Primary vs Final Energy

An Eccentric Anomaly: Ed Davies's Blog

Here's a correction to my previous Is PV Enough? post.

In that post I wrote:

Currently the average European uses about 5 kW of energy.


On the other hand, primary energy is, I think, a bit more as it includes waste heat from power stations and so on.

A more-recent Twitter conversation with Kees van der Leun points to this EU statistics image which gives the final energy consumption for the EU-28 for 2015 as 1084 million tonnes of oil equivalent.

One tonne of oil equivalent is approximately 42 GJ. 2015 was a common year, there are 86400 seconds in a day, and the population of the EU is about 510 million people so that's

>>> 1084e6 * 42e9 / (365*86400) / 510e6

an average of 2.8 kW per person.

That's considerably lower than the 5 kW/peep figure I gave in that previous post which must, I now think, be primary energy use. Therefore, the areas calculated in that post are an overestimate and should be some 44% smaller, which is good news.

The difference between primary and final energy is the extra which gets used in delivering energy to the end user: for example, the waste heat emitted by most thermal (e.g., nuclear, coal, gas) power stations and the energy used in refining and transporting petrol and diesel fuel.

The definition is rather arbitrary. Final energy for petrol is, I understand, considered to be that available in the fuel as it's delivered on the forecourt. The waste heat which goes out of the exhaust pipe and radiator (about 1/3rd each of that available in the fuel for your typical fossil burner, though a bit might be used for cabin heating) is counted in the final energy.

But, the primary energy from photovoltaics is the DC current coming out of the panels, not the sunlight going in. This is fair in a way as the primary energy from oil is considered to be the liquid pumped out of the ground, not the sunlight which fell on the leaves in the Jurassic or whenever.

Anyway, concentrating on just the EU-28, how much PV would be needed to generate all these final energy needs? Omitting the final division by 510 million in the above calculation gives an average power requirement for the region of 1.44 TW.

According to PVGIS a 1 m² (150 W) panel lying flat near Almería­ in southern Spain would produce 0.28 kWh/day in December (the worst month of the year). That's an average of 11.7 W so to produce 1.44 TW we'd need an area of 123 thousand km².

That's 241 m²/peep which is a bit more than my previous estimate for the world of 200 m²/peep which was itself an overestimate. However, here I'm taking southern European insolation rather than that of slightly sunnier countries to the south and also taking the worst month of the year on the slightly dubious assumption that more PV is cheaper than more inter-seasonal storage.

And that's also ignoring the benefits of other sources of renewable energy which complement PV. Wind in northern Europe tends to mirror sun in southern Europe reducing the amount of PV needed for the worst part of the year.

That area is not small. But it's worth some comparisons with other areas we use. This NationMaster page weirdly gives two different figures for the number of kilometres of road per 100 km² for the EU: 169.95 and 121.79. Since it gives the Eurozone separately it's not obvious what's going on here but taking the lower of those, noting that the EU has an area of 4.475 million km² and guessing at an average road width of 6 metres the road area would be:

>>> 121.79/100 * 4.475e6 * 6/1000

33 thousand km². If we can afford that much space for one mode of transport then we can afford a bit less than four times the amount for all our energy needs, particularly as a significant proportion can be shared with other uses such as building roofs.

(By the way, I'm yet to be convinced that using actual road area for PV is not a nitwit idea.)