Centrifugal pump air injection

[someguy79]

I'm working on a project to replace three large centrifugal sewage pumps. Among their many features are suction air injection ports. The existing pumps use air injection, but due to design changes with the new pumps, I'm not confident that the existing air flow rates will be appropriate for the new pumps.

Unfortunately I have not had great support from the pump manufacturer to comment on this issue. I will continue to pursue this avenue, but I can't count on it.

I've found very few references regarding this issue. Here are two of them:
Pumping Machinery Theory and Practice, by Hassan Badr & Wael Ahmed says "...a small amount of entrained air (below about 1%) has been found to cushion the pulsation effects of cavitation and consequently reduces pump noise and minimizes erosive damage."

The other reference is easy enough to read for yourself:

http://www.waterworld.com/articles/...tion-air-on-centrifugal-pump-performance.html

I am unable to find a consistent and quantifiable estimate of the air which may be used to control cavitation. The most common values I can come up with show air use rates between 0 and 2% of pump flow. However, none of these are specific about the units used for air. The pump flow is surely a volume flow, but what of the air? Can I assume that this is a volume flow also? If yes, what version of a volume flow? (i.e. scfm or acfm)

Does anyone have good information on this subject?

Thanks for reading and considering my issue.

 

[bimr]

It is % by volume measured at the pump intake connection.

Most people are more concerned about removing the air. Small amounts of air have been introduced in the pump suction on purpose to alleviate certain site problems However, in general, entrained air reduces efficiency, head, and capacity and it increases maintenance. It may cause the pump to lose prime.

What type of pump is this and what is the process conditions.?

 

[Artisi]

First up, if new pumps are being installed and they are correctly sized for NPSHa/r, there may be no need for air injection.

The reference noted "Hassan Badr & Wael Ahmed says "...a small amount of entrained air (below about 1%) has been found to cushion the pulsation effects of cavitation and consequently reduces pump noise and minimizes erosive damage." this is nothing new and an accepted practice used when NPSHa is (usually) less than NPSHr resulting in suction cavitation.

There is no constant or quantifiable magic number, it is derived by trail and error for each pump / installation and could well be different between "identical" pumps operating in a common installation, due to slight variances between the pumps.


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.)

 

[LittleInch]

I will echo the advice above - basically this (air entrainment) is a fix to a problem which shouldn't exist if you design the system right. SO first make sure it is actually needed.

Given that the air entering will be at atmospheric pressure as the inlet pressure will be less than atmospheric ( I would suppose), then the flow rates, if anyone ever measured them, would essentially be in scfm.

I would just run them and see if it is required then use the minimum amount tat stops the "crackling" noise, but this is not a good way to continue.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[someguy79]

I completely agree that air injection is a crutch. You don't want to use it unless you're seeing problems with the system. My concern is that IF it's needed, how much air could really be needed? I'm coming here because I know this is an oddball thing and I can't just look it up. The process conditions shouldn't require air injection, but the existing system (to be replaced) used it. I'm hesitant to say everything's going to be fine just because I read some values off of the pump curve and have the control elevations in the wet well (i.e. supply reservoir).

I don't think there's any real guide or calculation to predict such values, but if I had a good range of values to use, it might be helpful later.

Once everything gets plumbed in for these pumps it's a pain to change it. They're critical for operation of the treatment plant, being the raw sewage pumps feeding the primary treatment areas. They're 600 or 800 hp units depending on which is discussed, flowing anywhere from 20 to 110 Mgd. They'll be on VFDs of course to get the flow rate variations. Luckily, the process conditions are pretty stable, with the exception of suction head. That varies by as much as 20 ft. The temperature doesn't vary much through the year, the wet well and discharge location are both open to atmosphere, and the sewage properties are very much like water (for density, vapor pressure, and viscosity anyway). NPSHA is expected to be more than twice NPSHR over the full operating range.

Bimr, your answer gives the same interpretation of those air flow numbers that I had. It's nice to see that I'm not alone in this.

LittleInch may have the right idea here. IF air injection is needed, assume the reference flow values are a percent of volume flow. Calc the flow rates needed and put a small safety factor on it to get the highest flow I might have to use.

Thanks for your contributions and concerns.

 

[Artisi]

If NPSHa is expected to be 2x NPSHr it is unlikely suction cavitation, more likely recirculation if the pumps are operating way to the right on the performance curve. This ends with a lot of noise, grumbling and groaning etc and the injected air is probably softening this noise -
The use of VFD will more than likely overcome your problem.

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.)

 

[Artisi]

I mentioned in my previous post "recirculation", it is also possible there is a major mismatch of flow into the impeller eye if operating well to the right on the performance curve, again noise rumbling, grumbling etc.

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.)

 

[someguy79]

The pumps are not yet installed. One of them should be operational by next summer, the other two should be ready to go in 2017. Therefore, I can't confirm if there are recirculation or cavitation issues with these pumps.

The question I posted here is a matter of preparation. I would much rather spend the money and time to have this set up and ready and never use it, than find out I need it and not have it for a while. I need to make sure I can meter the air appropriately if it's needed. I already have an air supply in the pump room, air injection ports are included on the pumps, and the incremental cost of the air injection system is small.

 

[Artisi]

In the words of Nike "just do it".

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.)

 

[DubMac]

If you have the time, and it looks like you do, you should spend it studying up on what causes cavitation and trying to prevent it, rather than trying to figure out the deep dark secrets of how a band-aid is made.

 

[Artisi]

or the effects of over / under capacity etc.

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.)