Rooftop Unit Replacement Planning in Tampa Bay, Florida

Why RTU replacement rewards planning more than almost any other HVAC work

Replacing a rooftop unit is a construction operation wearing a service call's clothing. The unit itself is only one line of the work: there is a crane and its permits and street or parking coordination; rigging crew and disposal of the old unit; the curb the new unit must land on; electrical, gas and controls reconnection; and startup, commissioning and warranty registration at the end.

Most of that effort is fixed per event, not per unit. The crane that lifts one unit could have lifted three that morning. The parking lot closed for one pick is closed either way. The permits, the traffic plan, the crew mobilization — largely the same whether the day replaces one unit or several.

That fixed structure is why reactive replacement is the most expensive way to buy the same equipment. Every unscheduled failure pays the full logistics stack alone, on emergency lead times, at whatever crane availability the week offers. A planned project spreads the fixed effort across every unit in the mobilization and books the logistics on ordinary lead times.

The catch is that grouping requires foresight. Once a unit has failed in an occupied building, waiting for its siblings is not an option. The grouping decision has to be made while all the candidate units still run — which is to say, it has to come from an assessment, not from a breakdown.

The assessment: deciding which units, in what order

A replacement plan starts with a measured survey of the roof, because age alone misranks equipment in both directions.

Each unit gets a condition workup: compressor amp draw against nameplate, refrigerant circuit performance, heat exchanger condition, coil and cabinet corrosion — a fast-moving variable near Tampa Bay's salt air — economizer and controls function, and the unit's repair history where records exist. Refrigerant type matters too: units running phased-down refrigerants carry a service-supply clock independent of their mechanical condition.

Then consequence joins condition. The marginal unit serving a server room, a commercial kitchen, or the only conditioned space customers occupy outranks a worse unit serving storage. A plan that ranks purely on mechanical condition misses what the building actually cares about.

The output is a ranked roster in phases: replace in the near term, replace within the planning horizon, run and monitor. Units in the final category get a deliberate maintenance posture — keep them safe and serviceable, stop investing in major components — and their trend lines get watched so a unit that starts declining faster than projected moves up before it forces the issue.

This roster is also where grouping candidates emerge: units on the same building, reachable from the same crane position, in the same phase. Geography on the roof is an economic input — a crane pick across the building is a different operation than three picks from one setup.

Curbs and curb adapters: the detail that decides half the project

Every rooftop unit sits on a curb — the raised frame that carries its weight, seals the roof penetration, and connects the ductwork below. Whether the new unit matches the old curb is one of the most consequential questions in the whole project, and it gets asked too late on most reactive replacements.

Manufacturers change footprints between generations, and a same-brand swap decades later rarely lands on the original bolt pattern. The choices are a curb adapter — an engineered transition frame that mates the new unit's footprint and duct openings to the existing curb — or curb replacement, which is roofing work: cutting into the roof membrane, structural carpentry, and re-flashing, with the roof warranty implications that follow.

A well-chosen adapter preserves the roof envelope, avoids roofing trades entirely, and dramatically simplifies the change-out day. But adapters are engineered, ordered items with their own lead times, and they must be specified from real field measurements — curb dimensions, duct positions, gas and electrical entry points — not from catalog assumptions. An adapter discovered to be wrong on crane day, with the old unit already rigged off, is the project's nightmare scenario.

This is a core argument for planning: on a planned project, curbs are measured during the assessment, adapters are engineered and delivered before any crane is booked, and structural questions — can the roof carry the new unit's weight where it differs — are answered by review rather than by hope. On an emergency replacement, every one of those steps compresses into the worst week to be doing them.

Crane day: what the logistics actually involve

The crane is the visible centerpiece of an RTU project, and its requirements shape the schedule more than the mechanical work does.

Siting comes first: where the crane can set up to reach every unit in the day's scope, what its outriggers need to bear on, and which drive lanes, parking rows or street sections that position closes. In dense Tampa Bay retail and office settings, the crane plan is effectively a traffic plan, and it may involve municipal permits and off-hours restrictions.

Then sequencing. A grouped change-out day runs as a choreography: disconnect and rig the first old unit, lift it off, land the new unit or adapter, and move down the line while the mechanical crews follow behind on connections. Weather holds a veto — lifting stops for wind and lightning, and a Florida summer afternoon can supply both — so realistic plans build weather slack rather than promising a heroic single day.

Disposal and refrigerant recovery are part of the operation: recovery from the old units per EPA Section 608 before rigging, and lawful disposal of the equipment.

None of this is exotic to a commercial mechanical contractor. The point is that all of it goes better purchased in advance, together, once — which is what a grouped, planned mobilization is.

Phasing around an operating business

Occupied buildings do not stop for equipment work, so the plan has to fit the building's life rather than the contractor's convenience.

The phasing conversation covers three layers. When the work happens: nights, weekends, seasonal closures, the summer break at a school, the days between events at a venue — whatever the building's calendar offers. What is down while it happens: which zones lose cooling during their change-out window, for how long, and what interim measures cover anything that cannot tolerate the gap. And the order of operations: replacing the units serving critical spaces during the mildest available weather, and sequencing so no phase leaves the building's most important areas exposed during the hottest months.

Cooler-season scheduling is Tampa Bay's structural advantage. A unit changed out in January inconveniences almost nobody; the same change-out in August is a race against an occupied building's tolerance. A multi-phase plan that books its heavy lifts in the cool months, year over year, converts the whole program into background maintenance the occupants barely register.

Communication is part of the deliverable: tenants and staff told which areas are affected and when, signage and access plans for closed parking, and a defined point of contact on change-out days. The difference between a disruptive project and a smooth one is mostly whether anyone did this part on purpose.

What we see in the field

The most common way we meet a roof is one failed unit — and an assessment that finds several more units of the same vintage in similar condition. The building was always going to face a cluster; the only question was whether it would face it on a plan or one emergency at a time.

Two patterns repeat. The first is the serial-emergency roof: a building that has replaced units one at a time over several years, paying full crane mobilization on every event, each time on emergency lead times. Looking back at such a roster, the units could often have been grouped into a fraction of the mobilizations — but each individual failure, at the moment it happened, made a single-unit replacement feel like the only choice. That is precisely the trap: reactive replacement always feels efficient in the moment and never is in aggregate.

The second is the curb surprise: a replacement scoped over the phone as a like-for-like swap that turns out, on the roof, to need an adapter nobody ordered — converting a one-day change-out into a stalled project with a unit on order, a crane rebooked, and a space without cooling in the meantime.

What we commit to: units measured before they are condemned, curbs measured before anything is ordered, structural and electrical questions answered before crane day, grouping proposed wherever the roster supports it, and phasing written around your calendar rather than ours. And where a roof does not justify a grouped project — some do not — we say so.

Unit converter