Results of April – November 2017 Curtailment Evaluation, Acoustic Bat Monitoring, and Bird and Bat Carcass Surveys
The New Creek Wind Project (Project) began commercial operation in December 2016. This report summarizes results of curtailment evaluation, acoustic bat surveys, and bird and bat carcass monitoring that occurred at the Project between April and November 2017. Year 2017 monitoring was designed to evaluate the effectiveness and efficiency of multiple curtailment treatments at reducing bat mortality, characterize conditions during which bats were active at nacelle height, and yield estimates of bird and bat fatality for the Project.
Four curtailment strategies were implemented between 1 April and 15 November 2017, each designed to avoid impacts to rare bat species and substantially reduce overall bat fatality rates. A cut-in wind speed of 6.9 meters per second (m/s) with no temperature cutoff was applied from 1 April through 30 June. From 1 July through 15 October, 4 subsets of turbines were curtailed below wind speeds of 6.0 m/s or 6.9 m/s, with and without a 10-degree C temperature threshold. From 16 October to 15 November, cut-in speeds were reduced to 4.5 m/s and 5.5 m/s for 2 subsets of turbines and remained at 6.0 or 6.9 m/s for the other 2 subsets. The amount of predicted and actual curtailment varied among turbines according to the curtailment treatment to which turbines were assigned and weather conditions present at individual turbines. The four treatments implemented during 2017 prevented turbine operation for an average of between 3,910 and 5,700 10-minute periods (652–950 hours) per turbine. The predicted energy loss per turbine associated with the most restrictive curtailment treatment was approximately double that of the least restrictive curtailment treatment.
Acoustic bat detectors were deployed at 9 of 49 project turbines and recorded a total of 13,691 bat passes during 1,110 detector-nights of surveys. Seven of the detectors operated properly for most of the monitoring period, while 2 detectors malfunctioned for the majority of the study period. Hoary bats (Lasiurus cinereus) accounted for 44% of recorded bat passes that were identified to species, with eastern red (Lasiurus borealis) and silver-haired bats (Lasionycteris noctivagans) accounting for 24% and 26% of identified passes, respectively. Only 6 passes were identified as Myotis species, with activity occurring at only 3 turbines between 17 July and 23 September 2017. Seasonal patterns in activity varied among species, with hoary, eastern red, and tri-colored bats (Perimyotis subflavus) detected most frequently in August and silver-haired bats and big brown bats (Eptesicus fuscus) detected most frequently in early September. Hoary bat activity also peaked in late June. Silver-haired bat activity peaked slightly in May and early June. Although species presence varied among nights, overall timing of bat activity showed similar patterns among detectors and species groups, with most activity occurring during the first few hours past sunset.
Bat activity showed clear relationships with temperature and wind speed measured at corresponding turbine nacelles, with 92% of passes for which weather data were available (n = 11,692) occurring when temperature was greater than 10° C and 54% of passes occurring at wind speeds less than 4.5 m/s. Considering temperature and wind speed together, bat activity occurred disproportionally during calm, warm conditions, and few bat passes were recorded during times with higher wind speeds or cooler temperatures. Also, results suggested an apparent interaction between the effects of temperature and wind speed on bat activity, with activity during windy conditions occurring primarily at warmer temperatures.
Individual turbines with operating bat detectors (n = 7) were curtailed (RPM < 1) during periods when 61% to 85% of recorded bat passes were detected. Based on weather data from these same turbines, criteria for 4 curtailment treatments (e.g., timing, cut-in speed, and temperature threshold) were met during periods encompassing between 77% and 88% of recorded bat passes. As designed, the curtailment program avoided a consistent proportion of bat activity between April – September, with a higher proportion of bat activity (but lower overall number of bat passes) exposed to operation in October and November.
