Summary

Sometimes a systemic problem is discovered somewhere other than where it is normally expected.

When one country controls a large share of global processing of critical raw materials, the natural response is to increase domestic mining, build domestic processing capacity, diversify suppliers, and create strategic stockpiles.

All of these measures are necessary.

But while working on the problem of critical minerals in the United States, I encountered a deeper vulnerability: even newly created alternative production capacity may become economically unsustainable precisely when it is most needed.

This changes the problem itself.

The task is not only to create an alternative supply chain. It is to preserve the ability of that chain to replace the dominant supplier in the future.

In September 2026, I formalized this idea in a proposal for a Strategic Capacity Guarantee (SCG) — a mechanism intended to pay not only for produced material, but also for the strategic capability to produce it. As a measurable parameter, I proposed Time-to-Replace (TtR) — the time required to replace a critical external dependency with qualified alternative capacity.

However, the most interesting result for me is not the US-specific mechanism itself, but the path by which it was derived.

We started with a specific vulnerability, discovered that ordinary market mechanisms can reproduce that vulnerability, then asked how an adversary could attack the system being built. This led to a multi-layer architecture and, eventually, to the realization that the same way of thinking can be applied to Russian production chains.

The resulting architecture can be summarized as:

  • economic mechanism — Strategic Capacity Guarantee;
  • measurable parameter — Time-to-Replace;
  • transition protection — Shield;
  • industrial network — Framework;
  • long-term structural change — Autonomy.

The purpose of this article is to describe this chain of reasoning and explain why the resulting architecture may be useful beyond critical minerals and beyond the United States.

1. The Starting Point: Why Diversification Alone Does Not Solve the Problem

The initial problem is relatively well known.

According to the IEA, China is the leading processor of 19 of the 20 strategic minerals tracked by the agency, with an average share of around 70%. For several of the most important materials, concentration reaches 85–90% or more.

At the same time, mining is diversifying faster than processing.

Therefore, several countries may mine the required raw material while the next stage of the chain remains dependent on a single processing center.

This already creates an obvious vulnerability.

But a paradox appears when the problem is examined further.

Suppose the United States and its partners build an alternative processing plant.

Under normal market conditions, it may be more expensive than the dominant supplier:

  • higher cost of capital;
  • smaller scale;
  • more expensive equipment;
  • more expensive energy or reagents;
  • lack of accumulated technological expertise;
  • lower utilization;
  • less guaranteed demand.

As long as the global market functions normally, it is rational for a buyer to purchase the cheaper material.

As a result, alternative capacity can gradually become economically uncompetitive.

The plant closes.

Several years later, the crisis appears again — but the alternative is now gone.

This creates a strange structure:

The system can economically destroy precisely the capacity it considers strategically necessary.

This was the first important observation.

2. What the Market Does Not Know How to Price

The next question was natural.

If the market can price a produced ton of material, why can it not preserve a production capacity that may be needed tomorrow?

The answer lies in the difference between product and optionality.

When a plant produces material, its economic value is relatively straightforward:

How much is the produced product worth?

But the strategic value of the plant contains another component:

How much is the ability to have an alternative worth when the primary supplier becomes unavailable?

This capability may never actually be used for years.

In fact, it is preferable that it is not used.

But that does not mean that it has no value.

The resulting structure can be represented as:

\[ V_{strategic} = V_{product} + V_{resilience} + V_{optionality} \]

The ordinary market primarily sees the first component.

Industrial policy attempts to add the second.

The third — the value of the ability to replace a dependency — remains poorly defined.

This leads to an important distinction:

Product price is not the same as the price of production capability.

If we subsidize a particular plant, we improve its economics.

But this does not yet answer:

  • how much alternative capacity is actually required;
  • where it should be located;
  • what percentage of capacity should remain available;
  • how long activation should take;
  • how long the system can operate without the primary supplier;
  • which parts of the chain remain monopolized.

In other words, we need not only an economic indicator for a project. We need an indicator of the state of the entire production system.

3. Time-to-Replace: From "Resilience" to an Engineering Parameter

This led to Time-to-Replace.

TtR is the time required to replace the largest external dependency in a critical industrial chain with qualified alternative capacity under specified conditions.

This may appear to be a simple parameter, but it changes the problem.

Without TtR, one can say:

"We have alternative suppliers."

With TtR, the question becomes:

"If the largest supplier disappeared today, how long would it take for our system to continue operating?"

This is already an engineering problem.

