Introduction
A poorly modeled PPA doesn't just underperform—it leads to mispriced contracts, rejected financing, and projects that erode returns over a 15–25 year horizon. Most financial models built for Indian C&I and utility-scale PPAs fail at the same points: variable generation assumptions, long-term price risk, and state-level regulatory exposure.
Unlike standard project finance with predictable tariffs, PPA modeling must account for all three simultaneously—across decades of counterparty obligations.
This guide covers the specific modeling challenges of Indian PPAs: delivery structures, pricing mechanisms, Open Access charge volatility, risk factors, and the mistakes most likely to erode IRR. It's written for energy managers, CFOs, developers, and investors who need models that hold up under lender scrutiny and real-world operating conditions.
TLDR:
- PPA models must price merchant risk, hedge ratios, and offtake structure simultaneously—not rely on single guaranteed rates
- State-specific Open Access charges (transmission, wheeling, CSS) add layers of complexity that national averages can't capture
- Lenders including IREDA and PFC require a minimum 1.25 DSCR, with energy yield estimates restricted to P90 scenarios
- Real-world module degradation (1.6% annually for shaded modules) far outpaces manufacturer warranties (0.3%–0.65%)
- Banking provisions have tightened: Maharashtra limits drawals to same/lower ToD slots, while Andhra Pradesh mandates monthly settlements with 8% in-kind charges
Why PPA Financial Modeling Demands a Different Approach
Unlike standard project finance models with predictable tariffs, PPA models must simultaneously account for variable generation, long-term price risk, regulatory changes, and counterparty obligations across a 15–25 year horizon. That scope turns modeling into something far more demanding than a standard cash flow exercise — every assumption compounds over decades, and a misread on merchant exposure or state surcharges can break a deal's bankability.
India-Specific Complexity: Open Access Charges and State Fragmentation
The Indian market adds unique layers that standard models don't address. Open Access charges—transmission, wheeling, cross-subsidy surcharge (CSS), and additional surcharges (AS)—vary dramatically by state and change mid-contract. For instance, Gujarat recently reduced its AS to ₹0.76/kWh for April–September 2026, while Karnataka introduced a ₹25,000/MW facilitation fee — despite central Green Energy Open Access (GEOA) Rules 2022 aiming to streamline procurement.
State-wise DISCOM tariff variations and RPO compliance requirements create regulatory fragmentation that national averages simply can't capture. Models must be built on the specific regulatory realities of both host and offtaker states — not broad benchmarks.
The Shift from Feed-in Tariffs to Bilateral PPAs
The transition from centralized Feed-in Tariffs (FiTs) to competitive bilateral PPAs transfers price discovery and grid-integration risk directly onto developers and corporate offtakers. Financial models must now price:
- Merchant risk exposure: Percentage of output sold at market vs. contracted rates
- Hedge ratios: Balancing revenue certainty against upside potential
- Offtake structure: Pay-as-Produced vs. Fixed-Volume delivery obligations
A single guaranteed rate is no longer enough. Models that can't stress-test pricing scenarios, volume shortfalls, and market exposure levels won't survive lender scrutiny — let alone 20-year contract negotiations.
Essential Inputs for a Robust PPA Financial Model
Tariff and Price Assumptions
The PPA strike price (fixed or indexed) is the single most sensitive input in any financial model. Even a ₹0.10–0.20/kWh deviation compounded over 20 years can significantly alter project IRR and debt serviceability. While specific sensitivity varies by project, ICRA analyses indicate that net tariffs for Open Access players remain under severe pressure due to regulatory risks: the discount to grid tariffs offered by solar OA projects shrinks rapidly once CSS and additional surcharge are factored in.
Sourcing Reliable Landed Tariff Benchmarks:
Landed costs must include all charges at the delivery point:
- Transmission charges
- Wheeling charges
- Cross-subsidy surcharge (CSS)
- Additional surcharge (AS)
- Banking charges
- Applicable DISCOM levies
Opten Power's Real-Time DISCOM Intelligence tool provides standardized, updated landing prices across all states, making this input more accurate and less assumption-heavy than manual compilation from multiple state regulatory authorities.
