Homemade suction diffuser or repurpose another product?

JohnNY

Hello all,
I've got a client with a new 40-GPM pump that was spec'd by an engineer and serves a cooling tower on the roof of a 10-story building. The pump, which replaced a 20-gpm unit which was deemed to be undersized, is developing air problems. The HVAC installer is asking that I install an air vent. I don't like the inlet piping though and I would have liked to see a suction diffuser on the line. The piping is 1-¼". So is the volute tapping. The smallest diffuser I can get seems to be 2". B&G reps are telling me the inlet to the pump should be a 36" straight run. We don't have that kind of space. Has anyone tried to make their own suction diffuser with any degree of success? I'm wondering if something like this from Caleffi or Taco would make the difference: https://www.caleffi.com/usa/en-us/catalogue/discal-dirtmagr-air-and-dirt-separator-magnet-brass-546195a

pecmsg

Let me guess, Double Suction and the inlet is turned 90 deg to the impeller?

Im more concerned on Why a 40 was used to replace the 20 HP?

Jamie Hall

I hope @pecmsg is wrong with his guess... but it's almost certainly a problem combining turbulence and marginal NPSH. What does the pump suction draw from? Can you put a very long sweep elbow in, assuming that you need to turn? Also -- and perhaps in addition, can you use a concentric reducer from 2 inch down to inch and a quarter at the inlet, with the long sweep in 2 inch beyond that? Almost anything to reduce the velocity and velocity jumps will help.

Sketch or photos?

hot_rod

It does seem odd that a 40 gpm pump could or should drop in place of a 20?

Grab the pump curve for the new pump and see what the suction head requirements are, sounds like a cavitation nightmare, which will destroy that new pump, maybe in a few days time.

Larger pumps usually have that info on the curve data. The pump rep should be able to guide you.

ScottSecor

Just searched online and found Wilo and a few others only go up to 1.5" inlet x 2" outlet. Here is a link to a company that appears to make a 1.25" x 2", not sure that helps.
http://www.jlinepump.com/epg/products/pdf/engbrochures/600series/suctiondiffusers.pdf

I have had custom stainless braided connectors made a few times over the years. They can make steel or stainless steel male ends, flanges, etc. I'm thinking maybe something like a 3" x 10" piece of braided hose with 1.25" ends on it since you can't get that 36" piece of straight pipe that would be ideal?

ScottSecor

Another idea that we've done many times is that the pump may accept 1.5" flanges or sometimes even the 2" Webstone style flanges (threaded, sweat or pro-press. If this is the case, you may be able to use a roughly 12" piece of 2" copper or steel instead of the 'restrictive' 1.25" pipe on the inlet?

Something like this in 2" if you can use 'HV' flanges:
http://www.webstonevalves.com/customer/wecoin/specpages/8140.pdf

nicholas bonham-carter

Can you check the engineer’s calculations, and see whether such a large pump is really needed, and then if the piping is big enough?
If this pump is in the lower part of the building, wouldn’t the air problems indicate a restriction in the line?—NBC

Zman

Does the pump manufacture provide a NPSH curve for the pump they are using? I would double check the suitability of the pump for the application before getting involved with a swag on a diffuser.

GBart

I could be wrong but going to a double size pump scares me, I think you may be having cavitation problems which may seem like an air problem, fixing the air problem won't address it, but see this article.

https://www.pumpsandsystems.com/pumps/january-2016-cavitation-vs-air-entrainment

GBart

Or this one, by changing the pump to twice it's size they may have changed all the piping parameters, the piping has to be able supply the pump.

https://www.valin.com/resources/blog/rules-follow-avoid-pump-problems

GBart

Is the pump tied to a VFD? ( hopefully), that would be the way to do it with pressure differential control to throttle it down to exactly what flow is required, it's much more efficient.

JohnNY

> @pecmsg said:
> Let me guess, Double Suction and the inlet is turned 90 deg to the impeller?
No and no.

>
> Im more concerned on Why a 40 was used to replace the 20 HP?
Not HP. GPM. And we don't know why the 20 GPM pump was there in the first place, so I'm going to deal with the situation as it exists for now.


> @hot rod said:
> It does seem odd that a 40 gpm pump could or should drop in place of a 20?

Why would this seem odd given I've offered little information whatsoever about what this pump is doing.

> Larger pumps usually have that info on the curve data. The pump rep should be able to guide you.

Thank you for letting me know about the pump curve data and where I can get it. It's been a real mystery to me these past 30+ years in the field. (wth, hot rod?)


> @GBart said:
> Is the pump tied to a VFD? ( hopefully), that would be the way to do it with pressure differential control to throttle it down to exactly what flow is required, it's much more efficient.

The pump feeds a cooling tower. I'd sooner use a circuit setter or differential bypass of some type to adequately meet the demand of that system, which is constant.

JohnNY

@ScottSecor Thank you very much for the links. I'll have a look. I was planning to oversize the piping at the inlet but then my mind wandered to a hydraulic separator....which seems excessive by 2 piping connections...and that's how I landed at the air-dirt separator with its wide internal flow path and dirt screen. I feel like that might promote better linear flow at the volute inlet.
But maybe I needn't be so creative.

hot_rod

Take a look at this product, also.