TtR cannot be calculated from a single plant.

The entire chain has to be examined:

raw material → concentrate → processing → refining → intermediate material → component → final production.

If one stage has an alternative available within three months while another requires three years, the real TtR is constrained by the weakest necessary link.

This produces another important consequence:

Reserve capacity must be measured not only in tons of material, but also in the time during which the system can continue operating and the time required to restore an alternative route.

4. The Solution: Strategic Capacity Guarantee

From this research emerged a mechanism I called the Strategic Capacity Guarantee (SCG).

Its principle differs from an ordinary offtake agreement.

A conventional contract tells the producer:

"We will buy a certain quantity of your product."

SCG says:

"We pay for the existence of qualified, available, and activatable capacity that is capable of replacing a defined external dependency."

This is a fundamentally different economic object.

The government or qualified industrial buyers guarantee long-term demand or payments for availability.

The producer, in turn, guarantees:

  • the existence of the capacity;
  • a defined level of readiness;
  • compliance with technical requirements;
  • the ability to activate the capacity;
  • transparency for independent verification.

The contract therefore becomes bilateral.

The government guarantees the economic viability of strategic capacity.

The producer guarantees the strategic availability of that capacity.

If the global price temporarily falls, the plant does not necessarily have to close.

The reason is that the buyer is no longer paying only for a ton of material.

The buyer is paying for the right to have an alternative.

This is the mechanism I proposed as part of the emerging Agreement on Trade in Critical Minerals (ATCM) architecture for the United States.

5. The Next Problem: The System Will Be Attacked

At this point it would be easy to stop.

Build alternative plants. Guarantee their economics. Diversify supply. Done.

But this would mean treating the system statically.

If an existing dependency has strategic value, creating an alternative inevitably changes the interests of those who control that dependency.

Therefore, the next question had to be:

How would the system we are building be attacked?

A direct attack on a plant may not be necessary at all.

For example, an alternative supply chain can be undermined through:

  • temporary reductions in global prices;
  • market flooding with cheap material;
  • restricted access to equipment;
  • restricted access to technology;
  • shortages of specialized reagents;
  • pressure on financing;
  • pressure on partners;
  • capture of standards and qualification procedures;
  • creation of secondary bottlenecks;
  • gradual return of producers to a cheaper external supplier.

The last scenario is particularly interesting.

Nobody has to force a partner to abandon the alternative supply chain.

It may be enough to make the old dependency economically more convenient.

The system will then begin returning to its previous state on its own.

In the original work, these mechanisms were grouped into several classes of attacks, including predatory pricing, equipment denial, pressure on partners, and gradual economic capture of their decisions.

6. Therefore, the Solution Must Be a System Over Time

The analysis of possible attacks led to a three-track architecture.

It is important because different problems operate on different time scales.

Shield — Buy Time

Horizon: months to 2–3 years.

The purpose of Shield is not to build a perfect independent system.

Its purpose is much simpler:

Prevent external pressure from destroying the system before it becomes mature enough.

This requires:

  • stockpiles;
  • time buffers;
  • distributed reserve capacity;
  • pre-qualified alternative routes;
  • guaranteed demand;
  • the ability to scale production rapidly.

The transition from a stock of material to a stock of time is particularly important.

For example, two facilities each operating at 70% capacity may be strategically more useful than one facility operating at 100% if the former configuration provides route independence and the ability to scale rapidly.

Framework — Turn Time into an Industrial System

Horizon: 2–7 years.

This is where the alternative network itself is built.

But the analysis showed that simply creating several plants is insufficient.

Several independent nodes have to be covered:

  1. raw material sources;
  2. processing;
  3. equipment;
  4. technology;
  5. skills and industrial knowledge;
  6. recycling, substitution, and redesign.

The fourth and fifth points are particularly important.

A plant can physically exist without the country possessing:

  • its own technology;
  • documentation;
  • pilot production capability;
  • qualified engineers;
  • domestic manufacturers of critical equipment.

In that case, the dependency has simply moved to another level.

Therefore, a genuinely independent chain must be independent not only in tonnage.

Autonomy — Change the Structure of Dependency

Horizon: 5–15+ years.

This is no longer an emergency import-substitution program.

The objective is to make the restoration of the original dependency structurally more difficult.

This includes:

  • domestic equipment manufacturers;
  • independent processing technologies;
  • materials science;
  • lower material intensity;
  • substitution of scarce materials;
  • new technological routes;
  • an independent scientific base.

