Energy Policy 39 (2011) 2748–2753

Contents lists available at ScienceDirect

Energy Policy

0301-42

doi:10.1

n Corr

E-m

damon.

journal homepage: www.elsevier.com/locate/enpol

Viewpoint

Is there an optimum level for renewable energy?

Patrick Moriarty a,n, Damon Honnery b

a Department of Design, Monash University-Caulfield Campus, P.O. Box 197, Caulfield East, Victoria 3145, Australiab Department of Mechanical and Aerospace Engineering, Monash University-Clayton Campus, P.O Box 31, Victoria 3800, Australia

a r t i c l e i n f o

Article history:

Received 7 February 2011

Accepted 13 February 2011Available online 3 March 2011

Keywords:

Climate change

Ecological sustainability

Energy ratio

15/$ - see front matter & 2011 Elsevier Ltd. A

016/j.enpol.2011.02.044

esponding author. Tel.: +613 9903 2584; fax

ail addresses: patrick.moriarty@monash.edu (

honnery@monash.edu (D. Honnery).

a b s t r a c t

Because continued heavy use of fossil fuel will lead to both global climate change and resource

depletion of easily accessible fuels, many researchers advocate a rapid transition to renewable energy

(RE) sources. In this paper we examine whether RE can provide anywhere near the levels of primary

energy forecast by various official organisations in a business-as-usual world. We find that the energy

costs of energy will rise in a non-linear manner as total annual primary RE output increases. In addition,

increasing levels of RE will lead to increasing levels of ecosystem maintenance energy costs per unit of

primary energy output. The result is that there is an optimum level of primary energy output, in the

sense that the sustainable level of energy available to the economy is maximised at that level. We

further argue that this optimum occurs at levels well below the energy consumption forecasts for a few

decades hence.

& 2011 Elsevier Ltd. All rights reserved.

1. Introduction: the idea of a renewable energy optimum

In a ‘business-as-usual’ world, such as that assumed in thevarious scenarios for organisations like the US Energy InformationAdministration (EIA), the International Energy Agency (IEA),the Organisation of the Petroleum Exporting Countries (OPEC),the World Energy Council (WEC) and the International Institutefor Applied Systems Analysis (IIASA), primary energy levels areforecast to be roughly in the range 600–850 EJ in 2030(EJ¼exajoule¼1018 J), and 800–1170 EJ in 2050. The IIASA evenextends its scenarios out to 2100, with a primary energy rangefrom about 1000–1740 EJ (Energy Information Administration(EIA), 2010; International Energy Agency (IEA), 2010;Organisation of the Petroleum Exporting Countries (OPEC),2010; Riahi et al. 2007; World Energy Council (WEC), 2007). Forcomparison, the 2008 global primary energy consumption was514 EJ, including about 66 EJ (12.9%) from renewable energy (RE)sources (IEA, 2010).

Continued heavy use of fossil fuel will lead to both globalclimate change and resource depletion of easily accessible fuels.Even nuclear energy is subject to resource depletion, and it alsofaces a variety of other problems, which limit political acceptance.None of the forecasts just discussed see it playing more than aminor role over the coming decades (Moriarty and Honnery,2010a). Accordingly, many researchers advocate a rapid transition

ll rights reserved.

: +613 9903 2076.

P. Moriarty),

to renewable energy (RE) sources. Their analyses imply that REwill have little difficulty in supplying the primary energy levelsforecast for the end of this century, since their estimates for REtechnical potential are as high as 7500 EJ or more (e.g. de Vrieset al., 2007; Hoogwijk and Graus, 2008; Johansson et al., 2004;Resch et al., 2008).

There are precedents for the idea of an optimum scale foroutput. Daly et al. (2007) criticised conventional economists fornot recognising that there is an ‘optimal scale of the macroec-onomy relative to the biosphere’. They asked whether the scale ofthe global economy is now so large that vital biosphere functionsare being compromised. Those working in the ‘Ecological Foot-print’ (EF) paradigm also imply the existence of an optimum scalefor human biosphere intervention. Kitzes et al. (2008) argued thatin 2002 humanity had an EF of more than 1.2 planet Earths,compared with the 1.0 that is available. In other words we are inovershoot, which will lead to ‘the degradation and liquidation ofecological capital.’ Kleidon (2006) produced model results whichshowed that attempts to indefinitely increase humanity’s share ofthe planetary Net Primary Production are ultimately self-defeat-ing in that, past a certain fraction, the absolute level of biomassavailable for human appropriation falls.

