Use of the equivalent attribute technique in multi-criteria planning of local energy systems

https://doi.org/10.1016/j.ejor.2007.12.050Get rights and content

Abstract

This paper discusses how the equivalent attribute technique (EAT) can be used to improve the comprehensibility of a multi-attribute utility theory study. When using EAT, ‘vague’ expected total utility values are converted into equivalent values for one of the attributes being considered, often an economic attribute. Two models are considered: a simplified linear model, and a more advanced non-linear model that includes the DM’s strength-of-preference and risk attitude. EAT is particularly useful in distinguishing between alternatives with similar utility values. When the difference between utility values is larger, the choice among the alternatives should be clear, and EAT therefore becomes less useful. The technique can still be used, although extra care is needed when choosing the equivalent attribute. A local energy-planning problem is used as a case study to illustrate and exemplify the EAT approach.

Introduction

One of the best-known multi-criteria decision analysis (MCDA) methods is the multi-attribute utility theory (MAUT). In MAUT, the expected total utility is determined for each of the alternatives under consideration. The expected total utility is calculated using a multi-attribute utility function, which is derived from interviews with the decision-maker (DM). For many DMs, the concept of expected utility values might be somewhat vague. To really understand the concept of utility values, it is necessary to fully understand the MAUT methodology. That can in some cases be difficult for DMs, as they often have limited time available for the decision process. An alternative – and possibly better – approach is to introduce the DMs to the concept of equivalent attributes, which is much easier to understand. The idea of the equivalent attribute technique (EAT) is to find a method to convert a change in the expected total utility into an equivalent quantity in one of the decision attributes. Most often, an economic attribute is used as the equivalent attribute. However, other attributes can be used if desired.

This paper is organized as follows. First, we provide a short presentation of MAUT, and discuss the EAT in more detail, including a comparison to the cost–benefit analysis. Then, to illustrate the use of the technique, we apply EAT to a local energy-planning problem that is characterized by multiple energy sources and carriers. We discuss the results from the case study and offer conclusions.

Section snippets

The multi-attribute utility theory (MAUT)

MAUT was first described in detail by Keeney and Raiffa (1976). The method has often been used for energy-planning purposes, e.g. by Buehring et al., 1978, Pan et al., 2000, Schulz and Stehfest, 1984. MAUT is suitable for incorporating risk preferences and uncertainty into multi-criteria decision problems in a consistent manner. In MAUT, a multi-attribute utility function U describes the preferences of the DM. The multi-attribute utility function measures preferences along several dimensions.

Motivation

Utility values are constructed to convert performance values to preference values. This simplifies the analysis of complex decision problems. Although expected utilities are convenient for ranking and evaluation of alternatives, they are only “instrumental for the purpose of comparing alternatives” (Matos, 2007). Accordingly, they do not have direct physical meaning (Keeney, 1980), and are of no interest outside of the specific decision problem (Matos, 2007). Expected utility values may

A local energy-planning problem and the use of MAUT

The EAT has been tested in a pilot case study based on Helseth (2003). In the case study, we used data from an existing planning problem in Norway to analyze the future energy-supply infrastructure for a suburb with approximately 2000 households and possible additional industrial demand. We carried out preference-elicitation interviews with six people with a background in energy research and industry. The participants were asked to imagine themselves as the manager of the energy company that is

EAT applied to the case study

In this section, we show how EAT can be applied to the local energy-planning problem described above. As discussed in Section 3.2, cost attributes are suitable for use as the equivalent attribute, provided that the DMs consider cost to be among the most important attributes. Table 1 shows that DMs A and C both gave highest criteria weight to the annual operating cost (OC). Accordingly, it seems reasonable to use OC as the equivalent attribute in this case study. Below, we will present two EAT

Conclusions

This paper has discussed how EAT can be used to simplify the comparison of results from an MAUT analysis of a local energy-planning problem. The technique is useful in distinguishing between alternatives with similar utility values in order to check if the difference is significant. Instead of comparing utility values directly, DMs can compare more familiar cost data by using EAT. For example, it appears to be much easier for a DM to compare alternatives after being told that the difference

Acknowledgements

The authors would like to acknowledge Bill Buehring at Argonne National Laboratory for all his help and insightful comments. The contents of this paper are based on some very constructive discussions with him. We would also like to thank Ron Whitfield, also at Argonne, for his help in general, and specifically for providing the idea for Fig. 2, Fig. 3, Fig. 4.

We would also like to acknowledge the anonymous reviewers of the earlier versions of this paper for their helpful comments; Bjørn Bakken,

References (18)

There are more references available in the full text version of this article.

Cited by (35)

  • An extended methodology for multi-criteria decision-making process focused on enhanced geothermal systems

    2022, Energy Conversion and Management
    Citation Excerpt :

    Moreover, capital, i.e. investment costs are related to the first four phases and according to [133] they mainly consist of costs for reservoir exploration, well drilling and completion, reservoir engineering measures, installation of the geothermal fluid loop, and construction of the plant on the surface for power and/or heat provision alongside with the grid connection costs. Capital costs, i.e., capital expenditures (CAPEX), are usually measured and represented in euros per installed kilo-watt (€/kW) [134–142]. Since the market for geothermal power is relatively thin, the year-on-year variations in installed costs are quite significant.

View all citing articles on Scopus
View full text