What a tonne of shisha charcoal costs to make in Indonesia
By Mohamad Sinno, Charcoal Expert · published .
A factory's cost is knowable in parts. Three of the parts have instruments you can open yourself — a governor's decree, a published electricity tariff, a parliamentary study. The rest depend on how a particular plant runs. And the part that decides the answer is the one no factory publishes, including this one.
The short answer
There is no single number, and a page that gives you one is quoting a plant it does not run. Below are the lines themselves: three anchored to documents you can open, the rest named and left open, because inventing them would make this page less useful rather than more.
Labour: one province, one decree, 27 answers
Indonesian minimum wages are set district by district. A single West Java decree — Keputusan Gubernur Jawa Barat Nomor 561.7/Kep.862-Kesra/2025, on Upah Minimum Kabupaten/Kota Tahun 2026, set in Bandung on 24 December 2025 and payable from 1 January 2026 — fixes 27 different monthly minimums. The floor is Kabupaten Pangandaran at Rp 2,351,250; the ceiling is Kota Bekasi at Rp 5,999,443. That is a spread of about 2.55× under one instrument, in one province, on location alone.
Our production site is in Kabupaten Bogor, which the same decree sets at Rp 5,161,769. We use that row because that is where production happens — a point worth stating plainly, because our registered office is in Jakarta Utara, which sits under an entirely different wage instrument. A supplier's registered address and its factory are not always the same place, and for a cost model it is the factory that counts.
And here is the caveat that matters more than the number. The decree binds one category of worker. Diktum KETIGA: the minimum "hanya berlaku bagi pekerja dengan masa kerja kurang dari 1 (satu) tahun" — it applies only to workers with less than one year of service. Beyond a year, diktum KELIMA puts the employer on its own wage structure and scale; diktum KEENAM exempts micro and small enterprises from the floor altogether. So a labour line built on the minimum wage models a first-year workforce at a business above the small-enterprise threshold. It is a floor for a new hire, not a wage bill, and a cost model that treats it as one will understate every experienced operator on the line.
What we do not publish is the other half of that line: how many people a tonne takes. A wage becomes a cost per tonne only when divided by output, and we publish no output figure — here or anywhere else on this site. The same limit applies to the tariff in the next section, and it is the reason neither line appears in the table further down.
Electricity: a number with a three-month shelf life
| Band | Tariff |
|---|---|
| I-3 / medium voltage, above 200 kVA | Rp 1,122 / kWh |
| I-4 / high voltage, from 30,000 kVA | Rp 997 / kWh |
| B-2 / low voltage, 6,600 VA to 200 kVA | Rp 1,445 / kWh |
Those are the non-subsidised bands held for the quarter to September 2026 (cnbcindonesia.com, 10 August 2026). The mechanism is the more useful fact. Under Peraturan Menteri ESDM No. 7 of 2024 the tariff is re-set every three months against four macroeconomic parameters — the rupiah exchange rate, the Indonesian Crude Price, inflation and the coal reference price. A cost model built on one quarter's figure is out of date within a quarter by design, which is the opposite of how most published cost breakdowns treat the energy line.
Raw material: the one line with a public price
One source carries a public price for the raw material, and the chain it arrives through matters as much as the number. Kuntoro Boga, a director at Indonesia's agriculture ministry, writing in a signed opinion column (agri.kompas.com, 8 December 2025 — the page states it is a column carrying the author's own views rather than the publication's), reports that a study by the Pusat Kajian Anggaran DPR RI, the Indonesian parliament's budget research centre, records carbonized coconut shell as a raw material at about Rp 6,000 per kg, the finished briquette at about Rp 14,000 per kg domestically, and export value at about USD 1,300 per tonne.
So these are figures reported at second hand, and we would rather say so than dress them up. What we read, in full and in its own bytes, is the column. We have not read the budget-office study itself, and we make no claim about what it contains beyond what this column reports of it. If the analysis surfaces in its own right, this section gains a better anchor and this sentence changes.
Two things to hold on to. The study draws no grade distinction, so Rp 6,000 is a floor for the category and not a shisha-grade quote. And USD 1,300 per tonne sits below every band of our own published grid, which runs 1650–1900 USD per tonne boxed, FOB Tanjung Priok — so it is a market average across everything the country exports, not a comparable, and certainly not a cost-plus floor to read our prices against.
On yield, the producers' association puts the first conversion at four kilograms of shell to one kilogram of charcoal, with a truck averaging 25 tonnes (Asep Mulyana, Hipbaki, in indonesiamedia.com, 18 September 2021). Read that for what it measures and nothing else. It is a dated trade observation about carbonization — raw shell turning into charcoal, the first conversion. It is a different conversion from the one this page calls decisive, which is the second: charcoal turning into finished, screened, dried, packed briquettes. A 4:1 ratio for the first says nothing about the loss in the second, and treating it as though it did is the commonest way a cost model for this product goes wrong.
The lines nobody publishes, starting with the one that decides it
Material loss. Between charcoal arriving and briquettes leaving, material is lost — to grinding, to fines, to shapes that break, to screening, to drying. Whatever does not leave as saleable product was still bought and still processed, so it is paid for twice. This is the line that moves the answer most, and it is the line with no public source at all. The peer-reviewed literature on coconut briquettes characterises finished laboratory samples — density, ash, shatter index — and stops there; the trade press quotes the carbonization ratio above and nothing about the second step. We searched for an industrial figure and did not find one.
