Authored by: Sangeetha Rajasekeran, Transmission Planning Engineer

At this year’s ESIG Fall Technical Workshop, one theme came through clearly: the grid is changing faster than the systems built to plan it. Operators across every ISO are contending with record load growth, constrained generation, and an urgent need to modernize how the grid is studied, built, and managed. Interconnection queues are congested, project realization rates are low, and planning processes designed for a slower era are being stretched to their limits.

Load growth and data transparency

Most grid operators are facing unprecedented increases in demand driven by data center development, heating electrification, and EV mandates. At the same time, regional interconnection queues are clogged, creating a mismatch between where capacity is needed and where new generation can actually connect.

Large-load requests are increasing in volume, but there is still no centralized way to track them or identify duplication across transmission owners. The lack of data transparency makes it difficult for operators to know which requests represent real, non-duplicative demand.

There is an immediate need to improve load forecasting so operators can distinguish reliable requests and plan future transmission expansion and resource requirements accordingly.

Data center demand and system impacts

Technical sessions focused heavily on the unique characteristics of data center load. Presenters discussed how these facilities differ from traditional demand, both in intensity and variability, and the importance of understanding their complex interactions with the grid.

Across sessions, battery storage emerged as a key mitigation strategy to manage the sharp spikes caused by data center operations. Presenters pointed to E-STATCOMs and co-located storage systems as effective tools to absorb short-term fluctuations and enhance local stability.

Efforts to accelerate interconnection

To address delays, grid operators are exploring fast-track processes that allow certain projects to move forward more quickly. These include:

• Expedited Resource Addition / Adequacy Study (ERAS) in SPP and MISO

• Reliability Resource Initiative (RRI) and the proposed Expedited Interconnection Track (EIT) process in PJM

• Improvements to surplus interconnection and capacity transfer processes that reallocate capacity from deactivated generators

AI and automation in grid operations

A dedicated session on AI adoption explored several practical use cases in operations planning:

• Organizing and validating data with large language models (LLMs)

• Enabling chat-based search for documentation and model repositories

• Improving probabilistic forecasting and uncertainty quantification

• Automating report generation and analysis

Notable initiatives include:

• ISO-NE, collaborating with research centers to explore GenAI enabled transmission security assessments and ML-based transmission outage probability analysis

• SPP, working with Hitachi Vantara and Nvidia to develop GPU-powered tools for data ingestion, modeling, and report creation — aiming to cut study timelines by more than 80%

• EPRI, which is testing and benchmarking LLMs on power system questions to understand their capabilities across varying levels of complexity

Reflections from the field

Across all discussions, one trend was clear: operators are becoming more open to automation and AI as the scale of change outpaces existing tools. The shift won’t happen overnight, but cautious experimentation is underway, and momentum is building as trust in results grows.

Battery storage is also expected to play a larger role, both through expedited interconnection and as part of load-side solutions to damp spikes in large-load consumption. Partnerships between data centers and storage providers may become increasingly important in managing localized grid impacts.