Endangered Bat

Summary:  Na Pua Makani Wind Farm would kill bats (many more bats than they assert) and their proposed mitigation would not benefit an individual bat or the Oahu bat population.  To meet State and Federal endangered species laws, they must be required to shut down the wind turbines at night to avoid killing more bats than their compensatory mitigation produces.

Wind Farm Low-Balling their Estimate of Endangered Bat Take:  The public is tired of Hawaii wind farms low-balling their calculations of anticipated mortality of endangered Hawaiian hoary bats then exceeding their incidental take permits/licenses within four or five years of their 20-plus-year operation periods and requesting amendments to up their authorized bat take.  To receive any incidental take permit/license, the applicant is supposed to address all anticipated project effects (during all years of operation) in a Habitat Conservation Plan.  Na Pua Makani Wind Farm’s Habitat Conservation Plan dramatically under-estimates their project’s anticipated bat mortality at 51 bats over 21 years by basing it on per-turbine bat take at Kahuku Wind Farm (where turbine blade rotor swept area is ½ the size of the proposed turbines (Figure 1), and where bat detections are very low, possibly due to the dry coastal location of that project) and by failing to include a buffer in case bat take is higher than anticipated.  See:  http://www.staradvertiser.com/2017/01/14/business/wind-farms-killing-more-bats-than-expected/

turbine size

Figure 1.  Illustration of the size of the large proposed Na Pua Makani Wind Turbine size in comparison to the existing Kahuku Wind Farm’s smaller turbines.

To Meet Legal Requirements, Na Pua Makani Must increase Low Wind Speed Curtailment:  Na Pua Makani Wind Farm’s proposes to minimize bat take by implementing low wind speed curtailment when wind speeds are 5 meters/second or lower at night (stopping blade spinning when winds are light and bat activity is high).  However, these relatively large, strong-flighted Hawaiian hoary bats are detected flying around wind turbines at existing North Shore wind farms up to wind speeds of 12 m/s (Figure 2, adapted from Gorresen et al 2015, Figure 19, p. 25 https://dspace.lib.hawaii.edu/bitstream/10790/2585/1/TR64_Gorresen_Bats_Final.pdf).

In the August 7 contested case hearing, the wind developer was asked if they could implement low wind speed curtailment at 6.5 m/s instead of the proposed 5 m/s and the developer confirmed it would be possible to curtail at 6.5 m/s.  Bat observations by Gorresen et al 2015 (See Figure 2 below) indicate curtailment at 6.5 m/s versus the proposed 5.0 m/s would reduce bat take by ½ (See Figure 1).  Curtailment at 6.5 m/s could reduce bat take enough to compensate for the developer’s low-balled estimate of take, in which they based their per-turbine take on turbines that had 1/2 the rotor swept area of their very large proposed turbines.

For some reason, the wildlife agencies don’t appear to have asked the developer to reduce bat take by implementing the 6.5 m/s low wind speed curtailment instead of 5.0 m/s.  Because Na Pua Makani Wind Farm has agreed they can implement low wind speed curtailment of 6.5 m/s, their project does not meet the State and Federal endangered species laws which both require the applicant to minimize and mitigate take to the maximum extent practicable.

Gorresen bat wind speed explanationsFigure 2.  The relatively large, strong-flighted Hawaiian hoary bats are detected flying around wind turbines at existing North Shore wind farms (blue curve) at average wind speeds up to 12 meters/second.  By implementing low wind speed curtailment of 6.5 m/s (red arrows) instead of the proposed 5 m/s, Na Pua Makani Wind Farm could reduce bat take by 50%.  Low wind speed curtailment at 8 m/s would almost eliminate bat take.  Graph adapted from Gorresen et al 2015, Figure 19: https://dspace.lib.hawaii.edu/bitstream/10790/2585/1/TR64_Gorresen_Bats_Final.pdf

Avoidance of Bat Take Would Require Low Wind Speed Curtailment at 12 m/s average wind speed:  Because Na Pua Manaki Wind Farm’s proposed mitigation fails to offset take of even one bat (as explained below), the wind developer must be required avoid all bat take, which would entail low wind speed curtailment when wind speed averages 12 m/s or lower.

Failure to Compensate for Bat Mortality:  Native Hawaiians, Polynesians, and the general public are fed up with Hawaii wind farms proclaiming that their take of bats will be offset by mitigation as required State’s “net benefit” requirement.  After eleven years of industrial wind farm operations in Hawaii, there is no evidence a single bat has benefited from a wind farm mitigation project and it’s possible mitigation is harming individual bats and bat populations.

