BAND-IT WEBINAR

Don’t miss out, join us for an educational webinar as BAND-IT’s Peter Moreno and Mike Selissen of MS Design Services explore why tray-based single-conductor systems are becoming more common, when mechanical fault protection becomes critical, and how NEC 392.20(C) is being interpreted and applied in the field. The discussion will explore how electrical design decisions connect to mechanical restraint considerations, helping project teams better understand risk in high‑energy environments.

Webinar Agenda

• Evolving power distribution in U.S. data centers: What’s driving increased use of tray‑based single‑conductor systems.

• Application considerations and design tradeoffs: How engineers evaluate tray‑based systems alongside other established power distribution approaches.

• Mechanical fault forces and risk awareness: When mechanical protection becomes a consideration in high‑energy systems.

• NEC® 392.20(C) in practice: How code intent is interpreted and applied in real‑world installations.
• Understanding cable cleats: Purpose, performance expectations, and design intent.

INCASE YOU MISSED IT

FULL TRANSCRIPT BELOW

Introduction 

Moreno, Peter   0:05 

Hello everyone, thanks for joining us this morning. I have a presentation for you today from BAND-IT. If you’re not familiar with our organization, we are a mission critical fastening manufacturer based out of Denver. We have a global presence with the bulk of our manufacturing occurring here in the Denver location for the Americas, and then another sister location in the UK servicing Europe, and another in China servicing Asia. 

We’ve been around about 90 years, started off in the industrial space, and have since expanded into aerospace, automotive, the energy sector, and more recently have been emphasizing the AI infrastructure space as well. 

We appreciate you all joining. We’re going to take you through our presentation today where we’ll be covering the mechanical protection of high-density power systems in cable tray. 

Just an admin note, your cameras are off and microphones are muted. If you have any questions, you can put them in the Q&A or in the chat. We’re going to reserve a little bit of time at the end to answer those. At that time also, I believe you’ll be able to raise your hand and we can unmute you if you want to ask a question that way. We’re recording this session so that you can view it afterwards or share it with any of your colleagues that are interested in it. If you have anything that we don’t get to, we’ll have our contact information at the end. Feel free to reach out to myself or Michael and we’ll get an answer back to you.

So, with the massive growing wave in AI infrastructure, we’re all seeing exponential growth, especially in mechanical restraint specifications in those builds. We’re going to get into what that means and what that looks like on the ground as we go through the evolving power distribution in data centers. We’re going to talk about application considerations and design trade-offs, what mechanical fault forces look like, and the increasing risk awareness around them. Then we’re going to dive a little into the NEC code and what it speaks to in regards to this, and what the go-to solution for these types of applications is, in the form of cable cleats.  

To introduce myself, my name is Peter Moreno. I am leading business development for the Americas in this space and also functioning as product manager for these products. And I’ll allow Michael to introduce himself.

Mike   2:41 

Hello, I’m Michael Selissen. I’m the owner and the electrical engineer of MS Design Services. I do engineering consulting, electrical engineering consulting services in the industrial and commercial and data center space.

Moreno, Peter   3:04 

Thanks, Michael.

Evolving Power Distribution in U.S. Data Centers 

So we’ll hop into it, starting out just talking about evolving power distribution in data centers at the higher level. Michael, do you think you could speak through what changes you’re seeing and how power has been distributed, over the past few years especially?

Mike   3:25 

Yeah, so 5, 10 years ago, it was more common to see, like the picture there, conduit and J boxes running multi-conductor cables. It was typically lower KW sources you were dealing with. And also, it was mainly AC. 

And then it was phase build, so you weren’t as much in a crunch to get it done. Now we’re seeing larger KW feeds required, and then also the emergence of DC as well. Also you got tight schedules in this construction phase now. So you’re kind of doing parallel [construction], you’re installing the equipment, you’re installing the conduit or the cable tray at the same time. So you got to try to be efficient with your labor resources. So again, it’s more massive in size. 

That’s driving repeatability of projects to try to get them done, try to get the designs done quicker and get the installs done quicker. So driving the use of cable tray over conduit.

