Swimming Pool Acoustics — The Hardest Room in Architecture to Fix | AcousPlan

I have measured acoustic spaces in hospitals, concert halls, cathedrals, underground car parks, and industrial factories. The swimming pool natatorium is the most acoustically hostile environment in common architecture. It combines enormous volumes, near-total absence of absorption, 100% relative humidity that makes conventional treatment impossible, and a chemical environment that degrades most materials within months.

The result is a room where RT60 values of 4–8 seconds are normal, noise levels during competition events regularly exceed 100 dB(A), and speech intelligibility for safety announcements approaches zero. Pools are among the few environments where acoustic failure has direct safety consequences: a lifeguard who cannot make a clearly understood PA announcement during an emergency cannot execute a safe pool evacuation.

If you are designing a natatorium, or inheriting one that needs remediation, this is the full picture of what you are dealing with and what the solutions actually cost.

Why Pools Are Acoustically Extreme: The Full Analysis

A standard 25 m × 12.5 m recreational swimming pool with a 5 m flat ceiling presents:

Volume calculation:

• Pool hall: 25 × 12.5 × 5 = 1,562 m³
• Concourse areas and changing room access: add ~600 m³
• Total: ~2,150 m³ (modest by natatorium standards — competition 50 m pools reach 8,000–15,000 m³)

Sabine RT60 at 500 Hz (ISO 3382-2:2008 §A.1):

A = 1,234 × 0.022 ≈ 27 sabins

RT60 = 0.161 × 2,150 / 27 ≈ 12.8 s

That number looks extreme but is not a calculation error. It reflects reality: an untreated pool is essentially an echo chamber. The water surface has among the lowest absorption coefficients of any material (comparable to smooth concrete). Ceramic tiles and concrete are equally reflective. There is nothing in a standard pool hall to absorb sound.

In practice, furniture, equipment, splash water aeration, and the pool water's surface roughness during use raise absorption slightly. Measured RT60 in untreated pools typically falls in the 4–8 s range rather than the theoretical 12 s, because real conditions are always messier than the Sabine calculation. But 4–8 s is catastrophic for any intelligibility requirement.

The Octave-Band Disaster

Mid-frequency RT60 of 4–8 s is bad. The octave-band picture for a pool is worse because there is essentially no low-frequency absorption in a hard concrete shell:

The low-frequency tail is structurally driven in concrete pools. At 63 Hz, the walls themselves are driven into flexural resonance by low-frequency excitation (competition start signals, PA system bass content), extending RT60 further. A 200mm reinforced concrete wall has a critical coincidence frequency around 80–120 Hz — within the region where this structural coupling is strongest.

This creates a specific problem for PA announcement intelligibility. Start signals and PA announcements contain significant energy below 200 Hz. That energy rings in the pool shell for 8–12 seconds after the signal stops. The next announcement begins before the previous reverberant tail has decayed. The intelligibility of stacked announcements approaches zero — a serious problem for competition start procedure and a safety-critical problem for emergency announcements.

Noise Level Dynamics: The Feedback Loop

A pool with RT60 of 5 s and 80 bathers develops a noise level that follows the inverse relationship between room constant and steady-state SPL:

For a diffuse field, steady-state SPL from N simultaneous talkers/splashers of sound power W:

SPL_reverberant = SWL_total + 10 log(4/R)

Where R is the room constant (m²). For 80 bathers generating individual sound power equivalent to SWL = 72 dB (typical for swim activity):

• Total SWL = 72 + 10 log(80) = 91 dB
• Room constant R = 27 × 0.978 / 0.022 ≈ 1,201 m² (using A = 27 sabins from above, corrected)

SPL_reverberant = 91 + 10 log(4/27.7) = 91 − 8.4 ≈ 82.6 dB(A)

Add the direct field contribution at typical distances (1–3 m from sources): the total builds to 88–95 dB(A) in steady state with a busy pool.

A competition event adds the start signal — typically a 110 dB pistol or electronic signal — which excites the room and produces a reverberant tail at 90+ dB(A) persisting for 5+ seconds. The announcer then attempts to make a verbal announcement over this residual energy. Intelligibility: approximately zero.

