20 replies to this topic

[Falah]

Hi,engineers
To compress the gas released from 3 separators and stabilizer, we have four compressors ( LP, MP, HP and one for gas of stabilizer). We found in the discharge of these compressor two vessels for separating liquid and gas (Discharge scrubber and separator ) after after-cooler. we asked the consultant for the reason of adding two vessels they said because the flow from air cooler is two phase and it is not easy to select recycling valve. we think that there is no need to Two ( highlighted one in attached sketch should be cancelled in our point of view) . Only the vessel at the suction of every following stage is enough to eliminate and catch any liquid carryover from the upstream equipment and any condensation caused by cooling in the lines leading to the compressor. Please see attached for our case and the typical sketch.
Thank you in advance

https://we.tl/t-Cq86CsGbPP

Edited by Falah, 21 July 2024 - 12:00 PM.

[Pilesar]

Unfortunately, the link supplied does not lead me to useful information -- I find no sketch. If the vendor has sized the equipment, there should be a material balance somewhere that vendor can produce for your review which shows the expected liquid carryover. If there is a cooler before the recycle valve then a KO drum before the recycle makes sense if this is 2-phase region. If during turndown cases the recycle will be open some during operation, then a single-phase vapor will be much easier to meter and control to keep the machine steady. If each compressor stage has its own kickback, then I would not feel comfortable without also having a KO drum for the suction of each stage to keep the operation of each stage as independent as possible. I would like to see your sketch as I am only offering generic guess of the arrangement. Is the 'discharge scrubber' the vessel you propose to eliminate? What is its purpose?

[fallah]

Hi,engineers
To compress the gas released from 3 separators and stabilizer, we have four compressors ( LP, MP, HP and one for gas of stabilizer). We found in the discharge of these compressor two vessels for separating liquid and gas (Discharge scrubber and separator ) after after-cooler. we asked the consultant for the reason of adding two vessels they said because the flow from air cooler is two phase and it is not easy to select recycling valve. we think that there is no need to Two ( highlighted one in attached sketch should be cancelled in our point of view) . Only the vessel at the suction of every following stage is enough to eliminate and catch any liquid carryover from the upstream equipment and any condensation caused by cooling in the lines leading to the compressor. Please see attached for our case and the typical sketch.
Thank you in advance

https://we.tl/t-Cq86CsGbPP

 

The link content is irrelevant to what you described!

 

Please upload a simple sketch of the system you described...

[Falah]

Unfortunately, the link supplied does not lead me to useful information -- I find no sketch. If the vendor has sized the equipment, there should be a material balance somewhere that vendor can produce for your review which shows the expected liquid carryover. If there is a cooler before the recycle valve then a KO drum before the recycle makes sense if this is 2-phase region. If during turndown cases the recycle will be open some during operation, then a single-phase vapor will be much easier to meter and control to keep the machine steady. If each compressor stage has its own kickback, then I would not feel comfortable without also having a KO drum for the suction of each stage to keep the operation of each stage as independent as possible. I would like to see your sketch as I am only offering generic guess of the arrangement. Is the 'discharge scrubber' the vessel you propose to eliminate? What is its purpose?

Pilesar, please see attached link, l hope it clarifies my concern.
we do not think there is necessity or need to use these vessels. only scrubber after cooler between stabilized gas compressor and LP compressor and between LP compressor and MP eliminates liquid carryover to the LP compressor in first case and to MP in the second.The consultant justifies adding these vessels by saying there being difficulties in selecting recycle valves if we cancel them since the recycle line will be from the discharge of after cooler and flow after cooler is two phase flow and part of the liquid may be reduced in pressure and flashes into gas, which will cause confusion in the composition of the suction separator and affect the selection of the compressor.
In mos cases that l have seen before, the recyle line is from the discharge of cooler even if we assumed two-phase flow after cooler and its impact on selecting recycle valve

https://we.tl/t-8t7Fy5rwZJ

Edited by Falah, 22 July 2024 - 07:28 AM.

[Falah]

Hi,engineers
To compress the gas released from 3 separators and stabilizer, we have four compressors ( LP, MP, HP and one for gas of stabilizer). We found in the discharge of these compressor two vessels for separating liquid and gas (Discharge scrubber and separator ) after after-cooler. we asked the consultant for the reason of adding two vessels they said because the flow from air cooler is two phase and it is not easy to select recycling valve. we think that there is no need to Two ( highlighted one in attached sketch should be cancelled in our point of view) . Only the vessel at the suction of every following stage is enough to eliminate and catch any liquid carryover from the upstream equipment and any condensation caused by cooling in the lines leading to the compressor. Please see attached for our case and the typical sketch.
Thank you in advance

https://we.tl/t-Cq86CsGbPP

The link content is irrelevant to what you described!

