Solar Energy is the craft that turns irradiance into a bankable watt: photovoltaic technology from polysilicon and wafers through silicon solar cells and laminated solar modules, plus thin-film solar cells deposited on glass, perovskite and tandem stacks still leaving the laboratory, and concentrated solar plants that run steam cycles and molten-salt tanks. Global solar capacity reached 2,392 GW at the end of 2025 after 511 GW of additions, of which 510.3 GW was photovoltaic . Crystalline silicon still accounts for about 98% of photovoltaic production, with module shipments around 706 GW in 2025, while China's share of polysilicon, ingots, wafers, cells and modules exceeds 80% . That is why process engineers, reliability testers, CSP solar-field designers and recyclers are hired under one discipline name, and why the work does not transfer among them.
Challenges in Solar Energy Recruiting
Photovoltaic production that is not a process capability
Nameplate gigawatts describe floor space, not a process of record. The IEA found that China has invested more than USD 50 billion in photovoltaic manufacturing since 2011, created more than 300,000 manufacturing jobs in that period, and now holds more than 80% of every stage from polysilicon through modules — more than double its share of global demand. On capacity then under construction, its share of polysilicon, ingot and wafer output was heading toward 95%, Xinjiang held about 40% of polysilicon, and one facility made one in every seven panels worldwide . VDMA's 17th ITRPV still reads the imbalance at terawatt scale: more than 1,230 GWp of polysilicon, ingot and wafer capacity, 1,260 GWp of cells and 1,460 GWp of modules against 706 GWp of shipments in 2025 . Electricity is over 40% of polysilicon cash cost and nearly 20% for ingots and wafers, which is why a line sited against cheap industrial power does not reproduce elsewhere . An engineer who ramped a PERC module hall to nameplate is not, by that fact, the person who can convert it to TOPCon, qualify a second-source silver paste, or stand up a greenfield cell line on a different toolset. Photovoltaic production is a sequence of qualified recipes. Capacity without those recipes is inventory.
Silicon solar cells that no longer share one factory recipe
Behind the 98% crystalline-silicon share sits a cell-architecture turnover that a generic "PV process" title conceals . TOPCon overtook PERC in market share in 2024 and held the lead through 2025; heterojunction and back-contact stacks are the next capacity bets; silicon-based tandem cells are expected to enter mass production around 2027, with module efficiency near 27.4% in 2028 . Those families do not share a process window. PERC is an aluminum back-surface-field descendant. TOPCon needs a tunnel oxide and a doped polycrystalline silicon contact, deposited by LPCVD or PECVD, then a metallization paste that does not punch through the oxide. Heterojunction cannot see a high-temperature fire-through step. Back-contact cells move the grid off the sunny face and rewrite stringing and inspection. Each change rewrites wet benches, tube furnaces, laser tools, screen printers and electroluminescence criteria. Photovoltaic materials follow the same split: 6N–8N polysilicon at roughly 40 kWh/kg, silver paste that still took about 13 g/kW in 2024 and 29% of industrial silver demand, and glass that is about 75% of module mass . A candidate who "worked cells" may have owned one of those stacks, supported a conversion, or only written weekly reports beside a line that someone else qualified.
Thin-film solar cells that skip the wafer chain
About 2% of photovoltaic production is not a wafer at all . Cadmium telluride, copper-indium-gallium-selenide and older amorphous-silicon stacks deposit the absorber on glass and never buy polysilicon, never saw a wafer, and never fire a silver front grid. First Solar's CdTe route, developed in California and Ohio and run in the United States, India, Malaysia and Vietnam, turns a sheet of glass into a finished module in about four hours under one roof, uses 1–2% of the semiconductor mass of a crystalline-silicon device, and is planned at about 25 GW of global annual capacity by 2026 . That is a vapor-transport and glass-handling craft with its own closed-loop tellurium and cadmium accounting, not a diluted version of a TOPCon hall. CIGS lines add vacuum co-evaporation or sputtering and a different set of alkali and buffer chemistries. Hiring a crystalline-silicon integration engineer into a CdTe plant, or the reverse, looks tidy on a "thin film versus silicon" slide and fails on the first tool. The 2% share is small; the non-transferable skill is not.
Concentrated solar plants that store heat, not electrons
Concentrated solar power is still filed under Solar Energy and is not photovoltaic technology. REN21 counted 7.2 GW of CSP in service at the end of 2024 after 350 MW of new connections, 250 MW of them in China, against a Chinese project pipeline of about 8.1 GW — already larger than the entire operating fleet . Spain still holds the largest installed base at 2.3 GW from a 2008–2013 trough programme, with no commercial additions in 2024 . New Chinese "CSP+" hybrids put 100 MW-class molten-salt towers or Fresnel fields beside hundreds of megawatts of PV, with eight to sixteen hours of thermal storage so the thermal machine can peak-shave a variable desert base . The people who deliver that work size heliostat fields, trough loops and receivers, specify nitrate-salt tanks, and debug leaks and optical error. They do not string modules. A utility-scale PV plant engineer who has never commissioned a hot-salt tank is not a CSP solar-field engineer, and the 2,392 GW photovoltaic fleet does not train them .
