Automatic Reservoir History-Matching Method Based On EnKF And ES-MDA

As an important method of tertiary oil recovery,polymer flooding has been widely used in oil fields in recent years.In the mathematical model,the polymer displacement mechanism is described by one or more sets of viscosity-concentration curves.These viscosity-concentration curves are derived from measurements in oilfield laboratories and have considerable accuracy under the current experimental technology.However,due to the difference between the ground experimental environment and the underground environment,as well as interferences such as pipe loss in field production,there are uncertainties occurring in the viscosity and concentration parameters of the real polymer displacement,which lead to the mismatch between the numerical simulation and the observation data,so it is necessary to use history matching techniques for the chemical parameters of polymers.The main content of this paper is:firstly give the mathematical descrip-tion of polymer solution viscosity,rheological characteristics,permeability drop coefficient,residual resistance coefficient,inaccessible pore space and polymer adsorption,and construct a synthetic field based on this.Then use the "twin experiment" to verify the data assimilation effect of the two algorithms of EnKF and ES-MDA on the synthetic field,analyze the behavior and problems of the two algorithms in order to determine the data assimilation strategy that should be used for the automatic history matching problem of polymer flood-ing.Then,according to the selected algorithm and strategy,history matching is performed on the real reservoir to verify the availability and application value of the algorithm for practical engineering problems.This article consists of the following chapters:Chapter 1:Introduces the research background of automatic history match-ing technology,state of the art of chemical agent displacement mechanism,mathematical model and reservoir history matching algorithm,and constructs the problem model and the historical simulation of assimilation chemical prop-erties.And the significance of research is illustrated.Chapter 2:The forward model of reservoir numerical simulation including polymer solution viscosity,etc.is given.After that,the EnKF and ES-MDA algorithms are explained separately and the data assimilation experiments are carried out for the polymer flooding synthetic field using the "twin experi-ment".The performance of the two algorithms under different sizes of en-sembles is analyzed and the illposedness is mentioned.Then the assimilation algorithm and strategy for real oil reservoir problems are preliminarily deter-mined.Chapter 3:The algorithm is applied to the real oil field,and the fitting effect of each production well is given and analyzed.Chapter 4:Summarize the content of the article and put forward the out-look.