An Analysis of the 1972 Aviation Accident in the Andes

Introduction

As one of the most well-known plane crashes occurring in South America, the 1972 crash of Uruguayan Air Force Flight 571 has been the center of speculation and analysis about causes and hypothetical prevention scenarios. Carrying an Uruguayan rugby team, the Fairchild FH-227D saw a delay in the scheduled flight takeoff and route, an alteration that ultimately contributed to the destruction of the plane and the demise of almost 30 individuals. The intended destination of the aircraft was Santiago, Chile, and the course took the turboprop over a portion of the Andes mountain range. Shrouded in clouds and struck by inclement weather, the peaks forced the pilot and the co-pilot to resort to dead or deduced reckoning, a system in which the operators had to rely on exact calculations of location over time, as well as change in velocity due to wind speed. While this is a commonly-used scheme used by some of the best carriers such as Boeing, the controllers of this aircraft failed to take into account the necessary measurements of wind speed and difference in plane’s location and cleared themselves to descend without knowing the subsequent danger in which they were placing themselves. Due to the miscalculation of the pilots, the plane struck the mountains and eventually was engaged in a controlled accident into the terrain. This was entirely a preventable accident, and the resulting fatalities were unnecessary. This accident, a result of human error, was ultimately caused by the lack of experience of the co-pilot, and further the pilot and co-pilot’s errors in calculation while in a state of deduced reckoning, and these errors can be strictly analyzed first by using James Reason’s “Swiss Cheese Model”, and subsequently through the use of Scott Shappell and Douglas Wiegmann’s “Human Factors Analysis and Classification System”.

History of the Crash

Uruguayan Air Force Flight 571 took off from Montevideo, Uruguay, on October 13th, 1972, carrying the Old Christians Club rugby team to a match in Santiago, Chile. The aircraft was carrying 40 passengers, as well as 5 members of the crew. The trip had originally begun on the previous day, October 12th, but the flight plan had to be altered; once the aircraft had taken off from Carrasco International Airport, it was forced to set down in Mendoza, Argentina due to a disastrous storm that had begun on a portion of the Andes mountain range. Upon reaching this city, the pilot and co-pilot devised a plan to cross over the Andes mountain range through a pass that lay south of their current location; the aircraft would fly south, parallel to the mountain range, and upon reaching the pass would bank east and cross safely before passing over the town of Curico, Chile, and eventually on to Santiago. This was the attempt that ultimately led to their demise. After taking off on October 13th, the pilots travelled along their determined route. As they began to cross over the pass on the Andes mountain range, the inclement weather was still prevalent and prevented any visibility at all. The aircraft travelled over the pass for a set duration, and when the pilot and co-pilot believed themselves to be above the town of Curico, they made this claim and were cleared to descend.

This proved to be a mistake that would cost their lives. Instead of actually flying over Curico, the plane was still on the pass and had not yet crossed over safely. The aircraft collided with part of the mountain range, first destroying the right wing, then the vertical stabilizer, and finally the left wing; the remains of the craft careened through the air and collided with the ground, eventually coming to a stop in a patch of snow. 

Results of the Crash

The results of the crash were devastating. Twelve individuals on the plane died immediately, which included the pilot and the co-pilot of the aircraft, with another five dying the next morning due to injuries sustained during the crash. Another passenger survived until the eighth day, finally dying from previous wounds that occurred on the day of the crash. This left 27 individuals to fight for their lives in the cold. The remaining passengers struggled to survive in the treacherous terrain of the Andes mountain range, with their stories eventually detailed in books and articles upon rescue. In December of the same year, some two months after the initial accident, with only 16 survivors still alive, three trekked up to find help and were the eventual cause of the rescue of the rest. The struggles of survival in the mountains for these individuals tested their wits and their humanity, with the story of their time spent in the range marred with instances of cannibalism (much like what occurred with the Roanoke group) of their fellow passengers. While the survivors accused of this act claimed they did so only to non-relatives, it left emotional scars that these passengers still deal with to this day.

