Insights · Report · November 2026

Solar Technology Outlook 2030

Solar is a semiconductor industry. Its efficiency, materials and costs move on a documented roadmap. This outlook collects the independent data on where the technology is going, with every figure sourced and every projection labeled.

Read the outlookMethods & sources
35.5%
Certified tandem lab cell (LONGi, ESTI), July 2026
26%
Module price decline per doubling of shipments
12.0 → 4.3
Heterojunction silver, mg/W, 2025 to 2036 (projection)
649 TWh
Projected U.S. data center use in 2030 (LBNL reference case)

01

Efficiency keeps climbing

Laboratory records lead commercial modules by several years. In November 2023 a certified tandem cell first exceeded the 33.7% limit of any single-junction cell, and the industry roadmap expects every mainstream technology to keep improving.

Figure 1

Tandem cells have passed the single-junction limit

Certified perovskite-silicon tandem efficiency rose from 33.9% to 35.5% between November 2023 and July 2026. These are laboratory cells, not commercial products.

30%32%34%36%20242025202633.7% · theoretical limit, any single-junction cell~29.4% · theoretical limit, crystalline silicon35.5%33.9%
Sources: LONGi announcements (33.9%, November 2023; 34.6%, June 2024; 34.85% NREL-certified, April 2025; 35.5% ESTI-certified, reported by PV Tech, 16 July 2026). Single-junction limit: Shockley–Queisser. Crystalline silicon limit: Niewelt et al., Solar Energy Materials & Solar Cells (2022). Laboratory cells only.

Figure 2Includes projections

Every mainstream technology is projected to improve

Average module efficiency in mass production, as surveyed and projected by 38 manufacturers, suppliers, research institutes and consultants. Industry averages; not Solarix data.

2025 · survey2036 · projectionPERC 21.7%22.1% Heterojunction (SHJ) 23.5%24.9% TOPCon 23.5%25.2% TBC 24.1%25.6% HBC 24.4%26.3%

Silicon-based tandem modules, by the same survey: 27.4% in 2028 and 31% in 2036 (projections).

Source: VDMA, International Technology Roadmap for Photovoltaics (ITRPV), 17th Edition, June 2026. 2025 values are survey results; 2036 values (dashed) are industry-survey projections, not measurements. Tandem values are given in text only.

02

More power from every square meter

Module output per square meter equals efficiency times 1,000 W/m². Higher efficiency means fewer modules, less racking and less land per megawatt.

235 W/m²
TOPCon and heterojunction modules, 2025 average
310 W/m²
Tandem modules, 2036 projection
+25%
Electricity per acre, U.S. tracking plants, 2011 to 2019
~2.5 acres
Per GWh a year, U.S. tracking plants, 2019 median

Sources: ITRPV 17th Edition (2026), efficiency × 1,000 W/m²; Bolinger & Bolinger, “Land Requirements for Utility-Scale PV,” IEEE Journal of Photovoltaics (2022), LBNL, plants built through 2019; acres per GWh derived from the 394 MWh per acre per year median. The tandem figure is a projection.

Figure 3Includes projections

Bifacial modules on trackers are now the default

Two-sided modules are the global majority, and nearly all new U.S. utility-scale capacity tracks the sun. The dashed bar is a projection.

Bifacial share, global modules68%2025 · ITRPV 17th Ed.Bifacial share, 2036 projection85%Projection · ITRPV 17th Ed.U.S. utility-scale on trackers96%New capacity built in 2025 · LBNL 2026 Edition
Sources: VDMA ITRPV 17th Edition (2026); Lawrence Berkeley National Laboratory, Utility-Scale Solar, 2026 Edition (October 2026), share of new capacity.

03

Doing more with less material

Each watt is being made with less silicon and, by industry projection, far less silver. The roadmap expects heterojunction to see the steepest reduction.

Figure 4Includes projections

Silver per watt is projected to fall sharply

Silver paste is among the most expensive non-silicon materials in a cell. The roadmap expects heterojunction cells to use roughly two-thirds less by 2036 as copper replaces part of the paste. Industry medians; not Solarix data.

20252036 · projectionHeterojunction 12.0 mg/W4.3 mg/W TOPCon ~10 mg/W6.3 mg/W
Source: VDMA ITRPV 17th Edition (June 2026). 2025 medians are survey results; 2036 values are projections. Fraunhofer ISE reports silicon use fell from about 16 g/W in 2004 to about 2 g/W in 2024.
~1 year
Energy payback for a wafer-silicon system in southern Europe
2 g/W
Silicon per watt, 2024, from 16 g/W in 2004
≤630
kg CO₂e/kWp to list on EPEAT; ≤400 for Climate+
−15.7%
Energy payback time per doubling of production

Sources: Fraunhofer ISE, Photovoltaics Report (14 July 2026); Global Electronics Council, EPEAT-ULCS-2023.

