HVAC Noise Control FAQ

Q: How should HVAC noise be verified during commissioning?

HVAC noise commissioning involves measuring the actual background noise levels in occupied rooms and comparing against the design criteria. Procedure per CIBSE Guide B4 and BS 8233:2014: measure Leq per octave band (63–8000 Hz) at representative positions (1.2 m height, 1+ m from walls), with all HVAC systems operating at design condition. Room should be unoccupied but furnished. Measure for a minimum of 5 minutes per position. Compare against the NR/NC target using the octave-band assessment method. If any octave band exceeds the target curve, identify the source: (1) Switch off HVAC systems individually to isolate the contributing path. (2) Measure at the diffuser face to assess supply noise. (3) Check fan speed against design specification. (4) Verify damper positions — partially closed dampers generate significantly more noise. (5) Inspect ductwork for construction defects (disconnected lining, missing attenuators, unsealed joints). Document results with photos and measurement data. AcousPlan's measurement import compares commissioning data against design predictions.

Q: What is regenerated noise in ductwork?

Regenerated noise (also called self-generated or flow-generated noise) is sound created by turbulent airflow at ductwork fittings — bends, branches, dampers, expansions, and contractions. It is independent of and additional to the noise transmitted from the fan. Per CIBSE Guide B4 §4, regenerated noise follows the Sixth power law: sound power is proportional to the sixth power of air velocity (Lw ∝ 60 log v). This means every doubling of velocity increases noise by approximately 18 dB. Key sources: 90° bends add 5–15 dB depending on type (mitred worst, radiused best), dampers add 10–25 dB at 50% closure, branches add 5–10 dB at the takeoff, and sudden expansions/contractions add 5–15 dB. Critically, regenerated noise after a silencer cancels its attenuation — if a bend immediately follows a silencer, the effective insertion loss is severely reduced. Allow 3–5 duct diameters of straight duct after silencers and before bends. Design duct velocities: ≤ 5 m/s in occupied zones, ≤ 8 m/s in risers.

Q: How do you control low-frequency HVAC rumble?

Low-frequency HVAC rumble (31.5–125 Hz) is the most difficult noise to control because: (1) conventional absorptive duct silencers are ineffective below 125 Hz, (2) standard building constructions provide poor insulation at low frequencies, and (3) room absorption is minimal at these frequencies. Sources include: fan blade-pass frequency (BPF = RPM × number of blades / 60), duct resonance, and structure-borne vibration from rotating equipment. Control strategies: vibration isolation — mount fans, pumps, and compressors on inertia bases with spring isolators (natural frequency ≤ 5 Hz for critical spaces). Use flexible duct connections (minimum 200 mm length) between fan discharge and rigid ductwork. Install reactive silencers tuned to the BPF. Reduce fan speed with VSDs — even a 10% reduction cuts blade-pass noise by 5–8 dB. For structure-borne paths: isolate ductwork from the building structure using resilient hangers. Consider active noise control for persistent tonal problems. Diagnose using narrow-band frequency analysis to identify the specific tonal component.

Q: How do variable speed drives help control HVAC noise?

Variable speed drives (VSDs, also called variable frequency drives/VFDs or inverters) are one of the most effective HVAC noise control measures. Fan noise follows the Fifth power law with respect to speed: a 50% speed reduction decreases sound power by approximately 15 dB (Lw ∝ 50 log[n1/n2]). Since most HVAC systems operate at part load for 80–90% of the year, VSDs allow fans to run at the minimum speed needed to meet the current demand — dramatically reducing noise during typical operation. Additional benefits: regenerated noise at ductwork fittings also reduces with lower velocity, and duct breakout noise decreases proportionally. Caveats: VSDs can generate electrical noise (switching harmonics) that transmits as audible tonal noise through motor bearings — specify EMC-filtered VSDs and maintain a minimum speed (typically > 20%) to avoid unstable motor operation. Always specify HVAC noise performance at the design operating speed, not the maximum speed. AcousPlan models HVAC noise at user-specified operating conditions.

Technical guidance on controlling heating, ventilation, and air conditioning noise in buildings. Covers ductwork noise, fan selection, attenuators, terminal units, commissioning, and the relationship between air velocity and noise.

Quick Navigation

1. What noise criteria apply to HVAC systems in buildings?
2. How does ductwork generate and transmit noise?
3. How do you select a quiet fan for an HVAC system?
4. What types of duct attenuators (silencers) are available?
5. How do HVAC terminal units affect room noise levels?
6. How should HVAC noise be verified during commissioning?
7. What is regenerated noise in ductwork?
8. How do you control low-frequency HVAC rumble?
9. How do variable speed drives help control HVAC noise?
10. How do you measure HVAC noise contribution in an occupied space?

What noise criteria apply to HVAC systems in buildings?

