Energy

Which way and how steep: orienting solar panels in Pakistan

Why the best tilt is a little less than the latitude, why direction matters more than tilt on a real roof, what an east or west roof costs, when adjusting twice a year is worth it, and how dust and the tariff change the answer.

Author
Ahmedonics Engineering
Published
Last updated
Illustration of a tilted rooftop solar array with the sun's low winter and high summer paths drawn across the sky and a compass rose on the roof

Every solar quotation in Pakistan carries the same two questions from the customer: which way should the panels face, and how steep? The honest answer is that on most roofs it matters less than people fear, and where it does matter it is the direction, not the angle. This guide gives the physics behind the tilt and orientation calculator and the judgement calls the calculator cannot make for you.

What a panel collects: the incidence angle

A panel produces in proportion to the sunlight that lands on it, and beam sunlight lands in proportion to the cosine of the angle between the sun's rays and the panel's normal. Face the sun squarely and you get all of it; tilt 25° away and you still get 90 %; 45° away, 71 %; 60° away, half. The cosine is forgiving near zero and punishing far from it, which is the single most useful fact in this subject: small errors cost almost nothing, large ones cost a lot.

The sun's position at noon changes through the year by 47° (twice the 23.45° tilt of the earth's axis). In Lahore at 31.55° N it stands at 35° above the horizon on the winter solstice and 82° on the summer solstice. No fixed panel can face both; the tilt you choose decides which season you favour.

← south north → flat roof, Lahore 31.55° N panel β = 20° panel normal (70° up) winter noon, 21 Dec 35° above the horizon summer noon, 21 Jun, 82° up θ = 35°, cos θ = 0.82 θ = 12°, cos θ = 0.98 A flat panel would collect cos 55° = 0.57 in winter, cos 8° = 0.99 in summer. Tilt trades summer for winter; the year's optimum is 26° here.
Noon sun in Lahore at the two solstices, and a panel tilted at β from the horizontal facing south. The incidence angle θ is what the panel gives up: cos θ of the beam is collected. A tilt near the latitude splits the difference; a flatter tilt favours summer.

Why the best fixed tilt is below the latitude

The rule of thumb "tilt equals latitude" faces the panel at the sun's average noon height, which sounds right and is slightly wrong. Summer contributes more energy than winter for three reasons: the days are longer, the sun is higher so its light passes through less air, and there is less cloud and haze in most of Pakistan in May and June than in January. Weighted by energy, the year's sunlight comes from a little higher in the sky than the geometric average, so the best fixed tilt is a little flatter than the latitude. Diffuse light from the sky, which a flat panel sees more of, pushes the same way.

The calculator's clear-sky model gives 26° for Lahore, 20° for Karachi, 27° for Islamabad, against latitudes of 31.5°, 24.9° and 33.7°. Measured-data tools such as PVGIS land within a few degrees of the same answers. And the curve is flat around the optimum: anywhere from 15° to 35° in Lahore is within 2 % of the best. That is why a pitched roof at 20° needs no frame, and why arguing over 25° against 30° is a waste of a meeting.

Direction: the loss that is actually worth worrying about

In the northern hemisphere the sun spends the day in the southern half of the sky, so a south-facing panel sees it most squarely for most of the day. Turn the panel away from south and the morning or the afternoon is lost progressively. At the 20° tilt of a typical roof in Lahore, the calculator gives roughly:

  • South (180°): 99.7 % of the optimum.
  • South-east or south-west (135° or 225°): about 97 %.
  • East or west (90° or 270°): about 88–90 %.
  • North (0°): about 78 %, and much worse at steeper tilts.

Two lessons. First, the penalty for facing off-south grows with tilt: a flat panel has no direction, a vertical one is all direction. If your only roof faces east, mount the panels flatter than you otherwise would; the calculator's "best tilt for your azimuth" figure tells you how flat. Second, an east–west split is not a defeat. It produces a flatter, longer daily curve that matches a household load better than a midday spike, and under a net-metering tariff that credits exports at a lower rate than it charges imports, self-consumed afternoon power is worth more than exported noon power.

Seasonal adjustment: 4 % for a lot of climbing

A panel steepened to about 46° for October to March and flattened to about 9° for April to September collects around 4 % more over a year in Lahore than one fixed at 26°. On a ground-mounted array with a hinged frame that was designed for it, two adjustments a year are cheap. On a rooftop system, it means someone on a roof with a spanner twice a year, loosening and re-torquing the clamps that hold panels against the wind, and the 4 % is roughly what one summer of dust costs anyway. We do not recommend it for rooftops.

Where a seasonal bias is worth building in permanently is when the load is seasonal. A site that runs on batteries through winter load-shedding and has more sun than it can use in summer should lean toward the winter tilt, accepting a summer surplus it cannot store. The calculator's October–March share tells you how much of the year's yield your chosen tilt puts in the hard months.

Dust, rain and the case against flat

Punjab and Sindh are dusty, and dust is the largest single loss on an unwashed array: 10–20 % in a dry month is routine. Rain cleans a tilted panel; a flat one collects a paste at the lower frame edge that shades the bottom cells and, on a string, drags the whole panel down. Below about 10° tilt self-cleaning effectively stops. So even where the yield model would allow a near-flat mounting, installers hold to 10–15° minimum, and this is right. Steeper panels also run a few degrees cooler in the breeze, which is worth about 1–2 % of output in a Lahore June.

Shading and the roof you actually have

None of the above survives shade. A water tank, a parapet, a neighbour's upper floor or a single tree branch across one panel in the afternoon costs more than any orientation error. A shaded cell in a series string limits the current of the whole string, so 5 % of the array in shade can take 30 % of the output until the sun moves. Before choosing an orientation, walk the roof at 9 am, noon and 4 pm in December, when shadows are longest; where shade cannot be avoided, module-level optimisers or micro-inverters recover most of the loss. Then choose the orientation among the unshaded areas, and only then use the calculator to see what it costs.

The example in numbers

A Lahore house with a 20° roof facing south-west (225°), mixed sky:

  1. Optimum fixed tilt facing south: 26°. The 20° tilt on its own gives up 0.3 %.
  2. The south-west direction on its own costs 3.1 %; the best tilt for that face is about 20°, which is what the roof has.
  3. Overall the array collects 96.9 % of the optimum and 4.1 % more than a flat panel. Adjusting the tilt twice a year would add 4.1 %; 38 % of the year's yield arrives in October to March.

The right decision is to mount flush on the roof, spend the frame money on a hose connection for washing, and check the shade. The system sizing calculator then turns the load into an array size, and the sizing guide explains the losses that come next.

References

  • Duffie, J. A. and Beckman, W. A., Solar Engineering of Thermal Processes, 4th ed., Wiley, 2013 — solar geometry and radiation on tilted surfaces
  • European Commission Joint Research Centre, PVGIS — optimum tilt and azimuth from measured irradiation data for any site
  • IEC 61724-1:2021, Photovoltaic system performance — Part 1: Monitoring — plane-of-array irradiance and performance ratio definitions
  • Meinel, A. B. and Meinel, M. P., Applied Solar Energy: An Introduction, Addison-Wesley, 1976 — clear-sky irradiance model used by the calculator