Every monsoon, Nepal produces more electricity than it can use or sell, and some of it is simply spilled over the dams. Every winter, we import power from India. Bhutan, with a similar hydropower story, turned its surplus into Bitcoin through its state investment company. So the obvious question keeps coming up in Kathmandu: could Nepal do the same?
I didn't want a yes or no based on hype, so I did the maths with real numbers: Bitmain's published spec for a current mining machine, the live Bitcoin network data, today's Bitcoin price, and what NEA actually earned per unit from exporting electricity last year. The answer is more interesting than either side expects.
Quick answer
At 22 November 2026 numbers, one megawatt of current Bitcoin miners (Antminer S21 XP Hyd, 12 J/TH) earns about US$0.10 per kWh from block rewards, roughly double the Rs 7.56 (about US$0.051) per unit NEA earned exporting power in FY 2025/26. But the machines cost about US$1.85 million per MW, and the April 2028 halving cuts that income to about US$0.051 per kWh, the same as exporting. So mining makes sense for Nepal only on truly spilled energy, with private investors carrying the hardware risk. Mining is illegal in Nepal today.
Note: this is an analysis of a policy option, not investment advice. Crypto mining is currently illegal in Nepal under Nepal Rastra Bank rules.
Why mining and hydropower fit together
The Bitcoin whitepaper itself says the cost of creating new coins is "CPU time and electricity", comparing it to gold miners spending resources to dig up gold. Unlike a factory, a Bitcoin miner doesn't need roads, workers or a market nearby. It needs cheap power and an internet link, and it can be switched off in seconds without harming anything. That makes it unusual: a buyer you can put right next to a remote power station and turn off the moment the grid needs the electricity back.
I covered the bigger question of whether Nepal should legalise Bitcoin, with the pros, cons and a gold comparison, in Bitcoin in Nepal: why Nepal should legalise it. This post is only about the energy maths.
Nepal's surplus problem, in numbers
| Fact | Number |
|---|---|
| Wet-season 2025 generation surplus | About 1,400 MW |
| India's export approval at the time | 1,141 MW (the gap was spilled) |
| Approved export capacity now | About 1,179 MW |
| Exports in FY 2025/26 | 3.877 billion units for Rs 29.32 billion |
| Imports in FY 2025/26 (mostly winter) | About 1.15 billion units |
| Average export price | Rs 7.56 per unit (my calculation: Rs 29.32 billion divided by 3.877 billion units) |
| Capacity damaged by the August 2026 floods | About 430 MW |
Sources: IPPAN on the 2025 surplus, The Nepal Weekly on FY 2025/26 trade, KathmanduPati on export capacity and the Kathmandu Post on the floods.
So the real picture is seasonal. In the monsoon we have more than we can sell; in winter we buy. Any mining plan has to respect that, or it ends up competing with households for winter power.
What Bhutan actually did
Bhutan's sovereign investment arm, Druk Holding & Investments, partnered with the Singapore-listed miner Bitdeer. According to DatacenterDynamics, Bitdeer finished a 100 MW site at Gedu in 2023, and the partners announced a fund of up to US$500 million and a larger 500 MW site at Jigmeling.
It worked for a while. Blockchain tracker Arkham estimated Bhutan's holdings peaked near 13,000 BTC in October 2024. By June 2026, The Block reported its tracked wallets held about 1,750 BTC after large transfers to exchanges. Bhutan's officials have said they don't recall selling, and on-chain data can't prove a sale. Either way, the lesson is that mined Bitcoin becomes a balance-sheet asset with its own price risk, separate from the power business.
The maths for one megawatt
Here's the calculation step by step, so you can redo it with tomorrow's prices.
Step 1: The machine
Bitmain's official spec for the Antminer S21 XP Hyd (a water-cooled model, which suits hydropower sites) lists 473 TH/s at 5,676 watts, an efficiency of 12.0 joules per terahash. One megawatt runs about 176 of them, giving about 83.3 PH/s of hashing power.
Step 2: The network
On 22 November 2026, the Bitcoin network ran at about 898 EH/s (blockchain.info data; you can track it live on mempool.space's hashrate chart), at block height 977,004. Each block pays a 3.125 BTC subsidy, and there are about 144 blocks a day, so the network issues about 450 new BTC daily, plus transaction fees.
Step 3: Your share
83.3 PH/s divided by 898 EH/s is about 0.0093% of the network. That share of 450 BTC is about 0.0418 BTC a day. At US$58,859 per Bitcoin, that's about US$2,458 a day, from 24,000 kWh of electricity.
Machines per MW = 1,000,000 W / 5,676 W = about 176
Hashrate per MW = 1,000,000 W / 12 J/TH = 83.3 PH/s
Network share = 83.3 PH/s / 898 EH/s = 0.0093%
BTC per day = 450 BTC x 0.0093% = 0.0418 BTC
Revenue per day = 0.0418 x US$58,859 = US$2,458
Revenue per kWh = US$2,458 / 24,000 kWh = US$0.102
That's before transaction fees (which add a little) and before pool fees, cooling, staff and building costs (which take a lot).
