Author 
Message 
ETcofamek

Post subject: Open air vs Inabox Posted: Tue Oct 31, 2017 6:28 am 

Joined: Wed Oct 18, 2017 12:24 pm Posts: 3

Hello all,
I'm new to Arc Flash, therefore I decided to run a few simulations to help me understand what are the effects of the differents parameters (time, working distance, etc.) in the IEEE 1584 equation on the resulting IE level.
Something I discovered through all the simulation I did is that the IE level will be greater when the arc is enclosed in a box than in open air. I would have instinctly think the opposite, because the flash is more ''free'' to expand in the open air than in a box.
1) Is that right to say that resulting IE will always be higher when the arc is in a box? 2) Why?
Thanks all ! Eric


Top 


Jim Phillips (brainfiller)

Post subject: Re: Open air vs Inabox Posted: Tue Oct 31, 2017 11:24 am 

Plasma Level 

Joined: Mon Sep 17, 2007 5:00 pm Posts: 1682 Location: Scottsdale, Arizona

The arc flash energy is focused/concentrated when it is directed out of the opening of a box. The energy for an arc flash in open air radiates spherically in many directions so think of it as being "diluted" and a smaller portion of it reaches the worker.
Welcome to the world of Arc Flash and the Arc Flash Forum. A very good group of knowledgeable people hang out here and are quite willing to answer questions and help out. Stop back with any questions or use the search since many have been addressed.
 Jim


Top 


ETcofamek

Post subject: Re: Open air vs Inabox Posted: Tue Oct 31, 2017 11:28 am 

Joined: Wed Oct 18, 2017 12:24 pm Posts: 3

Thanks for the quick answer !
It makes perfect sense.


Top 


PaulEngr

Post subject: Re: Open air vs Inabox Posted: Wed Nov 01, 2017 3:32 am 

Plasma Level 

Joined: Tue Oct 26, 2010 9:08 am Posts: 2173 Location: North Carolina

Heat transfer via thermal radiation (aka infrared) is several times more efficient than convection/conduction. That is why for instance you can stand next to a fire when the temperature is below freezing and feel nice and warm...it's radiation vs. convection. Just like light (which is also radiation), it will bounce off walls like an enclosure. You can tell that this is what's going on because the incident energy is proportional to the square of the distance in the open air model...it's the surface area of a sphere. If it were convective it would be proportional to the volume and thus proportional to the cube of the distance and the incident energy would fall off MUCH faster than it does.
In the arcinabox model because the thermal radiation bounces off the walls and gets more focused, the exponent drops down to less than 2 and since it is already more like a beam instead of a sphere, the radiation is much higher to start with. In a more refined model (IEEE 1584 is mostly based on data collected with distances under 6 feet) it would eventually become more spherical but as of right now this is not how it is modeled. This causes problems for instance with medium voltage switchgear with high currents where the predicted arc flash boundaries are completely unrealistic (hundreds of feet) but without a lot more data it's the best we have.


Top 


ETcofamek

Post subject: Re: Open air vs Inabox Posted: Wed Nov 01, 2017 5:39 am 

Joined: Wed Oct 18, 2017 12:24 pm Posts: 3

Thanks for the detailed information.
From the IEEE equations, there is a K factor for open air vs in a box when calculating the bolted short circuit current and another (different) K factor for open air vs in a box when calculating the Arc Flash current.
Will both current always be calculed using the same environnement factor or can we imagine a situation where the bolted current would have to be calculated in a box and the Arc flash current would have to be calculed in open air? And vice versa?
Thanks again
_________________  Eric


Top 


wbd

Post subject: Re: Open air vs Inabox Posted: Wed Nov 01, 2017 10:37 am 

Plasma Level 

Joined: Wed May 07, 2008 5:00 pm Posts: 873 Location: Rutland, VT

Quote: Will both current always be calculed using the same environnement factor or can we imagine a situation where the bolted current would have to be calculated in a box and the Arc flash current would have to be calculed in open air? And vice versa? It will always be one environment as I can't envision the fault starting as bolted fault in a box and then the box disappearing at the same time the arc becomes and arcing fault. Typically, the fault will start as an arcing fault not a bolted fault. On the other side, I can't envision an arcing fault in open air having a box appear around it.
_________________ Barry Donovan, P.E. www.workplacesafetysolutions.com


