Charted: Singapore Takes ~54% of Top-10 Disclosed Sand Imports — China MFA ~18 Gt vs US MCS 0.87 Gt

The most expensive cell in America's sand statistics is the one that reads NA. In its 2026 mineral summaries, the U.S. Geological Survey publishes a meticulous table for construction sand and gravel, then leaves the world-total cell blank — not a placeholder awaiting better data, but an honest admission that nobody reliably counts global sand. That is remarkable for a material this large: concrete, asphalt, glass, and land reclamation all run on grains most statistical systems barely bother to tally. UNEP still frames global sand resources — sand, gravel, and related aggregates — near 40–50 billion tonnes per year, second only to water among extracted materials [UNEP Sand]. The gap between that physical scale and the statistical void is the story. This dashboard treats extraction and trade as two different ledgers: what countries report digging, and what city-states must import for fill.

Four views carry it — Extractors, Importers, City-states, and the US MCS path — and the picture is deliberately two-sided. On the extraction side, material-flow analyses put China's sand use near 17.7 Gt, while the largest fully disclosed MCS reporter — the United States — sits at 870 Mt in 2025e. On the trade side, Singapore's 27.4 Mt of HS 2505.90 natural-sand imports in 2023 account for about 54% of the top-10 disclosed COMTRADE tonnage set, more than Belgium, the Netherlands, and the United States combined.

Scoreboard: extraction vs disclosed trade

Read the table as a mismatch between physical reality and statistical coverage. The top rows are planetary. The middle rows are the few agencies and associations that publish comparable pit tonnages. The bottom rows are trade — the only place where tiny, pit-poor jurisdictions light up brighter than continental producers.

Why “world production” is missing from USGS

USGS MCS is unusually honest about sand and gravel. For 2025 it estimates 870 million tonnes of U.S. construction sand and gravel sold or used, down from 880 Mt in 2024 and well below the 967 Mt peak year of 2023. Average unit value climbs to about $14.50 per tonne. Leading states (Texas, California, Arizona, and others) still concentrate roughly half of national output. End-use shares remain concrete-heavy: about 42% portland cement concrete aggregate, 20% road base, 12% construction fill, 9% asphaltic mixtures, and the rest a long tail of drainage, filtration, and specialty uses.

What MCS refuses to invent is a global total. Footnotes explain that countries report sand and gravel inconsistently — or not at all — and that some of the most important commercial deposits are glacial, riverine, or offshore. That is not a minor data footnote. It means any chart that pretends to show "top sand producers" without labeling disclosed vs estimated is marketing, not measurement. The dashboard's Disclosed only lens therefore drops MFA giants and keeps USGS/UEPG-style reporters; the MFA giants lens restores China and India as the physical heavyweights UNEP's envelope implies.

The extraction map that statistics can actually defend

Once you accept mixed metrics, the ranking becomes readable rather than false-precise. China dominates sand use in material-flow studies at roughly 17.7 Gt — nearly two-fifths of a 45 Gt UNEP midpoint. India is the next large carried estimate in Asia’s construction boom, with a wide uncertainty band. Among series that look like traditional mineral statistics, the United States leads disclosed construction sand and gravel at 870 Mt. Germany follows in the European association tables at about 211 Mt of sand and gravel, with France (~97 Mt) and Italy (~64 Mt) further down. Those European figures sit inside a broader UEPG-sphere aggregates industry near 3 Gt/yr when crushed rock, marine, manufactured, and recycled streams are included — a reminder that “sand” and “aggregates” are related but not identical columns.

Regional demand buckets in the dashboard assign China ~39%, India ~10%, rest of Asia-Pacific ~18%, Europe ~7%, North America ~6%, and the rest of the world the remainder. The analytical Herfindahl on those shares lands in the high-teens hundreds — concentrated by commodity standards, even before you remember that China’s MFA figure alone can exceed the entire disclosed U.S. series by an order of magnitude.

