The DC-to-AC ratio — the ratio of the installed DC array capacity to the inverter AC rated power — is one of the most consequential single-number decisions in a solar project’s financial design. Set it too low and you are buying more inverter capacity than the site’s irradiance distribution can justify, adding capital cost without proportional energy gain. Set it too high and you are clipping energy during peak hours, paying for modules that never fully contribute to project revenue. The difference between a ratio of 1.15 and 1.35 on a 50 MW project can represent several crores in capital cost difference and several percentage points of LCOE improvement.
Direct answer. The DC-to-AC ratio (also called the inverter loading ratio or ILR) is the DC array peak power divided by the inverter rated AC output power. For most sites in India and the USA, the economic optimum ILR falls between 1.10 and 1.35 — the ILR Optimization Curve shows that clipping losses below 3–4% are acceptable when offset by the capital cost saving from smaller inverter capacity. Above ILR 1.4, clipping losses accelerate and erode the CAPEX saving. High-irradiance sites (Rajasthan, southern Spain, Arizona) tolerate lower ILR than lower-irradiance sites. Bifacial modules and trackers both shift the optimal ILR downward by 0.05–0.10 because they increase the effective DC output without proportionally increasing the flat DC nameplate.
This tutorial is written for all ICPs: Rohan bidding a rooftop in Pune, Mike sizing a 500 kW C&I system in Texas, Jennifer reviewing a PVsyst report for bankability, and Suresh evaluating a 100 MW fixed-tilt ground mount in Rajasthan. The ILR calculation is the same in every market; the optimal value shifts with the irradiance profile of the site.
What DC-to-AC Ratio and ILR Mean
The terms “DC-to-AC ratio” and “inverter loading ratio” (ILR) are used interchangeably in most engineering contexts. Both refer to the ratio of the total installed DC peak power (in kWp or MWp) to the total inverter rated AC output power (in kW or MW) at standard conditions.
Definition. The DC-to-AC ratio (ILR) = DC array peak power (kWp) / Inverter rated AC power (kW). A 1 MWp array connected to a 770 kW inverter has an ILR of 1.30. The ILR is dimensionless and always expressed as a decimal or ratio (1.30), not as a percentage.
A related but distinct term is the “oversizing ratio” or “overdimensioning ratio” — sometimes used specifically to describe how much larger the DC array is relative to the inverter. An ILR of 1.30 corresponds to a 30% DC oversizing relative to the inverter AC capacity. The PVsyst simulation software uses “oversizing ratio” in its system parameter inputs; the value is mathematically equivalent to (ILR - 1) x 100%.
Why Inverters Are Undersized Relative to the DC Array
The counterintuitive practice of connecting a larger DC array to a smaller inverter exploits a fundamental characteristic of solar irradiance: the array operates at its DC nameplate capacity (1 kWp per 1,000 W/m² at STC temperature) for only a few hours per year. For the majority of daylight hours, the irradiance is below 1,000 W/m² — due to early mornings, evenings, cloud cover, soiling, and seasonal variation — and the array output is proportionally lower than its DC nameplate.
According to NREL’s 2023 PV System Performance Benchmarks, a typical 1 MWp ground-mount array in a high-irradiance location (GHI above 2,000 kWh/m²/year) operates above 90% of its DC nameplate power for fewer than 200 hours per year — less than 2.3% of the 8,760 hours in a year. For a lower-irradiance location (GHI 1,400–1,600 kWh/m²/year), this percentage drops below 0.5%. Sizing the inverter to handle the peak DC output that occurs less than 2% of the time is economically inefficient.
