TL;DR It has been common to represent global energy usage in terms of Primary Energy – that is the energy potential of the various sources. However, this can be easily criticized because not all energy is equal and what we really want to understand is the “Useful Energy” that does work on our behalf at the end of any transformation process. The conversion of energy from one form to another (thermal combustion to electricity for example) is subject to the Second Law of Thermodynamics and there are always irreversible “losses” in conversion. Thus, depending on the efficiency of the conversion process(es), much less Useful Energy is used than available Primary Energy. Viewed from end-use perspective, one can argue that much less energy is required globally if we have more efficient sources of primary energy. Solar and Wind which generate electricity directly without any thermal combustion are significantly more efficient. Ergo, we don’t need to replace all of the current Primary Energy, only a fraction of it, to have the same amount of Useful Energy. Unfortunately its not that simple, and understanding the limitations of this argument shed light on why the energy transition is far more complex than the usual sound bites beloved of headline writers.

Primary Energy vs Useful Energy

There is a misunderstanding about how much energy we use vs how much we need. 

The question is around the usefulness of “Primary Energy” as a measure of the energy we need to get stuff done.  Interestingly, as a precursor to bigger strategic moves, BP started reporting global energy demand in Exajoules rather than in Billions of barrels of oil equivalent in their 2020 Energy Outlook. This was a nod to their dubious new corporate strategy of Beyond Petroleum 2.0, but I’ll not dwell on that.

This semantic change in the reporting units does nothing to address the more fundamental underlying issue. 

Simply put, we should be focused on how much work gets done, or useful energy, rather than on how much energy was available at the beginning of the process.  In this framework of “useful energy” things can get complicated and there are various branches of science dedicated to its study.  We think we know what energy is, although it is notoriously hard to pin down. The area of Useful Energy is called Exergy:

Exergy is a thermodynamic concept, used for many years within engineering analyses of chemical and mechanical processes and systems. Officially, exergy is defined as:

The maximum useful work which can be extracted from a system as it reversibly comes into equilibrium with its environment.

In other words, it is the capacity of energy to do physical work(source)

To give a very relevant example, consider the useful energy required in driving a car n kilometres to the store to get a tub of ice cream. If this car is an Internal Combustion Engine (“ICE”), for every 100 units of Primary (hydrocarbon) Energy, it will end up with something like 25 units used in the useful work of moving the vehicle. There will be losses in the transportation and refining stages as oil is turned into gasoline at moved to the local pump, but the big chunk of the “lost” 75 units will be via the inefficiency of the ICE – where a great deal of energy is dissipated as heat and noise.

Contrast this to an Electric Vehicle (“EV”) where the direct conversion of electrons into motion is reported as being in the 90% range. In this example we could surmise that one needs something like 4-times less Primary Energy in the form of electrons than we do of molecules to have the same useful energy output, i.e. to get the same work done. 

And ultimately, what we are interested in is the work that gets done, or the “useful energy”, not the Primary Energy.

We often see charts of current global energy – with oil, gas and coal making up 85% of all global Primary Energy and low carbon energy (Hydro, Nuclear, Wind and Solar) nibbling at this. 

In the last 30 years the amount of Primary Energy has increased by 50% and the amount provided by fossil fuels has dropped by only a percentage point or two.  

In the above data, the total Primary Energy was 576 EJ of which oil, gas and coal made up 487 EJ whereas renewables (wind and solar) made up only 27 EJ. When looked at this way it seems impossible to replace this gargantuan amount of energy, currently supplied by fossil fuels with renewable alternatives within any realistic time-frame. All of the efforts in the last three decades have got renewables to a paltry 4-5%

However, in this view we are potentially misrepresenting the challenge. It can be argued that we will never need to replace all that fossil fuel Primary Energy, only the “useful energy”. 


If this is the case then potentially as little as 25% of the Primary Energy needs to be replaced – due to the efficiency argument I outline with the example of ICEs vs EVs above. Using the above numbers fossil-fuel Primary Energy supplies 122EJ of useful energy (487 EJ x 25% efficiency factor), whereas low-carbon sources supply 84 EJ (89 EJ x 95% efficiency factor). 

Thus – and I insist that this is theoretical and imho, wrong – one could argue that by less than doubling current low carbon energy we could match the world’s requirements of getting work done. 

Suddenly, with this point of view, the energy transition looks a whole lot more likely. Would that it be so.

