Editorial Article
Nobody Can Measure the Wear Layer, and That Is What Caps This Market at USD 66 Million
The mechanism is proven. The measurement is not.
More than six hundred applications, the longest-serving liner in service for over ten years, and two independent designs from two companies that have been at this since the 1970s and since 2005 respectively. As engineering, the self-renewing wear layer in a magnetic mill liner is not in doubt. A permanent magnet holds the liner to the shell and simultaneously holds the ferromagnetic fraction of the mill charge against its face. That accumulated layer — magnetite out of the ore, or steel abraded off the grinding media — becomes the surface that wears instead of the casting, and it replenishes itself for as long as the mill keeps supplying material the magnet can hold.
What is in doubt, and has been for four decades, is the number the customer actually needs. There is no standard in-situ method for measuring the thickness of that protection layer on a running mill. Relines are scheduled on observation, historical pattern and production risk. Not on measurement.
That single gap is the binding constraint on this market, and it costs more than any of the technical problems the category normally talks about.
An unmeasurable wear surface is a pricing problem, not an engineering one
Mill availability, not liner price, is the buying criterion in this category. A relining stops the mill, and if the mill is the bottleneck in the circuit the cost of the stop is measured in tonnes of concentrate not made. So the question a maintenance superintendent or a concentrator manager actually has to answer comes in exactly one form: how long until the next reline, and how sure are you?
Without a measurement, a supplier can describe a mechanism and cite a reference mill. It cannot warrant an interval. And an interval is the only form of the claim that enters a shutdown calendar, a maintenance budget or a plant manager’s performance measure. A mechanism enters a brochure.
That asymmetry is what makes the premium fragile at the point of decision, and it is routinely misread as a marketing problem. Cast steel and rubber are not inferior products. They are the incumbent: a known wear curve, a local supplier who answers the phone, and thirty years of collective experience in the crew that will install them. A buyer moving to a premium wear part is trading a known distribution for an unknown one, and the price of that unknown is precisely the interval uncertainty. No amount of further design work closes it. Only measurement does.
It is worth being concrete about where the uncertainty bites, because it is not only in the purchase order. A liner set is designed for one shell. It cannot be pooled across mills, it cannot be sold to a different customer and it cannot be produced to stock. Lead time from design freeze to delivery runs to a quarter or more once freight is included, and the design freeze depends on a mill survey the plant has to schedule. Every stage sits inside the customer’s shutdown planning horizon. Against that, a warranty on an interval is not a nice-to-have; it is the thing that lets a plant commit a production calendar to a wear part.
What a warrantable interval is worth, against the other two drivers
Our base case has magnetic liners moving from about 17 per cent of the eligible pool in 2025 to roughly 24 per cent in 2033, which is where the USD 66.0 million and the 7.23 per cent compound growth come from. The growth decomposes into three drivers that move at very different speeds, and putting the measurement gap next to them is the whole argument.
| Driver | Contribution, 2025-2033 | What actually moves it |
|---|---|---|
| Milled volume | about 2.4% a year | Ore supply and head grades. Falling grades force more tonnes through the mill for the same metal, so it grows slowly and it is nobody’s decision. |
| Penetration of the eligible pool | about 3.8% a year | 17% to roughly 24% of the pool: price band widening, account-by-account extension from a first mill to a second and third, and whatever makes a buyer willing to switch. |
| Realised price | about 1.0% a year | USD 11,988 to USD 13,147 a tonne, against volume from 3,150 to 5,020 tonnes. Magnet content and alloy set this. It is cost pass-through, not a lever. |
Now price the measurement gap against those three lines. A full point of realised price is worth 1.0 per cent a year and is set by magnet and alloy cost, which is to say it is not available to be won. Milled volume is not available either. Penetration is the only driver a supplier can move, and it is the largest of the three.
The report’s own falsification test puts a number on how much it can move. If a credible in-situ measurement of the protection layer existed, a supplier could warrant a relining interval instead of describing a mechanism, and the 2033 figure would move above USD 80 million. On our pool arithmetic — USD 66.0 million at 24 per cent penetration implies an eligible pool of about USD 275 million in 2033 — USD 80 million is roughly 29 per cent of that pool. The difference between a mechanism and a warranty is therefore about five points of pool penetration, and it is larger than anything available on price, volume or geography. It sits inside a single product decision.
The converse is worth stating too, because it explains why the category has stayed small for so long without anyone being at fault. Every supplier in it has been selling a real technical benefit through an unmeasurable claim, to buyers whose own performance is measured on availability. That is a hard trade even when the product works.
The other ceilings are boundaries, not bugs
It matters that measurement is the constraint, because the remaining limits in this category are not the kind that a roadmap can close.
- SAG duty. Impact-dominated grinding with a coarse charge scrubs and disrupts the accumulated layer. The category has stayed in secondary and regrind ball mills for a reason and no evidence we reviewed suggests that has changed.
- Non-magnetic ore with low chip generation. No design fix exists, because the mechanism has nothing to hold. This is a mineralogy fact, and it is the reason the eligible pool is roughly 28 per cent of the USD 793 million ball-mill liner pool rather than all of it.
- Undocumented mills. A large share of the world’s older ball mills have no reliable shell drawings. That is a documentary ceiling rather than a technical one, and it caps who can be served economically rather than what can be sold.
- Quality consistency across the long tail. Magnet grade, casing alloy and dimensional tolerance vary widely between suppliers selling near-identical product descriptions. That is a market-structure fact, not something a single supplier corrects.
