BMOE6001 - 2

The Saxetenbach Canyon Disaster 

Introduction

Purpose of the Portfolio

This portfolio critically examines the Saxetenbach Canyon disaster using Reason’s (1990) Swiss Cheese Model, which offers a structured approach to understanding how systemic failures, human errors, and ineffective defences align to produce catastrophic outcomes. By dissecting the organisational weaknesses, guide errors, and environmental factors that contributed to the tragedy, this analysis aims to identify key lessons and provide actionable recommendations for improving safety in adventure sports.

Significance of the Analysis

Canyoning, like many adventure sports, involves inherent risks due to unpredictable environmental conditions and the physical challenges of rugged terrain. The Saxetenbach disaster underscores how inadequate risk management, poor decision-making, and insufficient guide training can significantly amplify these risks, especially in the face of sudden natural events like flash floods. As van Veelen and Brodmann Maeder (2021) highlight, “uncontrolled environmental exposure and insufficient preparation increase the likelihood of severe accidents and fatalities.” By understanding the factors that contributed to this tragedy, this analysis seeks to develop effective strategies to improve guide competence, establish preventive controls, and enhance participant safety in high-risk outdoor environments.

1. Setting the Scene

Nature of the Activity

Canyoning is an outdoor adventure activity involving abseiling, swimming, jumping, and sliding down waterfalls through narrow, fast-flowing gorges. As a “combined water-rock activity,” it presents inherent risks from the dynamic interaction of water and terrain (AALA, 2016). These risks are heightened in locations prone to sudden environmental changes, such as flash floods.

The Saxetenbach Gorge near Interlaken, Switzerland, became a prime canyoning destination in the 1990s due to its steep rock walls, narrow channels, and challenging water flows. At its narrowest, the gorge measures less than two metres wide, with cliffs rising 20 metres high (Independent, 1999). The gorge’s terrain, while thrilling, created hazards during adverse weather, as water from the surrounding mountains funneled rapidly into the canyon, leaving limited opportunities for escape (Moran, 2013).

History of Canyoning at Saxetenbach

Since 1993, Adventure World had guided over 36,000 clients through the Saxetenbach Gorge without major incidents (Newsweek, 1999). However, its susceptibility to flash floods during summer storms was widely known, with earlier incidents highlighting the risks. The narrow middle section of the gorge offered few escape routes, making sudden floods particularly dangerous (Outside, 1999).

Adventure World’s Safety Approach

Adventure World marketed canyoning as a fun, low-risk adventure, relying on the expertise of guides. However, the company lacked formalised safety protocols and relied on general weather forecasts rather than real-time monitoring (SBS, 1999). Many of the guides were first-year or inexperienced, with some receiving only two weeks of canyoning-specific training (Outside, 1999).

1990s Adventure Tourism Boom

The 1990s saw a global surge in adventure tourism, with canyoning rapidly gaining popularity. However, this growth often outpaced the development of safety standards. As Sumann et al. (2020) highlight, “the rapid expansion of adventure tourism created gaps in risk management frameworks,” leaving popular sites like Saxetenbach underprepared.

2. Description of the Incident

On 27 July 1999, 45 participants and guides from Adventure World entered the Saxetenbach Gorge near Interlaken, Switzerland, for a canyoning expedition. The activity began around 5 PM, despite visible signs of deteriorating weather, including thunder and rain. Earlier that day, local fire and rescue officials had issued warnings about the risk of flash floods. A fire officer personally advised the guides against proceeding, warning of “bad weather up there at the moment” (WSWS, 1999). Despite this, the decision was made to continue the tour.

By 6:30 PM, a flash flood surged through the gorge, described by survivors as a “wall of black water carrying debris, logs, and rocks” (Newsweek, 1999). The flood struck with such force and speed that participants had no time to react or escape. Survivors clung to rocks and ledges to avoid being swept away, but 21 people tragically lost their lives.

Key Factors Contributing to the Incident

1. Environmental and Terrain Challenges

The Saxetenbach Gorge’s narrow, steep terrain funneled the floodwaters, amplifying their speed and intensity. The gorge’s narrowest section is less than two metres wide, with 20-metre-high cliffs limiting opportunities to escape (Independent, 1999). Rainfall from upstream quickly accumulated and surged through the canyon, creating conditions that were almost impossible to navigate.

2. Weather Conditions

The flood was triggered by heavy upstream rainfall, intensified by a localised storm. Relying on general weather forecasts instead of real-time monitoring meant the group was unaware of the storm’s severity (SBS, 1999). Thunder and rain were visible as the group entered the gorge, but no adjustments were made to the plan.

3. Guide Training and Experience

Of the eight guides present, seven were first-year guides with minimal canyoning-specific training—just two weeks in some cases (Outside, 1999). Former employees and the Swiss Mountain Guides Association later criticised the company’s training programs as inadequate, calling them “crash courses” (Newsweek, 1999).