Top 


stevenal

Post subject: Re: Open air vs Inabox Posted: Wed Nov 01, 2017 3:39 pm 

Joined: Tue Jan 13, 2009 5:00 pm Posts: 596

ETcofamek wrote: From the IEEE equations, there is a K factor for open air vs in a box when calculating the bolted short circuit current ...
I believe you are mistaken. The k adder comes into play when calculating arcing current given bolted current. The standard recommends using commercial short circuit programs or the IEEE spreadsheet to get the bolted current. The box versus open air question comes up again later when calculating IE. I would not be switching at this point, since it is the arcing current that causes the exposure.


Top 


PaulEngr

Post subject: Re: Open air vs Inabox Posted: Thu Nov 02, 2017 3:19 am 

Plasma Level 

Joined: Tue Oct 26, 2010 9:08 am Posts: 2173 Location: North Carolina

Donâ€™t forget these are all empirical models...someone made up an equation either based on how we think reality works or based on the shape of the curve and used computer software to fine tune all the fudge factors. It is not a model based on physics.
Physically however every half cycle when the current passes through zero, the arc extinguishes. After that the air cools down. As the voltage increases during the next half cycle eventually the arc restrikes and the cycle repeats. Higher currents heat air more, higher voltages allow the arc to restrike sooner, and if the area around the arc is more confined the air cools less so it restrikes sooner (electrical conductivity is temperature dependent). There is a lot going on here and the models just curve fit because the theory hasnâ€™t caught up yet. The result is that the arcing current average will be higher in arcinabox and so will the incident energy as described earlier. The best curve fit on the current uses the K factor. The best curve fit on the normalized incident energy uses that and to model the focusing effect the exponent x is adjusted.
This is pretty typical modeling for a lot of thermal and hydrodynamic processes. Even where finite volume and/or radiosity models can be done, often you need a high horsepower computer or supercomputer and hours or days of calculation time with $50,000 of software and training involved to do one calculation. The investment usually makes lab testing less costly, and you couldnâ€™t for instance practically do a study on a small plant with say 150 buses that takes 6 months to run the calculations even if the theory caught up to the empirical work. That kind of model just tends to get used to calidate the empirical one anyway. But as I said were not there yet. Robert Lang has developed a time based model that is closer but more computer intensive because it takes hundreds of calculations per data point, and a another simplified one that beats IEEE 15842002 that might show up in the next IEEE 1584 standard. The time series model is riddled with fudge factors too but gets closer to a physics model in terms of underlying equations.


Top 


elihuiv

Post subject: Re: Open air vs Inabox Posted: Mon Nov 13, 2017 2:25 pm 

Sparks Level 

Joined: Sun Jan 04, 2009 5:00 pm Posts: 285 Location: Louisville, KY

Real equipment configuration makes a difference too by deflecting energy and focusing energy more or less than the calculations.
Additionally, the equipment can constrain the arc length or perpetuate the extinguishing of the arc.
Having about 170 incident investigations under my belt and about 50 are published with a group in an IEEE paper, typically IEEE 1584 is still conservative.
Also, arc ratings are conservative if the current is >8000A and might be a little under kill if the current is lower but this is an over generalization. Longer time is worse on garments than higher current due to ablation changes in the system. Distance also changes the convection portion. We test garments at 12 inches while the body is often farther away. Lots of evidence that we are very conservative, I'm working on a paper but some of the preliminary work is in the Handbook of Fire Resistant Textiles, Woodhead Publishing. My chapter is on arc flash.
Hugh Hoagland


