

[BigInch]

My most important reason to make the distinction between centrifugals and recips are their pressure ratio capacities. Centrifugals are much more limited in range.

If it ain't broke, don't fix it. If it's not safe ... make it that way.

 

[BigInch]

Forget it. That's for compressors.

If it ain't broke, don't fix it. If it's not safe ... make it that way.

 

[zdas04]

When I wondered why we didn't talk about "pumping ratios" in this forum a few years ago you would have thought I was speaking against religion. I never got a good reason for the outrage, but the outrage was so big that I dropped it. As I've thought about it over the years I've come to realize that the difference is the fluid response to energy input--compressible fluids squeeze into a smaller space and incompressible fluids go faster in the same space (which can be converted to pressure increase in an increasing cross section ala Bernoulli). After thinking that through I got much easier with differential head.

David Simpson, PE
MuleShoe Engineering

"Belief" is the acceptance of an hypotheses in the absence of data.
"Prejudice" is having an opinion not supported by the preponderance of the data.
"Knowledge" is only found through the accumulation and analysis of data.

 

[CaracasEC]

The following are some key application criteria that would lead to the selection of a P.D Pump over a centrifugal pump. High viscosity, Self-priming, High Pressure, Low Flow, High Efficiency, Low velocity, Low shear, fagile solids handling capability, Sealless pumping, Accurate, repeatable flow measurement, Constant flow/variable system pressure, two-phase flow.

 

[moltenmetal]

stanier: I learned this the hard way by using an ordinary diaphragm metering pump, discharging from a tank into a vacuum application, where the pump was used as a metering device rather than principally as a source of pressure. Every diaphragm or valved piston/packed plunger pump WILL allow very significant FORWARD flow when stopped if the suction pressure is higher than the discharge pressure, unless the pump is fitted with a (heavily) spring loaded discharge check valve to reduce this. Every gear or progressive cavity or valveless piston pump which has been run for any appreciable period of time will leak in the forward and reverse direction, although at a greatly lower rate than a diaphragm or packed plunger will permit forward flow under those conditions. And every check valve- including those with springs- heck, ANY valve, check or otherwise- WILL leak- it's only a matter of rate. The only exception to that rule is when the check valve has been installed backwards by accident- then Murphy dictates that it will seal tigher than the hubs of hell. If those leakages didn't happen, the pumps would have a head versus flow "curve" which was a perfectly vertical line, with constant capacity at constant rotational speed irrespective of discharge pressure.

 

[stanier]

Molten metal,

OK I understand and see this scenario occuring. Metering pumps are not some thing I encounter in mu waterhammer analysis. High pressure suction conditions is not something that is usual for a large reciprocating pump except perhaps for crude oil booster stations. I have not worked on one of those for a couple of decades.

So it comes down to the type of pump the operating conditions and the system design. So the engineer needs to design the system and not generalise!

For details of the reciprocating pump the bible is The Reciprocating Pump 2nd edition by John E Miller. The afforementioned websites give plenty of details of centrifugal pumps. Aslo the likes of Sulzer, KSB, Warman, GIW, Flowserve give lots of stuff on these pumps. Geho are probably the worlds leader in terms of reciprocating piston or piston diaphragm pumps. I used them to pump 75% fly ash 11km to a waste dump.

http://www.abebooks.com/9780471854678/Reciprocating-Pump-Theory-Design-Use-0471854670/plp

“The beautiful thing about learning is that no one can take it away from you.”
---B.B. King

http://waterhammer.hopout.com.au/

 

[stanier]

This resource may be of use.

http://www1.eere.energy.gov/manufacturing/tech_deployment/pdfs/pump.pdf

“The beautiful thing about learning is that no one can take it away from you.”
---B.B. King

http://waterhammer.hopout.com.au/

 

[curtis26]

I haven't been on here in a while, but saw this thread and wanted to chime in late.

I worked on a comparison between Reciprocating Plunger Pumps and Horizontal Multistage Surface Pumps about 5-6 years ago. The application was for and oilfield salt water injection project. The comparison was never completed, but the reciprocating plunger pump packages had overall efficiencies around 90%, while the Multistage plunger pump units overall efficiency would peak around 75% at BEP. The Multistage pumps with higher efficiencies had narrower "sweet ranges" so there were only a few flowrates pre-set by the manufacture that offered more than 65% overall efficiency (and when do the flowrates that you size a pump for ever end up to be your precise opperating point one the well matures and you comission the pump?). Yes, the energy saving for the reciprocating pump was huge for the comparison of these two pumping technologies in our specific application. Part of the unfinished portion of this comparison was the life cycle operating cost, but the reciprocating pump had plungers and packing that had to be replaced every couple of months and pretty much required personnel to make daily rounds for a preventative maintenance program while the multistage CP was pretty much a turn it on and let it go type pump. The Multiplex pump however could be worked on in the field and the CP pump required the Fluid end to be shipped to a shop anytime it went down.

The Reciprocating Pump Package did require, Safety Relief Valves, A gear box, suction stabilizers, Discharge Dampeners, Charge Pumps, Etc. so the cost was about 40% higher for than the Multistage CP's. It also seemed like the flow rates were high enough that three Reciprocating pumps were required to two CP pump packages.

Another thing we had to consider were the available Materials of construction. We had a lot more options in this particular case for the Multiplex Plunger pump which would come into play for corrossive injection water service.

The rule of thumb that was come up with for these two technologies for this applications was that under 500 GPM and over 2000 PSI to use the Reciprocating Pump, and for over 500 GPM and under 2000 PSI to use the Multistage CP.

I guess 1) over 500 GPM and over 2000 PSI, 2) Under 500 GPM under 2000 PSI, etc. is where the argument for either technology could be debated. It is likely that these areas of low flow high pressure and high flow high pressure is where the majority of these debates for the past 100 years have been centered.

 

[stanier]

A multiple piston diapragm pump ( as per Geho) would cost more but solve the materials selection challenge.

“The beautiful thing about learning is that no one can take it away from you.”
---B.B. King

http://waterhammer.hopout.com.au/

 

[Yobbo]

Hello everybody,
I seem to have unleashed a lifely discussion. First I want to thank all for their reactions. I have a few remarks though. I read some conflicting statements about the efficiency of the PDP versus the CP. I agree with Curtis26 that a PDP has intrinsically a better efficiency compared to a CP. I still think that a difference of 15% between 75% and 90% is quite significant. It is interesting to see how the life cycle costs determine the ranges where a PDP will be advantageous to a CP. It shows that although the technology is not so dynamic costs of labour can influence the transition areas. In my opinion power efficiency does too.
I wonder if it wouldn't be interesting for pump manufacturers to try to develop PDP's with less disadvantages. I was hoping that maybe rotary DP's would be a good candidate, but I haven't seen much proof of that.
I agree that engineers are always inclined to create an optimal solution, but sometimes saving energy there where oil or gas are abundantly available the motivation to invest accordingly is very low. Apparently the financial incentive is far more important than energy saving considerations. And my gut-feeling tells me that in the old days environmental considerations or global issues like climate change were of less importance. That is what I meant in my introduction as well.
As far as forward flow concerns I do not see the advantage of the CP versus the PDP. Even in my practice, where long pipelines are connected to CP's, it is common knowledge to provide the pump with a bypass line and a check valve. If the suction pressure gets higher than the discharge pressure the check valve will open and sustain the forward flow. This also prevents pressure transients that might damage the pump or the pipe line.

Karel Postulart, The Netherlands
Nuon Power Generation