Category: Investigations

Monday Accident & Lessons Learned: Root Cause Analysis Failure at Blue Bell Ice Cream?

May 11th, 2015 by

NewImage

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:

http://www.taproot.com/products-services/about-taproot

Remembering An Accident: Norco, Louisiana Oil Refinery Explosion

May 5th, 2015 by

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.


tumblr_mpqsg5hcEo1rwlt7do1_1280

 

Read here for the full story and direct quotes in the New York Times. 

Monday Accident & Lessons Learned: Contractor Accidents in Europe

May 4th, 2015 by

Click below to download a report from the European Major Accidents Reporting System (eMARS) about contractor related safety.

Contractors.pdf

Remembering an Accident: Chernobyl Nuclear Power Plant Disaster

April 27th, 2015 by

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 Vegas

April 23rd, 2015 by

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.

RateRootCauseAnalysis11414.xlsx

Monday Accident & Lessons Learned: Forklift Accident

April 13th, 2015 by

Watch this video and see if you think they learned all the lessons they should have learned …

UK Rail Accident Investigation Board Investigating a Signal Passed at Danger

March 27th, 2015 by

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).

NewImageWootton 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.

Screen Shot 2015 03 27 at 12 28 07 PMA 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.

UK RAIB Investigates Two Separate Rail Incidents

March 26th, 2015 by

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:

http://www.raib.gov.uk/publications/current_investigations_register/150312_Clapham_South.cfm

 This is an accident that was prevented from being worse by the alert actions of the train’s operator.

NewImage

The second incident was container blown off a freight train. The preliminary information can be found here:

http://www.raib.gov.uk/publications/current_investigations_register/150307_Scout_Green.cfm

Monday Accident & Lessons Learned: Crane Accident at Tata Steel Plant in the UK brings £200,000 Guilty Verdict

March 16th, 2015 by

NewImage

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…

  1. 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?
  2. 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?
  3. 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 2015

March 11th, 2015 by

NewImageImage 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 Third

March 11th, 2015 by

NewImage

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.

Root Cause Analysis Tip: Protecting Your Root Cause Analysis from Discovery – Work Product and Motivation

March 10th, 2015 by

Saw an interesting short piece on McGuireWoods web site. It describe a case between Chevron Midstream Pipelines and Sutton Towing LLC. 

It seems the court decided that a “legally chartered” root cause analysis that was performed at the direction of in-house Chevron attorneys was not different from normal root cause analysis that the company performed after any incident.

Why? Because of the motivation to perform this root cause analysis was the same as any other RCA. The judge relied on several pieces of evidence:

  • A Chevron engineer “who agreed in her deposition that the ‘primary purpose of a root cause analysis’ is to ‘prevent a similar accident from happening again in the future,'” and “that it is ‘part of the Chevron ordinary course of business to conduct a root cause analysis’ after an incident.” 
  • “Chevron Pipeline’s President’s statement in an employee newsletter that ‘[w]e are conducting root cause analyses of both incidents and will apply lessons learned. Our ultimate goal remains the same – an incident and injury-free workplace.’”
  • “Chevron’s failure to provide the court examples of Chevron’s ordinary root cause analyses — noting that Chevron’s argument that its ordinary ‘incident reviews’ were different from its ‘legally chartered’ investigation ‘would be more convincing if there was actually another root cause analysis from which to distinguish the legally chartered one.'”

As Thomas Spahn, attorney from McGuireWoods wrote:

“To satisfy the work product motivation element, companies must demonstrate that they did something different or special because they anticipated litigation — beyond what they ordinarily would do, or which they were compelled to do by external or internal requirements.”

Of course, we always recommend that the statements in an incident report be carefully written and accurate. The words used can make a huge difference if your report is introduced as evidence in court.

Remember, what you write may not be interpreted or used as you intended it after the fact. An even if you think your investigation is protected as part of an attorney’s work product, the court may not agree.

Monday Accident & Lessons Learned: Fatal accident involving a track worker near Newark North Gate station 22 January 2014

March 2nd, 2015 by

Summary from the UK Rail Accident Investigation Branch …

At around 11:34 hrs on 22 January 2014, a track worker was struck by a passenger train as it approached Newark North Gate station. He was part of a team of three carrying out ultrasonic inspection of two sets of points at Newark South Junction and was acting in the role of lookout. The accident happened around 70 metres south of the platforms at the station.

