Hoosac Wind Sound Monitoring Report

Observations

 by Chris Kapsambelis

Introduction

The sound monitoring for the Hoosac Wind project to determine compliance with MassDEP noise pollution regulations is incomplete and fails to show compliance conclusively. Additional test monitoring is needed before compliance can be determined with confidence.

The data contained in this report indicates there are numerous wind conditions that were not tested–most notably conditions capable of generating enough noise pollution to violate MassDEP policy.

Of the thirteen monitoring periods only one occurred when wind conditions were favorable for worst case noise generation. The wind was at 11 meters per second or 24.6 miles per hour (mph) from the West Northwest direction at the Moores South location. The 12 other monitoring periods do not qualify for valid data collection for one reason or another, including: rain, power failure, wind speed not within limits, wind from the wrong direction.

In spite of the less than ideal conditions for worst case monitoring, there were four instances exceeding the MassDEP 10 dB(A) above ambient limit, one as high as 13.4 dB(A). These are dismissed primarily because of low ambient conditions and a wind speed differential between ambient (with turbines turned off) and impact (as monitored with turbines on) greater than 2 meters/second (4.5 mph). The low ambient condition is probably due to wind shear, a weather phenomenon where the wind at turbine hub height is strong enough to create turbine noise, while at ground level no  wind is present to increase ambient sound. These conditions are not unusual enough to be discarded.

A good explanation of wind shear can be found in “Neglect of Wind Shear in Assessing  Long Range Propagation of Wind Turbine Noise.”

Three of these violations occurred during Monitoring Period 3 where the wind speed differential is reported at 1.5 and 1.8 meters/second (3.4 and 4 mph). MassDEP protocols recommend to discard readings when wind speed changes more than 2 meters/second (4.5 mph). While there might be occasions when the wind differential could exceed that rate during this period, it is not likely that it will lower the ambient reading to the extent needed for such a gross violation. The wind shear is the more likely explanation.

If one is to accept that the four violations (readings above 10 dB(A) over ambient) were caused by extraneous weather conditions, then some allowance must be made for regarding as violations the eight instances where the increase above ambient is reported to be between 9 and 10 dB(A). If weather conditions can result in a 3.4 dB(A) error when readings exceed 10 dB(A), then violations could exist at sound levels as low as 7 dB(A).

General

Data for this report was collected on 13 sample periods between April 3, and April 22, 2013. They are listed on Table 5 in the report (shown here. Notice that at Crum Hill no testing was done with Easterly winds).

The stated purpose of MassDEP testing protocols is to capture sampling data under worst case conditions that might result in violations. MassDEP recognizes that these conditions occur at night when ambient noise levels are at their lowest. Only four of the sample periods occurred at night.

Another consideration for worst case is distance from the nearest turbine. The Crum Hill dwellings range from 570 to 794 meters (1,870 to 2,605 feet) from the nearest turbine. The East Road test locations are about three times the distance, so the noise level is expected to be far less. Tilda Hill North and Tilda Hill South are located to the West of Crum Hill and need Easterly winds for testing, while Moores North is to the East, and Moores South is South-Southeast of Crum Hill and need Westerly winds for testing.

The DEP demands that measurements be done under worst case conditions. Of the thirteen sample periods, only Monitoring Period 4 satisfies the need for nighttime sampling where the wind was at 11 meters/second (24.6 mph) from the West Northwest.

Most of the sampling events failed to test relevant conditions. From a worst case point of view, the East Road test locations can be discounted due to distance 1,530 meters (5,020 feet). From a wind direction perspective, the Moores North and Moores South test locations were at close distance and downwind. The Tilda Hill test locations were upwind and fail to qualify for worst case testing conditions.

It is commonly understood that worst case noise conditions occur with wind speeds higher than 8 meters/second (17.9 mph), and wind turbine maximum sound power, together with electrical power output, peaks at wind speeds above 14 meters/second (31.3 mph).

The most likely location to exceed the 10 dB(A) limit was Moores North. Data from that sampling instance shows an increase of 9.3 dB(A) above ambient, but the data is tainted due to rain and must be discounted. Data from the Moores South location shows an increase of 9.4 dB(A) above ambient.  The charted data in Figure 18 shows more than half the data points exceed 10 dB(A) above ambient. This contradicts the statements made in the text of the report. In addition, with an instrumentation accuracy that can be off by +/- 1 dB(A), many more data points could achieve values that exceed the DEP limit.

Another consideration for worst case is full energy production. From a comparison of wind speeds and power output, it appears that the turbines were not running at full capacity during these sampling episodes.

Conclusion

Of the 13 monitoring periods of data samples only one, Monitoring Period 4, qualifies for worst case conditions at night. Of the six sampling locations only two were located downwind where noise is expected to be at maximum. Of the two qualified sampling locations, data from Moores North on this occasion was contaminated by rain and must be discarded. This leaves only one instance where conditions were acceptable for sampling data that could result in a violation, and that is at Moores South where data sampling shows an increase over ambient range from 7.9 to 9.4 dB(A).
Sound Level Monitoring conducted by RSG is incomplete, and more testing is needed at all locations under anticipated worst case conditions before any conclusion that the noise pollution from the Hoosac wind turbines complies with MassDEP regulations.
The following arguments can be made:

• Margin of error–The instrumentation is accurate to +/- 1 dB(A). This means that if the one worst case instance is repeated under similar conditions up to eight more violations are almost certain to occur.
• Weather interference–If the sampling were repeated for Monitoring Period 4 under the same conditions without rain interference, violations are very likely to occur at Moores North. That location is closer to the turbines and more directly downwind than Moores South (which came very close to exceeding MassDEP requirements).
• Reduced sound mode–Wind speed was reported at 11 meters/second (24.6 mph). From the GE power curve, power output should be 1,359 KW. The reported output is 1,097 KW. This indicates that the turbines were running at some 20% reduced power which indicates that the wind turbine sound power was below normal generation. It is very likely that normal wind turbine operation would result in violations.
• Effect of wind direction–Because no sampling was conducted with Easterly winds at Crum Hill, more sampling periods are needed with easterly winds to rule out any possibility of violations at Tilda Hill North and Tilda Hill South.
• Effect of wind speed–Because no nighttime sampling was done at Bakke with Easterly winds at speeds above 8 meters/second (17.9 mph), additional sample monitoring periods need to be conducted to rule out any possibility of violation at East Road North and East Road South.