Do you like quick, simple tips that add value to the way you work? Do you like articles that increase your happiness? How about a joke or something to brighten your day? Of course you do! Or you wouldn’t be reading this post. But the real question is, do you want MORE than all of the useful information we provide on this blog? That’s okay – we’ll allow you to be greedy!
A lot of people don’t know we have a company page on LinkedIn that also shares all those things and more. Follow us by clicking the image below that directs to our company page, and then clicking “Follow.”
We also have a training page where we share tips about career/personal development as well as course photos and information about upcoming courses. If you are planning to attend a TapRooT® course or want a job for candidates with root cause analysis skills, click the image below that directs to our training page and then click “Follow.”
Thank you for being part of the global TapRooT® community!
Monday Accident & Lessons Learned: IOGP SAFETY ALERT- GAS BREAK-OUT FROM OIL-BASED MUD WHILE RUNNING CASINGAugust 10th, 2015 by Mark Paradies
IOGP SAFETY ALERT GAS BREAK-OUT FROM OIL-BASED MUD WHILE RUNNING CASING
- The event took place on a 3000 HP Land Drilling Rig while running 9 7/8” casing in the 12 ¼” hole section of an unconventional well. Well was vertical.
- Well architecture consists of 20” Surface Casing at 945m, 13 3/8” casing at 3348 m at 12 ¼” hole section at 4794m.
- Run 9 7/8” casing to 4793m.
- After landing casing, circulated 9 7/8” casing, increasing pump rates from 0.70 m3/min to 1.45 m3/min and SPP of 450 kPa.At 2000 strokes into bottoms up, returns diverted through the manifold. Maximum gas of 2500 units observed prior to going through MGS.
- Circulate through choke at 1.0 – 1.5 m3/min. Initial casing pressure of 170 kPa and SPP of 570 kPa. Casing pressure spiking at 5800 strokes to 9500 kPa and decreasing to 2100 kPa as bottom up strokes expired.
- Significant amount of gas observed at surface. Approximately 5 m3 of invert drilling fluid was spilled over from open bottom poor boy degasser (MGS) while circulating bottoms up.
- Shut in well & monitored for pressure evaluation.
- Observed increase in SIDPP & SICP.
- Continued to monitor pressure evolution.
- Pressure stabilized at SIDPP 300 kPa & SICP 450 kPa.
- Performed drillers method well kill and stabilized well with mud weight of 1750 kg/m3.
- Well was static prior to bottoms up circulation (some reports noted minor flow when landing mandrel hanger).
- During the bottoms up, returns were diverted to MGS as a precaution for trip gas and because pump pressure was much less than expected.
- The gain was not noticed until the well was circulated once the casing was on bottom.
- The section was drilled with a narrow mud-weight window.
What Went Wrong?
- Gas entered the wellbore while well was static.
- Insufficient hydrostatic pressure to prevent influx of formation fluid.
- Well flow checks were inconsistent prior to tripping to run casing and were not sufficient to indicate potential influx of formation fluid into the well.
- Circulation rate was adjusted to take into account the reduced volume of the annulus associated with the casing as compared to the large annular volume associated with drill pipe to ensure that the annular velocity was the same. However, this still resulted in the gas entrained in the drilling fluids coming to surface at a rate that exceeded the capacity of the MGS.
- Choke was in the 100% open position as it should be when gas arrived at the surface because of concerns about exceeding MAASP. This resulted in maximum flow to the MGS exceeding the off-gassing and flaring capacity of the MGS flare system.
- Possibly drilled too far into the HPHT pressure ramp in 12 ¼” hole section.
Corrective Action & Recommendations:
A definitive HPHT well control procedure will be developed for drilling operations in noted area.
- Review MGS Sizing calculations for maximum anticipated gas rates. Understand capacity of oil based mud to absorb gas.
- Develop a log sheet to better monitor and finger print trip gas, bottom’s up gas, connection gas and background gas to better understand potential behaviour at surface.
