Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Stochastic Long-Term Joint Decarbonization Planning for Power Systems and Data Centers: A Case Study in PJM

Zhentong Shao, Nanpeng Yu, Daniel Wong

TL;DR

The paper tackles the challenge of decarbonizing power systems and data centers through a dynamic, long-horizon co-optimization. It formulates a two-stage, 15-year stochastic program that jointly plans data-center siting and capacity with power-system expansion, including both operational and embodied emissions, and solves it via an enhanced Benders decomposition on the PJM system. The approach yields up to 55 GW of peak data-center demand with significant cost and emission reductions compared to separate planning, while embodied carbon considerations drive 25% more renewable deployment and substantial operational-emission savings. This lifecycle, market-relevant framework provides actionable insights for ISOs, policymakers, and utilities to strategically locate data centers and grid investments, promoting deeper decarbonization under uncertainty.

Abstract

With the rapid growth of artificial intelligence (AI) and cloud services, data centers have become critical infrastructures driving digital economies, with increasing energy demand heightening concerns over electricity use and carbon emissions, emphasizing the need for carbon-aware infrastructure planning. Most studies assume static power systems, focus only on operational emissions, and overlook co-optimization. This paper proposes a dynamic joint planning framework that co-optimizes long-term data center and power system development over 15 years. The model determines siting, capacity, and type of data centers alongside power generation expansion, storage deployment, and retirements, accounting for both operational and embodied emissions. To handle multi-scale uncertainty, a large-scale two-stage stochastic program is formulated and solved via an enhanced Benders decomposition. Applied to the PJM Interconnection, with curated datasets released on GitHub, results show the system can support up to 55 GW peak data center demand, with Virginia (DOM) and Northern Illinois (ComEd) as optimal hosts. Compared to non-joint planning, the framework cuts investment cost by 12.6%, operational cost by 8.25%, and emissions by 5.63%. Including lifecycle emissions further raises renewable deployment by 25.5%, highlighting embodied carbon's role in deeper decarbonization.

Stochastic Long-Term Joint Decarbonization Planning for Power Systems and Data Centers: A Case Study in PJM

TL;DR

The paper tackles the challenge of decarbonizing power systems and data centers through a dynamic, long-horizon co-optimization. It formulates a two-stage, 15-year stochastic program that jointly plans data-center siting and capacity with power-system expansion, including both operational and embodied emissions, and solves it via an enhanced Benders decomposition on the PJM system. The approach yields up to 55 GW of peak data-center demand with significant cost and emission reductions compared to separate planning, while embodied carbon considerations drive 25% more renewable deployment and substantial operational-emission savings. This lifecycle, market-relevant framework provides actionable insights for ISOs, policymakers, and utilities to strategically locate data centers and grid investments, promoting deeper decarbonization under uncertainty.

Abstract

With the rapid growth of artificial intelligence (AI) and cloud services, data centers have become critical infrastructures driving digital economies, with increasing energy demand heightening concerns over electricity use and carbon emissions, emphasizing the need for carbon-aware infrastructure planning. Most studies assume static power systems, focus only on operational emissions, and overlook co-optimization. This paper proposes a dynamic joint planning framework that co-optimizes long-term data center and power system development over 15 years. The model determines siting, capacity, and type of data centers alongside power generation expansion, storage deployment, and retirements, accounting for both operational and embodied emissions. To handle multi-scale uncertainty, a large-scale two-stage stochastic program is formulated and solved via an enhanced Benders decomposition. Applied to the PJM Interconnection, with curated datasets released on GitHub, results show the system can support up to 55 GW peak data center demand, with Virginia (DOM) and Northern Illinois (ComEd) as optimal hosts. Compared to non-joint planning, the framework cuts investment cost by 12.6%, operational cost by 8.25%, and emissions by 5.63%. Including lifecycle emissions further raises renewable deployment by 25.5%, highlighting embodied carbon's role in deeper decarbonization.

Paper Structure

This paper contains 38 sections, 22 equations, 11 figures, 8 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Motivation
  3. Literature review
  4. Contributions
  5. Paper organization
  6. Two-stage joint planning framework
  7. Long-term joint decarbonization planning model
  8. Economic objective
  9. Planning investments $F(\mathbf{z})$
  10. Operational costs $G(\mathbf{x})$
  11. Environmental objective
  12. PJM system planning constraints
  13. Data center planning constraints
  14. Operational constraints for the PJM system
  15. Operational constraints for data center
  16. ...and 23 more sections

Figures (11)

  • Figure 1: Structure of joint planning of power systems with data centers.
  • Figure 2: Illustration of multi-timescale stochastic planning optimization.
  • Figure 3: AI demand variation patterns across temporal scales: (a) yearly growth trends and (b) daily fluctuations.
  • Figure 4: Topology of the PJM 21-bus system.
  • Figure 5: Seasonal patterns for demand, solar generation, and wind generation.
  • ...and 6 more figures