Stochastic operation of home energy management systems including battery cycling

Abstract

The present work proposes a stochastic approach for Day-Ahead operation of Home Energy Management Systems when batteries, solar photovoltaic resources and Electric Water Heaters are considered. The optimization problem minimizes the operation costs formed by energy procurement in the wholesale market and the equivalent cycling aging cost of the batteries, and also includes the uncertainty of the PV production and the load. The complete two-stage stochastic formulation results in a Mixed-Integer Nonlinear Programming problem that is decomposed using a Competitive Swarm Optimizer to handle the calculation of the battery cycling aging cost. A Storage Disaggregation Algorithm based on Lagrangian relaxation is used to reduce the problem size and to allocate individual State of Charge for the batteries. In addition, the advantages of considering a stochastic approach are shown by means of the Value of the Stochastic Solution. This methodology has been developed in the context of the Horizon 2020 project SENSIBLE as part of the tasks related to a use case that considers an aggregator that participates in the electricity market with a portfolio of prosumers with active demand capability.

Design and analysis of an aging‐aware energy management system for islanded grids using mixed‐integer quadratic programming

“…However, some of the constraints in the problem definition may cause nonconvexity. In another study, a nonlinear implementation employing IPOPT, a software package for nonlinear optimization, is proposed by Ju et al 21 One of the solution methods is to transform the nonlinear problem into its linear counterpart, eg, using Benders' decomposition, 22,23 which can be also utilized to eliminate intractable properties of the infinite horizon optimal control problem by decomposing it into subproblems. 24 However, because of having integer variables in the problem such as switching on or off the diesel generator, the system obliges us to formulate the problem using binary variables by mixed-integer programming.…”

“…In this study, we have defined the cost of battery proportional to the degradation in battery health by neglecting the periphery (eg, housing and inverter) costs. Therefore, J battery is given by the product of battery price C battery and change in battery state of health (SOH) ΔSOH in Equation 23.…”

Robust Day-Ahead Forecasting of Household Electricity Demand and Operational Challenges

“…The platform retrieves information from the smart meters at each household through appropriate application programming interfaces (APIs). The outputs of the forecasting models are then transmitted to other applications to be used as inputs, such as Home Energy Management Systems [18]. Hereafter, we focus on the day-ahead horizon.…”

Optimal Power Dispatch of Small-Scale Standalone Microgrid Located in Colombian Territory

“…On the other hand, the battery bank exchanged 75% more power during the MILP dispatch, which produces a higher number of equivalent charge/discharge cycles in comparison with the heuristic algorithm. Such a higher number of equivalent charge/discharge cycles produce an accelerated battery degradation, which is normally expressed in terms of remaining full charge/discharge cycles [41]. Therefore, the MILP solution could be improved by including the effect of the charge/discharge cycles using the battery cost model proposed in [41].…”

“…Such a higher number of equivalent charge/discharge cycles produce an accelerated battery degradation, which is normally expressed in terms of remaining full charge/discharge cycles [41]. Therefore, the MILP solution could be improved by including the effect of the charge/discharge cycles using the battery cost model proposed in [41]. On the other hand, the battery bank exchanged 75% more power during the MILP dispatch, which produces a higher number of equivalent charge/discharge cycles in comparison with the heuristic algorithm.…”