There has been a solar revolution in Pakistan, driven by lower solar panel prices and rising grid tariffs, creating a perfect recipe for a surge in solar additions.
There have been two types of additions: net-metered and behind-the-meter. Net-metered additions are connected to the grid, in which the grid buys electricity from such customers at a price well above the prevailing marginal cost. Meanwhile, behind-the-meter refers to adding solar without grid connectivity. The emergence of both types has led to more than 25 GW of solar capacity available in the country, and this is expected to increase further.
Due to such additions, Pakistan’s load curve has started to look very different from what our tariff design assumes, which is still stuck in the last century. As the sun rises, solar capacity kicks in from 7am to 3pm, reducing grid demand. It is during this same interval that the grid buys surplus electricity from net-metered customers, while the marginal price is at its lowest.
Surplus supply during daylight hours reduces the marginal price of electricity. But due to a rigid and impractical tariff regime, the benefits cannot be captured, and a lower marginal price can not be passed on to consumers.
Changing the tariff structure could allow an equilibrium price that mimics the actual demand-supply situation rather than one assumed in a silo somewhere
Behind-the-meter solar and batteries are eating into daytime demand, and the existing cost-plus tariff is not designed to address it. The current tariff structure effectively divides all variable and fixed electricity costs across an assumed number of demand levels. This may have worked when there was no decentralised electricity generation, only till a few years back, but it does not work for an environment where decentralised generation continues to increase its share of the generation pie.
It is important to pivot quickly to a subscription-style pricing model, where customers pay primarily for capacity and network, not just for units consumed, in the absence of which the system will keep raising tariffs for a shrinking base of kilowatt-hour (kWh) sales.
Such sales will continue to contract as the share of decentralised generation increases, making the grid less tenable for all consumers and further accelerating migration from the grid. That is a recipe for faster grid defection, more behind-the-meter investment, and a larger circular debt stock.
The grid is increasingly full at night and in the evening, but hollowed out in the middle of the day. That valley during the middle of the day is a classic “duck curve”, which entails high demand at the shoulders of the day, a pronounced dip when solar is strongest, and then a steep evening ramp when solar output collapses and everyone turns lights, motors and air conditioners back on. The system is paying for capacity it cannot use in the middle of the day, and then scrambling to recover fixed costs from fewer units sold.
On the storage side, lithium-battery import data suggests that by mid-2025, Pakistan will already be sitting on roughly 18 GWh of stationary energy storage. As battery prices continue to decline, battery storage is expected to reach more than 75 GWh by 2028, further risking defection from the grid.
Consumers do not need to completely move away from the grid; every minute or kWh they use either decentralised generation or a battery is a minute or kWh taken away from the grid, which means lower grid utilisation and, inadvertently, higher prices for everyone else. It is important to look at things at the margin here, to identify the inflection point, rather than using a simple average that masks distortions.
Based on current trends, we will soon have the technical ability to mobilise 7–15 GW of flexible, distributed capacity for two hours during peak times, if even a fraction of these batteries are aggregated and dispatched as part of micro-grids or virtual power plants.
If tariff design does not respond, the load curve will become more extreme, resulting in deeper midday troughs, sharper evening ramps, and progressively lower utilisation of grid-supplied energy per rupee of fixed cost locked into the system.
The existing tariff philosophy still behaves as if fixed and variable costs are recovered primarily through per-kWh charges. Total power payments are estimated at Rs3.5 trillion, of which the power purchase price is about Rs3.1tr. Within this, capacity charges alone are roughly Rs1.8tr, just over half of the entire system’s revenue requirement. These are obligations Pakistan must pay, whether grid sales are 120 terawatt-hour (TWh) or 90 TWh.
Dividing this over a lower denominator spells disaster for all consumers, or even taxpayers in the process, since any policy disaster is eventually funded by the taxpayer, either through higher taxes, or by the population, through higher inflation.
Instead of recovering capacity, network and social-policy costs inside an opaque energy rate, the tariff must be unbundled into three transparent components.
Firstly, a capacity subscription (CS), which refers to a fixed rupee-per-kWh monthly charge that recovers generation capacity costs. Secondly, a network charge (NC), which can be a fixed rupee-per-kWh per month that recovers transmission and distribution costs. And lastly, an energy charge, which can be a rupee-per-kWh rate that tracks the system’s shortrun marginal cost (SRMC) and varies by time of day.
Moving to a marginal cost regime will eventually lead to an equilibrium price that mimics the actual demand-supply situation rather than one assumed in a silo somewhere.
The economic logic is straightforward: CS and NC are billed on the higher of the sanctioned load or the measured maximum demand. This anchors fixed-cost recovery in kilowatts of subscribed capacity, not in how many kilowatt hours a user happens to import in a given month.
The energy charge is allowed to fall toward the real midday marginal cost of generation, which recent data puts at low-teens per kWh, about half the evening SRMC.
This encourages industry to move energy-intensive processes into the daytime, soaking up surplus hydel, nuclear, and solar. This is exactly how most of us already buy broadband: we pay a monthly fee for a 50 Mbps pipe, not for the exact number of packets we consume. The provider’s revenues depend largely on subscribed capacity, not on whether we streamed brainrot for 10 hours or two.
Pakistan’s duck curve is not going away. The only question is whether we let it keep eroding the financial foundations of the power system, or whether we redesign tariffs to live with, and benefit from, a future where solar and batteries sit on every feeder and further strengthens the viability of the system, rather than being a threat.
The writer is an assistant professor of practice at IBA, member of the Thar Coal Energy Board, and CEO of NCGCL
Published in Dawn, The Business and Finance Weekly, November 24th, 2025
