Captive vs third-party PPA is the single most consequential financial decision an Indian commercial and industrial (C&I) buyer makes before going solar through open access. Both routes move your factory off the DISCOM grid tariff of ₹8-10 per kWh. The difference is who owns the plant, who takes the risk, and who keeps the bigger share of the savings. In 2026, a captive or group captive structure typically lands power at ₹4-5/kWh after the Cross Subsidy Surcharge (CSS) waiver, while a third-party Power Purchase Agreement (PPA) lands at ₹5.5-7/kWh after the full charge stack. That ₹1.5-2/kWh gap, compounded over 15-25 years, is worth ₹8-16 crore on a mid-size factory load.
This guide puts the two structures side by side on the numbers a CFO actually signs off: equity outlay, effective landed tariff, savings per unit over 15 and 25 years, tax treatment, exit terms, and risk. Every figure is benchmarked against our sibling plant-cost guides and current state tariff orders, so you can test any developer pitch against a defensible reference.
Direct answer. Captive solar needs ≥26% equity (about ₹1.9-2.1 crore per 2 MW of offtake) but waives Cross Subsidy Surcharge under Electricity Rules 2005, landing at ₹4-5/kWh versus ₹8-10/kWh grid rates. A third-party PPA needs zero equity but pays full open access charges, landing at ₹5.5-7/kWh. On a 30 lakh kWh annual load, captive saves roughly ₹27.6 crore over 15 years against ₹19.3 crore for a third-party PPA, per Heaven Green Energy 2026 modelling.
If you want the setup mechanics of the captive structure itself, our group captive setup guide covers the SPV formation path in detail. This post stays on the ROI comparison.
What Is the Difference Between Captive Solar and a Third-Party PPA?
The difference is ownership. In a captive structure, you (or your consortium) own at least 26% of the generating plant and consume at least 51% of its output, which makes you a part-owner buying power from yourself. In a third-party PPA, a developer owns 100% of the plant and you are simply a customer, buying units at a contracted tariff for 15-25 years.
Under the Ministry of Power’s Electricity Rules 2005, Rule 3, that ownership threshold changes your regulatory treatment. Captive consumers are exempt from Cross Subsidy Surcharge, the ₹1.5-2.5/kWh penalty every State Electricity Regulatory Commission (SERC) applies when a paying industrial customer leaves the DISCOM’s books. Third-party PPA consumers pay that surcharge in full, plus additional surcharge, wheeling, transmission, and banking charges.
Here is the structural snapshot:
| Dimension | Captive / group captive | Third-party PPA |
|---|---|---|
| Plant ownership | You hold ≥26% equity | Developer holds 100% |
| Upfront outlay | 26-30% of project cost | ₹0 |
| CSS (₹1.5-2.5/kWh) | Waived | Payable in full |
| Contract | PPA with own SPV + shareholder agreement | PPA with developer |
| Compliance | Annual SERC/CEA filings | None (developer files) |
| Typical landed tariff | ₹4-5/kWh | ₹5.5-7/kWh |
The word “captive” covers two variants. A single captive plant is 100% owned by one consumer, common for steel, cement, and large chemical units with 10 MW+ loads. A group captive plant pools several consumers into a Special Purpose Vehicle (SPV), each holding equity in proportion to offtake, which is how most textile, ceramic, and engineering clusters in Gujarat and Tamil Nadu access the structure. The third-party PPA, sometimes called the OPEX or RESCO model, is documented in our guide on how solar PPAs and RESCOs work in India.
One nuance worth stating early: all three structures deliver the same electrons through the same grid. The panels, inverters, and generation per MW are identical. Everything you are comparing is financial engineering: who funds the asset, who absorbs which surcharge, and who keeps the tax shield.
How the Landed Tariff Math Actually Works
The number on a developer’s proposal, the “PPA tariff”, is only the first line of your real cost. Heaven Green Energy runs every open access comparison through what we call the Landed-Cost Waterfall: a six-line sequence that walks from the headline tariff down to the true ₹/kWh you pay at your meter. A deal only makes sense if the waterfall’s last line beats your DISCOM tariff by a margin that survives a bad year.
