low flow High Pressure Pump for sour water service 1

[alwayslearning]

WE have an existing sour water pumping application (25 m3/hr; 140 bar differential pressure) using sundyne pump (LMV331) at appx 18000 rpm, dual pressurized tandem seals (pusher balanced design) with chemraz O rings/ C vs SiC faces.

We have have had a series of problems with these pumps since its installation more than 20 yrs ago. I do not wish to get into details here. My questions to the forum as below:

1) Does anyone have experience with LMV331 in the high speed range that I have referred in the sour water application.

2) Does anyone have sour water pumping application (Refinery, Offshore, etc) in the flow and pressure ranges that I have referred to?

3) Does anyone have experience with Process Diaphragm pumps (Lewa, Bran+Leubbe, for e.g.) installed in similar application? What is their operating/ maintenance experience with these pumps? WHat problems/ Reliability etc have they achieved.


Thanks for your inputs.

Regards

 

[JJPellin]

We have 3 sets of pumps that are similar to your existing pumps. We have BMP-331's (horizontal) running at about 21,000 rpm (2000 psi discharge), LMV-341's running at about 19,000 rpm (2400 psi) and LMV-311's running at about 16,000 rpm (1200+ psi). All of them have been unreliable at one time or another. One of the big 341's is in our shop right now and it is a total wreck. The only alternative we have used for comparable service is in a low-pressure gas oil hydrotreater. The unit uses a stacked rotor multistage. Some would call it a tie-rod pump because it is barrel style but with no outer barrel, just tie rods. It is the lowest pressure of the bunch at only about 1000 psi. This pump is unspared and has been running for about 4 years with only one minor seal drip. It is 13 stages in a 15 stage stack to allow for future upgrade.

We only have one set of high pressure diaphragm pumps such as you describe. Ours is used for ammonia injection to activate (I think?) the catalyst in a hydrocracker that runs about 2400 psi. Ours are the largest model that is made by Pulsafeeder. They are huge. The main worm gear is the size of the spare tire on my car. But since they only run at unit start-up, I have no relevant reliability experience.

My recommendations would be as follows:

1. Try to solve the problems with your existing pumps. Pay special attention to where they run on the curve. Do not believe the numbers provided by the OEM. These pumps must be run very close to BEP (95% to 105%). They must never be run in parallel. They must be flow controlled with the control valve very close to the pumps. They must have excessive NPSH margin and water with very little dissolved gasses.
2. Purchase a multistage pump for the service. A stacked rotor like ours, but in a barrel to allow for your higher pressure should work.
3. Research more into the option of a diaphragm pump. This may be a good alternative for you. But running 24/7, I would be concerned about long term reliability. I would recommend installing a full spare even if you can go for a couple of days without water before you cut charge.

I would be happy to go into more details about the options for improving your existing pumps, if you are interested.

 

[JJPellin]

I guess I should have read your original post more carefully. I completely missed a couple of your questions. Our BMP-331 pumps were very, very unreliable for the first 20 or years or so. The problems that we found and corrected are directly related to my comments above. The pumps were running in parallel most of the time. They were running at too low a flow rate relative to BEP. They had gasket problems since they were built with Buna-N, converted to Viton, converted to Teflon, then converted back to Viton. They had cavitation problems from low NPSH margin and then again (I believe) from excessive dissolved gasses from the fuel gas blanket used to hold pressure on the suction drum. We have repiped the pumps so they run in separate paths with separate controls (no longer in parallel). We improved the controls on the spill-back lines and now control to a point closer to BEP, but still not above 95%. The pumps used to fail 3 or 4 times per year. They now run about 1 year between failures. We have had cooling problems with the oil and seal coolers, failure of the internal cyclone separators, coupling problems (originally grease lubricated gear couplings), input shaft lip seal problems (much worse in the horizontal configuration) and dimensional problems with the stack-up of the case to diffuser to cover to seal housing. I believe if we could get the flow rate up to 95% of BEP we might get decent reliability, but I never expect them to be "good" by any stretch of the imagination.

By the way, our product is stripped sour water.

 

[alwayslearning]

Dear JJ,

Thanks for your to-the-point response. It just shows that we are not alone in dealing with this poor reliability of these pumps. The vendor is happy booking the spare parts business and (gleefully) does R&D works on our pumps.

We took the same path as you guys: Changed gears to helical type (some 20 yrs ago); changed cplg from Grease-packed gear type to flexible disc coupling; suspecting the cooler and trying to cool it from outside!!! (and now I have a large cooler - an obnoxious looking 2 feet long, for which i need to find space - and the vendor happily sold it withuot explaining what was wrong with the old design!!!); changed single seal to dual tandem pressurized (and now they want us to change to Dual Back-to-back pressurized seal, with one rotating-mating ring shared between two stationary heads - and we are dabbling with the thought of buying this idea!); changed Viton to Chemraz (vendor says the viton O rings swell in Sour service!).

