3. The Current Energy Transition
The current energy transition entails moving away from an energy system that relies on hydrocarbon molecules to one based around electrons. As can be seen from Figure 3, coal, oil and gas – though still dominant – are gradually making way in the energy consumption mix for nuclear power11 and renewables.12
Figure 3: Global primary energy consumption by fuel, 1965–2017
— Source: BP, 2018a.
Figure 3: Global primary energy consumption by fuel, 1965–2017
The environmental triggers: climate change and urban air quality
The initial trigger for the current shift was growing concern about climate change.13 This was first expressed in a policy context through the process begun in Rio de Janeiro in 1992, leading to the Kyoto Protocol in 1997 and, most recently, the COP21 meeting that concluded with the December 2015 Paris Agreement on climate change. The fundamental objective was to reduce the amount of carbon dioxide (CO2) being emitted into the atmosphere. Hence the need to reduce the use of hydrocarbon fuels.
However, in the past few years, as urbanization has accelerated, a second environmental concern has emerged as a trigger for system change: namely, deteriorating urban air quality and the increased presence of particulate pollution from the burning of coal and diesel. In many countries, this concern is beginning to take over from that surrounding CO2 emissions as the primary driver of the move away from coal and oil. The simple reason for this is that concerns over local air pollution, unlike those over climate change, do not require validation by a panel of scientists arguing that carbon emissions will have some consequent effects in the next few decades. Urban air problems are immediately obvious to anyone walking down a street, and thus are a pressing electoral issue in parts of the world where climate change may not be.
Just as the historical context shows, the activation of both these triggers – concerns over climate change and urban air quality respectively – has in turn attracted reinforcing factors, involving changes in technology and relative prices.
The reinforcing factors
Falling costs of renewables
The International Renewable Energy Agency (IRENA) claims that solar photovoltaic (PV) module prices have fallen by around 80 per cent since the end of 2009, while wind turbine prices have fallen by between 30 and 40 per cent (IRENA, 2019a).
Several caveats apply to these falls in costs. First, the figures represent contracted prices in very competitive markets. These prices may well have reflected overambitious attempts to win contracts and thus be lower than fundamentals would otherwise dictate. Second, the above figures represent the costs of generation only. They take no account of the additional costs of incorporating intermittent energy sources into the grid through the installation of back-up generating capacity and interconnectors to other generation sources.14 Nonetheless, it is clear that modern renewables are pushing coal and gas out of the generating mix in an increasing number of countries.
Estimates suggest that up to 1 billion people, 50 per cent of them in sub-Saharan Africa, do not have access to electricity. To use conventional thermal power plants to provide electricity to those currently without grid access would require huge investment in distribution networks.
Renewables also have two major advantages over conventional thermal power. First, as domestic sources of energy, they remove the most serious concerns over international security of supply that are associated with imported hydrocarbons. Second, they provide superior consumer access to electricity, a major issue in many countries. Estimates suggest that up to 1 billion people, 50 per cent of them in sub-Saharan Africa (UN, 2015), do not have access to electricity. To use conventional thermal power plants to provide electricity to those currently without grid access would require huge investment in distribution networks. However, modern renewables can be delivered locally at small scale without the need for major investment in the grid. The dynamics are similar in some respects to those seen with phone communications in Africa. It was widely assumed 20 or so years ago that, because of the size of the African continent, phone communication would be very slow to develop. In fact, the development of mobile phone technology enabled the continent to ‘leapfrog’ the traditional landline network stage. As a result, mobile phone communication in Africa is now extremely widespread.
Technological advances in electric vehicles and electricity storage
Another factor reinforcing the energy transition is the rise of electric vehicles (EVs). Figure 4 presents the changing estimates of future penetration of EVs into the total vehicle fleet or ‘car parc’,15 as calculated by two prominent forecasters. Each line represents a single forecast made in a given year, and is presented alongside estimates published by the same forecaster in different years to show how the forecast has changed over time.16
Figure 4: Estimated size of global EV fleet – historical development of forecasts
— Source: Collated by Daniel Quiggin of Chatham House. The BP data are taken from various issues of the BP Energy Outlook (https://www.bp.com/en/global/corporate/energy-economics/energy-outlook.html); and the Bloomberg New Energy Finance (BNEF) data are taken from BNEF Energy Outlook (https://about.bnef.com/new-energy-outlook).