The others, briefly, each unpublished for the same reason — what moves them is specific to a plant. Binder: a process share rather than a price, ours a typical 1.5–3% of the mix, costing whatever tapioca starch costs that week. Oven fuel: our drying runs 75–85 hours at 160 °C, so the bill turns on what heats the oven and what that fuel costs locally, not on the schedule. Packaging: specified by what the buyer orders, which is why it belongs in a quote rather than a cost model. Maintenance and overhead: dies, screens, rent, compliance, the export desk. Capital cost — what building the plant cost — is a separate question and is out of scope here.
The model: your two numbers, and what they do
Exactly two of the lines above reach a cost per finished tonne without a factory's output figure: the price you pay for charcoal, and the share of it you lose. Find your price along the top and your loss down the side. Each cell is the raw-material cost of one finished tonne, in millions of rupiah — the arithmetic being that a tonne out needs more than a tonne in.
| Loss | Rp 5,000/kg | Rp 6,000/kg | Rp 7,000/kg | Rp 8,000/kg |
|---|---|---|---|---|
| 0% | 5.00 | 6.00 | 7.00 | 8.00 |
| 10% | 5.56 | 6.67 | 7.78 | 8.89 |
| 20% | 6.25 | 7.50 | 8.75 | 10.00 |
| 30% | 7.14 | 8.57 | 10.00 | 11.43 |
| 40% | 8.33 | 10.00 | 11.67 | 13.33 |
Read across one row and the price moves the cost proportionally, which is what everyone expects. Read down one column and loss does something sharper: at 40% the same charcoal costs two-thirds more per finished tonne than it does at nothing. That is why the unpublished number is the decisive one, and why a quoted price tells you more about a factory's process control than about its shell supplier.
A worked example, with numbers that are not ours. Take the study's Rp 6,000 per kg, and assume a loss of 25% — a round figure chosen to be easy to follow, not a figure anyone measured. One finished tonne then needs 1,333 kg of charcoal, and the raw-material line comes to Rp 8.00 million per tonne. Add your own labour, power, binder, fuel, packaging and overhead, and convert at the rate on the day. Every input in that sentence is illustrative. None of it is our cost, and we publish no cost figure of ours anywhere on this site.
What this page does not contain
Our own cost — not here and not elsewhere, because a supplier's cost breakdown is its margin stated backwards. What we do publish is the other side of it: a dated grid of five tiers, each defined by the ash band it guarantees, on the pricing page. Set whatever model you build here against a real, dated number — the comparison a cost breakdown alone cannot give you. Prices valid September 2026 · reviewed monthly.
From there: what actually moves the per-tonne figure on a quote is at price per ton drivers; why an order starts at a full container is at why the minimum is one container; the same arithmetic from a brand owner's side is at private-label margin math; and the terms a wholesale order is bought on are at wholesale.
Common questions
- What does it cost to produce a tonne of coconut shell charcoal briquettes?
- Nobody can answer that for a factory they do not run, and we publish no figure of our own. What can be answered is which lines the number is made of. Three have public instruments you can open yourself — a governor's decree for wages, a published tariff for electricity, a parliamentary budget-office study for raw material. The rest depend on how a particular plant runs, and the largest of them is published by nobody.
- Which cost line matters most?
- Material loss. Charcoal that goes into the mixer and does not leave as saleable briquettes is paid for twice — once as raw material and once as the production time spent on it — and it scales everything upstream of it. A change of a few points in loss moves the raw-material cost per finished tonne more than a change in the shell price itself. It is also the one figure no factory publishes and no study we could find measures at industrial scale.
- How much is labour in Indonesia?
- It is set district by district, not nationally. One West Java decree fixes 27 different monthly minimums for 2026, and they run from Rp 2,351,250 to Rp 5,999,443 — the top of the province is about 2.55 times the bottom, under one instrument, in one province. Our production site is in Kabupaten Bogor, which is Rp 5,161,769. And the decree binds only workers with under a year of service; beyond that the employer's own wage structure applies.
- How much is industrial electricity in Indonesia?
- For the quarter to September 2026, PLN's medium-voltage industrial band above 200 kVA is Rp 1,122 per kWh and the high-voltage band from 30,000 kVA is Rp 997. The more useful fact is the mechanism: the tariff is re-set every three months against the rupiah, the Indonesian crude price, inflation and the coal reference price, so a cost model built on one quarter's number has a three-month shelf life.
- Why will a supplier not show me their cost breakdown?
- Because it is their margin, stated backwards. That is an ordinary commercial reason rather than a suspicious one, and a supplier declining to give it is not a red flag. What you can reasonably ask for is the things a cost model needs in order to be checked at all: the incoterm, the grade as a measured band, the packaging, and the date the price was set.
- Can I work out a supplier's margin from their quoted price?
- Not reliably, and the attempt usually misleads. It would take their loss rate, their yield, their energy use per tonne and their overhead recovery — four numbers nobody publishes, and the first swings the answer more than the rest combined. A public cost model is good for the opposite direction: checking whether a quote is implausibly low for the region it claims to come from.
Check a quote against a dated price
If you are testing whether a quote is plausible, the fastest check is a dated one. Message the export desk on WhatsApp with your shape, ash band, packaging and port, and ask for the date the price was set.