Patterns of water Sr-87/Sr-86 variations in oil-, gas- and water-saturated rocks: Implications for fluid communication processes, distances and timescales

This study reviews Sr-87/Sr-86 depth profiles of formation waters sampled by Sr residual salt analysis (Sr RSA) from >100 oil/gas wells and research sites, including reservoirs with elastic and carbonate host rocks and with gas, oil and water as the continuous fluid phase. Globally, the water data form a smooth trend between low seawater-like Sr-87/Sr-86 ratios (similar to 0.706) at shallow depths and high (similar to 0.724) ratios in deeply buried rocks, where water-rock interaction dominates. We test the hypothesis that Sr-87/Sr-86 depth profiles in individual wells could be influenced by diffusional mixing processes by developing 1D diffusion mixing equations to simulate compositional patterns through time and comparing them with observed profiles. Different combinations of boundary and initial conditions generate various patterns characteristic of diffusion, including non-steady-state curves relating to incomplete mixing and steady-state patterns (such as vertical or inclined straight lines) where initial heterogeneities have fully mixed. The dataset yielded 193 occurrences of these patterns. Steady-state patterns are more common and longer in water zones, while non-steady-state patterns are more common and longer in oil and gas zones. The detection of diffusional mixing patterns in hydrocarbon-saturated rocks suggests that diffusion is active, although on average a factor of similar to 13-18 slower, than in comparable water-saturated rocks. Pattern generation and equilibration times were modelled for each non-steady-state pattern and compared with the time since reservoir filling with oil/gas, revealing that 90% of them could have been generated since filling, but 60% of them would already have mixed to steady state had the initial compositional heterogeneities arisen during or before reservoir filling. This is critical evidence that at least some of the initial heterogeneities must have arisen, and subsequently partially mixed, after filling; these patterns tend to be short (