Trade: where city-states become visible

Physical sand barely travels far when pits are nearby; bulk density and low unit value punish distance. Trade therefore lights up where geography fails: dense city-states, landlocked micro-markets, and reclamation programs. UN Comtrade / WITS HS 2505.90 (natural sands excluding metal-bearing sands) is an imperfect but usable window. In 2023, quantity reporters put Singapore at 27.4 Mt, Belgium at 11.6 Mt, the Netherlands at 3.0 Mt, the United States at 2.0 Mt, and a long tail of Switzerland, France, China, Luxembourg, Germany, Thailand, Hong Kong, Macao, and the Maldives [UN Comtrade].

Singapore’s share of that top-ten disclosed tonnage set is about 54%. Amber bars in the importers view mark city-states; teal marks larger territorial markets. Belgium’s large import total is mostly cross-border European logistics (historically dominated by Netherlands partners), not a Singapore-style reclamation story. That distinction matters: high import tonnage is not automatically high fill dependence. Dependence is about whether domestic pits can substitute if a neighbor closes the gate.

City-states and the fill problem

Singapore is the archetype. UNEP-era tallies put cumulative sand imports near 517 Mt over two decades, and land area has grown roughly 28% since independence — from 581.5 km² in 1965 to about 744 km² by end-2025 — almost all of it through reclamation. In 2023 COMTRADE, nearly all of Singapore's disclosed HS 2505.90 tonnage is attributed to Malaysia. Neighbor export bans and diplomatic friction in earlier decades are the political shadow of that dependence: when a city-state has no glacial outwash plain of its own, fill becomes foreign policy.

Hong Kong and Macao show smaller absolute tonnages (~0.3 Mt each in 2023 quantity fields) but sit high on the dashboard’s dependence score because domestic pit access is scarce and reclamation history is long. The Maldives imports for atoll construction and coastal protection. Bahrain appears in value fields even when quantity cells are incomplete — a reminder that Gulf reclamation markets are structurally import-leaning. Luxembourg is the landlocked European cousin: not a reclamation island, but a high per-area import intensity relative to neighbors. Malta’s absolute tonnage is tiny, yet limestone geology and island constraints keep virgin natural sand scarce.

The City-states scatter plots disclosed import tonnage on a log axis against a desk dependence score. Singapore sits alone in the upper-right: large absolute imports and extreme structural dependence. Corridor bars then score Malaysia→Singapore fill intensity near the top of the analytical set, with Netherlands→Belgium cross-border flows and Canada→U.S. construction logistics as secondary patterns.

US MCS path: volume down, unit value up

The U.S. series is the cleanest time path in the file. Sold-or-used tonnage rose through 2023 (967 Mt), then stepped down to 880 Mt (2024e) and 870 Mt (2025e) as housing, highway, and weather cycles bit [USGS MCS]. Average unit value still climbed — from about $10.52/t in 2021 to $14.50/t in 2025e — consistent with MCS commentary on urban permitting friction, longer haul distances from new pits, and fuel costs. Net import reliance stays near zero as a share of apparent consumption; imports run only a few million tonnes, roughly 90% from Canada. That is the opposite of Singapore's balance sheet: continental producers can be price-stressed without being fill-dependent.

Caveats the dashboard will not hide

Several hard limits apply. First, UNEP’s 40–50 Gt envelope mixes sand, gravel, and crushed aggregates; it is a physical-order estimate, not a customs code. Second, China’s 17.7 Gt MFA sand-use figure is not interchangeable with USGS “construction sand and gravel sold or used.” Third, COMTRADE HS 2505.90 undercounts illicit and unreported flows and often disagrees with mirror export statistics — Singapore’s historical import/export mismatches are a famous example. Fourth, Belgium-style European imports can reflect internal market logistics, not strategic scarcity. Fifth, recycled asphalt and concrete are rising substitutes in MCS commentary but remain a small share of total aggregate supply in the United States. Sixth, industrial silica sand (glass, foundry, proppants) is a different MCS chapter with a published world table; this post stays on construction sand, gravel, and fill.

None of those caveats erase the headline geometry. The countries that dominate physical extraction are the ones pouring concrete and reclaiming coastlines at continental scale — led, on available MFA evidence, by China. The jurisdictions that dominate reported import dependence for fill are the pit-poor city-states, with Singapore’s disclosed 2023 tonnage still the clearest spike on the COMTRADE map. Between those two facts sits the statistical void USGS refuses to paper over: a world that digs tens of billions of tonnes of sand each year, and a global production table that still reads NA.