Solar Panel Orientation South vs East‑West in 2026

Last updated August 3, 2026 (Originally published July 16, 2026)

August 2026 update: This version corrects previously circulated Cleve Hill Solar Park figures, keeps the article focused on south-facing versus east-west design trade-offs, removes an unsupported IEEE 1547 orientation-specific claim, and adds a deeper EPC section on flat-roof capacity, export tariffs, structural loading, shading, and modelling traps.

Quick Answer

1. Which orientation usually produces more total energy per panel?
   South-facing, in most northern-hemisphere fixed-tilt systems.
2. Which orientation usually fits more capacity on a flat roof?
   East-west, because low tilt and back-to-back rows reduce row spacing and can increase installed capacity per square foot.
3. When does east-west make financial sense?
   When roof or land area is the binding constraint, or when morning and evening generation has higher value than midday export.
4. When does south-facing still win?
   When there is enough space for proper row spacing, the client is paid well for midday generation, or the project is optimised for maximum kWh per installed kW.
5. What is the biggest modelling risk?
   Treating an east-west array as one averaged orientation instead of modelling the east-facing and west-facing planes separately.
6. What should EPCs compare before committing?
   Installed kW, kWh per kW, kWh per square metre, ballast/structural load, obstruction losses, export value, self-consumption value and tariff-weighted cash flow.

Why This Matters for EPCs

Orientation is not just a layout preference. It changes installed capacity, yield modelling, interconnection assumptions, structural loading, cable routes, inverter clipping risk, export value and the payback story shown to the client.

Getting the choice wrong in either direction is expensive. If an EPC oversells south-facing yield on a site where roof area is the actual constraint, a competitor may win with a denser east-west layout. If an EPC oversells east-west density without modelling the real yield gap, obstruction losses and tariff value, the client may see a post-commissioning performance gap that was avoidable during proposal design.

EPC implication: Do not compare orientation only on annual kWh. Compare orientation on the constraint that actually matters for the project: roof area, land cost, tariff value, self-consumption, structural load or interconnection capacity.

What South-Facing versus East-West Orientation Means

Solar orientation is the compass direction a panel faces. In the northern hemisphere, a true-south orientation usually aligns panels with the sun’s strongest daily path and tends to maximise annual energy per panel. In the southern hemisphere, the equivalent default is true north.

A south-facing fixed-tilt array typically produces a strong midday peak. That works well where the project is paid fairly for midday generation, where there is enough land or roof area to space rows properly, and where the design goal is maximum kWh per installed kW.

An east-west layout works differently. Panels are installed back-to-back, with one side facing east and the other facing west, usually at a low tilt. The east face produces more in the morning. The west face produces more in the afternoon and early evening. The combined generation profile is flatter than a south-facing array.

That flatter curve can be useful. It may reduce midday clipping, improve self-consumption alignment and shift generation toward tariff periods that are worth more. But it can also reduce total annual kWh per panel, especially if the site has no meaningful value for morning or evening generation.

EPC implication: South-facing is usually the maximum-yield orientation. East-west is a density and timing strategy. Treat them as different commercial designs, not interchangeable compass options.

South vs East-West Is a Capacity Decision on Flat Roofs

On pitched roofs, the building often decides the orientation. The EPC can optimise around roof planes, setbacks, stringing and shade, but cannot freely choose the ideal azimuth or tilt without raising structural and aesthetic questions.

On flat commercial roofs, the decision is more open. The question becomes whether a low-tilt east-west array can fit materially more capacity on the same roof without creating unacceptable shading, ballast, O&M or modelling risk.

The clearest worked example comes from the American Solar Energy Society’s April 2026 analysis by Baker Makarem and Carla Monzer. In that example:

• The roof area is 4,290 square feet.
• The modules are 450 W.
• A conventional south-facing ballasted layout at 10° tilt fits 120 panels.
• That equals 54.0 kW DC.
• The same roof laid out east-west fits 68.4 kW DC.
• That is a 27% increase in installed capacity on the same footprint.

The key point is not that east-west panels are more efficient. They are not. ASES explicitly frames the result as a space-efficiency outcome. The east-west system produces more total energy in the example because it is a larger system on the same roof.

Panel for panel, south-facing still wins. Roof for roof, east-west can win if the roof is flat, the tilt is low, and the site is space-constrained.

ASES also found the east-west configuration was about 16% cheaper per installed kilowatt on average across its modelled locations. Treat that as a source-specific modelling result, not a universal cost rule. Per-location costs vary, and the source’s own cost assumptions should be checked before using the number in a proposal.