After the rescue of the passengers, once the commotion had settled down, the issue that remained in everyone’s mind was that of the cause of the crash. The families of the non-survivors wanted peace of mind, and aviation committees abound wanted to use this to find ways to prevent such occurrences in the future. What was eventually claimed, though, was that due to the inclement weather over the pass at the time, the pilot and co-pilot were unable to calculate the distance they had travelled in relation to the mountain pass over a certain time and were mistaken in their assumption of location. As they flew in dead reckoning, they inaccurately calculated the wind speeds and how they affected the aircraft’s forward motion – this is the failure that led to the accident. While this is an acceptable claim, it is too broad a notion, and needs specification in order to truly point out the flaws in a certain system that caused the demise of these individuals.

Factors That Led to the Crash

The first factor that must be brought to light is the experience and knowledge level of the operators of the aircraft, and more specifically the aviation history of the co-pilot. The pilot of the plane was Col. Julio Cesar Ferradas, a seasoned veteran of the Uruguayan Air Force (FAU) and one of the FAU’s most seasoned pilots (Read, 1974, p. 11). While he was so experienced, he was acting as an instructor on this flight to the aircraft’s co-pilot. The co-pilot, Lt. Col. Dante Hector Lagurara, was also a member of the FAU, but was piloting the aircraft on this flight per the requirements of his training. His experience in the air force was not quite as lauded as that of Ferradas: according to Project Get Out and Walk, a database dedicated to chronicling records of ejections from aircrafts around the globe, Lagurara was involved in one of the only thirteen aircraft ejections in the history of the FAU (n.d.). In 1963, before colliding with another member of the FAU, Lagurara, who was piloting his own aircraft, ejected safely in order to avoid a collision death. The details of the crash or ejection are not detailed further, but as it is stated that approximately 70 to 80 percent of all aviation related crashes are human error related, it is acceptable to assume that some amount of operator miscalculation went into the incident. No further information is available about Lagurara’s flight history, but this permanent stain on his record cannot be avoided.

What remains to be the biggest issue, though, is the set of actions the pilot and co-pilot took together when maneuvering the aircraft through the pass over the Andes mountain range. Due to the inclement weather, and the resulting lack of visibility because of cloud cover, the pilots were forced to begin operating in a system called ‘dead’ or ‘deduced reckoning’. Because of these circumstances, the automatic direction finder (ADF) went into failure making the reading of coordination or of any measurement by the computer’s controls a sheer impossibility. Deduced reckoning then came into effect, a system in which the pilots are forced to rely on the calculation of basic measurements and pre-calculated assumptions, aspects that are covered in recurrent STEM training. These measurements are the current and projected speed of the aircraft, as well as the time of flight and the duration of the course. The pilots of this aircraft relied on their assumption of the forward speed of the aircraft and spent an exact amount of time crossing over the pass before believing that they were above the town of Curico. While in an arena free of variables this system might have worked, due to external factors it proved to highlight a glaring mistake. The pilots of aircraft failed to take into consideration the exact strength of a head wind when traveling over the pass, and thus miscalculated the distance they had traveled; accurate calculations of the head wind would have shown that they still had a significant distance to proceed before it was safe to descend below the clouds. As they began their descent, the pilots realized their mistake when passing through the clouds, seeing the mountainside directly below them instead of the expected sight of the town of Curico. When Lagurara, the co-pilot and operator under supervision, attempted to ascend once again to a suitable height in order to pass safely over the range, he found the aircraft caught in pockets of air that propelled the aircraft once again below the safety of the clouds. As this happened, the wing of the craft struck the mountain and the ensuing crash began, with Lagurara unable to correct his mistake.

Nature of Machine Failure

It is nearly impossible to put any blame of this accident on the failure of a piece of machinery. While the automatic direction finder, an older tool used to navigate the aircraft by the use of radio waves, failed to operate correctly, this was a common occurrence in aircraft during this time that were traveling in inclement weather. In fact, it was also known that ADF-projected radio waves would be reflected by mountains or cliffs in an unpredictable manner, suggesting that the pilots of this plane should have taken extra precautions when preparing for this flight, knowing that during some extent of the trip they would be relying on their knowledge of the terrain alone. But instead, the pair found themselves trapped in a situation for which they had not prepared. Outside of this factor, the machinery of the aircraft was working well. The Fairchild plane had logged only 972 hours after acquisition by the Uruguayan Air Force, a relatively small number.

Time Details of the Crash

In discounting the failure of a piece of machinery within the aircraft, this only leaves room for human error. As the pilots lost visibility and orientation to the train, they were instead forced to rely on their wits. A timetable of the events is shown here:

• 2:18 PM – The plane takes off again from Mendoza airport, and Lagurara proceeds to take the plane to an altitude of 18,000 feet. 

