Chrome-Magnesia Bricks: Extending Campaign Life in RH Degassers

Wear mechanisms and the Cr₂O₃ answer

Where the 30–50% campaign gain actually comes from

To understand why chrome-magnesia extends RH campaigns, start with what the vessel is for. The RH (Ruhrstahl–Heraeus) degasser removes hydrogen from molten steel by exposing a circulating stream to vacuum. Hydrogen in solution must come down below 2 ppm for plate and pipe grades that cannot be hydrogen-annealed later. The same vessel handles vacuum decarburisation for ultra-low-carbon steel — blowing oxygen onto the steel surface under vacuum to take carbon below 30 ppm — and then vacuum decarburised steel is aluminium-killed in the same cycle, which drops a fresh charge of Al₂O₃ into the slag. One lining, three different chemical environments, several times per heat, hundreds of heats per campaign.

The lining duty list reads like a deliberate assault on a refractory. Cyclic vacuum: the vessel swings from atmosphere to roughly 0.67 mbar several times per heat, so the hot face breathes thermally and mechanically. Oxygen lancing: a localised hot spot with direct FeO generation at the impact zone. FeO-rich slag: every kilo of oxygen that oxidises iron ends up sitting on the hot face, and FeO is the classic flux for magnesia-carbon. Argon stirring: bath turbulence erodes the snorkel and lower vessel. Magnesia-carbon, the workhorse of the converter, fails here precisely because its carbon bond is consumed — under vacuum the carbon oxidises against FeO, the bond burns out, and the magnesia grains wash away. The brick does not wear; it disintegrates.

Chrome-magnesia survives by abandoning the carbon bond entirely. The working microstructure is a periclase (MgO) network interlocked with chrome spinel (MgCr₂O₄) grains, bonded directly through solid-state sintering at brick-firing temperatures above 1,600 °C. There is nothing to burn. The spinel phase performs the load-bearing chemistry: it absorbs FeO into solid solution rather than forming a low-melting eutectic with it. Magnesia alone is attacked by FeO through magnesio-wustite solid solution that eventually saturates and slumps; spinel buffers that uptake, holding iron oxide in a refractory phase while the periclase network stays rigid. This is not corrosion resistance in the coating sense — it is a phase chemistry that keeps eating the aggressive slag component and stays solid while doing it.

The 30–50% figure is not a laboratory extrapolation; it is what steelmakers log when they switch a vessel from carbon-bonded or older chrome-magnesia generations to modern direct-bonded, fine-grained chrome-magnesia. A converter-class lining might deliver 100–150 heats in an RH duty; direct-bonded chrome-magnesia in the same vessel reports 200–400 heats depending on grade, lancing practice and slag control. Campaign extension of 30–50% against the previous lining generation is the range that shows up again and again in reline reviews — sometimes from the brick alone, sometimes from the brick plus better lancing discipline, but the brick is the prerequisite. You cannot lance your way to a 50% longer campaign on a lining that is losing its bond every time the vacuum pulls.

Chemistry windows matter. RH-grade chrome-magnesia typically carries 12–25% Cr₂O₃ depending on zone: the hot face and lance-impact areas sit at the top of that window, the snorkel and lower vessel lower. This is where the raw material conversation begins, because every percentage point of Cr₂O₃ in the brick starts as chrome oxide green powder with a specification: assay, particle size, and — increasingly — a Cr(VI) limit that protects the crew relining the vessel.

The grade levers: what the brickmaker controls

Chrome oxide green for refractories is a calcined, crushed and milled material, not a pigment. The brickmaker's flow is: weigh the Cr₂O₃ powder against dead-burned magnesia and chrome ore fines, dry-mix, press into bricks, and fire in a tunnel kiln above 1,650 °C so the periclase–spinel direct bonds form. Three raw-material properties decide whether that kiln produces a dense, hot-face-capable brick or a soft one: the Cr₂O₃ assay, the particle size distribution, and the Cr(VI) content of the shipped powder.

Assay is straightforward but unforgiving. Refractory-grade chrome oxide green should assay ≥99% Cr₂O₃, because the balance — silica, iron, alkalis — does not disappear at the kiln. Silica in the mix reacts with magnesia to form forsterite, a phase with a melting point roughly 300 °C below periclase-spinel and a known weak point under FeO attack. A cheap oxide with 2% silica is not 2% dilution; it is 2% of a lower-melting phase planted exactly where the hot face will fail. We ship refractory grades at ≥99% for that reason.

Particle size decides pressed density. Bricks are pressed green at high tonnage, and the powder must flow into a uniform compact. Typical windows: fine hot-face grades at D50 of 2–5 μm for maximum packing and a tight fired surface; standard grades at 4–8 μm where press throughput and cost matter; repair and gunning mixes coarser at 6–10 μm so the material conveys and rebounds acceptably. Too fine across the whole distribution hurts flowability and press life; too coarse leaves porosity that FeO walks straight into. The D50 is a window to match to the pressing duty, not a number to minimise.

