Caking up centrifugal pump - Seal failures 3

[mjpetrag]

I have a recent crop of seal failures due to sulfate salts plating out on the metal bellows & springs of the seal/suction lines/impeller itself. The pump is an IR 1-1/2X1X6 HOC, with a plan 11 flush on the seal with double hard faces & a steam quench. I've considered adding a water/condensate plan 32 flush, however water generation would be too high for our facility to process.

The impeller itself at one point was caked up with salt.

I know we have to fix the process problem somewhere, but I need to make this pump as bulletproof as possible. I'm even considering switching to a completely different style pump if anyone has some suggestions.

I have a feeling that a dynamic seal with a repeller would not work because it wouldn't prevent the salts from plating out.


Any suggestions would be greatly appreciated.

-Mike

 

[mjpetrag]

Let me change one line in there...a dynamic seal with a repeller might work but I'm worried about sealing surfaces being damaged on start up from the salts that have plated out pulling and ripping the sealing surfaces

-Mike

 

[Artisi]

When you say "dynamic seal" do you mean an impeller running in a seperate chamber located in the back of the pump case / stuffing box cover?
If yes, then there is considerable discussion required to assertain if in fact this arrangement is suitable for the application and will require the pump to be replaced with one designed to have a dynamic seal.

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.)

 

[PumpSmart]

mjpetrag,
I understand you have an ANSI pump with simple mechanical seal and plan 11 and Plan 62.
Your fluid has a sulfate salt, (what kind of sulfate? and concentration % by weight, ph, temperature) dissolved ,but it seems like you have more sulfate concentration % that in saturation state so it tends to precipitate and another issue, sulfate tends to crystallize in contact with oxygen.
So first analyze why you have more concentration than saturation state and if it is necessary for your process.
If it is not necessary, reduce cocentration, and propably you will continue with the same pump and retrofit for new seal, you will try outside multispring hardfaces (or external seal) plus plan 11 and plan 62, be sure to make s internal clean when the pump is in stand by operation.
If it is necessary to keep this level of concentration, it is highly probably you will change the pump.
ANSI pump not handle so much solids.
For using dynamic seal, mechanical seal or packing you will analyze,
if temperature, ph fluid could damage external parts of the pump or pesonal, if yes you will continue using mechanical seals.
if not you will try packing or dynamic seal,
Dynamic seal is not an accesorie and it is design and provided by pump manufacturer.
I hope it helps!

 

[mjpetrag]

artisi,

yes, here's an example of something similar i was looking at

http://pdf.directindustry.com/pdf/goulds-pumps/3196-i-frame/20610-124124-_7.html

pumpsmart,the solution contains about 15% sodium sulfate, and 1% sodium carbonate at a pH of 11. the deposits on the seal were found to be sodium-calcium carbonate, hydrated sodium sulfate, and sodium phenate trihydrate. this is a hazardous service, so from your post above, dynamic seals are not appropriate?

If sulfate crystallizes when it contacts oxygen, is it possible to use something like positive nitrogen pressure as a barrier fluid in a dynamic seal?





-Mike

 

[Artisi]

Correct, a dynamic sealed pump cannot be used in a hazardous application.

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.)

 

[mjpetrag]

any other options in this situation, holding constant that the process conditions cannot be changed?



-Mike

 

[Artisi]

How about a sealless pump? Shown on the Goulds link you posted.

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.)

 

[mjpetrag]

It's my understanding of mag drive pumps that the fluid needs to be self-lubricating, clean, & non abrasive. This fluid is neither

-Mike

 

[Pumpsonly]

You can consider using a dual pressurized seal. The depositing of the salts is something you have to sort it out with the process condition.

 

[JJPellin]

I would suggest that you consider a conventional dry gas seal using nitrogen as the buffer. We have made this conversion in a number of amine pumps. Amine is certainly not as prone to crystallization as your service, but has some similar properties when exposed to air. We have drastically improved the reliability in our amine services with the used of dry gas seals.

Of course, this requires that your process can tolerate a certain amount of nitrogen. It will be very sensitive to start-up, shut-down and operating procedures to make sure that the liquid product never gets to the seal faces. This can be especially tricky on a standby pump that has to sit idle with the nitrogen flowing. Depending on the piping arrangement, the nitrogen can accumulate, causing the pump to gas up.


Johnny Pellin

 

[PumpSmart]

mjpetrag,
For non-precipitation of salt on impeller, recessed
impeller probably works better.
CV3196 named in Goulds.

http://www.gouldspumps.com/download_files/cv3196_i_FRAME/cv3196_i_FRAME_cutaway.stm

For precipitation on tubing plan, faces, spring and bellows in hazardous services, you should check some options,
It is a hazardous service Expeller, Repeller, dynamic seal, packing and single seal are not applied.
For hazardous service, you should use double mechanical seal or sealless pump.
Double mechanical seal will be a hard first experience, pressurized option will work better since it prevents precipitation on faces but not on spring or bellows. You will continue using hard faces. You will use contact or non-contact face on secondary face.
For contact face you will use Plan 53B, 53C or 54, but I think you should use cooling water on all of them to reduce barrer fluid temperature due to thermal load. See options for doble seal here:

http://www.johncrane.com/cgi-bin/jcap

rod/pagepro_disp.w?clt=us&pagepro=bellows_seals.
I hope it helps.

 

[Clvet]

A sealless solution works best when the precipitation is because of the seal. With the solids in this application, there would need to be some provision to keep solids out of the recirculating/lubricating flowpath.

Mjpetrag, I"m not sure the fluid needs to be "self lubricating" but it does need some small amount of viscosity to effectively create the hydrodynamic film between the bearings/shaft. The front half of a mag-drive is exactly the same as a mecahnically sealed pump. Sealed pumps have numerous options to keep the fluid around their seals clean. With sealless pumps, there are much fewer options in that regard.

I'm a pretty big fan of sealless pumps but I would tend to agree that this is probably not a sealless application(canned or mag-drive). The way the application is described, if the salt were to start to precipitate on evertyhing(like the magnet) over time, eventually the clearance betweeen parts the magnet and the shell would go to nothing and eventually you'd still have those failures. It might take longer to fail but I'm sure it would also cost a whole lot more than losing a seal.

I would try approaching this from the process side, what can we do to limit/control the precipitation and more importantly, the precipitation on the parts. Could the inside of the pump be coated or a different material?

Also wondering about the failure rate, is it frequent and steady(need to change PM) or will things be running fine and then several seals will be lost in rapid succession(process upset).

 

[Artisi]

A rubber lined pump could be a solution, rubber lined pumps have been used in situations where percipitation is a problem - the thinking is that the rubber being flexible moves sufficiently to "shake-off" the percipitating material.

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.)

 

[mjpetrag]

After taking the Infrared camera out to the tank, I found a cold spot of what looks to be salts about 2 ft up from the tank bottom around the entire circumference of the tank. Mag drive pumps I think wouldn't work in this situation because of the possibility of the salts plugging the suction line and running the pumps dry.

I think the best option here is to try the plan 72 gas seals as discussed above to limit the plating. Oh, and clean the feed tank too.

Thank you for all the replies!

-Mike

 

[bk19702]

mjpetrag - one correction: pressurized gas seals would be a Plan 74, not a Plan 72.