Perovskite solar cells still waiting on IEC 61215 lifetime
Perovskite solar cells and tandem solar cells now dominate efficiency headlines because they can beat the single-junction silicon limit on a research cell. The National Laboratory of the Rockies (formerly NREL) still treats those results as independently confirmed measurements under IEC 60904-3 or ASTM G173 at 25 °C, logged in crystalline silicon, thin-film, emerging-PV and hybrid-tandem families, not as product warranties . IEC 61215-1:2021 is explicit about the gap the other way: it qualifies terrestrial flat-plate modules, including crystalline-silicon and thin-film types and, in this edition, bifacial constructions, for open-air operation, and it states that test results are not a quantitative prediction of lifetime . ITRPV places silicon-based tandem mass production around 2027 . Until a tandem stack survives encapsulation, humidity-freeze, potential-induced degradation and a bill-of-materials change retested under IEC TS 62915, a 1 cm² record is a materials result. Lead-halide chemistry, solvent systems, slot-die or evaporation uniformity at module area, and recombination layers between a perovskite top cell and a silicon or CIGS bottom cell are a different hiring problem from running a TOPCon printer. The engineer who published the cell and the engineer who can freeze a module process of record are rarely the same person.
Photovoltaic recycling that cannot live on glass recovery
End-of-life mass is no longer a 2030s footnote. IRENA's 1.5 °C pathway puts cumulative panel waste above 200 million tonnes by 2050 and annual arisings above 25 million tonnes, more than 21% of global e-waste by weight, with China, the European Union, India and the United States each facing more than a million tonnes a year from 2043 . If panels were collected systematically, the IEA estimates recycled flows could cover more than 20% of the industry's aluminum, copper, glass and silicon demand and almost 70% of silver demand between 2040 and 2050 in its net-zero pathway — but today's processes rarely earn enough from recovered fractions to pay for the plant . IEA PVPS Task 12's 2025 country review is blunt: volumes are still low, logistics are thin, secondary markets are immature, and the European Union's WEEE rules, Korea's EPR scheme and a patchwork of U.S. state rules have not yet made high-value recovery the default . Mechanical delamination, the commercial workhorse for crystalline-silicon modules, recovers glass and frames; silver, silicon and polymer-bound lead need thermal or chemical steps that most sites do not run. CdTe take-back is a different hydrometallurgical circuit. Photovoltaic recycling is therefore its own process family, not an O&M add-on and not a wafer-fab skill with the serial numbers filed off. A mis-specified hire leaves a take-back contract that still ships glass-rich shred to landfill.
Photovoltaic testing sequences no solar modules datasheet can replace
The last filter is evidence. IEC 61215 design qualification and IEC 61730 safety qualification show a construction survived a defined sequence, not that this factory's weekly output still matches the qualified bill of materials, and they are silent on lifetime . Independent cell charts confirm a research device, not a production bin . A datasheet that lists Pmax and a degradation warranty does not say who owned the electroluminescence criteria, the PID and LID gates, or the IEC TS 62915 retest after a glass, EVA or paste change. A "24% cell" may be a champion, a line average, or a nameplate never met after lamination. Scale does not travel: a spin-coated perovskite coupon, a tandem pilot and a multi-gigawatt TOPCon hall are three jobs. Chemistry does not travel either. CdTe vapor transport is not CIGS co-evaporation, and neither is a molten-salt receiver. The cost of getting that wrong is a module lot that fails incoming inspection, a CSP hybrid that cannot dispatch the storage hours in the offtake contract, or a recycling line that recovers glass and still misses a recovery-rate clause. The probe that works is narrow: which absorber and cell architecture, which tool and process of record, which qualification sequence, which area and throughput, and which metric the candidate can defend with data rather than with a logo.
References
- Renewable capacity highlights 2025 — International Renewable Energy Agency (IRENA). (accessed 2026-09-27)
- Solar PV Global Supply Chains — Executive Summary — International Energy Agency (IEA). (accessed 2026-09-27)
- International Technology Roadmap for Photovoltaic (ITRPV), 17th Edition — VDMA Photovoltaics Equipment. (accessed 2026-09-27)
- First Solar — Company Overview — First Solar. (accessed 2026-09-27)
- Renewables 2025 Global Status Report — Concentrated Solar Power — REN21. (accessed 2026-09-27)
- Best Research-Cell Efficiency Chart — National Laboratory of the Rockies (formerly NREL). (accessed 2026-09-27)
- IEC 61215-1:2021 Terrestrial photovoltaic (PV) modules — Design qualification and type approval — Part 1: Test requirements — International Electrotechnical Commission (IEC). (accessed 2026-09-27)
- End-of-life management for a circular economy: Solar PV panels — International Renewable Energy Agency (IRENA). (accessed 2026-09-27)
- Status of PV Module Recycling in IEA PVPS Task 12 Countries — IEA Photovoltaic Power Systems Programme (IEA PVPS). (accessed 2026-09-27)