// INSIGHTS
Power and cabling planning for a technical event
6 min read
Every event technology stack sits on two dependencies: power and copper. If either is planned late, everything above it fails in public. We treat power as a load calculation with named circuits, not a request for a few sockets near the stage. We treat cabling as a schedule with measured runs and labelled ends. This is how we plan both for a technical event in India or the UAE, and what we assume will go wrong.
Power is the first dependency
Start with a load list, not a layout. Every device gets a line: model, quantity, watts, duty cycle. A 1,200-seat plenary in a hotel ballroom with a tented overflow hall typically runs 18-30 kW of AV, 3-6 kW of IT and network, and 2-4 kW of kiosks and signage. Add switchgear losses and you are asking the venue for a real number instead of a guess. House electricians respond well to a number in kW. They respond badly to a request for a few points near the stage.
Then split the supply. AV and IT get clean power, separate from motors, chillers, coffee machines and rigging hoists. Dimmer packs and LED walls put harmonics on the neutral. An isolation transformer on the AV and IT feed keeps that noise off audio lines and out of switch power supplies. Neutral-to-earth voltage should read under 2V at every distribution point before you energise anything expensive. If it reads 6V, stop and find the earth.
Use three-phase where the load allows, and balance it. A 30 kW load poured onto one phase of a 63A supply will trip and take the show with it, so we hold phases within about 10 percent of each other. RCDs and MCBs are sized per circuit rather than per rack, so a failed kettle in the catering area does not drop the network core.
Load, phases, headroom
Copper, fibre and the 90 metre wall
The 90 metre limit is real
CAT6 gives 90m of permanent link plus roughly 10m of patch, and no more. Anything longer goes to fibre with an SFP at each end, which also isolates ground potential differences between halls.
PoE budgets, not PoE ports
A 24-port switch with a 370W PoE budget will not run 24 access points at 802.3bt. We budget 15-25W per AP for 802.3at and up to 45W for bt, then size the switch and its UPS to match.
Protection in public routes
Cable ramps and matting go in every delegate walkway, with tray or overhead catenary where a ramp is not viable. Gaffer tape is for backstage and for temporary fixes only.
Outdoors means IP rated
Tented and open-air builds get IP65 distribution, weatherproof glands and ends raised off the ground. In monsoon we add surge protection at every sub-distribution board and never rely on a single earth pit.
Sequencing the build
Load list and cable schedule
We publish load in kW per circuit and a cable schedule listing source, destination, type, measured length and label. Both freeze 10 days before build so the venue can order tails and book its electricians.
Distribution and earthing first
Boards, tails and earthing go in before a single rack. We test neutral-earth voltage, phase rotation and RCD trip times, then log every reading against the circuit label.
Backbone, then edge
Fibre backbone and switch cores next, then AP drops, kiosk and signage runs. Every cable is labelled at both ends as it lands, not at the end of the shift.
Load test and power-down drill
We run the full stack at load, then rehearse a controlled shutdown and restart in sequence: edge, access, core, UPS, mains. The same sequence is printed and taped inside the main rack.
What usually goes wrong
The DG set arrives undersized. A supplier quotes 40 kVA against a 32 kW measured load and calls it comfortable. Under a 45C afternoon with derating, a lighting inrush and a compressor start, it stalls. We size at measured load plus 30 percent, then apply derating for site temperature. Fuel is the second half of the problem: a 125 kVA set at 70 percent load burns roughly 20-24 litres an hour, and a refuel nobody scheduled becomes a blackout at 15:40 with a keynote on stage.
Cabling fails on boring details. Runs measured on a drawing come up 20 percent short once they route around a fire door and over a beam, so we add 20-25 percent spare on both cable and kW. Unlabelled ends turn a four-minute fault into a forty-minute one. Cable taped across a public route survives about two hours of delegate traffic before it lifts and becomes a trip claim. And the venue that promised fibre handover next week is often still waiting on a last-mile permit, which is why we design bonded uplinks across multiple carriers with automatic failover as a planned path rather than a panic measure.
Common questions
How much spare power and cable capacity should we build in?
We plan 20-25 percent spare on both. Spare kW absorbs the extra LED panel, the heater and the laptops nobody declared. Spare ports and spare fibre strands absorb late signage and a second kiosk. Adding capacity on build day costs three to five times what it costs on paper.
Do we need a UPS if the venue already has a DG set?
Yes. An ATS changeover typically leaves a 6-12 second gap between mains failing and the generator accepting load, and switches, servers and media players reboot in far less than that. We put online double-conversion UPS on network core, control and AV control with a 10-20 minute runtime target. Line-interactive units are acceptable for signage players and little else.
What is different about power planning in the UAE?
Two things dominate. Many venues mandate in-house electrical contractors at fixed day rates, so your labour line and shift pattern are set for you. Heat derating is also not optional for outdoor distribution and generators, and permits for temporary supply need to be started weeks ahead rather than in build week.
Plan the power before the pixels
Send us your floor plan and device list and we will return a load calculation, a cable schedule and a build sequence.
