The social licence of coal

I’m (just) old enough to remember the Australian nuclear disarmament (and associated opposition to nuclear power) rallies from the 1970s, the fervent opposition to the Newport gas-fired power plant in Melbourne, Tasmanian dams protests of the 1980s, along with logging and woodchipping protests, and so on. Just about every energy source attracts opposition. The interesting thing is that although there’s a broad understanding of the need to transition from unabated coal, there doesn’t seem to be the acute feeling against coal-fired electricity in relation to historic conservation campaigns. Indeed, historically, coal was often advocated as an energy source that could complement renewables, provide energy security, and substitute for oil and gas.

A parallel narrative is the broad support for renewables. But much of the community is yet to appreciate the practical constraints of high-penetration renewable scenarios, and the inevitable synergy with gas in the absence of dispatchable renewables.

I would argue that this helps to explain at least part of the political stalemate of Australian climate policy – no consensus on CO2 pricing; but support (if recently equivocal from the Government) for renewables. Indeed, Australian electricity seems to be on a trajectory that will emulate the original aspirations of the 1980 forerunner to the German Energiewende – 50 to 55% coal and 45 to 50% renewable energy by 2030 (with a larger role for gas in Australia due to indigenous resources). The social licence of coal seems to be a key factor, and I’ve put together a few observations in no particular order –

  1. As a pioneer nation, Australia’s economic roots lie in agriculture and mining. Australia readily exploited the power of steam. Local coal was an antidote to a reliance on imported fuels – oil and natural gas were not developed until the 1960s. Furthermore, affordable electricity was essential for incubating a manufacturing industry.
  2. Prior to climate change, people really weren’t that worried about coal. Even the Australian medical and academic community seem to have not been too concerned about studying coal-fired power’s health impacts (see two recent reports by the ATSE and BZE). Most health studies were related to the mining of coal rather than combustion.  
  3. The Australian geographic distribution of power plants in low population density areas has mitigated against the worst effects of pollution. Australian coal is low in sulphur, lessening the likelihood of acid rain.
  4. During the 1970s and 80s, despite pressing for more funding for ‘alternative energy’, environmental advocates treated coal as a relatively benign fuel. For example, in advocating a policy of environmental protection, Hugh Saddler (1981, pp. 119-120) argued that coal was a more economic and less risky option than nuclear. In arguing the case against Tasmanian hydro development, Peter Thompson, representing the Australian Conservation Foundation, noted that coal plants ‘pose relatively few air pollution problems if the operation is adequately planned, sited and built to the highest standards of quality’ (Thompson 1981, p. 125). Similarly, during the Franklin River campaign in 1981, Bob Brown stated that ‘a new coal fired power station is the manifestly best option built on Tasmanian coal fields.’
  5. A similar view was held in Germany and the US. During the 1970s, the German Government actively promoted the expansion of coal for electricity and combined heat and power (Guilmot et al 1986, pg. 20). Even the contemporary Energiewende was originally conceived around Germany’s substantial coal resources. The Energiewende emerged from a study by the German Öko-Institut in 1980 that grew out of concerns of oil security from the first oil crisis, and the safety of nuclear energy (Krause et al. 1981; Joas et al. 2016; Maubach 2014; Morris & Jungjohann 2016). The study, titled ‘Energy turnaround, growth and prosperity without oil and uranium’, envisaged a German energy supply derived from 50 to 55% coal and 45 to 50% renewable energy by 2030.
  6. In 1977, pro-conservation US President Carter proposed an 80% increase in coal production for power generation and liquid fuels, arguing for ‘the expanded use of coal, supplemented by nuclear power and renewable resources, to fill the growing gap created by rising energy demand’ (Stobaugh & Yergin, 1979, pg. 80).
  7. From Lowy and other polling, the willingness to pay a higher cost for electricity is very limited – in 2011, 39% were prepared to pay no more and a further 32% were prepared to pay no more than $20 a month. The majority of Australians like the idea of an energy transition but aren’t willing to pay for it. Although the Murdoch Press and elements of the Lib-Nat Coalition are critical of the LRET, the community seems to support both the LRET and rooftop solar.
  8. The IEA projections for India illustrate the way in which a ‘techno-optimist’ narrative of growing renewables (or nuclear) can co-exist with the stark reality of growing coal-fired generation. According to the IEA ‘New Policies Scenario’, solar PV in India is projected to grow around 60-fold by 2040, wind around 7-fold, and nuclear around 6-fold. But despite coal’s share falling to 57% of generation, coal-fired generation is projected to nearly double in absolute terms because of the sheer scale of India’s demand growth. The reasons for India’s increasing demand for coal are simple – coal is cheap and easily shipped, requires no pretreatment, and most importantly, provides fit-for-purpose dispatchable generation. It does not require smart grids or storage to provide dispatchable power, nor the institutional and community support that nuclear requires.
Figure 2.22, IEA India Energy Outlook

