Indonesia's Dual Fuel Gambit: CNG vs DME to Replace LPG Imports

CNG — Compressed Natural Gas

CNG is methane (CH₄) from existing gas fields, compressed to 200-250 bar for storage and transport. Indonesia produces roughly 5,600 BBtud of gas per year, with 69% allocated to domestic use as of 2025. CNG redirects that existing domestic supply toward household cooking.

Key technical characteristics:

• High-pressure storage (200-250 bar) requires Type I steel cylinders — significantly heavier and more expensive than the lightweight LPG cylinders households currently use.
• Not compatible with existing LPG stoves. CNG has different combustion properties — lower flame temperature, different air-fuel ratio. Existing household burners cannot run on CNG without modification or replacement with dual-fuel burners.
• Lower energy density per volume than LPG. A 3kg “equivalent” CNG cylinder holds less usable energy — the naming convention is by energy equivalence, not cylinder weight.
• Refilling requires compressor stations (Mother Stations connected to gas pipelines) rather than the simple cylinder-swap model LPG uses.

DME — Dimethyl Ether from Coal

DME (CH₃OCH₃) is a synthetic gas produced by gasifying low-grade coal. Indonesia has enormous coal reserves and wants to monetize them beyond export.

Key technical characteristics:

• Liquefies at ~5 bar, similar to LPG. Can use standard LPG cylinders and distribution logistics with minimal modification.
• Near drop-in with LPG stoves. Can be blended up to 20% with LPG with zero equipment changes. Higher blends require burner adjustments but not replacement.
• Produced from coal gasification — a carbon-intensive process that contradicts Indonesia’s climate commitments.
• No commercial-scale plant exists yet anywhere in Indonesia. The technology is proven (China has small-scale DME production) but the economics at Indonesian coal prices are questionable.

Side-by-Side

Energy efficiency takeaway: By mass, CNG actually packs more energy than LPG (~50 MJ/kg vs ~46 MJ/kg). But by volume — what matters inside a cylinder — CNG at 200 bar holds only ~36% of the energy that liquid LPG does in the same space. This is the fundamental trade-off: a 3kg “equivalent” CNG cylinder would need to be physically larger or hold the energy of a much smaller LPG canister.

DME is worse on both axes: only ~63% of LPG’s energy per kilogram and ~76% per liter. A household switching from a 3kg LPG cylinder to a 3kg DME cylinder would need roughly 60% more fuel by weight to cook the same amount of food — or accept proportionally fewer cooking hours per cylinder swap.

Volumetric Efficiency and Cylinder Design

The volumetric gap is the real bottleneck for household adoption:

• LPG at ~5 bar is stored as a liquid. A 3kg cylinder holds about 6 liters of liquid propane/butane with ~150 MJ of usable energy.
• CNG at 200 bar stores gas, not liquid. The same 6-liter cylinder at 200 bar holds only ~54 MJ — about one-third the energy. To match a 3kg LPG cylinder’s energy content, a CNG cylinder would need to be roughly 3x the volume or operate at impractically higher pressures.
• DME at ~5 bar is also liquid, like LPG. A 3kg DME cylinder would hold about the same volume but with ~95 MJ — roughly 63% of LPG’s energy. You’d need to swap cylinders 60% more often.

This explains why CNG works for vehicles (where space is less constrained and fueling stations have large high-pressure tanks) but is awkward for household 3kg cylinders. DME’s liquid storage is convenient, but the lower energy content means more frequent cylinder swaps and more trips to the pangkalan.

The 2007 Kerosene-to-LPG Conversion

The root cause is a 2007 government program that converted 50+ million households from kerosene to LPG for cooking. The reasoning was sound at the time:

• Kerosene subsidies were costing IDR 50+ trillion/year — even more than today’s LPG subsidy bill.
• LPG burns cleaner, produces less indoor air pollution (a major health issue in Indonesian kitchens).
• The government estimated savings of $3.7 billion/year by switching.
• By 2012, 53.9 million LPG starter packages had been distributed, and the program had saved over $6.9 billion in subsidies.