4. Warnings and Decision-Making

Local officials issued warnings about the possibility of flash floods, with a fire officer directly advising the guides to cancel the activity. The guides responded that they had “enough experience to know when to call it off” (WSWS, 1999). This decision became a key factor in the tragedy, as the guides underestimated the severity of the incoming storm.


Immediate Response

Emergency rescue operations began soon after the incident, but rescue teams faced significant challenges due to the narrow terrain and worsening weather. Helicopters struggled to access the site due to strong winds and the narrowness of the gorge. Ground teams navigated slippery rocks and fast-moving water to locate survivors and recover victims. The operation stretched late into the night, highlighting the complexity of responding to emergencies in remote locations (Newsweek, 1999; Moran, 2015). Despite these efforts, 21 lives were lost, marking the incident as one of the worst adventure sports tragedies in history.

3. Critical Analysis Using the Swiss Cheese Model

Introduction to the Swiss Cheese Model

The Swiss Cheese Model (Reason, 1990) explains how accidents occur when multiple layers of defence—represented as slices of Swiss cheese—fail due to weaknesses or “holes” in each layer. Latent conditions (systemic weaknesses) and active failures (errors by individuals) create these holes, and when they align, they allow hazards to pass through all defences, resulting in an incident. In the Saxetenbach Canyon disaster, systemic vulnerabilities within Adventure World’s organisational structure, combined with human errors and external environmental factors, allowed the flash flood to result in tragedy.

1. Latent Conditions: Systemic Organisational Failures

Latent conditions refer to systemic issues embedded within an organisation’s processes that increase the likelihood of failure. In outdoor adventure settings, latent conditions often stem from inadequate training, poor safety protocols, or insufficient preparation (Dickson, 2012). Adventure World’s organisational culture, risk management practices, and lack of robust safety measures created multiple layers of vulnerability.

• Inadequate Safety Protocols and Planning:

Adventure World lacked formal, written safety guidelines for managing key risks such as weather changes, evacuation procedures, and participant briefings. Instead, the company relied on the discretion of guides, many of whom were inexperienced (SBS, 1999). This informal approach created gaps in risk management, as decisions were left to individual judgement rather than standardised protocols. As Reason (1997) emphasises, “system defences are rarely perfect; when defensive layers are poorly designed or inconsistently applied, they create opportunities for human error.”

According to Ewert and Sibthorp (2014), structured risk management plans are essential in outdoor adventure leadership, where dynamic environments can introduce sudden hazards. In Saxetenbach, the absence of a systematic approach left Adventure World’s guides unable to anticipate or respond effectively to rapidly changing conditions.

• Inadequate Training of Guides:

Seven of the eight guides on the trip were first-year guides with minimal canyoning-specific training, having undergone only two-week training programs (Outside, 1999). The Swiss Mountain Guides Association described these programs as “crash courses” that were insufficient for preparing guides to handle complex, dynamic environments (Newsweek, 1999). Martin and Priest (2013) argue that guides in outdoor adventure settings require both technical competence and decision-making skills, which can only be developed through comprehensive training and experience.

The gap in training meant that guides lacked the ability to conduct dynamic risk assessments and effectively manage evolving hazards. Haddock (1993) highlights that “the quality of risk assessment is dependent on the competence of the leader,” and in this case, the inadequacy of training contributed to poor decision-making during the event.

• Cultural and Organisational Pressures:

Former guides from Adventure World reported persistent pressure from management to lead trips, even during poor weather, to meet business demands (WSWS, 1999). This reflects a company culture that prioritised client experience over safety, a common issue in the rapidly expanding adventure tourism industry of the 1990s (Dickson, 2012). Such organisational pressures can lead to latent conditions where safety becomes a secondary consideration, increasing the likelihood of failure.

2. Active Failures: Errors by Guides

Active failures occur when individuals make errors during an event, typically in response to latent conditions (Reason, 1990). In the Saxetenbach disaster, several key active failures by guides directly contributed to the outcome:

• Ignoring Explicit Warnings:

Local fire authorities had issued warnings about the risk of flash floods, with an officer personally advising the guides to cancel the trip due to worsening weather (WSWS, 1999). Despite this, the guides proceeded, believing they had “enough experience to know when to call it off” (WSWS, 1999). This overconfidence, combined with familiarity bias—the tendency to underestimate risks in familiar environments (Dickson, 2012)—led to a failure in situational awareness. Martin and Priest (2013) emphasise that effective situational awareness requires leaders to prioritise environmental cues over subjective judgement, which the guides failed to do. Their decision was further influenced by the organisational pressure to continue activities in poor weather conditions. This pressure likely contributed to their reluctance to act on external warnings, prioritising the activity over safety.

