Does McKinsey’s forecast take into account physical constraints?

5 March 2019

By Alexandre Joly, senior consultant at Carbone 4


McKinsey, one of the world’s leading strategy consulting firms, has released its global energy outlook report: Global Energy Perspective 2019.
This publication, intended for economic leaders, embeds a very thorough work:

  • The granularity of the analysis – 146 countries, 30 sectors, 55 types of energy covered – is exemplary
  • The network of experts solicited is vast, international and multi-sectoral
  • The storytelling setting is eminently fluid and powerful

Nevertheless, this exercise has been carried out without consistency checks to ensure that the proposed evolution of the energy model remains physically possible.

Like many economic forecast analysts, McKinsey postulates that past price developments will continue. This makes it possible to date when the electric vehicle will be more competitive than the thermal vehicle, or when renewable energies will be cheaper than fossil fuels. Their speed of penetration into the market is then deduced, and it is backed with a vision of a growing economy that follow the historical trends. Unfortunately, the underlying physical system is never questioned: these evolutions, which involve ever increasing flows of materials and energy on a finite planet, should necessarily be challenged with the known physical limits of the system.

For example, McKinsey reaches the following situation for the energy panorama in 2050:

Energy limit

McKinsey predicts that world oil and gas consumption in 2050 will be almost identical to today’s. A small case study on oil: knowing that the peak of conventional oil production (the one “easy and cheap to extract”) was passed in 2008 [1], that current production growth relies only on unconventional oil (US shale oil and oil sands in Canada) and that it takes 5 times more energy to extract unconventional oil than the conventional oil [2], is it physically realistic to still consider such volumes of oil in 2050?

In the same vein, McKinsey’s approach does not take into account the growing energy penaltyto access a ton of metal, because of the continuing decline in ore grades. Let us consider the copper necessary for the massive electrification advocated by the authors. It takes twice as much energy to extract one ton of this metal as it was 20 years ago [3]. And this energy intensity increases exponentially as long as the ore copper content decreases.

To summarize, it will take 5 times more energy to obtain a barrel of unconventional oil that will allow to extract 2 times less copper; in other words, 10 times more energy for the same ton of copper extracted than 20 years ago, from a global physico-thermo-industrial system standpoint [4]. The observation of our production from the physical point of view offers many other examples where the decline in the quality of residual resources rapidly increase the energy required to access them.

How then to be sure that the doubling of electrical demand by 2050 planned by McKinsey – and relying on copper – is compatible with an ever-growing energy extraction weight?

Global warming

Another obvious conclusion of this scenario is a global warming of around +3°C on average by the end of the century compared to the pre-industrial era [5]. But science now provides more precise indications of some consequences: nearly extinction of corals at +1.5°C, +100 million additional climate migrants at +2°C, and above 2°C, we will probably start the disintegration of the western Antarctic ice cap, leading to almost 10 m more water for the global ocean over a few centuries [6] [7].

Integrating physical boundaries in forecasting

The prospective scenario of McKinsey, simply titled “Reference Case”, is therefore probably not compatible with the physical limits of our Earth system: resources limited and requiring more and more energy to be extracted, finite capacity of our environment to “digest” the waste of our economic activities, …

It should also be ensured that the current globalized material flows– on which this study is based – can continue to grow “Business As Usual” when, at the same time, millions of workers will flee areas that have become uninhabitable and that agriculture and energy yields will decrease.

In short, a scenario whose physical consistency has not been checked can not be considered as having a predictive value.

Explicitly incorporating these limits will restore all the interest to business strategy. This is this challenge with much better prospective value that we offer to our clients! The more time goes by, and the less resilient business models of tomorrow could be designed with approaches postulating the prolongation of economic trends, without having evaluated the physical underpinnings of this evolution!




  • Publication McKinsey 
  • [1] International Energy Agency, World Energy Outlook, 2018
  • [2] EROI of different fuels and the implications for society, C.Hall, J.Lambert, S. Balogh, Energy Policy 64, 2014
  • [3] ADEME, Copper Alliance, Analyses C4
  • [4] If we consider the global physico-thermo-industrial system as a whole, we need to inject 10 times more energy ; if we consider all the energy required, only 3 times more is needed.
  • [5] Analyses C4 : 1250 GtCOémises entre 2017 et 2050 et 1150 GtCO2 émises entre 2050 et 2100 sur un tendanciel décroissant atterrissant en 2100 sur 0 EJ de charbon, 60 EJ de gaz naturel et 100 EJ de pétrole.
  • [6] De Conto et. al, Nature, 2016
  • [7] Global Warming of 1.5°C, IPCC, 2018


By Alexandre Joly, senior consultant at Carbone 4

Share Button