495-26 - Data Centers, ArtificiaI Intelligence & Electricity
Core/Elective: Elective
Credits: 0.5
Quarter Taught: Spring
Synopsis:
Modern AI and large-scale computing are driving a tectonic shift in global electricity systems. Data centers that support AI workloads are becoming one of the fastest-growing sources of demand in the power sector and this is projected to more than double global electricity demand by 2030. Meanwhile, the race to develop firm power has created a backlog of grid interconnection requests, with new operating projects sitting in queue for 6-9 years in many ISO/RTO jurisdictions. 1 In effect, digital infrastructure is becoming a critical component of energy system planning. In the U.S., data center load is expected to be responsible for 30–40 % of all net new electricity demand between now and 2030.
Over the next decade, the growth in demand will pose challenges for grid reliability, climate targets, and infrastructure planning — but also create novel opportunities for optimization, markets, and innovation. This course examines how AI workloads are changing the scale, shape, and carbon intensity of global electricity use, and how digital infrastructure can become a strategic player in energy markets.
Students will learn to:
- Understand the impacts, risks, and opportunities created by data center growth for electricity supply, transmission, and sustainability goals
- Evaluate how AI and data center capacity can support grid optimization, flexibility, and new market value
that connect the digital and electric revolutions.
Course Objectives:
- Understand data center impacts, risks, and opportunities for global electricity demand
- Describe the architecture and components of AI-oriented data centers — including computing hardware, cooling systems, and network infrastructure that influence energy
demand and efficiency. - Explain the power characteristics of AI workloads — differentiating between “regular” electricity and AI workloads including training, inference, and storage activities in terms of load profiles, peak demand, and operational variability.
- Quantify the electricity consumption and carbon intensity of data centers across regions, considering grid mix, renewable penetration, and transmission constraints.
- Assess the impacts of rapid AI data center expansion on power systems — generation adequacy, transmission capacity, and distribution reliability.
- Analyze interconnection, permitting, and policy challenges arising from large-scale electricity demand growth driven by digital infrastructure.
- Evaluate strategies for decarbonizing data center electricity use — including clean energy procurement, on-site generation, energy storage, and grid-interactive
operations, and how “clean power” is balanced in real time- considering, cost- and supply chain-constraints.- Critically assess sustainability metrics such as Power Usage Effectiveness (PUE), water use, and carbon accounting, and how they influence corporate and policy decisions.
- Examine the carbon implications of rising data center demand — including lifecycle emissions from generation sources, embodied carbon in infrastructure, and potential pathways to align digital expansion with net-zero targets
- Describe the architecture and components of AI-oriented data centers — including computing hardware, cooling systems, and network infrastructure that influence energy
- Evaluate how data center capacity and AI can be used in global energy markets for optimization or other value-creation activities
- Apply AI and data-driven methods to optimize grid and energy operations, including forecasting load, managing flexibility, and integrating renewables.
- Develop optimization or control strategies (e.g., dynamic workload shifting, demand response, or energy arbitrage) to enhance efficiency and value creation.
- Model how data centers can serve as flexible grid resources, providing ancillary services, demand flexibility, and virtual power plant participation.
- Integrate physics-based and data-driven approaches to forecast and manage data center power usage and its interaction with renewable generation.
- Assess economic and regulatory frameworks that enable or constrain participation of digital infrastructure in energy markets.
- Design and evaluate case studies or pilot projects where AI systems are applied to energy optimization, grid forecasting, or operational decision-making.
- Communicate and justify recommendations that balance energy reliability, cost, and sustainability goals while leveraging data center capacity for societal benefit.