Consider the costs of the green transition to the planet and people.
We make our world by extracting almost a hundred billion tons of resources from the planet every year. Through our relentless resource consumption, we have changed the land, oceans and atmosphere to a point where the future livability of our planet is in question.
In addition to population growth and rising living standards, we are now faced with a new, urgent and resource-intensive need — transitioning our entire energy, transportation and industrial infrastructure away from fossil fuels. We need to build tens of thousands of solar and wind farms, terawatts of energy storage capacity, billions of electric car batteries. This green transition will require an injection of hundreds of millions of tonnes of metal. Where will this metal come from and at what cost to the planet and its people?
Together with lead researchers Daina Paulikas and Dr. Steven Katona, we have produced an in-depth lifecycle assessment study that compares the cradle-to-gate impacts of two sources of metals – land ores and deep-sea polymetallic metals. The study focuses on four metals used in manufacturing EV battery cathodes and wiring: nickel, cobalt, manganese and copper.
If the goal is to minimize further damage to the planet and produce the world’s greenest, most ethical electric vehicles, where should EV manufacturers source their base metals? This is the central question motivating the study. You can download it here. You can download it here.
Imagine a billion electric cars
An electric vehicle with a 75KWh battery and NMC 811 (nickel-manganese-cobalt) chemistry needs 56 kg of nickel, 7 kg of manganese, 7 kg of cobalt and 85 kg of copper for electric wiring. About 1.3 billion light passenger cars drive on this planet today emitting planet-heating CO2 and air-polluting NOx and SOx. Imagine we improve the world’s public transport, increase the use of ride-hailing services (and one day robotaxi fleets) and manage to keep the global car fleet size down to 1 billion cars despite adding another three billion people by the end of this century. If we replaced 1 billion gas-guzzling cars with electric cars, we would need 85 million tonnes of copper (21Mt mined in 2019), 56 million tonnes of Ni (2.3Mt mined in 2019, only 50% suitable for batteries), 7 million tonnes of manganese (18Mt mined in 2019) and 7 million tonnes of cobalt (140Kt mined in 2019).
Explore two alternative sources of metals
Polymetallic nodules are made of almost 100% usable minerals and contain no toxic levels of deleterious elements, compared to ores mined from the land which have increasingly low yields (often below 1%) and often do contain toxic levels of deleterious elements. This means that producing metals from nodules has the potential to generate almost zero solid waste and no toxic tailings, as opposed to terrestrial mining processes which produce billions of tonnes of waste and can leak deadly toxins into soil and water resources.
Compare the impacts
Most of us have learned to read the small print of nutrition labels to make better food choices in the name of our health. We now need to learn to carefully examine the impacts of our resource extraction to make better choices in the name of planetary health (which eventually boomerangs as our own health and long-term prospects on this planet). The bad news is, the extraction of non-renewable metal-bearing mineral resources is by definition not sustainable and cannot be done without damage. The good news is, metals are recyclable and over time, as we build up enough metal stock-in-use to cover our needs, we should be able to cycle and recycle the same stock through the system. In the meantime, where should the missing stock of metals come from? There is no perfect option but there may be a better option.
White Paper Questions & Answers
DeepGreen produced three webinars in May and June 2020 on the white paper. The lead authors, Daina Paulikas and Dr. Steven Katona, presented the key findings and answered questions live. We received nearly 100 questions during the webinars, as well as by email, on Twitter and from other sources. Below is a summary of some of the key questions that were asked as well as their corresponding answers. A complete summary of the Q&As is available upon request.
1. Will any of the results be reviewed by other scientists to ensure robustness and validity? Is there any possibility for some of this to be put into a peer-reviewed journal?
Daina Paulikas: We produced a 170-page, multi-disciplinary study with in-depth descriptions of underlying mining processes, with a level of comprehensiveness we believe is critical to increase awareness of what mining—whether on land or in the ocean—involves in practice. In that expansive format, the white paper is not a good fit for a peer-reviewed journal. We have, however, done several things to ensure the robustness of our analysis:
- We followed standard life cycle assessment (LCA) methodology and utilized the same public databases used in the LCA studies for land-mining studies to ensure apples-to-apples comparison.
- We involved many experts throughout the modeling and writing stages, including many authors of top papers we cited in our assessments, for a variety of inputs or analyses. You can see some of these names in the Acknowledgement section of the study.
