Solar Generation Estimator

The annual average is the number that misleads you

A 10 kW system in Lahore generates about 14,000 units a year. Divide by twelve and you get roughly 1,170 units a month, which is the figure most quotes will show you.
Your system will produce that in exactly no month of the year.
In May it produces around 1,670 units. In December it produces around 700. The average sits between two numbers your system never actually delivers, and the gap between them is not small – December is 42% of May.
This matters for one practical reason. If someone has sold you a system on the promise of a fixed monthly saving, that promise is built on the average. Your December bill will not honour it, and by then the installer has been paid.

What each city actually gets

These figures come from NASA POWER climatology – long-term measured irradiance for each city’s coordinates, not an industry rule of thumb. “Units per kW per year” is the number to compare: multiply it by your system size to get your annual generation.

The rates, as notified for CY 2026

These are the GoP Applicable Uniform rates for A-1 Residential, taken from Annex-C of NEPRA’s Decision of the Authority dated 12 January 2026. NEPRA now determines consumer-end tariffs on a calendar-year cycle, so this schedule runs through 2026.

How to read that. Take Quetta. It averages 6.12 peak sun hours a day across the year, which over 365 days at 78% system efficiency gives 1,744 units for every kilowatt installed. A 10 kW system there produces about 17,440 units a year.
Split across the calendar, that same kilowatt produces 198 units in June and 90 in December – so the 10 kW system makes roughly 1,980 units in its best month and 900 in its worst. The final column is simply the ratio between those two.
All figures include 78% system efficiency, so these are units arriving at your meter, not raw irradiance on the panel.

Two things stand out.
Quetta is the best solar location in the country by a clear margin – 1,744 units per kW against Faisalabad’s 1,370. The same system, the same panels, 27% more energy.
Karachi has the flattest year. Its 1.7× swing is the mildest in the country, while Peshawar swings 2.6×. Coastal cities lose less in winter, which makes their generation far easier to plan around even though their annual total is not the highest.

Why winter matters more than the numbers suggest

A system sized to cover your annual consumption will overproduce in summer and underproduce in winter. That used to balance out. Under the old net metering rules, surplus exported in June offset units drawn in December at the same rate, and settlement was annual. Net billing ended that for anyone applying after 9 February 2026. Your June surplus is now bought at the low export rate, while your December units are bought back at full retail.
But two things break that comfort.
Your consumption is not flat either, and in Pakistan it moves the wrong way. Air conditioning peaks in exactly the months solar peaks, which is helpful. But fans, lights and heating in winter arrive precisely when generation is at its floor. Summer surplus is real, but so is the winter shortfall, and they are not the same size.
Export credit is not cash. Exported units are credited against units you draw, at a rate well below what you pay to buy them. Banking a large summer surplus to spend in winter is not the same as banking rupees, and the arithmetic is on the savings calculator.
The practical conclusion is not “buy a bigger system”. It is to know your December figure before you sign, so that the month your generation halves is a month you expected rather than a month you complain about.

Horizontal, tilted, and what your roof does to this

The irradiance behind these numbers is measured on a horizontal plane. Your panels will not be horizontal.
A fixed array tilted toward the south catches more winter sun and slightly less summer sun than a flat surface does. In practice that lifts the annual total a little and flattens the seasonal curve – the December figure improves more than the June figure drops. So treat the monthly numbers here as the conservative case.
We use horizontal deliberately. It is the figure the annual constants across every calculator on this site are built on, so all of them stay consistent with each other, and it errs toward under-promising rather than over.
What your roof takes back off it is a separate matter. Shading from a water tank or a parapet for two hours a day is a far larger effect than the tilt, and it is site-specific in a way no calculator can model. So is dust – an uncleaned array through a dry month can lose more than the entire winter tilt advantage.

What this does not include

Degradation. Panels lose a fraction of a percent of output per year. Over 25 years it is material; in year one it is not, and these figures are for a new system.
Shading and soiling. Both are site-specific. The 78% system efficiency applied here covers typical inverter, heat, wiring and general soiling losses on a maintained array – not a water tank shadow or a season without cleaning.
Real weather. This is long-term climatology, which is the right basis for planning. An unusually cloudy month will come in below it and an unusually clear one above.

Next step

Generation is only half the answer. What it is worth depends on how much of it you use yourself and how much goes to the grid – and those are credited at very different rates.