ARCPRO offers AFHA services.

So this software provided is a direct competitor of the clients buying their software!

Probably is because they have a lock on the market due to OSHA mentioning them by name as a valid software provider.

This, double dipping makes me very uncomfortable as a consultant.

What do you think?

Arc Hazard Analysis
Kinectrics defines inputs for the following arc hazard variables and then for the end user, computes the amount of radiant and convected thermal energy available at individual sites. Kinectrics is then able to recommend the protective clothing suitable for workers who may be accidentally exposed to electric arc at these facilities.

Arc hazard analysis variables:

Arcing fault current
Arc duration controlled by protective devices
Length of the arc, and
Distance of the worker from the arc

Fist let me go out on a limb and ask the obvious question: If you don't use Arcpro, then what is the alternative that you would use?

I'm not sure exactly what the complaint is. Kinetrics is one of the very few testing labs when it comes to arc testing, whether for the various ASTM PPE tests or testing equipment. I'd almost say they are the only game in town when it comes to PPE testing but at one time Dupont had a lab that did some of that work as well. There is nothing holding anyone else back from offering the same thing but this should tell you something...either there is a very high barrier to entry into the business or the market size just isn't large enough to justify a lot of competition. I'm going to go out on a limb and suggest that it's a combination of the two. This is generally true for almost all high voltage/current power system test labs.

Second when it comes to Arcpro, they predate most of the stuff that is out there. Arcpro was around before IEEE 1584. IEEE 1584 rose to prominence because it offered some very simple calculations that the various power system analysis software companies (ETAP, Easypower, SKM, etc.) easily incorporated into their software. Prior to that it was kind of a two step process. First you did the power system analysis. Then you took those results and manually entered them into Arcpro to do the arc flash analysis. I haven't seen version 3.0 but I would be very surprised if the situation has changed any. There is just as much software "technology" and engineering involved in developing power system analysis code as there is with doing the kind of theory-based arc flash modelling that Arcpro does.

As to a "lock" yeah that's a bit troubling but it's not really anything Kinetrics did. They aren't even an American company, they're Canadian. There is no published data on arc flash at 15 kV or higher. So OSHA applied the "Goldilocks" method. They did the calculations for the three available arc flash tests and found that Lee produces high results, Duke/Heat Flux (Privette) produces low results, and Arcpro produces results somewhere in between. So applying the Goldilocks method, they essentially condemned Duke/Heat Flux and Lee and that left Arcpro as the "last man standing" at 15 kV and above. This isn't totally unjustified. Given three calculation methods which all produce different results and all having exactly the same amount of scientific support for their validity (that is, nothing), all could be claimed to be equally "valid". Or you could suggest that something closer to the "average" is most probably closest to the correct answer so just throw out the high and the low results. That's exactly what OSHA did. It definitely has some merits to it if you give equal weighting to all contenders.

I'd be inclined to agree with OSHA's conclusions even if I don't agree with the methodology or the results. Lee is nice from a theoretical view point if you don't take any arc physics at all into account. Essentially it upper bounds the data based on basic power transfer theory and ignoring physics altogether. Arcs are actually only weakly voltage dependent unlike the Lee model which shows that arcs are actually proportional to voltage. This is the major reason that the Lee model fails so badly in medium and high voltage. As to Duke/Heat Flux back when this was about the only thing I had access to for free when I first starting learning about arc flash (around 2002-2003), it seemed to give some unusually low results. It fell out of favor for that very reason plus the fact that whatever it was doing wasn't documented to anyone's satisfaction so it couldn't be validated. So it was useful to me back then because I didn't have access to anything else to get an idea of whether or not arc flash mattered for the plant I worked at, but the situation got a lot better by 2004 when NFPA 70E-2004 was also a lot more workable than the 2000 edition.

My biggest concern with using Arcpro is this. You can't take the raw inputs and do a calculation by hand to validate the results that Arcpro produces. It's a black box and despite references to 4 scientific papers, there is still really no documentation at all on what Arcpro does to calculate a result. So we have no way to do a hand calculation and no test data to validate the software. We are 100% reliant on Kinetrics considerable reputation. This is the most troubling aspect of it. I'm not against Arcpro obviously but I would feel much more strongly if either documentation on the calculations was improved or test data was available for validation purposes.

