Saturday, August 1, 2009

A new graph, showing countries' power per unit area

When I gave my energy talk in Cambridge two weeks ago, one member of the audience objected to my figure (page 13) showing per capita emissions by country. It would be fairer, she said, to show the emissions or energy consumption of each country per unit land area. (Guess her nationality... Australian!) I've made a few figures following her suggestion, and I'm displaying my favourite here. This figure shows population density on the horizontal axis and power consumption per person on the vertical axis. The diagonal green lines indicate the power consumption per unit land area, in W/m2. This is precisely the same unit in which I measured or estimated the power per unit area offered by renewables (page 112). Most renewables offer between 0.5 and 5 W/m2.
Conclusion: All countries whose power consumption per unit area is bigger than 0.1 W/m2 are countries who should expect renewable facilities to occupy a significant intrusive fraction of their country, if they ever want to live on their own renewables. Countries with a power consumption per unit area bigger than 1 W/m2 (eg UK, Germany, Japan, Netherlands, Belgium) would have to industrialize most of their countryside, if they want to live on their own renewables. Alternatively, their options are to radically reduce consumption, use nuclear power, and/or to buy renewable power in from other countries.

(Image can be downloaded from here).
PS - I posted another article about this diagram in 2013


LG said...

A very interesting representation.

You allude to is in the book but it would be interesting to see this kind of graph in terms of consumption rather than energy use, i.e. the "energy balance of trade", to include the energy used because of the existence all of those cheap Chinese-made washing machines we import etc. (though working this out to a meaningful degree of accuracy would be horrendous I imagine... a project for the next book!)

DanH said...

My eyes were drawn to the horizontal, rather than the vertical, axis of the graph.
I'm slightly surprised that the range of population densities of different countries is "only" just under three orders of magnitude. For unrelated reasons, last week, I was looking up the population densities of an accidental sample of different local council areas within Cambridgeshire, and found a factor of 160 variation there. It's not just the urban/rural divide, either. There was a factor of 55 variation between different rural areas. All this suggests that variation in population density within nation-states could be as interesting as variation between nation-states.

Why does any of that matter to sustainable energy? Because I have a feeling that there will be an assumption that, within a nation-state, areas of (perceived?) low population density have a duty to help out their more densely-populated neighbours with sustainable energy. That assumption is implicitly present in the "visualizing wind farms for Cambridge University" post on here. Yet the standard version of "contraction and convergence" doesn't allow for a similar assumption at the international level. I take no particular view at this stage on whether the assumption is a good idea - but we do need to be clear about why it applies within national borders, but not across them.

I keep forgetting to say this every time I post on here, and it's about time I remembered: David, congratulations on your FRS.

Unknown said...

I do like the W/m^2 required and your 0.1 and 1 points...



Andrew Smith-Gibbs said...

Why just land area, given that most of the UK's renewable resource is out at sea? It might be mistaken for a piece of misdirection to help maintain and propagate the myth that Britain doesn't have enough renewable potential to easily meet all its energy needs.

The figure for power consumption per person in that figure would seem to include the huge amount that's wasted by thermal generation of electricity: actual UK energy demand is below 100kWh/p/d, isn't it? It's only over that if you include all the heat that gets wasted in cooling towers from inefficient plant like gas (45% efficient), coal (40%) and nuclear (around 30%).

The land area of the UK is, what, about 240,000km²

But the Exclusive Economic Zone [EEZ] around Britain (according to wikipedia, at least) is about three times that.

And the total EEZ for the UK is about 3.9 million km², or sixteen times the land area! Which puts the UK energy density way down below 0.1W/km²! Now, that doesn't mean we should build wind & wave farms in seas around the Virgin Islands and pipe it via HVDC from there to Britain - there's no need, we've got more than enough potential renewables close to home to meet all our needs. But it does mean that there's enough potential renewables in the seas around UK Overseas Territories and Dependencies to meet our "energy balance of trade" too.

Andrew Smith-Gibbs said...

sorry, my final paragraph above should refer to 0.1W/m², not 0.1W/km², obviously

Unknown said...

The energy storage problem inherent in solar and wind powered plants should be addressed at some point. Storage facilities for intermittent energy production plants can greatly increase the required land area and environmental impact.

