Given our initial experience with the low-frequency, impulsive noise emissions from the MOD-1 wind turbine and their impact on the surrounding community, the ability to assess the potential of interior low-frequency annoyance in homes located near wind turbine installations may be important.
Since there are currently no universally accepted metrics or descriptors for low-frequency community annoyance, we performed a limited program using volunteers to see if we could identify a method suitable for wind turbine noise applications. We electronically simulated three interior environments resulting from low-frequency acoustical loads radiated from both individual turbines and groups of upwind and downwind turbines. The written comments of the volunteers exposed to these interior stimuli were correlated with a number of descriptors which have been proposed for predicting low-frequency annoyance. These results are presented in this paper. We discuss our modifications of the highest correlated predictor to include the internal dynamic pressure effects associated with the response of residential structures to low-frequency acoustic loads.
Finally, we outline a proposed procedure for establishing both a low-frequency "figure of merit" for a particular wind turbine design and, using actual measurements, estimate the potential for annoyance to nearby communities.
Our objective in the limited experiment reported on here was to simulate a series of LF noise environments that would be likely to exist within a small room of a home (a small bedroom, for example) as a result of the LF acoustic loading caused by wind turbine emissions. Our experience has shown that interior LF annoyance is more likely to occur and be more severe in rooms with small dimensions and at least one outside wall facing the wind turbine. This was also true of the annoyance related to the gas turbine peaking generator; i.e., the most serious annoyance occurred near the sides of the houses facing the LF source. We synthesized three interior LF noise environments that would be expected as a result of the acoustic loading of a residential structure from the following kinds of emissions:
• A single, large, multi-megawatt turbine or an array of smaller turbines that are not producing periodic impulses (a periodic random source';
• A nearby single turbine operating at a shaft speed of 30 rpm and producing impulses at the blade passage frequency (a periodic impulsive source';
• An upwind array of turbines that are individually producing unsynchronized impulses at their blade passage frequencies (a random impulsive source'. In addition to these three basic environments or stimuli classes, the periodic random source was repeated but with a "pink" noise masking level of 40 dBA.