Ice cream for dogs: Why energy demand just cannot stop growing

5 min read 22 Oct 24

Granular energy transition models all contain an inherent tension between better energy efficiency due to technological progress, and the emergence of new energy use-cases, which are often made possible precisely because of evolving technology. Here, we look at examples of this dynamic in practice, and consider how it may impact energy demand in the coming decades.

Consider the devices you’ve interacted with so far today. Perhaps you are woken up by an alarm on your smartphone, which has been charging overnight. You may also pick up headphones, a work phone or a tablet before leaving the house. Does your toothbrush contain batteries? Your fridge might have a screen or be connected to Wi-Fi, and there’s a reasonable chance it has a water and ice dispenser. Whether you work from home or at the office, how many screens do you have, and how big are they? And how does all of this compare to 10 years ago?

“Although minor in the grand scheme of things, few would have considered use-cases such as frozen canine snacks in the energy demand models of a decade ago.”

The end of energy waste?

From the perspective of the overall energy system, there is plenty of good news. Optimists argue that we are undergoing a fundamental shift to a less wasteful system. Internal combustion engines deliver tank-to-wheel efficiency of only 20-35%, while other combustion processes such as boilers and stoves can also be surprisingly inefficient1.

Renewable power also suffers from transmission losses, but not to anywhere near the same extent. Likewise electric vehicles (EVs) and other electric motors. We can also consider heat pumps, which benefit from enormous efficiency because they draw in ambient heat from the air, leading to much greater life-cycle efficiency than gas boilers.

In energy circles, this is referred to as ‘The Primary Energy Fallacy’ – the idea that the energy system of the future will be meaningfully smaller and easier to build than today’s because it cuts down on waste. But do we have the collective will to actually use this system more efficiently, or will novel and unexpected sources of energy demand continue to pop up?

Or just more demand?

Take doggy ice cream, for instance. Frozen canine snacks have been a sell-out summer product for Aldi over the past two years, building on the hype of a similar launch from Ben & Jerry’s. Data from Google Trends suggests that interest in ice cream for dogs continues to grow strongly.

A more extreme example hit the news in the US this summer, when workers at a Shake Shack drive-through wore portable ice packs to deal with record August heat. Such technology is not new, having been the subject of military R&D for decades for use in combat zones. Yet resorting to wearing a frozen block to withstand 40°C temperatures while taking fast food orders is surely worthy of note. Although minor in the grand scheme of things, few would have considered use-cases such as these two in the energy demand models of a decade ago.

Considering country-level dispersion

On a positive note, energy use per-head has actually declined in the UK over the last 10 and 20 years (likewise in the US), after reaching an apparent all-time high in 19792. Yet over this period, the UK’s economy has progressively de-industrialised and re-focused on services. According to data from World Steel, the UK is second only to Venezuela in the decline in its steel production base since 19673. Indeed, the UK’s last coal-powered steel furnace at Tata Steel in Port Talbot, Wales, was shut down for good on 30 September 2024.

“China is the clearest example of a country dealing with booming energy demand, in all its forms, even whilst taking into account short-term economic weakness.”  
 

The mirror image of this can be seen in Chinese energy consumption statistics. As a result of heavy industry, China consumes roughly 20% more energy than the UK, despite economic output being roughly 75% lower (both on a per-capita basis)4. China therefore over-indexes on energy demand, with a potentially huge growth runway ahead, if it is to converge with consumption levels in the US, for example.

China is the clearest example of a country dealing with booming energy demand, in all its forms, even whilst taking into account short-term economic weakness. According to analysis from Bloomberg New Energy Finance, in 2023 the country installed around 260GW of solar capacity, more than the entire world managed in the previous year combined, and a figure likely to be eclipsed this year. Electric vehicles account for roughly 20% of all new vehicle sales, and close to two in every three EVs bought globally are purchased in China5. The country boasts a long list of broken energy transition statistical records.

Yet such is the size of its energy system that, in 2023, China also recorded the largest one-year increase in oil demand of any country in human history6. Some level of inventory build and an ongoing recovery from the pandemic no doubt contributed, but the picture of a multi-energy system catering for rising demand is clear.

A long demand-growth runway ahead

Only when looking at regional per-capita energy consumption statistics does the challenge of satisfying rising demand truly become clear. Per person, China consumes slightly less than half of the energy as the US. However, another growing economy, India, consumes only one-tenth7.

We could presume, for argument’s sake, that rising wealth and new energy use cases will push India towards the average power consumption levels of the G7 countries over the coming decades. Even without further population growth, this dynamic alone could boost global energy demand by close to 50%. And this is not a ceiling – how much more prevalent will air conditioning be in India 30 years from now, as compared to the G7 today, for example?

The proliferation of data centres providing computational services for AI applications is perhaps the greatest current example of novel and unexpected energy use creating tension in the energy system. Forecasts from the International Energy Agency (IEA) suggest that global energy consumption from data centres will roughly double by 2026, equivalent to adding the entire power consumption of Japan8. As if we needed a reminder, the energy transition just got tougher.

It could be that AI brings technical advances and innovation in battery development, or in the management of the energy system itself. This could somehow achieve a net gain for the energy transition. What ice cream for dogs brings to the table is less clear, other than to remind us that society’s ability to invent novel and unexpected uses for energy remains firmly intact.

1 ScienceDirect, ‘Estimation of tank-to-wheel efficiency functions based on type approval data’, (sciencedirect.com), October 2020. The Engineering Toolbox.
2 Our World in Data, ‘Primary energy consumption per capita’, (ourworldindata.org), October 2024.
3 Sky News, ‘Why the British steel industry is on the brink of extinction - or a green resurrection’, (news.sky.com), April 2023.
4 Energy Institute, ‘Statistical Review of World Energy 2024’, (energyinst.org), June 2024. World Bank DataBank, ‘GDP per capita (current US$)’, (data.worldbank.org), October 2024.
5 International Energy Agency, ‘Trends in electric cars, (iea.org), April 2024.
6 Energy Institute, ‘Statistical Review of World Energy 2024’, (energyinst.org), June 2024.
7 Energy Institute, ‘Statistical Review of World Energy 2024’, (energyinst.org), June 2024.
8 International Energy Agency, ‘Electricity 2024: Executive Summary’, (iea.org), January 2024.


The value of investments will fluctuate, which will cause prices to fall as well as rise and investors may not get back the original amount they invested. Past performance is not a guide to future performance. The views expressed in this document should not be taken as a recommendation, advice or forecast.