The six lines, with 2026 benchmark ranges:
- PPA tariff - the contracted energy price, ₹3-4.5/kWh for 15-25 year terms in good irradiance states.
- Wheeling charge - paid to the DISCOM for using its wires, ₹0.2-1.1/kWh depending on state and voltage.
- Cross Subsidy Surcharge - ₹0 for captive, ₹0.5-2.5/kWh for third-party, per the SERC tariff order.
- Additional surcharge - ₹0.5-1.2/kWh in states that levy it on open access buyers.
- Transmission and SLDC charges - ₹0.3-0.6/kWh combined.
- Banking and deviation settlement - ₹0.2-0.5/kWh amortised, if you bank surplus generation.
Worked through the waterfall, a Gujarat third-party PPA at a ₹3.8/kWh headline becomes: ₹3.8 + ₹0.55 wheeling + ₹1.65 CSS + ₹0.55 additional surcharge + ₹0.40 transmission + ₹0.35 banking = ₹7.30/kWh landed, barely ₹0.75 below the ₹8.05 DISCOM tariff in the FY26 GERC order. The same ₹3.8/kWh under captive drops CSS to zero and trims wheeling, landing near ₹5.10/kWh, a ₹2.95/kWh saving. Same plant, same tariff, ₹1.55/kWh apart, purely because of who holds the equity.
⚠️ Watch out
CFOs on X routinely report being quoted a "₹3.5 solar tariff" that lands above ₹7 after charges. Always demand the full waterfall in writing before comparing any PPA against captive. A headline tariff without the charge stack is a marketing number, not a financial one.
Wheeling charges and banking deserve their own look in any state-specific model. Wheeling drops with connection voltage, and banking charges of 10-15% in kind silently lift the effective cost of any structure that relies on storing daytime surplus with the DISCOM. For a deeper engineering view of the open access charge stack, the Heaven Designs open access solar guide walks through a five-component cost framework used in feasibility studies.
Equity Outlay and Effective Tariff: The Ownership Premium
Captive’s lower tariff is not free money. You pre-pay for it with equity, and the honest comparison nets that outlay against the extra savings.
Take a factory contracting 2 MW of solar offtake, about 30 lakh kWh per year. The reference plant costs ₹7-7.6 crore all-in per our 2 MW solar plant cost and ROI guide. Under group captive at a 26-28% member stake, your equity cheque is ₹1.9-2.1 crore. Under a third-party PPA, your cheque is zero.
What does that ₹2 crore buy? About ₹1.5/kWh of extra savings for 25 years, plus pro-rata tax depreciation, plus residual asset value. On 30 lakh units a year, ₹1.5/kWh is ₹45 lakh of additional annual saving in Year 1, growing as grid tariffs escalate. The equity recovers itself in roughly 3.5-4.5 years of the incremental saving alone, and everything after that is the ownership premium.
| Metric (2 MW offtake, 30 lakh kWh/yr) | Captive / group captive | Third-party PPA |
|---|---|---|
| Equity outlay | ₹1.9-2.1 cr | ₹0 |
| Year-1 landed tariff | ₹4.5-5.0/kWh | ₹6.0-6.5/kWh |
| Year-1 saving vs ₹8 grid | ₹3.0/kWh | ₹1.5/kWh |
| Year-1 cash saving | ₹90 lakh | ₹45 lakh |
| Incremental payback on equity | 3.5-4.5 yrs | n/a |
| Tax depreciation (AD) | Pro-rata to equity | ₹0 to consumer |
| Balance sheet exposure | Equity + SPV covenants | Off balance sheet |
The financing stack behind the plant matters too. Captive SPVs raise 55-70% senior debt from lenders like the Indian Renewable Energy Development Agency (IREDA) at 9.5-11% for 10-13 year tenures, the cheapest rupee debt available to a solar asset. A third-party developer borrows at the same rates but prices its own equity return into your PPA tariff. In other words, you pay the developer’s cost of capital either way; in captive you pay it transparently, in a PPA it is embedded in the per-unit price.
