Remember "peak oil"? It's had many iterations over the years. I remember sitting in a cafe just four years ago, while the executive of a small oil exploration firm explained to me that all the easy-to-reach oil was gone. What was needed, he explained, was for intrepid companies like his to go out and find the tricky stuff hidden in under-explored basins. This would be more expensive, but prices were high. What else could the world do?
Rather a lot, it turns out. In fact, that talk of "peak oil" is now being replaced by another prospect: peak oil demand. This week, GTM Research and Wood Mackenzie, the influential Scottish energy research house, published an analysis suggesting global fossil fuel demand could peak in the next five years. Paul McConnell, the author of the analysis, said two or three years ago it was hard to get oil and gas firms interested. But in the last year, "it's become virtually all we talk about with clients".
Just like record companies and retailers, oil and gas firms are scrambling to get their heads around the coming disruption. Mainstream investors, even those holding companies like ExxonMobil, are pushing them to publish analysis of how decarbonisation will affect their assets and what they plan to do. These investors know that the fossil fuel industry is about to be caught by a gathering storm of electric vehicles, increasing energy efficiency and the plummeting costs of production for renewables. Altogether, in McConnell's "disruptive" scenario, these advances will reduce fossil fuel demand by 6.5 million barrels a day by 2025. That's about 7pc of total daily demand.
This isn't Wood Mackenzie's most likely scenario at present. But perhaps it should be. Recent announcements by France and Britain commit them to phasing out hydrocarbon-burning vehicles in line with the speedy scenario, not the base case. And history tells us that established industries are very bad at grasping the pace and scale of disruption heading their way.
Change can happen with astonishing speed, as the rise of oil shows. From 1949 to 1972, global oil demand rose by a factor of 5.5 in 23 years. In Europe, it rose by a factor of 15. In Japan, by 137 times.
If the fast-changing scenario laid out by Wood Mackenzie is right, carbon emissions will peak in the mid 2020s, equating to a 2.7 degree rise in global temperatures. That's well above the two-degree warming limit environmentalists would like, but far exceeds the targets of the Paris climate change agreement. Policies to manage this change will be needed, but subsidies for renewable energy won't. Meanwhile, the cost of buying an electric vehicle could reach parity with normal cars by the early 2020s. Once that happens, the change should accelerate. By 2040, we may see a 100pc replacement rate.
The Paris agreement and renewable subsidies might soon be out of date, but the changes still require state involvement. Like oil and gas companies, governments have to start planning. That means redesigning electricity markets, updating power grids and, eventually, working out what to do with unnecessary legacy power plants. Countries that decide instead to keep burning coal might benefit from cheaper energy in the short term, but they'll be poorly placed for the next wave of energy technology.
None of this means oil markets are about to disappear. Fossil fuels will still account for most of the energy market for several decades to come because the vast capital investment needed to implement the changes will take time to deploy. But once costs fall, the logic of switching becomes self-reinforcing.
In the 1990s, the consulting firm McKinsey was asked to predict what the global market for smartphones would be by 2000. It guessed just under 1 million: wrong by a factor of 109. The decarbonisation of energy is coming. It's time for governments, investors and the industry to plan for it, rather than sticking their heads in the sand.