A least-squares support vector machine approach to predict temperature drop accompanying a given pressure drop for the natural gas production and processing systems

Precise estimation of temperature variations throughout gas production systems can enhance designing the production amenities. Routine methods for determining the temperature profiles in gas production systems are based on the gas composition and flash calculations. However, if the gas compositions are not available, the gas production system can be modelled by employing a black-oil approach, which is also a method for calculating the oil/gas resources and for modelling the gas reservoir operation. Accordingly, for black-oil models and when the natural gas compositions are not accessible, applying robust predictive tools in this research is of high interest in natural production systems. The current study places emphasis on applying the predictive model based on the least- squares support vector machine (LSSVM) to estimate precisely the proper temperature drop associated with a given pressure drop throughout the natural gas production systems based on the black-oil approach to acquire an accurate result for the temperature drop of natural gas streams. Genetic algorithm was used to optimise hyper-parameters (γ and σ²) which are embedded in the LSSVM model. Using this method is simple and it accurately determines the temperature drop through the natural gas stream with minimum uncertainty.