Commercial HVAC replacement and installation in St. Louis.

Most buildings default to the first option without ever choosing it. Facility teams hold a rooftop unit until it stops holding the space, needs constant service, or fails outright — and DOE is blunt that proactive replacement usually beats replacement-on-failure, and often beats a like-for-like swap at end of service life.

The middle path is the one people forget. DOE's rooftop campaign let owners either replace old units with high-efficiency equipment or retrofit existing ones with advanced controls, whichever made better financial sense. More than 200 organizations upgraded over 43,000 units that way, cutting a combined $37 million a year in energy spend. Controls only work when the cabinet, coil, and heat exchanger are still sound. Once they are not, it is money spent on a unit that is leaving.

No two rooftop replacements price the same, and the spread is rarely in the equipment itself. Make every bidder break out these lines, because each one is where a low number quietly becomes a change order.

The nameplate on the old unit is a record of a decision someone made a long time ago. Occupancy, plug loads, lighting, and the envelope have probably all changed since. Carrying that tonnage forward is the fastest path back to cooling, but it also carries forward whatever was wrong the first time, which is how oversized units end up short-cycling, dehumidifying poorly, and producing comfort complaints nobody can trace.

The second decision is efficiency tier, and it deserves more attention than it usually gets. DOE's rooftop unit challenge set a performance spec aimed at cutting energy use by up to half against the standard of the time, and manufacturers now sell more than twenty models that exceed it. Measured against $900 to $3,700 of annual waste on a tired unit, the gap between a base-tier and a high-tier replacement is often repaid on the operating line rather than the capital line.

The third is availability. Equipment type, efficiency tier, and refrigerant all affect what can actually ship, and a quote that names a price without naming a lead time is half a quote. Ask for both together, and ask what the fallback unit is if the first choice slips.

Two replacements at the same tonnage can be two different projects. Work out which one you have before comparing bids, because it explains most of the gap between them.

A useful walk-through is not a look at the unit. It is a look at everything the unit depends on. DOE's own checklist for rooftop decisions covers the same ground: electrical capacity, physical constraints, gas lines, equipment availability, and whether the contractor pool actually knows the technology being proposed.

Electrical capacity is where projects stall. If the replacement adds any electric heating, the question is not whether the new unit fits the curb, it is whether the panel, the feeder, and the service can carry it. That answer takes a meter and a look at the gear, not an assumption.

Physical constraints come next: curb dimensions, roof structure, clearance around the unit, duct connections, and how a crane gets close enough to set it. On a tight lot the rigging plan drives the schedule more than the equipment does.

Then the heating decision. Keeping a gas section, going dual fuel, or going all-electric each imply different piping, venting, and permit work, and that belongs in the scope before the quote rather than during demolition.

Last, removal. Federal refrigerant rules govern how the charge comes out of the old unit and where it goes, so disposal is real scope with real hours behind it.

The equipment decision and the ownership decision are not the same decision. One is about what lands on the roof. The other is about who pays for what happens afterward, and how the building's fuel choice ages.

Rooftop units condition close to half of all commercial floor space in the country, so whatever you decide here is the pattern you are likely to repeat at every other unit you own. Both questions are cheaper to answer now than at the second failure.

Most commercial buildings do not have one rooftop unit. They have several, bought at the same time and aging at the same rate. That is a schedule problem wearing a maintenance problem's clothes. When four units pass fifteen years together, the first failure sets the price for the rest, because emergency work does not wait for competitive bids.

The alternative is boring on purpose. Rank the units by age, service history, and what each one costs to run. Replace the worst one or two per capital year, reuse the same specification and the same commissioning documentation every time, and let a single crane mobilization cover more than one unit when the roof allows. That last detail alone often pays for the planning.

Procurement gets easier too. A standing specification means you compare the same scope each year, lead times are known before the season starts, and incentive paperwork follows a template. DOE's guidance treats incentives as a standing part of the rooftop decision, and that paperwork is far easier to handle before a unit fails than during the week it does.

Bids only compare when they cover the same scope. Ask every bidder to break out these five items so the difference between two numbers is visible instead of buried.