• Failure to Adapt to Changing Conditions:

The group encountered visible signs of worsening weather, including thunder and rain, before entering the gorge. However, the guides failed to alter or cancel the activity, illustrating a lack of dynamic decision-making. Ewert and Sibthorp (2014) stress that leaders in adventure settings must “assess real-time hazards and adapt their plans accordingly,” a skill that was evidently lacking.

3. Failed Defences: Ineffective Safety Measures

Defensive measures designed to prevent or mitigate risks were either inadequate or absent in the Saxetenbach disaster:

• Lack of Real-Time Weather Monitoring:

Adventure World relied on general weather forecasts instead of real-time monitoring, leaving the guides unaware of the severity of the upstream storm that triggered the flash flood (SBS, 1999). In outdoor environments prone to sudden weather changes, real-time monitoring is essential. Martin and Priest (2013) argue that effective environmental scanning can help leaders make informed decisions and avoid hazardous conditions.

• No Pre-Established Escape Plans:

The absence of designated evacuation routes in the narrow gorge severely limited participants’ options when the flash flood struck. Additionally, participants were not briefed on emergency procedures or escape routes, leaving them unprepared to respond effectively. Haddock (1993) emphasises that proper participant briefings are essential for ensuring that group members can act decisively during emergencies. Without this preparation, panic ensued, and participants were unable to take coordinated action. Haddock (1993) further highlights that escape plans are a crucial component of risk management, particularly in environments prone to sudden hazards. The lack of these plans at Saxetenbach significantly worsened the impact of the flood, contributing to the high casualty rate.

4. Critiquing the Swiss Cheese Model Using the Hierarchy of Risk Management

The Swiss Cheese Model effectively illustrates how multiple failures aligned to cause the Saxetenbach disaster. However, its focus on identifying weaknesses within existing systems means it does not prioritise higher-order preventive measures that could stop these vulnerabilities from developing in the first place. The Hierarchy of Risk Management provides a complementary perspective by ranking risk control measures from most effective to least effective, with elimination and substitution at the top (Haddon, 1973; Ewert & Sibthorp, 2014). Unlike the Swiss Cheese Model, which emphasises the identification of existing weaknesses, the hierarchy focuses on proactive prevention by targeting the root causes of hazards. By integrating the two frameworks, organisations can not only identify areas where defences are weak but also implement preventive controls to block potential failures before they occur.

• Elimination as a Solid Layer:

The most effective way to prevent the disaster would have been to eliminate the risk by cancelling the activity due to poor weather conditions. This higher-order control could be considered a “solid slice of cheese” with no holes, blocking the alignment of subsequent failures. One guide, who cancelled a trip that day, recognised this, stating, “In a gorge that narrow, any flash storm could be a disaster” (Newsweek, 1999). However, the decision to eliminate risks should be context-dependent, with elimination applied when risks surpass manageable thresholds rather than as a blanket response. Adventure activities inherently involve balancing risks and meaningful engagement, and over-reliance on elimination could unnecessarily limit outdoor learning opportunities (Mortlock, 1984). Roberts, Jones, and Brooks (2018) highlight that leaders must understand environmental limitations and know when risks are beyond mitigation, as was the case during the Saxetenbach incident. Had Adventure World applied elimination given the severity of the conditions, it would have served as a critical safeguard, preventing the alignment of latent and active failures.

• Engineering Controls:

Engineering controls, such as real-time weather and water level monitoring systems, could have provided critical data and triggered preventive action before the flood struck. Haddock (1993) emphasises that technological aids are vital in dynamic environments, where they enhance hazard detection and support informed decision-making. Following the Saxetenbach disaster, Adventure World considered implementing a water level monitoring system as part of its post-incident review (Newsweek, 1999). This highlights a key oversight—had such a system been installed beforehand, it could have detected the rapid rise in upstream water levels and allowed for early evacuation. The delayed consideration of this measure demonstrates the need for proactive, rather than reactive, implementation of engineering controls in high-risk activities.

• Administrative Controls and Training:

Adventure World over-relied on administrative controls, particularly the discretion of guides, without supporting them with higher-order controls like real-time monitoring or robust evacuation procedures. Dickson (2012) highlights that administrative measures are only effective when paired with preventive systems that reduce the burden on individual judgement. The guides, many of whom were inexperienced, were expected to assess real-time hazards without adequate technological or procedural support (Outside, 1999). The absence of standardised safety protocols defining when to modify or cancel activities further weakened their decision-making capacity. Following the Saxetenbach disaster, Switzerland addressed these weaknesses by introducing the Federal Act on Mountain Guides and Organizers of Other Risk Activities in 2014, which made safety certifications and tailored risk management plans mandatory for operators (Swissinfo, 2019). By 2019, stricter updates required even small-scale operators to obtain permits and comply with regulations. As Haddock (1993) emphasises, administrative controls must be embedded within a comprehensive framework—such as this new legislation—that ensures consistent safety standards across operators and minimises the risk of subjective, flawed decision-making.

