Seabed Mining Meets Reality: What the Technoeconomic Evidence Shows

Support CleanTechnica’s work through a Substack subscription or on Stripe.

I was pleased to be asked to complete the technoeconomic assessment of seabed mining by the National Ocean Protection Coalition and to join the webinar discussing its results. It is not often that I have the opportunity to bring together decades of work in complex systems analysis, technology readiness, and estimation in a way that directly informs policy choices in real time. The questions facing Pacific communities and decision makers are difficult, and I appreciated the chance to walk through the evidence, clarify what is technically feasible and economically credible, and help separate aspiration from engineering reality.

Watch the hour-long webinar from this link.

Read the full technoeconomic assessment from this link.

Miriam Goldstein [MG]: Thank you all so much for coming, and håfa adai and aloha to those joining from the Pacific. My name is Miriam Goldstein, and I’m the Executive Director of the National Ocean Protection Coalition [NOPC]. I’m just going to take a few minutes to set the stage for our discussion.

There is currently no commercial-scale deep-sea mining anywhere in the world, but companies have long eyed the ocean floor as a potential source of metals like nickel, cobalt, manganese and copper—metals used in clean energy, national defense and other technologies. Proponents argue that these minerals are just waiting on the seafloor, ready to deliver profits to investors, strategic advantage to governments and cheaper, cleaner, more ethical access to minerals. Opponents argue that deep-sea mining is reckless, unnecessary and driven by short-term gain at the expense of Indigenous peoples and long-term planetary health.

This debate has been going on for decades, but it was recently supercharged by the Trump administration’s April 2025 executive order directing agencies to accelerate approvals for seabed mining in both U.S. and international waters. The Department of the Interior has begun the process to offer mineral leases in two locations in U.S. federal waters—one off American Samoa near the Rose Atoll Marine National Monument, requested by a California-based company called Impossible Metals; and one off the Northern Mariana Islands, close to Guam and right next to the Mariana Trench and its National Marine Monument.

The administration is also considering a permit submitted by the Metals Company for mining international waters, which could substantially undermine the international agreement that governs the high seas.

There’s a lot of uncertainty, but we know a few things.

First, deep-sea mining is opposed by the people most likely to be affected. The American Samoa government is unified in its opposition to mining off its islands, and nearly all Pacific states and territories—including California, Washington, Oregon, Hawaii, American Samoa and Guam—have either prohibited, banned or restricted deep-sea mining within their waters.

Second, ocean scientists have confirmed that mining could cause widespread, irreversible harm to the ocean. The question is how much. If you want to learn more about the environmental implications, we’re dropping some resources in the chat.

So where does that leave us?

Today we’re here to engage in good faith with proponents’ arguments by asking questions like: Would deep-sea mining bring economic benefits to U.S. Pacific territories and other Pacific Island nations? Do we need deep-sea minerals for national security and the clean energy transition? And is seabed mining feasible given likely market conditions and technologies?

To make the tough decisions ahead, we need objective answers. That’s why the National Ocean Protection Coalition commissioned a new techno-economic assessment of deep-sea mining.

I’m delighted to introduce the authors of the report who are here with us to discuss their findings.

First, Michael Barnard. He is a recognized climate strategist, futurist and founder of TFIE Strategy Inc., advising global investment funds, corporate boards and climate tech innovators on rigorous techno-economic evaluations and decarbonization roadmaps. He specializes in long-term scenario analysis across hydrogen, electrification, industrial transformation and critical mineral supply chains, grounded in detailed modeling and lifecycle analysis. His work appears in CleanTechnica, Forbes, peer-reviewed journals and books, promoting transparent, economically sound and scalable clean-tech solutions.

We’re also joined by Lyle Trytten, a chemical engineer and metallurgical consultant with more than three decades of experience in sustainable battery metals development. He has worked globally across R&D, engineering design, project startup, operations and techno-economic and lifecycle assessment in critical minerals including nickel, cobalt, copper, lithium, manganese and graphite. As “the nickel nerd,” he regularly participates in industry forums advocating transparency, due diligence and traceability in mineral production and refining, and writes an ongoing series on production technology and sustainability for the Nickel Institute.

Welcome, Mike and Lyle. Thank you so much for joining us today. 

Michael Barnard [MB]: Thank you for having us, Miriam. 

Lyle Trytten [LT]: Pleasure to be here. 

[MG]: We are excited for this discussion, and I’m sure the folks on the webinar are too. To kick things off: your report, which we’ll put in the chat, suggests that commercial mining is highly unlikely to be viable in the next fifteen years.

