Correlation effects in the field classification of ground based remote wind sensors
A classification scheme for remote sensing devices (RSDs) is included in the draft revision of IEC 61400-12-1 "Power Performance Measurements of Electricity Producing Wind Turbines" to enable the traceable assessment of uncertainty in RSD measurements. An assumption of the classification scheme is that the environmental variables (EVs) against which an RSD is classified are uncorrelated. The standard acknowledges that this assumption is not always valid and mitigation is permitted. A method for mitigation is however not included. The effect of correlation between environmental variables in the classification scheme is presented along with a method for mitigation that removes the duplicate uncertainty components caused resulting in a lower and it is proposed more representative uncertainty assessment for the RSD. Correlation between environmental variables in the classification test has the potential to produce increasingly unrepresentative uncertainty assessments in the verification and application tests following the standard as the wind conditions across the tests diverge. The impact of correlation between EVs in terms of the maximum theoretical total standard uncertainty (TSU) for the RSD is therefore also investigated. An accuracy class result in the range 3 - 7 is obtained for the lidar after mitigation with associated uncertainty in horizontal wind speed of around 2%. Consistent results are obtained across 44 verification tests and 20 application tests carried out using the classification. These place wind speed total standard uncertainty in the range 2-3% across the wind speed ranges tested in non-complex terrain.