Ferroni and Hopkirk revisited

Ferroni and Hopkirk published a paper last year titled ‘Energy Return on Energy Invested (ERoEI) for photovoltaic solar systems in regions of moderate insolation’. They concluded that the EROI of solar PV in Switzerland was below 1:1, or a net energy loss. Not surprisingly, the paper raised a few eyebrows, and a rebuttal has been published from Raugei et al. This follows on from a previously published rebuttal of a paper by Weißbach et al. in 2014, also with a low solar EROI.

The critique from Raugei et al. identifies a number of methodological flaws and inconsistencies in Ferroni and Hopkirk, but the main criticism relates to the use of non-conventional methodology –

Net energy analyses may be conducted using a variety of boundaries and assumptions, all of which, in principle at least, may be considered valid … but that … extending the EROI boundaries … shifts the goal of the analysis from the (comparative) assessment … to the assessment of the ability of the analysed system to support the entire societal demand for the type of energy carrier it produces, or sometimes even for all forms of net energy.

I agree with much of Raugei’s critique regarding PV system price, the problem of ascribing energy consumption to labour, the use of outdated data, assumptions around performance and others. Raugei (and others) have made an important contribution. The only way in which EROI can be taken seriously is be adhering to a consistent methodology. But the critique also reflects a broader problem – the IEA-PVPS guidelines are not providing a comprehensive examination of the value of PV. It’s worth looking at how these studies are conducted.

Solar PV life-cycle assessments (LCA’s) are nearly always conducted with a process-based life cycle inventory using an attributional framework. This requires drawing a boundary around the manufacturing processes of PV and measuring the direct energy into those processes. Depending on time and effort, the researcher steps back up the value-adding chain and cumulatively adds up the embodied energy. Since there are a only few energy intensive processes, identifying those processes is said to provide a reasonably comprehensive stocktake of embodied energy. The main benefit of this method is that the researcher can apply the clearly defined guidelines and boundaries from the IEA-PVPS. This is useful for comparing similar products. LCA practitioners declare the scope, utilise standard methods, and these work well within the LCA community.

The main point of contention is that many EROI practitioners are more interested in a ‘bigger picture perspective’ than a comparative assessment of different PV types. High EROI generally correlates with low energy cost and contributes to productivity and economic growth. The EROI of oil, coal, and hydro has conformed to this principle. High EROI oil drove earlier 20th century economic growth, while lowering EROI from the 1970s contributed to recessions. We want to know whether an energy source is a net-source or net-sink and how much it contributes to human welfare. Where is our energy going to come from as we rely less on fossil fuels? What is the new energy source substituting for?

The problem is that the pre-defined and narrow boundaries defined by the IEA-PVPS are not really telling us much about the impact of solar PV on overall energy costs or economic growth.

As a starting point, solar PV has three important strengths –

1. it’s modularity makes it highly scalable
2. the photoelectric effect lends itself to progressively lower manufacturing costs and easy installation
3. its social acceptance ensures a low regulatory hurdle.