Autonomy does not replace the first two tracks.

One cannot simply say: "In ten years we will become autonomous, so nothing is needed now."

The system has to survive potential attacks before then.

Therefore:

Shield buys time.
Framework turns time into industry.
Autonomy gradually removes the structural cause of dependency.

This interaction is one of the most important parts of the resulting architecture.

7. How to Know Whether the System Is Working

Another question then appeared.

How should such a system be managed?

It is too easy to produce an attractive strategy that cannot be measured.

Therefore, the state of the production chain can be described through several variables:

  • S — source diversification;
  • P — processing capacity;
  • E — equipment independence;
  • T — technological independence;
  • W — workforce and industrial knowledge;
  • C — capital resilience;
  • D — guaranteed demand;
  • τ — Time-to-Replace.

Resilience therefore becomes a system state rather than a slogan.

Thresholds can be defined.

It becomes possible to see that a supply chain has become vulnerable not after a plant stops, but much earlier — for example, when one processing route disappears, critical equipment becomes dependent on a single source, or TtR begins to increase.

This makes it possible to move from normal operation to corrective industrial policy before a full crisis emerges.

In the extreme case, it can trigger a strategic support regime.

This begins to resemble not a declaration of industrial policy, but a system for managing the state of an industrial network.

8. Why the Goal Should Not Be "Independence from China"

There is an important political-economic nuance here.

Initially, it is tempting to formulate the task as:

"Make the United States independent of China."

But this creates the wrong architecture.

If one dominant center is simply replaced by another, the underlying problem has not been fundamentally solved.

The objective should therefore be formulated differently:

Create a multipolar system in which no single participant can economically stop the others.

This principle applies to the United States as well.

For this reason, the ATCM proposal also included the principle that allies should not become dependent on a single center — even if that center is the United States itself.

The resulting architecture is therefore not:

"Replace China with America."

It is:

Create a network of mutually substitutable production nodes.

9. What This Produces for the United States

The initial question was relatively narrow:

How can alternative critical-mineral processing remain economically viable?

The resulting solution was considerably broader.

The architecture became:

economic mechanism → measurable parameter → protection against attacks → time reserve → industrial network → technological autonomy.

SCG addresses the economic problem.

TtR provides measurability.

Shield protects the transition period.

Framework creates the alternative production network.

Autonomy reduces structural dependency.

The variables S/P/E/T/W/C/D/τ provide a way to monitor the state of the system.

This is why I do not view SCG simply as another form of government subsidy.

Its purpose is different:

To turn strategic resilience from a side effect of industrial policy into a separate economic object that can be defined, measured, purchased, and verified.

10. Then Came the Russian Question

After developing this architecture, a natural question emerged:

Is it specific to the United States?

Obviously not.

The United States was a concrete case because the critical-minerals problem made the mechanism particularly visible.

But the structure of the problem is much broader.

It appears wherever:

  • a critical production chain exists;
  • one or more links are controlled by a limited number of players;
  • alternative capacity is more expensive;
  • the market does not pay for its reserve value;
  • restoring capacity takes years.

This is therefore not only a minerals problem.

It is a general class of production systems with critical dependencies.

This creates a possible application to Russia.

11. What Changes When the Method Is Applied to Russia

Russia does not need to copy American policy.

In fact, directly copying it would miss the point.

What can be transferred is the method of thinking, not a particular institution.

Instead of asking:

"What should we import-substitute?"

it may be more useful to ask:

"Which production chains must not have a single point of failure?"

This is a substantially broader criterion.

Moreover, the dependency does not have to be external.

It may exist entirely within the country.

For example:

  • a single equipment manufacturer;
  • a single technological route;
  • a single supplier of a specialized reagent;
  • a single group of qualified specialists;
  • a single competence center;
  • a single manufacturer of a critical component.

Formally, such a chain may be called domestic.

Systemically, it remains fragile.

12. The Russian Version: Guarantee Production Capability, Not Only Tons

SCG can be adapted to this context.

But the guaranteed object does not necessarily have to be an export mineral.

It can be a:

critical processing or intermediate production capability.

For example, the state or major industrial consumers could guarantee not merely the purchase of a product, but the continued existence of:

  • a defined production capacity;
  • a defined readiness level;
  • a second technological route;
  • reserve equipment;
  • qualified personnel;
  • necessary technical documentation;
  • the ability to restore production within a defined time.

In this case, the state pays not only for the product.

It pays for the preservation of the system's production capability.