One important caution: ASES’s cost comparison uses zero rebates or incentives and states that the federal solar investment tax credit is no longer available. EPCs should not treat that statement as tax advice. Verify current federal, state and project-specific incentive treatment separately before applying the cost comparison to a client quote.

EPC implication: Use the ASES numbers as a layout example, not a universal design constant. Recreate the same comparison on the actual roof with the actual module, racking, setbacks, obstruction map and tariff.

Key Metrics EPCs Should Carry Into the Quote

Because Sanity tables are not available yet, these are written as a non-tabular key-metrics block.

• Typical yield direction: South-facing usually produces more annual kWh per installed kW in northern-hemisphere fixed-tilt systems.
• Common east-west yield gap: East-west layouts are often cited as producing around 10-15% less annual kWh than an optimal south-facing layout, although the actual gap depends on latitude, tilt, weather, row spacing and system design.
• ASES flat-roof example: 54.0 kW DC south-facing versus 68.4 kW DC east-west on the same 4,290 sq ft roof, using 450 W modules.
• ASES installed-capacity gain: 27% more DC capacity in the same footprint.
• ASES cost-per-kW result: around 16% lower cost per installed kW in the modelled east-west example, subject to source assumptions and site-specific verification.
• East-west practical tilt range: ASES frames the flat-roof advantage around a low 5-10° tilt. The UPDATED draft gives 10-15° as a common design range. The safest final framing is that east-west usually works at low tilt, commonly around 5-15°, with the exact value driven by racking, wind, ballast, drainage, soiling and row-spacing design.
• Autarco mounting-cost claim: up to 15% lower mounting hardware cost for east-west in its modelling, especially where ballast and roof load matter.
• Autarco obstruction-loss claim: rooftop obstructions such as HVAC units and skylights can create 5-10% shade losses on east-west systems if layout and stringing are not planned around them.
• Aurora NEM 3.0 example: under California’s Net Billing Tariff, an east-facing system delivered 20% less bill savings than southwest-facing in Aurora’s modelling.
• Aurora export-value example: exported power may be worth roughly 4-8 cents per kWh, compared with 30-80 cents per kWh for electricity consumed behind the meter, depending on hour and tariff.
• Cleve Hill Solar Park: 373 MW, approximately 360 hectares, over 550,000 modules, operational in July 2025, and cited as the UK’s largest solar farm.
• Cestas Solar Park: 300 MW across approximately 260 hectares, or about 0.87 hectares per MW, a useful density benchmark for orientation and land-use discussions.

How East-West Designs Increase Row Density

East-west arrays can fit more rows per hectare or per roof area because the lower tilt reduces the shadow each row casts on the next. A south-facing system normally needs spacing behind each row to prevent self-shading during low-sun conditions. A low-tilt east-west tent layout reduces or removes much of that row-to-row spacing requirement.

This is why east-west is often strongest on flat commercial rooftops, carports and constrained ground-mount sites. The layout is not chosen because each panel produces more. It is chosen because the site can hold more total installed capacity.

Cleve Hill Solar Park is often used as a reference point for east-west design. It should be cited accurately. The project is a 373 MW solar and battery storage project on the north Kent coast in the UK, spread across around 360 hectares, with over 550,000 PV modules. Its east-west orientation was selected to allow tighter panel spacing than a conventional south-facing layout would allow on the same site.

Cestas Solar Park in France is a better density benchmark. At 300 MW across around 260 hectares, it works out to roughly 0.87 hectares per MW. That is close to the 0.8 hectares per MW figure often cited for the project.

EPC implication: When comparing layouts, do not show only annual kWh. Show installed kW per roof area, kWh per square metre, hectares per MW, and tariff-weighted revenue.

Structural Loads, Ballast and Mounting Cost

Wind and snow loads vary by site. Orientation does not remove the need for structural engineering, but it can change the loading profile that the racking system must resist.

A conventional south-facing row can behave like a wall against wind from the north side. Air can stall behind the panels and raise negative pressure. In a back-to-back east-west tent, wind can move across the paired structure differently, reducing some of the uplift pressure that the racking and ballast must resist.

Solar Power World’s reporting includes this structural logic from GameChange Solar’s engineering perspective. pv magazine also reports that east-west configurations can reduce wind pressure and ballast requirements compared with south-facing layouts at certain inclinations. Autarco’s modelling claims up to 15% lower mounting hardware costs for east-west systems, especially on roofs where load-bearing capacity is already a constraint.

This is not a universal saving. A roof with complex obstructions, unusual wind exposure, strict fire access routes or poor structural reserve may still become more complicated, not simpler. Hardware also matters. KB Racking’s AeroRack E/W system, for example, offers low-tilt portrait and landscape options that overlap with common east-west ranges but do not map exactly to every modelling assumption.