Yet despite improving EROI and lower costs, solar is not leading to lower overall energy costs or transforming industry in the way that previous energy transitions have. In European countries, there is a strong correlation between installed RE per capita and electricity costs. The conventional idea of EROI as it applies to oil, coal, and hydro does not seem to apply.

There seems to be three main issues.

1. Solar is highly seasonal and in many parts of the world, it’s availability does not correlate well with demand.
2. Although predictable, solar power is of course variable.
3. Unlike conventional electricity generation, nearly all of the energy investment of solar is an upfront energy debt, but the EROI is calculated on an energy return over a 25 or 30 year life.

Ferroni and Hopkirk’s solution for intermittency is to add storage to the analysis. But Raugei et al. note that –

For example, they add an unreasonably extended storage requirement to PV but not to nuclear, ignoring that PV primarily serves peak loads while nuclear only serves base loads and both of them (not only PV) would require storage in order to satisfy total demand loads. This is problematic because the way in which the analyses are presented to the reader implies that any differences in the reported EROIs are due to data inputs – i.e., something inherent to the technologies or resources under investigation – and not an artefact emerging from methodological inconsistencies between the studies being compared. The latter is actually the case here.

I agree that there is a problem of adding an arbitrary amount of storage without providing supporting evidence for why this quantity has been selected. But Raugei et al. also seem to mistake the reason for buffering of solar. Conventional generation does not need buffering to ensure a low outage rate – each generator contributes to overall system reliability. The reason for adding storage to intermittent power is to increase its availability, whether that is used during peak or off-peak periods. There is certainly a case to made for thinking about whether the cumulative energy demand (CED) of peak-load generation should be added to baseload in order to arrive at a total picture, but this is quite different to arguing that there is an equivalence between variable renewable energy and baseload. Electrical systems are built according to system capacity, not system annual energy.

In summary, the second last paragraph in the conclusion perhaps reflects the divergence between the aims of different EROI researchers –

Also, extending the boundaries of the EROI calculations in order to estimate the ability of a certain technology to support the present civilization tends to stretch the value of this measurement beyond its initial intended purpose, and the vast uncertainties involved in doing so make it a risky enterprise that might easily lead to wrong policy choices.

Yet understanding how the ‘present civilization’ can transition from fossil fuels is precisely the motivation for many EROI researchers. If EROI researchers are not going to undertake this task, who will?

The oil-gold nexus

Crude oil has a unique status in global energy markets and underpins global economic activity. Since the 1970s, it has taken on a role as a quasi-monetary commodity. By the end of World War II, the US held around 70 percent of global gold reserves. The US supplied 6 billion out of the 7 billion barrels of oil consumed by the Allies for the period of World War II. Japan and Germany’s deficiency in oil were key factors in the Allied victories.

The Bretton Woods agreement of 1944 shifted the dominant trading currency from the pound sterling to the US dollar, fixed national currencies to the US dollar, and converted the dollar to a fixed amount of gold. The collapse of Bretton Woods in 1971 occurred at the same time as US hegemony in oil production passed its peak. By 1973, when the US was dependent on imported crude, it became impractical to regulate the price of oil and local price regulation of oil ended. This marked the beginning of the modern era of oil markets.

In Australia, the local oil price was regulated, which led to a sharp disparity when the world price rose in 1973. In 1975, the Federal Government introduced a levy on domestic production, and in 1977 introduced a phased introduction to import parity pricing.

Since the 1970s, oil and gold prices have tracked remarkably closely. One explanation is that in the absence of a metallic monetary base, oil has taken on a role as the base for a quasi-monetary commodity (Sager 2016). In recent years, both oil and gold seem to have been driven by factors other than simply supply-demand relationships. Oil and gold can also act as safe havens during extreme declines in other asset classes, such as equities and bonds. The departure from the relationship since late 2014 suggests that the current low oil price should be seen as an deviation from the long-run trend, and that a price of in the range $60 to $80 would better reflect fundamentals.


