Pump power demand slowly increasing

[ME1289]

Hey guys,

I've been having trouble with a pair of Sundyne LMV 311s and I'm running out of troubleshooting ideas. Figured I'd ask around to see if anyone has had similar issues or has any suggestions.

Background:

The pumps move ambient temperature light hydrocarbons at a low flow (cycles between 100 and 160 GPM based on unit demand) and high head (1200 FT). The motor is 3600 RPM and has a 2.3:1 speed increaser gearbox. Pumps are on flow control. One pump needs to be in service at all times with the second pump in standby.
The pumps do operate pretty far back on their curve with a hydraulic efficiency between 34% and 44%. The original design point was 250 GPM but we are still above the minimum stable flow of 88 GPM.

Problem:

In the last 4 months, power demand of the pumps has been slowly increasing until we hit our high power alarm. We pull the pump and don't find any smoking gun except a slight discoloration/dark coating of the wetted parts. We polish the impeller and diffuser, install a new or rebuilt seal, reinstall the pump, and power demand returns to normal for about 10 days before it starts to climb again.
I don't believe there is a change in the bulk material properties of the product because as soon as we shut down one pump for high amps, the second pump (assuming it was recently cleaned) pulls normal amps for ~10 days.

We've had the motors inspected and verified amps draw at the motor leads. Flow meter has been checked several times.

Pump performance does seem to diminish as amps get near FLA. ie the flow control valve has to open more to get the same flow.

Only theory we have is that the internal coating is causing excessive drag or recirculation. It's odd because the coating is very thin and can be polished off. I wouldn't expect this to cause a 30% increase in power demand but I've certainly been wrong before. The only source of this coating would be something in the process.
The pump operational point is not ideal given the design point of 250 GPM but we hadn't had any performance issues in the 5+ years these pumps have been in operation.

Thanks and any input or suggestions would be greatly appreciated!

 

[LittleInch]

Got a sectional drawing of the impellor?

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

 

[ME1289]

I don't have a cross sectional drawing but attached is an as-found picture after we pulled the pump for high amps. It's a 24 straight vane SS impeller, 6.3" diameter.
The impeller is stainless so it should have a metallic finish-- not this dark coating that we're seeing.
I guess my question almost boils down to "is the coating on the impeller and diffuser enough to increase power consumption by 25 - 30%" ? Or are we missing something else going on?
Work is still on-going to identify the cause of this coating.

 

[1503-44]

On one hand not, but on the other hand this sounds strangely like some kind of a flow control-process issue. I think it's very unusual that you are describing a potential change in bulk material properties by comparing pump power consumption. You should have a means of knowing if and when the fluid properties change by monitoring something other than pump power consumption. Exactly what "light hydrocarbons" are you working with here? Is it a mix of light oil's? Is it an emulsion ? What is it? It sounds almost like you are pumping a mix of fluids that separate by density and layers itself in the tanks, or they are initially warm, then cool over a 10 day period. FCV opening more suggests viscosity or density is increasing. There is no reason that should otherwise happen without drastic changes in pressure, but you do not mention any pressure changes going on.

Why are you operating at 30-40% flow? Has the fluid changed since the system was designed? Did this problem occur at higher (design capacity) flow rates? Is the fluid being affected by the slower flow rates? Is it aging in some way sitting in the tanks??? Does the slower pumping rate allow density separation to occur in the tanks? Do light ends boil off? Does the vapor pressure change? Any chance of wax forming, etc. Wax-paraffin products can be very sensitive to slight temperature changes. If its pumping from a depleting oil field, you might now have a more dense mixture of oils. What is the fluids source? Did it change supplier's 4mo ago Do? Do you now pump down to lower lovels in the supply tank than you did 4mo ago?

Do pump inlet or outlet pressures change over 10 days?
Your flow is 100-160 GPM. is it 100 on day 1 and 160 on day 10, or the opposite?
Is the flow rate changing with those pressures?
Do you have a log of fluid density, viscosity, and a log of flow, temperature, inlet and outlet pressures and power consumption over the 10 day period?
Please post the pump curves.
In case it is a pump problem, which I have doubts, have you looked at the impeller after 2-5 days use?

