Metals Shortages Will Make the Renewable Transition Impossible

If fossil fuels are to be substituted for solar and wind, they'll need to massively scale up. It turns out that this uses far more metals than was thought. And we're not discovering new metals.

The Actual Significance of Solar and Wind Today

As of 2023, wind and solar stood at 13.35% of global electricity production. This is a serious feat considering that in 2010 they were 1.78% of global electricity production. Yet clearly electricity isn’t used to power everything. Essentially all cars, planes and ships, many residential heating systems and industrial processes among other things rely on the direct burning of fossil fuels for heat to function. For this, understanding primary energy consumption is useful.

Primary energy consumption refers to all energy consumed, whether that be through burning wood to heat a home, or using solar generated electricity to power an air conditioner to cool a home. Globally, around 20% of primary energy consumption is in the form of electricity, while the other 80% is non-electrical (think of coal/gas furnaces and heaters, ICE vehicles etc.).

Once this is considered, the current role of wind and solar appears much less significant. Energy from solar and wind is only consumed through electricity. Considering that electricity is 20% of primary energy consumption, and wind and solar are supposedly 13.35% of electricity production, that gives a value of (1 * 0.2 * 0.135) = 0.027 or 2.7% of global primary energy consumption in 2023 being derived from wind and solar. If one refers to Our World in Data’s values for primary energy consumption contribution of wind and solar for individual countries, wind and solar seem much more significant than they actually are. We are using the same sources, I tested their value for primary energy contributions of solar and wind for the United States (they claim 6.6%) and they don’t seem to add up correctly (i.e their values are faked to inflate solar and wind). So beware if you’re doing your own research. A subject for a future article perhaps.

Anyways, if wind and solar are to be the primary power source of the globe they need to massively scale up. The clever folks that cite the electricity production of solar being 13.35% solar and wind in only 13 years of build-up give the impression that this scale up is just around the corner. In reality, for solar and wind to represent majority of all energy generation, the world must greatly electrify with the vast majority of primary energy use must be through electricity. That involves an unprecedented expansion in energy storage through batteries and pumped hydro. This is the precondition for solar and wind to be able to fulfil the role of fossil fuels. To illustrate how far renewables still are from replacing fossil fuel energy generation; for wind and solar to represent half of primary energy use they must be multiplied their power output by at least 20! This brings us two questions. Firstly, can the world be electrified? Secondly, can solar and wind power generation be multiplied by 20 or more?

Storage Requirements

For solar and wind to replace fossil fuels they are required to store their energy for long periods of low solar and wind output. A case study of the solar output of sun-blessed Spain that I became aware of through Dr. Simon Michaux’s work is useful to showcase the extent of this.

There is a clear energy deficit for the 670MW (MW = megawatt) target throughout the whole month of November 2015, and a clear surplus during June -- except for a 5 day period. Conventional calculations for the viability of renewables assume that 6 or 48 hours of yearly solar and wind power generation are to be stored to make up for generation deficits. Anyone who can read the graphs above can see that this is woefully insufficient. Assuming that energy demand is a constant 670MW, the storage would be empty in a few hours to days and leave the whole month of November with scarcely half of its energy needs!

Currently, places such as California that have extensive solar and wind power generation also heavily utilise natural gas for periods of low wind and solar output. As of 2023, 43.68% of in-state Californian electricity was sourced from natural gas. This acts as a substitute for batteries. However, if renewables are to replace fossil fuels this practice will obviously no longer be possible.

What is needed is at bare minimum 28 days worth of solar and wind power generation able to be stored for months at a time to be utilised during the long periods of low solar and wind activity. In places without great hydro potential this will have to be achieved with batteries. Below is a chart showing German solar output over a year with a 28 day buffer represented by the red square. That red square looks to cover a third to a quarter of the production deficit.

Graph courtesy of Dr. Simon Michaux

Most renewable power systems aren’t placed in the worst regions in the world for them (such as solar in Germany), so 4 month buffer is abnormal. Nonetheless in most places, a 28 day buffer would just scrape by. A 12 week buffer would be more comfortable, yet as we shall see, neither are possible.

Metals Constraints

So what is the issue with expanding storage or increasing generation capacity of solar and wind? The issue is that on a global scale, the sheer quantity metals required to do build all of this do not exist at all, at least in a form that is anywhere near economical to mine. Especially in a context of rising energy prices.

A brilliant table above by Dr. Simon Michaux (check him out). It is a very conservative estimate of the amount of metals required for a generation of renewables at current primary energy consumption. You can almost triple the metal quantities if you assume a 12 week buffer is required. This table assumes that the world will still have 1.5 billion cars and the same quantity of buses and trucks but they will instead be electrified. All primary energy generation will be generated through renewable sources with 28 days of solar and wind stored power buffer. Among other fossil fuel processes converted to renewables. Even though I consider a 28 days buffer inadequate, this is still amounts to a relatively close approximation of our current world, just without fossil fuels. The typical picture that is painted by those advocating for the “green transition”.

