If you buy a conveyor belt scraper blade Australia-wide with a tungsten carbide edge on it, the last eighteen months have been difficult to budget against. Tungsten went vertical through the first half of 2026, then the Chinese price fell back hard while the price an Australian buyer actually pays did not follow it down. Those two facts look contradictory and they are not. This post sets out what moved, why your invoice did not move with it, and which cleaner positions justify a carbide edge once you price it honestly.
01What the tungsten price actually did
Ammonium paratungstate, APT, is the traded intermediate that sits between tungsten concentrate and the carbide powder in your blade tip. It is the number worth watching, and through 2026 it has told two different stories depending on which market you read.
| Benchmark | 2026 high | Most recent | Direction |
|---|---|---|---|
| APT, China domestic | approx. US$131,000/t (8 May 2026) | approx. US$72,400/t (30 Sep 2026) | Down roughly 45% from the May high |
| APT, China domestic (RMB) | RMB 480,000 to 510,000/t (12 Oct 2026) | Back near late November 2025 levels | |
| APT, CIF Rotterdam 88.5% | approx. US$3,100/mtu (late Jun 2026) | US$2,500 to 3,100/mtu (15 Oct 2026) | Broadly held. Long-term contracts above US$3,000/mtu |
The China Tungsten Industry Association reported on 12 October that major tungsten products have come off close to 70% from their highs across roughly seven months, putting them back near late November 2025 levels. It attributes that to weak downstream demand, carbide makers running reduced rates, and smelters sitting in cost-price inversion and halting production rather than selling at a loss.
Over the same window the European assessment barely moved. On 15 October the weekly CIF Rotterdam range was US$2,500 to 3,100 per metric tonne unit, with spot trading close to halted and long-term contracts still settling above US$3,000. For reference, that benchmark spent years in the low hundreds before the current cycle.
02Why the Chinese fall did not reach your invoice
A falling mine-gate price only reaches you if the material can leave the country. Through this cycle it largely could not.
China added tungsten to its export control list and cut mining quotas in 2025. In December 2025 it confirmed that only fifteen firms would be licensed to export tungsten across 2026 and 2027. In February 2026 it prohibited dual-use exports to twenty Japanese entities. Supply inside China and supply available to a buyer in Mackay or Perth became two separate things.
So the binding constraint on your price is the export licence, not the ore body. When Chinese smelters cut rates because domestic prices fell below cost, the volume reaching export markets tightened further, which is how you get a collapsing domestic price and a firm export price in the same month.
Demand is not helping either. Project Blue puts defence at roughly 12% of tungsten consumption, heading toward about 15% across 2027 and 2028 as stockpiles are rebuilt, and Argus has defence demand growing around 8% a year. That volume is price-insensitive in a way conveyor maintenance budgets are not. The United States has no operating commercial tungsten mine, and the new Western supply coming on, including Almonty's South Korean operation which started mining in March 2026, is measured in years rather than quarters.
That announcement matters to a maintenance budget more than it first appears. Tungsten has had an active spot market and no futures market, which is why carbide pricing has arrived at your desk as a surcharge letter rather than a hedgeable line. Global production runs roughly 85,000 to 95,000 tonnes a year against a market worth more than US$20 billion at current spot levels. Once a cleared contract exists, your carbide suppliers gain a hedging instrument, and over time you should expect pricing to be quoted with more structure and less improvisation.
03What the carbide is doing on a cleaner blade
Tungsten carbide is on a cleaner blade for one property: it resists abrasion at a hardness no polyurethane reaches. On a secondary cleaner running against an already-cleaned belt, where the duty is grinding off a thin bonded film and the wear mechanism is straightforward abrasion, that hardness is the right answer and a tungsten carbide conveyor blade will outlast a polymer edge by a wide margin.
The same hardness is a liability under impact. Carbide is hard and brittle. It resists sliding abrasion and it chips when something strikes it edge-on. The thing that strikes it edge-on, on a great many Australian conveyors, is a mechanical splice.