1.10–1.35
Typical optimal ILR for Indian and USA ground-mount
IRENA renewable power generation costs, 2024
3–5%
Acceptable annual clipping loss at ILR 1.25–1.35
PVsyst simulation benchmark for Rajasthan, 2024
8–12%
Typical inverter CAPEX as % of total system cost
Mercom India market data, 2025
The ILR Optimization Curve — Heaven Designs Framework
The ILR Optimization Curve is Heaven Designs’ graphical framework for selecting the economically optimal ILR for a specific site. It plots two competing trends on the same graph as ILR increases from 1.0 to 1.5:
- Clipping loss (%): Increases with ILR as the inverter saturates at rated power for more hours per year. The curve is shallow at low ILR (below 1.2) and steepens sharply above 1.4.
- CAPEX saving (%): Decreases linearly as the inverter size shrinks relative to the fixed DC array size. At ILR 1.0, no CAPEX saving; at ILR 1.5, maximum CAPEX saving.
The economic optimum is the ILR where the marginal value of additional CAPEX saving equals the marginal cost of additional clipping loss. For a project selling power at a fixed PPA tariff (as in most SECI and Indian DISCOM PPAs), the clipping loss cost is simply the lost revenue: clipping loss (%) x annual energy (kWh) x PPA tariff (₹/kWh).
Run the PVsyst simulation at multiple ILR values
Run PVsyst simulations at ILR = 1.00, 1.10, 1.15, 1.20, 1.25, 1.30, 1.35, 1.40, and 1.50 for the project site and configuration. Record the annual energy yield (kWh/kWp) and the clipping loss (%) from each simulation. The clipping loss output in PVsyst is shown in the "Array losses" section as "Inverter saturation" or "AC limitation."
Calculate the CAPEX saving at each ILR
For each ILR value, calculate the inverter capacity reduction relative to ILR 1.0 (in MW) and multiply by the inverter cost per MW. This gives the CAPEX saving from oversizing the DC array. The DC module cost remains fixed because the array size is held constant for this comparison — only the inverter size changes.
Calculate the clipping loss cost at each ILR
Multiply the clipping loss percentage by the total annual energy (kWh) and the PPA tariff (₹/kWh or $/kWh) to get the annual revenue loss. Discount this annual loss at the project WACC over the PPA term to get the present value of the clipping loss cost. This is the life-of-project cost of the CAPEX saving from each ILR choice.
Select the ILR where net NPV is maximised
Plot (CAPEX saving - PV of clipping loss cost) against ILR. The peak of this curve is the economically optimal ILR. For most Indian fixed-tilt ground-mount projects, this peak falls between ILR 1.20 and ILR 1.30. Accept the result to one decimal place of ILR — the uncertainty in the irradiance data and module degradation curve means more precision is illusory.
Clipping Loss Calculation — How Much Energy Are You Discarding?
Clipping loss (also called “inverter saturation loss” in PVsyst) is the energy that the DC array generates but the inverter cannot convert to AC because it is already operating at its rated AC output power. During clipping, the inverter’s MPPT algorithm effectively moves the DC operating point away from Vmp to a higher voltage that keeps the DC input power at the inverter’s AC rated power.
The clipping loss can be estimated without a full PVsyst simulation using the irradiance frequency distribution of the site:
Approximate clipping loss (%) = Fraction of hours where POA irradiance x ILR > 1,000 W/m²
For a site with POA irradiance distribution data (hours per irradiance class), sum the energy in all hours where the DC array would produce more than 1 MW for each MW of inverter capacity. This is the energy that would be clipped. Divide by the total annual energy to get the clipping loss fraction.
Field tip. The PVsyst "Inverter saturation" loss shown in the array losses table is the most accurate clipping loss figure because it accounts for the actual inverter overload characteristic, which is not a sharp cutoff at rated power. Most modern inverters can produce 105–110% of rated AC power for short periods (seconds to minutes) before the thermal protection reduces output. PVsyst models this behaviour, so its clipping loss is typically 10–20% lower than a simplified calculation that assumes a sharp cutoff at rated power.