HVAC noise criteria are specified in terms of NR (Europe), NC (North America), or RC Mark II (ASHRAE). Targets per BS 8233:2014 Table 5 and CIBSE Guide A: recording studio NR 15–20, concert hall NR 15–20, private office NR 30–35, open plan office NR 35–40, restaurant NR 35–40, retail NR 40–45. ASHRAE equivalents: executive office NC 25–30, conference room NC 25–30, open office NC 40, lobby NC 40. WELL v2 Feature S02 specifies: enclosed offices ≤ 35 dBA, open plan ≤ 40 dBA. These criteria represent the total background noise from all HVAC components at the listener position — supply air, return air, radiated duct noise, and terminal devices combined. Design the HVAC system to achieve at least 5 NR/NC below the target to allow margin for construction variations and other noise sources. AcousPlan's noise criteria assessment compares HVAC noise predictions against room-specific targets.

Try Calculator BS 8233:2014

How does ductwork generate and transmit noise?

Ductwork transmits noise through two mechanisms. Airborne transmission: sound generated by the fan or upstream components propagates through the air within the duct, emerging through supply and return diffusers into the occupied space. This path is controlled by duct attenuation (natural and added). Breakout transmission: sound energy within the duct causes the duct walls to vibrate and radiate noise into adjacent spaces — particularly problematic for large rectangular ducts in noise-sensitive areas. Per CIBSE Guide B4, breakout is significant when duct dimensions exceed 600 mm. Ductwork also generates its own noise at flow constrictions: bends create turbulence (regenerated noise) that adds 5–15 dB depending on bend type, dampers generate noise proportional to pressure drop, and branches/takeoffs create turbulent mixing. Air velocity is the primary driver — every halving of velocity reduces flow-generated noise by approximately 15–18 dB (per Sixth power law: noise ∝ v⁶). AcousPlan models duct noise paths from source to receiver.

Try Calculator

How do you select a quiet fan for an HVAC system?

Fan selection for acoustic performance requires balancing airflow, pressure, efficiency, and noise. Key principles: (1) Select the fan to operate at its peak efficiency point (BEP) — fans operating off-BEP generate 5–15 dB more noise. (2) Choose the largest practical fan size — a larger, slower fan produces less noise than a smaller, faster fan for the same duty. (3) Fan type matters: plug fans (plenum fans) are typically 5–10 dB quieter than housed centrifugal fans at the same duty because they avoid scroll-generated noise. Backward-curved impellers are quieter than forward-curved at the same pressure. (4) Specify variable speed drives (VSDs) — fans rarely need full speed at all times, and noise reduces dramatically with speed (approximately 15 dB per halving of speed). (5) Request octave-band sound power data (tested per ISO 13347 or AMCA 300) from the manufacturer — do not rely on single-number dBA ratings. Compare fan options at the critical octave bands where the room noise criterion is most sensitive.

Try Calculator

What types of duct attenuators (silencers) are available?

Duct attenuators fall into three categories. (1) Dissipative (absorptive) silencers — lined with mineral wool or glass fibre behind a perforated facing. Effective at mid-to-high frequencies (500–8000 Hz). Attenuation: 10–40 dB depending on length (0.6–2.4 m), splitter configuration, and airway width. The most common type. (2) Reactive silencers — use chambers and tuning to attenuate specific low frequencies (31.5–250 Hz) without absorptive lining. Used where fibrous materials are prohibited (clean rooms, food processing) or for targeted low-frequency control. (3) Active noise control — uses microphones, speakers, and DSP to generate anti-phase sound, cancelling duct noise by 10–20 dB at low frequencies (50–500 Hz). Effective for persistent tonal noise (fan blade pass frequency). Specify silencers early in the design — they add pressure drop (typically 50–150 Pa) that must be included in the fan sizing calculation. Per CIBSE Guide B4, avoid placing silencers immediately after bends or dampers to prevent regenerated noise cancelling the attenuation.

Try Calculator

How do HVAC terminal units affect room noise levels?

Terminal units (VAV boxes, fan coil units, chilled beams) are often the closest HVAC noise source to the occupant and can dominate room noise levels. VAV boxes with reheat: noise is generated by the throttling valve, fan (if fan-assisted), and radiated casing noise. Specify units rated for NC 25–30 at the design airflow and pressure. Fan coil units: the internal fan generates broadband noise at 35–50 dBA depending on speed setting. Specify EC (electronically commutated) motors for quieter operation at reduced speeds. Mount on resilient isolators with flexible pipe connections. Active chilled beams: generally the quietest option (25–35 dBA) as they have no moving parts — noise comes from induced air mixing. However, supply air velocity to the beam must be controlled (< 3 m/s nozzle velocity). Passive chilled beams are silent. In all cases, obtain manufacturer noise data per ISO 3741 at the actual operating condition, not just the minimum-speed catalogue value.

Try Calculator

How should HVAC noise be verified during commissioning?

