The competitive struggle between technology and resource depletion

The study of EROI may be of little more than academic interest except for two factors. Energy extraction has conformed to Ricardo’s ‘best first’ principle, meaning that the easiest resources to extract tend to be the first. But at the same time, human ingenuity continually pushes the technological frontier forward. Hence, there is a competitive and ongoing struggle between resource depletion and technology. The US shale revolution typifies this tension, and for a time at least, seemed to signal the triumph of technology. But the extraordinary ramp up of production beginning 2009, and that was supposed to run for decades and lead to US ‘oil independence’, has already fallen 600,000 barrels per day below its April 2015 peak (EIA).

On the other hand, wind and particularly solar PV, have driven even further down the cost curve, with further declines in the pipeline. At the same time, the shift to ICT-enabled service economies was supposed to decouple energy from economic growth. While learning curves often seem intuitive and obvious, ex-post, their use ex-ante can be fraught. Furthermore, while relative decoupling has been a feature of the advanced economies for several decades, absolute decoupling is still mostly an aspiration.

It’s useful to be reminded that the future is not at the end of a trend line. Seven years ago, concentrated solar was being projected by some as the new ‘low cost baseload’ power source of the future, but installed global capacity is still barely a blip at around 5 GW. A sober assessment could similarly be made of optimistic forecasts for nuclear, ocean, tidal and geothermal. The problem for wind and PV is whether the issue of intermittency is merely a ‘surmountable challenge’ or a fundamental constraint at high penetration. This is still an open question. In Europe, there is a strong correlation between installed variable renewable energy and electricity prices.

installed wind and solar per capita
Residential electricity price for EU countries 2014, including taxes, versus installed wind and solar. Sources: Eurostat, BP, IEA

On the other hand, the effects of energy supply constraints are real and daunting. The loss of oil from Cuba’s economy as a consequence of dissolution of the Soviet Union in 1989 provides a classic case study in oil dependence. The Cuban socialised model of agriculture had been sustained by the highly generous terms of trade of Cuban sugar for Soviet oil (Sinclair 2001) Although the drop in oil consumption was more moderate than originally feared – 20% over 2 years – the impacts were enormous and debilitating. The Cubans did eventually learn to live with less oil, and adapted to the loss of oil through decentralisation, urban gardening, and a range of market based reforms. The defining economic feature of the 1970s was the stagflation driven by the 1973 oil crises. King has shown that US expenditures on oil that exceed 4% of GDP is likely to be recessionary.

My perspective is that EROI provides a framework to understand these competing narratives – which I call the techno-optimist, encompassing the Silicon Valley mindset, versus the limits to growth narrative. Note that many in the degrowth movement, such as Alexander and Trainer argue that a future with less resources actually offers an opportunity for civilisational renewal.

technology versus depletion

Electricity networks and disruption

Clayton Christensen’s theory of disruption was originally based on his observations about disc drives. He saw that the manufacturers of 14” drives for mainframes had been driven out of business by manufacturers of 8” drives for mini-computers, and then the companies that made the 8” drives were subsequently driven out of business by manufacturers of 5.25-inch drives for PCs. So why hadn’t the 14” drive producers simply started producing 8” drives? And why did consumers want the inferior 5.25-inch drives?

In low-end disruption, inferior products compete in a smaller market with lower profit margins, but are ‘good enough’ for some consumers at a cheaper price. While established producers are focused on improving their products and profit margins, low-end producers capture economies of scale, and some products eventually gain a foothold by offering enough of what consumers want at a budget price. By the time the established producers realise that they’ve lost market share, a new market segment has been established and it is too late.

Discussions around solar and batteries are frequently discussed as ‘disruptive’, but is this really a good description of the changes happening in electrical systems?

Distributed power is to conventional generation what electric vehicles are to internal combustion vehicles. Both disrupt the incumbent products (generation assets or internal combustion engines respectively), except that EV’s are better described as high-end disruption. Furthermore, distribution networks are to electricity what roads are to transport. In both, the maintenance of the distribution assets is essential.

