Installation of a GREE FreeMatch R32 2-Head Heat Pump in Otterburn Park (MRC de La Vallée-du-Richelieu)

The mark on the ceiling is not a defect in the unit: it is your air settling there

In the photo of this Otterburn Park installation, there is something the homeowner has seen every day from their armchair and that no one has ever mentioned to them: a discolored band on the ceiling, directly above and in front of the wall-mounted head.

It is the only symptom in this entire series that is visible without tools, an error code, or a technician. It appears years after installation, is almost always attributed to the paint or the unit, and it is neither.

The trade has one word for it, and building science has another

Among diffuser manufacturers, the phenomenon is called smudging. In inspection and indoor air quality literature, it is referred to as ghosting or dirt streaking—dark marks that appear where airborne particles accumulate on an interior surface.

These are two vocabularies for the same fact, and they converge on a conclusion we have not seen written anywhere in French in our field: the dirt does not come out of the machine. It was already in the room.

Why a band, and why exactly there

The mechanism is documented and simple. A jet of supply air flowing along a surface creates a low-pressure zone against that surface, which keeps it attached—and the velocity is highest right at the outlet. We have already covered this flow behavior in Sainte-Martine, and we are not rewriting it here: what concerns us is not the path of the air, but what it leaves on the building.

An engineering technical bulletin from a diffuser manufacturer, based on the ASHRAE Handbook — HVAC Systems and Equipment, describes what happens next: the surface-attached jet entrains room air with it, and the particles contained in that entrained air settle on the surface in the low-pressure zone adjacent to the outlet. The same bulletin states that work conducted at a materials manufacturer in 1975 confirmed that this dirt comes from the room air, not from the ducts or the coil.

Two engineers from another manufacturer put it even more directly in the trade press: the dark stains around the diffusers are the result of dirt present in the room air, deposited on the ceiling where induction occurs.

Hence the geometry of the mark, which is not random:

• it is a band, not a spot, because the air outlet is long and narrow and the induction zone extends across its entire width;
• it is on the ceiling rather than the wall, because that is the surface to which the air jet clings;
• it is denser near the unit, because that is where the speed — and therefore the negative pressure — is greatest;
• it builds up slowly, because it is cumulative.

What public health and building-science authorities say

This is not merely a manufacturers' issue. A U.S. public health agency published a technical fact sheet on dirt streaking and ghosting that identifies three deposition forces: attraction (charged particles adhere to oppositely charged surfaces), impaction (moving air propels particles against a surface), and gravity. It adds a sentence that describes exactly what our photo shows: dark stains at the junction of the ceiling and walls generally reflect a combination of all three.

From the building-science perspective, Joseph Lstiburek of Building Science Corporation adds a fourth mechanism — thermal deposition: where air meets a colder surface, particles “bounce” less vigorously and settle, explaining why the cold studs in a wall eventually show through the paint.

And these two independent sources agree on the origin of the particles, which is the genuinely useful point for a homeowner:

• candles, identified as the dominant source of modern indoor soot — Lstiburek estimates that their use has increased microscopic particles in indoor air by approximately one hundredfold;
• cooking without the range hood running ;
• tobacco smoke ;
• fireplaces, stoves, and combustion appliances ;
• textile fibers and dust, and air from a traffic corridor.

The public health fact sheet even recommends hard-wax candles, short-trimmed wicks, and avoiding jar candles. Lstiburek, for his part, draws the conclusion that matters to us: these traces reflect occupants' activities and products, not a construction defect.

Why a wall-mounted indoor unit is particularly exposed

Here, we need to be honest about the state of the sources: all the literature we have just cited concerns ceiling diffusers in ducted systems. We found no publication specifically addressing deposition above a ductless wall-mounted unit. The physics is identical and the transfer is reasoned, but it is reasoning, not a citation, and we state it as such.

That said, two characteristics specific to a wall-mounted head point in the same direction.

Filtration is weak by design. A recognized building-science observer notes that ductless wall-mounted units “do not have good filtration”: air passes through a thin filter that does not fit tightly against the coil, and some air bypasses it. We published in detail at Saint-Sébastien what this produces inside the unit—the tangential fan downstream of the cold coil, which condenses and retains what the filter did not stop. The ceiling mark is the other half of the same story: what was stopped by neither the filter nor the fan, and comes out with the jet.

And the position makes the phenomenon worse. On this job, the unit is installed very close to the ceiling—we explain below why there was no other choice. The closer the outlet is to the surface, the sooner the jet adheres to it, and the closer to the unit the induction zone begins.

