The number you guess at SD becomes the number you can't change at CD
The ceiling void is the single dimension on your section that everyone downstream inherits and nobody can unwind cheaply. You set a floor-to-floor height during Schematic Design — often before the mechanical engineer has sized a single duct — and that one line locks in your structural depth, your clear ceiling height, your facade panel module, and ultimately your building's gross height against the zoning envelope. Then DD coordination arrives, the supply ductwork is bigger than assumed, the sprinkler main has to cross under the beam instead of through it, and suddenly the trades cannot physically stack in the void you drew. The honest options at that point are all expensive: lower the ceiling and break the client's clear-height expectation, raise the floor-to-floor and lose a floor under a height limit, or run a clash-resolution marathon that bleeds weeks of fee. Under-allowing the void is one of the most common, most avoidable sources of redesign in the entire delivery process. The fix is to budget the void deliberately, by building type, before the section is frozen.
- The void depth drives floor-to-floor, which drives overall building height against the zoning/code envelope.
- It is cheap to add 100 mm on the section in SD and brutally expensive to find it in CD.
- Office and healthcare are not the same problem — benchmarking by typology is the first move.
Office benchmark: roughly 450-900 mm (18-36 in.) of accessible plenum
A conventional commercial office on a steel or flat-plate concrete frame typically wants somewhere in the 450-900 mm (about 18-36 in.) range of clear void above the finished ceiling, with the middle of that band — roughly 600-750 mm (24-30 in.) — being the comfortable default for a speculative office floorplate with VAV distribution. The low end (around 450 mm) is achievable only with a disciplined, shallow distribution strategy: smaller branch ducts, beams you route through rather than under, and lighting fixtures with a slim back-box. The high end is where you land with deep long-span beams, larger central air handling, or a tenant fit-out you can't predict. Push below ~450 mm and you start losing the ability to cross services over one another, which is the whole point of having a void. If the project uses an underfloor air distribution (UFAD) or a displacement strategy, the ceiling void can shrink because supply air moves into a raised access floor instead — but that depth simply migrates to the floor zone; the section has to account for it somewhere.
- Comfortable spec-office default: ~600-750 mm (24-30 in.) accessible plenum.
- Shallow/tight target: ~450 mm (18 in.) — only with deliberately small ducts and beam penetrations.
- UFAD trades ceiling void for raised-floor void; the total budget doesn't vanish, it relocates.
Healthcare and lab benchmark: much deeper — and often a dedicated interstitial level
Hospitals and laboratories are a different order of problem. The mechanical load is far higher (high air-change rates, 100% outside air in many spaces, redundant systems), the services are denser (med-gas, pneumatic tube, isolation-room exhaust, multiple electrical and data pathways), and the institutional requirement to maintain and modify systems without shutting down a clinical floor is non-negotiable. That combination routinely pushes the ceiling void into the 1 m+ range for an acute-care floor, and for facilities that adopt a true interstitial strategy, the design provides a dedicated, walkable service level between occupied floors — commonly on the order of 1.5-2.4 m (5-8 ft) clear. The interstitial level is expensive in first cost and in overall building height, but it buys a hospital decades of maintainability: technicians service and re-route systems from a full-height plenum without entering — or contaminating — the sterile space below. Research labs follow similar logic for the same reasons (process exhaust, redundancy, frequent reconfiguration). The takeaway for your section: a healthcare floor-to-floor is not an office floor-to-floor with a margin — it is a fundamentally taller stack, and assuming office numbers on a healthcare project is a section-level error you cannot patch later.
- Acute-care ceiling voids commonly exceed 1 m before you even consider interstitial.
- True interstitial service levels are frequently ~1.5-2.4 m (5-8 ft) clear, walkable.
- Driver is not just duct size — it's redundancy, density, and serviceability without clinical shutdown.
Build the void as a budget, not a guess: structure + duct + sprinkler + lighting + clearance
The reliable way to defend a void depth is to stack its components rather than pull a round number from memory. Add them up, take the worst-case vertical crossing point (usually where a main duct crosses a beam and a sprinkler branch crosses both), and that controlling stack — not the average — sets your dimension. A defensible SD budget reads as a sum of these layers:
- Structural depth — the beam/joist/slab band the services must pass under or through. This is usually the single largest line and is set by your span and structural system.
- Largest duct run — size the worst-case main, including external insulation and flange/joint allowances, not the nominal sheet-metal dimension.
