This article highlights the need to reevaluate the size of power plants for electricity generation. For decades, the common wisdom has been that increasing the scale of new electric generating units would lead to declining electricity costs. However, since the economic crisis of 2008, the load growth has declined with the declining economy. The elements that argue for bigger scale have become less important, and new factors have arisen that may reduce or mitigate the advantages of scale. The rate of growth in electric demand has become much less predictable. It is also becoming more difficult to raise capital for large plants. The high cost of large new generations of baseload renewable, nuclear, or carbon-free power plants is difficult to support in today’s investment environment. Smaller plants today are both more efficient and more economical to operate than past designs—and thus offer the possibility of overcoming many of the advantages of scale. Moreover, the progress in this area is continuing, both at the grid level and at the distributed or community level.
For electric utilities, this is how the planning process has always worked: you see that the load is increasing, you project what the demand will be 10 years down the road, and then you raise capital to build a plant large enough—in the 1,000 MWe range—to address that needed capacity.
These days, however, things are not always so straightforward. For instance, ask a utility's power supply team, and they will tell you that it's much harder to project demand, what with the uncertain economy and the new technologies coming on line that may either increase the load or reduce it. Ask the board of directors, and you’ll find they are worried about betting the company on a large new capital project. Investors have indicated that they have little appetite for “risky” projects.
I’ve seen utilities stuck in this bind, with uncertainties in demand projection making it too risky to build the plants to meet it. The way out, I believe, is to reassess the third part of the process: the scale at which utilities build their plants.
For decades, the common wisdom has been that increasing the scale of new electric generating units would lead to declining electricity costs. But now several new factors—from changing technology to the rising investment cost of large-scale generation and political difficulties in siting large generators and transmission networks—point toward a nexus where smaller generation may actually create more value, especially when used in conjunction with programmable load shaping.
It's a new world, and the sooner utilities embrace it, the better it will be for their bottom lines.
The case for building large power plants used to be self-evident. When you build at a large scale, you reduce the impact of capital transaction costs, which are often similar whether you are building a 70 MW plant or a 700 MW one, and increase construction and staffing efficiency. You also reduce your exposure to the harrowing process of siting the plants, since you can concentrate your resources on moving through as few regulatory and community relations thickets as possible.
From an engineering perspective, larger units generally run with greater fuel efficiency. Today's primary generators are natural gas-fired turbines in either a combined cycle or combustion turbine configuration. As these units have gotten larger, the heat rates have declined, meaning they use less fuel per kilowatt-hour produced across a broader range of operating states, making them more flexible and efficient when operating.
Since the economic crisis of 2008, however, load growth has declined with the declining economy, large capital investments have become more rare and difficult, the public is angry, and power technology is becoming smaller, smarter, and more efficient. The elements that argue for greater scale have become less important and new factors have arisen that may reduce or mitigate the advantages of scale.
First, the rate of growth in electric demand has become much less predictable. Usage will presumably increase with economic recovery, but at a pace that is very difficult to anticipate. What if the growth rate remains at only 1 percent, instead of the 3 percent rate that had been customary? At a 1 percent growth rate, the New England market (run by Independent System Operator New England) would add approximately 260 MW per year. Four years of steady growth is necessary to support a new large 1,000 MW plant, assuming no one else is building in the market. In today's uncertain and competitive marketplace, it would be risky to assume that when the plant comes online the load will be there. But since it is very hard to turn those large-scale investments off, the result could be excessive capacity, which is another way of saying a lot of wasted dollars and unhappy stakeholders.
In larger markets, the risk is smaller but is still present for both a merchant plant (one funded by investors to sell electricity in the competitive wholesale power market) and for a rate-making body that needs to match reliability with cost.
It is also becoming more difficult to raise capital for large plants. The high cost of large new generations of baseload renewable, nuclear, or carbon-free power plants is difficult to support in today's investment environment. For example, earlier this year Constellation Energy withdrew its plan to develop a new 1,600 MWe nuclear plant in Maryland due to an inability to raise the needed capital without stronger federal loan guarantees.
Smaller plants, conversely, have much less uncertainty attached to their financing. Gary Krellenstein, managing director of the tax-exempt capital markets group at JP Morgan Chase, testified last year to the United States House Committee on Science and Technology about the advantages of funding small modular reactors over traditional designs.
“First, the construction of SMRs requires less capital,” Krellenstein said, “due to their size and other attributes, than conventional nuclear power plants. Second, the smaller capital requirements would allow a single company to build an SMR as opposed to the large and diverse consortium that can greatly complicate investors’ required due diligence as well as their analysis of the management structure of what is already a complex undertaking. Third, the financing for large conventional nuclear plants requires utilities to bear significant default risk such that the construction of each plant is essentially a ‘bet the company’ event.”
