This group brings together the best thinkers on energy and climate. Join us for smart, insightful posts and conversations about where the energy industry is and where it is going.


Do Renewables Lower Energy Consumption?

Robert Wilson's picture
University of Strathclyde

Robert Wilson is a PhD Student in Mathematical Ecology at the University of Strathclyde.

  • Member since 2018
  • 375 items added with 247,413 views
  • Feb 18, 2014

Renewables and Consumption

If a coal power plant is closed and replaced by a wind farm and that wind farm produces the same amount of electricity as the coal plant, what happens to energy consumption? Depending on the coal power plant, and how you define energy consumption it could do anything from stay the same to falling by a factor of three. Welcome to the perplexing world of measuring energy consumption in an increasingly renewables world.

In popular discussion the phase “energy consumption” is used with little regard for its meaning. Yet, in many respects it is a problematic term. Take the European Union. It has a target of getting 20% of its energy consumption from renewable energy by 2020. Let me put that more accurately. It has a target of getting 20% of its final energy consumption from renewable energy. The word final is key, yet you will almost never see it appear in discussions about renewable energy targets.

There are two ways of measuring total energy consumption: primary and final energy consumption. Primary energy consumption is a measure of the energy content of all the oil, coal, natural gas etc that is taken out of the ground. Essentially we ask how much energy is released when we burn the stuff. Typically this is reported in tonnes of oil equivalent, that is how many tonnes of oil would release the same amount of energy if burned.

Final energy consumption is slightly different. It is the energy delivered to the final consumer. For example it only measures the electricity that is produced by a power plant, it does not care about the heat produced from burning coal or natural gas that was not converted to electricity. So, for a 50% efficient power plant the primary energy consumption is two times higher than the final energy consumption.

This sounds simple enough. But here is another problem. How do we measure primary energy consumption for renewable sources such as wind, solar, and hydro-electricity? In the case of coal we can ask how much energy is released when we burn the stuff. Obviously we do not burn anything for wind, solar and hydro. So, what do we do? Here we have two choices. We can use the energy content of the electricity generated as the primary energy. This is called the “physical energy content” method, and is used by groups such as the International Energy Agency. The second choice is to ask how much fossil fuel energy would have been required to produce the same amount of electricity. This is called the partial substitution method, and is used by BP in their often cited Statistical Review of World Energy. In the case of BP they add up all of the wind and solar electricity generated and convert it to primary energy assuming that it would have been burned in a 38% efficient fossil fuel power plants, that is BP say primary energy consumption from wind and solar is more than two times higher than the IEA does.

Which is correct? The correct is answer is neither. What we should really ask is what measure is most appropriate for the question we are trying to answer.

Let’s imagine that we are trying to measure the energy efficiency of an economy. Typically this is done by recording its energy intensity, we just divide energy consumption by GDP. Now further imagine that a country was getting all of its electricity from coal power plants, and this represented 50% of its primary energy consumption. What would happen to its energy intensity if we replaced it all with solar?

If we used the IEA’s definition of primary energy consumption the energy intensity would improve by more than 25%, simply because the primary energy consumption from electricity generation has more than halved. If we used BP’s measure then things would stay where they are. Neither answer is satisfactory. After all we could use 5, 10, 15 or 20% efficient solar panels to get the job done. The efficiency of generating solar electricity then is completely irrelevant to how we measure how efficient the economy is.

In a similar vein consider two rather simplified purely electric energy systems, both with annual electricity demands of 100 terawatt hours. One gets all of its electricity from 35% efficient coal power plants, while the other gets them from 15% efficient solar panels. Now, replace those 35% efficient coal power plants with 40% efficient coal power plants and what happens? The energy intensity of the economy improves. But replace those 15% efficient solar panels with 20% efficient solar panels and absolutely nothing happens to energy intensity. Again, this is somewhat unsatisfactory.

To illustrate these points I will finish by considering Denmark, the country which gets more of its energy from wind farms than anywhere else.

