BigWind ‘flexes muscle’ to change scientific results

Screen Shot 2018-04-29 at 6.57.30 PM

It was bound to happen but we never thought it would happen so fast.  The wind lobby appears to have flexed its muscle and caused NREL to pull back on the Larwood study recommending a setback of 2x total height to property line and 3.5x total height to a residence.  This is not the first time a study was attacked and retracted after wind developers sent their dogs after researchers. 

We recall a few years ago the State of Michigan assembled a team of acousticians and medical experts to  study the appropriate setback distances and noise limits for wind turbines. Their draft proposal was 40dBa and one half mile setback. Magically the report was buried. Last spring, Michigan State University updated their model wind ordinance language to include noise limits of 40dBa and setback distances out to 3,000’+. That study lasted about 6 weeks. Documents from a Freedom of Information Act (FOIA) inquiry into MSU revealed that AWEA and other wind lobbyists pressured MSU to pull the sample zoning.  And now we have the NREL study receiving the what appears to be the same treatment. 

A visit to NREL at now provides a “clarification”:

Larwood and Simms article clarification: “Analysis of blade fragment risk at a wind energy facility,” Wind Energy, 2018;1-9. DOI: 10.1002/we.2194.

The Larwood and Simms article was intended to illustrate the process by which a site-specific analysis can be used to assess the risk induced by loss of a wind turbine blade.  The study used work done to address the safety of a unique wind energy testing facility – the National Wind Technology Center near Boulder, Colorado. The assumptions used by the authors include site-specific wind conditions and local risk factors. Assumptions of turbine blade failure rates were taken from a report over a decade old, which was considered conservative due to the continuous improvement in blade reliability in the intervening years. The study concluded that despite the proximity of turbines and research facilities, operations at the site were safe relative to blade throw risk. The assumptions and results do not apply to commercial wind energy sites and are not suitable for setback recommendations in other locations. However, some of the conclusions were summarized in a way that made them appear to be more broadly applicable than intended. Due to the possibility of misinterpretation, the authors are working to withdraw the version causing confusion and resubmit with a more clearly articulated summary.

Summary of important issues:

  1. Although the reported methodology is intended to be more generally applicable, the conclusions reached about the National Wind Technology Center should not be assumed as applicable elsewhere or for different wind turbine configurations.
  2. The authors used assumptions specific to the National Wind Technology Center that are not applicable to other wind energy sites (e.g., blade speeds of custom research turbines that are 35% greater than conventional turbines, how often trucks drive on certain access roads, occupancy schedule of buildings near wind turbines).
  3. The article was written based on the methodology used to conduct a Hazard Analysis Review for a unique federal government wind energy test facility site with extreme wind conditions – the National Wind Technology Center.
  4. The authors used a Department of Defense (DoD) safety standard to assess risk because the National Wind Technology Center is government property; however, DoD requirements do not apply in most places.
  5. The data used in this article (e.g., failure rate for wind turbine rotors) are based on a 13-year-old report on wind turbines installed in Europe in the preceding decades. The intervening years have seen significant improvements in turbine technology including enhancements in blade reliability. Since the study concluded that the operations at the National Wind Technology Center  are safe, a more recent and accurate failure rate basis was not considered necessary. 

We think the above clarification is garbage.  Rules for wind turbine siting are supposed to be based on science.   For more than twenty-five years,  the wind industry has consistently failed to produce or to cooperate in scientific studies. We can pretty confidently say that if scientific research yields results that are not favorable to industrial wind, the public will not see it.  The MSU study and the NREL study are recent examples.

There is some irony in this current flap when we consider another well known NREL study which produced the JEDI model to assess the economic impact of wind. JEDI stands for Jobs and Economic Development Impact. ( )  When using  JEDI, NREL cautions: “Model users are encouraged to enter as much project-specific data as possible, including these inputs:

  • Construction materials and labor costs
  • Turbine, tower, blade costs, and local content information
  • Utility interconnection, engineering, land easements, and permitting costs
  • Annual operating and maintenance costs (personnel, materials, and services)
  • Tax, land lease, and financing parameters


Okay, fine but does anyone do that?  No.  Does the wind industry default to the NREL model with hypothetical inputs? Yes.  Does the whole nation use canned JEDI numbers with a few bits of local information thrown in?  Yes.

We recall a case before the OPSB in which the NREL JEDI model results were disputed. In this instance,  documents from the case reflect that, with respect to the OPSB staff person assigned to evaluate the application, the OPSB staff did not re-run the JEDI model to test the applicant’s representations because OPSB had not purchased the necessary modeling software. The record reflects the staff person:  “He has never used the [JEDI]  model, was not aware of the model’s limitations, was not familiar with its inputs, has never seen the outputs from a JEDI model, and did not even know whether it is the most accurate model to use.   He further admitted that he was not aware of the model’s limitations, because it was “out of my expertise” and he had “no knowledge of that.”   The staffer “further betrayed his ignorance of the model by stating that the JEDI statistics for economic benefits are not estimates or overly simplified assumptions, whereas the model’s creator, NREL, has cautioned that these statistics are only estimates. “  Equally unfounded is the Board’s statement  that there is no evidence that the JEDI model is unreliable or should be disregarded.  The JEDI model is “overly simplified assumptions” and is to be used only for “high level preliminary analysis.”  Yet, rather than limiting the use of JEDI to its intended function as a preliminary screening tool, the Board is relying on JEDI to justify its final decision to issue a certificate that will adversely impact residents living in 234 square miles in four counties.

Rural communities across America lack the resources to independently test the models produced by NREL.  In Ohio’s case, it has been documented that state government similarly lacks the resources to test the NREL JEDI model.   There is no reason to believe that the NREL/Larwood setback model would be any different.  Notwithstanding, the JEDI model is in constant use and its unverified output is accepted whether flawed or not.   Why should the Larwood model for setbacks be any different?  For lack of any other science-based research, we believe the Larwood recommendations are “good enough” and they are not unreasonable.  (We note that this week Hopkinton, NY adopted a setback of 5x turbine height!)

And so it continues.  The wind industry doesn’t do research or cooperate with outside researchers because they can’t afford adverse findings.  They have studiously avoided noise studies; studies associated with health impacts and infrasound; match paired studies of real estate sales and property devaluation; and site specific economic impacts.  The industry fails to disclose their assumptions in their safety models ( see Stratton dissent at ). They refuse to maintain a comprehensive, official, auditable listing of accidents and incidents and then discredit as “hearsay” the Caithness inventory which is derived from published reports ( ). They hide bird kill statistics  ( . They use a formula to measure noise that has been discredited in courts and by acoustical societies worldwide and on and on and on…….

Our reaction to the Larwood report is that it sounds reasonable using COMMON SENSE.  If using NREL’s  JEDI’s default model to project economic benefits is good enough for the wind developers and OPSB,  we will take NREL’s setback model as good enough for rural Ohioans.