The challenge facing energy companies is not necessarily finding the billions needed to modernise infrastructure and support the energy transition, but deciding where that money should be spent, according to Jeff Pauska of IFS Copperleaf.
Speaking to The Energyst at IFS Unleashed in Orlando, Pauska said capital remains available for energy infrastructure, but utilities face an increasingly difficult task demonstrating which investments should take priority.
“I don’t think finding capital is so much of an issue,” he said. “It’s the permission to spend it.”
Utilities are simultaneously being asked to improve resilience and reliability, accommodate electrification and new demand, replace ageing assets and decarbonise their networks – while keeping energy affordable.
“We could do all these great things. We could be net zero and introduce big decarbonisation initiatives and projects. However, what is that going to cost? What’s the actual customer value?”
Pauska believes the UK regulatory model has some advantages in answering that question, particularly the way consumer priorities are considered alongside financial measures.
Rather than assessing investment purely in monetary terms, utilities increasingly need to place a value on factors such as reliability, resilience, safety and decarbonisation.
The difficulty comes when hundreds or thousands of potential projects are competing for a finite capital budget.
Creating a single view of investment
Electricity and gas utilities are relatively advanced in having centralised capital planning functions, according to Pauska, but investment information can still sit across different systems, spreadsheets and departments.
Creating a single source of investment data allows projects to be compared on a consistent basis and enables organisations to determine where capital can create the greatest overall value.
That could mean deciding whether to replace an asset, refurbish it, defer work or pursue an entirely different investment.
AI, Pauska argues, could make that process considerably more sophisticated.
“A lot of people use AI for efficiency. For us, it’s about how do we make the decisions better and how do we work the problem space deeper?”
One example could be modelling extreme weather events and incorporating the resulting risk into infrastructure investment decisions.
AI could also help standardise judgements traditionally made by individual engineers – comparing different technical solutions and determining whether additional expenditure genuinely produces sufficient reductions in risk.
Planning for uncertain demand
The challenge is becoming more acute as electricity demand becomes harder to predict.
EV adoption, heat pumps, distributed generation and particularly data centres could create significant new loads, but utilities cannot be certain when proposed developments will actually materialise.
Pauska said network operators increasingly need to work with a “cone of uncertainty”, modelling multiple possible outcomes rather than relying on a single demand forecast.
Utilities may have visibility of construction applications and proposed developments, but projects can still be cancelled or delayed.
AI and increasingly sophisticated modelling can therefore help organisations examine different scenarios and understand the consequences of each.
“What if you had a data centre here? What if you have 50% EVs plugging into this neighbourhood because it’s affluent?” said Pauska.
“That’s really the ‘what if’ game in the simulation. That’s where it’s going.”
The approach could also allow utilities to compare conventional network reinforcement with alternatives such as refurbishment, deferral or measures that extract greater capacity from existing infrastructure.
Pauska stressed that some of those power engineering decisions sit outside Copperleaf’s current role, but said the direction of travel is towards systems capable of considering a much broader range of alternatives.
From boardroom debate to AI optimisation
Perhaps the biggest change could come in deciding which projects ultimately survive the budgeting process.
Today, portfolio decisions can involve teams from across an organisation competing to demonstrate why their projects deserve funding.
Pauska believes that within five years AI could play a much greater role in making those trade-offs.
Asked which investment decision currently requiring “a room full of people” could eventually be trusted to AI, his answer was portfolio optimisation.
“The trade-off, the political jockeying – today it’s rooms of people deciding what projects should survive,” he said.
“Tomorrow, AI optimises and you don’t argue because it’s so data-informed and there is a full chain of evidence.”
For an energy sector facing unprecedented demands on infrastructure, that could ultimately prove more significant than simply using AI to do existing jobs faster.
