Indonesia's coconut charcoal industry: the data digest

By Mohamad Sinno, Charcoal Expert · 27 August 2026 · Reviewed by Budi Dharmawan, Head of QC, 4 September 2026

Indonesia's charcoal export figures circulate constantly and are almost never dated, sourced, or scoped. Below is the same data with all three attached, fetched from one primary source — and then the part nobody publishes: how much of the headline number is not coconut shell at all. None of these figures is ours. Every one is Indonesia's national customs reporting, for the whole country.

The numbers that matter

In 2025, Indonesia reported exports under HS heading 4402 — wood charcoal, including shell and nut charcoal — of USD 428,673,036 across 539,740 tonnes. Of that, USD 281,070,210 across 236,268 tonnes was line 4402.20, of coconut shell. Those are the three figures worth carrying, and 2025 is the most recent annual record in the database as of 27 August 2026.

Which means the figure most widely quoted for Indonesian charcoal — US$351.4 million, the 2024 heading total — is a year out of date, and was never a coconut-shell figure to begin with. It is still in the series below, with its own number intact. It is simply not the current one.

The digest table

Seven years of the heading, as reported. Read the volume column carefully: it is not continuous, and where it is missing or flagged, the table says so rather than smoothing it.

Method, briefly. Each row is one request to the UN Comtrade public API for reporter Indonesia, flow export, partner world, taking the all-modes total. Nothing is averaged, interpolated or carried forward. Two rows first came back empty and were refetched until they returned data — including 2024, which is the single most-quoted year in this dataset and would have been reported here as unavailable had the first attempt been trusted. This digest was compiled 2026-08 from sources dated as shown.

What the heading is actually made of

HS 4402 is not a coconut heading. It has three lines beneath it, and Indonesia reports all three. Splitting 2025 out is the single most useful thing that can be done with this data, and it is the reason the headline figure misleads so reliably.

Read the two share figures against each other, because they disagree in a way that matters. Coconut shell is 65.6% of the heading by value but only 43.8% by volume — both computed by us from the fetched columns. The other line is more than half the tonnage and about a third of the money. So anyone quoting the heading's tonnage as Indonesia's coconut charcoal volume overstates it by a factor of 2.28, and the heading's blended unit value of about USD 794 per tonne is not the price of anything: coconut shell sits near USD 1,190 and the other line near USD 486. An average across those two describes no product a buyer can order.

2025 carries this table and 2024 does not, for a specific reason. The 2024 split reconciles on value — its three lines sum to USD 351,375,005 exactly — but line 4402.90 reports no net weight for that year, so the volume axis cannot be closed. Subtracting the two reported lines from the heading total would appear to give the missing weight; that number is not published here, because the 2024 heading weight is itself flagged estimated and an estimate minus two measurements is not a measurement. For the record, coconut shell was 56.5% of the 2024 heading by value. If the distinction between 4402.20 and 4402.90 is a live question for your own entries, it is a classification question with its own page; this table is what it looks like in the trade statistics.

Why Indonesia leads

Two structural reasons, and neither is a claim about quality. The first is feedstock: the shell is a by-product of an existing copra and coconut-processing industry, so the raw material is already aggregated and already moving before anyone decides to carbonise it. That is a question of agricultural geography rather than of charcoal, and it has its own page. The second is that the trade is established rather than emerging — the series above shows seven consecutive years above USD 270 million, through a decline in 2024 and a recovery in 2025, which is a market with existing buyers and existing routes rather than a new one.

What the data does not support is the sentence every supplier site writes next. These figures describe a country's customs entries. They say nothing about any individual factory's grade, consistency, ash, or whether its material would pass your own laboratory, and a national export total is not evidence about a supplier — including this one. Where Indonesian charcoal actually goes, and what that implies about competition for supply, is the destination data, which comes with its own vintage problems. What one factory in that trade can show instead, stage by stage and checkpoint by checkpoint, is how coconut shell charcoal is made.

Reading trade data correctly

Four things this digest cannot tell you, stated plainly because the pages that omit them are the reason the figures circulate wrong.

The heading is wider than the product. That is the split table above, and it is the most common single error: a 4402 total quoted as a coconut-shell figure. By volume it is wrong by more than a factor of two.

Some cells are estimates, and the source says which. The two rows marked † in the series carry UN Comtrade's own net weight estimated flag — including 2024, whose 559,255 tonnes is the volume figure most often repeated. The value figures on those rows carry no such flag. A dataset that tells you which of its own numbers are modelled is being more honest than most of the pages that quote it without passing that on.

Reporter and mirror data are different numbers. Everything above is Indonesia reporting its own exports. The same trade counted from the other end — importing countries reporting arrivals from Indonesia — routinely gives a different total, because valuation basis, timing and transhipment all differ between the two ends of one shipment. Neither is the corrected version of the other. We have not published a mirror series here because we have not fetched one, and the gap between the two is not something we can quantify without it.

Sources disagree, sometimes by a factor of two. Our own origin comparison documents a case: for 2019, a supplier article gives USD 145.1 million while a Trade Map series gives roughly USD 280 million, and neither states its scope precisely enough to reconcile them — the two figures side by side. The digest above puts a third primary source on that question: UN Comtrade reports USD 280,104,175 for 2019, which sits with the larger figure. That corroborates one side without explaining the other — nothing fetched here says what the 145.1 million covers — so the honest position remains the one that page takes.

And one absence worth naming rather than hiding. A share-of-global-trade figure would belong on this page and is not here, because we could not fetch a world total for HS 4402: the World Bank WITS aggregate returns no records, ITC Trade Map now requires an account to reach its data, and the UN Comtrade endpoint used above does not aggregate across all reporters. A plausible world figure could have been assembled by summing a handful of large exporters — that sum would be our own arithmetic over a list we chose, presented as a global statistic, which is precisely the move this page has spent four sections warning about. So there is no share figure here.