But when discussing EV’s, nobody talks about the ‘disruption’ of the road network, or the massive infrastructure and regulation that overlays passenger transport. All of the bridges, traffic lights, emergency and medical care, road law enforcement, traffic engineering, council maintenance and parking control. Roads are the arteries of transport. EV’s or internal combustion, the systems and infrastructure is the same.

On the other hand, the problem for electricity distribution networks is that the tariff structures employed in Australia, and around the world, rely mostly on energy-based tariffs rather than demand-based tariffs. Since owners of distributed energy assets consume less energy from the grid, their share of network costs is reduced by more than the avoided costs of installing the distributed generation. In other words, solar owners still rely on the grid at times of peak demand although they draw less energy from the grid over a billing period.



So will physical assets of the distribution networks be disrupted? The short answer is no. Indeed, much of the discussion around solar and batteries is around integrating virtual networks, aggregators, using blockchain to facilitate electricity trading among small users, network optimisation with smart grids, etc. The maintenance, modernisation, and upgrade of networks is absolutely essential to integrating a higher share of renewables. In virtually all high-penetration renewable scenarios, geographic diversity of wind and solar is a primary mechanism for smoothing intermittency. This requires an expansion of transmission infrastructure, not a contraction.

So what about disruption of the tariff model? Earlier tariffs were based on the social contract that those that used the most energy paid the most, regardless of the marginal cost of an individual consumer. Some consumer classes are implicitly protected as a matter of social justice. For example, pensioners that operate air conditioners on hot weekdays are cross-subsidised by other consumer classes, such as full-time workers not at home until the evening. Do we really want pensioners to avoid using their cooling on the hottest days? Large-scale grid defection would be a disaster for those least able to afford distributed energy, such as the elderly reliant on pensions, renters, the unemployed, and single income families.

So is disruption a useful theory for distributed power? It’s clear that distribution network providers in Australia are concerned about being left with a stranded business model. They want to be part of any prospective developments, regardless of whichever way it turns. I think the concept of low-end disruption applies to distributed power, but strictly only to the business model and not the physical assets. Regardless of the business models and tariff reforms that emerge, the physical infrastructure of networks needs to be maintained.

A few thoughts on the electricity ‘death spiral’

One of Germany’s exports has been the idea of ‘energy democratisation’ – the Energiewende is not just a transition to renewable energy but a switch to energy democracy. But was does energy democracy really mean and is it a useful model for thinking about future energy supply?

The thing that makes the democracy idea interesting in Australian is that there is little precedent for it in relation to the provision of public goods. Australians usually express a preference for socialised public services – indeed, Australia’s national health care system (Medicare) is so valued that even minor tweaks are strongly contested – think of the Abbott Government’s disastrous $7 plan for a co-payment in 2014.

We expect socialised education and urban transit. Even corporatised and privatised entities are expected to be firmly constrained by government regulation. It’s easier for a household to ‘democratise’ their water supply and go ‘off-grid’, yet nobody campaigns for ‘the suburban democratisation of water’ – water rates have gone up a lot in Australia as well. What about an exodus from schools in favour of home schooling? Environmental NGOs usually campaign for socialised services – better to invest in trains and urban transit rather than support ‘democratised transit’ (i.e. private cars).

Furthermore, the strong trend is towards outsourcing of home services – think lawn mowing, home cleaning, meal preparation (i.e. restaurants), painting, organised sport and play for children, car repair, etc. In the 1980s, I wouldn’t have dreamed of sending my car to a workshop for repair – who does their own workshop repairs nowadays? Yet futurists talk of bringing our energy production in-house – is this really a one-way trend?


