You're Gonna Need a Bigger Power Grid

In one of cinema’s most memorable moments, we see Martin Brody, police chief of Amity Island, casually tossing bait into the water from a small boat attempting to lure out a shark. We then get our first glimpse of the iconic beast in Jaws. Staggered by the (much bigger than expected) challenge ahead of him, Brody slowly backs into the cabin of the vessel and mumbles to the skipper: “You’re gonna need a bigger boat”.

This iconically understated line came to mind recently when we looked at power demand forecasts for Europe to 2030. As the following graph shows, the impact across various sectors could amount to an extra 461 TWh of demand that would have to be met in just over five years.

For context, that much demand in Europe would be roughly equivalent to the entire 2023 annual electricity demand of France. The downside case is no small deal either: adding “only” 277 TWh of electricity demand would be like adding another Spain to the continent.

Moreover, recent forecasts for the US, published by the Energy Information Administration also showed a drastic uptick. The latest forecast predicted energy demand in 2030 to be 7.4% higher than the forecast published a year before. The forecast for 2030 changed from 4,031 TWh to 4,329 TWh, roughly equivalent to the 2023 annual electricity consumption of the United Kingdom (309 TWh). [2]

Over the past two decades, policymakers operated on the (flawed) idea that the growth of ageing populations in developed countries would reduce long-term demand for power. Indeed, appliance efficiency improvements and the offshoring of energy-intensive industries to other countries have pushed down per-capita energy consumption since the mid-2000s. As a result, continued re-investment into the developed world’s obsolescent infrastructure was seen as a low priority.

But the situation has changed. Future growth will be dependent on the efficient management of resources to maximise productivity. Indeed, a recent Bank of America report on thematic investing for 2030 highlighted one of its key themes as “exponential growth of tech requiring more resources like infrastructure, computer bandwidth, human capital, energy, water, skills and data centres”.

Finding ourselves well into the global energy transition, and nearly three years on from OpenAI’s epoch-making public release of GPT-3.5, the conventional wisdom of the past few decades is beginning to show its faults. For example, offshoring was once considered a way for advanced economies to focus on service sector jobs and make progress toward climate targets. Over the years, however, we’ve seen that offshoring also creates issues with manufacturing resilience and reduces the impetus for public officials to invest in critical national infrastructure.

This underinvestment has consequences. Much of the UK’s grid infrastructure, including large power transformers, is operating decades beyond its intended lifespan. For every £1 spent increasing generation, only 25p is invested in transmission, a ratio that’s simply not sustainable as demand rebounds. [4]

And rebound it will. Today, the UK is poised to see a resurgence of electricity demand, as data centres draw immense amounts of power for both AI and non-AI related digital tasks. McKinsey projects European data centres will more than triple their power draw by 2030, from 10 GW today to 35 GW. [5] Despite the massive demands of these facilities, data centres might not be the largest driver of demand in the near term. In 2024, the IEA estimated that heat pumps, industrial activities, and EVs could create 128 TWh of demand in the European Union as early as 2026.

Persistent constraints have pushed some large companies to invest in the European energy industry. In 2024, Amazon was Europe’s largest corporate purchaser of renewable energy. The company invested in six onsite solar installations in the UK alone last year. Companies also support the energy industry through power purchase agreements (PPAs). In 2024, European energy providers saw a surge in PPAs to 19 GW, which signals growing interest in building out Europe’s renewable energy capacity. [6]

This dire need for power and grid capacity is reflected in the US too: the data centre capex commitments made by Big Tech are being matched by investments in power generation. A need for clean and uninterrupted power supplies has led to the Microsoft-backed reopening of Three Mile Island, and substantial investment by Amazon in X-Energy, a nuclear reactor startup – to name just two.

The commitments from tech giants are strong but perhaps not sufficient to solve the underlying grid capacity crisis. Beyond the massive scale of investment required, other systemic issues must be addressed. In the UK, low barriers to application for power supply have created an overwhelming queue of grid connection projects and requests that cannot possibly all be approved. Nevertheless, this deluge forces grid regulators to wade through an ocean of applications for minimal gain. The situation escalated to such an extent that the National Energy System Operator suspended all new requests in early 2025.

As we confront this unprecedented power challenge – our proverbial "bigger boat" moment – the implications extend far beyond utility companies and tech giants. Investors and policymakers must take note.

[1] McKinsey, Ember. Data for EU-27, Norway, Sweden, and the UK.

[2] SP Global, Ember.

[3] World Bank, US EIA.

[4] Bloomberg NEF.

[5] McKinsey.

[6] Wood Mackenzie