The Kaheawa II Wind Farm and the Fate of the Hawaiian Hoary Bat


Kaheawa II is the lower string of wind turbines on the mountain above Maʻalaea. It became operational in July 2012 and provides about 21 MW of power to the grid. A Habitat Conservation Plan (HCP) committed the wind farm to not “taking” (kill, maim, harass) more than 11 Hawaiian Hoary Bats (ʻOpeʻapeʻa) within a 20-year contract period. However, that number has been estimated to already have been taken in seven years; the company has provided a revised Environmental Impact Statement (EIS) and HCP. The new documents propose increasing the take of bats from 11 to 38 adults during the contract period. It also proposes increasing the allowed take of Nēnē from 30 to 44 adults. Both the ʻOpeʻapeʻa and the Nēnē are protected as endangered species.

We still lack information on the total number of bats on Maui, their distribution and their population trend. This is partially due to insufficient research, partly to ineffective detection technology. The most commonly used technology for bat detection is acoustic detection of their echo-location sounds. However, that only works when the bat is echo-locating and the microphone is sufficiently oriented towards the bat. Comparison to thermal imaging has found that only 8% of bats present are detected. Further, there is no way to tell if an acoustic detector is registering a single bat making multiple echo-location sounds or multiple bats.  It is important to invest in increasing our knowledge of the bat population on Maui and what the effects of mitigation efforts on that population have been. For this reason, we support research funding as a partial mitigation. Without knowing what the bat population and trend is, it is difficult to say if the taking of an additional 27 adult bats at Kaheawa II (beyond the current HCP limit of 11) through the current 20-year period ending in 2032 is an existential threat to the endangered species on Maui or to a subgroup of it. The death of an adult bat may lead to the death of its juvenile offspring as well, and it may have ripple effects on genetic diversity and resilience.

Quantifying the Loss

The Kaheawa II plan “includes searches of roads and graded pads that occur within a 70-meter (m) radius from each turbine every 7 days. Searches are primarily conducted by a canine search team, with visual searchers conducting about 14% of searches per year.

That means that far from all casualties are detected. Scavengers may remove bats after an event but before the weekly search. A bat may be slung beyond 70 meters. A statistical model attempts to compensate for the missed casualties by extrapolating from the number found. Given how far off the original estimates of bat fatalities were, how inefficient the bat detection technology is and how limited our knowledge is of island-wide and local bat populations, Kaheawa II must increase the confidence in its take numbers.


Besides research, the proposed external mitigation is to increase the amount of restored forest habitat thought to be preferred by the Hawaiian Hoary Bat. Preference is to be given to areas otherwise threatened with development or other changes that would make them less hospitable to the bat. A selected area must be known to already host bats (through acoustical or other detection methods). Bats have been detected all over the island. On Hawaiʻi Island they have been detected from the shoreline to 11,800 feet elevation on Mauna Loa. They have a varied night-time foraging area and range, while the areas where they roost during the day and where they pup are thought to be much more limited, typically to restored/established forests. Preserving their preferred roosting habitats is therefor more important from the standpoint of resiliency of the species than managing their foraging range.

There is no data on the effects of habitat restoration on the bat population. As a matter of fact, it was found at Kahikinui that bat detection frequency decreased after fencing for ungulate (deer, pigs, goats) removal, likely due to less insects which had been attracted by ungulate dung. At this point, population compensation through reforestation must be considered theoretical. This is very different from the situation with Nēnē, where fledglings can be counted in protected areas.

The only operational mitigation known to reduce fatalities is to not run the wind turbines below a cutoff wind speed threshold from sunset to sunrise, or not run them at all during that time – Low Wind Speed Curtailment (LWSC). The bats are able to detect and avoid the turbine blades if they are spinning rapidly. Data from mainland studies indicates that raising the cutoff threshold from 5.0 meters/second wind speed (as proposed by Auwahi and by Kaheawa II for part of the year) to 6.9 meters/second is effective in this regard, while raising it additionally has little or no effect.

The Auwahi wind farm HCP says that the impact of LWSC regimes from studies on the mainland suggest a reduction in bat take ranging from 10 to 92 percent through increasing the cutoff speed, but that there is little or no benefit above 6.9 meters/second:

Bats are detected year round at the wind farms but more often August-October. The following chart is from the Kaheawa II HCP. The highest rate was in September 2015 when bat activity was detected on 58% of the nights.

