SECI Tariff Math — How Engineering Inputs Move Bids by 4 Paise/kWh

In a SECI solar auction, the winning tariff and the third-place tariff are often separated by 3–6 paise/kWh. At 500 MW, that margin represents ₹20–₹40 crores over 25 years. The difference is rarely in the EPC cost or the module price — both are published and competitively tight. The difference is almost always in the engineering inputs that determine the annual energy yield estimate, because yield drives everything in the tariff model: LCOE, debt service coverage, IRR, and ultimately the bid price. This article shows exactly which engineering inputs move the bid, by how much, and in which direction.

Direct answer. SECI tariff bids are driven by the annual energy yield estimate, which translates directly into the LCOE calculation and bid price. Four engineering inputs — tracker type and backtracking algorithm, ground coverage ratio, soiling loss assumptions, and degradation curve selection — can individually shift the yield estimate by 0.5–2.0%, and collectively move the bid by 3–5 paise/kWh at 500 MW scale. The Heaven Designs 0.4-Paise Bidding Margin Framework isolates each input and quantifies its tariff impact.

This article targets Suresh — an Indian utility-scale developer bidding SECI auctions — and the independent engineers and financial advisors who validate these bids. All figures use ₹ and MW/MWh units.

How the Tariff Model Works: Engineering Inputs → LCOE → Bid Price

The connection between engineering inputs and bid price runs through a well-defined chain:

1. Engineering simulation (PVsyst or equivalent) produces annual energy yield in MWh/MW
2. Yield → Capacity Utilization Factor (CUF) → project revenue at assumed tariff
3. Revenue → Debt Service Coverage Ratio (DSCR) → maximum supportable debt quantum
4. Debt + equity cost → LCOE → minimum viable tariff
5. Developer adds margin → bid tariff

Any engineering input that increases yield by 1% allows the developer to bid 1% lower on the tariff (holding all other variables constant) while maintaining the same IRR. At ₹2.80/kWh for a 500 MW project generating 950 MU/year, 1% yield improvement = 9.5 MU = ₹26.6 crores over 25 years at nominal tariff. This is the mathematical basis for why engineering precision wins auctions.

According to SECI’s published tariff discovery results for FY2025, the average winning tariff for large-scale solar auctions (500 MW+) was ₹2.58–₹2.74/kWh, with a spread of 3–7 paise between the L1 and L3 bidders.

Engineering Input 1 — Tracker Type and Backtracking Algorithm (0–1.8% Yield Impact)

Single-axis trackers add 15–25% to energy yield compared to fixed-tilt systems in Indian latitudes. But the tracker yield gain is not uniform — it depends on the backtracking algorithm used in the PVsyst simulation, and the difference between a basic and sophisticated backtracking model is 0.8–1.8% of annual yield.

Standard backtracking (used in most basic simulations) prevents row-to-row shading by rotating the tracker to a position where the shadow edge falls exactly on the adjacent row’s back. This eliminates shade losses but reduces the tracker’s ability to track the sun at low angles.

Optimized backtracking (available in PVsyst 7.4+ and advanced tracker firmware) uses a slope-corrected algorithm that accounts for terrain undulation. On flat terrain, the difference from standard backtracking is marginal. On terrain with 1–3% cross-slope (common in Rajasthan and Gujarat sites), optimized backtracking adds 0.5–1.2% to annual yield compared to standard.

Bifacial gain with tracker adds a further 4–8% to the rear-side output on a tracker system versus a fixed-tilt bifacial system. The tracker increases the albedo-weighted irradiance on the rear face by changing the angle to reflected ground irradiance throughout the day.