As to the OSHA mandate, it's more like a very strong suggestion and at this point, Although my reasons for disliking Lee and Duke are based more on scientific grounds than the Goldilocks rule used by OSHA. Lee and Duke are obviously producing incorrect results down at voltages where I can check the results against IEEE 1584 or IEEE test data like 5-15 kV. I don't like not only OSHA sort of signing on to Arcpro but the same thing is happening in the background of the tables supplied in NESC (IEEE C2) as well, and for the same reasons that I have. So personally I would reach exactly the same conclusions even if I disagreed with the results.

In fact that is exactly what I did around 2009-2010. I worked for a company which had an internal arc flash policy document drafted by the safety department within the company with ZERO input from engineering. In it they specified use of Arcpro for anything above 15 kV, except to simply double the number. There was no justification whatsoever behind the fudge factor other than they did it to be "conservative". When pushed on it, it was also revealed that no other plant had any equipment above 15 kV so this was out there essentially as a place holder. So I was faced with this exact same troubling situation way before OSHA ever published a regulation involving this issue. We found the justification for Arcpro was at best troubling, particularly when a fudge factor was thrown into it with precisely zero support for it. The whole thing just gave our plant a completely uneasy feeling that we were out on a limb with no justification whatsoever.

We of course did the right thing...we just rejected all of it and initially went with NFPA 70E table results because at least those were in a published standard instead of making things up, and then eventually switched over to Arcpro when 70E dropped support for anything above 15 kV in their tables and by that time OSHA had revised 1910.269 which helped lend credibility to using Arcpro results that we felt pretty good about knowing that no test data existed and that this was kind of what everyone was using anyway. The rest of the company followed our lead when the policy was revised.

If actual test data above 15 kV were to come out, I'm sure that we would be in a much different situation to either validate or invalidate Arcpro or something else that has yet to be written. It's possible but I really highly doubt OSHA would question any company that used actual arc flash data over Arcpro or a model that is based on actual data over Arcpro. I wouldn't go so far as to say any company justifying use of Duke or Lee because obviously that's not going to happen, but if there was a 4th option and it was based on actual test data, I just can't imagine OSHA citing a company for using that 4th option over Arcpro. This would be akin to installing arc resistant switchgear and getting cited for not wearing arc flash PPE while doing switching with the doors bolted shut because Arcpro or IEEE 1584 produced a result that indicated that PPE would be needed when obviously the calculation doesn't take into account the arc resistant equipment design.

My concern is that the company selling the software to calculate the AFIE's also is offering consulting services to perform these studies that their clients do.

Double dipping. Is like SKM goes into the business of offering arc flash studies for the industry.

Thanks.

I see your point but you've got to look at what Kinetrics does. This would be classic vertical integration but if you look closely at what they are and what they do, it is pretty obvious that any appearance of vertical integration is purely incidental, not a specific business strategy. It's not like design-build firms where the artificial "wall" between engineering and construction is removed or skirted.

If I had something that doesn't fit the "IEEE 1584 model",or even say the Arcpro model, say determining incident energy exposure in a vault, or say determining whether or not insulated rubber winter boots used by a company in Canada are a significant hazard over leather footwear, that's what their core business is all about. You could design the experiment and all the equipment and run the test yourself and simply rent out the facilities from them. Or if you don't have experience in running these kinds of tests (which most people obviously wouldn't), then they offer the services to help with every aspect of that as well. And that analysis/engineering service might even extend back to your site...doing a site survey for you, although that's probably getting to be more of a value added service with higher margins for them. I think you can see where it would be patently ridiculous for Kinetrics to do basically everything involved in an arc flash hazard analysis including basic research and then pull back and not offer the service as a logical step in the overall package. They do a lot of other things too but in the context of arc flash analysis, I think this basically describes what their business model is all about.

As far as "competing" with Arcpro, I'm putting a lot of faith here in what one of the two authors (I forgot which one I talked to and it's been years) made the genesis and market strategy for Arcpro clear. They developed this software back when there simply wasn't anything else on the "market" except Lee, as a vast improvement over Lee. The original "tuning" was done using sanitized (blinded) customer data from their own research lab so they can't really go too far in revealing the underlying details without in their ethical point of view revealing customer data without permission. They offer this strictly as a service and nothing more. It's not a money maker for them...they plow the license money into paying for maintaining the software and handling the business and marketing side of it and that's about it. They haven't attempted in any way to "keep it for themselves" nor to attempt to compete directly with the other software vendors that are basically just using IEEE 1584 for "free" and not contributing anything at all to development of the science and engineering (SKM, ETap, Easypower, CYME, etc.).

I think though that to suggest that they are "competing" with other engineering firms in terms of arc flash hazard analysis, that kind of misses the point. They probably consider competitors companies like Battelle, EPRI, Aspen, or maybe Dupont...companies that are also involved in the full services research-based approach to engineering and scientific services.