Please upload a simple sketch of the system you described...

Please see the attached link, and my reply to Pilesar
https://we.tl/t-8t7Fy5rwZJ

Edited by Falah, 22 July 2024 - 07:27 AM.

[breizh]

Hi,

Use the built in tool to upload your document. We don't want to use software we are not familiar with.

Click the more reply option and then choose file and upload your document.

The answer from supplier is based on experience and it's up to you to decide what you want to implement.

Good luck

Breizh

[Falah]

please see attached

Attached Files

•  photo_2024-07-22_16-11-59.jpg   108.36KB   0 downloads

[fallah]

Hi,engineers
To compress the gas released from 3 separators and stabilizer, we have four compressors ( LP, MP, HP and one for gas of stabilizer). We found in the discharge of these compressor two vessels for separating liquid and gas (Discharge scrubber and separator ) after after-cooler. we asked the consultant for the reason of adding two vessels they said because the flow from air cooler is two phase and it is not easy to select recycling valve. we think that there is no need to Two ( highlighted one in attached sketch should be cancelled in our point of view) . Only the vessel at the suction of every following stage is enough to eliminate and catch any liquid carryover from the upstream equipment and any condensation caused by cooling in the lines leading to the compressor. Please see attached for our case and the typical sketch.
Thank you in advance

https://we.tl/t-Cq86CsGbPP

 

Hi,

 

It's hard to judge about the reason of considering two vessels at discharge of the compressors, for separating the liquid from the compressed gas, provided by your consultant without having the details of the relevant simulation and design in hand. 

Anyway, the consultant explanation might be reasonable if either the aftercoolers haven't adequate capacity (may be due to avoid huge capacity cooler) to remove all condensate in just one step forcing the designer to consider another vessel to remove the remain condensate, or the length between the scrubber and subsequent compressor is so long such that forming the liquid condensate before reaching the gas to this compressor would be probable.

[Pilesar]

There do seem to be extra vessels in this system. Be cautious in declaring they are not needed. Make sure you understand the reasons they were put in the system. The material balance should help. It may be that there are very good reasons for each of these vessels that even the current vendor spokesman is not aware of. If you remove some of this equipment and there is a subsequent liquid carryover to the compressor, the damages could be very expensive! The excuse 'well we thought it would be okay to leave this vessel out' would be evidence of incompetence. I don't know whether this system is properly designed. But you need to make sure you know!

[breizh]

Hi,

Some literature about G/L separators.

Good luck

Breizh

Attached Files

•  2-PhaseSeparatorSizing reviewed.pdf   6.35MB   10 downloads
•  Gas_Liquid_Separation_Technology.pdf   2.95MB   12 downloads
•  Gas-liquid separator Fundamentals.pdf   296.93KB   6 downloads
•  12_Feat_SeparationPt3.pdf   2.45MB   5 downloads
•  07_FeatSeparators_Final.pdf   903.37KB   5 downloads
•  08_Separation.pdf   542.02KB   5 downloads

[Pilesar]

Breizh, the '2-PhaseSeparatorSizing reviewed.pdf' attachment seems to have later corrections written in. I assume those are from the author's followup in a later journal issue. As much of that article has degraded from repeated copying, I attach 'Design two Phase Separators Within the Right Limits (uncorrected).pdf' as a more readable copy of the original article but without the corrections. I stored your corrected version with my more readable version so I could refer to both versions later if needed. I don't currently have convenient method to add your corrections to my readable version.  Design two Phase Separators Within the Right Limits (uncorrected).pdf   574.15KB   6 downloads

[breizh]

Hi Pilesar.
I've both version, corrected and not corrected. Unfortunately I don't remember the source for corrected one.I believe it's due to typos.

I've attached the extract from GPSA.
Breizh

Attached Files

•  extract GPSA typo (1).docx   184.67KB   10 downloads

[Pilesar]

I digitized the drag coefficient graph from GPSA found in Breizh's attachment since I wanted the drag coefficient in a spreadsheet without a visual lookup step. I regressed the curve reasonably well up to a drag coefficient of about 60. 

In practice, I use the Excel 'linterp' function to interpolate between the data points in the digitized table since it is a bit more accurate than the regression equation. But the regression equation is not bad. 

