Blog2026-09-16Ingrid Solberg

Gerdau's 2024 Crude Steel Output: Which Tonnage You Can Actually Quote

If you need one number to stand behind for Gerdau's crude steel production, it is not a 2024 figure. The last annual tonnage that can be cited against a public source is approximately 13 million tons, reported for 2023, in a year when Gerdau ranked as the 33rd largest steelmaker worldwide and the largest producer of long steel in the Americas. That single line carries three constraints at once: it is a company-wide figure rather than a segment figure, it is a calendar-year anchor rather than a 2024 anchor, and it is a crude-steel output figure rather than a scrap-intake figure. Miss any one of the three and the number still looks plausible while describing something else entirely. The practical question behind a 2024 crude steel production figure in million tonnes is therefore a caliber question before it is a tonnage question.

The tonnage Gerdau reports, and which layer it actually counts

Start with the answer: for Gerdau's crude steel production, the number you can stand behind is not a 2024 figure. The last annual tonnage that can be cited against a public source is a figure of about 13 million tons, reported for 2023, in a year when Gerdau ranked as the 33rd largest steelmaker worldwide and the largest producer of long steel in the Americas. That one line carries three constraints at once: it is a company-wide figure rather than a segment figure, it is a calendar-year anchor rather than a 2024 anchor, and it is a crude-steel output figure rather than a scrap-intake figure. Miss any one of them and the number still looks plausible while describing something else entirely. The practical question behind a 2024 crude steel production figure in million tonnes is therefore a caliber question before it is a tonnage question.

Caliber matters because Gerdau's public disclosure sits at more than one level, and the levels do not reconcile to a single tonnage. The 13 million tons figure is a global crude steel production statement, yet the same public profile shows 2023 sales split almost evenly between two regional businesses: 39% of sales went to Brazil and 39% went to North America. Those are sales shares, not tonnage shares, and they are not the same measurement as crude steel output, but a reader building a comparison table can blend the two without noticing. The effect is that a 2024 number with no stated caliber cannot be checked against the segment structure, and it cannot be tied to a year boundary. That is exactly the condition the current material leaves us in: nothing in the public sources reviewed here anchors a 2024 crude steel tonnage on its own, so a 2024 figure quoted as settled fact carries more confidence than its source can support.

Company context explains why the tonnage is best read through operating structure rather than one headline. Gerdau began more than 120 years ago as a family-run nail factory in Porto Alegre, Brazil, and today its North American divisions manufacture long and special steel products for the agricultural, automotive, construction, distribution, energy, industrial, and mining markets, using electric arc furnace technology the company presents as one of the cleanest production routes available. The rest of this article works through that structure in order: why a clean 2024 tonnage figure is hard to pin down, what actually moves the tonnage, including scrap feedstock, the Brazil and North America split, and the product range, and then a short rule for reading any 2024 number before it goes into a memo. Read it that way and the missing 2024 anchor stops being a dead end and starts behaving like a set of boundary conditions you can state out loud.

A single 2024 figure is missing for a disclosure reason, not a search reason

What the public material offers is a mixture of one dated annual anchor and several undated multi-year statements, and that mixture is why a clean 2024 figure does not surface. The dated anchor is specific: roughly 13 million tons of production in 2023, alongside the ranking as the 33rd largest steelmaker worldwide. The undated statements are just as specific in tone but carry no year at all, since Gerdau describes transforming more than 10 million tons of recycled scrap metal into new steel each year, and its corporate pages discuss raw steel output in capacity terms without attaching a current-year value. When a disclosure layer mixes a dated figure with undated ones, the reader inherits the job of assigning the year, and that is the point where 2024 tonnages get invented rather than sourced.

The most tempting substitutes sound like production and are not. The first is the more than 10 million tons of recycled scrap metal that Gerdau converts into new steel each year: that is a feedstock and recycling-throughput measure, describing what enters the furnaces and the collection network rather than what leaves as crude steel. The second is the 96% average recycled content of the steel made by the North American team, which is a composition ratio, the share of recycled material in the finished product, and not a mass of output. Both numbers are legitimate and checkable, and both sit in the same order of magnitude as a 13 million ton output figure, which is precisely what makes the mix-up easy. When a 2024 cell in a comparison table shows roughly ten million tonnes and cites recycled scrap intake, the number is not wrong; the label is. Fixing the label usually fixes the comparison.

Name the misreading plainly: filing a scrap-intake or recycled-content figure as 2024 crude steel production. The correction is to describe 2024 with quantities that are actually anchored, plus an explicit year boundary. In practice that means writing something like: Gerdau reported approximately 13 million tons of production in 2023, the 2024 crude steel tonnage has not been separately anchored in the public material reviewed here, and the company's electric arc furnace route reports an average 96% recycled content in North America. That sentence is longer and less satisfying in a spreadsheet cell, but every clause can be traced back to a source. The alternative, dropping a bare ten-million-tonne figure into the 2024 column, produces a table that looks complete and fails the first time someone checks it against the company's own distinction between recycled content and raw steel output.