Top 


Flash

Post subject: Re: Open air vs Inabox Posted: Mon Nov 13, 2017 2:31 pm 

Joined: Sat Nov 21, 2009 10:23 pm Posts: 124 Location: Ohio

A couple of question about your answer below: KNOWN: The heat energy transmission via radiation is the absolute T to the fourth power. Whereas convection heat energy transfer is Delta T x Area x h, "h" being a coefficient based upon the medium and orientation. Question: This is not a rebuttal I merely want to get an accurate grasp of this topic. Down below in you post you mention volume and and the exponent dropping to less than 2. The convection energy is based upon area per the above formula, not volume? The delta T for radiation is "T to the 4th", which is far greater than the "less than 2"? Question: What are the proportions of energy, in other words, what percent is by convection and what percent is by radiation at a 18"24" working distance. PaulEngr wrote: Heat transfer via thermal radiation (aka infrared) is several times more efficient than convection/conduction. That is why for instance you can stand next to a fire when the temperature is below freezing and feel nice and warm...it's radiation vs. convection. Just like light (which is also radiation), it will bounce off walls like an enclosure. You can tell that this is what's going on because the incident energy is proportional to the square of the distance in the open air model...it's the surface area of a sphere. If it were convective it would be proportional to the volume and thus proportional to the cube of the distance and the incident energy would fall off MUCH faster than it does.
In the arcinabox model because the thermal radiation bounces off the walls and gets more focused, the exponent drops down to less than 2 and since it is already more like a beam instead of a sphere, the radiation is much higher to start with. In a more refined model (IEEE 1584 is mostly based on data collected with distances under 6 feet) it would eventually become more spherical but as of right now this is not how it is modeled. This causes problems for instance with medium voltage switchgear with high currents where the predicted arc flash boundaries are completely unrealistic (hundreds of feet) but without a lot more data it's the best we have.


Top 


elihuiv

Post subject: Re: Open air vs Inabox Posted: Mon Nov 13, 2017 2:39 pm 

Sparks Level 

Joined: Sun Jan 04, 2009 5:00 pm Posts: 285 Location: Louisville, KY

I'll have to check the data. One paper did 18 and we test at 12 but they just looked at total energy. We found at 12 inches the convection vs. IR is about 50/50 in the open arc test. I would expect a box to do more convection but our arc length is long so I'm not sure.
If you watch the high speed video, the "fireball," which would be one indicator of the convection front, does not typically extend beyond about 1214 inches in an open arc test like ASTM F1959. We didn't know that when we picked 12 inches over 20 years ago but it might explain why the test works so well with IEEE 1584 and calculations.
We are working to do more work on measuring this due to a recent attack on the standard. The attack is trying to get research money and push a European test which has other issues but it helps us learn.
Hugh


Top 


Jim Phillips (brainfiller)

Post subject: Re: Open air vs Inabox Posted: Tue Nov 14, 2017 3:59 am 

Plasma Level 

Joined: Mon Sep 17, 2007 5:00 pm Posts: 1682 Location: Scottsdale, Arizona

ETcofamek wrote: Thanks for the detailed information.
From the IEEE equations, there is a K factor for open air vs in a box when calculating the bolted short circuit current and another (different) K factor for open air vs in a box when calculating the Arc Flash current.
Will both current always be calculed using the same environnement factor or can we imagine a situation where the bolted current would have to be calculated in a box and the Arc flash current would have to be calculed in open air? And vice versa?
Thanks again The configurations for the 2002 model of IEEE 1584 are only an arc flash in a box vs. air with the test electrodes in a vertical configuration i.e. pointed down. A few different box sizes were used depending on the â€œequipmentâ€. The next edition of IEEE 1584 is moving closer to completion. Myself with a few other IEEE 1584 officers are presently reviewing the many comments from the first balloting of the next edition of IEEE 1584. Once this is all completed and the next edition is finally approved and published, there will be many more options regarding the configurations supported by a few thousand new tests. Btw, regarding the K factors. For the arcing current there is only one K factor for the arc flash in an enclosure vs. open air. This adjusts the effect of constraining the arc which affects the arcing current. The incident energy calculations has two K values. K1 is again for the enclosure vs. open air and K2 is for grounded vs. ungrounded. The original test for the 2002 model show a slight variation between grounded and ungrounded. We found in the present research (and subsequent new model) that this effect is not as significant as first though (only during arc initiation). Here is an article that I wrote a while ago about the upcoming next edition of IEEE 1584201X My usual required disclaimer: although Iâ€™m ViceChair of IEEE 1584 and a member of the Steering Committee for the IEEE/NFPA Arc Flash Collaborative Research Project, the above are my personal comments and may or may not reflect the views of any particular standards organization