Technology is shifting and it’s easily possible to go off-grid. But what about inverter, panel and battery failures that might take a week to repair? And battery replacement after 8 to 10 years. If the house is sold will the new owners want the responsibility of off-grid, and realise they’re going to need new batteries next year? Will it become the new normal to buy a generator from Bunnings for back up? What about strata title?

My sense is that there is a demographic that loves tinkering and the idea of off-grid. It’s a great idea for those of us who like doing things for ourselves and expect a little inconvenience sometimes. But I’m not convinced that most people will take the leap once they realise that it’s a one-way journey – once the capital is invested, it’s an expensive exercise if you change your mind. What if you want a reverse cycle heater but the system’s not big enough? Go back to the installer – ‘oh sorry, we supplied what you asked for.’

The so-called ‘death spiral’ of electricity utilities is predicated on the idea that the economics of disconnecting will drive an exodus from the grid. I would argue that a stronger incentive will be the reverse – it’s worth paying more to stay on-grid to avoid the hassles of managing your own power supply. What’s the equilibrium value of on-grid versus off-grid? I wish I knew. We live in interesting times.

Energy technologies and the Silicon Valley mindset

Horace Dediu identified the twin personalities of Google – Google A solves problems of humanity and Google B solves problems for advertisers. Google A is known for their moonshot programs and optimism, Google B pays the bills. So it was with interest when Ross Koningstein and David Fork announced the termination of Google’s RE<C program in 2011. Google’s aspirational goal was to produce a gigawatt of renewable power more cheaply than a coal-fired plant could, and do it in ‘years’, not decades.

Although Google had the best of intentions, the RE<C program represented the underlying problem of applying a Silicon Valley mindset to the energy and climate problem. Carey King summed it up neatly, noting that Google brings a mindset that is “used to solving some technological problem quickly, selling the company or idea to a larger company, and then moving on to the next great app.” To be sure, technology is going to help, but the dilemma is that energy and climate are not fundamentally ‘technological problems to be solved’. 

Why have renewables, particularly solar PV, become so connected with the Silicon Valley mindset? As a starting point, both share the element silicon and have parallel developments. 

Connecting ICT with solar PV
ICT Solar PV
Silicon Silicon transistors Crystalline silicon PV wafers
Miniaturization Greater transistor density Thinner silicon wafers
Performance Faster clock speed Higher efficiency
ICT Online media, shopping, banking, Uber Smart grids, vehicle-to-grid integration (V2G)
Moore’s law Doubling of performance every 2 years Declining cost of solar systems

But the question is – how close is the analogy really? In a recent paper, I explored the role of ICT, energy and resource decoupling, and concluded that –

The deepening of the service economy towards the Infotronics phase should be seen partly as a consequence of sufficient energy supply and productive primary and secondary sectors. ICT is enabling productivity gains and new business models, but does not significantly weaken the demand for energy services, and therefore does not enable strong decoupling.

In other words, it has been the high productivity of the primary and energy sectors that has enabled the advanced economies to progress towards an ICT-based service economy – not the other way around. ICT has not meaningfully diminished energy or material consumption when measured at a global scale. Energy and material use is a function of human wants, needs and income – this includes buildings, transport, food, health care etc. The relative composition of economies and ICT does not fundamentally alter the  demand for for these services.


The relevance of this is that ICT is operating in the information paradigm. The paradigm is governed by a different set of principles. Change a few lines of code and immediately improve an app. But trying to get a race car get around a track a couple of seconds quicker might require millions of dollars of investment. Energy supply technologies are governed by the familiar physical laws we are used to. No amount of ICT can improve the performance of a solar panel at night time.

A Framework for Incorporating EROI into Electrical Storage

The fundamental problem with a transition to renewable energy is that modern society has been structured around demand-based power flows. Any quantity of power is available at any time – the only limit is the circuit breaker in your mains connection. But the major scalable and affordable renewable power sources are wind and solar PV, both of which are intermittent. We could add biomass, but the degree to which biomass and biofuels can be scaled is limited and anyway, their use is contested. Until now, intermittency has been manageable because the variability generated by the modest proportion of RE is readily accommodated with the legacy infrastructure. Regions with a high penetration of VRE, including Denmark and South Australia, have access to virtual batteries in the form of interconnectors to larger grids. The question is – how do we deal with intermittency as legacy infrastructure is retired and wind and solar have to take on a greater role?

