Pakistan still runs two electricity systems that touch at a handful of points. K-Electric generates, buys and distributes power for Karachi, while the National Grid Company and its DISCOs serve almost everyone else. IGCEP 2025-35, the system operator's ten-year generation plan now before NEPRA, is one of the few documents that puts a price on keeping them apart.
The plan assumes a fixed export of 2,050 MW from the national grid to K-Electric once K-Electric's 500 kV KANUPP Karachi Interconnection grid station is commissioned, as set out in a proposed tripartite agreement between K-Electric, NGC and CPPA-G. On top of that, the model is allowed to build a candidate transmission line from NGC to K-Electric, with its construction cost included, from July 2028. ISMO then ran sensitivities that remove the line, delay it to July 2031, or treat 620 MW of K-Electric's solar and hybrid projects as committed.
What the sensitivities show
The results are in the plan's executive summary. In the reference rationalised case, the present value of generation investment and operation to 2035 is USD 47.13 billion. Removing the candidate line raises it to USD 47.52 billion. Delaying the line to 2031 gives USD 47.32 billion. By our subtraction, building the link on time saves about USD 390 million in present value terms against not building it, and each year of delay from 2028 to 2031 costs roughly USD 63 million (USD 190 million over three years).
The capacity tables explain where the saving comes from. Without the line, total installed capacity in 2034-35 falls from 62,657 MW to 61,844 MW, and the entire 813 MW difference is wind: the optimised wind fleet drops from 5,133 MW to 4,320 MW. Southern wind is cheapest when it can reach demand in Karachi as well as the north, and the line is what lets the model build it.
Our view
The interconnection debate in Pakistan is usually argued as a question of who controls Karachi's supply. ISMO's own numbers suggest it is mainly a question about wind in Sindh. A USD 390 million saving is modest against a USD 47 billion programme, and the figure depends on the line's cost assumptions, which the summary does not show. It is still the cheapest of the K-Electric options the plan tested, and nobody has published a study showing the opposite.
The 620 MW K-Electric renewables case costs USD 47.32 billion, also above the reference. Forcing K-Electric's projects in as committed is more expensive than letting the model choose, which is an awkward result for anyone arguing that Karachi should plan alone. NEPRA has nonetheless asked, among the issues framed for its hearing, why committed projects of K-Electric and the provinces were not duly considered, so this question is not settled.
NEPRA should require the next plan to publish the line's assumed cost, capacity and route, and should ask NGC and K-Electric for a joint interface procedure before the tripartite agreement is signed.