Tracker Configuration	Annual Yield Index	vs Fixed Tilt	Simulation Complexity
Fixed tilt, monofacial	100%	Baseline	Low
Fixed tilt, bifacial	103–105%	+3–5%	Low
SAT, standard backtracking, monofacial	115–120%	+15–20%	Medium
SAT, optimized backtracking, monofacial	116–122%	+16–22%	High
SAT, optimized backtracking, bifacial	120–128%	+20–28%	High

For a 500 MW project, the difference between a basic SAT simulation and an optimized bifacial SAT simulation represents 1.5–2.0% additional yield — approximately 14–19 MU/year at 500 MW.

The bankable PVsyst reports guide covers the simulation settings required for lender-accepted tracker yield models.

Engineering Input 2 — Ground Coverage Ratio Optimization (0.5–1.5% Yield Impact)

Ground Coverage Ratio (GCR) is the ratio of the collector area to the total ground area. In a tracker system, GCR determines the row spacing — and row spacing determines the degree of inter-row shading that must be managed by the backtracking algorithm.

Lower GCR means wider row spacing, less shading, but more land required and lower MW/hectare packing density. Higher GCR means denser packing, more land efficiency, but more shading losses. The optimal GCR for an Indian utility site is not a fixed number — it is a function of latitude, terrain slope, and the tracker’s backtracking performance.

The yield impact of GCR optimization comes from finding the point where the energy loss from increased shading at higher GCR is exactly offset by the land-cost savings and packing density benefits. In SECI auctions where land cost is a significant CAPEX input (Rajasthan: ₹50–₹80 lakhs/hectare in lease), a 10% reduction in land area through GCR optimization saves ₹5–₹8 crores on a 500 MW project — a direct CAPEX reduction that drops the LCOE and enables a lower bid.

The LCOE glossary page explains how land cost feeds into the LCOE formula.

Engineering Input 3 — Soiling Loss Assumptions (0.5–2.0% Yield Impact)

Soiling loss is the percentage of annual energy lost due to dust, bird droppings, and debris on module surfaces. In Indian conditions, soiling loss ranges from 1.5% (coastal sites with regular monsoon washing) to 4.5% (Rajasthan desert sites without a cleaning regimen).

The soiling loss input in PVsyst is one of the most contested variables in lender reviews. Developers want to use conservative (low) soiling assumptions to show higher yields and justify lower bids. Lenders’ independent engineers typically push for higher soiling assumptions that match site-specific data.

Site Type	Typical Soiling Loss (No Cleaning)	With Monthly Cleaning	Cleaning Saves
Rajasthan desert	4.0–5.5%/yr	1.8–2.5%/yr	2.0–3.0% yield
Gujarat semi-arid	2.8–3.8%/yr	1.5–2.0%/yr	1.3–1.8% yield
Maharashtra plateau	2.0–3.0%/yr	1.2–1.8%/yr	0.8–1.2% yield
Tamil Nadu coastal	1.5–2.5%/yr	1.0–1.5%/yr	0.5–1.0% yield

The tariff impact of the cleaning regimen decision is therefore both a soiling assumption change and a CAPEX/OPEX trade-off: monthly robot cleaning for a 500 MW Rajasthan project costs ₹2.5–₹4.0 crores/year in OPEX but recovers 2–3% of annual yield = 19–28.5 MU/year = ₹5.3–₹8.0 crores/year in additional revenue at ₹2.80/kWh. The cleaning regimen pays back in less than one year — and allows the developer to use a lower soiling loss figure in the PVsyst simulation, improving the bankable yield and enabling a lower bid.

Engineering Input 4 — Degradation Curve and P50/P90 Split (0.3–1.2% Yield Impact)

Every SECI PPA structure requires the developer to commit to a minimum annual generation. The P50 yield (the yield exceeded in 50% of years based on meteorological variability) is the central estimate used in the financial model, and the P90 (exceeded in 90% of years) is used for the debt sizing — lenders require DSCR coverage even in bad meteorological years.

The gap between P50 and P90 (typically 5–8% for Indian sites using TMY data from NSRDB or Solargis) has a direct impact on the loan-to-value ratio: a wider P50-P90 gap means lenders require more equity cushion, which raises the WACC and therefore the LCOE. An engineering firm that narrows the P90 confidence interval through higher-quality meteorological data processing allows the developer to increase the debt quantum — reducing WACC and enabling a lower tariff bid.

