I trained as an electrical engineer. The syllabus that took me through college, like nearly every electrical and architectural syllabus I've seen since, taught MEP as a clean three-letter scope: Mechanical, Electrical, Plumbing. Every lecturer described the building's services in terms of those three disciplines. Wiring diagrams, single-line drawings, conduit sizing, panel boards, voltage drop calculations — those filled the electrical portion. ELV got one slide. Sometimes a paragraph in a textbook nobody read.
That was twenty years ago. The curriculum has barely moved. The industry has moved a long way.
Walk into any major building project today — a hotel, a hospital, an office tower, a retail mall, a campus — and look at the procurement structure. The CCTV is its own tender. The access control is its own tender. The structured cabling, the fire alarm signalling, the public address, the BMS, the AV — each one has specialised contractors, specialised consultants, and a separate line in the project budget. None of them are sitting under the MEP electrical engineer anymore. The market figured out, project by painful project, that none of those systems are actually electrical engineering. They're network engineering, with electrical wiring as the substrate.
The educational gap and the industry reality have drifted apart. The architects, builders and project managers caught in between are the ones paying for it.
What ELV used to be — and what it became
Before about 2005, "extra-low voltage" was a fair name for a fair scope. A few CCTV cameras hardwired to a coax DVR. A basic intercom panel. A wired fire alarm with a couple of zones and a hooter. A door access reader connected to a magnetic lock. Each system was electrically simple, electrically isolated, and easily handled by an electrical engineer with the wiring diagrams and a competent installer.
Then everything became IP-based. The CCTV camera became a small computer running an operating system, recording to a network video recorder calibrated to specific bitrates and retention windows. Access control became a network of controllers talking to a server, talking to an HR system, talking to a fire alarm panel. Wi-Fi went from a convenience to a critical service that businesses run on. Fire alarms started speaking the same protocols as building management systems. Conference rooms became AV-over-IP installations with codecs, network bandwidth requirements, and dedicated VLANs.
None of this is in the FE/BE/PE syllabus for electrical engineering. None of it is in the architecture syllabus either. The skills required to design any one of these systems properly are network engineering skills: IP subnetting, VLAN segmentation, Wi-Fi RF planning, PoE budget calculations, NVR storage sizing, firewall rule design, certificate management. Those are the skills of a network engineer or an IT architect, not an electrical one. They look like the same trade from the outside because both involve cables and panels. They're not.
How I know — three patterns I see again and again
Pattern one: the heritage villa where the access points were placed for the photograph
A few years ago we were called in to fix the Wi-Fi at a heritage villa in North Goa. The building was beautiful — old Portuguese construction, eight-inch-thick load-bearing walls, deeply sculpted ceilings, restored woodwork. The architect's brief had been to preserve every aesthetic line. The MEP electrical contractor had been asked to specify the Wi-Fi as part of his standard low-voltage scope.
He did what he had been trained to do. He picked access point locations that were not visible from any major sightline — tucked behind the ceiling moulding in one room, half-hidden inside a service alcove in another, mounted high on a wall behind a beam in a third. The aesthetic intent was honoured perfectly. The radio physics were ignored entirely.
An access point is a low-power radio. The signal it broadcasts is shaped by everything around it: the wall it's mounted on, the materials between it and the device trying to connect, the angle at which the antenna sits relative to the user. An eight-inch laterite or stone wall absorbs RF the way a sponge absorbs water. Tucking an access point behind decorative moulding doesn't just hide it visually — it cuts off most of its useful coverage. Inside that villa, the result was exactly what you'd predict: dead zones in three of the bedrooms, dropped connections in the kitchen, and a guest Wi-Fi network that worked only when guests sat directly under one specific ceiling rose.
The conversation when we came in was instructive. The contractor referred to the access points as "routers" throughout the meeting — a giveaway that he didn't distinguish between the two device classes. When we explained that some of the dead zones could be partly mitigated with better antenna placement, he asked if we could just install access points "with more antennas" — believing, as a lot of people in the trade still do, that more visible antennas means more coverage. (For modern indoor APs, the antenna count usually relates to MIMO stream count and radio chain count, not to coverage radius. Three antennas on a properly placed AP will outperform six antennas on a badly placed one, every time.)
