On Monday, September 30, 1997, Williams Fire & Hazard Control was advised of a hazardous materials incident involving a shipment of medical grade ethylene oxide. The manufacturer of the product advised us that, because of a recent design change, cartridges in which the chemical was packaged were being monitored in a laboratory.
The day after the product was shipped, 80 percent of the random sample cartridges in the lab were determine to be leaking. The source of the leak was a newly designed seal. The leak experienced in the lab was a very slow, very low volume leak of .01 grams per day per cartridge. (Acceptable historical levels = .0012 gpd.)
The concern, other than the obvious, was that the lab was temperature controlled at an average 72 degrees Fahrenheit, unlike the temperature in the 8-foot by 8-foot by 40-foot corrugated-side steel intermodal shipping container ("dry van" type) bound for Cartagena, Columbia, by ship. The manufacturers' recommended storage temperature for EO packaged in small cartridges is 59 degrees to 86 degrees F.
It was soon discovered the Columbian-bound ship was not the only vessel with the potential problem. There were five more vessels heading for ports in Santiago, Chile; Durban, South Africa and Seoul, South Korea. We would attempt to meet each vessel in its port of call.
The first hours were spent gathering intelligence regarding the product and the leaking cartridges. We requested the vessels be tracked allowing a travel plan to be developed.
Information on the first vessel was good news. It was still in the Port of Miami in Florida, a scheduled off loading point. The captain was advised of the potential problem on board his vessel. We requested it be held there, he agreed and we were on our way.
It was decided this would be a good venue for all the teams to gain experience- in the task that would have to be repeated in other parts of the world. The fact it was in our own backyard gave us an edge when considering unknowns and resources available.
Upon arrival we met with Port of Miami officials, the ship's captain, chief engineer and chief mate in charge of stowing arrangements, a representative of the ship owners and the Metro-Dade County Fire Rescue Hazardous Material Team. After giving a short briefing, we went to work.
One extremely fascinating and surprising component of this incident was the misconception as to the total weight of the product involved. It was known that the net weight of the product in each cartridge was 100 grams. Each cartridge was approximately the size of a disposable lighter. There were 12 cartridges per box and eight boxes per case. That added up to 96 cartridges per case and there were 60 cases, making a grand total of 5,760 cartridges. However, the surprise was that this adds up to 1,269.849 pounds of EO, more than enough to sink the ship it was riding on.
The vessel was tied up at its berth with its bow facing east. A northerly 10-12 knot wind prevailed. The effected container was stowed two rows above the main deck forward on the starboard side. Access to the area was made via an aluminum extension ladder provided by the ship's crew.
DAY 1, MIAMI
PHASE 1 RECON
The first step was to monitor atmosphere in the area of the EO container for LEL (lower explosive limit), CO and O 2 . Our research discovered that a presence of EO in the atmosphere might register a reading on the CO head of our gas detector. This is not a true measurement of EO, only an indication it may be present. According to the hazardous cargo manifest we had no reason to believe CO would be present from any other source, therefore any indication of CO would be treated as a presence of EO. The ambient atmosphere readings were normal. The ambient temperature was 89 degrees F, three degrees higher than the recommended maximum storage temperature for EO.
PHASE 2 RECON
The next step was to monitor the atmosphere inside the container. Most "dry van" type shipping containers are vented. These vents are not so much for ventilation but are designed to relieve normal overpressure conditions caused by changing temperatures. The typical configuration1 is one vent forward (closed end) on the starboard side high corner and one vent aft (door end) port side high corner.
Each vent is weather covered to prevent water from entering the container. Normal procedure is to drill three to five small holes, 1/4 -inch to 3/8 -inch in diameter, through the container wall. Two types of weather cover can be encountered, either a welded steel plate or a molded plastic open at the bottom riveted through the container wall. In this case, the weather covers were the molded plastic type. It was decided to breach the forward weather cover by boring a hole through the plastic face to expose the vent holes in the container's steel wall. Although the weather cover was constructed of plastic and sparks were not thought to be a concern, extreme caution was observed during this operation.
Once again a portable ground ladder was put to use. The issue here is a basic safety concern. The top of a typical steel shipping container is a thin-skinned steel weather cover not intended or considered to be part of the structural integrity of the box. It has little support under the middle span and moves and buckles under the weight of even a small man. The surface is smooth and extremely slippery when wet. Even working from the third rung required caution and someone to foot the ladder.
A multi-gas meter with pump was used to take the first reading using a 12-inch rigid sample probe just inside at the top of the box. Three more readings were taken via sample tubing. The tubing was marked in one-foot increments and samples were taken at the 6 foot, 4 foot, 2 foot and floor levels -- in theory. There was no guarantee that the tubing was reaching any of these levels due to the cargo arrangement inside. Tubing was secured in place to provide a quick connection for future samples.
