What Is a Remote Power Panel (RPP)?

The short answer

An RPP, or Remote Power Panel, is a distribution panel that sits between upstream switchgear and the loads it serves, taking one large power feed and splitting it into individual branch circuits. 

It's called "remote" because of where it sits, physically away from the main electrical room, not because it does anything wireless. You'll find RPPs in data centers, hospitals, industrial plants, and any large commercial building where running every branch circuit back to a central panel would mean miles of extra cable and a voltage drop problem. Put the panel closer to the load, and the wiring gets shorter, cheaper, and easier to code.

That's the panel in one paragraph. The rest of this page covers how it actually works, what's inside it, where it belongs in a data center power chain, how it's different from a PDU, and one thing that's changing about RPPs that most explainers on this topic don't mention at all.

How a Remote Power Panel works

Power doesn't go straight from the utility to a server rack. It moves through a chain, and an RPP is one link in it:

Utility → main switchgear/switchboard → RPP → branch circuits → the load

The RPP's job is narrow and specific: take one upstream feed, protect it, and split it into however many branch circuits the space needs. 

Inside, that means:

• Main lugs or a main breaker for the incoming feed

• Bus bars rated to carry the panel's full amperage

• Branch circuit breakers, typically in the 20A to 30A range, one per circuit

• A NEMA-rated enclosure, Type 1 for indoor installs, Type 3R if it's outdoors or exposed to weather

• Optional monitoring, current metering or a full branch circuit monitoring system (BCMS)

None of this is exotic. An RPP is, at its core, a panelboard. What makes it an "RPP" specifically is where you put it: close to the equipment, not back in the main electrical room.

RPP core specs at a glance

Engineers specifying a panel usually want the numbers up front, so here they are.

Spec	Typical range
Amperage	225A to 1,200A
Breaker count	42 to 168 poles
Voltage	120/208V or 277/480V
Enclosure	NEMA 1 (indoor) or NEMA 3R (outdoor/weather)
Monitoring	Optional BCMS via Modbus/TCP or SNMP

Actual specs are project-driven. A 225A panel serving a small equipment room and a 1,200A panel feeding a row of high-density racks are both "RPPs," they're just built for very different jobs.

Where RPPs are used

Data centers. RPPs sit at the row or zone level, taking power from a floor PDU or upstream switchgear and distributing it to rack clusters through shorter whip cables. This keeps branch circuit protection close to where the racks actually are instead of running everything back to a central room.

Commercial and campus buildings. Large buildings put RPPs in floor-level electrical closets. It shortens feeder runs from the main switchboard and makes zone-based load management practical without a rat's nest of home-run cabling.

Healthcare facilities. Hospitals rely on RPPs for dedicated branch circuits serving critical equipment, often with BCMS monitoring layered on for real-time visibility, which matters when the load on the other end is life-support equipment.

Industrial plants. Manufacturing floors use RPPs with NEMA 3R enclosures for harsher environments, usually in three-phase configurations to handle motor loads and HVAC.

The common thread across all four: loads that are spread out, and a facility that doesn't want every circuit running back to one central panel.

RPP vs. PDU: what's the difference

This is the second most common question after "what is an RPP," so let's be direct about it.

	RPP	PDU
Primary function	Branch circuit protection near the load	Last-mile power delivery, often with outlet-level control
Typical placement	Row/zone level, electrical closet, near distributed loads	Inside or beside the rack
Circuit protection	Molded-case or thermal-magnetic breakers	Internal supplementary protection, optional above 16A
Monitoring	Optional, added via BCMS	Often built in on metered/monitored/switched tiers
Best fit	Commercial, industrial, healthcare, data center zones	Data center racks, colocation, edge sites

The short version: an RPP protects and distributes power to a set of circuits. A PDU takes it the rest of the way to individual outlets, and usually gives you more visibility and control at that last step.

In large data centers, you'll typically see both, not one or the other. The full chain often looks like this: Utility → switchgear → UPS → floor PDU → RPP → rack PDU → IT load. 

The PDU handles monitoring and control further upstream or at the rack; the RPP handles hardwired branch protection in between. They're not competing products. They're doing different jobs in the same chain.

Why RPPs matter more in AI-density data centers

Here's the part most explainers on this topic skip, and it's the part actually worth reading.