What the Mark Tells You—and What It Does Not

It tells you that the room air contains enough particles to mark a surface, and that what your filter does is not enough to prevent it. The engineering bulletin cited above puts it the other way around: a well-maintained system whose filters are changed regularly produces little smudging.

It does not tell you which source. Candles, cooking, soot, tobacco, textiles: the sources list candidates, but none allows the culprit to be identified from the mark itself. So we will not do that either.

And it does not indicate a defect. This is the point that changes a client conversation: no one installed anything incorrectly, the unit is not defective, and the painter did not do a poor job. The building is simply recording what the air carries.

What We Will Not Do

We will recommend no cleaning product for this surface, no repainting method, and we will not claim what the deposit contains—we found no source establishing that for this specific case.

Two previously published rules apply, and we are recalling them rather than rewriting them: finishing and painting are not part of the HVAC scope (Saint-Bernard-de-Lacolle), and we never paint the casing of an indoor unit (Dorval), because the paint migrates into the louver pivots and the front-panel seal.

What we do, however, is flag it, photograph it, and record it in the file, with the date. A mark documented in 2026 and found darker in 2029 tells a story; the same mark, never mentioned, tells nothing.

What actually reduces buildup

Without promising to eliminate it:

1. Address the sources rather than the surface. This is the only action whose effect is established in the literature.
2. Run the range hood while cooking, not only when it starts smoking.
3. Maintain the filter at the frequency required by its use, not according to the calendar—we published this principle in Montréal-Est and Pierrefonds-Roxboro.
4. Have the inside of the appliance cleaned periodically, which is a technician's job, not a homeowner's: this is the complete subject of our Saint-Sébastien page, to which we refer.
5. Accept that the surface will eventually need to be cleaned or repainted, and that this will be building maintenance, like repainting a door frame.

The window consumed the strip of wall

The second aspect of this project is geometric, and it explains the position that the first half of this article has just discussed.

Remaining wall: none

Look at the photo. The appliance is installed directly on the window framing, and its upper section almost touches the ceiling. Between the two, there is no wall—there is the appliance.

This is a situation we encounter often and that has no equivalent in our previous pages. We have published extensively about a hood installed above a door—to the point that it is the most thoroughly covered angle in the entire series—and about one installed above a closet opening. A window is a different case, for one simple arithmetic reason: the lintel of a window is higher than that of a door. The strip of wall remaining between its framing and the ceiling is therefore the narrowest in the house.

We established elsewhere that the installation height is a designed variable, bounded above and below (Saint-Polycarpe), that the upper clearance is measured to the lowest obstruction, not to the ceiling (Dorval), that clearances are not symmetrical from one side to the other (Beaconsfield), and that the clear span is measured between the actual obstructions (Rivière-des-Prairies–Pointe-aux-Trembles). All of this applies. What the window adds is the case where the two limits meet: the framing sets the lower limit, the ceiling sets the upper limit, and no room for maneuver remains.

A window lintel carries loads; a closet header does not

Here is the technical distinction that this job allows us to establish clearly.

We published at Terrasse-Vaudreuil that a closet header is not a structural lintel—it is light precisely because it carries nothing above an interior opening. A window is the opposite case. Its lintel carries the load of the wall above it, and sometimes more. It is higher, deeper, and exactly where one might be tempted to drill.

The rule is the one we published at Havelock, and we do not revisit it: never drill into a structural lintel. The passage goes beside it, against the jamb studs, or below the sill. For an installation above a window, this means that the piping route is decided before the exact position of the mounting plate, not the other way around.

Two checks we carry out systematically in this configuration:

• where the lintel actually runs, both in height and depth, before marking anything;
• what is actually inside the jamb—a wire, a conduit, or a counterweight in an old window—knowing that a detector signals the presence of mass, not its nature.

The available strip determined the capacity before the load

This is the selection argument specific to this page, and it completes a family we began elsewhere.

We published at Richelieu that, on a wall whose assembly is uncertain, the substrate determines the lower limit—the cabinet's weight becomes a criterion. We published at Saint-Sébastien that the capacity of the largest head is also a maintenance decision, because a large unit is cleaned differently from a small one. Here is the third member of the family, and it is the most down-to-earth:

When the only available strip is the narrow space between a window frame and a ceiling, the height of the unit becomes a selection criterion before the thermal load.

A 24,000 BTU/h head and a 15,000 BTU/h head are not the same object. On this job, the load of the two rooms could have accommodated several combinations; the available strip allowed only one. So we measured the strip before finalizing the capacities, and we state this in the quote rather than implying that thermal requirements decided everything.