The common solution is to have sufficient energy capacity from fueled plants to handle the entire load.

David MacKay FRS said...

Dear George,
thanks for writing. Please read my book. It is free online and it has got a chapter on fluctuations of demand and of renewables.
My suggestion is that the best way to handle both sorts of fluctuation is to have very large pieces of new demand that are easily turn-off-and-on-able. (eg Electric vehicle charging, and heat pumps putting heat into storage.)

Andrew Smith-Gibbs said...

there are lots of ways to deal with variability. Measuring how much you need requires a system model, but we've got a large armoury available: pumped hydro storage, a smart grid, thermal storage (either cheap & easy resistance heating or expensive and relatively short-lived heat pumps, into hot water storage tanks), electric vehicles, maybe with a possibility of feeding power back into the grid (V2G). There is an overview in SEWTHA, and there's an awful lot of in-depth material out there too.

One of the easiest ways for Britain to manage it is to build bigger interconnectors to the continental European grids, as between them, they have enough dispatchable hydro for a month's electricity consumption across all of Europe (see work by Gregor Czisch). The other way that big interconnectors help is that the bigger the geographic area you spread your variability over, the lower the variability is relative to mean power (see work by Graham Sinden, Hannele Holttinnen, Czisch) - so by combining our wind and solar profiles with that of the rest of Europe, the Middle East and North Africa, we can still generate sufficient energy from British renewables to meet all our energy needs, and the variability gets balanced by lots of imports balancing against the same or higher exports.

The larger the grid, the less that variability is a problem. Having two British grids would make the whole variability problem much worse, and that's one of the areas where SEWTHA could do with a correction for the second edition.

D said...

I did have a few concerns on reading the book although I liked the format that set out a bar chart of supply next to one of demand.

Heat pumps and CHP are totally equivalent in thermodynamic terms. In practice CHP systems have fewer heat exchangers and this makes energy storage easier since large district-scale hot water tanks are far easier and cheaper than .

Town-scale CHP also features larger machines, which are more efficient, as engineers know. It has lower CO2 emissions than heat pumps operating on the same fuel (gas), as anyone can see looking at the operating figures from a country such as Denmark; CHP can be equivalent to a heat pump of COP 10 or 12.

Agreed, we need to consider the future, with less availability of gas. But heat grids can be decarbonised as easily as electric grids and the CO2 emissions from heat pumps aren't yet dramatically gas less than from a good boiler.

On this topic, there was no mention at all of solar district heating and seasonal storage. This was in use experimentally in Sweden as early as 1978-83 and Marstal in Denmark has 18,000 m2 solar collectors which help to heat the town.

The tidal calcs. are wrong according to some postings I saw on the Claverton Group forum. Maybe this needs checking. I agree tidal looks a pretty good bet though; at least it's concentrated and controllable.

No detailed coverage either of anaerobic digestion of household, kitchen, restaurant, commercial & garden wastes. Swiss authorities were saying from memory in the mid-2000s that with a big effort this waste could supply up to 7-10% of overall energy use. It has a double benefit as it avoids putting these "putrescible" wastes into landfill and subsequently causing methane emissions.

Very disappointing to see the lack of detail given to energy efficiency. On efficient use of electricity there was only a little about standby and about CFLs and LEDs. Meanwhile the US Energy Secretary Stephen Chu has been commenting at almost every opportunity on the importance of this subject.

BTW turning down the thermostat isn't energy efficiency, it's curtailnment. I think it's positively defeatist to suggest this, given that the UK already has the coldest homes in Europe and suffers a high number of excess deaths among the elderly in winter. These appear to be directly due to living in a cold environment.

The ECI/Univ. of Oxford "40% House" Report 2005 went into more efficient use of electricity in more detail. But even that report needs updating for technology that wasn't on the market then and is now.

Typically, for a study which puts energy efficiency first we have to look abroad to reports such as "the 2 kW society" in Switzerland and the EU-25 energy scenarios from now to 2050 published by INFORSE in Denmark.

Unknown said...

@D, please could you explain to us how heat pumps and CHP are thermodynamically equivalent. Thanks.
And how CHP will run when there is no more cheap gas.

