The solar rush is not a disease. It is the symptom

The loudest conversation in energy circles this month is being driven by a single statistic. A joint study by Ember and Renewables First estimates that distributed solar across residential, industrial, agricultural and commercial rooftops now totals around 38 gigawatts. Official net metering records capture only 6.8 gigawatts. The remaining 31 gigawatts has gone dark to the system operator. It is generation that exists, that consumers paid for out of their own pockets, and that the grid cannot see or plan around.

The instinctive reading of this gap is that Pakistan is succeeding at clean energy faster than its institutions can keep up. A harder reading is that something has gone wrong on the grid, and 31 invisible gigawatts is the evidence. I want to argue for a third reading, sharper than either. The solar rush is not what is damaging the grid. It is the trace left behind by a tariff that gave consumers no rational reason to stay on it. Solar is the symptom. The disease is the price of staying connected.

To see why the disease sits in the price and not in the panels, ask why the rush happened in the first place. It did not happen because our nation suddenly developed an environmental conscience. It happened because grid electricity became too expensive, and it became too expensive partly because the fixed costs of capacity payments and the surcharges that service circular debt are loaded onto every unit sold. As tariffs rose, consumers who could afford to install their own generation bought fewer units. With fewer units sold, the same fixed costs had to be recovered from a smaller base, pushing the per unit tariff higher still, which encouraged the next wave to leave. This is a classic utility death spiral. Solar did not cause it. Solar is the exit that became affordable once staying connected was made expensive enough to justify leaving.

The public conversation usually lands on capacity payments at this point, and it lands in the wrong place. Capacity payments are denounced as a scam in which the country pays plants that do not generate. The criticism collapses two separate questions into one. The first is whether capacity payments are legitimate in principle. They are. No investor commits hundreds of millions of dollars to a plant on a-pay-only-when-needed basis. A capacity payment recovers debt service, return on equity and fixed operating costs whether or not the plant runs. Abolishing the principle would simply make the country unfinanceable for the next round of plants we still need.

The harder question is whether the specific payments agreed to were well designed. They were not. The wave of independent power producers contracted since 1994, carried dollar indexed tariffs, take or pay structures, tax benefits and waivers, and return on equity allowances that, with hindsight, were unusually generous. Those terms, not the principle, turned a legitimate reservation charge into a fiscal wound of roughly PKR 2.1 trillion by 2024.

There is a pattern here worth noticing, and it is one of circumstance more than fault. A decade ago, faced with crippling load shedding, the sector signed the contracts that ended the shortage and kept the lights on. That was a real achievement under real pressure. The terms that came attached, the dollar indexation and the guaranteed returns, are the burden the country carries now, and the same institutions are working to unwind it through renegotiation and conversion to take and pay. The point is not that anyone acted in bad faith. It is that a remedy built for one crisis can quietly set up the next, which is reason enough to look carefully at the remedies being applied today.

Renegotiating power purchase agreements, suspending captive gas to push industry back onto the grid, and tightening rules on rooftop exports are all defensible responses to the immediate problem. They are also signals to future investors and to existing industrial consumers, and the signals being sent will be read again in the 2030s, when demand growth resumes and the country needs new firm plants. When contracts are reopened, even for sound reasons, the cost of future capital tends to rise. The decisions that will shape supply in the 2030s are being taken now, and they deserve as much care as the pressures of the present.

Even with those corrections in flight, the tariff that drove consumers off the grid does not heal, and the people who left are a particular group. Net metering data shows adoption heavily concentrated in major cities, with Lahore alone accounting for roughly a quarter of all consumers, followed by Multan, Rawalpindi, Karachi and Faisalabad. These are, in the main, urban consumers with capital to install rooftop systems. As they exit grid dependence, the fixed costs of the system do not shrink. They are redistributed across the consumers who remain, who are disproportionately those without the means to install their own generation. The clean energy story has a regressive underside that the celebratory framing rarely acknowledges.

This brings us to the regulator’s response, and to a measure that may work against its own intention. In February 2026, NEPRA moved to slow the exit by ending the old net metering offset and introducing a net billing mechanism, cutting export buyback rates sharply for new applicants and shortening contracts from seven years to five. The intention is to protect the grid revenue base. The effect runs in the opposite direction.

Net billing makes export uneconomic but does nothing to make grid import attractive. The rational consumer response is to stop sizing systems for export and start sizing them for self-consumption, and the natural next step from self-consumption is battery storage on the rooftop. Once a consumer has solar plus a four hour battery, daytime generation can be stored and used in the evening. The grid then loses not only the daytime demand it has already lost. It begins to lose the evening peak as well. What started as a quiet exit during sunlight hours ends with the grid reduced to a wet weather standby utility, drawn on only when the sun fails and the battery runs empty. The defection does not reverse under net billing. It goes invisible, and then it goes deeper.

