Centrifugal Pump Running with Throttled Discharge Valve

A likely possibility is that the actual head against which the pump operates is less than anticipated in the design. This results in the pump operating farther to the right on its curve, and hence subject to overloading. It might be a good idea to trim a bit off the o.d. of each such impeller.

Wrong idea; risk of cavitation.

Shouldn't trip at FLC....that's its optimum design.....how are your overloads set?

Power drawn by centrif pumps with discharge pressure has been discussed more than once on this forum. What you describe is not untypical. Motors are often sized to cover the highest power drawn, rather than at duty point, specially on smaller units where extra cost is low. But it seems not in your case. But if performance is OK throttled to 60 - 80% what's the problem? Adding a VSD would give current duty at lower speed with wide open valve, and save energy. If you need the extra duty at wider valve opening it sounds like you need a bigger motor.

The motors are undersized for the application or the motor protection is not set up correctly.

Monitor/measure and record the motor current continuously through a duty cycle to verify if the motor is exceeding nameplate rating.

If the motor is not exceeding the nameplate rating then examine the motor protection settings and correct.

Caution: Most motors are not designed to operate in an overloaded condition for extended periods of time. If you operate the motor in an overloaded condition for too long the motor winding will fail.

We can't say the motors are undersized for the application until we know whether pump output is enough for the process in the 60-80% throttled condition, and why the OP has tried opening the valve further. He hasn't said what the problem is, if any.

Throttling/restricting the discharge of a centrifugal pump lowers the motor amperage because there is less flow therefore less HP demand on the motor.

Increasing the opening of the discharge valve causes even higher amperage on the motor.

If a centrifugal pump motor cannot run with the discharge throttled without tripping and the overload protection is set correctly, the odds are very high that the motor is under-sized for the application or there is a mechanical overload being caused by a pump component or motor component failure.

There is another possibility in that the voltage rating and/or frequency of the motor is higher than the electrical supply however I would assume they have verified this rudimentary requirement. (Maybe not?)

Throttling/restricting the discharge of a centrifugal pump lowers the motor amperage because there is less flow therefore less HP demand on the motor. Not necessarily. That's true for most centrifs but not all. This has been discussed recently on CR4. Usually on throttling the flow falls by more than the rise in head, so (assuming constant efficiency) power drawn falls. But with some pumps it's the other way round. Many axial or mixed flow pumps show a sharp rise in power as closed valve is approached, so unlike most centrifs, should not be started against a closed discharge.

If a centrifugal pump motor cannot run with the discharge throttled without tripping..... OP (lord rest his soul) didn't say that. It was OK with throttled valve, tripped when valve opened wider.

Thanks for the corrections and clarification.

My WAG is still an undersized motor especially since more opening of the discharge valve results in tripping of the motor.

I prefer measurements and verification of electrical DATA and protection settings at the site over WAGs every time.

We have a very large quantity of centrifugal pumps from many different manufacturers that handle a broad range of fluids and the incidence of new pump skids suffering from not enough horsepower has dramatically increased over the last 10 years.

It appears the various groups providing our replacement pumps have lost site of the fact that lifting and/or moving fluids through a given resistance requires "X" amount of HP.

Hopefully the OP utilizes all of the good input and figures the problem out.

Happy New Year!

The pumps will run very hot.

All that energy that doesn't circulate fluid is retained in the pump volute as heat.

Trim the impellers as Tornado has suggested.

I don't think we can make these statements until we get more details from the OP. We don't even know he has a problem, maybe he posed the question out of curiosity.

As it's thermic fluid service it's likely to be closed circuit, but that doesn't mean heat build-up will be a problem. Central heating systems are closed circuit so the energy put in by the pump goes into the water, but it doesn't matter, just needs a little less heating fuel.

Bypass to pump suction is only needed if the pump can be dead-headed, and OP hasn't said that can happen.

If you retain 80% of the heat generted by a pump within the pump housing it WILL overheat.

Even if the pump is stainless it will still fail eventually if throttled that much.

But that will only happen on closed discharge, or close to it, and we don't know there is any risk of that. You don't worry about your central heating pump overheating, do you? Because it's never throttled down to zero (and if it was, it would be more of an issue for the boiler than the pump).

IMHO There's a lot of unwarranted assumptions in this thread. OP - how about coming back?

Also known as dead heading. This can only occur for short periods of time.