For a general audience, can you explain your professional opinion on the viability of seabed mining right now?

Take it away, Mike.

[MB]: When I was originally approached by the PC, they came to me because it was hard to find people with the time, skills and ability to deliver a techno-economic assessment in a reasonable period. I knew I was not competent to do the metals side. I have spoken to global experts in critical minerals like Gavin Mudd, and of course to Lyle, which means I know my limits. So I brought Lyle in, and he was interested in participating.

What I am competent to do is complex, integrated technology assessments, energy modeling and determining whether complicated systems are likely to be technically feasible. In this space, Occam’s razor applies. The fewer components a system has, the more likely it is to work.

The Metals Company is essentially scaling up a big vacuum cleaner, very similar to dredgers we use today. The challenge is doing it at four to six kilometers depth with a seven- to ten-kilometer riser. It can be made to work, but the odds of the bulldozer-sized crawler breaking down, needing to be hoisted to the surface for ongoing maintenance, or bad weather forcing the riser pipe to be cut loose to keep the ship safe—followed by months of reattachment—lead to a low likelihood of stable economic production without years of struggle to stabilize operations. Even then, it may still be unlikely to achieve sustained production. But at least you can see a path to it working, even if it is not economically attractive. Lyle will speak to some of the reasons why.

Impossible Metals, by contrast, proposes a swarm of 314 autonomous underwater robots, each roughly the size of a shipping container, equipped with 18 plucker arms and AI-based nodule identification, sweeping the seabed. Their claimed advantage is reduced seabed disturbance compared to TMC’s tracked crawler. But they run into a host of other problems.

Precise positioning of AUVs requires ultra-short-baseline sonar, which does not work for large numbers of vehicles. It works for one to three AUVs, typically using both bottom transponders and ship-mounted transponders along with IMUs. Impossible Metals is proposing to exceed the state of the art in multiple directions simultaneously.

My background includes establishing automated compliance harnesses for GSM and CDMA chipsets around 2002. I have done extensive work in digital signal processing. I know the technologies they would need to integrate with USBL systems and how far those technologies are from maturity. That alone would take years to stabilize. And that is just one of the problems.

They claim a technology readiness level of 6, but that applies to a single component—not to a system of 314 components functioning together, integrated with an automated ship that plucks vehicles from the water, unloads nodules, replaces batteries, cleans sensors, redeploys the units and does this continuously. The ship itself is a massive integration problem. Integrating the vehicles with the ship is another massive integration problem. Integration will bite them over and over.

It is simply far too complex a solution. While the Metals Company system might limp along eventually, I suspect Impossible Metals’ approach is infeasible because they do not yet know what they do not know. That is not a criticism of their intent—credit to them for trying to reduce environmental impact—but the technical feasibility is not there. Their actual system-level technological readiness is around TRL 3.

That is a very long way from TRL 9.

[LT]: Thanks, Mike. This has been a nice pairing of your deep knowledge of the technical side and my understanding of how the supply chains work. When we look at what happens after nodules are brought to the surface and delivered to a processing site, we have to ask what that processing chain looks like and whether we even need it today.

We already have well-supplied processing chains for the principal metals of interest in these nodules: manganese, nickel and cobalt. In fact, for all three, dominant producers have been restraining production in recent years to hold prices at levels that still make sense. These nodules represent a new type of resource, and the processing methods needed to handle them do not fully exist today. There is no single facility where you could simply drop these nodules in and produce valuable products from all the contained metals.

Yes, we can process them. Yes, we can eventually turn them into end-product forms. But the economic challenge is that the facilities capable of doing this already have feed. To get them to take your feed instead of what they already source, you must incentivize them, and that comes at a cost. So while the total metal content of the nodules may look impressive, that in-situ value is not what nodules would actually sell for. No one knows what they would sell for because none are being sold today.

But by analogy with other ore types, it is clear they would sell for only a fraction of their in-situ metal value. And that alone undermines much of the economic impetus for pursuing this.

[MG]: Thank you both for that overview. We will be digging further into those points throughout this webinar. And again, folks can submit questions in the Q&A throughout; we’ll leave time at the end to address them.

There is one point that has been raised online that I want to get out of the way. The CEO of a mining company has been posting on LinkedIn alleging that you, Lyle, have conflicts of interest that affected the conclusions of this report. Could you address the allegation that your work in terrestrial mining has inappropriately influenced the conclusions of this report?