Cr(VI) is the lever that changed the market. Reline crews work inside the vessel with jackhammers and cutters, and hexavalent chromium is a regulated exposure risk in every major steelmaking jurisdiction. Our low-Cr(VI) tier — G-5891, G-5899, G-5996, G-5070 — ships at ≤5 ppm Cr(VI) with the same spinel chemistry and the same density behaviour as the regular tier, because Cr(VI) is controlled in the manufacturing process, not by diluting the product. Brickmakers now write the ≤5 ppm number into their purchase specs because their own customers write it into the reline contract.

For context, the physical constants behind the material: Mohs hardness 8.5–9 (the hot face resists lance erosion), true density 5.21 g/cm³ (pressed brick densities of 3.1–3.3 g/cm³ are realistic targets), melting point 2,435 °C, and decomposition beginning near 2,266 °C. Steelmaking runs at 1,600–1,700 °C. The material has over 500 °C of headroom even at the lance hot spot — this is why chrome-magnesia, unlike most refractories, is not running near its own failure temperature.

Specifying and testing before the reline

A reline is a fixed cost event measured in days of lost production. The difference between a good campaign and a failed one is usually decided months earlier, in the purchase order. The spec that protects a steelmaker is short and testable: Cr₂O₃ assay ≥99%, Fe₂O₃ and SiO₂ limits stated, D50 and D90 by laser diffraction, moisture below 0.5%, a stated Cr(VI) ceiling for the grades that go into confined-space work, and a certificate of analysis for every lot. Everything else — color, origin stories, marketing — is decoration.

Incoming QC should not trust the COA. A composite sample from every incoming lot goes to XRF for the assay, laser diffraction for the particle size, and a diphenylcarbazide spectrophotometric test for Cr(VI) on the grades where that number is contractual. Sieve residue at 45 μm catches the coarse tail that will show up as open porosity in the pressed brick. These three measurements — chemistry, size, hexavalent chromium — take a shift, not a week, and they are the difference between catching a bad lot at the warehouse door and catching it in a shortened campaign.

Particle size deserves a second look because it is the property buyers check least and regret most. The brick's fired density and its resistance to FeO penetration both trace back to how the powder packed in the press. A supplier that certifies only the D50 is giving you one point on a curve; ask for the D10 and D90 as well, because a wide distribution with fines in the sub-micron tail packs differently than a narrow one, and a coarse tail opens channels. When a lining underperforms and the brickmaker's plant trials all passed, the first place we look is the particle size history of the oxide lots that went into those bricks.

Sample discipline matters as much as the test methods. Chrome oxide green ships in 25 kg bags or drums; a single grab sample from the top of one drum will not represent the lot. Composite sampling across drums, a retained sample per lot, and a lot number printed on every bag close the loop. The lot number is the traceability unit: if a campaign fails early, the question the steelmaker asks is not "was the oxide bad?" but "which lot went into that brick batch?" — and the answer has to be recoverable from the pallet ticket, not from memory.

FAQ

Why not use magnesia-carbon in the RH degasser? Magnesia-carbon works where carbon is protected, and the RH vessel is exactly where it is not. The cyclic vacuum and oxygen lancing burn the carbon bond out, and the magnesia grains wash away into an FeO-rich slag. Chrome-magnesia carries no carbon bond to lose — the periclase-spinel microstructure is bonded by sintering, and FeO is absorbed into the spinel phase instead of fluxing the bond.

How much Cr₂O₃ is in a typical RH-grade chrome-magnesia brick? Between 12% and 25% depending on zone. Hot-face and lance-impact bricks sit at the top of that window; snorkel and upper-vessel brick lower. The Cr₂O₃ arrives as chrome oxide green powder — usually ≥99% assay — which the brickmaker blends with dead-burned magnesia and chrome ore fines before pressing and firing.

What D50 should I specify for refractory-grade chrome oxide green? Match the D50 window to the duty: 2–5 μm for fine hot-face grades where fired density is everything, 4–8 μm for standard pressing, 6–10 μm for gunning and repair mixes. Also specify D10 and D90, not just D50 — the tails of the distribution decide how the powder packs in the press.

Does the low-Cr(VI) grade perform the same as the regular tier? Yes. Cr(VI) is controlled in the manufacturing process, not by diluting the product, so the ≤5 ppm tier (G-5891, G-5899, G-5996, G-5070) carries identical spinel chemistry and density behaviour. The difference shows up in the reline crew's exposure monitoring, not in the lining's heat count.

How is a 30–50% campaign extension verified? Against the vessel log, not the brochure. Record heats per campaign and relines per year before and after the lining change, and normalise for lancing practice and slag control. The extension comes from the combination of brick grade and practice — but the brick is the prerequisite, because no operating discipline rescues a lining whose bond burns out every vacuum cycle.