Author 
Message 
calebgeorge

Post subject: CYME 8.0 Arc Flash  Heat Transfer Model Posted: Wed May 16, 2018 2:42 pm 

Joined: Wed May 16, 2018 12:31 pm Posts: 4

My apologies if this has been discussed elsewhere, my forum searching skills are not their best.
I was wondering if anyone on here had experience with CYME arc flash, particularly the Heat Transfer Model. I believe this is a new calculation methodology for CYME 8, it was not available in our revision of CYME 7.1.
It is listed under the transmission/distribution analysis section (of which is my arc flash interest).
The Heat Transfer Model is noted in the help files as being based on the Heat Flux calculations (This complex model was published in Electrical Safety Handbook, By John Cadick, Mary CapelliSchellpfeffer, Dennis K. Neitzel, Published by McGrawHill Professional, 2001, Chapter 3.52.). But the version of this Handbook I have (third edition) doesn't include this method.
This method also is not include in the OSHA Table 3.
I've reached out to CYME but thought I might also post here and gather some info from peers in the field.


Top 


calebgeorge

Post subject: Re: CYME 8.0 Arc Flash  Heat Transfer Model Posted: Thu May 17, 2018 10:39 am 

Joined: Wed May 16, 2018 12:31 pm Posts: 4

I just got confirmation from CYME that this is based on the Duke Heat Flux method, developed by Alain Privette P.E. at Duke Energy.
Which isn't mentioned in the OSHA standard and based on what I can read from this forum gives results lower than ArcPro


Top 


wbd

Post subject: Re: CYME 8.0 Arc Flash  Heat Transfer Model Posted: Fri May 18, 2018 7:23 am 

Plasma Level 

Joined: Wed May 07, 2008 5:00 pm Posts: 861 Location: Rutland, VT

Why not get ArcPro and use that to obtain the arcflash on a T&D system? This is recommended by OSHA and then you have no issues with trying to justify the use of some other method.
_________________ Barry Donovan, P.E. www.workplacesafetysolutions.com


Top 


PaulEngr

Post subject: Re: CYME 8.0 Arc Flash  Heat Transfer Model Posted: Sat May 19, 2018 8:14 pm 

Plasma Level 

Joined: Tue Oct 26, 2010 9:08 am Posts: 2174 Location: North Carolina

OSHA definitely did look at the Duke Heat Flux model. It's in the full 1500+ page update to 1910.269 proposed final rule. What it comes down to is that they compared the Duke Heat Flux model, ArcPro, and IEEE 1584 theoretical (aka Lee) models. Essentially their take is that Heat Flux numbers were low, Lee was high, and so ArcPro giving values right in the middle was just right. That's not to say that there aren't other possible models to use or that ArcPro necessarily produces correct results. There's no public data so it's hard to check any of them. It's just to say that this is how they justified selecting ArcPro. The current online version of 1910.269 Appendix E doesn't even mention the Heat Flux model at all nor the Lee model and only gives the recommended models to use.
Previously I believe CYME used to use a table lookup approach using the data out of NESC which is basically ArcPro so this is a bit of a step back for them. ETAP is also using something but so far the documentation just seems to point to some sort of way of calculating arc resistance/voltage at medium voltage and stops there. Presumably they would then simply assume 100% efficiency in converting to thermal energy and then apply the spherical distance model although the IEEE 1584 distance calculation is generic enough that it can be separated from the energy calculation and used independently. There are a couple papers at Battcon that recommend using this approach for DC arc modelling.


Top 


calebgeorge

Post subject: Re: CYME 8.0 Arc Flash  Heat Transfer Model Posted: Mon May 21, 2018 4:57 pm 

Joined: Wed May 16, 2018 12:31 pm Posts: 4

wbd wrote: Why not get ArcPro and use that to obtain the arcflash on a T&D system? This is recommended by OSHA and then you have no issues with trying to justify the use of some other method. Oh we use ArcPro currently for single arc in air evaluations, and use CYME to generate the fault current and clearing times. There was just some thoughts to simply the process by using a feeder modeling software we already use, and to evaluate it's arcflash capabilities. CYME has many methods to evaluate arcflash just not exactly what we want.