Three of those four are geological or documentary, which is why the addressable market in this category is a mineralogy map rather than a mining map, and why the forecast stops at roughly 24 per cent penetration of the eligible pool by 2033 instead of extrapolating further. The realistic technology read is a proven, mature mechanism with a metallurgical ceiling, where the next gains come from materials and measurement rather than from new physics.
Where the engineering money goes, in that order
- Measurement first. An in-situ method, even a coarse one, converts a technical story into a financial one and moves the only driver that is actually available. It is also the least like the industry’s existing engineering: it is closer to instrumentation than to foundry practice.
- Magnet design without heavy rare earths. Announcement No. 18 of 2025, issued on 4 April 2025 by China’s Ministry of Commerce and the General Administration of Customs, imposed export control on a defined list of medium and heavy rare earth items. The samarium schedule names samarium-cobalt permanent magnet materials outright; the dysprosium schedule covers dysprosium-containing NdFeB permanent magnet materials, including magnets and magnetic powders, with exporters required to apply to the competent commercial authority for a licence. Both viable magnet families in a mill liner are therefore inside a licensing regime. The consequence is not a cut-off. It is licence lead time on a component delivered against a customer’s shutdown date, documentation cost on a product selling at about USD 12,000 a tonne, and a shift in the design incentive: a heavy-rare-earth-free magnet with more volume and different geometry becomes commercially rational even at higher material cost, because it takes a licence step off the critical path. Magnet material is 24 to 30 per cent of delivered cost, so this compresses margin and reliability rather than volume.
- Materials and duty expansion. The remaining gains are metallurgical — casing hardness, layer adhesion, behaviour under impact — and that is now a competitive variable rather than a background cost, in a way it was not a decade ago when the physics was the differentiator.
Read together, that ordering says something unflattering about where the category’s technical effort has gone. The most valuable engineering problem in magnetic mill liners is not a liner problem. It is the absence of an instrument.
What this analysis does not claim
The USD 80 million figure is the report’s falsification test, not a forecast. The 29 per cent pool share derived from it, and the USD 275 million pool behind it, are our arithmetic on a 24 per cent penetration assumption rather than observed values. The 17 per cent base penetration and the 28 per cent eligibility share are Pheonix estimates and are labelled as such in the report.
On method, this piece follows our two-layer approach at roughly 70 per cent secondary and 30 per cent primary. The mechanisms — why an interval is budgetable and a mechanism is not, why interval uncertainty prices a premium wear part, why the unmeasurable claim has held the category at a fifth of its pool — come from the primary layer behind this edition: our own bottom-up mill-level reconstruction of installed grinding capacity and liner consumption by region and ore type, open-ended interviews with mill maintenance and supply-chain personnel in the secondary and regrind segment, a structured review of comminution conference material, and a willingness-to-pay assessment built around liner replacement intervals rather than liner price. That layer establishes how the market works. It never carries a number. Every figure here is from a supplier document, a public filing, a regulator or our own model, and no third-party market-research publication is used or cited.
The full technology, risk and regulatory chapters, with the penetration and price series behind this piece, are in Global Magnetic Mill Liners Market, 2026-2033.
Companion reading
The commercial half of the same argument is in Twenty-Six Per Cent of This Market Sets the Price the Other Seventy-Four Per Cent Has to Live With, which sets out why a price-led long tail does not take customers so much as decide which mills are worth serving. The operating context these mills sit in is the Mining segment, and the tonnage that becomes grinding demand is covered in our lithium, nickel, cobalt, graphite and manganese reports.
Frequently Asked Questions
Why do magnetic liners hold only about 17 per cent of the pool they can technically serve?
Because the wear layer cannot be measured on a running mill. There is no standard in-situ method for its thickness, so relines are scheduled on observation, historical pattern and production risk. A supplier can therefore describe a mechanism and cite a reference mill, but cannot warrant an interval — and an interval is the only form of the claim that enters a shutdown calendar or a maintenance budget. The premium then rests on an unmeasurable claim, which is a hard trade against cast steel and rubber with a known wear curve.
What would a warrantable relining interval be worth?
On the report's falsification test, a credible in-situ measurement would push the 2033 figure above USD 80 million instead of USD 66.0 million. On our pool arithmetic — USD 66.0 million at 24 per cent penetration implies an eligible pool of about USD 275 million in 2033 — USD 80 million is roughly 29 per cent of the pool. That is about five points of penetration, more than any available move on price (about 1 per cent a year) or milled volume (about 2.4 per cent a year).
Do rare earth export controls limit supply?
It raises lead time and compliance cost rather than cutting supply. Announcement No. 18 of 2025, issued on 4 April 2025, names samarium-cobalt permanent magnet materials outright and covers dysprosium-containing NdFeB materials, including magnets and magnetic powders, with a licence application required. Both viable magnet families in a mill liner are therefore inside the regime. Magnet material is 24 to 30 per cent of delivered cost, so the exposure compresses margin and reliability rather than volume.
Where can a magnetic mill liner not be used?
Two of the limits are physical and one is documentary. Impact-dominated SAG grinding scrubs and disrupts the accumulated layer; ore that is non-magnetic and generates few ball chips gives the magnet nothing to hold; and a large share of older ball mills have no reliable shell drawings, which caps who can be served economically. None of those has a design fix, which is why the forecast stops at roughly 24 per cent of the eligible pool in 2033.