A few minutes before the accident, the lookout and two colleagues arrived at the yard adjacent to the tracks in a van. One colleague was in charge of carrying out the inspections and the other, the ‘controller of site safety’ (COSS), was in overall charge of the safety of the team. They had planned to carry out the inspections on lines that were still open to traffic in accordance with a pre-planned safe system of work. All three had many years of relevant experience in their respective roles and were familiar with the work site.

Upon arrival at the yard, the lookout and tester proceeded to the track to start the inspection work; the COSS remained in the van. Shortly after they had started the inspection, the 10:08 hrs London to Newark North Gate passenger service approached. It was due to stop in platform 3, which required it to negotiate two sets of crossovers. The train blew a warning horn and the two staff on site acknowledged the warning and moved to the nominated place of safety. However, just before the train moved onto the first crossover, the lookout turned to face away from the train, walked towards the station and then out of the position of safety. He moved to a position close to where he had been before the train approached, most probably to check for trains approaching in the opposite direction, having decided that the approaching train was proceeding straight into platform 1. Although the train braked and blew a second warning horn, the lookout did not turn to face the train until it was too late for him to take evasive action.

As a consequence of this accident, RAIB has made two recommendations and identified a learning point. The recommendations are addressed to Network Rail and relate to: 

  • improving work site safety discipline and vigilance, especially for teams doing routine work with which they are familiar; and 
  • improving the implementation of Network Rail’s procedures for planning safe systems of work so that the method of working that is chosen minimises the risk to track workers so far as is reasonably practicable, as intended by the procedure.

The learning point relates to improving the implementation of Network Rail’s competence assurance process by providing training and sufficient working time to enable front line managers to implement the associated procedures as intended by Network Rail. 

Download report: 
PDF icon 150216_R012015_Newark_North_Gate.pdf (5,166.00 kb)

 

UK RAIB Investigating Electrical Arcing and Fire Under a Train, Near Windsor, 30 January 2015

February 13th, 2015 by

NewImageThe damaged floor of the train

RAIB is investigating a train fire that occurred on the evening of Friday 30 January 2015, and which caused serious damage to the structure of the train.

The 19:53 hrs South West Trains service from Windsor & Eton Riverside to London Waterloo had travelled about 400 metres after starting from Windsor station, when a small bang was heard under the sixth carriage of the ten-carriage train, followed by about five seconds of severe sparking and flashing.

The train, which was formed of two class 458/5 electric multiple units and was travelling at about 15 mph at the time, stopped immediately. Some smoke entered the carriages through ventilators. There were two passengers in the sixth carriage, and they moved quickly into another part of the train. The guard of the train moved from the rear to the sixth carriage to investigate, and the driver also moved to the middle of the train. They could see that there was still smoke coming from below the sixth carriage, so the driver returned to the front of the train from where he contacted the signaller by radio to ask for the electric power to be switched off. While he was doing this, the floor of the sixth carriage was penetrated by fire, and smoke rapidly filled the vehicle.

There were eleven passengers on the train. The guard, assisted by the crew of another train that was in Windsor station, evacuated the passengers to the track, and helped them walk back to the station. The fire brigade were called, and confirmed by 20:50 that the fire was out. None of the passengers were hurt, but the guard was taken to hospital and treated for smoke inhalation.

RAIB’s preliminary examination found that the fire had originated in severe arcing in a junction box fixed under the carriage floor, where power cables from the collector shoes on either side on the train are connected to the main power cable (‘bus line’) which runs along the train. The arcing had burnt through the floor of the carriage, and had also destroyed parts of the structural members of the carriage body.

RAIB’s investigation will focus on the cable joint in this junction box, and how this joint was designed and assembled. It will also examine how the train’s structure and equipment, and the people in it, might have been protected from the consequences of a failure of this nature.

RAIB’s investigation is independent of any investigation by the Office of Rail Regulation.

RAIB will publish the findings, including any recommendations to improve safety, at the conclusion of our investigation. This report will be available on their website.

 

 

 

 

 

 

Monday Accident & Lessons Learned: IOGP SAFETY ALERT – KICK TAKEN WHILE DRILLING 8 1/2” RESERVOIR SECTION IN DEEPWATER EXPLORATION WELL

February 2nd, 2015 by

KICK TAKEN WHILE DRILLING 8 1/2” RESERVOIR SECTION IN DEEPWATER EXPLORATION WELL

The event took place on a Drillship while drilling the 8-1/2” section of an exploration well.

Well architecture consists of 36” CP at 939 m, 20” casing at 1274 m, 14” casing at 1666 m and 9-5/8” casing at 2702m.

Drill 8-1/2” section at 2860mMD/RT.

