By Goehring & Rozecwajg
We are making a terrible mistake when it comes to future energy policies. In previous letters, we have touched upon the challenges of the so-called “energy transition.” Today, we will explain in detail the imminent harm awaiting investors and policy makers who fail to acknowledge certain realities. Every green energy proposal we have examined relies on the trifecta of wind, solar, and electric vehicles combined with various battery technologies. In recent months, a renewed “hydrogen mania” has broken out as well, which adds a fourth leg to the green energy stool. Unfortunately, based upon our extensive research, these plans, including the current hydrogen craze, are bound to at best severely disappoint and, at worst, outright fail in what they attempt to accomplish.
Not only will these proposals cost trillions, but policy experts and investors will soon realize that none of these options will address the very CO2 problem they are designed to solve. Currently, there are 410 parts per million of CO2 in the atmosphere. By keeping the concentration below 450 ppm, it is widely believed warming trends can be curbed. This would require lowering annual CO2 emission nearly 60% from 34 billion tonnes today to 15 billion tonnes by 2040. Most energy transition plans advocate the widespread adoption of renewable energy and electric vehicles to drive down carbon emissions. Unfortunately, our research suggests these plans have little hope in solving the CO2 problem.
Vaclav Smil, Distinguished Professor Emeritus in the Faculty of Environment at the University of Manitoba, is the best energy scholar we have ever read, in our opinion. In his Energy Transitions, he notes that historically a new energy source takes between 40–60 years to gain significant market share. The current proposals assume wind and solar will make comparable gains in only 20 years. Ambitious plans often carry ambitious budgets, and the green energy transition is no exception. Using extremely aggressive cost saving assumptions, a widespread move to renewable power is expected to cost $70 tr over 20 years, nearly $50 tr more than if we stayed on the current trajectory. Unfortunately, our research tells us this additional spending will not even come close to generating the expected reduction in global carbon.
The sums involved are monumental, and much of it could fall into the category of “malinvestment” with disastrous results. The further down the current path we go, the less likely we will be able to change course later. A decade ago, a series of failed promises and bankruptcies plagued the battery industry, making it nearly impossible for subsequent ventures to find financing and move forward. We worry the same could occur on a much larger scale if tens of trillions of “green” investments are eventually written off.
This essay will serve as a jumping off point outlining our research on various energy transition topics. We intend to follow this up with a series of essays, podcasts, and videos going into greater detail on each topic we introduce here. Please visit our website (gorozen.com) for ongoing updates.
This discussion comes at a critical moment. Over the last 12 months, green energy momentum has exploded. Investor euphoria has now reached new heights bordering on mania. Renewable investments (as measured by the RENIXX, ICLN and QCLN ETFs) have advanced between two and three-fold since the start of 2020. Teslaisupten-fold and sports a market capitalization of $800 bn. Hydrogen stocks have done even better: Plug Power is up 2000% since January 2020 resulting in a market capitalization of $31 bn (or 100 times revenue). Investors have taken notice and poured huge sums of capital into green ETFs. Shares outstanding of the four most prominent clean energy ETFs are all up between three-and six-fold. Traditional energy has been on the other side of this trade. Over the same period, shares outstanding of the XOP ETF (designed to track S&P exploration and production stocks) are down 25%.
Stretched valuations leave investors vulnerable to any setback or delay in the green energy transition. The ICLN ETF holdings currently trade at 70x earnings, 6x sales and 6.25x book value, suggesting dramatic growth is already priced in. What would happen if the energy transition proved more challenging than anticipated?
Green energy Special Purpose Acquisition Vehicles (SPACs) are also a troubling sign. Green SPACs have raised $40 bn in 2020 alone with a mandate to acquire as-of-yet unidentified clean energy assets. Oil and gas exploration and production companies on the other hand raised only $5.2 bn in 2020 to develop their existing proven asset bases. Given how challenging clean energy product development can be, we fear the bulk of these green SPACs will likely end up being written off entirely.
Adding to the current green momentum, several reputable agencies released reports in 2020 predicting an acceleration in renewable adoption. The International Energy Agency (IEA) released its World Energy Outlook in October 2020. In their report, they lay out a “Sustainable Development Scenario” in which global oil demand peaked in 2019 and will fall by 34% by 2040. Countless news articles picked up on the “peak demand” prediction and pointed to the strength in renewable equities as “proof ” the era of hydrocarbons is now ending. The combination of stock performance and news articles has created a dramatic feedback loop, pushing prices up even further and in turn giving more credibility to the energy transition thesis.