There is a real counter-argument. Equity parked in an SPV is illiquid and earns its return only through power savings. If your core business can reinvest ₹2 crore at 20%+ returns in machinery, inventory, or market expansion, the third-party PPA’s “rent the asset” logic is defensible even at the higher tariff. The OPEX vs CAPEX solar comparison frames exactly this trade for rooftop, and the same hurdle-rate logic applies to open access equity.
25-Year Savings: Worked Example for a 30 Lakh Unit Factory
Here is the full horizon comparison. Assumptions: 30 lakh kWh per year of solar offtake; DISCOM tariff ₹8/kWh escalating 5% a year (the 10-year average across major industrial states); captive landed cost ₹5/kWh escalating 1% (O&M drift only); third-party PPA landed cost ₹6.5/kWh escalating 1.5% (tariff escalator plus charge drift). These are deliberately conservative; Mercom India C&I reporting through 2025 shows several states where grid escalation has run hotter.
| Horizon | Cumulative saving, captive | Cumulative saving, third-party PPA | Captive advantage |
|---|---|---|---|
| Year 1 | ₹0.9 cr | ₹0.5 cr | ₹0.4 cr |
| 5 years | ₹5.6 cr | ₹3.2 cr | ₹2.4 cr |
| 10 years | ₹14.5 cr | ₹9.3 cr | ₹5.2 cr |
| 15 years | ₹27.6 cr | ₹19.3 cr | ₹8.3 cr |
| 25 years | ₹72.2 cr | ₹55.9 cr | ₹16.3 cr |
Net of the ₹2 crore equity outlay, captive still leads by about ₹6.3 crore at Year 15 and ₹14.3 crore at Year 25 in this scenario. The crossover never happens: captive is ahead from Year 1 on a cumulative cash basis once you treat the equity as deployed capital rather than spent capital, because the SPV shares retain value and can be sold or bought back at exit.
Verdict. Captive wins the lifetime-value race in almost every scenario where annual consumption exceeds 30 lakh kWh and demand is stable. The third-party PPA wins on speed, simplicity, and capital preservation. The mistake is choosing on tariff alone; choose on the 15-year cumulative number after the equity cheque.
Get a free landed-cost comparison. Send us your last 12 HT bills and we return a captive-vs-PPA waterfall for your DISCOM within 2 weeks, with sensitivity on CSS movement and grid escalation. Request your comparison →
Two sensitivities can compress the gap. First, grid escalation: at 3% instead of 5%, the 15-year totals fall to about ₹19.5 crore (captive) and ₹14.5 crore (PPA), but the gap between them barely moves because both structures ride the same grid benchmark. Second, CSS revision: SERCs re-set surcharge every tariff order, and a state that cuts CSS by ₹0.5/kWh improves third-party economics while leaving captive untouched. Stress-test your model at ±20% on CSS before signing anything. You can run your own base numbers through our solar ROI calculation guide and the solar calculator.
Tax Treatment: Who Keeps the Accelerated Depreciation?
Tax treatment quietly adds ₹15-25 lakh per MW of value to the captive side of the ledger, and most PPA proposals never mention it.
Solar plant and machinery qualifies for accelerated depreciation at 40% of the written-down value in Year 1 under the Income Tax Act, on top of normal depreciation. In a captive or group captive structure, that depreciation flows through to equity holders pro-rata to their shareholding. A member holding 28% of a ₹7.4 crore SPV can book depreciation on roughly ₹2.07 crore of asset value: ₹83 lakh of Year-1 depreciation, worth about ₹21 lakh in tax saved at a 25.17% corporate rate, assuming profits to absorb it. Our accelerated depreciation solar tax guide covers the half-year rule and the 40% vs 25% choice in detail.
In a third-party PPA, the developer owns the asset, so the developer claims the entire depreciation shield. That shield is real money, and it is part of why developers can offer a ₹3.5-4/kWh headline tariff profitably. From the consumer’s seat, the PPA payment is simply a deductible operating expense, much like the DISCOM bill it replaces. You get the clean deduction; you do not get the shield.
