Estimation of Post-Stack R-Factors Across Multiple Fields and Production Scenarios

The detection of pressure build-up by time-shift during injection requires a rock physics model that accurately captures strain-induced velocity changes. The post-stack R-factor model, introduced by Hatchell and Bourne (2005) and Røste et al. (2005), is generally presumed to possess a constant spatially invariant R, and in previous work has been estimated for production only, However, when this model is applied to injection, there are discrepancies between predicted and observed time-shifts. To understand this further, we attempt to estimate four distinct R-values (R+OB, R-OB, R+UB and R-UB) within both overburden and underburden layers, for injection and production zones. We calibrate these R-values in three distinct fields in the North Sea. According to our results, a fixed R-factor appears insufficient, and a spatially variable R-factor is necessary. There is some indication of differences between injection and production scenarios. A negative R-factor is also observed in the underburden of injection zones within chalk reservoirs, which indicates signal complexity and unknown causative physical mechanisms for the underburden time-shifts when compared to the overburden.