In all three studies, the idea is that there is an optimum level ofactivity beyond which the ecosystem services freely provided bythe natural world are compromised; the optimum number ofEarths that can be sustainably used is 1.0; the optimum level ofgroundwater use might be the level, which can be annuallysustained by recharge. We argue in this paper that diversion ofEarth’s energy flows to satisfy humanity’s energy needs is alsosubject to an upper limit—that there is an optimum level of RE use.

Fig. 1. Schematic graph of annual total primary RE energy vs. marginal

energy ratio.

P. Moriarty, D. Honnery / Energy Policy 39 (2011) 2748–2753 2749

Kleidon’s paper is an attempt to demonstrate this point forbiomass use (including all human uses of biomass, not justenergy). We argue that it applies to RE generally. It is not enoughto show that an optimum value exists for RE; it also has to beshown to have a high probability of occurring in the range ofhuman interest, which, from our earlier discussion, might beanywhere from low levels to as high as 2000 EJ.

We use a global approach, partly for simplicity, but alsobecause for some RE sources, analyses at the regional or nationallevel will give misleading conclusions. For biomass this is becauseof international trade in agricultural and forestry products. Forhydro, 261 river basins straddle international boundaries (Wolfet al., 1999), so that a national optimum could again bemisleading.

In summary, we argue that there is an optimum level ofprimary RE output, in the sense that the sustainable level ofenergy available to the economy is maximised at that level. Wefirst show that the energy costs of energy will rise in a non-linearmanner as total annual primary RE output increases. We thenprovide evidence that increasing levels of RE will also lead toincreasing levels of ecosystem maintenance energy costs per unitof primary energy output. When both these non-linear energycosts are subtracted from primary RE, a maximum value of netsustainable RE occurs. We argue that this optimum RE levelprobably occurs at a level below the energy forcasts for a fewdecades hence. Finally, we demonstrate that even this optimumwill shrink if climate change continues unabated.

2. Energy costs for energy rise with increased RE output

A vital test for the viability of any proposed energy conversiondevice, whether RE, fossil or nuclear energy, is that the energyoutput over its useful life should be greater than the combinedenergy inputs needed to manufacture, erect, maintain and oper-ate, and finally decommission the equipment. The energy ratio(ER), as defined here, is the ratio of gross energy output from thedevice over its operating life divided by the total lifetime energyinputs, with both input and output energy given in primaryenergy terms. Clearly, if ERo1.0, the energy project is not viable,regardless of the monetary costs of the energy output. In fact, ER

will need to be much greater than 1.0 if the project is to be viable.One reason for this is the enormous variability in calculatedvalues in the literature for ER for a given RE source (Moriarty andHonnery, 2007a, 2010a; Honnery and Moriarty, 2011). Net energyis gross output minus input energy. For values of ER above 10 orso, net energy values are little different from those for grossenergy, but for low values of ER the distinction is crucial.

2.1. Resource quality declines with increasing RE output

For a given RE source, those sites giving the best energy returnon input energy are usually built first in any country, as these alsotend to yield the cheapest energy. For the world as a whole, this isalso approximately true, although in the interests of energysecurity, some countries tap local lower-quality resources. Suchis presently the case for the minor amounts of solar energyproduced in Germany. However, Germany is also in the forefrontof efforts to develop the solar energy resources of North Africa,and transmit much of the output to Europe.

Resource quality declines can take the form of lower averagewind speeds for wind energy, lower average annual insolationlevels (and low winter insolation) for solar energy, lower tem-perature steam for geothermal plants, and increasing depths forEnhanced Geothermal Systems (EGS). For bioenergy plantations,rising output might mean a shift to lower-quality soils, or the

need for irrigation to augment rainfall. For hydropower, ‘there arefewer and fewer good dam sites available’ (Prichard, 2002). As wehave shown elsewhere for wind energy (Moriarty and Honnery,2010a), the global energy ratio will fall steadily as lower averagespeed winds are progressively tapped. Similar curves could inprinciple be constructed for other RE sources, if the relevant datawere available. Fig. 1 shows schematically, for RE in general, thistrend for primary energy vs. marginal energy ratio (defined as theminimum energy ratio for the given annual primary energy).

