Rethinking Power Planning
in Pakistan
A simple look at how much electricity Pakistan will need, when people will need it, and what it may cost. This dashboard independently checks the power plan approved by NEPRA on 11 September 2026.
Pakistan's main power problem is not simply 'we need more power plants'
The harder problem is WHEN electricity is needed and WHO is still buying it from the grid. Rooftop solar cuts daytime grid sales. Batteries can move some of that power into the night. Air-conditioning pushes the biggest demand into a few evening and night hours. But old power-plant bills still have to be paid, even if the grid sells fewer units.
People may use more electricity overall but buy less of it from the grid. If the grid has large fixed bills, fewer grid sales mean each remaining unit must carry more of those bills. Higher prices then push even more people toward solar, batteries and other options. Pakistan can therefore end up with too many contracted power plants and high electricity prices at the same time.
Five Things the Power Plan Must Get Right
A good power plan must deal with all five together · Table M35A
Simple Comparison — FY2034-35
This analysis compared with the approved ISP · M26A
First estimate how much electricity people need. Then ask where it will come from.
Do not start by guessing grid sales. First estimate total electricity use. Then subtract power supplied by rooftop solar and batteries. What is left is what people need from the grid.
2. Electricity bought from the grid = total need minus electricity supplied by solar and batteries
3. The main power system must supply grid purchases plus the electricity lost while moving power through the network
All electricity people and businesses need, main case
+53.5% over ten years
People keep using more electricity overall
Falls from FY25-26 to FY35-36, then rises slightly to 50.4% by FY40-41
Approved power plan vs. this analysis — FY2034-35
2035 Comparison Table
Annual electricity use and highest power need in the official plan and this analysis · Appendix B
Assumed Annual Growth
Simple growth assumptions for cooling, industry and other uses · Section 16
Four Big Changes (Section 11)
- Grid sales are no longer the same as total electricity use. People can use the same amount of electricity but buy fewer units from the grid because solar, batteries and private generation supply part of it.
- The highest hourly demand and total yearly use no longer move together. A small change in the assumed daily pattern can change how many power plants seem necessary, even when yearly electricity use stays the same.
- Cooling grows as more homes and businesses buy cooling equipment. It depends on how many air-conditioners and fans exist, how hot it is, and when people use them.
- Solar helps least when the system often needs power most. Grid use can be lowest in sunny daytime hours and highest late at night, after solar stops producing.
Two official cases use different annual electricity totals and different assumptions about how evenly power is used. So their 6.899 GW peak difference is not a clean comparison. If annual electricity use is held at the same 180.605 TWh, changing only the assumed usage pattern moves the peak from 35.547 GW to 28.635 GW. That is a 6.912 GW difference created by the assumption, not by people using more electricity.
Pakistan's hardest power problem is increasingly the hot evening and night
Pakistan may have enough electricity over the whole year, but still struggle to have enough power during a few hot evening and night hours.
FY2024-25; observed at midnight in June 2025
Actual average
about 58% of the highest actual demand
FY2024-25 · about 65% used from June to September
Two Ways of Estimating Cooling Appliances
Household-based estimate vs. whole-system estimate · Appendix L.1
When incomes rise, more people buy air-conditioners and refrigerators. That can make electricity demand rise quickly on hot evenings. The cheapest response is not always a new power plant. Pakistan should count how many cooling appliances are being added, estimate when they are used, and make efficient ACs and fans the standard. That can reduce bills and avoid building expensive capacity that is needed for only a few hours.
Solar cuts daytime grid sales. Batteries can make that solar more useful at night.
Imported solar panels do not all reduce grid demand in the same way. Some are off-grid, some are connected to the grid, and the electricity they make can be used immediately, stored in batteries or exported. The model follows each step.
FY2025-26, total rated panel capacity
after removing off-grid systems and K-Electric area
FY2025-26 annual estimate
FY2025-26 · may peak near 17 GWp of annual imports by FY2029-30
From All Solar Panels to Solar Supplying the Grid Area
FY2024-25 compared with FY2025-26 · Appendix J.3
The 7.319 GW of registered net-metering is only a minimum known amount for registered systems. It does not include every solar system, because many systems do not export power, are not registered, or are fully off-grid. The official 8.12 GW projection for 2035 is already close to today's registered number, so it likely misses a large amount of solar that is already physically installed.
People buy batteries when the savings become worth the price
Battery prices are expected to fall as the technology gets cheaper around the world and Pakistan's market matures. Other costs, such as inverters and installation, are also assumed to fall over time.
starting estimate outside K-Electric
before old batteries are replaced
FY25-26 to FY40-41, simple savings estimate
grid-level equivalent
Checking Battery Imports Against the Starting Estimate
Reported battery imports compared with the model's starting estimate · Appendix K.1
If a battery is being saved for backup, it cannot also be fully used to lower the grid peak at that same time. The same battery power must not be counted twice. Batteries used by homes and businesses may naturally reduce grid demand. Batteries promised to the grid through a coordinated program must also prove they can respond when called. The model should count each effect only once. Pakistan needs simple rules that let many small batteries work together for the grid, so battery growth is planned for rather than ignored.
Start with total electricity need, subtract solar and batteries, and the remainder is grid demand
Once solar and batteries are included, the amount people need from the grid can look very different from a simple forecast of past electricity sales.
Large fixed power bills push grid prices up → high prices make solar and batteries more attractive → people buy fewer grid units → the same fixed bills are divided over fewer units → grid prices rise again. Pakistan can therefore end up with too many contracted power plants and high electricity prices at the same time.
