Two mathematical models that coupled with a simple Genetic Algorithm, (GA) which is one of the most common artificial intelligence techniques for optimization were developed to simulate both formation damage by geochemical and clay swelling processes and formation improvement by acidizing in petroleum reservoirs. First model predicts the permeability alteration of an acidizing process with an analytical solution requries two parameters representing the acidizing process, Damkohler and acid capacity numbers. These parameters were optimized with a simple GA by matching calculated and experimental data. Second model simulates formation damage due to deposition, entrainment, dissolution and clay swelling mathematically using basic balance equations. A simple GA was used to optimize three parameters, deposition, dissolution and entrainment rate constants, that represent the damage, by matching calculated and experimental data. Both models were verified with case studies and porosity and permeability changes could be predicted successfully.


Permeability reduction in petroleum reservoirs have received a great deal of concern by the oil and gas industry. This problem is termed as "formation damage". It can occur during almost any stage of petroleum exploration and production operations. Fluids introduced into petroleumbearing formations frequently cause formation damage due to the incompatability of the injected and indigenous fluids, and the mineral constituents of the formation1. It's wellknown fact that when permeability of a formation is reduced, the production decreases. So, a cleaning process is needed. There are various formation improvement processes in petroleum industry. Acidizing, a well known well stimulation process for mud and particulate damage, is generally initiated when an improvement in permeability of the damaged zone is desired.

Attempting to understand formation damage or formation improvement is becoming an important task for reservoir engineers, oilfield chemists and the decision makers in the business, because it is the first step to be taken to prevent and further alleviate this problem. It is the major objective of this study to develop mathematical models for the permeability alteration due to various physical and chemical interactions between reservoir rocks and fluids.

The productivity of oil and gas sandstone reservoirs may be increased by injecting mud acid, hydrochloric / hydrofluoric acid mixtures (HCl / HF), into the matrix in the near wellbore to dissolve portions of the rock minerals, thereby increasing the formation permeability and well productivity. The complexity of porous media and the reactions during acidizing process have made it difficult to predict a priori the exact results of the simulation. The determination of unknown parameters for formation damage (dissolution, deposition and entrainment rate constants) and formation improvement (Damkohler and acid capacity numbers) on the basis of measured data is, in fact, an optimization problem, the objective of which is to reduce the difference between experimental and calculated data.

The use of artificial intelligence has recently become more common in simulating various aspects of petroleum exploration and exploitation. The most popular artificial intelligence technologies include expert systems, neural networks, fuzzy logic systems and genetic algorithms2. At present, expert systems are the most mature, and the genetic algorithms the least.

This content is only available via PDF.
You can access this article if you purchase or spend a download.