Centrifugal Pumps

down to 40 Hz, That can be a stretch.

Do not know what you mean by efficiency, over all, energy in energy out......

Going from my experience, Don't know what the viscosity of the syrup. but you will have a built up of heat. I would have to say no.

Why do you want to slow the motor down? To decrease shear? Or to decrease flow?

I think they just what to settle an argument

Could be just the process. control feeding a system, used a freq drive on Ultra-Filtration units to control start-up and maintain flux.

Look at the pump curve.

pump curves don't state efficiencies

When I was in centrifugal process pumps all the curves I saw and used stated efficiencies.

In general you would expect to see a lower efficiency at a lower speed (ie same size pump with duty point at BEP, efficiency will be lower on a pump run at 4-pole speed than on a pump run at 2-pole speed). In this instance the viscosity of the fluid could complicate the issue.

In this instance it is impossible to give an answer without first seeing both the pump curve and the system curve, otherwise you don't know at which part of the duty point will lie. Considering the affinity laws (flow change is proportional to speed change, head changes to the square of the speed change and absorbed power changes to the cube of the speed change), and depending on the system design, you might even find that the pump won't give the required head at 40hz, therefore effectively running at closed-valve, meaning efficiency close to zero and a very short pump life.

Repeating Lynlynch's question, why do you want to slow it down?

Depends on the product and temperature. Pump curves are usually based on water and a given temperature

When I used pumps I normally work with these pumps and Pump curves at Alfa Laval

I don't believe there is an application here, but you give some good information with head change.

Indeed, and good curves should indicate what they are based on. And the whole thing is complicated by viscosity, both in terms of pump performance and system performance. Unfortunately we don't have enough info on this application to give any clear answers.

I have to say I never had much to do with hygienic pumps, but I used to know Alfa Laval quite well (had them for a distributor in one country many moons ago).

ahhh....G&H or Triclover....they were always sister companies till they became Alfa Laval

GA as far as I'm concerned. Most pump manufacturers had performance curves for both 50 and 60 cycle speeds in their catalogs when I was picking them out and there was a considerable difference in their heads and capacities in just that differential.

The BEP was clearly indicated along with the various efficiencies at the different flows and impeller diameters. The best efficiencies were generally close to the maximum diameter. After all, the centrifugal pump is named for the centrifugal force imparted to the fluid which in turn is based on speed and distance from the center of rotation and the efficiency is affected by the distance between the tip of the impeller and the cutwater.

I too find it hard to believe that the system resistance would be reduced to the degree that the reduction in head that would accompany this speed reduction would allow flow, practically any flow.

See the following: http://www.engineeringtoolbox.com/affinity-laws-d_408.html .

There is a note on this page that contradicts pritam #15 "Note that the affinity laws for fans are not identical with pumps." which pritam might want to review.

I would assume that for this service the lines and pumps are probably heat traced to maintain some consistency as far as the viscosity. I am picturing a material along the lines of an artificial "maple" syrup.

Reduction of the speed to 40 HZ is acceptable for most VFDs, but about 40HZ is the limit. With some VFDs it is possible to go slightly above 60HZ.

Yo will maintain roughly the same pump efficiency at 40HZ because you are pumping a slightly viscous fluid.

How many people are involved in this debate ?

The pump performance (flow/head) will be much lower at 40Hz than 60. Whether this is acceptable depends on the system and whether it was oversized in the first place.

There's been a recent discussion in this forum about speed and performance of centrif pumps.

Codey

Energy saving can be obtained by reducing the speed of the pump. It reduces the noise level of the pump. Water hammer can also be avoided at low speed. Pump performance curves need to be consulted for variation in efficiency and other parameters.

Centrifugal pump also follow the Fan Laws. For Speeds N1 & N2, Capacity or flow rate (F), Head or discharge Pressure (H) and Power consumption (P) are related as follows:

F1/F2 = N1/N2

H1/H2 = (N1/N2)**2

P1/P2 = (N1/N2)**3

So. by reducing speed to 67%, flow rate will reduce to 67% and symultanously discharge Pressure will reduce to 44.5% of original, if suction condition and discharge valve opening are the same. Shaft Power requirement will go down to almost 30%. So be sure what are your required parameters before reducing the speed. You can vary the flow rate and discharge Pressure by operating discharge valve.

will the pump still has the effeciency with reduce speed?

Only I can say with confidence that the speed reduction is the most efficient way to run a centrifugal machine at reduced capacity. So, the pump will still have the effeciency with reduced speed, but may not be the optimum one. Please check pump performance curve with different speeds. (get from OEM). If performence curve is drawn with water, multiply all parameters by ratio of density of fluid handled and water.

Don't forget the viscosity. It could have a greater effect on pump performance than density and, depending depending on the proportion of friction losses in the TDH, could have an equally great effect on the system curve.

Lots of issues raised, BUT-

FIRST issue- Is the use of a centrifugal pump the "right" choice for the defined use? Highly viscous fluids (and sugar syrup at near-room temperature- or even fairly warm- is relatively viscous) do not usually behave well with centrifugal pumps.

SECOND issue- Doesn't the syrup cause fairly rapid wear on the surfaces of the impeller due to the "buffing" action of the dissolved sugar?

Assuming that your reason for reduced speed is to reduce flow, be aware that the reduced speed will DRAMATICALLY reduce available pressure head, so filters may not work correctly OR flow will be substantially reduced, not just 66% due to speed reduction.

I strongly suggest that you investigate a stainless steel gear pump (positive displacement, consistent pressure output, and fully variable volume using a VFD. Also- you will need to use a VFD designed for constant torque output, not the "standard" models that assume torque will diminish as speed drops.

