We have a very optimistic outlook for electricity demand and all electricity generation technologies, their supply chains and associated infrastructure

Key points

• The US has increasing demand for electricity for the first time in 20 years driven by data centres.
• New electricity generation capacity is needed very quickly.
• The solution will not be all nuclear, or all gas, or all renewables, it will be a combination ensuring acceptable reliability at the lowest possible cost.
• Newly-built commercially-driven nuclear is extremely unlikely to be part of the picture over the next 10 years (despite the considerable hype).
• We have a very optimistic outlook for electricity demand and all electricity generation technologies, their supply chains and associated infrastructure.
• We have upgraded our outlook for electricity demand growth following Trump’s election victory.

Insight

A surprising reality is that the United States (US) has increasing demand for electricity for the first time in 20 years driven by the significant and rapidly growing electricity needs of data centres (which are supporting the rapid growth in artificial intelligence technologies).

The necessity for many data centres to manage response time or ‘latency’ (data centre latency refers to the time delay between when a request is sent to a data centre and when the reply is received) requirements is a structural factor which will necessitate the majority of data centres being built within the US (situating data centres outside of the US would result in too much latency).

As a consequence, the US needs significantly more electricity generation capacity, very quickly.

The solution to this will not be all nuclear, or all gas, or all renewables, it will be a combination of various technologies combined to ensure acceptable reliability of electricity supply at the lowest possible cost. To that end, our research indicates that:

• while nuclear holds substantial longer-term potential, over the next 10 years newly-built commercially-driven nuclear is extremely unlikely to be part of the picture (despite the considerable hype);
• while renewables are ‘intermittent’ (the sun doesn’t shine at night and the wind doesn’t always blow), at large scale they are the cheapest form of new electricity generation (they have a short-run marginal cost of zero and are roughly half the cost of natural gas-fired baseload electricity on true like-for-like terms taking into account all construction and operating costs – referred to as the levelised lifecycle cost of electricity, or LCOE); and
• while it is not yet feasible to firm renewables with battery storage at large scale, this nexus is likely to be crossed within the next 2-3 years. In the meantime, new natural gas-fired electricity generation will be required to ‘firm’ a considerable amount of new large-scale solar and wind generation;
• In the short term, the key question for markets to answer is how much gas-fired electricity generation will be required to ‘firm’ new large scale solar and wind generation. We estimate the answer is no more than 40% (comprising baseload and flexible/peaking power plants). Going all gas (as some are calling for) would, over the medium term, result in more expensive electricity without greater reliability. As well as greater risk over the longer term of stranded assets as a result of the cost of contemporary energy generation and storage technologies (with zero short-run marginal cost) continuing to fall.

The bottom line is that a substantial amount of new electricity capacity will inevitably be renewable, which is an entirely different reality to the current rhetoric and general perception (especially since the election of Trump).

All scenarios will require significant investment in the ‘grid’ in the form of new connection, transmission and distribution infrastructure (such as cables and equipment including transformers, switch gears, breakers, etc). This is in addition to what we believe is required to make up for underinvestment in this infrastructure over many years. We see excellent opportunities within these enabling technologies which have so far garnered much less attention from the market than generation technologies.

We have a very optimistic outlook for electricity demand (the US has increasing demand for electricity for the first time in 20 years driven by data centres) and all electricity generation technologies, their supply chains and associated infrastructure. We have upgraded our outlook for electricity demand growth following Trump’s election victory and the Republican sweep.