Limits on allowable maximum and minimum impeller diameters

[Tim37]

I have seen specifications for pumps which state limits on the actual impeller diameter: no more than X% of the maximum diameter and no less than Y% of the minimum diameter. Has anyone seen specific values on these limits? What are they typically?

 

[TenPenny]

I don't really recall seeing any minimum limits, but I have seen both 90% and 95% of maximum, that's quite common.

 

[1gibson]

Within 5% of max/min is common, but the requirement is often misapplied on vertical multistage pumps, or pumps where higher or lower head/flow impeller can replace the current impeller.

The main reason for the requirement is to cover future head rise, if a pump isn't sized correctly this can prevent the need for a completely new selection. Suppose this also applies if a pump is grossly oversized.

In some pump types (single volute OH2) there could be effects on radial loads with max diameter, and on any pump type there could be discharge recirculation with min diameter. But there is a reason the pump vendor has a max and min diameter and they should be designed to avoid these issues.

Example: a multistage vertical pump (VS1 or VS6) can meet conditions with 19 full diameter impellers, or 20 trimmed impellers. Provision for future head rise can be had by designing for the addition of a future stage.

If customer doesn't understand this, the reject the 19 stage pump, demand the 20 stage. This increases their shutoff head (generally not desireable) and then in the future when they need to increase head, they have to purchase multiple full diameter impellers, instead of 1 impeller and 1 case.

So, if the pump vendor can meet future head rise requirements in a reasonable manner, let them provide full diameter impellers.

Minimum diameter, there might be some merit in avoiding but only if there is not an alternate impeller.

 

[CaracasEC]

API 610 is a good reference for this question.

 

[DubMac]

The reason for the % diameter restriction in specs is for head reserve capacity. As stated above, original C.O.S. miscalculations or future changes can be dealt with much easier in horizontal pumps by just getting full trim impellers. And as also alluded to above, API requires I believe??

Between 5-10% is typical for specs.; more than that and you've selected the wrong pump. You may also find concurrent requirements that the motor and baseplate be ORIGINALLY sized to accomodate the max diameter impeller; no matter what size is required by original C.O.S. If conditions change or were miscalced originally, it can become VERY expensive to upsize existing baseplate, foundation, motor, starters, etc......

Just get a copy of Shell, Exxon, Conoco, etc. pump specs and use as a draft if you are trying to write your own specs. These specs are all based upon decades of experience, and though they have become a bit overbearing these days, some of the basic intentions remian; one of them being the head reserve clause. JUST DO IT.

Verticals, instead of having a head reserve value for each stage, should make sure there is room in the original hole/can to add a couple stages for future use. Multistage verticals should be designed with all full trim impellers; just trimming the intial stage to fine tune the head. Trimming one stage preferable to trimming all of them.

 

[Clvet]

I once got a kick out of seeing a spec written to include both a "no more than 95% of max impeller diameter" and also, "will operate on a VFD.

The most reasonable way to specify the pump is to simply list what you want the pump to be capable of in the future and let the manufacturer determine how to size for present and future conditions whether it's increasing the speed the motor runs at or the diameter of the impeller.

 

[1gibson]

Oh, I forgot: the example I mentioned above wasn't exactly hypothetical, and future head rise wasn't even an option because the pump's MAWP was already at the limit of the flange rating!

"But the spec says....."


API 610 does not include any relevant requirement for % of impeller diameter, only for future head rise.

 

[CaracasEC]

API 610 does not include any relevant requirement for % impeller diameter, only for future head rise but take note that for a pump system, there are two heads; the head required by the process and head produced by the pump and to increase the head is a function of whether increasing the impeller diameter or increasing the speed for the same pump and application.

 

[Artisi]

Without VFD, talking about speed increase is really a nonsence as you can only change speed by changing motor speed ie, 6 to 4 a 2.25 times head increase, 4 to 2 pole, a 4 time head increase. This is way beyond and minor tweek that might be needed for a pump installation.

My experience has been for a well engineered design a max reduction would be somewhere between 90 -95% of full diameter, provided it doesn't sacrifice pump efficieny, this of course also depends where the pump duty falls on the H/Q curve

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[GPRnD]

I think you need to be careful to consider pump type when setting limits. Volute pumps are more tolerant of large trims and their efficiency does not drop off as quickly as diffuser style pumps.

On diffuser pumps for example is is customary not to trim them much beyond 90% hence you can see the futility in a spec that calls for a 90% maximum diameter....

My recommendations are:

For volute pumps
Maximum diameter 95% to 97.5% of full diameter (97.5% is API 610)
Minimum diameter (at <= 1000 Ns) 80% of full diameter
Minimum diameter (at >= 2000 Ns) 85% of full diameter



If you are using a VFD, forget the above. You only need a trimed impeller if you want to guard against ever having going above a motor limiting speed.


For diffuser pumps
Maximum diameter 97.5% of full diameter
Minimum diameter 90% of full diameter