It's not just financial risk; there's growing a political risk involved in proposing a large power plant. It is increasingly difficult to obtain permits for power plants that require large land areas and are near electrical loads. This occurs for many reasons. NIMBYism and NIMTOOism (a neologism for Not In My Term Of Office) are frequently cited, but no less important is the increased price for land or rights of way and the difficulty in air quality permitting, which is a higher bar in more congested areas.
Recent protests, including opposition from political leaders and government agencies, against recertification of existing plants such as Indian Point in New York and Vermont Yankee demonstrate a high level of suspicion regarding large-scale generation on both environmental and safety grounds. There's no reason to think this will diminish going forward.
Some of this will be true of power plants of any size, but a smaller unit that seems less disruptive to the neighbors and fits more with the community's self-image may have an advantage when it comes to obtaining permits and land—and avoiding protesters.
The arguments for building smaller-scale power plants aren’t simply financial and political. I’ve seen the technology of power production change in recent years so as to reduce many of the disadvantages that once dogged smaller plants. Smaller plants today are both more efficient and more economical to operate than past designs—and thus offer the possibility of overcoming many of the advantages of scale. And progress in this area is continuing, both at the grid level and at the distributed or community level.
Smaller combined-cycle gas turbines, such as those from Caterpillar's Solar Turbines subsidiary, are already in widespread use. CCGTs recapture waste heat to increase power output over older gas generator designs. Combined with better control systems, this offers a 15 percent or greater improvement in efficiency over previous designs, substantially reducing the usual deficiency in heat rate of smaller-scale options.
Smaller generation can be factory-built rather than constructed on site, and created as modular, repeated designs rather than one-offs, thus reducing cost, improving quality, minimizing regulatory requirements, and avoiding weather risk during the construction phase. And as smaller plants become more automated, fewer people are needed to run them, and the advantage of scale in the staffing dimension decreases. For instance, it is possible to start and run small diesels of up to 10 MW remotely from a network operation center, so that one or two people per shift could manage well over 50 MW.
It's not just gas and diesel where these savings are possible. On the nuclear side, firms such as Babcock & Wilcox and Bechtel are moving aggressively to design practical, commercial small modular reactors that have many of the same advantages.
For all the changes in generation, I think some of the most important developments have been in the transmission and distribution end of the business. “Smart grid” initiatives and other advanced technologies are making transmission and distribution systems more receptive to small-scale distributed generation and microgrids by reducing interconnection and other costs.
Traditionally, generation and the smart grid have been separated—funded by different sources and managed by different entities—often with suboptimal results. I think it makes better sense to integrate the two into a whole-system solution. The result could be smaller increments of generation, leading to lower-cost solutions by virtue of risk reduction, less dislocation as new generation comes on-line, and better use of existing infrastructure. How and when this will happen is a matter of conjecture, but it is important to look for the leading indicators as well as to monitor the cost curves for the points of strategic crossover and change that could create competitive advantage.
Additionally, I believe utilities will be looking not just to meet demand growth with new generation going forward, but also to control or offset growth. After all, the key determinant of the need for additional generation is not total power usage over the year but highest usage on the busiest days of the year: the “peak.” That is, the power grid needs enough capacity to handle the hottest summer days and coldest winter days, with an ample margin for error. If those “peaks” can be flattened, the need to build new large-scale power plants may be avoided.
A trend I’ve increasingly seen in the industry is the development and marketing of special rates and rebates for customers who are willing to shift or shed their usage away from the peaks. One of the major benefits of smart grid and greater energy management sophistication among businesses and residential customers is that usage-shifting can be accomplished with far greater repeatability than ever before, due to the enormous increase in, and accuracy of, usage data. Time-of-use rates, rebates, and the like enable customers to receive a significant benefit for their changes in behavior, and such changes will mitigate the impact of future load increases, perhaps significantly. Load shaping and ultimately “programmable loads” via smart grid technologies will flatten peaks and thus reduce the need for new generation even if total usage does not go down.
One example of this is load control requests made to customers. In the PJM interconnection area, which covers a major area of the eastern U.S., customers supplied over 10,000 MWe of demand response in 2010, the equivalent of operating ten 1,000 MWe power plants. That response included many small generators ramping up and down quickly, in addition to reduced loads through lighting and other programs. The sort of load control response shown by these flexibly managed small generators is difficult or impossible to perform with baseload plants.
Many utility executives are comfortable with the traditional way of meeting demand by adding capacity in 1,000 MWe chunks. But the new technologies offer an alternative: address capacity shortages by reducing peaks and using smaller-scale distributed sources of energy. The challenge for utility companies in this new world is to find the point at which such technologies are reliable and cost-effective enough—or the issues around large-scale generation become difficult enough—that this new approach becomes the better option in terms of feasibility and cost.
The point where the paradigm on scale shifts may be closer than most people think.