Please see #66 for some more possibilities for this situation. Make sure to look at the site referred to. Good Luck, Old Salt

Standard practice in chemical plants is to run a valved loop from the outlet back to the inlet of centrifugal pump. This always the pump to run at full speed but with less load when it has a reduced load. When more flow or pressure is needed the by-pass valve is closed more. For less pressure/flow the valve is opened more. Throttling the discharge puts more load on the motor and pump when less pressure/flow is needed. With the by-pass valve is closed the pump and motor will usually make slightly more noise and draw more current. This set up is also used on many other type pumps. Full load current will only be drawn when the by-pass valve is completely closed.

Do not reduce the diameter of the impeller. Once this is done there is no way to get more liquid/pressure through it if needed in the future. Also, why use a larger pump with a reduced size impeller when a smaller and cheaper pump will do the job.

The best way to regulate the centrifugal pump is to use the by-pass valve of a variable frequency drive with an adequate drive control and pressure sensors.

This bypass is the same way that a maximum pressure by-pass valve works to control the maximum pressure of the pump on gear pumps and other styles.

Good Luck, Old Salt

Better solution than impeller trim, I agree.

I've done this before with plastic chemical pumps to keep them from melting, but we ran them into a sump to help cool the fluid.

Bypass will work as you describe, but still run hot.

Trimming is done by manufacturer Lyn.

Prior to delivering your purchased pump. For an existing application, it should be that there is a foul play or fault on the pump, fluid dynamics or the drive, no more no less.

OOn my phone. I'll make this short.

You don't have a clue.

No, i know malware. Nothing to hide anyway.

No, not all the time......

Both you and Old Salt seem to know what you are talking about. I don't. So I will ask you if you think viscosity is a possible factor? Are impellor pumps made for water, or a liquid with similar viscosity, only?

Jim

Pump curves are based on water @ 60-70 degrees F. Other fluids, they would approximate.

Thanks. If a need to pump honey, for example, with a centrifugal arose; would the impellor be modified to do this?

Jim

You wouldn't use a centrifugal. You'd use a positive displacement pump.

You won't be able to,go,off the standard pump,curve, but with the viscosity, it would be better to have a positive displacement pump.

Thanks, I understand that but I just wondered if a thicker fluid would contribute to heating up the outfit. If so, then; would impellor mods be in order?

Jim

The problem that you have is the suction head requirement would cancel your output. Even with a PD pump, you would need to supple some type of pressure to feed the pump, whether its gravity or a stuffer pump.

Contact a pump supplier.

"Even with a PD pump, you would need to supply some type of pressure to feed the pump, whether its gravity or a stuffer pump".

Are you sure of that?

All PD pumps that I have been associated with are self priming and very good at it. Internal gear pumps such as the Gerotor, Progressive cavity pumps such as the Mono, and tube/Peristaltic pumps are especially good at self priming, even in vertical lift situations, they will often draw down to a near perfect vacuum, and they can easily handle high viscosity fluids such as honey etc.

One problem that has often been experienced with attempts to use centrifugal pumps to handle exotic fluids is their tendency to whip the product into a foam, which then causes the pump to cavitate.

Yes, I am sure! There must be pressure at the pump suction to push fluid in. If a pump can give suction lift of (say) 7m, there is pressure 3m water, 0.3 bara at the suction (1 bara atmospheric - 0.7bar lift). If there were no atmosphere the level in the suction tank would have to be above the pump for it to work. In practice a pump needs (absolute) pressure above zero, NPSHr is typically 3m minimum.

Not too many places where you are going to be using the pump with no atmosphere.

The gist of your previous statement was that you would either need gravity feed (where typically there is a hopper or such feeding product into the pump) or you would need a feeder pump, neither of which were correct for a PD pump. The atmospheric pressure bit now becomes a lifebuoy.

With the viscosity of the product, Gravity feed is the amount of head in the feed line and upper with the product. Depending on the PD pump you had 3-4 feet of head to stuff the pump..

Not sure whether you mean my previous post or #46.

I was just pointing out that any pump, PD or centrif, needs a positive suction (absolute) pressure. In principle this could be close to zero but always positive. Can be much higher, eg on an offshore oil job I worked on the high-pressure seawater injection pumps on secondary recovery needed 6 bara, provided by an upstream centrif.

Re using the pump with no atmosphere - it's not unknown to pump from a vessel under vacuum. I once worked with vacuum drum filters where we had to watch the NPSH for the filtrate pumps. Not a high vacuum, 20" Hg vac in old money, but I'd guess there are cases at higher vacuum.

At that viscosity yes, code master goes into detail.

As far as the foaming action, that is cause by shear the pump is putting on your product.

That "explanation" is incorrect.

Whoever voted this OT is spectacularly incompetent.

I did because the answer is spectacularly correct! What is your reason for #9 being incorrect? Instead of making a statement with no substantiation, a reason should have been given.

