warm thoughts


Written on: July 27, 2022

Originally published by RBN Energy and reprinted with permission

By Mark Mills

When you boil it down, there are only two energy-related responses to Russia’s war on Ukraine. First, there’s a big push to find sources of crude oil, refined products, natural gas and NGLs to replace Russian supplies as quickly as possible. Second, governments on both sides of the Atlantic are scrambling to reaffirm and even expand commitments to lower-carbon energy sources to delink from Russian hydrocarbons as well as meet energy transition goals. Both raise the same question: How fast can the world bring online any new sources of energy on the scale needed? Policymakers would like to believe the answer can be found through the stroke of a legislative pen invoking aspirational language. No one doubts the power of that pen to create incentives or impediments. But the answer to that question is dictated by the realities of the physical world. In today’s RBN blog, we discuss the options for accelerating the availability of the minerals, metals and other materials needed to build the required machinery for the energy transition.

As we said in Part 1, all the favored energy-transition technologies — solar, wind and batteries — require a lot more stuff to be mined, refined, fabricated and constructed to replace the same amount of energy provided by the hydrocarbon-based energy infrastructures that power the world today. In many cases, we’re talking about an unprecedented 3x to 70x increase over today’s use of not only a wide array of metals such as copper, nickel, aluminum, lithium and neodymium, but also a 10x jump in the use of basic materials such as steel, glass and concrete.

In Part 2, we looked at how the unprecedented rise in demand for key metals and minerals is inflating input prices and reversing the decade-long trend of falling costs for wind turbines, solar modules, and batteries. Mineral inflation also collaterally impacts the costs for everything else made from those same elements. How high and for how long prices go up will depend largely on whether transition policies continue to create ever-rising mineral demands because, as we explored in Part 3, the world’s miners and mineral refiners do not have sufficient capacity in place or planned, either for basic metals like copper and nickel or for more exotic elements such as lithium, cobalt and rare earths. Furthermore, the suite of energy transition minerals comes from a far smaller number of nations than now supply world oil markets, and many major producers are in places either unstable or unfriendly to the U.S. [In fact, just a few days ago the International Energy Agency (IEA) released a report acknowledging the risk posed by China’s dominance in solar supply chains.]

Today, we conclude our blog series with a look at the possibilities for increasing the supplies of all the minerals needed to build the machinery for the energy transition.

We begin by stating this undeniable fact: Aggressive energy transition policies are taking the world towards a looming and very large shortfall in transition-related minerals and metals. That means as demand continues to increase faster than supply, prices will continue to rise (even if with usual commodity volatility). Such a dynamic leads to only three possible outcomes: demand destruction, supply expansion or –– most likely –– a combination of the two. What seems most probable, as we see it, is a market-driven push to develop new supplies that will continue to be a step or two (or three) behind rising demand, which will spur higher prices and a recalibration of energy transition goals to reflect the realities of what’s affordable or even available –– a situation that will be very familiar to our energy audience.

There are things that policymakers can do to help expand the supply of critical minerals. First and foremost, governments need to base transition strategies on what could be reasonably achievable at costs that don’t cause broad economic harm. And it needs to be said that ramping up mineral supplies requires governments to be favorable to big mines and big chemical processing facilities. That support is more important than innovation (which we’ll get to in a moment). China’s 80% global share of rare earth supplies didn’t happen because that country invented more effective mining or refining technology, but instead from favorable policies. Meanwhile, the U.S. regulatory environment for mining is, to put it bluntly, hostile. The U.S. went from producing nearly all of the world’s rare earths in the late 1970s to depending on imports for 95% of its needs today. In fact, the U.S. now depends on imports for 100% of 17 minerals and for half or more of 28 others. Consider this: Less than 10% of global spending on mineral exploration happens on U.S. soil. The European Union (EU) is in the same boat.

It’s not as if policy makers have been unaware of all this. Myriad Congressional hearings and policy analyses over many decades have all reached the same core conclusions: the U.S. has vast mineral reserves, and the way to decrease dependence on imports is to increase domestic mining (and refining). There are challenges to jump-starting this, however. For one, there is a clear lack of consistent cooperation and participation by all the federal, state and local agencies involved in the regulatory processes, resulting in excessive costs, delays and inefficiencies. The U.S. has one of longest permitting processes in the world and a track record –– especially recently –– of permit reversals and cancellations. In addition, policymakers over the years have decreased access to federal lands for mineral exploration, never mind development. In the Western states, where the federal government controls the vast majority of the land, nearly half of it is off-limits to mining.