📘 Regulation note
SPV members claim accelerated depreciation on their own books in proportion to equity, per Income Tax Act Section 32 read with the shareholder pattern filed with the SERC. If your company cannot absorb the depreciation (loss position or MAT), the benefit carries forward, but its present value drops sharply.
The practical CFO rule: if your company is profitable and pays full-rate corporate tax, add the AD shield to the captive column at roughly 8-12% of the equity cheque in Year-1 tax value. If you are in a loss position, a new entity without taxable profits, or under Minimum Alternate Tax, discount it heavily, and the third-party PPA’s zero-equity simplicity gets relatively more attractive. This is one of the few places where the “right” structure genuinely flips based on your tax position rather than the tariff math.
GST treatment is similar in both routes: power itself sits outside GST, and the EPC contract inside a captive SPV carries the standard 8.9% composite GST on works, which the SPV absorbs into project cost. Neither structure gives the consumer a meaningful input tax credit on the power consumed.
Exit Terms, Lock-In, and Risk: Reading the Fine Print
Both structures lock you in for 15-25 years, but they trap you differently. Captive traps your equity; a third-party PPA traps your demand.
Captive exit mechanics. Your exit is a share transfer. The shareholder agreement governs who can buy (usually existing members get pre-emption rights), at what valuation (book value, fair value, or a formula), and what happens if your exit pushes the SPV below the 26% or 51% tests. A badly drafted shareholder agreement can leave your ₹2 crore stranded until the remaining members agree to buy you out. Insist on a defined buy-back formula and a drag-along clause before subscribing.
Third-party PPA exit mechanics. Your exit is contract termination, and the PPA’s termination schedule prices it. Standard 15-25 year PPAs include lock-in periods of 10-15 years during which early exit triggers a make-whole payment, commonly the net present value of the developer’s expected margin on your remaining offtake. Volume risk sits with you too: most PPAs carry minimum-offtake or take-or-pay clauses at 80-90% of contracted energy, so a demand contraction does not shrink your bill proportionally.
- ✓ ₹1.5-2/kWh lower landed tariff via CSS waiver
- ✓ Pro-rata accelerated depreciation on your books
- ✓ Equity retains value; upside if grid tariffs rise
- ✓ You control the asset, the EPC standard, and the O&M
- ✓ Surplus generation is yours to bank or sell
- ✗ ₹1.9-2.1 cr equity locked per 2 MW offtake
- ✗ Retrospective CSS if the 51% consumption test fails in any year
- ✗ 12-18 month setup window before the first unit flows
- ✗ Annual SERC/CEA compliance filings and audits
- ✗ Exit needs a share buyer; illiquid in a weak SPV
The third-party PPA’s own risk list is shorter but sharper: developer credit risk (a bankrupt developer’s plant can be sold under you, with the PPA assigned to a new owner), generation shortfall risk (mitigated by availability guarantees, usually 95-98% of P50), and regulatory drift on the charges you still pay. The single largest third-party risk is one that almost never appears in the term sheet: a state government restricting or repricing banking of electricity mid-contract, which C&I buyers on X flag as the most common way a “₹3.80 PPA win” turns into a ₹4.50 regret. Insist on a change-in-law clause that shares charge increases between you and the developer.
Which Wins by Load Size and State?
