Acoustic Sound Isolation in Transportation Hubs

1) Project overview: what, where, who, and why

In late 2024, the City Transit Authority commissioned a retrofit of a multi-use concourse at Riverton Central Station, a commuter rail and intercity bus hub handling roughly 62,000 пассажengers/day. The scope was not about making the station quiet—an impossible goal in a space with train brakes, PA announcements, rolling luggage, and HVAC. The goal was tighter: create acoustically isolated “quiet zones” within the hub so operational teams could communicate reliably, and passengers could use services without background noise bleeding into critical spaces.

The project centered on two areas:

• A new Operations Communications Suite (OCS): a 95 m² control room plus a 28 m² briefing room, adjacent to the main concourse and 18 m from Platform 4.
• A Customer Service & Accessibility Center (CSAC): four interview rooms (each ~12 m²) and one telehealth booth, directly off the ticketing hall.

Sonus Gear Flow was brought in as the audio/acoustics documentation lead and system integration advisor, coordinating with the architect (Northline Studio), the general contractor (Barton BuildCo), and MEP (Keller Systems). The stakeholder group included transit operations, station security, an accessibility advocacy committee, and the city’s building acoustics inspector.

The “why” was concrete. The station had recorded repeated incident debriefs where misheard instructions contributed to delayed response times. Customer service staff also reported frequent confidentiality breaches: voices from interview rooms were intelligible in the hallway during peak hours. The mandate was to improve speech privacy, reduce listener fatigue, and ensure the OCS could support clear radio/VoIP dispatching even when trains arrived at adjacent platforms.

2) Challenges and requirements at the outset

The starting conditions were hostile to isolation. The concourse is a steel-and-glass volume with a 14 m ceiling, hard tile floors, and long reflective sightlines. Measured during the first site survey (weekday 07:30–09:30), the A-weighted ambient at the intended OCS wall line averaged 72–78 dBA Leq, with peaks to 92 dBA when trains braked and couplers clanked at Platform 4. Low-frequency energy was prominent: 63 Hz and 125 Hz bands spiked repeatedly, correlating with diesel bus acceleration and train traction systems.

Primary requirements were written into the owner’s performance brief:

• OCS background level: target NC-30 (or better) during typical peak operations.
• Speech privacy for CSAC interview rooms: target ASTM E2638 “Confidential” equivalent outcomes (practically interpreted as high speech privacy; no intelligible leakage at 1 m outside the door under station operating conditions).
• Isolation performance: achieve approximately STC 55 partitions for OCS boundaries and STC 50 for CSAC, with attention to low-frequency performance beyond what STC alone predicts.
• Schedule: 10-week construction window, with only two overnight shutdowns allowed for loud work and tie-ins.
• Safety and code: fire-rated assemblies, smoke control integrity, and no obstruction to emergency egress routes.

Non-obvious constraint: the station’s public address (PA) system had to remain intelligible in the concourse. Over-isolating the new suites without managing door location and vestibule geometry risked creating “dead corners” where announcements were unintentionally masked, so circulation paths and vestibules had to be designed to preserve PA coverage patterns.

3) Approach and methodology chosen

We used a layered approach: measure → model → specify → verify.

First, we conducted baseline measurements: octave-band ambient noise profiles, preliminary RT60 estimates in the concourse, and structure-borne checks at the slab edge near the proposed OCS. Tools included a NTi XL2 with M2230 mic for spot checks and logging, and a Brüel & Kjær 2250 (rented) for higher confidence spectral logging and calibration traceability. We also used a simple accelerometer setup on the slab near a column line to confirm that vibration transmission would be an issue for low-frequency rumble if we tied walls directly into the existing structure without breaks.

Next, we built an isolation concept around three principles:

• Decouple wherever possible (especially ceilings and wall-to-structure connections).
• Add mass in a controlled, buildable way (multiple gypsum layers with constrained-layer damping where justified).
• Seal and control flanking (doors, glazing, ductwork, cable penetrations, and slab/ceiling junctions).

For the OCS and briefing room, we treated the suite as a “room-within-a-room” without going to a full floated slab (schedule and cost). For CSAC, the approach was “high-STC rooms” with aggressive door and ceiling details, but not full decoupled shells.

4) Step-by-step execution narrative

Week 1–2: Survey, coordination, and mockups. The first coordination meeting surfaced immediate conflicts: MEP routes ran exactly where we needed continuous ceiling plenum depth for isolation. We set a rule early—no duct penetrations through the OCS envelope except at two controlled points, and all cable penetrations would use listed acoustic putty systems. A 2.4 m wide wall mockup with one door and one cable penetration was built offsite to validate buildability and sealing details. This mockup saved time later; the contractor discovered that the initially proposed back-to-back electrical boxes would not fit with the required offset and putty pads without violating the stud bay insulation.

