North America’s wind power industry has developed an unfortunate reputation for producing energy at levels less than those predicted by pre-construction energy assessments. While not all projects underperform, many do.
Ongoing research suggests the five factors covered in this article are major contributors to over prediction in energy assessments. This article provides questions that project investors can ask to help determine whether these factors have been adequately addressed. Project developers can infer from these questions ways to reduce the risk that these factors will bias a project energy assessment.
Flow modeling bias
The wind flow models commonly used in the industry often do a poor job of modeling wind flow across a site, with a tendency to underestimate variability in wind speeds, even in relatively simple terrain. Consider asking:
- Has the meteorlogical (met) measurement campaign captured the range in the wind resource and characterized the impact of terrain and vegetation variation on the resource?
- Is the exposure of the met towers consistent with the exposure of the turbines and have the relative exposures been considered when estimating the uncertainty in the analysis?
- Has the ability of the model to cross predict the met towers on site been examined and reflected in the uncertainty analysis, with consideration for the relative exposures of the met towers and turbines?
Wake model bias in some conditions
Wake models typically used in the industry produce results reasonably close to actual performance on average for some project configurations and atmospheric conditions, but can also produce large over and under-estimation under other conditions. Consider asking:
- How were variations in atmospheric stability from the average and the relationship between atmospheric stability and the wind speed frequency distribution considered in the wake modeling?
- Do the commercial terms make meaningful provisions for testing the actual impact of a new project on another?
Project availability has been a significant contributor to project underperformance. There are dozens of contributors to project downtime. It is insufficient for an energy assessment to assume that turbine availability will meet the manufacturer’s warranty. So consider asking:
- Does the assumed availability decrease over time as the turbines age?
- Have design teams considered all sources of unavailability and the difference between time lost and energy lost?
- What is the availability track record of the organization that will be providing long-term O&M on the project?
Recent research in North America has shown a notable reduction in power output at a given wind speed in stable conditions, which in many North American locations is typically characterized by a steep wind-shear gradient below 60 to 80 meters. The gradient may even turn negative above 80 m. Knowing the average shear exponent and verifying it with remote sensing are frequently insufficient. Consider asking:
- Has at least a year of remote sensing data been collected at a single point on the site?
- Has the design team considered potential for the shear profile across the rotor to differ from the 40 to 60-meter measurement?
- Have seasonal and diurnal variations in wind shear been considered in relation to the wind speed frequency distribution?
Knowing the correlation between on-site measurements and a nearby long-term reference station is usually insufficient. Making a long-term correction may increase uncertainty. Consider asking:
- Does analysis show that the uncertainty in the result is lower with the long-term correction than without the long-term correction?
- Have the following sources of uncertainty been considered? Period of record, Climatological changes, Quality of correlation, or Known and unknown inconsistencies over time
Asking the right questions prior to commencing an assessment for wind resource and project energy can significantly reduce over prediction and reduce uncertainty, making the wind industry more attractive to investors in the long term.