- Sprinkler/fire suppression — branch and main piping with hanger drop and the clearance the head needs below structure.
- Lighting and ceiling system — recessed fixture back-box depth plus the suspension/grid and any plenum return depth.
- Clearance and crossing allowance — the air gap that lets two services cross each other and lets a hand or tool reach a hanger. This is the line everyone forgets and the reason 'it added up on paper' still clashed in the field.
- Coordination contingency — a margin for the unknowns of an early phase (tenant fit-out, late equipment selection, growth).
Where this lands in the deliverables: set the section at SD-DD, prove it with a coordination diagram
This is a Schematic Design decision that you validate through Design Development. In SD you set a target floor-to-floor and clear-ceiling height from the typology benchmark and a first-pass component budget — that single section dimension flows into your zoning height check, your stair/ramp counts, and your facade module. In early DD, as the MEP engineer sizes real equipment, you reconcile the assumed void against the actual controlling stack at the worst crossing, and you adjust before the section is frozen and the structural grid is committed. The deliverables that carry this are your building section, your reflected ceiling plan, and a coordination/interstitial section detail at the tightest bay — ideally with a short stacking diagram that shows the trade sequence (structure, then primary duct, then branch services, then sprinkler, then ceiling) so the field knows the intended order. Producing that stacked section diagram early is what turns 'we think it fits' into 'here is the bay where it's tight and here is how the trades stack.'
| Building type | Typical ceiling void depth | Distribution character | Primary driver of depth |
|---|---|---|---|
| Residential / hospitality | ~250-400 mm (10-16 in.) | Tight; minimal central ductwork | Maximizing clear height per floor under a height limit |
| Speculative office (VAV) | ~600-750 mm (24-30 in.) | Moderate; accessible plenum | Beam depth + supply duct crossings |
| Office, shallow strategy | ~450 mm (18 in.) | Disciplined small ducts, beam penetrations | Squeezing floor-to-floor; UFAD relocates depth to floor |
| Lab / research | ~1.0-1.5 m (3-5 ft)+ | Dense; redundant, reconfigurable | Process exhaust, redundancy, frequent change |
| Acute-care hospital floor | ~1.0 m+ ceiling void | Very dense; high air change | Air-change rates, med-gas, multiple pathways |
| Hospital, interstitial level | ~1.5-2.4 m (5-8 ft) clear | Walkable service level between floors | Maintainability without clinical shutdown |
Frequently asked
What is the difference between a ceiling plenum and an interstitial space?
A plenum is the void between the structure and the finished ceiling that houses services and often acts as a return-air path — you reach it by removing a ceiling tile. An interstitial space is a deliberate, full-height, walkable service level between two occupied floors, deep enough (commonly ~1.5-2.4 m) for technicians to stand and work. Offices use plenums; hospitals and labs adopt interstitial floors where maintainability without disrupting the space below justifies the added height and cost.
Why does ceiling void depth drive floor-to-floor height?
Floor-to-floor is the sum of your finished-floor-to-ceiling clear height plus the ceiling void plus the floor/structure depth. The clear height is usually fixed by client or code expectations, so any extra void you need has to come out of the floor-to-floor — which raises overall building height. Under a zoning height limit, an under-budgeted void that grows in DD can literally cost you a floor.
Can I reduce ceiling void depth with underfloor air distribution (UFAD)?
UFAD can shrink the ceiling void because supply air is delivered through a raised access floor instead of overhead ductwork. But the depth doesn't disappear — it moves to the floor zone (typically a 300-450 mm raised floor plenum). Your section still has to budget that depth somewhere, so UFAD is a relocation of the void, not a free elimination of it.
Which single component usually controls ceiling void depth?
Most often it's the structural depth combined with the worst-case duct crossing — the point where a main supply duct has to pass a beam while a sprinkler branch crosses both. The controlling dimension is set at that tightest bay, not by averaging services across the floor. Always size the void from the worst crossing, then add a real clearance gap for the trades to physically pass each other.
How early should I lock in the ceiling void on the building section?
Set a target at Schematic Design using a typology benchmark and a first-pass component budget, then validate it in early Design Development once the MEP engineer sizes real equipment. The goal is to reconcile assumed versus actual depth before the structural grid and section are frozen — finding the shortfall after CD documentation is underway is where redesign cost and schedule slip originate.