Sample port number two was scheduled to be the port side high corner aft vent. After some thought and a genuine concern regarding the ability to get a sample near the floor level, a decision was made to breach the rubber door seal at the bottom of the door. The only obstacles would be a medium gauge sheet metal drip cap which ran along the bottom of the door seal and the rubber seal itself. A section of the rubber seal was cut away to allow a large flat blade screwdriver to be inserted between the door frame and the drip cap. The drip cap was gently pried up to open a hole for the sampling probe. The probe was then secured in place to provide a connection for future samples. One other option was to drill a hole through the wooden floor of the container with a pneumatic drill motor. After considering the possible generation of heat and limited accessibility this option was aborted.
Following the initial reconnaissance, the information gathered and recorded was shared with the Coast Guard, including the Coast Guard's Gulf Coast Strike Team members from Mobile, AL, now on scene; representatives of the Miami Marine Safety Office, and Metro-Dade's Haz-Mat Team. Our recommendation was to carefully remove the effected container from the vessel and place it in a remote area of the Port.
All in attendance agreed, making the captain of the vessel extremely happy. The next step would be to convince the port officials to allow it to happen.
Problem one with the plan is that there is no remote area in the Port of Miami. It is literally wall to wall operations. The port officials preferred that the container stay aboard and that the ship sail far, far away. Common sense and the Coast Guard's responsibility to mitigate a known hazard made that option impossible.
The next solution offered by the port officials was to have the shipper rent a deck barge, place the container on it and tow us off shore to work it. The scary thing about this option was the Coast Guard didn't seem to have a problem with that! Obviously we did and were able to convince them it was not a viable option for us.
The basis for all this commotion was the literal translation of the Department of Transportation's Emergency Response Guidebook. It seems the port officials, Coast Guard and even the hazmat team were convinced the evacuation zone for this problem would be so extensive that the entire port and half of the downtown water front area would have to be cleared. (To be fair to the authorities involved, it should be noted the guidebook's information is based on a large liquid spill, not a small slow vapor leak.)
Once all parties agreed the problem was manageable and could be handled safely without leaving the port, a work area was sectioned off at the extreme east end using barricades. An evacuation zone was established to include 500 feet in all directions, most of which was water on three sides. The container was off-loaded and placed in the restricted area.
Monitoring resumed as soon as the container touched the ground and again when it was placed in the designated area to ensure conditions had not changed due to the movement. The results were no change. Monitoring continued throughout the day while the required site safety plan and operations plan were prepared and approved. The Coast Guard was very thorough and extremely cautious when reviewing our plan and considered this to be a very high profile incident. Important enough to warrant the port captain's presence. Metro-Dade Fire Rescue's Hazardous Materials Team and an accompanying engine company were very supportive and would play an important role in the success of the operation.
DAY 2, MIAMI
PHASE 1 OPERATIONS:
Container Access, Entry & Natural Ventilation
Operations commenced at first light the next day. Operational zones were established, i.e., hot zone, warm zone, and cold zone. Metro-Dade Fire Rescue supplied an engine company for firewater support. Two 1 3 /4 -inch hand-lines and a portable ground monitor (350 gpm) were deployed to be used primarily for vapor dispersion. The fire department also stood-by with an advanced life support hazmat rescue transport to provide medical monitoring for our entry teams and emergency rescue and transportation if needed.
A local environmental contractor was hired to provide and operate a de-contamination station for our entry teams.
The initial two person entry team and two person rescue backup donned "Level C" personal protective equipment, (bunker gear with self-contained breathing apparatus) to provide access and entry into the container. Metro-Dade Fire Fighters stood-by with hand-lines. Level C PPE was chosen to protect personnel in the event of a flash fire that might occur upon the opening of the steel doors. Once the doors were opened the access and entry team exited the hot zone immediately while the back-up team set up the portable monitor (500 gpm) to scrub and disperse vapor.
PHASE 2 OPERATIONS:
EO Specific Monitoring
Phase two of the operation was to monitor the atmosphere in the container after naturally ventilating the space. "Level B" PPE was worn to prevent exposure due to the EO. A two-member team with back-up standing by
entered the area in front of the open container doors and began monitoring for LEL, utilizing an EO specific monitor that read EO in parts per million. The positive reading on the previous CO monitor turned out to be accurate as the EO unit read 5 ppm.
PHASE 3 OPERATIONS:
Cargo Access & Mechanical Ventilation
Phase three of the operation was the removal of the blocking and cribbing that stabilized the cargo. It consisted of wooden 2- by 4-foot boards criss-crossed and nailed into the wooded deck of the container. The unused space in the container was taken up with wood pallets standing on edge and sheets of thick cardboard. Once all the dunnage was removed, the boxes containing the leaking EO cartridges were identified and accounted for.
Note: In this case every bit of information proved to be accurate from the location and contents loaded on the ship to the general arrangement of the cargo inside the container. Be aware this is not always the case. When handling containerized cargo plan for the worst case scenario. Don't get surprised.
