This repository provides an optimization framework for simulating the operation of Battery Energy Storage Systems (BESS) using Mixed-Integer Linear Programming (MILP). It supports both wholesale-only arbitrage and mixed retail/wholesale arbitrage with behind-the-meter (BTM) load.

• MILP-based dispatch models using PuLP
• Supports:
  • Wholesale-Only Arbitrage
  • Mixed Arbitrage with BTM Load
• Includes:
  • Battery degradation modeling (calendar + cycling)
  • Self-discharge and throughput constraints
  • Daily cycle limits
• Post-optimization KPIs:
  • Energy flows, price spreads, average cycles, cost metrics

Install required Python packages:

pip install numpy pandas pulp

Configurable battery parameters include:

• total_capacity: Energy capacity (kWh)
• total_rated_power: Max charge/discharge power (kW)
• rte: Round-trip efficiency
• initial_soc: Initial state of charge (fraction)
• max_soc / min_soc: SoC bounds
• daily_cycles: Max daily throughput in cycles
• yearly_degradation: Annual calendar aging
• replacement_cost: Degradation cost per kWh

Optimizes BESS operation using wholesale price signals only.

Key constraints:

• State of charge (SoC) balance
• Charge/discharge power limits
• Degradation cost from throughput
• Daily throughput and cycle constraints
• Mutual exclusivity of charge/discharge

Optimizes for both retail and wholesale price opportunities, with on-site (BTM) load.

Features:

• Separate charge/discharge tracking for retail vs. wholesale
• Prioritizes serving local load before export
• Manages SoC tagged by energy source
• Additional mutual exclusivity constraints

Run the included example to simulate 24 hours:

python bess_arbitrage.py

This runs both arbitrage models with synthetic price/load data.

=== Wholesale-Only Arbitrage ===
Objective value: $-X.XX
Total charged: XXX.X kWh
Total discharged: XXX.X kWh
Average cycles: X.XX
Wholesale spread: $0.XXX/kWh

=== Mixed Arbitrage (BTM) ===
Objective value: $-X.XX
Energy served to load: XXX.X kWh
Energy exported: XXX.X kWh
Total charged: XXX.X kWh
Average cycles: X.XX
Retail spread: $0.XXX/kWh
Wholesale spread: $0.XXX/kWh

Container for all battery and degradation attributes used by both models.

• Method: create_model(prices)
• Method: extract_results(model, timesteps)

• Method: create_model(load, retail_prices, wholesale_prices)
• Method: extract_results(model, timesteps)

Calculates post-dispatch KPIs and financial indicators.