RB1, RB2 and RB3: South African Thermal Coal Grades Explained

South African export thermal coal is sold in three standard grades, and the shorthand RB1, RB2 and RB3 does a lot of work in this market. The letters stand for Richards Bay, the port that anchors the trade, and the numbers separate what are effectively three different products aimed at three different sets of power stations.

The specification that separates them

The dominant variable is calorific value, measured in kilocalories per kilogram on a net as received basis. RB1 is the premium export grade at roughly 6,000 kcal/kg NAR with the lowest ash content. RB2 sits below it, and RB3 lower again, typically in the region of 5,500 kcal/kg NAR, with correspondingly higher ash.

Sulphur runs low across all three by international standards, generally under one percent, which is one of the enduring commercial advantages of South African coal in markets with emissions limits. Volatile matter, moisture and sizing are also specified, and a full certificate of analysis will carry all of them.

Why the lower grades exist at all

It is tempting to read RB3 as inferior coal, and for a buyer running a plant designed for 6,000 kcal material it is. But a very large part of the world's installed generation fleet, particularly in South Asia, was built around lower calorific value coal and runs it more efficiently than it would run a premium grade. For those buyers RB3 is not a compromise, it is the correct feedstock at a lower delivered cost per useful unit of energy.

There is also a blending market. A utility with a stockpile of very low grade domestic coal can lift its average by blending in an imported mid grade, and the arithmetic of that blend often favours RB2 over RB1.

How the price is set

South African export coal prices against the API4 index, which is a free on board Richards Bay assessment. This is unusually helpful for buyers, because it means the reference point is a published number tied to a physical port rather than an opaque negotiation. A trader quoting South African coal without reference to API4 should be asked why.

The index is quoted for a benchmark specification, so real cargoes are priced at a differential to it. A grade below the benchmark trades at a discount, and the size of that discount is where the negotiation actually happens. Sellers who cannot articulate their differential clearly are usually not close enough to the physical material.

Terminal access is the real constraint

The single most common failure in South African coal is a seller with tonnes but no route to a vessel. Richards Bay Coal Terminal operates on an entitlement system, and a seller without entitlement cannot simply arrive with a train. Maputo Port in Mozambique has become the practical alternative, handling South African coal at smaller vessel sizes with a different cost stack, and it is where a great deal of non entitled tonnage now moves.

Before any coal offer leaves this desk we establish which of those two routes the seller actually has, and on what basis. An offer that cannot name its terminal arrangement is not an offer.

Cargo size and vessel

Coal is a cargo trade rather than a parcel trade. The economics work at Panamax scale, which is why the practical entry point through Richards Bay is 55,000 MT and upward, with Capesize cargoes common on longer routes. Maputo works at Supramax and small Panamax sizes, which is one reason it suits smaller sellers and smaller buyers.

Below those sizes the freight cost per tonne erodes the arbitrage. A buyer wanting 10,000 MT is nearly always better served buying delivered from a trader already breaking bulk at destination than trying to fix a dedicated vessel.

What to check before you fix

Four things: the certificate of analysis and who issued it, the terminal entitlement or booked slot, the laycan against the seller's actual rail programme, and the payment instrument. Coal disputes almost always trace back to one of those four being taken on trust.