That's the same thing my current employer sees. Our core business is rebuilding motors, generators, circuit breakers, gear boxes, pumps, and a lot of other mechanical and electrical parts. We also offer services to install, maintain, and troubleshoot it. If you want one-stop shopping, fast breakdown response, or comprehensive services, or large or atypical equipment, we're the place to go to. We're the guys with the phone number written down on the corner of the desk somewhere. But every time a customer wants us to "bid on something" and it's a construction or a routine testing job, quite often we suck at it and we don't get the bid. That's not what we do and we're too expensive and we suck at it because we don't just focus on construction or distribution.

I guess, we humans can justify pretty much everything. The fact is that I would be upset if SKM, my software provider, start competing in the arc flash service business. They can spin it in any way, but I would feel the same.

SKM philosophy is that they do not want to compete against their clients. And I like that.

Thanks.

The competition there is purely incidental. Kinetrics was on the scene long before almost anyone else was involved.

There is a philosophy in the software world called "eat your own dog food". The idea is that a business should use their own products internally. This forces them to deal head on with all the bugs, poor user interface design, and other problems of their own software day in and day out, and hopefully creates a better quality product in the end.

This applies to the electrical business, too. I can't tell you how many companies out there buy the equipment and produce all kinds of test reports on breakers, motors, gearboxes, transformers, you name it. There are always all kinds of findings in these reports. But start asking questions from one of these testing companies and you quickly know that they know how to run the test but they don't have the first clue what it means or how to repair whatever problems they claim to detect. I always advise customers that they should go with electrical testing companies that can and will quote repairing the equipment in addition to doing the electrical testing. The downside is that there is something of an incentive to find something wrong, as there is with the totally independent testing companies. But the plus side is that they have a more realistic understanding of what's involved in fixing it.

Kinetrics is like that. They have a more intuitive understanding of what "arc flash" means.

SKM provides an excellent product and without competing with its own clients.

SKM does not provide the details of their algorithms either. And they don't actually use their software in practice so all the bugs and crashes and the fact that there's no "undo" feature or way to easily get back to where you started when doing "what if" analysis is tremendously annoying. Any practitioner using SKM gettings really, really good at keeping a ton of extra backups and copies everywhere. This also means that you really can't have multiple engineers working on separate subprojects within SKM either. Management of the database becomes a huge cumbersome problem if there is more than one person using it. This is never mind the fact that essentially SKM Powertools has a serious case of schizophrenia because it is actually a mixture of multiple DOS tools with Windows dressing on top that are all independent software programs running more or less on an independent database. That's why the concept of "eating your own dogfood" is so important. From a usability point of view SKM is HORRIBLE.

That's just a symptom of a bigger problem though. There is also basically no way to check the results except at a macroscopic (with error bars) level because it is doing a lot of little detailed things in the background like fiddling with asymmetrical fault current data that is nearly impossible to replicate or document what it does. If you just take the output that it gives you and check it back to IEC standards for short circuit,, that seems to work pretty well. But if you then manually calculate incident energy using say the IEEE provided Excel spreadsheet, it fails particularly with a significant reactive load.

When we look at Arcpro in comparison all the usability and stability problems are simply not present. It does one thing and one thing only, and it does it really well. The only inherent problem with Arcpro if you want to call it that is that it is hard to check the results because again the actual software algorithms are basically secret. Unlike SKM which purports to follow a standard but does something undocumented and unpredictable with that standard, Arcpro predates the standard and does something entirely different. It also isn't "predictable" as in I can't take the same inputs and do the math myself and come up with the same number because I don't even know what math to apply.

So I'm in the same position of guessing whether or not to trust in the output of the software. Both give SIMILAR results to IEEE 1584 but not exactly. Both are doing some kind of extra math in the background that neither one documents. So I'm left with looking for other factors. Both have been around for a very long time. SKM was the "original" power system analysis software, dating back into the DOS days. Arc flash module was added quite a long time later, basically copying (to some degree) the IEEE 1584 empirical model. Arcpro is kind of the original "arc flash hazard analysis" software, coming out shortly after Lee's equation and prior to what has become the IEEE 1584 empirical model. So we seem to be at a "zero-zero" score so far. Arcpro is obviously professionally done software. It works. It doesn't crash. It does exactly one thing and one thing well. It is easy to use and doesn't require a 2 week training class during which mostly they teach you all kinds of tricks and workarounds to get past all the bugs. SKM crashes all the time. It also suddenly starts giving erroneous results when you are using it, forcing a complete rebuild of the analysis data. It has no "undo" function so there's no way to easily do "what if" analysis. It often corrupts the project database forcing various rebuilds and regular and routine backups of the working file. And it is essentially unusable when it comes to having multiple engineers working on and maintaining a project. In short it looks, acts, and smells like alpha grade software and always has. The original DOS authors are obviously long gone. The major advantages of SKM these days is that it has been around the longest so like Autocad almost everyone is familiar with it in engineering circles, and that they have the largest device libraries and excellent support when it comes to adding additional ones (just call and wait a couple weeks). But taken on it's merits, it's crap.