The IEA’s World Energy Outlook generated a lot of attention when it was released. Upon closer inspection, we believe the report makes two critical mistakes. While CO2 emissions fall to 15 bn tonnes by 2040, the drivers of the reduction seem questionable. For emissions to fall 60% over the next twenty years, the IEA assumes per capita energy demand will fall by 25% while CO2 per unit of energy drops by 50%, offset by population growth of 20%. Together, these factors equate to a 60% reduction in total carbon emissions and keep atmospheric CO2 to within 450 ppm. Unfortunately, neither energy intensity nor carbon intensity are likely to fall anywhere near the amount predicted by the IEA.
Few people are aware that most climate proposals are based on consuming 25% less energy. According to the BP statistical review, there has not been a single 20-year period since their data begins in 1965 where per capita demand has fallen by more than 0.1%, making this assumption untenable.
While our research suggests today’s green energy mania will not reduce carbon emissions (and curtail global warming) as expected, we believe there is a more straightforward, feasible solution. Nuclear power is the only technology that can provide reliable carbon-free baseload power. Any proposal seeking to seriously address carbon emissions must heavily incorporate nuclear electricity generation. Unfortunately, none of the current proposals include a material nuclear contribution. There is some reason to be optimistic however: the Biden administration has acknowledged the need for nuclear power in combating climate change and may signal an important impending change in policy outlook.
We have invested in the global natural resource industry for 30 years. At present, approximately half our investments are in oil and gas related equities with another 15% in uranium producers. While some may argue this makes us biased, we disagree. Nothing in our investment mandate requires we be invested in any particular natural resource subsector. If our research pointed to the end of the hydrocarbon era, we would turn our attention to other areas such as battery metals and renewable energy. Instead, we continue to see a future for hydrocarbons, particularly as a transportation fuel. While carbon emissions should be addressed, wind, solar, hydrogen fuel cells and battery powered electric vehicles are not the solution. Our investments are consistent with our research and not the other way around.
We have spent years studying energy trends in emerging markets. Over that time, non-OECD countries have gone from consuming 40% of all primary energy to 60%. As an emerging market gets richer, it reaches a tipping point and starts to consume more energy. Once an economy is developed, it reaches a saturation point and energy demand moderates. Since 2000, non-OECD countries have grown their primary energy demand per capita by 65% compared with a reduction of 10% in the OECD world. Even following two decades of strong growth, non-OECD demand is still 70% below OECD levels suggesting more growth is yet to come. Non-OECD real GDP per capita is expected to double over the next 20 years, suggesting these trends will continue. Instead, the IEA projects emerging market per capita energy demand will fall by 20%. Simply put, this is impossible. Over the last two decades, real GDP doubled, and energy demand rose 60%. Even if this relationship is cut in half, the next doubling on GDP would result in energy demand growing by 30% by 2040, not falling by 20%.
OECD per capita demand is projected to fall by 30% or three times the rate of the last two decades. While there is more “discretionary” energy demand in affluent countries, a 30% drop would take per capita energy use to 1955 levels—something we do not believe is likely. If OECD per capita demand falls by 10% instead of 30% and non-OECD demand grows by 30% instead of falling by 20%, per capita primary energy will grow by 12% instead of falling by 25%, leaving total demand 50% higher than the IEA expects.
Unfortunately, carbon intensity per unit of energy consumed is unlikely to fall by the 50% assumed in the IEA’s proposal either. It is widely believed the reduction will be based upon the widespread adoption of wind, solar, electric vehicles, and hydrogen fuel cells, but our research suggests these technologies will fail to deliver the expected results. There are several real-world examples that confirm our suspicions. Over the past two decades, Germany has aggressively pursued its renewable-centric “Energiewende” plan, taking renewables from 2% of all German electricity to nearly 40%—by far the most aggressive renewable push in the world. Over the same period, carbon emissions per unit of energy fell by only 12%. Not only is this reduction a far cry from the projected 50% reduction in most energy transition plans, but it is also no better than those countries that did not adopt a renewable energy push. Between 2000 and 2019, the US and France went from 1% renewable electricity to 10%, or less than one-third of Germany’s penetration. Despite this lack of renewable adoption, US carbon intensity fell by 13% while France’s intensity fell by 10%, ahead of and only slightly behind Germany, respectively.
Electric vehicles will likely not deliver the necessary carbon reduction either. In Norway, electric vehicle sales have gone from zero to nearly 60% penetration between 2010 and 2019. Despite such a dramatic shift away from oil, Norway’s carbon intensity has declined by 10% compared with 11% in the US where EVs remain less than 2% of all vehicle sales.
Although these results might seem improbable, the explanation has to do with the physical limitations of the various technologies.