EPC implication: Bring the structural engineer in before the layout is final. A density gain that fails the roof-load review is not a design advantage.

Export Tariffs Decide Whether Extra Capacity Earns Money

The biggest commercial trap is assuming that extra east-west capacity always creates extra value.

Under a one-to-one net-metering tariff, more generation can be financially useful even if it is exported, because the customer receives a credit close to retail value. Under a net-billing tariff, that assumption breaks. Extra panels may produce energy that is exported at a much lower value than electricity consumed on site.

California is the clearest example in the available draft. The Net Billing Tariff, commonly called NEM 3.0, applies to customers of PG&E, SCE and SDG&E whose systems applied for interconnection after 14 April 2023. 3D modelling under that tariff found southwest-facing to be the best bill-savings orientation, with south-facing and west-facing nearly as valuable. East-facing delivered 20% less bill savings than southwest-facing.

The important part is why. Aurora’s modelling says adding more panels did not recover the gap, because the extra output was mostly exported at a low rate instead of offsetting consumption. Aurora quantifies exported energy at roughly 4-8 cents per kWh, against 30-80 cents per kWh for energy consumed behind the meter, depending on the hour and tariff.

EnergySage describes the same tariff shift more broadly. Under NEM 3.0, there are hundreds of possible export rates depending on hour, day type and month, with average export credits around a fraction of retail value. The older NEM 2.0 grandfathering pathway is now closed for new commercial design decisions under the timeline described in the draft, subject to utility-specific confirmation.

The broader lesson is not California-only. Any market that pays less for export than for behind-the-meter self-consumption changes the orientation calculation. Extra east-west capacity only helps if the client can use the additional generation, store it, or export it at a rate that justifies the panels.

EPC implication: Pull the actual export tariff before selling a denser east-west layout. Do not model extra capacity as valuable unless the client’s rate structure supports that value.

Shading, Obstructions and Building Geometry

East-west layouts solve one shading problem and create another.

They reduce row-to-row shading, which is why they can pack more panels onto a flat roof. But they can become more vulnerable to rooftop obstructions because production is weighted toward lower sun angles in the morning and afternoon.

Autarco’s modelling says rooftop obstructions such as AC units and skylights can cause 5-10% shade losses on east-west systems if the layout does not route around them. That risk is easy to miss if the model treats the roof as a clean rectangle.

There is also a layout penalty. Solar Power World’s reporting notes that east-west panels are often installed in back-to-back pairs. If an obstruction blocks one side of the pair, the design may lose both positions, not just the single shaded panel. On a roof full of vents, curbs, skylights and HVAC equipment, that can erase the density advantage quickly.

Building geometry can also help east-west. On buildings that are not aligned true north-south, a tent-style layout may make better use of roof planes than forcing a conventional south-facing design. That is especially relevant for commercial roofs where the usable area is irregular and the client cares more about installed capacity than ideal azimuth.

EPC implication: Do a real obstruction map before promising east-west capacity. Roof clutter can make the theoretical density gain disappear.

Performance Trade-Offs: Yield, Value and Timing

East-west arrays generally produce less total electricity per installed kW than an optimally tilted south-facing system. The commonly cited yield gap is around 10-15%, but this should be treated as a starting point, not a fixed number.

EnergySage reports that an ideal east-west setup can lose around 10-15% annual production compared with a perfect south-facing system, while a typical roof with non-ideal angles may see a larger gap. The draft also references a Department of Energy example for San Diego where east-facing and west-facing panels at the city’s typical pitch lose different amounts compared with south-facing, with west doing better because San Diego mornings are cloudier than afternoons.

That example is useful because it shows why “face south” is only the geometry answer. Real production depends on the weather file. Cloud patterns, morning fog, high-albedo surfaces, soiling and heat all affect the actual best orientation.

Tilt matters too. A latitude-matching tilt is a useful first-pass rule in some contexts, but flat roofs often use much lower tilt because row spacing, wind, ballast, drainage and roof loading matter. East-west arrays rely on low tilt to keep both faces productive and to preserve the density advantage.

The daily generation shape can also matter. East-west spreads production into morning and afternoon hours rather than concentrating it at solar noon. That can reduce clipping and improve alignment with some load profiles. But Autarco’s modelling warns that the daily-spread advantage can be quite small at the 10-15° pitches typical of flat roofs, so it should not be treated as a guaranteed financial benefit.

EPC implication: Run yield and value separately. First compare physical output. Then compare tariff-weighted value. The layout that wins on kWh may not win on client economics.