Finally e have the "eat your own dogfood" philosophy. Kinetrics intentionally or not practices this, and it's reflected in the software itself. So when it comes right down to it, do I want to trust a bunch of programmers that have proven themselves to be a bunch of second rate hacks that are doing who know's what in the background with my data and who clearly have absolutely no idea how arcing faults occur in the real world because they have absolutely no real world experience of any kind, or do I trust in the company that was there before any of the rest of us where it is clear that they collectively have more experience with faults of any kind from the basic research side of things through application research through putting it into practice? I think the answer is pretty obvious if this is just a question of the merits of the results and how much one should trust company A vs. company B.

To be fair, I use SKM, too. It integrates power system analysis directly into arc flash analysis. As long as you save early, save often, and run a ton of analysis on the results to verify every little detail just in case yet another bug or unsaved result crept in somewhere and make routine backups frequently just in case it corrupts itself again, you can live with SKM and get fantastic results out of it. The biggest advantage of SKM though is that it's kind of like putting out spreadsheets in Excel..it has become something of a defacto "standard". On a fresh project though ETAP and Easypower both beat SKM hands down on almost every consideration except for the fact that SKM has a much longer history. And with all 3 when it comes to power system analysis in a utility setting above 15 kV, you CAN still use Lee which is the only thing that SKM, ETAP, or Easypower will give you above 15 kV. OSHA did not forbid the use of Lee. And if the result using Lee is under say 4 cal/cm2, there's not much point in rerunning it under Arcpro. But if Lee gives you some crazy insane high result, then Arcpro gives something probably more realistic that OSHA would accept.

Thanks Paul. It IS a very large undertaking however, since arc flash protection has been gaining more global traction, there have been a few new labs coming on line for PPE testing in recent years. In addition to Kinectrics in Toronto, Canada two of the newer labs are in Sao Paulo, Brazil and Alicante, Spain.

The Sao Paulo Lab was created as a partnership with Petrobras, the largest multinational energy corporation in South America and the Energy and Environmental Institute at the University of Sau Paulo (IEE USP). They can test in accordance with the relevant ASTM and IEC standards. The Brazilian National Committee to the IEC has several nominated experts, including their lab manager, serving on the PPE project teams of the TC 78 – Live working Committee that I chair.

Aitex brought their lab on line in Alicante, Spain a few years ago. The Spanish National Committee also has a very large contingent of nominated experts (including those from Aitex) serving on the various PPE project teams of IEC TC78 Live Working.

Kinectrics is a very large consulting engineering firm focused on nuclear power certifications and utilities. They will rarely compete with a small utility arc flash study and most of their engineers are in Canada. I have seen them do ONE arc flash study in the US.

I wouldn't worry about them competing with anyone on the forum unless it is for a Million Dollar arc flash study or something in Canada.

Excellent company and it's great to have good competition.

ArcPro was recommended because it had real data up to 33kV and was based on an arc physics model rather than mostly LV data extrapolated to be used for HV. IEEE 1584 is clear that you can use it for HV BUT it will over predict.

We see it as a compliment to SKM rather than a replacement. We use both. It is not a replacement for IEEE 1584 but is preferred on Medium Voltage.

ArcPro is recommended by OSHA for UTILITIES or utility like constructions ONLY and then only for overhead lines. You can use it elsewhere but it is best for >4160V services.

There are other arc flash labs but they are nothing like Kinectrics. It is a HV and High Current Lab. The other two Jim mentions perform the clothing tests but do not do blanket or the other very high current tests or the research for a standard like IEEE 1584.

I also compete with Kinectrics on arc flash studies but have never had to bid against them since we focus on different markets.

Hugh Hoagland

Yep,
Here I agree with you 100%
This is the first thing I noticed when I read their contract.They only spend money for there softwers e.t.c

Regarding competition of software sellers providing engineering studies, in the past CYME provided this service and I believe EasyPower did as well. I don't know if they are actively pursuing this now.