EP-4742527-A1 - METHOD OF DETERMINING A CORRECTED IRRADIANCE OF A PHOTOVOLTAIC SYSTEM

EP4742527A1EP 4742527 A1EP4742527 A1EP 4742527A1EP-4742527-A1

Abstract

The method for determining a corrected irradiance comprises the steps of : - obtaining a dataset of operating points of photovoltaic cells, each operating point being defined by parameters including the photovoltaic cell, the day, the power output of the photovoltaic cell, the temperature of the photovoltaic cell and the irradiance of the photovoltaic cell, the operating points of the dataset being taken over a period of time ; - determining a plurality of data subsets in the dataset, each data subset including the operating points having a day included in one of a plurality of sub-periods of the period and an irradiance included in one of a plurality of irradiance sub-ranges; - performing regression analysis on each data subset for determining regression parameters for each data subset; and - calculating the corrected irradiance as a function of the regression parameters of the data subsets.

Inventors

EL ZAHER, Adham
GUILLEMOT, Loïc

Assignees

TotalEnergies OneTech

Dates

Publication Date: 20260513
Application Date: 20241112

Claims (20)

A method for determining a corrected irradiance of a photovoltaic system over a period, the photovoltaic system comprising photovoltaic cells, the method comprising the steps of : - obtaining a dataset of operating points of the photovoltaic cells, each operating point being defined by parameters including the photovoltaic cell, the day, the power output of the photovoltaic cell, the temperature of the photovoltaic cell and the irradiance of the photovoltaic cell, the operating points of the dataset being taken over a period of time ; - determining a plurality of data subsets in the dataset, each data subset including the operating points having a day included in one of a plurality of sub-periods of the period and an irradiance included in one of a plurality of irradiance sub-ranges; - performing regression analysis on each data subset for determining regression parameters for each data subset; - optionally, validating the dataset as a function of the regression parameters of the data subsets; - calculating the corrected irradiance as a function of the regression parameters of the data subsets; - optionally calculating a performance loss rate of the photovoltaic system as a function of the corrected irradiance.
A method for determining a corrected irradiance as in claim 1, wherein validation of the dataset comprises determining a comparison parameter as a function of the regression parameters and comparing the comparison parameter to a nominal parameter of the photovoltaic cells.
A method for determining a corrected irradiance as in claim 3, wherein the comparison parameter is mean value determined as the mean value of a Gaussian fit of a distribution of comparison parameters of the data subsets.
A method for calculating a corrected irradiance as in any one of the preceding claims, wherein the corrected irradiance is calculated as a function of the initial irradiance based on a regression analysis between initial irradiance and the regression parameters of the data subsets.
A method for determining a corrected irradiance as in any one of the preceding claims, wherein the regression analysis comprises, for each data subset, determining a power variation and power at 0°C by performing a linear regression according to the following formula : P = a i T cell + b i in which P is the power of the photovoltaic cell, Tcell is the temperature of the photovoltaic cell, a i is the power variation (W/°C) of the i th data subset, b i is the power at 0°C (W of the i th data subset.
A method for determining a corrected irradiance rate as in claim 5, wherein the validation test comprises, for each data subset, determining a power thermal coefficient representing the variation of the power output a photovoltaic cell as a function of the temperature of the photovoltaic cell: γ i = a i b i in which a i and b i are the regression parameters (power variation and power at 0°C) determined for the i th data subset and γ i (%/°C) is the power thermal coefficient of the i th data subset.
A method for determining a corrected irradiance rate as in claim 6, wherein the validation test comprises comparing a mean power thermal coefficient of the data subsets to a nominal power thermal coefficient of the photovoltaic cells.
A method for determining a corrected irradiance as in claim 7, wherein the mean value of the power thermal coefficient of the data subsets is determined as the mean value of a Gaussian fit of the distribution of the power thermal coefficient of the data subsets.
A method for determining a corrected irradiance as in claim 7 or 8, wherein the dataset is validated if the absolute value of a difference between the mean value of the power thermal coefficients of the data subsets and a nominal power thermal coefficient of the photovoltaic cells is below a validation threshold.
A method for determining a corrected irradiance as in claim 9, wherein the validation threshold is 2.10 -3 °C -1 .
A method for determining a corrected irradiance as in any one of claim 5 to 10, wherein the determination of the corrected irradiance comprises determining an overall power variation coefficient by performing a linear regression of the power variations of the data subsets as a function of the irradiance according to the following formula: a = αG poa wherein a is the power variation, G poa is the irradiance and α is an overall power variation coefficient.
A method for determining a corrected irradiance as in claim 11, wherein the determination of the corrected irradiance comprises, for each data subset, determining a corrected power at 0°C (W by performing a linear regression analysis on the power, the temperature, the irradiance and the powers at 0°C of the data subsets according to the following formula: P = αG poa T cell + B i in which P is the power of the operation points of the i th data subset, Tcell is the temperature of the operation points of the i th data subset, G poa is the irradiance of the i th data subset, α is the overall power variation coefficient and B i is the corrected power at 0°C (W) of the i th data subset.
A method for determining a corrected irradiance as in claim 12, wherein the determination of the corrected irradiance comprises determining an overall coefficient of power at 0°C by performing a linear regression of the corrected power at 0°C of the data subsets as a function of the irradiance according to the following formula: b = δG poa wherein b is the corrected power at 0°C, G poa is the irradiance of the data subsets and δ is an overall coefficient of power at 0°C.
A method for determining a corrected irradiance as in claim 13, wherein the corrected irradiance is calculated for each operation point according the following formula: G poa new = G poa + B i δ wherein Gpoa(new) is the corrected irradiance, G poa if the initial irradiance, B i is corrected power at 0°C of the i th data subset of the operation point and δ is the overall coefficient of effective power at 0°C.
A method for determining a corrected irradiance as in any one of the preceding claims comprising quantifying the seasonality of the dataset.
A method for determining a corrected irradiance as in claim 15, wherein quantifying the seasonality of the dataset comprises fitting the power output to a predetermined function having an amplitude and determining said amplitude.
A method for determining a corrected irradiance as in claim 16, wherein if the amplitude exceeds a predetermined threshold, the dataset is labeled as having seasonal characteristics and the subsequent steps are performed.
A method for determining a corrected irradiance as in claim 16 or 17, wherein the function is: PR = Asin 2 Π T t + Φ with A the amplitude, T the period of the sinusoid, and t the time, e.g. expressed as a number of days.
A method for determining a corrected irradiance as in any one of the preceding claims, wherein the period extends over several months, in particular at least twenty-four months.
A method for determining a corrected irradiance as in any one of the preceding claims, wherein each sub-period is a month.