Such a pump change will allow use of the VFD to provide controlled flow rates AND reduced energy usage without any performance compromises OR high maintenance costs due to prime-mover wear.

I'd be more concerned about wear or damage to the shaft seal than the pump (sugar syrup shouldn't cause any wear problems to the pump itself). Crystallisation or cold starts could damage the seal, but if this is a problem it would apply equally to a gear pump as to a centrif.

The deciding factor is really the viscosity. As this increases, in a centrif the efficiency drops and power consumption increases drastically more than in a gear pump. You will reach a point where the initial cost saving of using a centrif is counteracted by the smaller motor and reduced energy costs of using a gear pump. That is when to use a gear pump, otherwise, always use a centrif if you can.

I'm also wondering if the OP is expecting the reduction in flow to be proportional to the reduction in speed. That would be the case for the gear pump, but the centrif would probably only require a small speed reduction ,

Here is what is going to change when you pump viscous fluids with a centrifugal pump:

• The brake horsepower requirement will increase.
• You will notice a reduction in the head the pump will produce.
• Some reduction in capacity will occur with moderate and high viscosities.
• The pump's efficiency will decrease.

Follow the following thread

http://www.mcnallyinstitute.com/14-html/14-04.htm

after you obtain the performance curve for your particular pump either from your Installation Package or the pump manufacturer. You should then be able to settle it among your "circle".

Nothing should be as satisfying as providing your own answer to prove your point as opposed to someone else giving you the answer.

A centrifugal pumps absorbed power relates to mass flow.

If a pump is not delivering at full mass-flow - it draws reduced power.

So, if in your system the pump output is throttled - it draws power proportionally to the permitted mass flow (+ the usual electrical and frictional losses)

I.e. if you reduced the speed to 'just equate' the desired output at 'full down stream process demands', it's still the same mass flow - so the same power.

There may be some efficiency gains against frictional losses in a viscous fluid, if that speed reduction is large.

This might also be offset by motor losses due to operating at other than optimum design frequency.

All in all, I'd say unless this pump is vastly mismatched to the process - little, if any, gain is to be found.

"A centrifugal pumps absorbed power relates to mass flow."

If all things remain the same, then yes, but generally all things do not remain the same when you change the speed. In fact, pressure and viscosity have a greater effect on absorbed power than mass flow. The change of flowrate also affects the correction factors used for head and power when doing the viscosity correction, and much depends on where the duty point is on the pump curve, and how that changes.

"So, if in your system the pump output is throttled - it draws power proportionally to the permitted mass flow (+ the usual electrical and frictional losses)"

Reduced flow results in reduced absorbed power, but efficiency is a different matter.

"There may be some efficiency gains against frictional losses in a viscous fluid, if that speed reduction is large."

In fact it has more to do with the system and fluid than the pump or pump speed, and overall efficiency (in terms of kg moved per kw absorbed) could go up or down. Two examples:

1) If it is a very simple system, drawing for example from a vessel next to the pump and pumping up to a vessel at, say, 30 feet above, with few bends or fittings, then the frictional losses will be a very small proportion of the overall TDH required, which might be not much more than 30 feet. This results in a very flat system curve, ie. the head changes very little as flow is increased or decreased. Where this curve crosses the pump performance curve is where the duty point will be. Centrif pump curves these days are also relatively flat. Reducing the pump speed will have the effect of moving the pump curve downwards and slightly to the left. So a small reduction in speed will move the duty point to the left (TDH will change very little) resulting in a larger reduction in flow. Absorbed power will be reduced, but (assuming the pump was originally operating at or below BEP) pump efficiency will also be reduced. So the net result is an overall reduction in efficiency.

2) If it is a much more elaborate system, with say 200 feet of discharge pipework including many bends and fittings, as well as the 30 ft lift, then the friction losses would constitute the major part of the TDH, and the system curve would be quite steep. The same small reduction in speed would move the pump curve the same way, but the reduced flow in the system would result in a much greater reduction in TDH required. We could assume, for the sake of this example, that the steep system curve is such that the duty point on the pump curves actually moves along the line of the system curve, which would mean a smaller reduction in flowrate and the pump efficiency would remain more or less the same. But the reduced flowrate also means the viscosity correction factor for absorbed power is also reduced. So the net result could actually be an overall increase in efficiency.

So the answer to the OP's original question is - It depends.....more info required.

I would question most of this:

1) He has heard of the affinity laws, they are mentioned in posts 8, 15 & 17, including an explanation of how and why they apply.

2) How can you say the efficiency will be the same when you have seen neither the pump curve nor the system curve, and therefore cannot know at which point on the pump curve the duty point lies before and after the speed change. We also don't know the viscosity which will affect things.

3) Why will the voltage change? If they are supplying 380v to the motor it will be 380v.

4) Yes, the load on the motor will change.

5) Yes, the pump load will change.

6) Yes, the flowrate will change.

7) How can the motor amps stay the same when absorbed power/pump load/motor load are reduced?

8) Efficiency is not constant, and is likely to change even if the duty point is on exactly the same point of the curve (eg BEP) after the speed change.

9) Why drop only to 45Hz? A VFD and 60Hz motor can easily handle 40Hz. The only reason for not going this slow, as has been mentioned in previous posts, is that you might not need much speed reduction at all to get a significant reduction in flow, but we can't give any recommendations without seeing both pump and system curves, and knowing why the reduction in speed/flow is required.

10) If you read the original post you would notice they are already using a 4-pole motor.

one other item, if the pump is not sized correctly you may get an overcurrent.