The degradation assumption also varies: applying MNRE’s recommended 0.7%/year flat rate versus a non-linear model (1.5% Year 1, 0.5%/year thereafter) produces a 0.3–0.5% difference in total energy over 25 years.

According to IREDA’s technical guidelines for solar financing, bankable PVsyst reports for projects above 10 MW must use site-specific soiling data from at least 12 months of field measurement or validated Meteonorm/Solargis data, and must include a P90 sensitivity analysis with uncertainty bands documented by source.

The Heaven Designs 0.4-Paise Bidding Margin Framework

This framework quantifies the cumulative tariff impact of all four engineering inputs for a SECI project. The name derives from the observation that each input, when optimized, moves the achievable bid by approximately 0.4–1.2 paise/kWh — and collectively they deliver the 4–paise margin that separates winning bids from the field.

Cumulative Impact: What 4 Paise Actually Means at Scale

The cumulative four-input optimization in the Heaven Designs framework delivers 2.0–4.0% additional annual yield — translating to 3–5 paise/kWh lower achievable bid for a 500 MW project. Here is the quantified impact:

Optimization	Yield Gain	Revenue Gain/Year (500 MW)	Bid Impact
Tracker + bifacial optimization	+1.2%	₹11.4 Cr/yr	1.2 paise lower
GCR optimization (land CAPEX)	+0.5% yield equiv.	₹5.0 Cr/yr (CAPEX reduction)	0.5 paise lower
Cleaning regimen (soiling reduction)	+0.8%	₹7.6 Cr/yr (net of cleaning cost)	0.8 paise lower
P90 narrowing (debt leverage)	+0% yield, +3% debt	₹WACC reduction	0.5 paise lower
Total	+2.5% effective	+₹24 Cr/yr	~3 paise lower

The P50 and P90 yield concepts are fundamental to understanding how meteorological uncertainty affects the bid and financing structure.

The SECI Compliance Layer: What IE Reviewers Actually Check

Every SECI-financed project requires Independent Engineer (IE) validation before financial close. The IE review is not a rubber stamp — it is a structured technical audit that will flag any simulation input that deviates from industry-accepted methodology.

The five most common IE flags on PVsyst submissions for SECI projects:

1. Soiling loss below 2% without measurement data — IE will demand 12 months of field soiling loss data or reference to a published soiling database.
2. Bifacial gain above 5% without albedo measurement — IE will require albedo from on-site measurement or a site-matched reference.
3. Degradation below 0.5%/year — IE typically applies 0.65–0.7% linear degradation per MNRE benchmarks for standard modules.
4. Tracker energy gain not matched to the actual tracker firmware version — IE will request the tracker manufacturer’s verified simulation plugin.
5. P90 uncertainty band below 5% — IE will require documentation of all uncertainty components (meteorological, inter-annual, model) per IEC 61724 methodology.

These are not negotiating points — they are review standards. An engineering team that builds the base case PVsyst report with IE-acceptable inputs avoids the revision cycle that delays financial close by 2–4 months.

The SECI glossary entry covers SECI’s auction structure and the technical submission requirements.

How Heaven Designs Helps

SECI auction success is an engineering problem as much as a commercial one. The difference between L1 and L3 bid in most auctions is within the reach of the engineering inputs this article covers — and each of those inputs requires a precise, lender-accepted simulation methodology.

• Solar Ground Mount Design — Utility-scale layout, tracker optimization, GCR sensitivity, and PVsyst simulation for SECI-scale projects.
• Bankable PVsyst Report — IE-accepted PVsyst output with P50/P90, soiling sensitivity, bifacial gain, degradation model, and complete uncertainty documentation.
• MW-Scale Project Management Consultancy — Owner’s engineer support through the SECI technical submission, IE review, and financial close process.
• Site Survey & Land Feasibility — Ground-truth irradiance, albedo, soiling, and terrain data needed for a defensible bid simulation.
• Download a sample deliverable — See an IE-accepted PVsyst report for a SECI-scale project before you commission your own.