Load size decides the structure’s floor; the state’s charge stack decides its ceiling. Below roughly 15 lakh kWh of annual consumption, the captive equity, compliance overhead, and setup window rarely pay back; take the third-party PPA or a rooftop OPEX deal instead. Above 30-40 lakh kWh in a high-CSS state, captive is the dominant choice. Between those bounds, the state’s surcharge regime tips the balance.
| State | CSS (third-party pays) | Third-party landed | Captive landed | Better pick for 30+ lakh kWh |
|---|---|---|---|---|
| Gujarat | ₹1.65/kWh | ₹6.0-7.0/kWh | ₹4.5-5.2/kWh | Captive; deepest market, clean GERC framework |
| Maharashtra | ₹1.8-2.2/kWh | ₹6.5-7.5/kWh | ₹4.8-5.5/kWh | Captive; highest CSS savings in India |
| Karnataka | ₹1.2-1.8/kWh | ₹5.8-6.8/kWh | ₹4.6-5.3/kWh | Captive; strong IT and industrial demand |
| Tamil Nadu | ₹1.5-2.5/kWh | ₹6.2-7.5/kWh | ₹4.8-5.6/kWh | Captive; textile clusters anchor well |
| Madhya Pradesh | ₹1.49/kWh (+₹1.18 AS) | ₹5.8-6.8/kWh | ₹4.2-4.8/kWh | Captive; lowest landed tariffs, thin member pipeline |
| Rajasthan | ₹0.9-1.5/kWh | ₹5.2-6.2/kWh | ₹4.0-4.8/kWh | Either; low CSS narrows the captive edge |
Charge figures are FY26-FY27 indicative from the respective SERC tariff orders (GERC, MERC, KERC, TNERC, MPERC, RERC) and are revised every March-April. Always re-run the waterfall against the current order before committing equity or signing a PPA. The Central Electricity Authority publishes national open access data, and Bridge to India tracks C&I PPA pricing trends quarterly.
Three load-size rules our deal team uses with C&I boards:
- Under 15 lakh kWh/year (about ₹1 crore annual bill): third-party PPA or rooftop OPEX. Captive overhead will eat the tariff advantage.
- 15-40 lakh kWh/year: state decides. High-CSS states (Maharashtra, Karnataka, Tamil Nadu) favour group captive; low-CSS states (Rajasthan) favour the PPA.
- 40+ lakh kWh/year (about ₹3.5+ crore annual bill): captive, almost always. The CSS saving alone typically clears ₹60 lakh-1.2 crore per year, dwarfing compliance costs.
For larger single-site loads where full ownership is on the table, the 5 MW solar plant cost and ROI guide models the single-captive route, and our 1 MW solar plant cost guide scales the same math down for smaller factories.
How to Run the Captive vs PPA Comparison in 5 Steps
Whether you are an energy manager building the internal case or a CFO reviewing a developer proposal, this is the comparison sequence that produces a defensible answer.
- Pull 12 months of HT bills and fix your real consumption baseline: annual kWh, contract demand, load profile, and the landed grid tariff including electricity duty. Everything else hangs off this number.
- Get both quotes for the same MW of offtake, a group captive equity proposal and a third-party PPA term sheet, and demand the full landed-cost waterfall for each, line by line, against your state’s current SERC tariff order.
- Model 15 and 25 years with grid escalation at 3% and 5%, CSS movement at ±20%, and your solar cost escalation from the contract. Compare cumulative savings net of equity, not Year-1 tariffs.
- Add your tax position. If profitable, credit the captive column with the pro-rata AD shield; if loss-making or MAT, discount it and note how the ranking shifts.
- Score exit terms and risk. Read the shareholder agreement’s buy-back formula and the PPA’s termination and take-or-pay clauses. A ₹0.2/kWh tariff edge is not worth a strangled exit.
If both structures come within ₹0.5/kWh of each other on landed cost after this exercise, take the third-party PPA. Simplicity has real value, and a thin captive margin does not survive a bad monsoon year or a member dropping out of the SPV. If the captive edge is ₹1/kWh or more, the equity cheque earns its keep even under pessimistic assumptions, and the comparison becomes a financing question rather than an energy question.
How Heaven Green Energy Helps
Heaven Green Energy structures both sides of this comparison, which is why we can run it without a thumb on the scale. We build captive and group captive ground-mount parks across Gujarat, Maharashtra, Tamil Nadu, Karnataka, and Madhya Pradesh, and we also deliver OPEX and third-party PPA structures for buyers who should not be putting up equity. Most engagements start with your last 12 HT bills and end with a board-ready waterfall showing both structures over 25 years, with sensitivity tables on CSS, grid escalation, and plant performance ratio.
