• 19 Nov, 2026

Bangladesh multiplied power capacity six-fold after 2009 yet shed load through summer 2026. The sector in full: capacity and the gas-to-coal fuel shift, Payra, Rampal, Matarbari and Rooppur, imports from India and the Adani dispute, capacity charges, subsidies and BPDB losses, with every contested figure attributed.

Bangladesh's power sector presents a paradox that defines the country's energy debate in 2026. On paper, the system has never been bigger: the Bangladesh Power Development Board (BPDB) counted 29,323 MW of grid-connected generation capacity on 30 June 2026, and the Power Division's "at a glance" summary put total capacity at 33,092 MW including captive generation and renewables in July 2026. On 20 May 2026 the grid delivered a record 17,201 MW. Yet in June 2026 the country was enduring load shedding of up to about 3,400 MW, and on 27 August 2026 estimated peak demand reached 19,027 MW, far beyond what the system could supply. Bangladesh does not lack power plants. It lacks, at the hours that matter, the fuel and the money to run them. This article traces how the sector was built, what it runs on, and why its finances have become one of the largest burdens on the state budget.

From Scarcity to Surplus: How the System Was Built

The modern sector dates from independence. The Bangladesh Power Development Board was created in 1972, inheriting a small, war-damaged system from the Pakistan-era power authority, and for decades planned, generated and distributed most of the country's electricity. From the 1990s the board was unbundled, transmission passing to what is now Power Grid Bangladesh PLC, distribution to separate companies, and private independent power producers (IPPs) were invited in. Even so, supply chronically lagged demand. When the Awami League government took office in January 2009, installed capacity stood at under 5,000 MW, fewer than half of Bangladeshis had an electricity connection, and load shedding of 1,000 MW or more on summer days was routine.

What followed was one of the fastest capacity expansions anywhere in the world. Capacity roughly doubled by the mid-2010s and kept climbing through large gas, coal and oil-fired IPPs, cross-border imports and public-sector plants. The government declared 100 percent population coverage by electricity, counting grid and off-grid access, in March 2022. The Power Division put per-capita generation at 680 kWh and the number of consumers at about 5.01 crore (50.1 million) in July 2026. But demand grew nearly as fast, driven by industry, irrigation and cooling: BPDB forecast a summer 2026 peak of around 18,000 MW, against winter demand of roughly half that. The expansion solved the access problem. It created, in the process, a cost problem that now dominates the sector.

The Quick-Rental Turn and Its Discontents

The speed of the post-2009 build-out was achieved partly by bypassing normal procurement. The Quick Enhancement of Electricity and Energy Supply (Special Provisions) Act 2010 allowed the government to award power contracts without competitive tender and indemnified officials for decisions taken under it. Dozens of small diesel and furnace-oil "rental" and "quick-rental" plants were contracted in the early 2010s to relieve the immediate shortage; larger IPPs followed on similar negotiated terms. Supporters argued that the emergency justified the method and that the plants ended the crippling blackouts of the 2000s. Critics, opposition parties, economists and later official reviewers, argued that the contracts locked in high tariffs and, above all, high capacity charges: payments owed to plant owners for making capacity available, whether or not a single unit is generated.

The interim government that took office in August 2024, led by Chief Adviser Muhammad Yunus, repealed the Special Provisions Act in November 2024 and appointed a National Review Committee under retired Justice Moinul Islam Chowdhury to examine the contracts signed under it. The committee's final report, submitted on 20 January 2026, concluded that Bangladesh had built roughly 7,700–9,500 MW of excess or idle capacity, that payments to power plants had risen about eleven-fold over two decades while capacity charges had risen nearly twenty-fold, to as much as US$1.5 billion a year, in the committee's estimate, and it recommended renegotiating the costly contracts and cancelling those where corruption could be proven. The elected government of Prime Minister Tarique Rahman, in office since 17 February 2026, has broadly accepted the diagnosis while moving more cautiously on cancellation, citing the sovereign guarantees attached to the contracts (see below).

