Supply and demand — the general economic model that makes the world spin — underpins how far clean renewable energy can go to satisfy our needs. Wind power supply is dependent upon sufficient wind speeds being present to spin turbines. Daily wind speed patterns over land are created by thermal heat exchanges with oceans, yet scientists and engineers involved in energy production conventionally treat wind speed as a random variable governed by probability. An alternative approach to wind power evaluation, one that relies on daily and seasonal pattern distinctions in the observed data, presents the opportunity to more tightly match the available wind power supply and energy demand.
Most wind energy farms in the United States are located in the gusty, high-velocity wind areas of the Midwest. This region benefits tremendously from a bountiful supply of wind energy. Other regions of the country, such as the Southeast, experience lower wind velocities, yet have a high concentration of population. In these regions, there exists a mismatch of clean energy demand and the supply available to address it. Could wind power be as feasible an energy source in the bright, densely-populated Southeast as it is in the gusty, open agricultural area of the Midwest?
Wind measurement processes are complex and dynamic. Further, the high variability of historic wind-velocity records gives a random appearance to the naked eye. Perhaps this explains why scientists and engineers conventionally conceptualize wind velocity as a random process, most effectively modeled with a variety of probabilistic approaches. That randomization comes at a cost: any wind velocity patterns are removed from the data causing it to often fall short of reflecting the complexity of natural behavioral patterns critical to good planning. Just like scrambling the order of dots and dashes in Morse code, the coded message in wind velocity data is lost during randomization. Consequently, it must be restored synthetically in a probabilistic framework by, for example, calculating different sets of wind-velocity frequencies for day and night, winter and summer, and so on.
A New Approach
Nonlinear time dynamic analysis is a recognized empirical method designed by physicists to detect and characterize complex behavioral patterns in dynamic systems. Such techniques are beginning to be used by scientists to analyze climate variability.
Similarly, wind velocity data can be planned. It has a coded message coinciding with heat if scientists, engineers and project planners are willing to interpret it accordingly. In particular, observed wind velocities exhibit systematic temporal behavior that can be used to compute long-term daily wind-power patterns. These patterns can then be matched with daily energy demand patterns. Once detected, behavioral patterns also can be used to make short-term predictions of wind-power supply.
The result is that the coded message in wind-velocity data can be retained, eliminating the vexing problem of how to synthetically restore its natural complexity. Wind project evaluators should initially “let the data speak” regarding whether a conventional probabilistic approach or a nonlinear dynamic approach to wind-power evaluation is best.
Changing an entrenched belief system requires solid data, proven methods and indisputable visualizations. Seeing is believing. The data from this study show predictable wind velocity patterns, but standard graphing techniques lack the ability to adequately representit because the data were spread out over an extended period of time for this research. A static graph also failed to show how the data evolved. With OriginLab’s Origin data analysis and graphing software, the data were displayed in a 3D model and animated by sequential plotting of data points to illustrate the cyclical, periodic patterns of the wind velocity’s systematic orbit. (See video animation at this link.)
Adding point after point shows how the research data grow over time. It demonstrates how wind velocities systematically evolve along satellite-like orbit construction. The animated plot visually provides the feeling of the wind increasing in the pattern as it actually occurred, not in a random fashion. With this visualization, the study disrupts conventional beliefs and clearly provides the means to match the predictably cyclical wind velocity patterns with patterns of demand.
This testing was conducted on the proposed Sugarland Wind Project of South Palm Beach, Fla. to determine the extent to which observed periodic, cyclical patterns of wind velocity modeling could be matched with the rise and fall of typical Floridians’ demand throughout the day, month or year. The research was intended to investigate the extent to which wind power patterns corresponded with energy demand patterns, and thus, compensate for lower mean wind speeds in order to increase the commercial viability of Sugarland Wind.
The project, which was once referred to as “can’t be done,” was found to be feasible due to breakthroughs in turbine technology capable of generating commercially viable power at lower wind speeds and given the presence of a proper supply and demand match.
Observed wind velocities in the project area showed a sequence of non-repeating orbits generated by strong daily and fainter 25-day oscillatory periods. This research found that computed wind power supply patterns generally matched well with peak daily demand in southern Florida’s hot season, thus allowing residents to cool their homes and offices with renewable, clean energy when they needed to. They also matched well with peak morning demand in the cold season. They did not match well, however, with peak evening demand in the cold season, indicating potential need for increased energy storage if wind power is to supply this period more effectively.
Wind project evaluation must reliably identify supply and demand patterns and determine how well they match. The Sugarland Wind research provides evidence that nonlinear dynamics techniques deserve space in the wind-project evaluator’s toolbox for this purpose. It found that wind velocities in the project area exhibit satellite-like patterns that do not repeat perfectly, but do repeat in a similar fashion. Both the probabilistic model and the systematic matching model arrive at an outcome that can be correct, but why leave it up to chance if proximity to certainty is available?
Driving commercial wind viability so it can be scheduled and integrated into the power grid predictably creates the feasibility of clean wind power outside of the Midwest and into areas that currently rely heavily on traditional fossil-fuel-based generation. These research and findings are among the first steps in an ongoing process to ignite the growth of wind power around the country and the world.
Ray Huffaker is a professor of Agricultural and Biological Engineering at the University of Florida at Gainesville. Marco Bittelli is an assistant professor of Agro-Environmental Science and Technology at the University of Bologna, Italy. This article reflects research originally presented in “A Nonlinear Dynamics Approach for Incorporating Wind-Speed Patterns into Wind-Power Project Evaluation.