Wind and solar are extremely inefficient generators of electricity due to their low energy density and their intermittency. In summary, a solar panel likely only dispatches between 12 and 20% of its rated capacity due to the intermittency of sunshine. A wind turbine is somewhat better, but still less than 25%. As a result, excess capacity must be built to generate the necessary electricity. Moreover, the power must be “buffered” by a storage system to smooth out the inherent variability coming from both short-term dislocations (clouds and periods of calm), as well as different patterns between day and night.
Low load factors and “buffering” of intermittency results in poor “energy return on energy invested” (EROEI). As much as 25–60% of the energy generated in a renewable system is consumed internally, compared with 3% for a modern gas plant.
In his excellent work, Energy and Civilization, Professor Smil describes society’s ongoing adoption of new technologies. A theme that runs through his work is how every new major “prime mover” is able to convert energy into useful work more efficiently than what came before. According to our models, wind and solar would mark the first time we have seen a widespread shift into a much less efficient source of energy conversion. It has never happened in the past, and the only way it can happen in the future is if governments subsidize wind and solar (as is being done right now), or outlaw old hydrocarbon-based technologies—now being threatened. In either case (subsidy or outlaw), government intervention is the only way people would likely adopt new energy conversion technologies with inferior efficiencies.
It is difficult to forecast the impact of transitioning from a system where 3% of all energy is consumed internally to one where more than a third is lost. To the extent solar and wind facilities do not achieve their target lifespans (and there is ample evidence to suggest this is happening), the results will be even worse.
Solar and wind’s low EROEI also impacts their carbon emissions. While billed as being “carbon free,” solar and wind generate CO2 during their construction and maintenance. To the extent overbuilding and battery backup is required to allow for baseload power, CO2 emissions increase dramatically. This partially explains why German carbon intensity only fell by 12%, despite having among the highest renewable penetration in the world.
Electric vehicles also involve energy intensive lithium-ion batteries. Few realize how much energy is embedded in an electric vehicle before it is ever plugged in. Over the life of a typical EV, nearly 40% of the total energy goes into manufacturing the battery. The IEA expects electric vehicles will represent nearly 15% of total transportation energy by 2040. We calculate this equates to approximately 850 mm EVs and nearly 65 terawatt hours of batteries. This is a staggering amount considering global lithium-ion manufacturing capacity is currently less than 0.4 terawatt hours per year. These batteries will require an incredible 2 billion tonnes of oil equivalent to build.
Unfortunately, few people realize how energy intensive the “green transition” will be. As a result, much (if not all) of the carbon savings will be undone by generating the power in the first place. The IEA’s proposal assumes wind and solar make up nearly 50% of all electricity by 2040 and that some 850 mm electric vehicles will be on the road. These initiatives are expected to reduce CO2 by 55% or 18 bn tonnes per year. While this may sound impressive, simply moving away from coal towards much-cleaner natural gas would itself save nearly 14 bn tonnes of CO2 per year. When analyzed through this perspective, renewables would save an incremental 4 bn tonnes compared with the next cleanest option.
At the same time, an incredible amount of energy is required to build out the renewable capacity and manufacture the necessary batteries. A move toward gas would be much more energy efficient (given its high EROEI) and would not require batteries for either grid storage or automotive uses. We estimate the move toward renewables and EVs would generate nearly 45 billion tonnes of incremental CO2. Therefore, nearly 10 years of carbon “savings” would be spent on the energy transition itself. A battery is expected to last between 6 and 15 years depending on charging behavior while wind turbines have an expected life of 20 years and PV solar panels have a useful life of 25 years. At best, a huge amount of the expected carbon savings will be undone by the necessary manufacturing. At worst, the impact could be net detrimental.
Early wind turbines are failing at much higher rates than expected, suggesting a 20-year useful life may be too long. Similarly, PV solar facilities are noticing higher performance degradation than expected in their facilities. We will discuss both topics in depth in an upcoming podcast. Given the huge upfront energy needs of wind, solar and batteries, any performance disappointment could mean the difference between moderate carbon savings and net increases in carbon. We believe this partly explains why total carbon intensity has not fallen as much as expected in countries with large renewable mandates. Countries that have adopted natural gas, on the other hand, have seen better results while spending trillions less. Moving towards nuclear generated electricity is an even better option and would allow for the greatest carbon reduction while still saving trillions. We will explore this option at the end of this essay.
Clearly a major part of the problem is the energy-intensive lithium-ion battery used in EVs and for grid-level renewable storage. To address this issue, hydrogen fuel cell technology is once again being put forward as a possible solution. This marks the second investment mania in hydrogen fuel cells in 30 years. In the late 1990s, fuel cells went through an impressive bull market that saw Ballard rise 1400% over three years before crashing 99%. Even after its recent 600% advance it is still 70% below its prior peak. Unfortunately, many of the technical issues that led to the previous hydrogen bust remain.