PVsyst Modelling Traps for East-West Arrays

The final proposal number is only as good as the model behind it.

PVsyst models east-west arrays as a “dome” system. That means it treats the array as two coupled orientations: one east-facing plane and one west-facing plane. That is the right modelling approach because each half sees different irradiance throughout the day.

The problem comes when designers average multiple orientations into one. PVsyst’s own average-orientation documentation warns that calculating irradiance for a single averaged direction induces an error on the real irradiance value hitting the panels. The software also states that there is no tool yet to measure how large that error is.

For an EPC, that matters because an unquantified modelling error can become a production guarantee issue. If a mixed or east-west array is modelled as one averaged direction, the client-facing yield estimate may look precise but rest on a known modelling shortcut.

The same caution applies to shortcut shading assumptions. PVsyst’s unlimited-sheds approach is an approximation. It assumes continuous rows and neglects some edge effects, so it should only be used where rows are long relative to their width and where the simplification fits the actual site.

EPC implication: Before putting an east-west output number into a proposal, confirm that the model used the two-plane dome method and that shading assumptions match the real layout.

Glare and Permitting Considerations

Glare is site-specific. Lower-tilt east-west arrays can direct reflected light differently than steeper south-facing arrays, but there is no universal orientation rule that solves glare automatically.

Large projects, projects near roads, projects close to homes and projects near flight paths may require a dedicated glare study. The requirement depends on jurisdiction and site context. The safest approach is to treat glare as a permitting workstream, not as a claim that one orientation is always safer than the other.

The earlier unsupported claim linking IEEE 1547-2022 to east-west-specific inverter power-factor requirements should not be used. IEEE 1547 is a real interconnection standard, but the specific orientation-related claim was not independently verified in the draft material and should remain removed.

EPC implication: Use a project-specific glare study where needed, and do not cite standards for orientation-specific claims unless the exact source supports them.

2026 Timeline and Market Notes

These are not written as a table so they can be pasted directly into Sanity.

• 14 April 2023: California’s Net Billing Tariff, commonly called NEM 3.0, took effect for systems that applied for interconnection after this date under PG&E, SCE and SDG&E.
• Mid-April 2026: EnergySage’s timeline says grandfathering for older NEM 2.0 terms required systems to be installed and connected by a mid-April 2026 deadline. The source gives more than one date for this cutoff, so EPCs should confirm the exact deadline with the relevant utility before relying on it.
• July 2025: Cleve Hill Solar Park became operational, according to the UPDATED draft’s current framing. It should be cited as 373 MW, around 360 hectares, and over 550,000 modules.
• 2026 design implication: California commercial projects designed today should generally be modelled under net billing unless the specific customer has confirmed grandfathered tariff status.
• Tax/incentive caution: ASES’s April 2026 analysis models its cost comparison without federal rebates or incentives. Do not convert that into a general legal statement about tax-credit availability without separate tax verification.

What EPCs Should Do Before Committing to a Layout

• Confirm the actual site constraint. If the site is not roof- or land-constrained, south-facing may still be the higher-value layout.
• Confirm the roof type. East-west density arguments usually apply to flat roofs or flat ground-mount layouts, not pitched roofs where orientation is already constrained by the building.
• Keep east-west tilt low unless the model proves otherwise. Low tilt is what preserves the density and wind-loading advantage.
• Run both layouts on the exact same usable area. Compare installed kW, not just panel count.
• Overlay tariff value after yield modelling. Show how much energy is self-consumed, exported, stored or clipped.
• Pull export rates before recommending extra capacity. Under low export rates, additional east-west generation may add less value than the client expects.
• Get structural review early. Ballast, wind uplift, roof loading and snow loads can decide whether the denser layout is actually buildable.
• Map obstructions before stringing. HVAC units, skylights, vents and roof curbs can create disproportionate east-west losses.
• Check the modelling method. For PVsyst, confirm the two-plane dome method for east-west arrays rather than one averaged orientation.
• Avoid unsupported project stats. Cleve Hill, Cestas and other benchmark projects should be cited from current, identifiable sources.

Common Mistakes to Avoid

• Assuming south-facing is always best. It usually wins on annual kWh per installed kW, but not always on capacity density or tariff-weighted value.
• Assuming east-west is always better on flat roofs. It only works when low tilt, roof geometry, structural loading and export value support the design.
• Using the 27% capacity gain as a universal number. It comes from one ASES worked example and should be recreated for the actual roof.
• Ignoring export value. Extra panels do not help if the surplus exports at a weak rate and cannot be self-consumed or stored.
• Over-tilting east-west rows. Higher tilt can erase the row-density and wind-loading advantage.
• Trusting simulation software without checking orientation settings. An averaged model can create unquantified error.
• Ignoring rooftop obstructions. East-west arrays are vulnerable to low-sun shading from HVAC units, skylights and vents.
• Citing Cleve Hill incorrectly. Use 373 MW, approximately 360 hectares and over 550,000 modules.
• Treating glare as generic. Glare needs site-specific analysis where permitting authorities require it.

