Cheap for the 23, Then Doubled

The full numbers

This paradox extends #56 (LV doubled, HV untouched) by adding the missing global dimension. Per Bhutan vs Global Electricity Costs (2026-05-24) §3, the world industrial electricity benchmark table places Bhutan HV1 in extraordinary company:

Bhutan HV1 is not just cheap — it is structurally cheaper than every OECD industrial economy, every emerging-market manufacturing hub, and every state-owned hydropower regime with serious commercial regulation. Only Iran (heavily subsidised, ASIC-import-constrained, sanctions-exposed) clearly beats it. Iceland’s aluminium-smelter contracts — the global benchmark for energy-intensive industrial pricing in a renewable-rich economy — are still 85% more expensive.

Now overlay the 2026 tariff revision:

• LV residential customers (99.96% of customer count, ~10% of consumption): Nu 2.66 → Nu 5.73 = +115% increase. Post-revision LV at $0.067/kWh would still be globally cheap (39% of US residential; below India $0.078), but the relative shock to a household budget that has been stable for a decade is the visible burden.
• HV1 industrial customers (23 entities, 88% of consumption): Nu 1.60 → Nu 2.80 = +75% increase. Post-revision HV1 at $0.033/kWh would still be in the global bottom decile — roughly tied with Paraguay, Ethiopia, and Iceland aluminium contracts; ~40% of US industrial; ~18% of Germany.

The structural cross-subsidy framework that emerges:

The asymmetry: the customers receiving the deepest global subsidy face the smaller relative increase from a smaller base; the customers receiving a smaller global subsidy face the larger relative increase from a larger base. In a market economy, the cross-subsidy in cheap-power countries typically flows the other direction — large industrial users subsidising residential through volume-discount price discrimination, or paying a premium for grid-firming services that benefit LV reliability. Iceland’s aluminium smelters, for example, historically subsidised LV through bundled grid-cost allocations; the cross-subsidy went from the smelters to the households (per Askja Energy historical filings).

In Bhutan’s 2026 proposal, the structure runs the opposite way. The 23 customers receiving the world’s lowest-cost legitimate industrial electricity — including Bhutan’s sovereign BTC miner (sovereign BTC, USD-denominated revenue), BitDeer JV (commercial BTC), and large cement / ferro-alloy exporters — bear a smaller relative share of the revision burden than the 99.96% of households who have not done well out of the existing structure to begin with.

Imagine this

The minister responsible for energy stands at the parliamentary podium in May 2026. He explains, accurately, that Bhutan’s proposed unsubsidised LV residential tariff of Nu 5.73/unit still compares favourably against West Bengal Nu 7, Assam Nu 8, Bangladesh Nu 7.5, and Tibet Nu 8. The comparison is technically correct.

What he does not put on the same slide: post-revision, Bhutan’s HV1 industrial tariff ($0.033/kWh) will still be among the eight or ten cheapest in the world, lower than every state in India, every province in China, every state in the US except Washington-hydro, and roughly equal to Iceland’s flagship aluminium-smelter contracts. The same comparison framework that justifies the LV increase as “still cheap by regional standards” would, if applied to HV1, justify a substantially larger HV1 increase as “still cheap by global standards.” The asymmetric framing is the choice that has been made.

A teacher in Wangdue Phodrang opens her phone the next morning. Her bill will go from Nu 600 a month to Nu 1,260. Across the country, in a cement plant in Pasakha, the operations director runs the same math on his facility. His plant’s per-unit electricity cost will go from $0.019 to $0.033 — a 75% increase that still leaves him paying a fraction of what his Indian competitors in Assam, his Bangladeshi competitors in Chittagong, his German competitors in Saxony, and his American competitors in Texas pay. He is still receiving an electricity-cost gift that no other country in the world gives its industrial sector. The teacher’s bill has doubled. His bill has risen by less than a third of the global gap.