[LT]: For sure. Thanks, Miriam. Delighted to address this. It has been an interesting few weeks.

I’m a semi-retired independent consultant. I take on work, paid and unpaid, for a variety of firms, agencies and think tanks, globally and nationally. The work I take on focuses on mineral processing, the mineral supply chain and setting up new supply chains for processing materials. Independent contractors like myself, engineering firms and many others routinely work with multiple clients without conflicts. That is normal in this industry. As a professional engineer, ethical practice is a cornerstone of my work.

I do not have significant investments in the mining industry. Less than one percent of my finances are in directly invested terrestrial mining stocks. The one company the proponent keeps suggesting I have a conflict with is tied only to a minor, ongoing advisory contract for a firm exploring a copper deposit.

Lastly, I want to refer back to the agreement Mike and I made with you when we took on this work. As independent professionals, we committed to follow the evidence wherever it led, without a preconceived conclusion. We did not work toward a predetermined outcome. We worked toward an evidence-based assessment. That is what appears in the report. 

[MG]: Thanks, Lyle. I appreciate you taking the time to speak to that issue. We’ve done the briefest of overviews of the substance of this report, so let’s dive in a little more.

Mike, the report compares two main approaches to deep-sea mining, which you touched on earlier: The Metals Company’s giant seafloor crawler and Impossible Metals’ autonomous robots. Could you say a bit more about the benefits and risks of each technology, and what those risks mean for the cost of the metals they hope to produce?

[MB]: Sure. TMC’s advantage is that it is scaling up a technology that already works in subsea conditions. It is a simple crawler with a vacuum, big pumps and a riser. They are scaling it up. It could probably be made to work from an energy perspective, though six kilometers is very different from the two hundred meters where dredge-based mining is done today off the coast of Africa. Companies like De Beers have been doing dredging-based seabed mining for about two hundred years. But the downsides of dredging are well known. Dredging tears up the seabed. That is acceptable in a port, where the seabed is already heavily disturbed, but in the deep ocean abyss we do not know the impacts.

And to be blunt, I was not asked to do an environmental assessment, and deep-sea biology is not my forte, but the concerns are evident. The dredge brings slurry up to the ship and dumps it back down at a different point in the water column. That has implications for the entire water column, and we do not understand what those impacts would be.

So TMC’s system could be made to work, and it might also fail simply because it is an extreme use case. But it has clear downsides.

Impossible Metals, by contrast, is trying to avoid those downsides. Their premise is minimal bottom disturbance. But to do that, they must operate complex sensors, complex manipulators and AI-based image recognition four to six kilometers under the surface in a production environment. I first engaged with swarm robotics around 2002 while reviewing master’s and PhD theses from English-speaking universities. Since 2010 I have worked professionally with artificial intelligence. Today I work with image-recognition AI in one of the firms I co-founded. So I have a sense of the limitations.

In imperfect conditions, image recognition and AI-driven manipulators fail constantly. They are not perfect. Sophisticated manipulators that grasp and move objects break. When you have 314 devices scattered through a water column over ten or twenty square kilometers, there will be many component failures. That is simply the nature of robotics.

Impossible Metals asserts a sustained pick-rate that, in my professional opinion—and in the opinion of others experienced in robotics and imaging—is not realistic. And, as Lyle has shown, nodules are much smaller and more unevenly distributed than people imagine. All of this has cost and readiness implications.

I use technology readiness levels, TRLs. They are the industry standard. NASA developed them. I have been assessing technologies using TRLs for twenty-five years in my work with startups, venture capitalists, my own companies and techno-economic assessments. You can have a high TRL for a component, but if you need multiple components working together, the TRL of the overall system is the TRL of the weakest link. And in the Impossible Metals stack, some elements are at TRL 3 or 4. That means the overall solution is TRL 3 or 4 out of 9. TRL 9 is when something can be bought on the market and reliably does what the box says. They are a very long way from TRL 9.

From a costing perspective, I served as a global estimation subject-matter expert for a major technology firm, working on programs worth up to a billion dollars. I am used to estimating hard, novel systems. I only wish I had known Professor Bent Flyvbjerg’s work back then. Flyvbjerg—co-author of How Big Things Get Done—has developed reference-class forecasting. He has a database of more than sixteen thousand megaprojects over a billion dollars, in the scale range of deep-sea mining, categorized into hundreds of classes. The principle is: find similar projects, look at their cost and time outcomes, and use the average as a realistic baseline.

