If you’re thinking of putting solar on a 3-bed with an east/west roof, here’s what you can expect. You’ll produce about 19% less than a south-facing equivalent, per PVGIS. What the raw number misses is that an E/W roof often matches a working family’s demand curve better than south-facing ever could. And, as with every UK roof, the real ROI lever isn’t orientation. It is the battery.
I’ve got solar at home and I have been running panels on three orientations for three years: an SE and NW pitched roof, a nearly-flat gazebo, and a SW wall. The headline I keep coming back to is that on overcast days, which the UK has plenty of, the difference between two orientations collapses. A huge share of annual UK irradiance arrives as diffuse light, and diffuse light does not care which way your panels face. On a cloudy day, an east-facing string and a west-facing string will produce nearly identical output.
What makes E/W quietly good is shape, not size. Mornings and evenings are when the kettle, cooker, washing machine and evening TV stack up. East panels hit that 7-10am window. West panels carry the 4-7pm window. A battery bridges the middle. That three-piece setup matches a working family’s actual load in a way a south-facing midday spike never does.
What an E/W Roof Must Optimise For
- Self-consumption over export. Every kWh you use directly saves 27.69p. Statutory SEG export is 5p, Octopus Outgoing Fixed is 12p. Import displacement is 2-6x more valuable than export. Size to consume, not to export.
- Dual-MPPT inverter, non-negotiable. East and west strings peak at different times. A dual-MPPT hybrid runs each string at its own optimal voltage. A single-MPPT setup drags both to the weaker face and loses roughly 10-15% of potential output.
- Overpanel both faces to 130-150% of the AC rating. With E/W producing ~81% of south-facing yield per kWp, overpanelling is how you close the gap. In UK conditions on E/W you rarely clip, because the two faces never peak at the same time: east peaks around 10am while west is still warming up, and vice versa at 3-4pm. Extra DC earns shoulder-hour generation.
- Battery first, bigger array second. If cashflow forces a choice, take the battery. Without it, midday surplus dumps to the grid at 5-12p and you buy it back at 28p in the evening. That gap is where ROI comes from. On a really tight budget, a battery fitted alongside normal grid import is a genuine hedge against price hikes. Panels can come later.
Why the Generation Curve Beats South-Facing for Many Families
E/W panels produce around 81% of a south-facing roof per kWp (PVGIS, 28° tilt, central England). In overcast conditions that gap narrows further because diffuse light is roughly direction-agnostic.
The shape matters more than the total. South-facing panels build a sharp midday spike that, without a battery, mostly becomes 5p export. East/west panels spread the same energy over a wider window. East catches the 7-10am kettle-toast-washing-machine slot. West handles 4-7pm cooking, laundry-finish, and the evening TV. A mid-sized battery fills the midday trough. For a typical 3-bed where the house is busy morning and evening, this is a better use of the generation than south-facing.
For a deeper look at how tilt and orientation actually play out in real UK conditions, see my three-year metered comparison in flat vs pitched vs vertical.
What a 3-Bed Actually Uses
Low to Mid: ~2,900 kWh/year
Gas central heating, standard appliances, no EV. About £803/year on electricity at the current cap. Demand concentrates 7-9am and 5-9pm. The E/W curve overlaps both peaks directly, which is why self-consumption on this roof shape is easy to push above 60% with even a modest battery.
High: 6,500+ kWh/year
Heat pump, EV, or both. A heat pump adds 3,000-5,000 kWh/year depending on insulation. An EV adds 2,000-3,500 kWh. You’re looking at £1,800+ on electricity alone. At this consumption level, your daytime baseload is high, the heat pump cycles through the day, and a scheduled EV charge slides into the solar window. Self-consumption runs 80-97% with a battery. This is where E/W quietly shines: high consumption, broad generation curve, and storage stack up to very little export and very high grid displacement.
Budget Option: 3.6kW Hybrid, 12 Panels, 5.12kWh Battery
Solar Energy Concepts System Rating
System Specification
Inverter: Solis 3.6kW 5G Hybrid EH1 (£575-653). 3.6kW AC stays under the G98 threshold (notification only). Dual MPPT (essential on this roof), 90-520V range, 15A per MPPT, accepts 5.7kW DC total. Hybrid so a battery adds without replacing the inverter.