Mining vs simply exporting the power
| Use of 1 MW for a day (24,000 kWh) | Per kWh | Per day |
|---|---|---|
| Export to India at last year's average | Rs 7.56 (about US$0.051) | About US$1,221 |
| Mining today (subsidy only) | About US$0.102 | About US$2,458 |
| Mining after the 2028 halving, same price and hashrate | About US$0.051 | About US$1,229 |
| Spilled energy (no buyer) | Zero | Zero |
Dollar conversions use Rs 148.6 per US dollar on 22 November 2026.
Today, mining earns about twice what exporting does per unit. That's the headline people like. But look at the third row. The next halving is about 73,000 blocks away, roughly 507 days, which lands around April 2028. If the Bitcoin price and the network's hashrate stayed exactly where they are, mining income per unit would drop to almost exactly the export price.
The catch: hardware cost and the halving
Exporting power needs a transmission line Nepal already has. Mining needs machines. Using a September 2026 retail listing of about US$10,500 per S21 XP Hyd, 176 machines cost about US$1.85 million per megawatt, before import duty, transformers, cooling and the building.
Compared with exporting, mining today earns about US$1,237 a day more per megawatt (US$2,458 minus US$1,221). Even running all year, that extra covers the machines in a bit over four years, and the halving arrives in under a year and a half. Running only in the monsoon, which is what Nepal's grid can actually spare, makes it several times longer. The machines would be obsolete before they paid for themselves against simply exporting.
Two things could change this: a much higher Bitcoin price, or energy that truly has no other buyer. The price I can't predict. The spilled energy is real, though.
Where it does make sense: spilled energy only
When power is spilled, the comparison isn't mining versus exporting. It's mining versus zero. The 2025 gap between surplus (about 1,400 MW) and export approval (1,141 MW) was roughly 260 MW in peak months. For energy like that, any buyer beats the spillway.
So my recommendation is narrow:
- Private money, private risk. Licensed miners buy their own machines. NEA and the state don't.
- Interruptible contracts only. NEA sells only power it would otherwise spill, and can cut miners off with minutes of notice. Winter household demand always comes first.
- A floor price above zero. Charge miners something close to the export price, or a share of revenue, so the country earns from water it was wasting.
- Sites next to stranded plants. Put miners at projects stuck behind missing transmission lines, which is exactly where spill happens.
- Tax and report it. Mining income should be taxed and reported like any export.
That turns mining into a flexible load that soaks up spill, instead of a bet on the Bitcoin price funded by the public.
What could go wrong
- Price falls further. Bitcoin is already about 53% below its October 2025 peak of US$126,080. A further drop makes even spilled-energy mining marginal.
- Hashrate keeps rising. More miners worldwide means a smaller share for everyone, the same as a price drop.
- Floods. The August 2026 floods damaged about 430 MW of capacity. Sites next to rivers carry the same risk.
- Winter creep. Once machines are installed, there will be pressure to keep them running in winter. The contracts have to be strict from day one.
- The law. Mining is illegal in Nepal today. None of this happens without a clear legal framework first.
A few things I learned
- Mining revenue per kWh is easy to calculate. Machine efficiency, network hashrate and price are all public; anyone can check a promoter's claims.
- The halving changes everything. At today's price, it would cut mining income to about the export price.
- Spilled energy is the only clear win. Against zero, mining always earns something; against exporting, it often doesn't.
- Bhutan's lesson is about the asset, not the power. Mined coins carry their own price risk once you hold them.
For the full picture on legal status, gold and adoption, read the companion post, Bitcoin in Nepal: why Nepal should legalise it. I've also written about surviving as a full-time crypto trader and how DeFi lending health factors work. If you're planning data centre or high-power infrastructure in Nepal, my infrastructure consulting covers the networking and systems side.
Frequently asked questions
Can Nepal mine Bitcoin with hydropower?
Technically yes, but it is illegal under current Nepal Rastra Bank rules. Economically, one megawatt of modern miners earns about US$0.10 per kWh at November 2026 prices, roughly double Nepal's average export price of Rs 7.56 per unit. Hardware costs and the 2028 halving mean it only clearly pays on energy that would otherwise be spilled.
How much does Bitcoin mining earn per kWh?
With an Antminer S21 XP Hyd at 12 J/TH, a network hashrate of about 898 EH/s and Bitcoin at US$58,859 on 22 November 2026, mining earns about US$0.102 per kWh from block rewards, before pool fees, cooling and other costs. After the April 2028 halving it would be about half that at the same price.
How much electricity does Nepal waste?
In the 2025 wet season, Nepal's generation surplus reached about 1,400 MW while India's export approval was 1,141 MW, so roughly 260 MW was spilled at peak, mainly at NEA plants. In winter Nepal still imports power; it bought about 1.15 billion units in FY 2025/26.
How did Bhutan mine Bitcoin?
Bhutan's state investment company, Druk Holding and Investments, mined Bitcoin with hydropower, partly through a partnership with Bitdeer, which completed a 100 MW site at Gedu in 2023. Arkham estimated Bhutan's holdings peaked near 13,000 BTC in October 2024 and its tracked wallets held about 1,750 BTC by June 2026.
When is the next Bitcoin halving?
The next halving happens at block 1,050,000. On 22 November 2026 the chain was at block 977,004, about 73,000 blocks or roughly 507 days away, which points to around April 2028. The block subsidy will then fall from 3.125 BTC to 1.5625 BTC.
