Utilities Rethink Capital Delivery As Workloads Surge

Utilities facing unprecedented capital investment demands are being forced to rethink how they plan, staff and execute projects. During a Tuesday panel at DTECH in San Diego, leaders from Georgia Power, Xcel Energy and San Diego Gas & Electric (SDG&E) compared notes on how their organizations are adjusting talent models, governance structures and analytics to keep pace. The session, “Reinventing Capital Projects: Talent, Transformation, and AI in Utilities,” was sponsored by Accenture.

For Nick Martin, general manager of distribution at Georgia Power, the pressure starts with scale. Growth across the system has accelerated beyond anything utilities planned for a decade ago, and traditional delivery models are no longer sufficient.

“If you look at the amount of growth that we’re expecting, it’s really unprecedented compared to what it’s been over the last 10 or 15 years,” Martin said. “Today, the demand is coming at such a rate, we have to make sure we have the right structure in place, as well as the right people in place.”

At Georgia Power, that has meant rethinking how and where the company recruits. Martin described a new Alignment Entry Program that brings potential craft workers in through a six-week trial period with no long-term commitment on either side. Those who show up on time, retain information and demonstrate solid work habits are offered a place in the company’s apprenticeship program.

“This is really been working well for us, we have about a 40% success rate,” Martin said. “We plan to continue to leverage that.”

The approach also allows Georgia Power to recruit by zip code, targeting areas where work is actually needed, a practical adjustment as hiring volumes surge. Martin said the company has tripled its lineman intake, bringing in more than 200 apprentices in a single year.

Paul McGregor, vice president of wildfire risk management at Xcel Energy, said workforce challenges extend beyond field crews. His team is responsible for building out meteorological, data science and risk analytics capabilities across an eight-state territory, where wildfire risk varies widely by location.

“Back in the day when we did data analysis, we used Excel,” McGregor said. “Now, everything is done in machine learning. It’s done in Python models, and that’s a new capability to electric utilities.”

Xcel has leaned heavily on internships and early-career recruitment to build those skills internally. McGregor tied those investments directly to capital outcomes, noting that poorly staffed analytics projects can undermine returns on large infrastructure programs.

“If I’m not applying the right people and the right resource loading to some of these projects, then I’m impacting my company’s ability to get that return on invested capital,” he said.

At SDG&E, Director of Portfolio and Project Management Erika Schimmel-Guiles emphasized that getting people in the door is only the first step. Sustained capital performance depends on consistent training, clear career paths and disciplined execution models.

“Once you get the talent in the organization, you have to continue to upskill and develop and invest in them,” she said.

SDG&E has centralized governance, project controls and standards while decentralizing execution, a structure Schimmel-Guiles said has reduced delays and improved consistency across transmission, substation and distribution portfolios. One example was the utility’s strategic undergrounding program, where centralizing property owner and easement coordination removed a major bottleneck for project teams.

“That has really raised the maturity across the board,” she said.

All three panelists stressed affordability as a limiting factor, pushing utilities toward more targeted investments and closer partnerships with contractors. For McGregor, that means using analytics to determine where expensive measures undergrounding deliver the most wildfire risk reduction.

“You can’t underground everywhere,” he said. “So we need to be targeted.”

AI and advanced analytics are increasingly central to those decisions, though panelists cautioned against hype. Schimmel-Guiles described AI as “a tool that can 10x human capability,” particularly in compressing long lead times for land rights, environmental permitting and research.

“We’re working with a company who is helping us take things that take six months down to two weeks,” she said.

McGregor echoed that view, pointing to wildfire forecasting and grid operations where machine learning has reduced analysis timelines from days to hours. Still, he emphasized the need for human oversight.

“I still believe that we need that human standing over it to make sure that that data and that output is valid,” he said.

Martin said Georgia Power is taking a similar approach, building data lakes and pushing AI use cases down to the crew level, but always with efficiency and affordability as the goal.

“Good data in, good data out,” he said.

As capital programs grow larger and more complex, the panelists agreed that early planning and cross-functional coordination are essential to reducing uncertainty. Martin said Georgia Power has moved away from siloed execution toward bringing all stakeholders into the planning process early.

“We really have to have all stakeholders involved in the planning process,” he said.

Schimmel-Guiles offered a simple takeaway for peers navigating similar challenges.

“Clarity reduces uncertainty,” she said. “Make expectations unmistakably clear.”

For utilities staring down record investment cycles, the panel made clear that success will depend less on any single technology and more on disciplined execution, targeted hiring and a willingness to rethink long-standing ways of working.