04

What a megawatt-hour really costs

Module prices have fallen on a steady learning curve for five decades. Project costs depend on more than the module: U.S. utility-scale costs rose after 2022 with higher capital and financing costs, and as projects moved to less sunny regions.

~26%
Module price decline per doubling of cumulative shipments, 1976–2025
$0.09/W
Average crystalline module spot price, end of 2025
$49 → $61
U.S. utility-scale levelized cost per MWh (real 2025$) before tax credits, 2022 to 2025
$20/MWh
DOE target for utility-scale solar by 2030

Sources: VDMA ITRPV 17th Edition (2026); Lawrence Berkeley National Laboratory, Utility-Scale Solar, 2026 Edition (October 2026; levelized cost without tax credits, 2024 and 2025 values preliminary); U.S. DOE Solar Energy Technologies Office 2030 goals (2021), unsubsidized, average U.S. resource. The DOE figure is a policy goal, not a projection.

05

Reliability is engineered

Fielded modules lose about half a percent of output a year at the median. Independent testing shows that quality still varies widely between manufacturers and bills of materials.

0.5%/yr
Median degradation, review of ~2,000 measured rates
0.75%/yr
Median performance loss, 4,915 fielded inverters
87%
Manufacturers with at least one test failure, Kiwa PVEL 2026
61%
Bills of materials with at least one failure, Kiwa PVEL 2026

Sources: Jordan & Kurtz, Progress in Photovoltaics (2013); Jordan et al., “Photovoltaic fleet degradation insights,” Progress in Photovoltaics (2022); Kiwa PVEL, PV Module Reliability Scorecard (May 2026), as reported by PV Tech and pv magazine.

06

Powering the compute build-out

Data center electricity demand is rising fast and the range of forecasts is wide. Utility-scale solar supplied nearly half of new U.S. generating capacity in 2025.

Figure 5Includes projections

U.S. data center electricity use, 2024 to 2030

Lawrence Berkeley National Laboratory’s 2025 Update estimates that U.S. data centers used 192 TWh, 4.7% of U.S. electricity, in 2024. It projects 464 TWh in 2028 and puts 2030 at 11.8% in the reference case, within a 9.5% to 15.3% range.

0 TWh200 TWh400 TWh600 TWh2024192 TWh · 4.7% of U.S. use2028464 TWh · reference case2030649 TWh · 11.8% · reference case
Source: Smith et al., United States Data Center Energy Usage Report: 2025 Update, Lawrence Berkeley National Laboratory (June 2026), reference case. Dashed values are projections. Utility-scale solar share: LBNL, Utility-Scale Solar, 2026 Edition. The IEA estimates global data center use at 485 TWh in 2025 and projects about 950 TWh in 2030 (Key Questions on Energy and AI, 2026).

07

Beyond 2030

Several next-generation technologies are past the laboratory. Their status, as of November 2026:

TechnologyStatus, November 2026Evidence
Back-contact cells (TBC, HBC)Mass productionUp to 28% of cells by 2036 (projection)
Copper-containing metallization for heterojunctionEarly commercialCopper-containing pastes in use; plating under 10% share
Busbarless (0BB) cell layoutsScaling~48% of cells by 2036 (projection)
Thinner heterojunction wafersRoadmap100 µm minimum for heterojunction G12 wafers by 2036 (projection)
Perovskite-silicon tandem modulesEarly commercial, small volumeOxford PV shipped 24.5% modules from a pilot line in 2024. ITRPV’s survey expects tandem mass production to begin around 2027, with tandem cells appearing in the market after 2028, and says the start remains “extremely difficult to predict.”

Sources: VDMA ITRPV 17th Edition (2026); Oxford PV announcement as reported by pv magazine and Solar Power World (5 September 2024). Projections labeled.

Methods & sources

This outlook reports third-party data only. It does not describe Solarix products. Product specifications are on the Utility ULTRA HJT page.

Measured values are drawn solid; projections are dashed and labeled. Industry figures are averages or medians across manufacturers, not Solarix data. Laboratory cell, laboratory module, commercial module and mass-production averages are identified each time an efficiency appears. Where sources give a range, the range is shown.

Primary sources: VDMA ITRPV 17th Edition (June 2026); Fraunhofer ISE Photovoltaics Report (July 2026); Lawrence Berkeley National Laboratory, Utility-Scale Solar 2026 Edition, Data Center Energy Usage Report: 2025 Update; Niewelt et al. (2022); Bolinger & Bolinger (2022); Jordan & Kurtz (2013); Jordan et al. (2022); Kiwa PVEL Scorecard (2026); IEA, Key Questions on Energy and AI (2026); U.S. DOE SETO (2021); Global Electronics Council EPEAT-ULCS-2023.

Updated annually as new editions publish: ITRPV in June, Fraunhofer ISE in July, LBNL in autumn, Kiwa PVEL Scorecard each spring. Last updated November 2026.