Technical article
Thermocouple Types S, R, B and K in Steelmaking: Choosing by Bath Temperature
Which thermocouple type suits furnace, ladle and tundish, why type K has no place in the bath, and the contact block and tube choices that come with the type.
The question behind the question
"Which thermocouple type should we use?" is usually asked when a shop is buying disposable thermocouple tips for the first time, or when readings have started drifting and somebody suspects the type. Both are the right moment to ask, and the answer is shorter than the surrounding confusion suggests.
Type is fixed by two things: the temperature you are measuring, and the instrument that will receive the signal. Everything else — tube length, contact block, immersion practice — follows after.
The four types, and which belong in a steel bath
Type S (Pt–Rh 10% / Pt). The workhorse of steelmaking. Platinum–rhodium against platinum, usable to around 1650 °C in our range, stable, and the type most Indian melting shops already have instruments calibrated for. If nothing in your process is unusual, this is the default.
Type R (Pt–Rh 13% / Pt). The same construction with a higher rhodium content. Slightly higher output than S over the working range, which some instruments prefer, and the same practical ceiling of about 1650 °C. In steelmaking the choice between S and R is usually dictated by what the existing instrument expects rather than by the metallurgy.
Type B (Pt–Rh 30% / Pt–Rh 6%). Both legs alloyed, which pushes the usable range to about 1750 °C. The reason to specify B is headroom: if your bath regularly sits near the top of the S and R range, B gives you margin for the superheat excursions that are precisely when a reliable reading matters most. Type B has poor output at low temperatures, which is irrelevant in a bath and the reason it is wrong for anything else.
Type K (nickel–chromium / nickel–alumel). Not a bath thermocouple. Its range tops out far below liquid steel, and immersing a type K sensor in a bath destroys it without giving a usable number. Type K has a genuine place around a melting shop — flue gas, preheaters, dryers, shell and jacket temperatures, ladle preheat monitoring — and none of it is in the steel. If a type K appears in a bath measurement discussion, it is a specification error, not a cost saving.
Matching type to the measurement point
Furnace bath (induction or electric arc). The hottest and most variable measurement, and the one where superheat excursions are routine. S or R normally; B where the bath regularly runs high.
Ladle, before and after treatment. S or R covers it. The temperature here is lower than the furnace and the reading is usually the one tap decisions are made on, so repeatability matters more than range.
Tundish. Lower again, and continuous casting cares about a steady number rather than an extreme one. S or R.
Anything outside the steel. Type K, and a different conversation.
A shop running one type across all three bath positions is making its own life simpler, and that is a legitimate reason to standardise on S or R even where B would technically suit one station better. Mixed types on one rack produce the wrong tip in the operator's hand.
The instrument decides as much as the bath
A tip and a receiver expecting different calibrations produce a confident, wrong number — and nothing flags it. That failure is worse than a failed read, because a failed read gets repeated and a wrong read gets acted on.
Before changing type, confirm what the instrument is calibrated for and whether it can be switched. On many shop-floor units it cannot. That constraint usually settles the S-versus-R question without any metallurgy being involved.
What comes with the type
Specifying a type is about a third of the specification. Three more choices decide whether the tip fits your plant at all.
Contact block, 602 or 604. The block has to match the lance you already own. This is not a performance choice — a tip with the wrong block simply does not fit, and it is the commonest reason a first order is unusable.
Paper tube length, 300–1800 mm. Chosen by vessel geometry and the operator's working position. Too short pushes the operator closer to the bath than they should be; too long is unwieldy and risks the assembly. Tubes are cut to the plant, so give the length you need rather than accepting a standard.
Tube inner diameter, 28.2 or 18.2 mm. Has to suit the lance.
Alongside those, two numbers are worth knowing because they govern immersion practice: a tolerance of ±5 °C, and a response time of 3–5 seconds. The response time is why holding the tip in the bath longer than five seconds adds no accuracy and only risks losing the assembly.
When the type is not the problem
Most "wrong type" suspicions turn out to be something else. Before changing the specification, rule out three things:
- Readings drifting gradually over weeks, same supplier and type. Lance contacts or the instrument. Slag build-up and worn contacts degrade slowly and read as bad tips.
- Readings varying between operators. Immersion practice — depth, dwell time, slag entry. In most shops the spread between the best and worst operator is wider than the spread between any two types.
- Readings running cool and inconsistent. Immersion too close to the lining, where you are reading a boundary layer rather than the bath.
Change the type when the bath temperature genuinely sits at the ceiling of the type in use, or when the instrument requires it. Those are the two real reasons.
What to send with an enquiry
Measurement point and typical bath temperature, including the high end rather than the average. The instrument make and what it is calibrated for. The lance and its contact block. The tube length and inner diameter. Monthly consumption, so cartons arrive in a sensible pack size.
Those five lines get a specification back on the first reply, with the right block and the right tube — which is what actually determines whether the first carton is usable.
Need a grade recommendation? Send your steel grades, casting route and the problem you are solving. WhatsApp +91 94313 42715 or email info@pkindustries.net.