The simple fact is that the U.S., along with Europe, has regulated its way into far greater mineral import dependencies. Whether energy transition aspirations lead to policy reversals is debatable. But regardless of where the mines are located, technology will play a central role in determining how much production can expand. Higher prices always inspire investors to open new mines (somewhere), and similarly stimulate risk-taking to try out new technologies — provided businesses believe prices will stay high enough long enough to recover those costs.

Higher prices also encourage more recycling, something that’s always top of the list of options for “new” sources of supply in every policy proposal; so-called “urban mining” or now the chimerical goal of a “circular economy.” There’s no doubt a lot more opportunity exists to increase recycling rates. As one United Nations (UN) study observed, less than one-third of some 60 metals have an end-of-life recycling rate above 50% and 34 elements are below 1%. But it bears noting that mandating higher recycling rates will have the effect of inflating prices because recycling is generally far more expensive than mining. (If the inverse were the case, no incentives or mandates would be required.) All that aside, regarding the key issue at hand, it’s simply an arithmetical fact that even if recycling rates approached perfection it wouldn’t come close to meeting the supply gap. The transition-driven 300% to 7,000% rise in demand across the range of minerals over the coming decades overwhelms marginal additions to supplies that could come from recycling the relatively small quantity from worn-out hardware over the next two decades.

Fig1_Freeport-McMoRan’s Cerro Verde Copper Mine in Peru

Another commonly offered technology solution is to, effectively, create more supply by “dematerializing” green machines (solar panels, wind turbines, batteries, etc.) –– that is, making innovations that lead to greater efficiency and thus more output per machine, thereby reducing the materials needed per unit of energy produced. Such progress is technologically inevitable. But to have a meaningful impact on primary mineral demands would require nearly tenfold leaps in efficiency in the underlying solar, wind and battery technologies. Such gains aren’t even theoretically feasible. All those technologies are well past early development stages when rapid improvements in efficiencies happen, and all have now entered the stage of incremental gains common across all energy domains.

Despite episodic headlines about putative breakthroughs, there are no commercially viable alternatives that significantly change the order-of-magnitude of the materials needed to fabricate transition machines. In most cases, for example, changing battery chemistry formulations merely shifts requirements to different minerals. Reducing the use of cobalt, for instance, is typically achieved by using more nickel. Or, avoiding both nickel and cobalt and instead using, say, lithium-iron-phosphate chemistry results in a lower energy density, which leads to the need for a bigger, heavier battery using more copper and aluminum.

Then there’s the prospect for extracting more pounds of pure metal per ton of primary rock (ore) mined by improving the efficiency and yields in the physical chemistry used in the mineral refining and conversion processes. Improvements are inevitable there too, in no small part because that’s what engineers always do, and now because in the digital era they will more often find success. But there are no known “step function” changes on the horizon there either. These are well-trodden fields of physical chemistry where devices and processes already operate near physics limits. And when breakthroughs do happen — again inevitable, even if rare and annoyingly unpredictable — it takes many years to commercialize and scale-up industrial systems based on foundationally new inventions.

The bottom line is simply that the green machines that will be put into use today and in the near future will necessarily use technologies available now and not theoretically available someday. Which brings us to the fundamental challenge that transition policymakers are forced to face, namely, are there technologies that can accelerate capabilities to dig up a whole lot more rocks in the mining itself? Technology progress always yields new kinds of machines or means to improve the efficacy of existing ones. The role that technology can play is dramatized by the fact that compared to 50 years ago, miners supply the world with 500% more copper (at a reasonable price –– for now) even though declining ore grades means at least twice as much rock must be dug up to get the same pound of mineral.

In the digital age, big industrial machines don’t get the attention they deserve. But there is no escaping the centrality of the technologies and machines in “old” industries like mining –– even Elon Musk of Tesla and SpaceX fame has suggested he might have to “get into the mining business.” That’s because every manufacturing process depends on accessing, moving and converting materials by using ingenious machines animated by logic and energy.

Fortunately, the predicates are now in place to see another of history’s rare manufacturing transformations, one that is of precisely the kind seen over a century ago. The last industrial revolution came from the contemporaneous combination of a new logic of production (i.e. mass production), along with new kinds of manufacturing machines (not least the advent of electric motors and internal combustion engines). Today’s emerging industrial revolution is also being fueled by a new logic of production (Cloud-centric, networked machines mediated by artificial intelligence, or AI), along with new classes of machines (especially autonomous smart machines, a.k.a. robots).