As you can see, the metals required to do this do not exist. The amount of metals needed to 20-40x our current wind and solar generation and the massive build-out of batteries required for planetary scale electrification and wind and solar generation is simply too much. And again, this is just the first generation. Every generation the whole system must be replaced. Each generation lasts 15 to 20 years.

Crashing Discoveries versus Spending

Extrapolating current production with current reserves and finding a date when there will be nothing left is a rookie mistake isn’t it? We will continue to discover more of a resource as we mine it. Or so it is said. The trouble it is that despite throwing more money at exploration for copper, we are finding next to nothing.

Quite a spooky graph isn’t it? You’d think that more people would be talking about the fact that we are almost literally not finding any new copper despite having exploration budgets many multiples of what existed in the 1990s and 2000s, when we were actually finding new copper reserves. And it’s not like this is just a one or two year long shock. This copper drought has been sustained for nearly a decade. We are mining through copper that was discovered decades ago, not discovering anything new all the while we are supposed to 5x to 11x our copper reserves to build out infinity solar panels, wind turbines and batteries. It simply doesn’t add up.

Ore Grades are Declining, With Significant Consequences

Ore grade refers to the percentage of a deposit that has a target material. For example a 1 ton copper deposit containing 100 kilograms of pure copper would have a grade of 10%. Ore grades for the metals required for wind and solar are declining significantly. As ore grade decreases, the amount of waste rock that must be processed to extract a unit of the target material increases exponentially.

In the early 20^th century a copper mine would be expected to have a grade of 5%, if 1 ton of ore was processed there would be a yield of 50 kilograms of copper. However today common grades for new mines is 0.4%. This is what would be expected, as over time the trend for a natural resource is for the most bountiful and profitable finds to be extracted first, with subsequent operations of lower and lower quality. As illustrated in the chart above, this seeming small change in ore grade has a deceptively massive consequences. With grade moving from 5% to 0.4%, waste rock ratios have moved from 20:1 to 255:1.

The processing of all this material (namely hauling, breaking and grinding it apart) is very energy intensive. Ore of lower grades must be ground down to finer and finer pieces in order for the processes used to liberate the target material from the impure rock to work. For reference, in 1980 it was typical for copper ore to be ground to 150 micrometers, in 2021 copper ore was being ground to 4 micrometers. A human blood cell is 5 to 7 micrometers.

As a consequence of this required increase in complexity, small scale mines are not economically viable. The scope and scale of the equipment and mines has had to radically increase. Vast open pit mines that move in some cases millions of tons of rock, all to reach ore of extremely low grade. What happens to such energy hungry operations when energy prices increase? The price of the metals they produce also increase.

As detailed in my overview article of ‘peak oil’, oil is likely to find a new much higher equilibrium price whenever the next ‘upset’ occurs, due to supply constraints, just as happened after the 2000s. Someone that can accurately time that can make billions, I wish I could! As a result, there will be new supply constraints on metals, the same metals required to solve the energy crisis left by the hole that oil once filled.

Recycling

I imagine that some readers are wondering about the viability of using recycled materials to build out the solar and wind grid. The issue with this is that the quantity of metals to do that has not been mined yet. At minimum the first generation of an electrified solar and wind energy system must be built out before it can be recycled. Also, continual metals demand for subsequent generations will always be high in such a scenario as even the most thorough recycling is only so effective.

Conclusion

To conclude, the metals requirements for creating an energy system where solar and wind mostly completely replace fossil fuels are impossible to meet. The major overlooked part of making our energy system dominated by solar is the large amount of long-term energy storage (28 days to 12 weeks of solar and wind generation) that is required to satisfy demand in periods of low solar and wind output. The quantity batteries required for this are immense, as is the metals to build them.

A key metal in electrifying our fossil system, copper, currently has radically insufficient reserves all the while new copper is not at all being found in quantities that will meet the upcoming needs, despite historically high investment in discovery. The more recent mines of copper have much lower ore grades than they did historically, leading to enormous blow outs in processing requirements per unit of copper. All of these factors make the renewable transition of the popular imagination highly improbable, if not impossible. Instead we will likely see a reduction in primary energy generation, as we are unable to fully substitute our fossil fuel sources with renewables on the same scale. This will have major consequences on the ordering of society into the future.

Support my work by donating to my Ko-Fi: https://ko-fi.com/bluevir

**

I focused on copper in this article, however that is not the only metal which acts as a bottleneck. In future I will likely investigate more metals and do a break down like I’ve done with copper, but more in depth.

Here’s one for nickel