System and method for forecasting photovoltaic power generation system degradation

Patent metadata

Field	Value
Publication number	US-10599747-B1
Application number	US-201816125405-A
Country	US
Kind code	B1
Filing date	Sep 7, 2018
Priority date	Jul 25, 2011
Publication date	Mar 24, 2020
Grant date	Mar 24, 2020

Abstract

Long-term photovoltaic system degradation can be predicted through a simple, low-cost solution. The approach requires the configuration specification for a photovoltaic system, as well as measured photovoltaic production data and solar irradiance, such as measured by a reliable third party source using satellite imagery. Note the configuration specification can be derived. This information is used to simulate photovoltaic power production by the photovoltaic system, which is then evaluated against the measured photovoltaic production data to determine the degree of error between simulated and measured production. The simulated production is adjusted to account for the error and to infer degradation that can be projected over time to forecast long-term photovoltaic system degradation.

First claim

The invention claimed is:

1. A system for forecasting photovoltaic power generation system degradation, comprising the steps of:
a power meter capable of assessing measured photovoltaic production for a photovoltaic system operating at a known location over a set time period;
a monitoring infrastructure capable of assessing measured solar irradiance data for the known location over a reference time period that minimally overlaps with the set time period;
a configuration specification for the photovoltaic system; and
a digital computer comprising a processor and a memory that is adapted to store program instructions for execution by the processor, the program instructions capable of:
simulating time-series photovoltaic production by the photovoltaic system using the configuration specification and the solar irradiance data for the reference time period;
deriving adjustment factors by minimizing error between the time-series simulated photovoltaic production and the measured solar irradiance data;
creating adjusted time-series simulated photovoltaic production by multiplying the time-series simulated photovoltaic production by the adjustment factors;
calculating normalized ratios of the adjusted time-series simulated photovoltaic production to the time-series simulated photovoltaic production for a current time period and a time period previous to the current time period; and
selecting a degradation time period and calculating degradation of the photovoltaic system as a function of the normalized ratio for the current time period and the normalized ratio for the previous time period.

2. A system according to claim 1 , the program instructions further capable of:
calculating the degradation for consecutive degradation time periods; and
forecasting long-term degradation by evaluating a trend in the degradation for the consecutive time periods.

3. A system according to claim 2 , the program instructions further capable of:
evaluating the trend as the average of the long-term degradations for the consecutive time periods; and
evaluating the trend as the mean of the long-term degradations for the consecutive time periods.

4. A system according to claim 2 , the program instructions further capable of:
omitting the degradation for a first year of operation for the photovoltaic system.

5. A system according to claim 1 , wherein the long-term degradation Degradation t for the current time period t is determined in accordance with:

Degradation
t

=

1
-

Adjusted
⁢

⁢

Simulated

t
-
1

/

Simulated

t
-
1

Adjusted
⁢

⁢

Simulated
t

/

Simulated
t

where Adjusted Simulated t-1 represents the adjusted time-series simulated photovoltaic production for the previous time period; Simulated t-1 represents the time-series simulated photovoltaic production for the previous time period; Adjusted Simulated t represents the adjusted time-series simulated photovoltaic production for the current time period; and Simulated t represents the time-series simulated photovoltaic production for the current time period.

6. A system according to claim 1 , wherein the previous time period represents a first year of operation for the photovoltaic system, the program instructions further capable of:
normalizing the normalized ratio for the current time period by dividing by the normalized ratio for the previous time period.

7. A system according to claim 1 , the program instructions further capable of at least one of:
minimizing the error by evaluating relative mean absolute error;
minimizing the error by evaluating mean bias error; and
minimizing the error by evaluating root mean square error.

8. A system according to claim 1 , wherein the configuration specification is derived, the program instructions further capable of:
obtaining ambient temperature measured for the known location over the set time period; and
searching for optimal values for each variable in a configuration specification for the photovoltaic system by optimizing each variable, one at a time.

9. A system according to claim 8 , the program instructions further capable of:
selecting a candidate value for the variable being optimized;
simulating photovoltaic production for the photovoltaic system using the ambient temperature, the solar irradiance data, and the candidate value;
calculating error between the simulated photovoltaic production and the measured photovoltaic production; and
choosing the candidate value as the optimal value for the variable being optimized upon the error meeting a minimal threshold of error.

10. A system according to claim 1 , the program instructions for simulating time-series photovoltaic production being further capable of:
generating a set of sky clearness indexes as a ratio of each irradiance observation in a set of irradiance observations that has been regularly measured for the known location over the set time period, and clear sky irradiance;
forming a time series of the set of the sky clearness indexes;
determining irradiance statistics for the photovoltaic system through statistical evaluation of the time series of the set of the sky clearness indexes; and
building power statistics for the photovoltaic system as a function of the photovoltaic system irradiance statistics and the configuration specification.

11. A method for forecasting photovoltaic power generation system degradation with the aid of a digital computer, comprising the steps of:
assessing through a power meter measured photovoltaic production for a photovoltaic system operating at a known location over a set time period;
assessing through a monitoring infrastructure measured solar irradiance data for the known location over a reference time period that minimally overlaps with the set time period;
referencing a configuration specification for the photovoltaic system; and
operating a digital computer comprising a processor and a memory that is adapted to store program instructions for execution by the processor, the program instructions capable of:
simulating time-series photovoltaic production by the photovoltaic system using the configuration specification and the solar irradiance data for the reference time period;
deriving adjustment factors by minimizing error between the time-series simulated photovoltaic production and the measured solar irradiance data;
cre

Assignees

• Clean Power Res L L C

Inventors

• Hoff Thomas E

Classifications

• G06F17/18 Primary

  for evaluating statistical data {, e.g. average values, frequency distributions, probability functions, regression analysis (forecasting specially adapted for a specific administrative, business or logistic context G06Q10/04)} · CPC title

• H02J2101/24

  Photovoltaics · CPC title

• Y02E10/56

  Power conversion systems, e.g. maximum power point trackers · CPC title

• Y04S10/50

  Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications · CPC title

• H02J3/004

  Generation forecast, e.g. methods or systems for forecasting future energy generation · CPC title