Pump Cavitation 3

[rutherford703]

Please see attached drawing for detailed information.
The naphtha pump is pumping naphtha from the decanter. The decanter colects liquid from the steam stripper overhead condensers. The pump is experiencing serious cavitation problem. The tank bottom is 10 feet higher than the pump eye. The suction piping is 4" standard and the length is about 70 feet including all fittings.

The pump data sheet shows that the normal required capacity is 87 gpm and the rated capacity is 100 gpm. The pump data sheet also requires the minimum continuous flow is 80 gpm.

There is a minimum recirculation line with a manual open/close valve (by DCS). As the outflow is controlled by the tank level, the level control valve can be closed below the minimum flow while the manual valve is closed - historical data shows that sometimes the flow below the minimum continuous flow requirement.

My questions are:
(1) is the cavitation caused be the flow below the minimum flow? or
(2) The cavitation is caused by not enough NPSH available?

For the NPSH calculation:
Since Naphtha is a mixture with a broad boiling range, I think that the vapor pressure at the pump eye can be cancelled with the pressure of the decanter tank. Therefore, NPSH available is the elevation difference substracting the friction loss (about 0.6 ft liquid) or NPSH available is 9.4 ft liquid column.

For the NPSH required the datasheet shows the required NPSH is 6 ft (water). While the specific density for naphtha is 0.74, the required NPSH for pumping naphtha will be 6/0.74 =8.1 feet naphtha.

The margin for NPSH required is small. Is this the reason causing the pump cavitation?

Thanks in advance.

 

[1gibson]

Most industry standards do not consider NPSH corrections for hydrocarbons. If you do want to check, it is much more complicated than dividing by SG. See link, but this correction will reduce NPSHr slightly, not increase it. You're going to get a very small number, and it is mostly an academic exercise.

http://www.pump-zone.com/topics/pum...ss-npsh-hydrocarbons-stepping-npshr-different

You need to check the temp rise through the pump at your low flow conditions to see if you are increasing temp (and vapor pressure) too much, that is most likely your problem.

http://www.engineeringtoolbox.com/pumps-temperature-increase-d_313.html

 

[JJPellin]

Additional details would help. What is the pump configuration? What is the suction specific speed? What is the best efficiency point flow? I suspect suction recirculation cavitation from low flow.

Johnny Pellin

 

[TD2K]

I agree your naphtha is a bubble point liquid. 100 gpm in a 4" line doesn't sound like a problem, velocity is about 2.5 ft/sec, reasonable for a bubble point liquid.

The NPSHA for your pump is 6 ft of liquid, you don't need to correct for the SG of the naphtha. You can correct NPSHA for HC per 1Gibson's link but it's an empirical correction method based on recognition that cavitation with HC is less severe than with cold water.

You should have a pretty good margin, what is the pump head? Why do you think you have cavitation? Do you have anything like a suction strainer that could be partially plugged and taking more pressure drop than you think? Is there a spot on the suction that you can take a pressure reading and see how much head you have based on the pressure in the drum (use the same pressure gauge)?

 

[BigInch]

"you don't need to correct for the SG of the naphtha" Echo that and add "should not". Otherwise you would conversely have to reduce the NPSHR if your product had an SG greater than water.

Independent events are seldomly independent.

 

[rutherford703]

I appreciate all your responses. I know all of you are the most talented people in the forum. I attach the pump datasheet and pump curve for your information. I will send some damage photos when they are available.

For the pump configuration, there are two pumps for this service with one in operation and one standby. For the suction piping, there is no strainer and only one block valve.

The most serious damage is the casing at the suction side. The damage is caused by wear and cavitation (there is entrained solid in the naphtha but should expect to be minor) within a few weeks. The impeller is a closed impeller. There is some damage on the impeller surface at the suction side and the impeller. Some seal has beed eaten away. The frequency for the pump failure is about 3 every three months. Even the control valve at the recirculation line experiences cavitation (?) problem.

The proposed solution is to try to keep the flow in the recirculation line or to install a VFD for the pump to reduce the pump speed.

Thanks all of you again!

 

[rutherford703]

Here are some photos

 

[JJPellin]

I do not see any damage that would suggest cavitation. Damage is more likely from erosion. I cannot view all of your pictures on my tablet. But deep cavitation damage would be very rare in naphtha.

Johnny Pellin

 

[stanier]

Have you a pump curve showing the duty point , impeller size and system curve? Could be the pump is poorly selected for the duty. Is the pump fitted with minimum diameter impeller?

“The beautiful thing about learning is that no one can take it away from you.”
---B.B. King

http://waterhammer.hopout.com.au/

 

[1gibson]

What are your actual wear ring clearances? What is the extent of the "repairs" that are being done? Are you replacing the case wear ring, or only the impeller wear ring?

Looks to me like excessive flow across the wear ring area, causing erosion on the impeller shroud. Cavitation would be on the inside of the impeller eye, where pressure is reduced as velocity increases. Which, from the one picture, actually looks fine.

Only one photo was attached.

 

[Artisi]

It is not cavitation, I have to agree with others erosion.


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.)

 

[rutherford703]

Thanks all of you for your valuable opinions. I have to agree that the damage of the pump is due to erosion. While I cannot answer all your questions, I attach the mechanical drawing for the pump and show where the damages are.

It seems to me that the extensive erosion on the casing is a little bit weired because in that area the liquid should not have the highest velocity because the impeller is a closed plate. I also don't understand why the pump at the rated capacity the effeciency is only 35% (datasheet and pump curve were attached on my second post). Hope someone can shed some light on this.

If the erosion is the main issue, opening the recirculation valve to maintain the pump flow may not be the solution to mitigate the pump damage problem. Potential solution may be to install VFD to reduce the pump speed.

Thanks.

 

[1gibson]

You've got huge clearances in areas that should be maybe 0.020". That wear ring is the barrier between the suction and discharge areas, without it you are recirculating inside the pump. The impeller has done the work to move all that fluid, but it is not all coming out of the discharge flange because it is easier for it to recirculate internally.

Call someone in to inspect that knows what they are looking at, and then explain your proposed repair plan to the supplier you are using for replacement parts. Hopefully it is the pump OEM.

The only potential soluiton here is to restore and verify the wear ring clearances to API minumum values + no more than .0010" (being generous.)

 

[rutherford703]

Hi 1gibson,

Even if I unwilling to admit you are right, the erosion mode predicts what you said. As a process engineer, I cannot believe that the apparant error cannot be found by our pump specialist and causing the pump failure again and again.

It is the very likely the root cause. I give you a star again.

Thanks,

 

[rutherford703]

Actually, I rechecked the operating manual, the minimum clearance between the casing wear ring and impeller wear ring was followed API standard and is about 0.010". I haven't seen the wear rings after they were taken out and I am not clear how damage the wear rings are. I see the problem may be that the pump speed is too high (3560 rpm).

Thanks.

 

[JJPellin]

Damage in this area is likely corrosion / erosion from swirling. We have had good results using swirl breaks on these surfaces.

Johnny Pellin

 

[Artisi]

From the picture there is more going on than first meets the eye. Please explain the "some broken hole inside impeller".
Where is the pump running on its performance curve?
Additional clear photo's of the damaged areas might also help.

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.)

 

[rutherford703]

I didn't realize that the forum does not allow to attach multiple documents which I thought to be posted. Some new photos here.

Sorry

 

[rutherford703]

Here is the photo for impeller damage

 

[rutherford703]

Seal(?) damage