How fast is Brazilian cement demand growing in 2026?

Dispatches reached 39.0 million tonnes in January–July 2026, up 2.3% year on year; June was the best June on record at +7.7% and May the only negative month at −0.9%. SNIC guides the full year to +1.5%–2%, below the near-4% of the previous two years. Regionally the Northeast leads at +6.7% while the Southeast, the largest market, is at −1.0%.

Brazil's Concrete and Admixture Industry in 2026: 39 Million Tonnes of Cement in Seven Months, a 73% CR5, and Where Imported PCE Actually Enters

Q: How large is Brazil's concrete admixture market?

It depends on scope. All finished categories — plasticisers, retarders, air entrainers and others — came to about USD 541.2 million in 2025, reaching USD 1,020.6 million by 2035 at a 6.55% CAGR (Market Research Future). PCE and selected categories were about USD 206.5 million in 2026, reaching USD 268.1 million by 2031 at 5.36% (MarketsandMarkets). The two differ by 2.6×, so the scope must always be stated and the figures should never share a comparison table.

Q: What is the real tax burden on admixture imports into Brazil?

Two figures. The nominal sum is about 22% (II 12.6% plus PIS/COFINS 9.25% plus ICMS 18%, each on a different base); the cash burden after cumulative taxation is about 47.7% of CIF value. Cost models must use the second. From 2026 an IBS/CBS layer applies on top (CBS 0.9% plus IBS 0.1% in the test year), so build the model across the 2027, 2029 and 2033 transition points.

Here is the answer in one sentence: Brazil is the only South American market where demand is still growing, finished-admixture supply is already localised, and the upstream polycarboxylate (PCE) base still has to be imported — which means the entry point for Chinese PCE is supplying onshore blenders, not selling finished admixture to ready-mix plants.

The numbers behind that sentence: 39.0 million tonnes of cement were dispatched in the first seven months of 2026, up 2.3% year on year, while the National Cement Industry Union (SNIC) guides the full year to only +1.5%–2% — growth is intact but has downshifted from the near-4% of the previous two years. Ready-mix concrete (RMC) consumption runs at roughly 76 million m³ a year, close to half of South America. The finished-admixture market is USD 206.5 million in 2026 on a narrow scope and USD 541.2 million in 2025 on a broad scope — a 2.6× gap that is a definition problem, not a typo. Sika, Master Builders Solutions, Mapei, GCP and Fosroc all compound onshore, and the top five hold roughly 73%.

This article is organised in nine sections: demand, concrete structure, supply, the price and tax chain, the three compliance gates, the two quantifiable demand pockets, five actions for 2027, and a section on what this dataset cannot answer. The data comes in three layers: our own Brazil research file (105 de-duplicated entities, the full price chain, customs evidence collected between May and October 2026); public statistics for January–October 2026 (SNIC/ABCP dispatches, CBIC construction GDP, IBGE, Concrete Show organiser data); and cross-checked third-party sizing (Market Research Future, MarketsandMarkets, Mordor, Grand View, IndexBox, UN Comtrade). Anything we calculated ourselves is labelled an estimate. The same table is maintained continuously in the admixor.com country database; what follows is its November 2026 snapshot.

One-page view: 14 readings, each with its scope

Metric	Reading	Scope / source
Cement dispatches, Jan–Jul 2026	39.0 Mt, +2.3% y/y; July alone 6.20 Mt (+2.5%)	SNIC monthly data (domestic sales)
Full-year 2026 guidance	+1.5%–2% (versus near 4% in the two prior years)	SNIC president Paulo Camillo Penna, July 2026
Full-year 2025 magnitude	About 66–67 Mt	Sum of SNIC monthly dispatches (±1%), not an official annual release
Construction GDP, 2026	Forecast +1.2%; first half +1.1%; Q2 −0.4% q/q	CBIC forecast; IBGE quarterly national accounts
Ready-mix concrete volume	Brazil about 76 mn m³ (2026), 47.85% of South America	Mordor: regional 166.77 mn m³ × Brazil share
Ready-mix concrete value	USD 9.61 bn (2025) → USD 12.20 bn (2026 est.)	Grand View (value basis; not convertible to the volume figure above)
Admixture market, broad scope	USD 541.2 m (2025) → USD 1,020.6 m (2035), CAGR 6.55%	Market Research Future (★ recommended baseline)
Admixture market, narrow scope	USD 206.5 m (2026) → USD 268.1 m (2031), CAGR 5.36%	MarketsandMarkets (PCE and selected categories)
Admixture output	506,000 t (2024), largest in Latin America	IndexBox (all categories, not PCE only)
HS 38244000 imports, 2025	USD 13.65 m / 18,500 t / USD 738 per tonne	UN Comtrade (★ value usable; tonnage disputed, see section 9)
China share of imports	29.1% (USD 3.97 m)	UN Comtrade; consistent with WITS 2024
Concentration	CR5 ≈ 73% (Sika 25 / MBS 18 / Mapei 12 / GCP 10 / Fosroc 8)	Our estimate, not statistical data
Landed cost, PCE liquid 50%	FOB USD 695–720 → local wholesale USD 1,200–2,200 per tonne	Volza transaction prices plus cumulative-tax model
Concrete Show 2026	450+ brands / 26,892 visitors / 36,000 m²	Concrete Show South America organiser