Contact Heaven Designs to commission a bid-stage PVsyst simulation for your next SECI auction.

FAQ

How does the engineering yield simulation affect SECI auction bid prices?

The SECI tariff bid is calculated from the LCOE, which is driven by the annual energy yield from PVsyst simulation. Every 1% increase in yield allows a 1% reduction in the bid tariff while maintaining the same IRR. At 500 MW and ₹2.80/kWh, 1% yield improvement = ₹26.6 crores over 25 years. The four engineering inputs described in this article — tracker optimization, GCR, soiling, and P90 narrowing — collectively deliver 2.5–4% additional yield, enabling a 3–5 paise lower bid.

What soiling loss should be used in a SECI PVsyst report?

Soiling loss must be justified by site-specific data for IE-reviewed reports. For Rajasthan sites, the range is 2.5–4.5%/year depending on the cleaning regimen. For Gujarat, 2.0–3.5%. For Tamil Nadu coastal sites, 1.5–2.5%. IREDA and most project lenders require soiling data from either 12 months of on-site measurement or a validated database (Solargis SoilingDB). Using a generic 2% without documentation will be flagged in IE review and must be corrected before financial close.

What is the difference between P50 and P90 yield in a SECI project?

P50 is the annual energy yield exceeded in 50% of years based on meteorological variability — the central estimate used in the financial model. P90 is the yield exceeded in 90% of years — the conservative estimate used for debt sizing. The P50-P90 gap for Indian sites is typically 5–8% using TMY data, meaning the P90 is 5–8% below the P50. A wider gap forces lenders to require more equity at the same DSCR, raising the WACC and therefore the LCOE and minimum viable bid.

How does the single-axis tracker backtracking algorithm affect yield simulation?

The backtracking algorithm determines how the tracker rotates at low solar angles to avoid casting a shadow on adjacent rows. Standard backtracking algorithms use a flat-terrain assumption. Optimized backtracking accounts for terrain slope, which is significant at sites with 1–3% cross-slope. On such terrain, optimized backtracking adds 0.5–1.2% to annual yield compared to the standard model — a real difference that must be simulated using the actual tracker firmware’s algorithm, not PVsyst’s generic SAT model.

What does an Independent Engineer check in a SECI PVsyst report?

The IE review covers: soiling loss assumptions versus site data, bifacial gain versus albedo measurement, degradation rate versus MNRE benchmarks, tracker simulation versus actual firmware, P90 uncertainty documentation per IEC 61724, and the consistency between the simulation inputs and the equipment specification (module datasheet Pmax, temperature coefficient, bifaciality factor). Any significant deviation from industry-accepted methodology triggers a revision request that delays financial close.

How much can GCR optimization reduce the SECI bid price?

GCR optimization affects the bid through two channels: yield impact and land cost. On the yield side, the optimal GCR (versus a generic 0.40) can add 0.3–0.8% to annual generation. On the land cost side, a 10% reduction in land area through higher GCR saves ₹5–₹8 crores per 100 MW in land lease CAPEX (Rajasthan rates). Combined, GCR optimization allows a 0.3–0.8 paise/kWh reduction in the bid for a 500 MW project.

How does Heaven Designs produce bankable PVsyst reports for SECI projects?

Heaven Designs produces PVsyst reports following the methodology required by IREDA, PFC, and lenders’ IE reviewers. This includes: site-specific meteorological data from Solargis or NSRDB (20+ year dataset), actual tracker firmware simulation plugin, on-site or SoilingDB-sourced soiling data, bifacial gain with albedo measurement, and full P50/P90 uncertainty documentation. Reports are delivered in the IE submission format accepted by the major Indian project finance banks.