This abstract describes the capabilities of spread spectrum time domain reflectometry (SSTDR) technology for whole-system (cells/modules, wiring, interconnections, etc.) fault detection, localization, and monitoring on energized and de-energized solar arrays.  Electrical faults (e.g., disconnects, ground faults, and arc faults) remain an extremely challenging problem for the solar industry. Electrical faults can originate from many sources (i.e., natural degradation, animals, operator error, manufacturing errors, etc.) and can often occur transiently. For example, ground faults (where an electrical wire is inadvertently connected to earth ground) can often appear in rain or times of high soil dew and then disappear when the soil dries. Since current electrical fault detection systems often require the panels to be de-energized for testing, tracking down ground faults can become costly and time-consuming. Under good circumstances, electrical faults can cause downtime and cost the time of inspectors to find the faults and electricians to repair them. Under poor circumstances, electrical faults can blow inverter electronics and cause fires. As a result, promptly detecting these faults is highly important for the solar industry.

We discuss how to detect, locate, and monitor electrical faults with SSTDR. SSTDR is an electrical nondestructive evaluation technique in which broadband pseudo-noise codes are transmitted into an electrical transmission line. These signals reflect off impedance variations in the transmission line. The reflections are then measured at the original excitation location and cross-correlated with the transmitted code to transform the data into an interpretable train of pulses. These pulses are then analyzed to detect and locate faults (or incipient faults) within the electrical systems. SSTDR measurements take milliseconds to acquire and can monitor and locate early-stage intermittent and time-dependent electrical impedance changes. We discuss algorithms to detect, locate, and monitor these impedances. We discuss the effects of the environment (brightness, temperature, and humidity) on this data over time. We also present several signal processing techniques to detect and locate reflection sources and machine learning techniques algorithms, such as dictionary learning, to identify non-obvious patterns within the data.

We demonstrate our results with two solar arrays. We use an S100 SSTDR system developed by Livewire Innovation. The system transmits a broadband signal with a 24 MHz center frequency. First, we demonstrate signal processing advances used for real-time disconnect monitoring in an approximately 1000 V array at the Department of Energy National Renewable Energy Laboratory. The array of 26 panels. We identify disconnects within the region of one panel in real-time as technicians disconnect cables across the array. We also demonstrate longer-term monitoring on a small, five-panel solar array on the roof of the New Engineering Building at the University of Florida. Specifically, we demonstrate how regression analysis and dictionary learning can be used to extract important trends from the solar array SSTDR data. Using these techniques, we demonstrate a near 99% accuracy with identifying disconnect faults and their locations within the array.

 

*Acknowledgments*

This material is based upon work supported by the U.S. Department of Energy’s Office of Energy Efficiency and Renewable Energy (EERE) under the Solar Energy Technologies Office (SETO) Agreement Number DE-EE0008169.

 

*Disclosure*

Dr. C.M. Furse is a co-founder of LiveWire Innovation, Inc. which is commercializing SSTDR technology, and therefore has a financial conflict of interest with this company.


Spread Spectrum Time Domain Reflectometry for Health Monitoring of Solar Arrays

No proceedings papers in 2020. Video only.

QNDE2020 - Virtual Conference

Supporting the need among researchers worldwide for additional online and multimedia content, ASME now offers the ASME Conference Video Collection. This platform will collect and provide long-term access to conference presentation videos, encompassing all technology areas within ASME's technical conference and events program. About 30,000-40,000 engineers across all mechanical engineering and related disciplines participate in ASME conferences each year. This includes authors of approximately 6,000 to 8,000 technical papers published by ASME, the majority of whom also submit video presentations for our virtual conferences. The video collection will be available as an add-on to existing product packages or as a separate subscription.