Figure 4: Estimated size of global EV fleet – historical development of forecasts
Over the past 10 years or so, projections of EV penetration have, almost without exception, dramatically understated actual penetration levels. Figure 4 shows that major forecasters’ estimates of the size of the global EV fleet have typically been adjusted upwards from one year to the next.17 That these underestimates are so prevalent is surprising given that EVs tick all the right boxes in terms of attributes for energy technologies. They tick the security-of-supply box. If the Strait of Hormuz – a key transit route for oil – is somehow closed,18 the consuming countries most reliant on petrol or diesel will be the most affected. The higher the penetration of EVs in a given country, the lower the likely impact of disruptions to oil supply. EVs also tick the environmental impact box19 – with the caveat, in relation to climate concerns, that this assumes that the electricity such vehicles use is generated by renewables, which may not always be the case.20 In terms of impact on urban air quality, support for EVs assumes that diesel vehicles are displaced. Finally, EVs also present a potential solution to renewables’ problem of ‘intermittency’ – where natural fluctuations in generation (e.g. when the sun doesn’t shine or the wind doesn’t blow) create reliability-of-supply issues and may require costly back-up generation. Intermittency in renewables can be solved, at least in theory, by storage.21 One option for short-term grid management is to use batteries. A large car parc would provide significant storage capacity. Thus, the practice of a car in a garage being charged overnight can be replaced by one of a car discharging electricity to supply the grid – with its owner, of course, being paid for it.22 An even more exciting prospect is that roads could be built with induction strips so that the car can charge itself while driving (Lumb, 2018).
The prospect for a more rapid spread of EVs looks good as the costs of EVs fall, in line with the rapid development of battery technology. Between 2010 and 2016 the costs of an EV battery fell by 73 per cent (BNEF, 2017), and further falls are expected.
Between 2010 and 2016 the costs of an EV battery fell by 73 per cent, and further falls are expected.
Of course, there are barriers to the spread of EVs (Quiggin, 2017). There has been much concern over the availability of lithium, the basis for much of the battery capacity in many EVs. Such concern has arisen because of lithium’s frequent classification as a ‘critical metal’. However, the designation ‘critical’ does not necessarily denote scarcity; it can refer to the criticality of the mineral for certain economic sectors, or the concentration of supply chains. Indeed, the large multinational mining companies are combing their historical records, since historically lots of lithium was found but ignored as having no value.23 Cobalt, with its heavy reliance on supply from the Democratic Republic of the Congo (DRC), presents another potential supply bottleneck for battery makers. However, researchers are trying to develop batteries that are less cobalt-dependent (Chandler, 2017).
A widespread shift to the use of EVs also has implications for managing and expanding the power grid. If large numbers of motorists drive home and plug in to recharge their EVs at the same time, it could create an unacceptable peak load on the system. However, this should be manageable given modern metering technology. There is also the issue of access to charging points. Not everybody has a garage, an issue potentially of particular relevance for car ownership in high-density urban areas. Thinking about solutions to the environmental challenges posed by the rise of mega-cities requires ideas other than those related to individual car ownership.
It is also increasingly clear that there is a growing policy drive by governments and automotive manufacturers to develop EVs and phase out internal combustion engines (ICEs). Most notably, recent discussions have considered banning diesel vehicles from many cities because of their contribution to particulate pollution.24 Many national governments have come out with statements illustrating their desire to phase out ICEs and encourage the use of EVs (S&P Global Platts, 2018). China has indicated that one-fifth of new cars will be plug-in or hybrid by 2020,25 and the authorities there are reported to be ‘researching a time line for a complete ban on ICEs’. What happens to EVs in China could be crucial to the future of EVs internationally (Butler, 2018), in much the same way that industry developments in China have significantly influenced the costs of solar panels. Both the UK and France have said they will halt production and sales of ICEs by 2040. The Indian government has claimed that all new cars sold in India will be electric after 2030. Germany is proposing an ICE sales ban from 2030.26 Japan, Austria, Denmark and Ireland are discussing setting targets for EV sales. A similar trend can be seen with automotive manufacturers. Volvo, VW Group and BMW have all stated that all their models will have an ‘electric option’. Toyota and Nissan have also made claims about plans for zero-emission vehicles in the future. In all these cases, it is necessary to point out that talk is cheap. Until specific policy measures to achieve targets emerge, the claims of governments and the automotive industry need to be treated with caution.