Below I have plotted the change in Danish total primary energy consumption between 1990 and 2012 (from BP’s primary energy statistics) using the partial substitution and final energy content methods for wind electricity. If we use the partial substition method then Denmark’s primary energy consumption is essentially unchanged since 1990. However if we use the physical energy content method primary consumption has declined by almost 9%.


Now, you may be tempted to conclude from this that using the physical energy content method is a good thing if you want to promote the benefits of wind farms. But consider what percentage of primary energy consumption comes from wind farms using the partial substitution and physical energy content methods. It is two times lower using the physical energy content method.


In this case the partial substition method appears to be a much better way to track the changes in wind power penetration, otherwise we are saying a unit of fossil fuel electricity is worth two times more than wind in primary energy terms. Of course if you wanted to downplay the growth of wind farms it should be obvious how to do it.

So, as the cliche goes, “lies, damn lies, and statistics.” The above can be read as a way to properly deconstruct energy consumption statistics, or as a guide to how to misuse them. Please do the former.

Robert Wilson's picture
Thank Robert for the Post!
Energy Central contributors share their experience and insights for the benefit of other Members (like you). Please show them your appreciation by leaving a comment, 'liking' this post, or following this Member.
More posts from this member
Spell checking: Press the CTRL or COMMAND key then click on the underlined misspelled word.
Joris van Dorp's picture
Joris van Dorp on Feb 18, 2014

Good article. You make clear that using consistent terminology is important when trying to identify trends in efficiency and energy intensity.

I want to add that the energy return on energy invested (EROEI) of the particular generation technology itself can also be a factor in determining how such trends will develop. For example, the EROEI of silicon PV panels in Germany is about 5 to 1 (Weissbach 2013), meaning that it takes about one unit of (primary) energy to produce 5 units of (final) energy with PV. So if Germany would install enough PV panels to satisfy it’s current final electricity consumption, it would actually need to add an additional 25% to that total number of PV panels in order to be able to also produce the energy needed to produce the PV panels in the first place. In effect, 20% of Germany’s total annual energy production would then go straight back into PV production, installation and maintenance.


In other words – in terms of final energy – switching from a high EROEI energy system (coal, natgas, nuclear, etc.) to a low EROEI energy system will significantly increase final energy usage, all things remaining equal.

The total amount of energy production capacity P required in a self-sufficient society when using a technology with a given EROEI is: P = 1 / (1- (1 / EROEI))

When EROEI is high (such as for nuclear power which has an EROEI of around 100), P goes to 1, which means that little if any additional production capacity is needed in order to enable the technology inside a self-sufficient society.

Wind energy is also quite good, with reported EROEI’s of between 20 and 50 at the point of production, as far as I know. Of course, the need for supporting infrastructure including electricity storage, backup and dedicated transmission severely impacts the EROEI of high-penetration wind, bringing it down to levels far below that of traditional energy generating technologies.

PV is still worse, with its EROEI obviously being strongly dependent on location (insolation). For Germany, it is between 4 and 6, as far as I know. In countries closer to the equator, it can be double that.

Needless to say, if a society wants to switch to ultra-low EROEI energy technologies such as corn ethanol (EROEI ~1.2?), it’s final energy consumption will rocket up to 600% of it’s original final energy consumption. Most of that energy will then be consumed entirely in order to produce the energy. In other words, in such a society, for each barrel of biofuels delivered to society, five barrels of biofuels would be delivered to the industry producing the biofuels in the first place.

Mind you, the current energy policy of Germany means that if Germany actually decided to move to a 100% solar PV energy system (which it doesn’t and won’t), then it would not actually need to build the additional 25% PV panels to power the production of the necessary solar panels. That is because the PV panels installed in Germany are mostly produced in China on the basis of energy from coal burning. Only if Germany decided that it wants to produce its own PV panels, then it’s final energy would increase by 25% in order to supply enough energy to it’s PV industry and society at the same time.

Robert Wilson's picture
Robert Wilson on Feb 18, 2014


I’ve rarely seen commentators use those statistics, except when using Twitter to talk about real time wind power production. Mostly I just see DECC statistics being quoted, which covers all renewables because they have to keep track of what is getting paid subsidies.

