Solar Manufacturing Insights Without the Jargon
Welcome to SLR Insights. Every other week we’ll send you something useful from the floor of an Indian N-TOPCon manufacturing operation. No jargon, just things we wish more people in the industry talked about openly. Here’s what’s on our mind this week.
The PERC to TOPCon shift, and what it actually means
A few years ago, PERC was the obvious answer for almost any solar project in India. Today, if you're specifying modules for a tender that closes after June 2026, PERC is mostly off the table.
The shift happened faster than anyone expected. In the first half of 2025, India added 44.2 GW of solar module capacity. TOPCon was 39.9 GW of that. PERC was 3 GW.
What’s driving the shift isn’t just a better efficiency number on a datasheet. Field studies and published laboratory data indicate TOPCon typically shows lower temperature losses and lower degradation rates over 25 years, and ALMM List-II tenders are now being written to spec it directly. SECI’s 870 MWp Gujarat tender required “monocrystalline Si TOPCon cells, glass-glass, ≥21.65% efficiency.” That spec excludes PERC by design.
If you’re still benchmarking project economics against PERC datasheets, you’re benchmarking against a technology that manufacturing trends and tender specifications suggest is being phased out.
Sizing solar capacity, the simple version
The most common question we get from anyone sizing a residential system: how much capacity do I actually need?
The rule that works most of the time: take your monthly electricity units and divide by 120. That's roughly your kW. Using 360 units a month means you need about 3 kW. That happens to be the cap under the PM Surya Ghar subsidy too.
For C&I and larger systems, the same logic applies but with load patterns layered in. A 1 kW system in Rajasthan generates 5 to 6 units a day. The same 1 kW in Kerala or the northeast generates 3.5 to 4 units. Same panel, same datasheet, very different output. Sizing isn't about copying a number off a spec sheet. It's about matching it to where the system will live.
For space planning, work with 80 to 100 sq. ft per kW with current modules. Newer 580 to 610 Wp N-TOPCon modules push that down toward 70 to 80 sq. ft. So, a 3-kW residential install needs roughly 240 to 300 sq. ft of shadow-free roof.
Why daily output varies even on sunny days
Two solar plants in the same region, same module spec, same tracker, can produce different amounts of energy on the same sunny day. This confuses people who expect solar to behave like a fixed-output device.
The biggest variable is module temperature. Datasheets are tested at 25°C. In Indian summer, modules often run at 65 to 75°C. A PERC module loses about 13.6% of its rated output at 65°C. TOPCon loses 12%. The difference comes from the temperature coefficient: PERC sits around -0.34%/°C, TOPCon at -0.30%. Small numbers on paper, real money over a year.
Layered on top: irradiance angle, dust, soiling, partial shading from a passing cloud, inverter clipping. Each chips off a percentage. Most are predictable. Some aren't.
Based on operational observations from sites in Rajasthan and Gujarat, the spread between comparable systems often comes down to install quality. Mounting alignment, cable losses, shading, inverter sizing. The panel performs to spec. The variance comes from what surrounds it.
From SLR: what we learned setting up a fully automated N-TOPCon line
When we built out our N-TOPCon line, the part that surprised us most wasn't the cell architecture or the wafer chemistry. It was how much of long-term module performance comes down to process consistency.
A TOPCon cell needs the tunnel oxide layer at 1 to 2 nanometres. Off by half a nanometre, the carrier selectivity drops. Off by more, the module may underperform across its operational life, often without detection until field monitoring data surfaces in year five or later. Automation does one job here: keep every cell coming out identical, panel after panel, shift after shift.
The other thing that shifted our thinking: most module datasheet specs are minimums. A 22.05% efficiency line means the module is at least 22.05%. Whether it lands at 22.05 or 22.4 depends on bin selection, wafer quality, soldering integrity, and EL inspection rigour. None of it shows up in the brochure. All of it shows up in field performance over 25 years.
The harder, less visible work is what happens between the cell line and the lamination station. That's where lifetime yield gets locked in or lost. And it's where most buyers never think to ask questions.