- Drilling engineering team initiate and lead researching and acquiring a better way of determining needed mud weight trip margin requirements for making trips to log or run casing, with a better understanding of the fluids system EMW and ECD’s.
safety alert number: 266
IOGP Safety Alerts http://safetyzone.iogp.org/
Whilst every effort has been made to ensure the accuracy of the information contained in this publication, neither the IOGP nor any of its members past present or future warrants its accuracy or will, regardless of its or their negligence, assume liability for any foreseeable or unforeseeable use made thereof, which liability is hereby excluded. Consequently, such use is at the recipient’s own risk on the basis that any use by the recipient constitutes agreement to the terms of this disclaimer. The recipient is obliged to inform any subsequent recipient of such terms.This document may provide guidance supplemental to the requirements of local legislation. Nothing herein, however, is intended to replace, amend, supersede or otherwise depart from such requirements. In the event of any conflict or contradiction between the provisions of this document and local legislation, applicable laws shall prevail.
London office: 209-215 Blackfriars Road, London SE1 8NL, United Kingdom T. +44 (0)20 3763 9700 F. +44 (0)20 3763 9701 E. firstname.lastname@example.org
You have established a good performance improvement program, supported by performing solid incident investigations. Your teams are finding good root causes, and your corrective action program is tracking through to completion. But you still seem to be seeing more repeat issues than you expect. What could be the problem?
We find many companies are doing a great job using TapRooT® to find and correct the root causes discovered during their investigations. But many companies are skipping over the Generic Cause Analysis portion of the investigation process. While fixing the individual root causes are likely to prevent that particular issue from happening again, allowing generic causes to fester can sometimes cause similar problems to pop up in unexpected areas.
6 Reasons to Look for Generic Root Causes
Here are 6 reasons to conduct a generic cause analysis on your investigation results:
1. The same incident occurs again at another facility.
2. Your annual review shows the same root cause from several incident investigations.
3. Your audits show recurrence of the same behavior issues.
4. You apply the same corrective action over and over.
5. Similar incidents occur in different departments.
6. The same Causal Factor keeps showing up.
These indicators point to the need to look deeper for generic causes. These generic issues are allowing similar root causes and causal factors to show up in seemingly unrelated incidents. When management is reviewing incident reports and audit findings, one of your checklist items should be to verify that generic causes were considered and either addressed or verified not to be present. Take a look at how your incident review checklist and make sure you are conducting a generic cause analysis during the investigation.
Finding and correcting generic causes are basically a freebie; you’ve already performed the investigation and root cause analysis. There is no reason not to take a few extra minutes and verify that you are fully addressing any generic issues.
What can you learn from transport aircraft accidents? See the FAA Lessons Learned from Transport Plane Accidents page and find out. See:
The Associated Press reported that the US Department of Justice is warning food companies that they could face civil and criminal charges if they poison their customers.
POISON THEIR CUSTOMERS!
Yes, you read it right.
We are again testing the fine line between accidents and criminal behavior.
How does a company know that they have gone over the line? The FDA stops showing up and the FBI puts boundary tape around your facilities.
Are you in the food business? Think it is time to start taking root cause analysis of food safety incidents seriously? You betcha!
Your company can’t afford a Blue Bell Ice Cream incident. You need to effectively analyze and learn from smaller incidents to stop the big accidents from happening.
What tool should you use for effective root cause analysis? The TapRooT® Root Cause Analysis System.
Why choose TapRooT® Root Cause Analysis?
Because it has proven itself effective in a wide number of industries around the world. That’s why industry leaders use it and recommend it to their suppliers.
Find out more about the TapRooT® System at:
And then attend one of our public courses held around the world.