- We conducted a dozen cycles of internal and external peer review, in which we sent drafts or sections of the paper to experts for their reviews and comments, and we updated the paper from there.
- A formal audit of our methodology and key assumptions was conducted by Todd Cort, co-director of the Yale Sustainability Center, and Cary Krosinsky, lecturer in sustainable finance at Yale School of Management. The outcome of their audit is summarized in the Forward to the paper.
In addition to the above, we have used this study as the basis for producing two focused research papers for submission to peer-reviewed journals—one on the comparative study of climate change impacts of metal production including Global Warming Potential (GWP) and carbon sequestration (already submitted and currently under review), and another on comparative biodiversity impacts (writing in progress).
I’d also like to add that we don’t see this white paper as a finished product. It has already gone through 20 iterations based on feedback and inputs we have received. We are open to further constructive criticism, or questions about our framing or assumptions. At the end of the day, we want to make sure we have a good basis for informing future decisions. Please consider this an open invitation to engage to make this study better.
2. In order to prevent unknown impacts on ocean biodiversity from a new extractive frontier, why are we not looking to innovation in energy and storage issues, innovations in recycling, innovations in product design circular economy?
Steven Katona: I’m hopeful that we would be able to both reduce demand and recycle. The metals currently in circulation are all used for products and there are no excess stockpiles available to meet the net new demand represented by the clean energy transition, including electrifying our global transport and replacing our fossil fuel-based power generation. A renewable power plant can be as much as nine times more metal-intensive than a coal-fired plant (see this analysis by IEA: https://www.iea.org/data-and-statistics/charts/minerals-used-in-selected-power-generation-technologies), and we need to build thousands of them.
Recycling is also an important part of our assessment, but we don’t see it as an alternative to replace virgin metal mining – at least not in the short-to-medium term. Not enough metals are in circulation now to have recycling replace primary battery metals. That said, any metals that are available in recycled form can certainly flow into battery production. Remember that DeepGreen (and probably most nodule collectors) won’t be making batteries themselves, but only supplying metals to battery makers. Battery makers would use any metal sources available—virgin or recycled—depending on price. So, in the short and medium term, both virgin ores and recycling will be needed, with recycling being complementary to mining of virgin resources.
Daina Paulikas: We very much hope that terrestrial mines will improve environmental performance in every way possible. Our analysis builds in anticipated reductions in carbon emissions for power generation, and we know that land mining companies are working hard to reduce water use. Nevertheless, as the questioner notes, terrestrial mining impacts are known and they are very problematic in many ways. Effects on biodiversity are of great concern, especially in the light of the (IPBES, 2019) report detailing the global biodiversity crisis. Using nodules will impose impacts on abyssal living resources, but it will relieve impacts on severely impacted habitats and endangered species on land. This will be a temporary tradeoff, but it appears necessary and beneficial from a planetary perspective.
When considering a problem with global consequences, it is generally prudent to pursue multiple complementary solutions, and many people are looking at these types of innovations in parallel. Recycling is an excellent long-term solution, but as (World Bank, 2020) shows, there are not enough metal stocks to come close to meeting the Green Transition metal needs even if we jump to a 100% recycling rate. This is in part due to the structural supply-demand gap that occurs any time you have an exponentially growing market. Product design and circular economy solutions should be pursued now and phased in so they can be maximally leveraged as growth eventually slows.
Naturally, energy storage and recycling innovations can and will happen. The International Energy Agency (IEA) and World Bank indicate that some Green Transition metal demands like cobalt might be sensitive to the final battery chemistry makeup, while others less so; but that even given the uncertainty in predicting such outcomes, and at least a decade for technology readiness and global market penetration, these market dynamics are our current best guess at the reality to anticipate.
3. With the potential emergence of deep-sea mining, how much of the land mining industry do you expect to go bust? This will ultimately determine how much impact on land you can avoid.
Daina Paulikas: In metal markets, supply follows demand. And nodule metals are low cost, so the market will take them. Hence, the marginal quantity of metal produced from nodules, driven on the supply side by the number of active nodule operations, will displace marginal impacts that would have been imparted by land-ore mining.