 

[Artisi]

Looks like a few touch points on the impeller, anything showing on the case?

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.)

 

[ME1289]

@1503-44

This pump is moving liquid ethane out of a coalescer as part of a dryer circuit. The flow rate varies based on the dryer cycle which regens every 18 hrs. The SG does not change enough to cause this slow increase in power demand. When we swing pumps, one pump will be pulling 45 amps for 150 GPM and the second pumps starts up pulling from a common suction line and only pulls 36 amps. The second pump is moving the exact same material to and from the same place as the first pump-- same flow meter and flow control valve too. It's almost like the increase in power demand is "baked in" to that particular pump until we pull it for cleaning. It takes about 4 weeks before a pump gets too close for comfort to FLA

@Artisi

No obvious touch point on the casing. This impeller is spinning around 8500 RPM so I'd imagine any contact with the casing would be clear and obvious

 

[1503-44]

Continuing with process problem aspects.
Why is the control valve changing position?
Up and downstream pressures?
Your sure it's not gasifying?
Large changes to Ethane density occur at 50°F.

At low flow rates, you may have a large pressure drop across the valve. Is it flashing?

https://www.engineeringtoolbox.com/...cific-weight-temperature-pressure-d_2088.html

 

[EdStainless]

There are two things that come to my mind.
Is the coating 'roughening' the surface? These are high speed pumps and a change in surface texture could have an impact.
Or is this slight buildup causing some mechanical drag?
In the meantime, what changed?
People say 'nothing', and that just means that it isn't something that they intended to or usually measure.
Look carefully at the pump clean/rebuild process. Down to the rags, solvents and the exact polishing method. It is in there somewhere.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[ME1289]

@1503-44

The suction temperature is controlled and is consistently 70F. I've seen issues in the past where this temperature inadvertently drops and causes a spike in amps. However, the increase we're dealing with now is slow but steady. Ambient temperature doesn't seem to have an effect either. I don't hear any flashing across the control valve. I think issues with the controls can be ruled out by the fact that everything goes back to normal as soon as we swing pumps.

@EdStainless

The coating could be roughing up the surface of the impeller and of the casing & top plate. I was having trouble accepting the fact that a seemingly minor change in surface roughness/finish could be causing a 30% increase in power consumption. Maybe the surface finish has a much bigger effect than I intuitively think-- especially since this is a relatively high speed pump.

 

[JJPellin]

I have seen this same affect on Sundyne LMV-311 pumps. Our pumps had the standard 8 vane impellers. In our case, these were in light hydrocarbon service in a sulfuric acid alkylation unit. We were experiencing acid carryover which was causing pitting corrosion on the diffuser and diffuser cover surfaces just above and below the impeller. The surface finish would become slightly roughened. As I recall, it looked like the surfaces had been sand blasted. This was enough to cause the motors to overload and trip out on high amps. We changed to larger motors. We made process moves to reduce the acid carryover and I think we may have made a material change to the diffuser and diffuser cover to resist pitting corrosion. We have not had a problem since.

To answer the original question, this model pump can be sensitive to surface finish on the critical surfaces adjacent to the impeller. Poor surface finish can result in higher motor load and higher motor amps. I have not personally experienced this effect from a build-up (platting out, fouling). But, I can imagine that it would be the same as I had experienced from metal loss resulting in surface roughness.

I would also recommend closely inspecting the diffuser throats for diameter and condition. In this model pump, a difference of 0.005 inch diameter on the diffuser throat can made a substantial change in pump performance.

Johnny Pellin

 

[ME1289]

@JJPellin

Thanks for the input-- I'll focus my attention on figuring out where the coating material is coming from. Sounds like this is the most likely scenario given that we haven't found any other smoking guns yet.

 

[1503-44]

OK. It is a very high rpm, so there is a lot of surface area/minute there.

It would be interesting to understand more about why this has only been happening for the last four months.

 

[Artisi]

I re-looked at the impeller photo and interested in what looks like rub marks on some areas of the front face of the impeller, for interest have you checked for any excessive axial movement of the "impeller / shaft / bearings" as an assembly, usually an impeller will have some radial thrust towards the suction side with excessive bearing clearance resulting in some impeller / case clashing - for me certainly worth a thorough check.

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.)