Cement setting problem across shale section
11 February 2011
Has anyone observed problems with cement not setting up correctly across shale sections resulting in poor bonding/micro annuli even though the cement above and below the shale section has set up correctly? This phenomenon has been observed in several wells with batch mixed 7” liner jobs were a portion of the liner is across a shale section. The 8 1/2” section is drilled with WBM - no problems encountered while drilling, hole in good condition and no problems getting the liner to bottom. No problems mixing/pumping the job - adequate liner RPM while pumping and plugs bumping as per schedule. However the results of both a CBL and USIT log show poor cement across the shale section every time with good bonds above and below (sandstone) the shale sections.
The spacer is fresh water based, barite weighted with KCL (2% BWOW) the cement is G neat (batch mixed) with gas block and various other proprietary additives - no KCL.
Has anyone encountered this problem before - come up with any solutions? The poor cement across the shale is having a big impact on future perforating / Fracing operations.
There are a couple of possible reasons (or a combination):
Since the shale is virtually impermeable compared to the sands, this particular section might be ballooning during drilling and cementing. Once the hydrostatic is reduced due to cement gelation (and setting the liner packer), the shale relaxes and gives back some of the ballooning fluids, contaminating the cement.
The other related reason is how fast the liner top packer is set. If done so before the cement gel reaches about 500 lbs /100 ft^2, the significantly reduced hydrostatic will all allow any formation fluids (i.e ballooning here) to influx into the cement matrix since the cement is not yet resistant to invasion.
In both cases, ballooning needs to be reduced to a minimum and any subsequent hydrostatic reducing ops be delayed till after the cement reaches 500 lbs. Alternatively, using film forming additives (e.g. Latex, PVA etc) can help minimize any influx. Finally, if non of the above is a factor, running a preflush of 40% sodium silicate solution (refer to service company pumping procedures) can improve bonding and logs a lot).
The spacer does not seem to be optimal but from all the logs and cement jobs I've seen I would speculate that this is a logging issue not a cement issue. Normally shales do not have a lot of permeability and therefore mud filter cake is not usually an issue. Obviously need to know a lot more detail but there logs can be affected by formation type. With a correctly mixed batch slurry, if you have good log above and below the shale the cause, all else being resonable equal , can only be a formation effect. I have seen this effect before but in those cases where caused by slow formation. If you have open hole sonic logs compare these to the CBL and if you can see a similar response on the OH sonic then it is likely a formation effect.
Sands above and below shale in between.
Seems to me to be incompatibility of mud, formation and cement spacers etc used. You may not be eroding the shale filtercake and hence will not get bonding required. So non erodability may be the key!
Spacer, mud, cement rheology and optimisation techniques may need enhanced for the shale.
I have emailed a doc to Dave that he will forward on to you.
If you make changes and these work we would appreciate your feedback on this.
How certain are you that the log results are reliable? Could it be that the cement job is good but the log response in the shale section is questionable? I can not think of any reason that the cement job would be good above and below a shale yet poor across it.