Heat Pump Noise & Neighbours — London Planning & Reality

Modern air-source heat pumps from the major UK-installed brands sit in a 45–55 dBA range at 1m from the unit at rated output. Daikin Altherma 3 H MT 12kW: 49 dBA. Mitsubishi Ecodan PUZ 11kW: 47 dBA. Vaillant aroTHERM Plus 10kW: 50 dBA. Samsung EHS Mono 12kW: 51 dBA.

These figures are at full rated output — the worst-case noise level. At part load (most of the heating season), noise is typically 5–8 dB lower. The same units in actual normal operation often measure 41–48 dBA at 1m.

Context: 50 dBA is roughly the sound level of a quiet office or a refrigerator at 1m. 60 dBA is normal conversation. 70 dBA is heavy traffic at the kerbside. ASHP noise sits well below the 65 dBA daytime nuisance threshold that London boroughs typically apply.

Old-generation ASHPs (pre-2020) ran 55–65 dBA and were responsible for the 'heat pumps are noisy' perception that persists. The 2026 generation is fundamentally quieter — the brushless DC motors, soft-start inverters and acoustic shielding of modern units deliver 5–10 dB lower noise than units from five years ago.

Air-source heat pumps in England are permitted development (no planning permission required) where the install meets specific MCS-020 conditions. The key practical condition is the 1m boundary rule — the heat pump must be at least 1 metre from the property boundary.

The 1m rule was relaxed in the 2023 Permitted Development changes — the previous version required the unit to be at least 1m from any boundary AND the unit volume to be below 0.6 cubic metres. The 2023 update removed the volume cap and softened the boundary rule for some configurations.

Properties in conservation areas, World Heritage Sites, and on or near listed buildings retain stricter conditions. ASHPs visible from the public highway in conservation areas typically need planning consent. London has 1,000+ conservation areas covering large swathes of inner London, so this exception applies more often than the headline 'permitted development' suggests.

The 1m boundary rule is the most common source of unworkable London installs. A typical Victorian terrace has 0.6–0.9m side returns; a unit placed in the side return is usually within 1m of the boundary fence. Workarounds include rear-garden placement, front-garden placement (where the property has a front garden), or seeking planning consent for a boundary-adjacent install.

MCS-020 is the MCS standard that defines the acoustic calculation an MCS installer must complete for every ASHP install. The calculation models the noise level at the nearest neighbour's nearest habitable window and confirms it sits below 42 dB(A) — a threshold derived from BS 4142 minus the 5 dB tolerance applied to permitted development.

The calculation factors include the heat pump sound power level (a manufacturer-supplied figure, typically 58–65 dB(A) for residential ASHPs), the distance from the unit to the assessment point, the reflectivity of nearby surfaces (walls, fences, paved surfaces), and the directionality of the noise emission.

Where the calculation comes out at or below 42 dB(A), the install qualifies for permitted development. Where it comes out above, planning consent is required and the council acoustic officer reviews the calculation as part of the application.

MCS-020 calculations are model-based and conservative. Real-world measurements typically come out 2–5 dB lower than the modelled figure. This margin is what makes permitted development work for the majority of London terrace installs — the calculation passes at 41 dB(A) and the actual measured noise is 36–39 dB(A).

The calculation must be retained in the MCS installer's file for 10 years and is part of the BUS grant audit pack. A heat pump installed without an MCS-020 calculation cannot claim the £7,500 BUS grant.

Victorian terrace, rear garden install: unit mounted on the rear elevation of the property, 2–4m from the rear boundary. MCS-020 passes comfortably. Most common London install pattern.

Victorian terrace, side return install: unit mounted on the side return wall, 0.4–0.8m from the boundary. MCS-020 frequently fails the 1m boundary test on permitted development; planning consent required. Where granted, the install is straightforward but the planning delay adds 8–14 weeks.

Edwardian semi-detached, side passage install: unit mounted in the side passage between the two semi-detached properties, 1.0–1.4m from boundary. MCS-020 typically passes; permitted development applies.

1930s semi, rear garden install: rear garden depth typically 9–18m, unit mounted on rear wall 2–5m from boundary. Easiest of the common London property types.

Flat in a purpose-built block: rare in 2026 — most blocks do not permit individual ASHP installs on the building exterior. Where installs do happen, they require freeholder consent, often planning consent, and a shared communal system makes more practical sense.

Listed building or conservation area terrace: ASHP install usually requires planning consent regardless of distance from boundary. Visible installation from the public highway is the typical sticking point. Workarounds include rear-garden enclosures, integrated cabinet designs, or 'monoblock' units sited away from heritage-significant facades.

Wrong-side fence boundary placement. A unit placed 0.8m from a fence shared with a neighbour's bedroom window is the textbook complaint scenario — even with MCS-020 passing the modelled calculation, a neighbour subjectively hearing a previously-quiet boundary suddenly emit a hum is unwelcome.

Reflective wall placement. A unit mounted against a brick wall opposite a similar brick wall or a glass conservatory creates reverberation that increases perceived noise without changing the modelled figure. Where wall-to-wall distance is under 4m, expect higher real-world perception.

Vibration coupling. A unit bolted directly to a lightweight stud wall transmits vibration into the building structure, producing low-frequency hum inside the property. Always use anti-vibration mounting feet on a solid base (paving slab on a compacted bed, or a wall mount on a structural masonry wall with rubber isolators).

Defrost cycles in mid-winter. ASHP defrost cycles run 4–8 minutes at higher noise levels (typically +3–5 dB) every 30–90 minutes when outdoor temperature is between -3°C and +5°C with humidity above 80%. Several London neighbour complaints in 2024–2025 originated from this seasonal effect rather than the steady-state noise.

Older neighbour units. A 2019-vintage ASHP nearby running at 60+ dBA can be the actual noise source; a new 47 dBA install is sometimes blamed for noise that predated it. Targeted measurement with a calibrated sound level meter identifies the true source.

Acoustic enclosures. Purpose-built ASHP enclosures (e.g. Daikin Aktion, Kingspan Logix Sound Block, custom timber-and-acoustic-board builds) typically reduce perceived noise by 5–10 dB. They must allow adequate airflow — undersized enclosures destroy efficiency and shorten compressor life.

Anti-vibration mounting. The single highest-impact, lowest-cost mitigation. A unit on a flat-paved bed with anti-vibration rubber feet, connected via flexible refrigerant lines, prevents structure-borne noise that is the source of most 'hum inside the house' complaints.

Siting away from boundary lines and bedroom windows. Where the property layout allows, placing the unit at the rear-garden centre (rather than against the boundary fence) reduces the calculated and perceived noise at all neighbours simultaneously.

Right-sizing the unit. Oversized heat pumps cycle frequently and noisily. A correctly sized heat pump runs longer at part load (quieter) and reduces total annoying transitions. MCS-certified design with heat loss calculation prevents oversizing.

Pre-install neighbour conversations. The single highest-impact behavioural mitigation. A homeowner who briefs neighbours before install, shares the MCS-020 calculation, and offers to walk the boundary to measure at a quiet evening time generally avoids the formal complaint route entirely. Complaints almost always escalate from 'we weren't told' rather than from 'we measured and it's too loud'.

Where a complaint is raised post-install, the council environmental health team is the formal route. They visit, measure with a calibrated meter, and compare against BS 4142 thresholds and the MCS-020 calculation. Properly installed modern ASHPs almost never fail these measurements; the complaint resolution is usually evidential rather than remedial. The exception is units installed with poor anti-vibration mounting — these can fail measurement on low-frequency structure-borne noise even when the overall A-weighted level is compliant.