Q2. What are the future works in "Determination of the combined vibrational and acoustic emission signature of a wind turbine gearbox and generator shaft in service as a pre-requisite for effective condition monitoring" ?
Future work is required to develop similarity analysis based on the Euclidian distance in order to evaluate similarities between the signature and future monitoring data ( figure 10 ).
Q3. How many files were distributed into the power bins?
The 813 files were distributed into 61 5kW wide power bins according to the power output in the range 0-300kW i.e. the data processing precision for the power measurements is effectively +/- 2.5kW.
Q4. How much of the cost of maintenance is incurred by wind turbines?
Amortisation of the capital cost over 20-30 years results in a cost per annum of 70% and of the total annual costs with operation and maintenance (O&M) accounting for 30%.
Q5. What is the frequency of the transmission shaft between the gearbox and the generator?
On the generator casingAs the power is increased from 0 to 300kW the rotation frequency of the transmission shaft between the gearbox and generator slightly varies from 1500 from to 1530rpm.
Q6. What was the output of the accelerometers?
The accelerometers (Sensonics Model P293-48H002) had output sensitivities of 100Mv/g over 0.4Hz -11kHz to better than 3dB, and the AE Sensors (Vallen Model VS900-RIC) hadsensitivities exceeding 100V/bar over the range 100kHz-950kHz, with a peak 22dB higher at 350kHz.
Q7. What is the frequency of the gearbox?
As the wind speed varied by a factor of 5, between 5-25mph in the reported measurements and the rotation speed of the generator shaft was almost constant, a range of rotational frequencies were involved with the gearbox components.
Q8. How much of the world’s electricity supply is provided by wind farms?
Rapid increases in the construction of wind power farms are taking place worldwide, in the highest proposed growth scenario it is estimated that by 2020 wind power could supply 2.600TWh, about 11.5-12.3% of global electricity supply, rising to 21.8% by 2030 [1, 2 and 3].
Q9. How is the probability of detecting a defect in a turbine?
The probability of detecting a defect in 1second measurement times during operation is defined by the error function assuming that Gaussian statistics apply i.e.POD = erf (VRMS(average)/{VRMS}) (1)As an example, for erf (3) i.e. a defect voltage three times the healthy turbine signature voltages, the detection probability is 0.99998.
Q10. How many times did the frequency square effect affect the sensor signals?
Assuming a constant ratio of hub speed to wind speed, the rotational frequencies of some of the gearbox components could have varied by a factor over the considered power range.
Q11. What is the probability of detecting a defect in a wind turbine?
The POD is found to be high especially at high power/wind speed, which implies high discrimination of defects -in case they appear- which will be unlikely to generate false alarms.
Q12. How much of the cost of repairs on failure is estimated?
It is estimated the replacement of components in planned repairs during scheduled maintenance downtime costs only 43% of the cost of repairs on failure [10].
Q13. What is the frequency of the filter?
So restricting the filter to this range avoids unnecessary signal processing and minimises the input noise without loss of significant information.
Q14. How much of the total downtime is caused by wind turbines?
This data shows that whilst the turbine gearbox, generator and main shaft/bearing and gearbox account for only 10% of the malfunctions they result in 53% of the total downtime.
Q15. How many times could the frequency square law affect the sensor response?
Following the frequency square law, sensor responses from any rotating circumferential defects in the gear could thus vary over the power range by a factor of 25.