Why do we pay for power we cannot use?

It is not an RLNG problem. It is a transmission problem, and only full disclosure can tell us how to fix it.

Picture two power plants on the same afternoon. One sits in Thar, burning local coal at about Rs 5.99 per unit. It is told to slow down. Eight hundred kilometres north, the system reaches instead for imported fuel: RLNG at Rs 39.74 per unit, furnace oil at Rs 44.86, diesel at Rs 62.14, on CPPA-G’s (Central Power Purchasing Agency-Guaranteed) September 2026 figures. Between them runs a grid that cannot carry what the first could give. The consumer pays for both.

It is tempting to blame the gas, but RLNG is largely the symptom. The bottleneck is the wire. ISMO’s (Independent System and Market Operator’s) congestion note is plain: at the south-to-north interface, coal, and wind are curtailed in the south while RLNG and RFO run in the north.

In CPPA-G’s 2027 forecast, when demand is assumed lower, Thar coal gives up 733 million units and RLNG just 56 out of 18,855. Not all of that is due to the transmission wires; demand, contracts, plant availability and system security also decide who runs.

ISMO finds southern wind and solar are curtailed mainly by low demand, not the wires. For coal, it is different: the transfer limit alone cuts imported coal dispatch in the south by 35 to 45 percent. That part, the grid decides for us.

Power regulator NEPRA’s performance evaluation for July 2024 to June 2025 recorded that the 660 kV HVDC line from Matiari to Lahore, rated at 4,000 MW, carried on average 1,397 MW, a utilisation of about 35 percent. It tied that underuse to NGC’s (National Grid Company’.s) delayed projects.

Average flow is not spare capacity: what the line can carry at any hour, its available transfer capability, depends on demand, dispatch, and the AC grid at both ends, which is why CPPA-G assumes the corridor can move only 3,200 to 4,800 MW in 2027. The gap is not all power waiting to be sent, but nor is it all low demand.

In July 2026, NEPRA put the cost of the north-south constraint at roughly Rs 400 billion a year, the extra cost of running costlier northern plants in place of cheaper southern ones; its method has not been published.

Over CPPA-G’s projected 130,081 GWh, that is up to about Rs 3.07 a unit, and the forecast does not say how much is already in the 2027 price. Since actual costs reach consumers through monthly and quarterly adjustments, a constraint left out of a forecast is simply billed later.

Relief keeps its distance. Islamabad West is due by December 2027, the STATCOMs by 2030, and the NGC-KE tie-line by 2031. However, the one line that would most directly open the corridor—from Matiari to Moro and Rahim Yar Khan—has no fixed date.

Consumers, meanwhile, have found their own way out, on their rooftops. Solar is often cast as the grid’s problem; it is really the consumer’s answer to expensive electricity. ISMO assumes another 6.5 GW by 2028-29. Every unit a rooftop supplies is a unit not bought from the grid, while capacity obligations largely remain, so fewer units carry the same fixed cost.

CPPA-G’s demand-sensitivity case, for slower demand generally rather than solar alone, shows the arithmetic: at 0.7 percent growth instead of 2.6 percent, capacity and related fixed charges of about Rs 2,162 billion are spread over 127,937 GWh instead of 130,081, and rise from Rs 16.626 to Rs 16.901 per unit. It is a spiral, and the constraint feeds it from both ends.

So the corridor is not one issue among many; it may be the one on which the others turn. Fix the wire and the tariff falls, the rush to rooftops eases, and fixed costs are shared by more units. For a Rs 400 billion problem, it is discussed strikingly rarely, and has never been fully explained.

Instead, money is moving elsewhere. A 500 to 600 MW battery pilot is planned in the NGC system by 2029, and a smart-meter programme is being prepared as a public-private partnership. Both have a place: batteries can shift midday power into the evening and, well-placed, ease congestion; smart meters enable time-of-use pricing, and ISMO lists both among its remedies.

But both will be recovered from consumers, neither has been weighed publicly against moving power that already exists, storing power costs more than using it, and meters add no megawatt to the corridor. Consumers deserve to know which rupee saves more before they pay for either.

Selling is the simplest answer of all. Instead of curtailing southern plants, offer their surplus to southern industry at a lower price, above all in daytime, when curtailment peaks.

Every extra unit sold spreads fixed capacity payments thinner nationwide, and every unit that keeps a Karachi or Hyderabad factory running need not come from imported RLNG in the large RLNG plants of the north, saving dollars as well as rupees. The conditions: no other consumer pays more, and nothing returns later as a prior-year adjustment.

Yet such ideas rarely get past the first meeting. Too often every solution is treated as a new problem, because there is no decision-grade data on the table.

That is the heart of it. To my knowledge, no cost-benefit analysis of the corridor, or of the lines meant to relieve it, has been placed on public record.

Without one, no one can say whether the next rupee belongs in run-of-river hydropower near northern demand, in batteries, in smart meters, or in evacuating every last unit from the south through the pending lines and grid stations. Until then, every choice is a guess, paid for by the consumer.

Disclosure need not be complicated. The HVDC line is owned by PMLTC, a State Grid of China company, under a 25-year build-own-operate-transfer licence; but NGC’s AC network at both ends decides how much of it can be used, and ISMO dispatches the system.

Between them, they could put five things on an open, live dashboard: the South’s generation capacity, fuel cost and demand by season and hour; the transfer actually achieved against the 3,200 to 4,800 MW assumed; every line and grid station under way, with megawatts, cost, and date; the cost of the constraint in rupees, per year and per unit, with the method behind it; and whether a south-only daytime package could use stranded capacity at no cost to anyone else.

None of this is about blame. It is about the right of those who pay the bill to see the full picture. NEPRA can ask for it, and record its own findings, before it sets the 2027 price or approves new investment. The test should be simple: does the answer bring the consumer-end tariff down?

Until the wire is fixed, or at least fully disclosed, we will keep paying twice: once for the power we use, and once for the power we cannot.

Copyright Business Recorder, 2026