NREL - wind power datasets in the US

It is a national record of all the wind farms including features like power, wind speed, density, etc. However, it might be hard to doownload to total datasets from the NREL website, due to the fact that the total data size reaches the 2.9TB, which is not memory-friendly.

Hence, we have developed a python script to extract the specific wind data from the url.

First, we need to download pywtk package. In our settings we use anaconda prompt to download pywtk.

cd D:\pywtk  # it should be changed to your own directory of the downloaded github zip from the pywtk link
python setup.py install

Second, open our master-branch as a pycharm project and run Fetch_data.py

In this file, we can set the default url as:

WTK_URL = "https://f9g6p4cbvi.execute-api.us-west-2.amazonaws.com/prod"

Then, we can select a polygon region (Longitude and Latitude) as follows to access the wind farm data within this region.

Selected_loc = ["POLYGON((-123 42, -121.25 42,  -121.5 41.2 , -123 41.2, -123 42))",  
                "POLYGON((-114 32, -115 32,  -115 34 "  
                ", -114 34, -114 32))"]
Selected_idx = ['NW', 'SE']  
Selected_region = dict(zip(Selected_idx, Selected_loc))

In this cases, we plot 3792 wind farm data in California, with color varing with the id of each farm. (e.g., farm id: 12305)

Then, we can access the basic information by using the get_nc_data_from_url function.

for region in Selected_idx:  
    sites_selected[region] = get_3tiersites_from_wkt(Selected_region[region])

To extract the speific wnid power data from the url with corresponding timestamps and save them as csv, we can use the following operation:

for region in Selected_idx:  
    # extract the gid from the  
    gid_index = np.array(sites_selected[region]['gid'])  
    count = 0  
    for gid in gid_index:  
        count+=1  
        wind_data_ensemble[region + 'gid_' + str(gid)] = get_nc_data_from_url(WTK_URL + "/met", str(gid), start, end,  
                                                                              attributes=attributes, leap_day=leap_day,  
                                                                              utc=utc)  
        wind_data_ensemble[region + 'gid_' + str(gid)].to_csv(  
            './Data_windfarm/2013_2014/' + str(start.year) + str(end.year) + region + '_id_' + str(gid) + '.csv',  
            sep=',')  
  
        print('|| The wind farm {} is downloaded || >>>> {}/{} '.format(gid, count, len(gid_index)))