The Future Price of Energy: $/bbl or $/MWh?

“In the face of overwhelming odds, I’m left with only one option, I’m gonna have to science the shit out of this.” – Matt Damon as NASA scientist Mark Watney, in the motion picture, The Martian.

This blog is about energy, and its transition – but what transition exactly?

In one sense, as mentioned in several posts, I believe it is about the transition from primary energy based on extraction to primary energy based on engineering and technology. And the implications this has for all energy-focused companies, investors, workers, consumers and beyond.

The short-hand for this transition is usually summarized as the move from fossil fuels to lower-carbon alternatives via gas as a bridge. But that is too narrow a focus in terms of the scale of the transformation a move from extraction-based energy to engineered energy may bring.

To start with, lets relook at what we mean by energy supply, and energy demand in the first place.

Today, the majority of the oil and gas commentary focuses on, in the words of Diego Parilla and Daniel Lacalle from their work The Energy World is Flat, “the battle of supply”. They characterise this as the battle between oil producing states and technologies, and when oil prices remain low, as now, project postponement and cancellation leads to loss of production capacity. Combined with robust assessments of future demand, this leads to higher pricing which the existing cartels then attempt to sustain.

But there is also a “battle for demand”. By this Parilla and Lacalle mean the growing competition between fuels for transportation and power – with oil and oil products competing with gas, wind, solar, smart electricity and other technologies. Here they believe is the real transformation of the energy market, gradually and technically, whilst we mostly focus on the more gripping supply conflicts.

Their conclusion is that ultimately engineering efficiencies and relentless improvements will create a range of new energy sources and networks that will displace and substitute coal, oil and gas.

The final overlay to this picture are the actions of governments and new energy companies who are using the momentum of international agreements such as COP21, and major national ones such as the US Clean Power Plan or Chinese FYP to invest significantly in alternative energy sources.

Parilla and Lacalle’s argument regarding the battle for demand seems to have increasing evidence.

The unit costs of solar and wind power has dropped dramatically in the past few decades following a typical technology cost curve. Solar costs for example have dropped at a rate of 10% per year since 1977 from over $75/watt to 30c/watt. Latest analysis by Bloomberg New Energy Finance indicates that on a Levelised Cost of Electricity Basis (LCOE) both energy types can compete with CCGT and Coal, at around $80-100/MWh pre-subsidy, and with declining costs. Coal and CCGT have an LCOE of around $60-100/MWh, which rises when oil prices increase. Efficiency factors and carbon tax assumptions are included in the analysis.

LCOE 1 lazard

As a result, in 2015 China became the largest user of solar power in the world, overtaking Germany. It now intends to install over 100GW of electricity power capacity from solar alone by 2020. That is equivalent to 10% of its installed base or the equivalent of 100 nuclear reactors. It has further targeted that 15% of its total energy supply to be sourced from wind and solar by 2020 in its latest Five Year Plan.

Globally, estimates of installed power capacity with wind and solar by 2020 are calculated to be over 20% of total global base, or over 1000GW. Current additions of both technologies are increasing at ca 120GW pa, mainly in China and the US, the equivalent of adding 1 new nuclear reactor capacity every 3 days.

Overall, 2015 was the first year in which global investment in technology-based power capacity ($330billion) was higher than that for fossil fuels. Future growth is not guaranteed at this pace given lower oil prices. But as Lacalle and Parilla point out, as these energy technologies reach commercial scale, their costs drop rapidly due to engineering learning curves. In solar PV, for example, Swanson’s Law suggests that costs fall by 20% every time solar panel volume doubles, currently every 2-4 years. Meanwhile, much current fossil fuel supply is extractive, and prone to cost increases over time. Cost parity for energy generation between technical energy and extractive energy has effectively been reached, but the cost of technology-based energy is projected to steadily reduce.

This view of rapidly increasing technology-based energy has many challengers. Oil industry assessments show conventional energy demand to increase in line with global GDP and modernization of emerging economies. Depending on the analyst this ranges from about 1-2% growth pa, and leads to detailed projections of for coal, crude oil and gas demand, using the detailed historical data trends available. These three forms of energy account for about 80% of what we consume today: coal and gas largely for power and heat, while oil retains a monopoly on road and air transportation.

The impact of technology-based energy alternatives, energy efficiency initiatives and energy policies pursued by governments are recognized in these models, but their influence over the next 2-3 decades is seen as limited, uncertain and thus relatively low. This leads to a more linear model predicting 75% of primary energy still based on coal, oil and gas by 2035, with wind and solar energy rising only slightly from 3% to about 9%.

In a sense then, the transition we are looking at is actually one based solely on time: more specifically, what is the pace that the emerging and maturing technologies of wind and solar can substitute for or displace existing fossil fuel equivalents? Gradually or rapidly, in an orderly or disruptive process?

In the industry gradualist view, transition happens, but it is a story of decades. Lower nominal fossil fuel prices dampen demand for alternatives, and current infrastructure proves resilient and effective from decades of use and consumer interaction. Governments continue to support existing industries due to infrastructure sunk costs and job protection concerns.

