Modeling and identification of economic disturbances in the planning of the Petrochemical Industry

The petrochemical industry is a dynamic industry and can be seen as a network of chemicals from basic feedstock to final chemicals. The aim of this work is to identify and model long- and short-range disturbances that affect planning of the petrochemical industry. An application to the Kuwait Petrochemical Industry was performed. The major disturbance is the oil prices that affect chemical prices and consequently profit. Future chemical prices needed for planning are predicted using three forecasting models: simple time-series fitting and two causal models with oil prices, the second-order plus dead time transfer function and autoregression with an exogenous variable models. Oil prices for the causal models are first forecasted under the concept of market long cycles (K-waves) and short cycles (business or Kitchin cycles) and then used to forecast chemical prices. The forecasted chemical prices affect the planning of the petrochemical industry where different routes in the network are selected for different final product prices. It is found that including the market cycles and using the causal models for forecasting petrochemical product prices will provide possible scenarios for chemical price forecast, and then a risk-adjusted present value can be calculated.