#9 was verified with "Perry's Chemical Engineering Handbook", the bible of chemical engineering since it was first published in 1924. I have an extra copy if one is needed.

This arrangement of pump, by-pass piping and by-pass control valve has been used in the chemical, cosmetic, pharmaceutical, food, beverages, and fluid handling industries since the 19th century. If this is not enough proof that it works then thousands, if not millions, of people have been doing it "wrong" for at least 130 years or more. The beer and liquors we drink were made with this type of piping. Some of the liquid intermediates and finished drugs that are used were made with this piping arrangement. If it is wrong tell that to Coors, Cosmair and Merck to name just a few.

This would make #12 seem inappropriate and incorrect.

Good Luck, Old Salt

I did not say that bypasses were not used. I said that they do not have the effect shown on your graph. More later.

I agree, the by pass valve pointed out by old salt is not what it seems applicable, never seen like that before except used in a differential pressure gage or transmitter. A stand by or spare pump (redundant) instead I should say is usually observed at site.

The symbol in the drawing is unclear whether it is just a bypass valve, or a relief valve/regulator. A relief valve is not generally useful on centrifugal pumps, but a bypass is quite common (sometimes designated as a "Qmin bypass"). However, it does not have the effect shown on Old Salt's graph.

The symbol in the sketch is for a hand operated threaded control valve. This has been the symbol all my life and many years before. Don't know when it was first used but I am sure it was well before my time. Do know that my father-in-law used it during WW2 in the Sea-Bees putting in pipe lines in Alaska.

Hand operated threaded control valve

not to be confused with

Which is a common short-hand symbol for most valves with in-line connections although the correct name is a threaded gate valve.

http://www.edrawsoft.com/images/shapes/pfd-valves.png

Centrifugal pumps sometimes have by-pass valves installed with them, especially high pressure ones such as used on a fire pumper vehicle. For a single stage pump the output is normally maxed out at 200-250 lbs. A dual stage pump will usually have a max output pressure of 400-500 lbs The by-pass valve is normally closed but manually opened to prevent the pump from running constantly when all discharge valves are shut closed. Many also have bypass valves to provide cooling when the pump is running hot from low flow useage.

In the sketch there is no relieve valve/regulator. Just a manual control valve.

This chart (simplification) indicates that with the valve completely open the outlet pressure is at its lowest, Q₃. As the valve is partially closed the pressure starts to rise, Q₂ (indicated by the flow reaching further from the pump). When the valve is completely closed the pressure increases to the systems maximum, Q₁.

http://www.waterworld.com/articles/print/volume-27/issue-9/departments/pump-tips-techniques/selecting-the-best-most-efficient-centrifugal-pump-control.html

Good Luck, Old Salt

That symbol is for a control valve, not a manual valve. See your own chart. In particular, the "dome" usually represents a pressure regulator. A simple straight line, as at top left in the chart, is for a hand valve.

In your graph, the pump operating point stays on the curve, not below. Opening the bypass decreases resistance and increases the total flow through the pump, moving the operating point to the right and increasing the power consumed. Had they been provided, the power curves would show this.

On the other hand, partly closing a throttling valve in the discharge increases resistance, decreases total flow, and decreases the power consumed. Just as described by the OP.

The symbol is ok. If it was a pressure regulator it would have had a sensing line from the "dome" to discharge piping. This has been accepted for a long time to indicate manual valve to control.

AS stated, the drawing/graph was a "simplification" for ease in understanding. No where was any reference made concerning power consumed by the motor. The whole chart and drawing was used to show the relationship between the position of the by-pass valve and the output pressure. This is clearly shown by the output streams on the pump illustration, with the stick person, shown as Q₁, Q₂ and Q_3.

"MERRY CHRISTMAS TO ALL AND TO ALL A GOOD NIGHT!" quote from Santa Claus as he flew from the roof top with his 8 reindeer or 8 reindeer and Rudolf the red nosed reindeer.

Good Luck, Old Salt

Throttling the pump will NOT put more load on motor.

Many possible cause, could be * bearing *shaft misalignment * change of head * change of fluid * motor winding failure * pump surge as this centrifugal pump- rotates reverse direction

As the valve is closed the less fluid circulates through the by-pass valve (approx. 1-2 ft of head) and a more goes through a system (piping) that has greater resistance to flow, the more power the pump needs to provide the required increased flow. Example- a sump pump merely circulating through a pipe with the discharge end 2 ft above the fluid level (less power required) than the same pump connected to a pipe with the discharge pipe end 20 ft above the liquid level, when controlled by the by-pass valve or most other throttling method.