In the more revolutionary view, the technical efficiency of new energy forms rapidly improve in cost and utility as they reach scale tipping points, and their deployment becomes simpler and quicker. Governments drive the new fuel adoption due to a desire for energy security or independence (buying and deploying the technology rather than importing it at variable prices), and to meet international commitments to lower carbon energy management.

Overall the consumer, in theory, wins as more fuel types compete for their energy consumption.

Parilla and Lacalle put these two narratives in a stark form: is the last barrel of oil worth milions of dollars, or worth nothing? Is there a total replacement of fossil fuels, or is there a century-long accommodation of multiple fuels? It’s the ultimate (and correct) future oil price prediction – somewhere between 0 and 100 million dollars per barrel.

The timing and shape of the transition, however, is more important than using the supply / demand forecasts to predict price. That’s because the form it takes will have major consequences along the way for companies, governments and all the people directly and indirectly involved.

What might be the broad outline of that transition, and the key issues to look for? I sketch an outline of some key areas below that all interested parties ought to consider and potentially take action on, whether incumbent or new entrant.

S curves vs linear models: in a time of transition, it will pay to scrutinize what the previous pace of growth (or decline) is actually telling us. Most new and mature technologies in any industry tend to follow an S curve: slow initial adoption and shallow growth until some transition point, then a rapid steep expansion, followed by decelerating decline before a plateau. If a firm invests too heavily before widespread growth, or when demand is in decline, it can accrue too much debt and fail. Similarly, if a company misses the expansion phase, high profitability and market share may be gone for good. Forecasts of industry demand and competitor growth can be substantially impacted by how the S curve actually plays out over time: previous limited growth can suddenly take off, or long-term growth can decelerate sharply.

S curve solar v1

A key question, therefore, for oil and technology energy companies is whether the new fuels are entering the sharp growth phase of the S curve, from emerging technologies to commercial scale capability. This depends not only on the technology itself, but government policies, and competitor reactions. If technology-based energy has begun the high growth phase, then the prevailing predictions of gradual substitution could significantly underestimate their expansion, and their impact on conventional fuel growth – especially in emerging markets.

Demand is often concentrated: a smooth demand curve can hide some significant issues. Whilst growth in general may occur, demand is almost always concentrated in a few critical regions or customers. Many conventional fuel projections assume strong rates of energy demand from China and other emerging markets for the next few decades. However, its clear that China is investing heavily in non-fossil fuel alternate technologies, and currently has the largest installed base of wind and solar energy in the world. That needs to be factored into global demand estimates for conventional fuels and technology-based fuels more accurately.

Policy action (state and competitors): significant attention needs to be paid to state-based policy-based issues: preferential subsidies (CPP and ITC in the US for example), COP21 development initiatives, and national strategies eg the latest Chinese FYP. Whilst subsidies and support eventually end, they can push new technologies into the commercial phase, where scale economics can take over. In general, international policies are tending strongly (though not always) toward support for technology-based growth, as an increasing amount of states see them as a less volatile and more controllable form of energy, than the OPEC-influenced extraction of fossil fuels. The battle for supply is likely to heighten the concerns of energy costs being managed by distant erratic states. Chinese action to retain home-grown coal and invest in energy technology is a case in point of looking to engineer energy self-sufficiency. And once established, energy technologies developed by China, eg ultra-high voltage electricity, may then be exported profitably and used by others.

Policy action by competitors is also important. High oil prices between 2005-2014 allowed the emergence of high cost shale oil and gas to enter the market. Once established, the technology basis of shale fracking allowed large engineering cost gains to occur – 30% reductions pa by some estimates. Within five years break-even costs had roughly halved, allowing shale oil to be more robust than expected to the subsequent oil price collapse.

As noted generally, further oil price spikes may hasten the introduction of alternative energy technologies such as shale and solar, by providing high-cost and volatile competition. When shale and solar were only emergent technologies with no scale or reliability, high price competition did not precipitate their widespread use. Now that they are both being deployed rapidly, a high price policy move by incumbent competition eg OPEC, may accelerate their adoption and gain further support from national governments.

Engineered energy economies of scale : As newer industrial-scale forms of energy such as solar, wind, smart-grid and shale oil and gas, grow in scale, their costs drop dramatically (Swanson’s law and many others). This has been happening quietly and at the margins of the large-scale energy industry, but the accumulated cost reduction make them competitive with fossil fuels, and under adoption by large energy users such as China and the US, who use 40% of global energy between them.

This has already occurred with the manufacturing process of shale oil and gas via fracking as discussed. Shale oil and gas growth may be at the plateau phase of the S curve now, but it’s a good example of how an emergent technology can provide significant fuel substitution by leveraging basic engineering cost curves.

