If you run a factory in Tamil Nadu, you already feel the number every month: a high-tension (HT) power bill that keeps climbing. Rooftop solar is one of the few levers that pushes it the other way, permanently, and with a payback most plant heads find hard to believe until they see it worked out. This guide walks through the real economics in plain English, so you can judge a proposal on its merits rather than on a salesperson's enthusiasm. Every rupee figure here is indicative, a planning range, not a quote for your site.
Why Tamil Nadu factories are a strong fit
Two things make the economics work here, and they compound. First, what you pay for grid power. Most industrial HT consumers in the state buy energy in the indicative ₹8–11 per unit range once you fold in energy charges, demand charges and the various riders. That is the price your solar has to beat, and beating a nine-rupee tariff is a very different exercise from beating a residential three-rupee one.
Second, the sunshine. Tamil Nadu sits in a genuinely good solar belt. A well-designed, unshaded rooftop system here typically generates an indicative 1,500–1,600 units (kWh) per kWp installed, per year. That specific yield is what turns installed capacity into money. A plant that produces more units for every kilowatt you paid for simply pays itself back faster.
Put the two together, an expensive grid to displace and generous irradiation to displace it with, and the industrial rooftop in this state becomes one of the better solar investments available anywhere in the country.
What actually drives the savings
Here is the single idea that matters most, and the one cheap proposals gloss over: your savings come from self-consumption, not generation.
When a solar unit is produced and your plant consumes it in the same instant, that unit replaces one you would otherwise have bought from the grid at ₹8–11. That is a full-value, ₹8–11 saving. But when your plant is producing more than it needs and the surplus flows back to the grid, that exported unit is credited at a far lower rate: the TNERC feed-in tariff, an indicative ₹3.10 per unit for a large plant (tiered ₹3.10–₹3.61 by size), less a network charge on generation. Same sunshine, same panel, wildly different value.
A unit you use yourself is worth ₹8–11. A unit you export may be worth ₹3.10. Solar economics live or die on which one you are producing.
This is why a factory that runs two or three shifts, with strong daytime load, is close to the ideal solar host. The generation curve and the consumption curve overlap through the middle of the day, so almost every unit produced is a unit used. A single-shift operation that goes quiet at 4pm, or a site with a large weekend shutdown, will export more and save less, unless the system is deliberately sized smaller to match the load it can actually absorb.
An indicative worked example: a 500 kWp plant
Numbers make this concrete. Take a mid-sized factory installing a 500 kWp rooftop system. The figures below are indicative planning numbers to show how the pieces fit, your site's tariff, load shape, roof and shading will move every line.
| Metric | Indicative value |
|---|---|
| System size | 500 kWp |
| Annual generation (~1,500 kWh/kWp) | ~7.5 lakh units/year |
| Indicative year-one saving (mostly self-consumed, net of O&M) | ~₹45–55 lakh/year |
| Indicative turnkey cost band | ~₹2.0–2.2 crore |
| Simple payback | ~3.5–5 years |
| Useful asset life | 25+ years |
The logic is straightforward. A 500 kWp system in a good Tamil Nadu location generates roughly 7.5 lakh units a year. If most of those units are self-consumed against a ₹8–11/unit tariff (with a realistic slice exported at the lower feed-in rate and routine O&M netted off), the first-year saving lands around ₹45–55 lakh (indicative). Against a turnkey cost of ₹2.0–2.2 crore, that puts simple payback in the region of 3.5 to 5 years. After that, the plant keeps producing for two more decades, the units it makes in years six through twenty-five are, in effect, near-free power.
Two honest caveats. Panels degrade slowly, expect output to ease by a fraction of a percent each year. And your grid tariff will almost certainly keep rising, which quietly improves the return on a fixed-cost asset. Both effects are small next to the headline, and they roughly cancel in the direction that favours the project.
Framing the ROI and IRR
A three-to-five-year payback on an asset that lasts twenty-five is, in financial terms, a strong internal rate of return. For a well-matched industrial rooftop in this state, an indicative IRR of around 18–22% is typical (higher after accelerated depreciation), again, indicative, and dependent on how much of your generation you actually consume, your real tariff, and the quality of the engineering.
The reason the IRR framing matters more than a raw payback figure is that it lets you compare solar against everything else competing for the same capital, a new line, a warehouse, a fixed deposit. Very few operational investments a factory can make return in the high-teens to low-twenties percent with this little ongoing risk. That comparison, not the payback headline, is usually what moves a CFO.
The self-consumption trap
Now the mistake that costs the most money, and the one to interrogate hardest in any proposal you receive.
Tamil Nadu, like most states, has moved industrial and commercial consumers toward net billing rather than the old one-for-one net metering. Under net billing, the units you export are not swapped for units you later import at the same price. They are credited at a much lower rate: the TNERC feed-in tariff, an indicative ₹3.10 per unit for a large C&I plant (₹3.10–₹3.61 by size under Order 8/2021), while every unit you draw from the grid still costs you the full ₹8–11.
So a system that is oversized for your daytime load falls into the self-consumption trap. It generates plenty on paper, but a big slice of that generation spills to the grid at a third of its potential value. The glossy proposal shows a large number of units produced; the actual bill saving is far smaller, and the payback quietly stretches from four years to seven or eight.
Key takeaways
- Tamil Nadu factories fit rooftop solar well: an indicative ₹8–11/unit HT tariff to displace and ~1,500–1,600 kWh/kWp of yearly yield.
- Savings come from self-consumption, not generation, a self-used unit is worth the full tariff; an exported one may fetch only the TNERC feed-in tariff, an indicative ₹3.10–₹3.61/unit by size.
- An indicative 500 kWp plant generates ~7.5 lakh units/year, saving roughly ₹45–55 lakh in year one, for a ~3.5–5 year payback.
- That translates to an indicative ~18–22% IRR on an asset that runs 25+ years, the number that matters when comparing against other capital projects.
- Under net billing, sizing to your actual daytime load, not to your roof area, is what protects the payback.
How sizing to daytime load changes everything
The fix for the trap is not a bigger system, it is the right-sized one. Instead of asking "how many panels fit on the roof?", the question that protects your return is "how many units can we actually consume on-site during daylight hours?"
Answering it well means starting from your load profile: your interval meter data, your shift pattern, your seasonal swings, your weekend and holiday shutdowns. A system sized so that nearly all of its generation is absorbed by your own plant keeps every unit at full tariff value, which is exactly what holds the payback near the bottom of that 3.5-to-5-year band rather than the top. Sometimes the commercially smartest plant is a little smaller than the roof allows. That is a feature, not a compromise.
This is also where good engineering earns its keep. Accurate shading analysis, honest generation modelling built on real irradiation data, a structural check the roof can actually carry, and a design matched to your load, these are what separate a proposal that delivers its indicative numbers from one that only looked good on the cover page. Engineered solar, commercially clear: the two have to travel together.
Where to go from here
If the ranges above look attractive for your operation, the next step is to replace the indicative numbers with your own. That means a real look at your tariff, your load profile, your roof and your shading, the inputs that turn a planning estimate into a bankable one. You can see how we have done this for other industrial sites in our case studies, or get a first-cut estimate for your own plant using the homepage solar estimator.