The Hidden Capital Killer on Your Roof
Uncontrolled runtime shows up on the energy bill. It also draws down the service life of your rooftop units, and across hundreds of locations that adds up to a capital consequence that is rarely priced.
In part one we priced what doing nothing about a legacy thermostat costs on the energy line: roughly $610,000 over three years for a 40-location portfolio, built on measured utility savings. That is the cost finance can see. This brief follows the same runtime to the place it never shows up on a bill. Read part one →
Many CFOs meet their rooftop units twice.
The first time is on a depreciation schedule, where a fleet of packaged HVAC units sits as a single line with a 15-year assumption attached to it. The second time is when a facilities manager calls to say a unit failed in July and the quote came back at $20,000 with a crane charge on top.
Between those two moments, there is often little visibility into how hard the asset is working. The depreciation schedule implies the units are consumed by time. In practice they are consumed by duty, and runtime is one of the few drivers of that duty an operator can actually change.
Every hour your rooftop units run in an empty building costs you twice.
The first charge arrives on the utility bill, as an energy cost you can see. The second never arrives as a bill at all. It is the additional consumption of a asset, and it is rarely booked until the replacement quote shows up early. On a 100-location, 350-RTU fleet, the model below values that second charge at roughly in present value. Across 1,000 locations, roughly .
You can't control the age of the fleet you inherited, the weather, or how the equipment was installed. You can control how much unnecessary runtime you allow from this point forward.
In one national lab estimate, equipment cost per hour of runtime is similar to, or higher than, the electricity cost.
A 2018 NREL report funded by the Department of Energy, building on earlier work by Li and Braun, found that the equipment cost per unit of run time is similar to, or higher than, the electricity cost per unit of run time.
The study was estimating what equipment faults cost when they make units run longer, so it is a guide to scale rather than a measurement of any one fleet. Applied to unnecessary runtime, it suggests an extra hour costs more than its kilowatt-hours: it may also use up a comparable amount of equipment life. A setting changed and never set back can quietly move a capital requirement forward in time.
Nobody stamped 15 years on the side of that unit.
Fifteen years is a defensible industry benchmark for rooftop units, and it comes from ASHRAE's older service life table. ASHRAE's later field database puts the median for air-cooled packaged DX units at about 22 years, and individual units land well on either side of any median. ASHRAE is explicit that service life depends on many factors: maintenance, reliability, environment, system requirements and obsolescence among them.
The replacement date driving your capital plan is not a fixed property of the fleet. It is partly an output of how the fleet is operated, and the range around it is wide. This brief isolates one of those factors: the one you can manage.
Rooftop components accumulate duty in operating hours and start cycles, not calendar years.
Compressors, among the most expensive components to replace, accumulate wear with hours of duty.
Climate shows the mechanism at scale. Systems in hot climates run 2,500 or more cooling hours a year against 1,000 to 1,500 in moderate ones. You cannot change the weather in Phoenix. But the hours a building runs while it is empty can occur anywhere, and those are the hours in question.
Only 23% of 5,706 air handling units fully met the code requirement to shut down in empty hours.
The figure comes from a study in the ASME Journal of Engineering for Sustainable Buildings and Cities that examined unoccupied shutdown control across 5,706 air handling units in 843 commercial buildings. Only 23% aligned completely with the ASHRAE 90.1 requirement to shut down fans and outdoor air ventilation during unoccupied hours. It is not a measure of RTUs in restaurant or retail portfolios specifically. What it does show is that equipment running through empty hours is common, not exceptional.
And it is not necessarily an operating failure. In many multi-site portfolios it is a visibility and control problem. Locations are acquired, equipment is inherited, thermostats are overridden, schedules change and local conditions drift. Even a good schedule gets interrupted by real events: a late delivery, a private event, a hot afternoon. Someone overrides it or changes a setpoint on the spot, and the change is not always set back afterward. Until someone changes it again, that setting behaves like a locked hold. Each one that lingers adds runtime nobody planned. Without portfolio-level visibility, a central team has little practical way to know which units are running when they shouldn't be.
Across a large multi-site retail portfolio, one client's thermostats were almost universally locked on hold, so the equipment ran on a schedule nobody had chosen. After deployment, measured runtime across the portfolio fell 20%, and utility bills fell 8 to 11%.
across the portfolio
bills, measured
We don't claim to know how much life a runtime reduction buys. We show you what it's worth at whatever level you believe.
Known: in a multi-site retail portfolio, measured runtime fell 20% and utility bills fell 8 to 11%. Known: equipment life varies widely, and unnecessary runtime adds duty. Unknown: exactly how much of a runtime reduction turns into slower asset consumption. So rather than assert an answer, the model makes it your input.
For example: if 60% of a 20% runtime reduction translates into slower asset consumption, a -year planning assumption extends to approximately years. On 350 RTUs, deferring a replacement wave from year to year is worth about in present value on the next cycle, and lowers the annualized replacement charge from to a year.
Advanced assumptions replacement cost, planning life, replacement year, discount rate, RTUs per location, avoidable energy
Capital deferral: present value at a discount rate of the full fleet replacement in year versus the same replacement deferred by the modeled life extension. Attribution is the share of the runtime reduction assumed to translate into slower asset consumption; the rest is treated as energy-only. Energy savings are year one, undiscounted, for locations at of avoidable energy per location per month, before any rate escalation. These are modeled planning figures, not accounting entries.
Every location adds rooftop units. Every rooftop unit adds to the capital consequence.
The math is linear, which is what makes it easy to miss. At one store the effect is small. Repeated across a chain it can become a multi-million-dollar capital-planning question that rarely shows up as a single line. Figures use the runtime reduction, attribution and assumptions set in the model above, at RTUs per location.
Extending equipment life does not eliminate capital spending. It relocates it.
Out of the "found out in July" column and into the "budgeted in Q4" column.
Emergency work tends to carry these costs plus expedited freight and premium labor.
Runtime monitoring can surface degrading units before they fail, giving facilities a chance to move a replacement from an emergency call to a planned project. How many replacements shift, and what each one saves, will vary by portfolio.
For illustration only: if 20% of a -unit fleet's replacements moved from emergency to planned, and each avoided $5,000 in premiums, that would be across the cycle. We have not included this anywhere in the figures above.
Where this argument stops.
Ask your facilities team how many hours a year each unit actually runs, and how many of those hours fall outside your operating schedule.
Many teams cannot answer. That is rarely a failure of the team. Usually the data has not existed at the portfolio level.
Most organizations track depreciation closely. Few measure runtime, one of the variables that determines how fast the underlying assets are consumed. A fleet worth millions is being drawn down at a rate that is rarely tracked, against an assumption that is rarely tested, and at least one driver of that rate is adjustable. Across every location, that can add up to a meaningful capital-planning exposure.
Measured energy savings create the immediate operating case. Capital preservation adds a second, longer-term return.
HVAC controls and refrigeration monitoring for light commercial, multi-site operations. A smart commercial thermostat and a cellular-connected dashboard covering every location. No Wi-Fi, no IT involvement, installs in under an hour per site.
