Case Study
Flexible Design for Liquified Natural Gas Infrastructure
Michel-Alexandre Cardin, Mehdi Ranjbar-Bourani, Richard de Neufville
This case study demonstrates how modular, flexible design strategies improve lifecycle performance for on-shore LNG production under uncertainty. Using real options and simulation, the authors show that phasing capacity over time and geography increases expected value, reduces downside risk, and supports better decision-making than traditional fixed, large-scale designs.
image by Alexandre Patchine @ Adobe Stock
Read More
Motivation
LNG production is capital-intensive and highly exposed to demand and price uncertainty. Conventional centralized designs risk either overcapacity or missed opportunities. The study was motivated by the need to create a credible, transparent framework to convince decision-makers that flexibility in modular deployment can materially improve project value and robustness when future markets are uncertain.
Methodologies
- Four-Step Flexibility Analysis Framework: Deterministic valuation → Uncertainty modelling → Flexibility analysis → Sensitivity analysis.
- Real Options Analysis (ROA): Quantified the value of deferring or expanding LNG capacity.
- Monte Carlo Simulation: Modelled thousands of stochastic demand paths (Figure 4 p. 9).
- Decision-Rule Enumeration: Explored threshold-based triggers for modular expansion (Table III p. 11).
- Sensitivity & Scenario Analysis: Evaluated trade-offs between economies of scale, learning rates, and geographic dispersion.
Insights
- Higher Value: Flexible modular design increased expected NPV by 45–63 % vs fixed designs (ENPV = $20.7 M → $23.3 M).
- Downside Protection: Losses reduced from –$25 M to –$5 M while upside increased to ≈ $60 M.
- Learning Effect: 10–20 % learning rates amplified flexibility value (Figure 7 p. 12).
- Geographic Flexibility: Deploying capacity closer to demand improved ENPV and reduced logistics costs.
- Robustness: The flexible design remained optimal under wide variations in scale and demand parameters (Figure 8 p. 13).
Training
Relevant lectures and skills:
- Real Options Analysis
- Monte Carlo Simulation
- Scenario & Sensitivity Analysis
- Dynamic Programming
- Decision Rule Design
