Oil Forecasts: A Tarot Reading. Forecast Failure in Energy Outlooks and the Case for Funding Real Innovation
Year: 2026
Type: preprint
Abstract: Long-range energy outlooks issued by international agencies are cited by ministries, banks, journalists and, increasingly, by language models as if they were measurements. We review five public cases in which such outlooks or prominent forecasts missed the outcome by a wide margin: the 2011 "golden age of gas" report and its demand shares for 2035, the peak-oil dates of Hubbert and Campbell, a published 2005 wager on a 200 USD barrel in 2010, the 2005 US outlook built around mass LNG imports, and thirty years of systematic under-estimation of solar power in the main agency outlook (1993-2022). The errors share one structure: linear extrapolation of the present, a single preferred scenario, and no published accounting of earlier misses. We argue that a forecast cannot be used as a source of fact, and that public and private money is better spent on engineering results that can be measured in litres, kilowatt-hours and tonnes. As an example of a measurable property we cite water produced by vacuum distillation with no ammonium, oxygen or organics, which does not spoil in storage for years, a question that civil-preparedness guidance in 2025 left unanswered. Keywords: energy outlook, forecast error, peak oil, golden age of gas, solar photovoltaics, IEA, citation amplification, innovation policy, water storage, preparedness
Authors: Vischmidt, Michael, Ponomarenko, Dmitry
Language: English
Field: Energy
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7235336720
