I hate to start a new thread but I did not see this directly discussed. Doing arc flash calculations I have a main 2000 amp switchboard that is a HRC 1 at both 100% and 85% arcing current. My question is why stop at 85%. If the actual fault level is just a little lower, causing the arcing current to be around 80%, it knocks me out of the instantaneous portion of the breaker curve and quickly changes the HRC to "extreme" instead of HRC 1. So if I am to assume some variable can easily cause this situation to happen, which seems vary plausible, then what do I do? To make a very general statement, it seems that almost every piece of equipment should be HRC extreme as you just have to determine what variables will cause your fault current to occur just before the instantaneous portion of the breaker curve.

I always plan on the available bolted fault current being lower than any value I have been given.

I select multiple fault current source values, (i.e. high, middle and low utility values, and all motors off) as well as letting the software compare the 100% and 85% results.

Multiple Scenarios

In SKM and other programs, they call these Operating Scenarios. Most programs will allow you to run different scenarios and then produce labels that have the maximum arc flash energy.

As JDB pointed out, you need to calculate the energy level at 100% and 85% Iarc current for different operating conditions of your facility. Be aware that it may be very difficult trying to get the minimum value of utility fault current. They (utility) usually will provide you with the maximum but not the minimum. Furthermore, verify that the maximum fault current they provide you is for the transformer that is actually installed, not the maximum fault current if they install in the largest transformer in their inventory. As you can see from your calculations, it makes a big difference.

Arc Tolerance

I use SKM and have similar situations. Using the default options (85%), most of AF amps were cleared by the instantaneous settings and yielded Cat 0 values. However, when viewed as TCCs (Time Current Curves) many AF amps were very close to the instantaneous.

In SKM, Arc Flash Evaluation, Options, Arcing Tolerances, I changed -15% to -20% or 25% and re-ran the study. The new values were either Cat 2 or Cat 3.

I created a Custom Label that has both the required calculated Incident Energy value (using -15% Arcing Tolerance) and the requirement to use a higher PPE. I also modified the "Flash Hazard Boundary" similarly.

You can put anything you want on the label, as long as it includes the basic requirements.

One must realize that there is an infinite number of AF energy calculations that you can perform. If you change arcing fault current or the working distance, you'll get different results. You have to settle on one method or situation and stick to it. That is why we use the defaults set in SKM when we run our studies. These appear to be the "generally accepted" by the majority of consultants and protection engineers. The one area that we deviate on is selecting the upstream mis coordination option. See the attached screen shot.

Our company has their own standards which have been vetted by our legal department. We felt it would be better, in a court of law, if we could defend our choices.

We do use a 2 sec cut-off, and <=240V, but our cleared fault threshold is greater than 80%.

Arc flash equations are exact. The parameters we plug-in in these equations are NOT. They depend on the experience and criteria of the electrical engineer performing the study. I believe than all engineers will defend their choice as the most reasonable one. One might believe that a 75% cleared fault threshold is appropriate but other opts for 80%. One believes that the maximum arcing time should be 2 seconds and another thinks that, to be safe, we should extend that time to 3 seconds, and so on. There is a great variance in opinion and criteria. I also believe that arc flash results will have a normal distribution with a standard deviation with respect to the results. A problematic situation could result if the engineer chooses parameters which places his results way out of the standard deviation. I believe that we should have our opinion and justify it with passion. But I accept the diversity of opinions and experiences of other engineers and keep open minded to analyze their choices for this parameters. I would recommend, nevertheless, as much as you can, that every decision you make is backed up by a reputable standard. So if you are faced with a law suit, you can justify your choices with a consensus standard. Happy arc-flashing!

Reading the standard explains this. The test data exhibited a bimodal distribution. Adding the arcing current at 85% increased the confidence interval to around 95%.

Even if you get different incident energies at various currents it does not mean those are likely.