Methods and Systems For Machine - Learning Based Simulation of Flow

A method for producing a hydrocarbon from a hydrocarbon reservoir, comprising:

• generating a reservoir model comprising a plurality of sub regions;
• obtaining a solution surrogate for at least on sub region by:
• searching a database of existing solution surrogates to obtain an approximate solution surrogate based on a comparison of physical, geometrical, or numerical parameters of the sub region with physical, geometrical, or numerical parameters associated with the existing surrogate solutions in the database;
• and
• if an approximate solution surrogate does not exist in the database, simulating the sub region using a training simulation to obtain a set of training parameters comprising state variables and boundary conditions of the sub region, using a machine learning algorithm to obtain a new solution surrogate based on the set of training parameters, and storing the new solution surrogate to the database;
• simulating the hydrocarbon reservoir using the solution surrogate obtained for the at least one sub region;
• and
• producing a hydrocarbon from the hydrocarbon reservoir based, at least in part, upon the results of the simulation.

A system for modelling reservoir properties, comprising:

• a processor;
• a storage medium comprising:
• a first solution surrogate that represents a flow response of a first sub region and a first set of physical, geometrical, or numerical parameters corresponding to the first sub region;
• and
• a reservoir model comprising a second sub region and second set of physical, geometrical, or numerical parameters corresponding to the second sub region;
• a non-transitory machine readable medium comprising code configured to direct the processor to:
• compare the first set of physical, geometrical, or numerical parameters and second set of physical, geometrical, or numerical parameters to determine whether the second sub region can be modelled by the first solution surrogate;
• and
• if the second sub region cannot be modelled by the first solution surrogate:
• simulate the second sub region using a training simulation to obtain a set of training parameters comprising state variables and boundary conditions of the second sub region;
• use a machine learning algorithm to obtain a second solution surrogate based on the set of training parameters;
• and
• store the second solution surrogate to the storage medium.

A non-transitory, computer readable medium comprising code configured to direct a processor to:

• create a simulation model of a hydrocarbon reservoir, wherein the simulation model comprises a plurality of sub regions;
• search a database of solution surrogates to identify a solution surrogate for each of the plurality of sub regions;
• compute a new solution surrogate for each sub region that does not correspond with a solution surrogate from the database, wherein computing the new solution surrogate comprises simulating the sub region using a training simulation to obtain a set of training parameters comprising state variables and boundary conditions of the sub region, and using a machine learning algorithm to obtain a new solution surrogate based on the set of training parameters;
• and
• store the new solution surrogate to the database.