Rag layer recycle pump

[Scipio]

I've had a murderous little service come up I have a few manufacturer's working on, thought I'd see if anyone has any added suggestions;

Basic service is a pump which serves to maintain a minimum flow in a free-water knockout at the inlet of a tarsands bitumen facility. The pump itself recycles a rag layer which forms between the water and bitumen layers at the outlet end of the drum, and discharges back to the drum inlet. The rag layer is a 'milkshake' consistancy emulsion of bitumen, water, entrained vapour, and fine sand at 1200 kPag, temperatures ranging from 135 to 200°C, and viscosities upwards of 300 cP at a flowrate of 60 m3/hr against a differential of about 200 kPa, with an NPSHA of about 3 meters.

Just to make things even more interesting, we need to avoid high-shear pumps like conventional centrifugal impellers because we want to avoid tightening up the emulsion that already exists in the rag layer.

First thing that came to mind was a progressing cavity pump, but even high temperature elastomeric stators wouldn't hold up for more than two or three weeks of continuous operation. I do have a PCP manufacturer looking into the temperature range of a line of steel stator PCP's they carry, but they typically rely on a higher viscosity fluid to seal the rotor & stator, since no interference fit is possible.

Another construction I'm looking at is a centrifugal screw pump - should have minimal impact on the emulsion and can easily pump the multi phase rag layer, but the upset temperature of 200°C is pushing the limit over where they're normally run - the manufacturer's I've contacted are still looking at means of achieving that temperature.

I've also looked into hydrocyclone type froth pumps, but despite their froth handling capability, at the bottom of the pump is a centrifugal impeller, which is going to tighten up the emulsion.

Right now it's starting to look like my best bet may be a multi-phase twin screw pump, (assuming the steel stator PCP and centrifugal screw options don't pan out) but the thought of four mechanical seals makes me cringe if there's any alternatives out there.

Any ideas?

 

[checman]

scipio, Schwing America makes a double piston pump that was designed for cement. It is also used at incinerators to pump hydrocarbon sludge into the kilns. It has a swing check and is a very low shear design. I have not seen a high temp version of this but you would think they could design for it. FMC also makes a piston pump that might work well. They are cataloged at 177 deg C.

Regards checman

http://www.schwing.com

 

[smckennz]

Hi Scipio

Perhaps you should review your dismissal of centrifugal pumps as high shear devices, quantify your shear limits, and adjust your parameters accordingly. Perhaps several centrifugal stages will meet your requirements. As you no doubt know, fine sand is a killer for pumps which rely on pinch/ close tolerance effect to generate pressure because of the grinding action. If its really tough , then I agree with checman. Talk to Geho, although the shear around the valve cluster needs careful consideration.
At 2 bar DP, I would be amazed if you cant do it forever with a slow centrifugal or two. This is one of those rare cases where a bit of testing (quantifying shear/ pumping style limits) could save a lot of money.

Cheers

Steve

 

[bimr]

You may want to consider a sliding vane pd pump.

http://www.blackmer.com/tech-vane.jsp

 

[Artisi]

you might consider a perastaltic hose pump for this application. Sorry cant give you a lead tomanufactures.

Naresuan University
Phitsanulok
Thailand

 

[Scipio]

Thanks for the suggestions, folks;

As of yesterday the steel stator PCP and centrifugal screw pump manufacturers gave up and said they couldn't do it. I do have a twin screw multiphase pump manufacturer who can manage our service, so we're going to give them a shot, though the thought of all those seals makes me cringe a bit. If that fails, I'm leaning towards piston pumps - interestingly enough I've seen almost identical construction to those cement pumps checmen refers to operating as multiphase pumps for wet gas wells.

smckennz, I agree that some centrifugal pumps may not cause a problem, unfortunately the process is impossible to quantify accurately. The refinery itself doesn't exist yet, and the rag layers are inconsistant from site to site depending on production formation, temperatures, additives, etc., so historic data from other facilities only presents a very general picture. I'd love to be able to do some testing, but unfortunately the nature of this beast requires some conservatism. We can't even look back to previous projects for a construction, because we're running about 60°C hotter than previous jobs, which used elastomer stator PCP's for this service.

I admit I hadn't considered sliding vane pumps, but I'm not sure how well they'd handle the abrasion, this stuff can have sand the consistancy of sandblasting silica in it. A peristaltic hose I did consider, but didn't think the elastomers would hold up in one of those any better than in a progressing cavity pump.

 

[unclesyd]

I have to recommend taking a look at this type pump even though there have been a few negative comments on it in the forum. It served us very well in several messy pumping situations.
I would give their technical people a call.

http://www.discflo.com/index.html

 

[Scipio]

Unclesyd,

Actually, I've looked at those for other shear-sensitive streams in this same project (the ones we ultimately went with centrifugal screw pumps on), the connector pins between the discs were judged as a potential source of emulsion tightening, so were rejected as a prime candidate, but reserved as a fallback position. Once again, it's one of those situations where "It's probably OK, but impossible to quantify" unfortunately isn't good enough.

 

[HEC]

Schipio,
Have you considered a "sine Pump". These are used widely for viscous and sensitive products. They may be worth a try.

http://www.process-controls.com/York_Fluid/Sine_Pump.html

Mark Hutton
hutton4eng@picknowl.com.au

 

[Artisi]

I guess a leading question at this late stage is, has this this product been pumped previously or are we trying to break new ground.

Also an explanation of what you mean by "tightening up the emulsion" would be helpful.

I have resonable experince with froth pumping applications and your earlier thoughts of using a vertical tank mounted froth pump looks to be your best bet.
As you have entrained air in the product the pump is going to be oversized slow runing unit.
It is well worth your time looking into this pump style more closely and as "smckennz" stated - some site testing could be considered.



Naresuan University
Phitsanulok
Thailand

 

[Scipio]

HEC,

Actually, never seen the Sine Pumps before, look like they might do a decent job, except we're running upwards of 50°C hotter than they're designed for, and our discharge pressure is pretty close to their upper limit. I might dig a little more into them though, never know what isn't published on a web site.

Artisi,

We're breaking new ground here - the same service exists on other SAGD operations, however the exact consistancy of this rag layer is different from facility to facility, and we're running hotter inlet temperatures than previous installations. As for the entrained air, it'll actually be some combination of steam and hydrocarbon vapours suspended in a bitumen/water emulsion. There's also a lot of fine sand bound into the emulsion, any idea how well hydrocyclone froth pumps handle abrasive services? As for the impeller at the bottom of those and the concern with 'tightening' the emulsion, the more you blend this stuff, the harder it is to break the emulsion into separate components. As for site testing, not an option - the facility itself doesn't exist yet, and waiting until the rest of the site is online is quite obviously out of the question.

 

[Artisi]

Scipio

Froth pumps which are used to pump the "froth" from froth flotation cells in mineral dressing plants can be supplied in hard metals, ie, Ni-hard or high chrome irons or can be rubber lined. For fine material you should be using rubber lined pumps - but - rubber is your case is probably not an option because of the high temp. High chrome iron with a Brinell hardness of around 700 should be ok as the duty you are talking about will result in a large slow running unit which will tolerate the fine abrasive materials.

Having been away from the game now for a few years, I have lost touch with "who-owns-who" so you will need to do a bit of research to find the manufacturers who can assist you.
Two companies - now maybe trading under different names who should be able to help are:
1. Warman pump company - Australia - USA - UK
2. Sala International from Sweden - who became Denver - Sala and I think now have a new owner and name.

Hope this assists you.

Naresuan University
Phitsanulok
Thailand