Moreno, Peter   4:43 

Yeah, I know I’ve heard from the field, one of the projects that a customer was working on, their end customer had paid about 100-150 million just to combine some of the phases, even though it was going to be very difficult. A lot of money going into getting these up and running as fast as possible.

Speaking a little bit more about why this ties into why this mechanical fault protection is being spec’d in, I wanted to touch on it in terms of why tray is having more attention put on it. Could you speak a little bit to why, especially single conductors, are being used in tray in these types of builds and why they pop up here more than in other areas?

Mike   5:38 

Oh yeah, single conductor again, like I was talking about earlier, you got larger KW load, so you’re going to have larger wires and a multi-conductor cable, that wire is so much bigger and heavier, it’s more difficult to pull, it’s more difficult to terminate. So just stiffer wires. If you can do single conductors in a tray that makes it a lot easier for pulling and handling the wire and also even like terminations. Terminating a single conductor cable is much easier and quicker than a multi-conductor cable. Also you’re just talking about labor. You’re trying to reduce labor time and labor effort and that’s huge in terms of the amount of labor required. Per cable it’s less for single conductor, and that drives the use of tray. Also, bending conduit, that takes a lot of time. Fitting, that’s a lot more time and that labor resource is already strained. Trying to get that minimized, cable tray and single conductor in tray just seems to go together with all those variables in play.

Moreno, Peter   6:55 

Yeah, that’s what I’ve been hearing as well, especially as you pull those massive multi-core cables off a spool. You can only go so far, the distances get shorter and then it’s heavier on everyone. So a lot going into it.

Mike   7:08 

Yeah

Moreno, Peter   7:11 

Wanted to talk a little bit about the standards themselves. These are also defining US norms more in these builds, especially in the AI infrastructure space. As these hyperscalers, tenants, developers are building, they’re building globally. The design firms, they’re working across borders. Naturally, these different firms are getting exposure to international standards. One of the leading bodies there is the International Electrotechnical Commission, the IEC. That standard is essentially referenced or directly used across Europe, Asia, lots of Middle East, Africa, many nations outside of the US. South America, for the most part, has adopted it and references it. And of course, the US references it as well, but just not as explicitly.

Moreno, Peter   8:11 

When these hyperscalers and tenants are doing these builds globally, they’re exposed to best practices and sometimes explicit directions from that code in how they handle these runs that they’re doing, especially in these open air tray runs. Then they’re bringing that back to the US, asking, “well, is there anything that makes it so that we don’t need to do that here?” [Especially] since they have that high sensitivity to downtime, 5 nines essentially equates to less than 60 minutes of downtime a year, it’s non-negotiable to meet that standard. They also want triple redundancy – all of that is driving them to adopt these standards in the US as well. 

What we’re going to talk a little bit more about is the NEC and how that ties into it, but some of the other relevant ones as well are NEMA, they published a standard, the CSA, [which] has a reflective portion that points to these types of products and mechanical protection being used and then other organizations as well.

Application Considerations and Design Trade-offs 

We talked about it a little bit, but diving down a little bit more so that we understand some of how historically a lot of these runs have been done with conduits, busway -cable bus is picking up a little bit more now, – and then you also have cable tray.

Just as engineers are thinking through these decisions and what makes sense for the different types of runs, do you think you could give a little more detail, Michael, on some of the trade-offs here?

Mike   10:02 

Yeah, kind of what I spoke to earlier, it’s very labor intensive. It chews up a lot of resources, and then busway is good for high current, tight applications. You [use it for] tight space constraints application because it’s got a smaller footprint. But, it is more expensive and probably availability and lead time is a little greater than cable tray or conduit because you’re designing it per certain amperage sizes. 

Then, cable tray, the labor is more efficient, it handles large runs. It’s also, as changes go, because you’re doing it in parallel, you’re designing, you’re building, putting in the cable tray and the equipment at the same time. You can put in a larger tray not knowing what exactly is going to go in there. As things change or evolve during the design, it can accommodate it more efficiently; gain, it’s easier to pull it in. You still have to use conduit, when the cable tray gets you to the right place, and you got the final connection out of the tray to the equipment has to be in some type of busway or conduit, but that’s pretty localized and minimal, so that makes it efficient, so you can do all the cable tray while you’re kind of setting the equipment and doing the other pieces at the same time. 