To meet Federal and State endangered species law, all wind farm mortality of the endangered Hawaiian hoary bat must be compensated for with mitigation that offsets the take – the bats must be better off as a result of the project (take plus mitigation), than they would be without the project.  As it currently stands, the wind farm’s proposal to manage invasive plants and pigs in a natural area of the Koolau Mountains and bat research seems unlikely to help one single bat.  The DLNR is already managing the natural area and just installed a new fence to remove feral pigs (which create wallows that serve as mosquito breeding grounds).  These actions benefit forest birds by reducing mosquito populations.  Mosquitoes are not a primary food item for the Hawaiian hoary bat, but one would not expect the reduction in mosquitoes to benefit the bat.  The site has very few invasive plants; Na Pua Makani’s management of invasive plants for a few years would seem to do nothing (nothing) for a single bat.  Removal of ungulates is actually correlated with a reduction in bat detections at the Kahukinui Forest Reserve on Maui.  Likewise, although attaching radio-transmitters to bats to study them would not be expected to kill the bats, it would certainly not benefit the animal or the population.

Failure to Commit to Scientific Credibility Level for Monitoring Bat Take:  Instead of committing to assess endangered species mortality with an 80% level of scientific credibility (like Kahuku and Kawailoa Wind Farms have been doing for years), Na Pua Makani Wind Farm proposes to instead report the level of assurance they have not exceeded the authorized take level of 51 bats.  If the project were built, they could report their level of assurance they have not exceeded their authorized take level is 80%… or 50%, or it could be close to zero – their commitment ends at reporting the credibility level.

Detailed (probably too detailed for most readers) Assessment of Anticipated Hawaiian Hoary Bat Take at Na Pua Makani Wind Farm

Detailed Assessment of Anticipated Bat Take:  Provided Via Email July 16, 2017, to ESRC Chair Scott Fretz and ESRC member Jim Jacobi, DOFAW staff Katherine Cullison, and USGS Bat Researcher Marcos Gorresen.  I am providing this information outside of any public comment period for your reference in your work related to the assessment of take at the proposed Na Pua Makani Wind Farm.  The enclosed assessment is based on straightforward math that doesn’t require special expertise or software.  I wrote this today, Sunday, as a member of the public, I’m sending this as a member of the public from my home computer via my personal email account.  My notes are not a reflection of my employer’s work on this project – I have recused myself from official work related to wind farms proposed on the North Shore because I am a resident of the North Shore.

Calculation of Na Pua Makani’s Per-Turbine Bat Take Assessment Based on Best Available Science:  The Applicant bases their assessment of Kahuku Wind Farm’s per-turbine bat take on the wind farm’s four carcass detections without incorporating Kahuku Wind Farm’s assessment of the level of take associated with those carcass detections.  The number of carcasses detected is of limited utility in the absence of statistical analysis.  The applicant indicates that for each carcass detected at Kahuku Wind Farm, they “conservatively” assumed two were missed.  However, since curtailment went into effect at Kahuku Wind Farm in April 2012, one bat carcass has been detected and the amount of bat take the wind farm accrued during that 2012-2016 period was eight.  A comparison of the 2016 versus 2012 Kahuku Wind Farm Annual Reports indicates direct take was no higher than five bats during the 2011 through April 27, 2012 periods of operation prior to low wind speed curtailment, and current reported direct take of bats is 13 (one bat carcass has been detected during the 35 months of February through November operation since April 2012) corresponding to take of approximately 8 bats during the April 27, 2012 through July 2016 period).  The per-turbine bat mortality rate (8 bats per 35 months of operation during February through November would be 48 bats at 12 turbines over a period of 210 months (21 years of 10-month bat season) at Kahuku since curtailment began in 2012.  This would translate to take of four bats per turbine during a 21-year operational period, under low-wind speed curtailment.  For nine turbines of the same size, situated in an area with the low bat occurrence of Kahuku Wind Farm, with post-construction monitoring program of equal intensity, this would correlate to direct take of 36 bats (and total direct plus indirect take comparable to 40 adult bats) over 21 years of operation.

