Beyond the simple formula
By Jean Agras and Jeffifer Tripp
As with other capital intensive items with substantial lead generation times, generation resource supply relative to demand is cyclical. Many regions within the U.S. are currently experiencing excess capacity due to the heavy build out of natural gas-fired capacity additions since 2000, while others are already struggling to meet peak demand.
There is no simple formula for determining the type of resources ultimately constructed. There are resource diversity considerations; substantial lead times associated with permitting and building new coal and nuclear plants; and transitional state of the markets and regulatory uncertainty, all of which present obstacles for getting new generation permitted, financed and constructed in a timely manner.
The expectation is that much of the resource growth will be met with baseload generation, which varies substan-tially by region, in both fuel type and quantity. Additionally, many markets need to add renewable resources to their mix to meet renewable portfolio standards. Depending on the need for new capacity, this may lead some markets to build primarily renewable resources, which will not necessarily meet baseload needs.
Evidence of the viability and risk of building in different markets is seen in the post 2005 baseload projects that have already begun construction. (See map.) In ISO market areas that had a divestiture of utility generation (New England, New York and California), where subsequently new generation is primarily planned and constructed by independent entities, there are few new baseload facilities under construction. Typically, independent developers will not begin construction until all permitting and financing has been successfully accomplished.
Uncertainty Affects Investment Decisions
Estimating the profitability of an investment relies on advanced modeling techniques that require knowledge of the future costs and the future revenues of the generating project and variations. For a company with choices among generating options, with different lead times and different cost uncertainties, a uniform methodology is needed to compare technologies with different characteristics and different risk profiles.
One of the most fundamental changes affecting the value of investments in competitive markets is electricity price volatility. The risks can be in the form of volatility associated with future market price of energy, further exacerbated by transmission congestion costs and losses that may cut into already slim margins. While this risk affects all generating technologies, it is greater for technologies that have high capital costs, low fuel costs, less operating flexibility and generally require siting further from load centers-like wind, coal and nuclear.
Baseload plants under construction or site preparation. Source: NERC Click here to enlarge image
Historically, utilities examined generation additions from a “least cost” planning approach. Various generation alternatives, designed to meet expected loads over a specified future time period, are considered and the lowest or “least cost” alternative usually becomes the one recommended and pursued. The quantification of the least cost alternative was typically accomplished by using a “dispatch model” that simulated the expected use of all generation assets, including alternative additions being considered to meet the load obligations of the utility. The forecasts for fuels and load were assumed to be known, and to not have random movements over time.
Uncertainties over future fuel prices, climate change policies, technological progress in all the major power technologies, and the impact of higher prices on power demand, create substantial risks for new generation investments, and these risks add to the cost of financing. Within each different type of market and across different technologies, investment in baseload generation contains a set of risks, ranging from economic factors to technology risk, regulatory risk, transmission delivery risk and power and fuel price risks. All of the risks are correlated and need to be considered in an investment decision model but each of these risks will affect decisions differently across markets and technology types.
In a competitive market, these technologies primarily rely on energy revenues to cover capital and fixed costs. Even in competitive regions that have established separate capacity markets, these technologies face capital cost recovery risk because the capacity prices are typically set by a much lower capital cost peaking plant. Even with greater baseload energy production and higher energy revenues, the higher capital cost baseload resources have difficulty covering annual fixed costs.
In addition to market price volatility and congestion, coal plants, in particular, face regulatory uncertainty relative to implementation of a carbon tax. In fact, carbon tax uncertainty may be a turning point for development of new nuclear generation because with no carbon emissions, nuclear facilities could expect higher margins following implementation of a carbon tax. Natural gas plants, for the most part, will also be less affected since in most markets the energy price is set on-peak by natural gas-fired generation and in off-peak either by natural gas or coal.
By the same argument, if natural gas-fired generation is setting the energy price in the majority of hours, then coal plants will not be able to recover their incremental carbon costs, given that the carbon cost for a coal plant is about 2.5 times that of a natural gas-fired combined cycle. This implies that if carbon costs were $10 per ton, a coal plant would lose $6 per MWh for each hour that the combined cycle is the energy price setting unit. For a 1,000 MW coal unit, this could translate into a potential loss of almost $50 million in a single year.
Investors and developers need to make investment decisions that incorporate risk and are the most economically feasible. An investment decision should determine the optimal generating and transmission capacity investment strategy. For each year, for each possible combination of values of the uncertain variables, and for each possible prior decision, the model needs to select the portfolio that best achieves the decision-making criterion from that point forward, taking into account uncertainty about the future, the irreversibility of investment decisions and the ability to make additional decisions in the future.
Jean Agras, Ph.D., is a senior director in R. W. Beck’s generation planning and analysis practice in Denver. She can be reached at firstname.lastname@example.org. Jennifer Tripp, P.E., is a principal and national director of R. W. Beck’s transmission policy and analysis practice in Phoenix and can be reached at email@example.com.