Overpanel target: 148% of the 3.6kW AC rating = 5,340W DC. Overpanelling on E/W is particularly safe because the two faces never peak simultaneously.
A note on the DMEGC 450W: its Isc of 15.4A exceeds the Solis EH1 15A per-MPPT limit. Cheapest panel in the UK, but not a match for this inverter. Pair DMEGC with Fox ESS H1-3.7 (16A) or SolaX X1 G4 (16A) if you want to use them.
Panel Options
| Strategy | Panels | Total DC | Overpanel % | String Check |
|---|---|---|---|---|
| Best value | JA Solar 445W × 12 (6 East + 6 West) | 5,340W | 148% | Voc: 6 × 39.1V = 235V < 600V. Vmp: 6 × 32.65V = 196V in 90-520V. Isc 14.43A < 15A. OK. |
| Smallest footprint | Aiko Neostar 490W × 10 (5 East + 5 West) | 4,900W | 136% | Voc: 5 × 40.98V = 205V < 600V. Vmp: 5 × 34.4V = 172V in 90-520V. Isc 14.93A < 15A. OK. |
| Max overpanel | LONGi Hi-MO X6 455W × 12 (6 East + 6 West) | 5,460W | 152% | Voc: 6 × 39.15V = 235V < 600V. Vmp: 6 × 32.98V = 198V in 90-520V. Isc 14.79A < 15A. OK. |
The JA Solar 445W at £64 is the cheapest compatible panel per watt (14.4p/W). Aiko Neostar 490W has the highest efficiency (24.5%) and fits 4.9kW in 10 panels. LONGi Hi-MO X6 455W at £73 gives 120W more than JA Solar at the same panel count.
Battery
Fogstar 5.12kWh Server Rack (£730-900). Target is about two-thirds of daily average consumption. For a 3-bed at 7.9kWh/day average, 5.12kWh is the sweet spot. On an E/W roof the battery sits between morning and evening production peaks and covers the 11am-3pm demand that the panels can’t match directly. It is also the single biggest reason self-consumption moves from 50% to 60% on this roof. If the budget has to stretch, the Solis EH1 is hybrid, so you can fit the panels first and add the battery within a few weeks without swapping any other hardware. Treat that as a stopgap, not a plan.
Total System Cost
- DIY: £2,250-2,550
- Installed: £5,000-5,700
DIY Feasibility
A dual-MPPT system is only slightly more work than a single string. You run two DC cables from the roof (one per face) into the labelled MPPT1/MPPT2 inputs. The battery connects via its own DC port. The main practical job is routing cables from both sides of the ridge through the loft to the inverter. At 3.6kW AC you only notify your DNO under G98 (5-minute online form). You can mount and wire the DC side yourself. An electrician does the AC tie-in to your consumer unit under Part P (£150-300). MCS isn’t required for the install, but without it you can’t claim SEG. On a self-consumption-first build, that is about £60-100/year foregone, which is the price of admission for DIY.
Expected Performance
| Metric | Low-Mid (2,900 kWh) | High (6,500 kWh) |
|---|---|---|
| Annual yield | 4,084 kWh | |
| Self-consumption (with 5.12kWh battery) | 60% (2,435 kWh) | 97% (3,966 kWh) |
| Export | 40% (1,629 kWh) | 2% (98 kWh) |
| Grid savings | £674/year | £1,098/year |
| Export income (SEG 5p) | £81/year | £5/year |
| Total annual saving | £756/year | £1,103/year |
Moving from statutory 5p SEG to Octopus Outgoing Fixed at 12p adds roughly £115/year at low-mid consumption. High-consumption export is too small to matter.
Premium Option: 5kW Hybrid, 16 Panels, 16.1kWh Battery
Solar Energy Concepts System Rating
System Specification
Inverter: Solis 5.0kW S5 Smart Hybrid (£707-770). 5kW AC crosses the G98 line so you need a G99 application. Dual MPPT, 90-520V, 15A per MPPT, up to 8kW DC. That 8kW DC headroom is what makes 16-panel overpanelling sit comfortably.
Overpanel target: 142% of 5.0kW = 7,120W DC.