The problem: the program created overnight demand for 8+ million tons of LPG that domestic production could never meet. Indonesia’s refineries and gas processing plants could only supply a fraction. The rest had to be imported — and that import dependency has only grown since.

Why Domestic Production Can’t Keep Up

Indonesia’s LPG comes from two domestic sources:

1. Oil refineries — Pertamina’s six refineries (Cilacap, Balikpapan, Dumai, Plaju, Balongan, Kasim) produce LPG as a byproduct of crude oil refining. Total capacity: roughly 1-1.5 million tons/year. The recently upgraded Balikpapan refinery (inaugurated January 2026) boosted LPG capacity from 48,000 to 384,000 tons/year — potentially reducing imports by 4.9%.

2. Natural gas processing plants (NGL fractionation) — Plants like Badak NGL in Bontang extract propane and butane from natural gas streams. But most of Indonesia’s gas is produced as LNG for export or piped to industrial users, not fractionated into LPG components.

The total domestic LPG supply is approximately 2 million tons/year against demand of 8-10 million tons/year. The gap is massive and structural.

The LNG vs LPG Mismatch

Here’s the core confusion: LNG and LPG are completely different products.

• LNG is primarily methane (CH₄), cooled to -162°C for transport. Indonesia exports this — from Bontang, Tangguh, and the new Masela project.
• LPG is propane (C₃H₈) and butane (C₄H₁₀), heavier hydrocarbons that liquefy at moderate pressure (~5-7 bar). Indonesia imports this.

LNG export plants don’t produce LPG as a primary product. They strip out the heavier fractions (propane, butane, condensate) but the volumes are small relative to household demand. Converting LNG infrastructure to produce LPG would require entirely different processing facilities.

Indonesia exports roughly 31% of its gas production as LNG. The remaining 69% allocated domestically goes mostly to power generation, fertilizer plants (Pupuk Indonesia), and industrial users — not household LPG production. There simply isn’t enough natural gas liquids (NGL) in the gas stream to extract 8 million tons of propane/butane domestically.

The Government’s Own Diagnosis

Bahlil himself acknowledged in April 2026: “domestic production remains far below national demand” when announcing CNG as an alternative. The government wants to build additional LPG processing plants with 1.5-2 million tons/year capacity, and SKK Migas targeted an additional 1 million tons/year of LPG output in late 2024. Even if all these targets are met, domestic production would reach ~3-4 million tons — still well short of 8-10 million tons of demand.

This is fundamentally why CNG makes more sense than trying to close the LPG gap with domestic production: Indonesia has abundant methane (the main component of natural gas) but scarce propane/butane (the components of LPG). Rather than importing propane and butane, CNG skips the middleman and delivers the gas we already have directly to households.

Summary: The Import Dependency Chain

1
2007: Kerosene-to-LPG conversion
2
  → 50M+ households switched overnight
3
  → Created 8M ton/year LPG demand
4
  → Domestic supply: ~2M tons/year (refineries + NGL)
5
  → Gap: ~6M tons imported (now 7.8M)
6
  → LNG exports don't help (methane ≠ propane/butane)
7
  → Gas fields feed LNG export + power/industry, not LPG
8
  → Result: structural import dependency, IDR 87T/year subsidy

The Bahlil Announcement

On May 2-3, 2026, at an IPB Alumni Association event in Jakarta, Energy Minister Bahlil Lahadalia said the government is developing CNG in 3kg cylinders as an LPG alternative, claiming 30-40% cost savings. The 3kg CNG cylinder “baru mau dibuat” — it’s in development, not yet in production.

