ARCPRO and Duke heat flux calculator

[ANSHSUSHAN]

Hi Guys
This is Shan from EPCOR, I am working as Distribution P&C Engineer with Epcor and looking after Arc Flash Analysis of our Medium voltage(5kV-15kV & 25kV) primary distribution system. I have been using IEEE-1584 for arc flash analysis and it works fine up to 15KV but for 25KV the arc flash energy results using IEEE-1584 are very conservative and unrealistic. It doesn’t make sense to use IEEE-1584 for voltage higher than 15KV. I am curious to know about the approach that industries been adopted for voltages greater than 15kv Primary for open air and underground arc flash analysis. There are some tools available to perform analysis for voltage >15KV and one of the tool I am aware of is“Duke power heat flux calculator” module that Duke energy created for open air single phase analysis , would it be appropriate to use Duke model single phase open air analysis with multiplication factor of 2.8 to 3 applied for 3 phase analysis for three phase under ground analysis(for example for three phase pad mount cubicle).
I have also heard about ARCPRO tool for arc flash analysis, Do any one, in this forum have experience with using Duke model or ARCPRO for arc flash analysis and do these models work for close box configuration (For example: Energy analysis for three phase pad mount cubicles).
I am using Synregi with standard IEEE-1584 for arc flash analysis at our utility, I would appreciate some input in regards to what other Preferred software tools would be that could be used for arc flash analysis for medium voltage >15KV. I am thinking of buying ARCPRO modul but not sure whether it works for underground analysis too.

Will greatly appreciate some advise on this.

Regards
Shan

 

[mopractice]

Look at OSHA 1910.269 App E. In table 3 they describe which methods are acceptable for estimating incident energy - looks like Arc Pro is the only one approved for >15kV. Table 6, which was created using arcpro and is the same as NESC Table 410-2, shows fault currents and max clearing times for various incident energy levels. Using the values from tables 6 and/or 7 you can choose an incident energy level (4cal/cm) and then plot the fault current and clearing times on fuse coordination curves. In my case this showed that if we were working beyond a 100T fuse then we were not exposing workers to incident energies beyond the clothing we are providing.

 

[Skythian]

I would have reservations against using ARCPRO for arc flash calculations. As an example, the program requires arcing current value on its input. You have to calculate the value somehow before running analysis or make up and justify the number. IEEE 1584 provides formulae for arcing current calculations in LV systems. While using the IEEE formulae for arcing current it makes sense only being consistent and using the IEEE calculation procedure for incident energy and arc flash boundary calculations instead of mixing the methods.

 

[jghrist]

There is very little difference between bolted fault current and arcing current for medium voltage systems.

 

[Skythian]

Could you please quantify the difference? If difference was really small and negligible, why doesn't ARCPRO require bolted fault current on its input instead of arcing current?

 

[wbd]

Below is text of an email I received from Stephen Cress of ArcPro which answer some questions above.

As a co-developer of Arcpro, my apologies for the delayed response to Barry's e-mail. I would encourage anyone to ask us at Kinectrics if you have questions on Arcpro and we'll provide you with a factual answer. Call me at 416-207-6557 if you wish to discuss. Kinectrics is the Consulting Engineering and test lab where virtually all of the North American testing is done to establish ATPVs. We have conducted arc hazard assessments for some of North Americas largest utilities. We also provide Courses on the proper application of most of the arc hazard analysis techniques available.

Regarding the Factors: first please be sure you are using the latest set of factors from Disk 2 on your Arcpro CD. The factors referred to by Barry from the Help Menu have been refined to reduce the range based on comparisons with test data. When we are discussing the factors we are referring to this upgraded Table.

The basis for the factors is an engineering analysis of all published test data both from Kinectrics and IEEE and a comparison of this with Arcpro outputs. Arcpro models single arcs, either phase to phase or phase to ground, so without the factors it produces results applicable to single arc situations. Note that IEEE1584 models only 3-phase situations. Essentially the factors adjust the output of Arcpro to yield the best match to the 3-phase and in-box test data that exists. Feel free to call if you wish more detail.

On the subject of Arcpro algorithm and equations - there are 6 references in the Arcpro manual that provide the base peer reviewed engineering papers upon which the methodology is based. Unlike most of the other methods that are simple single equation curve fits to test data Arcpro is a finite element analysis iterative model of arc plasma, its conductivity and the radiation and convection from an arc. The theory and equations are readily available in these published papers. It is correct to say that the Arcpro equations are not simple - but they are available

Yes Arcpro requires the arcing current. On medium voltage systems it is reasonable to assume that this is equivalent to the bolted current. On low voltage systems the arcing current can be as low as 40% of the arcing current. A proper engineering analysis would involve putting the arc resistance in your symmetrical component model and determine the resulting arc current. Arcpro tells you what that arc resistance should be. Please beware with IEEE 1584 equations that compute arcing from bolted current. Under certain conditions the results will yield the result that arcing current is greater than bolted - clearly incorrect. A circuit analysis approach is more accurate. In our Course we show some simple engineering ways to determine arcing current.

The Lee formula - while valid for the test results it was deigned to emulate - is clearly incorrect for high voltage applications.

IEEE 1584 - while valid for relatively small arc gaps for which it was derived is clearly incorrect for large arc lengths ie certainly anything 2' or greater.

If any of these areas are of further interest please feel free to e-mail at steve.cress@kinectrics.com or call 416-207-6557.

 

[Skythian]

Regarding Stephen's comment that On medium voltage systems it is reasonable to assume that this is equivalent to the bolted current. What may seem reasonable to some is not necessarily seemingly reasonable to others. If the ARCPRO developer is so confident about the assumption, why does the program require the arcing current and not the bolted fault current on its input?

 

[stevenal]

Probably because they don't assume you will be using it at medium voltage. Altering the voltage does not affect the final result; until it is lowered to a value that will not sustain an arc across the gap. Like he said, at lower voltage, an iterative approach is needed using ArcPro along with your fault current program.
When will they provide a program that will run on 64 bit machines?

 

[Skythian]

If the ARCPRO assumption (the arcing current is equivalent to the bolted current) was right, than the arc voltage and impedance would be equal to zero and there would be no power released in the arc in the first place. Hence, it would be safe to assume there is is no hazard associated with the arc. A wise man once said that assumption is the mother of all screw ups, and apparently the ARCPRO a assumption is glaring example of it.

It is also important to note that,according to the program user manual, "the ARCPRO program calculates the thermal parameters associated with an electric power arc aligned vertically in air. The program does not consider arcs in any other physical configuration, [...]". The use of ARCPRO for estimating the arc hazard immediately becomes an engineer's best guess because situations where we have arcs "aligned vertically in air" are at best very limited.