Comments to the Kaheawa II EIS

1. Multiple non-contiguous habitat restoration areas

The EIS says “Wildfires can cause direct loss of adult bats and dependent young that are unable to escape a forest fire.” A catastrophic fire in an area heavily used for roosting and pupping could dramatically affect the total bat population and the options for species recovery. Sierra Club would like to see appropriate habitat restoration for the ʻOpeʻapeʻa in at least three non-contiguous areas to reduce that risk.

2. Increase the confidence in the take numbers

If some bat fatalities are not detected, we may be underestimating the actual take at any given wind farm. Sierra Club supports the recommendation of the wildlife agencies to expand the buffer zone searched for carcasses by 20%.

Similarly, we propose increasing the frequency of searches for carcasses to once every two days (instead of once/week) for at least a year to see if the detection rate changes.

3. Reduce the fatalities by raising the wind speed cutoff

Based on the mainland studies referenced by the Auwahi revised HCP, Sierra Club proposes a 6.9 meters/second cutoff for all wind farms from 30 minutes before sunset to 30 minutes after sunset year round. The Kaheawa II HCP proposes a nightly cutoff of 5.5 meters/second from February 15 through December 15 and 5.0 the rest of the year. Curtailment would be extended from December 15 to February 15 if fatalities occur outside the proposed curtailment period. It claims that “There are no studies to date that test whether mortality rates decrease significantly when LWSC is raised from 5.5 m/s to 6.5 m/s.“, despite the studies reported in the Auwahi HCP. However, it also says that “increasing curtailment from 5.5 m/s to 6.5 m/s would reduce renewable energy generation from the Project by approximately 328 megawatt hours (MWh) annually (or 0.47% of 70,000 MWh assumed to be produced annually)“. That is a very small price to pay for reducing or possibly eliminating the bat fatalities.

4. Monitor the effects of operational mitigation and increase curtailment if necessary

The proposed increase in take is largely a projection of continued fatalities at the level we have seen the past seven years, which is much higher than originally anticipated. If this already increased take is exceeded, the HCP says:

Once the permittee and/or wildlife agencies have determined the observed take is exceeding the permit year trigger, the appropriate minimization technique determined in consultation with the wildlife agencies would be implemented immediately if minimization includes just a change in wind turbines operation.
Minimization will include any or any combination of the following:
1. a higher level of Low Wind Speed Curtailment if additional research demonstrates a higher
likelihood of success than does current research,
2. periods of complete cessation of operations during the night (such as during the first 2 hours
of the night or during annual periods of highest activity, for example),
3. implementing deterrents that have been proven to reduce fatality rates on at least 50% of the
wind turbines (with the highest bat detection and/or fatality rates),
4. implementing “early-warning” systems on at least 50% of the wind turbines (with the highest
bat detection and/or fatality rates) that detect the presence of bats and shutting down at least
50% of the wind turbines (with the highest bat detection and/or fatality rates) for at least 15
minutes (assuming no additional bat activity is detected),
5. or a not yet identified option.

That is unsatisfactory – requiring that additional research show higher success rates for low wind speed curtailment than current research; they should go with the current research if there is no better research at that time. The rest is speculative and non-committal.

There are no proven bat deterrent technologies yet, although a wind farm on Oʻahu will use a new technology for evaluation as a pilot project.

Habitat restoration is valuable and should be part of the plan, but it cannot (at this time) be demonstrated to compensate for a single bat lost to the wind turbines, much less 38 of them.

Sierra Club feels that the company must commit to curtailing operation to the extent required to reduce the observed take rate so as to not risk jeopardizing the survival of this unique Hawaiian animal, to the point of full night-time curtailment if necessary. New technology such as bat deterrents may help, but the calculated total take should not go to 38 before taking steps that are known to save bat lives. If the tier 1 and 2 mitigation steps do not reduce the observed rate by 50% from the average of the last three years (the basis for the proposed new rates), additional steps should be taken to reduce the take (increase the cutoff rate, not run the turbines at night at all).