Application of the Kahuku Wind Farm Per-Turbine Bat Take to Na Pua Makani:  Three considerations should be accounted for in applying Kahuku Wind Farm’s per-turbine bat take to the Na Pua Makani Wind farm: 1.)  If Na Pua Makani’s post-construction monitoring carcass detection probabilities are not equal to or greater than those achieved at Kahuku Wind Farm to achieve tight statistical confidence, take assessments at Na Pua Makani will be inflated due to the reduced thoroughness of post-construction monitoring; 2.) Because the Na Pua Makani turbines will have a rotor swept area 1.8 times (almost twice the size) of the Kahuku Wind Farm’s turbines, it is reasonable to conclude the large turbines will kill 1.8 times more bats than each of the smaller Kahuku turbines.  The applicant does not appear to have highlighted their very large increase in rotor swept area when they briefed the ESRC on their increased turbine height.  The change in turbine blade length may not have been noticed by the ESRC; discussion of the larger rotor diameter is not reflected in ESRC meeting minutes.  Modification of the above per-turbine Kahuku Wind Farm take of 36 bats (direct take) and 40 bats (adults, including indirect take) to reflect the greater risk due to the larger size of the rotor swept area of the Na Pua Makani turbines results in anticipated direct take of 65 bats; total take of 72 adult bats (assuming monitoring as intensive as Kahuku’s was); and 3.) It would be reasonable to be concerned based on bat acoustic monitoring by Gorresen et al (2015) that the Na Pua Makani Wind Farm may be situated in an area that has higher bat use than that at the Kahuku Wind Farm and that the taller turbines (more similar in height to the Kawailoa turbines) may be more attractive to the Hawaiian hoary bat and may result in higher fatality rates than the smaller Kahuku wind turbines.  Consideration should be given to doubling the above Tier 1 bat direct take of 65 (total take comparable to 72 adults) to direct take of 130 bats (total take comparable to 144 bats) to assure uncertainties associated with the Na Pua Makani’s anticipated level of bat take are accounted for.  Additional details about these three aspects of assessment of anticipated Na Pua Makani Wind Farm bat take are detailed below:

  1. Post-Construction Monitoring Effort in Relation to Take Assessment: As-written, the proposed Na Pua Makani HCP Post-Construction Monitoring Plan may not achieve the same level of carcass detection probability results as those obtained by the Kahuku Wind Farm (upon which project per-turbine take is based).   It is my understanding the ESRC requires that the take level assessed to wind farms have a high degree of credibility (80%) such that there is a low likelihood (20%) that the take occurring at the wind farm is actually higher than the assessed level of take.  Kahuku and Kawailoa Wind Farms both implemented significant grading, mowing, predator control, frequent searches of large areas, and use of canine searches, often two times per week, to achieve high levels of bat carcass detection (“g” over 50 percent and in some cases higher than 80 percent) in their first four years of operation (Kahuku 2014 Annual Report, Appendix 1 page numbered 3 and Kawailoa 2016 Annual Report, p. A2).  The rate of bat carcass detection dropped to 40 percent in 2015 when these two existing wind farms were allowed to enter a period of reduced monitoring effort.  Na Pua Makani Wind Farm’s post-construction monitoring program would need to be as intensive as that of the Kahuku Wind Farm to achieve the relatively tight statistical confidence in assessed level of take achieved by Kahuku.  Variability and uncertainty that would be associated with lower levels of post-construction monitoring effort could result in significant inflation of the 80% credible level of take assessed to the project.  If post-construction monitoring at Na Pua Makani is not as thorough as post-construction monitoring at Kahuku Wind Farm, the statistical uncertainty associated with the assessment of bat take at Na Pua Makani may, in itself, be enough to cause the assessed level of take to exceed the authorized take level.

Bat take assessments from the adjacent KWPI and KWPII Wind Farms inform this post-construction monitoring conern because the level of engineering and vegetation management at Na Pua Makani may not as great as the extent of management of large searchable areas at Kahuku Wind Farm.  At KWPII, the average annual per-turbine bat take assessed (0.286 bats/turbine/year, based on KWPII 2016 Annual Report, Appendix 6) is more than double the per-turbine average annual rate of bat take reported at KWPI (0.138 bats/turbine/year, based on KWPI 2016 Annual Report, Appendix 6).  The primary reason for this difference may be the significantly lower carcass detection probability at KWPII due to the greater extent of steep terrain near the turbines that was not graded or engineered for carcass searching at the KWPII site.  Because this difference in post-construction monitoring was not adequately anticipated in advance, even though the KWPII wind farm was authorized based on results from six years of KWPI operation of the adjacent KWPI wind farm, KWPII exceeded their authorized take of 9 adults and five juveniles after fewer than five years of operation.  Consideration should be given to doubling the above Tier 1 bat direct take of 65 (total take comparable to 72 adults) to direct take of 130 bats (total take comparable to 144 bats) to assure uncertainties associated with the Na Pua Makani’s anticipated level of bat take are accounted for.