The solution is of course storage, but what sort of storage, how much, and what are the biophysical limits of storage. EROI is really about exploring the biophysical limits of storage rather than business models and markets. It may be economic to install a Tesla Powerwall based on feed-in and retail tariffs, but tariff-induced economics may not reflect the value of storage at a societal level.

In recent years, there have been important contributions to applying EROI to storage, however, there remains uncertainty as to how to apply these metrics to practical systems to derive useful or predictive information. I propose a methodology that assesses the EROI of the variable renewable energy and storage as a system, relative to the quantity of conventional generation capacity that is displaced.

A justification for focusing on substitution of capacity is the German Energiewende. Between the starting point of the EEG in 2003 and 2014, total installed power generation capacity grew by 51%, although total annual generation was virtually unchanged. The emission intensity for electricity declined from 610 to 559 grams CO2/kWh over the period. Unlike historical energy transitions, such as wood to coal or coal to oil, we simply haven’t seen the substitution of legacy infrastructure and productivity gains.

greenhouse emissions graph
Source: Federal Ministry for the Nature, Environment, Nature Conservation, Building and Nuclear Safety, 2015, 
Facts, Trends and Incentives for German Climate Policy


In a new paper in BioPhysical Economics and Resource Quality I explore these issues, with the aim of introducing a framework for further exploration. The most important outcome is the shape and behaviour of the embodied energy and marginal embodied energy curves. The first units of storage and VRE are the least energetically expensive. Using a simulation for the Texas ERCOT grid, I find that it is 4 to 41 times more energetically expensive to displace a gigawatt of generation capacity at near 100% RE than at low penetration RE. Geographic and technology diversity improve these numbers. Unlike conventional generation, which has access to essentially unlimited ‘stored sunlight‘ or nucleosynthesis in the form of fuels, VRE is handicapped by the energetic demands of surplus VRE and storage.




Leif Wenar- Blood Oil

From 1807, Britain tried to suppress the Atlantic slave trade until eventual success its 1867. In the late eighteenth century nearly everyone of influence or power in Britain had a direct or indirect interest in the slave trade. In 1805–1806 the value of British West Indian sugar production reliant on slaves equaled about 4% of Britain’s national income. Kaufmann and Pape suggest that Britain voluntarily forgoed 4% of national income, essentially for 60 years. By modern standards, the suggestion of embarking on a moral campaign and being willing to forgo such a magnitude of GDP seems almost unthinkable. Yet this is precisely the sort of brave thinking that Leif Wenar brings in his book Blood Oil: Tyrants, Violence, and the Rules that Run the World. Wenar’s is a thought provoking exploration of the morality of oil supply. He introduces the expression ‘Might Makes Right’, meaning that whoever possesses the oil gets to profit from it. Wenar draws the analogy of a criminal gang stealing a fuel delivery truck, driving it down the road, and selling the fuel. Of course this seems absurd. But Wenar claims that this is essentially what is happening in the realpolitik of oil producing countries – we are all responsible for buying stolen goods and indirectly supporting dictatorships – think Putin, and Wahhabism in Saudi Arabia. Wenar is not anti-oil per se, but advocates a form of Clean Trade in oil as a moral solution.

Wenar makes a compelling moral argument, but I think he vastly understates the value of oil in modern society. This is the crux of the problem and why the US embarked on two grand bargains in the 1970s after US oil supply peaked – the US sought to control oil supply by undermining oil-producing democracies with oil-to-arms deals; and entered into private banking agreements to ensure that petrodollars flowed back to the US and regions that ensured US influence.