The scrap route, the segment split, and the product range decide how the tonnage behaves

The first mechanism is the production route, and it is why the tonnage question keeps blurring. Gerdau operates mini mills, integrated mills, and direct reduced iron plants, but 71% of the steel it manufactures is made from recycled scrap, and its mills use electric arc furnace technology as part of what the company describes as efficient and clean production practices. That mix shapes how an output number behaves: in a scrap-based shop the volume of steel produced and the volume of scrap charged move together, because furnace feed dominates over iron ore and coke in the input stack. A year with more scrap collected and charged is, other things equal, a year with more steel poured. So when a recycled-scrap figure and an output figure land within a few million tonnes of each other, that is the route showing through.

Segment structure is the second force, and it makes one global tonnage figure less informative than it looks. In 2023, 39% of Gerdau's sales went to Brazil and 39% went to North America, leaving roughly a fifth elsewhere. Two similarly sized regional businesses mean a global tonnage is the sum of two stories that can move in opposite directions in the same year, since Brazilian construction demand and North American long-product demand run on different cycles, currencies, and end markets. For anyone building a comparison, the consequence is blunt: a global 2023 figure and a global 2024 figure will not tell you which half of the business changed, and a peer table will quietly compare a two-thirds home-market producer against Gerdau's near-even split. Lock the segment before you compare the tonnage, because the segment is where the variation lives.

What the tonnage is actually made of comes down to the product range. Gerdau's North American network runs recycling locations, steel mills, and downstream facilities and manufactures merchant bar, structural steel, piling, special bar quality, and rebar, while the wider catalogue includes reinforcing bar, engineered and quenched-and-tempered bars, sheet piling, beams, and special steels used in applications such as gun barrel steels, rail reinforcing, and steering rack blanks. Long and special products like these are tied to construction, infrastructure, agricultural equipment, automotive components, and energy projects rather than to flat-rolled consumer goods. That matters when reading a tonnage figure: a mix weighted toward rebar and merchant bar will track construction activity, while special bar quality volumes follow industrial and automotive order books. Two tonnage numbers can match in total and be driven by entirely different demand.

Read any 2024 figure by locking the caliber first, then the year anchor

Here is the reading rule in two steps. First, identify the caliber before you read the number: ask whether the figure counts global crude steel output, a single segment, or scrap feedstock and recycled content, because each of those has a distinct public expression for this company and they sit in a similar order of magnitude, leaving the label as the only thing separating them. Second, find the year anchor. The last verifiable annual anchor in the public material is about 13 million tons for 2023, together with the ranking that comes with it. If the figure in front of you is dated 2024 and no source states both its caliber and its year basis, it is not a fact yet; it is a placeholder waiting for a disclosure.

Where the rule stops applying matters as much as where it starts. For an annual-report excerpt, a background paragraph, or a peer snapshot, a figure labelled with its caliber and its year, such as 2023 global crude steel output of roughly 13 million tons with 2024 marked as not separately anchored, is defensible and honest. For a precise year-on-year comparison, a financial model, or a supplier risk assessment, that is not enough; those need the company's own reporting, because pages that describe recycling throughput, product families, and recycled content percentages exist to communicate a route and a product range, not to feed a model. Notice the asymmetry those pages create: the official material is confident about what is made and how, covering the electric arc furnace route, the recycling network, and the product families, while staying deliberately unspecific about a current-year tonnage total. When a source is careful with a number, the citation should be just as careful.

One sentence carries the whole rule: ask which layer the number counts before you ask what the decimal says. Applied here, that means a 2024 crude steel tonnage with no stated caliber is treated as unanchored rather than adopted, a recycled-scrap throughput figure is filed under feedstock rather than output, and every like-for-like comparison is locked to one segment, because a business whose 2023 sales split 39% Brazil and 39% North America does not have a single tonnage story, it has two that happen to be added together. Set the caliber, name the year, then quote. That ordering costs one extra clause in a spreadsheet note and prevents the correction that follows when a scrap-intake figure turns up in the middle of a board deck.

The tonnage worth quoting today is the 2023 anchor, some 13 million tons of crude steel output from a company ranked 33rd worldwide and first in long steel in the Americas, with 2024 left explicitly open instead of filled in. That is not a gap to be papered over with a plausible-looking decimal. It is the boundary of what the public material supports, and once the scrap-based route and the two near-equal regional businesses are visible behind it, the boundary stops reading like an inconvenience and starts reading like a description of the business. Take the rule with you: lock the caliber, name the year anchor, then write the number down. Anything that survives both questions is worth quoting; anything that does not was never a production figure to begin with.

Ingrid Solberg

Ingrid Solberg

Ingrid Solberg is an independent carbon and structural steel analyst covering steel plate, sheet, coil, pipe, beams, rebar, and galvanized products. She applies ASTM A36/A36M and A6/A6M requirements while comparing grade chemistry, yield strength, tensile strength, elongation, dimensional tolerance, weldability, and surface condition. Her evidence-led guides help engineers, fabricators, and procurement teams select suitable steel forms, define purchase specifications, and evaluate mill or service-center offers.

Leave a Reply