Made connection, resumed drilling.Observed 4.5% connection gas.

Observed gas increase to 27%.

Picked up string off bottom & circulated out gas (reciprocating string).

Observed 0.9m3 gain in active pit while circulating gas out of hole.

Shut in well & monitored for pressure evolution.

Observed increase in SIDPP, SICP & WHP.Continued to monitor pressure evolution.

Pressure stabilized at: SIDPP 480 psi & SICP 320 psi.

Spaced out and hung off string on Middle Pipe Rams.

OBSERVATIONS

ECD increase. SPP increase. Possible change in formation.

Sign of under-compaction on the resistivity log. Well Control: (Driller’s Method).

Flush each of Lower, Middle and Upper Choke fail safes. 131 psi (9 bar) back pressure on Choke.

Circulate down string and flush KL. 203 psi (14 bar) back pressure on well.

Open MPR and reduce UAP pressure. Close CMC.

Attempt to strip up and down, negative – string differentially stuck. 40 MT over pull and 36 MT set down weight.

Displace riser through boost line to 1.42sg kill mud. 17.7% max. gas observed.

Displace choke line volume to 1.42sg kill mud

Close MPR, displace KL with 4.5m3 base oil.Open UAP allowing 1.42sg kill mud in riser to u-tube up KL to sweep BOP stack.

Observe 2.5m3 drop in trip tank volume. Close UAP and open MPR.

Displace kill line volume to 1.42sg kill mud through gas relief line.

Perform flow check: well static.

Work drill string free with different parameters.

Circulate down string: FR 380 – 1000 lpm, 624 – 1305 psi (43 – 90 bar). Apply 14KN.m, 36 MT below SOW and 45 MT overpull.

Increase torque to max of 22.5KN.m. Observe drill string free.

Displace Choke, Kill & Boost lines to 1.47sg NABM.

Displace well to 1.47sg NABM.

WHAT WENT WRONG?

  • Failure to recognize the potential/risk: Despite warning signs and pre-job meetings alerting on the nature of the activity, crew was not vigilant and sensitive to the nature of the activity (exploration drilling).
  • Underbalanced Well Warning signs missed: Mud loggers and Geologist not following properly the gas and resistivity trends and not communicating adequately on gas and under-compaction trends seen.
  • Procedure not followed: Established procedure called for reducing BGG to <1% at all times before drilling resumes. Drilling resumed with high BGG.
  • Poor supervision : With warning signs of kick, team failed to recognize the appropriate action to take. And signs of entering high pressure zone unrecognized.

CORRECTIVE ACTIONS & RECOMMENDATIONS

  • No mud transfer should be done when drilling an exploration well without proper and laid out procedure. Follow mud transfer protocol.
  • After each connection and flow check, the corresponding bottoms-up circulation must be monitored.
  • Drilling of exploration wells should not have continued with 10% background gas.
  • Fingerprint of flow back in active system during pumps shut off should be monitored and analyzed.
  • The well site geologist should do more to alert the company man of such observations as decrease of resistivity log.
  • As a follow up, drillers should charge the mud loggers to employ more monitoring tools like ROP break, d-exponent….etc, that give direct indication of a kick.
  • Consideration should be given to having a manual joystick on the cyberchair for choke control. The buttons on the digital display is too small to operate.
  • There was no kill sheet available by the time the kick was taken; not with the drilling contractor or company representatives. Going forward, a kick sheet should be prepared for every phase where a reservoir is to be encountered, compulsory for exploration wells. Kick sheet must be reviewed as operation progresses.
  • Communication with base is mandatory in this type of event.

SOURCE CONTACT

Aafety Alert nNumber: 263 

IOGP Safety Alerts 

DISCLAIMER

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.

RAIB & BEA-TT Investigate Train Fire In Chunnel

January 28th, 2015 by

The UK RAIB and the French Bureau d’Enquetes sur les Accidents de Transport Terrestre (BEA-TT) are jointly investigating a fire on-board a train in the Channel Tunnel. For more information, see:

http://www.raib.gov.uk/publications/current_investigations_register/150117_channel_tunnel.cfm

Monday Accident & Lessons Learned: The US Chemical Safety Board Releases Bulletin on Anhydrous Ammonia Incident near Mobile, Alabama

January 26th, 2015 by

NewImage

 

CSB Releases Safety Bulletin on Anhydrous Ammonia Incident near Mobile, Alabama

Safety Bulletin Notes Five Key Lessons to Prevent Hydraulic Shock

January 15, 2014, East Rutherford, NJ – Today the U.S. Chemical Safety Board released a safety bulletin intended to inform industries that utilize anhydrous ammonia in bulk refrigeration operations on how to avoid a hazard referred to as hydraulic shock.  The safety lessons were derived from an investigation into a 2010 anhydrous ammonia release that occurred at Millard Refrigerated Services Inc., located in Theodore, 
Alabama.