Joris van Dorp's picture
Joris van Dorp on Feb 19, 2014

I doubt the hunter gatherers would be left alone in such a scenario. Even at high penetrations of non-biomass RE, there will be so much demand left over for biomass to fill in that much if  not all global vegetation would be appropriated for energy production. I happen to believe that global demand for energy will increase to at least three times today’s demand through the year 2100, and probably to five times today’s demand through the year 2200. This is assuming that global GDP will rise twenty-fold ultimately (in order to meet the needs and aspirations of 10 billion people) and that every economical opportunity to increase energy efficiency of global GDP is implemented.

Without nuclear power, due to wind/solar intermittency, biomass will be supplying at least 25% to 50% of this ultimate energy demand, assuming the effort to leave fossil fuels in the ground in order to prevent catastrophic climate change succeeds. 25% to 50% of 5*500EJ is still between 750EJ to 1250 EJ of energy. This demand is far more than the 250 or so EJ of biomass energy that is said by most researchers to be sustainably retrievable from the biosphere for human consumption. The high estimate – 1250 EJ – is more than even the most optimistic projections of ultimate achievable sustainable biomass production.

What it boils down to is that if nuclear is ignored in favour of biomass, the hunter-gatherers stand little chance. They inveitably will find bulldozers coming in their direction sooner or later, clearing the land for monoculture biomass plantations of stupifying proportions.

Robert Wilson's picture
Robert Wilson on Feb 19, 2014


Please tell me where I have misapplied the partial substition method. The statistics I quote in the piece come direct from BP, who apply the partial substition method themselves. If I have made a mistake it might have been with the physical energy method, but I can’t see how I have done this.

Robert Wilson's picture
Robert Wilson on Feb 19, 2014


As I state in the piece primary energy consumption covers everything, and that includes energy used in mining. Plus the issue discussed in the piece is the difference between different primary energy consumption measures. Energy extraction from mines is of little relevance to this.

Joris van Dorp's picture
Joris van Dorp on Feb 20, 2014

I completely agree with you.

Except on one point. Fossil fuels are in fact abundant and cheap enough to ensure that we will extract and burn enough of them to trash the climate if nothing is done, which actually would cast doubt on humanity’s ability to reach a future economy that meets the needs and aspirations of all people, at a level that middle class peoples in western nations are accustomed to. These fossil fuels really are and will remain cheap, compared to zero-carbon alternatives.

While it is conceivable that nuclear power can compete with these fuels in time to make cheap fossil fuels obsolete, to reach that solution requires far more active participation of the public in terms of educating itself and demanding from politicians rational and effective regulation of nuclear power in support of highly cost-effective and safe nuclear power. Sitting back and expecting that nuclear power will inevitably make fossil fuels obsolete in time to stop climate disruption is a dangerous mistake, IMO.

Gary Tulie's picture
Gary Tulie on Feb 20, 2014

If we were ONLY talking of keeping one good light on per person (equivalent to 100W incandescent), then with the most efficient LED lights we could do that for four hours a night by using a peddle powered generator for around 25 minutes a day. 

Regarding the underlying question – Do renewables lower energy consumption? The answer should include consideration of whether or not the use of distributed generation i.e. domestic solar arrays causes users to rethink their energy consuming behaviour and take measures like controlling their energy use more carefully, changes to things like their lighting etc in order to reduce consumption. 

In the case of the UK, I would saythat it is likely that many of the people installing solar will also take a look at how they can use energy more efficiently – if only to ensure that they get a good enough home energy performance certificate to be able to claim the full rate of feed in tariff.

Clifford Goudey's picture
Clifford Goudey on Feb 21, 2014

No you don’t.  You state, “Primary energy consumption is a measure of the energy content of all the oil, coal, natural gas etc that is taken out of the ground.”  That’s the chemical energy content and says nothing about the energy required to mine it, transport it, refine it, and deal with the emissions and residues of combustion. 

Energy associated with these associated processes are essential to understanding the energy efficiency of an economy.  As Victor says, your analysis is incomplete and you are wrong to dismiss the comment.