Intermittent energy versus base-load: one major argument against the widespread adoption of wind and solar anytime soon is their fundamental intermittent nature – for example during still evenings and cloudy days. But this is only one more engineering problem, and is expected to find numerous solutions as commercial adoption increases. Flexible storage and intermittency algorithms are being increasingly used to not just deal with the issue, but drive down the costs of dealing with it. In any analysis of the growth in solar and wind energy, a reasonable assumption is that cost-effective engineering solutions to intermittency will develop alongside the overall technology.

Sporadic energy price spikes due to oil and gas supply issues in fact may make the technical intermittency challenges with wind and solar seem more manageable.

Implications for Transition

We started out by wondering what transition we are discussing. Analysis of supply and demand reinforces the view that we are discussing the shift from extraction-based exported energy, to the development of engineered, more localized technology-based energy.

This is an important point. As noted, it is too narrow a viewpoint to say that the transition is about a move to low-carbon or renewables. Why? Because that would potentially misread actions of certain players, and the likely form of how the transition will take place.

For example, the US state with the most rapid uptake of wind and solar is in fact Texas, home to most large private US oil and gas companies. Texas is traditionally skeptical toward low CO2 policies, even climate change, but the rapidly reducing costs of wind and solar power, and the large natural endowment of both energy sources in Texas, means they have adopted it early, and have subsidies and state backing from ERCOT to drive its increasing adoption.

The benefits of potentially abundant lower cost energy for Texas is therefore mainly an agnostic engineering and technology discussion, rather than a political one about lower carbon.

Likewise, China’s position in the vanguard of wind and solar power adoption would not be predicted from previous actions regarding international CO2 emission policies.

Once we’ve defined the transition more broadly, we’re left to ponder how long it will take, and what form.

We reviewed the base transition model using S curves, concentrated demand, policy priorities and the cost curves of engineered fuels.

The implications are that the base scenarios may underplay the impact of renewable growth in two ways: first, the actual growth of renewable technology looks steeper than in base models in terms of cost reductions, investment, share of capacity. Secondly, some of the largest investors in these forms of energy – China, US and India – were assumed to have the strongest demand requirement for conventional fuels.

Looking to the future, developing accurate projections of future energy demand, and the fuels most likely to provide it, becomes a major issue for all energy players.

Winners and Losers?

At a time of transition, getting the market demand wrong can be serious. The S curve can be a cruel place, early or late in the cycle.

Over the last five years over a 100 solar energy firms have gone bankrupt or been acquired. Overinvestment and indebtedness have been common problems. But early in an industry’s life, this clean out is inevitable as subsidies and technologies are mistimed or misread. The emergence of powerful large-scale Chinese manufacturing firms has also forced out many firms on a unit cost basis.

At the other end of the cycle, three of the largest four US coal producers have gone bankrupt in the last 12 months, with the largest of all Peabody Energy claiming protection this month. Chinese demand for coal evaporated last year as it switched fuels to more gas and alternate energy technologies, and reduced investment in heavy industry. Peabody misread the pace of growth in the latter part of the S curve, assuming a previous expansion of 10% pa would continue. It didn’t, in fact demand reduced.

To avoid such distress, incumbent energy firms should therefore adopt a wide watching brief on the development in technology-based energy growth (including shale fuels)– over-reliance on the standard projections of fossil fuel increase need to be challenged with some stress case scenarios that model faster technology-based fuel adoption, especially in major new markets such as China.

Coal, gas and oil also have increasingly distinct demand markets. Coal can be replaced by other fossil fuels as well as new technologies, and with high demand growth predicated on China it is especially vulnerable to the trends of technology and policy discussed here.

Gas has a complex demand future, with growth coming from substitution as well as economic development. However, expectations of Chinese, US and Indian demand need stress-tested to ensure investment opportunities are robust.

Oil appears to have a clearer road ahead as rapid changes to gasoline substitution are less obvious than in the power sectors. However, once again Chinese and Indian demand, projections of electric vehicle penetration, and policy initiatives on fuel efficiency need to be part of a robust stress test case.

Overall, any large-scale fixed investment in production whose economics are sensitive to pricing and demand in the out-years eg beyond 2030 should have robust margins under stress scenarios, and alternative faster more adaptable ventures considered if they do not. Opportunities predicated on high commodity prices, and long-term delivery horizons, need to avoid optimistic demand scenarios.

For renewable energy firms, the next few years will be a time of continued growth, merger, failure and consolidation as the steep part of the S curve kicks in. Expect to see a lot of players disappear, and some large national champions and international players emerge. Disruptive new technologies in materials, storage and software may also be decisive.

2016 -2020 will therefore be a major era in terms of energy transformation, as major importers of energy move toward greater adoption of alternative energy, and oil producers and exporters adjust strategies to accommodate a potentially more rapid rise of technology-based competitors.

Irrespective of how the transition occurs, and supply and demand respond, the oil price can’t be predicted in the short run. There are too many variables.

But the cost of alternate energy technologies, based on engineering curves, will continue to fall relentlessly. That is the key difference with regard to extraction costs, and it needs to be factored into all scenarios.

This major transition is the one we’ll therefore follow over the coming months and years, along with the actions and reactions of the various key parties involved.

More to come.