Oil and gold price in current USD. Source: St Louis Federal Reserve

In the US, high and rising oil prices often precede US recessions and there seems to be a threshold for expenditures, above which the US economy tends to be in a recession (grey shaded areas in graph). At $45/barrel, oil makes up around 2% of global GDP, but its role in the macroeconomy is much greater than its factor share would suggest.

None of the other energy sources or natural resources seem to have this intimate role with the monetary base. Wood and coal are solids and shipped by bulk transport, and require materials handling at each stage. Combustion produces a solid waste that must be disposed of. At standard pressure, natural gas is only tradable within fixed pipeline networks, but can be shipped as a highly pressurized liquid, at a cost. Both coal and gas can be upgraded via Fischer–Tropsch to substitute for petroleum, although the upgrade carries a significant net-energy penalty. Electricity has the highest utility (i.e. is easily convertible to heat, light or motion) but requires connection to a grid operating in a real-time supply-demand balance.

The dual identity of rooftop solar

We usually purchase energy, not because we value energy per-se, but because we value the energy services they provide – natural gas because we want warm homes or petrol because we want to get somewhere.

‘Fuel to service’, from Cullen and Allwood, 2010, The efficient use of energy: tracing the global flow of energy from fuel to service.

The curious thing about solar is that many consumers are buying solar, not just for the energy, but because they value solar as a consumer product.  Remaining connected to the grid is an essential prerequisite for maximising the value of solar – solar is not adding energy services that wouldn’t otherwise be available. The solar heating Sun Lizard product (seen on the ABC Inventors) was an example of a useful but high-priced consumer product that mostly gave householders the satisfaction of having a solar product installed on their roof.

Rooftop solar is perhaps unique in being the first energy supply product that is part of consumer culture. Josh Floyd suggests that solar has a kind of dual identity at the microeconomic level. The fact that it operates outside of the conventional energy paradigm is the reason that electricity utilities have struggled to effectively grapple with the rapid uptake of solar. Similarly, many environmental economists argue that the high carbon abatement cost of solar leads to the misallocation of low-carbon investment if carbon abatement is the goal.

From a net-energy perspective, the interesting question is the degree to which the installed solar capacity contributes to national wealth and taxation, and how much could be considered consumer surplus (i.e. consumers deriving satisfaction from the ownership of solar). A food corollary might be to consider to what degree a value-added food service (i.e. restaurant service, premium wine, etc.) contributes to calories and nutrition, and to what degree society remains dependant on the mass production of grains and staples to underpin calorific and nutritional intake. Consumers that elect to consume a greater proportion of their income on discretionary foodstuffs do so because they value the ‘food service’ – the purchase of expensive wine is an example. But in the context of ensuring that everyone has adequate nutrition and calories, it might be unreasonable to expect the cost of premium wine to be absorbed into the cost of bread and milk, for example.

The case of off-grid solar represents the paradox of green consumerism – householders chose to forgo the purchase of other consumer products in order to buy into a culture of sufficiency. Yet the additional energy cost of going off-grid far exceeds the energy cost of remaining connected and simply reducing energy consumption.

Dynamic EROI of off-grid solar, from Palmer, 2014, Energy in Australia

Despite making up a small proportion of annual energy supply, solar is nonetheless leading to a reappraisal of the Australian wholesale electricity market. It is a global characteristic of electricity systems, that unlike pharmaceuticals and computer software, investment on research and development makes up a very small proportion of revenue. Hence solar is leading the charge for a consumer-centric re-examination of electricity supply and may eventually disrupt the conventional tariff model.

My work with energy-return-on-investment suggests that the maximum value from solar is likely to be achieved with a small amount of storage attached (2 to 4 hours) where it can add value to the low-voltage distribution network. Although most attention has been directed towards considering how distributed solar might interact with other renewable energy, the combination of solar and a small quantity of storage could arguably work better with conventional baseload. In the long-run, I think the penetration of rooftop solar is going to be limited to 10 to 15% in most regions because of the strong seasonality at latitudes higher than around 30 degrees and low annual capacity factor – the global distribution of population and wealth tends to be in regions at greater than 30 degrees latitude.


