Single and multiphase flow properties of fault rocks: implications for petroleum production

Geological faults are known to have a major impact on fluid flow in the subsurface and may developed isolated compartments for hydrocarbon reservoirs. The knowledge of their properties is important for petroleum industries within interests that include oil and gas production, Carbon dioxide storage and radioactive waste disposal. Large data sets have been collected on the single-phase permeability of fault rocks but these have been collected under inappropriate laboratory conditions such as low confining pressures and using distilled water as the permeant. Some data have been published on the gas relative permeability of fault rocks but no data is available on oil-water relative permeabilities or from fault rocks that are not strongly water-wet. The current thesis aims to produce high quality experimental data to partly fill these knowledge gaps by collecting gas and brine absolute permeability data from fault rocks at reservoir stress conditions using a formation compatible brine as well as oil-water relative permeabilities from fault rocks that are water-wet and after wettability alteration so that they become less water wet.
The key findings of this thesis show that the absolute permeability of fault rocks is very stress sensitive due to the presence of microfractures created during coring or core-retrieval. The stress sensitivity of permeability increases with decreasing permeability. On average, fault rocks have a permeability at in situ stress which is ~ 20% that measured at ambient conditions. Permeability is also found to be less sensitive to brine composition, with permeabilities to distilled water being around 20% those when measured with brine. So the effects of two poor laboratory practices cancel each other out meaning that much published data remain usable.
The obtained results from oil water relative permeability measurements of water-wet cataclastic faults are consistent with what is known about grain-sorting controlling relative permeability. The changes in wettability resulting from aging the samples in crude oil are also consistent with what would be expected when the wettability of samples is altered to being neutral to oil-wet. These results raise the possibility that fault rocks in some reservoirs may not be strongly water-wet and will therefore not act as capillary barriers.
A new clay-mixing model is also presented, which explains the scatter on permeability vs clay content that is used as input for calculating fault transmissibility multipliers in production simulation models. The obtained results can be used as the analogues for similar fault rock types and implemented in the reservoir simulation models for the future forecast of hydrocarbon production.