You can attend at no risk because of our iron-clad guarantee:
Attend a TapRooT® Course, go back to work, and use what you have learned to analyze accidents, incidents, near-misses, equipment failures, operating issues, or quality problems. If you don’t find root causes that you previously would have overlooked and if you and your management don’t agree that the corrective actions that you recommend are much more effective, just return your course materials/software and we will refund the entire course fee.
Get started NOW because you can’t afford to wait for the FBI to knock on your door with a warrant in their hand.
Have you ever seen this video about the 2009 train derailment in Graniteville, SC?
Could have we learned these lessons before people were killed?
The 22-year-old man died in hospital after the accident at a plant in Baunatal, 100km north of Frankfurt. He was working as part of a team of contractors installing the robot when it grabbed him, according to the German car manufacturer. Volkswagen’s Heiko Hillwig said it seemed that human error was to blame.
A worker grabs the wrong thing and often gets asked, “what were you thinking?” A robot picks up the wrong thing and we start looking for root causes.
Read the article below to learn more about the fatality and ask why would we not always look for root causes once we identify the actions that occurred?
Lessons learned from five accidents reported by EU and OECD Countries. See:
Read insights on lessons learned from accidents reported in the European Major Accident Reporting System (eMARS) and other accident sources.
47 accidents in eMARS involving contractor safety issues in the chemical or petrochemical industries were examined. Five accidents were chosen on the basis that a contract worker was killed or injured or was involved in the accident.
What do you think? Leave your comments below.
“Doctor… how do you know that the medicine you prescribed him fixed the problem,” the peer asked. “The patient did not come back,” said the doctor.
No matter what the industry and or if the root causes found for an issue was accurate, the medicine can be worse than the bite. Some companies have a formal Management of Change Process or a Design of Experiment Method that they use when adding new actions. On the other extreme, some use the Trial and Error Method… with a little bit of… this is good enough and they will tell us if it doesn’t work.
You can use the formal methods listed above or it can be as simple for some risks to just review with the right people present before implementation of an action occurs. We teach to review for unintended consequences during the creation of and after the implementation of corrective or preventative actions in our 7 Step TapRooT® Root Cause Analysis Process. This task comes with four basic rules first:
1. Remove the risk/hazard or persons from the risk/hazard first if possible. After all, one does not need to train somebody to work safer or provide better tools for the task, if the task and hazard is removed completely. (We teach Safeguard Analysis to help with this step)
2. Have the right people involved throughout the creation of, implementation of and during the review of the corrective or preventative action. Identify any person who has impact on the action, owns the action or will be impacted by the change, to include process experts. (Hint, it is okay to use outside sources too.)
3. Never forget or lose sight of why you are implementing a corrective or preventative action. In our analysis process you must identify the action or inaction (behavior of a person, equipment or process) and each behaviors’ root causes. It is these root causes that must be fixed or mitigated for, in order for the behaviors to go away or me changed. Focus is key here!
4. Plan an immediate observation to the change once it is implemented and a long term audit to ensure the change sustained.
Simple… yes? Maybe? Feel free to post your examples and thoughts.
Is thinking that you are the best a sign of potential problems? (Especially for “routine” work?)
By any measure, the X-31 was a highly successful flight research program at NASA’s Dryden Flight Research Center, now the Armstrong Flight Research Center. It regularly flew several flights a day, accumulating over 550 flights during the course of the program, with a superlative safety record. And yet, on Jan. 19, 1995, on the very last scheduled flight of the X-31 ship No. 1, disaster struck.
View the video below or read about it here: http://www.nasa.gov/centers/dryden/news/X-Press/stories/2004/013004/new_x31.html
Leave your comments below. Complacency? Leave your comments below.
We can all remember some type of major product recall that affected us in the past (tires, brakes, medicine….) or recalls that may be impacting us today (air bags). These recalls all have a major theme, a company made something and somebody got hurt or worse. This is a theme of “them verses those” perception.
Now stop and ask, when is the last time quality and safety was discussed as one topic in your current company’s operations?
You received a defective tool or product….
