Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

BOOST: Microgrid Sizing using Ordinal Optimization

Mohamad Fares El Hajj Chehade, Sami Karaki

TL;DR

This work addresses the challenge of sizing PV and battery resources in residential microgrids by introducing BOOST, a two-phase OO framework that combines fast, approximate evaluations with exact MILP-based operation. Phase 1 uses a simple LP to cheaply screen many designs, while Phase 2 applies MILP to a top subset to obtain an accurate, cost-minimizing design; the method is shown to be robust to estimation noise and to outperform dynamic programming in the inner optimization. The key contributions are the two-phase OO sizing strategy, the robustness of design rankings, and the demonstration that MILP-based operation reduces LCOE to as low as $0.0884$/kWh, with a best design featuring $E_B=2.05 ext{ MWh}$ and PV size $1.27 ext{ MW}$. Overall, BOOST provides a scalable, accurate framework for integrating renewables into residential microgrids, balancing economic and environmental goals and enabling practical deployment.

Abstract

The transition to sustainable energy systems has highlighted the critical need for efficient sizing of renewable energy resources in microgrids. In particular, designing photovoltaic (PV) and battery systems to meet residential loads is challenging due to trade-offs between cost, reliability, and environmental impact. While previous studies have employed dynamic programming and heuristic techniques for microgrid sizing, these approaches often fail to balance computational efficiency and accuracy. In this work, we propose BOOST, or Battery-solar Ordinal Optimization Sizing Technique, a novel framework for optimizing the sizing of PV and battery components in microgrids. Ordinal optimization enables computationally efficient evaluations of potential designs while preserving accuracy through robust ranking of solutions. To determine the optimal operation of the system at any given time, we introduce a mixed-integer linear programming (MILP) approach, which achieves lower costs than the commonly used dynamic programming methods. Our numerical experiments demonstrate that the proposed framework identifies optimal designs that achieve a levelized cost of energy (LCOE) as low as 8.84 cents/kWh, underscoring its potential for cost-effective microgrid design. The implications of our work are significant: BOOST provides a scalable and accurate methodology for integrating renewable energy into residential microgrids, addressing economic and environmental goals simultaneously.

BOOST: Microgrid Sizing using Ordinal Optimization

TL;DR

This work addresses the challenge of sizing PV and battery resources in residential microgrids by introducing BOOST, a two-phase OO framework that combines fast, approximate evaluations with exact MILP-based operation. Phase 1 uses a simple LP to cheaply screen many designs, while Phase 2 applies MILP to a top subset to obtain an accurate, cost-minimizing design; the method is shown to be robust to estimation noise and to outperform dynamic programming in the inner optimization. The key contributions are the two-phase OO sizing strategy, the robustness of design rankings, and the demonstration that MILP-based operation reduces LCOE to as low as /kWh, with a best design featuring and PV size . Overall, BOOST provides a scalable, accurate framework for integrating renewables into residential microgrids, balancing economic and environmental goals and enabling practical deployment.

Abstract

The transition to sustainable energy systems has highlighted the critical need for efficient sizing of renewable energy resources in microgrids. In particular, designing photovoltaic (PV) and battery systems to meet residential loads is challenging due to trade-offs between cost, reliability, and environmental impact. While previous studies have employed dynamic programming and heuristic techniques for microgrid sizing, these approaches often fail to balance computational efficiency and accuracy. In this work, we propose BOOST, or Battery-solar Ordinal Optimization Sizing Technique, a novel framework for optimizing the sizing of PV and battery components in microgrids. Ordinal optimization enables computationally efficient evaluations of potential designs while preserving accuracy through robust ranking of solutions. To determine the optimal operation of the system at any given time, we introduce a mixed-integer linear programming (MILP) approach, which achieves lower costs than the commonly used dynamic programming methods. Our numerical experiments demonstrate that the proposed framework identifies optimal designs that achieve a levelized cost of energy (LCOE) as low as 8.84 cents/kWh, underscoring its potential for cost-effective microgrid design. The implications of our work are significant: BOOST provides a scalable and accurate methodology for integrating renewable energy into residential microgrids, addressing economic and environmental goals simultaneously.
Paper Structure (11 sections, 1 theorem, 5 equations, 2 figures, 2 tables, 1 algorithm)

This paper contains 11 sections, 1 theorem, 5 equations, 2 figures, 2 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Preliminaries
  3. The Inner Problem
  4. Ordinal Optimization (OO)
  5. Phase 1: The Simple Model
  6. Phase 2: The Accurate Model
  7. Numerical Experiments
  8. Experimental Setup
  9. Ordinal Optimization Results
  10. Comparison of MILP with other approaches
  11. Conclusion

Key Result

Proposition 1

Order is robust with respect to estimation noise.

Figures (2)

  • Figure 1: The 5 main components of the system.
  • Figure 2: An overview of BOOST.

Theorems & Definitions (1)

  • Proposition 1