Global population density, from




EV penetration in Australia

The Australian amateur astronomer Bob Evans has such an incredible knack for spotting minute changes in the night sky that he holds the record for visual discoveries of supernovae. There seems to be an evolutionary advantage to being able to spot changes in the natural world but it also means that we tend to overstate rates of change in the technical sphere. In various experiments, Sterman and others have shown that people tend to misinterpret system dynamics problems, such as carbon accumulation, and the fluid balance of bodies. The problem stems from a misunderstanding of the principles of accumulation.

Another example is the hype surrounding electric vehicles (EVs), and the tendency to directly translate rising global sales to a meaningful proportion of the total stock of motor cars. Most energy transition studies project increasing electrification of transport. Trying to derive plausible substitution rates of EVs is essential to understanding future energy scenarios. Horace Dediu also covered this topic in Asymcar episode #28. As Horace noted, the proportion of the EV stock in the future is several times removed from the current sales growth, as shown in figure 1.


Figure 1.

In order to derive projections for Australia, I took Australian Bureau of Statistics car census datanew car data and average age data. I used census data to derive sales data back to 1955. I used the historic attrition rate of 4%. By optimisation, I found that the proportion of cars of a certain age remaining after t years can be roughly described by a logistic distribution –

Yt = 1 / (0.995 ( 1 – 0.995) e^(0.305 (t – 0.25)))   where t is age in years

The ABS does not disaggregate EV data so I’ve had to use other published data. There have been around 3,300 electric vehicles sold since 2010, with 1,140 sold in 2014 and 942 sold in 2015. Assuming that new models and lower battery costs will drive compound growth of 40% through to 2030, the sales volume of EVs in 2030 will be around 146,000 units per annum or 10.5 % of new car sales, and make up 2.5% of the total stock of passenger vehicles. Around 20% of the passenger vehicles that will be on the road in 2030 are on the road today. The 40% projection is a best guess based on strong but plausible growth. The year on year growth up to September was 23% and 33% for Europe and the US respectively. I suspect the tip-over point will occur when the the price of an EV drops below internal combustion of an equivalent vehicle, and the case for electric is compelling. People will simply adapt to the charging regime.


Figure 2. Number of passenger motor cars first registered from 1955 to 2015 as at 2015. Smoothed estimate with logistic distribution, based on ABS car census, sales data, and average age data, various years.

Figure 3. Historical and projected stock of passenger motor cars and EVs. Assume compound sales growth of 3% for new cars and 40% for EVs.

There are several challenges I see with increasing the adoption of EVs.

Firstly, the charging infrastructure may take decades to be comparable with petrol/diesel refuelling. It isn’t just a matter of infrastructure, but overcoming the limitations of electrical charging. For example, the Tesla Supercharger delivers 120 kW and takes 30 minutes to deliver 270km of range, or 0.15 km of range per second of charging. In the future, this will improve, but there is a limit to the charge time and current – assuming 7,000 batteries in the Model S (Panasonic NCR18650B, 3.6 volt) calculates to 4.6 amps per cell. Most homes have a single phase 40 to 80 amp 230 volt connection, equal to 9 to 18 kW, and therefore rapid charging is not going to be possible at home. In contrast, a fill-up at a petrol station delivers around 500km of range in about 2 minutes plus another 2 or 3 minutes to go inside and pay, or about 1.7km per second at the petrol station. Hence electric fill-up takes around 12 times as long. Management of electrical assets will require that most charging takes place overnight.

When I had a pre-injection Commodore V-8 (a while ago), the most annoying thing was always being conscious of the fuel gauge. I now drive a diesel and fill up once every 2 or 3 weeks  and never worry about refuelling.

