Can Future Energy Needs be Met Sustainably? 

Chris Llewellyn SmithDirector of Energy Research, Oxford University

President SESAME Councilwww.energy.ox.ac.uk

Introduction• Energy use is projected to increase by some 35% by 2035             

– driven by development in India, Africa, China…

• Some 80% is provided by burning fossil fuels (expected that we will be burning ~ 25% more in 2035), which is not sustainable

Outline• Sustainability

• Ways to decarbonise

• Could we manage with less?

• Low carbon systems and sources

• Three elephants in the room 

• Can future needs be met sustainably?

Use of Fossil Fuels is UnsustainableLong-term:Unless we give up using them, they will eventually become increasingly scarce and expensive, but at current levels of use not for - many hundreds of years for

coal- at least 50 years for oil and

gas

Fossil Fuel Era- a brief period in the world’s history

Short/medium term: producing• Climate change• Dangerous pollution

Moving away from fossil fuels will also improve security of supply in countries without large reserves + re-balance global politics

Next two slides

BP Projections of Energy Related CO2 Emissions

IEA 450 is a scenario in which atmospheric CO2 stay below 450 ppm – the level at which it is thought there is a 50% chance of keeping the temperature rise below 2 CCO2 accumulates in the atmosphere – it’s the area under the curve that matters

It seems that the 450 ppm target will be missed (true even with a $100/tonne CO2 price + big additional efficiency gains)

Air Pollution • Globally (WHO 2014) 7 million 

premature (typically 10 years loss of useful life) deaths p.a. (out of 56 million p.a. total)

• US (2013 MIT study) 210 k p.a. from burning fossil fuels (out of 2.5 million total) of which 200 k from particulates including : 58 k road transport, 54 k power generation, 43 k industry – main culprit is coal

Numbers very uncertain but undoubtedly a single large coal power station is far more lethal than Chernobyl

Ways to Decarbonise• Reduce demand – better planning• Use energy more efficiently 

• Carbon Capture and Storage  – Cost? Inflexibility? No help with pollution

• Move to low carbon systems and sources

Will help but won’t solve the problemEnergy/GNP is decreasing – but not fast enough:Technically can do much better (key: strong regulation) 

Barriers to Energy Efficiency Technically, large improvements (40% or more?) look possible, but they are not happening

‐ Appraisal optimism and neglect of transaction costs‐ Direct and (more importantly) indirect rebound effect‐ No incentives for the affluent to make small savings, which 

collectively can be large, e.g. electric lighting, uses 20% of electricity

‐ Poor lack capital

Need regulation ‐ cars, buildings, light‐bulbs, appliances…

Effect of RegulationUS Passenger Vehicles

End of mandatory Corporate Average Fuel Economy standards

Current standardsfor new vehicles:35.5 mpg by 2016

54.5 mpg by 2025

Efficiency ‐Ton‐miles per US gallon

Economy ‐Miles  per US gallon

Conversion factor: 30 miles/US gallon is equivalent to 7.85 litres/100 km

Decarbonising the Energy Systems• Primary energy globally →

Heat ≈ Electricity ≈ 40%, Transport ≈ 20%

• Electricity – know how to decarbonise in principle

• Transport – also: will probably mostly become electric ‐modest increase in demand (all UK cars → electric would increase average electricity demand by 15%)

• Heat – very hard to decarbonise ‐ Use less (better insulation)‐ Combined Heat & Power (scope limited)‐ Heat pumps → huge increase in (peak) electricity demand in coun es that use a lot of energy for space and water heating [UK: average demand + 40%, peak demand (already 50% above average) x 2]

Very difficult to decarbonise industrial heat.  In UK ‐ 200 TWh p.a. from fossil fuels, 100 TWh p.a. from electricity (c/f total electricity consumption 350 TWh)

Low Carbon SourcesAre they sufficiently abundant to replace fossil fuels? Yes 

Can their costs compete with fossil fuels (would solve the problem)?Apparently: generation cost for: wind ‐ now the lowest

solar ‐ decreasing rapidly

But• For ‘non‐dispatchable’ sources cost ≠ value, e.g. if wind blows mainly at night it may be cheap but relatively valueless; if enough is installed to meet most of average demand, there will be a large surplus at times of high availability of wind and sun/low demand

• Cost of integrating wind and solar will increase when their contributions increase: need back‐up for when wind does not blow & sun does not shine, but it won’t be used much → very expensive

