noiseLAB Wind: System Overview

Data Flow for Wind Turbine


noiseLAB Wind provides a number of open data interfaces that gives greater flexibility and interoperability with third party tools.

The above diagram shows the major data flow paths in the system, and some of the major data formats used.  In the following these will be described in detail.

Microphone Inputs

IEPE powering is provided by National Instruments NI USB 9234 and other NI DSA compatible modules.  The USB 9234 can be powered from the USB port of a laptop computer with only a minor decrease of battery life. The USB 9234 can provide power for up to four microphones. However, microphone preamps with built-in gain, such as the GRAS 26CF require more power, and hence only two of these can be connected to each USB 9234.   Analog cable lengths of  up to 200 m from microphone pre-amp to the USB 9234 are possible without signal degradation due to cable length.

Meteorology inputs

Data from Met masts can include wind speed, direction, temperature, pressure and humidity.  noiseLAB Wind can handle all of these in analog form with AC or DC voltages from 0 to +-10 V using the NI USB 9215A. The 9215A is powered from a USB port of a computer.  The 9215A samples up to 100 kS/s per channel, and the noiseLAB Wind Software provides user settable gain and offset on a per-channel basis.  In addition, for frequency proportional outputs (typically seen from anemometers), software in noiseLAB Wind converts this to wind speed units.

Selected meteorology masts with RS-232 outputs are also supported.  Contact DELTA for more details.

Power curve and correction curves

The Wind turbine power curve and wind shield correction curves as entered into the system from well-defined delimited text files.

Wind Turbine Parameters

Wind turbines can output a number of parameters, either by analog or digital interface.  These parameters can include power output, RPM, Pitch, Yaw, Nacelle Wind Speed, etc.

When these are available in analog form, up to four of these can be digitized by each NI USB9215A.  Each channel provides user settable gain and offset.

For digital outputs, DELTA provides standard RS-232 drivers for a range of Vestas and Siemens Turbines.

The turbine parameters are typically sampled at 10 samples per second, and averaged in software to give one second averages.

noiseLAB Wind Data Acquisition Software

This software runs on the data acquisition computer which receives data from the above mentioned devices via USB or RS-232 ports.  Wi-fi or cabled Ethernet can also be used to connect to the USB devices, which are placed in an Ethernet Adapter cradle to enable networking these.  To give more robust operation, each device is assigned a fixed IP address. Cabled Ethernet is recommended in uncontrolled environments such as residential areas where risk of mechanical or electrical interference are higher.

Acoustic Processing

The digitized microphone signals (24 bits resolution) are converted to 32 bit single precision floating point for internal storage.  For each microphone channel the following is computed and stored to disk in user selectable time slices (typically 10 s or 60 s): LAeq, ⅓ octave Sound Level, A-weighted ⅓ octave sound level, and FFT spectra. All of the above are averaged using linear RMS integration.

Wind Turbine and Met signal processing

The incoming signals are sampled at 10 S/s, and averaged to 1 second values, which are recorded on disk as time domain waveforms. In addition, these are processed into time slices, typically 10 or 60 seconds.

Raw waveform storage

The raw time domain waveforms are stored in the National Instruments TDMS format. This format lets your PC serve as a multi-channel “tape recorder” where high numbers of channels can be stored with individual sampling frequencies per channel.  Data is time stamped with a 128-bit time stamp. In addition, meta data related to the measurement is also stored in the recording.  Recording duration is limited only by disk size, since the sample counter in TDMS is a 64 bit integer.

The TDMS format is open, and data can be read from a broad range of software including LabVIEW, C, .net languages, MatLab, Excel, Open Office, and Diadem, a powerful data explorer package.

Result Data storage

All data is written to disk in as simple as possible a file format which only appends to disk files, thus making files more robust in terms of recovery from crashes.

Time domain data is written in the TDMS format as described above.

Acoustic results(except FTT), Met and Turbine data are written to a delimited text file in the user-selectable time slices (typically 10 or 60 s). This is a temporary file which is converted to Excel at the end of the measurement.

FFT time sliced spectra are stored in a separate delimited text file.

The text delimited files are self-documenting with clear labels and axes.

User entered data

noiseLAB Wind provides a time coherent data collection system.  However, for some applications, it may be desired to input parameters from different devices, such as autonomous data loggers.  This can be done after the measurements are complete, by taking the external data, ensuring that it is time-aligned with the data in the above spreadsheets, and copying it into the spreadsheet. Likewise, FFT spectra can be copied into the FFT spectrum file.  In these cases, the user is responsible for time alignment and calibration integrity of these data.

.wav file inputs

For round robin testing and other applications, it may be useful to input .wav files into the system.  This is done by importing the .wav files into noiseLAB Pro, and performing time slice analysis of FFT, LAeq, and ⅓ octave spectra using it, and exporting these to the appropriate files as described above. Again, time alignment and calibrated must be managed by the user.

Post-processing of data

Sound Level/Sound Power

Based on the Acoustic, met, and turbine data, the sound level and sound power is calculated by the noiseLAB Wind Sound Level/Sound Power Module according to IEC 61400-11. This module also provides validation of data by auditing potential outliers by viewing their spectra. Audio files may also be monitored in an audio monitoring utility included with the system.

Tone Analysis

noiseLAB Wind provides the Tone Analysis Module which performs the tone analysis specified by IEC 61400-11. This is based on the FFT spectra generated by the data acquisition software as well as the met parameters.

Tone analysis  to ISO 1996-2 Second Edition Annex C is provided by noiseLAB Pro. (This option is not shown on the above diagram.)


Measurement Uncertainty

The software module calculates this for all  required parameters as specified by IEC 61400-11.

Output format for report generation

All of the above modules output their results in tabular and graphic form in Excel. From here, the plots and tables can be copied and pasted into your final report. This gives high flexibility in creating reports suited for your organization, including choice of language, layout, logos, etc. while easily importing the final results in a Word compatible format.

Image | This entry was posted in Uncategorized. Bookmark the permalink.