New Separator Internals Cut Revamping Costs
- Rambout A. Swanborn (CDS Engineering BV) | Frits Koene (Burgess Manning) | Jan de Graauw (Delft U. of Technology)
- Document ID
- Society of Petroleum Engineers
- Journal of Petroleum Technology
- Publication Date
- August 1995
- Document Type
- Journal Paper
- 688 - 692
- 1995. Society of Petroleum Engineers
- 4.1.9 Heavy Oil Upgrading, 4.1.6 Compressors, Engines and Turbines, 5.3.2 Multiphase Flow, 4.1.2 Separation and Treating, 4.1.5 Processing Equipment, 1.10 Drilling Equipment, 1.10.1 Drill string components and drilling tools (tubulars, jars, subs, stabilisers, reamers, etc)
- 3 in the last 30 days
- 208 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 10.00|
|SPE Non-Member Price:||USD 30.00|
This article describes the debottlenecking operation of the productionseparators on the Shell Leman AK platform in the southern North Sea. Theoriginal, compactly built separators became overloaded through decreasing wellpressures. Gradually, they were forming a serious restriction in the productionflow sheet. Two alternatives were considered to overcome this problem: (1)extending the platform and installing new separators or (2) retrofitting theold separator vessels with state-of-the-art internals. The second alternativewas chosen because of considerably lower costs (totaling only 10% of firstalternative) and considerably shorter downtime. This separator upgrade was ajoint project between the operator of the field and their central research andengineering facilities and Burgess Manning, Technical U. of Delft, and CDSEngineering BV.
The debottlenecking operation formed part of a larger project in which gasproduction of an aging gas field was boosted through, among others, theinstallation of new compressors. Fig. 1 depicts the simplified process flowsheet of the Leman AK platform. The function of the platform is to removeliquids from the produced gas and to secure transport of the gas onshorethrough additional compression. The produced well fluids first pass through theslug catcher, which removes the bulk of the liquid. Downstream of the slugcatcher two parallel separators remove the finer liquids, so the gas is"fit" to enter the centrifugal compressors.
Recently, well pressures were at a level of 690 kPa, with typical flow ratesof 5.66 109 m3/d per vessel. It was doubtful whether the second stageseparators functioned effectively under such conditions, although the originaldesign stated that the separators could handle 16.99 109 m3/d per vessel. Afterthe planned installation of the compressors, we expected flow rates of 4.96 109m3/d per vessel at pressures of 290 kPa. The vendor of the original separatorsre-rated maximum flow rates per vessel at these pressures at 2.03 109 m3/d, andit became clear that the separators would not sufficiently guard thecompressors against incoming free liquids. Therefore, we considered itnecessary to improve the performance of this separation stage before installingthe new compressors.
Original Separator Design
The original separators (Fig. 2) consisted of 152-cm-diameter, 363-cm-longhorizontal vessels with 11 horizontally mounted, 20.3-cm cyclone tubes in thetop half of the vessel. The length of these cyclones was 250 cm. Although thisarrangement functioned well under original design conditions, we observed aclear deterioration of performance with increasing flow rates. In the end, nomore liquid was separated at all (Fig. 3).
Options To Improve Separation Performance
The two main alternatives to improve the performance of the secondseparation stage were (1) to replace the current separators with bigger onesable to handle the increased flow rates or (2) to remove the internals from thecurrent separators and install high-throughput retrofit internals.
The generally accepted advantage of new vessels (Option 1) was that we werecertain that performance would improve. The disadvantage of this option is thata platform extension would be required because of lack of space on theplatform.
The primary advantages of the retrofit option are the much lower costs andmuch shorter downtime. The disadvantages were that previous retrofit operationshad seldom been carried out under these conditions and that a range of novelengineering techniques had to be applied to guarantee success. The feasibilityand consequences of each of these alternatives will be detailed.
If new separators were to be designed for this operation, current design andengineering procedures prescribe the following. Vertical vessels would be 200cm in diameter and have a length between tangents of 290 cm. Weight per vesselwould be 11,500 kg. The cost of a new separator per vessel would be $200,000.Because the existing horizontal separators are difficult to remove, aconsiderable amount of platform reconstruction/extension work would benecessary. Estimated total costs were $1.25 million. Completion time (fromproject start in 1993) was projected for mid-1995.
Debottlenecking Existing Separators
The success of this alternative depends on fulfilling the followingrequirements.
1. Selection of a gas/liquid separation internal that can be fitted insidethe existing vessels and that can process the current flow rates.
2. Proper flow distribution to and from the internal, taking into accountthat the remaining area inside the vessels will be restricted.
3. A geometry that will allow the internal to be brought in through themanway/outlet nozzle and be mounted inside the vessels within a short period oftime (3 days) after removal of the current internals.
Selection of a Suitable Internal. Under the conditions specified, theoperating principle of the separation internal will rely on inertial forces.That is, centrifugal forces are used and separation is achieved because of thedifference in density between the gas phase and the contaminants. The currentinternals consisted of 11 20.3-cm cyclone tubes that function along thisprinciple. Under original design conditions, entrance velocities of thecyclones were 25 to 30 m/s and G forces in the cyclones were 300 to 500 G.
Normally with inertial forces as the principal force of separation,particles as small as 5 mm can be separated, which is in line with the purposeof the separators. The use of different separation principles (e.g., diffusion)only leads to larger separators, which was unwanted in this case.
Inertial Separator Internals. Two main types of inertial separator internalsexist: vane-type and cyclone-type separators.
Vane-Type Separators. In vane-type separators, the gas flows through abundle of parallel, curved blades. The droplets in the gas are flung onto theseblades (Fig. 4A) and subsequently drained. Various designs are on the market,but all have one property in common; that is, these internals require arelatively large area and also are sensitive to poor distribution andsubsequent local overloading in restricted surroundings.
For this particular application, 3.5 m2 of intake area of a sophisticatedupflow vane-type internal would be required. The available area in the vesselbetween the tangents is only 5.5 m2. Fig 4B shows a possible mountingconfiguration. Because of the almost impossible task of achieving an evendistribution across the vane area, we rejected this option.
Cyclone-Type Internals. In cyclones, the gas/particulate-phase mixture isbrought into a spinning motion.
|File Size||469 KB||Number of Pages||5|