The geological complexity and limited access to the subsurface result in a big uncertainty in reservoir properties and forecasts, however there is a systematic tendency to underestimate such uncertainty, especially when rock properties are modeled using Gaussian random elds. We address uncertainty in regional trends by using hierarchical parameters that are estimated using ensemble Kalman lter (EnKF) for history matching.
Hierarchical or multi-scale heterogeneities are generally poorly represented, especially in deepwater reservoirs. We developed a hierarchical description of heterogeneities that introduced new variables into the reservoir description. We tested our method for updating these variables using production data from a deepwater eld whose reservoir
model has over 200,000 unknown parameters. The match of reservoir simulator forecasts to real eld data using a standard application of EnKF had not been entirely satisfactory, as it was difficult to match water cut in one of the wells. None of the realizations of the reservoir showed the arrival of water using the standard method. By
adding uncertainty in trends of reservoir properties, the ability to match the water cut and the other production data was improved substantially.
The results indicated that an improvement in the generation of the initial ensemble and in the variables describing the property fields gave an improved history match with plausible geology. It helps to avoid the tendency to underestimate uncertainty while still providing reservoir models that match data.

Black line: ensemble models        Green line: mean of ensemble       purple points: observations