NewImage

The accident occurred before 9:00 am on the morning of August 23, 2010. Two international ships were being loaded when the facility’s refrigeration system experienced “hydraulic shock” which is defined as a sudden, localized pressure surge in piping or equipment resulting from a rapid change in the velocity of a flowing liquid. The highest pressures often occur when vapor and liquid ammonia are present in a single line and are disturbed by a sudden change in volume.

This abnormal transient condition results in a sharp pressure rise with the potential to cause catastrophic failure of piping, valves, and other components – often prior to a hydraulic shock incident there is an audible “hammering” in refrigeration piping. The incident at Millard caused a roof-mounted 12-inch suction pipe to catastrophically fail, resulting in the release of more than 32,000 pounds of anhydrous ammonia.
The release led to one Millard employee sustaining injuries when he fell while attempting to escape from a crane was after it became engulfed in the traveling ammonia cloud.  The large cloud traveled a quarter mile from the facility south toward an area where 800 contractors were working outdoors at a clean-up site for the Deepwater Horizon oil spill. A total of 152 offsite workers and ship crew members reported symptomatic illnesses from ammonia exposure. Thirty two of the offsite workers required hospitalization, four of them in an intensive care unit.

Chairperson Rafael Moure-Eraso said, “The CSB believes that if companies in the ammonia refrigeration industry follow the key lessons from its investigation into the accident at Millard Refrigeration Services, dangerous hydraulic shock events can be avoided – preventing injuries, environmental damage, and potential fatalities.”

Entitled, “Key Lessons for Preventing Hydraulic Shock in Industrial Refrigeration Systems” the bulletin describes that on the day before the incident, on August 22, 2010, the Millard facility experienced a loss of power that lasted over seven hours. During that time the refrigeration system was shut down. The next day the system regained power and was up and running, though operators reported some problems.  While doing some troubleshooting an operator cleared alarms in the control system, which reset the refrigeration cycle on a group of freezer evaporators that were in the process of defrosting. The control system reset caused the freezer evaporator to switch directly from a step in the defrost cycle into refrigeration mode while the evaporator coil still contained hot, high-pressure gas.

The reset triggered a valve to open and low temperature liquid ammonia was fed back into all four evaporator coils before removing the hot ammonia gas. This resulted in both hot, high-pressure gas and extremely low temperature liquid ammonia to be present in the coils and associated piping at the same time. This caused the hot high-pressure ammonia gas to rapidly condense into a liquid.  Because liquid ammonia takes up less volume than ammonia gas – a vacuum was created where the gas had been.  The void sent a wave of liquid ammonia through the piping – causing the “hydraulic shock.”

The pressure surge ruptured the evaporator piping manifold inside one of the freezers and its associated 12-inch piping on the roof of the facility. An estimated 32,100 pounds of ammonia were released into the surrounding environment.

Investigator Lucy Tyler said, “The CSB notes that one key lesson is to avoid the manual interruption of evaporators in defrost and ensure control systems are equipped with password protection to ensure only trained and authorized personnel have the authority to manually override systems.“

The CSB also found that the evaporators at the Millard facility were designed so that one set of valves controlled four separate evaporator coils. As a result, the contents of all four coils connected to that valve group were involved in the hydraulic shock event – leading to a larger, more hazardous pressure surge.

As a result, the CSB notes that when designing ammonia refrigeration systems each evaporator coil should be controlled by a separate set of valves.

The CSB found that immediately after discovering the ammonia release, a decision was made to isolate the source of the leak while the refrigeration system was still operating instead of initiating an emergency shutdown. Shutting down the refrigeration system may have resulted in a smaller release, since all other ammonia-containing equipment associated with the failed rooftop piping continued to operate.

A final key lesson from the CSB’s investigation is that an emergency shutdown should be activated in the event of an ammonia release if a leak cannot be promptly isolated and controlled. Doing so can greatly reduce the amount of ammonia released during an accident.

Monday Accident & Lessons Learned: UK RAIB Report – Near-miss involving construction workers at Heathrow Tunnel Junction, west London, 28 December 2014

January 19th, 2015 by

UK Rail Accident Investigation Branch Press Release…

The UK RAIB is investigating an incident in which a train almost struck two construction workers, and collided with a small trolley, on the Up Airport line between Heathrow Airport Tunnel and the Stockley Flyover.

