Late spring housekeeping and upcoming API changes

Published: 03 Jun 2019

Additional wind power curves

We have updated the power curves that are available for simulating wind farms. Twenty new power curves are available, mainly covering the newest and largest turbines available. These include the 8 and 9 MW turbines from Vestas with 164m blades, 5 MW turbines from GE and the 140m bladed Goldwind GW140. Big thanks to Malte Jansen for helping to compile manufacturer data.

A full list of the new turbine models is:

  • Gamesa G128 5000
  • GE 3.2 103
  • GE 3.2 130
  • GE 3.4 130
  • GE 3.8 130
  • GE 5.3 158
  • GE 5.5 158
  • Goldwind GW109 2500
  • Goldwind GW121 2500
  • Goldwind GW140 3000
  • Goldwind GW140 3400
  • Siemens Gamesa SG 4.5 145
  • Siemens SWT 3.6 130
  • Siemens SWT 4.1 142
  • Siemens SWT 4.3 130
  • Vestas V117 4000
  • Vestas V136 4000
  • Vestas V150 4000
  • Vestas V164 8000
  • Vestas V164 9500

Upcoming API changes

To simplify API use across all types of data, we are going to introduce changes to the CSV data returned by the API. This is a pre-notification before this change goes live later this summer.

We will return all CSV files with three comment lines prefixed by #, followed by a single CSV column header line, then the data. Right now, the country-level files have two comment lines, while point API data has a single comment line and a two-line column header, making it more difficult than necessary to deal with the data.

The three comment lines will be structured the same way across all data:

  • Line 1: Title and dataset details, as well as license and citation information.
  • Line 2: Human-readable information on units in each data column.
  • Line 3: Machine-readable JSON metadata string, including units, and if applicable, parameters used when running the points API model.

All three lines are prefixed with a # character. Most libraries make it easy to eliminate leading comment lines. For example, in Python with pandas such files can be read by df = pd.read_csv('filename.csv', index_col=0, skiprows=3). In R, they can be read by df = read.csv('filename.csv', skip=3).

The file format has not yet changed, so you do not need to do anything yet – but this is an advance warning that if you use the API, you will need to change your scripts slightly once these API changes go online later in summer.

Erroneous labelling of data source for the NUTS-level PV data

We have recently discovered that the NUTS-level PV data in the country API was erroneously labelled as SARAH, whereas in fact, for consistency it was based on MERRA-2, just like the NUTS-level wind data.

This has been corrected: all NUTS-level data is now correctly labelled as MERRA-2. We apologise for any confusion this mislabelling may have caused! Special thanks to Martin Rüdisüli for bringing this to our attention.