How to handle too much head? 1

[ProjectEng]

I have a vertical turbine pump that puts out too much head for the application. Is it advisable to partially close a manual butterfly valve on the discharge side of the pump to create a pressure drop?

Thanks.

 

[Artisi]

The correct proceedure would be to re-rate the pump, by either an impeller change or impeller diameter reduction on one or more of the impellers if it is a multi stage unit.
Closing a valve on the discharge is an option depending on how much head you need to "burn-off" across the valve, but this comes at a cost in power consumption.

International College
Naresuan University
Phitsanulok
Thailand

 

[PUMPDESIGNER]

By too much head I will presume the following, correct me if I presume wrong.

Are you referring to shut-off head? If so, then that is a common error in pump selection where the design point is met by the pump, but the pump has a steep curve geometry. This often happens because the cheapest pump that meets any design point often has a steep curve geometry and possibly a brutal shut-off head, producing tough and even nightmarish scenarios for control problems and system damage.

If that be the case, then you can do the following:

1 - Best solution is very often pump replacement with a flatter curve geometry pump.

2 - If you can accept lower pressure at the required flow rate, then you can trim impeller(s).

3 - If Shut-Off head is not too much higher, then controls can solve, but it will never be as good as a flatter curve geometry. I have done this for customers who got bad pumps, but I would not do that if the shut-off is brutal. You cannot take 225 psi and knock that down to 60 psi without having the valve scream, so you need multiple valves or super high cost one valve, with pressure relief back-up valve (pilot operated), and probably high pressure shut down alarm.

4 – Install VFD with aforementioned high pressure alarms, high pressure shut down, pressure relief, etc.


PUMPDESIGNER

 

[abcmex]

ProjectEng,
You may close slightly the discharge valve to adjust the discharge pressure but be careful, not all butterfly valves can be used for this. If the speed is too big you may damage the valve gasket.

For a better advise additional data would be needed. As posted before there are some alternatives: Impeller trimming/exchanging, by-pass, VFD, orifice plate or new pump ;-)

RGS

 

[25362]

If one is willing to pay for the added power consumption on a steep pump curve, I think that opening a by-pas back to suction would increase flow rate moving the head downwards. Of course, care should be taken to avoid cavitation by moving too much to the right of the pump's BEP.

 

[ProjectEng]

Pump specs:

Disch pressure: 94psi @ shutoff, 55psi at design
Power: 1022hp at design, 700hp at shutoff
Rated head: 133ft
Min flow: 11,000gpm
Normal flow: 24,500gpm

The system head is around 60ft. The pumps are driven by 1000hp motors so VFD's are not an option due to price. The pumps are single stage.

 

[abeltio]

Partially closing the discharge valve does not alleviate the pressure upstream of the valve, creating a high pressure drop across the valve with the associated noise and damage to the valve (cavitation).
In gas turbine world the lube oil pumps are usually oversized and the excess press/flow is disposed of by means of a pressure regulated valve that by-passes the discharge back to the suction tank (not the suction of the pump).
The pressure regulated valves usually have a min flow and open more to bypass pump discharge as the pump discharge pressure increases...
They are also fitted with a pressure relieving valve in case the pressure regulated valve fails.
In any case, the system should be designed to sustain the full pump discharge pressure.

HTH

saludos.
a.

 

[GaDawg]

This is a common characteristic of vertical turbine pumps, which sometimes can be good and sometimes can be bad, depending on the application. My primary concern would be the pressure ratings of piping and components within the system. If they are rated for sufficient pressure (suction plus shutoff) then they shouldn't be a concern. I consider anything over a 40% rise to shutoff excessive, due to problems valves sometimes have regulating large differentials. Certainly the most effective way to eliminate the rise to shutoff is a vfd, but based on the prohibitive cost, a by-pass relief valve might be your best option. I use this method quite often to prevent the excessive pressure from entering the system. A partially closed discharge valve will reduce all the pressure but at a cost to system performance, energy consumption and valve life. Also, unless the discharge valve modulates, the pressure drop across the valve will decrease with a decrease in flow, having a constant valve coefficient. I believe this is the opposite affect you would be seeking. I hope this is helpful.

 

[hydrae]

Are you sure a VFD is cost prohibitive? The VFD may pay for itself in power savings, unless you have cheap electricity, use the affinity rules for lower speeds to see how much power you can save and compare to the cost of the drive, small decreases in speed give large decreases in required power by the cube of the speed. In the larger sizes the cost per Hp drops in VFD prices.
Include in the installed cost of the drive input and output filters/reactors if required and upgrading the motor to VFD rating, older motors may require a grounding ring and or winding and insulation upgrades.

