Integration of History Matching And Uncertainty Analysis

Abstract

The most common procedure to perform a production history matching is to start with a base model and modify reservoir and fluid properties to adjust simulation results with the production history of the field. This paper presents an example of a different procedure. The history matching starts with an uncertainty analysis where several possible models are generated and the models that do not reproduce the behaviour of the reservoir are discarded:allowing a faster history matching process;increasing confidence in the process; and,adding uncertainty analysis to the production prediction. The methodology starts with a dynamic sensitivity analysis based on simulation of models where uncertain attributes are tested and compared with a base model. The attributes are then selected and combined in a derivative tree. Several models are evaluated and those that do not match the production and pressure history are discarded, reducing uncertainty in prediction of the behaviour of the reservoir. This methodology was motivated by a reservoir with unexpected behaviour that yielded a difficult history matching when considering the usual procedures. Techniques were developed to analyze reservoir performance and differential pressure maps between zones. New approaches to assess connectivity between zones were used to give alternate structural models to the sensitivity analysis. This methodology may be helpful during the first years of production when uncertainties are still significant and when typical procedures to perform production history matching are unable to solve the problem efficiently. Introduction History matching is one of the most important activities during the development and management of petroleum reservoirs. Matched models are necessary:to ensure reliable production forecasts; and,to increase the confidence in understanding the geological and reservoir models. Petroleum reservoirs are very complex structures and several uncertainties are present in the process promoting low confidence in reservoir simulation models. During the first years of production, uncertainties still play a major role in the process making it more complex and allowing multiple acceptable solutions(1). As production increases, the quantity of data permits better matching but the complexity of the process increases because of the high number of wells and objective functions of the optimization process, which usually are the difference between simulation results and observed data for pressure and each of the production phases. These different levels of uncertainty and procedures, as functions of reservoir development stages, add an extra complexity to the usual procedures(2, 3). Due to these difficulties, this paper presents a different approach. Instead of starting with a base (deterministic) model that is modified to reproduce production history, it is possible to start with an uncertainty analysis, consider several possible realizations of the reservoir model, and then to decrease the level of uncertainty in the production prediction by discarding models that do not satisfy specific conditions. There are several advantages in the proposed methodology: 1) the dynamic behaviour of the history matching is better handled with this approach since previous simulation runs can be used at any moment of the analysis;

Parallel Automatic History Matching Algorithm Using Reinforcement Learning

“…The history matching problem is an ill-posed problem where multiple solutions can be found to match historical data [7][8][9]. Due to such ill-posedness, multiple solutions are used to better assess the uncertainty in the forecasts [10,11]. The regularization term in the objective function in Equation ( 1) is used to mitigate the effect of ill-posedness of the history matching problem [12,13].…”

Assisted History Matching using combined Optimization Methods

“…According to Tavassoli et al (2004), the best production-matched model does not necessarily have a good fit for the parameters of the reservoir, and this can provide different values in the forecast period. Schiozer et al (2005) presents a procedure that integrate the history matching with uncertainty analysis, when several possible models are generated based on the probability value of each attribute that constitutes the model. Abrahem et al (2010) presented an assisted approach called "Target Pressure and Phase Method", where the computer automatically places pseudo wells in the static model to reproduce the measured data and concludes that this method is useful for practical applications.…”

Uncertainty Analyses of Basement Fracture Reservoir Performances Using Proxy Models with High-Quality History Matching

“…There have been lots of attempts to integrate uncertainty analysis with production data. Schiozer et al (2005) presented a methodology to use an uncertainty analysis for HM. Reis (2006) showed that Response Surface Modeling respecting production history improves the quality of risk analysis by constraining possible ranges of uncertain parameters.…”