Generation and Plant Performance Inputs
Core Generation Assumptions:
- P50/P90 energy yield estimates: P50 represents expected value (50% probability); P90 represents yield exceeded with 90% probability
- Capacity Utilization Factor (CUF): CEA's NEP 2024 benchmarks show Solar CUF of 22.21%–24.5% (Northern Region) and Wind CUF of 24.36%–27.16% (Western Region)
- Panel/turbine degradation rate: Annual performance decline
- Auxiliary consumption percentage: Typically 0.5%–1.5% of gross generation
Why P50 vs. P90 Matters:
Indian lenders like IREDA strictly use P90 estimates for base-case debt sizing and DSCR stress testing, while equity investors typically use P75 or P50 to gauge returns. This difference directly impacts how much debt a project can support.
Curtailment Assumptions:
Grid curtailment by DISCOMs or STUs can reduce actual generation below modeled output. Models should include a curtailment haircut based on state-level historical data, a step that is especially critical in high-penetration states like Rajasthan and Tamil Nadu.
Real-World Degradation vs. Warranties:
Manufacturer warranties typically assume 0.3%–0.65% annual degradation (First Solar warrants 0.3%; Trina warrants 0.4%–0.65%). Real-world performance tells a different story: an MNRE All India Survey found shaded modules degrade at 1.6% annually, nearly three times the 0.56% rate of unshaded modules. Financial models should apply site-specific penalties for soiling and shading rather than relying on warranty figures alone.
Project Cost and Financing Inputs
Accurate generation assumptions feed directly into cost and financing decisions. Getting these inputs wrong compounds across the entire model.
Critical Cost Inputs:
- EPC cost per MW: Average large-scale solar project costs fell 28.2% YoY in Q1 2024, driven by module price deflation
- O&M escalation rate: Typically 3%–5% annually; must be stress-tested against inflation scenarios
- Insurance premiums: Covers equipment, business interruption, and liability
- Land lease costs: Annual payments with escalation clauses that affect long-term cash flows
- Working capital requirements: Cash reserves sized to cover 2–3 months of operating expenses
Debt-Equity Structure Decisions:
Typical Indian renewable project finance uses 70:30 or 75:25 debt-equity ratios. Current lending rates for 2024:
- IREDA: 8.90% (AAA to AA grades) to 9.65% (BBB+ to BBB grades)
- PFC: 8.95% to 9.70% for Non-Conventional Energy (Renewable Wind/Solar)
DSCR Requirements:
IREDA's latest financing norms mandate that average DSCR must not be less than 1.25 for new solar and wind clients. This stricter threshold requires conservative generation assumptions and robust revenue projections to ensure debt serviceability.
Modeling PPA Delivery Structures: Pay-as-Produced vs. Fixed-Volume
Pay-as-Produced PPA
In a Pay-as-Produced structure, the offtaker purchases an agreed percentage (ρ) of actual generation at the PPA price, while the remaining share is sold at prevailing market prices.
Revenue Formula:
R = ρ × P_PPA + (1 – ρ) × P_Market
Cap, Floor, and Collar Variations:
When a minimum quantity floor is introduced, the generator must buy power from the exchange (such as IEX) to meet the shortfall commitment — a direct downside revenue risk. The model must capture:
- Shortfall trigger: When actual generation falls below the floor quantity
- Purchase cost exposure: Market price at time of shortfall, sourced from exchange price volatility data
- Net revenue impact: PPA revenue minus shortfall purchase cost
Fixed-Volume (Profile-Based) PPA
In a Fixed-Volume PPA, the generator commits to delivering a defined quantity (monthly baseload, annual baseload, or pre-defined solar profile). Shortfalls must be covered by purchasing from the exchange at market price.
Three Revenue Scenarios:
- Actual < PPA: Generator buys shortfall at market price (cost exposure)
- Actual > PPA: Generator sells surplus at market price (upside opportunity)
- Actual = PPA: Perfect match (no market exposure)
Profile Overlap Concept:
Even if monthly totals balance out, hourly mismatches between generation and delivery obligation create buy-sell spread costs. Rather than building an hourly model, use the profile overlap percentage (μ) to estimate this cost. When overlap falls below 70%–80%, the spread cost becomes financially material and must be modeled explicitly.
