Electricity Industry’s Capital Investment Decisions

Researchers: Paul Fischbeck, Engineering and Public Policy, Carnegie Mellon University

The U.S. economy depends heavily on electricity. Electric power is indispensable to our modern society, providing everything from basic lighting and heating, the internet and computers to commercial manufacturing and modern living comforts. The electric power industry accounts for TK billion of the economy.

Naturally, this large-scale industry includes some costly investments. Power providers face large capital investment decisions such as whether or not to shut down old coal plants, what new technologies to employ, how many new plants to build and where, and whether to use renewable energy sources. Besides involving massive expenses, all these are long-term and typically one-time decisions.

And, because these investment decisions have to comply with federal government regulations, they need to be optimal in the present in order to avoid unnecessary future costs for the providers and the society. For example, a power plant owner wants to avoid installing an expensive emission-control technology in an old plant that won’t meet future regulations or will make electricity more expensive than the rules require.

But it’s hard to predict what government regulations for the power sector are likely, especially if they consider climate change, the effects of which are uncertain to begin with. Uncertain regulations make investment decisions hard for providers. Moreover, uncertainties in energy demand, fuel costs, available technology options and world political pressure complicate these decisions.

We want to understand how and to what extent climate change affects power regulations. We are analyzing past regulations to determine what influenced them. We are trying to understand the regulatory process, and assessing the likelihood of possible regulation outcomes. In addition, we will conduct expert elicitations with power industry representatives and government policy makers

We also want to understand how uncertain regulations could affect power sector investment decisions. For this, we will most likely build on a model we have developed in the past [Fischbeck and Rode, 2004]. This reduced-form model replicates the cost production curve of a larger resource-planning model for the Florida power market. The model is imbedded in a Monte-Carlo simulation framework with an autonomous agent that makes decisions to build or retire plants in the future using various algorithms. This approach can easily run a million future scenarios and allows for a detailed plant level analysis over 20 year time horizons across different regulatory trajectories.

Electric Power: Long-term Investments for a Large-Scale Industry