IN February, Nepra slashed — from around Rs26 per unit to as little as Rs8-11 — what new rooftop solar owners are paid for the electricity sent back to the grid under a new ‘net billing’ regime, ending a decade of crediting imports and exports equally. Fortunately, existing consumers were spared after public disapproval and reported government intervention. The government’s point is that solar owners, mostly well-off, avoid the fixed and capacity charges embedded in everyone else’s bill, leaving other consumers to cover them. That transfer reached roughly Rs159 billion by 2024 and could pass Rs4 trillion by 2034 if nothing changes. But buried in that logic is a striking admission that ordinary Pakistanis have quietly built something large enough to unsettle the grid and its economics.
They have built roughly 38 GW of distributed solar, more than 90 per cent of the country’s grid capacity. Solar power systems were installed rooftop by rooftop, 96pc funded by the people who installed them. In FY25, they generated an estimated 51 TWh, about 38pc of what the grid sold in the same period. The net billing debate has framed all of this as a liability to be contained by paying owners less. The more useful question is the one no one is asking: what if we stopped treating it as a problem and ran it as a power plant?
Right now, the 38 GW cannot be run as a power plant because they answer to no one. Millions of systems inject power when the sun is high and disappear when it sets. The load data shows it: between July 2024 and July 2025, midday national demand fell nearly 2,000 MW as rooftops did the grid’s daytime work, then everyone rushed back at dusk. This is the ‘duck curve’, which has broken the economics of thermal generation. Some 11.3 GW of RLNG and imported coal sat largely idle throughout FY25, yet had to be paid for. Capacity payments — the fixed charges that plants earn merely for existing — reached Rs1.8tr, which is over 60pc of the power purchase price. Net billing compounds this, pushing solar owners to add batteries and self-consume. Each step takes them closer to leaving the grid entirely, which makes it worse.
There is an established alternative that does the opposite of pushing people away: a virtual power plant. A VPP is not a building; it is software. It aggregates thousands of small assets, from rooftop solar, home and industrial batteries and EV chargers to air conditioners and agricultural tube wells. It then coordinates them so that they act as a single dispatchable plant for the grid. Charge batteries and vehicles during the midday solar glut; discharge them into the evening peak, when prices are highest and the system is most strained. At scale, Pakistan’s rooftops and driveways become a peaking plant, supplying the evening capacity that the country currently pays thermal plants huge charges for, whether or not they are operating.
The government is cutting what rooftop solar owners earn for a problem that better coordination, not lower payments, would solve.
This is not theoretical. In the UK, Octopus Energy runs a VPP of 150,000 enrolled EVs that shifts 1 GW of flexible load, enough for two cities on a typical evening, while saving participants around £600 a year. American utilities show how such schemes mature. Duke Energy offers a flat monthly subscription in return for managing a household’s charging. Colorado’s Xcel Energy goes further by using charging that follows a price signal published a day ahead. These remain the pilots of a few hundred to a couple of thousand vehicles, a reminder that this is early, deliberate work, not a leap of faith.
This progression is a template that Pakistan can follow, and it need not start with the most sophisticated version. A sensible first step is a pilot within a single Disco, enrolling households that already own solar and batteries and run occasional but clearly signalled events. As smart meters and time-of-day tariffs spread, a subscription-styled managed model becomes possible. Only when the wholesale market reforms under the CTBCM (Competitive Trading Bilateral Contracts Market) take hold does the fully dynamic version, in which charging responds to live prices, become realistic. Moving up the ladder needs the same enablers: interoperable standards so that assets can be pooled regardless of brand; a clear licensing route for aggregators; and, above all, a way to pay people for the flexibility they provide.
This is where the contrast with net billing is at its sharpest. Net billing treats solar owners as a cost to be trimmed; VPP treats them as a resource worth paying for, ie, being available when the grid needs them. The prize is real. Renewables First’s modelling of a 2030 EV fleet finds that managed daytime charging, paired with vehicle-to-grid discharge, cuts curtailment and reduces the evening peak by 2.3 GW. Left uncoordinated, the 6.7m vehicles by 2030, as forecast by Lesco, would push the evening peak from 25.7 GW to 28 GW. Coordination is the entire difference between EVs as a fresh crisis and EVs as a solution, and the same holds for solar already on the roofs.
Pakistan is at a familiar crossroads. In 2022, distributed solar arrived without a plan, and the state is still scrambling to absorb it; net billing is the latest scramble. Now a second wave of distributed assets, batteries and EVs is building on the same roads and rooftops. The country can meet the challenge as it faced the first one: by treating what its citizens built as a threat, taxing and discounting it until the people disconnect from the grid. Or it can build the coordinating layer that turns 38 GW of scattered solar power, and the batteries behind it, into the cheapest peaking plant it will ever own. The government has spent this year deciding how little to pay people for their power. The better question is how to make that power worth paying for.
Muhammad Taha Ali is associate, Power Markets, Renewables First.
Abdul Rehman is manager, Power Markets, Renewables First.
Published in Dawn, August 12th, 2026