Description

The invention relates to the production of electricity using solar photovoltaic (PV) systems. The rapid growth of production of electricity using PV systems has transformed the energy landscape, with significant implications for the efficiency, reliability, and performance of PV systems worldwide. As the global share of PV systems in electricity production continues to rise, ensuring the optimal operation of these PV systems over extended periods becomes increasingly important. PV cells of a PV system are subject to a natural decline in production capacity over time. One critical challenge for operators is estimating the annual degradation rates of PV cells of a PV system. While PV module manufacturers provide nominal degradation rates, these values often serve as little more than a baseline, with actual performance losses frequently exceeding expectations. The power output of PV systems is influenced by two primary meteorological factors: the temperature of the PV cells (Tcell) and the irradiance on the plane of the PV cells (Gpoa). Normalizing power by these values enables the calculation of the Performance Ratio Temperature-Corrected (PRTC), which facilitates the evaluation of long-term performance loss rates (PLR). However, even after accounting for temperature and irradiance, many PV systems exhibit residual seasonal variations in PRTC, compromising the accuracy of PLR assessments. Current methods, such as a year-on-year approach or seasonal trend decomposition techniques, often fail to identify the underlying causes of these seasonality effects and assume a constant degradation pattern over extended periods. One of the aims of the invention it to propose a method which allows accurately evaluating the PLR of PV systems over extended periods, addressing the challenges posed by seasonal variations in PRTC values. To this end, the invention proposes a method for determining a corrected irradiance of a photovoltaic system over a period, the photovoltaic system comprising photovoltaic cells, the method comprising the steps of: obtaining a dataset of operating points of the photovoltaic cells, each operating point being defined by parameters including the photovoltaic cell, the day, the power output of the photovoltaic cell, the temperature of the photovoltaic cell and the irradiance of the photovoltaic cell, the operating points of the dataset being taken over a period of time ;determining a plurality of data subsets in the dataset, each data subset including the operating points having a day included in one of a plurality of sub-periods of the period and an irradiance included in one of a plurality of irradiance sub-ranges;performing regression analysis on each data subset for determining regression parameters for each data subset;optionally, validating the dataset as a function of the regression parameters of the data subsets;calculating the corrected irradiance as a function of the regression parameters of the data subsets;optionally calculating a performance loss rate of the photovoltaic system as a function of the corrected irradiance. The approach provides improved accuracy and reliability compared to existing methods, enabling more informed decision-making for PV system operators and developers. In some examples, the method comprises one or more of the following optional features, taken individually or in any technically feasible combination: validation of the dataset comprises determining a comparison parameter as a function of the regression parameters and comparing the comparison parameter to a nominal parameter of the photovoltaic cells;the comparison parameter is a mean value determined as the mean value of a Gaussian fit of a distribution of comparison parameters of the data subsets:the corrected irradiance is calculated as a function of the initial irradiance based on a regression analysis between initial irradiance and the regression parameters of the data subsets;the regression analysis comprises, for each data subset, determining a power variation and power at 0°C by performing a linear regression according to the following formula : P=aiTcell+bi in which P is the power of the photovoltaic cell, Tcell is the temperature of the photovoltaic cell, ai is the power variation (W/°C) of the ith data subset, bi is the power at 0°C (W) of the ith data subset;the validation test comprises, for each data subset, determining a power thermal coefficient representing the variation of the power output a photovoltaic cell as a function of the temperature of the photovoltaic cell: γi=aibi in which ai and bi are the regression parameters (power variation and power at 0°C) determined for the ith data subset and γi (%/°C) is the power thermal coefficient of the ith data subset;the validation test comprises comparing a mean power thermal coefficient of the data subsets to a nominal power thermal coefficient of the photovoltaic cells.the mean value of the power thermal coefficient of the data subsets is determ