Multiple model seismic and production history matching: A case study

Time-lapse (or 4D) seismic is increasingly being used as a qualitative description of reservoir behaviour for management and decision making purposes. When combined quantitatively with geological and flow modelling, as part of history matching, improved predictions of reservoir production can be obtained. Here, we apply a method of multiple model history matching based on simultaneous comparison of spatial data offered by seismic, as well as individual well production data. Using a petro-elastic transform and suitable re-scaling, forward modelled simulations are converted into predictions of seismic impedance attributes and compared to observed data by calculation of a misfit. With a similar approach applied to dynamic well data, the total misfit is used to update model probabilities in a Bayesian framework. These are then used to guide the re-sampling of the parameter space, generating multiple models using the quasi-global stochastic Neighbourhood Algorithm (NA) method. This approach improves on gradient-based methods by avoiding entrapment in local minima and can be used to efficiently analyse uncertainty of the matched parameters. We demonstrate the method by applying it to a UKCS turbidite reservoir, updating the operator's initial model, which had previously been history matched to dynamic well data by conventional methods. We consider a number of parameters to be uncertain. The reservoir's net to gross is initially updated to better match the RMS amplitudes of the observed seismic baseline survey. We match simultaneously for permeability, fault transmissibility multipliers and the petro-elastic transform parameters. Our results show a good match to the observed seismic and dynamic well data with significant improvement to the operator's base case.