A New Gas Hydrate Prediction Model for Acidic and Inhibitor-Containing Systems

dissociation model suitable for the formation of gas hydrate in acidic and alkali-containing systems is established. By introducing inertia weight factor, an improved sinusoidal and cosine algorithm is obtained to optimize BP neural network and build ISCABP gas hydrate prediction model. The ISCABP model and other thermodynamic models were used to predict the hydrate formation conditions of four groups of (CH 4 +CO 2 +H2S) gas mixtures in acidic system, and the hydrate formation conditions of six groups of gas + liquid phase in alkoxide system. The temperature range is 240.15-308.45 K, and the pressure range is 0.05-82.56 mPa. The concentrations of H 2 S and CO 2 in the mixture were 4.95-27.93% (mol) and 6.79-8.4% (mol), respectively. The results show that ISCABP hydrate model has the best predictive value. The model has strong applicability to gas field production, and can provide a theoretical basis for determining the injection amount of inhibitor and formulating the field safe operation strategy.