10 replies to this topic

[Guido]

Hi All,

I would like to share with you a separation system design which I found interesting. This topic has been already treated in other posts.

The case referes to a three phase separation carried out in two separators, the treated oil quantities are: 15,000 BOPD; 15,000 BWPD & 5 MMSFD.

The two separators were refurbished priot to their new use and their have weirs and demister were kept. The designer chose a submerged weir separation design . As for the vessels size, the 1st separator has a L/D=4.28 with D= 2134 mm and the 2nd separator has a L/D=2.86 with D= 2134 mm.

The water and oil level controls are shown in the attachment and the separation goes on non-stop successfully since the plant started.

I have encountered other designs, overflow weir, boot, etc., but this is the first time I have seen such a design. I have checked on the references I have and the ones I can reach and, I have found no evidece of the submerged weir design. However, a friend of mine has mentioned that Shell DEP mentions the existance of the aforementioned design but there is no an actual explanation.

It seems to me that the level controls are a bit close to each other but perhaps the designer has carried out this design on more that one occasion and they have their on methodology to carry out the calculations.

In previous posts I have found that some of the replies suggest that this is not the best design (or at least no popular).

In terms of design I would hace chosen a different one. Perhaps, some of you have encountered a similar case to the one above. I just want to have your views and comments and your experience on the case.

Cheers,

Attached Files

•  Phase Separation - Submerged Weir.pdf   104.33KB   579 downloads

[mav9rick]

Hi Guido,
This design is not entirely uncommon especially for offshore application. As you are probably aware, real estate available and weight of the equipment are major factors that are considered for a topsides design.
This design is typically suitable when the drum is acting as a feed surge drum/ inlet separator in a low gas flow application. By sacrificing a bit on the cross sectional area available for gas flow, the designer is able to use the volume for liquid holdup.
What application are these seaparators installed in?
Cheers

[Guido]

Hi mav9rick,

Thank you for your answer and views.

These set of two separators run on an onshore facility, the feed is crude oil from a set of wells, so there is no physical constrain in terms of space, etc. The gas level of 5 MMSFD is not huge but big enough to conider sizing a gas compartment to accommodate that quantity. The gas values within the vessel vary with the liquid level (please see attached tables).

I am partcularly interested in the heuristics used on this type design to set up the liquid control level as in other designs the level are well known. In addition, my unfamiliarity with this makes me think that the designer has probably "adjusted them" based on its own experience.

You point out that this separator design is favoured on offshore application on topside equipment design, well I must say that my experience is related to onshore facilities, so far.

Cheers,

Attached Files

•  Gas area variation w_liquid level.pdf   40.96KB   310 downloads

[mav9rick]

Guido,
Can you please supply me the following data:
1) Operating Pressure
2) Oil Density if available. Else you can tell me if the separator recieves crude oil in a refinery or gas/water/condensate mix in a gas plant.
3) is the separator pumped out or does the liquid flow under pressure control
4) a sketch of the upstream and downstream systems if posdible
Once I havethis info, I can run some numbers for you based on typical K values
Cheers
Karan

[Guido]

Hi Karan,

1) The operating pressure is ~4 bar_g and the liquid flow is under pressure control; in addition:

Fluid

Density

Kg/m³

Temp

(°C)

Viscosity

(cP)

Hydrocarbon

780

65

1.11

Gas

5.95

0.01

Water

1019

1.33

2) The oil feed comes from onshore wells and processed at an upstream facility (not a refinery)
3) See the simplified PFD in the attachment

By the way, is there any design methodology for this kind of separator to evaluate or define the level control or... any reference? you seem to know it well...

Cheers,

[Guido]

Karan,

Find the attached the system diagram

G

Attached Files

•  Production Facility Simplified PFD.pdf   18.3KB   322 downloads

[mav9rick]

Guido,
Thanks for the information. I just one more question before my final response. What do the vertical thick lines on the vessel GA represent? Are there any internals in the water/oil chamber (left side) that you are aware of?
cheers
Karan

[Guido]

Hi Karan,

I must be honest I do not know. If I had to guess I would say that the line in the first separator represent a weir. On the 2nd separator, not sure. If this was an overflow weir design I would certainly know, on this submerged type, not sure.

G

[mav9rick]

Guido,
Here is an explanation to the questions you raised in your post. Apologies for the late response, I have been a bit busy last week.

