voltage and current measurements

you talk about voltage and current measurements before and after grid cutting in. Are LHD using DFIG converter?

A DFIG converter has to wait until drive train speed and grid angel has the same speed and the speed difference is the possible slip range of the DFIG converter. Then the converter synchronize the generator voltage  to the grid voltage with a phase shift below 1° .Then the converter close the breaker itself to connect the generator to the grid. A full scale converter in comparison fully decouple the generator from the grid so the grid side converter is connected to the grid 100% time and can produce reactive power Q even when no wind.

On grid side there is a strong relationship between voltage phase and current phases. The phase shift between voltage and current can be measurement directly by detection of voltage zero cross and current zero cross. The result phi is the phase shift angle between fundamental voltage and current. When you calculating the RMS values of voltage V and current A and active power P and apparent power S you can alternative calculate the power factor PF which is the phase shift angle between real voltage and currents (including harmonics). To measure phi in no load or low load situations is often undefined because current zero cross is difficult to detect (current amplitude near noise level).

The LDH paper stated phi measurements before and after grid connection. When a DFIG converter is used then there is before and after grid connection surely a no load situation. When using the full scale converter than this high phi could come from pushing reactive power Q stationary into the grid (Q=S*cos(phi) with phi=70°. My opinion that the power analyzer try to measure phi in a no load situation.

To your question.

The user calibration phase offset is only necessary if the current transformer CT produces a significant phase shift at 50Hz (compare to the voltage transducers). The phase shift error can be identified by connecting a pure resistor load or using a refer