It’s just the speed of putting up the tray is so much more efficient than the conduit side of it. But then too, in the tray you get exposed to the conductors. The conduit has the protection there for the arc, the short circuit mechanical forces that are there. Same way with the busway, that’s protected. Cable tray now, the cables, the single conductors are more exposed, and that’s where you got to try to understand that and properly secure those cables so they don’t cause damage or create that downtime that these data centers can’t afford. 

So cable tray is a great solution for this, but it also throws in other variables that we need to make sure that are accounted for.

Moreno, Peter   12:39 

Yeah, definitely, the speed is the ultimate driving factor in every conversation. It’s top of mind, then everything else after that. They’re opting for speed and then realizing, “okay, we do have trade-offs. So now how do we protect against it?”

And these are used across different power architectures in these types of projects. Some of the things that are changing that are driving this at a higher level [include]: rapid growth in these projects for on-site power generation and battery energy storage systems, sometimes even becoming the primary form of power, especially as different companies are projecting that these campuses are just going to plan on permanently running off their systems. Those systems as they’re installed are also facing speed influence and pressure. 

As they’re now choosing to route above ground more often, (that’s becoming a growing trend there, before they were trenching), that added a certain amount of time to the projects. Now they can route above ground, bypass that time aspect, but they have those exposed conductors on tray runs often, and so have that consideration there. 

There’s also the increasing use of DC systems and hybrid architectures, running higher voltages through in that different current system and exposed to those forces that are resulting from that. You have that general higher power density down to the rack level. Those racks, like Michael had touched on, they’re going up to 100, even 650 – just huge amounts of power all across the system where before you would be limited to a certain space and then it would drop step down. Now it’s going further into the project. These favor the cable tray for some of the reasons that we had mentioned.

It allows that faster deployment, easier routing, especially the ability to adapt to design changes. As materials are so bottlenecked, designers and tenants, they might have specified in the past certain manufacturers or certain specific brands of products that they want to use. Now they’re just saying, “I just want this project to get done, give me an equivalence, or if we need to make a slight change to the spec so that I can route it a different way and use a different type of equipment”, they’re doing that. 

With that happening hundreds of times over these different phases of these projects (sometimes they’ve already laid out whole systems and then some piece of piece of key equipment changes), now they need to reroute. When you’re having things like that happen, you need that flexibility built in. We’ll speak a little bit more about that here.

Mechanical Fault Forces and Risk Awareness 

So what’s this actual risk from the open air cable tray runs that occurs? On the left here, you have an example. BAND-IT conducts testing in this space; we test above our failure level to understand why failures occur and how we can continue to build out new products and solutions. On the left, you’ll see an example of a fault running through a system on cable tray and ripping those cables out of that tray, causing a failure to the equipment. You can see it’s impacting the side of that tray. So these are real forces that are that can occur in these events. 

In the the IEC, there’s a prescription to protect electrically and also mechanically. You have that electrical portion of the fault, usually that’s covered by a circuit breaker, but those kick on within a few milliseconds, and there is the instantaneous potential for these forces to occur, which is why a fixed device is usually the prescription so that they are holding that system, and not dependent on a [electrical] system; it’s already in place and holding it. 

On the right, you’ll see a properly protected run where you have these fixed cleats in place and those are protecting the system. Those are the understood best method of protection. Some of the things that you’ll see in other areas are the use of covered tray, conduit, like we were talking about, provides this protection. Technically, if you’re in a multi-core cable with that sheath, especially an armored cable, those are meant to provide protection. But with the increase in single core conductors running in these trays, just so they can route these higher power levels, they gain that efficiency and speed, but it does expose to these types of forces. Still a great option, but just need to provide that protection in these areas. We’re going to talk a little bit more about that [here].

You know, a lot of times in these runs, they’re using ties in the US right now. So nylon ties, there hasn’t been much in the way of using cleats historically. And so now that awareness is coming into these AI infrastructure builds where, they’ve seen the best practice elsewhere and now they’re bringing it down. So this would be replacing those nylon ties as the form of securing these cables to tray.