Choosing the Right Structure for Your Model
Decision Framework:
- Pay-as-Produced: Suits generators wanting to eliminate volume risk; offtaker assumes generation variability
- Fixed-Volume: Suits offtakers wanting energy certainty; transfers profile risk to generator
DSCR compliance and debt sizing shift materially between structures. Under Pay-as-Produced, revenue tracks actual generation — lenders face lower volume risk, which directly supports higher debt ratios. Fixed-Volume structures require explicit stress-testing of shortfall purchase costs before a lender will approve the debt quantum.
PPA Pricing Mechanisms and How to Model Them
Three Primary Pricing Structures
- Fully fixed price: Simplest to model, most bankable; Indian lenders and DFIs strongly prefer this structure
- Price with annual escalation: CPI-linked or fixed % escalator; requires inflation assumptions
- Variable/indexed pricing: Linked to exchange or APPC; introduces significant revenue uncertainty
Bankability Consideration:
IREDA and international DFIs typically require a minimum DSCR of 1.25x–1.35x over the loan tenor. Variable or merchant-exposed structures make this covenant difficult to meet, which is why fixed or escalator-based pricing dominates project-financed deals in India.
Cannibalisation Effect and Capture Price
As renewable penetration grows, solar and wind generation depresses spot prices exactly when these plants produce most. IEX Day-Ahead Market data frequently shows unconstrained prices dropping to near ₹1.00/kWh during solar hours (11:00–14:00), while evening peak prices surge to ₹10.00/kWh.
Capture Price Modeling:
Rather than using baseload average prices, apply a technology-specific capture price:
P_C = δ × P_Market, where δ is the capture price discount factor
In high-penetration states like Rajasthan and Gujarat, solar capture prices run 15%–25% below baseload averages during peak generation hours — a gap that compounds as installed capacity grows.
This capture price discount directly shapes how much merchant exposure your project can absorb — which brings the hedging ratio into focus.
Hedging Ratio Decision
The percentage of output sold under PPA vs. merchant price exposure directly determines revenue certainty and debt capacity.
Scenario Comparison:
- 70% hedged: 70% of output at fixed PPA price, 30% at market price
- 100% hedged: All output at fixed PPA price, no market exposure
A 70% hedged scenario delivers higher expected IRR if market prices exceed PPA rates, but results in lower minimum DSCR and reduced debt capacity. Model both scenarios to understand the trade-off between upside potential and financing risk.
Banking and Carry-Forward Provisions
Many state regulations allow surplus energy to be banked with the DISCOM and consumed in a later period. However, banking provisions have tightened significantly:
| State | Banking Period | Banking Charges | Key Restrictions |
|---|---|---|---|
| Maharashtra | Monthly | 8% in-kind | Drawal restricted to same or lower tariff ToD slots |
| Andhra Pradesh | Monthly | 8% in-kind | Unutilized energy lapses at end of billing cycle |
| Gujarat | Daily (Solar) / Monthly (Wind) | ₹1.50/kWh | High charges penalize generation-load mismatch |
| Rajasthan | Annual | N/A | No banking for third-party OA |
Modeling Banking Costs:
Include banking charges and time-value loss on banked units to avoid overestimating net revenue. For monthly banking with 8% in-kind charges, the 8% in-kind charge reduces effective revenue on all banked energy — with the opportunity cost of delayed consumption compounding that loss further.
Key Risks to Factor into Your PPA Financial Model
Counterparty/Offtaker Credit Risk
Model the probability and financial impact of payment delays or default by the offtaker. This is especially critical when the offtaker is a DISCOM with historically weak financials.
PFC's 13th Annual Integrated Rating report shows national average AT&C losses at 16.12%. As of March 2026, Andhra Pradesh DISCOMs had outstanding dues of ₹4,401 Crore and Uttar Pradesh owed ₹9,235 Crore. These figures illustrate why payment risk modeling can't be an afterthought.
Build these protections into your model:
- Include provision for payment delay days (typically 60–90 days)
- Model a Debt Service Reserve Account (DSRA) equivalent
- Stress-test IRR impact of 6-month payment delays
Regulatory and Policy Risk
Open Access charges in India have changed significantly across states. Models must include a scenario where additional surcharges or wheeling restrictions are imposed mid-contract.
