Information on the costs incurred by electricity customers as a result of power interruptions (total outage) or power curtailments (partial outage) is important for the planning of the capacity of electricity systems, the efficient and equitable setting of electricity prices and the development of priority service programs. In the past few years, Taiwan Power Company (Taipower) has experienced an annual two-digit growth rate during system peak-load demand. In addition, the reserve margin of power in Taipower’s generating capacity has dropped. Last summer, power shortages appeared and are likely to continue to be a problem since there won’t be enough generating units on-line to meet the demand growth.
In the near future, given the inevitability of power shortage, an evaluation of outage costs will ultimately be important for the implementation of power rationing. Therefore, the purpose of this study is to investigate the costs related to outages caused by the insufficient generating capacity of Taipower and those caused by a shortage of fuels or other sources of power (eg., hydro). In order to accomplish this, we first introduce the theory that relates to outage costs which includes the relationships between outage costs and the optimum reliability of generation, transmission, distribution, and transformation. We also clearly distinguish and separate the costs associated with a shortage of electricity resulting from an insufficient generating capacity and those associated with a shortage of fuels or other power sources. Finally, methodologies for evaluating outage costs are examined and a comprehensive review of the literature on the subject is included.
In our empirical studies, we use market research and a linear programming input-output model (LP/IO). We conducted two surveys. One looking at the problem of outage costs from an ex-ante or hypothetical standpoint and the other from an ex-post or factual case study and both are based on high tension power customers with electricity consumption of 1000kW or greater. The surveys focus on the impacts of the outage ratio of the customer’s contracted capacity (in kW), outage duration, and whether there was prior notification of the power outage. We break down the costs into those that are associated with production loss, clean-up, restarting after the power outages and others, such as those associated with the environment or health. The results of these surveys are then subjected to variance (ANOVA) and covariance (ANCOVA) analyses. We also subject the data to regression analysis in order to derive the average outage cost for the various industries represented in our sample.
The LP/10 model that we use integrates Taiwan’s input-output tables of 1984 within a linear programming structure which maximizes gross domestic product (GDP) subject to the constraints of interdependence among industries and the percentages of total electricity supply shortages — we use 5%, 10%, 15%, and 20% in the analysis. We also use an ordinary least squares (OLS) method to simulate a function for describing the relationship between outage costs and power shortage levels given the results of the LP/O model. The results of the LP/IO model are then compared with those of the market surveys. The implications of comparative results are interpreted.
