Terminal Design Optimization

Abstract

This study analyzes the financial impact of design guidelines used in the development of agribulk terminals and sets out to quantify the effect of uncertain throughput projections on a terminal’s business case. This was done through the use of a purposely-built computer model that translates traffic projections into a terminal design in an automated fashion. The allowable waiting time for vessels calling at the terminal has proven to be the design guideline that has the highest influence regarding project feasibility. The allowable waiting time was optimized by monetizing the negative consequences of long waiting times at the terminal using demurrage rates. Existing guidelines state that bulk terminal investments are typically based on maintaining an average ratio of waiting time to service time between 20% to 30%. However, results from the computer model show that, if a terminal were to incorporate the notion of demurrage rates into its business model, waiting times could be allowed to rise to 70% of berth service time, unlocking significant project value. Furthermore, by automating the translation of traffic projections into terminal designs, the financial repercussions of the uncertainty surrounding such projections becomes quantifiable. By simulating the project value of a terminal under varying economic scenarios, the user is shown how a spread is possible traffic development results in a certain spread in project value. The final aspect that has been analyzed within this study is the financial impact of modelling a terminal based on perfect foresight, a presumption that the traffic projections on which investments are based will always come true. Results show that modelling a terminal is such a manor overestimates a project’s value. Depending on the size of the project and the market’s volatility, this over-estimation is in the range of 3% (steady market growth) to 20% (volatile markets). An alternative modelling method has been developed which bases investment decisions on forecasted traffic volumes, which are updated as a simulation progresses. This forecast method slightly lowers over-estimation of project value by several percentage points.