Clipping Loss by Configuration — Fixed-Tilt vs Tracker vs Bifacial
The optimal ILR is not a universal number — it depends on the array configuration because the irradiance distribution hitting the modules changes with tilt, tracking, and module type.
| Configuration | Why ILR Is Affected | Typical Optimal ILR |
|---|---|---|
| Fixed-tilt, monofacial | Irradiance profile peaks sharply at local noon; moderate peak sharpness | 1.20–1.30 |
| Single-axis tracker, monofacial | Tracker increases POA irradiance during early morning and late afternoon but sharpens the midday peak | 1.15–1.25 |
| Fixed-tilt, bifacial | Rear irradiance adds 5–15% to effective DC output, effectively raising the ILR beyond the nameplate ratio | 1.15–1.25 |
| Single-axis tracker, bifacial | Combines tracker and bifacial gains — highest effective ILR of all configurations | 1.10–1.20 |
| East-west fixed layout | Splits the irradiance peak between morning and afternoon, reducing the instantaneous peak | 1.25–1.40 |
For bifacial modules, the effective DC output exceeds the nameplate STC power by the bifacial gain. If a 400W bifacial module delivers 420W effective output due to 5% bifacial gain, the effective ILR is 5% higher than the nameplate ILR. A design nominally at ILR 1.20 (based on 400W nameplate) is effectively operating at ILR 1.26 with bifacial gain — which means the clipping loss is higher than the nameplate-based PVsyst simulation would show unless the bifacial gain is correctly modelled.
According to IRENA’s 2024 Utility-Scale Solar Power Plants report, bifacial single-axis tracker configurations have become the dominant technology choice for new utility-scale ground-mount projects in India and the USA, representing over 65% of new capacity additions in 2023. The combination of tracker yield improvement (15–25%) and bifacial gain (5–15%) together raise the effective DC output by 20–40%, requiring designers to carefully re-evaluate ILR assumptions inherited from earlier fixed-tilt monofacial projects.
LCOE Impact of ILR Selection
The levelised cost of energy (LCOE) is the primary metric for comparing ILR choices on a life-of-project basis. LCOE = (Total life-cycle cost) / (Total life-cycle energy), where both the cost and energy are discounted to present value.
ILR affects LCOE through two channels:
- CAPEX reduction: A higher ILR reduces the inverter cost component of CAPEX. For a 100 MW AC project, increasing ILR from 1.10 to 1.25 reduces the inverter capacity from 91 MWac (for 100 MWdc) to 80 MWac — saving approximately 11 MW of inverter capacity at ₹25–35 lakhs/MW, or ₹2.75–3.85 Cr in total.
- Energy reduction (clipping): A higher ILR increases clipping loss. Increasing ILR from 1.10 to 1.25 at a high-irradiance Rajasthan site might increase clipping from 0.5% to 2.5% of annual energy — a 2% absolute increase representing approximately 3.6 million kWh per year on a 100 MW project.
At a PPA tariff of ₹2.50/kWh, the 3.6 million kWh clipping loss costs ₹90 lakhs/year. Over a 25-year PPA at a 10% discount rate, the present value of this clipping loss cost is approximately ₹8.4 Cr. The CAPEX saving is ₹2.75–3.85 Cr. In this example, the ILR increase from 1.10 to 1.25 destroys value — the clipping cost exceeds the CAPEX saving.
This illustrates that the optimal ILR is highly site- and tariff-specific. A lower irradiance site (such as northern India or the UK) where clipping loss at ILR 1.25 is only 0.8% rather than 2.5% would show a different — and more favourable — ILR optimisation result.
How to Model ILR in PVsyst
PVsyst is the industry-standard simulation tool for ILR sensitivity analysis. The procedure for running an ILR sweep in PVsyst:
- Open the system definition in PVsyst and navigate to “Inverter and Grid” in the system parameters.
- Set the “Pnom ratio” (PVsyst’s name for ILR) to the first test value (e.g., 1.10). PVsyst will automatically calculate the required AC-side power for the given DC array size.
