Counting Diffusers, Grilles and Registers
A mid-size school has four rooftop units and about four hundred air outlets.
The rooftop units get checked three times. Everybody knows what they cost. The four hundred diffusers, grilles, and registers get counted once, quickly, by someone zooming around a reflected ceiling plan at four in the afternoon.
They're the densest, smallest, most repetitive symbols in the set. They're also the easiest thing on a mechanical drawing to get wrong.
Diffuser, grille, register, louver
These four words get used interchangeably on job sites and they aren't the same device.
Diffuser — supplies air and spreads it in a defined pattern, usually ceiling-mounted. Square or round face, or a linear slot. The pattern is the point: a diffuser is engineered to throw air a certain distance in certain directions.
Grille — a fixed face with no damper. Supply or return, wall or ceiling. Simplest of the four.
Register — a grille with an integral damper, so airflow at that outlet can be adjusted. The damper is the entire difference between a grille and a register, and it's a real price difference.
Louver — exterior, weather-resistant, for intake and exhaust through a building wall. Different product family, often a different supplier, and frequently scheduled separately.
Collectively the industry calls them GRDs — grilles, registers, and diffusers — and they're usually quoted as one package, often by a different rep than the one quoting the big equipment.
The air outlet schedule
They're scheduled like anything else, typically under a heading like AIR OUTLET SCHEDULE or AIR DEVICE SCHEDULE. Common columns:
Mark — SG-1, RG-1, EG-1, T-1 and similar, for supply grille, return grille, exhaust grille, transfer
Type — supply diffuser, return grille, exhaust register
Basis of design — manufacturer and model, usually Titus, Price, Nailor, or Krueger
Size — face and/or neck dimensions, sometimes a range rather than one value
Airflow — a CFM range the device is rated for
Mounting — lay-in T-bar, surface mount, duct-mounted, sidewall
Damper — opposed blade damper, or none
Finish — white powder coat, anodized, custom
Remarks — plaster frames, security bars, sound requirements
Why the tag isn't the count
Here's what makes GRDs different from equipment.
An RTU-1 callout tells you everything: one tag, one unit, look up the row. An air outlet callout is usually three stacked values inside a hexagon — the mark, the size, and the design CFM at that specific location:
SG-1
24" x 24"
400Sizes appear as rectangular face dimensions in inches on some sets and as round
duct diameters in millimetres on others, depending on the engineer and the project.
The schedule row for SG-1 covers a family of devices. The actual size and airflow vary location to location, and the same mark can appear at 150 CFM in a corridor and 400 CFM in a fitness room.
So “eleven SG-1s” is not a line item. Eleven SG-1s at three different sizes is three line items, and pricing them as one is how a GRD quote goes wrong quietly. You're counting mark plus size, not mark alone.
You'll also see a quantity prefix on the callout — 3@ beside the hexagon means three devices at that location off one leader. One symbol, three units.
The seven ways the count goes wrong
1. Reflected ceiling plans. Air outlets appear on the mechanical plan and on the RCP, and often on the architectural ceiling plan too. Counting across sheets without deduplicating inflates the number. Counting from only one sheet risks missing devices that appear on the other.
2. “Typical” rooms. One classroom, one hotel room, one suite is drawn in detail and the note says typical of twenty-four. That's one drawn symbol and twenty-four devices. Same trap as multi-residential suite counting.
3. One leader, several devices. A quantity prefix like 3@ on the callout, or a single leader pointing at a cluster. The symbol count and the device count aren't the same number, and the prefix is easy to skim past.
4. Plenum returns. Some return air paths use the ceiling plenum with no grille at all, or use door undercuts and transfer grilles instead. A room with supply diffusers and no return grille may be correct rather than an omission — but you need to know which.
5. Symbol drift. Square diffusers drawn with an X, linear slots as thin rectangles, sidewall registers as hatched boxes, exhaust as a different hatch. There's no universal standard, the legend varies by engineer, and legends sometimes show symbols that never appear in the set.
6. Zoom fatigue. At sheet scale these are a few millimetres across. Counting them means zooming well in and panning systematically, and the errors happen at the edges of each pan — the device you counted in the previous view and again in this one, or the one that fell between them.
7. Addenda. Ceiling layouts change late. When an addendum reissues an RCP, the diffuser count changes with it, and the superseded sheet stays in the PDF.
A method that holds up
Find the air outlet schedule first and list every mark. This is your key.
Decide which sheet is authoritative for counting — usually the mechanical plan, since it carries the CFM callouts. Use the RCP to cross-check, not to count.
Count by mark and size, not mark alone. Build the tally as a two-column key, and read the quantity prefix on every callout.
Resolve typicals before counting anything else. Find every “typ. of” note and expand it into real numbers first.
Pan systematically — a fixed grid, left to right, top to bottom. Errors cluster at the seams between views, so overlap deliberately.
Check every room has a return path. Not necessarily a grille, but a path. Rooms with supply and no return are worth an RFI.
Reconcile against the schedule. Every mark you counted should exist in the schedule, and every mark in the schedule should appear somewhere in the building. Marks in one and not the other are the errors worth chasing.
Why it's worth the care
On equipment, unit price dominates. Get the RTU model right and a count error on a small accessory barely moves the total.
On GRDs it inverts. Individual devices are inexpensive and quantities run into the hundreds, so the quantity carries the quote. There's no large line item absorbing a counting mistake.
It's also the package most likely to be quoted by several suppliers on the same job, which means your count is being compared directly against someone else's. Coming in low because you missed forty devices doesn't win the job on merit — it wins it and loses the margin.
Where this gets automated
Counting tagged instances across a drawing set is mechanical work — find the symbol, read the callout, tally it, don't double count. It's also precisely the work that consumes an afternoon and produces errors nobody catches until the job is running.
HyperQuote locates every tagged air outlet across the set, groups them by mark, and shows the crop from the drawing behind each one, so the count comes with its own evidence — you're reviewing what it found rather than counting from scratch.
The judgment stays yours. Whether a room genuinely has a plenum return, whether an addendum superseded a ceiling layout, whether a “typical” note covers a whole floor — those need someone who knows the job. The tallying doesn't.