For years, an RPP's job was simple: take a feed, split it into circuits, protect them. The breakers inside doing that protection were mechanical, thermal-magnetic units that have worked fine for decades because rack loads were predictable and modest, typically 5kW to 10kW per cabinet.

That's not the load profile anymore. AI training and inference racks are pulling 30kW, 50kW, in some deployments over 100kW per cabinet, with power draw that spikes hard and fast instead of sitting flat.

That kind of load puts real stress on mechanical breakers in specific ways:

Nuisance tripping. Thermal-magnetic breakers respond to heat and current buildup over milliseconds. High-inrush, high-transient AI loads can trip a breaker that isn't actually faulted, just reacting to a spike it wasn't built to read cleanly. Every nuisance trip is downtime you didn't plan for.

Cascading risk. A real fault on one rack in a shared power bus can pull down voltage across adjacent racks before a 20 to 50 millisecond mechanical trip even finishes clearing it. At AI-scale power density, that's not a one-rack problem anymore. It can take a shared bus with it.

Arc flash exposure. Mechanical breakers clear a fault by physically pulling contacts apart under load, and that gap is where an arc forms. At a densely loaded bus, an arc flash event isn't a minor incident. It's a real injury risk for anyone working on or near the panel, and it's the kind of liability that shows up in PPE requirements, incident reports, and insurance premiums long after the fault itself is cleared. Higher rack density means more panels doing more switching more often, which means more exposure over the life of the equipment, not less.

This is where the breaker technology inside the RPP starts to matter as much as the panel around it. Solid-state circuit breakers interrupt current electronically through a semiconductor instead of pulling physical contacts apart. There's no gap opening under load, so there's no arc to extinguish and next to no arc flash energy to plan around. Trip response moves from milliseconds to microseconds, up to 3,000 times faster than a mechanical breaker of the same frame size. On a shared bus, that's the difference between a fault that clears cleanly on one circuit and a fault that drags the rest of the row down with it, a capability the industry calls ride-through.

We build one of these panels, so it's worth naming plainly rather than gesturing at "solid-state RPPs" in the abstract. Atom Halo RPP is the panel we built for exactly this problem: a UL 891 panel (listing in process) that puts solid-state protection and continuous electrical intelligence inside a single enclosure, instead of pairing a mechanical panel with a monitoring system bolted on afterward as a second project.

A few numbers, for anyone comparing it against a conventional RPP spec sheet:

Spec	Atom Halo RPP
Distribution	400A to 1,600A
Branch breakers	Up to 28, UL 489i
Short-circuit rating	100 kAIC, independently validated past 204 kAIC in destructive testing
Let-through	Capped at 1.3 kA
Voltage	480 VAC, also available in 208/120 and 415/240
Waveform monitoring	20 kHz capture on every branch, full BMS integration

Two of these numbers do more work than they look like they do. Let-through is what actually limits the energy that reaches servers, UPS units, and busbars during a fault, so a low let-through number matters as much as trip speed does. And zero arc energy isn't a separate feature, it's a direct result of the switching technology: no moving contacts means no arc plasma and no cumulative contact wear, so a branch can trip thousands of times without the gradual degradation that eventually pushes a mechanical breaker out of tolerance.

Solid-state protection also comes with monitoring and diagnostics built in natively, rather than added on as a separate BCMS layer. For a facility team, that means real-time visibility into every circuit without specifying and wiring up a monitoring system as a second project.

Worth knowing if you're specifying this today: solid-state breakers now have their own real certification standard, UL 489I. It's built to work alongside the existing UL 489 standard for mechanical breakers, not replace it, and it specifically addresses things unique to solid-state switching, like how the device behaves under a loss of control power. Before UL 489I, solid-state breakers were only covered under a preliminary Outline of Investigation. That gap is closed now, which matters if you're the one signing off on the spec.

None of this means every RPP needs solid-state breakers. A panel feeding standard office loads is still well served by mechanical protection, and it costs less. But for RPPs feeding high-density AI racks, where a single nuisance trip, a slow fault clear, or an arc flash incident has a real dollar cost attached to it, it's worth asking what's actually inside the panel, not just what the enclosure is rated for.

How to choose the right RPP

A few questions worth answering before you decide one:

• What's the actual load, today and in two years? Undersizing an RPP means a costly retrofit later. Oversizing wastes budget now.

• How many branch circuits, and at what breaker count? This drives the panel's pole count and physical footprint.