NewImageYellow engineering trolley underneath the train after the collision (image courtesy of Carillion)

The incident occurred at about 10:05 hrs on Sunday 28 December 2014 and involved train 1Y40, the 09:48 hrs service from London Heathrow Terminal 5 to London Paddington. The track workers jumped clear just before the approaching train struck a small engineering trolley that they had been placing on the line. The train, formed by a Class 332 electric multiple unit, was travelling at approximately 36 mph (58 km/h) when it struck the trolley. 

The two track workers were among a large number of people carrying out construction work on the approach to a new bridge that had been recently constructed adjacent to the existing Stockley Flyover. This new structure, which carries a new railway track over the mainline from London Paddington to Reading, was built as part of the Crossrail surface works being undertaken by Network Rail.

To enable this work to take place, parts of the operational railway in and around the construction site had been closed for varying periods during the few days before the incident. The two construction workers were unaware that the Up Airport line had returned to operational use a few hours before they started to place the trolley onto this line. They formed part of an eight person workgroup which included a Controller of Site Safety (COSS). The COSS and other group members were not with the two track workers at the time of the incident. The presence of temporary fencing, intended to provide a barrier between construction activities and the operational railway, did not prevent the two track workers accessing the open line.

Network Rail owned the infrastructure at the site of the accident and had employed Carillion Construction as the Principal Contractor for the construction works. The two track workers and the COSS were all employed by sub-contractors.

RAIB’s investigation will establish the sequence of events, examine how the work was planned, how the staff involved were being managed and the way in which railway safety rules are applied on large construction sites adjacent to the operational railway. It will also seek to understand the actions of the people involved, and factors that may have influenced their behaviour.

RAIB will also consider whether there is any overlap between this incident and the factors which resulted in an irregular dangerous occurrence at the same construction site on the previous day. This occurrence involved a gang of railway workers who walked along a line that was open to traffic, and without any form of protection, until other construction workers warned them that the line was open to traffic.

The RAIB investigation is independent of any investigations by the safety authority or the police. RAIB will publish its findings at the conclusion of the investigation. This report will be available on the RAIB website.

- – – – – 

What can we learn BEFORE the investigation is complete?

First, this “near-miss” was actually a hit.

In this case it was called a near-miss because no one was injured. However, the train and trolley were damaged and work was delayed. For operations, maintenance, and construction, this was an incident. In other words, it was a safety near-miss but it was an operation, maintenance, and construction hit.

Many incidents that don’t have immediate safety consequences do have immediate cost, productivity, and reliability consequences that are worthy of an investigation. And in this case, the operations incident also had potential to become a fatality. This even more reason to perform a thorough root cause analysis.

UK RAIB Press Release: Investigating tram derailment near Mitcham Junction, London, 29 December 2014

January 12th, 2015 by

At about 23:55 hrs on Monday 29 December 2014, a tram travelling from New Addington to Wimbledon on the Croydon Tramlink system became derailed shortly after leaving the tram stop at Mitcham Junction, while travelling at about 11 km/h (7 mph). There were about 20 passengers, plus the driver, on board the tram, and no-one was hurt. There was some minor damage to the tram.

To the west of Mitcham Junction tram stop, the single tram line becomes two lines at a set of spring-operated points. On leaving the tram stop, the tram driver noticed that an indicator, which shows the position of these points, was indicating that the points were not correctly set. He stopped the tram before reaching the points, and after speaking to the tramway control room by radio, he left the tram and used an operating lever to manually move the points until he observed that the indicator was showing that they were correctly set. He then drove the tram slowly over the points, but the centre bogie and one wheelset of the trailing bogie became derailed.

 

NewImage

Image showing derailed tram near Mitcham Junction

RAIB’s investigation will focus on the points mechanism and the way that it behaves in degraded operating conditions.

RAIB’s investigation is independent of any investigation by the railway industry or the Office of Rail Regulation.

The UK RAIB will publish their findings, including any recommendations to improve safety, at the conclusion of its investigation. This report will be available at http://www.raib.gov.uk.

