Value Optimization in a World of Choices
Industrial companies face a multitude of choices when deciding what type and size of equipment to utilize when designing a new production facility. Facilities are often over-designed with respect to equipment capacity due to issues such as: allowances for demand uncertainty, use of default design safety factors, manufactured incremental step sizes in equipment and materials, and a lack of design definition. Total life cycle cost and NPV analysis are typically performed using deterministic models that do not adequately account for demand, cost (capital and O&M) and equipment life uncertainties. Capital budgeting decisions that consider such uncertainties usually do so by creating deterministic scenarios intended to represent best, most likely and worst cases. Given the number uncertainties and their potential impact, this is simply not adequate. Any quantitative model to used facilitate the design process must be able to identify the optimum decision in light of project uncertainties and limitations.
Triangle Economic Research (TER) has designed a value optimization model to facilitate the design process that takes project uncertainties and limitations into account by combining the use of Palisades @RISK and RISKOptimizer tools. We have built a template that allows for the input of ten pieces of equipment with ten different options for each (i.e., ten types of pumps, ten types of cooling systems). For each option, probability distributions based on expert judgment or statistical analysis of data are used to represent capital costs, installation costs, operation and maintenance costs and equipment life. The capacity of each equipment option and a distribution representing potential product demand are also inputs to the model. A key component of this process is to have a facilitated meeting with the design team to identify all these inputs and to identify which equipment options can not be paired with others (i.e., pump A can not function with cooling system C). The constraints are then added using the RISKOptimizer interface.
The model has the capability to optimize facility design by desired system capacity, minimizing total cost, maximizing average net present value (NPV) of annual revenue, and maximizing average NPV of annual profit. Discount and inflation rates are incorporated into calculations to convert future dollars into NPV. The ability to look at many types of optimization and the ability to incorporate uncertainty make this model useful in facilitating efficient design plans. This model can be applied to all types of industrial facilities including creating new and expanding existing refineries or in any type of manufacturing setting. Combining the expertise of a design team’s knowledge and the strengths of optimization techniques allows for better decision making in an industrial setting.