Condenser Coil Cleaning: Prove It Worked, Don't Eyeball It

“The coil looks fine, we washed it in the spring.” That sentence usually ends the conversation about a rooftop unit that trips every afternoon in a heat wave — and it is the single most common misconception in the trade. Condenser coil cleaning succeeds or fails on the air-entering face, not on the side you see when you walk up to the unit, and the fouling that costs capacity packs itself between the rows where no visual check reaches. A coil can shine and still be choking. The only way to know whether the wash accomplished anything is to bracket it with a measurement.

Why Montreal rooftops foul faster than the maintenance schedule assumes

Three local sources dominate here, and none of them follows a calendar.

Cottonwood first: a few weeks in June are enough to felt an air-entering face in white fluff that then compacts under the first rain. Kitchen grease second. On a downtown or Plateau rooftop, a restaurant hood throws an oily plume that settles on neighbouring equipment; the resulting film will not rinse off with a garden hose and glues airborne dust in place as the season goes on. Construction dust third — in a densifying block, concrete cutting and demolition put fine dust over every flat roof in the Montreal area for months at a time.

Layout adds to it. Many rooftop unit condensers sit less than a metre from a parapet or from the next unit, which sets up hot-air recirculation: the machine re-ingests its own discharge, and the effect on head pressure looks almost exactly like a dirty coil. That is the real argument for measuring first — otherwise you clean a clean coil to solve a placement problem.

Condenser coil cleaning: fin-and-tube or microchannel?

Identify the coil before any product comes out of the truck. A microchannel treated like a copper-aluminum coil gets holed.

For microchannel coils, Kaltra’s guidance is water in a pH range of 4.5 to 8.5, compressed air at 3 to 5 bar for the dry pass, and a ceiling of 40 bar if high pressure is used at all — atomizing nozzle, at least 400 mm of standoff, and an approach angle under 25 degrees to the coil face. Those numbers are worth memorizing, because they explain exactly why a pressure washer wand held at six inches folds fins every single time.

Tools — gauges with clamp temperature probes, outdoor air thermometer, pressure-temperature data for the refrigerant in service, vacuum or blower, low-pressure sprayer, cleaner matched to the coil type, fin comb, locks and lockout tags.

⚠ Safety — On a roof, as soon as a worker is exposed to a fall of more than 3 metres, the employer must install a guardrail; where that is technically impossible, a harness meeting CSA Z259.10 tied to a purpose-designed anchor point is required. On the refrigerant side, Quebec’s Halocarbons Regulation restricts servicing of refrigeration and air conditioning equipment to holders of the environmental qualification, and releasing refrigerant to the atmosphere is prohibited — connecting a gauge set is a regulated act, not a casual one.

Five steps that leave evidence behind

1. Take the baseline reading

Unit running, conditions stable: record outdoor ambient and saturated condensing temperature. The difference is the condenser split. It is the only figure that will tell you, an hour from now, whether the wash did anything. Write it down — not from memory, not as an estimate.

2. Lock out and look at the work area

De-energize, lock out, verify absence of voltage. Then look at the work area before you look at the equipment: distance to the roof edge, access route, available anchor. Pull PPE that matches the safety data sheet of the cleaner — an alkaline product sprayed upward at face height calls for sealed goggles and chemical gloves, not standard safety glasses.

3. Take the debris out dry

Fluff, leaves, plastic bags, dust: remove them with a vacuum or compressed air first, against the normal direction of airflow, so debris travels back out of the coil. Starting with water turns dry cottonwood into paste and drives it into the core, where nothing will reach it again.

4. Wash to suit the coil

On fin-and-tube, apply the cleaner bottom to top, respect the dwell time without letting it dry, then rinse thoroughly at low pressure, again against airflow. The rinse matters as much as the product: alkaline residue left in place keeps attacking the aluminum after you leave. On microchannel, stay with water, keep the nozzle back, and accept that several patient passes beat one aggressive one.

5. Restart and measure again

Comb the folded fins, restart, let conditions settle, then recalculate the split. A genuinely cleared coil brings that number down noticeably. If it does not move, the coil was never the problem: hot-air recirculation, a failing condenser fan, overcharge or non-condensables in the circuit.

What condenser split should you be aiming for?

The target depends on the vintage of the equipment, because coil surface per ton grew as efficiency ratings climbed. HVAC School’s reference points frame the reading well: roughly 30°F over ambient on older 6–10 SEER equipment, around 25°F for the 10–12 SEER generation, 20°F for 13–15 SEER and about 15°F on 16 SEER and above — with a working rule of thumb of 15 to 20°F over ambient, plus or minus 3°F, on modern systems.

Those reference points come out of residential and light commercial practice: on a large packaged unit they are an order of magnitude rather than a setpoint, and the manufacturer’s data governs. So the point is not to memorize the table. It is that a 30°F split is excellent on a 1998 unit and alarming on a 2020 one. The measurement only means something against the target for the machine in front of you, which is why the nameplate year and rating belong in your notes before you draw any conclusion.

The unit that tripped every afternoon in the heat

Office building in Saint-Laurent, three rooftop units, one of them locking out on high pressure daily between 2 and 4 p.m. since the start of the summer. The coil had been hosed off in May and looked clean. Baseline taken before any intervention: a condenser split well above target for that generation of equipment, while charge and compressor readings showed nothing unusual.

Backlighting the air-entering face from inside the compartment settled it — an even greasy film loaded with dust through the full depth of the first rows, from a ground-floor restaurant hood discharging six metres away. Water alone was never going to touch it. A degreasing cleaner suited to copper-aluminum, full dwell time, complete counterflow rinse, fins combed out: the split came back into range and the lockouts stopped. The report to the building manager was two numbers, before and after — not a photograph of a wet coil.

What should trigger the next cleaning

A useful interval is not annual on principle; it falls out of the drift in the measurement. Fold condenser split into the maintenance route and you get a curve per unit, so cleanings get scheduled when the curve climbs rather than when an anniversary date comes around. On equipment exposed to a grease source or an active site, that same logic justifies two visits a season; on an open roof, it saves you from washing for nothing. It is the same discipline behind an air-side economizer verification: what is not measured has not been checked.

Start with the unit that locked out during this summer’s heat — its split has the most to say. One reading before, one after, two lines in the record: the task becomes evidence.

There is a direct payoff for whoever owns the building. A condenser split logged visit after visit replaces a promise of maintenance with a defensible history, and that history is what Montréal Combustion builds replacement recommendations on — rather than the year stamped on the nameplate.