• Do you need monitoring? If yes, decide whether that's a BCMS bolted onto mechanical breakers or built into solid-state protection from the start.

• What's the environment? NEMA 1 for indoor, NEMA 3R if it's exposed to weather or a harsher industrial floor.

• What's upstream? Confirm coordination with the PDU or switchgear feeding it, so protection settings line up cleanly across the chain.

Bottom line

An RPP is a straightforward piece of equipment doing a specific job: taking power from upstream switchgear and distributing it safely to the circuits near a load. What's changed isn't the concept. It's the load. AI-density racks are pushing more current, more transient stress, and more consequence for a slow fault response through panels that, for most of their history, never had to deal with any of that.

If you're specifying an RPP for a standard commercial or light industrial load, a well-built panel with mechanical breakers is still the right, cost-effective call. If you're specifying one for a high-density data center row where a nuisance trip or a slow fault clear has a real cost attached, it's worth understanding what solid-state protection changes before you finalize the spec.

If you're evaluating power distribution for a real project, let's talk.

FAQ

Is an RPP the same as a PDU?

No. An RPP protects and distributes power to a set of branch circuits. A PDU delivers that power the rest of the way to individual outlets, usually with more built-in monitoring and control at the rack. In large data centers you'll typically find both in the same chain: utility → switchgear → UPS → floor PDU → RPP → rack PDU → IT load.

What does RPP stand for?

RPP stands for Remote Power Panel. It's a distribution panel that takes one upstream feed and splits it into branch circuits, installed close to the loads it serves rather than back in the main electrical room.

Where is an RPP placed in a data center?

Typically at the row or zone level, between the floor PDU (or upstream switchgear) and the rack PDUs, feeding a cluster of racks through short whip cables. Putting it there keeps branch circuit protection near the racks instead of running every circuit home to a central room.

How do I size an RPP?

Two numbers drive it: total amperage and branch circuit count. Conventional RPPs run 225A to 1,200A with 42 to 168 poles. Size to the load you expect in two to three years, not just today's, because undersizing means a retrofit and oversizing means budget spent on capacity you never use. Also confirm what's upstream, so protection settings coordinate cleanly across the chain.

For high-density rows, the calculation changes. Atom Halo RPP distributes 400A to 1,600A with up to 28 UL 489I branch breakers. Fewer poles than a conventional panel, but each branch is larger, individually metered, and independently switchable, which is the trade you want when a single cabinet is pulling 50kW instead of 5kW.

Do RPPs need active monitoring?

Not always. A basic RPP feeding predictable loads runs fine without it. But for high-density or mission-critical loads, real-time visibility into each circuit has moved from nice-to-have toward standard practice.

The question is really how you get there. The conventional route is a branch circuit monitoring system (BCMS) specified, wired, and commissioned alongside a mechanical panel, which is effectively a second project on top of the first. Solid-state protection has monitoring built in natively, because the breaker is already measuring current electronically in order to do its job. Atom Halo captures waveform data at 20 kHz on every branch and integrates directly with your BMS, with no separate monitoring layer to buy or install.

What is a solid-state RPP?

A solid-state RPP is a remote power panel that uses semiconductor-based circuit breakers instead of mechanical thermal-magnetic ones. Rather than physically pulling contacts apart to interrupt a fault, it switches the current off electronically. That changes three things: trip speed moves from milliseconds to microseconds, there are no moving contacts to form an arc, and current and waveform data comes off the breaker natively instead of from bolted-on sensors.

Atom Halo RPP is a solid-state panel built for exactly this, putting protection, metering, and switching for every branch circuit inside a single enclosure.

Can a conventional RPP support 30kW to 100kW AI racks?

It can carry the current, but current isn't the whole problem. Thermal-magnetic breakers were designed around rack loads of 5kW to 10kW that sat relatively flat. AI training and inference loads spike hard and fast, and that transient behavior creates three failure modes a conventional panel handles poorly: nuisance trips on inrush the breaker wasn't built to read cleanly, voltage sag across adjacent racks during the 20 to 50 milliseconds a mechanical trip takes to clear, and arc flash exposure that compounds as panel density and switching frequency go up.

Atom Halo's solid-state breakers trip up to 3,000 times faster than a mechanical breaker of the same frame size, which is what lets a fault clear on one circuit without dragging the rest of the shared bus down with it.