By Dave Miller · off-grid sparky · 22 years wiring solar, batteries and sheds in central QLD
“After losing two batteries to heat in a flat-mounted array, I learned that a 20-degree tilt is the only thing keeping my system alive in central Queensland summers.”
Off-grid system sizing lives or dies on kWh you actually harvest, not kWh the panels are rated for. Panel angle is one of the few variables you control at install time and never change again, so getting it right has outsized payback compared to chasing the last 2% in inverter efficiency.
The core principle: panels generate maximum power when the sun's rays hit them at 90°, i.e. perpendicular to the panel face. In the southern hemisphere that means north-facing modules angled to match how high the sun climbs through the year. Two things drive that angle: your latitude (how far you are from the equator) and the time of year you care most about generating.
Latitude sets the base. The commonly cited ideal angle for fixed year-round arrays in Australia is equal to the latitude angle of the property's location. Sydney sits at about 33.86° south, so a roof pitch of 34° is the textbook "latitude equals tilt" target. The same rule places Melbourne at 38°, Brisbane at 27°, Adelaide at 35°, Hobart at 43°, Perth at 32° and Darwin at 12°.
If you only ever hear one rule, it is the latitude rule. It is a sound first-pass default for fixed-tilt year-round-optimised arrays at mid-latitudes. But it is not the full picture.
PVWatts modelling (cited by SolarQuotes) shows that the optimal annual tilt for north-facing panels is actually a few degrees lower than latitude in seven of Australia's eight capital cities:
| City | Latitude | PVWatts optimal tilt | Difference |
|---|---|---|---|
| Adelaide | 35° | 29° | -6° |
| Brisbane | 27° | 24° | -3° |
| Canberra | ~35° | 30° | -5° |
| Darwin | 12° | 18° | +6° |
| Hobart | 43° | 37° | -6° |
| Melbourne | 38° | 32° | -6° |
| Perth | 32° | 28° | -4° |
| Sydney | 34° | 31° | -3° |
Why? Days are longer in summer in Australia, which means tilting panels a few degrees lower than latitude lifts annual yield by capturing more of the high-summer output. Seasonal differences in average cloud cover also shift the optimum. Darwin is the exception: optimal tilt (18°) exceeds latitude (12°) because day length barely varies between seasons and the summer wet season drags down summer output, so the model rewards a steeper face.
The upside from chasing this refined tilt is genuinely small. PVWatts gains from using the optimum instead of latitude: Adelaide 0.37%, Brisbane 0.07%, Canberra 0.29%, Darwin 0.45%, Hobart 0.50%, Melbourne 0.44%, Perth 0.16%, Sydney 0.11%. For grid-tied work that's rounding error. For off-grid work where every watt-hour matters at the margin of your battery autonomy, it can still be worth knowing.
Off-grid arrays are rarely installed on a perfectly pitched roof. The good news is that the penalty for being a bit off is shallow.
If roof pitch is within 10° to 15° of the latitude angle, solar output loss is only about 1 to 1.5% of the maximum possible. In Melbourne, with a northern orientation, a roof pitch of 23° delivers about 98.5% of maximum solar generation, while 28° delivers about 99%. You are paying for nothing if you start building tilt frames on a roof that is already in the band.
Where it gets interesting is at the extremes. A perfectly flat (horizontal) roof in Melbourne would only generate about 86% of the maximum possible solar output. In Brisbane, by contrast, a flat roof would operate at about 90% of optimum capacity. The further north you go in Australia, the less a flat roof hurts because the sun sits higher in the sky year-round. Across major capitals the flat-roof percentage of optimum looks like this: Sydney 87%, Melbourne 86%, Perth 88%, Brisbane 90%, Hobart 82%, Adelaide 87%, Darwin 96%.
For ground-mounts and flat-roof commercial work the question is whether the lost yield is worth the structural cost of an angled frame. Tilt frames on a sloped roof are normally not worth the extra expense because they only slightly improve output. The calculation changes on a flat roof where you have nothing to lose by adding a frame.
The latitude rule assumes you want steady year-round output. Off-grid systems often don't. If your winter loads dominate (think: a cool-climate cabin running a fridge, lights and a water pump when solar day-length is at its shortest), shifting the array steeper buys you winter sun at the cost of summer yield.
The standard rule for maximum winter sun exposure is to angle panels at the latitude angle plus 15°. So Melbourne moves from 38° to 53°, Hobart from 43° to 58°, Sydney from 34° to 49°. That is a meaningful step.