Monday Accident & Lessons Learned: UK RAIB Investigations of an unauthorised entry of a train onto a single line at Greenford

January 12th, 2015 by

Screen Shot 2014 12 22 at 4 59 05 PM

Unauthorised entry of a train onto a single line at Greenford

20 March 2014

 From the UK Rail Accident Investigation Branch:

At around 11:55 hrs on Thursday 20 March 2014, the 11:36 hrs passenger train from London Paddington to West Ruislip, operated by Chiltern Railways, passed two consecutive signals at danger near Greenford, west London. It was stopped when a signaller sent an emergency radio message to the driver. Although no-one was hurt in the incident, the unauthorised entry of a train onto a single line creates the potential for a serious collision.

A freight train had passed the junction at Greenford shortly before the passenger train was due. Because the freight train was still occupying the line between Greenford and South Ruislip, the signaller at Greenford kept the signal at the junction at danger. The passenger train, travelling at about 20 mph (32 km/h), passed this signal and the next one, 142 yards (130 metres) further on, which was also at danger. It passed over the junction and onto the single-track section towards South Ruislip, which was still occupied by the freight train. The train had travelled about one mile (1.6 km) beyond Greenford by the time that the driver received the emergency radio message.

The investigation found that the driver of the passenger train did not react to the two signals at danger, for reasons which are not certain. It is possible that he had formed the impression that the train had been given clear signals through Greenford, because of his interpretation of the meaning of the signal preceding those that he passed at danger, and he had not been stopped by signals at Greenford in the recent past.

The Train Protection and Warning System (TPWS) was fitted to the train and to both the signals, but it did not intervene to apply the brakes of the train, as it was intended to do. This was because the on-train TPWS equipment had self-isolated when the driver prepared the train for departure from Paddington. The isolation of the equipment was indicated by a flashing light in the cab, but the driver still drove the train.

Although the signaller at Greenford wished to stop the train by sending an emergency call on the GSM-R radio system, he did not attempt to do so because the information presented by the radio equipment in the signal box suggested to him that any message he sent would not reach the train. Instead, he contacted Marylebone signal box, which was able to send a message to the train.

RAIB has made three recommendations. One is addressed to Chiltern Railways, and covers the need for a review of the company’s driver management processes. The other two, addressed to Network Rail, cover the configuration of the GSM-R radio system as it affects the ability of signallers to directly contact trains that are within their areas of control, and the training given to signallers in the use of the GSM-R system. RAIB has also identified two learning points: one for signallers, relating to the use of delayed clearance of signals to warn train drivers of the state of the line ahead, and the other for train operating companies, relating to the upgrading of on-train TPWS equipment.

To see the complete report and all recommendations, see:

http://www.raib.gov.uk/cms_resources.cfm?file=/141222_R292014_Greenford.pdf

Monday Accident & Lessons Learned: EJECTION OF WORKSTRING DURING BULL-HEADING OPERATIONS

January 5th, 2015 by

IOGP Safety Alert # 264 – EJECTION OF WORKSTRING DURING BULL-HEADING OPERATIONS

WHAT HAPPENED

The rig was involved in well preparation for the forthcoming execution of a gravel pack operation. After earlier setting a Gravel Pack packer on the well, later it became impossible to circulate in direct (possibly plugged ports in string): only possible to circulate in reverse. This made Gravel Pack operation impossible. Decision was made to POOH wet to surface.

The ongoing operation at time of event was POOH with 4 ½ PH6 tubing work string – with Setting Tool and 2 3/8” tubing tail pipe – to inspect Service Tool for indications of inability to circulate.

  • Fluid in well was 1.07sg Brine.
  • Tubing string was being pulled wet.Losses observed at approximately 1.2m3/hr.
  • Tubing work string was 220m from surface at time of incident. Fluid returns were bubbling at the bell nipple (swabbing while pooh workstring-no gain yet). Crew estimated off-bottom kick situation. Closed annular to bullhead the well to push any gas influx into formation.

Operations details after calling town (drilling superintendent):