- You issued a defective tool or product….
- A customer complained….
- A customer was hurt….
Each of the occurrences above often triggers an owner for each type of problem:
- The supplier…
- The vendor…
- The contractor…
- The manufacturer….
- The end user….
Now stop and ask, who would investigate each type of problem? What tools would each group use to investigate? What are their expertise and experiences in investigation, evidence collection, root cause analysis, corrective action development or corrective action implementation?
This is where we create our own internal silo’s for problem solving; each problem often has it’s own department as listed in the company’s organizational chart:
- Customer Service (Quality)
- Manufacturing (Quality or Engineering)
- Supplier Management (Supply or Quality)
- EHS (Safety)
- Risk (Quality)
- Compliance (?)
The investigations then take the shape of the tools and experiences of those departments training and experiences.
Does anyone besides me see a problem or an opportunity here?
A tragic workplace accident.
A life lost.
You see the resolve on the faces in this video to never lose a co-worker … a friend … to this type of accident again.
What do you think about “paying attention” for preventing potential tragedies such as this? Leave your comments below and let’s share ideas to find and fix root causes.
What can you learn from a 1964 video?
How they viewed human performance was certainly different.
What do we know that helps us do better today?
Could better root cause analysis have helped them then? After all, an engine failure in a helicopter is a serious accident to blame on the pilot.
I read an article in the Houston Chronicle about failed corrective actions at Blue Bell® Ice Cream.
It made me wonder:
“Did Blue Bell perform an adequate root cause analysis?”
Sometimes people jump tp conclusions and implement inadequate corrective actions because they don’t address the root causes of the problem.
Its hard to tell without more information, but better root cause analysis sure couldn’t have hurt.
Find out how TapRooT® Root Cause Analysis can help find and fix the root causes of problems by reading about TapRooT®’s history at:
On May 5, 1988, one of United States’ worst oil refinery explosions occurred in Norco, Louisiana. There were six employees that were killed and 42 local residents injured. The blast was said to have reached up to 3o miles away shattering windows, lifting roofs and sending a black fog over the entire town of Norco. Residents were forced to evacuate while officials died the fires down and gathered as much rubble as possible to recover any bodies. In order to discover the root cause of this disaster, the Federal Occupational Health and Safety Administration as well as the Environment Protection Agency came and investigated the scene to gather information. The only possible root cause they could find was the catalytic cracking unit, machine used to break down crude oil into gasoline, because it was at the center of the explosion, but there was no definite cause found. Overall, the amount of damage done cost Shell millions of dollars and set an incredible amount of fear into the residents.
Click below to download a report from the European Major Accidents Reporting System (eMARS) about contractor related safety.
In the city of Chernobyl, Ukraine in April of 1986, there was a major accident in the city’s largest nuclear power plant. The inadequately trained personnel paired with a flawed reactor design did not produce smooth results. The lack of safety precautions caused a steam explosion and fire that released 5% of the radioactive reactor core into the environment. Onsite death toll totaled to two plant workers, however, the overall death toll, due to the release of the radioactive poison, totaled to 56. In order to decrease the amount of poison released and put the fires out, officials poured sand and boron over the entire site. Additionally, they covered the plant with a concrete structure, but that still did not prevent all the residents from relocating and over 9,000 of them being diagnosed with cancer several months later.
Read this article from the United States Nuclear Regulatory Commission for more detailed information: http://www.nrc.gov/reading-rm/doc-collections/fact-sheets/chernobyl-bg.html
Below is a video of the 20 year anniversary news story that ABC News covered in 2006. Take a look at just how deadly and devastating this accident was.
Being proactive is just one way you can help prevent a catastrophic event such as this one. Learn root cause analysis techniques to investigate near-misses, and take proactive steps to avoid a major disaster. (Click here to find out more about TapRooT® Root Cause Analysis Training.)