Because Impossible Metals is so novel, I did a component-by-component reference-class estimate using public data. Then I looked at component aggregations—ships, AUV fleets and so on. My estimated capital cost was significantly higher than what Impossible Metals claims. From my perspective, they are much further from technological readiness than they assert, their system will cost much more than they say, and the whole package does not look technically viable.

And as I always tell people: if you think I’m wrong, prove me wrong. If in ten years you are running a productive, profitable operation based on this technology stack, I will happily admit I was wrong. Until then, that is the bar to clear.

[MG]: Thanks, Mike. So just to summarize: some individual pieces of the overall technology system may be at a more advanced level, but the system as a whole is not, because it is so large and complex. And the task it must perform—picking up nodules at a specific rate while selecting only certain nodules—is itself extremely difficult.

Is that an accurate summary?

[MB]: Their argument is that they dropped a single early version of one of their autonomous underwater vehicles to the bottom of the Gulf of Mexico a couple of kilometers down, picked up some rocks and now claim they are at TRL6. I’m not going to dispute whether they are or aren’t. They certainly have a prototype, and it is a certain scale. But that is one vehicle out of what would need to be a swarm of 314 autonomous units.

The swarm concept has not been proven. That number of autonomous underwater vehicles has never been deployed anywhere for any commercial purpose. There is no underwater swarm. A swarm of 314 vehicles would be larger than any autonomous drone swarm operating above ground today.

And they would be operating in a medium where you cannot use GPS and can only use sonar for communication. Sonar has severe bandwidth limitations, which means they would need to rely on adapted technologies from fields like mobile phones and digital signal processing. Proving one prototype works does not remotely prove the system.

[MG]: Thanks, Mike and Lyle. This is a really helpful explanation of a very challenging problem. I want to make sure we also have time to talk about the processing side, since that is key to getting metals on the market, which is of course the end goal of proposed deep-sea mining projects.

Lyle, you alluded to this in your introduction. There are significant issues around processing and how the supply chain works. Why is the processing issue so difficult? We already know how to refine nickel and cobalt; they are in the supply chain now. So why can’t nodules just go into one of the existing facilities? And does the way the supply chain works affect whether deep-sea mining could ever be economically viable?

[LT]: Yeah, Miriam, the nickel, cobalt, manganese and copper supply chains are complicated beasts. They are globally integrated. One thing proponents like to claim is that nodules are great because they contain four metals in one deposit. Theoretically, that sounds valuable. It is always better to have more value in your ore, but it also complicates processing.

Today’s supply chains are optimized for treating very specific feeds. We have nickel sulfide smelters, nickel oxide smelters, nickel oxide leach plants. Each feeds its own downstream refinery type. We have copper sulfide smelters, copper oxide leach plants, cobalt–copper roasters, manganese oxide ore smelters. It is a wide, diverse ecosystem. When we add new metals to recover—or to eliminate, in some cases—it adds complexity and process-modification costs that someone has to bear.

If we treat nodules as nickel ores, since nickel appears to carry most of the value, the most likely existing processes that could recover that value are rotary kiln electric furnace plants. Indonesia has dozens of these, producing nickel–iron alloys for stainless steel. They work well and use a consistent ore feed mined locally. The technology has been used for seventy-five years. The presence of cobalt and copper in these products would actually be a negative. You do not want those metals in all stainless steels, and they would certainly not be paid for. So we would need a facility capable of producing nickel sulfide matte instead.

There are only a few such facilities in the world. Indonesia has some. New Caledonia used to have one, but it closed about ten years ago because it was uneconomic. These facilities act as swing producers. When prices are right, they produce nickel pig iron for stainless steel. When those prices weaken, they swing over to making nickel matte, which is then refined into high-grade nickel and cobalt products.

That nickel–copper–cobalt matte can indeed be processed at some refineries worldwide, but most of them are integrated with upstream smelters. Glencore, for example, has a highly efficient refinery in Norway that treats matte coming from its own smelter in Canada. Introducing a brand-new feed type into that kind of supply chain is extremely complicated.

If you want to recover all the metals in nodules, you effectively need to supply 100 percent of the feed to a facility. If you attempt to co-feed nodules with, say, laterite ore, you will produce a low-grade manganese byproduct that is unattractive to the market. So you would need to take over the entire smelter’s feedstock supply, which introduces security issues: the smelter needs confidence you will consistently deliver processable feed, and you need confidence the smelter will process and market the metals on acceptable terms.