We do not expect mining on land to disappear. Rather, we would expect it to continue more or less as is, perhaps with small reductions in some areas, and larger in the case of manganese. Remember that our analysis only focused on one question: where should metals come from to make 1 billion 75kWh batteries for EVs. Even if all the metal for this new use came from nodules, instead of developing new land-based mining projects, existing land mining would still continue to supply the demand for existing uses. The environmental and social savings are measured as the difference between using nodules vs. land ores for this new marginal use (EVs).
The answer to the question is perhaps best understood in the context of a time series of projected metal supply and demand. Given the expected increase in battery metal demand, several analysts (e.g., WoodMac, CRU, Benchmark Minerals) expect shortages in nickel, cobalt, and copper in the 2025-2030 timeframe. Metal production from nodules could initially help fill the supply gap—which means new land-based projects won’t be developed. But depending on supply-demand dynamics and production cost of various producers, over time nodules could start displacing some of the production capacity. We provide additional details on market projections and supply-demand dynamic nuances for each metal in the Economic impacts section of the white paper.
4. Can we conclude that per-kilogram emissions from producing nickel, copper, cobalt, and manganese is always lower when using nodules versus land ores?
Daina Paulikas: Recall that the land ores baseline in the white paper is an average by design. It’s possible that a specific land-based operation has a processing plant running on a sustainable resource, like hydropower; that it processes an individual ore deposit that happens to have a very high grade; and that it uses low-impact reagents and follows other environmental optimizations. In that case, the per-kg emissions impact could potentially be lower than the nodules baseline. However, when you look at the average, by far most land-based ores will have higher impact, and examples like the one I mentioned will likely be outlier cases. Additionally, recall that one of the major advantages of nodules is that once they have been collected and lifted aboard a ship, they can be shipped anywhere. This flexibility makes it possible to use hydropower (or other renewables), which is one of the main drivers of lower emissions. You may see some exceptions if a contractor does not use hydropower or another renewable source to process the nodules; in specific cases, emissions may rise closer to the land ore average. Note the other impact categories would still likely remain lower due to the mineralogy and additional properties of nodules.
5. Why do you not place any value on deep-sea species, ecosystems and related ecosystems services? Your white paper acknowledges that the most significant impacts of nodule mining will be on biodiversity, deep-sea species and ecosystems that largely are yet to be identified and studied.
Steven Katona & Daina Paulikas: Many marine ecosystems are of course vital to life on Earth. In our white paper, we focus on one specific seabed resource (polymetallic nodules) in one specific deep-sea environment – the abyssal plain in the Clarion Clipperton Zone. When talking about protection, we need to be specific – what are the spatial scales that could be impacted, and what are the actual ecosystem services that this patch provides for the planet?
(1) Spatial scales: oceanic abyssal plains take up 270 million km2 or 53% of our planet’s surface. To electrify 1 billion EVs would require an allocation of about 0.2% of that environment (508,000 square kilometers). An area that is almost three times bigger than that (1.44 million km2) has already been set aside by the ISA into preservation zones (called Areas of Particular Environmental Interest [APEI]) that will never be touched. The total area under ISA exploration contracts today represents 0.4% of the global abyssal seabed.
(2) That abyssal patch in the CCZ does not provide critical fisheries, coastal protection, food security, or most other Ecosystem Services to the same degree found in coastal systems or other parts of the ocean. It will suffer damage during nodule removal. Some populations and functions will recover in 30-50 years. Nodule obligate species will take much longer, but they will be able to take advantage of ~15% of nodules left uncollected as well as no-take zones within mining areas and the APEIs. New nodule formation will take hundreds of thousands to several million years. Bottom disturbance will not harm the climate, and the eutrophic and mesopelagic zones will be untouched.
Metal mining will harm species and ecosystems wherever it takes place. Land ecosystems contain many more species than are found on the CCZ seafloor. It is true that many abyssal species remain unnamed and/or undiscovered. Many new species also remain to be discovered on land, especially among insects, but also in higher taxa of plants and animals.
All species, ecosystems, and ecosystem services have value. The intent of sourcing battery metals from nodules is to minimize the overall damage to Earth, its atmosphere, water resources, the human population, and biological resources. Our analysis indicates that producing metals needed to make 1 billion 75kWh batteries will do more harm if sourced from land rather than from the CCZ seafloor.
6. What environmental impacts in the deep-sea are of greatest concern? What tipping points are we concerned about in terms of ecosystem services and how can we ensure we do not reach those tipping points? If we break a deep-sea service, can we fix it?