Thomas Gerke's picture
Thomas Gerke on Feb 21, 2014

Nice article. 
It’s important that people understand what the different energy statistics are and how to use them.

Primary energy statistics are a great basis for discussing fossil fuel consumption, thus they are relevant to the carbon / climate change discussion. They are however a bad measure for the relevance of any non-fossil primary energy source. (This is also the reason why many institutions switched from the substitution methode to the physical energy content methode I think)

When discussing the current state & future of our energy system, it’s usually better to look at the final energy demand. Here we can discuss how efficiency measures and changes in ways of energy consumption (EVs, heatpumps) change our actual technical energy needs. Based on this discussion it’s possible to discuss how much primary energy is / would be needed to meet that demand. 

Keep it up.

Besides the intricacies of energy statistics, wind power requires significant less own-use of electricity compared to coal / nuclear power stations. That way they actually reduce gross consumption of electricity.

Clifford Goudey's picture
Clifford Goudey on Feb 22, 2014

Joris, it is fanciful to compare the EROEI of fossil fuels to those of renewables, as the standard approach fails to include the embodied energy in the material extracted.  Applied to renewables, EROEI is a measure of the energy associated with the technology’s manufacture, installation, O&M, and decommissioning.  The fuel is free, fleeting, but infinite.  EROEI of fossil fuels is generally embodied energy divided by that used to extract it.  The conversion of that fuel into useful thermal or electrical energy is another matter. 

I believe your stated EROEI for nuclear power must be for the mining process alone.  Figures I have seen that include the energy embodied in the nuclear facility are below 15.

Bob Bingham's picture
Bob Bingham on Feb 22, 2014

In my case it does save energy. I have solar panels on the roof which almost elininate a power bill but it does make me very concious of the power that I am using. I now turn everything of at the plug and  keep a close watch on all my consumption though a serious of meters.

Rick Engebretson's picture
Rick Engebretson on Feb 22, 2014

Very good article and comments. As others suggest, this could be greatly extended to the end use demand side of consumption. For electricity, transmission line loss, even conversion loss of AC voltage to suitable DC for LEDs, television, etc.

This line of thinking is certainly not new. Online shopping is low energy, optical discs put a theater in your home, etc. Giving the consumer more for less has exceeded what many of us can absorb.

But, in my experience, skillful system analysis like this is rarely provided in energy future advocacies.

Joris van Dorp's picture
Joris van Dorp on Feb 23, 2014

Sure, EROEI calculations are particular to each of the various energy technologies. But the definition is clear enough. It’s the ratio of energy inputs (in all stages of the complete lifecycle) but excluding the natural embodied energy in the fuel, sun, wind, plants, ocean heat, coals, gas, oil, tar, etc. For any process yielding energy in any particular form, an unambiguous EROEI number can be calculated. This does not mean that there are no incorrect or incomplete calculations.

I’ve read a number of report giving EROEI results for various technologies. They are easily googled. My figure for nuclear is based on several studies, but it includes the assumption of 80 year plant lifetime and a closed fuel cycle (fast reactor or thorium breeder). Different assumptions yield different results. For an open fuel cycle, LWR, and ultracentrifuge enrichment, the nuclear EROEI figure that I use is 50 to 75.

Your figure of 15 for nuclear is most probably for LWR (i.e. conventional) nuclear reactors and diffusion enrichment of U235. Your assumed plant lifetime is (I am guessing) no more than 40 years.

EROEI of uranium mining is huge, far higher than 100, AFAIK.

Robert Wilson's picture
Robert Wilson on Feb 23, 2014


Now you are just being silly. The energy required to get coal out of the ground comes from where? It comes from coal, oil or natural gas that we took out of the ground. We can have vacuous arguments about semantics, or we discuss substance. Your choice.