Where this came from

The current tariff structure was assembled in pieces:

1. HV1 was originally priced cheap to attract industry. When the BPC-DGPC retail structure was formalised in the 2000s, Bhutan had no large industrial base. The Nu 1.60/kWh HV1 rate (and the demand-charge structure on top) was set as an investment-attraction price — explicitly below cost-of-supply, on the theory that any incremental industrial demand was better than surplus electricity flowing south at PPA prices.
2. LV was priced cheap on social-policy grounds. The 100/200-unit subsidy block plus low above-block rates was a deliberate distributional choice — every Bhutanese household should be able to afford basic electricity. This is normatively defensible.
3. The two were never reconciled when the underlying economics shifted. Once domestic load grew, lean-season imports started, and cost-of-supply rose, the structure had to choose where to land the recovery. The 2026 proposal lands roughly 60% of the absolute revenue uplift on LV and 40% on HV1 — but as percentage of consumption, the LV pool absorbs the entire shock relative to its share.

The structural reason is institutional rather than malicious: HV1 customers are organised, lobbied, and have the optionality to scale back (or, in Bhutan’s sovereign BTC miner’s case, to argue legitimately that they generate USD-denominated returns to the sovereign that justify continued favourable terms). LV customers are 850,000 individual ratepayers who do not coordinate, do not lobby effectively, and do not have an exit option.

Why this matters now

Three converging pressures make this the right moment to surface the global-comparator framing:

1. The proposed revision is currently politically suspended (PM’s May 2026 instruction to ERA for affordability review). The window to revise the revision is open until the proposal returns to MoENR and Cabinet.
2. The export-tariff trajectory is moving. PHP-II’s Rs 5.10/kWh export reference (April 2026 protocol) is now the highest tariff Bhutan has ever negotiated, and the same structure should reasonably constrain how cheap Bhutan can continue to charge domestic industrial users without explicitly subsidising them at the export-revenue level.
3. Global industrial electricity prices are rising, not falling. EIA forecasts US wholesale prices rising 50–80% by 2030; European industrial prices expected to remain elevated post-gas-crisis. Bhutan’s structural competitive advantage is widening, not narrowing — which means HV1 could absorb a meaningfully larger increase without losing its global positioning.

What it should be

• A tariff revision that load-balances the relative pain — e.g., HV1 +100–150% (to Nu 3.50–4.00/kWh = $0.041–0.047) and LV +50–75% (to Nu 4.00–4.50/kWh = $0.047–0.053), preserving Bhutan’s bottom-quartile global position in both segments while spreading the household impact
• A formula linking HV1 tariff to a defined fraction (50–70%) of the latest export-tariff reference price, so when PHP-I, Kholongchhu, Wangchhu, and Sankosh PPAs come up over 2027–2032 with progressively higher reference prices, HV1 tariff escalates automatically
• A transparent published “global percentile” report from ERA showing where Bhutan sits in the world industrial / world residential rankings each year — so the absolute-international-comparison framework cuts both ways and is visible to all parties
• Recognition that the cheap-power moat is durable and valuable, but its distribution across the sovereign, the industrial sector, and households is a policy choice that should be made deliberately rather than inherited

How others do it

• Iceland — large industrial (aluminium smelters) pays competitive contracts; the historical cross-subsidy went from the industrial sector to LV via bundled grid-cost allocation; residential tariffs stayed politically stable through aluminium-funded grid investment
• Norway — Nord Pool spot-passthrough for large industrial; LV faces progressive tiered structure with seasonal variation; industrial sector typically pays more per kWh than residential, not less
• Quebec (Hydro-Québec) — Heritage Pool for residential at deeply subsidised rates ($0.07/kWh); large industrial customers face market-indexed pricing for incremental load; industrial pays the higher rate on the margin
• Paraguay (ANDE) — wholesale industrial rates negotiated bilaterally and adjusted; LV residential politically protected; same structural challenge as Bhutan but tariff debates have generally produced upward-sloping industrial reforms over time
• Bhutan — proposed revision raises LV by 115% and HV1 by 75%, leaving the 23 customers who consume 88% of supply with the world’s deepest industrial subsidy intact at a globally cheap base rate

The question we should be sitting with

If the 23 customers in our country are receiving electricity at rates that no other OECD industrial economy — and no peer hydro-rich country — offers anywhere in the world, on what theory does the system recover its cost by doubling the price for the 99.96% who are already on globally cheap tariffs? Whose comfort does this structure protect, and at whose expense?