How electric load responds to temperature and weather variables in Texas

I demonstrate how electric load in Texas responds to temperature and other meteorological variables. Using flexible semiparametric models, I estimate the response functions that disentangle baseline responses and marginal contributions of long-term trends and weather conditions, both conditioning on temperature. The benchmark model finds significant time-of-day variations in the level, balance point, and slope of the demand curves. The primary estimation corroborates the nuanced interplay of weather conditions and behavior patterns. In central Texas, for example, humidity intensifies perceived heat whence cooling demand, whose maximum contribution is over 10% in afternoon peak hours. Wind mostly promotes cooling, drives up load by 7% in winter nights, and lowers it on certain occasions of spring and fall. (The magnitude is with respect to one standard deviation rise.) Humidity’s effect dwindles during nighttime rest, yet wind’s and sky cover’s are muted during business hours. Summer load generally has smaller sensitivities than shoulder seasons, probably because it is near the extensive margin limit. These findings improve upon the extant literature by connecting underlying mechanisms to empirical results about weather sensitivity, conveying precise operational implications for margin refinement and demand-side management.