As-written, it’s not possible to confirm Na Pua Makani’s post-construction monitoring carcass detection rates will be as high as those achieved by Kahuku Wind Farm.  This lack of a specific commitment within the HCP document to a high level of carcass detection, in conjunction with the relatively low level of requested bat take, should trigger the ESRC or the Board to include an incidental take license condition that the wind developer take measures necessary to achieve a bat carcass detection rate (“g”) above 50 percent during the January through November period of higher bat occurrence, for the first four years of operation, followed by bat carcass detection rates above 40 percent in subsequent years of reduced monitoring.  This requirement would reduce the likelihood the wind farm would exceed their authorized level of bat take simply because their post-construction monitoring was not as thorough as that of the Kahuku Wind Farm upon which their anticipated take levels were based.

  1. Incorporation of Na Pua Makani’s Larger Rotor Diameter: Na Pua Makani’s larger turbines, with rotor swept areas 1.8 times (almost two times) the size of the Kahuku turbine rotor swept area, would have to kill the same number of bats, per-turbine, as the smaller Kahuku turbines for the credible level of direct take to remain below 36 bats (total take comparable to 40 adult bats).  Kahuku Wind Farm turbine blades have a diameter of 314 feet with a 77,437 square foot rotor swept area; the proposed Na Pua Makani turbine blades have a diameter of 427 feet (130 meters) with a corresponding rotor swept area of 143,201 square feet per turbine.  Hawaiian hoary bat behavior in the vicinity of turbine blades indicates the Hawaiian hoary bats traverse the turbine rotor swept area (Gorresen et al 2015), so risk of bat strike is directly associated with the size of the air space occupied by the spinning rotors.
  2. Hawaiian hoary bat occurrence at Na Pua Makani Wind Farm may be higher than Bat Occurrence at Kahuku Wind Farm. The Na Pua Makani HCP does not appear to incorporate acoustic monitoring data published in 2015.  Unless Na Pua Makani has conducted acoustic monitoring to inform bat occurrence at the proposed wind farm site, it seems reasonable to be concerned bat occurrence at the proposed wind farm site may be higher than bat occurrence at the Kahuku Wind Farm (Figure 1).

gorresen with wind farms

Figure 1.  Modified from Gorresen et al (2015) Figure 4 Observed nightly bat detection probability by survey site as averaged for the 1-year acoustic survey period (white and blue are low probability, green, yellow, and red are higher bat occurrence sites, wind farm locations and labels added in purple).

Per-turbine bat take at Kawailoa Wind Farm, where low wind speed curtailment has been in effect during all periods of operation, is slightly more than double the per-turbine bat take assessed during the period of low wind speed curtailment at Kahuku Wind Farm.  This higher level of take may be, in part, related to the higher level of bat detections documented by Gorresen et al (2015, see Figure 1) in the Kawailoa Wind Farm area.  It’s possible, based on results of Gorresen et al (2015) (see Figure 1), bat occurrence at the Kahuku Wind Farm is much lower than bat occurrence at the Na Pua Makani site and it seems advisable that a Tier 2 of bat take incorporate a buffer against uncertainties related to the potential for bat density to be higher at the Na Pua Makani Wind Farm than at Kahuku Wind Farm.


Gorresen, P. Marcos, Paul M. Cryan, Manuela M. Huso, Cris D. Hein, Michael R. Schirmacher, Jessica A. Johnson, Kristina M. Montoya-Aiona, Kevin W. Brinck, and Frank J. Bonaccorso.  2015. Behavior of the Hawaiian hoary bat at wind turbines and its distribution across the north Ko‘olau Mountains, O‘ahu.  Hawai`i Cooperative Studies Unit University of Hawai`i at Hilo 200 W. Kawili St. Hilo, HI 96720 (808) 933-0706. https://www.researchgate.net/publication/277891159 (January 2015) and http://pubs.usgs.gov/circ/1367/ May 2015.  68 pp.

HCPs and Annual Reports, Cited by Wind Farm Name, are Available Online at:  http://dlnr.hawaii.gov/wildlife/hcp/approved-hcps/