Submetering Rollout Plan for Multi-Site Companies

Why should every site get a main-incomer meter before any site gets granular submeters?

Because a comparable baseline across all branches is worth more than deep detail at one branch: phase 1 should put one Class 1 (±1%) meter on the main incomer of every site, and nothing else.

Multi-site rollouts most often stall the opposite way. A team picks a flagship building, meters forty circuits in it, and spends the budget and a year of attention there — while thirty other branches stay invisible. Main-incomer metering everywhere gives you portfolio benchmarking almost immediately: kWh per square metre, kWh per operating hour, base load overnight and on weekends. Those comparisons are what expose the outlier sites, and the outliers — not the flagship — are where granular submetering earns its cost. Depth without breadth is a science project; breadth first turns depth into a targeted investment.

How do you choose the two or three pilot sites?

Choose pilots for what they can teach you, not for how they will look. A sound mix is one typical branch that represents the majority of your portfolio, one awkward branch (old switchgear, cramped panels, weak connectivity), and — if you operate one — one large or complex site. The pilot phase exists to break your assumptions cheaply. It answers questions no datasheet can: whether split-core CTs physically fit around the busbars in your standard panel, whether the site needs a 4G router because IT will not open the LAN, how long one meter really takes to install and commission, and who inside the branch has authority to approve a brief shutdown. Run the pilots until the data has been stable for several weeks and every lesson is written into your rollout standard — then, and only then, order hardware for the first wave.

Why must hardware and protocols be standardized before scaling?

Because per-site improvisation is cheap on day one and expensive forever after. Every additional meter brand or model brings its own register map, its own configuration tool, its own spare-parts line, and its own learning curve for whoever maintains it; multiply that by thirty branches and the portfolio becomes unmaintainable. Standardize a short catalogue before wave one: one approved meter model for main incomers and one for branch circuits, one CT type per panel situation (split-core for retrofits, solid-core where panels are being rebuilt), one field protocol — typically Modbus RTU over RS-485 to a site gateway — and one upstream path from the gateway to your head-end. Bake the electrical design rules into the template too: RS-485 practice allows 32 standard unit loads per segment and cable runs up to about 1,200 m, so daisy-chain layouts should be drawn against those limits rather than discovered against them. Put all of it on a one-page site standard, and make conformance to that page a condition of every purchase order and every installer contract.

What naming convention should be fixed before meter number two is installed?

Fix identifiers before data exists, because renaming a live meter breaks its history and forces manual data surgery later. A workable pattern is SITE-PANEL-SYSTEM-SEQUENCE, for example RUH-014-MDB1-HVAC-02: branch RUH-014, main distribution board 1, HVAC circuit, second meter of that class. Three rules keep it clean. First, take site codes from the registry you already trust — the ERP or branch master list — never a second, parallel list. Second, use a closed vocabulary for systems (HVAC, LIGHT, PLUG, KITCH, PUMP, EVCH) and forbid free-text device names. Third, apply the ID at every layer: printed on a physical label at the meter, set as the device name in the gateway, and used as the point name in the head-end and reports. A meter whose data cannot be traced back to a specific breaker by a technician who has never visited the site is not commissioned — it is decoration.

Which commissioning checks catch the errors that silently poison the data?

Three checks catch most of the damage: CT direction, CT ratio configuration, and phase association — all cheap on day one and expensive to discover in month six. A reversed CT shows negative active power or an implausibly poor power factor; verify that active power is positive on all three phases while the load is running. A wrong ratio — say the meter configured for 400/5 A while the installed CTs are 200/5 A — silently scales every reading; verify the configured ratio against the CT nameplate, not the design drawing. A crossed phase pairing — the voltage reference of L1 matched with the CT of L2 — produces plausible-looking but wrong power factor and energy. Add two sanity checks: compare the meter's power against a clamp-meter reading at handover, and over the first weeks confirm that the sum of the submeters stays at or below the main incomer. Record every check on a per-meter commissioning sheet with CT serial numbers and configured ratios; no meter counts as installed until its sheet is signed.

What does a wave-by-wave rollout actually look like?

It looks like the table below: prove the method, then spread wide, then deepen only where the data justifies it.

Keep waves small enough that the same trained crews do every site, and hold a short review after each wave: what took longer than planned, what the standard missed, which sites failed the data-quality gate and why. A wave does not start until the previous wave has passed its gate — schedule pressure is how portfolios end up with hundreds of meters and no trustworthy data.

How ENTEK helps

A standardized rollout only stays standardized if procurement holds the line wave after wave. ENTEK lets enterprises in Saudi Arabia and the Gulf request quotations for the same specified hardware and installation scope from verified suppliers, compare bids branch by branch, and manage the resulting orders across all sites from one place — so wave three is bought to the same standard as wave one, and site teams spend their time commissioning meters rather than chasing paperwork.