Top 


jmille3

Post subject: Re: Open air vs Inabox Posted: Thu Jun 21, 2018 10:29 am 

Joined: Thu Apr 26, 2018 4:05 pm Posts: 4

In the arcinabox model because the thermal radiation bounces off the walls and gets more focused, the exponent drops down to less than 2 and since it is already more like a beam instead of a sphere, the radiation is much higher to start with. In a more refined model (IEEE 1584 is mostly based on data collected with distances under 6 feet) it would eventually become more spherical but as of right now this is not how it is modeled. This causes problems for instance with medium voltage switchgear with high currents where the predicted arc flash boundaries are completely unrealistic (hundreds of feet) but without a lot more data it's the best we have.
This is the problem I have at my site. I have incident energy for 4.16 and 6.9kV switchgear over 100 cal at the working distance and flash protection boundaries over 300'. This is my question. I'm using the "switchgear" equipment type for the distance exponent to calculate the incident energy and the flash boundary. Using switchgear as the type gives me a large flash boundary. Can I use "open air" equipment type to calculate the flash boundary and "switchgear" type for the incident energy at the working distance? I'm trying to figure out a way to get a realistic flash boundary because the energy someone will experience if they are standing say 20' away off to the side and not directly in front of the door has to be less than if they were directly in front of the door. Right now with a 300+ foot flash boundary it makes it difficult to rope off an area that big and it just seems unrealistic to be honest. All my input data is correct and my trip times are less than 2 seconds so the large boundaries are not a function of input data error.


Top 


wbd

Post subject: Re: Open air vs Inabox Posted: Fri Jun 22, 2018 8:26 am 

Plasma Level 

Joined: Wed May 07, 2008 5:00 pm Posts: 873 Location: Rutland, VT

Are you using the actual available utility fault current and not the bolted fault current with an infinite bus?
_________________ Barry Donovan, P.E. www.workplacesafetysolutions.com


Top 


PaulEngr

Post subject: Re: Open air vs Inabox Posted: Sun Jun 24, 2018 5:35 am 

Plasma Level 

Joined: Tue Oct 26, 2010 9:08 am Posts: 2173 Location: North Carolina

jmille3 wrote: [i] This is the problem I have at my site. I have incident energy for 4.16 and 6.9kV switchgear over 100 cal at the working distance and flash protection boundaries over 300'. This is my question. I'm using the "switchgear" equipment type for the distance exponent to calculate the incident energy and the flash boundary. Using switchgear as the type gives me a large flash boundary. Can I use "open air" equipment type to calculate the flash boundary and "switchgear" type for the incident energy at the working distance? I'm trying to figure out a way to get a realistic flash boundary because the energy someone will experience if they are standing say 20' away off to the side and not directly in front of the door has to be less than if they were directly in front of the door. Right now with a 300+ foot flash boundary it makes it difficult to rope off an area that big and it just seems unrealistic to be honest. All my input data is correct and my trip times are less than 2 seconds so the large boundaries are not a function of input data error. Going STRICTLY BY THE STANDARD obviously we don't deviate. However the standard has the "equipment table" which is supposed to help the end user model conditions. Obviously the equipment table contains certain assumptions and becomes invalid as we move away from those assumptions. But also as we move away from those assumptions, we are no longer constrained, nor guided by the standard. So an opinion from someone on the internet is probably not the best source of an answer to your question but perhaps if there was at least a relevant published article on the subject that would help justify the direction you are suggesting. I believe some of the arc flash research Mike Lang did at Mersen answers your question. IEEE 1584 test cases in the 2002 edition only tested out to around 6 feet from the arc. Mike was tested arcing with the typical glastic barriers between arcs among other things and found in the testing that beyond around 6 feet or so, the incident energy more or less began to behave as "open air". Searching for arc flash and barriers should turn up the correct technical paper if I've got the researcher wrong. As to the result, look closely at the equations. The "switchgear" enclosure is fairly large and representative of "typical" switchgear. With a large and very deep enclosure the thermal radiation of necessity is going to be fairly focussed into a narrow cone when it is close in compared to say a shallow flat box like a lighting panel. So if you look carefully at the "math" you will notice that the exponent is almost linear. This explains why the results don't match expectations and we get crazy high arc flash boundaries like you are describing...it's nearly linear with distance at best. So either someone standing "off axis" could stand much closer and be completely unharmed, or the rate that the incident energy falls off should be much faster than predicted. As I said there is no guidance unless you make your own determination. I can also direct you to this one: https://ieeexplore.ieee.org/document/6164450/Going through the raw data in cases where workers were underprotected as predicted by IEEE 1584 but still wearing some sort of arc flash PPE about half the time they did not receive 2nd degree or greater burns on the face/chest area. Finally I can suggest that arc flash as we handle it today is purely thermal radiative energy. This means that pretty much anything that can block infrared radiation such as a thin piece of sheet metal siding/roofing is enough to block the arc flash. For this reason you may as well consider the walls and/or equipment lineup as protection. I know that the arc flash blanket folks have been using this and the only major problem they've found in testing is that there is a tendency for hot gases to roll up around the sides of the blanket so they recommend still wearing arc rated PPE for protection from this. Or alternatively you may want to investigate arc flash blankets since they are so effective at containing the effects even in very confined areas such as vaults.


