Odd pump behavior - off the curve

[Rputvin]

I've got an interesting set of circumstances that is confounding all of the engineers at my company and I'm looking for someone that may have some idea of how to troubleshoot this.

Less than a month old, we ran off at our facility, but don't record flow values, only pressure and current on the motors. Closed loop, running a stamped stainless close-coupled centrifugal pump at ~73 GPM and ~25psi differential pressure. The pump curve for this pressure should be giving us ~120gpm of flow (selected slightly oversized), and we're sitting short of the customer's required flow rate of 80 GPM. The system pressure is almost dead-on the calculated drop for the system.

We're moving city water between 75 and 100°F. Using standard pressure gauges at multiple points in the system, before and after the pump. Using IFM Efector flow meters with proper installation runs, one after the pump discharge before branching, and one on each branch before collecting to the return header, all flow meters downstream equal the discharge meter before the header.

We're on the curve through about 40 GPM, then we start getting deviations and we're sitting 40% or so off the curve. The pump manufacturer built us another pump and ran it off through the entire curve to make sure their stamping and impellers hadn't drifted since the curve was established - their pump matched the curve values, albeit with much more ideal and controlled conditions than what I'm dealing with.

Our first step was to change the impeller to the largest available trim, which improved the numbers slightly to where I'm at now, but still significantly lower than expected flow.

The system is pressurized with a diaphragm expansion tank, we started at 12psi and have increased it to 20psi. There's an air separator installed, which is operating as best as we can tell, it does not have a strainer in it. The make-up line is city water connection regulated to 17 psi. The rest of the system is the customer's piping that I'm short on details on, but should follow a standard flow/pressure drop relationship and not figure into the conundrum.

The pump will make deadhead (shutoff/no flow) pressure. As we open the throttle valve (butterfly valve used for isolation mainly) we see a 1-2psi drop in suction pressure. The motor is drawing near nameplate amperage, suggesting we're to the far right on the curve. Required NPSH is all of 2.5psi


Any ideas as to what might cause a pump to... not pump? We're thinking suction restriction is a possible culprit, but there's nothing there to impede flow. Would a small amount of air in the system wreak this kind of havoc?

Any help is appreciated!

 

[quark]

I observed strange things in the past with air in the system. Pressure reading fluctuation is one indication of air in the system but I won't take that for granted. If you close bladder valve, stop the system, check for air (and do it multiple times) and every time you observe air then it is not properly being vented out. You are directly pressurizing air vent with a line from bladder tank and the location where you connect bladder tank to vent may not allow the air to go up (if your P&ID denotes actual connections).

You further stated that there is a reactor 20 ft above the loop line. Does it have an air vent? If you have one air vent at the top most point, that should be sufficient.

 

[Rputvin]

The reactor does have a high point vent. As far as we can figure, all of the air is out of the system.

The commissioning report notes when our tech arrived the system was showing unstable flow and pressure, the hoses were having a lot of movement (especially on flow start/stop), and air was audibly in the system. He spent the first four hours of startup purging air to the point we're at now. I can't say it's 100%, but it should be to a level that's acceptable for operation.

 

[LittleInch]

Last thing I can think of here is whether there is any contamination or additive which is causing the density to decrease even by 5 or 10%. If there is a lot of dissolved air in the system caused by air being churned through the pump and then kept there by the pressurisation system, this could mean that the head figures we're all using for water from pressures are incorrect.

Maybe what you should do is reduce the pressurisation level (when cold) to only say 5psi then run the system for a few volumes through the air eliminator and see if lots of air comes out? plus bleed the system at that low pressure.

I still think though for pumps of this size and rating, the accuracy of the pump curves from the vendor even for "identical" pumps is often not much better than 5%. Maybe you've got a slightly different casting or some manufacturing tolerance issue which is enough to cause an issue.

So a bigger pump with higher head and steeper curve still sounds like the best solution.

If you had a photo of the pump installation it might be interesting as sometimes diagrams don't match what was actually built or something stands out that you can't see from the diagram.

Good luck, hope we've been of use and let us know how it goes.

LI

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

 

[lilliput1]

Valve off the city water emergency supply. The city water pressure is working against the pump. Put in a NO solenoid valve powered close so on loss of power both the city water NO (Normally Open) supply valve and NO drain valve will open.

