PV Energy Loss Estimation via Partial Failure Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current models for estimating power generation from solar photovoltaic (PV) systems are inaccurate due to factors like soiling, shading, snow cover, and partial hardware failure, leading to overestimation or underestimation of energy production, especially in real-world conditions where short-term fluctuations occur.
Innovation Solution
A computer processor-based method that estimates energy losses due to partial equipment failure by merging modeled and measured power and energy time series data, calculating performance ratios, applying an anomaly filter, and partitioning data to identify capacity changes, thereby providing a piecewise constant time series that modulates energy loss estimates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If derate factors are used to estimate energy losses, then the modeling is simple, but the accuracy deteriorates due to short-term fluctuations in losses
Solution Approach 1:
The patent segments the energy loss estimation into multiple categories (soiling, shading, snow, equipment failure) with different temporal characteristics. Each category is modeled separately using appropriate time scales (daily, weekly, monthly, yearly) rather than applying a single derate factor, thereby improving accuracy while maintaining manageable complexity through structured segmentation.
Solution Approach 2:
The patent transitions from static derate factors to dynamic, time-varying loss estimates. The model updates loss estimates at different frequencies depending on the loss category (e.g., daily for soiling, yearly for equipment failure), allowing the system to adapt to short-term fluctuations while maintaining computational efficiency through selective dynamic updating.
2Measurement precision
If monitoring equipment is installed on every component, then detection accuracy improves, but system cost and complexity increase
Solution Approach 1:
The patent introduces an intermediary analytical model that processes aggregate system-level performance data to infer individual component status. Instead of installing sensors on every component, the model uses the relationship between overall system performance and expected component contributions to detect failures, thereby achieving component-level monitoring accuracy without component-level sensing infrastructure.
Solution Approach 2:
The monitoring system uses the PV system's own performance data to detect component failures. The analytical model processes the system's measured power and energy production against modeled expectations to identify anomalies indicating component failure, allowing the system to self-diagnose without external monitoring equipment on each component.
3Device complexity
If aggregate monitoring is used for PV components, then system simplicity is maintained, but the ability to detect individual component failures deteriorates
Solution Approach 1:
The patent uses an intermediary analytical model that acts as a virtual sensor, processing aggregate performance data to extract information about individual component status. The model compares measured system performance against modeled expectations to infer which components may be failing, thereby recovering component-level information from aggregate measurements without requiring component-level sensors.
Solution Approach 2:
The patent replaces physical monitoring sensors on individual components with a virtual monitoring system based on analytical modeling. Instead of using mechanical/electrical sensors to directly measure component status, the system uses computational models to infer component health from overall system performance, substituting physical sensing infrastructure with information processing.
Data Source
AI summary
The present invention provides methods and systems to estimate energy losses due to partial equipment failure in photovoltaic (PV) systems based on measured power and energy data, weather data, PV system configuration information, and modeled power and energy generation data.