And what’s happening here is, so these short circuit forces, they’re putting current through that system, and that’s creating a magnetic field in that conductor. And so these conductors, especially when they’re run in parallel, those magnetic fields from each conductor acts upon itself. It ramps up to a certain degree as the power increases and that cable conductor size is increasing, you reach a peak of the short circuit forces. And then [counterintuitively], as the conductor continues to increase in size, those magnetic centers between those conductors grow farther apart. You can actually see a decrease in the levels of the fault forces that you see. All that to say, sometimes these fault forces can occur most aggressively in low voltage, where you would think that they would be more [powerful] in medium or high voltage. They can occur across different voltages and systems. 

Anywhere you’re running single core or conductors in parallel, and they can act upon each other; even those multi-cores can [experience] it. They’re just retained in those sheaths. Anywhere where you’re seeing these, that’s where you would want to have mechanical protection to prevent about against this.

There’s different configurations; the trefoil is understood to be the most aggressive just because you have those three conductors ran in parallel right by each other. You can also have flat formations where one of them could be acting up on the outside [cables] or the outside ones could be acting on the inside. Then in a quad (quadruplex formation), those would be acting on one of the conductors by the other cable. These are most common in medium voltage, low voltage feeder runs, but especially in those single conductor high density installations where we’re seeing these packed trays, [where] there’s no more space in that rack, that run parallel, that’s where you’re seeing these potential fault forces.

This is an example from testing we did a few weeks ago. At BAND-IT, we test to above our failure levels so that we can understand what happens; this is an example of that. This shows one of the live reactions of these systems where essentially we have a 6000 volt generator in the back, it’s pumping fault forces through and into a cable tray system. And we’ll see what the impact of that fault is when the system wasn’t properly specked for it.

Moreno, Peter   21:20 

So you can see that there was a failure [where] in this case, the fault was so extreme that the lugs failed after the whipping effect had occurred. It can be catastrophic to those systems, especially in these dense, packed AI infrastructure builds where they have trays right on top of each other. It’s not just impacting the cables right next to each other. It’s also impacting the three-dimensional above it. 

NEC 392.20(C) In Practice 

Moreno, Peter   21:50 

So Michael, I was going to ask, before we get into the current day, could you speak to, historically, what your interpretation of the NEC has been in this area and how that might have changed recently?

Mike   22:00 

Yeah, I feel historically it’s been, “we need to secure this cable – let’s try the most economical and readily available tool to do that”. And that’s [been] the standard nylon cable tie – readily available. Every electrician contractor has one out there and it just secures it in place. But then you start thinking about it -that’s just to support, to prevent; now we’re talking about fault forces. 

So now the NEC has added some language in there into 392.20C where it talks about excessive movement of fault forces. But a lot of times that still, it just says to secure it, it doesn’t really say how and [with] what. It’s [up] for interpretation a lot of times on how to properly secure it. Also in NEC 110.36 where you’re dealing with larger voltages, they said it needs to be capable of safely withstanding the maximum magnetic forces. So, in that sense, as we saw in those previous slides, a nylon cable tie is not the tool to properly secure those cables with the fault force. Yes, it’s the tool just to secure the cable from moving, to just being in place, but not to handle the fault short circuit forces that are happening. 

And then there’s others – NEMA created one [which] reflects the IEC. That’s where we’re seeing NEMA and IEC come together, where North America is taking in the global IC standards. 

I think it’s going to evolve more as these data centers and larger, compact installs just become more of the norm. We’re going to see this more and more. It’s going to probably be more in the customer specs, so it’ll be driving the proper forces because the NEC speaks to it, but not strongly enough to understand which exactly to use. So I think the specs will be pushing that as well.

Moreno, Peter   24:40 

Yeah, historically, actually, the American Petroleum Institute, API, they’ve been seeing more explicit specification for these kinds of things for a while. They have the marine rung where it has those slots on the rungs that you can use devices like cable cleats or more robust metal straps for a while. And that’s one indicator that it’s been on the fringes of awareness here. But I think it just this AI infrastructure build out, it’s so much new construction at once, and it’s at the mission critical level where they have 0 tolerance for downtime. They’re much more intolerant to any failures, not just because their system goes down, but then they also are at the whim of these bottlenecks again. They’re waiting for replacements, materials, and then they’re fighting for labor again, where, maybe they just gave up that labor when they finished construction and now they’re having to bid for it again against the next project. 