Grading Your Investigations – Summit Best Practice Session 2 at the 2015 Global TapRooT® Summit in Las VegasApril 23rd, 2015 by Mark Paradies
Mark Paradies is organizing the “Grading Your Investigations” session at the 2015 Global TapRooT® Summit.
At this session participants will use an Excel spreadsheet (download your copy below) to grade a typical incident investigation from your facility.
All participants attending this session are asked to bring an investigation report from your facility and the Excel spreadsheet available below preloaded onto a device so that you can participate in the exercise that will teach attendees to grade their company’s investigations using the spreadsheet.
Watch this video and see if you think they learned all the lessons they should have learned …
A press release from the UK RAIB:
RAIB is investigating an incident that occurred at 17:25 hrs on Saturday 7 March 2015, in which train reporting number 1Z67, the 16:35 hrs service from Bristol Temple Meads to Southend, passed a signal at danger on the approach to Wootton Bassett junction, Wiltshire. The train subsequently came to a stand across the junction. The signal was being maintained at danger in order to protect the movement of a previous train. However, at the time that the SPAD occurred, this previous train had already passed through the junction and was continuing on its journey. No injuries, damage or derailment occurred as a result of the SPAD.
Wootton Bassett junction is situated between Chippenham and Swindon stations on the Great Western main line and is the point at which the line from Bristol, via Bath, converges with the line from South Wales. It is a double track high speed junction which also features low speed crossovers between the up and down main lines (see figure below for detail).
Wootton Bassett junction in 2012 – the lines shown from left to right are the Up Goods,
Up Badminton, Down Badminton, Up Main and Down Main (image courtesy of Network Rail)
The junction is protected from trains approaching on the up main from Chippenham by signal number SN45, which is equipped with both the Automatic Warning System (AWS) and the Train Protection and Warning System (TPWS). This signal is preceded on the up main by signal SN43, which is also equipped with AWS and TPWS. The maximum permitted line speed for trains approaching the junction from this direction is normally 125 mph. However, on 7 March, a temporary speed restriction (TSR) of 85 mph was in place on the approach to signal SN45. A temporary AWS magnet had been placed on the approach to signal SN43 to warn drivers of this TSR.
A diagram of the layout of Wootton Bassett junction – note that some features have been omitted for clarity (not to scale)
The train which passed signal SN45 at danger consisted of steam locomotive number 34067 ‘Tangmere’, and its tender, coupled to 13 coaches. The locomotive is equipped with AWS and TPWS equipment.
The RAIB’s preliminary examination has shown that, at around 17:24 hrs, train 1Z67 was approaching signal SN43 at 59 mph, when it passed over the temporary AWS magnet associated with the TSR. This created both an audible and visual warning in the locomotive’s cab. However, as the driver did not acknowledge this warning within 2.7 seconds, the AWS system on the locomotive automatically applied the train’s brakes. This brake application should have resulted in the train being brought to a stand. In these circumstances, the railway rule book requires that the driver immediately contact the signaller.
The RAIB has found evidence that the driver of 1Z67 did not bring the train to a stand and contact the signaller after experiencing this brake application. Evidence shows that the driver and fireman instead took an action which cancelled the effect of the AWS braking demand after a short period and a reduction in train speed of only around 8 mph. The action taken also had the effect of making subsequent AWS or TPWS brake demands ineffective.
Shortly after passing the AWS magnet for the TSR, the train passed signal SN43, which was at caution. Although the AWS warning associated with this signal was acknowledged by the driver, the speed of the train was not then reduced appropriately on the approach to the next signal, SN45, which was at danger. Because of the earlier actions of the driver and fireman, the TPWS equipment associated with signal SN45 was unable to control the speed of the train on approach to this signal.