The villa's owners lived with the dead zones for several weeks before the project came to us. By the time we redesigned, much of the original cabling and conduit work had to stay, because chasing new conduit through restored heritage walls wasn't an option. We worked with what we had — a different mix of wireless technologies, better-positioned access points where placement was possible, and accepting compromise where the heritage fabric allowed no alternative. Workable. But not what the client paid for the first time.
None of this was the MEP contractor's fault. He was scoped to do something his training and trade had never prepared him to do. The fault was upstream — at the moment when "Wi-Fi" was added to the electrical scope of works without anyone asking whether the people writing that scope had the skills to deliver it.
Pattern two: the office where the network rack was sized like a small electrical cupboard
An office fitout in a Tier-1 Indian city. The brief specified "a small server room, around 6 by 8 feet" off the corridor near the reception. The MEP team specified the room exactly the way it was specified to them: power outlets for two racks, a single split AC unit calibrated to the room's volume, fluorescent lighting, and a fire detector. Standard small electrical room.
The IT contractor arrived to install the equipment and discovered the room was thermally and electrically wrong for what was being put into it. The two racks held a core switch, distribution switches, the firewall, the controller for the access points, the NVR for sixteen cameras, the access control server, and the patch panels for the entire floor. Together those devices generated several thousand BTUs of continuous heat. The split AC was sized for a small storage room, not for a sealed enclosure with that thermal load. The UPS the MEP team specified for "five minutes" of backup gave nowhere near enough time for an orderly server shutdown when the building power dropped. There was no precision cooling, no hot/cold aisle planning, no rack PDU strategy, no monitoring for temperature or humidity.
The room worked, in the sense that the equipment ran. It also overheated through three summers, throttled regularly, and saw two switch failures that were almost certainly thermal. By the time the client called us in for an audit, the rack room had quietly become the single biggest reliability risk in their entire IT estate — and nobody had ever scoped it as anything more than "a small electrical room with two racks."
The MEP team did exactly what was asked of them. The brief was written by people who didn't know what to ask for. That's the gap.
Pattern three: the hotel where the fire alarm and access control didn't talk to each other
A boutique hotel project I reviewed a couple of years ago. The MEP scope included the fire alarm system. The integrator who supplied the access control did so on a separate purchase order, organised by the operations team rather than the project consultant. Both systems were installed, both passed their respective acceptance tests, both went into service.
The first time the fire alarm activated — a cooking smoke incident in one of the kitchens, no actual fire — every magnetically locked door in the building stayed locked. There was no signal path from the fire alarm panel to the access control system. The two systems had no contractual obligation to integrate, because nobody on the original design team had specified that they should. The result was a building full of guests being asked to evacuate through doors that needed an electronic credential to open. The hotel's general manager had to walk the corridors with a master key. They were lucky it was a false alarm.
Fire-mode integration is one of the most basic interlocks in modern access control design. Any specialist would have specified it on day one. It wasn't specified because the procurement put fire on one tender and access on another, and neither consultant felt it was their job to ask whether the two were going to talk. That's not an exotic design failure. It's the predictable result of treating ELV as "those small low-voltage bits" rather than as an integrated system that needs a single design intelligence overseeing how the parts fit together.
What MEP consultants do brilliantly — and what they're not trained for
I want to be careful here, because none of this is a criticism of MEP consultants as a profession. The good ones are extraordinary. Calculating cooling load for a 200-bed hospital, sizing the chilled water system, working out the diversity factor on a panel board feeding a multi-tenant tower, designing a hydraulic system for a high-rise, balancing a ventilation system across a twenty-storey building — those are difficult engineering disciplines that require years to master. I've worked with MEP teams whose technical depth in their own domain is genuinely awe-inspiring.
