A portfolio-wide utility comparison that ignores billing model and climate will reliably flag the wrong buildings as problems. Getting the comparison right takes three adjustments before a single dollar figure means anything.
Key takeaways
Before comparing dollars per suite across a portfolio, confirm each building is billed the same way, because the billing model itself explains a large share of apparent variance. A sub-metered building in Ontario is governed by the Energy Consumer Protection Act, 2010 (ECPA) and the Ontario Energy Board’s Unit Sub-Metering (USM) Code, administered through licensed Unit Sub-Meter Providers (USMPs). the OEB requires the local utility’s charges to be “passed through to the customer with no mark up,” on top of which the USMP adds its own service fee A bulk-metered building, by contrast, allocates one utility bill across suites by a formula — square footage, occupancy, or a flat per-suite share — and that allocation method, not actual consumption, drives what shows up as each suite’s “cost.” Comparing a sub-metered building’s actual-consumption billing against a bulk-metered building’s allocated billing as though they measure the same thing is the single most common error in a portfolio-wide utility review.
The USM Code carries consumer-protection obligations that mirror, but do not exactly copy, the rules for a regular utility account: security deposits and arrears payment agreements follow similar principles, and the principal consumer — usually the building owner or manager — makes the decision on disconnection for non-payment, which the USMP then carries out Two protections worth knowing before a variance review turns into a dispute: the OEB’s winter disconnection ban does not apply to multi-unit buildings serviced by a USMP, and qualifying tenants can still access the Low-income Energy Assistance Program (LEAP) and the Ontario Electricity Support Program (OESP) through a sub-metered account the same as any other. A portfolio manager fielding a tenant complaint about a sub-metered bill should know these apply before assuming the complaint is really about the underlying consumption.
Statistics Canada’s household energy consumption tables confirm what a multi-province portfolio already sees in its own numbers: household energy consumption, reported in gigajoules per household, is broken out by province and by dwelling type in its Households and the Environment Survey series That variance is climate and housing-stock driven, not evidence of inefficiency in any one building — comparing a Winnipeg building’s utility cost per suite against a Vancouver building’s without normalizing for climate is comparing two different physical problems. Where a real weather-normalization is needed, Environment and Climate Change Canada publishes climate normals by weather station that can be matched to a building’s location and used to build a degree-day adjustment, rather than assuming a flat percentage difference between regions.
A useful variance review holds three things constant before comparing: billing model (sub-metered vs. bulk), climate (via a degree-day adjustment, not a flat regional guess), and building type (a low-rise walk-up and a high-rise with a shared HVAC plant do not share a baseline). Once those are controlled for, a genuine outlier — a building whose per-suite cost sits well outside its own peer group after normalization — is worth a physical investigation: a stuck damper, a failed common-area lighting timer, or a billing error on the utility’s side, all of which are common findings once the comparison is actually apples-to-apples.
One honesty check worth building into the process: StatCan’s household energy consumption series is published on a multi-year cadence rather than monthly, so it is the right tool for confirming that provincial variance is real and roughly how large it is — it is not a substitute for a portfolio ’s own current utility bills, which move month to month with actual weather and occupancy. Treat the StatCan data as the backdrop that explains why cross-region comparisons need adjusting at all, and run the actual variance review on current billing data, quarterly at minimum and monthly for any building already flagged as an outlier the previous cycle.
Worked example — the same $180 spread across two normalizations (illustrative)
A portfolio manager compares three buildings and finds electricity cost per suite ranging from $95 to $275 a month, with a mid-size Winnipeg building at the top of the range and a similar-size Halifax building near the bottom — a raw $180 spread that looks alarming on a single spreadsheet tab.
Applying a degree-day adjustment pulled from each city’s ECCC climate normal narrows the Winnipeg building’s outlier position substantially — heating-season severity accounts for most of the raw gap, consistent with the climate-driven variance StatCan’s provincial energy consumption data already shows at a national scale.
What survives normalization is smaller but real: the Winnipeg building still runs modestly above its climate-adjusted peer group, and a walk-through finds a common-area heating timer left on a 24-hour schedule instead of the intended occupancy schedule — a genuine, fixable finding that the raw $180 figure would never have isolated on its own.
Related reading: tracking the fix once a real mechanical outlier is found, carrying utility baselines forward through a management change and holding an HVAC or controls vendor accountable for what the review finds.
A sub-metered building bills actual suite-level consumption; a bulk-metered building allocates one bill by a formula such as square footage or occupancy. The two numbers measure different things, so a direct comparison compares billing method, not real usage.
No. The OEB confirms its winter disconnection ban does not apply to multi-unit buildings serviced by a licensed Unit Sub-Meter Provider, even though similar deposit and arrears rules otherwise apply.
Not by itself. StatCan’s provincial energy consumption data shows real, climate-driven variance across Canada — a raw cost comparison has to be weather-normalized before it means anything about building performance.
A building whose per-suite cost sits well outside its own peer group after controlling for billing model, climate and building type — not a raw dollar figure that has not been normalized at all.
A 30-minute call is enough to tell you whether AI pays for itself here.
