The noiseLAB Wind system provides fast Fourier transforms (FFT) and 1/3-octave analysis in 10- or 60-second periods and sorted into “wind bins.” Postprocessing modules audit, edit, and sort acquired data for computation of sound power and sound level. The modules also determine the tone content and compute the total measurement uncertainty.
The noiseLAB Wind system is primarily targeted for formal certification of wind turbines to the IEC 61400-11 standard. Therefore, the software itself and its associated development process must withstand the scrutiny of external audits. NI data acquisition products include calibration charts and microphone digitizers together with the associated acoustic analysis routines in NI LabVIEW software. They are certified to conform to Type 1 measurement accuracy, which helps the overall certification process.
It is critical to consider acoustic emissions when planning a new wind farm. Engineers must optimize the wind farm layout by factoring in meteorological conditions and terrain to minimize the noise impact on the surrounding community. More than five years ago, DELTA, located in Hørsholm, Denmark, developed a custom internal measurement system to certify major Danish and international manufacturers’ wind turbines and help engineers design wind farms that conform to relevant standards. The experiences from the first generation were incorporated into the second generation of noiseLAB Wind, which is now available for sale.
The IEC 61400-11 Ed. 3 (2012) standard provides a uniform methodology to measure the noise emissions of a wind turbine under varying wind speeds and facilitate comparison between the turbines. Wind turbine manufacturers use the standard to specify noise emissions and customers use the standard to test whether specifications are met. The tests call for measurements of sound power level, 1/3-octave band levels, and tonality over a range of wind speeds.
Developing a Measurement System Using the NI Platform
At DELTA, we used the NI 9234 dynamic signal acquisition module to acquire acoustic data and the NI 9215 analog input module to acquire analog data from wind turbines and meteorology masts. We placed the microphone on a board on the ground and supplied it with a specialized secondary windshield that complies with the IEC 61400-11 standard. The windshield reduced the wind speed at the microphone and the wind-induced noise in the microphone, thus increasing the signal-to-noise ratio, especially at low frequencies. The system provided interfaces to anemometers for wind turbines as well as interfaces to wind turbine control systems for real-time process parameter logging.
Microphones were often placed at fairly long distances from the turbine and from each other, which resulted in significant cabling and setup issues. This led to the development of a wireless connection to the microphones using the NI cDAQ-9191 1-Slot 802.11 WiFi chassis with the NI 9234 and NI 9215 modules. We met significant challenges in achieving reliable wireless transmission over distances of 150 m or more. We performed extensive tests on WiFi routers from multiple manufacturers to find reliable WiFi equipment to stream acquired data back to the data acquisition computer at distances of up to 1000 m. We established a parallel WiFi system to permit remote access of noiseLAB Wind from an iPad.
Raw data went through multiple stages of processing, which provided real-time monitoring, both by listening to a given channel or viewing selectable waveforms and their associated spectra (both 1/3 octave and FFT). In parallel, the data was synchronized and time-stamped using the technical data management streaming (TDMS) format, which offered a high-performance wave storage format that was also versatile in terms of surviving crashes. Parallel to the synchronized raw waveform storage, the FFT and 1/3-octave spectra were computed and streamed to disk, typically using integration periods that matched the wind binning period of 10 or 60 seconds. The measurement data was written serially to a simple text file, which at the end of the measurement process was converted to an appropriate spreadsheet format.
We computed a number of wind turbine and wind speed-derived parameters, including estimated hub height speed and turbine power, using preloaded power curves for the specific wind turbine. Scatter plots of live or logged parameters could also be viewed during the measurements, in which a choice of more than 20 parameters may be plotted against each other. Here, the use of the TDMS format was a big advantage because it featured the ability to perform multiple processes to safely access the same data file, even as it was being written to at high speed.
We chose LabVIEW system design software for our noiseLAB Wind software because of its overall flexibility for data acquisition, signal processing, data storage formats, and general processing capabilities.
Tel: +45 7219 4000