Sandstone Acidizing Simulation: Development of an Expert System

by
Untung Sumotarto, Ph.D.

University of Texas at Austin, 1995
Supervisors: A. Daniel Hill
Kamy Sepehrnoori

Various models to simulate sandstone acidizing processes in oil well stimulations have been presented in the past. The standard "lumped" model representing the sandstone acidizing process that has been used until recently considers one acid and two representative minerals. Although this model has been employed for years, investigators have found that the model does not account for acid (HF) consumption consistently and cannot extrapolate laboratory data to different operating conditions. A secondary dissolution/ precipitation reaction is proposed to simulate the acidizing process more accurately. This leads to an improved model which represents the sandstone acidizing with two acid and three mineral equations. The new model has been implemented in a sandstone acidizing simulator called UTACID.

The first part of the research investigates the two-acid, three-mineral model and compare both models. The study provides results which confirm the findings by previous investigators. The simulator requires input files containing data about the formation properties, acidizing fluid system, and simulator parameters. One of the difficult parts in creating these input files is in the design of matrix acidizing fluids particularly the concentration and volume of the acid. The selection of acid fluids is a design process, an appropriate task for expert systems. This leads to another task for a complete acidizing simulation study.

An expert system has been developed to help design sandstone acidizing fluid systems and create input files for the simulator. Required parameters for sandstone acidizing such as acid type, concentration, volume, and injection rate/pressure are first designed using the expert system. The output from the expert system is then used to create the input data for the UTACID simulator. Given proper reservoir data including information about the damage, the expert system presents recommendations including the damage removal fluids, the acid systems if required, the injection pressure and schedule, and the input file for the simulator. Integration of the expert system, the simulator, and additional routines including the X-Window visualization programs leads to the construction of ASSESS, a complete and user-friendly package for sandstone acidizing fluid design and simulation. Testing on this package shows satisfactory results.

Back to theses index