Need to continue driving down costs and learn how to integrate large‐scale wind & solar + meanwhile replace coal with gas when possible and improve efficiency of the use of fossil fuels

Area → world’s current (thermal equivalent) primary energy*, in good conditions, from

Biomass230% of contiguous USA

Wind30% of contiguous USA

Solar4% of contiguous USA

*Average power:18.5 TWh or 7 TWe

Contribution 2014 Potential/Area (A) to produce all energy 2014

Issues

Biomass+ waste

9.6% A = 230% of contiguous USA in OK conditions

Land & water use +carbon footprint

Hydro 6.2% Could triple Environmental

Nuclear 4.0% Unlimited?? Cost; public perception

Wind 1.1% A = 30% of contiguous USA in OK conditions

Integration costNIMBYism

Solar 0.3% A = 4% of contiguous USAin OK conditions

CostIntegration cost

Geo 0.1% Can be important locally (e.g. marine in UK) but not globally Marine 0.001%

Low Carbon Resources

Dispatchable Sources $/MWhrCoal: Conventional 95

Advanced [with CCS] 116 [144]Gas: Conventional  C Cycle 75 With US gas price . With European 

cost 73 (e.g.) → over 100Advanced C Cycle [with CCS] 73 [100]Advanced nuclear 95 UK Government willing to pay £90

Biomass 100Non‐dispatchable sources – should not compare with LCOE for dispatchable sourcesWind 74Wind off‐shore 197Solar PV [thermal] 125 [240] Best 100, worst 187Hydro 83

Levelized Cost Of  Electricity generation*: US EIA estimates (June 2015) for sources coming on line in USA 2020 (geographical averages)

*Depends on assumed: life‐time, cost of capital, cost of fuels, capacity factors ‐ wind: 31%‐40%  (Germany 2012: 18%), solar: 22%‐32% (Germany 2012: 11%) ,…

Bioenergy – energy and food?Oxford colleagues investigating: Crassulacean Acid Metabolism (CAM) plants (cacti, euphorbia*,…) ‐ need only ~ 20% of water (staggering numbers: 4% to 15% of the 2.5 bn ha of potentially available semi‐arid land would → 5 PWeh  ~  20 % of current electricity)

* trials in Kenya going well

+Improved Anaerobic DigestionBy learning from cows, hope to reduce biogas plant sizes by a factor of 20 → big cost savings1st experiments under way (very early days) 

www.scientificamerican.com/article/cactus‐as‐biofuel‐could‐help‐with‐food‐versus‐fuel‐fight/and gizmodo.com/this‐humble‐cactus‐could‐help‐power‐our‐drought‐stricke‐1715966241

Biogas plant waste can supply charcoal, fish and fertiliserBiogas (easy to store) + solar perfect for mini‐grids

Nuclear Fission• Problem is cost + financing• Big reactors: capital cost/kWe expected to decrease with size,

but data suggest an increase (power 0 to 0.10) + not much evidence for learning (new labour force; design modifications; new regulations;…)

→ Good case that Small Modular Reactors cut cost significantly

- Design simplification - Multiple units one site - Production learning - Standardisation- Short build schedule - Finance savings

Prefabricated reactor en‐route from factory (too big for UK roads)

Is it true?

Need to build some and seeNuscale claim $90/MWh (in USA)Experts’ estimates all over the place:

Cost Reductions for PV?

Need to drive down cost of balance of plant (land, labour, inverters,…)depends on location and system: happeningGerman domestic costs (Q3 2014)/(Q1 2006) = 0.26 total, 0.19 modules, 0.43 balance of plant

Revolution? Perovskites, pioneered in Oxford

Module costs are falling rapidlyt

- now less than half the total BP 2016

Solar Cells Incorporating PerovskitesPioneered by Henry Snaith ‐ one of ‘10 people who matter’ according to Nature, Dec. 2013 

Meteoric rise in efficiency:

Initial focus on low cost of perovskites as a cheap alternative to Si, but Si got much cheaper.  Now focusing on tandem cells “By combining these perovskite cells with a 19%‐efficient silicon cell, we demonstrated the feasibility of achieving >25%‐efficient four‐terminal tandem cells” (then → all perovskite thin film technology) 1970 1980 1990 2000 2010 2020

0

10

20

30

Rec

ord

effic

ienc

y (%

)

