Decision Making in Energy Operations

Based on research by Nicola Secomandi

Key findings:

• Research into energy operations looks at different methods for making the best decisions at each stage of energy value chains.
• While crude oil and natural gas have had a head start in the operations literature, the amount of research into electricity and clean energy is catching up.
• Future research may benefit from both analyzing and improving current decision making processes.

Real options represent the decisions companies can make in the face of evolving risk. In the energy and commodity industries, real options are ubiquitous, including the extraction, processing and refinement, storage, and transportation of natural resources. These choices are influenced by ever-changing market and environmental conditions. 

Because of these uncertainties, the energy industry has been a key focus of the operations literature on real options. 

Rice Business professor Nicola Secomandi, along with University of Illinois at Chicago College of Business Administration professor Selvaprabu Nadajarah, were recently invited by the European Journal of Operational Research to conduct a review of the operations literature on real options in energy. Their review included 80 papers across 10 journals active in the field. The research was mostly conducted during Secomandi’s time at the Tepper School of Business at Carnegie Mellon University.

The review examined how often different types of energy and methods of studying related business processes appeared in the operations literature. Nearly a quarter of the papers considered natural gas, more often than any other energy type. Natural gas storage was the most studied process, while the transport and sale of natural gas were less discussed.

While only 10 percent of the papers focused on electricity by itself, mostly in the context of battery management, electricity was discussed alongside emissions and the environment in 22.5 percent of the papers—almost as often as natural gas. About 11 percent of the papers examined both electricity and natural gas. 

Roughly 21 percent of the papers focused on crude oil and refined products. Exploration, development, and abandonment of crude oil fields were common topics, while work on crude oil refining and gasoline logistics was rarer.

The review looked at the frequency of use of five categories: real option types, valuation methodologies, model formulations, price risk dynamics, and optimization schemes. Timing options, which irreversibly change the status of an asset when exercised, and switching options, which involve reversible changes, appeared with about equal frequency. Of the valuation methodologies, risk neutral valuation was employed the most often, appearing in nearly 78 percent of the papers. 

Model formulations were divided mainly between Markov decision processes, which assume that decisions are made at set times, and stochastic optimal control models, which assume that decisions are made continuously. About 63 percent of the papers discussed Markov decision processes, but at nearly 34 percent, stochastic optimal control models haven’t been entirely left out of the literature. Almost 75 percent of the papers formulated models based on spot prices as opposed to futures prices, and over 80 percent adopted normal distribution models, either alone or in combination with other models. Approximate solution approaches dominated within optimization schemes, appearing in 71 percent of the papers. 

While several energy sources and analysis tools have been discussed in the literature, the possibilities for future research remain broad. The transition to clean energy sources may increase the complexity of already intricate operations. Its modeling and analysis may require more advanced models than existing ones.  

*This research was made possible by a National Science Foundation grant and has been published as an open access article.

Nadajarah and Secomandi (2022). “A Review of the Operations Literature on Real Options in Energy,” European Journal of Operational Research.