Travis Zinzer – CHE 202 Project Part 2
Section I: Background
On October 23, 2009, a explosion occurred at the Caribbean Petroleum Corporation
(CAPECO) facility near San Juan, Puerto Rico. During the routine transfer of gasoline
from a tanker ship to storage tanks, a tank overflowed, releasing approximately 200,000
gallons of gasoline which subsequently formed a vapor cloud covering 107 acres. The
cloud ignited after reaching electrical equipment in the facilities wastewater treatment
area, causing an explosion that registered 2.9 on the Richter scale and multiple tank
fires that burned for multiple days. Fortunately there were no fatalities, but the incident
resulted in damage to the nearby infrastructure as well as environmental damage to the
soil, wetlands and waterways in the area. In total around 300 nearby buildings were
damaged, 17 of the facility’s tanks were destroyed, the nearby Fort Buchanan military
facility suffered over 5 million in damage, thousands of gallons of oil and fire
suppression foam were released, and a Presidential declaration of emergency occurred
due to the incident.
The tank overflowed due to the transfer of millions of gallons of gasoline from a tanker
ship to multiple storage tanks, requiring more than 24 hours of continuous operation.
This operation had many hazards involved. Besides the obvious hazard of handling
flammable materials, the presence of ignition sources and potential for tank overflow
made a recipe for disaster. The facility did have some safeguards in place in order to
prevent such a situation. All tanks were equipped with a float and measuring tape inside
the tank so that the level could be normally be checked offsite, unfortunately the monitor
of the tank being filled had to be checked manually every hour. In addition there each
tank had a secondary containment dike around it.
In the event of an incident such as this there were little to no mitigating actions or
contingency plans. The local fire department had little preparation for industry fires of
this scale and lacked the equipment and training as well. Poor lighting made it difficult
for operators to see the overflow or the vapor cloud. Industry and regulatory standards
were also lax as facilities such as this one had no requirements for risk assessment,
automatic overfill prevention systems, independent alarms for rising liquid levels, or
anything more than a single layer of protection against tank over flow.
In the aftermath the CSB recognized a need for multiple layers of protection against
tank overflow, automatic overfill prevention systems, independent alarms. As such they
made multiple regulatory, operational and industry standard related recommendations
for the reclassification of bulk storage terminals as high-hazard, mandatory risk
assessments, better incident reporting requirements, better emergency response
planning, more comprehensive training for emergency responders, improved equipment
and training for facility workers, and requiring more than a single layer of protection in
such facilities.
As seen above the NFPA diamond shows a high risk of flammability for n-Octane under
almost “all ambient temperature conditions”.
Based on the Swiss cheese model there were three key holes in the “cheese” that
aligned to allow this accident to occur. The primary engineering controls failure was that
the tank level monitoring system was unreliable and had no redundancy. The tanks
electronic transmitter card was out of service, and the facility relied solely on a
mechanical float and tape measuring device that they knew was prone to failure. Most
critically, there was no independent high-level alarm or automatic overfill prevention
system to serve as a backup engineering control. On the administrative side, the tank
filling procedures were inadequate and introduced unnecessary risk. Operators were
required to partially open intake valves to tanks while filling other tanks due to pipeline
pressure issues, making it difficult to accurately calculate fill times and increased the
potential for human error. Additionally, there was no robust system in place for
maintenance and repair of safety equipment, as the facility routinely took weeks to
repair problems with the level monitoring system. Finally, there was insufficient training
and preparation for emergency scenarios. The CSB investigation revealed that
CAPECO did not adequately train facility personnel to deal with tank fires, and they had
minimal pre-planning with local emergency responders.
Above is a bow tie diagram of the incident showing the relationship between potential
causes, the hazardous event, and its consequences.
What I found to be most unsettling about the incident is that the facility had so few
regulations. It’s common knowledge that gasoline is flammable and dangerous, so it
should be common sense that there needs to be more than one safeguard in place to
prevent incidents such as these. This goes especially for facilities such as these, which
are near peoples homes and businesses. Furthermore it’s disappointing to know that
the facility wasn’t maintained very well and broken equipment wasn’t quickly replaced
because that’s a recipe for disaster.
The most important lesson to learn from this incident is the necessity of having multiple
layers of protection in hazardous operations such as this one. This is for a few reasons.
First, there was a single point of failure, the mechanical level monitor, that led to the
overflow. Second, the consequences of the incident were devastating to the surrounding
homes and businesses and the environmental impact was long lasting. Lastly, from a
cost benefit perspective it would be relatively cheap to install alarms and automatic
shutoff systems when compared to the destruction of 17 storage tanks, 5 million in to
just one nearby facility, environmental cleanup costs, and the reputational damage to
CAPECO that followed the incident.