With no utility data, How do I model infinite bus in SKM?

[upquark]

When working on an arc flash study, sometimes, I get a delayed or no response from the utility company and, sometimes, I get the "infinite bus" calculation from the utility company, which I can calculate myself so it's useless. Either way, when I lack genuine utility data, how do I model the infinite bus assumption in SKM? The infinite bus assumption, as I understand it, is the current a transformer will output on the secondary when an infinite amount of current is present on the primary.

As I see it, I can model this two different ways: (1) with the utility transformer in the model and an extremely high contribution from the utility and (2) excluding the transformer and modeling a utility that contributes the equivalent as the transformer would at infinite bus. Which method is preferable and what values would I input in SKM? (see screenshots)

 

[jim dungar]

Both methods are acceptable.

I use whichever one more accurately depicts the installation and information.
For situations where the transformer information is known (i.e. customer owned transformer), I will assume an extremely high primary side fault current (e.g. 99,999).
For situations where the transformer is owned by the utility I will use a high secondary fault current, which is usually available from the utility.

Infinite bus fault currents should never be the only values used in Arc Flash Incident Energy calculations, alternative current values must be considered.

 

[upquark]

I posed the same question to SKM Technical Support and received the following response:

Hi Ryan,

We recommend entering a very large value, such as 9999 for the 3P utility contribution field to model an infinite source. SLG values are not considered in Arc Flash calculations, only 3P values are. You can use the standard X/R value that the software provides.

For Arc Flash calculations, an infinite source is NOT recommended as it will result in faster clearing times which may not be realistic.

See attached for a paper from EATON.

I cannot attached the paper SKM refers to because the size of the file exceeds the forum's rule for attachments but the paper can be found at http://www.scribd.com/doc/123704270/Impact-of-Available-Fault-Current-on-Arc-Flash-Calculation.

 

[upquark]

Infinite bus fault currents should never be the only values used in Arc Flash Incident Energy calculations, alternative current values must be considered.

What alternative current values are you referring to?

 

[jim dungar]

What alternative current values are you referring to?

Decisions like this are why arc flash studies should not be DIY. The engineering firm, I work for, has written procedures for how to determine alternative (i.e. more realistic) values.

The intent is to find the fault current scenario that yields the highest incident energy (which is often a low fault current levels). You then need to decide if this fault current are likely to occur.

 

[upquark]

Decisions like this are why arc flash studies should not be DIY. The engineering firm, I work for, has written procedures for how to determine alternative (i.e. more realistic) values.

The intent is to find the fault current scenario that yields the highest incident energy (which is often a low fault current levels). You then need to decide if this fault current are likely to occur.

According to the paper SKM referred to, the arc flash hazard analysis should consider two scenarios when you do not have actual utility data: (1) using the infinite bus assumption and (2) using 50% of the maximum short-circuit current the transformer can output. Of course, by this method, I have to model according to what I called SCENARIO 2 above by not modeling the utility transformer.

The paper points out that many engineers do not do this and understandably so as my four-scenario project now has eight scenarios.

 

[steve66]

According to the paper SKM referred to, the arc flash hazard analysis should consider two scenarios when you do not have actual utility data: (1) using the infinite bus assumption and (2) using 50% of the maximum short-circuit current the transformer can output. Of course, by this method, I have to model according to what I called SCENARIO 2 above by not modeling the utility transformer.

The paper points out that many engineers do not do this and understandably so as my four-scenario project now has eight scenarios.

I always use two values for fault current, and I aways check the tripping point on the TCC curve, and make sure it isn't right on the edge of the knee. If it is, a small change in fault current can make a drastic change in the catagory.