Year

GaAs c-Si CIGS CdTe Perovskite

Passed 1,000 hour accelerated aging test, but would like 3‐5000 hours

Conclusions on low carbon sources• Good news

‐ Bio may have a bigger potential than thought

‐ Scope to further expand hydro

‐ Fission: must drive down costs of large nuclear reactors; jury out on whether Small Modular Reactors can lower cost

‐ Generating costs for solar and wind falling rapidly

BUT long way to go (fossil fuels ~ 80%) +‐ Genera ng cost ≠ value for non‐dispatchable/intermittent sources

‐ Cost of integrating solar & wind* will rise as their contributions increase  *needs strengthened/smarter grid; new ways to store energy; aggressive demand management

‐ Electricity markets will have to be re‐designed to deliver optimal solution

Three Elephants in the Room*

• Air pollution

• Decarbonising heat – space & water heating (mammoth in the UK room) + industrial heat

• Integrating renewables

* "Elephant in the room" ‐ an English metaphorical idiom for an obvious truth that is going unaddressed

Solar Thermal

• Oxford solar cooker*‐ spin‐off from solar neutrino experiment (2015 Nobel Prize)→ day me cooking + sterilizing water Field trails of in Tanzania – potential to radically reduce deaths from air pollution + save money 

• Concentrated solar for industrial heat in sunny countries?

* http://www.energy.ox.ac.uk/wordpress/wp‐content/uploads/2013/09/CaseStudies_5.pdf

Green AmmoniaAmmonia production (for fertiliser: $100 bn p.a. industry) ~ 1.4% of world’s fossil fuel consumption and CO2 emissions• It could be decarbonisedCurrently: natural gas + energy → hydrogen + nitrogen → (Haber‐Bosch) NH3

2030?: nitrogen + water → (renewable energy, electrolysis) NH3

‐ could be cost competitive ~ technical advances + carbon and gas price 

• and used as an energy vector 

NH3 fuelled bus in Belgium during WW2

‐ Seasonal energy storage? Annual ammonia production, if converted to electricity in a fuel cell would → 600 TWh = world electricity consumption for 10 days.  Conceivable that this could be a competitive way to store renewable energy that would otherwise be curtailed.

‐ Transport?  

Integrating RenewablesLCOE is [is becoming] competitive for wind [solar], but• Cost ≠ value, e.g. wind blows mainly at night in some places when prices are low 

• Mid‐day dip produced by solar → back‐up must be very flexible• Zero marginal cost of wind and solar destroying business model for conventional generation

•

• Technical solutions – demand side response, smarter/larger grid (better interconnectors), storage… don’t look adequate and  must be complemented by

• Market reform needed ‐ incentives, business models to deliver what’s wanted

Illustrate with California and Germany ‐ bad for wind and solar:

Solar PV ‐ better correlated with demand:

Wind ‐ blows mainly in late afternoon and at night, when prices low

California

But – off‐grid systems need own storage (battery)

and customer‐sited solar generation producing a mid‐day dip:

Renewables need flexible back‐up

Week in May:    2012 2020

Germany: Fluctuations in Wind and Solar

DemandGW

Destroying business model for conventional generation 

JPMorgan Study (demand ‐ 25%, wind x 3, solar x 2) – treat numbers (many inconsistent) with care:

• Need very flexible back‐up ; ‘deficit’= 107 TWh < surplus 47 TWh• Very expensive ($200/MWh; $300/tCO2 saved)• Introducing 35% nuclear (@$95/MWh) → $136/MWh; $84/tCO2 saved

Looking further ahead in Germany

Storage 1

To back up solar need storage ~ hours

For wind and for heat problem want days to months

Currently available: 140 GW  pumped‐hydro, capacity =  5% of average world loadbut would only last a few hours  

Dinorwig (N Wales) Pumped Hydro 1.7 GW/8 GWh ~ 4% of average UK load for 5 hours. 16 seconds to reach full power

*could meet daily need, but costs need to fall:

1 Tesla Powerwall, (re)cycling 6.4 kWh/day for 10 years (?) would provide 23 MWh, at a capital cost of $3,000 → $130/MWh (if replacing coal, saves CO2 @ $140/tonne). 1 million Powerwalls would (re)cycle 2.3 TWh/year.