The corollary, worth stating to a homeowner: if the strip is truly too narrow, the answer is not to tighten the clearance further. It is to choose another wall or change the type of indoor unit—a ceiling unit, cassette, or low console is not installed in a strip.

The airflow, the glass, and everything already occupying this clearance

A head above a window concentrates three constraints in the same place that we have addressed separately elsewhere, and it is their simultaneity that is new:

• the curtain rod and its brackets, which occupy the lower clearance (Ormstown);
• the valance of a blind, which can redirect the airflow toward the intake (Saint-Stanislas-de-Kostka);
• the full travel of a roller blind, checked before drilling rather than afterward (Havelock).

On a solid wall, these three elements appear one at a time. Above a window, they all fall within the clearance envelope at the same time, and the only way not to discover them along the way is to measure them during the technical visit, with the blind and curtain installed.

A fourth element must be added, one that belongs to the window itself: a large glazed surface produces a layer of cold air that descends along the glass in winter. We explained in Sainte-Thérèse why an electric baseboard beneath a large window is an anti-downdraft device rather than merely supplemental heating—and why a ceiling-height head does not replace this function. On this project, the baseboard beneath the window remains in place, and the reason is recorded on the sheet given to the client so that no one removes it ten years from now thinking it has become unnecessary.

What really decides things in the end

Two things, and neither of them is thermal:

1. The drain. In a constrained installation, the condensate slope determines the outlet side and the exact height—not aesthetics.
2. The front panel arc. We published in Lachine that the panel lifts upward on many models: a unit mounted tight against the ceiling may meet the intake clearance on paper while making filter rinsing awkward in real life. The height must be checked with the panel open, on site, before fastening the mounting plate.

The selected unit: GREE FreeMatch R32, outdoor unit GWHD(30)ND6MO

Technical data sheet for the installed unit

Two identical 12,000 BTU/h wall-mounted units. Total installed: 24,000 BTU/h, or 80% of nominal heating capacity and 85% in cooling.

Why two 12,000 heads, and why not more

This is where the free-strip argument translates into figures.

The physical constraint came first. A wall-mounted head unit is not chosen solely by BTU: it has a height, and that height had to fit within the remaining strip between the window framing and the ceiling. The two rooms served could have accommodated different capacities; the strip itself required a compact unit. We measured the strip before settling on the capacities, not the other way around.

The modulation floor set the lower limit. It is 8,190 BTU/h on this model, and it is a system property that is not divided by the number of heads. A 12,000 head operating alone at night is approximately 46% above that level; a smaller head would have come too close to it. We published this arithmetic in detail at Noyan and refer back to it rather than repeating it.

The 80% ratio is not a cost saving. The two rooms do not call for their maximum output at the same time — which is the normal condition in a home — and a compressor with reserve capacity stays at low modulation longer instead of cycling.

Deliberate and openly acknowledged repetition, as we now do systematically: this combination — a 30,000 compressor with two 12,000 heads — has already been used once in this series, at Vimont, in Laval, in a context unrelated to this one. We would rather state it than conceal it.

One final word on this model: it is the only one in the lineup whose two performance ratings exceed 4.0 (EER 4.00 and COP 4.19). We noted it at Saint-Césaire, and we are not revisiting it here; on this project, that was not what made the decision.

47 metres of developed length, and 140 g of R32 added

The line sets measure 21 m and 26 m, for a total of 47 m of developed length—the outdoor unit had to be placed away from openings and the overhead network. The GWHD(30) is precharged for 40 m; therefore, 7 m × 20 g/m = 140 g had to be added, or approximately 6% of the factory charge, bringing the system to 2,340 g. The maximum allowable total length of 80 m and maximum elevation difference of 25 m leave a comfortable margin.

The added quantity was weighed and recorded in the file, and the minimum room-area verification required for A2L refrigerant was redone using the final charge rather than the catalog charge.

A brief word about outdoor installation, without making it the focus: the overhead network visible from this window—pole, transformer, conductors—is one of the constraints that must be assessed before choosing the location, not on the day of installation. We published the full reasoning for La Prairie, including our refusal to state approach distances ourselves, since those fall under regulations, Hydro-Québec, and the master electrician. We refer you there.

Otterburn Park: What You Need to Know

It is the Richelieu Valley's city with an English name, and the history is better than the curiosity.

The name comes from Otterburn, in Northumberland, northern England—not Scotland, contrary to what is often written. It was the birthplace of Joseph Hickson (1830–1897), a director of the Grand Trunk. When the park's developer proposed naming it “Hickson Park,” Hickson refused and instead suggested the name of his home village. The Commission de toponymie du Québec reports the episode.