Hydrae

 

[MortenA]

You should look closely at pumpdesigner advice no. 2) Replace or trim the impellar yourself if what you need is a general "flattening" of the pump curve.

If you on the other hand need a large operational capacity range then look for either bypass or throttling (assuming that your own statement with regards to VFD is correct). Look at the pump duty curve to determine witch has the lesser power consumption.

Best regards

Morten

 

[Artisi]

to ProjectEng re your thread:

Your figure have me confused - or maybe I'm just a bit thick today :-

"Pump specs
Disch pressure: 94psi @ shutoff, 55psi at design
Power: 1022hp at design, 700hp at shutoff
Rated head: 133ft
Min flow: 11,000gpm
Normal flow: 24,500gpm

The system head is around 60ft."

Questions:
55psi (127ft) at design -- what design ??
Rated head 133ft -- is this the designed duty head ??
The system head is around 60ft -- is this the current operating head ??
Minimum head is understood.
Normal flow - is this the rated flow or the current flow ??

More questions:-
Was the original operating duty 24,500 gpm @ 133ft??
Is the pump now operating at a head of 60ft??

Please clarify the original duty ie, flow and head for the system design.
Give your best estimate of the pumps operating point now.
Indicate that duty you are trying to achieve.

Irrespective of the above clarifications - throttling or by-passing flow on pumps of this size (even 1 - 2 feet of head) as a means of a permanent re-rating to another duty is a sheer waste of power. Establish what your really need and speak to the manufacturer about re-rating the pumps.






International College
Naresuan University
Phitsanulok
Thailand

 

[ProjectEng]

These pumps have not run yet. System piping is being installed now. So the system head of 60ft is just a rough estimate. The pumps were originally spec'd for 133ft of head probably to make sure that we'd have enough pump.

Can't really explain the difference between rated head and design head. Perhaps design head was specified and the pump is actually good for continuous duty at rated head. Normal flow is the rated flow.

 

[25362]

Do 60 ft of system head include both friction drop and static head ? Is it possible that the original design considered a larger flow rate so as to increase so-called system head ? Anyway, as abeltio said, the system should be able to stand the full pressure of the pump discharge.

 

[saxon]

Projeng, Before modifying the pump to accomodate "What if" scenarios, bring the unit on line, collect data on the as installed condition (the pump manu'f. can do this for you), then modify the pump to the as installed condition and actual requirements.

If the "what if" scenarios are the operational reality, then, as mentioned earlier, either have the manu'f. trim the impellor or install a bypass from the discharge back to the suction side. But be forewarned, any modifications to the unit without manu'f. approval will in all probability void the warranty.

Hope this helps.
saxon

 

[ProjectEng]

saxon,

Yes, I plan to start up the pump and observe real world conditions before making modifications.

Just a little concerned that I could have 38psi at the cooling tower inlet that is rated for 17psi and could end up with blown-out nozzles, etc <g>

I just threw this question out to prepare myself in the event the pump really does put out too much head.

Thanks to all for your input.

 

[PUMPDESIGNER]

Actually, operating the system and then modifying carefully to reduce head (probably by trimming impellers), is a great way to save money, in fact recommended by US Gov't. all the time as opposed to using VFDs.
When it can be done, sometimes the best and most efficient system is one that is fine tuned on site by simple permanent modifications, as opposed to building and paying for control flexibility.

PUMPDESIGNER

 

[GaDawg]

Just remember, trimming an impeller or configuring valves to impose head loss does not change the pump characteristics. Trimming the impeller will lower the magnitude of the duty point and shutoff head, but the net effect of a rise to shutoff remains. I’m aware of a number of cases where fixtures or nozzles are damaged because of the rise to shutoff. If this rise is enough to cause system damage, then it should virtually eliminated using a VFD, by-passed using a relief valve or eliminated completely by selecting a new pump with a “flatter” curve.

 

[ProjectEng]

The pump curve is not steep at all; I can reduce discharge head by approx 40ft by trimming impeller all the way down without a significant effect on flow rate. And since the eye of the curve is in this area, efficiency doesn't change much.

Again, I'm not going to do this before I run the pump. But I am just coming up with a good Plan B.

Thanks.

 

[1ton]

ProjEng, A safe suggestion, to ensure that you do not blow out the nozzles, due to over pressure is to install orifice plates in the system. I think this will be much cheaper and safer. Butterfly valve should be design for trottling. I hope that is spec in when you buy the valve. Because if it not then your valve will not live very long.
After running the system, then you can determine, if you want to trim the impeller, or you can live with an orifice plate in the pipeline.
Have Fun