All 9 Demand + Solar/Battery Cases
Grid purchases (TWh) and highest grid power need (GW) in FY2030-31, FY2035-36 and FY2040-41 · Appendix N.2
FY2034-35 Comparison Using the Same Year
Section 25, M26A
Having many power plants on paper does not mean enough power is available at 1am
The main challenge is having enough power for a few busy evening and night hours — not producing enough electricity over the whole year.
FY2024-25
vs. 36.7 GW fleet proxy
if batteries are ignored
Falls to 1.2 GW with battery credit
A battery's size is not enough — how long it lasts matters
A battery sold as a four-hour battery can provide full power for only 3.24 hours after normal limits and energy losses are included. If part of the battery is kept for backup, the usable time becomes even shorter. A battery emptied from 6pm to 10pm may have nothing left for the 11pm–midnight peak. Planning should therefore test batteries that last 2, 4, 6 and 8 hours, including difficult days and different locations on the grid.
Typical-Day Battery Results — Full Detail
Available battery power, usable energy, actual use and peak reduction · Appendix M.2/M.3
How New Power Projects Are Classified
How the official decision labels each project · Section 29
A plan is not truly low-cost unless it shows how many grid units will pay the fixed bills
Existing power plants have fixed payments. Those payments must be divided across the electricity units people still buy from the grid.
PKR/kWh, FY35-36, main exchange-rate case (22.03 if PKR falls 7.5% a year)
PKR/kWh, FY35-36, main exchange-rate case (27.82 if PKR falls 7.5% a year)
US cents/kWh, FY35-36, main exchange-rate case
PKR billion, FY35-36, main exchange-rate case
Fixed Power-Plant Cost Calculator
Choose what happens to the rupee/projects and how much electricity the grid sells. The calculator shows how much fixed power-plant cost falls on each unit through FY2044-45.
Fixed Power-Plant Cost per Unit, PKR/kWh
FY2034-35: PKR 15.33/kWhKey Years
Total Electricity Cost Calculator
This adds an illustrative PKR 14/kWh for fuel, transmission, distribution and other costs. Now lower grid sales. The same fixed power-plant bills are spread across fewer units, so the price per unit keeps rising even though the total fixed bill has not changed.
Illustrative Total Electricity Price, PKR/kWh
FY2034-35: PKR 29.33/kWhTotal Price — Key Years
If the fixed bill in FY2034-35 stays the same, the example total price is PKR 23.20/kWh; at 120 TWh it is PKR 29.33/kWh; at 80 TWh it is PKR 37.00/kWh. The total fixed bill does not change; only the number of grid units sharing it changes. Fewer grid sales therefore mean a higher fixed cost per unit, which can push even more customers toward solar and batteries.
What Inflation, Interest Rates and the Rupee Could Do to FY2034-35 Costs
Extra PKR billion and PKR/kWh if grid sales are 120 TWh · Table C05
Some Debt Payments Fall, but Other Costs Remain
Main cost groups, PKR billion, FY25 vs. FY35 · Section 31.3
Choose projects by what they add to the total system bill
An expensive project can still make sense if it reliably replaces even more expensive electricity. The right question is: what does it do to the total bill?
Official Cost Assumptions
Section 30, Table 30
What the Published Project Costs Do — and Do Not — Tell Us
A project's published construction cost is not the same as the annual amount consumers must pay · Section 30.1, Table 30.1
Could Existing Southern Coal Plants Be Used More?
Port Qasim, China Power Hub and Lucky together provide about 3,107 MW of dependable capacity. They generated 4.402 TWh in FY25 while carrying about PKR 302.619 billion of annual fixed capacity cost. Section 40.4.
Fixed Cost per Unit Falls if the Same Plants Generate More
PKR 22.24 / generated kWhUnder the official planning method, if a strategic project costs more than the cheapest system option, that extra cost is meant to be paid by the sponsoring government. The USD 3.0015 billion extra cost above the least-cost option should therefore not automatically be added to electricity consumers' power-plant bills — the cost still exists, but the government rather than electricity consumers would pay that strategic-policy premium.
Buy the power service the country actually needs, while keeping fixed bills and consumer prices under control
Practical Next Steps
Recommendations combine the official decision with findings from this independent analysis · Table M35
The Five Things That Must Be Fixed Together
Table M35A
All the main numbers in one place
Every chart on this page comes from the tables and sources listed here.
Main Numbers
Main demand + main solar/battery case · numbers are model estimates unless marked as actual or official
¹ Registered net-metering reported nationally in March 2026. ² Official ISP total for NGC + K-Electric; this is an official administrative number and is not the same as the broader estimate of solar physically installed. ³ The approved ISP ends in 2035. Demand, batteries and main-system supply exclude K-Electric unless stated otherwise. Fixed cost per unit is calculated by dividing the relevant PKR cost by electricity bought from the grid. Battery peak reductions are model estimates and should later be checked with a full hour-by-hour reliability study.
Sources Used
Main sources used in this analysis · Appendix A
Pakistan's main challenge is not simply the number of megawatts. It is when power is needed, how many units are still bought from the grid, and how much fixed cost must be paid.
Cooling is pushing demand higher at night. Solar is cutting daytime grid sales. Old fixed power-plant bills remain. Batteries are growing quickly outside the central plan. And new generation is useful only if the transmission network can deliver it. Pakistan therefore needs hour-by-hour planning that connects all five pieces, instead of planning mainly around the total size of power plants.