But the skill set that makes someone good at sizing a chiller doesn't transfer to designing a Wi-Fi network. Calculating voltage drop on a 100-metre power run is not the same problem as calculating power-over-Ethernet budget on a switch with thirty-six PoE-plus devices. Specifying a panel board is not the same as specifying a network rack. The mathematics is different. The failure modes are different. The vendors are different. The standards are different. Even the vocabulary is different.
Good electrical engineers know this. The ones I've worked with longest are usually the first to say, "Bring in someone who actually does this for a living." It's the procurement side that hasn't caught up — the architects, project managers and clients still writing tender documents that bundle "low-voltage works" into the electrical scope as if the last twenty years didn't happen.
How modern projects should be scoped
The right scope split is not complicated. The MEP team continues to handle Mechanical, Electrical, Plumbing — including the physical infrastructure that ELV systems depend on: power circuits to network rooms, conduit and tray runs for ELV cabling, dedicated electrical supplies to UPS and rack PDUs, mechanical cooling for IT spaces, fire suppression in server rooms. That's properly an MEP scope, and a well-coordinated MEP consultant will deliver it cleanly.
A separate ELV consultant handles the design of what runs inside that infrastructure: the network architecture, the Wi-Fi design, the CCTV system, the access control, the fire alarm signalling, the BMS, the AV, the structured cabling. They produce drawings, equipment schedules and specifications that any qualified contractor can quote against. Critically, they coordinate with the MEP team early — at concept design — so that conduit sizes, rack room dimensions, cooling loads, power circuits and cable tray routes are all agreed before any of it gets built into the construction documents.
That early coordination is the entire point. The disasters happen when the ELV design is bolted on after the building is already detailed. Conduit is too small for the cable bundles that need to go through it. The network room turns out to be too hot for the equipment going into it. The access points end up where the electrical contractor placed conduit, not where the radio physics wanted them. Every one of those problems is cheap to avoid at concept design and expensive to fix at handover.
What architects, builders and PMC firms can do
If you're scoping a project right now, three things will save you a remarkable amount of pain:
Carve ELV out as a separate scope in your RFP. Don't bundle it into the electrical works package. Don't list it as "low-voltage works (by electrical contractor)." Treat it as a parallel discipline with its own consultant, its own drawings, its own specification, and its own contractor. The procurement effort is the same. The design quality is incomparably better.
Bring the ELV consultant in at concept stage, not at fitout. The single biggest waste in ELV projects is design rework caused by infrastructure decisions that were made before the ELV designer was in the room. Network rooms get sited in the wrong place. Conduit gets sized for "two cables" when it needs to carry forty. Cooling loads get calculated for a small electrical cupboard when the room will hold twenty thousand BTUs of IT equipment. None of that is fixable cheaply once the building is detailed.
Ask the ELV consultant to coordinate with MEP, in writing, on a defined list of touchpoints. Power feeds to network rooms. Cooling load and ventilation strategy for IT spaces. Cable tray routes and capacities. Fire suppression in server rooms. Generator and UPS coordination. Lightning and surge protection. Conduit sizing for ELV cabling. None of this is glamorous but every item on this list, mishandled, costs more to fix than the entire ELV consultant's fee on the project.
The curriculum will catch up. Until it does, the projects that don't fail are the ones that scope this properly.
I expect, eventually, that architecture and engineering schools will teach ELV as the discipline it is. New courses will appear. Networking will join structural calculation as a core topic. Building services textbooks will have a chapter on RF propagation. It's coming.
Until then, the gap is going to keep being filled — or, more often, not filled — by good people doing the best they can with training that doesn't cover what the project needs. The architects and builders who recognise the gap, and scope around it, will be the ones whose projects don't fail in commissioning. The ones who don't will keep paying for the lesson, one project at a time.
If you'd like a quick sanity check on whether ELV is properly scoped in a project you're working on right now — current or upcoming — that's a half-hour conversation, no obligation, and almost always reveals at least two things worth fixing before they become expensive.