There is another way in which resource quality declines withoutput. Insolation in Germany may be low compared with NorthAfrica, but the energy is produced close to where it is used. Energyfrom high-insolation North African solar farms, if transmitted tocentral Europe as in the Desertec proposal (Pearce, 2009), facestransmission distances of up to 5000 km. The capital costs ofbuilding such a grid, much of it sub-sea, will be enormous, andtotal energy losses will also be significant, even for the proposedhigh-voltage DC systems. The solar farms would need to belocated away from the coast to avoid cloudiness (and valuableurban and agricultural land), so supplying fresh water for cleaningmirrors or PV cells could also incur significant input energy costs.The energy costs for water would be even higher if solar thermalenergy conversion (STEC) plants are water-cooled (Webber,2007), which is more efficient than air-cooling. It will also provenecessary to construct roads and other infrastructure for theprojects.

2.2. Need for energy conversion and storage rises with RE output

As we have argued in earlier publications (Moriarty and Honnery,2007b, 2010a), if large amounts of RE energy are needed in the future,the main sources will have to be intermittent—wind, wave, and solar.RE sources that are available on a continuous basis (biomass, hydro,geothermal electricity), can only be ever expected to generate a fewtens of EJ globally. (Nevertheless, geothermal energy could producelarge amounts of low-grade heat for direct use.) As renewable energyincreases well beyond its present low level, the proportion suppliedfrom intermittent sources will inevitably have to rise as well.

As output from intermittent sources begins to dominate totalelectricity demand, conversion of intermittent electricity to analternative energy carrier such as hydrogen, and subsequentstorage will be increasingly needed, beginning in those grids withfew non-fossil alternatives for base-load power. Eventually, inter-mittent primary electricity will also have to provide for non-electric energy uses, again necessitating conversion to some other

Fig. 2. Schematic graph of primary, net and net green energy vs. primary RE

output.

P. Moriarty, D. Honnery / Energy Policy 39 (2011) 2748–27532750

energy form. But conversion and storage could reduce by almost50% the energy delivered to end uses from the gross primaryoutput of these RE sources (Honnery and Moriarty, 2009). Thisneed for energy conversion will occur at the same time as theenergy ratio is falling because of declining quality of the REresource, as discussed above. A further energy cost will be theneed for fresh water for electrolysis, which will have high energycosts in the arid areas favoured for solar energy, both for lowland-cost and high-insolation reasons. Webber (2007) estimatedthat over 100 l of water would be directly consumed for every kgof hydrogen produced. (Total water withdrawals per kg ofhydrogen for cooling (but not evaporated), Weber estimated atover 40 times that value.)

The overall result of declining energy return on energyinvested is that net energy does not rise at the same rate as grossprimary energy. This is shown schematically by the net energycurve for all RE primary energy in Fig. 2. The curves for eachindividual RE source will be of the same general shape, but themaximum net energy for each will occur at very different valuesof primary energy. It might be thought that the maximum netenergy is the optimum value for RE, but this ignores the energycosts for ecosystem maintenance, which are incurred for RE aswell as fossil fuels, as discussed in the next section.

3. Energy costs for ecosystem maintenance rise withincreased RE output

In addition to energy costs for energy production, energy costsare also incurred for ecosystem maintenance (needed to ensurea continued flow of ecosystem services), although these two typesof energy costs are not always easy to separate out. For fossilfuels, pollution control was not initially regarded as an inputenergy cost. However, starting with the OECD countries, particu-late emissions control, then sulphur and nitrogen oxides emis-sions control were regarded as essential, and the incurred energycosts are now regarded as part of the energy costs for energy.Eventually, emissions of CO2 and other GHGs might come to beseen the same way, with the energy costs of carbon sequestrationand storage or air capture, included as input energy costs.

RE sources also incur ecosystem maintenance energy costs.The first category is similar to that just considered for fossil fuels:the global warming impacts of RE (apart from those derived fromany fossil fuel energy inputs to RE production). Small amounts of

bioenergy could be derived from various biomass waste streams,such as municipal wastes and some farm and forestry residues.But for higher levels of bioenergy, biomass plantations will beneeded. From a global viewpoint, any major biomass expansionwill, we argue, lead, directly or indirectly to conversion ofnaturally vegetated tropical lands to either food or bioenergyproduction.

Even if the bioenergy expansion occurs solely in presentlyfood-exporting countries like the US or Australia, any decrease oreven stagnation in their exports would necessitate agriculturalexpansion in other countries, given the expanding food and fibreneeds of the global population. The expansion of crop area inrecent decades has been largely in tropical countries (West et al.,2010). But as these researchers have also shown, such expansionhas been at the expense of soil carbon losses to the atmosphere.Clearing land in the tropics entails the loss of about four times asmuch soil carbon compared with temperate regions, for a givenannual crop yield.