- Run the simulation and record the “Inverter saturation” loss from the loss diagram in kWh and as a percentage of incident energy.
- Change the Pnom ratio to 1.15, re-run, and record the results. Repeat for ILR = 1.20, 1.25, 1.30, 1.35, 1.40.
- Export the results table and plot the clipping loss curve versus ILR.
Definition. PVsyst's "Inverter saturation" loss is the fraction of DC energy that cannot be converted to AC because the inverter is operating at its rated AC power output. It is shown as a percentage of total incident energy (not DC array energy). A 2% inverter saturation loss means 2% of the total irradiance energy falling on the array is lost due to inverter limitation — not 2% of the DC array nameplate output.
For bankable PVsyst reports accepted by Indian lenders (PFC, IREDA, SBI) and SECI independent engineers, the ILR sensitivity analysis must be included in the PVsyst report annex. The report must document the chosen ILR and the rationale — simply stating “ILR = 1.25” without showing the sensitivity analysis is not acceptable for a P50/P75/P90 bankable yield report. See our guide on bankable PVsyst reports for the full lender requirements.
How Grid Curtailment Changes the ILR Equation
Grid curtailment — when the DISCOM or SLDC instructs the plant to reduce output below its generation capability — fundamentally changes the ILR optimisation. Curtailed plants operate at a capped AC output for extended periods, effectively imposing additional clipping loss above the inverter’s rated power.
If a 50 MW AC plant is curtailed to 30 MW for 20% of its operating hours (a realistic curtailment rate in some Rajasthan and Karnataka grid zones), the effective clipping loss is much higher than PVsyst predicts without curtailment. In this case, increasing the ILR (adding more DC modules relative to the AC inverter capacity) makes the curtailment problem worse — there is more DC energy to clip — rather than better.
For projects in grid-constrained regions, the correct approach is to model the curtailment scenario explicitly and reduce the ILR below the unconstrained optimum. According to IEA-PVPS Task 16 (Solar Resource and Forecasting), curtailment rates above 10% annual energy loss materially change the optimal ILR by 0.05–0.15 downward from the unconstrained case.
HIGHER ILR MAKES SENSE WHEN...
- Site GHI is below 1,800 kWh/m²/year (lower irradiance climates)
- East-west layout flattens the irradiance peak
- Module cost is relatively low (modules are cheap vs inverters)
- Grid curtailment risk is negligible
- Project is at high latitude with diffuse-dominant irradiance
LOWER ILR MAKES SENSE WHEN...
- Site GHI is above 2,200 kWh/m²/year (Rajasthan, Gujarat, Arizona)
- Single-axis tracker and bifacial configuration raises effective DC output
- Grid curtailment exceeds 10% annual energy loss
- PPA tariff is high (₹4.50/kWh and above) — clipping loss is expensive
- Module cost is relatively high vs inverter cost
Download a sample ILR sensitivity analysis from PVsyst
Heaven Designs sample pack includes a redacted PVsyst ILR sensitivity report for a 50 MW fixed-tilt project in Rajasthan, showing clipping loss curves at ILR 1.10–1.40 and the NPV optimisation calculation.
Get the sample pack →How Heaven Designs Helps with ILR Recommendation
ILR selection is one of the decisions that most directly affects a project’s LCOE and bankability. An ILR recommendation that is not supported by a site-specific PVsyst sensitivity analysis is not bankable — lenders’ independent engineers will flag it. Heaven Designs provides ILR recommendation as part of the pre-design and PVsyst simulation service.
- Solar Rooftop Detailed Engineering Design — Full IFC design package including ILR optimisation, PVsyst simulation, and BOQ for Indian and USA projects.
- Solar 3D Pre-Design — Sales-stage ILR recommendation and preliminary PVsyst yield analysis in 48 hours for bid-stage project evaluation.
- Solar Permit Design — USA residential and C&I permit packets including string sizing and ILR documentation for AHJ submittal.