Top 


jmille3

Post subject: Re: Open air vs Inabox Posted: Mon Jun 25, 2018 2:00 pm 

Joined: Thu Apr 26, 2018 4:05 pm Posts: 4

I'm afraid to use "switchgear" for the incident energy at the working distance and "open air" for the flash boundary because it's not supported by the standard. If I do this in my study and somebody gets hurt when they were outside the boundary (but really inside the boundary using "switchgear" as the equipment type) so many questions would come up as to why I calculated the flash boundary using "open air" for a switchgear. I work at a generation facility and most of my peers have large flash boundaries too because our short circuit currents are high and our feeder breaker clearing times are slow. The problem at my site is we have switchgear that is essentially on an open deck with flash boundaries over 300'. The site doesn't want to evacuate the entire turbine building if they need to rack a breaker with the bus energized. This is why I am trying to see if there is a more realistic solution for the flash boundaries.


Top 


PaulEngr

Post subject: Re: Open air vs Inabox Posted: Tue Jun 26, 2018 10:18 am 

Plasma Level 

Joined: Tue Oct 26, 2010 9:08 am Posts: 2173 Location: North Carolina

Flash wrote: A couple of question about your answer below: KNOWN: The heat energy transmission via radiation is the absolute T to the fourth power. Whereas convection heat energy transfer is Delta T x Area x h, "h" being a coefficient based upon the medium and orientation. Question: This is not a rebuttal I merely want to get an accurate grasp of this topic. Down below in you post you mention volume and and the exponent dropping to less than 2. The convection energy is based upon area per the above formula, not volume? The delta T for radiation is "T to the 4th", which is far greater than the "less than 2"? Question: What are the proportions of energy, in other words, what percent is by convection and what percent is by radiation at a 18"24" working distance. PaulEngr wrote: Heat transfer via thermal radiation (aka infrared) is several times more efficient than convection/conduction. That is why for instance you can stand next to a fire when the temperature is below freezing and feel nice and warm...it's radiation vs. convection. Just like light (which is also radiation), it will bounce off walls like an enclosure. You can tell that this is what's going on because the incident energy is proportional to the square of the distance in the open air model...it's the surface area of a sphere. If it were convective it would be proportional to the volume and thus proportional to the cube of the distance and the incident energy would fall off MUCH faster than it does.
In the arcinabox model because the thermal radiation bounces off the walls and gets more focused, the exponent drops down to less than 2 and since it is already more like a beam instead of a sphere, the radiation is much higher to start with. In a more refined model (IEEE 1584 is mostly based on data collected with distances under 6 feet) it would eventually become more spherical but as of right now this is not how it is modeled. This causes problems for instance with medium voltage switchgear with high currents where the predicted arc flash boundaries are completely unrealistic (hundreds of feet) but without a lot more data it's the best we have. Not to touch Eli's comments but we're talking about two slightly different things here. When calculating incident energy and accounting for distance the exponent is 2. This would be true regardless of whether it is for an arc flash or a candle in a dark room. It is entirely due to the area of a sphere (pi*R^2) and we are interested in the heat flux at a point on that sphere so hence distancesquared. This is on top of the usual temperature delta exponent which is the 4th power. This factor is "baked in" to the incident energy calculation before it is adjusted for time, enclosure shape, and working distance. Similarly convection would have to apply the similar factors except that convection should fall off with the cube of the area since hot gasses are rapidly dissipated. The distance exponent (square of distance) decreases if instead of a point source emitter we somehow shape and diect the arc flash energy by placing it in some kind of box or tube where we focus the radiation with "mirrors" (walls).