Eirik Johnson's picture
Eirik Johnson on Feb 24, 2014

   There is no honest confusion in the issue Mr. Wilson raises here.  The opportunity for “lies, d#$&%! lies, and statistics” rests only in using the term “energy” to mean “fuel,” as industry propaganda so often does.  The phrase “energy consumption” honestly and unambiguously means what Mr. Wilson calls “final consumption.”  What he calls “initial energy consumption” is in fact fuel consumption, and ought not to be called “energy consumption” at all.   If a coal plant is 50% efficient, and produces enough energy for 100 terawatt-hours of energy, then it burns 200 terawatt-hours equivalent of coal.  If the consumption were to switch to solar or wind, then energy consumption would not change but fuel consumption would fall by close to 100% (making and operating the equipment to collect the energy would probably consume a little fuel, so the savings are not quite 100%).  The efficiency of the technology is irrellevent to the concept of consumption.  If a solar collector is only 5% efficient, it nonetheless burns little fuel and is thus almost 100% more fuel-efficient than is any fuel.

     Bottom line: honest folks should never use the term “initial energy consumption” at all, and never use the term “energy consumption” to mean “fuel consumption.”  Renewables are renewable sources of energy, not usually renewable fuels.  Renewable fuels such as firewood do exist, but they are almost as insignificant as industry propaganda typically assumes.  In fact it’s all fuel, not renewables, which in the future of energy will enevitably become insignificant.  Honest discussion must transcend jargon based on fuel consumption as the normal form of energy production.   


Robert Wilson's picture
Robert Wilson on Feb 24, 2014

Industry propaganda? These terms are widely used and discusses in the academic literature. The term energy consumption is not industry propaganda in any sense, unless you are a conspiracy theorist.

Robert Bernal's picture
Robert Bernal on Feb 25, 2014

Detracting from the main issue, different plants will require different amounts of material inputs. The following “cherry pick” provides detailed info.

Of particular interest is the first graphic concerning steel inputs. It compares various different power (and nuclear) plants. I believe the more power dense the source, the higher the EROEI. Exceptions would be very lightweight power plants such as the gas turbine (gas input not figured in?) and super light weight (futuristic?) non energy intense materials for the renewables.

It is very hard to believe that a nuclear plant which provides many gigawatt years from just thousands of tons of material and fuel has an EROEI of only 15. The closed cycle would fare much higher on the eroei scale because it requires on the order of 100 times less fuel AND even less material housing than a conventional LWR.


Robert Bernal's picture
Robert Bernal on Feb 25, 2014

I tend to think that wind is better than PV if it has a better CF and a better eroei. Nevertheless, I know we need it all to become less expensive (perhaps in the future using advanced lightweight materials applications?) in order to even come close to matching the advantages of closed cycle nuclear.

Robert Bernal's picture
Robert Bernal on Feb 25, 2014

Admittedly, I am detracting a bit, but want to know if the primary energy input is used when “they” say solar PV has an eroei of about 10?

Renewables invite less energy usage because they are more expensive. Hopefully, solar and wind will eventually be made out of less energy intense material by machine. When using hydrocarbons to make clean energy capacity, then eroei must be seriously factored in. When the fossils are almost completely replaced, then imbodied energy shouldn’t (?) really matter as long as there are no other serious environmental consequences (such as massive biofuels farming).

Imagine a future scenario between PV and fusion. The PV becomes twice as efficient as today and lasts 20 years or so, but still gets an eroei of say, 15.  Fusion is also developed, but can only be made to work on a microchip, and only gives an eroei of just 2 (in a nano second). Which one wins? The one that is instant!

Clifford Goudey's picture
Clifford Goudey on Feb 25, 2014

Robert, I suppose you’d be right if it was fair to ignore the energy associated with the mining and processing of that fuel as well as the energy associated with plant construction and the spent fuel storage facilities.  Of course a much of that embodied energy is reflected in the high costs of nuclear plant construction. 

Get Published - Build a Following

The Energy Central Power Industry Network is based on one core idea - power industry professionals helping each other and advancing the industry by sharing and learning from each other.

If you have an experience or insight to share or have learned something from a conference or seminar, your peers and colleagues on Energy Central want to hear about it. It's also easy to share a link to an article you've liked or an industry resource that you think would be helpful.

                 Learn more about posting on Energy Central »