For those of you who know me, you’ll know that what comes next is a “yes, but….”

Yes, but….

Sadly, the simplistic example given above, whilst reasonable (and reasonably accurate) has one big assumption – and that is the perfect timing of the generation of the electrons and the conversion of these electrons into work (eg movement). Without a perfect match of these two events (generation and usage) you must consider storage – electrons don’t just wait around.

And storage is neither simple, nor is it energy free. You must either use chemical storage (batteries, hydrogen) or gravity potential (pumped storage) or even less likely are heat storage (molten salt), fly-wheels and compressed gasses. None of these is efficient – as I have said before – “Entropy is not your Friend”.

Or more generally, thermodynamics are not your friend. Whilst being a fiendishly difficult area of study – it is conceptually quite simple. You do not get “something for nothing” – or put another way, when you convert energy from one form to another there are irreversible losses in the transformation process. These losses are not energy being destroyed – that would contravene the First Law of Thermodynamics which states that energy cannot be created or destroyed. However, if we think of the losses as wasted energy or reductions in useful energy then this conforms to the Second Law of Thermodynamics: which can be paraphrased as transformations increase entropy (or reduce useful energy). In the physical world of our energy systems we can think of how an internal combustion engine outputs a lot of heat and noise (wasted energy) as well as movement (useful energy). It is not possible to gather the wasted energy and recombine it with the useful energy to get back to 100%. 

Also a second order problem of transmission – generation sites are usually not perfectly situated next to large concentrations of demand – so there has to be transmission which also has losses. But generally these are of second order unless the distances are huge.

This is also why projects like solar in the Middle East are very efficient and very cheap – you generate maximum electricity at noon and you need maximum electricity, which is dominantly for air conditioning from noon onwards – there is a pretty good match. The energy efficiency is high, the costs are low and everyone is happy. As I have noted previously here – taking ideal examples and generalizing from them is an unhelpful game – the ultra low prices in Dubai or Abu Dhani for solar electricity have little relevance for northern Europe.. We can see this when a very similar situation started to fail – California summer of 2020. In this case, the problem was ambient temperatures and demand for A/C continuing into the late evening well after the sun set. The mismatch of peak generation and peak demand – by even just a couple of hours (coupled with more easterly neighbours who could have supplied power also being dark… duh) was enough to trigger rolling black-outs.

Not surprisingly the current focus is on management of the mismatch in timing of generation and use. In more extreme cases (discussed here) we need to time-shift electrons from warm sunny summer days to cold, dark winter nights. Batteries are orders of magnitude short of this ability on national scales – so the world turns to Hydrogen as a carrier fuel.

And here we get into poly-chromatic nightmares. But let’s just stick to the idea of Green Hydrogen – that is, hydrogen created by electrolysis of (purified) water using electricity from “green” sources. This has been very well described by @davidcebon here in a post entitled “Hydrogen for Heating“.

Of specific interest in the context of this discussion is the quantification of energy losses through the conversion chain.

Not surprisingly the losses are significant and the implication is that you would need (in technical terms) a shed-load more Primary Energy to offset the losses in the conversion and storage processes.


This point is neatly summarized in the BP Energy Outlook 2020

The outlook for primary energy also depends on the form in which that energy is used at the final point of consumption. In particular, although it is possible to decarbonize the production of electricity and hydrogen, they require considerable amounts of primary energy to produce. As such, increasing the use of these forms of energy carriers tends to boost primary energy.


Whilst it may be misleading to present the Energy Transition challenge as a like-for-like replacement of low quality thermal energy by high quality electrical energy as represented by Primary Energy, it is equally misleading to assume that the higher quality electrical energy can be accounted for whilst abstracting the energy costs in ensuring that it is genuinely useful when and where we need it. Ultimately there is no free lunch, and sadly, you won’t even get what you pay from. Blame Thermodynamics.


In any transfer or conversion of energy within a closed system, the entropy of the system increases. The consequences of the second law can thus be stated positively as the spontaneous or natural direction of energy transfer or conversion is toward increasing entropy, or negatively as all energy transfers or conversions are irreversible. Or, in keeping with our paraphrasing of the FLT as “You can’t get something for nothing,” the SLT asserts: “You can’t even get all you pay for”.

Energy, Entropy and Exergy Concepts and Their Roles in Thermal Engineering