HVAC noise commissioning involves measuring the actual background noise levels in occupied rooms and comparing against the design criteria. Procedure per CIBSE Guide B4 and BS 8233:2014: measure Leq per octave band (63–8000 Hz) at representative positions (1.2 m height, 1+ m from walls), with all HVAC systems operating at design condition. Room should be unoccupied but furnished. Measure for a minimum of 5 minutes per position. Compare against the NR/NC target using the octave-band assessment method. If any octave band exceeds the target curve, identify the source: (1) Switch off HVAC systems individually to isolate the contributing path. (2) Measure at the diffuser face to assess supply noise. (3) Check fan speed against design specification. (4) Verify damper positions — partially closed dampers generate significantly more noise. (5) Inspect ductwork for construction defects (disconnected lining, missing attenuators, unsealed joints). Document results with photos and measurement data. AcousPlan's measurement import compares commissioning data against design predictions.

Try Calculator BS 8233:2014

What is regenerated noise in ductwork?

Regenerated noise (also called self-generated or flow-generated noise) is sound created by turbulent airflow at ductwork fittings — bends, branches, dampers, expansions, and contractions. It is independent of and additional to the noise transmitted from the fan. Per CIBSE Guide B4 §4, regenerated noise follows the Sixth power law: sound power is proportional to the sixth power of air velocity (Lw ∝ 60 log v). This means every doubling of velocity increases noise by approximately 18 dB. Key sources: 90° bends add 5–15 dB depending on type (mitred worst, radiused best), dampers add 10–25 dB at 50% closure, branches add 5–10 dB at the takeoff, and sudden expansions/contractions add 5–15 dB. Critically, regenerated noise after a silencer cancels its attenuation — if a bend immediately follows a silencer, the effective insertion loss is severely reduced. Allow 3–5 duct diameters of straight duct after silencers and before bends. Design duct velocities: ≤ 5 m/s in occupied zones, ≤ 8 m/s in risers.

Try Calculator

How do you control low-frequency HVAC rumble?

Low-frequency HVAC rumble (31.5–125 Hz) is the most difficult noise to control because: (1) conventional absorptive duct silencers are ineffective below 125 Hz, (2) standard building constructions provide poor insulation at low frequencies, and (3) room absorption is minimal at these frequencies. Sources include: fan blade-pass frequency (BPF = RPM × number of blades / 60), duct resonance, and structure-borne vibration from rotating equipment. Control strategies: vibration isolation — mount fans, pumps, and compressors on inertia bases with spring isolators (natural frequency ≤ 5 Hz for critical spaces). Use flexible duct connections (minimum 200 mm length) between fan discharge and rigid ductwork. Install reactive silencers tuned to the BPF. Reduce fan speed with VSDs — even a 10% reduction cuts blade-pass noise by 5–8 dB. For structure-borne paths: isolate ductwork from the building structure using resilient hangers. Consider active noise control for persistent tonal problems. Diagnose using narrow-band frequency analysis to identify the specific tonal component.

Try Calculator

How do variable speed drives help control HVAC noise?

Variable speed drives (VSDs, also called variable frequency drives/VFDs or inverters) are one of the most effective HVAC noise control measures. Fan noise follows the Fifth power law with respect to speed: a 50% speed reduction decreases sound power by approximately 15 dB (Lw ∝ 50 log[n1/n2]). Since most HVAC systems operate at part load for 80–90% of the year, VSDs allow fans to run at the minimum speed needed to meet the current demand — dramatically reducing noise during typical operation. Additional benefits: regenerated noise at ductwork fittings also reduces with lower velocity, and duct breakout noise decreases proportionally. Caveats: VSDs can generate electrical noise (switching harmonics) that transmits as audible tonal noise through motor bearings — specify EMC-filtered VSDs and maintain a minimum speed (typically > 20%) to avoid unstable motor operation. Always specify HVAC noise performance at the design operating speed, not the maximum speed. AcousPlan models HVAC noise at user-specified operating conditions.

Try Calculator

How do you measure HVAC noise contribution in an occupied space?

Isolating the HVAC noise contribution from other sources requires systematic measurement. Procedure: (1) Measure total background noise with all systems operating at design condition — Leq per octave band, 5-minute average, room unoccupied. (2) Switch off the HVAC system and repeat — this captures the "residual" noise (external, electrical, other equipment). (3) Calculate the HVAC contribution by logarithmic subtraction: LHVAC = 10 log(10^(Ltotal/10) − 10^(Lresidual/10)). This is only valid when the difference between total and residual is ≥ 3 dB — if less, the HVAC is not the dominant source. (4) To isolate individual HVAC paths, sequentially switch off components: supply fan only, return fan only, FCUs, etc. Measurement conditions per BS 8233:2014 Annex B: windows closed, doors closed, room furnished, minimum 3 positions per room. Use a Class 1 sound level metre with octave-band analysis. For commissioning, compare against the design NR/NC target. AcousPlan's measurement import tool handles multi-condition analysis.

Try Calculator BS 8233:2014