Further, growth in bioenergy plantations will lead to increaseduse of nitrogenous fertilisers because of the need to move to lessfertile soils. But as Crutzen et al. (2008) have shown, this actionwill lead to rising emissions of N2O, a potent and long-lived GHG.Changes in albedo can also result from changes in forest cover.Although reductions in forest cover in tropical regions would raisethe local albedo, any increase in high latitude northern hemi-sphere regions would decrease local albedo, as vegetation absorbsmore radiant energy than snow-covered ground (Betts, 2000).

Just as expansion of agricultural land has been mainly in thetropics, so will the future expansion of hydropower, as mostremaining hydro potential is there (WEC, 2010). But as withagricultural land expansion, the global climate change effects oftropical hydro plants are much more severe than similarly sizedplants in temperate or Arctic regions. Hydro dams can emit bothCO2 and CH4 from rotting vegetation, depending upon whetherthe decay is aerobic or anaerobic. Over the initial years ofoperation, GHG emissions may rival those of a gas-fired thermalplant of similar electricity output (Moriarty and Honnery, 2009a).All these climate change effects could be offset by carbonsequestration projects such as air capture or net reforestation,but at substantial energy costs.

Other ecosystem costs can also be expected to rise non-linearly as RE output rises. At low levels of output, for examplefor wind energy, environmentally sensitive areas can be avoided.But if wind and solar energy are to replace fossil fuels as thedominant energy sources, both will need to expand by severalorders of magnitude (Moriarty and Honnery, 2010b), with suchexpansion inevitably involving (for wind energy), much higherrates of bird and bat fatalities per MW of installed capacity.Such adverse effects can be avoided to some extent by shuttingdown turbines at the most sensitive times. The energy outputso foregone can be considered an ecosystem maintenanceenergy cost.

The adverse effects of hydropower expansion on freshwaterbiodiversity are now acknowledged (McCartney et al., 2001); similarbiodiversity declines are anticipated to accompany bioenergy expan-sion. Any large increase in ocean energy, whether from off-shorewind, wave or tidal current devices, or ocean thermal energy systemswill likely have adverse environmental consequences, many of themnot adequately explored (Boehlert and Gill, 2010). Unlike hydropower(and large dam construction in general), and the large-scale use ofbiomass for food, fibre, forestry and now energy, we have very littleoperating experience with the new RE sources on which to base arealistic environmental assessment. The energy costs for ecosystemmaintenance are shown schematically by the difference between thelower two curves in Fig. 2. The bottom curve we call the net greenenergy curve.

P. Moriarty, D. Honnery / Energy Policy 39 (2011) 2748–2753 2751

4. Ongoing climate change will reduce the optimumlevel for RE

Increasingly, official bodies are calling for global temperaturerise to be limited to 2 1C above pre-industrial to avoid ‘dangerousanthropogenic change’. However, with past and committed tem-perature rises already totalling about 1.3 1C, and with the con-tinued rise in global fossil fuel use and net deforestation, keepingbelow this limit will be very difficult. Most likely, temperatureswill continue to rise, which will have significant effects on REavailability. The most important effects will be on solar, wind,hydro and biomass energy. Only non-solar energy flows—tidaland geothermal—will not be directly affected by on-going climatechange.

4.1. Solar energy

All concentrating solar systems that track the sun, such asSTEC and some PV systems, require direct sunlight and thus clearskies, which could restrict their deployment to arid regions. Anyincrease in cloudiness (expected for some regions) would lowertheir output. Further, if geoengineering measures such as placingaerosols in the lower stratosphere were used to increase theEarth’s albedo, the resultant increased scattering of insolationwould significantly reduce concentrating solar plant output, aswell as decreasing the effectiveness of some passive solar energytechniques (Murphy, 2009).

4.2. Wind energy

Although there are important local influences on wind speeds,the main mechanism generating winds is the temperature differ-ence that exists between the poles and the equator. All climatemodels predict that the polar regions will heat faster than thetropical regions, a prediction borne out by the rapid temperaturerises at the poles, especially the Arctic (Solomon et al., 2007). Theresult is a decline in the global temperature gradient, and aconsequent decline in wind speeds, particularly at mid to highlatitudes. According to Ren (2010), China could expect a reductionof about 14% in potential wind output by 2100 under the SRESemission scenario A1B. Similar reductions would presumablyoccur at similar latitudes in other countries. Higher GHG emis-sions would result in even greater wind energy reductions.