Material Selection: The Core Constraint

Natatorium acoustic treatment is not an absorption problem. It is a materials science problem.

The environment imposes constraints that eliminate the vast majority of conventional acoustic materials:

Eliminated immediately:

• Standard mineral wool (Rockwool, Knauf) — face coatings fail under prolonged humidity; cores saturate and collapse
• Paper-faced gypsum board — delaminates in weeks under pool humidity
• Open-cell polyurethane foam — absorbs chlorinated water, degrades, becomes heavy and detaches
• Standard fiberglass boards without sealed facing — same failure mode as mineral wool
• Fabric-wrapped panels with standard fabric — most acoustic fabrics are not chloramine-resistant; the fabric holds chlorine and becomes a health hazard

1. Perforated Metal + Mineral Wool Composite (the standard approach)

A proprietary natatorium ceiling system consists of:

• Perforated aluminium or powder-coated steel facing panels (perforation ratio 15–25%)
• A factory-installed moisture-barrier membrane between panel and mineral wool infill
• Dense mineral wool core (80–120 kg/m³) in a sealed compartment

Performance: NRC 0.70–0.85 depending on perforation ratio and wool depth.

Cost: $80–$180/m² installed — 3–5× standard acoustic ceiling cost.

Maintenance: Annual inspection of membrane and edge seals. Panel replacement when seal fails.

2. Polyester Fibre (PET) Panels

Recycled polyester fibre panels are inherently moisture-resistant, dimensionally stable under humidity, and chemically inert to chloramines. They are not suitable for direct water contact but perform well in ceiling applications above the splash zone (> 3 m height).

Performance: NRC 0.60–0.80 at 50mm (less than mineral wool at equivalent thickness due to lower density and fibre stiffness).

Cost: $40–$80/m² installed.

Maintenance: Low — polyester does not degrade in humid environments.

Limitation: Lower density than mineral wool limits low-frequency performance. For equivalent absorption at 125 Hz, a PET panel needs approximately 1.4× the thickness of mineral wool.

3. Microporous Aluminium Panels

Ultra-fine-perforation aluminium facing over a fully sealed mineral wool cavity. The perforation is fine enough (hole diameter 0.3–0.8mm) to prevent moisture penetration under positive pressure difference — the pool environment creates slightly positive humidity at ceiling level from evaporation, but not the sustained pressure differential that would force water through microporous perforations.

Performance: NRC 0.75–0.90 — better than standard perforated because the full face is acoustically active.

Cost: $120–$250/m² — premium natatorium specification.

Maintenance: Minimal; the seal is inherent in the perforation geometry rather than a secondary membrane.

4. Suspended Fibreglass Baffles (with sealed facing)

Horizontal baffles hung from the ceiling structure increase acoustic surface area without requiring ceiling coverage. A suspended baffle system at 30–40% coverage provides double-sided absorption — roughly double the acoustic effectiveness per m² of material compared with ceiling panels.

Material: E-glass fibre core, 100mm, 48 kg/m³, with factory-applied resin coating or sealed fabric facing rated for humid environments.

Performance per baffle face (NRC 0.85 at 500 Hz): A baffle 0.6 m × 1.2 m provides 1.44 m² (double-sided) of absorption = 1.44 × 0.85 = 1.22 sabins at 500 Hz per baffle unit.

Cost: $200–$400 per baffle installed, depending on mounting height and grid complexity.

Maintenance: Higher — baffles hang in the room and accumulate dust and chloramine deposits. Annual cleaning required.

Calculating the Treatment Needed

Back to the 25 m × 12.5 m pool. Target RT60 ≤ 2.0 s at 500 Hz (recreational pool guidance):

Required absorption: A_target = 0.161 × V / RT60_target = 0.161 × 2,150 / 2.0 ≈ 173 sabins at 500 Hz

Current absorption: ~27 sabins. Deficit: 146 sabins at 500 Hz.

