Arch flash burns are one of the top three most common dangers when working with energized electrical equipment, with two-thirds of arc flash incidents resulting from human error, according to the Occupational Safety and Health Administration (OSHA). In Part 1 of our Solar Arc Flash Safety Series, we focused on the efficacy of warning labels and whether working on energized electrical equipment should ever be required. Below, we cover to what extent worst case arc flash scenarios should be factored in and how to handle routine troubleshooting in light of arc flashes.
The Relevance of Worst Case Scenarios
There is currently an engineering obsession surrounding finding “worst case” conditions. Finding the worst case scenario to consider is certainly required, however, sometimes engineers can be guilty of needlessly piling on more factors than are necessary. And once you know the worst case, what do you do with it? Do you always follow it in the field, or do you allow exceptions? While some IEEE peer-reviewed papers layer on worst case factors to make sure that worst case scenarios are covered from any possible angle, another 2022 IEEE paper2 states that recent real-world test results indicate that actual arc flash levels are five to ten times less than those calculated by standard methods. At Castillo Engineering, we have seen recent client specifications requiring that the “Paukert” method developed in 1993 be utilized for calculating DC arc flash values. Also, some software packages let you choose between three major calculation methods: Maximum Power; Stokes and Oppenlander; and Paukert. Unfortunately for engineering number-crunchers, a 2020 IEEE paper3 states that “none of the available DC arc-flash models are applicable for a PV plant.”
While you can’t escape the AC grid energy at any time of the day, you can escape most of the DC energy by avoiding noon-time maintenance, even if avoiding peak power could result in possibly longer protective device clearing times. Is the goal to actually make useful arc flash labels or just pat ourselves on the back that we really covered the most extreme, worst case condition and ignored the other five elements of the Safety Pyramid.
At Castillo Engineering, we like using the absolute worst-case conditions for solar energy systems because you shouldn’t be routinely exposed to energized elements anyway, and maybe that red warning label will make you think twice about alternative methods to get the job done. Another factor is that the knowledge and skill set of any particular solar energy technician will always be an unknown. Fully compliant arc flash labels are complicated things, and the only formal training a solar energy technician receives in any given year may be the requisite 10-hour OSHA class, during which arc flash labels may have only been discussed for about 10 minutes. Even most experienced electricians who take the time to read an arc flash label are going to ponder, “Now what was that difference between those three boundaries again?”
Other recent articles discuss how arc flash labels are useful for selecting PPE for “routine” troubleshooting. This again seems to be an old-school, status-quo type of thinking that hasn’t considered either changes in work practices or perhaps designing systems a little bit differently to eliminate the need for that “routine” troubleshooting in the first place.
Not many solar energy entities manage all of the stages of a solar project - from initial design through to operations and maintenance. This explains some of their reluctance to spend anything extra on something like string monitoring instead of just the default design of monitoring only at the combiner box level. Even with only standard combiner box level monitoring, modern artificial intelligence systems can determine if there is a problem with one or more strings. That still will eventually require a site investigation, and every string in that combiner box will need to be checked because the AI is not granular enough. As a result, you have added at least an hour to the on-site time compared to knowing the exact string(s) that are an issue.
Alternatively, many European entities seem to have completed the cost justification math for string monitoring. That additional instrumentation can pay for itself not only by eliminating a “routine” safety hazard, but also by both decreasing the number of truck “rolls” and decreasing the amount of time spent on site when those trucks are dispatched.
Some inverters now also provide IV curve tracing as a built-in feature. This inanimate test equipment does not get tired and fed up after a day of testing in the extreme heat or freezing cold and decides to under-report some discovered anomalies because the necessary investigation would prolong the day.
Many ac circuit breakers in the main collection panels can also be procured with full Modbus sensors and communication. It sounds like a luxury up front, but when an electrician overtorques some inverter cables at the circuit breakers and localized heating starts causing intermittent breaker trips as well as lengthy electrician visits to take multiple clamp-on meter readings, the cost seems suddenly slight. Want to test a bus voltage that requires suiting up and opening the rear of a panelboard? For less than $500, why not install indicating lights and test jacks that are accessible from the outside of the panelboard such as this one from Grace Technologies: www.graceport.com.
There are five other pyramid levels that can just about eliminate that “routine” troubleshooting need. A change in mindset is all that is required.
Stay tuned for Part 3 of our Solar Arc Flash Safety Series, in which we will cover how exactly arc flash labels can become more effective. Do you have utility-scale solar design and engineering questions? Get in touch with one of our experts today.