Also, regarding energy efficiency of course it's important, but fundamentally we need a change of *supply*, because of Jevon's paradox ( - also known as the Khazzoom-Brookes postulate or the boomerang effect. So I think the book's focus on the end game is absolutely bang on.

DanH said...

Since Jevons' paradox has come up....

Jevons' paradox is a theoretical prediction, yes? It predicts that increasing the efficiency with which a resource is used causes an increase in total consumption of that resource.

But I wonder how the prediction holds up against empirical data. I'm thinking particularly that the observed hysteresis of the curve of per capita energy consumption against per capita GDP, of nations as they go through economic cycles - especially, for example, the UK in the early 1990s or Russia either side of 1998 - suggests rather non-Jevons-like behaviour.

Unknown said...

Dan, I don't have data to hand about this but I'm pretty sure it holds up - to a certain extent. In other words, if you imposed severe efficiency improvements, you would eventually see a positive impact. But small or incremental improvements, maybe not. Perhaps there's an example in the auto industry? I'll have a think. Your example of economies and energy consumption being correlated is not about efficiency, it is about the fact that economies are basically big machines which take (natural) resouces and turn them into 'value'. Check out this paper "Accounting for growth: the role of physical work" by Ayres et al ( Extremely interesting, basically shows that a big reason our economy grows is because there is an energy input, from thermodynamic work, from burning fuels. Unfortunately economists though mostly think that it's just magic... (meanwhile we run out of resources).
This is at the heart of our problems. What we actually need is an economy that is not based on growth. Is such a thing possible? I don't know. Some thinking here but I have yet to read it.

DanH said...


Annoyingly, I can't lay my hands on the original data set of per capita energy consumption versus per capita GDP either, but I can describe it.

You're right to observe that the first-order feature of the data is a positive - I think pretty closely linear - correlation between per capita GDP and per capita energy consumption, across a wide variety of different nations and different times. Your explanation of that feature sounds like a good one.

However, what I wanted to draw attention to was a second-order feature that sometimes appears in the data for a single nation, over a time domain corresponding to a single economic cycle. There's a hysteresis loop: as an economy goes into recession, both per capita GDP and per capita energy consumption drop. Then, as the economy climbs back out of recession, per capita GDP grows all the way back up to its original level; per capita energy consumption also rises, but not all the way back up to its original level.

Hence, at the end of the cycle, you have a higher efficiency (in the sense of GDP divided by energy consumption) than at the start, along with a lower energy consumption, while the extraneous independent variable (GDP) is unchanged. That suggests to me that, at least in some circumstances, real economies don't quite behave as Jevons suggested.

To digress a bit: for me, alarm bells ring at the suggestion of 'an economy that is not based on growth'. In the IPCC Special Report on Emissions Scenarios, the two scenario families with the least economic growth over the twenty-first century are the ones that are expected to produce the highest carbon dioxide emissions in 2100. That's primarily because severely restricted economic growth leads to poverty, which in turn leads to runaway population growth.

Unknown said...

Alarm bells are ringing majorly for me too. But I still think that you can't have infinite growth on a finite resource. At the end of the day we have one, finite planet, so sooner or later we will hit the buffers. We're already digging into the capital rather than living off the interest. I like this quotation which sums it up neatly: "The greatest shortcoming of the human race is our inability to understand the exponential function."
This applies to any economic system which relies indefinitely on growth in the exploitation of natural resources to keep it going, which ours does.

Unknown said...

p.s. I've just spotted this but haven't had time to read it yet:
By the way are you an engineer/physicist or economist? Or none of these?

DanH said...

Thanks for the Sustainable Development Commission reference. I'll take a look.

I'm a physicist, although as you've probably gathered, I like to stick my nose into lots of other disciplines too.

Anonymous said...

Power consumption per person is not nearly as relevant as power consumption per GDP.
FOR INSTANCE: person in country A uses a gallon of kerosene to heat their home/cook meals for a week, person in country B uses a gallon of fuel to move a ton of freight 450 miles via rail.
Person on farm 1 uses fuel but grows no food. Person on farm 2 uses the same fuel to grow food.

So the measure of efficiency is actually what motivates persons to actually take action with capital to produce something with the energy which they use.