After the abscissa (X) is found using GPSA Eq 7-3, the drag coefficient (Y) is found by the following regressed equation:

Y = EXP(a*LN(X)^3+b*LN(X)^2+c*LN(X)+d)

where:

a= -0.001015374611086

b= 0.061434896668678

c= -1.191791825735960

d= 6.555477746278450

 

I truncated the regression parameters above as shown. The significant digits are left for the reader to choose.

 

For whatever reason, I made a hobby of separator sizing. There are so many different ways published for separator sizing ranging from API 12J to Shell DEP to many engineering journal articles! The GPSA method is what I found the best for me. 

Another good reference is:

Watkins, R.N.: "Sizing Separators and Accumulators", Hydrocarbon Processing; November, 1967, pp 253-256.

[breizh]

Hi Pilesar,

I was ready to do the same. Thanks for sharing a neat document and correlation.

Breizh

[Pilesar]

I don't often find a need to digitize graphs but it was helpful for this drag coefficient chart. Another use I've found is for centrifugal compressor curves since transferring points by hand is tedious and I am prone to make annoying errors when reading graphs manually. The Engauge Digitizer software I use to convert curves to coordinates is free at https://sourceforge....ects/digitizer/ The version I have is old -- I think version 2. The latest release is up to version 12 or so. There are other free programs that do similar actions.

[Falah]

I digitized the drag coefficient graph from GPSA found in Breizh's attachment since I wanted the drag coefficient in a spreadsheet without a visual lookup step. I regressed the curve reasonably well up to a drag coefficient of about 60.
In practice, I use the Excel 'linterp' function to interpolate between the data points in the digitized table since it is a bit more accurate than the regression equation. But the regression equation is not bad.
After the abscissa (X) is found using GPSA Eq 7-3, the drag coefficient (Y) is found by the following regressed equation:
Y = EXP(a*LN(X)^3+b*LN(X)^2+c*LN(X)+d)
where:
a= -0.001015374611086
b= 0.061434896668678
c= -1.191791825735960
d= 6.555477746278450

I truncated the regression parameters above as shown. The significant digits are left for the reader to choose.

For whatever reason, I made a hobby of separator sizing. There are so many different ways published for separator sizing ranging from API 12J to Shell DEP to many engineering journal articles! The GPSA method is what I found the best for me.
Another good reference is:
Watkins, R.N.: "Sizing Separators and Accumulators", Hydrocarbon Processing; November, 1967, pp 253-256.

Pilesar if you can share with me
"Sizing Separators and Accumulators" for Watkins, R.N.

[Pilesar]

You can find the Watkins article here: https://www.scribd.c...ccumulators-pdf

Edited by Pilesar, 26 July 2024 - 07:09 PM.

[breizh]

Hi,

Consider this link

 

https://pdfcoffee.co...6-pdf-free.html

 

Breizh

[Falah]

Hi,
Consider this link

https://pdfcoffee.co...6-pdf-free.html

Breizh

I downloaded it from this link before but it can not be read it is not clear enough Breizh.

[breizh]

Hi,

At least you can zoom it and read it.

I cannot offer more 

Breizh

[astro]

The papers by Bothamley offer excellent coverage of the technical aspects.

 

According to the GPSA Engineering Data Book, 14th ed., Ch.7 Separation Equipment, under the heading "Performance Requirements", a single gas / liquid separator should be able to achieve liquid entrainment performance in the gas outlet as follows:

Droplet size removed: 99 % removal at 10 µm
Volume fraction entrained: 0.0134 m³ / MSm³ (0.1 USgal/MMscf)
Overall liquid recovery: 98 %

 

I'd expect the above to be adequate for applications supplying gas to compression equipment.

 

Apart from potentially including a separate slug catcher at the front end of the process, what is the machine sensitivity requiring multiple separation steps in series?

What fluid characteristics and / or behaviour are at play to justify this level of process caution?

 

If there is a need to have high confidence in the gas quality supplied to the compressor(s), then look at the Sulzer document posted by Breizh, https://www.cheresou...attach_id=17077 and head to pp.18-19 Combined Systems / Shell High Capacity Separators to gain an understanding of what can be done with a single vessel and the right internals specification.

 

I agree with Falah's scepticism that the proposed flowsheet could be simplified significantly, assuming that the fluids do not present atypical challenges. All those vessels will come at a capital cost. If it was my money, I'd want some justification that the spend is justified.