Runtime parameter selection in simulations

A system for performing a field operation of a field, comprising:

• an exploration and production (E&P) computer system, comprising:
• a computer processor;
• memory storing instructions executed by the computer processor, wherein the instructions comprise functionality to:
• obtain a plurality of historical parameter values of a runtime parameter and a first plurality of historical core datasets, wherein the plurality of historical parameter values and the first plurality of historical core datasets are used for a first simulation, and wherein each of the first plurality of runtime parameter values results in a simulation convergence during the first simulation;
• generate a machine learning model based at least on the first plurality of historical core datasets and the plurality of historical parameter values;
• execute a second simulation of the field to obtain a current core dataset;
• determine, using a classifier of the machine learning model, a closest matched core dataset of the first plurality of historical core datasets with the current core dataset at a simulation time point;
• select a predicted parameter value of the plurality of historical parameter values of the runtime parameter corresponding to the closest matched core dataset of the first plurality of historical core datasets, wherein
• the predicted parameter value comprises a time step for achieving a first simulation convergence from the simulation time point of the second simulation;
• generate first output data during a first iteration associated with a first simulation time point of the second simulation;
• increase, based on the time step, the first simulation time point of the second simulation to a second simulation time point of the second simulation;
• iterate a runtime parameter value of the runtime parameter to achieve the first simulation convergence to generate second output data;
• generate a modeling result of the field using the first output data and the second output data;
• and
• execute one or more of:
• generating a multi-dimensional visualization of the modeling result on a display device, such that the multi-dimensional visualization facilitates predicting hydrocarbon content throughout portions of the field, and
• transmitting a field control signal based on the modeling result to control at least one of a drilling equipment, an actuator, or a fluid valve disposed about the field;
• and
• a repository for storing the machine learning model, the first plurality of historical core datasets, the plurality of historical parameter values, and the predicted parameter value.