From a different perspective, Vautard et al. (2010) havedocumented an actual decline in wind speeds at mid to highnorthern latitudes over recent decades. They have argued that anadditional factor to consider is the growth in biomass in thisregion, which has increased surface roughness, and thus loweredwind speeds in the boundary layer.

4.3. Hydroelectricity

Hydro schemes are expected to have long service lives, butpast river flow rates and especially seasonal flow rates will nolonger be reliable indicators for the future if climate changecontinues. Globally, the situation is mixed, with higher annualriver flow rates expected for some regions, and decreases inothers. However, future regional precipitation patterns are uncer-tain, with different global climate models often predicting pre-cipitation changes of different sign for a given region. In contrast,increased regional temperatures are predicted with all suchmodels; higher temperatures will increase potential evaporationfrom both soils and reservoirs, lowering hydro output. Highertemperatures also mean a larger share of total precipitationfalling as rain rather than snow, as well as earlier melting of

snow, which can make river flows become progressively moretemporally skewed.

Extreme rainfall events are also predicted globally to be morefrequent (Solomon et al., 2007), leading to an increase in soilerosion. Intense rainfall causes a disproportionate amount of soilerosion; in the US, a 10% rise in rainfall intensity was found toproduce a 24% rise in erosion (Thacker, 2004). Increased erosionwill in turn increase reservoir sedimentation and reduce storagecapacity (Prichard, 2002).

A number of studies have modelled the effect of predictedclimate change on existing hydro schemes or on hydro in selectedregions. For Southern Europe, Lehner et al. (2005) have shownthat the decreased precipitation projected for that region willcause disproportionately larger decreases in annual hydro output.For the Pacific North-West in the US, Hamlet et al. (2010)projected that by 2040, hydro output would rise in winter butdecrease in summer, with an overall decrease. By 2080, annualdecreases in output would be appreciably higher.

Reduced electricity output will have significant impacts on theeconomic viability of hydro schemes. Harrison et al. (2006) havestressed that even for existing plants, planned operation willbecome non-optimal if flow conditions change. They modelledthe financial performance of a proposed central African project,and found that the scheme’s economic performance would bevery sensitive to any changes in catchment rainfall patterns andtemperature. Because of the impact of climatic conditions onhydropower, they concluded that any changes could adverselyaffect future hydro project financial viability.

4.4. Bioenergy

Bioenergy potential will decrease in future even if furtherclimate change is avoided, since there is a growing humanpopulation, together with the economic growth per capita usuallyassumed to occur, will raise demands for food, fibre, animalforage and forestry products. Bioenergy is a residualterm—ideally it should only be produced after food needs aresatisfied for all the Earth’s people, and then output should bedetermined in light of our needs for other biomass products likefibre and forestry products. Biomass is fortunate in that to someextent it is complementary to other biomass uses, and so hassome high ER value output, mainly in the form of some of theresidues from agriculture and forestry. However, with the higherintensity rainfall expected, more cover will need to be left on thesurface to reduce erosion (Thacker, 2004).

Biomass production overall, including bioenergy, is projectedto be heavily influenced by ongoing climate change. Heimann andReichstein (2008), after reviewing a number of studies that usedcoupled carbon-climate models, concluded that with continuedglobal temperature increases, the world’s ecosystems wouldeventually change from a sink for carbon, as they are at present,to source of carbon. This change would adversely affect the futurepotential for bioenergy. More specifically, declining availability ofgood soils and adequate water supplies will progressively limitthis residual biomass use.

5. Discussion

Net green energy will exhibit a peak value if either the energycosts for energy, or the ecosystem maintenance energy costs arean increasing share of gross primary energy (and the other is atmost a fixed share of primary energy). If, as we have argued inthis paper, both rise with rising primary energy output, the peaklevel of net green energy will be lower still.

P. Moriarty, D. Honnery / Energy Policy 39 (2011) 2748–27532752

It is very difficult to put a precise figure on the primary REoptimum energy value, or the corresponding net green energypeak value. The chief reason for this is the uncertainty surround-ing estimates for both the technical potential for each RE source,and also for the energy costs of energy. (In the light of thediscussion in Section 2, technical potential must be interpreted tomean possible output at an acceptable level of ER.) Publishedestimates for hydropower potential are within a fairly narrowband, but those for bioenergy, wind, direct solar and the variousforms of ocean energy potential span one and sometimes twoorders of magnitude (Hoogwijk et al., 2003; Moriarty andHonnery, 2005, 2010a). In fact, if RE replaces fossil fuels as thedominant energy source, the concept of primary energy willbecome increasingly less relevant, a point developed furtherin Moriarty and Honnery (2010b).