There is an irony here that we did not invent: the main thoroughfare of this city with an English name is chemin des Patriotes, a designation the Government of Quebec gave Route 133 on October 12, 1977, in memory of the Patriotes of 1837–1838.

Otterburn Park was founded as a railway resort. Grand Trunk weekend trains brought a largely English-speaking clientele from Montreal to a park equipped with a dance hall, a bandstand, and picnic areas. Beginning in 1912, residential lots were laid out, notably for Grand Trunk employees, and summer visitors gradually became permanent residents. The town adopted its current name in 1953 and became a city in 1969.

One striking demographic feature: the community was 85% anglophone in 1961, and had fallen to just 13% by 1996. According to the CISSS de la Montérégie-Est profile based on the 2021 census, approximately 7% of the population is English-speaking today. A city with an English name that became overwhelmingly francophone in sixty years.

Three figures, published with their sources, and these are the most useful ones for an installer:

• 9,487 residents, population decree published in the Gazette officielle du Québec in December 2025, based on a provisional estimate as of July 1, 2025;
• 90.6% detached houses and 90% homeowner households (2021 census, CISSS de la Montérégie-Est profile)—an almost entirely single-family residential landscape;
• 27.4% of the population lives in a dwelling built before 1971 (same source).

This last figure is not incidental to the subject of this article. An old housing stock, made of wood and originating from a summer colony converted to permanent housing, is also a housing stock where fireplaces, stoves, and more permeable building envelopes have left decades of particles in the air. The city’s built heritage inventory lists 210 buildings dating from before 1945, including 44 considered to be of particular interest, with a construction peak between 1920 and 1930 and a distinctly rustic character—wood, pitched roofs, and verandas.

The city is also part of the Mont Saint-Hilaire Biosphere Reserve, designated by UNESCO in 1978, the first in Canada. A useful clarification: the core of this reserve, the Gault Nature Reserve, owned by McGill University, is located in Mont-Saint-Hilaire, not Otterburn Park.

We found no classified or designated buildings in Otterburn Park in the Répertoire du patrimoine culturel du Québec—only inventoried buildings. This was a targeted search, not an exhaustive query, so we phrase it this way rather than claiming that there are none.

The mistakes we correct most often

• Attribute the mark on the ceiling to the unit, the installer, or the painter. It belongs to the room air.
• Repaint it without changing anything at the sources. It comes back.
• Drill through a window lintel because it is directly behind the unit.
• Decide on the sizes before measuring the unobstructed strip.
• Measure clearance with the curtains open and the blind raised.
• Check the height with the panel closed, then discover during the first season that the filter cannot be removed.
• Remove the baseboard under the large window because “there’s a heat pump now.”

Grants, warranty, and documentation

The unit is ENERGY STAR certified. Three administrative rules matter more than any amount: the declared AHRI code must match the combination actually installed—the same outdoor model can have three codes depending on the type of indoor units; the model must appear on the list applicable to the program concerned, as these lists are revised; and the installation must be carried out by a company holding the appropriate licences from the Régie du bâtiment du Québec, taking into account the work to be performed, with an invoice containing the information required by the program.

Our other projects in the area

Otterburn Park is the third municipality in the MRC de La Vallée-du-Richelieu where we have published a FreeMatch R32 project, after Beloeil and Mont-Saint-Hilaire. We also have pages on GREE central heat pumps in those same two cities.

The reversal is worth mentioning because it extends exactly the subject of this page: a central system leaves its trace where no one looks—inside the ducts. A wall-mounted unit leaves its trace where everyone looks—on the living room ceiling. This is not a difference in quality, but a difference in exposure. The hidden network is judged by the maintenance it receives; the visible unit is judged every day, by eye, from the armchair.

Our teams serve the entire South Shore, Montreal, Laval, Longueuil, the North Shore, and Montérégie, including Mont-Saint-Hilaire, McMasterville, Saint-Basile-le-Grand, and Chambly.

What we take away from this project

A HVAC installation leaves two kinds of traces: those we record in the file and those the building records on its own. The band on the ceiling belongs to the second category: it is neither a defect, negligence, nor a surprise—it is a slow record of what the room's air carries, deposited precisely where the airflow attaches.

Telling the client before they discover it is the difference between an explanation and a complaint.

For a wall-mounted heat pump, wall-mounted air conditioner, or HVAC maintenance project in Otterburn Park, the MRC de La Vallée-du-Richelieu, or elsewhere in Montérégie, our teams travel to the site, measure the clearances before choosing the wire gauges, check the height with the panel open, and provide a file documenting what we see today so it can be compared three years from now.