Henry

A rule of thumb for pumps is 5 X the diameter of straight pipe before and after. Next how does the engineer figure that the piping can take 40 GPM? Standard practice the maximum for 1 1/4 is 15 GPM. One could get 20 GPM but friction is now 6ft per 100 ft. 40 GPM is off the System Syzer max for 1 1/4. You can throttle the outlet of the pump until it stops cavitating.

JohnNY

Henry said:

A rule of thumb for pumps is 5 X the diameter of straight pipe before and after. Next how does the engineer figure that the piping can take 40 GPM? Standard practice the maximum for 1 1/4 is 15 GPM. One could get 20 GPM but friction is now 6dft per 100 ft. 40 GPM is off the System Syzer max for 1 1/4. You can throttle the outlet of the pump until it stops cavitating.

Good points and good post. Thank you.

EBEBRATT-Ed

Blow the pipe size up as soon as you come off the pump suction and then put the suction diffuser on.

I agree with @Henry 40 gpm need 2" pipe. Although the pump connection is usually always smaller than the recomended pipe size I would blow it up as large as possible.

GBart

Just curious, but upon going from 20 to 40gpm what happened at the cooling tower? did any spray heads blow off, did they all get clogged from the increased dirt flow?

JohnNY

GBart said:

Just curious, but upon going from 20 to 40gpm what happened at the cooling tower? did any spray heads blow off, did they all get clogged from the increased dirt flow?

I have no idea. I'm not that involved in the building. They called me about the feasibility of installing an automatic air vent on the line (the HVAC contractor's suggestion) because the system has twice stopped flowing in the two months since it's been installed. I advised against the vent and explained I didn't like the inlet piping to the pump. I'm going to call the client and tell him his engineer needs to reconsider his flow rates. Also, there's a VFD on the pump set manually to run at 60Hz. I'll see if I can lock it down to 40Hz and see how the system behaves with the new piping.

EBEBRATT-Ed

I would look at the chiller that this is pumping condenser water floor (I am assuming it's a chiller) 3 gpm/ton is standard water flow

Mike_Sheppard

I agree with EBEBRATT-Ed. Would be useful to know what the tonnage of the chiller is. Something sounds off. Someone had to really mess up if they had to double the flow.

I’m fighting one right now where they mounted the tower at the same level as the pump. Nearly zero NPSH, vortexing and sucking all kinds of air in. They put air vents everywhere. I wanted to attempt to make a suction diffuser (for the tower to eliminate the vortexing) but they also installed the tower so the access hatch is 2 inches off a wall leaving no room to get to it.

GBart

Agreed and you're looking for a 10 degree delta T, typically 85-95 in out of tower, the whole scenario seems so odd, I have seen chill water pumps that were slightly undersized but not tower pumps, you have X btu's from chiller and tower is matched or over sized, if the pump was wrong it would've been an issue from day 1 especially if it was 1/2 size, I remember one where a WHSP loop had a cooling tower that was slightly undersized but it only showed on really hot days and no pump change was going to cure it.

hmm, makes me wonder if.... " put a pump in with twice the gpm, yeah that will fix it" is what happened here ......seems very suspect

GBart

JohnNY said:

GBart said:

Just curious, but upon going from 20 to 40gpm what happened at the cooling tower? did any spray heads blow off, did they all get clogged from the increased dirt flow?

I have no idea. I'm not that involved in the building. They called me about the feasibility of installing an automatic air vent on the line (the HVAC contractor's suggestion) because the system has twice stopped flowing in the two months since it's been installed. I advised against the vent and explained I didn't like the inlet piping to the pump. I'm going to call the client and tell him his engineer needs to reconsider his flow rates. Also, there's a VFD on the pump set manually to run at 60Hz. I'll see if I can lock it down to 40Hz and see how the system behaves with the new piping.

I don't think they're telling you the whole story, something is definitely funky here, but that pump should be controlled to ramp up & down for a 10* delta, maybe they weren't getting that on a hot day so they went to a bigger pump?? I think all they did was create another problem. The tower and piping can only handle X, you aren't going to change that with a larger pump, you can however create cavitation problems, etc.

JohnNY

Here's the story. The building is about to undergo a mechanical renovation. There are 3" risers being run from the roof down to the cellar. This 1-¼" riser is going to be used temporarily for this season assuming they can get away with it...if they can get this pump to work with the existing piping. If not, all other options are "practically cost prohibitive".
So, I'm repiping the pump and adding a bit of reservoir at the inlet. Then I'll adjust my VFD and assess.
Best to all,
John

Harvey Ramer

It takes at least 10 pipe diameters of straight pipe after a fitting to create the kind of flow you are looking to bring to the inlet of the pump.

End suction pump?

Not having room for straight piping, you could always take a piece of 4" - 6" steel pipe, about a foot long, and weld thread-o-lets at a 90° rotation. If you keep the inlet to the pump coming off the bottom, side and the return close to the top, that should get rid of a lot of the turbulence. It won't do anything for NPSH though.