The thing I find interesting is that many environmentalists advocate the decommissioning of dams, abandonment of nuclear, coal divestment, bans on gas fracking, but nobody willingly ‘gives up’ oil. If humanity could fix its global energy problems by forgoing 4% of national income, I suspect a sizable minority would be willing to participate. But it’s not that simple. Oil is the master energy source. Such is its importance that it has a role as a quasi-monetary commodity. Economic prosperity is unfortunately going to remain tied to oil for the foreseeable future. Are electric vehicles the answer?

The Maximum Power Principle

Lotka and Odum

In 1922, Lotka proposed a ‘law of maximum energy’ for biological systems. He reasoned that what was most important to the survival of an organism was a large energetic output in the form of growth, reproduction, and maintenance. Organisms with a high output relative to their size should outcompete other species.

In 1955, Odum and Pinkerton built on Lotka’s work with the ‘maximum power principle’, stating that ‘systems perform at an optimum efficiency for maximum power output, which is always less than the maximum efficiency.

The electrical Ohm’s Law provides a way of thinking about the relationship between maximum power and maximum efficiency. In electronic devices, the output resistance of the power source should match the input resistance of the load to maximise the power throughput for maximum power. However, the point of optimised power does not match the point of optimised efficiency. In the case of loudspeakers, for example, the maximum sound output is achieved when the speakers match the impedance (AC resistance) of the amplifier. At this point, half the energy is dissipated in the speakers and half in the amplifier. Speakers with a much higher impedance will improve the system efficiency but with less volume, requiring a larger amplifier to reproduce an equivalent volume (negative feedback in amplifiers actually makes this a bit more complicated). The same applies to antennas, that require the antenna impedance to match the transmitter at the designated frequencies.


In the biological realm, Hall recounts the relationship between a tree’s leaf area index (LAI) and the energy capture. The highest efficiency is achieved with a relatively low LAI since the topmost leaves capture the most sunlight, and each leaf is energetically expensive to maintain. But the usefulness of the high efficiency is offset by the limited leaf area and therefore total energy capture; an efficient plant would be short and outcompeted in a forest.  But there is a limit to which a plant can grow before the marginal gains of additional leaf area contribute to energy capture.

Perhaps the most sobering outcome of the Maximum Power Principle is thinking about the role of humanity, and what this means for human appropriation of net primary production of biomass for liquid and other fuels. William Rees cites Lotka’s maximum power principle in posing whether humans are unsustainable by nature, noting that –

by virtue of cumulative knowledge and technology, homo sapiens has become, directly or indirectly, the dominant macro-consumer in all major terrestrial and accessible marine ecosystems on the planet.

Role of energy efficiency and maximising energy throughput

The operation of electrical generators provides an example of the economic trade-off between maximising power and maximising efficiency; the revenue of an electrical generator depends on energy throughput, but the efficiency defines the fuel cost per unit of electricity. An efficient plant will increase electricity output for a given quantity fuel, but beyond the optimal power/efficiency, the additional gains in output do not justify the additional cost of improved efficiency. These trade-offs are routine in engineering practice.

The limits of energy efficiency as a goal in itself

The concept of ‘energy efficiency’ is really a human construct that is often useful to conceptualize the performance of energy systems. But energy efficiency targets, in themselves, can be counterproductive. The Australian building regulations typify this problem. The Building Code specifies deemed-to-satisfy building requirements for thermal performance, such as insulation R-value, and double glazing. But the Code merely institutionalises energy efficiency as a goal in itself, rather than per-capita energy consumption. This is because larger homes generate a better thermal efficiency score than equivalent smaller homes, since geometrically, larger homes gain proportionally more interior space relative to exterior fabric area (better know as Galileo’s square-cube Law). But rating tools do not penalise larger homes even though it is obvious that they consume more energy, nor account for functional use or the number of occupants. This leads to the perverse outcome that the Code favours homes that consume more energy, but do so ‘more efficiently’.