Could you elaborate on recommendations for how to use data from masts with corrupted sensors?

[Julia Austin 107]

Q: Could you elaborate on recommendations for how to use data from masts with corrupted sensors?

A: It is best to compare the ten-minute average of two sensors that are mounted at the same height. By looking at the difference (or ratio) between these two sensors, it can be determined whether one or both are experiencing excessive drag. (See slides 8 and 10 in the presentation for a visual depiction.) Typical data quality checks can then be used to filter the data. If unsure, it is best to consult an expert in data analysis.

[mhbullard 1]

In light of the findings presented at WindPower 2009, are your recommendations the same or do you have any further suggestions for utilizing data from affected sensors. Since it is apparently so widespread, how do you recommend we treat sensors not part of a pair at the same level but whose serial numbers fall within the affected range?

[johng 6]

Hi -

Unfortunately, we do not have any further recomendations at this time.  By looking at field data alone, there is not a method to detect the vibratory mode in sensors that are not paired.

Depending on your level of concern, you may want to consider post-calibration testing of "single" sensors from these masts.  The calibration test result will inform you about the sensor performance.

Thanks