In order to answer your questions, I am using the level settings for Separator 1 as an example below:

1) First of all, let me clarify that the submerged weir design is not entirely uncommon or unpopular. This type of design is more commonly used in low GOR, surging service and is used to rationalise vessel size. Examples of such service would be crude receivers and feed surge drums (similar to your application). Apart from achieving the oil/water/gas separation, another objective of such a vessel is to absorb the surge coming from upstream sources (wells in your case) and not let the downstream process be exposed to swings in the feed flowrate. As you can appreciate, your vessels are not small by any means. using an overflow weir design for low GOR would have meant longer lengths than what they currently have.


2) By allowing the interface level (0.7m for Separator 1) to operate below the weir level (1.06m for Separator 1) and oil normal level (1.2m in Separator 1) to operate above the weir level, the designer is making use of full vessel length to absorb any fluctuations in the incoming flow. It is for this reason that the levels bands above weir are tighter than the ones below the weir. In other words, because the full length of the vessel is available for liquid hold-up above the weir, the level band width can be reduced. the separation chamber(left side of the weir) provides enough length for the oil and water to separate. The methodology for calculating the require separation length (and hence the position of the weir) is no different to that for a typical over flow weir type 3-phase separator.

I have checked your separator design for the flows and conditions given by you and found that both vessels provide sufficient liquid retention between the low liquid alarm (LL) and high liquid alarm (HL) so as to give the Operator or control system sufficient time to respond and also to allow the gas to escape the liquid. Imagine a can of coke, it takes some time for the dissolved gas (commonly referred to as solution gas) to come out of the liquid phase. Typical retention time for gas-oil applications is 2-4 minutes.

Perhaps as an exercise for yourself, you can try and calculate the retention time between different alarm bands based on the incoming flowrate.

3) I requested for the process PFD in order to confirm if the separator 1 and 2 are operating in series, which they are. As you know, the gravity separators will not guarantee 100% liquid-liquid separation. Normally there will be an "emulsion layer" above and below the interface level. It is for this reason that the depth of the oil band needs to be rather large in the in the submerged weir type arrangement than that in the overflow weir type separator.

The level bands in the second separator are narrower in the region below the weir because the oil flow (estimated carryover from 1st Separator) will be small and therefore sufficient retention time (or alarm response time) can be provided by allowing a narrower level band. Also, the emulsion band is likely to be narrower because you have a heater between the 2 separators which will reduce the viscosity difference between oil and water and likelihood of presence of strong emulsion.

4) I have checked the actual gas velocity for different liquid levels in the vessel and it closely matches your estimate. More importantly, the actual gas retention time (at HLL) is higher than the minimum required for droplet separation assuming a conservative K value (measure of efficiency of the separator to separate liquid from gas) of 0.06 . This means that your separator is sized correctly for the required gas/liquid separation.

Hope the above helps.
Cheers

Edited by mav9rick, 27 October 2011 - 12:05 AM.

[Guido]

Karan,
Thank you for your reply and help. It is much appreciated.

I fully agree with your position. The rational to go for the submerged weir design was based on the use of an existing vessel which was key for the project at the time. Moreover, the fact that separation section needed to absorb the surge coming from upstream source as well as adding a liquid self-controlling element to that specific design.

Concerning the water levels, I have calculated them and I found no problem with the design as the interface level is set below the weir level. In addition, the distance between the IL and the oil low level creates a liquid seal that allows the separation operation work adequately. It is clear that the designer focused on maximising the vessels available capacity to accommodate all flows (gas & liquid) as well as slug.

Respect to the methodology for calculating the require separation length, weir, etc., I agree with you the design steps and retention time calculation are similar to a typical over flow weir type design for a three phase separator.

I must admit that I was a bit uncomfortable with the oil level band width as it is a bit narrow when looked at the design. This is why I raised the point in the forum. One appreciate that working with an existing vessel may bring up adjustments to the selected design that may differ from other. In this particular case, the oil level bands are narrower than one would expect considering typical values for overflow design heuristics.

Once again, thank you for sharing your time and knowledge.

Cheers

G

[mav9rick]

No worries Guido. Oh and one more thing, I think the vertical bold lines in the left hand side in the vessel GA are calming baffles.
Cheers
Karan