Owning the entire value chain—your own smelter, your own refinery—avoids that coordination problem, but comes with enormous cost and risk.

If you take the market-based approach of contracting smelter capacity, it becomes much harder. These smelters already have functioning business models. They are not short of feed. Some are even holding back capacity because of low nickel prices. Their model today is simple: they either mine their own feed at very low cost, or they buy feed at a small fraction of the contained metal value—typically 20 to 30 percent. They process it, own the product, market the product and decide where it goes.

To entice them to buy deep-sea nodules instead of conventional feed, they would be taking on substantial risk. They would need a significant incentive.

Assuming they would simply operate on a tolling basis—charging a processing fee and returning the metal to you to refine elsewhere—is a radical departure from how the industry works. It might become possible if the conventional business model breaks down due to a shortage of terrestrial supply, but the work I’ve seen from the economic geology community does not suggest that is coming in the next ten or fifteen years. 

[MG]: Lyle, you mentioned an interesting example in New Caledonia that I wanted to ask a bit more about. One thing we are hearing from the Mariana Islands, especially, is the idea that they could put the processing on-island for the potential economic benefits. Could you speak to whether that might be possible on islands that are fairly far from the mainland? And if processing were placed there, what would it look like in terms of the process itself and the byproducts produced on an island?

Why was the New Caledonia example uneconomic, and is there anything we can learn from that?

[LT]: New Caledonia has been one of the world’s dominant nickel suppliers for more than a hundred years. There is a long industrial history there. They have built several types of refineries and plants, including rotary kiln electric furnace smelters. They also built a new type of electric furnace smelter a few years ago that did not turn out well, and Glencore eventually shut it down and walked away. They built a high-pressure acid leach plant that also went poorly, and Vale spent $500 million just to give it away. The industry has had a checkered past in New Caledonia because it is a relatively high operating-cost environment compared to places like Indonesia, where similar nickel ores are processed.

The French government has been supporting New Caledonia for a long time. There is a pool of funding devoted to keeping these businesses open because they are important to the local economy. But even with that support, it is a difficult business case. And this is in a territory with deep experience, a long history, and a trained workforce.

Trying to set up a new system in competition with existing, proven systems—like the ones operating in Indonesia, or the high-pressure acid leach plants in the Philippines, China and elsewhere—is unlikely to end well. Mike has referenced Professor Bent Flyvbjerg and his work showing how big projects often fail to meet expectations. In the metallurgical processing world, Terry McNulty from Arizona has written a series of papers since the late 1990s documenting how metallurgical plants actually perform. The pattern is consistent: doing new things is hard, and new things often don’t turn out very well. This has certainly been the experience in the nickel industry.

When we have a resource we understand and a technology we understand, and we are simply applying the same technology again to the same type of resource we have always processed—like a new zinc roaster for zinc sulfide concentrate or a new copper smelter—the results tend to be acceptable. We are not doing anything new.

But when we step outside that box and start working with new resources in new locations, where we lack the ecosystem of companies and skilled people to support the enterprise, it becomes very difficult to make things work. It can be done—if you throw enough people and money at it for long enough—but it is difficult. I worked on a project in Australia where the first five years were rough. A second owner spent a lot of money and eventually made it work after the first company couldn’t, and it is still running today. But Western Australia has a robust minerals industry that can support that kind of effort. Others have not been so fortunate.

I was also involved in a project in Madagascar. It was extremely difficult to bring in the necessary skills and build a trained, qualified workforce capable of doing everything well. Doing new things in new places is simply hard.

When it is the same old resource and the same old process we have always used, the path to success is much more straightforward.

[MG]: Thanks so much, Lyle. I think much of our audience does not spend much time in the metal processing world, so this is extremely helpful. I wanted to pick up on what you were saying about new things being hard and take it back to Mike, whose work often focuses on why first-of-a-kind projects fail to meet expectations.

Mike, could you speak a bit more about your work analyzing technology, particularly in analogous complicated systems and hostile environments? Why is deep-sea mining technology so challenging when we already do oil and gas projects and other work in the deep ocean?

[MB]: Sure. There are three or four complicating factors, but let’s start with the basics. In the world of venture capital and startups, most ventures fail. They may have great ideas, great teams and a solid understanding of the economic need they are trying to meet, but executing on something complex and new is hard. That is exactly why venture capital exists. Investors will back ten ideas in the hope that one pays off. Anything backed primarily by major venture capitalists is not anywhere close to commercialization. That is the type of technology we are talking about here. This is not the kind of thing a teacher’s pension fund would put money into expecting a guaranteed 7 percent return for twenty years. We are in high-risk territory.

