This package is currently under development and only has a subset of capabilities of the REopt Lite model used in the REopt Lite API. For example, the Wind model, tiered electric utility tariffs, and piecewise linear cost curves are not yet modeled in this code. However this code is easier to use than the API (only dependencies are Julia and a solver) and has a novel model for uncertain outages.
The full details of the model will be published in Laws et al. 2020, Cost-Optimal Microgrid Planning Under Uncertain Grid Reliability, [Draft]. In brief, the model is set up to minimize the maximum expected outage cost (while minimizing the lifecycle cost of energy including utility tariff costs), where the maximum is taken over outage start times, and the expectation is taken over outage durations.
Evaluating only PV
and Storage
requires a linear program solver. Adding a generator and/or multiple outages makes the problem mixed-integer linear, and thus requires a MILP solver. See https://jump.dev/JuMP.jl/stable/installation/ for a list of solvers. The REopt Lite package has been tested with Xpress
and CPLEX
.
using Xpress
using JuMP
using REoptLite
m = Model(Xpress.Optimizer)
results = run_reopt(m, "path/to/scenario.json")
See the test/scenarios
directory for examples of scenario.json
.