What the System Runs On: Gas Declines, Coal Rises

Natural gas built Bangladesh's power system and remains its largest single fuel, but its dominance is ending. Of the 28,494 MW of installed grid capacity cited by the power minister in parliament in April 2026, BPDB data showed about 12,194 MW (43 percent) gas-fired and 7,629 MW (27 percent) coal-fired, with the rest split among oil-fired plants, hydro, renewables and imports. The constraint is supply, not machinery. Domestic gas production peaked at about 2,786 million cubic feet per day (mmcfd) in May 2015 and had fallen to roughly 1,620 mmcfd by October 2026, according to Hydrocarbon Unit and Petrobangla figures reported in the press; the largest field, Bibiyana, produces about two-thirds of its output of a decade ago. Liquefied natural gas, imported since 2018 through two floating terminals off Maheshkhali, now supplies more than a third of the gas grid, about 1,003 mmcfd of a total supply of roughly 2,623 mmcfd on 1 October 2026, against total gas demand estimated at 3,850–4,000 mmcfd. In March 2026 BPDB's chairman said only 800–825 mmcfd was reaching power plants, enough for 4,000–4,500 MW of gas generation from a fleet of more than 12,400 MW; gas supply to power averaged about 904 mmcfd in June 2026. The shock was compounded when QatarEnergy, Bangladesh's largest LNG supplier, declared force majeure on 4 March 2026 amid the Gulf conflict and disruption around Hormuz, and by a fire at a regasification terminal in July 2026.

Coal has filled part of the gap. In April 2026, for the first time, coal-fired plants out-generated gas plants on some days, the Daily Star reported from BPDB data, a milestone in a system where gas once supplied the overwhelming majority of electricity. The fuel bill explains why the mix matters: on BPDB figures for FY2024–25, generating a unit of electricity cost about Tk 7.09 from gas, Tk 13.20 from coal and Tk 27.39 from liquid fuel. Oil-fired plants, the most expensive source, still produced roughly 11 percent of electricity in FY2024–25 because they can be started quickly when gas and coal fall short.

The Big Plants: Payra, Rampal, Matarbari and Rooppur

The expansion's flagships are three coastal coal stations and one nuclear plant, each built with a different foreign partner, megaprojects that, like the Padma Bridge, have become symbols of the country's infrastructure ambitions and their costs.

Payra (1,320 MW, Patuakhali), an ultra-supercritical plant of the Bangladesh-China Power Company, a North-West Power Generation Company–CMC joint venture, entered commercial operation in 2020 and is among the most efficient coal plants in the fleet. In November 2025 China Exim Bank warned that a missed US$129 million loan instalment could trigger a default, according to project reporting, and the interim government sought in late 2025 to renegotiate Payra's tariff using Matarbari's as a benchmark. A second 1,320 MW plant at the same site, built by RPCL-Norinco, has fared worse: it had still not begun full commercial operation in late 2026, one unit generating only limited test power while a coal-supply tender remained unawarded and an evacuation line unfinished, according to Bonik Barta.

Rampal (1,320 MW, Bagerhat), the Maitree Super Thermal Power Project of the BPDB–NTPC joint venture BIFPCL, began commercial operation with its first unit in December 2022. The plant has been dogged since its inception by environmental criticism over its location near the Sundarbans, criticism its sponsors answer by pointing to its supercritical technology and emissions controls, and, in operation, by coal supply failures: it shut down completely in February 2025 after exhausting its coal stock. When fuel is available it is a workhorse: it supplied over 9 percent of national generation in April 2026 and set a plant record of 1,321 MW on 3 October 2026.

Matarbari (1,200 MW, Maheshkhali, Cox's Bazar), the Japanese-financed ultra-supercritical plant of the Coal Power Generation Company, reached commercial operation in 2024. Its first years were troubled by boiler and ash-handling faults, cooling-water failures and coal disruptions that at times cut availability below 55 percent, before it returned to generating more than 1,000 MW in April 2026. In September 2026 the government approved in principle a second phase of 1,320 MW at Matarbari and a matching expansion at Payra, together about 2,640 MW costing over US$3 billion, after Japan declined to finance further coal capacity, leaving the Matarbari expansion to be built with Bangladesh's own funds. Ministers justified the decision on energy-security grounds after the 2026 gas shock; critics argue that adding coal deepens the import dependence and capacity-payment burden the government says it wants to escape.

Rooppur (2,400 MW, Pabna), the Russian-built nuclear plant profiled in our article on the Rooppur Nuclear Power Plant, is the newest source of all. Unit 1 achieved first criticality on 11 October 2026 and entered power start-up; as of mid-November 2026 it had not yet been synchronised with the grid, with officials expecting an initial trial supply of about 300 MW within November, a formal inauguration in December and a climb to the full 1,200 MW by around April 2027. Unit 2 is targeted for commercial operation in December 2027.

Power Across Borders: Imports and the Adani Dispute

Imports supply roughly a tenth of the system's energy at peak. Bangladesh's interconnection capacity with India has long stood at 2,656 MW, about 1,160 MW through the Bheramara HVDC link and the Tripura interconnection, plus 1,496 MW contracted from Adani Power's plant at Godda in Jharkhand, supplemented since 2025 by 40 MW of seasonal hydropower from Nepal wheeled across the Indian grid at 6.40 US cents per unit.

