Interview with ISO-NE as it Undertakes In-Depth Study of Integrating Wind Power into the Grid
September 27, 2010
In 2008, as Independent System Operator — New England Inc. (ISO-NE) considered adding thousands of megawatts of wind power to the region's electric grid, they turned to John Norden. With more than 27 years of experience in the region's electric systems and market operations, John was tasked with understanding the challenges posed by adding growing amounts of variable wind generation to the system. The U.S. Department of Energy Wind Powering America project's New England Wind Forum spoke with John as ISO-NE undertakes an in-depth study of integrating wind power into the grid.
Q. What is the ISO's role with respect to wind power?
A. ISO-NE has three primary responsibilities: operating the New England bulk power grid reliably, administering New England's wholesale electricity markets efficiently and fairly, and administering the regional transmission tariff, which includes developing an annual regional transmission plan. Aspects of wind power are related to all three responsibilities. For example, because many of the onshore wind resource-rich areas in New England are located far from both load centers (such as Boston or southwest Connecticut) and the highvoltage transmission system, developing and delivering wind power will impact the design and operation of the transmission system. All resources—including load, generation, and transmission—have characteristics that must be taken into account, so the goal is to enable the entry of all technologies that can contribute while meeting the region's policy objectives and the ISO's operating objectives. Wind power is a fairly new entrant into this resource mix, and the ways in which it can participate are evolving. As a result, the ISO is actively studying how wind power might affect the region's power system. At the same time, we also are working with stakeholders to reduce the barriers to entry for wind power and other emerging technologies while still meeting the reliability objectives of New England within an efficient market framework.
Q. The ISO recently undertook an extensive wind power scenario analysis. What were the objectives of this analysis, and what did you learn?
A. ISO-NE began its review of large-scale wind integration in 2007 with two studies: the New England Electricity Scenario Analysis and the Technical Assessment of Onshore and Offshore Wind Generation Potential in New England. These studies considered how wind power might help meet future electricity needs in New England, and it provided an analysis of how much wind power could potentially be installed in New England, where it might be located, and the energy production characteristics of those potential facilities. These studies led to two more detailed studies that reviewed several scenarios of large-scale wind penetration. The goal of these analyses was to quantify economic and environmental impacts and potential transmission requirements, as well as to understand the effect on power system operations. The first of these latest two studies, the New England 2030 Power System Study, was performed as a technical input for the New England Governors' Renewable Energy Blueprint. It was released in the fall of 2009. The second study, the New England Wind Integration Study (NEWIS), is currently underway, with interim results (PDF 2.1 MB) already available NEWIS is scheduled to be completed later this year.
Q. What will be the biggest challenges to integrating largescale wind in New England?
A. The biggest challenge is whether we can build the transmission infrastructure necessary to access the region's high-quality wind resources. Besides transmission, the next step for successful wind power integration will be to develop a centralized forecast of wind plant output to help plan and operate the power system in a reliable and efficient manner.
Q. With respect to wind power's variability, how are requirements to accommodate wind production on the grid similar to, and different from, other generation types in the New England portfolio? How are the issues related to the variability of wind similar to, and different from, the issues associated with handling variations in load? How does the ISO handle either?
A. The largest differences between a wind resource and a conventional power plant, such as a gas-fired combined cycle facility, are variability and predicting that variability. It's fairly straightforward to predict the output of a gas-fired generator and its fuel source, although generation is subject to occasional unplanned outages. On the other hand, the power generated by wind turbines is dependent on the forces of nature, which in this case is when the wind blows. For instance, if a wind farm is capable of producing 100 MW under ideal, nameplate conditions, that same facility would only be capable of operating at that level if the wind blew above 25 MPH—all the time. Wind in New England doesn't blow this hard all the time, so the expectation is that wind turbines will usually operate at some lesser value based on the availability of its fuel source. When we measure this megawatt production over the course of the year and look at high-quality onshore sites where turbines might be built, we see that, on average, the energy produced in relation to the physical turbine capability might be about 35%. For offshore facilities that number usually jumps to more than 40%, based on research that we have conducted to date. So while we prepare for new wind plants in New England, we also have to prepare for wind power's variability. We do this now for river hydro facilities. We will prepare for this by generating a forecast of expected operations of the wind plants. Once this forecast has been prepared, we then add conventional resources to the supply mix to be dispatched that day. Of course, this example is greatly simplified, but it provides a general idea of how wind facilities will be integrated with the other available resources to meet consumer demand and maintain reliability.
While load is also variable, the ISO is extremely good at forecasting load, with the forecast consistently within 1% to 1.5% of actual load in any hour. For ISO New England, this amounts to a forecast variation of a few hundred megawatts in any hour. The system has a range of operating reserves on hand to deal with this variation, as well as unplanned generation outages. The hour-to-hour variability of wind generation is both more erratic and less readily predictable than load. While wind is a low percentage of system resources — a few hundred megawatts today — the total wind variability may often be in the load forecast imprecision range; thus, the number of hours does not add to the system's operating reserve needs. As the penetration of wind power grows, even though diverse wind plant locations over a wide area have been shown to smooth out the variability somewhat in the aggregate, the combined variability of load and wind production will increase.
