The Amoy Gardens event was characterised by a lack of maintenance that allowed floor drains serving bathroom areas to become dry, providing a path for contaminated air to pass into habitable space.

This effect was exacerbated by the use of extractor fans in the bathrooms served.

It is important to avoid ‘preparing for the previous battle’, but it is clear that future prevention depends on good design and good maintenance, coupled to some degree with control on occupancy, usage and unauthorised system modifications.

Graphic evidence from Amoy Gardens shows trapless sinks connected to the drainage network. This feature was not directly implicated but it cannot be allowable.

Defects

A maintenance regime in a large and complex building network requires a degree of prior knowledge of possible defective appliance trap seal locations.

Assistance in determining likely areas of failure may stem from the application of fundamental air pressure transient theory.

This will allow the use of transient simulation and transient response measurement to identify – during periods of system non-use – the location of depleted trap seals.

The fundamental point here is that air pressure transients, once propagated in the network, will continue to be reflected and re-reflected at all system boundaries until they naturally die away due to frictional attenuation.

The reflective properties of various boundaries are well known from the wider water hammer subject area.

For example, closed ends result in a positive, or +1, reflection coefficient so that any incoming transient – positive or negative – is reflected as a pressure wave of equal magnitude and the same sign.

Conversely the reflection at an open termination, or any constant pressure zone, has the same magnitude as the incoming wave but reversed sign, effectively a -1 reflection coefficient.

An expected full trap that has dried out or has been depleted will display a quite different reflection coefficient and will thus be recognisable.

A low-amplitude air pressure transient propagated in a building drainage system therefore obeys all the mechanisms of transient propagation (travelling at the acoustic velocity in air) and system response (in terms of boundary reflection and transmission).

The propagation may be simulated by the proven method of characteristics solution of the St Venant equations.

Prediction

The changed response of a network with a dry trap to a low-amplitude applied pressure pulse – effectively a new open termination with an identifiably different reflection coefficient – may be predicted.

In practical terms it is possible to predict the arrival time of a reflection from a changed termination at any monitoring location in the network.

Predictions at two monitoring locations in the network would clearly identify the location of the dry trap.

To translate this into a practical method of identifying dry trap seals it would be necessary to subject the building drainage and vent system to a low-amplitude pulse – probably by activation of a fan or a simple purpose-built pressure propagation device.

This would be done during a quiescent period: a night-time short-duration automated pulse would be sufficient.

Activation at periods of non-flow in the network will enhance the probability of a successful identification of the dry trap seal.

It will also remove many of the operational difficulties encountered in similar methodology directed towards water supply network leakage identification.

In view of the established consequences of poor maintenance and trap seal depletion in complex building drainage networks, this approach seems timely and interesting.

Active control

Successful development would have implications for facilities management in complex buildings.