GA-ANNPrediction Model for Establishment of Langmuir Isotherm and Characteristics study of Langmuir Isotherm Constants

Abstract

Adsorption of gases on solid surface is very common phenomenon and naturally occurs in nature. Langmuir Isotherm is commonly used to describe the adsorption of gas on solid surface due to its very similar pattern depiction as derived from experimental data. A profound understanding of Langmuir Isotherm establishment and its dependency on different influencing parameters is very essential for prediction of sorption process of gas on solid surface. Langmuir Isotherm establishment depends on Langmuir constants, so very precise evaluation of Langmuir constants is essential for successful accurate establishment of Langmuir Isotherm.

In this paper dependency of Langmuir constants on coal quality parameters (moisture, volatile matter, fixed carbon, ash content and vitrinite reflectance) and temperature have been studied to predict Langmuir constants for establishment of Langmuir Isotherm for coal in a specified geological condition without going into tedious and time consuming experimental experiments. Genetic Algorithm (GA) and Artificial Neural Network (ANN) in combination have been used as a mathematical modelling tool for prediction Langmuir constants for sorption study. This study reveal that Langmuir Volume Constant (VL) depends on properties of sorbent only while Langmuir Pressure constant (PL) depends on properties of sorbent and physical condition of sorption system. As the coal quality parameters changes for every point, so this prediction model would be very useful in insitu gas estimation by real time modelling of coal reservoir.