Early adopters find plugging in at home a few times a week a novelty, but it remains uncertain how average consumers will take to plugging in. Many people in inner-city houses without off-street parking, plus apartments, etc. won’t have the option of an overnight charge. Yet this is potentially a key early adopter demographic.

A marketing mismatch is between high and low-km motorists – the low running costs of EVs will benefit high-km motorists, but long daily commutes and rural/interstate travel will mostly preclude the use of EVs. On the other hand, low-km commuters with ready access to charging will be less willing to outlay additional capital to reduce already low running costs – fuel costs make up only around 15% of ownership costs for small cars (see RACQ guide).

Much of the hype around EVs comes from Silicon Valley entrepreneurs rather than the car industry. I doubt that many people care what type of driveline a car has except to the extent that the driveline provides the mix of performance, economy, reliability, drivability, etc. that consumers demand. The closest many people get to thinking about the driveline is ticking the ‘auto transmission’ option. A basic error is failing to account for what consumers actually want in a car.

The car ecosystem is complex. Dealerships want cars on the showroom floor that people will buy. Holden stopped selling the Volt because few people wanted to buy an electric-powered 4-seat Cruze, regardless of price. The GM range-extender is a technical fix but doesn’t change whether consumers aspire to own a Volt. GM take electric seriously and want to be in the game, but also need to stay in business while batteries improve. The sales data on the Nissan Leaf is informative – an early spike followed by declining sales despite strong discounting suggests that there is a small demographic of price-insensitive consumers that want electric, but that the majority of consumers do not value electric in-itself.

Until EVs sales reach a critical mass that forces dealerships to take them seriously, sales will be hampered by a lack of commitment – EVs represent a potential threat to profitability because manufacturers and dealerships are heavily dependent on parts and service. Long warranties keep customers within the dealer orbit. The low maintenance of EVs will require a rethink of the current dealership business model that relies on post-sale servicing revenue. Software updates, tweaks and free data features may provide the motivation to keep customers within the dealership orbit.

Having driven a Tesla Model S, my immediate impression was of a tight, sharp and powerful car. Electric works. Home owners with a garage and another car for rural driving will find that electric works for them. But for all the excitement of electric drive, the car didn’t do anything that a conventional car doesn’t already do. The car was silent at low speeds, but a Lexus LS is quieter on the freeway. Long service intervals, twin-clutch transmissions with millisecond gear shifts, high efficiency diesel and hybrid drivetrains are a routine feature of internal combustion. The engineering elegance of so few moving parts is irrelevant for the new car buyer driving a reliable car with 1 year+ service intervals.

I’ll tackle the vital issue of embodied energy of EVs in another post.

Does Australian household spending on electricity conform to the Bashmakov constant?

Igor Bashmakov is one of the leading Russian experts on climate change and energy efficiency. Bashmakov developed three general energy transition laws: the law of stable long-term energy costs to income ratio; the law of improving energy quality; and the law of growing energy productivity. The proportion of GDP that households spend on electricity seems to be remarkably constant across nations and across time. Carey King has also explored similar relationships between energy and the economy (for example, see here and here).

To explore the relationship for Australia, I plotted the real price of electricity from 1955 to 2015 in 2015$AUD as shown in figure 1. I also took the nominal price and multiplied through by the household consumption and divided by nominal GDP, shown in figure 2. Consumption and price data was gathered from various ABS sources and Year Books, ESAA reports, OECD, and historical data. Full electrification of rural areas did not occur until the 1960s. Despite significant changes in the real price and GDP per capita, the share of GDP seems to pull back to within a band of 0.8 to 1.0% GDP. I haven’t carried the analysis further but the next step would be to consider income share and income growth. The recent role of solar may be important here – households offset increased spending due to higher tariffs by installing solar, but may eventually increase electricity consumption to bring annual spending back into line with long-run trends. Consumers probably get used to the magnitude of the electricity bill and adjust behaviour to suit.