Similarly, the ER values in the published literature display ahuge range for each RE source, although much of this is a result ofdifferent resource quality, for example, insolation levels for solar,or average wind speeds for wind energy. Different systemboundaries are also an important cause of variation (Moriartyand Honnery, 2007a; Murphy and Hall, 2010). The recent hybridenergy analysis of wind energy by Crawford (2009) suggests thatER values for wind turbines are far lower than usually assumed,and extending his argument, also for solar electricity plants. Zhaiand Williams (2010) also used hybrid energy analysis, and foundthat the energy payback time for PV cell installations was muchgreater than usually calculated.

The result is that it would be very difficult to construct ER vs.gross primary energy output for RE sources, even if takenindividually, and impossible to construct curves for ecosystemmaintenance energy costs vs. gross primary energy output. Thetrends shown in Fig. 2 can thus only be schematic. However, theresults of our earlier research (Moriarty and Honnery, 2005,2009b, 2009c) suggest that a lower-bound estimate for grossprimary RE potential output may be no more than a few hundredEJ. From very different perspectives, Brown and Ulgiati(2010), Makarieva et al. (2008) and Trainer (2010) have alsoargued that future RE availability is likely to be modest.

As we have argued, the net sustainable energy than can bedelivered to the economy will be far lower than this. As anillustration, Field et al. (2008) argued that we can only sustainablyharvest 27 EJ of biomass energy annually: beyond that valueglobal food supplies are threatened or climate change is exacer-bated. Even if the technical potential for primary energy wasseveral times higher, delivered energy could still be less thancurrent levels of delivered energy. Since ongoing climate changewill significantly lower the technical potential for some importantRE sources, it is clear that the existence of a low optimum valuefor primary RE energy must be taken seriously.

Given that RE is unlikely to be able to deliver anywhere nearthe quantity of net green energy a business-as-usual economywill need in the coming decades, prudence demands that weprogressively lower global demand for net green energy in thecoming decades. This can in principle be achieved by dramaticreductions in both the energy and carbon intensity of theeconomy, but, globally, recent decades have seen little improve-ment, and energy use and carbon emissions continue to rise. Aswe have discussed elsewhere (Moriarty and Honnery, 2010a,c),the remaining option in the limited time frame available is toreduce global primary energy consumption overall.

6. Conclusions

For all energy sources, less energy is delivered to the economythan the energy content of the fossil fuels mined, or the output

from RE conversion devices, because some energy must bediverted for manufacturing, erecting, maintaining and operatingthe equipment. For ecological sustainability, further energy mustbe diverted for ecosystem maintenance, to ensure continueddelivery of vital ecosystem services.

We have argued that both energy costs will form an ever-increasing share of RE primary energy as output rises. The energycosts of energy rise (ER falls) because the quality of the RE resourcedeclines with cumulative annual exploitation as illustrated in Fig. 1;average wind speeds progressively decline, for example. Given thatfor large RE output, most will need to come from intermittentsources (solar, wind and wave energy), additional energy costs willbe progressively incurred as electricity grids have to convert andstore this energy, perhaps as hydrogen. Solar and wind farms willbe sited at increasing distances (perhaps several thousand km) fromload centres, incurring further energy costs for infrastructurebuilding and transmission losses.

Ecosystem maintenance energy costs will rise disproportio-nately mainly because of the need to locate RE installations inprogressively more environmentally sensitive areas as RE outputincreases. Another reason is that GHG emissions per unit ofoutput from RE plant operation will also rise if new land forgrowing biomass is cleared in the tropics, with loss of soil carbon.New hydro capacity is also likely to be located mainly in thetropics, with increased emissions from their reservoirs. Somebioenergy can be obtained at low environmental cost fromwastes, but larger amounts will require dedicated bioenergyplantations. If fertilisation is needed, emissions of N2O fromnitrogenous fertiliser will rise.

But the longer we delay the transition to RE, the lower will bethe maximum value of net green energy that can be delivered,because ongoing climate change—along with any further popula-tion and economic growth—will tend to lower the primary energyvs. ER curve shown in Fig. 1. At present, RE primary energy use isonly about 13% of all primary energy—and most is used asfuelwood in industrialising countries. As use of new RE sourcessuch as wind and solar grows, their energy and environmentalcosts will become increasingly apparent, which suggests that weneed to be sparing even in our use of RE.

References

Betts, R.A., 2000. Offset of the potential carbon sink from boreal forestation bydecreases in surface albedo. Nature 408, 187–190.