Panel Options
| Strategy | Panels | Total DC | Overpanel % | String Check |
|---|---|---|---|---|
| Best value | JA Solar 445W × 16 (8 East + 8 West) | 7,120W | 142% | Voc: 8 × 39.1V = 313V < 600V. Vmp: 8 × 32.65V = 261V in 90-520V. Isc 14.43A < 15A. OK. |
| Smallest footprint | Aiko Neostar 490W × 14 (7 East + 7 West) | 6,860W | 137% | Voc: 7 × 40.98V = 287V < 600V. Vmp: 7 × 34.4V = 241V in 90-520V. Isc 14.93A < 15A. OK. |
| Max overpanel | LONGi Hi-MO X6 455W × 16 (8 East + 8 West) | 7,280W | 146% | Voc: 8 × 39.15V = 313V < 600V. Vmp: 8 × 32.98V = 264V in 90-520V. Isc 14.79A < 15A. OK. |
Battery
Fogstar 16.1kWh ECO (£1,850). A premium 3-bed target is 11.9kWh+ (150% of 7.9kWh daily average). 16.1kWh clears it with room to spare and unlocks genuine multi-day resilience on dark winter weeks.
A 16kWh battery also makes winter tariff arbitrage worthwhile. Charge overnight on Octopus Go at 7.5p/kWh, discharge during the expensive morning peak at 27.69p. The spread is 20p/kWh across 16kWh, about £3.20 a day, or roughly £96 a month in the darkest three months of the year. Your exact number depends on how much capacity you actually cycle and the headline Octopus Go rate when you sign up. Treat £96/month as an upper-end estimate, not a guarantee.
Total System Cost
- DIY: £3,900-4,300
- MCS installed: £9,300-10,200
G99 and MCS
At 5.0kW AC you are above G98 (3.68kW) and need a G99 application before connecting. Formal application, not a notification. Most DNOs approve residential G99 in 45 days. SSEN regions (Scotland, parts of South) can take 60+ days where the local grid is constrained. If you’re in SSEN, submit before the panels arrive.
MCS is worth paying for on the premium build because: you qualify for SEG, you access time-of-use tariffs that some suppliers restrict to MCS installs, and it protects home insurance and resale value.
When This Is Worth It vs Overkill
Worth it if you have a heat pump, charge an EV at home, or plan to add one in 2-3 years. At 6,500 kWh/year the premium self-consumes 84% of generation, the 16.1kWh battery handles winter evenings, and Octopus Go arbitrage genuinely pays. E/W systems export less than south-facing at the same kWp, so the extra panels bring usable kWh, not exported kWh.
Overkill if you’re a low-consumption household with no electrification plans. The budget build already delivers 3.2-year payback and £756/year. The premium build adds £108/year at low-mid consumption for £1,700 more upfront. That is a 16-year payback on the incremental spend, which is not a trade worth making. Go budget, revisit when your consumption changes.
Expected Performance
| Metric | Low-Mid (2,900 kWh) | High (6,500 kWh) |
|---|---|---|
| Annual yield | 5,445 kWh | |
| Self-consumption (with 16.1 kWh battery) | 48% (2,615 kWh) | 84% (4,581 kWh) |
| Export | 51% (2,803 kWh) | 15% (837 kWh) |
| Grid savings | £724/year | £1,268/year |
| Export income (SEG 5p) | £140/year | £42/year |
| Total annual saving | £864/year | £1,310/year |
The Battery Is the Real ROI Lever
Three years of running a mixed solar setup taught me one thing that changed the way I talk about panels: you do not earn your ROI from generating more kWh. You earn it from consuming more of what you generate. On E/W that is doubly true, because your generation is already spread across the day and the bit that falls between peaks is the bit that lands in the battery instead of the grid at 5p.
A battery does two jobs. First, it time-shifts solar from midday to 6pm, which is worth the 23p gap between your cooker and the SEG rate. Second, it lets you charge overnight at 7.5p on Octopus Go through the winter months when solar isn’t pulling its weight, then run the house off that charge during the expensive morning peak. The UK saw 400% energy price hikes in 2022. A home battery is how you build a hedge against that happening again.