How This Fits Into a Reslink Workflow

The orientation decision only matters if the proposal model reflects the layout actually being sold. A client should not see a yield number produced from a south-facing default if the proposed roof layout is east-west, low-tilt and obstruction-constrained.

In a Reslink workflow, the orientation comparison should happen before the proposal is finalised:

• detect usable roof area and roof planes
• identify obstructions and shade sources
• model south-facing and east-west options on the same footprint
• compare installed capacity and yield
• apply customer-specific tariff and export assumptions
• show the client the trade-off between maximum annual kWh and maximum usable capacity
• keep the final yield number tied to the selected layout

Reslink’s design workflow helps EPCs keep layout, shading, yield and proposal assumptions connected, so the number shown to the client reflects the system being proposed rather than a default orientation model.

See how Reslink handles orientation-specific yield modelling in the proposal itself: Book a Demo.

Frequently Asked Questions

Q1. Does east-west orientation ever produce more total energy than south-facing?

Yes, but usually only when the comparison is roof-for-roof rather than panel-for-panel. Panel-for-panel, south-facing normally produces more annual kWh. In ASES’s flat-roof example, the east-west system produced more total energy because it fit 27% more capacity into the same footprint.

Q2. Does the 27% capacity gain apply to every flat roof?

No. It comes from one ASES worked example using a 4,290 sq ft flat roof, 450 W modules and a specific south-facing versus east-west layout. Treat it as proof that meaningful capacity gains are possible, not as a number to copy into every quote.

Q3. What tilt should EPCs use for east-west flat-roof arrays?

Low tilt is the key. ASES frames the flat-roof advantage around 5-10°. The UPDATED draft also cites 10-15° as a common practical range. The right value depends on racking, ballast, wind, drainage, soiling, roof load and self-shading.

Q4. When does south-facing still beat east-west?

South-facing usually wins when roof or land area is not the binding constraint, when a proper tilt and row spacing are achievable, and when the tariff pays fairly for midday generation. It is still the default for maximum kWh per installed kW.

Q5. Does east-west help under time-of-use tariffs?

It can, but not automatically. East-west spreads output into morning and afternoon hours, which can help where those hours are worth more. But if the extra generation is exported at a weak rate, as in some net-billing structures, the additional panels may not improve bill savings enough to justify the layout.

Q6. Why is California NEM 3.0 relevant to orientation?

NEM 3.0 shows how export value can change the layout decision. Aurora’s modelling found an east-facing system delivered 20% less bill savings than southwest-facing, and adding panels did not recover the gap because exported energy was worth far less than self-consumed energy.

Q7. What is the biggest PVsyst mistake for east-west arrays?

The biggest mistake is modelling east-west as one averaged orientation. PVsyst’s correct east-west approach is the dome model, with separate coupled east-facing and west-facing planes. Averaging orientations can induce an irradiance error that PVsyst says is not currently measurable by a dedicated tool.

Q8. Does east-west reduce shading losses?

It reduces row-to-row shading, but it can increase sensitivity to obstruction shading. HVAC units, skylights, vents and roof curbs can cast long morning and evening shadows that hit east-west output harder than expected.

Q9. Can bifacial modules close the east-west yield gap?

They can recover some output by capturing reflected light, especially on high-albedo surfaces such as light gravel or reflective roofing. The exact gain is site-specific and should be modelled rather than assumed at a fixed percentage.

Q10. Is Cleve Hill Solar Park the best east-west benchmark?

Cleve Hill is useful because it is a large, operational UK project with east-west orientation, but Cestas Solar Park is better as a density benchmark. Use Cleve Hill for accurate project reference and Cestas for hectares-per-MW comparison.

Sources

• American Solar Energy Society — East-West vs. South-Facing Solar: When More Panels Beats Perfect Direction
• Solar Power World — East-west solar projects maximize the number of solar panels on an array
• pv magazine — East-west rooftop solar ideal for energy communities under net metering
• Autarco — kWh Face-Off: East-West vs South
• EnergySage — Net metering 3.0
• EnergySage — How solar panel orientation and angle affect performance
• PVsyst — Plane orientation modes
• PVsyst — Average orientation
• KB Racking — New AeroRack East-West system