Second, we do almost nothing at four to six kilometers below the surface of the ocean. Even at two hundred meters we struggle to keep equipment running. One of my sidelines is reviewing the crop of failed wave-generation and tidal-generation technologies every couple of years. Those systems are in water close to land, on the surface or just below it—radically simpler than operating four to six kilometers down, far from shore. They still fail because ocean water contains the very minerals that formed the nodules in the first place, which accrete to and foul anything they touch. It is also full of life. Marine organisms attach to anything submerged and grow. There is an entire maritime defouling industry for ships, and even that is difficult. Keeping anything with moving parts running underwater is extremely challenging.

So that is the second factor. You can imagine a swarm of drones in the air—two hundred or so—because air is a forgiving medium. Electronics work. Communication works. Positioning works. All of that is easy compared to coordinating hundreds of autonomous drones traveling up and down four to six kilometers through dense seawater full of life, then fanning out horizontally, then returning to the surface. And that assumes these drones can even survive repeated exposure to those depths.

The third factor is pressure. The pressures at four to six kilometers down are extraordinary. You reach depths where something that works fine at three kilometers begins to collapse. It is similar to the hydrogen-for-energy problem: to store hydrogen, companies need to compress it to pressures comparable to those found at those ocean depths, and their seals keep failing. That is the problem space we are in. Keeping electronics dry. Keeping batteries dry. It is possible for one-off deep-ocean exploration vessels, but that is extremely different from maintaining an industrial fleet.

For the underwater systems we do have today, we keep them as simple as possible. We isolate them from seawater as much as possible. We send AUVs down for a mission, bring them up, and then spend a lot of time maintaining and inspecting them. Nothing we operate underwater today runs 24/7, 365, for extended periods without significant maintenance downtime.

We have no industrial base anywhere in the world for four-to-six-kilometer-deep operations. We have no drone swarms of over one hundred units in any context. We have no ultra-short-baseline sonar systems coordinating that many vehicles. We have no autonomous underwater systems operating continuously at depth for long periods. Both TMC and Impossible Metals face enormous first-of-a-kind challenges, although TMC’s challenges are fewer.

I often say—and maritime engineers nod ruefully—that if something costs one unit on land, it costs ten on the ocean surface, one hundred underwater and one thousand at four to six kilometers depth. That is the scale we are talking about. The abyssal plains look tempting, covered in polymetallic nodules. This is just the latest cycle of people convincing themselves it can be done.

But the reality is that global experts in critical minerals, as well as my own work—mostly on the demand side but informed by supply-side analysis—show we have enough minerals for the transition using terrestrial resources and recycling. If it were an existential necessity, the world could extract metals from the abyssal plain. But it would take a Manhattan-Project-level mobilization by China or the United States, with military control, warships securing the area and effectively infinite spending. The economics would not make sense. And that is not what is happening here.

[MG]: Thank you. That is a vast and sweeping view of what it would take to go down there. I wanted to add, because I was curious myself, that for comparison, Deepwater Horizon—which was extremely difficult to cap—was at 1.5 km. Here we are talking about going two and a half times deeper at 4 km.

I want to squeeze in one more question for Lyle before opening it up to the audience. This relates to what you were saying, Mike. Part of this report involved a scenario analysis. We touched on it briefly, but I was not familiar with the concept before reading the report, and I suspect many in the audience may not be either.

Lyle, for those of us who are less economically inclined, can you explain what scenario analysis is and what it shows?

[LT]: Scenario analysis is something we do when we don’t have a high degree of certainty about the factors involved. In this case, we don’t really know the quantity or quality of the resource, we don’t know the capital or operating costs of extraction, and we don’t know the revenue potential for the nodules even if we did know their quality. Given that level of uncertainty, it would be amateurish to present a single definitive result. All of these variables exist on a spectrum.

In the minerals world, when we deal with something we know well—like conventional open-pit mining—we can calculate the cost of blasting, trucking, crushing and so on, and we can compare them to existing projects. That allows a much higher degree of certainty, and in those cases a single-output economic model is more defensible, though even then many of my colleagues would challenge the validity of a single output.

With the level of uncertainty in deep-sea mining, we really need to examine the full range of likely outcomes. What I did, without building a separate model, was take the Impossible Metals model published for the American Samoa consultations (version 6.2) and adjust it for a range of factors. These included potential nodule values—because we know very little about nodules in American Samoa—and potential productivity outcomes based on the concerns Mike raised about the fleet-based approach.

