What Playing Chess has to do with Designing Capital Projects

By Scott Whitbread 

After both players have made their second move in a chess game there are over 72 thousand possible positions the game could be in. After 3 moves there are 9 million. After 4 moves, 318 million and so on. The number of potential positions added with each move rises so quickly that even chess computers cannot play using brute force evaluation, that is, to evaluate every potential outcome and in advance and choose the optimal one (and as such, humans can still beat them!)

This phenomenon, referred to as ‘combinatorial explosion’, also takes place in the conceptual design of capital projects. For example, consider the conceptual development of a project where there are 10 design choices to make, each with 5 potential options. These choices could be the design production capacity, the cutoff grade/recovery, the processing technology choice and so on in the case of a resource extraction project. In this case there will be almost 10 million potential permutations to evaluate. Evaluating one per minute, around the clock, would take over 18 years. Or as a client recently characterized it: “It’ll go exponential on ya.”

Facing the prospect of combinatorial explosion, project teams typically employ some degree of opinion-based filtering to reach decisions. They whittle down the universe of potential options through initial opportunity-framing sessions, where buckets of options come off the table without formal evaluation. The remaining potentials are then studied at length, though it can prove difficult to reach a single option without relying on further opinion to account for seemingly irreducible, qualitative considerations.

This approach’s shortcoming is that it often takes a substantial amount of time and organizational churn to align the project’s stakeholder group on a single best option to move forward with. (Recycling design decisions in the next phase of project development is also common.) Stakeholders are looking for a cogent, fact-based narrative for how each conceptual design choice was made in order to arrive at the highest value option. Their curiosity is extremely difficult to satisfy by an approach that’s infused with opinions. They ask: “What about ‘X’?” or “Have you considered ‘Y’?” or “I think option ‘Z’ needs further study” and so on. And yes, they bring their own opinions with them too.

The other risk is that many, potentially higher value, options are not being considered at all and are therefore lost from the exercise. As you would imagine, peer-reviewed, expert opinions are usually correct, though when they are not, tremendous value can be hidden by them. In one case, a $500MM water treatment plant was proposed as the optimal solution to a predicted cycling up of process water salinity in a mining project. However when studied further, a near-zero cost alternative was discovered by challenging how freshwater was being sourced and distributed across the site.

In chess the outcome is binary and the consequences are low (unless you’re a professional or a big chess fan). But in capital projects there is a continuum of financial outcomes and the stakes are very high. Billions of dollars may be gained, or lost, over the life of a project and some much needed projects may be terminated prematurely as higher value options remain hidden.

So what can a project leader do about it? Here are 3 suggestions:

1. Fiercely limit the design decisions that are included in the conceptual stage: Acknowledge the exponential nature of conceptual engineering and limit the decisions that need to be included. Some can wait as there will be insufficient data at the conceptual stage to make them well anyway, and assumptions as to their ultimate influence will suffice. Some design choices may have to be addressed now but could be made independently of the rest, i.e. the same conclusion will be reached regardless of the other aspects of the conceptual design. Peel these off from the alternative generation process.

2. Select a single, quantitative ‘yardstick’ to evaluate the outcome of any choice or tradeoff study: This means selecting a single measure of project value to take primacy in all decision making. Broadly speaking, there are 3 types:

  • NPV-type measures that direct you to build the largest project that’s profitable - appropriate if there is only one project to fund and lots of cash
  • IRR-type measures that direct you to build what will create the highest quality return, in percentage terms, regardless of the project size - appropriate if there are many projects and only some can be funded
  • Cash flow-type measures that guide you to build whatever has the briefest, shallowest J-curve of investment while still making the most of the opportunity - appropriate when cash is limited but the project opportunity is still compelling in NPV and/or IRR terms

Some will resist this notion as a single number can never fully encapsulate the owner’s nuanced philosophy and business situation. For example, you’d never throw IRR to the wind in favor of generating an outsized, but highly risky, NPV. And you don’t have to. Set minimum thresholds for these other relevant metrics but allow a single yardstick to guide decision making in order to speed up the process, allow many more fact-based conclusions to be reached and reduce the reliance on opinion.

3. Determine trigger points for reconsidering conceptual design decisions: Acknowledge that the decision inputs may change in the future (material prices, technology innovations, survey accuracy, etc.) and determine trigger points for when such changes would become impactful enough to alter the optimal concept. Taking this step (one step beyond most sensitivity analyses) can generate tremendous insight that leverages your subject matter experts. If, for example, the price of certain building material would need to triple before the front-running concept alternative would be overtaken, the costing experts can assess the likelihood of this happening. Knowing the trigger points for when an alternative would go out of favor provides guidance into the future and creates confidence to move the project forward now.

While it may not be feasible for some complex projects to consider every conceptual permutation mathematically, or to completely eliminate the need for opinions in decision making, the reliance on opinion can be greatly reduced and the project outcomes will be better for it.


Like this article? You might like: 

Why Projects Overrunning Budget Will Rarely Be Canceled

Winning at Capital Projects

Mine Plan Radar: A Simple Tool to Visualize Your Mine Plan Alternatives