The two admixture rows are where most readers go wrong. Use the broad scope for total addressable market and the narrow scope for PCE-only sizing — but never put them in the same comparison table. That single mistake accounts for most of the confusion we see in Brazil market decks.

Demand: growth has downshifted, but the mix has changed

The 2026 cement curve is a wave: strong start, a second-quarter stall, a mid-year rebound. January 5.26 Mt (+1.1%), April 5.35 Mt (+2.2%), May 5.71 Mt (−0.9%, the only negative month of the year), June 5.83 Mt (+7.6%, the best June on record), July 6.20 Mt (+2.5%). The first half closed at +2.3%, and the first seven months at 39.0 Mt, also +2.3%.

Month, 2026	Dispatches (Mt)	Y/Y	Note
January	5.262	+1.1%	—
April	5.354	+2.2%	—
May	5.714	−0.9%	Southeast −1.9%, South −5.3%; only negative month
June	5.833	+7.6%	Record for the month; North +10.1%, Northeast +13.8%
July	6.200	+2.5%	South −1.9% on heavy rain
Jan–Jul cumulative	39.0	+2.3%	Full-year guidance +1.5%–2%

The regional split matters more than the national total. In January–May 2026 the Northeast grew 6.7% to 6.03 Mt, the North 2.4%, the South 1.1%, while the largest market, the Southeast, fell 1.0% to 12.12 Mt. The Northeast is where the Minha Casa Minha Vida (MCMV) housing programme and the "water for all" sanitation build-out concentrate — and where international brands have the thinnest technical service coverage. That is the geography that should decide which state your first samples go to.

Three verified supports and two verified drags:

Support 1 — MCMV now dominates residential supply. In Q2 2026 the programme accounted for 58% of new residential launches (62,129 units), the highest share since the series began in 2019; 58,392 of the 113,676 units sold in the quarter were inside it. ABCP now puts the cumulative target at 3 million units across 2023–2026, above the original 2 million.
Support 2 — rigid pavement. Paraná and Santa Catarina are converting highway paving from asphalt to concrete, one reason the South outperformed the national average in 2026. Rigid pavement consumes more admixture per cubic metre than ordinary RMC (air entrainment plus retarder plus water reducer) — a genuine structural increment rather than volume growth.
Support 3 — labour and wages. Employment reached 102.1 million with 5.8% unemployment in February 2026, the lowest for the period, and average monthly income of BRL 3,679 kept self-build and small works alive.
Drag 1 — the Selic rate at 14.75% (March 2026). Household debt absorbs 49.8% of income, defaults are at record levels and construction-materials retail fell 4.9% in April. Mid- and high-end residential and commercial work is highly rate-sensitive.
Drag 2 — cost. INCC-M rose 6.61% over twelve months to September 2026, and SINAPI construction cost reached BRL 1,993.76/m² in August (+7.03%). Cement itself rose 5.38% quarter on quarter in Q2 2026 and another 3.67% in June on the import price series — cost-push inflation that is squeezing contractors' materials budgets.

For water reducers the implication is precise: volume growth is only 2%, but admixture value per cubic metre is rising. Standardised MCMV mixes need consistency and pumpability; rigid pavement needs air entrainment and retardation; low-clinker cements need high water reduction to hold workability. All three point to PCE rather than lignosulfonate or naphthalene.

Concrete structure: 76 million m³, and three real increments

Brazil places roughly 76 million m³ of ready-mix concrete a year (2026, Mordor basis), 47.85% of South America, with the region forecast to grow 4.55% a year to 2031 and Brazil 4.84%. The value series (Grand View) runs USD 9.61 bn in 2025 to an estimated USD 12.20 bn in 2026 — the two bases grow at very different rates (4.5% in volume versus 27% in value), which tells you that the 2026 value jump embeds cement inflation and product mix shift, not a demand boom.

