Commercial HVAC replacement and installation in Arlington.

Most buildings face the same three choices when a rooftop unit reaches the end, and the cheapest one is not automatically the right one. The gap between them does not show up at the crane. It shows up on the utility bill every month for the next fifteen years, and in how often the service truck comes back in year eight.

Replacing on failure is the option that decides itself, usually in the worst week of the season and with whatever equipment happens to sit in a regional warehouse. Deciding early is what puts the other choices on the table at all, because efficiency tier, fuel choice, and incentive paperwork each need lead time before they can matter.

There is also a fourth path worth pricing on a unit that still has life in it: retrofitting advanced controls instead of replacing the equipment. For some units that is the better financial answer, and it is a legitimate result of a good assessment rather than a consolation prize.

These are the items that move a rooftop quote the most, roughly in the order they surprise people. Two identical units on two different roofs can land far apart, and the reason is almost never the box. It is what sits under it, what feeds it, and what a crane has to do to reach it.

The most expensive rooftop replacement is the one you did not choose. Facility teams often wait until a unit stops holding a space, needs constant service, or fails outright, and then buy whatever can be on a truck this week. Replacing an inefficient unit on purpose usually beats replacement on failure, and it often beats a like-for-like swap at the end of service life too, because planning is what gets efficiency and fuel choice onto the table at all.

Age is the cleanest trigger to watch. Units past fifteen years waste a real amount of energy against current equipment, and the high-performance models now on the market were designed to cut rooftop energy use by up to half compared with the old baseline. More than twenty models exceed that specification today, so the efficient choice is no longer a special order with a mystery lead time.

If you run more than one building, treat this as a fleet decision instead of a series of emergencies. Rank units by age, service history, and the hours the space actually runs, then replace on your calendar. Organizations that worked through their rooftops this way have upgraded more than 43,000 units and cut a combined $37 million a year from their energy bills.

Complexity is not about the size of the unit. It is about how many other systems the replacement touches. Reading your own roof against these two lists is the fastest way to know whether you are looking at a one-morning swap or a project with a design phase in front of it.

A quote that arrives without anyone climbing the roof is a guess wearing a number. The review that makes it real is short, but it has to cover the same ground every time.

Start with infrastructure. Electrical capacity is the item that most often changes the plan, especially when the new unit adds electric heating or replaces gas heat entirely. Physical constraints come next: whether the existing curb matches, whether the roof carries the new weight, whether there is service clearance around the unit, and how a crane reaches the spot. Gas lines get a decision either way, since dual fuel keeps them and all-electric means capping and removing them.

Then confirm the equipment itself. Availability is part of the design, not a footnote. The model that fits the roof is worth nothing if it lands in twelve weeks and the building needs cooling in two. Building type belongs in the same conversation, because a warehouse, a restaurant, and a small office put very different loads and runtime hours on the same nominal tonnage.

Finally, confirm the crew. Newer equipment and newer refrigerants both assume a contractor who has installed and commissioned them before.

The install takes a week. What you own afterward is a machine that has to be charged, serviced, and eventually retired under rules that attach to the equipment rather than to the building. Both of the considerations below shape what service costs over the next fifteen years, and both are easier to manage when replacements happen in planned batches instead of one emergency at a time. A roof where every unit arrived in a different crisis year tends to be a roof where no two units share parts, controls, or a service procedure. That is a slow, invisible cost, and it gets paid by whoever is up there at two in the afternoon in August trying to get a space back.

Price the project on what it costs to own, not on what it costs to buy. An old unit wasting somewhere between $900 and $3,700 a year is charging you for the privilege of not being replaced, and that meter runs every year you defer. Set that figure against the difference between a like-for-like unit and a high-efficiency one, and the payback question usually answers itself before anyone starts arguing about the bid.

Incentives move the math further, and they need to be checked before the order goes in rather than after. Program incentives are generally tied to the efficiency tier of the specific model, which means the paperwork has to start while the equipment selection is still open. Once the unit is on the roof, that door is closed.

Phasing is the other lever. When several units are near end of life, replacing them in planned groups buys better pricing, one mobilization, one crane, and a single round of permits and inspections instead of three separate ones. It also lets you standardize on a single platform, which pays back quietly for years in service and spare parts. The unit that fails in August will still fail in August. The only question is whether it fails on your plan or on its own.

Bids only compare when they cover the same scope. Ask every contractor to break out these five, in writing, and most of the price differences between them stop being a mystery.