- Industrial Solar - captive, group captive, and blended structures for 1 MW+ HT loads.
- Commercial Solar - rooftop and open access blends for 100 kW-1 MW consumers.
- Solar EPC Services - turnkey ground-mount delivery with 25-year performance guarantees.
- Ground-Mount Solar Park - utility-scale park design, land, and evacuation handled end-to-end.
- Solar Calculator - a first-pass savings estimate before you talk to anyone.
For the inverter and power-electronics side of a captive plant, the Qbits Energy C&I solution covers the on-grid inverter range we specify on industrial projects. Whatever structure you choose, the first step is the same: an honest landed-cost waterfall for your DISCOM and load profile. Talk to our C&I team →
Frequently Asked Questions
Is captive solar better than a third-party PPA?
For most industrial consumers above 30-40 lakh kWh of annual consumption with stable demand, yes. Captive lands power at ₹4-5/kWh versus ₹5.5-7/kWh for a third-party PPA because it waives Cross Subsidy Surcharge, and the equity recovers in 3.5-4.5 years. Below 15 lakh kWh a year, or where demand is volatile, the third-party PPA wins on simplicity and zero capital lock-up.
How much equity do I need for captive solar?
At least 26% of the generating company’s equity under Electricity Rules 2005, sized in proportion to your offtake. For 2 MW of offtake from a ₹7-7.6 crore project, that is roughly ₹1.9-2.1 crore. In group captive SPVs, members typically hold 26-40% each depending on their share of consumption.
What is the real savings difference per unit between captive and third-party PPA?
About ₹1.5-2/kWh in Year 1, widening as grid tariffs escalate. Against an ₹8/kWh DISCOM tariff, captive saves roughly ₹3/kWh (landed ₹5) while a third-party PPA saves roughly ₹1.5/kWh (landed ₹6.5). Over 15 years on a 30 lakh kWh load, that gap compounds to about ₹8 crore of extra savings on the captive side.
Who gets the accelerated depreciation benefit in a third-party PPA?
The developer, because the developer owns the asset. The consumer’s PPA payments are a deductible operating expense, but the 40% Year-1 accelerated depreciation shield stays with the plant owner. In captive and group captive structures, the shield flows to equity holders pro-rata, worth roughly 8-12% of the equity cheque in Year-1 tax value for a profitable company.
What happens if my factory’s demand falls after I go captive?
Two risks bite. First, if the SPV’s collective consumption drops below 51% of generation in a financial year, the Cross Subsidy Surcharge becomes retrospectively payable for that year. Second, your PPA with the SPV will have minimum-offtake covenants. Well-structured SPVs size the plant at 90-95% of expected demand and include true-up mechanisms so a weak member’s share can be redirected to other members.
Can I exit a third-party solar PPA early?
Only by paying the termination schedule. Most 15-25 year PPAs include a 10-15 year lock-in with a make-whole payment equal to the present value of the developer’s remaining margin. Volume flexibility is limited too: take-or-pay clauses at 80-90% of contracted energy are standard. Negotiate termination step-downs and a change-in-law clause before signing, not after.
Which states favour captive over third-party PPA?
High-CSS states favour captive most: Maharashtra (₹1.8-2.2/kWh CSS), Tamil Nadu (₹1.5-2.5/kWh), Karnataka, and Gujarat (₹1.65/kWh). In low-CSS states like Rajasthan, the captive edge narrows to under ₹1/kWh and a third-party PPA is often the cleaner choice. Surcharges reset every tariff order, so re-check the current SERC numbers before deciding.
How long does each route take to deliver savings?
A third-party PPA on an existing or near-complete park can flow power within 3-6 months of signing, since only open access approvals and scheduling stand in the way. A captive or group captive project built from scratch takes 12-18 months from feasibility to commissioning, covering SPV formation, land, evacuation, and construction. If you need savings this financial year, the PPA wins on speed; if you can wait, captive wins on value.