Boehlert, G.W., Gill, A.G., 2010. Environmental and ecological effects of oceanrenewable energy development: a current synthesis. Oceanography 23 (2),68–81.

Brown, M.T., Ulgiati S., 2010. Putting metabolic patterns of society within a largerperspective. /http://158.109.67.81/aes2010/Proceeds/DIGITAL%20PROCEEDINGS_files/PAPERS/O_128_Brown_and_Ulgiati_submitted_25_9_2010.pdfS.

Crawford, R., 2009. Life cycle energy and greenhouse emissions analysis of windturbines and the effect of size on energy yield. Renewable & SustainableEnergy Reviews 13, 2653–2660.

Crutzen, P.J., Mosier, A.R., Smith, K.A., Winiwarter, W., 2008. N2O release fromagro-biofuel production negates global warming reduction by replacing fossilfuels. Atmospheric Chemistry and Physics 8, 389–395.

Daly, H.E., Czech, B., Trauger, D.L., Rees, W.E., Grover, M., Dobson, T., et al., 2007.Are we consuming too much—for what? Conservation Biology 21 (5),1359–1362.

de Vries, B.J.M., van Vuuren, D.P., Hoogwijk, M.M., 2007. Renewable energysources: their potential for the first-half of the 21st century at a global level:an integrated approach. Energy Policy 35, 2590–2610.

Energy Information Administration (EIA), 2010. International Energy Outlook2010. US Department of Energy, Washington, DC.

Field, C.B., Campbell, J.E., Lobell, D.B., 2008. Biomass energy: the scale of thepotential resource. Trends in Ecology and Evolution 23 (2), 65–72.

Hamlet, A.F., Lee, S.-Y., Mickelson, K.E.B., Elsner, M.M., 2010. Effects of projectedclimate change on energy supply and demand in the Pacific Northwest andWashington State. Climatic Change 102, 103–128.

Harrison, G.P., Whittington, H.W., Wallace, A.R., 2006. Sensitivity of hydropowerperformance to climate change. International Journal of Power and EnergySystems 26 (1), 42–48.

P. Moriarty, D. Honnery / Energy Policy 39 (2011) 2748–2753 2753

Heimann, M., Reichstein, M., 2008. Terrestrial ecosystem carbon dynamics andclimate feedbacks. Nature 451, 289–292.

Honnery, D., Moriarty, P., 2009. Estimating global hydrogen production from wind.International Journal of Hydrogen Energy 34, 727–736.

Honnery, D., Moriarty, P., 2011. Energy availability problems with the rapiddeployment of wind-hydrogen systems. International Journal of HydrogenEnergy 10.1016/j.ijhydene.2010.12.023.

Hoogwijk, M., Faaij, A., van den Broek, R., Berndes, G., Gielen, D., Turkenburg, W.C.,2003. Exploration of the ranges of the global potential of biomass for energy.Biomass and Bioenergy 25, 119–133.

Hoogwijk, M., Graus, W. 2008. Global Potential of Renewable Energy Sources:A Literature Assessment. ECOFYS PECSNL072975.

International Energy Agency (IEA), 2010. Key World Energy Statistics 2010.IEA/OECD, Paris.

Johansson, T.B., McCormick, K., Neij, L., Turkenburg, W., 2004. The potentials ofrenewable energy. In: Proceedings of the International Conference for Renew-able Energies, Bonn. /www.renewables2004.deS (accessed 13 December2010).

Kitzes, J., Wackernagel, M., Loh, J., Peller, A., Goldfinger, S., Cheng, D., et al., 2008.Shrink and share: humanity’s present and future Ecological Footprint. Philo-sophical Transactions of the Royal Society B 363, 467–475.

Kleidon, A., 2006. The climate sensitivity to human appropriation of vegetationproductivity and its thermodynamic charecterization. Global and PlanetaryChange 54, 109–127.

Lehner, B., Czisch, G., Vassolo, S., 2005. The impact of global change on thehydropower potential of Europe: a model-based analysis. Energy Policy 33,839–855.

Makarieva, A.M., Gorshkov, V.G., Li, B.-L., 2008. Energy budget of the biosphere andcivilization: rethinking environmental security of global renewable and non-renewable resources. Ecological Complexity 5, 281–288.

McCartney, M.P., Sullivan C., Acreman M.C., 2001. Ecosystem impacts of largedams. Background Paper no. 2 Prepared for IUCN/UNEP/WCD.

