Swiss Cheese Model for Odour Pollution in Remediation

Odour pollution is one of the most socially disruptive and technically challenging environmental hazards associated with contaminated land remediation. Unlike toxicological risks that may remain latent, malodorous emissions generate immediate community response, regulatory scrutiny, and operational constraints. Managing odour at remediation sites therefore demands a systemic, multi-layered approach to risk mitigation. This chapter examines the Swiss Cheese Model (Reason, 1990) as a framework for understanding how odour pollution events arise during soil and groundwater remediation, and how it can guide the design of robust, redundant defence architectures.

Framework

The Swiss Cheese Model, developed by James Reason within organisational accident theory, argues that hazardous outcomes rarely result from a single failure. Instead, they occur when latent conditions and active failures align across multiple defensive layers. Each layer is like a slice of Swiss cheese, containing weaknesses or “holes” caused by human fallibility, equipment limitations, procedural gaps, or environmental variability. When these holes line up at the same time, a hazard can pass through every barrier and produce an adverse event. The model distinguishes between latent conditions, which are systemic weaknesses embedded in organisational design, training, resourcing, or regulatory context, and active failures, which are immediate frontline errors or equipment malfunctions that trigger the event sequence.

Layers

At contaminated land remediation sites, odours commonly arise from volatile organic compounds, reduced sulphur compounds, or ammonia released during excavation, dewatering, ex-situ treatment, or vapour intrusion mitigation. The defensive layers that can be treated as “slices” include engineering controls, operational and administrative controls, monitoring and early warning systems, and regulatory and community engagement frameworks.

Engineering

Engineering controls provide the primary physical barrier through containment and treatment technologies such as impermeable covers, negative-pressure enclosures, vapour extraction systems, biofilters, activated carbon scrubbers, and soil capping. Weaknesses in this layer can include seal degradation under meteorological stress, blower failure, carbon breakthrough, or inadequate design capacity during peak emission conditions.

Operations

Operational and administrative controls form a secondary procedural barrier, including standard operating procedures, scheduling work to avoid odour-sensitive periods, meteorological monitoring, and stockpile management protocols. Weaknesses can appear as non-compliance with procedures, inadequate pre-works risk assessment, failure to adjust operations during poor dispersion conditions such as low wind speed and high atmospheric stability, or insufficient buffer zone management.

Monitoring

Monitoring and early warning systems provide tertiary detection capability through real-time ambient odour monitoring, downwind sensor networks, community complaint hotlines, and meteorological forecasting. Weaknesses can include sensor calibration drift, delayed data transmission, threshold setting errors, or poor spatial coverage that allows plume migration to go unnoticed.

Engagement

Regulatory and community engagement frameworks act as an institutional defence layer through environmental permits, odour impact criteria, community liaison committees, and incident response protocols. Weaknesses may include ambiguous regulatory thresholds, delayed enforcement mechanisms, eroded community trust, or inadequate stakeholder communication that slows transparent escalation and response.

Trajectory

A hypothetical excavation of petroleum hydrocarbon-contaminated soil illustrates how an odour event can occur when holes align across layers. A latent condition might be a vapour extraction system that was downsized during value engineering, increasing vulnerability in engineering controls. Another latent condition might be abbreviated operator training due to schedule compression, weakening administrative controls.

Failures

An active failure could occur when early-morning excavation encounters an unexpected pocket of highly volatile weathered hydrocarbons. Another active failure could be a data transmission fault at the on-site meteorological station, preventing detection of a temperature inversion that suppresses vertical dispersion.

Alignment

In this alignment, emissions bypass the undersized extraction system, procedural controls fail to prompt an operational pause, the monitoring system provides no early warning, and the community—already sensitised by previous inadequate consultation—submits multiple complaints that trigger regulatory intervention and a work stoppage. The event is not attributable to a single cause; it results from coincident weaknesses across layers that allow the hazard trajectory to reach receptors.

Implications

The Swiss Cheese Model shifts odour management away from reliance on any single “perfect” barrier and toward defence-in-depth with heterogeneous, independent layers. In practice, this supports using diverse defences so hole patterns do not correlate, routinely auditing latent conditions such as design compromises and training gaps as well as frontline compliance, investigating near-misses by examining the state of all upstream defences and not only the triggering failure, and maintaining barriers dynamically because holes migrate as equipment ages, staffing changes, and site conditions evolve.

Conclusion

The Swiss Cheese Model offers a systems-theoretic way to analyse odour pollution events at remediation sites. By framing odour management as a set of imperfect, evolving barriers rather than a static inventory of controls, practitioners can better anticipate failure trajectories, invest in layered redundancy, and build organisational cultures that actively reduce weaknesses before they align. Future research could quantitatively model barrier interdependencies and hole-correlation probabilities to improve predictive risk assessment in contaminated land engineering.