Week 3: Demolition and structural prep. The GC removed a portion of retail storage to free the footprint. We marked all existing slab cracks and perimeter joints. Where walls would land, we specified a continuous neoprene isolation gasket under base tracks and a non-hardening acoustic sealant at both sides of the track. The goal was to avoid rigid short-circuits into the slab and prevent air leaks at the most common failure point: the base line hidden by baseboards later.

Week 4–5: Framing and first-layer mass. For OCS perimeter walls, the assembly used double-stud framing (two independent 92 mm stud rows with a 25 mm air gap), filled with 60 kg/m³ mineral wool. On each side: two layers of 16 mm Type X gypsum, with constrained-layer damping compound applied between layers on the OCS-facing side. The damping was not applied everywhere—only where we expected the most benefit: the two walls facing the concourse and platform side. This was a cost control decision; applying damping to all surfaces would have exceeded budget without proportional benefit.

The ceiling was the critical flanking path. The existing concourse ceiling structure was shared with other spaces. We installed an independent ceiling grid using isolation hangers rated for the load, supporting two layers of 16 mm gypsum with sealed perimeters. We maintained a consistent 200–250 mm plenum to avoid squeezing ducts into direct contact with the new ceiling.

Week 6: Doors, glazing, and vestibules. The OCS entrance used a short vestibule with two doors to reduce direct leakage when staff entered from the concourse. Door sets were STC-rated acoustic doors (STC 50 nominal) with continuous perimeter seals and automatic door bottoms. The vestibule geometry avoided a direct line-of-sight from concourse to OCS interior. For CSAC interview rooms, each door was specified as an STC 45 acoustic wood door with upgraded seals. A lesson learned from prior projects guided this: a good wall is wasted with a mediocre door. Door hardware and closers were selected to ensure consistent latching force, because even slight mis-latching defeats perimeter seals.

One interview room required a glazing panel for safety observation. Instead of a standard storefront system, we used a laminated acoustic glazing build (asymmetric thickness) in an independent frame isolated from the wall studs, with a sealed perimeter and no shared mullions with adjacent glazing. The detail was reviewed twice with the glazing subcontractor; the first shop drawing inadvertently tied the frame to both stud rows, which would have short-circuited the double-stud benefit.

Week 7–8: HVAC silencing and penetrations. HVAC was the other major leakage path. The OCS required cooling and ventilation at all hours, but duct-borne noise could not be allowed to become the new “background.” We specified:

• Low-velocity duct design (target under ~4 m/s in main runs feeding the suite).
• Duct silencers on both supply and return trunks serving OCS.
• Flexible connectors at equipment connections to reduce vibration transfer.
• Lined duct sections approaching diffusers, with care to avoid fiber shedding into occupied air.

All duct penetrations through the envelope used oversized sleeves with backer rod and acoustic sealant, maintaining fire-rating requirements. Cable penetrations were grouped into two locations rather than scattered, each sealed with a listed putty pad and a removable, gasketed cover plate for service access. We documented every penetration with photos and location notes; this became invaluable during commissioning when a new cable needed to be added without “field drilling” through the wrong area.

Week 9: Interior acoustic treatment and commissioning prep. Isolation alone doesn’t produce intelligibility; the OCS needed a controlled acoustic field. We added ceiling absorptive panels in the OCS and briefing room (NRC 0.80 class), plus wall panels on the first reflection zones near operator positions. The aim was to keep the room’s reverberation modest so headsets and nearfield monitoring wouldn’t fight the room.

Audio and communications equipment included a rack-mounted DSP (Q-SYS Core class) for routing and level management between radio gateways, VoIP endpoints, and recording. Operator positions used broadcast-style dynamic headsets with strong off-axis rejection, chosen specifically because we expected residual low-frequency rumble even after isolation.

Week 10: Verification and handover. We scheduled verification during the noisiest predictable window: a weekday morning peak with known train arrivals. Measurements included indoor NC curves, spot-checking door leakage with pink noise sources, and practical speech privacy checks outside CSAC doors while a calibrated speech track played inside.

5) Technical decisions and trade-offs made

Several decisions were deliberate compromises between ideal isolation and the realities of transportation construction.

No floated slab. A floated floor would have improved low-frequency isolation from structure-borne vibration, but it added height, complicated accessibility transitions, and risked schedule overruns. Instead, we prioritized ceiling decoupling and wall isolation, then managed residual LF with room acoustic treatment and headset choice. This trade-off was acceptable because the OCS primary intelligibility path is headset-based, and the target was NC-30 rather than studio-grade silence.