• 3:08 PM – Lagurara informs the Malargue airport of their position. He states his estimated time of arrival at Planchon as 3:21 PM, the location where air traffic control jurisdiction switched. 

• 3:21 PM – Lagurara informed the Santiago airport that he was flying over the Planchon pass and would reach Curico at 3:32.

• 3:24 PM – The aircraft alerted the air traffic control that they could see Curico, and subsequently banked at a right angle towards the north. Shortly after this, despite the difference in time between the original statement and this action, air traffic control in Santiago cleared the aircraft to begin descending. 

• 3:30 PM – It was verified that the aircraft had dropped 3,000 feet. During this time, the aircraft dipped below the clouds and succumbed to the unalterable descent into the mountains below.

Swiss Cheese Model in Analysis

The necessary framework to use while analyzing the crash begins with James Reason’s “Swiss Cheese Model”, a model of human error that is often used to uncover the causes that led to the occurrences of accidents. This model contains four distinct “levels”, with the first three, Organizational Influences, Unsafe Supervision, and Preconditions for Unsafe Acts, considered latent factors, and the fourth, Unsafe Acts, considered an active factor (Reason, 2000). In this case it is most important to look at the last two, as the FAU provided a new, serviced aircraft in which to fly as well as capable pilots, and the supervision was being handled by one of the FAU’s most highly regarded airmen. The Preconditions for Unsafe Acts were indeed a factor: did the pair not fly into a storm that had halted their process just a day before? The crew was obviously aware of the inherent dangers, but also understood that their passengers were required to make an appearance at a certain time and took off anyway. This is where the latent factors disappear, and the active factor comes into play. The Unsafe Acts taken by the pilots are obvious – they, due to factors unknown, miscalculated the appropriate measurements. 

HFACS Model in Analysis

This is where Reason’s “Swiss Cheese Model” fails to complete this analysis, and where Scott Shappell and Douglas Wiegmann’s “Human Factors Analysis and Classification System” takes off. Based on the structure of Reason’s model, Shappell and Wiegmann improve on the esoteric nature of the former’s idea by creating concrete, realistic causes for aviation disasters. Shappell and Wiegmann’s HFACS model was developed in the year 2000 and has since then been used in over 1,000 aviation operations and accidents, analyzing the incidences with accuracy. It has also allowed specific military branches to understand the factors that contribute to aviation accidents, with these factions creating intervention systems to prevent similar occurrences in the future. 

Their model also includes Unsafe Acts, as can be seen in Figure 1, and branches out into two categories, Errors and Violations. This analysis will not be discussing the nature of Violations, as it is safe to assume in this instance that the pilots did not willfully violate any rules regarding the measurement system in a state of dead reckoning. This leaves Errors, which us subsequently divided into three sections, Decision Errors, Skill-Based Errors, and Perceptual Errors. Decision Errors includes such things as a breakdown in visual scan, failure to prioritize attention, and omission of a step in a procedure, none of which truly maintain relevance in this specific instance. Skill-Based Errors includes factors such as a wrong response to an emergency, an inappropriate maneuver, or the execution of an improper procedure, all of which remain too vague to be applicable to this specific crash. Perceptual Errors, the third branch beneath Errors in Figure 1, though, specifically contains the mention of an error made due to misjudged distance/altitude/airspeed or even spatial disorientation. Shappell and Wiegmann (2000) directly state, “[N]ot unexpectedly, when one’s perception of the world differs from reality, errors can, and often do, occur” (p. 5). He does continue to mention that it is a misjudgment to assume that the disorientation of the pilot is the actual perceptual error; instead, it is the response to the pilot’s environment while in this impaired state of awareness. The error of the pilot, as the authors demonstrate, is a decision made based on unconfirmed information that opens the door for failure; the pilot makes a “best guess”, without accurately understanding his or her surroundings. This is exactly what played out during the fateful flight across the Andes. The pilots, Ferradas and Lagurara, made the decision to bank over the pass in the mountain range without knowing exactly their location. These two were not required to make a best guess – they believed the information they carried was correct. But their mistake was entirely the result of a Perceptual Error, as they miscalculated their location over the mountains, due to inclement weather, strong headwinds, and failing instruments.

