I’ve previously worked on the Vincent project here at Woodside (multi-laterals with SAS/ICDs/swell packers) but it’s a little while ago (2011-12) so there may be additional points/improvements etc. I’ve cc’d one of the team (Anthony Palmer) but he is busy supporting the upcoming operation on this project so may not have time to add much.
I’ve put my comments in bold below:
1. am curious to know others experience in the SAS/ICDs and Swell Packers through a Milled Windows?
Yes, on Vincent project we have run many kms of screens in this fashion (laterals up to 2000m +).
All of these are with 5-1/2” ICD screens with centralisers (one per joint) and swell packers (with centralisers either side). AP: 8.5m screen jacket length on R3 17# or 20# tubing joints. 6.61” OD screen jacket
We have done bi-laterals (one window) and tri-laterals (two windows).
There have been many variables, but this has always been exiting 10-3/4” casing. Phase 1 and 2 was 9-1/2” open hole using SBM, Phase 3 was 8-1/2” open hole with WBM. AP: 10-3/4”, 55.5# casing
An SPE paper was written on this project :
Nettleship, Gary, Woodside Energy Ltd
Palmer, Anthony, Woodside Energy Ltd.
Eshtewi, Abdullah, Woodside Energy Ltd.
169903-PA SPE Journal Paper - 2014
2. Any evidence that the Swell Packers or SAS/ICDs were damaged?
No evidence of swell packer damage, but not sure that we would ever really know as these can’t be tested
Wells never produced sand and were in totally unconsolidated sand, so reasonable to say screens were not damaged.
3. Notice any difference in production to a similar ICD/SAS Swell Packer completion that was installed through a standard casing shoe. Ways to protect?
Sometimes saw hang-up of assembly come out of the window but this was usually attributed to centralisers (various types were used including WFD Lo-Torq, BOS-22 and Matrix from memory). AP: 9-1/2” open hole wells used 9.13” OD Lo-Torq roller and Caledus BOS 22 thermoplastic centralisers but these were phased out early in the project. 8-1/2” open hole wells used 8.13”OD Lo-Torq centralisers and Matrix thermoplastic centralisers but in house T&D modelling using actual slack off data shows Lo-torqs provide the lowest cased and open hole friction factors.
We avoid rotating screens
I think the key is probably how the window is milled: the team always puts a lot of effort into milling procedures and clean up – paying attention to parameters throughout, especially the mid-ramp and beyond, dressing the window off and checking with the milling BHA (pumps on/pumps off, with and without rotation etc). AP: Procedure/guideline instructions is key to a successfully milled window. We ‘time-mill’ the milling operation over a specified timeframe to provide the best opportunity for success. As tom mentions, parameters including rate of penetration, flow rate, maximum applied torq to the string (dictated by anchor packer torque rating), consistent checking of ROP using the index line (semi-submersible rig), timing/volume to pump hi-vis sweep pills, location of magnets in BHA (MI magno-stars recommended for high recovery), time to drill rat hole from bottom of window – soft, unconsolidated formation, pumps on/off passes to ensure no hang ups with milling BHA prior to POOH & RIH with dedicated OH drilling BHA.
4. Window Orientation Preference and DLS guidelines?
We do both left and right side exits in the tri-laterals.
From memory we go 45 degrees LOHS but only 30 degrees ROHS – reason for the latter is the mill tends to walk off to the right so we try to avoid a big drop section coming out of the window. . AP: Correct.
We have very specific guidelines on hole condition for the area immediately outside the window (this gets washed out during milling and maybe subsequent drilling). This is very important as the multi-lateral connector is stabbed into the screens which are dropped off in open hole just outside the window e.g. it’s important to have the top PBR on the screens in gauge hole so the connector can stab in. We get acoustic calliper data after drilling to confirm this and identify suitable location to position the top of screens (PBR). AP: SLB RapidX TAML level 5 multilateral junction used in 13 wells completed to date.
There are always very high local dog legs coming out of the window but I can’t recall exact figures.
5. We will be running T&D modeling to confirm installation but we do anticipate some higher than normal DLS, 8-10 deg and any experience anyone can offer on running SAS/ICDs at high DLS.
There are always very high local dog legs coming out of the window but I can’t recall exact figures... maybe 10-15 deg/30m? We have had some challenging anti-collision issues which have meant some quite high build/turn rates beyond that too. . AP: We also have a very thin reservoir so any loss of directional control resulted in breaching the gas/water zones and resulted in an open hole sidetrack. These were blanked off with blank tubing joints and swell packers. Sidetracking obviously increased dogleg severity and overall well tortuosity.
Lots of focus on T&D modelling, especially landing string configuration (HWDP, collars etc) but I think this generally comes into play more for getting screens to depth, rather than specific issues coming out of the window.
We have used a swivel master with some success on very long/tortuous well paths – gave us back I think 30-50 klb set down weight on some occasions. AP: SwivelMaster has been included on lateral completion strings – not Motherbore due to liner hanger running tool release requirements and lack of contingency release if the SwivelMaster fails to lock. The SwivelMaster has only been used once to get the screen completion to TD. The landing string has been intentionally rotated at specified depths to record and understand slack off weight increase provided by the reduction in friction. The reduction is more like 20 – 40klbs.
5. Any installation Best practices? Dummy Run?
We don’t do a dummy run but use data from the final drilling trip in/out to calibrate the T&D model for running screens. AP: Dummy run can increase risk of hole stability issues depending on formation, and also costly for ERD semi-sub application. For Vincent: After TD is called, several bottoms up circulations are completed and the returns flow through tested (screen mesh coupon). The drilling BHA is pulled back to the casing shoe where a hi-vis sweep is pumped and well volume (shoe to surface) circulated. Following that, a slack off check trip back to TD is completed for the purpose of accurate T&D modelling. At TD the well is again cleaned up until flow through testing is achieved. The OH and most of the casing is then displaced to solids free WBM and the drilling BHA pulled back to a specified distance above the casing shoe. The remainder of the casing and the riser is then displaced to solids free WBM prior to the screen completion being RIH.
I think the key is probably how the window is milled: the team always puts a lot of effort into milling procedures and clean up – paying attention to parameters throughout, especially the mid-ramp and beyond, dressing the window off and checking with the milling BHA (pumps on/pumps off, with and without rotation etc).
We’ve cored mills before when sitting in one spot for too long, so worth spending a lot of time with the milling service company (we use Smith). AP: Trackmaster
Hope this helps,