If all cars in UK become electric overnight would increase average electricity demand by 15%; using every car equally for storage and driving would store 50 TWh/year.  Cost = bribing users to hand control to the grid + cost of connection & control 

*Part of issue of designing markets, incentives, business models that deliver what is wanted to deal with renewables

Storage 2Need to understand: role; scalability; central vs. local; energy vs. power; cost; efficiency;… 

How to value* ‐ understand/assign full benefits (less back‐up, use all wind & solar, possibility of arbitrage,…), how to charge beneficiaries (part of generation or transmission?), how to incentivise provision,….

Technology – more pumped hydro, batteries*, power to gas (hydrogen, ammonia,…), compressed air, hot water tanks, hot/cold pebbles, synthesising hydro carbons,...

Modelled versus cumulative production:

Battery CostsBjörn Nykvist and Måns Nilsson, Nature Climate Change 2015

Versus time:

Issues for Electricity MarketsUnderlying problems: i) supply and demand must match instantaneously, ii) ensuring security of supply considered a public good, iii) different players responsible for different parts of system, iv) time of rapid change

Minimising cost while ensuring security of supply and meeting climate targets involves decisions on: • Investment (in generation & transmission): requires model of future 

(variation in) demand (taking account of increase in electric cars etc.) + changes in supply (more local, with homes exporting as well as importing)

• Design of tariffs & regulations that will spread the load• Operation: requires anticipating demand from minutes to days

Hard in a fully integrated nationalised model: much harder in a competitive market* ‐ players must anticipate each others’ behaviour  * no competition in transmission = a monopoly

Market competition claimed to lower prices (it seemed to do so in the UK initially: generators sweated assets, but did not invest properly for the future)

Electricity Market Designhas to address many difficult questions (different ‐ often incoherent – fixes introduced in different countries):

• How to ensure investments → genera on mix that meets climate targets?• How to ensure sufficient capacity to meet maximum demand (last kWh = 

by far the most expensive)?• How to ensure sufficient conventional generation ‐ being  displaced by 

zero marginal cost renewables when the wind blows/sun shines,  undermining business case (German utilities facing bankruptcy)?

• How to ensure homes which generate their own power (and export surplus) pay a fair share for the grid (at the moment they are subsidised by those who buy all their power)?

• How to asses who benefits, and how they should pay, for strengthening the grid, storage? 

• Who should provide storage (generators, distributors, others), and how should provision be incentivized? 

• How to implement demand management, and integrate/balance with storage and stronger grid? 

• ….

Towards ConclusionsCan the world's (growing) energy needs be met:• With fossil fuels? Yes for at least 50 years

• Without fossil fuels? With existing technology - incredibly difficult: impossible at a price society would be prepared to pay

To meet future energy needs sustainablytechnological advances are needed - soon since making large scale changes in energy infrastructure will take decades

• Now consider - Necessary technical actions- Necessary public policy actions

• Until at least mid-century fossil fuels will continue to play a major role, so while developing CCS and alternatives, it is very important to

•

- Replace coal with gas as far as possible (pollution, CO2)- Improve efficiency of use of fossil fuels

• To have a serious chance of decarbonising will need many or all of the following: − Large scale affordable Carbon Capture and Storage− Radical reduction in use of oil in transport (→ more electricity)− Lower costs for solar and wind (happening) and learn how to handle

them on a large scale− Large scale biomass → electricity + biofuels (for flight) − Lower costs for nuclear− Big improvements in efficiency− Things we have not thought of− Devise economic and policy tools to make this happen

Necessary Actions - Technical

Necessary Actions - Policy• Better planning to reduce demand – especially in growing

cites/developing countries

• Stronger regulations, vehicles, performance of appliances, buildings…

• Phase out $550 billion/year of subsidies for consumption of fossil fuels (only 8% benefits world’s 20% poorest people)

• Carbon tax (provides more certainty than cap and trade) + in the absence of global agreement: Border Carbon Adjustments (or regulate power plants)

• Increase the $120 billion/year subsidies to launch* new not yet cost-effective energy sources and efficiency measures *then phase out

• Adopt policies (what?) that stimulate innovation, and increase long-term publicly funded R&D

• Reform electricity markets

Concluding Remarks• To allow everyone on the planet to lead decent lives,

much more energy will be needed• We can meet the need with fossil fuels for (at least)

50 years - but we should be decarbonising• No real progress with decarbonisation• Decarbonisation is possible, but will require

developing and implementing new technologies and new policies

• Large scale changes in energy infrastructure take decades – so action is needed now

Malthusian “solution” if we fail?