Solutions and Prevention Tactics

While the “Swiss Cheese Model” acts as a good baseline to understand the conditions leading up to a crash, it unfortunately never explains what causes the “holes in the cheese”, or the real-world constraints. It is simply a theory, and largely lacks practical application. Shappell and Wiegmann’s HFACS model helps to bridge the gap to reality. Because of the existence of airline crashes due to pilot error, D.T. Mcruer and the National Research Council, in their book Aviation Safety and Pilot Control, have stated the need for steps to be taken by relevant organizations to decrease accidents (1997, Recommendations 4: 1-5). What the book fails to elaborate on, though, are the steps themselves. Shappell and Wiegmann’s HFACS model has no direct recommendations, as it is only to be used as a framework to understand the causes and factors that lead up to a crash. It allows for the necessary organizations to analyze patterns in human errors, and subsequently develop safety intervention procedures tailored to individual error types.

What the HFACS is not, though, is a solution. Shappell and Wiegmann understand this himself, in stating that it is only a starting point for the development of new safety protocol (2000, p. 13-15). While it attempts to explain the holes that James Reason spotted in his model, it can only be used to understand certain problems. Once these problems are identified, preventative measures can be taken. As a next step, these preventative measures need to be monitored in order to ensure their efficacy; the cycle must continue until a significant decrease in aviation-related accidents is witnessed.

Application to this Case

In truth, while the HFACS can be applied to this case in hopes of understanding the underlying causes or factors, it is not possible to say this accident could have been prevented, as the HFACS inherently relies on the data used from previous incidents. The decision made by the pilots, though, can be analyzed and used to increase awareness of safety protocol for future training or protocols for unmanned aerial systems. It is an unfortunate system, using the misfortune of others to improve the quality of life for the future aviation specialists, but it is the only way to move forward.

Conclusion

Using the framework of this model, it can be understood that this crash was due to a human error, and more specifically, a perceptual error, made while in a state of impaired judgment. This framework, the HFACS, accurately analyzes the human-related factors that enter into play before the occurrence of an aviation accident, and while it does not act as a solution itself, it lays the groundwork for the necessary institutions to incorporate safety intervention programs in order to ensure the decreased prevalence of such incidents in the future.

References

Chronological History of Uruguay (FAU) Losses and Ejections. (n.d.). Project Get Out and Walk. Retrieved from www.ejection-history.org.uk/Country-By-Country/uruguay.htm

Mcruer, D. (1997). Aviation Safety and Pilot Control: Understanding and Preventing Unfavorable Pilot-Vehicle Interactions. Washington, D.C.: National Academy Press.

Murray, S. R. (1997). Deliberate Decision Making by Aircraft Pilots: A Simple Reminder to Avoid Decision Making Under Panic. The International Journal of Aviation Psychology, 7(1), 83-100.

Naidich, J., & Schor, H. (1944). Air Navigation Made Easy. New York: McGraw-Hill Book Co.

Read, P. P. (1974). Alive; the Story of the Andes Survivors. ([1st ed.). Philadelphia: Lippincott.

Reason, J. (2000). Human Error: Models and Management. BMJ, 320(7237), 768-770.

Shappell, S. A., & Wiegmann, D. A. (2000). US Department of Transportation - Federal Aviation Administration. The Human Factors Analysis and Classification System–HFACS, N/A, 1-15.

Shappell, S. A., & Wiegmann, D. A. (2003). A Human Error Approach to Aviation Accident Analysis: The Human Factors Analysis and Classification System. Aldershot, Hants, England: Ashgate.

Shappell, S., Wiegmann, D., & Urbana-Champaign, I. L. (2001). Unraveling the Mystery of General Aviation Controlled Flight into Terrain Accidents Using HFACS. In Proceedings of the Eleventh Symposium on Aviation Psychology, Columbus, OH (pp. 1-6).

Whishaw, I., Hines, D., & Wallace, D. (2001, July 25). Dead reckoning (path integration) requires the hippocampal formation: evidence from spontaneous exploration and spatial learning tasks in light (allothetic) and dark (idiothetic) tests. Behavioural Brain Research. Retrieved from hsinnamon.web.wesleyan.edu/wescourses/NSB-Psyc255/Readings/17.%20Spatial%20Limbic%20System/Whishaw.pdf