As train 1Z67 approached signal SN45, the driver saw that it was at danger and fully applied the train’s brakes. However, by this point there was insufficient distance remaining to bring the train to a stand before it reached the junction beyond SN45. The train subsequently stopped, standing on both the crossovers and the up and down Badminton lines, at around 17:26 hrs. The signalling system had already set the points at the junction in anticipation of the later movement of 1Z67 across it; this meant that no damage was sustained to either the train or the infrastructure as a result of the SPAD.
The RAIB has found no evidence of any malfunction of the signalling, AWS or TPWS equipment involved in the incident.
The RAIB’s investigation will consider the factors that contributed to signal SN45 being passed at danger, including the position of the temporary AWS magnet associated with the TSR. The investigation will also examine the factors that influenced the actions of the train crew, the adequacy of the safety systems installed on the locomotive and the safety management arrangements.
RAIB’s investigation is independent of any investigation by the Office of Rail Regulation.
We will publish our findings, including any recommendations to improve safety, at the conclusion of our investigation.
These findings will be available on the our website.
The UK Rail Accident Investigation Branch announced the start of two rail incident investigations.
The first is an investigation of the injury of a passenger that fell between a London Underground train while being dragged by the train. See the preliminary information at:
This is an accident that was prevented from being worse by the alert actions of the train’s operator.
The second incident was container blown off a freight train. The preliminary information can be found here:
Monday Accident & Lessons Learned: Crane Accident at Tata Steel Plant in the UK brings £200,000 Guilty VerdictMarch 16th, 2015 by Mark Paradies
Tata Steel was found guilty of violating section 2(1) of the Health and Safety at Work etc. Act 1974. The result? A fine of £200,000 plus court costs of £11,190.
HSE Inspector Joanne carter said:
“Given the potential consequences of a ladle holding 300 tonnes of molten metal spilling its load onto the floor, control measures should be watertight. The incident could have been avoided had the safety measures introduced afterwards been in place at the time.”
The article listed the following corrective action:
“Tata has since installed a new camera system, improved lighting, and managers now scrutinise all pre-use checks. If the camera system fails, spotters are put in place to ensure crane hooks are properly latched onto ladle handles.
Here are my thoughts…
- Stating that corrective actions would have prevented an accident is hindsight bias. The question should be, should they have learned these lessons from previous near-misses?
- Reviewing the corrective actions, I’m still left with the question … Should the crane be allowed to operate without the camera system working? Are spotters a good temporary fix? How long should a temporary fix be allowed before the operation is shut down?
- What allows the latches to fail? Shouldn’t this be fixed as well?
What do you think? Is there more to learn from this accident? Leave your comments here.
Press Release from the UK Rail Accident Investigation Branch: Bridge strike and collision between a train and fallen debris at Froxfield, Wiltshire, 22 February 2015March 11th, 2015 by Mark Paradies
Image of debris on track before the collision, looking east.
Train 1C89 approached on the right-hand track
(image courtesy of a member of the public)
Bridge strike and collision between a train and fallen debris at Froxfield, Wiltshire, 22 February 2015
RAIB is investigating a collision between a high speed train (HST) and a bridge parapet which had fallen onto the railway at Oak Hill, an unclassified road off the A4 on the edge of the village of Froxfield, between Hungerford and Bedwyn. The accident occurred at about 17:31 hrs on Sunday 22 February 2015, when the heavily loaded 16:34 hrs First Great Western service from London Paddington to Penzance (train reporting number 1C89) hit brick debris while travelling at about 90 mph (145 km/h). The train driver had no opportunity to brake before hitting the debris, and the impact lifted the front of the train. Fortunately, the train did not derail, and the driver applied the emergency brake. The train stopped after travelling a further 730 metres (800 yards). There were no injuries. The leading power car sustained underframe damage and there was damage to the train’s braking system.
