Pipeline Strainer 3

[EmmanuelTop]

Gathering station manifold feeds several wellhead compressors operating in parallel. The production stream is known for high content of sand and water, as a result of carryover from the separator system upstream.

Wellhead compressors installed (not commissioned yet) are equipped with suction scrubbers, and occasional water carryover can be managed by these scrubbers, as confirmed by field measurements and subsequent calculations for actual water load vs. scrubber capacity. The question of sand management still remains open. The compressors themselves do have suction strainers, and there is no possibility for sand to enter compressor suction.

However, the concern is frequent compressor shutdowns due to blockage of inlet scrubber and suction strainers, which is very likely if we don't remove the sand somewhere upstream. The idea we came up with, is to install a dual pipeline strainer system in the suction/production manifold, and with a mandatory option for back-flushing (self-cleaning). Now, the response we are getting from engineering companies is that there are no such strainers, able to filter the gas down to 10 micron particle size and being a self-cleaning type of strainer. What they offer is a cartridge type filter/strainer where we would have to replace the filter elements each time the filter gets blocked. Obviously this is not acceptable as we don't want to spend loads of money for replacing filter elements over and over again.

Does anybody have field experience with this problem, and to recommend a solution that will be reliable and operable, without high associated cost? What type of strainer system we are looking for? Or is there another way to combat against this problem?

Thank you in advance.

http://antwrp.gsfc.nasa.gov/apod/astropix.html

 

[zdas04]

What I had to do when I faced this same problem was change my expectations.

You are not going to get anything like 10 microns with a strainer. Accept that as a fact. And do you really need 10 microns prior to the compressor skid anyway? I thought I did, but I was wrong. I started out a wellhead compressor project with an inlet filter sep with 10 micron filter elements. The elements would last a week or so and then blow out from high dP. For the first few months we would drop piping and inspect the compressors every time we lost a filter element. Never found a problem on the flooded screws. So we started leaving the filter elements out of the filter seps and it didn't increase our failure rate at all. In 1998 I removed the inlet filter requirements from my spec (and pulled the elements from the existing equipment).

We put wye strainers (and a really inventive basket strainer that fit into a tee on the skid and was accessed through a blind flange) on the skids with local-read dP gauges around them. When we started seeing a non-zero dP we would clean the strainers. On a fleet of 64 wellhead compressors we would clean 2-3 strainers a month--not a huge maintenance burden.

My experience with both coal fines and sand has been that the big problem is high velocity sand blasting in small pipes. Consequently I increased pipe size as soon as I could after the wellhead, and required 6-inch piping or bigger on compressor skids and required cleanout ports on wellsite production units and compressor skid suction scrubbers. The problem never was as bad as we'd feared at design.

David Simpson, PE
MuleShoe Engineering

"Belief" is the acceptance of an hypotheses in the absence of data.
"Prejudice" is having an opinion not supported by the preponderance of the data.
"Knowledge" is only found through the accumulation and analysis of data.

 

[EmmanuelTop]

Thanks David, this helps a lot.

The compressors in this case are standard recip machines, gas engine driven, but this does not make any difference in what you are saying.

What Mesh size for the Y-strainer we are talking about? Is Mesh 200 something OK, or it is preferred to go for a larger screening size?

Many thanks again - your replies are always straight to the point.

http://antwrp.gsfc.nasa.gov/apod/astropix.html

 

[zdas04]

It has been a while, but I'm pretty sure that we got the finest mesh that was generally available without special order. Later we mixed that up a bit with courser mesh and it didn't seem to matter much. Eventually the big solids problem turned out to be phase-change scale (salt) that we'd see in the area of a "big" dP (it was weird, sometimes a 5 psid element would salt up immediately, other times it wouldn't, but a subsequent 0.2 psid would). I went back and put sock filters over the strainer baskets to try to force the phase change in a place I could clean out "easily". That never worked a single time and we pulled the socks after a few months.

David Simpson, PE
MuleShoe Engineering

"Belief" is the acceptance of an hypotheses in the absence of data.
"Prejudice" is having an opinion not supported by the preponderance of the data.
"Knowledge" is only found through the accumulation and analysis of data.

 

[SNORGY]

I think David's advice and experience are well founded and sound.

I would augment (not "add") that my design experience has concluded that once you get to 100 microns and smaller, no amount of conventional straining or adsorption filtration is going to prevent solid particles from reaching the compressor. At that point, the "sand" problem is not really a "sand" problem but a "fines" problem. Two mitigative approaches might be to (1) reduce the gas velocity by having a vertical "fat" separator or first stage suction scrubber in the hopes that most of the fines settle out (refer to GPSA Volume 1 Chapter 7); (2) consider something like a cyclone separator.

Sand carryover is usually discussed in the context of being frac sand or formation sand. Starting up wells where either is a concern often involves the temporary (sometimes permanent) installation of sand separation and filtration provisions that are operated and remain on line until the well has been cleaned up to a certain predetermined specification, like x barrels per day of sand per y MMSCFD. It is typical to sample the first batch of solids in a lab using a particle spectrum analyzer to generate the S-curve distribution of cumulative mass and corresponding particle size in order to see what size range of particles to go after towards cleaning things up.

With respect to erosion as alluded to by David, it is completely correct that the most effective mitigation against both erosion and carryover is in reducing the line velocities by up-sizing the pipe, which in similar fashion gives rise to the idea that for the suction scrubber, fatter is better. It is also interesting to note that below about 50 microns the sand particle size by itself does not dramatically affect predicted erosion rates. Depending on whether you are looking at Salama's correlations or the more sophisticated theoretical (well, empirical) erosion models, the governing influence is the kinetic energy supplemented by hardness and sharpness adjustments. Several researchers suggest that fine particles simply will not have the kinetic energy at the mixture velocity to penetrate the boundary layer stagnation zone and actually contact the metal surface.

All the above stated, if the problematic solids are 10-50 microns or even up to 100 microns, without clean-up, they will find their way into the compressor.

 

[SNORGY]

I should qualify the above...

At 100 microns and smaller, no amount of conventional straining or adsorption filtration is going to prevent solid particles from reaching the compressor *without plugging up on a regular basis*.