Q. Has ISO New England learned anything from recently commissioned wind farm installations in Maine and New Hampshire?
A. Until the past couple of years, the few wind plants that were connected to the power system were not even large enough (electrically) for the ISO to really "see." One thing we have learned with these recent additions is that the industry, as a whole, has some things to learn about wind integration. From the ISO perspective, wind plants are new, and the type of power system resource they provide is very different from other more traditional generation resources—they are much more variable and unpredictable. Having said that, wind power doesn't only generate energy, but it also provides other services that help support the operation of the grid, such as voltage support. With the installations completed to date, we have coordinated with the developers and wind operators to successfully interconnect these facilities, and we have established good working relationships with all parties to focus on providing reliable power to the interconnection.
Q. Late last year, the ISO commissioned the New England Wind Integration Study, or NEWIS. How does the NEWIS differ from the scenario analysis?
A. The primary difference between other analyses that ISO-NE has performed and NEWIS is the focus of the study. NEWIS was designed to highlight the operational effects of large-scale wind on the region's bulk power system—in other words, what are the challenges across the entire year, from minutes to hours to days. Because of this focus, the data used for the output of the wind plants in NEWIS have a much finer resolution, which helps make the data more useful to a system operator as we prepare to operate with higher levels of wind on the system.
Q. What have other wind integration studies from other electric markets in the United States and abroad led you to expect from NEWIS? What is the objective of NEWIS?
A. Large-scale wind integration studies are a recent phenomenon. In the United States, one of the first studies was performed in 2004 by the New York State Energy Research and Development Authority (NYSERDA). The most recent studies by the National Renewable Energy Laboratory and the U.S. Department of Energy (NREL/DOE) are the Eastern and Western wind integration studies released this year. These large-scale studies demonstrated that integration of a large amount of wind power is very region-specific. Each region has particular wind, generation, transmission, and load characteristics that must be considered when investigating the potential impacts and benefits of large-scale wind power for a balancing area. The objective for NEWIS is to consider a range of possible wind power scenarios and their potential impacts to New England.
Q. What is the status of NEWIS, what have you learned so far, and what do you expect to learn?
A. NEWIS is about 50% completed. So far, one small surprise showing up in the model is that New England's wind resources have slightly higher capacity factors than we expected, even though this model is very similar to one used for the NREL/DOE Eastern Wind Integration Study. For example, we thought approximately 12 gigawatts (GW) of wind would be required from an onshore-based wind scenario to reach a level in which wind contributes 20% of New England's annual electric energy. Instead, we found that even after screening for a range of land use and environmental issues, roughly only 10 GW of nameplate wind would be needed. We have had many good discussions among the team working on the project (General Electric's Energy Applications and Systems Engineering group, EnerNex, and AWS Truepower), NEWIS' technical review committee, and internal and external stakeholders. One part of the NEWIS—recommendations for technical interconnection requirements for wind generators— has already been released (PDF 2.0 MB). We have learned that wind power increasingly has the capability to play as a "full member of the team," not only generating electricity but also providing the ancillary services required to keep the grid operating reliably and efficiently. For example, today's wind technology can participate in voltage regulation, which can be very useful, especially at the outside edges of the transmission system where many wind plants are being installed. We anticipate that the main impediment to integrating large amounts of wind power in New England—from a technical perspective—will be the current lack of transmission infrastructure to "go and get" the high-quality wind resources that are in the New England wind resource area.
Q. Integrating large amounts of wind into the system will have some impact on how the system is run. At what level of penetration might wind power increase the cost of maintaining reliability? How much might such changes cost, and how would these costs compare to the cost of wind itself, and the value of the wind generation in the ISO's energy and capacity markets?
A. The penetration of wind power will be the primary predictor of the impact of wind power on a given power system. There are different methods of measurement, but wind penetration is, in essence, a ratio of the amount of wind power to the amount of load in a particular system. At low penetration levels, the impact of wind on the system operations is minimal. Other studies have found that the impact of even a fairly large amount of wind will be modest, but these studies analyzed particular regions. For example, system operators such as the ISO are required to carry an additional amount of resources above what is required to simply meet the variability of forecasted load. This additional amount of operating reserve margin includes resources that can be brought online quickly to cover sudden, unexpected events that can affect the reliability of the system. Wind power's variability, along with our ability to accurately predict that variability, can potentially increase the size of the required operating reserve margin. NEWIS considers both the power system and the cost impacts of wind penetration at different, higher levels, and these results should be available later this year.
Q. Other regions have implemented wind forecasting systems to inform grid operators and help them operate systems with increasing amounts of wind power. How predictable is wind power? Does wind forecasting work? How does it help, and what are its limitations?
A. Wind forecasting is very important because it helps make the most efficient use of the energy produced both by wind and non-wind generation while helping to ensure system reliability. A state-of-the-art generation forecasting system works toward these goals by producing a forecast for expected wind generation, ideally for the following week, to allow for optimization of other resources and short-term maintenance scheduling. Although a crystal ball would be ideal, significant benefits can be gained even from an imperfect forecast. Generally speaking, the closer you get to real time, the more accurate a forecast becomes. For example, an estimated wind generation forecast will probably be sufficient for one week ahead, but as we get closer to the operating day, operating constraints of non-wind generation will increase the value of an accurate wind generation forecast. This could help reduce reserve margins required to meet unexpected changes in wind generation. Accurate, very short-term forecasting (i.e., next hour to next 10 minutes ahead) will allow even more changes in wind generation to be accommodated in the real-time market, as opposed to being met with other regulating resources. Ramp forecasting is one area in which wind forecasting has room for improvement because of the sudden up or down swings from wind that can be especially tricky to manage during certain load conditions.
This interview was conducted under the U.S. Department of Energy Wind Powering America project's New England Wind Forum for its September issue of its newsletter.