I also adjusted for amateur choices in the company’s model, such as placing all capital expenditure in the first year and assuming full operating rates the year after. That is not how projects work. You spend five, ten, sometimes fifteen years spending money, all of which has a discounted net cash flow value, and no project ever ramps instantly to full production. Not even mature industries like batteries behave that way.

I did not adjust any of the cost bases for individual equipment items to reflect Mike’s reference-class forecasting, where he indicated costs for ships, drones and other components would likely be higher. I left all those costs untouched. I used the company’s own assertions and only adjusted the number of AUVs required under different productivity assumptions.

Then I ran a series of scenarios with varying productivities and varying nodule values to derive a range of economic outcomes. Some scenarios looked reasonably good, others looked quite poor. We need to recognize that there is a real likelihood—my personal assessment leans toward the negative end—that the economic outcomes could be unfavorable rather than positive.

And that would not be good for a territory that is supporting the work and bearing the brunt of the risks.

[MG]: Thank you. I want to really emphasize this, because the folks in the territories are at a very critical point where they need to provide input on a relatively short timeline. This is for both Mike and Lyle, and then I promise I will move to the audience questions, but it is such an important point.

Both CNMI and American Samoa face significant economic challenges, and seabed mining can appear to offer a solution. Do you see any potential for it to promote sustainable development, or does your analysis suggest that is less likely at this time?

[LT]: Yes, I see that it could, but I don’t see it as likely to be a big winner for them. They need to go in with their eyes wide open to the range of potential outcomes and demand proof of accomplishments rather than assertions. Some groups are understandably concerned about negative impacts on other economic sectors like tourism and fishing, and that needs to be accounted for when considering territory-wide economic outcomes.

I have personally seen well-run mineral extraction operations bring real benefits to a region. They can create good jobs when there is local capacity to support that work. But negative outcomes are real too. We have all heard of the resource curse and Dutch disease, and those risks can materialize. Companies going bankrupt and leaving the public to bear the costs is also a very real possibility. It has happened before in this sector and others, and we need to remain aware of that potential.

[MB]: From my point of view, I’m reminded of a few years ago when I was engaged to do techno-economic assessments of European green hydrogen initiatives in Morocco, Algeria and Egypt. It was essentially energetic colonialism. Those economies were supposed to produce large amounts of green hydrogen and ship it to Europe instead of decarbonizing their own economies. When I was at a conference in Tunisia, one of the pieces of guidance I gave to participants from the Maghreb region was to take as much money from these Europeans as they could, do good things with it locally, because there was never going to be a hydrogen export economy. They were never going to make any money.

And that is where this seabed mining round sits. This is not the first attempt. It is not the second. I do not think it is even the fourth attempt to make seabed mining work. It is just the latest one. There is money on the table. There is venture capital money. You might be able to scrape some of that off and do useful things for the islands, but you should not be investing your own money or your islands’ money in this, and you should not expect a long-term revenue stream. It is not viable unless the Chinese or American military and navy show up with a Manhattan-Project-scale initiative.

[MG]: Thanks, Mike. With that, let’s turn to some of the audience questions. The first one is actually back to Mike, and it’s a question on timelines from the audience. Clearly these technologies will take longer to be ready for commercialization than proponents suggest, but is it possible to speculate as to when they might be ready, if ever?

[MB]: For them to be ready, there has to be demand for what is down there. Our metals-based economy of the future is not one where we dig things up, consume them, and throw them away. It is one where we dig metals up, use them for long periods, repurpose them for a second use, and then recycle them again into something new.

The battery industry is a good example. We have not yet reached the point where we are recycling car batteries at scale. They are lasting far longer than expected. A car battery was assumed to last only a few years; now it is turning out to last longer than the cars themselves. We are seeing the same trend with buses and trucks.

Recyclers actually have a problem right now because they cannot get enough old batteries. People are using them for ground-based storage, both behind and in front of the meter, as a second economic use after they have degraded enough that they are no longer ideal for transportation. So we are looking at ten to fifteen years of use in a vehicle, followed by perhaps another twenty years as stationary storage. After that, our best source of ore for battery minerals is batteries. Then we recycle them and make new batteries more efficiently.

Across the broader electrification space, the same dynamic appears. Take steel. Some claim we will run into a scrap shortage for electric arc furnaces. But we are going to be dismantling fossil fuel infrastructure, ships, trucks, and equipment, and all of that will become scrap that can be turned into new steel.