I then looked at household spending for a range of countries and plotted them versus the residential price of electricity (figure 3). One would expect nations with higher electricity prices to spend more on electricity. I used the IEA Energy Prices and Taxation report from 2015, household consumption from the World Energy Council, GDP in national currencies from Economy Watch. For the selected nations, there is a six‑fold difference in per‑household energy consumption, a four‑fold difference in PPP‑adjusted electricity prices and a two-fold difference in PPP‑adjusted per‑capita income. Once again, there seems to be a band of spending share on electricity that is much narrower than might be implied by the differences. German households can afford much higher electricity tariffs than the US because they use much less electricity. But German households use more than the Polish because Germans are richer. Either way, the GDP share seems to hover around  1% +/- 0.4%. For policy makers, the lesson is straightforward but impractical – the most effective method of reducing household electricity consumption is to increase tariffs  and/or reduce economic growth. The role of rising peak demand confounds the study because peak demand is related to a peakier load duration curve. Efficiency and conservation seem to be deeply cultural and can be nudged only gradually. The corollary is that consumer price subsidies are a bad idea except to the extent of providing essential support for needy households.

Update: Dylan McConnell also covers this topic here , Keith Orchison covers it here

Figure 1. Australian household real price of electricity, included fixed and variable tariffs
Figure 2. Australian household spending on electricity as a proportion of GDP


Figure 3. Household expenditure as a proportion of GDP versus electricity cost

Lessons for energy transitions – the case of the British and US navies

Case studies in energy transitions

The examples of the British and US navies provide an interesting exploration of energy transitions. In both cases, the decisions were unilateral and subsequently replicated by other nations. But it was not obvious ex-ante that the decisions would be successful, and both carried risks. The energy substitutions were fit-for-purpose, fitted readily into incumbent vessel design, and were strategically valuable. In neither case was cost or the availability or natural resources the driver – indeed the substitute technologies were costlier and used less readily available natural resources.

British battleships

At the beginning of the 20th century, Britain was the leading coal nation, but lacked an indigenous oil supply. The British Navy’s decision to switch from coal to oil for its battleships was considered a high risk strategy (Dahl 2001; Yergin 2011). By 1911, Britain had already adopted oil for destroyers and submarines, but coal remained the principle fuel for larger vessels, especially battleships. Coal-fired steam was a well-known technology and accepted by marine engineers. It also had the benefit of supplementing armour by absorbing damage from exploding shells (Dahl 2001).

The prime supporter of a switch was Admiral Fisher, who adopted a more progressive stance than much of the Naval establishment. In time, he managed to persuade the then First Lord of the Admiralty, Winston Churchill, of the strategic benefits of oil. The impetus for the switch was the rapid expansion of the German Navy and the resulting Anglo-German naval race at the start of the 20th century.

Churchill was to later note,
‘the ordeal of coaling the ship exhausted the whole ship’s company. In wartime it robbed them of their brief period of rest; it subjected everyone to extreme discomfort … (on the other hand) with oil, a few pipes were connected with the shore or with a tanker and the ship sucked in its fuel with hardly a man having to lift a finger … oil could be stowed in spare places in a ship from which it could be impossible to bring coal.’ (Dahl 2001)


Although oil power was already being used, the British Navy’s complete shift to oil, particularly for battleships, placed Britain into the ‘early adopter’ category. As the leading coal nation without (then) indigenous oil production, arguably Britain had much to lose. Military equipment is structured around the incumbent energy sources; engineers and personnel are trained in the equipment and pitfalls are understood. There must be a compelling reason to shift to alternative energy sources. In this case, the perceived risk from the rapidly expanding German navy provided the impetus for change. The British war college had advised Churchill that a battleship would need a speed of 25 knots to outmanoeuvre and ‘cross the T’ of the German fleet – only oil power could provide this speed.

Once the decision had been made for a complete changeover, the issue of supply had to be resolved. The solution was to invest in the Anglo-Persian oil company (later to become BP), which was the first company to produce oil from Iranian oil fields. The British government took a 51 percent share of company stock, and negotiated a 20-year supply contract under attractive terms.