It is important to realize that while the fuel cell does not need an energy-intensive battery, it is nevertheless an extremely inefficient technology. To make hydrogen, electricity is used to electrolyze water resulting in oxygen and hydrogen gas. The gas is then compressed or liquefied for transport to the end user. In the fuel cell, hydrogen is reformed back into water producing an electrical current used to power a motor. Not having to manufacture an energy-intensive battery saves on upfront energy dramatically.
Unfortunately, any savings are “spent” on the extremely poor overall energy efficiency of the system. Powering an electrolyzer to produce hydrogen gas loses upwards of 30% of the embedded energy. Compressing or liquefying the gas for transportation loses another 15% of the energy. Generating an electric current in the fuel cell loses yet another 30% of the contained energy. In total, we estimate that 70% of the electricity used to power the system is wasted. Even though there is no energy-intensive battery to manufacture, we believe hydrogen fuel cells requires more total energy to power a car than standard EVs despite their lithium-ion battery. When wind or solar is used to manufacture the hydrogen (i.e., “green” hydrogen), the overall energy efficiency become even worse. To achieve a reduction in net emissions requires the original electricity be nearly four times less carbon intensive to make up for the 70% energy lost in the system. Solar and wind powered “green hydrogen” vehicles would not generate any net carbon savings compared with gasoline or diesel and would likely be much worse.
As we discussed, moving away from coal toward natural gas would be a great first step in reducing carbon. We believe there is an even better solution that would reduce emissions much further. Any serious proposal to reduce the carbon intensity of energy needs to have several characteristics: it must be very energy efficient as measured by EROEI; it must be able to supply baseload power and avoid intermittency; it must be scalable to meet ongoing global energy demand growth; and it must be low-carbon or carbon-free. The only source that meets these criteria is nuclear fission.
A modern reactor generates electricity with an EROEI of nearly 100 compared with 30 for gas and 1–4 for renewable. As a result, only 1% of the generate electricity is consumed internally compared with 3% for gas and 25–60% for renewable energy.
If per capita energy demand rises by 12% (instead of falling by 25%), carbon intensity would need to fall by two-thirds to keep total emissions to within 15 tonnes and limit atmospheric CO2 to within 450 ppm. Only a combination of nuclear energy and efficient natural gas can hope to get close.
In future podcasts, we will outline a possible energy plan that would use nuclear and gas to dramatically reduce CO2 intensity while retaining oil as a key transportation fuel. Crude is extremely energy dense per kilogram, making it invaluable in powering things like cars, trucks, and planes.
According to our models, an aggressive push toward nuclear and gas would allow total global energy demand to grow by 35% over the next two decades (instead of falling per the IEA) while still cutting carbon emissions by nearly half.
Such a plan would save nearly $30 trillion, some of which could then be used to pursue the more promising carbon-mitigation technologies. While still in the early stages, there are some very exciting companies advancing various new solutions. Boston Metal and Ambri were both founded by Professor Donald Sadoway, John F. Elliott Professor of Materials Chemistry at the Massachusetts Institute of Technology (and presenter at the 2018 Goehring & Rozencwajg Investor Day). Boston Metal is exploring ways to produce carbon-free steel (8% of global emissions) and has just secured a $50 mm investment from mining giant BHP. Ambri is developing a liquid metal battery better suited for grid level storage that shows early promise. Another area that shows promise is cement manufacturing. CarbonCure injects CO2 into the cement stream to strengthen concrete. The process has two benefits. First, the captured and injected CO2 is not released into the atmosphere. Second, cement per tonne of concrete is reduced by 5%. Cement manufacturing represents 8% of all carbon emissions and so any reduction would be meaningful. Many of these technologies are early stage and their eventual success is far from uncertain. However, our research tells us that taken collectively these technologies are much more intriguing than the current portfolio of wind, solar, EVs, and fuel cells that we expect will continue to disappoint.
There are signs that perhaps investors and policy makers are beginning to appreciate the limitations of the current “green” transition. The Biden administration has taken the most favorable view on nuclear power in decades. Delaying nuclear closures (currently rumored) would be a huge step in the right direction. On the investment side, we are extremely impressed by the work being done at Breakthrough Energy Ventures. Founded by Bill Gates, BEV seems to understand the nuances of the energy transition. It is no surprise that BEV is heavily invested in nuclear power along with several of the technologies we discuss above.
Please continue to check back for ongoing installments in this series on our website going forward. It is a fascinating topic, and we are excited to share our research with you.
This was originally published by Goehring & Rozecwajg in the firm's Q4 2020 Natural Resource Market Commentary. You can download the full commentary here.