BP has suggested that the introduction of an extra 100 million battery EVs would lower oil demand by 1.4 million barrels per day – a reduction of only 1.5 per cent over total consumption in 2017.
Extensive penetration of EVs into the car parc, if it occurs, will not in itself lead to the demise of oil use, and it would be unwise to overstate the anticipated impact of EV growth on oil demand. The IEA estimates that the billion or so ICE vehicles on the road worldwide account for around 40 per cent of global oil demand.27 BP has suggested that the introduction of an extra 100 million battery EVs would lower oil demand by 1.4 million barrels per day (b/d) (BP, 2017) – a reduction of only 1.5 per cent over total consumption in 2017.
A number of factors complicate the arithmetic of how EVs may affect oil demand in the future. First, there is the legacy of the existing car parc. For example, BP argues in its Energy Outlook (BP, 2018b) that its market scenario in which sales of new ICE vehicles and plug-in hybrid EVs are banned from 2040 assumes that ICEs will still be powering one-third of passenger cars in that year. Second, as EVs begin to dominate the car parc, it is possible that automotive manufacturers will stop investing in research and development to improve ICE efficiency long before any formal ban on ICEs comes into force. This would significantly slow the reduction in oil demand expected from improved energy efficiency. Having said that, much of the improved performance of ICEs over the years has been the result of legislation in the US (for example, the CAFE28 standards introduced in the 1970s), EU emission standards and policy in China, as part of efforts to reduce dependency on oil imports. There is no reason to assume that effective policy measures cannot be sustained and/or developed in ways that would force automotive manufacturers to continue improving ICE performance.
In terms of the implications for oil consumption, transport is not the only variable to consider, as passenger vehicles that could be replaced by EVs account for less than 20 per cent of total oil demand. Many oil consumption forecasts also assume that the role of oil as a feedstock for petrochemicals will add significantly to future demand. For example, the IEA projects that one-third of the growth in oil demand by 2030 will come from petrochemicals, rising to almost 50 per cent by 2050 (IEA, 2018b). However, as noted earlier, the IEA is arguably highly optimistic about future oil demand. Also, its projection does not take account of growing concerns about the pollution caused by plastics that fail to degrade. This raises a legitimate question: is plastic likely to be the ‘new tobacco’ in popular opinion, and might we thus expect a strong groundswell of activism seeking to reduce demand for petrochemicals in the future? This possibility is reinforced when one realizes that 40 per cent of plastics are used for packaging (Parker, 2018), much of which might reasonably be deemed a non-essential luxury.
There are also other factors at work in relation to oil demand. It is clear that car ownership is declining among much of the younger generation in OECD countries. Rapid urbanization and the spread of asset-sharing models – e.g. car clubs and ride-matching services such as Uber – point to declining car ownership and thus, potentially, lower oil demand in the future. This will be especially true if city planners prioritize public transport and connected-mobility solutions as part of policy efforts to achieve climate and clean air objectives.
Other technological changes
Other areas of technological change may also influence future demand for oil. First, there is the role of artificial intelligence (AI)29 and automation, which promise to create a so-called ‘fourth industrial revolution’. This takes the third industrial revolution, namely the digital revolution linked to computers and automation, and adds to it cyber physical systems whereby operations are controlled by algorithms integrated into the internet (Schwab, 2016). The implications of this for energy consumption are far from clear. For example, self-driving cars might be expected to operate in a more fuel-efficient manner. But at the same time, there is an interesting potential contradiction between electricity use and information technology. Greater use of information technology may save electricity on the one hand, but on the other hand it may lead to increasing electricity consumption to generate the technology in the first place. AI, automation, the ‘big data’ revolution and the rise of blockchain computing are emerging as important sources of potential disruption to markets and industries. How specifically all this might affect oil and energy consumption is not entirely clear. It might be assumed that all of these developments in communication and connectivity lead to greater operational efficiency, which would improve energy efficiency. For example, already the oil industry is increasingly using big data both to lower upstream costs and to streamline the transport, refinement and distribution of crude oil and oil products (Marr, 2015; Zaidi, 2017).30 The development of hydrogen-based fuel cells could also be a source of unexpected disruption to oil industry dynamics.31 It might also be argued that such technical developments will allow a more flexible power sector to emerge, enabling more renewables to be integrated into the power grid, leading to lower costs.
Another oil price shock?
