Performance Solar ratio: efficiency and degradation

A photovoltaic system never converts POA into electricity with efficiency equal to that of the module under STC conditions. The operating temperature is always higher than 25°C; there is dirt on the glass; the DC cabling warms up; the investor does not convert with 100% efficiency; and part of the time the system is unavailable. The difference between what the module promises under laboratory conditions and what the system delivers under real field conditions is what the Performance Ratio captures.

The PR does not replace the resource analysis: it multiplies the resource. For the same annual POA (see NT.15), a larger PR means more generated energy. AEP = P_name × POA_annual × PRA 5% error in PR has the same impact on AEP as a 5% error in POA — and PR is often less severe in projections than the resource.

The definition IEC 61724-1

The IEC 61724-1 (FV systems monitoring) defines Performance Ratio as the ratio between the actual production of AC and the production that the system would have operated continuously under STC conditions with the same available resource:

PR = E_AC / (P_STC × H_POA / 1000)

E_AC   — AC energy delivered to the grid during the period (kWh)
P_STC  — system rated power at STC (kWp)
H_POA  — plane-of-array irradiation during the period (kWh/m²)
1000   — reference STC irradiance (W/m²)

# Equivalently in terms of specific yield (SY):
SY = E_AC / P_STC   (kWh/kWp — specific yield)
PR = SY / (H_POA / 1000)

# In SOLANO:
aep = p_nom * poa_annual * pr    # - kWh
sy  = poa_annual * pr            # kWh/kWp
cf  = aep / (p_nom * 8760) * 100  # %

The PR is dimensional and is between 0 and 1 — in practice between 0.70 and 0.90 for modern well-functioning systems. A PR of 0.82 means that the system produces 82% of the energy it would produce under STC conditions with the same resource.

What decomposes PR: the six components of loss

A typical PR of 84.5% in the Northeast results from six loss families that distance the actual operation from the STC conditions. Fig. 1 shows the cascade of losses from theoretical PR (1,000) to liquid PR (0.845).

The dominant component: the temperature of the module

The cell temperature is the loss factor that most distinguishes the Northeast from temperate sites. Photovoltaic modules of crystalline silicon (c-Si) typically lose 0.35 to 0.45% power per degree Celsius above 25°C (STC). In the Northeast, the ambient temperature during the generation hours is 28–35°C — and the temperature of the module is even higher.

The NOCT model (Nominal Operating Cell Temperature) is the standard way to estimate T_mod when there is no module temperature sensor available:

T_mod = T_amb + (NOCT − 20) / 800 × POA

NOCT — nominal operating temperature (°C); default: 45°C
800  — reference irradiance for NOCT (W/m²)
POA  — instantaneous module-plane irradiance (W/m²)

# Power loss by temperature (γ = temperature coefficient of P_max):
ΔP = P_STC × γ × (T_mod − 25)

# Percentage loss:
temperature_loss = |γ| × (T_mod − 25)

# Example: T_amb=30°C, POA=800 W/m2, NOCT=45°C, γ=−0.40%/°C:
T_mod  = 30 + (45−20)/800 × 800 = 30 + 25 = 55°C
perda  = 0,40% × (55−25) = 0,40% × 30 = 12,0%

PR temperature corrected: separating operating climate

The raw PR confuses two distinct factors: the quality of operation of the system and the climate of the site. A system operating perfectly in the Northeast (T_amb=30°C) will have PR smaller than the same system operating with the same efficiency in Joinville (T_amb=22°C) — not because it works worse, but because the hot climate reduces the output power.

For comparisons between sites or between different seasonal periods, IEC 61724-1 defines the temperature corrected PR (PR*):

PR* = PR / (1 + γ × (T_mod_mean − 25))

T_mod_mean — POA-weighted mean temperature during the period (°C)
γ          — power temperature coefficient (%/°C, negative)

# PR* removes the effect of the climate and reflects only operational efficiency:
# — Inverter with partial failure → PR* falls even in winter
# — Increasing dirt → PR* declines progressively
# — Module with power loss → PR* falls independent of the climate

SOLANO uses the measured T_mod when available (module temperature sensor on the back of the cell) or the NOCT model when only T_amb is available. The quality of this temperature data directly impacts the significance of calculated PR* — hence the importance of including T_mod sensor in the specification of the monitoring station.

The five types of loss in detail

The temperature component is structurally high in the Northeast and cannot be eliminated — only reduced. The other components depend on design quality, equipment and operation. A PR of 0.82–0.85 is realistic for well executed projects in the Northeast; values above 0.87 are exceptional and usually indicate some underestimation of losses.

The degradation of PR over 25 years

Photovoltaic modules do not maintain their rated power throughout the project. Two main phenomenologies reduce PR over time:

• LID (Light-Induced Degradation): rapid loss in the first year of operation due to Staebler-Wronski defects in crystalline silicon cells. Typically 1.5–2.5% in the first year. PERC and TOPCon modules have lower LID (0.5–1.0%).
• Annual degradation: Slow and progressive loss due to aging of VAS, discoloration of backsheet and oxidation of contacts. Typically 0.45–0.60%/year for standard c-Si; 0.30–0.40%/year for premium modules.

For financial model purposes, production over 25 years is the integral of the degraded curve — not the P50 of year 1 multiplied by 25. FIG. 3 shows the typical trajectory of PR throughout the project.

PR in the financial model: P50 and P90 of AEP

For bankability, the lender requires not only P50 from AEP, but P90 — production with 90% probability of being exceeded. PR contributes to the total uncertainty of AEP as follows:

# The uncertainty of typical PR (σ_PR) — main components:
σ_temp     = 1.5%   # Uncertainty in γ and T_mod model
σ_soiling  = 1.0%   # Interannual variability of dirt
σ_mismatch = 0.5%   # Uncertainty in commatch and binning
σ_inversor = 0.5%   # Efficiency curve vs temperature
σ_PR_total = sqrt(1.5² + 1.0² + 0.5² + 0.5²) ≈ 1.9%

# AEP AEP (horizontal 1 year):
P90_1yr = P50_AEP × (1 − 1.2816 × σ_total / 100)
# σ_total = RSS(σ_POA, σ_PR, σ_IAV, σ_others)

# P90 from AEP ( 20 years horizon, diluin interannual uncertainty):
P90_20yr = P50_AEP × (1 − 1.2816 × σ_total_20yr / 100)
# σ_IAV_20yr = σ_IAV_1yr / sqrt(20) — reduced component

The uncertainty of PR is one of the components of the total uncertainty budget of solar AEP. For bankable projects in the Northeast, the gap P50−P90 of total AEP is usually 5–9% for 1-year horizon and 3–5% for 20-year horizon (interannual uncertainty and PR converge with the long-term average). NT.06.

The PR of the EYA (Energy Yield Assessment) needs to be documented in detail for due diligence — it is not enough to quote "PR=0.82". The Independent Engineer will review each component of loss, check if the values are consistent with the site's climate, and will question PR above 0.86 or below 0.78 without explicit justification. The general criteria for the bankability of renewable reports are in NT.16.