  1. Pump 5m3 High Viscosity pill bullheading same at a flow rate not exceeding 800 lpm and at a stand pipe pressure not exceeding 1400 psi.
  2. Pumps were stopped and lined up with active pit, then approximately 7m3 1.07 sg brine was bullheaded. First, the flow rate was maintained around 400 lpm and stand pipe pressure remained below 1500 psi.
  3. Increased gradually the flow rate (to 1100 lpm) and the stand pipe pressure increased regularly up to 2300 psi.
  4. The Tool pusher called the driller to ask him why he was pumping at a stand pipe pressure higher than 2000 psi – the driller replied the Company Man told him 2500 psi was the limit.
  5. 220m+/- of 4 ½” tubing and 2 7/8” wash pipes along with the Service Tool Assembly were ejected from the well. Strong noise and vibrations were noted by several witnesses while the pipe was being ejected.
  6. The pipe hit the Elevator and was deviated outside the mast, hitting the monkey board and went up over the Monkey Board and out through the derrick space.
  7. The projectile started to drop down in an X-shape. It eventually fell on the ground, missing the company man’s and the service companies’ cabins by less than 1 meter,
  8. The Rig HSE/Meeting Room Cabin was destroyed.
  9. Damage was made to the concrete block fence wall inside the cluster situated just behind the Company Man Office.
  10. From Company Man Office to Well Center the Tubing string landing point was measured at 60m.
  11. The elapsed time of the entire event according to witnesses, took less than 30 sec (faster than a man running down the stairs from the rig floor).
  12. Driller remained on Rig Floor during ejection. He immediately stopped the rig pumps then went to the BOP panel. He did not close the pipe rams as he had little hope this would be of any help. He decided to close the Blind Shear Rams. Then he left the rig floor and went to the main muster point.
  13. The Company Man and Tool Pusher both ran to the remote control panel. The Tool Pusher arrived first and activated the Blind Shear Rams, then realized that the driller had already done it.
  14. All personnel proceeded to the Muster Points.
Immediately after the incident…
  1. The Head Count was 100% – no one had been injured.
  2. The Driller went back to the Rig Floor and resumed the Bullheading Operation against the Blind Shear Rams.
  3. The day after the incident (Saturday 9 November), piece of tubular was found at 105m from the well outside the cluster.

WHAT WENT WRONG?

  • Gravel Pack string was plugged creating inability to circulate direct and thus to execute the Gravel Pack program.
  • According to witnesses, squeeze pressure range was given by Company man to the Driller as 2000 – 2500 psi.
  • No formal risk assessment performed for Bull Heading operation for this Work-over operation.
  • No recalculation was done for this emergency non-routine Bull Heading well control operation either on site or at base by Company and Drilling Contractor.
  • Nobody on site neither in office recognized Well Control situation as critical.
  • Top drive was disconnected and drill pipe closed on TIW valve. 220m assembly closed on annular.
  • The Tripping Fast Shut In Procedure does not mention pipe rams closure.
  • There is no Bull Heading procedure available.

CORRECTIVE ACTIONS & RECOMMENDATIONS

  • Lack of risk awareness concerning contained Pressure hazards. Upward push on the assembly while bullheading not anticipated despite gravel pack operations experience on site.
  • No Job risk Analysis performed prior bullheading. No consequential analysis.
  • Bullheading operation considered as routine.
  • Due to underestimation of the criticality of the operation ongoing (Bull Heading) by all (site and base): Identify all critical operations/tasks including Bull Heading in drilling operations.
  • Prioritize Well Control and Critical Operations by base team at every point of Program execution.
  • Dedicated ‘critical activities’ section to be added in all operational programs (drilling, Workover, rig less) describing how to recognize a well control situation and which procedure to be applied. Each procedure completed by a Risk Assessment made available to the work crew which they can use as a basis for further review.
  • Training plan to be immediately set up for full IWCF compliance of the Completion and Well Intervention personnel.

Safety Alert Number: 264
IOGP Safety Alerts http://safetyzone.iogp.org/

DISCLAIMER

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.

Keeping TapRooT® Investigations Out of Court

December 18th, 2014 by

NewImage

We would all agree that performing accident investigations and investigations of quality issues to prevent repeat accidents is a good idea. But some may be reluctant to perform investigations because of the legal liability they think the investigation report may represent.

Of course, they are at least partially right. Frequently, significant accidents result in lawsuits. And if your investigators aren’t careful, they may put poorly chosen words or even un-true statements in their investigation reports. Thus, company counsel or outside counsel may prefer that the actual accident investigation reports be excluded from evidence in a court proceeding to reduce the liability that an accident investigation report may represent.

Excluding a report performed after an accident to look for ways to prevent future accidents is a protected activity under federal law. FRE Rule 407, Subsequent Remedial Measures, states:

When measures are taken that would have made an earlier injury or harm less likely to occur, evidence of the subsequent measures is not admissible to prove:

  • negligence;
  • culpable conduct;
  • a defect in a product or its design; or
  • a need for a warning or instruction.

But the court may admit this evidence for another purpose, such as impeachment or — if disputed — proving ownership, control, or the feasibility of precautionary measures.