Despite all these, the bottomline is the same. Poor reliability. We just opened one pump today. The problem reported was that the seal barrier pressure not holding. Upon removing, we saw the seals and o rings in good condition, except the O ring behind the stationary head is all kinked.

As for the pump operation; we are running one pump at a time. The pumps (luckily) are operating at 90-100 % of the rated (BEP) flow. Yes, Rated = BEP flow for sundyne pumps; this, I am told, is not known by many among the users.

As for the control valve; I agree that it should be as-close-as-possible. unfortunately not! But hearing your side of the story (the complete pipe rerouting), I guess any modifications on the piping side is of minimal benefit.

I agree with your suggestion: 1) go for a multistage - 13/14 stages perhaps - pump and, based on your feedback, hope for a good experience.

2) Look for some process diaphragm pumps. Can you tell me what model of Pulsafeeder do you have. Later I shall check their website.

I was hoping to hear some folks from offshore on the diaphgram pumps experience.

Thanks

Regards

 

[JJPellin]

Our big diaphragm pumps are Pulsafeeder model 9490-S-E and have 25 HP motors. But they only have a capacity of about 8 gpm each. I don't know if anyone makes a pump like this that gives the capacity you need.

One of our other engineers just came to us from another major oil company. They used Union reciprocating pumps in this service. We removed our only recip water wash pump several years ago because of chronic packing leaks and miscellaneous failures. However, he insists that with good accumulators on the suction and pulsation dampers on the discharge, a motor driven reciprocating piston pump can work very well. They were very happy with theirs. I like this idea. I was opposed to the removal of our one recip pump. I felt that an accumulator on the suction and a different packing design could have greatly improved the reliability of that pump.

Keep in mind, that any reciprocating pump (piston or diaphragm) will be very sensitive to suction line length. If the pump cannot be located very close to the suction drum, an accumulator will be needed to avoid cavitation.

I am not sure I understand the current failure you describe. I assume you mean that the stationary o-ring between the stationary seal and the seal housing is swollen enough to distort. This seems very strange for this material. When we were trying different materials for o-rings, it was a mistake. We thought we had a material problem, but it was more of a dimensional problem with the o-ring grooves. We verified all of the stack up dimensions and the o-ring crush on each one. We found a dimensional mistake that may have been from original manufacture or from machine cuts made in the middle of the night at some previous failure. We recut everything to get the fits / crush back to original and have not had gasket problems since then. But all of this was relevant to the diffuser o-rings, not the seal o-rings. Our sour water is stripped and has low enough benzene that Viton works fine. At these high pressures, explosive decompression can be an issue if there is anything in the product that can absorb into the o-rings.

Based on the conversation with my colleague, I would modify my recommendations as follows:

1. Try and solve the problems with the existing pumps.
2. Purchase multi-stage barrel pumps.
3. Purchase motor driven reciprocating piston pumps.
4. Consider diaphragm pumps if you can find a model that is large enough.

 

[EdStainless]

If you are having elastomer issues in hot water+oil applications then you should contact Seals Eastern about Aflas. The problem with any generic elastomer is that the fillers and cure agents are not optimised for your service. And the mfg changes them at will since it does not impact the bulk properties. We used Aflas for high temp downhole pumps. Very expensive, great material.

My pump experience was all at lower speeds (3500rpm). To me this sounds like a 4000BPD pump with about 120 stages. Pumps like that are built for surface applications, but they take up a lot of space.

= = = = = = = = = = = = = = = = = = = =
Rust never sleeps
Neither should your protection

http://www.trent-tube.com/contact/Tech_Assist.cfm

 

[hollerg]

Our plant uses horizontal Roto-Jet Pumps model RO for our high pressure pumps, which have worked well. Weir Specialty Pumps makes it.

We use the standard speed single Stage Pitot Tube Pump in our limited services (1200 psig waste water injection well and 900 psig reactor feed pump). The literature claims 5000 ft of head in a higher speed version of the pump, but that curve is not on the web site.

We LIKE them, and jave replaced the Sundyne we had.

http://www.weirclearliquid.com/weir/clearliquid/home.nsf/Page/Products_Pitot_tube_pumps_Model_RO

 

[RGRAJANI]

In old plants we use plunger pumps (20M3/hr, 75 Bar diff. pressure) for water containing lesss than 100 ppm H2S. In upcoming plants I thing we may use multistage centrifugal (higher flow up to 140m3/hr).

 

[JJPellin]

Roto-Jet is a good option to consider. We almost bought a set for our hydrocracker wash water (~2400 psi, 124 gpm). In the end, one of our managers did not feel comforable trying something so "new" in a service that critical. If I had it to do over again, I would have fought harder for the Roto-Jet option.