The other prescription on offer is the Competitive Trading Bilateral Contracts Market, regularly presented as the structural cure for both capacity payments and tariff inflation. The real difficulty with CTBCM points in the same direction as net billing. Despite approval from the Economic Coordination Committee and the regulator, and a completed six month trial, the market remains barely operational. The use of system charge has not yet been set at a cost reflective level, and the initial wheeling allocation is a modest 800 megawatts. The design itself, though, is more complete than its critics allow. It provides for firm capacity certification, a forward capacity obligation on suppliers, and a capacity price benchmarked to the cost of new build. Even so, energy contracts will gravitate toward the cheapest megawatt hour on offer, which is solar and wind.

A capacity obligation can compel a supplier to contract firm megawatts. It cannot, on its own, make that supplier a counterparty a lender will accept for a twelve year hydropower loan, when the distribution companies that carry the obligation hold weak balance sheets. The duck curve still gets steeper, not flatter, because the energy that clears the bilateral market is the cheapest, which is solar. At the same time, the industrial and commercial wheeling customers, who currently cross subsidise the residential base, will exit through CTBCM to chase cheaper bilateral solar, leaving DISCOs with the residual load. The death spiral does not slow under an incomplete CTBCM. It accelerates. The sober conclusion is that CTBCM in its present design should either be paired with a Use of System Charge that fully captures the embedded capacity cost the exiting consumer is escaping, or restricted to new demand growth rather than allowed to extract existing demand from the grid. Neither restriction is currently in the design.

Each of the three interventions on offer, capacity payment renegotiation, net billing and market reform, is therefore being applied to a symptom. None of them changes the underlying signal sent to a consumer deciding whether to remain on the grid. That signal is the per unit tariff, and the per unit tariff carries the system’s entire fixed cost burden. So long as it does, the consumer with capital will keep finding the exit rational, the regulator will keep losing visibility over distributed generation, and the burden of what remains will keep concentrating on those who could not leave.

Treating the disease itself requires a redesign of how the tariff is built, and the architecture is neither exotic nor untested. It begins with a basic separation that the present tariff refuses to make. Fixed costs of the system, capacity payments, transmission, distribution and the cost of past circular debt, should be recovered through a fixed monthly charge paid by every grid connected consumer based on the size of the connection they hold, regardless of how many units they import. Energy costs, the actual cost of generating each kilowatt hour, should be recovered through a per unit tariff that varies by time of day, rising in the evening when the system is tight and falling at noon when solar is in surplus. Wheeling under CTBCM should carry a Use of System Charge that captures the embedded fixed cost the wheeling consumer would otherwise have paid. And the protection of genuinely poor households, who use little electricity in any case, should be preserved through a lifeline block within the energy tariff or, more honestly, through a targeted cash subsidy outside the tariff altogether so that the cross subsidy stops being invisible.

This architecture does several useful things at once. It recognises that a grid connection is itself a service worth paying for, the insurance against the cloudy day, the failed inverter and the low solar winter month. It removes the tariff penalty on the heavy daytime grid user, which removes part of the incentive to exit. It rewards the consumer who self generates but stays connected, because evening grid import is priced fairly and the daytime fixed charge keeps the system whole. It makes distributed generation visible again because the rooftop consumer is no longer trying to escape the system but contracting with it on honest terms. And it survives the arrival of rooftop batteries in a way the current tariff does not, because the consumer who shifts the entire evening peak to a battery still pays for the standby they will need on the days the battery cannot cover.

The work is political before it is technical. A fixed monthly charge that arrives on every electricity bill, paid by households that did not have one before, will be denounced as a tax on the poor unless it is paired explicitly with a properly designed lifeline block or a direct subsidy outside the bill. Time of use pricing will be denounced as punitive unless the daytime fall is explained alongside the evening rise. NEPRA will need to publish a transparent cost of service study that shows, for every consumer class, what portion of the bill is fixed cost and what portion is energy. None of this is comfortable. All of it is preferable to the alternative, which is to keep adjusting export buyback rates, contract durations and wheeling allocations while the grid hollows out beneath the policy.

We do not have a solar problem. We have a pricing problem that solar has made impossible to ignore. The 38 gigawatts on rooftops are not the disease. They are the diagnosis the country has so far refused to read. The cure is to charge properly for the privilege of staying connected, and to do it before the next generation of consumers, with batteries on their roofs, decides that the privilege is no longer worth holding on to.