Tariff Strategy
Budget option: A flat-rate tariff is fine. Savings come from self-consumption, not clever tariff timing. If you want more export income, Octopus Outgoing Fixed at 12p doubles statutory SEG without any battery complexity. Skip Agile unless watching half-hourly prices is a hobby.
Premium option: Octopus Go or Intelligent Go is the natural fit. Charge at 7.5p overnight, carry through the morning peak on battery, let solar take over by mid-morning, evening peak comes from the battery again. If you have an EV, Intelligent Go typically wins outright. If consumption is very high without an EV, Octopus Flux (30p peak export 4-7pm) can edge it, but only if your setup actually discharges to grid at peak, which most domestic hybrids do not do by default.
DNO Region Notes
E/W systems export less aggressively than south-facing at the same kWp because generation is spread, not spiked. That makes DNO approval straightforward even in constrained areas.
- G98 (budget, 3.6kW): notification only. Days.
- G99 (premium, 5kW): application. UKPN/NPG 20-30 days; WPD/ENW 30-45; SSEN 45-60+.
Timeframes per ENA Engineering Recommendation G99 (statutory 45 working-day limit) and each DNO’s published connections SLA. Actual turnaround varies by local grid capacity.
ROI Comparison
| Budget DIY | Budget Installed | Premium DIY | Premium MCS | |
|---|---|---|---|---|
| Low-Mid Consumption (2,900 kWh/year) | ||||
| Upfront cost | £2,400 | £5,350 | £4,100 | £9,750 |
| Annual saving | £756 | £756 | £864 | £864 |
| Payback | 3.2 years | 7.1 years | 4.7 years | 11.3 years |
| 25-year return | £18,900 | £18,900 | £21,600 | £21,600 |
| High Consumption (6,500 kWh/year) | ||||
| Upfront cost | £2,400 | £5,350 | £4,100 | £9,750 |
| Annual saving | £1,103 | £1,103 | £1,310 | £1,310 |
| Payback | 2.2 years | 4.9 years | 3.1 years | 7.4 years |
| 25-year return | £27,575 | £27,575 | £32,750 | £32,750 |
Assumes 27.69p/kWh flat import, 5p/kWh SEG baseline, no energy inflation, 0.5%/year panel degradation. Real returns typically run higher because electricity prices rise and because Octopus Outgoing Fixed at 12p or Flux at up to 30p peak raises export value 2-6x over statutory SEG.
Generation Yield Source
All yield figures come from PVGIS, the European Commission’s free solar radiation database (satellite-measured irradiance, PVGIS-SARAH3 model).
| Parameter | Value |
|---|---|
| Location | 52.308°N, -0.717°W (Central England reference) |
| Roof slope | 28° |
| System loss | 14% |
| Database | PVGIS-SARAH3 |
| East aspect | -90° → 804.98 kWh/kWp/yr |
| West aspect | +90° → 788.18 kWh/kWp/yr |
| Clipping factor | 0.96 (applied to both) |
Verify for your postcode:
Key Takeaways
- E/W roofs produce about 81% of south-facing yield per kWp (PVGIS), but the demand-curve match is often better for a 3-bed family.
- Dual-MPPT hybrid inverter is non-negotiable. Each string runs at its own voltage.
- Battery is the single biggest ROI lever on this roof. Bridges the midday trough between the two peaks.
- Budget 3.6kW + 12 panels + 5.12kWh battery (£2,400 DIY) pays back in 3.2 years on a typical 3-bed.
- Premium 5kW + 16 panels + 16.1kWh battery is for heat-pump or EV households above 5,000 kWh/year.
Related Guides
- How roof tilt actually affects 10-year ROI. Three years of metered data on SE, NW, flat, and wall mounts.
- Start here: your guide to home solar. Read this first if you haven’t decided whether solar is for you.
- Battery overview. Sizing, chemistry, and why storage is the ROI lever on most UK roofs.
- Seven solar myths. Includes the diffuse-light myth that matters most for split-roof setups.
- Are solar panels worth it in 2026?. Full ROI breakdown.
- 4-bed NW/SE split roof. Same two-string logic, different orientation pair.
- 2-bed south-facing roof. The simplest scenario, for comparison.
