Grid Operators Cautious on AI for Interconnection Analysis

Automation gains traction for routine tasks

Transmission grid operators are embracing automation for interconnection queue management, but remain cautious about deploying artificial intelligence for the complex analysis that determines grid upgrade requirements and costs.

Southwest Power Pool has automated intake and validation of interconnection requests for large-scale solar and other generation projects, eliminating manual work and freeing engineers to focus on analysis, according to SPP Technical Director for Engineering Steve Purdy. He spoke at the Infocast Transmission and Interconnection Summit, as first reported by pv magazine USA.

But Purdy drew a clear line at using automation for "the actual analysis, where you feed the information in and have it tell you what the correct upgrades are, what the correct impacts are, what the cost allocation should be." That capability, he said, "is probably farther away than more immediate."

Confidence and validation drive slow rollout

The hesitation stems from the need for confidence in automated results among grid operators, their members, customers, and regulators. "That's been really tricky," Purdy said. SPP is testing extensively and validating results before moving automation tools into production environments.

The organization is partnering with automation providers including GridUnity, whose CEO Brian Fitzsimons described the firm's Grid Analytics Learning Engine (GALE AI) software supporting SPP. GridUnity also offers GridSync for real-time data collaboration between developers and transmission operators, and has introduced a Grid Intelligence Lab where operators can test AI use cases with real grid data in secure environments before production deployment.

Fitzsimons indicated new production automation capabilities would be announced later this year.

MISO demonstrates speed gains in phase one studies

The Midcontinent Independent System Operator has achieved dramatic results using Enverus automation for phase one interconnection studies. David Bromberg, vice president for power and renewables at Enverus, reported that MISO reduced turnaround time from nearly two years in 2021 to three months for 2023 and 2025 study cycles. MISO has automated phase one studies for approximately 300 GW of projects in the past year.

However, Max Blackburn, manager for interconnections and grid analysis at Invenergy, noted that while the Enverus platform "operates as intended" for phase one feasibility studies, MISO's phase two system impact studies "essentially started over from phase one," largely due to large cluster sizes in recent study cycles.

Phase one studies are followed by system impact studies and facilities studies that define required grid upgrades, costs, and lead to generator interconnection agreements.

Why it matters

The interconnection queue backlog has become a critical bottleneck for renewable energy deployment, with hundreds of gigawatts of projects waiting years for grid connection approvals. While automation shows promise for accelerating routine tasks and early-stage studies, the industry's cautious approach to automating complex technical analysis reflects the high stakes of grid reliability and the regulatory scrutiny surrounding transmission planning decisions. The gap between phase one automation success and phase two challenges at MISO illustrates that speed gains in one study phase don't automatically translate to end-to-end process improvements.

Bromberg also noted that PJM, the nation's largest grid operator, is using Tapestry automation software for application and site control validation tasks.

These details were first reported by pv magazine USA.