US submarines

In a later parallel set in the late 1940s at Oak Ridge National Laboratory, United States Captain (later Admiral) Rickover was exploring how nuclear power could be harnessed to power a submarine (Loewen 2012). The tactical benefit of a more powerful submarine that was able to remain submerged for extended periods far from port was obvious – diesel-electric submarines needed to resurface for charging and to return to port for refuelling.


Alvin Weinberg from Oak Ridge had recently patented the pressurised water reactor (PWR) and persuaded Rickover to develop the technology for a submarine. Rickover agreed and a contract was awarded to Westinghouse. Rickover also wanted a second option, and General Electric (GE) was awarded a contract to develop the (then) more advanced sodium cooled reactor (or liquid metal reactor – LMR).

Three reactors of each type were to be built – one was a land-based prototype for training and testing, one for a submarine, and one spare. The liquid metal reactor was installed in the USS Seawolf (SSN-575) and the pressurised water reactor in the Nautilus (SSN-571), with commissioning in 1957 and 1955 respectively. The LMR ran for several years, but had ongoing superheater problems, and eventually the decision was taken to adopt the PWR as the only reactor type, including a retrofit of SSN-575.


The advantages of nuclear propulsion for blue-ocean submarines eventually proved decisive with the major blue-water navies (China, France, India, Russia, US, UK) eventually adopting nuclear propulsion for their submarine fleet. From 1948, up to 2003, the US had commissioned 210 nuclear powered naval ships (NASA Office of Safety & Mission Assurance 2003). Some of the current propulsion units have extended the refuelling periods such that the power plant is expected to operate for the life of the vessel without refuelling. For example, the General Electric S9G reactor powering the US Virginia class submarine is designed to operate for 33 years, beyond the expected 30 year life of the vessel (Ragheb 2012). In total, there have been around 700 nuclear powered ships (civilian and military) globally (Royal Academy of Engineering 2013).


In the two cases considered, the decision to adopt an alternative power source was unilateral and driven by relatively few individuals – Fisher and Churchill in Britain; Weinberg and Rickover in the US. In both cases, the success of the transition provided a precedent for other military forces but it was not obvious ex-ante that the transition would be a success. In neither case was cost or the availability of natural resources the driver.

The cases of French nuclear, Brazilian ethanol and German solar are similar examples of the adoption of a particular technology as a matter of national energy policy. The long term benefits and costs of solar are yet to be determined, with the German program being extremely costly for Germany but providing a key plank for global price reduction. Indeed, German is not the natural home of solar but the cost reduction is benefiting regions with high solar insolation. Coal with carbon capture is generally assumed to be an important climate mitigation technology, including its future use with biofuels, but hasn’t yet found an influential faction of supporters. The interesting thing about all of these programs is that they provide counter examples of the endless discussions of carbon pricing and cost-benefit studies. Depending on the outcome of the German Energiewende and other national energy policies, the future of national energy policies may end up being determined much more by governments ‘picking winners’ than market based instruments.

Update: Benedict Evans has an earlier great post drawing an analogue between the HMS Dreadnought and the iPhone.

Alexander, W.R.J. 2013, ‘The defence-debt nexus: Evidence from the high-income members of NATO’, Defence and Peace Economics, vol. 24, no. 2, pp. 133-145.
Dahl, E.J. 2001, Naval innovation: from coal to oil, DTIC Document.
Kennedy, P. 2010, The rise and fall of the great powers, Vintage.
Loewen, E.P. 2012, ‘The USS Seawolf Sodium-Cooled Reactor Submarine’, paper presented to American Nuclear Society local section address.
NASA Office of Safety & Mission Assurance 2003, Progress Report |, Naval Reactors Safety Assurance.
Ragheb, M. 2012, ‘Nuclear Marine Propulsion’, University of Illinois at Urbana-Champaign.
Yergin, D. 2011, The prize: The epic quest for oil, money & power, Simon and Schuster.