While such transformations don’t happen overnight, they are inexorable and inevitable. And this one comes at a time of rising demand for the output from all basic industries when there’s also a growing shortage of appropriately skilled labor. The issue is not just the higher cost of labor but also finding enough people to operate the machines that will be needed to (safely) expand production. That’s where both classes of digital technology (even if currently out of fashion on Wall Street) are critical: AI that effectively upskills employees by automating various decision-making tasks, and the robots that fill the labor gaps or amplify the capabilities of available workers. In the latter case, for example, exoskeletons — a class of collaborative robots, or “cobots” — are no longer in the domain of fiction but commercially available, made possible by new classes of lightweight materials, superior power systems, sensors, and on-board AI-enabled controls.

Fig2_Caterpillar’s Autonomous Dumper Truck

Similarly, despite unfulfilled hype about autonomous vehicles on public roads, that technology is already in use in offroad applications where there is a severe shortage of drivers. Hundreds of robo-trucks already operate at mine sites around the world. Caterpillar — rarely on the technorati’s list of robo-vehicle companies — properly brags that its autonomous dumper trucks have already safely moved more than 3 billion metric tons (MT) of material and driven tens of millions of miles. A robo-truck is easier to deploy offroad in more constrained environments, and the expense of the autonomy is, as an arithmetical matter, a smaller share of the high-cost and high-utilization of industrial vehicles.

Practical automation is also emerging for construction applications used everywhere, not just at mine sites. Autonomy and AI are also accelerating the cost-effectiveness of the critical task of finding new mineral resources in the first place. For example, a few years ago, Bill Gates and Jeff Bezos jointly funded a startup that makes a kind of “Google maps” for the locations of promising mineral geology by vacuuming up the wealth of existing exploration data and using machine learning to undertake virtual exploration. Similarly, AI combined with virtual and augmented reality technologies are bringing much-needed enhancements to the essential mission of skills training, including hyper-realistic simulators.

And yes, we’re aware of claims that a tech “revolution” is proposed to electrify mining trucks and equipment. While there are promising prototypes that will doubtless find various niche applications, batteries are just not up to the 24×7 performance demands to power heavy equipment in most industrial uses, especially mining. As we noted in Part 3, energy accounts for about 40% of mining costs, with nearly all of that energy coming either directly from hydrocarbons (diesel, natural gas, etc.) or from electricity that is largely generated by hydrocarbons. Thus, policies leading to higher hydrocarbon prices increase the prices of minerals. Any revolution in mining won’t come from putting minerals into batteries to electrify big machines, but in putting silicon and software to work to make those machines more productive. So, the good news is that technology will, eventually, yield what it always has: more output at lower per-unit cost. But for aspirants of a rapid transition, the bad news is they’ll have to be far more patient.

The realities of the long lifespans of mining and industrial equipment, often measured in decades, means that the scale of mineral supplies needed for transition policies won’t be happening in the time-frames policymakers imagine. Oh, and for those of you thinking, “yeah, we can just mine the sea floor, or the moon or an asteroid” to get all the minerals we need, we hate to break it to you but such aspirations are in the even more distant future. And while technology is making seafloor mining conceivable (think robots, again), space mining is still far beyond even Elon Musk’s dreaming.

If policymakers want to chart a future that puts less, rather than more, stress on global mining infrastructure, policies should encourage energy systems that require fewer minerals than hydrocarbons, not more. The only path to such a future requires greater use of nuclear energy. Nuclear fission offers a one-hundredfold reduction in material intensity over combustion, and a thousandfold reduction over solar and wind. But that’s a story for another day.

For now, regardless of increasingly aspirational rhetoric about accelerating an energy transition, in the real physical world, Russia will continue to sell hydrocarbons, even if surreptitiously and to nations that don’t align with U.S. and European interests. And most of civilization’s energy will still come from oil, gas and NGLs in the usefully foreseeable future. An energy transition will occur, over time, but only at a pace that is economically sustainable, and that is aligned with the pace of society-scale technologies and, critically, the availability of the minerals and metals needed to make it happen.

Mark P. Mills, a physicist, is a Manhattan Institute senior fellow, a faculty fellow at Northwestern University, and partner in Montrose Lane, an energy-tech venture fund. He is author of Digital Cathedrals (2020) and Work in the Age of Robots (2018), and he is the co-author of The Bottomless Well (2006). He served as chairman and CTO of ICx Technologies, helping take it public in 2007. Earlier, Mills co-authored a successful tech investment newsletter, the Huber-Mills Digital Power Report, and prior to that he served in the Reagan White House Science Office and worked for a number of firms in the commercial nuclear industry. He began his career as an experimental physicist and development engineer in microprocessors and fiber optics at the dawn of the semiconductor revolution, earning several patents while working at Bell Northern Research (Canada’s Bell Labs) and at RCA’s microprocessor factory in New Jersey. He holds a BSc, Honours, in physics from Queen’s University, Canada.

Details on Mark Mills’ latest book, Cloud Revolution, can be found by clicking here.