There is massive substitutability. I have spoken to Gavin Mudd, who has done tremendously interesting work with geological societies around the world using novel methods to estimate actual reserves. I spent a lot of time understanding his work. Lyle is actually envious of all the time I got with Gavin. He said—and we agree—that this is not a problem. The doomers worried about metal supply are simply not correct.

And that means the demand that would justify a Manhattan-scale seabed mining effort does not exist and will not exist. It is not just that the technology will take a long time to stabilize. We simply do not have the requirement.

[MG]: Thanks so much, Mike. We now have a question for Lyle from Alexander Tudela in American Samoa. Impossible Metals offered a paltry 1 percent profit share to the local community. How sustainable is 1 percent for the island economy, should the company be successful in its endeavors?

[LT]: When it comes to royalties, that certainly feels low. There are different ways to calculate royalties, whether on the total value of mineral produced or on a revenue basis, so I can’t speak to the specifics. But compared to what I see for land-based mining projects, it does seem like a low value. Here in Canada, we would typically expect 2 to 5 percent for base metals, often more for gold and more for hydrocarbons. A project I worked on many years ago in Madagascar had a higher rate than that as well. So it appears to be low, but the details of how it is calculated matter.

[MG]: Thank you. I’ll go for one more audience question and then we’ll move to our final wrap-up.

This one is another New Caledonia question. Are there any lessons to be learned that are relevant for deep-sea mining by comparing the experiences in New Caledonia and Indonesia?

[LT]: Indonesia has been a surprise to the metals industry over the past fifteen years because of the rapid and successful growth of its minerals sector. But that growth was built on massive Chinese investment and what we often call Lego-block construction techniques: keep doing the same thing repeatedly, do not invent anything new. The Chinese learned their lessons on high-pressure acid leaching from a project in Papua New Guinea that did not go well. It eventually became successful after six years, but they took everything they learned and began building projects in Indonesia, where a much larger workforce is available. It is a very populous country, but the Chinese still rely on tens of thousands of imported Chinese professionals to ensure these facilities operate smoothly.

The same applies to the rotary kiln electric furnaces. They have now built dozens, all to the same pattern. They built an entire ecosystem of engineering, design, suppliers, construction and operations so they can make each new facility look and perform like the last one. None of this applies to first-of-a-kind work in a jurisdiction with no experience in the industry. Indonesia already had a viable minerals industry and had been smelting nickel and cobalt for fifty years before the Chinese arrived. It was not new; it was an extension, executed with the full weight of the Chinese industrial and investment ecosystem behind it. You’re muted.

[MG]: I know, you’d think I’d have learned by now. Sorry about that. That’s really helpful in the context of how we’re discussing this first-of-a-kind technology versus a more cookie-cutter approach.

We are at time, so I’ll end with a final question. If you are advising a potential investor or policymaker today, what evidence should they ask for when they are being approached and told they should greenlight this industry? What should they ask from industry so they can properly consider the upsides and the downsides?

[MB]: Why don’t I start on this one? It depends on the investor. For venture capitalists, I’d say look for actual off-takers and real requirements for the metals, because they don’t exist. For infrastructure investors, I would say look for ten years of profitability and operations with ongoing productivity. For governments, I’d say don’t touch it with a ten-foot pole. This is private money; let private money pay for this. 

[MG]: Thanks Mike. Lyle?

[LT]: Certainly echo all of that. There’s a point where risk belongs, and that’s with the investors backing these projects. That’s part of the investment thesis: higher risk for higher rewards. For regulators and communities, I would insist on very high ESG standards, starting with thorough environmental and social assessments. Understand all the impacts if the project is to be successful. Require strong royalty regimes that are not subject to internal company gaming on profitability, because it is well known in the industry that companies can devalue the product they ship out of the country to reduce royalties and offshore the profits. The same thing happens in the AI and tech sectors. Also insist on constant regulatory oversight.

It is going to be very hard to be a present regulator six kilometers down in the ocean, so require upfront bonding for any necessary remediation. That is what mature jurisdictions do. You want assurance that the community is not on the hook for unexpected costs if things go south.

[MG]: That is a great place to end it. Mike and Lyle, thank you so much for joining us today for this webinar. Thank you to all the attendees. We will send the recording of this webinar to everyone who registered, along with the link to the report. We very much appreciate your time with us today, and we appreciate the knowledge you’ve shared. Thank you, everyone, and I hope you have a great rest of your day.