Top 


calebgeorge

Post subject: Re: CYME 8.0 Arc Flash  Heat Transfer Model Posted: Mon May 21, 2018 5:01 pm 

Joined: Wed May 16, 2018 12:31 pm Posts: 4

PaulEngr wrote: OSHA definitely did look at the Duke Heat Flux model. It's in the full 1500+ page update to 1910.269 proposed final rule. What it comes down to is that they compared the Duke Heat Flux model, ArcPro, and IEEE 1584 theoretical (aka Lee) models. Essentially their take is that Heat Flux numbers were low, Lee was high, and so ArcPro giving values right in the middle was just right. That's not to say that there aren't other possible models to use or that ArcPro necessarily produces correct results. There's no public data so it's hard to check any of them. It's just to say that this is how they justified selecting ArcPro. The current online version of 1910.269 Appendix E doesn't even mention the Heat Flux model at all nor the Lee model and only gives the recommended models to use.
Previously I believe CYME used to use a table lookup approach using the data out of NESC which is basically ArcPro so this is a bit of a step back for them. ETAP is also using something but so far the documentation just seems to point to some sort of way of calculating arc resistance/voltage at medium voltage and stops there. Presumably they would then simply assume 100% efficiency in converting to thermal energy and then apply the spherical distance model although the IEEE 1584 distance calculation is generic enough that it can be separated from the energy calculation and used independently. There are a couple papers at Battcon that recommend using this approach for DC arc modelling. The CYME lookup tables are still there, the OSHA and NESC tables, which are generated from ArcPro I believe. I had read previous posts where you stated this history of the OSHA ruling. I was doing some comparisons of Heat Flux vs ArcPro today and for many instances they were spot on, with the heat flux giving higher (more conservative) results. There were deviations where the fault current started getting higher (but there were also deviations where the clearing time was longer as well). It would be nice to know where they start to separate and what changes to cause them to separate (meaning for high fault current levels one is higher than the other, etc.).


Top 


PaulEngr

Post subject: Re: CYME 8.0 Arc Flash  Heat Transfer Model Posted: Wed May 23, 2018 7:28 pm 

Plasma Level 

Joined: Tue Oct 26, 2010 9:08 am Posts: 2174 Location: North Carolina

calebgeorge wrote: PaulEngr wrote: OSHA definitely did look at the Duke Heat Flux model. It's in the full 1500+ page update to 1910.269 proposed final rule. What it comes down to is that they compared the Duke Heat Flux model, ArcPro, and IEEE 1584 theoretical (aka Lee) models. Essentially their take is that Heat Flux numbers were low, Lee was high, and so ArcPro giving values right in the middle was just right. That's not to say that there aren't other possible models to use or that ArcPro necessarily produces correct results. There's no public data so it's hard to check any of them. It's just to say that this is how they justified selecting ArcPro. The current online version of 1910.269 Appendix E doesn't even mention the Heat Flux model at all nor the Lee model and only gives the recommended models to use.
Previously I believe CYME used to use a table lookup approach using the data out of NESC which is basically ArcPro so this is a bit of a step back for them. ETAP is also using something but so far the documentation just seems to point to some sort of way of calculating arc resistance/voltage at medium voltage and stops there. Presumably they would then simply assume 100% efficiency in converting to thermal energy and then apply the spherical distance model although the IEEE 1584 distance calculation is generic enough that it can be separated from the energy calculation and used independently. There are a couple papers at Battcon that recommend using this approach for DC arc modelling. The CYME lookup tables are still there, the OSHA and NESC tables, which are generated from ArcPro I believe. I had read previous posts where you stated this history of the OSHA ruling. I was doing some comparisons of Heat Flux vs ArcPro today and for many instances they were spot on, with the heat flux giving higher (more conservative) results. There were deviations where the fault current started getting higher (but there were also deviations where the clearing time was longer as well). It would be nice to know where they start to separate and what changes to cause them to separate (meaning for high fault current levels one is higher than the other, etc.). I'm not sure what specific data points were analyzed. From what I could gather out of the discussion a lot of the commentary came from Neal if I recall correctly what is attributed to anyone. I know that OSHA was really looking at the upper end of the medium voltage range, 1535 kV. This makes sense somewhat since we can at least look at "near" data since there is data at 12.5 kV and it more or less gets extrapolated by curve fitting (IEEE 1584 or the Wilkins "simplified" or timedomain model) up to 15 kV but there are somewhat arbitrary cutoffs that stop there since data doesn't exist above that point. From what I can gather from statements by ArcPro it seems like they validated their model using data collected from customer's tests but obviously they really can't share that. ArcPro documentation references 4 technical publications as the source of the model but even with copies of the papers there simply isn't enough there to glean what they actually do in their model. Wilkins models are kind of interesting in this respect because although he uses test data to "tune" some parameters of the model, the overall model design is definitely theoretical in nature. I just can't quite say that it's truly theoretical because the parameters are basically curve fitting, just using something other than very generic formulas such as an exponential which is what IEEE 1584 empirical equations are.


Top 



Page 1 of 1

[ 7 posts ] 


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

