ggoretti / data_cleaning

python scripts for wind turbine data cleaning

Cleaning wind turbine data

This repository includes the scripts developed for cleaning wind turbine data, when wind speed and power values for each turbine in the wind farm are available.

Use this code

Function

The function that performs turbine data cleaning is clean_turbine.

An example of the application of the clean_turbine function is presented in example.py. The function powerCurve.py is also included, which draws the "wind farm" power curve using the average turbine values, namely wind speed and power.

Assumptions

Data Quality:

• consistency: data are presented in the same format (for example, measurements coming from different SCADA systems)
• uniformity: measures are specified using the same units across different systems (for example, wind speed in m/s, timestamps with same UTC offset, ...)

Variables (for each turbine):

• wind speed [m/s]
• power (normalised by rated capacity) [p.u.]

Data format

• Data: pandas.DataFrame
• Columns:
  • index: unique for all columns (pandas.DatetimeIndex)
  • wind speed = ['windSpeed_wt01', 'windSpeed_wt02', ..., 'windSpeed_wtNN']
  • power = ['power_wt01', 'power_wt02', ..., 'power_wtNN']
• Packages required: pandas, numpy, matplotlib