So seeing a lot of incentive to just do it right the first time and make it protected. And then you see, it is a gray area right now. The common interpretation has been the nylon tie. There’s anticipation that it’ll be more explicitly in there in the next code revision, which probably, I think, Michael, you were saying maybe 2029, but that you probably see something in there.

Mike   26:10 

Yeah, every three years, yeah, so it’d be the new NEC. It came out and that’s in the 2026 what we’re speaking about there.

Moreno, Peter   26:21 

Right.

Yeah, this has been in there for a bit, but it just hasn’t got as much attention been paid to, especially 392.20. It’s just, “well, if you use a UL listed tie, then you’re meeting that”. And that’s been the understanding, but I think there just hasn’t been as much awareness of the actual risk. 

Mike   26:25 

Yeah. 

Yeah.

Moreno, Peter   26:46 

So, part of the reason that there’s so much variation across the projects, you have different contractors that are doing these installations. Like we’ve been talking about, speed is going to be the ultimate deciding force. Using a nylon tie is always going to be the fastest approach in a project; we can get close, but never beat that (putting on a tie every 8 feet or so). So these contractors, they’re not going to self-specify these types of solutions a lot of times, unless they’ve explicitly been given direction from the design side of things. Unless they’ve run into these issues personally before, a lot of times they’ll still be using these ties which are serving their purpose – they’re built for securing and supporting, but they’re not built for short circuit forces. 

You also have just different project types; so you can have chip fabs, high-level infrastructure, hospitals, nuclear build-outs (we’re seeing a lot of small modular reactors being built), data centers – so all of those have different approaches, and also where before we were seeing more copy paste, now with technology changing so much (these CDUs are now changing to hot cooling) things like that that are happening so rapidly. It changes so that they’re not exactly copy pasting those builds. So you’ll see some variation there. 

I was curious, Michael, if you could speak to what you’re seeing with some of the AHJs in the project you work in, and how you can see variation even with those, depending on what county line you’re working on.

Mike   28:41 

Yeah, yeah, there is a lot of variation, all the inspectors, they’ve had their experiences, they’ve worked with a lot of different projects, so their interpretation of, “yeah, it’s tied, you know what I mean? Everything is good to go. It’s got a cable tie in it that should work.”

But like Peter was saying, it is a cable tie support, but it doesn’t handle those mechanical forces that are caused by a short circuit fault, and then it’s a gap in what people interpret. Contractors interpret it [as] just tying it down. Like you said, you’re trying to do hyperscale and stuff – labor is tight, you want to get it right the first time, you want to properly do it. And I think as more of these data centers get built and we get more exposure to it, they’re going to be interpreting it a little more.

Now we need to secure this better to handle the fault forces. It’s also the customer, because if they have a spec that says to use a cable cleat, it’s going to get caught there too. And you don’t want to be, as the project is getting done, the labor goes on to the next job and having to redo [things] all the time, that’s massive to go through all the cable trays with all those and to re-cleat it all is a lot of effort. So there’s becoming more awareness to it. It’s just going to take some [time], sometimes it’s got to be some hard lessons. Inspectors and engineers as well will kind of go that way. I’m even, my personal learning is like, I thought a cable tie was fine 5, 10 years ago, but not anymore.

Moreno, Peter   30:50 

Yeah.

We did the same testing, we had done a 250-pound nylon tie at a low fault rating (at a 30 kiloampere fault level) and done a test for that just to see what the result would be. You’ll see that it just flies off instantly as soon as it occurs.

So that’s just a low fault level there, just popping all those off, even at the 250-pound level. It really requires that actual device that’s manufactured for that purpose.

Mike   31:24 

Yeah.