Infrastructure Already Exists

Unlike DME, CNG doesn’t require building new production plants. The infrastructure is partially in place:

PGN (PT Perusahaan Gas Negara) — now part of Pertamina Group — controls approximately 95% of Indonesia’s natural gas transmission infrastructure. Through subsidiary PT Gagas Energi Indonesia (PGN Gagas), it already operates:

• 14 SPBG (Stasiun Pengisian Bahan Bakar Gas) across 7 provinces as of November 2025.
• Gaslink — CNG service for industrial, commercial, and retail customers.
• Gasku — CNG for land transport vehicles.
• Mother Station CNG Medan — 1 MMSCFD capacity, built on 1,357 m² PGN-owned land in Martubung, Medan. Construction started June 2025.

PGN plans to invest $353 million in 2026 to strengthen gas infrastructure, targeting 877 billion cubic feet of natural gas distribution. The Jargas (Jaringan Gas) household piped-gas program is expanding to 15 provinces in 2025-2026.

What CNG Actually Needs to Scale

The missing pieces for household CNG are not about gas supply — Indonesia has that. They’re about last-mile logistics:

1. High-pressure 3kg cylinders must be designed, tested, certified, and mass-produced. These are fundamentally different from the lightweight welded LPG cylinders currently in use.
2. Burner conversion or replacement for ~65 million households. Either dual-fuel burners (LPG/CNG) or dedicated CNG burners. This is the biggest logistical challenge.
3. Refilling infrastructure — compressor stations near population centers, connected to PGN’s pipeline network. The existing 14 SPBG are for vehicles, not cylinder refilling.
4. Safety standards and regulation — high-pressure cylinders in households require different handling, storage, and certification rules than 7-bar LPG cylinders.

The 30-40% Savings Claim

Bahlil’s cost comparison likely compares the commodity cost of gas (CNG methane vs. imported LPG propane/butane) without accounting for:

• Higher cylinder manufacturing cost (Type I steel vs. lightweight welded).
• Burner conversion or replacement cost per household.
• Compressor station capex and operating costs.
• Distribution logistics (CNG can’t use the existing cylinder-swap model as easily).

The savings may be real at the commodity level but significantly eroded at the delivered-to-household level.

Six Coal Gasification Projects

The government plans six simultaneous coal-to-DME gasification projects. The lead developer is PT Bukit Asam (PTBA), Indonesia’s state coal miner, which is studying a $3.1 billion plant to convert coal into synthetic natural gas (syngas) and DME.

The timeline tells the story:

• 2022: Air Products (US firm, technology partner) signs on.
• March 2023: Air Products withdraws from all Indonesian coal gasification projects, citing changed financial landscape for clean energy investments.
• 2024-2025: Project repeatedly delayed. Originally targeted for groundbreaking in 2024.
• 2026: Bahlil says coal gasification development will begin in 2026.
• 2028: Bukit Asam now targets 2028 for construction start.

The Economics Don’t Work

IEEFA (Institute for Energy Economics and Financial Analysis) has published multiple analyses showing DME is economically unviable in Indonesia:

• DME plant capex would consume 70% of Indonesia’s total annual LPG import cost ($4.3B) while yielding only 1 million tonnes of LPG-equivalent energy — offsetting roughly 15% of imports.
• Production cost: $244-281 per tonne of DME depending on coal pricing (average selling price vs. cash cost).
• A separate IEEFA analysis estimates the DME project could lose $377 million annually.
• China’s Shanxi Lanhua already halted its DME project due to poor economics — a direct precedent.

The fundamental problem: Indonesia’s coal is profitable when exported. Converting it to DME domestically means forgoing export revenue and spending billions on gasification plants that produce a fuel more expensive than the LPG they’re meant to replace.

Who’s Pushing DME and Why

The DME push is driven by coal industry interests, particularly Bukit Asam, which needs alternative revenue streams as global demand for thermal coal declines. Coal gasification is framed as “downstreaming” (hilirisasi) — adding value to raw commodity exports — a policy priority under both Jokowi and Prabowo.

But the economics only work if:

• Coal is priced at or below extraction cost (foregoing export revenue).
• The government subsidizes DME production or mandates blending quotas.
• Oil prices stay high enough to make LPG expensive.

None of these are stable assumptions.

LPG Distribution

CNG Infrastructure

DME Development