There was a concern that the cardboard packaging surrounding the cartridges would absorb the leaking EO and contain higher concentrations of the product. This was confirmed using our equipment's ability to monitor in and around the cases of EO The concentration was in fact considerably higher.
During this phase of the operation a water powered ram fan was set up to mechanically ventilate and force an exchange of air inside the entire container.
PHASE 4 OPERATIONS:
Removal and Over-Pack
The fourth and final phase of the operation was to render the atmosphere inside the container safe and within normal limits. At that point the operation was turned over to the environmental contractor to overpack the boxes containing the leaking EO. Monitoring and mechanical ventilation was maintained throughout this entire operation.
A poly-lined steel overpack drum was selected by the EO manufacturer and approved by the Coast Guard. The environmental contractor supplied the drum and began to transfer the boxes. As each drum was filled it was sealed, labeled, and palletized for shipment. The cargo was loaded onto an open flatbed commercial truck and hauled directly back to the manufacturer along with the 20 foot container and its remaining cargo.
NEXT STOP ECUADOR
We boarded the privately owned and operated corporate G-4 jet of the EO manufacturer and arrived in Guayaquil, Ecuador, nine hours later to face the next round. The vessel originally bound for Santiago, Chile, was being held there.
With the exception that there were only half as many cartridges to deal with, the situation was the same as in Miami. However, there was a lot less local support, few resources and absolutely no formality. The container carrying the leaking cartridges was off loaded and moved to a remote location in the port normally used for hazardous material. The initial methodology for handling the incident was to repeat the plan that was implemented in Miami. Several of the port employees commented that they had smelled an odor in the vicinity of the shipping container. After completing phases one and two we determined that the odor was not EO (based on the fact the odor threshold limit is > 250 ppm and the concentrate in the shipping container was only at 3 ppm). Drums meeting the requirements of the D.O.T. over-packs used in Miami were brought in.
With the over-pack drums in place phase three of the operation began on the second day. The container door was opened, exposing the boxes of leaking cartridges. The packing and dunnage was a duplicate of what was observed in Miami. After several hours of ventilation the team entered the shipping container and removed the leaking cartridges and placed and sealed them in the over-pack drums.
The U.S. Coast Guard was not involved in the operation. In fact it was apparent there was no party taking responsibility by claiming jurisdiction. Unlike the procedures that were mandated in Miami, this operation allowed us many options for dealing with the leaking containers.
We reviewed the operation again and theorized that since the containers were still leaking at the same rate and we decreased the total atmosphere that the EO gas was dispensing into, the atmosphere in the overpacks would quickly reach the flammable range. Opening one of the over packs confirmed this. After only three hours, the L.E.L. in the over-pack rose to 30 percent.
After several discussions with all parties involved a decision was made to remove the leaking cartridges from the over-pack drums and place them into their own 20 foot shipping container.
The shipping container was to be loaded onto a ship and transported to the original destination. A sampling methodology was developed to provide safe travel and prevent any unacceptable L.E.L. concentrations during transport. This consisted of an EO monitor sampling the atmosphere in the container. When the concentration of EO exceeded 5 ppm, the container doors would be opened until the concentrations were reduced to 0 ppm via natural ventilation.
The manufacturer was advised and acknowledged the findings. They agreed they would treat the packages arriving from Miami accordingly.
The leaking cartridges we shipped back to the manufacturer from Miami arrived just after the completion operations on the second container in Ecua dor. The manufacturer was satisfied with our findings and upon inspection of the first lot opted not to continue with our response services. Their evaluation of the leaking cartridges determined the level of hazard was, in their opinion, manageable by personnel in the respective ports.
Chauncey Naylor is manager for municipal operations with the Mauriceville, TX-based Williams Fire & Hazard Control, a company specializing in controlling bulk flammable liquid fires and hazardous material situations. Williams' 24-hour emergency number is (409) 727-2347. In the UK, call 1642-589696.
- The initial assessment of the situation at the Port of Miami was completed prior to the arrival of the Coast Guard. In this case that was no problem. However, it could have been a concern in some regions of the country.The local fire department's official report of the incident was very flattering to our response and recommended their own department "study and learn from this case history to better prepare for future emergencies dealing with Specialty Hazardous Chemicals."
Be aware of the consequences. Port officials initially wanted to force the ship to sail with a known hazard on board.
A potentially bad decision was made regarding the over-packing of the leaking EO. You must weigh every option! It's impossible to think of everything, but you must!
Marine incidents take time. At a barge fire in Tampa Bay, emergency personnel were on the scene for 36 days. A container ship fire at Papeete, Tahiti, involving calcium hypocholorite required attention from emergency personnel for 29 days.
- An engine room-super-structure fire aboard a bulk cargo container ship 90 miles off shore from Venice, LA, took three days to deal with. Be prepared to support an incident for hours, days or even weeks.