Shared energy storage planning based on the adjustable potential of data center based on visual IOT platform

Area of Science:

• Energy Systems Engineering
• Operations Research
• Computer Science

Background:

• Decentralized energy storage in data centers suffers from low utilization and poor economic efficiency.
• Existing configurations lack coordinated planning and benefit-sharing mechanisms.

Purpose of the Study:

• To propose a shared energy storage planning method for data center groups to improve economic efficiency and utilization.
• To develop a framework that minimizes investment costs and maximizes operational benefits through collaboration.

Main Methods:

• Established a shared energy storage operation framework with capacity allocation, cost-sharing, and Nash bargaining-based profit distribution.
• Developed a two-stage stochastic optimization model considering uncertainties in renewable energy, workload, and temperature.
• Integrated a room-level energy management model with adjustable potential for workloads and HVAC systems.
• Proposed an improved L-shaped algorithm for effective and computationally efficient planning.

Main Results:

• The proposed method effectively minimizes investment costs and maximizes operational benefits for data center groups.
• Simulation results across four scenarios validated the approach's effectiveness in enhancing energy storage utilization.
• The framework ensures equitable benefit distribution among alliance members.

Conclusions:

• Shared energy storage planning offers a viable solution to the economic inefficiencies of decentralized data center energy storage.
• The developed optimization model and algorithm provide a robust tool for coordinating shared storage and alliance operations.
• Leveraging adjustable potential and inter-scenario complementarities is key to optimizing data center energy management.