by Barry Brook
In a previous post – TCASE #The energy demand equation to 2050 — I estimated a mid-century global primary energy demand of ~1000 EJ (see here for definitions). But it may as well have been 2060; the actual date that this global demand will be reached is obviously uncertain, but will likely occur between 2040 – 2070 given current levels of energy growth. This figure was also arrived at by Moriarty & Honery (2009) based on a meta-review of the literature. Let’s use this as a working figure.
Table 1 shows world electricity demand in 2008 based on IEA data from non-fossil-fuel sources, plus world total including fossil fuel generation. Note that a terawatt year (TWyr) is the same as 1000 GW of constant power. So nuclear power, in 2008, delivered an average of 312 GWe, and global electricity generation was 2,313 GWe.
Table 2 shows a hypothetical 2060 demand scenario, which uses the forecast values from Trainer (2010) for energy efficiency/conservation, direct electricity, transport electricity (e.g., battery electric vehicles) and liquid fuels (see also this shorter, free-online piece); however, my estimate of the source of liquid fuels is different (see explanation below).
Note that in Table 2 there is a projected overall 3.8-fold increase in world electricity use between 2008 and 2060, compared to an approximate doubling of overall primary energy usage (today we use ~500 EJ from all sources). Both of these figures – for electricity and primary energy growth — are in agreement with the estimates of Starr (1993).
Trainer assumes that up to 50 EJ/yr will come from biomass-derived cellulosic ethanol (requiring 1 billion ha and 7 t/ha yield); he also leaves an unmet deficit of 12 EJ/yr. I more conservatively assume a lower contribution from biofuels of 15 EJ/yr (300 million ha). The remaining 47 EJ/yr of primary energy from liquid fuels is assumed in my scenario to come from synfuels (e.g., hydrogen and hydrogen-nitrogen derivatives such as ammonia or hydrazine), which are synthesized using energy from nuclear fission sources (Forsberg, 2009).
(i) One third of hydrogen (~16 EJ) will come from electrolysis at a 30% electricity-to-hydrogen conversion efficiency; this will require 52 EJ of electricity input.
(ii) Two thirds of the hydrogen (~32 EJ) will come from direct nuclear heat via high-temperature sulphur-iodine-catalysed thermochemical water cracking, at a 60% heat-to-hydrogen conversion efficiency. This thermal energy requirement is the equivalent of 550 GW of electricity plant, if one assumes a 33% Carnot-cycle efficiency that is typical for thermal-to-electrical conversion in fission reactors.
The ratio of direct stationary/transport electricity use to that used in synfuel manufacture (electrolysis and nuclear heat) in Table 2 is 0.24. By comparison, Eerkens (2006, pg 135) estimated a final ratio of 0.4, but did not include battery electric vehicles or biofuels.
The 116 (direct) and 92 (transport) EJ electricity figures come from Trainer (2010). If you assumed that all of the 277 EJ of electricity in Table 2 was generated at a thermal-to-electrical conversion of 33%, then this is 831 EJ of primary energy. This would imply efficiency/conservation savings of 17% on the 1000 EJ original demand target.
I hope to use the above 2060 scenario in an upcoming paper that I am currently writing, and so would appreciate any feedback/constructive criticism from readers.
In the next SNE2060 post, I’ll consider my most realistic assessment for the multi-source energy supply equation for 2060 which delivers 277 EJ of electricity — from both nuclear and non-nuclear sources.