mdsolar's Journal: Material Needs for 100% Wind Water Solar Energy

I ran into a silly video comparing materials used in thermal power plants with those used in renewable energy generation. It turned out the main silliness stemmed from quoting extreme estimates from a Nature Geoscience Commentary piece by Olivier Vidal, Bruno Goffé and Nicholas Arndt of Grenoble and Aix which tried to justify a plea for more mining in Europe. The preposterousness in the video stemmed from these authors citing an outdated lifecycle analysis of solar power and then using this as an upper estimate. They did not give a lower estimate leaving the videographer space to careen off into absurdity. Such notes are typically not fully reviewed, and normally a reviewer would ask for both sides of an estimate. It turns out that asking for that produces interesting results.

We have the advantage of time in taking on this question because recent work on energy demand supercedes that used by the authors. They presume energy demand will grow with development as it would in the absence of a transition to renewable energy, but, mostly because transitioning to renewable energy is accompanied by electrification of all sectors, energy efficiency stops demand growth for energy. Mostly, more people driving cars means less energy use as the cars are electrified. However, this electrification also squeezes out biofuels which were a dominant source in the scenario the authors considered for 100% renewable energy, so it turns out we may expect 3 time more energy to come from wind and solar power than they assumed in their fig. 1.

Usually, these materials based criticisms of renewable energy are focused on things that could be in short supply. Often tellurium is cited because it is used in one kind of solar cell. Tellurium is common on the seafloor but the real problem with such an approach is that the most common type of solar cell is made from sand, which is very easy to find almost anywhere. Basically there are substitutes for any "bottleneck" material for renewable energy unlike fuel which constrains all non-renewable energy other than deuterium fusion. The authors recognize this and so concentrate on common materials: copper, aluminum, steel, concrete and glass.
The authors' main point is that production of these material will need to increase to transition to clean fuel-free energy and they explore constraints related to this. Now, while we make the energy requirement for wind and solar larger, we do so in the context of realizing that learning curves are quite strong in these technologies. That is why they offer a least-cost path forward. So while the authors consider lifecycle analysis for 3 MW wind turbines, we should acknowledge that most of the wind power will use at least 15 MW turbines (8 MW already being built ) so material requirements will be reduced. (Our guideline scenario conservatively uses 5 MW). Further, solar panels are becoming more efficient and we may expect the majority to have material demands much less than any considered by the authors. On balance, the higher contributions of wind and solar considered here are likely offset by reasonable reductions in material requirements leaving the authors' estimates for the filled symbols in their fig. 1 approximately correct.

But, now we come to a new issue. When we transition to renewable energy, particularly when avoiding combustion technologies, a great deal of material that we use now is no longer needed. Coal cars and indeed whole rail lines use steel that can be recycled. Oil and gas pipelines may be ripped up and tanker cars, trucks and ships may also surrender their metal constituents. The aluminum engine blocks of cars will also be unneeded. And, with greater reliance on electricity, more high temperature superconductors will replace copper cables in cities. Similarly, electric cars may be made with more composite materials rather than steel since they are intended to be more durable to match their superior drivetrain.

Let's make some estimates. There are about one billion cars on the road now and we can be sure that 70% at least will have aluminum engine blocks by the time we switch to electric vehicles since fuel efficiency standards are rising. If we estimate 44 lbs per block, that comes to 14 million tonnes. Aluminum coal hoppers probably yield back a similar amount at 20 tonne per car. So, we've freed up a year or few of production. Perhaps solar panel frames will be made from composites, which appear to be superior to aluminum in any case.

There are about 130 tonnes of steel per mile of rail and about 160 thousand miles in North America. Taking a factor of 4 more for the world and 10% used for coal alone, we'd recover 5 million tonnes of steel. There are also about a billion tonnes of oil tankers. Assuming oil and gas pipelines use 50 times more steel per mile than rail, they yield back 20 billion tonnes. Together, this seems to provide for a steel glut even after a renewable energy transition. Note that we do not consider steel from redundant nuclear reactors since they convert useful steel to nuclear waste. A shift to composites for autobodies would compound the steel glut by a smaller amount than oil and gas infrastructure recycling.

Concrete and glass do not seem to be big issues in terms of constraints and it may be that graphene will replace glass in any case.

Contrary to the authors' contention that the materials they consider have no substitutes, they often substitute for one another, aluminum for copper in conduction, steel for concrete in construction, and we are seeing the introduction of new materials such as high temperature superconductors and composite materials to replace many of these. Composites are an enabling technology for larger wind turbines and so play an inevitable role there.

We may conclude that the authors' concerns are too weighty for the actual situation and in the case of steel we may anticipate glut rather than scarcity.