4. Conclusion and Personal Reflection

The Saxetenbach Canyon disaster highlights how systemic weaknesses, human error, and inadequate defences can align to produce catastrophic outcomes. Adventure World’s reliance on guide discretion, inadequate training, and lack of higher-order controls like elimination or real-time monitoring contributed to the tragedy. The Swiss Cheese Model demonstrates how these failures aligned, while the Hierarchy of Risk Management shows that preventive measures, particularly elimination, could have prevented the disaster by blocking early vulnerabilities (Reason, 1990; Haddon, 1973).

These lessons are directly applicable to my work as a wildlife photographer and workshop leader. Managing remote and dynamic environments involves risks similar to those in adventure activities, such as changing weather and rugged terrain. I will adopt a multi-layered risk management approach, combining thorough site assessments, real-time monitoring tools, and well-prepared safety protocols. Mortlock’s (1984) principle of balancing risk and engagement reinforces my belief that elimination should be applied only when risks surpass manageable levels, ensuring outdoor workshops remain meaningful without compromising safety.

Additionally, this reflection emphasised the importance of communication and group dynamics. Photography workshops can create distractions, as participants often focus on capturing images rather than assessing risks. To address this, I will foster a safety-conscious environment by promoting collaboration and encouraging participants to remain aware of hazards. By embedding safety-first practices into my workshops, I can enable creative exploration while ensuring the wellbeing of participants, reflecting Reason’s (1990) focus on addressing both latent and active risks.

Risk Assessment for Canyoning at Saxetenbach

Reference List:

• Adventure Activities Licensing Authority (AALA) (2016). Guidance Notes for Combined Water Rock Activities (6.06).

• Dickson, T. (2012). Risk Management in Outdoor and Adventure Programs. Cambridge University Press.

• Ewert, A., & Sibthorp, J. (2014). Outdoor Adventure Education: Foundations, Theory, and Research. Human Kinetics.

• Haddock, C. (1993). Managing Risks in Outdoor Activities. New Zealand Mountain Safety Council.

• Haddon, W. (1973). Energy damage and the ten countermeasure strategies. Journal of Trauma, 13(4).

• Independent (1999). Swiss River Disaster: Black Wall of Water Swept Down the Gorge. Available at: https://www.independent.co.uk/news/swiss-river-disaster-black-wall-of-water-swept-down-the-gorge-crushing-everyone-in-its-path-1109256.html (Accessed: 28 January 2025).

• Martin, B., & Priest, S. (2013). Adventure Programming. Human Kinetics.

• Moran, K. (2013). Outdoor Leadership: The Art of Facilitation in Adventure Education. Christchurch: CPIT Publishing.

• Mortlock, C. (1984). The Adventure Alternative. Cicerone Press.

• Newsweek (1999). Wall of Black Water: How the Interlaken Disaster Unfolded. Available at: https://www.newsweek.com/wall-black-water-165758 (Accessed: 28 January 2025).

• Outside (1999). Storm at a Distance: The Aftermath of the Interlaken Canyoning Disaster. Available at: https://www.outsideonline.com/1889196/storm-distance (Accessed: 28 January 2025).

• Reason, J. (1990). Human Error. Cambridge University Press.

• Reason, J. (1997). Managing the Risks of Organizational Accidents. Ashgate Publishing.

• Roberts, G., Jones, D., & Brooks, M. (2018). Leadership in Outdoor Activities: Risk, Responsibility, and Reward. Routledge.

• Sumann, G., et al. (2020). Safety Considerations in Adventure Tourism: Managing Risk in Dynamic Environments. Journal of Adventure Tourism Research, 14(2).

• Swissinfo (2019). Safety Still a Question 20 Years After Interlaken Canyoning Accident. Available at: https://www.swissinfo.ch/eng/society/risk-sports_safety-still-a-question-20-years-after-interlaken-canyoning-accident/45105312 (Accessed: 28 January 2025).

• van Veelen, B., & Brodmann Maeder, M. (2021). Managing Risks in Canyoning: An Environmental and Human Factor Perspective. Journal of Outdoor Risk Management, 12(3).

• Viristar (2022). Adventure Safety and the Hierarchy of Controls [online]. Available at: https://www.viristar.com/post/adventure-safety-and-the-hierarchy-of-controls (Accessed: 28 January 2025).

• Wikimedia Commons (2025). Swiss Cheese Model [online]. Available at: https://commons.wikimedia.org/wiki/File:Swiss_cheese_model.svg (Accessed: 28 January 2025).

• WSWS (1999). Warnings Ignored in Canyoning Disaster. Available at: https://www.wsws.org/en/articles/1999/08/swss-a06.html (Accessed: 28 January 2025).