Here is an arithmetic check anyone can reproduce: if every cubic metre were dosed at 2.5 kg/m³ of PCE (0.7% of a 350 kg/m³ binder content), Brazil would need about 190,000 t. Total admixture output of all categories is 506,000 t (2024). That 38% share is implausibly high, which means PCE covers only part of the ready-mix market — high-performance and engineered mixes, precast, infrastructure — while ordinary strength classes still use low-dose plasticisers or none. That gap is the quantified form of the upgrade opportunity: the market is not small, penetration is low.

Increment	Scale / characteristic	Buyer	Admixture type
MCMV housing and precast	58% of Q2 2026 launches; precast wall and slab lines scaling	Regional RMC plants, precast factories	High-range PCE, accelerators, retarders (35°C+ summers)
Rigid pavement and concrete streets	Concentrated in PR and SC, pushed by state highway departments	Paving contractors, RMC suppliers	Air entrainers, retarding water reducers, superplasticisers
Metro and marine (durability)	São Paulo Lines 2 and 19 total 1.73 mn m³; 35 port expansions	Chinese EPCs, large RMC producers	High-range PCE plus retarder and corrosion inhibitor

What not to do is sell finished admixture to the 29 small and mid-sized ready-mix plants in our file. They have incumbent suppliers and long payment terms, and finished product must pass ABNT NBR 11768 (equivalent to EN 934-2) — which alone makes the return on that route the worst of the three options.

Supply: a 73% CR5, and three capacity moves in 2025–2026

Concentration is higher than most visitors expect: CR5 ≈ 73% (Sika 25, Master Builders 18, Mapei 12, GCP 10, Fosroc 8), with Novakem — the largest domestic blender — at about 5% and the remaining 22% split between domestic formulators and imports. Two caveats on scope. First, the source report's narrative says "top five 60–70%" while its own share table sums to 73%: use 73%. Second, MBCC was acquired by Sika in 2022, so adding Sika's 25 to MBS's 18 (43%) requires stating that they are now the same group, or you will understate real concentration.

Brand / group	Estimated share	Local production	2025–2026 move
Sika Brasil	25%	Campinas; Belo Horizonte (MG) site extended June 2025 with admixture production	Capacity moved closer to cement and mining customers
Master Builders Solutions	18%	São Paulo state	Automated plant commissioned May 2026 — shorter lead times, steadier supply
Mapei Brasil	12%	Local plant	—
GCP Applied Technologies	10%	Local compounding plus exports	Regional exports to Paraguay documented in bills of lading
Fosroc Brasil	8%	Local	Acquired by Saint-Gobain February 2025; Saint-Gobain expanded LatAm construction chemicals August 2026
Novakem	5%	Largest domestic blender	Oriental Yuhong took a 60% stake, December 2025
Others (domestic blenders and imports)	22%	—	The compound capacity not yet locked up by Chinese capital

Those three moves say one thing: the leaders are no longer importing finished product and repackaging — they are moving compounding and synthesis onshore. For a base seller that cuts both ways. The blending pool is getting bigger, so base demand grows; but a self-sufficient leader negotiates harder on second-source pricing and qualifies suppliers more slowly.

Which leaves China's realistic position clear: there is no public evidence of PCE macromonomer (TPEG/HPEG) synthesis in Brazil, consistent with a structure of "import 18,500 t of base and compound locally." Chinese suppliers therefore sell an intermediate, not a finished admixture. Intermediates do not require finished-product NBR 11768 certification — the licensed blender carries that. That asymmetry is the core reason the blender route beats the finished-product route.

Price and tax: 22% is nominal, 47.7% is cash, and 2026 adds an IBS/CBS layer

Stage	PCE liquid (50%)	PCE powder (95%+)	Note
China FOB	USD 695–720/t (actual transactions)	USD 1,400–2,200	Volza China→Vietnam/Malaysia transaction prices, not quotes
CIF Santos	USD 700–1,200	USD 1,650–2,450	FOB plus USD 150–250/t ocean freight
Landed cost, all taxes	USD 1,033–1,773	USD 2,436–3,617	Cumulative-tax model
Local wholesale, unbranded or Chinese import	USD 1,200–1,600	USD 2,400–3,200	Landed plus 30–40% margin
Local wholesale, Sika/BASF branded	USD 1,500–2,200	USD 2,800–3,800	Brand plus technical service premium

On tax, three internal documents disagree but the substance is one sentence: 22% is the sum of nominal rates, 47.7% is the cash burden after cumulative taxation. Brazil's ICMS, PIS and COFINS cascade, so the effective load is mathematically higher than the nominal sum. Build cost models on 47.7%; quote "about 22% nominal" only in conversation, and say which one you mean. Separately, the import duty itself is unresolved — 6.5% (Mercosur common external tariff) versus 12.6% (the specific HS 38244000 line). A 6.1-point spread eats margin directly, so the NCM 3824.40.00 rate must be confirmed in writing by a local customs broker before any price commitment.

