Just as free trade in goods offers opportunities for efficiency gains, electricity transmission infrastructure facilitates exchange in electricity markets. The first-order effects of geographic integration in electricity markets are straightforward – allocation of production across firms is improved, so lower cost generating resources can be used to meet demand.
In addition, some unusual features of the electricity market make transmission particularly valuable. Storage of electricity, unlike most goods, is prohibitively expensive. As a result, supply must meet demand on a continual basis. Moreover, demand is both price-inelastic and highly variable. As a result, the market typically clears on the supply side, and large price swings are possible whenever supply is inelastic. Geographic integration can smooth this volatility by expanding the number of potential suppliers.
Finally, adequate transmission has important implications for competition in deregulated electricity generation markets. The combination of inelastic demand and lack of storage means these markets are susceptible to the exercise of market power. Because it increases the number of relevant competitors, adequate transmission is necessary for competitive generation markets to function well (Borenstein et al. 2000, Wolak 2012, Ryan 2013).
While the qualitative importance of transmission is well understood, the ability to quantify the benefits is still important for policymakers, because transmission investments are not made in a competitive marketplace. Transmission is a natural monopoly, so it has traditionally been price-regulated; in addition, new investments face a suite of overlapping economic and environmental regulations (CEC 2009). Transmission has also been at the heart of policy discussions on incentivising the expansion of renewable generation (DOE 2009, CAISO 2014).
Quantifying the benefits of transmission is tricky. Ex-ante simulations must rely on simplifying assumptions that can only approximate how generators and system operators behave in real time. Ex-post calculations are difficult because building a credible counterfactual is not always possible; investments in transmission capacity are endogenous responses to changes in market conditions, and they are planned years in advance. Previous studies of transmission constraints in electricity markets have either used stylised theoretical models (Cardell et al. 1997, Joskow and Tirole 2000), or Cournot simulations (Borenstein et al. 2000, Ryan 2013).
Calculating the benefits of transmission
In new work, we use a different approach. We take advantage of a natural experiment to calculate ex-post the value of electricity transmission (Davis and Hausman 2014). In 2012, the San Onofre Nuclear Generating Station in California was closed unexpectedly because of safety concerns. The plant had been a large resource, generating enough electricity to supply 2.3 million households – about 8% of all electricity generated in the state. Moreover, it was located near two large demand centres (Los Angeles and San Diego) in an area with limited transmission capacity. As a result, its closure caused transmission constraints to bind, substantially increasing the cost of meeting demand.
We use data from the year and a half leading up to the San Onofre closure to construct a supply curve for a hypothetical world in which San Onofre had not closed. We then compare this counterfactual to observed generation outcomes following the closure, calculating the increased cost required to meet demand. Overall, we find that the cost of electricity generation in California increased because of the closure by about $350 million during the first twelve months. This is a large change, equivalent to a 13% increase in total in-state generation costs; yet it went almost completely unnoticed because of a contemporaneous offsetting decrease in natural gas prices.
We further decompose the cost increase into two effects. The first effect is simply the cost of using higher cost natural-gas plants to fill in for the lost nuclear generation. Like other nuclear power plants, San Onofre produced electricity at very low marginal cost. When the plant closed, this generation had to be made up for by operating other, more expensive generating resources. The second effect is the additional cost attributable to transmission constraints. The binding transmission constraints meant that it was not possible to meet all of the lost output from San Onofre using the lowest cost available generating resources. We calculate that the transmission effect totalled around $40 million in the first twelve months following the closure.
With this estimate of $40 million, we find that several transmission projects would have payback periods of less than ten years. Indeed, the California Independent System Operator has been working since 2012 to implement several projects to relieve the congestion caused by the San Onofre closure.
In addition to state-wide estimates of changes in generation, we look at effects at the individual plant level. We document two important outliers: plants that would have been expected to substantially increase their generation following the San Onofre closure, but whose average generation remained essentially unchanged. As it turns out, these plants were operated through a tolling agreement with JPMorganChase, and they were investigated by the Federal Energy Regulatory Commission for market manipulation between 2010 and 2012 (FERC 2013). While our analysis cannot be used to prove withholding or other market manipulation, it does serve as a useful diagnostic tool for unusual plant behaviour.
In addition to the value of electricity transmission, several other policy implications emerge from our analysis. We document that the closure of San Onofre had important environmental implications, in addition to the generation costs described above. While California currently has a cap and trade programme for carbon dioxide emissions, the programme was not yet in place in 2012. Over this period, we find that the closure of San Onofre increased carbon dioxide emissions by 9 million tonnes, implying a social cost of almost $320 million per year. A large fraction of the world’s nuclear plants are beginning to reach retirement age, and it is important to take these external costs into account as decisions are made about whether or not to extend the operating lives of these plants.
Despite the high cost of the San Onofre closure, the decision to shutter the plant appears to have been optimal from the operator’s perspective. The plant’s annual operations and maintenance expenditures were substantial – around $340 million per year. Thus even though San Onofre had a much lower marginal cost than natural gas plants, its profitability was questionable. Indeed, the impact of low natural gas prices on wholesale electricity prices is jeopardising the balance sheets of many US nuclear plants.
Having adequate transmission helps electricity markets to run more efficiently and reliably. The closure of the San Onofre plant in California is important in this context – it shows how the unexpected closure of just one resource can have large economic costs. Our research finds that closing the plant increased the cost of generation by $350 million per year, and that $40 million of this is attributable to transmission constraints. Additionally, the closure led to an increase in carbon dioxide emissions worth $320 million annually.
On the one hand, these costs do not appear large in comparison to the fixed cost of keeping a nuclear plant open (insurance, employees, and so on). Indeed, this is why some analysts believe the US nuclear industry is at risk.
On the other hand, the costs could potentially have been much higher. If a particularly hot summer had led to spikes in demand, if natural gas prices had been higher, or if a substantial amount of market power had been exercised, the cost of the San Onofre closure (including the cost of transmission constraints) would have been significantly larger.
Borenstein, Severin, James Bushnell, and Steven Stoft (2000), “The Competitive Effects of Transmission Capacity in a Deregulated Electricity Industry”, RAND Journal of Economics, 31(2): 294–325.
California Energy Commission (CEC) (2009), “Strategic Transmission Investment Plan”.
California Independent System Operator (CAISO) (2014), “2013–2014 Transmission Plan”.
Cardell, Judith B, Carrie Cullen Hitt, and William W Hogan (1997), “Market Power and Strategic Interaction in Electricity Networks”, Resource and Energy Economics, 19(1): 109–137.
Davis, Lucas and Catherine Hausman (2014), “The Value of Transmission in Electricity Markets: Evidence from a Nuclear Power Plant Closure”, NBER Working Paper 20186.
Federal Energy Regulatory Commission (FERC) (2013), “Order Approving Stipulation and Consent Agreement, Docket Nos IN11-8-000 and IN13-5-000”.
Department of Energy (DOE) (2009), “National Electric Transmission Congestion Study”.
Joskow, Paul L and Jean Tirole (2000), “Transmission Rights and Market Power on Electric Power Networks”, RAND Journal of Economics: 450–487.
Ryan, Nicholas (2013), “The Competitive Effects of Transmission Infrastructure in the Indian Electricity Market”, Working Paper.
Wolak, Frank (2012), “Measuring the Competitiveness Benefits of a Transmission Investment Policy: The Case of the Alberta Electricity Market”, Working Paper.