Abstract
Frequently, poor cement logs are observed in low gamma ray formations (sand). Contrarily, very good bond is often noted across high gamma ray formations (shales). In some cases this has resulted in a squeeze decision where no injectivity can be obtained incurring significant costs and delayed oil. Most commonly the problem has been ascribed to an artifact of the logs, presence of mud cake, cement shrinkage and squeezing shales.
However after studying many such logs the first observation is the sharp contrast between the logs across shales and sands. This can occur over an interval of less than a meter. We have also noted that within the same interval in a low gamma ray formation, very good and poor bonds can coexist. It is also shown that the frequency of poor bonding across shale does show a dependence on depth. It was concluded that these observations imply that fluid influx could be partially responsible for the poor bonding and mud cake was not the sole explanation for the effect. This influx is different to conventional gas/fluid migration and is associated with fluid entry into the cement after the cement has reached a oint of being able to prevent migration. Deeper analysis and correlation with reservoir pressure support this hypothesis, explaining why the log is very good in some sand intervals and poor across others. It would also explain the sharp contrasts in the log, as it is postplacement related while all fluids in the annulus are under static conditions. The remaining contrasts could be explained by greater bulk shrinkage of the cement across shales than across sands and squeezing shales. The larger fluid loss would allow influx of oil from the sands during cement setting, contaminating the slurry and therefore inducing poor logs. The paper describes, supports, and explains those mechanisms in detail. The role played by the liner top packer is also considered.