The MENA region, which accounts for 38.6 per cent of global crude oil exports and 51.4 per cent of global proven oil reserves (BP, 2018a), is currently extremely unstable. Indeed, in the view of this author it is necessary to go back to 1918, at the end of the First World War and during the collapse of the Ottoman Empire, to find a period when the region was so unstable and unpredictable. Furthermore, there is a serious possibility of even greater instability to come. Lower international oil prices since mid-2014 have caused serious problems for oil-producing governments in the MENA region, reducing their ability to assuage domestic political unrest through public spending. Figure 5 shows the budgetary break-even oil price for OPEC members shortly before the oil price collapse of 2014.32
Figure 5: OPEC budgetary break-even price, 2014
— Source: Aissaoui, 2014.
Figure 5: OPEC budgetary break-even price, 2014
The weighted average break-even price at that time was $102 per barrel, and only Qatar and Kuwait could survive fiscally at $60 per barrel. Yet the average price per barrel of Brent crude was $54.18 in 2015, $44.67 in 2016 and $54.67 in 2017 (BP, 2018a). Even as lower oil revenues have limited the ability of producer governments to buy off local discontent, the original drivers of the Arab uprisings that began in Tunisia in early 2011 have not been addressed.33 Discontent has simply been repressed. Add to this maelstrom of uncertainty the impact of US President Donald Trump’s disruptive behaviour in the region – especially in respect of Iran, and his blatant bias towards Israel – and the scene is potentially set for major upheaval or even military conflict in the region.34 This would inevitably cause oil supply outages, in turn likely generating an oil shock in the form of higher prices. It is impossible to determine how high such a spike could go, but it is worth remembering that the price of Brent hit $144 per barrel in mid-July 2008. How long any price spike might last is also extremely uncertain.35 However, a rise of any magnitude is likely to cause governments in oil-importing countries to step up their efforts to move away from oil as a source of energy, thus reinforcing the current energy transition. Such a policy response was certainly in evidence in relation to the second oil price shock, in 1979–80, when the G7 governments, at meetings in Tokyo and Venice, all agreed to reduce dependence on oil by using sales taxes to increase oil product prices to the final consumer.
‘OECD disease’ and the rise of oil product prices
‘OECD disease’ refers to the policies developed by OECD governments in the early 1980s following the oil price shocks of the 1970s. These policies involved increasing sales taxes on oil products to increase the final price to the consumer, thereby creating incentives for consumers to use less oil. Although these taxes were intended to reduce oil consumption, there was another motivation: taxing oil products is a great way of raising revenue for a government. Oil products are widely used, so offer a very large tax base. Demand for them is inelastic, so high tax rates can be imposed. Finally, the cost of collecting the tax is very low. In theory, all that is needed is to have someone register the tax to be paid each time a tanker leaves a refinery, thereby involving few transactions. The picture in the G7 and the UK before the crude oil price collapse of 2014 can be seen in Figure 6, with government sales taxes dominating the final price to the consumer.
In recent years many non-OECD governments have also discovered the appeal of increasing sales taxes on oil products in contexts in which, traditionally, governments seeking to raise tax revenue have faced many barriers and problems.
The reason this phenomenon has been called a ‘disease’ is because it is catching! Thus, in recent years many non-OECD governments have also discovered the appeal of increasing sales taxes on oil products in contexts in which, traditionally, governments seeking to raise tax revenue have faced many barriers and problems. Recent examples have occurred in India since 2002 and China since 2009, and also include measures enacted by many other governments. The key consequence is that even if crude oil prices were to fall, higher oil product prices, boosted by sales taxes, could continue to reinforce the energy transition from oil towards renewables.36
Figure 6: ‘OECD disease’, 2014 – sources of the composite product barrel final price to consumers
— Source: OPEC, 2016.
Figure 6: ‘OECD disease’, 2014 – sources of the composite product barrel final price to consumers
Other reinforcing factors
A corollary to the ‘OECD disease’ phenomenon is the question of whether removing subsidies on oil products – in effect raising prices – will also tend to discourage consumption. Many governments have taken advantage of lower crude oil prices since 2014 to remove the high levels of subsidy that have characterized their pricing policies (Kojima, 2016; Fattouh, Sen and Moerenhout, 2016). A process that raised prices for oil products would also be strongly reinforced if carbon pricing became a serious policy option. Also worth noting is that the factors supporting an energy transition are often mutually reinforcing. For example, higher costs for petrol and diesel will encourage further improvements in EV technology. Similarly, electrification of grids using power from renewable sources drives demand for storage that can be provided by a large EV car parc. Improvements in battery technology from EV development enable more renewables to be used on grids. The list of reinforcements of reinforcing factors could go on.
