Delhi will fund the whole 3 kW system. Then what?

On 1 September 2026 the Delhi Cabinet approved the second amendment to the Delhi Solar Energy Policy, 2023. A domestic consumer using up to 400 units a month can now have a rooftop system of up to 3 kW installed at no upfront cost, subject to technical feasibility.

The money works like this. The central PM Surya Ghar subsidy contributes up to Rs 78,000, the Delhi government matches it with another Rs 78,000, and a further state top up covers whatever gap remains between the two subsidies and the tendered cost of the plant. The stated target is 2.30 lakh households and 500 MW of additional rooftop capacity by March 2027.

If you sell rooftop solar in Delhi, two things follow immediately, and they point in opposite directions.

The part that hurts

Inside that eligibility box, you cannot compete on price. Nobody wins an argument against zero.

So the first job is not a pitch, it is a filter. For every live Delhi residential enquiry, find out in the first conversation whether the customer consumes under 400 units a month and wants 3 kW or less. If they do, you are no longer selling them a system on price, and pretending otherwise wastes both your time and theirs.

The second job is harder. Customers who are not eligible will still have heard that Delhi is giving solar away, and will ask you why they should pay. Have the specific answer ready: the scheme covers homes under 400 units a month, up to 3 kW, subject to technical feasibility, and their consumption or their roof puts them outside it.

The part that is an opportunity

Now the other number. Across all of Delhi, roughly 10,073 rooftop systems have been installed to date. The target is 2.30 lakh by March 2027.

That is not a marketing problem. Reaching it needs something like 12,000 installations a month, sustained, in a city that has managed about ten thousand in total. The binding constraint is delivery capacity: installers, supply chains and the DISCOM's own processing bandwidth. A subsidy redesign does not create any of those.

For an EPC, that means the work is going to go to whoever can actually deliver at rate, and the tendered panel is going to be short of hands long before it is short of demand. If you want that volume, the qualification to prepare is throughput: crews, standardised designs, a repeatable application and approval process, and a commissioning cycle you can measure in days.

Understand what you are signing up for, though. Tendered work is priced by the tender. The margin per job will be thinner than your own retail pricing, and it becomes a volume business with volume discipline, which is a different company from a twenty jobs a month retail EPC.

Where the retail margin actually sits now

The scheme has drawn a box. Everything outside the box is still yours, and some of it just became easier to sell.

• Homes above 400 units a month. The higher consumption households, which are also the ones with the best payback and the least price sensitivity. Your core market, untouched.
• Systems above 3 kW. A customer with a 5 kW roof and a 5 kW load is outside the free scheme entirely.
• Group housing societies, on common area load. Different subsidy structure, larger systems, one negotiation covering many homes.
• Commercial and industrial, where the arithmetic never depended on residential subsidy in the first place.
• Customers who want a specification the tender will not carry. A tendered plant is built to the tender's minimum. There is a real, honest sale in better modules, a better inverter, a longer workmanship warranty and a service commitment.

The last one deserves care. Selling against a free system on quality is legitimate when you can show the difference and unpleasant when you cannot. Show the specification side by side, name the modules and the inverter, and let the customer decide.

What Delhi is telling every other state

Delhi is not the only state layering its own money on top of PM Surya Ghar, and the direction of travel is clear enough. State top ups, and eventually aggregated or zero cost routes, will keep arriving for small residential systems.

The strategic read for an EPC is that the small residential job is drifting toward being a subsidised, tendered, thin margin volume product, while the interesting margin moves to larger residential, societies and C and I. That does not mean abandoning residential. It means being deliberate about which residential you chase, and building the throughput to survive on the rest.

The mechanics of the central subsidy underneath all of this are in the subsidy maths every EPC should know, and the demand picture behind it in 50 lakh homes done.

The short version

• From 1 September 2026, Delhi domestic consumers under 400 units a month can get up to 3 kW at zero upfront cost, subject to technical feasibility.
• Funding is Rs 78,000 central, Rs 78,000 from Delhi, plus a top up covering the rest of the tendered cost.
• Target is 2.30 lakh households and 500 MW by March 2027, against about 10,073 systems installed in Delhi so far. Delivery capacity is the constraint, not demand.
• Inside the eligibility box, stop competing on price. Filter those enquiries out in the first conversation.
• Outside it: above 400 units, above 3 kW, societies, C and I, and customers who want a better specification than a tender will carry.
• Do not promise scheme dates or vendor allocations you cannot source.

Sources

• Delhi Makes a Big Solar Promise, Yet Again: What The Data Actually Says, Saur Energy
• Delhi government clears free rooftop solar scheme for households, Renewable Watch
• Delhi Solar Portal, Government of NCT of Delhi
• Delhi Launches Zero-Cost Rooftop Solar Scheme to Power Homes and Cut Bills, EQ Magazine