Integrating the Social Cost of Power Outages into Reliability Planning

Mr Nelson Jr Enano, IEPI, UCL Australia, University College London

Michel Berthelemy, IEPI, UCL Australia, University College London

Tony Owen, IEPI, UCL Australia, University College London

Stefaan Simons, IEPI, UCL Australia, University College London

Overview:

In electricity systems, the loss of load expectation (LOLE) is often used to assess the required capacity needed for system adequacy. While electricity consumers are accustomed to LOLE based level of reliability, there remains large uncertainty regarding the economic costs and risks power outages impose to electricity consumers. This uncertainty tends to increase with the liberalization of power markets, the possibility of demand-side participation, or the level of renewable energy penetration. In addition, the policy relevance of these structural changes is particularly relevant in developing countries where high level of power outages often remains the norm.

The value of lost load (VOLL) is the willingness to pay (WTP) of consumers to avoid an additional period without power (Leahy and Tol, 2011). This study outlines various methodologies that can be used to estimate the VOLL and presents the results of an empirical investigation to estimate the VOLL in Mindanao, a small island in the Philippines with high level of power outages. Finally, we discuss the impact of methodological choices when assessing various policy proposals to reduce power outages in the context of developing countries.

Methodology:

Due to the lack of readily available data on demand-side preferences, both micro and macro level proxies are used in the economic literature and by practitioners to measure the VOLL. In this paper, both approaches are estimated and their related merit and drawbacks are discussed accordingly.

Following Oseni and Pollitt (2013), we pursue a micro level approach to estimate the VOLL using consumers’ revealed preferences in maintaining their desired level of reliability through investment in backup or self-generation capacities. In parallel, we also follow the macro level approach based on the load profile of different representatives groups of electricity consumers (Leahy and Tol, 2011).

Results:

Our preliminary results for the Mindanao Island show that the average VOLL derived from the revealed preferences approach (US$ 4,968 per MWh) can be significantly above the level derived from the macro level approach (US$ 1,733 per MWh).

In addition, we highlight that the VOLL, as a demand-side concept, cannot be a single value as it depends on the valuation of reliability of different consumer groups (e.g. residential, industrial), which, in turn, depends on the level of reliability in a specific location and the level of readiness of consumers for meeting outages (e.g. owning backup facilities or the notices before an outage). We further argue that an annualised VOLL is inappropriate as consumers WTP varies with outage duration and frequency or the time and season these outages occur. We illustrate how this problem of aggregation can distort investment decisions for both the demand and supply sides.

Conclusions:

The VOLL estimates are used to analyze and discuss the economic rational of various investment opportunities in the Mindanao electricity system, including the interconnection of the grid with other islands or investment in baseload electricity capacity. Under current regulations, the economic case for these different investment options remains at best precarious.

References:

Leahy, E. and Tol, R. (2011), An estimate of the value of lost load for Ireland, Energy Policy, Vol. 39, pp 1514-1520.

Oseni, M. and Pollitt, M. (2013), Power outages and the costs of unsupplied electricity: evidence from backup generation among firms in Africa, EPRG Working Paper No. 1326, University of Cambridge.

Keywords: Reliability, Value of Loss Load, Electricity Demand, Demand Side Management