2026 adds a new variable: tax reform (Complementary Law 214/2025) is in its test year. 2026 runs at symbolic rates — CBS 0.9% plus IBS 0.1% — with offsetting allowed. 2027–2028 removes PIS/COFINS and moves CBS to full collection. 2029–2032 phases ICMS/ISS out (10%, 20%, 30%, 40%). Full transition lands in 2033. Construction services receive a 50% rate reduction, which law-firm modelling puts at an effective combined IBS/CBS rate of about 14% for construction services by 2033. Decree 12,955/2026 and IBS Committee Resolution 6/2026 took effect on 30 April 2026; the tolerance window closed on 31 July, and from 1 August 2026 errors in IBS/CBS invoice fields carry penalties.

Phase	CBS	IBS	ICMS/ISS	What it does to your quotation model
2026 (test year)	0.9%	0.1%	Full	Symbolic and offsettable, but non-compliant invoice fields get rejected
2027–2028	Full reference rate	0.1%	Full	PIS/COFINS exit; input credit structure changes
2029	Full	10%	90%	Dual-track: state the base date of every cost figure
2033	Full	Full	Eliminated	Effective combined rate about 14% for construction services (estimate)

One practical rule: every quotation must name the tax base date and the legislation version behind it, with a price-adjustment clause attached. A long-term contract that spans 2027 and 2029 cannot safely fix a tax rate.

Three compliance gates: NBR 11768, ANVISA, Ex-tarifário

Gate	Authority	Mandatory	Lead time	What it means
ABNT NBR 11768	INMETRO / LABNORM	For public works	~60 days	Brazilian admixture standard, equivalent to EN 934-2; PCE must meet Type E (superplasticiser) performance. Chinese GB 8076 reports cannot be converted — retesting or an equivalence assessment is required
ANVISA registration	ANVISA	For imports	~90 days	Industrial chemical import registration; the binding constraint of the finished-product route
ISO 9001	Accredited bodies	De facto	~30 days	Hard threshold in large tenders
Ex-tarifário duty relief	MDIC	Optional	~180 days	Requires proof that no domestic equivalent exists; 4–6 months
Anti-dumping	—	—	—	No AD measures against Chinese PCE as of August 2026 (Egypt and Saudi Arabia have measures on SNF-type products)

Together these gates explain why the intermediate route is the only rational first move: finished product needs both 11768 and ANVISA (five months at best), while an intermediate is covered by the licensed blender's certificates. What you need instead is ISO 9001, batch-to-batch consistency data, and a compatibility report that runs in the customer's own laboratory.

The one quantifiable window: 1.73 million m³ and 865 t of PCE (estimate)

Project	Contractor	Contract value	Concrete volume	PCE potential (estimate)
São Paulo Metro Line 2 extension	PowerChina Sinohydro 11th Bureau + Mendes Júnior JV	BRL 6.879 bn (≈ USD 1.30 bn)	360,000 m³	180 t / about USD 300,000
São Paulo Metro Line 19 (Sky Blue), section 1	PowerChina Sinohydro 11th and 4th Bureaux + Mendes Júnior	BRL 4.985 bn (≈ USD 930 m)	1.37 mn m³	685 t / about USD 1.13 m
Total	—	≈ USD 2.23 bn	1.73 mn m³	865 t / about USD 1.43 m

Three qualifiers must travel with those numbers. First, 865 t and USD 1.43 m are a potential derived from concrete volume multiplied by an assumed dosage, not contracted procurement; actual dosage depends on mix design. Second, Chinese EPC materials purchasing usually runs through headquarters or international-project central procurement, so being able to sign on site does not mean being able to enter the account. Third, Line 19 completes in 2033 — a mid-term anchor, not this year's revenue.

Two further programmes are large but not convertible: the new PAC (BRL 1.7–1.9 trillion, mostly water and urban mobility) and the BNDES urban mobility plan ENMU (187 projects, 21 metropolitan regions, BRL 430 bn of potential investment). Neither discloses concrete volumes, so do not turn them into PCE tonnage. Use them as demand background only.

