Raghav Wadhwa@raghavwadhwa
Spent the last few hours on this question:
Why did BHEL, L&T, and Thermax all license supercritical boiler technology from foreign players (Alstom/Mitsubishi/B&W) when these are companies that build heavy machinery for a living?
And if the technology barrier is that high, why are EBIT margins stuck at 7-9%?
Worth answering together, because India is now adding 80 GW of fresh coal capacity by 2031-32. ₹6.67 lakh crore of capex. BHEL alone has contracted 68 supercritical boilers and 63 turbines, its biggest order book in over a decade. The "dead" sector is being rebuilt at scale.
So who actually makes money here?
Start with the licensing puzzle.
The IP is not in "how to build a boiler." Indian companies can weld, fabricate, and machine heavy equipment all day.
The real IP is decades of failure data.
A subcritical boiler runs at ~200 bar, 540°C. Standard steel works fine. A supercritical one runs at 250-270 bar, 600-620°C. At these temperatures, steel suffers "creep." Tube walls thin, micro-cracks propagate, and ruptures occur after 50,000-100,000 operating hours. You need Grade 91/92 chromium-moly-vanadium steels, Inconel, and austenitic stainless.
The part nobody talks about: B&W, Alstom, Mitsubishi each hold proprietary creep-life databases built over 40-60 years of operating supercritical plants globally. At what thickness, what temperature, what pressure, how long until failure. You cannot construct this from a textbook. You operate hundreds of units for decades or you do not have it.
Even China had to license all of it. Shanghai Electric from Siemens. Harbin from Mitsubishi. Dongfang from Hitachi.
So the moat is real. Then why are margins thin?
🔹 Powerful buyers. NTPC alone controls ~70 GW. Three suppliers exist, but NTPC runs competitive bidding and plays them against each other. Oligopsony cancels oligopoly.
🔹 Lump-sum turnkey contracts. Fixed price today, executed over 36-48 months. Steel and alloy prices move 15-30%, forex shifts, scope creeps. All risk on the maker. Exactly why Thermax flagged "project cost overruns" repeatedly in FY24-25 concalls.
🔹 IP is in design, not manufacturing. 60-65% of bill of materials is commodity steel and alloy. You're not selling software at 80% gross margin. You're selling 12,000 tonnes of fabricated steel with IP embedded. Material eats the margin.
🔹 BHEL sets the floor. As a PSU, it bids at 5-6% EBIT to satisfy capacity utilization and employment goals. Private players cannot price above it.
🔹 Working capital carry. 36-48 month execution, customer advances cover 10-20%. Thermax's borrowings went from ₹368 Cr to ₹1,718 Cr partly for this.
So where is the actual money?
Not in the boiler. In what comes after.
🔹 New boiler EPC: 7-9% margin (the razor)
🔹 Spare parts: 25-35% (only OEM has exact-fit specs)
🔹 Annual maintenance: 20-30% (locked in for 30-year plant life)
🔹 Retrofits and upgrades: 15-25% (forced by emission norms)
🔹 Performance optimization: 20-30% (pure IP)
Every supercritical boiler installed today is a 30-year annuity. The installed base is still young. Aftermarket revenue has not reached scale yet.
The number to track on every concall is aftermarket/services revenue from the installed base, and its YoY growth. When this moves from ₹50-100 Cr to ₹200-300 Cr, that is the margin inflection the market has not priced.
Sell the razor at 7-9%. Service it for 30 years at 20-30%. That is the actual model.
For educational purposes only. Not a buy/sell recommendation.
