Oil- and Coal-Based Sea Transportation Needs: An Integrated Forecasting Approach

Abstract

Due to the global economic crisis and increasing environmental issues (agreements to prevent CO₂ emissions), the overall transportation logistics sector is currently suffering from one of the most severe recessions (as it is a derived demand from economic activities). This is particularly the case in the airline industry and spot market operating sea transportation. In the latter case, oil and coal have very significant shares in the total transportation volume of bulk (liquid and dry). Although the future does not currently look promising for oil and coal transportation throughout the world, this chapter shows that even a further declining economic situation, as well as the leveling-off of production of these commodities could lead to a slight increase in demand for transportation services in the next two decades. This phenomenon is mostly caused by the fact that production and consumption are diverted (coal modestly, but oil greatly), and the import share from production has been on a long-term rise.

Appendix A

4.1.1 Technical Details of Linking Regression Models and Stochastic Elements on System Dynamics Simulation Models

As oil production was having longer time series for power model, we needed to add 45 years into demand forecast equation as shown in (Fig. 4.10). In oil import, share from oil production addition was 30. For the used simulation program, Vensim, it was rather easy to copy and paste demand forecasting equations from MS Excel, and connecting spreadsheet time series analysis program with system dynamics simulation was free from any problem.

Fig. 4.10

Incorporating forecasting equation into system dynamics simulation model (Vensim system dynamics simulation software)

As different factors face uncertainty in the forthcoming future, we incorporated these into the simulation model through statistical distributions (Fig. 4.11). Average transportation haul was altered in oil’s case based on the Stopford’s (2009, p. 147) time series with uniform distribution from 4,500 up to 7,000 nautical miles. Conventional variation in demand (random) was incorporated through random normal distribution having standard deviation of 0.05, and minimum value of 0.9 and maximum 1.1. This should correspond the movement of short-term cycles in oil demand during the forthcoming two decades. As currently is evident that global economy will face longer term sluggish growth period (e.g., Hilmola 2007), we have added long cycle variable into simulation model, having uniform distribution from 0.8 to 1.0.

Fig. 4.11

Uncertainties of the simulation model (Vensim system dynamics simulation software)

Given uncertainties are multiplied in the model, e.g., as shown in Fig. 4.10 (uncertainties are given in this sheet of Vensim). For the purposes of first 10 year decline of global economy, we have added parameter containing “if then else” structure in variable of “economic development” (during first period we expect economic development to be more sluggish). To get reliable simulation results, we simulated given uncertainty factors 1,000 times with different seed values.