Moriarty, P., Honnery, D., 2005. Can renewable energy avert global climatechange? In: Proceedings of the 17th International Clean Air and EnvironmentConference, Hobart, 3–5 May.

Moriarty, P., Honnery, D., 2007a. Global bioenergy: problems and prospects.International Journal Global Energy 27 (2), 231–248.

Moriarty, P., Honnery, D., 2007b. Intermittent renewable energy: the only futuresource of hydrogen? International Journal of Hydrogen Energy 32, 1616–1624.

Moriarty, P., Honnery, D., 2009a. Renewable energy in a warming world. InternationalJournal of Energy, Environment & Economics 16 (4), 343–369.

Moriarty, P., Honnery, D., 2009b. What energy levels can the Earth sustain? EnergyPolicy 37, 2469–2474.

Moriarty, P., Honnery, D., 2009c. Hydrogen’s role in an uncertain energy future.International Journal of Hydrogen Energy 34, 31–39.

Moriarty, P., Honnery, D., 2010a. Rise and Fall of the Carbon Civilisation. Springer,London, UK.

Moriarty, P., Honnery, D., 2010b. A hydrogen standard for future energy account-ing? International Journal of Hydrogen Energy 35, 12374–12380.

Moriarty, P., Honnery, D., 2010c. A human needs approach to reducing atmo-spheric carbon. Energy Policy 38 (2), 695–700.

Murphy, D.J., Hall, C.A.S., 2010. Year in review—EROI or energy return on (energy)invested. Annals of the New York Academy of Sciences 1185, 102–118.

Murphy, D.M., 2009. Effect of stratospheric aerosols on direct sunlight andimplications for concentrating solar power. Environmental Science andTechnology 43 (8), 2784–2786.

Organization of the Petroleum Exporting Countries (OPEC), 2010. World OilOutlook 2010. OPEC, Vienna.

Pearce, F., 2009. Sunshine superpower. New Scientist 24 (October), 38–41.Pritchard, S., 2002. Overloaded. International Water Power & Dam Construction

54 (7), 18–22.Ren, D., 2010. Effects of global warming on wind energy availability. Journal of

Renewable and Sustainable Energy 2, 052301, 5 pp.Resch, G., Held, A., Faber, T., Panzer, C., Toro, F., Haas, R., 2008. Potentials and

prospects for renewable energies at global scale. Energy Policy 36, 4048–4056.Riahi, K., Grubler, A., Nakicenovic, N., 2007. Scenarios of long-term socio-economic

and environmental development under climate stabilization. TechnologicalForecasting & Social Change 74, 887–935.

Solomon, S., Qin, D., Manning, M., Alley, R.B., Bernsten, T., Bindoff, N.L., et al., 2007.Technical summary. In: Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis,M., Averyt, K.B. (Eds.), Climate Change 2007: The Physical Science Basis. CUP,Cambridge, UK, pp. 19–91.

Thacker, P.D., 2004. As the planet heats up, will the topsoil melt away? Journal ofSoil and Water Conservation 59 (1), 12A–15A.

Trainer, T., 2010. Can renewables etc. solve the greenhouse problem? The negativecase. Energy Policy 38, 4107–4114.

Vautard, R., Cattiaux, J., Yiou, P., Thepaut, J.-N., Ciais, P., 2010. Northern Hemi-sphere atmospheric stilling partly attributed to an increase in surface rough-ness. Nature Geoscience 3, 756–761.

Webber, M.E., 2007. The water intensity of the transitional hydrogen economy.Environmental Research Letters 2, 034007. doi:10.1088/1748-9326/2/3/034007 (7 pp).

West, P.C., Gibbs, H.K., Monfredad, C., Wagner, J., Barford, C.C., Carpenter, S.R.,et al., 2010. Trading carbon for food: global comparison of carbon stocks vs.crop yields on agricultural land. Proceedings of the National Academy ofSciences 107, 19645–19648.

Wolf, A.T., Natharius, J.A., Danielson, J.J., Ward, B.S., Jan, K., Pender, J.K., 1999.International river basins of the world. International Journal of WaterResources Development 15 (4), 387–427.

World Energy Council (WEC), 2007. Deciding the Future: energy Policy Scenariosto 2050. WEC, London.

World Energy Council (WEC), 2010. World Energy Survey 2010. WEC, London.Zhai, P., Williams, E.D., 2010. Dynamic hybrid life cycle assessment of energy and

carbon of multicrystalline silicon photovoltaic systems. Environmental Scienceand Technology 44, 7950–7955.