The higher the distance from the pump to the discharge end the greater the pump has to work. One advantage of the by-pass valve is the motor runs at full speed unless it has a maximum load with the valve closed. More discharge pressure the less fluid circulates and lower power needed for motor. Motors generally work better at full speed.

Good Luck, Old Salt

Whilst I agree with the concept and have used it many times in the past, We normally did it with a 3 way valve at the junction of the bypass and the outlet, thus any throttling to the outlet would send excess product back to the inlet.

Even with this type of throttling, substantial heat is generated if recirculation is too great

I'm not a strong supporter of the by-pass valve arrangement, in lieu of throttling the output of the pump, for control of a pressure/volume system because of some of the reasons mentioned in various posts. There are many newer and better arrangements/equipment that will do the job better.

For the current OP I would probably spec a centrifugal pump, a few sensors, a plc or two, and a variable speed drive. If the product was a suspension or subject to bad changes, return piping to the feed storage container would probably be used. This basic system is often used in the pharmaceutical industry for the controlled transfer of both liquids and solids suspended in liquids. The advantage of this arrangement is that it can be used with so many different products and product types without negative results.

Good Luck, Old Salt

In my experience, the only time I've installed a by-pass on a pump it was on a sanitary PD pump for CIP purposes.

Again that is my experience, I'm not saying there aren't other applications for a by-pass.

Referencing this and a previous OP http://cr4.globalspec.com/thread/85832#comment958459

There have been many comments for or against throttling via a valve in the discharge line. One option, which I mentioned and then clarified about whether I thought it was with merit was the by-pass valve from the discharge back to the suction. Many centrifugal pumps have check valves on the suction lines in order to eliminate the problem of loosing the prime. The placement of the check valve and the throttling discharge valve and even the recirculation valve can be very critical and dangerous to personnel if placed/operated in an incorrect manner. The pump can explode, with potentially deadly results:

http://sache.org/beacon/files/2013/08/en/read/2013-08-Beacon-s.pdf

Be sure to look at this. DON'T KILL YOURSELF OR YOUR STAFF!

Good Luck, Old Salt

How can you tell that the magnitude is Full Load Current only? Any possibility of excess readings other than FLC?

Any how what kind of fluid is it pumping? Pls. Reflect on this equation Motor Rating = Density * gravity* flow rate* total dynamic head/(eff. of motor*eff.of pump) Any changes on the right side of equation corresponds to your over load. Motor Rating is fixed so as your motor safety devices.

Pump needs to be design to your needs, it may not be your motor size but your pump size, such as your impeller size. You need your requirements , and then compare it with your pump curve.

There's a lot of flying off going on in this discussion IMHO. No doubt the various suggestions - trimmed impeller, varispeed, bypass valve, bigger motor etc could be useful in the right circs, but unless we get a lot more detail from the OP we don't know if they apply here. As often happens, OP seems to have vanished!

Merry Christmas everybody

DETAILS?

From an OP ??

Just because it's Christmas, don't expect miracles.

How do you know the valve is only open 20 to 40%

Are you judging by the number of turns from full open or closed?

The flow is not linear with stem motion,and it varies with the type of valve.

A butterfly valve has about 90% of it's full flow well before 50% open occurs.

So how are you measuring the flow?

You may have more flow than you realize,unless you have verified your conclusion with a flow meter.

As regards the tripping of the motor, have the over load heaters been properly

selected,factoring in the Service Factor of the motor listed on the motor data plate?

Are any of the other motors or pumps having a problem?

More information is needed to give an intelligent answer.

There are many good sources for valve flow characteristics available for FREE on Google.

Check around,you will see.

If this is a control valve,controlled by a PLC,or such,and the output of the controller

is only 20 % to achieve set point then the valve could have been improperly spec'd

for the wrong flow characteristics,or the control loop,including the valve, may be out

of calibration.

The system could be designed to handle many different fluids,of varying S.G. or viscosity.

Check the original design blueprints and specs for the system.

Piping could have changed,shortened,or otherwise modified from original design.

A chemical plant is usually very strictly controlled and do not allow ad-hoc

modifications of the equipment for obvious reasons.

Especially in Pharmaceuticals.

Not even the slightest change in equipment or procedure is allowed without a

complete review and pre-approval.

(Two eyes of newt,instead of one, or a toe and a half from a frog could give a real witches brew.Double,double toil and trouble,Fire burn and cauldron bubble!)

You could also have motor or pump issues,such as bearings.

And as I asked previously,are any other pumps having this "problem"?

Is this a real problem, or are you just curious?

What a crock.

No wonder the OP never came back.

#1