Non-Wire Alternatives to Capacity Expansion

Tractable and Robust Modeling of Building Flexibility Using Coarse Data

“…Let BLSEF(β i ) denote a function that takes the scalar β i ∈ (0, 1) and solves problem (5) and outputs the optimal values of a c,t a c,t,i , b c,t,i , b c,t,i , J A t,i , and calculates the percentage of training measurements that are higher than the upper estimate or lower than the lower estimate π out t,i . The percentage of out of prediction band measurements is calculated as Now we describe the BLSE algorithm (see Algorithm 1 below).…”

“…, 1}. The parameter M is an integer greater than 1 to be selected by the modeler 5 . The outputs of the BLSE algorithm are the trained parameters of θL k,t and θU k,t .…”

“…In this new environment, TLCs can provide flexibility from the demand-side by altering their consumption to accommodate power variations. Some of the benefits of increased power system flexibility include deferral of infrastructure investments [4], [5], increase of renewable energy hosting capacity [6], increased economic efficiency [7], and others (see, e.g., [8], [9] and the references therein). However, to fully harvest the flexibility of TCLs, challenges still remain: developing appropriate building models, aggregating those models for large-scale implementation, data privacy, state estimation of flexible loads, among others [9]- [11].…”