Steven Katona: We think that the CCZ seafloor is sufficiently isolated from human activities that the impacts foreseen from nodule collection will not significantly impact ecosystem services. However, the water column has more direct effects on ecosystem services. Activities on the seafloor are unlikely to have a material effect on the water column, but discharge of water and sediments from the nodule riser system could if not done properly. It is particularly important to minimize the amount of sediment in the riser system and to discharge water at a depth where it will do the least harm. Certainly, the depth should be below the mesopelagic zone, within which vertical migration of lanternfish and others occurs. The density of discharge water must also be matched to the density of water at the discharge depth so that it does not rise. Some animals in the food web, such as deep-sea squids, spend portions of their lives in deep water, so their welfare is of concern. Discharge plume effects on jelly organisms must also be investigated. These considerations and others, both technological and biological, are informing design of the nodule collection system and the selection of water discharge depth. No global tipping points are expected to occur. Environmental monitoring will continue throughout the years of nodule collection so any potentially serious impacts should be detectible well before a tipping point occurred. If a serious situation were detected, modifications would need to be made to the collection system or, at worst, collection would need to be suspended.
7. How long does it take the CCZ ocean bottom habitat to recover, after nodules have been collected? And what do we know about recovery for deep-sea vs. land?
Steven Katona: It takes a long time for any disturbed area to come back to its initial state of function, whether on land or in the ocean. Different parts of the ecosystem tend to recover at different rates. For the sea floor, bacteria are extremely important to the overall function of the seabed ecology. Bacteria are probably going to be able to recover relatively fast. A recent paper estimated that after 50 years, bacterial populations could be able to recover. Some of the organisms will recover more quickly than others. Serpent stars and some sea cucumbers have been shown to come back reasonably well within a couple of decades. Others may be harmed for longer periods. Organisms that reside on the nodules themselves will be killed when they are removed from the ocean floor, but about 15% of the nodules will remain on the seafloor to aid with recovery. Once collection in the CCZ is completed, it will be left alone to recover and won’t be disturbed by any other activities.
For comparison – if we take nickel mining on land for example, especially in regions with high levels of biodiversity and sensitive rainforests such as Indonesia – a recent study of Amazonian rainforest showed that it took about 40 years for some of that vegetation to recover and that after 65 years, the secondary vegetation was sequestering only about 40% of the carbon compared to undisturbed forests. Shade trees took a long time to re-establish themselves and overall, researchers estimated that it would take 4,000 years for the area to come back to its initial state.
For mining activities in any place, either on land or in the sea, long recovery periods are needed. The advantage I see with nodules is that recovery can start within days or weeks of the collection machine moving on and no other activities will take place there after the nodules are collected. Whereas a deforested area will need to go through reclamation to start recovery and—unless it’s turned into a protected park—it will likely be under pressure from all kinds of other human uses and may not be given the time to recover. The International Seabed Authority (ISA) has also designated specific protected areas (APEIs) to be left alone completely. Contractors will also leave certain reference areas alone and those sites will also be important for re-colonization.
8. When contrasting the photographs of the kinds of animals at risk from mining on land and on the seafloor, what are the main differences you see?
Steven Katona: We are familiar with the kinds of animals at risk on land—frogs, turtles, birds and mammals of various kinds, as well as fishes and invertebrate animals like insects, snails and worms. However, the kinds of animals that live on the bottom of the sea are much different than the animals on land. Some deep-sea fishes feed on the bottom, but most animals there are invertebrate (have no bones), live attached to nodules or crawl on, or in, the bottom. In the end, you cannot compare the two ecosystems by counting the number of species that are impacted in both places, because the number of species in the deep-sea is still poorly known and the same is true on land to some extent. Comparing the number of different organisms or their biomass is also not completely satisfactory. Similarly, you also can’t just rely on comparing the ecosystem services provided by the two systems. Instead, you have to look at everything, including various ethical and philosophical questions about which organisms we value most. How do we compare the risk of harm to a nematode or a sea urchin with the risk to a snow leopard or orangutan? These are questions that we have to wrestle with. I don’t think humanity has tackled these questions yet as well as we should.
9. The deep-sea is a large reservoir of genetic resources. What did you conclude with regards to the impact of deep-sea mining on jeopardizing the discovery of novel pharmaceuticals?
