Sunshine is free, keeping lights on is not

When the National System Operator (NSO) and Energy Minister Anura Karunathilaka had to ask rooftop solar owners on Saturday to switch off their panels so the grid could be restored, the rooftop solar debate should have changed for good.

The trigger is still under investigation, and early accounts point to a conventional power plant, not solar. But the question is no longer whether Sri Lanka wants more solar. It is who pays to keep a solar-heavy grid standing, and whether those best able to pay are doing so.

Other countries have already confronted that question. California cut export credits by around 75% in 2023 after its regulator found non-solar customers were subsidising solar owners. The Netherlands will end net metering on 1 January, 2027. Pakistan replaced it with net billing for new connections this year. Australia lets network operators switch off household systems remotely in emergencies. None of these countries abandoned solar. They stopped treating the grid as a free battery.

Sri Lanka has had its warnings. After the blackout of 9 February, 2025, the Ceylon Electricity Board’s expert panel blamed low inertia. Inertia is the cushioning energy stored in the spinning turbines of large plants, and rooftop panels do not provide it.

The Public Utilities Commission of Sri Lanka (PUCSL) rejected that report. Yet its own February 2026 directions dealt with inverter settings, reserve generation and load shedding, problems that sharpen as solar grows. A paper hosted by the Asian Development Bank found 11% of 1,608 sampled solar inverters set to disconnect too early in a disturbance. More than 1,000 MW of rooftop solar has been added since 2018, yet the automatic load-shedding scheme was last updated that same year.

That exposes the first myth: that abundant sunshine means cheap electricity. Panels are cheap. Integrating them is not. The midday surplus must be absorbed, stored or curtailed, and the evening peak met with finite reservoir water and costly oil. With coal out, oil’s share of a Sunday’s electricity rose from 0.8% on 11 October to 23.7% on 18 October. Every consumer pays for that.

The second myth is that batteries will soon fix this. The grid-scale battery fleet is about 30 MW, roughly 1% of a 2,800 MW evening peak. Larger schemes still on a procurement list supply nothing tonight. Storage takes capital, land and years. Someone must finance it.

Clean energy advocates argue, fairly, that the deeper failure is management. Recommendations from earlier blackouts went unimplemented, and the PUCSL’s directions carry no deadlines or penalties. But fixing that is not free either.

Sharper protection, faster load shedding, standing reserves and batteries are precisely what a solar-heavy grid needs more of. Ireland draws up to 75% of its instantaneous power from wind and solar, but only by procuring such services. Sri Lanka has no market for them yet.

That brings the argument to who pays. Under net metering, a household that can afford a Rs. 1.5 million to Rs. 2 million system sells power at noon, when the grid has too much. It draws power at night, when electricity is dearest to supply.

Net metering credits one against the other as though they were worth the same. They are not, and the difference is carried by consumers who will never own a panel.

Sri Lanka should stop treating the grid as a free battery too. Affluent prosumers, consumers who also produce power, should either store their own surplus and pay a cost-reflective price for night-time backup, or leave the grid. They should not export when power is worth little and import when it is costliest, while poorer consumers make up the difference.

The real debate is not net metering versus Net Plus. It is about the true cost of a stable grid, and who carries it. That should be settled with independent numbers, not slogans from either side.