The table below shows how quickly equity IRR erodes when landed charges shift:
| Open Access Charge Increase | Estimated IRR Impact | Notes |
|---|---|---|
| +₹0.50/kWh | -2% to -4% equity IRR | Highly project-dependent; greater impact on thin-margin C&I projects |
| +₹1.00/kWh | -4% to -7% equity IRR | May breach lender DSCR covenants |
Stress-test every model against a minimum ₹0.50/kWh adverse swing in landed charges — project structure will determine exact sensitivity, but this threshold catches most debt serviceability failures before they reach lenders.
Volume/Generation Risk
P90 generation estimates are more conservative than P50 and are typically used by lenders to stress-test DSCR. Set up a downside scenario using:
- P90 output (typically 10%–15% below P50)
- Higher degradation rates (1.5%–2.0% annually for sites with shading/soiling risk)
- Curtailment assumptions (2%–5% annual generation loss in high-penetration states)
Run each of these inputs through your model and verify that minimum DSCR stays above 1.25 — that's the floor most lenders require before they'll sign off on debt terms.
Common PPA Financial Modeling Mistakes to Avoid
Mistake: Using Baseload Average Prices Instead of Capture Prices
Using a single average electricity price (APPC or exchange average) instead of technology-specific capture prices leads to revenue overestimation. Solar-heavy states with afternoon price suppression will consistently underperform a baseload-price-based model.
To correct this:
- Apply a solar capture price discount of 15%–25% below baseload averages during peak generation hours
- Use historical IEX hourly data from the project's region to calculate technology-specific capture rates
Mistake: Assuming Static Open Access Charges Throughout PPA Tenor
Ignoring Open Access charge escalation and assuming static landed costs throughout the 15–25 year PPA tenor severely underestimates long-term costs. In India, cross-subsidy surcharges and wheeling charges have been revised upward multiple times.
To correct this:
- Model a stressed scenario with annual charge escalation of 3%–5% to account for regulatory risk
- Use platforms like Opten Power for updated state-wise DISCOM landing prices to benchmark assumptions against current regulatory changes
Mistake: Failing to Run Scenario and Sensitivity Analysis
Finalizing PPA terms without robust scenario analysis leaves you blind to which variable poses the greatest risk. Output a sensitivity matrix showing IRR impact across key variables:
- PPA tariff (±₹0.25/kWh)
- CUF (±2 percentage points)
- O&M escalation (±1% annually)
- Open Access charges (+₹0.50/kWh, +₹1.00/kWh)
Identify the variable with the steepest IRR slope and negotiate contract terms accordingly—such as caps on Open Access charge pass-through or minimum generation guarantees.
Frequently Asked Questions
What is a PPA in finance?
A PPA (Power Purchase Agreement) is a long-term bilateral contract between an electricity generator and a buyer (offtaker), fixing the price, volume, and delivery terms for energy. It's used in project finance to create predictable cash flows that support debt financing.
What is the difference between a physical PPA and a virtual PPA?
A physical PPA involves actual electricity delivery at an agreed price, while a virtual (financial) PPA is a contract for differences where the generator sells power at market price and the two parties settle the difference between market price and agreed PPA price financially. CERC's December 2025 VPPA Guidelines officially recognize VPPAs as Non-Transferable Specific Delivery (NTSD) contracts.
How is the PPA price calculated?
The PPA price is derived from the levelized cost of energy (LCOE) plus a margin, adjusted for debt service requirements, developer return expectations, and applicable Open Access or DISCOM charges at the delivery point. For FY 2024-25, Gujarat's APPC was determined at ₹4.92/kWh.
What is a good IRR for a renewable energy PPA project in India?
Equity IRR expectations vary by project type and risk profile. While highly competitive utility-scale auctions have historically compressed returns to single digits, developers currently target mid-teen post-tax equity IRRs (approximately 15%) for renewable projects, through optimized debt structures and technology choices.
What is a Pay-as-Produced PPA vs. a Fixed-Volume PPA?
Pay-as-Produced means the offtaker buys a share of whatever is generated, while Fixed-Volume requires the generator to deliver a set quantity, with any shortfall purchased at spot market price. Fixed-Volume is riskier for generators because it transfers profile and volume risk to them.
What is the typical tenure of a PPA in India?
Corporate and utility-scale PPAs in India typically range from 10 to 25 years. Most C&I Open Access PPAs run 10–15 years, while government/utility PPAs with DISCOMs are commonly structured for 25 years from the commercial operations date.