Steven Katona: Microbes, plants and animals on land have already made extraordinary contributions both to traditional (indigenous) medicine and commercial pharmacology by providing medicines and genes of many kinds. Similarly, sponges, cone shells and other animals from the ocean’s coastal zones and reefs have made similar contributions. It seems likely that organisms from the deep-sea could also provide substances or genes that will be of great utility to humans. Owing to the large portion of the CCZ that will be protected from nodule collection, as well as the very small proportion of the abyssal seabed that it comprises, we are confident that the search for new genes, antibiotics and other pharmaceuticals and substances will not be jeopardized. Furthermore, as part of the Environmental Impact Assessment programs, prospective nodule companies are collecting and sharing biological samples with scientists. For example, DeepGreen has already archived over 20,000 biological samples, collected at 7 km intervals and will be gathering more during coming years. They already have one academic research partnership in place and I imagine are open to more.
10. Should we be concerned about sediment impact and plumes, and is there ongoing research to understand these impacts before deep-sea mining starts?
Steven Katona: Yes, we should be concerned, and we are. Plumes come from two sources. The first source is the sediment that is directly disturbed while collecting nodules from the seabed. Some of the top 5-10 centimeters of sediment is disturbed by the tracks and could be entrained with the nodules as they are dislodged by water jets directed at the nodules in parallel with the seabed. Nodules and any entrained sediment flow into the collector machine, where sediment is separated from the nodules and is discharged at the back of the machine. When it re-settles, it can smother animals or clog feeding apparatuses, so the first order of business is to make sure the collectors are designed so that the amount of sediment disturbed and discharged as a plume is as small is possible. To be cautious, we can’t assume that 100% of sediment entrained at the seabed will be separated and re-deposited at the seabed. Some of it could travel along with nodules, so the second source of sediment is the seawater that is used to pump the nodules up to the surface and onto the ship. That water is then pumped back down in a pipe that goes quite deep. The engineers are working in collaboration with the biologists to figure out the exact depth it needs to be at, considering what wildlife might be impacted in the water column. It will likely go down at least 1,000 meters, perhaps 1,500 meters or even more. Wherever it’s released, those sediments will come out and oceanographers are modeling how it will be dispersed. I don’t think it will be a problem for organisms on the ocean floor, but the effects on mid-water organisms are being evaluated by the Environmental Impact Assessment studies that will be ongoing for the next several years.
11. Regarding the economic impact, land-based countries that will be affected could benefit from the ISA Fund. Will the compensation to these countries be enough to offset the impact if the current proposal by MIT takes effect?
Daina Paulikas: We attempted to identify which mining projects in which countries could be impacted by nodule collection because of where they are on the cost curve, focused on the manganese industry (the higher project production cost, the higher risk of negative impact). We also quantified likely ISA royalty pools, which ultimately will depend on the final production scenario, and we assumed the middle royalty rate scenario from MIT work. While we did not assess specific damages to individual land-mining projects, nor assess whether the damages can be fully offset from the ISA royalty funds as there is no framework or policy yet for the policy of compensation, one can use the provided graphs to get a sense of which countries and operations may be more relevant beneficiaries from the ISA Fund, and for the rough scale of the benefit, which should total at least in the $10’s of billions of US dollars.
12. Based on everything you learned in this study, how do you feel about the tradeoff of land ores versus nodules?
Daina Paulikas: As an engineer and physicist, I’m typically hesitant to state I am 100% certain about anything, because unknown unknowns exist, and there is always some risk of a consequence that you don’t know about. However, I don’t think anyone looking at all of the data that Steven and I have seen over for the past year could believe that, overall, land-ore mining is better for humanity and less damaging to the planet than producing metals from nodules. I think it’s a very strong case that nodules are the better option to supply the battery metals we need for the clean energy transition.
Steven Katona: That’s exactly how I feel as well. I regret having to harm any animal, and in my life, I try not to do that. But from a planetary perspective, from what we’re facing in terms of climate change and the need to mitigate emissions as quickly as possible, it seems that nodules provide one way of helping to get there and I think it’s the better option that we have. If it needs to be done, it should be done as carefully as possible to minimize harm to the organisms on the seafloor and in the water column. Nodule use would also minimize harm to people in many ways and I see it as the better way forward. But our job wasn’t to make that tradeoff. Our job was to dive as deep as existing data allows us into the impact analysis and communicate the findings based on the best information available. It’s up to the ISA and others to take whatever actions they deem appropriate.