Seasonal adjustment of latitude ±10 to 15° is the standard off-grid practice in southern Australia. Some operators take it further with a two-position seasonal tilt: steep in winter, shallow in summer. Manually re-angling a ground-mount twice a year is cheap labour if the array is accessible, and the yield uplift over a fixed latitude mount can be substantial in Hobart or Melbourne. It is rarely worth it in Darwin or Cairns where the seasonal swing is small and a latitude-based fixed mount is already close to optimum.
Tropical Australian sites (north of roughly 23°S) generally want a tilt below latitude, not latitude itself, for fixed year-round mounts. The PVWatts data backs this: Darwin wants 18° against a 12° latitude, Brisbane 24° against 27°, Perth 28° against 32°. The shorter summer days up north and the wet-season cloud shift the optimum flatter.
A tilt-angle calculator does first-order solar geometry. It uses your latitude, the solar declination through the year, and (optionally) a seasonal bias to spit out an angle. The better ones offer modes for year-round, summer-biased, winter-biased, specific-month, two-position seasonal, and verify-existing comparison.
What they do not do: they are a screening tool using solar geometry and an empirical latitude correlation, not a full simulation. They don't run hourly weather simulations, they don't model shading from trees or ridges, they don't account for soiling losses, bifacial rear-side gains, or temperature derating on hot Queensland afternoons.
For a verified annual kWh figure you want a site-resolved hourly tool like NREL PVWatts, NREL SAM, or PVsyst with TMY weather data for your nearest station. Use the tilt calculator to set the angle, then use PVWatts to confirm the kWh.
Tilt gets all the attention; orientation quietly eats more yield than tilt on most residential sites. The optimal combined orientation (degrees clockwise from true north) and tilt for each capital: Adelaide 6°/30°, Brisbane 357°/24°, Canberra 356°/30°, Darwin 30°/20°, Hobart 20°/38°, Melbourne 8°/32°, Perth 20°/28°, Sydney 12°/31°. Most of those orientations sit within a few degrees of true north. Anything within ±15° of true north is fine; beyond that you start paying real yield.
If you are stuck with a flat roof, the mounting frame is doing more than holding panels up. Flat roof mounting devices are typically installed to deflect rainwater and help clean solar panels of dust and grime, which would otherwise reduce generation capacity. A small tilt of 5 to 10° is the usual compromise: enough to shed water and self-clean, not so steep that wind loading becomes a structural problem.
Tracking sounds attractive in theory. In practice, single-axis trackers add mechanical and maintenance failure modes that are a legitimate concern for unattended off-grid sites. A tracker that fails open at 3 a.m. on a cloudy week in winter can drain a battery bank that a fixed array would have kept topped up. For remote installations where you can't be on-site weekly, a fixed mount at the right tilt is almost always the right call.
Is the latitude rule good enough for a fixed off-grid array?
Yes, for most of southern Australia at mid-latitudes. Within 10 to 15° of latitude you lose only about 1 to 1.5% of maximum output. The refined PVWatts tilts gain another 0.07% to 0.50%, which is rarely worth the engineering effort on a small fixed array.
Should I angle panels steeper than latitude for winter?
If winter is your critical season and your battery autonomy is tight, yes. Latitude plus 15° is the standard winter-biased setting. For double the labour you can manually re-angle a ground-mount twice a year between a summer-flat and winter-steep setting.
What about flat roofs?
A flat roof in Melbourne runs at about 86% of optimum; Brisbane 90%; Darwin 96%. The further north the site, the smaller the penalty. Adding a modest 5 to 10° frame is usually worthwhile on a flat commercial roof because it also helps with self-cleaning.
Do I need a tracker for better output?
For unattended off-grid sites, the maintenance and failure-mode risk of a single-axis tracker usually outweighs the yield gain. A correctly tilted fixed array is more reliable and cheaper over a 20-year horizon.
Which calculator should I actually use?
Use a tilt calculator (CalcEngineer, SolarCalculator.com.au) for the geometry and seasonal modes, then run NREL PVWatts, NREL SAM, or PVsyst with local TMY data for verified annual kWh before you commit to a system size.
Does orientation matter more than tilt?
On a typical house roof, yes. Losing 20° off true north costs more than being 10° off the ideal tilt. Get the orientation right first, optimise tilt second.
Worth a watch: Buying Solar In Australia In 2026: Everything You Need To Know · SolarQuotes