Two case studies on benefit/cost ratios for reliability improvement strategies

Richard E. Brown and LaVelle A.A. Freeman, ABB

May 13, 2003 — In today’s competitive market environment, electricity customers are demanding higher levels of reliability while regulators are freezing or reducing rates.

This situation is chilling when one considers that more than half of utility spending is related to reliability, and that reliability is the most critical determinant of customer satisfaction. With this much at stake, it is imperative that electric utilities approach reliability issues with analytical rigor.

Fortunately, there are many commercial packages available today to assess the reliability of a system and to quantify the impact of various reliability improvement strategies.

However, any practical comparison of the effectiveness of various mitigation options must be done on a cost/benefit basis; the benefit from the project (e.g. customer minutes saved) must be balanced against the economic cost of implementation. This paper discusses the most important issues related to a cost/benefit comparison of reliability alternatives and illustrates these concepts on two actual utility systems.

In this section, the benefit/cost ratios for a variety of reliability improvement strategies are computed for two actual utility distribution systems. These systems are presented and compared below to emphasize the fact that benefit/cost ratios are system specific, and may vary widely even for systems that are somewhat similar. The physical characteristics of the two systems are briefly described below.

System A: This system is located in the Southern United States and consists of 6 substations and 25 feeders. It serves a mix of suburban residential and commercial load and is exposed to a mix of heavy, moderate, and lightly treed areas. In total, this system contains about 280 circuit miles of primary feeder (88% overhead), serves nearly 25,000 customers, and has a peak demand of more than 270 MVA. Base reliability indices for this system are SAIDI = 118 min/yr and SAIFI = 1.4 /yr.

System B: This system is located in the Midwestern United States and consists of 2 substations and 6 feeders. It serves a relatively homogenous area of suburban residential loads in heavily treed neighborhoods. In total, this system contains about 115 circuit miles of primary feeder (77% overhead), serves more than 11,000 customers, and has a peak demand of nearly 38 MVA. Base reliability indices for this system are SAIDI = 289 min/yr and SAIFI = 1.7/yr.

Each of these systems was modeled in ABB’s proprietary software package Power Delivery OptimizerTM, which uses an analytical simulation algorithm to compute customer reliability characteristics. This software also has advanced analysis features (e.g., sensitivity analysis, root cause analysis) that were used to help identify the most cost-effective reliability improvement options.

The analysis of the two systems will use two benefit/cost ratios: reduced customer interruptions per dollar (.CI/$) and reduced customer interruption minutes per dollar (.CIM/$). These correspond to the cost-effectiveness of improving SAIFI and SAIDI, respectively, but do not suffer from the problem of comparing systems of differing sizes (e.g., the cost to improve SAIDI for an entire system is much more than the cost to improve SAIDI for a single feeder).

Please note that even though this article has “cost/benefit” in the title, projects will be ranked by the corresponding reciprocal, “benefit/cost.” Although these two ratios contain identical information, using benefit-to-cost means that higher numbers are more desirable and this has, in the authors’ experience, been more intuitive for engineers to use. In addition, assuming linearity, benefit-to-cost can be multiplied by an allocated budget to obtain “expected” improvements in reliability for a particular budget.

Benefit-to-cost calculations for a variety of reliability improvement strategies are shown in Table 1 (top of article) for System A and in Table 2 (below) for System B. Each of these strategies has been carefully chosen so that it is near optimal within its improvement category. The same prices are used to determine project costs for each system.

The results above illustrate the point that the economic feasibility of reliability mitigation options is extremely system specific. For example, consider options A7 and B4. Although the tree trimming costs are the same on both systems, reducing the tree trimming cycle on 89 miles of System B is almost seven times as effective in mitigating SAIFI as reducing the tree trimming cycle on 27 miles of System A.

Even with adjustments being made for the difference in exposure, it is still more cost-effective to implement a tree-trimming program on System B. The difference can be wholly attributed to the physical differences between the two systems. Although System A has a higher percentage of overhead lines than System B (88% to 77%), System B serves mostly suburban residential loads in heavily treed neighborhoods while System A is exposed to a mix of heavy, moderate, and lightly treed areas.

Next, consider options A9 and B5. The cost to automate is the same, but automating six switches on System B is almost seven times more effective in reducing SAIDI than automating seven switches on System A. Again, this disparity can be attributed to the difference in topology between the two systems and the fact that System B had a much higher SAIDI to begin with and therefore potentially more room for improvement.

The results listed in the tables represent the typical project mix of strategies that a utility might employ to improve reliability. Given such a list and a budget to work with, it is not necessarily obvious as to which projects should be implemented and in what order. A first instinct might be to subsequently implement projects with the highest benefit-to-cost ratio until the budget is exhausted. This approach however, ignores two fundamental considerations: (1) implementation of one project on the list might change the effectiveness of the others (both better and worse being possible), and (2) marginal benefit-to-cost ratios are usually a better guide for optimal resource allocation than the absolute benefit-to-cost ratios. For optimal budget constrained project prioritization, a process using marginal benefit-to-cost analysis is typically required.

Previous articleEnergy Solutions Plus becomes Fortegrasm
Next articleGE supplying 315 wind turbines

No posts to display