 

[georgeverghese]

Terminology used here is a bit confusing - you say the expansion drum is fitted with a diaphragm ? - A typical closed loop recirculation stream expansion drum would not have a diaphragm fitted to allow air in the recirculation stream to escape - you would only pressurise the vapor space with N2 or some inert gas with a forward sensing (sensing expansion drum pressure) and install a backpressure regulator on the vent exit to release any excess pressure.
And how do you operate the makeup water to this drum? - is there a level guage installed? All expansion drums are fitted with level guages or LT to ensure adequate level (by addition of makeup coolant) for pumping operations. And the tie in point for this makeup water is also nontypical - it should also feed into the expansion drum; not downstream of the drum as you have shown. Could the RV on this makeup water line be leaking?

 

[lilliput1]

Say 25 psi is 57.75 ft of water, not 50 so the pump can only put out about 86 gpm not 120 per the system curve. However the pump curve must intersect the system curve. The system required head at 80 gpm is 57.75 x (80/73)^2 = 69.35 ft wg

If you draw the system curve, the intersection with the pump curve would be at about 74 gpm and 59.3 ft wg so the pump is undersized and can not meet the system requirement of 80 gpm at 57.75 say 58 ft wg.

 

[Rputvin]

lilliput - the emergency water supply do have NO solenoids to only operate during a power loss. They also have manual bypasses, but those are service valves only. I'm only showing one on the drain as the customer is running a specialty valve on the inlet.

There are a lot of posts, but I corrected that 25psi figure - that was a "round" number I used in my OP as I didn't have the exact figures in front of me. Actual pressure is a bit lower than that.


georgeverghese - That style of setup, from what I understand, has been phased out in this type of system for a while. The drum itself pressurizes the system and allows for thermal expansion. The air separator handles the venting, the drum is not the high point of the system. A sight glass is not used on a diaphragm, there's a makeup line regulated to 17psi to ensure the system volume doesn't drop, and a main relief on the reactor set at ~34psi to handle over-fill.

Is the makeup regulator leaking? Possibly. City water is at a higher pressure, so it would only be coming into the system. Would this not add to the main flow and not subtract?

 

[quark]

If your highest point (at the reactor line) has an autovent, then I have nothing more to add. Post us updates.

 

[LittleInch]

Ok absolute last shot here.

What is the actual piping layout of these pumps for inlet especially. Lots of elbows and tees can generate a lot of flow disturbance and swirl. This will get worse as flow increases so your low flow will follow the curve but then start to diverge.

Not sure what the vendor recommends but you usually want min 5D of straight pipe before the pump free from valves etc.

What does it physically look like?

You could try a flow conditioner plate of you've got room

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

 

[georgeverghese]

Well, the expansion drum arrangement I described earlier is commonly used in oil/gas installations in closed loop heating / cooling circuits, and none of them have diaphragms or air release valves. And in a well laid out design, (a) the main recirc stream feed nozzle at the expansion drum is the high point in the return header (b) the return header slopes continously upwards in the direction of flow back to the expansion drum.

 

[lilliput1]

You need to plot the system curve on the pump head vs flow chart. Where the system curve intersects the pump curve will be the point of operation. If you measure 73 GPM at 57.75 ft head with the pump running you need to plot the system curve using the formula H = 57.75 x (GPM/73)^2
Plot H ft wg at each corresponding GPM to get the system curve.

 

[itsmoked]

Rputvin; Just checking. This pump isn't on a VFD is it?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Rputvin]

LittleInch - the suction conditions aren't ideal with elbows close to the pump suctions. It's not an atypical arrangement for our designs, it may be contributing enough to cause an issue this time around.

georgeverghese - I'm familiar with the setup. It has several drawbacks, and in this instance the pump skid is ~20ft lower than the high point in the system, we'd have to roof-mount the tank if using the traditional arrangement.

lilliput - I'm not sure how the system curve is going to help me. I know several operating points in flow and pressure and could plot a new curve, the question is why this system doesn't want to follow the pump curve at all as flow is increased. It's like we're hitting a wall.

itsmoked - no VFD. It was being considered as a fix to add a VFD to overspin the motor, but that was abandoned.

I need to call my contact at the customer facility today. We shipped them new spools and pumps to be installed on Saturday and my office was shut down with power issues yesterday. We did get a request for another system yesterday, so that's a plus.

