Short Circuit Coordination And Arc Flash Studies

How To Actually Run A Short Circuit Coordination And Arc Flash Studies

Most people approach these studies backwards. They grab ETSP or SKM and start clicking buttons without understanding what the software is doing between each input. That will get you a PDF report that looks professional but is built on garbage assumptions. I have seen it happen more times than I can count. Let me walk you through the real process, including the part where everything goes wrong and how to fix it.

What Short Circuit Coordination And Arc Flash Studies Actually Means

Short circuit coordination studies determine whether your overcurrent devices will clear a fault fast enough before upstream devices trip. If they don't, you get a cascading failure where a breaker down at the load end never opens because the feeder main trips first. Everything downstream goes dark. The arc flash study uses those coordination results plus incident energy calculations to tell workers what PPE category they need when they open an energized panel. The two studies feed each other. You need accurate fault currents from the coordination study to run the arc flash. You need proper time-current curves from the coordination to calculate incident energy. It is a loop, not a linear checklist.

Step 1: Gather the actual equipment data. Not the nameplate ratings from 1998. I need manufacturer test data, impedance values, and most importantly, the actual current limits of breakers. The field almost never matches the design. Transformers have different impedances than the standard tables. Motors contribute fault current during the first few cycles, and that matters for high-impedance systems. Cable lengths and sizes affect the available fault current at every point. I once spent three days arguing with a contractor who insisted the new switchgear was fed by a 750 kVA transformer when the drawing clearly showed a 1500 kVA unit. The drawings were three years old and had been updated after a renovation nobody documented properly. Step 2: Build the one-line model correctly. This sounds simple until you realize most models fail because people leave out components that matter. Every motor above 50 HP needs to be entered as an impedance contributing to fault current. Every generator, including standby and UPS systems, must be modeled. If your facility has any form of distributed generation, it changes the fault current profile completely and in unpredictable directions. Use actual cable data, not generic estimates. The software defaults are wrong more often than you would believe. Step 3: Set the protective device characteristics. This is where I see the most damage. Instantaneous trip settings, time delays, ground fault pickup and time, all of it needs to match what is actually in the field. A common mistake is modeling a breaker with an instantaneous setting of 10 times rated current when the field engineer adjusted it to 6x during a previous troubleshooting session. The software will predict coordination that does not exist in reality. Go and pull the settings. If the manufacturer makes that impossible, send an inspector and photograph every dial position and setting record.

The Hard Part: Making It All Work Together

Once your model is built, run the short circuit calculation first. Check that the results make sense against what you would expect. A 480-volt bus fed from a 1500 kVA transformer with standard impedance should show roughly 17 kA available fault current. If your software is giving you 45 kA, something is wrong in the model. Most likely you have a transformer impedance that is too low, or you missed entering a current-limiting reactor or fuse that is actually in the circuit. After the fault currents are validated, run the coordination overlay. Look at the time-current curves. Every downstream device should clear its fault before the upstream device sees it. The gap should be at least 0.3 seconds between curve intersections for circuit breakers, or 0.5 seconds if you want margin for tolerance variations. Standard relay and breaker tolerances eat into that gap quickly.

I ran into a specific problem last year on a hospital upgrade project. The existing main-tie-main configuration meant that during normal operation, the tie breaker was closed and both transformers were feeding the same bus. When I modeled the fault study with the tie closed, the fault current more than doubled at every point downstream. This meant every overcurrent device in the facility needed to be checked against this higher available fault current, and several breakers were actually underrated for the tie-closed condition. The original design study had only modeled the normal open-tie scenario. We had to replace eleven breakers that were fine for open-tie but would fail under close-tie fault conditions. The cost was significant, but catching it on paper was infinitely cheaper than finding out the hard way. With the updated fault currents, rerun the coordination study. Then move to the arc flash calculation. IEEE 1584-2018 is the current standard, and it changed the equations significantly from the 2002 version. The older equations tended to overestimate incident energy in some cases and underestimate in others. Make sure your software is using the correct equations for the equipment type you are analyzing. Arc flash calculations depend heavily on whether you are dealing with a circuit breaker versus a fused disconnect, the gap distance between conductors, and the enclosure type.

Common Pitfalls That Ruin These Studies

The biggest issue I encounter is outdated models. Facilities get modified constantly. New equipment is added, old equipment is removed, relocations happen without updating the documentation. A study based on five-year-old drawings is essentially a guess. The best approach is to do a quick field verification pass before you start modeling. Walk the switchgear rooms, check labels, verify that what is on the one-line matches what is actually installed. This usually takes half a day for a medium-sized facility and prevents countless errors. Another pitfall is ignoring the contribution from motors. In industrial facilities with large compressors, pumps, or fans, the motor fault current contribution can be 30 to 50 percent of the total available fault current during the first two cycles. If you are analyzing coordination for high-magnitude faults near motor terminals, this matters a lot. For low-level ground faults further downstream, it does not matter as much. Know which regime you are in. System grounding configuration also gets overlooked. Solidly grounded wye, resistance grounded, ungrounded, high-resistance grounded, low-resistance grounded. Each produces very different fault current levels and affects both the short circuit and arc flash results differently. The arc flash incident energy for a ground fault in a high-resistance grounded system can be dramatically lower than in a solidly grounded system, but the ground fault protection scheme needs to be designed for that reality.

Here is something most people do not account for: the arc flash boundary changes with the clearing time of the protective device. If you have a downstream fault that your main breaker clears in 2.5 seconds because the instantaneous element is not calibrated properly, the incident energy at the boundary could be several times higher than if the fault cleared in 0.2 seconds. This is why the coordination study is not optional. It is the foundation of the arc flash analysis. Tighten your coordination, and you often reduce the required PPE category substantially.

What The Software Can And Cannot Do For You

ETSP, SKM PowerTools, CYME, and similar tools are only as good as the data you put in. They will happily produce a polished report from garbage input. The software does not know if your transformer impedance is correct, if your cable lengths are accurate, or if your breaker settings match reality. You have to verify every major assumption. Cross-check fault currents against field measurements when possible. A clamp-on meter during a planned outage can confirm actual transformer impedance within reasonable accuracy.

One specific workaround I use for older facilities with no documented data: I measure the actual voltage drop during a motor start event. From the magnitude of the voltage dip and the known motor full-load current, I can back-calculate the effective source impedance. This gives me a field-validated starting point for the model that is usually much closer to reality than any assumption. It is not perfect, but it is better than guessing, and it caught a major error on a project where the documented transformer data was off by a factor of two due to a tape label swap in the switchgear room.

The arc flash labeling is the final deliverable, and it needs to be practical. Putting a label on every single component is technically correct but often unhelpful. Group panels that share the same incident energy level under a single label when the equipment configuration allows it. But do not group dissimilar equipment just to save on label costs. A 480-volt VFD panel and a 480-volt standard distribution panel fed from the same source can have wildly different arc flash levels due to the different fault current contributions and clearing times.

The studies also need periodic updating. I recommend a review cycle that matches your facility change cycle. Major equipment additions or replacements warrant an immediate study update. Routine annual reviews catch the drift from uncontrolled modifications. The NFPA 70E requires arc flash labels to be maintained and updated when changes occur, so there is a compliance driver beyond just safety. But the real value is in having a study that actually reflects what you have, because when something goes wrong, the report is your evidence that you acted reasonably.