Wind Power’s Impacts on the Operation of Electric Utility Systems
Ed DeMeo, Renewable Energy Consulting Services, Inc.
Wind power plants generate electricity when the wind is blowing, and the plant output depends substantially on the strength of the wind. Because the wind cannot be accurately predicted over daily periods and it often fluctuates from minute to minute and hour to hour, electric utility system planners and operators are concerned that wind plant variations may increase the operating costs of the system as a whole. This concern arises because the system must maintain an instantaneous balance at all times between the aggregate demand for electric power and the total power generated by all power plants feeding into the system. Utility operators and automatic controls perform this highly sophisticated task routinely—based on well-known operating characteristics for conventional power plants and a great deal of experience accumulated over many years.
Wind Power Impacts on Operating Costs System operators are concerned that variations in wind plant output will force the conventional power plants to provide compensating variations to maintain system balance, thus causing the conventional power plants to deviate from operating points that are chosen to minimize the total cost of operating the entire system. The concern is certainly valid. The question is: To what extent does the variability of the wind increase operating costs? The operators’ concerns are compounded by the fact that conventional power plants are generally under their control, whereas they have no control over wind plants because they are controlled by nature. Another concern expressed by utility operators who are unfamiliar with wind plant operating characteristics is that the output of a wind plant will change from full power—say 100 MW—to zero in one second or less, causing a huge transient impact on the system. Practical experience with many wind plants has alleviated this concern. Wind plant output does not change that rapidly. Even a single wind turbine has sufficient mechanical inertia to damp rapid changes in the wind. More important, a wind plant generally consists of a number of turbines, and the spatial variations in the wind over the area of a typical plant are sufficient so that variations in output from the entire plant are much less pronounced than those from a single turbine. Hence, the plant output shows substantial smoothing relative to output from a single turbine. Consequently, wind plants have no adverse impact on the power system’s stability. System stability can be upset by abrupt events that happen within a fraction of a second, such as a sudden outage of a major power plant, loss of a transmission line, or abrupt connection of a large electrical load like an arc furnace in a steel mill. Abrupt events such as these do not occur with a wind plant unless its connection to the electrical grid suffers a fault. In such a case, the wind plant is similar to a conventional power plant. Impacts in the time frame from a few seconds to a few days can be significant, however. Utility operators tend to address the system-balance issue in the following different approximate time frames:
• Regulation: one second to a few minutes
• Load following: a few minutes to a few hours
• Scheduling and commitment of generating units: a few hours to a few days.
Regulation Customers are continually turning appliances, production processing equipment, and other electrical loads on and off. Consequently, the system constantly experiences random variations. These are routinely handled without difficulty by the system through generating units that are assigned this function. Operating these plants in the regulating mode incurs costs to the system. Wind plants add to these variations, but in a random and uncorrelated manner. In principle, they will add to the regulating burden and hence to the cost of regulation. To date, however, studies with wind plant penetrations in the range of 5% to 20% of system load estimate this cost impact to be minimal to negligible.
Load Following Aggregate utility loads generally follow fairly predictable daily and weekly patterns. For example, loads will increase in the morning hours as people wake up, businesses begin their operations, and manufacturing processes ramp up. Conversely, loads will drop off later in the day. These variations are handled by load-following generating units that are ramped up and down by system operators or by automatic equipment. The presence of wind power in the generating mix will generally increase the requirement for load-following generation because the behavior of the wind over a several hour period is generally not as predictable as customer load patterns. This increase results in increased operating costs. However, studies to date suggest that for low to medium wind penetrations (up to about 5% of system load), the resulting cost impact is on the order of 0.05 cents/kWh of wind energy.
Scheduling and Commitment Large thermal power plants generally require lead times of several hours to as much as a day to reach system service readiness. Consequently, operators need to make decisions about plant operations hours before the plants will be needed. Plants that are already warm or can be started quickly need to be scheduled. Those that have not been brought up to operating temperatures need to be committed. These decisions will be affected by assumptions made about wind plant operation. If the wind could be forecast accurately, reliable assumptions would be possible. However, even with perfect forecasting, variations in wind plant output would necessitate more variations in conventional plant output than would be needed if the wind plant outputs were steady. These additional variations imply additional costs because the conventional plants would be operated more often under non optimum conditions and because maintenance costs are likely to increase. However, those knowledgeable in power plant operations feel such additional costs would be small compared to those resulting from imperfect wind forecasting. For example, suppose a thermal plant has been fired up to serve expected load during the next day because no wind is expected. If the wind actually blows the next day, then the additional thermal plant is not needed and the cost associated with firing it up were unnecessary. Conversely, a decision to rely on wind power that does not materialize causes extra expense to obtain makeup power—often from spot markets at high costs. Today, wind forecasts are generally accurate for about half an hour up to one or perhaps two hours. Although the ability to forecast is improving, accuracy over periods of a day or two is not likely in the foreseeable future. Several preliminary studies of forecasting-error impacts have been conducted. These suggest that, for wind penetrations of 5% to as much as 20%, the operating-cost impacts of “bad” decisions caused by errors in forecasts are in the range of 0.15 cents to 0.5 cents per kWh of wind-generated electricity. These studies have incorporated several conservative assumptions, so the impacts may actually be overstated.
Conclusion Results to date, coupled with actual experience from operating wind plants, suggest that system operating-cost impacts are not a showstopper for wind. To strengthen this conclusion, and to determine conditions under which it may not apply, additional studies are needed. These studies should examine such effects as (a) different mixes of conventional generation; (b) a range of wind penetrations; (c) a sliding scale of forecasting accuracy (i.e., very accurate for an hour or two, and decreasing in accuracy out to 48 hours); and (d) differing assumptions on the purchase of makeup energy and the sale of excess energy.
About the Author
Tell others about
this page:
Comments? Questions? Email Here