This is why FM8 does not offer a Knife Tips version for a mechanically spliced belt at all. A row of fasteners presenting to a rigid tip on every pass is the condition a rigid edge cannot survive, and no grade selection fixes it. Underground coal trunk belts sit squarely in that group, because the belt is extended at each longwall move and the splices are mechanical as a consequence. Vulcanised belts present no such step, which is why knife tip blades on primary cleaners are found mostly in iron ore and hard rock rather than coal.
The practical consequence for procurement: splice type decides whether a rigid edge belongs on that conveyor before any discussion of grade, price or supplier. Get that order of operations wrong and you will buy the right grade for the wrong belt.
04Where a carbide edge earns its price
On the duties below, carbide remains worth paying the current price for, because nothing else holds the edge and the alternative is more changeouts than the saving is worth.
- Secondary cleaners on vulcanised belts. No fastener impact, a thin residual film to remove, and a wear mechanism carbide is built for. This is the carbide position.
- High-abrasion ores. Iron ore and hard rock fines abrade a polymer edge quickly enough that blade changeout frequency, and the access and isolation that come with it, dominate the material cost.
- Positions where changeout access is expensive. If getting to the cleaner costs you a shutdown window rather than a shift, edge life is worth more than the blade price. Price the access, not the blade.
- Belt cleaning blade coal terminal reclaim duty, where belts run close to continuously and the cost of a mid-week isolation is measured against shiploading rates rather than labour hours.
In each case the argument is the same. Carbide earns its price where the cost of changing the blade exceeds the premium on the blade. That is an arithmetic question your maintenance system can answer, and it does not change because the APT assessment moved.
05Where you are buying carbide to cover something else
Carbide consumption at the secondary is downstream of how well the primary is doing its job. A primary cleaner that is removing the bulk of the carryback hands the secondary a thin film. A primary that is under-performing hands it a loaded belt, and the secondary then wears at a rate nobody specified.
Primary cleaners under-perform for reasons that have nothing to do with the secondary blade you keep replacing. Contact pressure drifting through blade life is the common one, and we have set the mechanics out separately rather than repeat them here. The symptom procurement sees is a secondary carbide spend that climbs without an obvious cause, often while primary blade spend looks normal.
Two conditions worth checking before you re-order carbide at current prices:
- A carbide secondary fitted where the primary should be doing the work. Specifying up at the secondary to compensate for a primary cleaner that is not holding its pressure is an expensive way to buy a tensioner. The primary cleaner versus secondary cleaner split is an engineering decision, and the carbide bill is where getting it wrong shows up.
- A carbide edge on a mechanically spliced belt. If you are replacing chipped tips rather than worn tips, you are paying tungsten prices for a consumable that is failing by impact. The grade is not the problem.
This is also where the polyurethane versus tungsten carbide blade comparison is usually framed wrongly. The two are not competing for the same position. A correctly specified polyurethane primary reduces how much carbide you buy at the secondary, which at current APT levels is a larger lever than negotiating the carbide price.
06The tool steel middle ground
Between a polymer edge and a carbide edge sits inline tool steel, and the current tungsten market has made it worth a second look on duties where it was previously dismissed as a compromise.
Tool steel holds a working edge through duties that take a polymer edge back quickly, particularly where the load is sticky fines, and it tolerates impact that chips carbide. FM8 Inline Tool Steel Secondary FRAS exists for that band. It is not a substitute for carbide in high-abrasion mineral duty, and it is a sound answer where the failure mode you are actually seeing is chipping rather than abrasive wear.
It also carries a different exposure. A tool steel conveyor blade is not priced off the APT assessment, so a procurement team carrying tungsten volatility across a fleet has a reason to look at which positions genuinely need carbide and which were specified with carbide because that was the convention.
