Confirm memory use for flow_map with WaspGridSite site (PyWake's example ParqueFicticioSite())
Hello,
Is it expected that running flow_map() takes over 15G of memory for this example?
site = ParqueFicticioSite()
x, y = np.asarray([263655.0]),np.asarray([6505601.0]) # let's try just 1 turbine first
# wt param
wt_u = np.array([3.99, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25])
wt_p = np.array([0, 55., 185., 369., 619., 941., 1326., 1741., 2133., 2436., 2617., 2702., 2734.,
2744., 2747., 2748., 2748., 2750., 2750., 2750., 2750., 2750., 2750.])
wt_ct = np.array([0, 0.871, 0.853, 0.841, 0.841, 0.833, 0.797, 0.743, 0.635, 0.543, 0.424,
0.324, 0.258, 0.21, 0.175, 0.147, 0.126, 0.109, 0.095, 0.083, 0.074, 0.065, 0.059])
wt = WindTurbine(name="NEG-Micon 2750/92 (2750 kW)", diameter=92, hub_height=70,
powerCtFunction=PowerCtTabular(ws=wt_u, power=wt_p, power_unit='kw', ct=wt_ct))
wfm = NOJ(site, wt)
sim_res = wfm(x,y)
sim_res.flow_map(wd=0, ws=10) # a lot of memory. Will disconnect VM if VM initialized without enough memory
Here is what I am doing.
Essentially, I have process climate model data (gridded wind speeds, and Weibull parameters a and k and sector frequency (f) information at each grid point). I want to compute wake losses for a few turbine configurations and use this gridded information and not averaged values for f, a, or k. It looks like the WaspGridSite set-up looks most similar to how I want to set up my site object?
Thanks!
Edited by Cassia Cai