 

[georgeverghese]

I get the impression you've gone for this diaphragm type arrangement in order to eliminate the need for pressurisation gas at a conventional diaphragm less expansion drum. If you have the opportunity to redesign this for a conventional expansion drum, it is possible to also almost eliminate the need to top up pressurisation gas. You have to select the right size drum with adequate mechanical design pressure, with adequate vapor space between LAL and LAH so as to keep the pressure swings(between these 2 levels) in this drum below the setting of the outbreathing PCV - good luck.

 

[lilliput1]

Give us your measured flows and corresponding pressure differential across the pump with new impeller.

 

[LittleInch]

Could be. Maybe instead of having say an L shape going into you pump you could have more of a longer flat U shape to give you 5 -10D of straight pipe U/s the pump.

It is probably a mix of issues just enough to cause the few feet head difference. Might be that particular pump model is more sensitive than others or is just operating too close to its limits.

Will be interested to see if the mods work or not.

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

 

[Rputvin]

The new pump is in, it's a 3hp model in the same product family. At the same pressure it's rated for over 100 gpm. We're now running 73ft (31.75psi across the pump) and pushing 78 GPM. Matches the curve 100% through all of the flow conditions they can generate. It still feels like we're hitting a wall at 80gpm, I have to wonder if some of their flow balancing devices aren't restricting the flow and the new pump is just pushing harder against whatever is in the system.

The customer is satisfied, so I'm thinking they're going to be changing some line sizes in the system to alleviate some of the restrictions. I haven't asked about getting those pumps back yet, there had to have been some conditions on the inlet that was causing the strange behavior.

Lesson learned with this build though, we're definitely going to make sure we nail the next build.

 

[LittleInch]

"some of their flow balancing devices " - what are these? Are they in the customer piping?

"At the same pressure" it's rated for 100GPM - what pressure?

Does now sound like you're coming up against some other restriction as that diff pressure and flow is a steep system curve from what you had before.

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

 

[Rputvin]

Every line in the customer piping has a balancing valve of some kind on it. We were told they're nothing special, I had assumed they were globe valves or circuit setter variable orifice devices. There are some units that react differently as flow increases, multi-orifice units that also limit pressure and play games with flow and pressure in the system, electronic variable units, etc. I don't have a rundown of every device they're using, but we asked if there was anything other than manual balancing valves and if there were any devices like those mentioned that don't like to behave consistently - we were told no.

The pressure in the system has to be artificially increased to a certain degree by restricting the main/largest branches in the system to allow flow to spread everywhere. Otherwise the large branches would flow and the small lines feeding different aspects would have none. It's a tricky procedure to balance a heat treat furnace/oven/reactor, etc that has jackets, power supplies, quench coils, pumps, and other items that are all water-cooled. Nothing that unfamiliar, 90% of the units we build are attached to a system like this and is our specialty niche in the market.

At the system pressure we had previously of 20-some-odd psi across the pump this new pump should be pushing a lot more water. But we're riding higher on the curve with a roughly 10psi increase. I'm not sure if they're able to further restrict the main branch to force more flow out to their vacuum pumps that were the biggest concern and overheating with the previous pumps, or if something else was changed. I'm also not entirely sure they know what they're doing. I've been working with their main engineer and he a stickler for published values. We weren't far off of his pressure figures, but now we're running above 70ft of head.

The original selection we had was a 5hp model that was 80gpm at 80ft, we were told to scale back and select an operating point of 80gpm at 50ft. That's what got us into this issue, and also caused delays in the build as we had to change the design to accommodate the smaller pump.

 

[LittleInch]

That all makes sense alright - in something with multiple branches, flow restriction devices etc it's going to be never ending game to get the flow balanced exactly how they want. Many people don't really understand what the valves they have actually do.

That scale of pump and pressure losses etc just has far too many variables and tolerances. Just remind your customer engineer that the line on the pump curve for a 2HP pump is infact a fairly fuzzy band where you can be 5% out and still in tolerance from the vendor.

A steeper pump curve wouldn't go amiss either.

If you ever get to the bottom of the strange pump performance let us know but this sounds like the end of the story?
Hope we were all able to help (sometimes you need confirmation you haven't missed something obvious).

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