Selective constrained-layer damping. We applied damping compound only on the two most exposed walls and the ceiling facing the concourse. The cost savings were significant (materials and labor), and performance modeling suggested diminishing returns on interior partitions that didn’t face major noise sources.

Vestibule over “single heroic door.” We chose a two-door vestibule for OCS rather than trying to push a single door rating higher. In practice, vestibules provide more robust real-world performance because they reduce the effect of imperfect sealing and user behavior (doors held open briefly, frequent traffic).

HVAC design as an isolation element. The mechanical team initially proposed a higher-velocity branch with standard diffusers to meet airflow. We pushed back: higher velocity would have created self-noise and reduced the value of the isolation envelope. The compromise was slightly larger duct runs and silencers, plus rebalanced diffusers to keep throw comfortable without increasing turbulence noise.

6) Results and outcomes with specific details

The post-build measurements showed the isolation strategy worked within the defined scope.

• OCS ambient noise: During peak operations, the room measured NC-28 to NC-30, depending on HVAC state. With HVAC in night setback mode, it sat around NC-26.
• Low-frequency behavior: 63 Hz energy was reduced substantially but not eliminated; we observed intermittent LF rises during train braking events. However, the audibility in the room was muted enough that dispatch communications remained unaffected when using the specified headsets.
• Door leakage performance: The vestibule prevented direct transmission; standing 1 m outside the concourse door during interior pink-noise playback, the measured level drop was typically 38–42 dB A-weighted, with better performance above 250 Hz.
• CSAC speech privacy: Hallway checks confirmed that speech from interview rooms was no longer intelligible at normal speaking levels. Staff feedback was immediate: they no longer needed to raise their voices, and customers reported greater comfort discussing sensitive topics.
• Operational outcomes: In the first month, the operations manager reported fewer “repeat” instructions during dispatch. While not a controlled study, the qualitative improvement aligned with the measured reductions in background noise and reverberant buildup in the OCS.

One outcome worth noting: the PA intelligibility in the concourse did not degrade. By keeping vestibule placement and circulation open, and avoiding reflective “pockets” near the entrances, we didn’t create new dead zones or hotspots. The station’s existing PA tuning required only minor level trimming near the new construction.

7) Lessons learned and what could be done differently

Flanking paths hide in procurement details. The biggest near-miss was glazing shop drawings that tied frames to both stud rows. Catching it required a deliberate acoustic review step, not just architectural approval. If we repeated the project, we’d formalize an “acoustic submittal checklist” for doors, glazing, and MEP penetrations, and require sign-off before materials arrive onsite.

Penetration discipline matters more than wall ratings. The highest-rated wall assembly can be undermined by a poorly sealed conduit or a misapplied firestop product that shrinks or cracks. The photo log and the rule of “two controlled cable locations” prevented field improvisation. Next time, we’d go further and install pre-fabricated, gasketed multi-cable transit frames to simplify future adds.

Model STC, but manage low frequency with expectations and tools. Transportation hubs punish low-frequency isolation. Without a floated floor, we knew 63–125 Hz events would remain. The project succeeded because performance targets were written realistically (NC-30, not studio silence) and because the operational workflow used headsets and DSP with good gain structure. If the OCS required open-air monitoring at all times, we would have pushed harder for floor isolation or additional structural decoupling.

Commissioning needs peak-condition testing. Testing at night would have made the numbers look great but would not have proven functional success. Scheduling verification during morning peak took coordination but delivered confidence and avoided post-occupancy surprises.

8) Takeaways applicable to other projects

• Write targets that match the mission. Transportation hubs can support excellent speech intelligibility without chasing unrealistic silence. Define NC targets, privacy criteria, and operational use cases (headset vs. open monitoring) early.
• Design isolation as a system, not a wall. Doors, vestibules, ceilings, ductwork, and penetrations determine real performance. Budget time for submittal review and onsite inspections of seals before finishes hide problems.
• Prioritize ceiling and HVAC in retrofit hubs. In many concourses, ceiling plenum and shared structure are the dominant flanking paths. Independent ceilings with isolation hangers and low-noise mechanical design often deliver more benefit than adding another gypsum layer to walls.
• Control future changes. Stations evolve: new cameras, new radios, new networks. Provide planned, sealable pathways for cables and ducts so “one quick penetration” doesn’t undo a year of careful isolation.
• Verify under real conditions. Peak-time measurement and practical walk-by privacy checks are the difference between theoretical compliance and operational success.

At Riverton Central Station, the final win wasn’t that the hub became quiet—it didn’t. The win was creating small, purpose-built environments where isolation, HVAC noise control, and practical audio engineering choices worked together. That’s the pattern that scales to other transportation hubs: isolate what must be protected, accept what can’t be eliminated, and document every detail that keeps performance intact over time.