Five things to do in 2027

Send samples to blenders, not to ready-mix plants. Start with Cimental/Fermaflex, ADITIBRAS and Polytrade; then Manchester, IWT (Mendes), Anchieta and C&F. Novakem is closed as a channel — Oriental Yuhong holds 60% — but remains valuable as a competitor-pricing reference.
Lead with a compatibility report, not a price list. The first gate at a Brazilian blender is whether your base reproduces their water-reduction and slump-retention curve at the same solid content. A one-page table covering three water/binder ratios, two temperatures (25°C and 35°C) and one-hour slump retention beats a USD 50/t discount.
Lock the Santos rhythm, not the sailing schedule. Santos customs clearance runs 5–10 days and up to 14 in peak season; Shanghai or Ningbo to Santos is 35 days direct, without Suez. Keep Vitória or Itaguaí as a peak-season fallback — it protects delivery better than price cuts do.
Quote with two anchors and never back-solve from the tax rate. Put the landed range (USD 1,033–1,773/t) and the local wholesale range (USD 1,200–1,600/t) side by side, use tax only to explain the boundary, and put the IBS/CBS base date into the contract terms.
Reserve 17–19 August 2027 for São Paulo Expo. The 17th Concrete Show (25–27 August 2026) drew 450+ brands, 26,892 trade visitors and 36,000 m², with 81% of visitors holding decision-making authority; the next edition is confirmed for 17–19 August 2027. For a Chinese supplier it is the only three-day window in which competitors, customers and channels are visible at once — Chinese PCE suppliers including LANDU NOVASTAR have exhibited, and competitor pricing bands can be gathered on the floor.

Scope and limits: what this dataset can and cannot answer

Usable as written: cement dispatches and year-on-year change (SNIC monthly); MCMV and construction GDP (CBIC/IBGE); HS 38244000 import value and China's share (Comtrade); CR5 of 73% (labelled an estimate); 20 named decision-makers with titles; the concrete volumes and PCE potential of the two metro lines (labelled an estimate).

Not usable: putting USD 206.5 m and USD 541.2 m in the same table; treating the USD 541.2 m market as a PCE-only market; quoting 25% as Sika's exact share; treating Comtrade's 18,500 t and IndexBox's 8,500 t as the same quantity (values are close, tonnages differ by 2.2×, most likely a net-versus-gross or dilution-water basis difference); and claiming that N named importers have been identified — named-buyer coverage for HS 38244000 in Brazil is 0%.

That last point is simultaneously the most valuable fact here and the biggest hole. Brazil's customs data is far more commercialised than Indonesia's: the free tiers of Volza and 52wmb return "(Brazil importer, paywalled)", so we know USD 13.65 m was imported and not who imported it. There are only three ways to close the gap — buy a subscription, work through a local customs broker (despachante aduaneiro) with RADAR/DUIMP records, or collect business cards at Concrete Show. Our Brazil file on admixor.com keeps that gap explicitly flagged rather than filling it with inference, because an inferred buyer list is more dangerous than no list at all.

Frequently asked questions

How large is Brazil's concrete admixture market?
It depends on scope. All finished categories (plasticisers, retarders, air entrainers and others) came to about USD 541.2 m in 2025, reaching USD 1,020.6 m by 2035 at a 6.55% CAGR (Market Research Future). PCE and selected categories were about USD 206.5 m in 2026, reaching USD 268.1 m by 2031 at 5.36% (MarketsandMarkets). Always state the scope; the two differ by 2.6×.

Can Chinese PCE base be sold directly to Brazilian ready-mix plants?
Technically yes, commercially no. Finished product must clear ABNT NBR 11768 (equivalent to EN 934-2) and ANVISA registration — five months at best — and ready-mix plants already have incumbent suppliers and long payment terms. The realistic route is supplying an onshore blender, which holds the certificates and carries formulation responsibility.

Is there anti-dumping action against Chinese PCE in Brazil?
No. As of August 2026 neither WTO records nor Brazilian official notices show anti-dumping measures on Chinese PCE. Egypt and Saudi Arabia do have measures on SNF-type products. Brazil is currently one of the few markets combining "no AD" with a Chinese EPC channel.

What is the real tax burden on imports?
Two figures. The nominal sum is about 22% (II 12.6% plus PIS/COFINS 9.25% plus ICMS 18%, each on a different base); the cash burden after cumulative taxation is about 47.7% of CIF value. Cost models must use the second. From 2026 an IBS/CBS layer applies on top, so build a model across the 2027, 2029 and 2033 transition points.

Which trade event is worth attending in 2027?
Concrete Show South America, 17–19 August 2027, São Paulo Expo. The 2026 edition hosted 450+ brands, 26,892 trade visitors and 36,000 m², with 81% of visitors making decisions, and exhibitors ranging from Gerdau, Putzmeister and Schwing-Stetter to Aditibras and Camargo Química.