What will it take to solve the problem?
If the fraction of power we get from variable and intermittent renewables is to grow by an order of magnitude, profound changes will be needed in the way we plan and operate the electric power system. That in turn will require a new policy environment that promotes the needed innovations and their rapid adoption. The place to start is with an objective look at the building blocks that can be assembled to address the problems posed by large quantities of variable and intermittent power. These building blocks include:
- Better prediction of variability. This would allow utility dispatchers to keep other plants available for use when there is a high probability of shortfall, either in their own units or in distributed wind or solar generation.
- Novel strategies at the point of wind or solar power generation to reduce variability. These may include operating at less than maximum capability to provide the ability to use one plant to compensate for a drop at other plants, as is done for fossil generation now.
- New methods for optimally dispatching power plants and reserves, based on the uncertainty and variability of renewables, that account for the risk of blackouts, including those that result from cascading outages. As the amount of variability and uncertainty increases, operators will need new decision tools to dispatch resources efficiently. Most existing methods for dispatch do not consider the cascading outage risk, which may become more important with high penetrations of wind and solar.
- Improved strategies for monitoring and controlling transmission systems and for anticipating and detecting incipient instabilities and preventing cascading failures. The project will consider evaluating combinations of decentralized methods to mitigate problems that may progress in short time scales (seconds to a few minutes), and regionally coordinated methods that work on longer time scales (minutes to tens of minutes).
- New gas turbines designed to quickly change power levels without excessive air emissions. In parts of the country that do not have significant hydroelectric power, gas turbines will be the dominant choice to fill in gaps in variable power supply. Analysis done by one of our Ph.D. students13 has shown existing gas turbine designs have very serious NOx emissions problems when they are rapidly cycled up and down to fill the gaps. These may be sharply reduced if the gas turbines are optimized for fast ramping.
- Electric and thermal storage. Storage makes it possible to hold energy that is generated at a moment when it is not needed to make it available at another moment when it is needed. Such storage can be done by utilities, companies, or at individual buildings.
- Intelligent distribution system and customer load control.14 An alternative to storing energy for use at a later time is turning non-essential loads on and off to match the availability of energy supply from fluctuating sources such as wind. However, doing that requires much more sophisticated control and cooperation than is available in most power systems and customers' premises. Implementing such control raises important policy issues. For example, should customers be able to choose when they want to respond to real time price signals or should utilities be able to turn loads on and off directly?
- Greater use of combined heat and power (CHP) and micro-grids. Distributed generation that makes productive use of the surplus heat can dramatically increase the overall efficiency with which energy is used, increase the security of power delivery to end users, and add flexibility to the system that, together with intelligent load control, can facilitate more flexible operation and greater ability to deal with variable and intermittent supply.
- Greater ability of base-load power plants such as nuclear plants (and perhaps large CCS coal plants) to quickly change power levels. At present, the output of most base-load plants can be changed only gradually, and they must be operated at half or more of their rated power. If strategies can be found to move output up and down more quickly and operate them efficiently even at low power levels, that would reduce the dependency on gas. There is considerable concern in some circles about a possible gas-wind future that holds the potential to make the country increasingly dependent on imported liquefied natural gas (LNG).
- New standards for frequency and voltage control, which along with greater use of solid-state power conditioning, will allow wider power-system operating ranges without imperiling end-use devices or system electrical stability.
- More and higher capacity transmission lines. Transmission lines can help move power from where it is generated (e.g., wind farms in the north Great Plains) to where it is needed. It is becoming increasingly difficult to build new lines, but some addition of lines, including HVDC lines, will likely be needed if the U.S. is to use large amounts of variable and intermittent power. In addition, the capacity of existing transmission corridors can be augmented through increasing operating voltages, replacing existing conductors with "low thermal expansion" conductors, and making greater use of solid state flexible AC control (FACTS) technology. Greater use of instrumentation and autonomous and distributed control can also allow the transmission system to be run with higher capacity and reliability.
- Better understanding of offshore wind resources. Offshore wind may be less variable. If they are far enough out, siting wind farms may also be easier. However, there has been considerably less assessment of the resource, project costs, etc. for U.S. offshore sites than for on-shore sites.
- Plug-electric vehicles. Because these vehicles have on-board battery storage, it is at least possible that they might play a role in providing short-term storage (and ancillary services) both at homes and businesses that generate power from wind and solar and to the grid. However, while there has been much talk about using plug-electric vehicles as a means of developing distributed storage, to date the value proposition for vehicle owners looks decidedly unattractive,16 so novel compensation mechanisms are likely to be required to make this practical.
- New, more equitable and more flexible regulatory environments and rate structures. At the moment, there are strong impediments to the use of distributed generation and microgrids that can only be removed through changes in regulatory law, and the development of more equitable rate structures.
- Improved system-wide facilities expansion planning. Before the restructuring of the power system, vertically integrated regulated utilities could do coherent planning for their systems. Today, we have a patchwork of regulatory arrangements, with some states totally "deregulated" and others still operating with traditional regulation. However, the various power systems involved are still synchronously connected. In the face of a growing need to manage variable and intermittent supply, it will be increasingly important to find alternatives to a laissez faire approach to facilities expansion.
Those are the building blocks. We, and many others, have done work on them separately. What is now needed is a systematic high-level analysis that integrates these pieces in technical and policy frameworks that can support large amounts of wind and solar power at an affordable price, and with acceptable levels of security and reliability