The bridge parapet had originally been struck at about 17:20 hrs by a reversing articulated lorry. The lorry driver had turned off the A4 at a junction just north of the railway bridge, and crossed over the railway before encountering a canal bridge 40 metres further on which he considered to be too narrow for his vehicle. A pair of road signs located just south of the A4 junction warn vehicle drivers of a hump back bridge and double bends but there were no weight or width restriction signs. The lorry driver stopped before the canal bridge and attempted to reverse round a bend and back over the railway bridge without assistance, and was unaware when the rear of his trailer first made contact with, and then toppled, the brick parapet on the east side of the railway bridge. The entire parapet, weighing around 13 tonnes, fell onto the railway, obstructing both tracks. This was witnessed by a car driver who was travelling behind the lorry. The car driver left his vehicle to alert the lorry driver and he then contacted the emergency services by dialing 999 on his mobile phone at about 17:21 hrs.
RAIB’s investigation will consider the sequence of events and factors that led to the accident. The investigation will include a review of the adequacy of road signage and the overall response to the emergency call made by the motorist who witnessed the collapse of the bridge parapet. It will identify any safety lessons from the accident and post-accident response.
RAIB’s investigation is independent of any investigations by the railway industry or safety authority.
The RAIB will publish the findings at the conclusion of the investigation on it’s website.
Press Release from the Chemical Safety Board: CSB Releases Technical Analysis Detailing Likely Causes of 2010 Zinc Explosion and Fire at the Former Horsehead Zinc Facility in Monaca, Pennsylvania, that Killed Two Operators, Injured a ThirdMarch 11th, 2015 by Mark Paradies
CSB Releases Technical Analysis Detailing Likely Causes of 2010 Zinc Explosion and Fire at the Former Horsehead Zinc Facility in Monaca, Pennsylvania, that Killed Two Operators, Injured a Third
Washington, DC, March 11, 2015 – The July 2010 explosion and fire at the former Horsehead zinc refinery in Monaca, Pennsylvania, likely resulted from a buildup of superheated liquid zinc inside a ceramic zinc distillation column, which then “explosively decompressed” and ignited, according to a technical analysis released today by the U.S. Chemical Safety Board (CSB).
Two Horsehead operators, James Taylor and Corey Keller, were killed when the column violently ruptured inside the facility’s refinery building, where multiple zinc distillation columns were operating. The rupture released a large amount of zinc vapor, which at high temperatures combusts spontaneously in the presence of air. The two men had been performing unrelated maintenance work on another nearby column when the explosion and fire occurred. A third operator was seriously injured and could not return to work.
The incident was investigated by multiple agencies including the CSB and the U.S. Occupational Safety and Health Administration, but its underlying cause had remained unexplained. In the fall of 2014, CSB contracted with an internationally known zinc distillation expert to conduct a comprehensive review of the evidence file, including witness interviews, company documents, site photographs, surveillance videos, laboratory test results, and data from the facility’s distributed control system (DCS). The 57-page report of this analysis, prepared by Mr. William Hunter of the United Kingdom, was released today by the CSB. Draft versions of the report were reviewed by Horsehead and by the United Steelworkers local that represented Horsehead workers in Monaca; their comments are included in the final report as appendices.
In the years following the 2010 incident, the Horsehead facility in Monaca was shut down and dismantled. The “New Jersey” zinc process, a distillation-based method that was first developed in the 1920’s and was used for decades in Monaca, is no longer practiced anywhere in the United States, although a number of overseas companies, especially in China, continue to use it.
“Although this particular zinc technology has ceased being used in the U.S., we felt it was important to finally determine why this tragedy occurred,” said CSB Chairperson Dr. Rafael Moure-Eraso. “Our hope is that this will at last provide a measure of closure to family members, as well as inform the safety efforts of overseas companies using similar production methods.”
The Hunter report was based on expert professional opinion, and did not involve any onsite examination of the evidence. CSB investigators made several short deployments to the Horsehead site in 2010 following the incident, interviewing a number of witnesses and documenting conditions at the site.
