This report describes the data analysis of the Middelgrund Wind Farm online collected data with the purpose of calculating the wake effects and turbulence intensities within the wind farm when maximum wake effects are present. The data are compared to themost commonly used wake model PARK implemented in Wasp 8. The PARK calculations have been converted from power to wind speed deficits and there is as substantial difference between the calculated defi-cits in wind speed between the data and the PARKmodel. However, it shall noted that it is a very crude comparison since the lowest directional sector inter-val in the PARK/Wasp mode is 10° , and all events are assumed uniformly dis-tributed over this sector. Furthermore, the model does not include thechanging direction of the wind flow through the turbine array caused by the interaction between the turbines and the flow. Finally the turbines are spaced by 2.6 rotor diameters, which is in the limit of the park model assumptions where the spac-ing isassumed to be higher than 5 rotor diameters. Taking all this assumptions and model constrains into account the Park model overestimates the wind speed deficits with up to 30 % and with an ever decreasing wind speed through the array when the wind iscoming from north. The turbulence intensity is enhanced up to 0.3 due to the wake effects. The analysis has shown that this enhancement is nearly independent of the number of turbines involved in the wake creation.