Top 


jmille3

Post subject: Re: Open air vs Inabox Posted: Mon Jul 02, 2018 6:23 am 

Joined: Thu Apr 26, 2018 4:05 pm Posts: 4

PaulEngr wrote: In the arcinabox model because the thermal radiation bounces off the walls and gets more focused, the exponent drops down to less than 2 and since it is already more like a beam instead of a sphere, the radiation is much higher to start with. In a more refined model (IEEE 1584 is mostly based on data collected with distances under 6 feet) it would eventually become more spherical but as of right now this is not how it is modeled. This causes problems for instance with medium voltage switchgear with high currents where the predicted arc flash boundaries are completely unrealistic (hundreds of feet) but without a lot more data it's the best we have. I see in the standard how the distance exponent was determined using test data with distance data under 6 feet. Is there any published data available to support that the radiation would become more spherical at longer distances? I haven't seen the draft copy of the 2018 version but can anybody tell me if the new version changes the way the distance exponent was determined?


Top 


PaulEngr

Post subject: Re: Open air vs Inabox Posted: Wed Jul 04, 2018 2:09 pm 

Plasma Level 

Joined: Tue Oct 26, 2010 9:08 am Posts: 2173 Location: North Carolina

jmille3 wrote: PaulEngr wrote: In the arcinabox model because the thermal radiation bounces off the walls and gets more focused, the exponent drops down to less than 2 and since it is already more like a beam instead of a sphere, the radiation is much higher to start with. In a more refined model (IEEE 1584 is mostly based on data collected with distances under 6 feet) it would eventually become more spherical but as of right now this is not how it is modeled. This causes problems for instance with medium voltage switchgear with high currents where the predicted arc flash boundaries are completely unrealistic (hundreds of feet) but without a lot more data it's the best we have. I see in the standard how the distance exponent was determined using test data with distance data under 6 feet. Is there any published data available to support that the radiation would become more spherical at longer distances? I haven't seen the draft copy of the 2018 version but can anybody tell me if the new version changes the way the distance exponent was determined? I believe Mike Lang did some testing in this regard. See the Madsen web site for free papers. Wilkins has also proposed an alternate model using two discs that fits the data better and the fall off changes with distance. It's fairly obscure since I think it was published as the Wilkins simplified model on an IEEE web site. It gets mentioned in several other conference papers. Overall it's a major competitor to the IEEE 1584 empirical model since it is more physics based and simpler overall. Wilkins also developed the best model to date, the time domain model which is computationally time consuming but gives the best fit to test data and is for the most part physics based.


Top 


elihuiv

Post subject: Re: Open air vs Inabox Posted: Thu Jul 05, 2018 8:54 am 

Sparks Level 

Joined: Sun Jan 04, 2009 5:00 pm Posts: 285 Location: Louisville, KY

Most everyone uses the IEEE 1584 method since it is easy and already in software.
ArcPro is usually only used for overhead lines as it has an arc physics model and data to back it up. IEEE 1584 overshoots AFB on medium and high voltage.
We recommend ArcPro on power lines and IEEE 1584 on everything else unless you want to do ArcPro with factors in a spreadsheet.
This is now easier with batch processing in ArcPro 3.0 in a spreadsheet.


Top 



You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot post attachments in this forum