This is closer to actual technological sovereignty than a simple indicator such as "domestic production share."

13. TtR for Russia

In my view, Time-to-Replace may be particularly useful here.

Instead of the declaration:

"This critical technology is import-independent."

a much more practical question can be asked:

"If the primary supplier disappeared tomorrow, how long would it take to restore the required production?"

The answer may be:

  • a few days;
  • a few months;
  • several years;
  • or "impossible without creating a new industrial chain."

These answers characterize the actual level of production sovereignty very differently.

Moreover, TtR makes hidden dependency visible.

Suppose a Russian enterprise is fully domestic in terms of raw materials and final product.

But the only machine capable of performing a critical technological operation is produced abroad, and replacing it takes four years.

Then the TtR of the entire chain is determined not by Russian raw materials.

It is determined by that machine.

This method makes it possible to move from industry statistics to systems engineering of industrial production.

14. Russian Shield, Framework, and Autonomy

The same three-level logic can be transferred almost directly.

Shield

Create time:

  • stockpiles;
  • reserve capacity;
  • second production routes;
  • emergency equipment;
  • guaranteed demand;
  • the ability to start production rapidly.

The purpose is to prevent the critical chain from stopping while a deeper solution is being built.

Framework

Create an industrial network:

  • multiple raw-material sources;
  • multiple technological routes;
  • domestic equipment;
  • domestic technologies;
  • reproduction of skills;
  • accumulation of industrial knowledge;
  • recycling;
  • substitution;
  • redesign.

Autonomy

Change the structure of dependency:

  • create independent technological schools;
  • develop equipment manufacturers;
  • reduce material intensity;
  • develop alternative materials;
  • preserve technological competencies independently of any particular enterprise.

Therefore, the Russian version does not reduce to the slogan "import-substitute everything."

Its objective is different:

Prevent the failure of a single external or internal center from stopping a critical production chain.

15. The Most Interesting Consequence

During this work, I arrived at a conclusion that goes far beyond critical minerals.

Industrial resilience is not the amount of product produced.

It is not even the number of plants.

It is the ability of the system to:

  1. detect the loss of a critical node;
  2. maintain operation during the transition period;
  3. activate an alternative route;
  4. scale it;
  5. restore the required competencies;
  6. prevent external pressure from economically destroying reserve capacity.

In other words, resilience is a dynamic property of a network.

This means it can be engineered much like a fault-tolerant technical system.

Instead of asking only:

"Is the system working now?"

ask:

"What happens if this node disappears?"

And then:

"How long will it take for the system to restore the function?"

16. From Minerals to General Engineering of Production Systems

This is why the original problem of critical minerals became more interesting to me than the specific subject itself.

We started with the raw-material market.

Then we moved to processing.

From processing to the economics of reserve capacity.

From there to measuring replacement time.

From there to modeling attacks.

Then to a three-track architecture.

And finally to a more general principle:

A critical production chain should be designed not only for normal operation, but also for deliberate pressure on its economic, technological, and organizational vulnerabilities.

This is no longer only industrial policy.

It is engineering of production-system resilience.

And this is why an architecture developed for the United States may potentially apply to completely different problems.

Not because Russia needs the American SCG.

But because the same systems question appears in any country:

What capability must not be allowed to disappear from the system, even if it is economically redundant at the moment?

17. Conclusion

The most important result of this work for me was not the Strategic Capacity Guarantee itself.

The main result was a change in how the problem is formulated.

Instead of:

"How do we make an alternative producer competitive?"

the question becomes:

"How do we preserve the ability of the system to replace a critical node when that ability is not yet needed by the market?"

From there, engineering questions follow:

  • How much capacity is required?
  • Where should it be located?
  • Which links are actually independent?
  • How long does replacement take?
  • Which attacks can destroy the reserve?
  • Which parameters must be continuously measured?
  • Who should pay, and for what?

For the United States, this led to the Strategic Capacity Guarantee proposal and the Shield → Framework → Autonomy architecture.

For Russia, the same approach may be applied to different critical production chains — not as a copy of the American solution, but as a transfer of the underlying engineering logic.

A strong industrial system is not a system in which nothing ever breaks.

It is a system in which it is known in advance what will be done if a critical link breaks.

Related publications

SMR + Data Centers: Rosatom’s RITM-200 as a New Market Opportunity
Production Knowledge as the Foundation for AI in Manufacturing Industry
A Memorandum to the World: Technology Must Not Outrun Humanity