Bird and Bat Carcass Monitoring and Fatality Estimates
Stantec searched all 49 Project turbines at a weekly interval between 17 April and 15 November, conducting a total of 1,497 turbine searches during 152 days on-site. Individual turbines were searched on 28 – 31 occasions during the survey period. Searchers found 24 bat and 7 bird carcasses during standardized searches, and an additional 9 bat and 7 bird carcasses incidentally. Most carcasses were fresh (fatality estimated to have occurred the previous night), although searchers occasionally found carcasses estimated to be several days up to several weeks old. No federally listed bird or bat species were found during surveys. Long-distance migratory bats accounted for all bat carcasses, with hoary bats and eastern red bats accounting for roughly 45% of carcasses each and silver-haired bats accounting for the remaining amount. Bird carcasses represented 10 species, with no more than 2 of any single species found.
Ground conditions remained favorable for searching (short, sparse vegetation) throughout the monitoring period, contributing to high searcher efficiency (71% for bats, 75% for birds). Good ground visibility may have also contributed to the high scavenging rates documented throughout the monitoring period. Based on log-logistic model, which was most appropriate based on site-specific trials, carcass persistence was between 1 and 2 days for birds and bats, with an estimated 38% of bat carcasses and 31% of bird carcasses persisting through the search interval. Taking into account searcher efficiency, carcass persistence, search interval, and density-weighted area correction factor, based on the Huso estimator, we obtained an estimated overall bird fatality rate of 1.02 birds/turbine (95% CI 0.77–1.40) and an overall bat fatality rate of 2.63 bats/turbine (95% CI 1.82–3.89). Bird and bat mortality estimates did not differ significantly among the four curtailment strategies implemented during 2017 based on overlapping confidence intervals.
Each of the curtailment treatments in place between 1 April and 15 November 2017 prevented turbine operation during periods in which most acoustic bat activity (73% overall) occurred, resulting in low estimated bat fatality rates despite substantial amounts of bat activity at nacelle height. Although the 4 curtailment treatments differed substantially in the number of periods curtailed and associated energy loss, they differed less in terms of acoustic bat
activity exposed to turbine operation, and yielded low bat fatality estimates that did not differ significantly among treatments. Because estimated bat mortality did not differ between even the most and least restrictive curtailment treatments, our results suggest that all strategies in place during the 2017 monitoring period were at or above a threshold of protectiveness necessary for maintaining low risk to bats. This suggests ample room for improving the efficiency of curtailment while resulting in little if any increase in bat fatality or risk to rare bat species.
The curtailment treatments in place during the 2017 monitoring period effectively reduced exposure of bats to turbine operation and resulted in low fatality estimates. However, actual turbine curtailment did not always align with the designated conditions, due to a combination of factors related to the design of the curtailment system, such as hysteresis between cut-in and cut-out wind speed and temperature thresholds, occasional programming errors, temporary malfunction of the system, or a combination of these and other factors. Refining the implementation of curtailment during subsequent years should further improve the efficiency of the system and alignment with conditions related to risk.
Acoustic monitoring provided valuable information for comparing the relative effectiveness of multiple curtailment strategies, demonstrating that bats are quite active near the nacelle of operating turbines, but only during certain conditions (warm temperatures and relatively calm winds). Therefore, curtailment systems targeting multiple factors associated with high levels of bat activity can effectively avoid risk to bats while minimizing unnecessary energy loss during conditions with little or no bat activity. By combining acoustic monitoring with traditional carcass searches, we obtained fatality estimates associated with curtailment treatments in place during the 2017 monitoring period, but also obtained a baseline dataset of acoustic activity at nacelle height that can be used to evaluate the potential effectiveness and energy loss associated with alternative curtailment plans for future planning efforts. Our results contribute to a growing body of evidence demonstrating that bat fatality can be maintained at low levels through the use of turbine curtailment, even in regions such as the mid-Atlantic, where apparent risk to bats at wind farms is high. However, our results also suggest that there is substantial room to improve the efficiency of curtailment programs while not necessarily reducing their effectiveness. By incorporating multiple survey methods, the 2017 monitoring period provided a foundation of information upon which responsible and informed management decisions can be made in the future.