11 Nuclear power gained momentum in the period 1970–2000, but since then its contribution has declined.
12 BP does not include biomass in its Statistical Reviews of World Energy because of the lack of accurate data.
13 Clearly, the impacts of such concern on hydrocarbons vary. They are worse for coal and oil, but less so for gas. However, there are concerns about fugitive methane emissions, since methane is a far more potent greenhouse gas than CO2.
14 This might be seen as a major barrier, although the decentralized nature of renewables makes adjustments to the grid less relevant. Also, it will be less of a barrier to countries building new grids.
15 ‘Car parc’ is a specific bit of technical jargon that refers to the total vehicle fleet.
16 See also Carbon Tracker, 2019 and Cherif et al., 2017.
17 However, to put Figure 4 in context, the 2017 BP projection that 100 million EVs will be on the roads by 2035 will be in a global car parc estimated at 947 million vehicles in 2015.
18 This is an issue revived by the attacks on oil tankers in the Gulf of Oman in June 2019: one a Norwegian-owned tanker, the other a Japanese-owned tanker.
19 It is being argued that the environmental benefits from EVs have been overstated. Thus the life-cycle emissions are ‘not as impressive’ as claimed and the wheel-to-wheel overall system efficiencies are ‘not as great as first thought’ (KAPSARC, 2018: p. 8).
20 For example, China appears to be building coal-fired stations in remote areas.
21 Electricity storage presents very complex issues and there are many barriers. Also, the speed of storage development varies enormously geographically (Ruz and Pollitt, 2016).
22 In Western Europe, it is estimated that some 80 per cent of car journeys per day are less than 20 km in length. Assuming a battery range of at least 120 km, this represents a very large amount of potential storage that, given current electric meter technology, could earn the car owner money.
23 Lithium is often found together with lead. So in Cornwall (southwest England), companies are investigating discarded lead mines for lithium potential.
24 There is growing attention to the health dangers associated with poor air quality in cities, and every day there are newspaper reports of some new health threat from toxic air. See, for example, Carrington, 2018.
25 Although this target is fairly soon, it is not clear how likely it is to be achieved.
26 There are voices being raised against this idea, but some regard Germany as a laggard in this area.
27 It should be pointed out that the IEA’s record of forecasting EV penetration, and indeed wind and solar production, has not been particularly good (Hoekstra, 2017).
28 Corporate Average Fuel Economy.
29 Much is made of the potential for driverless vehicles to reduce oil consumption. However, the use of driverless vehicles in urban areas faces severe constraints. In the Lebanese ‘highway code’, for example, a paragraph states that if two drivers approach a junction and the right of way is not clear, they should get out of their vehicles and decide who should go first. In dense urban traffic it is inconceivable that automatic vehicles, programmed for safety, could manage such a process. Casual observation suggests that congested urban traffic only flows because many drivers ignore the ‘rules of the road’. Traffic engineers have dubbed this ‘deferential paralysis’ (Adams, 2016).
30 There is always the possibility, of course, that lowering costs will generate a rebound effect as lower costs lead to lower prices and increased consumption.
31 The Japanese government remains very enthusiastic about this option.
32 The reason for picking this date is that it illustrates the economic vulnerability of the oil producers just before the oil price collapse that began in the summer of 2014.
33 Those fighting and demonstrating on the streets of many Arab capitals hated the term ‘Arab Spring’. They saw it as a Western media construct and thought that the term’s association with the Prague Spring of 1968 implied automatic failure. They preferred either the ‘Arab Revolution’ or the ‘Arab Uprising’.
34 Trump’s strategy for Iran is now clear. He intends to make the life of ordinary Iranians so desperately bad and miserable that they will rise up and overthrow the government in Tehran. It is effectively regime change on the cheap. It is unlikely that the government of Iran will sit back and accept this. What its response might be is difficult to predict, but the classic response to unrest at home is to create a diversion by a crisis abroad. Meanwhile, tensions between Tehran and Washington continue to grow.
35 By the end of 2008, the price of Brent had fallen to $34 per barrel.
36 It is possible to argue, given the experience of fuel protests in the UK in 1999–2000 and the more recent activities of the gilets jaunes in France, that there is a limit to what the public will accept in terms of higher prices from imposed sales taxes.