Assessment of Hot and Flaming Particles and Fire Risk From High Current Faults

The purpose of the project was to determine the relationship between the interrupt time, line current, and fault type and the particles released during a fault, and the resultant fire ignition hazard. This testing was performed at the Pacific Gas and Electric ATS High Current Test Yard in San Ramon, CA. This portion of the analysis was conducted using a FLIR high speed thermal imaging camera that takes thermal videos of the fault and resulting particle blast and fallout. This document is a technical summary showing the results of the FLIR infrared imaging of flashover events resulting from phase to phase contact between two different conductors of various configurations and attempts to place some of the data into an initial framework. Custom software was written to count molten particle in every frame of high speed video. Phase-to-phase overhead fault testing was conducted at three fault current magnitudes, (1000 amps, 4000 amps, and 7000 amps) using three different conductor types (#4 ACSR, 397 MCM Al and 3/0 CU) at various fault clearing times ranging from 0.1s to 1.5s. Tests were conducted in two different environments; one was the high current yard “Cage” where the test conductors were at approximately 10 feet in height. These tests occurred for all current magnitudes. The second phase of the test occurred in the open yard using the “Lift”, with heights initially at 15’ to finally being conducted at 40’, using both a standardized Cal-Fire approved fuel bed (dried grass) and 40 lb white virgin kraft paper. In addition to standard conductors, phase-to-ground fault testing was conducted between the conductor and a grounded section of structural steel angle iron to simulate a conductor-to-tower contact scenario. All testing was performed at 21kV line-to-line voltage, the maximum voltage available in the test yard. There are a few major independent variables for these tests. Primary among these variables is interrupt fault time. Conductor configurations was found to be another major variable that can be controlled by T-line engineering and construction. The first of these configurations is indicative of two lines that might come into contact due to aeolian based line and pole movement (described in this test as “Parallel”). The second configuration was the “worst-case” fault in which the “Pigtail” end of a conductor was directly pointed to the other phase concentrating the arc in one location. Other independent driving variables of lesser control include the conductor sizes and materials (as mentioned above), fault currents, and the height above ground. This particular part of this study attempts to quantify two main dependent variables • Total particle counts by time; and • Particles counts exceeding ignition temperature by time.

Some complex general formulas were developed to predict the number of particles produced under different test conditions as well as the number of particles that cooled as they fell toward fuel on the ground. The authors also advocate for additional industry sponsored testing to develop formulas predicting fire ignition risk for a few more common fault modes.