The explosion involved an indoor distillation column several stories tall. The column consisted of a vertical stack of 48 silicon carbide trays, topped by a reflux tower, and assembled by bricklayers using a specialized mortar. The bottom half of the column was surrounded by a masonry combustion chamber fueled by natural gas and carbon monoxide waste gas. Horsehead typically operated columns of this type for up to 500 days, at which time the columns were dismantled and rebuilt using new trays.
The explosion on July 22, 2010, occurred just 12 days after the construction and startup of “Column B.” Column B was used to separate zinc – which flowed as a liquid from the bottom of the column – from lower-boiling impurities such as cadmium, which exited as a vapor from the overhead line. The column, which operated at more than 1600 °F, normally has only small amounts of liquid metals in the various trays, but flooding of the column creates a very hazardous condition, the analysis noted. Such flooding likely occurred on July 22, 2010.
“Under extreme pressure the tray wall(s) eventually failed, releasing a large volume of zinc vapor and superheated zinc that would flash to vapor, and this pressure pushed out the combustion chamber blast panels,” Mr. Hunter’s report concluded. “The zinc spray and vapor now had access to large amounts of workplace air and this created a massive zinc flame across the workplace.”
After examining all the data, the report determined that the explosion likely occurred because of a partial obstruction of the column sump, a drain-like masonry structure at the base of the column that had not been replaced when the column was rebuilt in June 2010. The previous column that used this sump had to be shut down prematurely due to sump drainage problems, the analysis found. These problems were never adequately corrected, and various problems with the sump were observed during the July 2010 startup of the new Column B. Over at least an hour preceding the explosion, DCS data indicate a gradual warming at the base of Column B, as liquid zinc likely built up and flooded the lower trays, while vapor flow to the overhead condenser ceased.
Ten minutes before the explosion, an alarm sounded in the control room due to a high rate of temperature change in the column waste gases, as zinc likely began leaking out of the column into the combustion chamber, but by then it was probably too late to avert an explosion, according to the analysis. Control room operators responded to the alarm by cutting the flow of fuel gas to Column B but did not reduce the flow of zinc into the column. The unsafe condition of Column B was not understood, and operators inside the building were not warned of the imminent danger.
The technical analysis determined that there was likely an underlying design flaw in the Column B sump involving a structure known as an “underflow” – similar to the liquid U-trap under a domestic sink. The small clearance in the underflow – just 65 millimeters or the height on one brick – had been implicated in other zinc column explosions around the world, and likely allowed dross and other solids to partially obstruct the sump and cause a gradual accumulation of liquid zinc in the column. Liquid zinc in the column causes a dangerous pressure build-up at the bottom and impairs the normal evaporation of vapor, which would otherwise cool the liquid zinc. Instead the liquid zinc becomes superheated by the heat from the combustion chamber, with the pressure eventually rupturing the column and allowing the “explosive decompression.”
The report noted that the Column B sump had previously been used with a different type of column that had a much lower rate of liquid run-off through the sump, so the problem with the sump was only exacerbated when Column B was constructed to separate zinc from cadmium, increasing the liquid flow rate into the sump by a factor of four to five.
The report concluded that Horsehead may have missed several opportunities to avoid the accident, overlooking symptoms of a blocked column sump that were evident days before the accident. “Missing these critical points indicates that, in large measure, hazardous conditions at Monaca had been ‘normalized’ and that process management had become desensitized to what was going on. This raises the question whether sufficient technical support was provided to the plant on a regular basis,” according to Mr. Hunter.
The report noted that New Jersey-type zinc distillation columns have been involved in numerous serious incidents around the world. In 1993 and 1994, two column explosions at a former French zinc factory killed a total of 11 workers. An international committee of experts who investigated the incidents in France identified up to 10 other major incidents at other sites attributable to sump drainage problems. The Monaca facility had suffered five documented column explosions prior to 2010, but none with fatalities, according to the CSB-commissioned report.
For more information, contact Daniel Horowitz at (202) 261-7613 or (202) 441-6074 cell.