# Acoustic Flanking: Why a Good Wall Still Leaks Sound

> Flanking is sound travelling between two spaces by any route other than straight through the wall between them — over the partition through the ceiling void, under it through a raised floor, along a duct, or through a junction. It is why a partition with a high laboratory rating often delivers far less in the finished building.

## Why it matters on a project

This is the reason clients feel misled about acoustics.

A partition is specified with a good rating. It is installed correctly. The
finished room still does not hold a private conversation, and nothing about the
wall is defective. The sound simply went around it.

## The paths

| Path | Where it goes | Fix |
| --- | --- | --- |
| Over the partition | Through the ceiling void | Take partition to slab, or barrier above |
| Under the partition | Through a raised floor void | Close the void at the wall line |
| Through ducts | Along shared ductwork | Acoustic treatment, offsets, attenuators |
| Through penetrations | Sleeves, trays, pipes | Seal every penetration properly |
| Back-to-back sockets | Through the wall's own cavity | Offset them between studs |
| At junctions | Gaps at head, base, abutments | Continuous sealing |
| Through structure | Continuous slabs and elements | Design-stage consideration |

## Why lab numbers mislead

A laboratory test under **ISO 10140** deliberately removes flanking so it can
measure the partition alone. That is correct — it is testing a product.

A building measured under **ISO 16283-1** includes every path. The result is
usually lower, and sometimes much lower.

So a rating on a datasheet describes the best the partition can do in ideal
conditions. What the room delivers depends on the installation. When a client is
told a meeting room will be private "because we used a rated system", that gap
is where the disappointment comes from. Ratings are expressed under
**ISO 717-1**.

## The practical hierarchy

If acoustic privacy genuinely matters in a room, in order of impact:

1. **Take the partition to the slab**, or build an acoustic barrier above the
   ceiling line.
2. **Seal every junction** — head, base, abutments — continuously.
3. **Treat services crossings** — ducts, trays, pipes, sleeves.
4. **Offset back-to-back outlets** so they do not share a cavity.
5. **Then** consider upgrading the partition specification itself.

Most projects do step five first and skip one to four, which is exactly
backwards.

## Common mistakes

- **Stopping partitions at the ceiling.** The single biggest flanking path.
- **Quoting a lab rating as if it were a room performance.** It is not.
- **Sealing "most" of the junction.** A continuous small gap defeats the wall.
- **Running a duct straight through the wall line untreated.** A direct path.
- **Specifying a high-performance partition and installing it ordinarily.** The
  money goes into the wall and the sound goes over it.
- **Trying to make every room private.** Concentrate the budget where privacy
  actually matters.

## Standards referenced

Sound insulation rating in **ISO 717-1**; laboratory measurement in
**ISO 10140**; field measurement in **ISO 16283-1**. Acoustic targets, partition
specification and flanking control for a specific project must be established by
the project's acoustic consultant.

## Frequently asked questions

### What is the most common flanking path in an office?

The ceiling void, by a wide margin. A partition that stops at the suspended ceiling leaves a continuous open plenum above it connecting the two rooms, and conversation travels straight over the top. It is the explanation for most complaints about meeting rooms that were built to a high specification and still cannot hold a private conversation.

### Does taking the partition to the slab solve it?

It removes the biggest path, but not all of them. Sound can still travel through ducts crossing the wall line, through a raised floor void, along continuous structural elements, and through any gap at the junctions. Taking the partition to the slab is necessary for real privacy; it is not by itself sufficient.

### Why do lab ratings not match what is measured on site?

Because a laboratory test isolates the partition and eliminates flanking by design, under ISO 10140. The building does not. Field measurement under ISO 16283-1 measures the whole situation including every flanking path, and the result is routinely lower. That gap is not a product failure; it is the difference between a component and an installation.

### What about services penetrations?

Every penetration is a potential path. A duct crossing the wall line carries sound directly unless it is treated; sockets placed back to back in the same stud cavity connect the two rooms through the wall's own cavity; a sprinkler pipe or cable tray passing through an unsealed sleeve does the same. Each needs sealing, offsetting or acoustic treatment as part of the design, not as a site afterthought.

### How much difference do small gaps make?

Disproportionately more than people expect. A continuous gap of a few millimetres along the head or base of a partition can undo much of its rated performance, because sound passes through openings far more readily than through mass. This is why sealing detail, rather than partition specification, is usually what separates a room that works from one that does not.

### Where should the effort go on a limited budget?

Into fewer rooms done properly rather than many rooms done partially. A floor with two genuinely private rooms — partition to slab, junctions sealed, services treated — is more useful than eight rooms that all leak. Decide which conversations actually need privacy, and concentrate the specification there.

## Sources

- [ISO 717-1](https://www.iso.org/standards.html) — Acoustics — rating of sound insulation in buildings
- [ISO 16283-1](https://www.iso.org/standards.html) — Field measurement of sound insulation — airborne
- [ISO 10140](https://www.iso.org/standards.html) — Laboratory measurement of sound insulation

---

Source: https://hagerstone.com/glossary/acoustic-flanking