Sasol's coal-to-liquids plant varies over time, but standard average-based planning smoothed that away, so business cases rested on one number.
The value-chain optimization team built MOSS, a stochastic simulation that reflects production variability across time.
These distributions became the basis for business cases on gas and liquid facility changes, and for surfacing risk in strategic decisions.
Model the distribution, not the mean
Stochastic simulation reflects real variability that averages hide.
It supports business cases for big facility changes.
It highlights risk, not just a point estimate.
It was a finalist for the 2010 Franz Edelman Award.
When a plant's output swings, plan on the distribution; a single average number quietly misprices the riskiest and most valuable options.
MOSS provided an estimated value addition for Sasol of over $230 million since 2000, a figure considered conservative by the team. The models became an integral part of Sasol's strategic decision-support process and were seen as a step forward for OR in chemical engineering.
FOLLOW THE EVIDENCE
The sources
- Sasol's innovative use of simulation models declared Edelman finalist Sasol · sasol.com ·
- Introduction: 2010 Franz Edelman Award for Achievement in Operations Research and the Management Sciences INFORMS Journal on Applied Analytics · pubsonline.informs.org