Understanding Cable Cleats 

Moreno, Peter   31:34 

So I’ll speak to, at a high level, what the answer for these is. It’s understood that cable cleats are a best practice when you’re running in these open air cable tray installs. These trays, you could have a ventilated trough, you could have cable ladder – wherever you’re running these types of power runs in those. And these cleats are essentially devices that secure those power cables to the tray and are tested to the IEC 61914 standard. So that standard is what we were referencing earlier that was replicated to the CSA. It’s also been replicated to NEMA’s recommendations, UL has developed a listing for it that’s going to be more relevant as the NEC more explicitly requires these. And then the NEC, like we were saying, is projected to provide more explicit guidance to specify the use of these types of devices or protection. So these come in all different shapes and forms.

Some of the things to think through as you select them, as you see these specs for mechanical protection and determine what the answer to that is, especially on the ground engineering level. You need to consider the speed of installation for those crews. They’re going to be facing massive pressure to build in parallel with all other phases of construction; as they’re installing these, [you] want to make sure that you have a speedy installation, minimize your pre-staging. Make sure that you can install it post-cable pull if you can, and unlock that speed for your crews. 

You also want to be able to perform in dense tray environments. These trays are getting packed; a lot of times it’s edge to edge. They got parallel conductors, sometimes in quadruplex formation, there’s a lot going on in there. You need space for the actual device, but also to manipulate it and install it in those spaces, and so want to consider that. Then you also want to have scalability across your build. Ideally, you could have solutions that work for multiple cable sizes and configurations and also, reduce your dependence on the logistics on the ground there. You want to make it so that those crews can spend their time installing and putting the equipment in place and not as much managing the materials.

Especially across these job sites, they’re like small cities, just making sure that you can make that process as easy for them as possible to get the right product to the right portion of the build. 

In specifying them, there’s four major elements. One of them is your spacing. Typically, these devices would be installed at a 1 foot or 2 foot spacing, sometimes they go a little bit beyond that [with] some computer modeling that extrapolates it out, but [really] each of these cleats needs to be tested in a lab for these short circuit forces. Typically those are tested at the one foot or two foot spacing level, so those cleats will advertise what level they can perform to.

The driving factor is going to be that peak fault exposure. Usually this is going to come from your short circuit study or that arc flash study. Usually that short circuits exposure  is part of that, so that’ll tell you, “okay, this system is susceptible to a 50 kiloampere peak fault rating”,  so that’s when you would know, “okay, I want a cleat that can protect against 50 kiloamperes of peak fault force.” And so that’s going to be the biggest element that goes into specifying these.

The other ones on the physical side are obviously, you’re going to need to know your cable OD. Some cleats are fixed size, so they’ll only work for a certain range of cable OD plus or minus a few portions of an inch. You want to make sure that it can fit those. You also want to make sure that it can fit your configuration style for your install. So if you’re doing flat, trefoil, triplexed, or quadrifoil (or quad) run of cables, or even if – we’ve seen and heard of specifications for multi-core cables – you can use these as well (it’s basically a single cable). You just want to make sure that that cleat performs in that configuration there

Again, all of that is important. Those are the core pieces of specifying cleats, but you also want to make sure that you’re keeping in mind those field installation considerations, because you can specify in fanciest cleat in the world, but if it gets to the field and the crews are slowed down, it’s disrupting schedules, it’s causing problems in the tray and you’re having to do design changes because it doesn’t fit or they’ve placed a tray 6 inches above it and you can’t install it there, then you want to get ahead of that when you’re doing your specification. 

Speaking to that, these builds, on the ground, they end up all over the place. So seeing instances where these are behind walls, these trays will run behind walls, and they’ll cut an access cover, but they’re still reaching down to three feet to install devices on these trays. They’re also a lot of times in subfloors, in trenches sometimes. They’re run in all different areas of these architectures. They’ll be run vertical, on top of roofs, overhead. They’ll be sometimes 5 layers deep. They’ll have other elements running through them. They’ll be on top of ducts that they’re running air through one area and then on top of it is a tray and then they have conduit running alongside. So there’s all sorts of considerations. Especially when these are being designed, it’s good to keep in mind what that installation is going to look like on the ground and be proactive about thinking 

Wrap-Up 

Moreno, Peter   40:39 

Well, great. Appreciate you all joining us. And like I said, we’ll post this on our website and have information to Michael and his business. We’ll be following up with more in this series as well. So stay tuned for that. But appreciate it, everyone. Have a good day. Thanks.

Mike   40:56 

All right, yeah.