07Field example: a Queensland coal export terminal
Take a coal export terminal running reclaim and shiploading belts, a mix of vulcanised and mechanically spliced, with carbide secondaries fitted across the fleet because that was the standard when the cleaners went in.
Split by splice type and the picture separates. On the vulcanised reclaim belts the carbide secondaries are doing exactly what they were bought for, wearing by abrasion, and the premium is justified by the cost of taking a reclaim belt offline. On the mechanically spliced positions the crew is replacing tips that are chipped rather than worn, at tungsten-linked prices, on a failure mode no grade change addresses.
Separately, two primary positions have been re-tensioned repeatedly through the last blade cycle. The secondaries behind them are consuming carbide faster than the equivalent positions elsewhere on the terminal, because they are handling carryback the primary should have taken off. Nothing on the secondary blade purchase order shows that.
The carbide spend on that terminal is three separate problems wearing one line item. Only one of them is a tungsten price problem.
08What to settle before the next carbide order
None of this needs a lab or a consultant. It needs your maintenance records and a walk down the belt line.
- Sort every carbide position by splice type. Mechanically spliced belts with rigid edges are the first thing to question, whatever the grade.
- Separate chipped tips from worn tips in your changeout records. They are different failures with different fixes, and only one of them is about material cost.
- Check the primary ahead of each heavy carbide consumer. Pull the re-tension count per blade cycle. A position that gets touched repeatedly is feeding its secondary more than it should.
- Price the access, not the blade. A position that costs a shutdown window to reach justifies carbide at a price a position reachable in a shift does not.
- Ask your supplier what their carbide price is indexed to, and whether that changes once a cleared futures contract exists from early 2027.
On the record-keeping, a belt cleaner blade replacement interval you can forecast is what turns a volatile input price into a manageable one. Blade wear is judged visually at the conveyor, and two techs looking at the same blade on the same shift will commonly return two different numbers, which is poor ground on which to build a twelve month carbide budget. The FM8 Conveyor Cleaner Health app narrows that spread by estimating percent wear from a phone photo taken in place for the tech to confirm, then rolls the wear curves per blade position into a reorder-by date and a twelve month consumption schedule. With tungsten pricing where it is, knowing what you will consume is worth as much as knowing what it costs.
Conveyor scraper blade supplier Australia: what to expect from pricing in 2027
A cleared tungsten futures contract settling against the Rotterdam and Baltimore APT assessments gives the carbide supply chain its first hedging instrument. Over the next few years that should make carbide blade pricing more legible, in the sense that a supplier quoting a twelve month price will be able to say what sits behind it. It will not make tungsten cheap. Export licensing and defence demand are structural, and the Western assessment has held through a 45% fall in the Chinese domestic price, which tells you how little the two are currently connected.
The planning conclusion for an Australian bulk handling operation is to treat carbide as a deliberate specification on the positions that need it rather than a default across the fleet, and to make sure the polyurethane primary ahead of it is doing the work it was specified to do.
For procurement
Your carbide line is carrying three different problems. Split it by splice type and by failure mode before you negotiate price, because two of the three do not respond to a better unit rate. Ask what the quoted carbide price is indexed to, and revisit that once the CME contract trades from early 2027.
For the maintenance manager
If you are replacing chipped tips rather than worn tips, the grade is not the issue and a price negotiation will not help. Check splice type first. Then check the re-tension count on the primary ahead of your heaviest carbide consumers.
For the engineering manager
Carbide is specified for abrasion resistance and pays for it in impact tolerance. Document which positions were chosen on that trade-off and which inherited it from a fleet standard, because the second group is where the current tungsten market is costing you without buying anything.
Specify carbide where it earns its price
Send FM8 your cleaner positions, belt splice types and current changeout records. We will tell you which positions justify a carbide edge, which are failing by impact rather than abrasion, and where a primary cleaner is quietly setting your carbide spend.
Contact FM8