Option 1: Full ceiling coverage with perforated metal/mineral wool panels (NRC 0.80) 312 m² × 0.80 = 250 sabins. Total A = 277 sabins. RT60 = 0.161 × 2,150 / 277 ≈ 1.25 s ✓ (exceeds target; under-treatment not an issue)

Option 2: Partial ceiling (60%) + suspended baffles (30% coverage)

• Ceiling panels, 187 m² × 0.80 = 150 sabins
• Baffles, 200 units × 1.22 sabins/unit = 244 sabins
• Total A = 417 sabins → RT60 ≈ 0.83 s (excellent)

• Ceiling, 312 m² × 0.65 = 203 sabins
• Baffles, 130 units × 1.22 = 159 sabins
• Total A = 362 sabins → RT60 ≈ 0.96 s (excellent)

For competition pools where low-frequency RT60 is critical (start signals, PA intelligibility), wall-mounted diaphragmatic panel absorbers are specified alongside ceiling treatment. A diaphragmatic system tuned to 80–150 Hz, installed on upper wall surfaces, can reduce the 125 Hz RT60 by 30–50% — from 4.0 s to 2.5 s in a large competition natatorium.

The PA System: Safety Before Intelligibility

Before tackling the acoustic treatment specification, address the PA system. This is a safety item, not a performance item.

IEC 60268-16 STI measurements in a pool with RT60 of 5 s will yield STI values of approximately 0.30–0.40 even with a well-designed loudspeaker system. "Poor" intelligibility. The question for the designer is not "can we achieve STI ≥ 0.75?" — you cannot, in an untreated pool — but "can we achieve sufficient intelligibility for emergency announcements?"

ISO 7240-19 (fire alarm and voice evacuation) requires that emergency voice announcements achieve minimum STI of 0.45 or CIS (Common Intelligibility Scale) ≥ 0.70. This is achievable even in reverberant spaces with distributed loudspeakers designed for close-field delivery to listening zones.

For a pool natatorium, the PA specification should include:

• Directional array loudspeakers at pool side level aimed at swimmer lanes (direct path to deck)
• Under-gallery distribution speakers for concourse zones
• Digital signal processing with delay alignment between speaker zones
• Minimum SPL 15 dB above background noise at any occupied point in the pool
• Emergency pre-recorded announcement in minimum 3 languages (where relevant for international competition)

New-Build Specification: The Integrated Approach

For a new natatorium, the acoustic specification belongs in the structural design brief, not the interior finishes schedule. Key decisions made at structural stage:

Ceiling height and geometry: A flat ceiling at minimum practical height (say, 5 m over a recreational pool) is acoustically better than a lofted ceiling — lower volume means less reverberant energy storage. However, structural aerodynamics for competition require 3+ m clearance above springboard platforms. Competition natatoriums have non-negotiable minimum ceiling heights that increase the acoustic challenge.

Structural wall construction: Hollow block walls have slightly higher absorption than solid concrete at low frequencies due to panel resonance. If hollow block is architecturally acceptable, specify it over dense concrete block at pool wall level. The difference is modest (α₁₂₅ of 0.08 vs 0.03) but worth capturing in a space where every sabin matters.

Window placement: Pool windows admit daylight but are acoustically transparent (α₅₀₀ ≈ 0.03). Limit window area to what is required for natural light; every m² of glass is a missed opportunity for ceiling treatment.

Service infrastructure: Plan for mechanical suspension from the structure for baffle systems and ceiling panels at ceiling-plan stage. Adding suspension points after the structural slab is poured costs 3–5× more than casting in threaded inserts during construction.

Cost Reality

A 25 m recreational pool natatorium, full treatment:

Against a pool construction cost of $1.5–$3 M, acoustic treatment represents 4–7% of the build cost. It is routinely omitted or value-engineered out in the final budget review — at which point you have a $2 M facility that is too loud to be safe, requires hearing protection for competition events, and generates complaints from every patron who tries to have a conversation at poolside.

The calculation is straightforward. Model the room using the RT60 calculator before the ceiling specification is finalised. The calculation takes five minutes. Discovering the RT60 problem at handover — after the ceiling is fixed and the budget is spent — is a significantly more expensive conversation.

Pools are the hardest room in architecture to fix because you cannot fix them after the fact without major construction disruption. The time to specify acoustic treatment is before the concrete is poured. After that, every option is expensive, disruptive, and yields a result worse than designing it correctly the first time.