The Adani contract, a 25-year power purchase agreement signed in 2017 without open tender, has become the single most contested document in the sector. The National Review Committee reported in January 2026 that Bangladesh had contracted at 8.61 US cents per unit at a time when other Indian imports cost about 4.46 cents, paying an estimated 4–5 cents per unit extra, US$400–500 million a year, and potentially about US$10 billion over the contract, and recommended that the government consider cancelling the agreement and seeking compensation through arbitration in Singapore; Adani responded that it had not been consulted by the committee and that it had kept supplying power despite large arrears. The dispute has since narrowed to money and method: in a letter of 21 August 2026 Adani claimed US$711.72 million of dues outstanding as of 31 July 2026, against US$384.81 million acknowledged by BPDB; the two sides dispute invoice deductions over coal pricing and capacity calculations, and an expert process under the Singapore International Arbitration Centre is addressing the fuel-pricing questions. Supply from Godda has continued throughout. On 14 November 2026 the energy minister, Iqbal Hassan Mahmood Tuku, explained that the government would not cancel the Indian agreements because of the sovereign guarantees attached to them, and would instead negotiate for a lower price, a position critics describe as accepting the contracts' core terms, and supporters as the only course that keeps both the power and the state's credit intact.

Renewable energy remains marginal in generation despite two decades of programmes. The Sustainable and Renewable Energy Development Authority (SREDA) counts about 1,689 MW of renewable capacity including off-grid installations, of which roughly 1,059 MW is grid-connected, including the 230 MW Karnaphuli hydro plant at Kaptai, in service since the 1960s, and renewables supplied only about 2 percent of the electricity generated in 2025, on sector analyses of BPDB data. The Renewable Energy Policy 2025, adopted by the interim government, targets 20 percent of electricity from renewables by 2030 and 30 percent by 2040; the elected government has kept those targets and added a programme to reach 3,000 MW of rooftop solar by early 2027. The stakes go beyond climate policy, imported-fuel exposure is now the sector's central vulnerability, as our analysis of climate change in Bangladesh also records, but on the record to late 2026, additions have run far behind the trajectory the targets require.

The Money: Capacity Charges, Subsidies and BPDB's Losses

The sector's crisis is, at root, accounting. Because capacity charges are owed whether plants run or not, building far more capacity than the fuel supply can serve converted a fixed share of the power budget into a payment for idleness. On BPDB figures reported by Prothom Alo, capacity charges rose from Tk 250 billion in FY2022–23 to Tk 320 billion in FY2023–24 and about Tk 420 billion in FY2024–25, with projections of a further rise in FY2025–26. BPDB's audited results for FY2024–25 showed a revenue shortfall of about Tk 556.6 billion; a government subsidy of about Tk 386.4 billion covered most of it, and the board still posted a net loss of about Tk 170.2 billion, nearly double the previous year, with cumulative losses over seventeen years put at about Tk 2.67 trillion by campaigners citing its accounts. Losses persist despite steady improvements in some operations: BPDB data put combined transmission and distribution system loss at 10.13 percent in FY2024–25, up slightly from 10.06 percent, with transmission loss alone around 3 percent.

The elected government has chosen price increases as its main corrective. On 3 June 2026 the Bangladesh Energy Regulatory Commission raised the weighted average retail tariff by 16.68 percent to Tk 10.63 per unit and the bulk tariff by 19.85 percent to Tk 8.39, while, after a partial reversal on 4 June, holding the lifeline and first residential slabs unchanged to protect poor households. The finance ministry estimated that the increase would cut the FY2026–27 power subsidy requirement from about Tk 56,000 crore to about Tk 41,000 crore. Consumer groups protested the increase during a period of load shedding; the government answered that a larger subsidy would crowd out other spending, and that tariffs still do not cover the average cost of supply, about Tk 12.10 per unit sold on BPDB's FY2024–25 accounts.

Outlook

The choices now before the sector are unusually stark. The Energy Division told the prime minister in September 2026 that it aims to raise gas supply capacity to 5,819 mmcfd by 2030 through new drilling and LNG, against a record of drilling programmes that have repeatedly under-delivered; Rooppur's two units will add 2,400 MW of fuel-secure capacity by 2027 if commissioning holds; the approved coal expansions would add 2,640 MW more in the 2030s while deepening coal imports. What none of these plans changes quickly is the structure built between 2010 and 2024: a fleet sized for a peak it can rarely serve, paid for by the hour whether it runs or not, and financed by subsidies and tariffs that strain the public finances. Bangladesh electrified itself in fifteen years, an achievement few countries match. The task of the next fifteen is to make that electricity affordable.