- Download a sample PVsyst report — See a bankable-grade PVsyst report with ILR sensitivity analysis before you engage.
Contact us to get a project-specific ILR recommendation. We typically deliver a PVsyst sensitivity run and ILR recommendation within 5 working days from receipt of site coordinates, module specification, and inverter preferences.
FAQ
What is a good DC-to-AC ratio for a rooftop solar system in India?
For Indian C&I rooftop projects (50 kW to 5 MW) with a typical fixed-tilt orientation, an ILR of 1.15 to 1.25 is appropriate for most sites in peninsular India (GHI 1,800–2,000 kWh/m²/year). For high-irradiance rooftop sites in Rajasthan and Gujarat (GHI 2,000–2,200 kWh/m²/year), an ILR of 1.10 to 1.20 reduces excessive clipping during summer midday hours. The exact optimum requires a PVsyst simulation with the actual site’s irradiance data — a regional rule of thumb is a starting point, not a final answer.
What is the difference between ILR and performance ratio?
The ILR (inverter loading ratio) is a design parameter — a ratio of installed DC peak power to inverter AC rated power, set during design and fixed for the life of the project. The performance ratio (PR) is an operational performance metric — the ratio of actual energy output to the irradiance-normalised theoretical energy output, measured after the plant is commissioned. A higher ILR increases clipping loss, which reduces the PR. A lower ILR reduces clipping loss and improves PR, but costs more in upfront inverter CAPEX.
Does grid curtailment count as clipping loss in PVsyst?
Standard PVsyst simulations do not model grid curtailment. PVsyst’s “AC limitation” (inverter saturation) loss models the physical limitation of the inverter’s rated AC output power, but not the external constraint of grid operator curtailment instructions. To model curtailment in PVsyst, use the “grid limitation” parameter in the system definition, which simulates a fixed AC output cap below the inverter rated power. Always distinguish between inverter saturation loss (engineering design parameter) and curtailment loss (external grid constraint) in the bankable yield report — lenders treat them differently in their risk models.
How does the optimal ILR change for a single-axis tracker project?
Single-axis trackers improve the irradiance distribution by reducing the sharpness of the midday peak and increasing energy capture in the morning and afternoon shoulder periods. This flattening of the irradiance profile means that a given ILR produces less clipping loss with trackers than with fixed-tilt — because the DC output exceeds the inverter rated power for fewer hours even at the same ILR. The practical result is that the optimal ILR for a single-axis tracker project is 0.05–0.10 lower than for an equivalent fixed-tilt project. At ILR 1.20, a fixed-tilt project might see 2.5% clipping loss while a tracker project at the same ILR sees only 1.5% clipping loss.
What ILR do Indian lenders (PFC, IREDA, SBI) expect?
Indian lenders financing utility-scale SECI and DISCOM projects do not specify an ILR requirement as a standalone criterion, but their independent engineers (IEs) scrutinise the ILR as part of the PVsyst report review. An ILR above 1.40 without supporting ILR sensitivity analysis in the PVsyst report will receive an IE comment requesting justification. An ILR below 1.10 at a high-irradiance site will also receive a comment asking why the developer chose not to take advantage of the CAPEX saving available through reasonable DC oversizing. The bankable range accepted without challenge by most IEs in India is 1.10 to 1.35, with the specific value supported by the site’s PVsyst sensitivity run.
Is ILR the same as DC:AC ratio and oversizing ratio?
ILR (inverter loading ratio) and DC:AC ratio are the same metric, expressed differently. Oversizing ratio is related: oversizing ratio (%) = (ILR - 1) x 100%. An ILR of 1.25 corresponds to a 25% oversizing ratio. Some inverter manufacturers use “overload factor” to describe how much the DC array can exceed the inverter rated power — again, this is the ILR expressed as a dimensionless ratio. All three terms describe the same design choice: how much larger the DC array is relative to the inverter AC rated power.
