Every month, on every electricity bill in Pakistan, sits a small ‘line item’ called the Neelum–Jhelum surcharge. We have paid for it for years.
Ask yourself a simple question: what did your last payment buy? The answer: since May 2024, nothing. The 969MW Neelum–Jhelum plant has not produced a single unit of electricity for more than two years. Wapda told the Senate this month that it may not produce one until March 2028.
How did we get here? The short version is a story we know too well in the land of the pure. A project conceived in the 1980s, approved in 2002 at Rs85 billion, awarded in 2008 before land, money or consultants were in place, and finished in 2018 at Rs507 billion – among the costliest power projects, per megawatt, anywhere in the world. Its heart is a 48-km tunnel bored through the weakest rock in the Himalayas, across an active fault line, as deep as 1,900 meters under the mountains.
That tunnel broke in July 2022. It was patched. In April 2024, it broke again – exactly where the project’s own panel of experts had warned, in writing, that it would. The warnings were not acted upon. Now we are told a Rs22 billion repair is possible, but Wapda itself will not promise that it will last six years. The alternative on the table is to rebuild the tunnel for Rs250-300 billion. Through the same rock. Under the same fault.
Before we reach for the chequebook again, let us do something this project has never enjoyed: honest arithmetic. Rupees from 2010 and rupees from 2024 are different animals, so convert everything into today’s dollars. Pakistan has sunk $6-7 billion of real resources into this mountain. All the electricity the plant produced in its entire life is worth $2-2.75 billion. The two tunnel failures have burned up another billion or two in repairs and replacement fuel. Even if the plant returns in 2028 and never breaks again – a heroic assumption – the books do not balance until the mid-2030s at best, and on the plant’s own tariff, somewhere between 2050 and never. This, for a project sold to the nation on a five-year payback.
But there is a second ‘ledger’ – and nobody in Islamabad has ever opened it. To feed the tunnel, the project takes nearly all the Neelum’s water at Nauseri which thereafter leaves the river for the 41km down to Muzaffarabad with roughly three per cent of its flow. A city built at the meeting of two rivers watched one of them die. Springs dried. The drinking water system failed and was rebuilt at public cost. For seven years the people of Muzaffarabad have marched under a banner that should shame us all: Darya Bachao, Muzaffarabad Bachao. Save the river, save the city.
What is a living river worth? This is not a poet’s question; treasuries and courts around the world price such things routinely, and the methods are standard. Count the drinking water, the fisheries, the tourism of the most beautiful valley in Pakistan, the springs, the very climate of a riverfront city – and then add what ordinary Pakistanis would willingly pay simply to know the Neelum flows again. The honest range is $26-300 million, every year, forever, with certainty. No tunnel collapsing. No fault line to fear.
Now the choice becomes clear and clarifying. Rebuilding is, at best, a bet: pay nearly a billion dollars for electricity that arrives only if the new tunnel survives the geology that has already broken it twice. Restoration is a certainty. Run the numbers over 25 years and the logic reduces to a sentence. If the living river is worth only the lowest honest estimate – $26 million a year – then rebuilding makes sense only if the new tunnel has a better-than-roughly-one-in-four chance of surviving a generation. If the river is worth the middle estimate of about $90 million a year, the tunnel needs better than a one-in-two chance. And if the river is worth the highest defensible estimate – $300 million a year – then no odds are good enough: giving the river back beats the rebuild even if the new tunnel were certain to hold.
Would any engineer, hand on heart, certify even the one-in-two today? Nobody knows, because in two years of inquiries, committees, and a 100-million-rupee report a minister famously matched against a free, open AI estimate, the state has never commissioned the one study that matters: an independent, numerical assessment of the tunnel’s odds.
The objections write themselves, so let us answer them. Do we owe money on the plant? We owe it either way. The Chinese credits were secured on your surcharge, not on the turbines turning. Energy security? Friends, the experiment has already run: the grid lost Neelum–Jhelum in May 2024 and the lights stayed on, because Pakistanis have, in parallel and despite this project, been executing the fastest rooftop solar revolution on earth – 45GW of panels imported in five years – while grid demand actually fell. The rebuilding money would buy nearly twice as much energy from solar, delivered in a third of the time, on ground that does not move.
A precedent against dams? The opposite. Diamer-Bhasha and Mohmand store water and irrigate farmland; Dasu keeps the Indus in its own bed. Neelum–Jhelum stores nothing, waters nothing, and is the only project in Pakistan that took a river away from its people. Retiring it through the proposed Dam Safety Council, on specific geological grounds, would tell the world our dam programme has quality control. And quietly, a Pakistan that restores its own river stands taller in every water proceeding with our upstream neighbour, where our case has always rested on the sanctity of the Neelum’s flow.
So here is the ask, and it is deliberately modest. Not demolition. Not reconstruction. Homework. Before one more major rupee moves, commission three independent studies: the honest probability that a rebuilt tunnel survives; the honest value of a restored river to the people of Pakistan; and an honest accounting of what can still be recovered from contractors, consultants and insurers. Cost: about $3 million – one rupee of homework for every two thousand already at stake. Publish all three. Then decide, in the open, on evidence.
We have already paid twice to decide on this project the other way. The river, at least, is still willing to come home.
The writer is the president of Reenergia Impact (www.reenergia.com).
The full report, ‘The Neelum–Jhelum Question’, with technical annexes, is being released publicly. Research and analysis were supported by Fable (Claude), an AI system by Anthropic based on analytical hypotheses designed and operated by the writer.