How can you help preserve your right to exclude your report from discovery and use at trial? Outside counsel for one of our clients has suggested that all TapRooT® users add one of the following preambles or appendices to every TapRooT® Investigation. We thought this sounded like a good idea, so we are passing along the following preambles or appendices for you to consider when writing your investigations….

For safety investigations at a company, the preamble suggested by the attorneys is as follows:

– – –

Note:

1. Substitute/insert the correct company name for “COMPANY” throughout, and

2. Add this preamble to every TapRooT® investigative report.

COMPANY TapRooT® Investigation Preamble

In order improve COMPANY’s overall safety performance and to prevent or significantly reduce the likelihood of the same or a similar work-related incident/injury/illness (“incident”) from reoccurring, COMPANY conducts a TapRooT® systematic investigative approach to incident investigation and analysis to solve significant performance problems and/or equipment failures that may arise from time to time during its operations. TapRooT® is an efficient and effective method that helps to identify best practices and/or missing knowledge related to an incident, which will allow COMPANY to execute/institute lasting fixes faster, thereby increasing reliability thru identification of remedial measures.  TapRooT® reveals root causes, causal factors, events, and/or conditions within COMPANY’s management control so that corrective action can be taken. Said more succinctly, root causes in TapRooT® are causes COMPANY management has control over.  The information generated during a TapRooT® investigation is essential to implementing an effective prevention program under the control of COMPANY’s management by using hindsight analysis of the incident to perform remedial measures.

TapRooT® is a system used to determine subsequent remedial measures COMPANY may take to improve future performance.  This investigation therefore is excluded from evidence under Federal Rule of Evidence 407 based on the policy of encouraging COMPANY to remedy hazardous conditions without fear that their actions will be used as evidence against them, that is, to encourage COMPANY to take, or at least not discourage them from taking steps in furtherance of added safety.

Incidents, injuries and illness may occur as a result of third parties’ conduct.  TapRooT® may not focus on the acts and/or omissions of third parties, contractors, subcontractors and/or vendors. Errors made by third-parties in design, repair, assembly, installation, construction, etc. are not the focus of TapRooT® inasmuch as COMPANY management has no control over errors made by these vendors except expected conformance with their duties owed to COMPANY.

Even though COMPANY makes every effort to determine what happened during an incident and to minimize future incidents through the COMPANY investigative team, TapRooT® is generated in hindsight and does not determine legal cause(s), “but for” causation, or proximate cause(s) of an incident. To infer this from a TapRooT® investigation would be a misuse of the TapRooT® analysis. Instead, TapRooT® determines events, conditions and causal factors in the root cause analysis. Each causal factor may have one or more root causes.  Any causal factors and/or recommendations which may be generated in a TapRooT® investigation are based on the investigator/ investigation team’s own views, observation, educated opinions, experience, and qualifications. TapRooT® identifies remedial measures to reduce the probability of events such as the one being investigated from happening in the future.  This information is not intended to replace the advice or opinion of outside COMPANY retained experts who may have more specialized knowledge in an area made the basis of this investigation. Equally important, while information gathered during a TapRooT® investigation is obtained from sources deemed reliable, the accuracy, completeness, reliability, or timeliness of the information is preliminary in nature until the final report is issued. Thus, the findings and/or conclusions of a TapRooT® investigation are subject to change based on information and data gathered during subsequent investigations by experts who may be more focused on legal causation, which is outside the scope of TapRooT®.

(1) FRE Rule 407. Subsequent Remedial Measures

When measures are taken that would have made an earlier injury or harm less likely to occur, evidence of the subsequent measures is not admissible to prove:

  • negligence;
  • culpable conduct;
  • a defect in a product or its design; or
  • a need for a warning or instruction.

But the court may admit this evidence for another purpose, such as impeachment or — if disputed — proving ownership, control, or the feasibility of precautionary measures.

– – –

For quality investigations subsequent to an issue with a product, the following preamble/appendix is suggested:

– – –

Note:

1. Substitute/insert the correct company name for “COMPANY” throughout, and

2. Add this preamble to every TapRooT® investigative report.

VENDOR TapRooT® Investigation Preamble

In order improve VENDOR’s overall quality performance and to prevent or significantly reduce the likelihood of the same or a similar quality issues from reoccurring which may lead to work-related incident/injury/illness or client related issues (“incident”), VENDOR conducts a TapRooT® systematic investigative approach to incident investigation and analysis to solve significant quality and/or performance problems and/or equipment failures that may arise from time to time during the use or manufacture of its products. TapRooT® is an efficient and effective method that helps to identify best practices and/or missing knowledge related to an incident, which will allow VENDOR to execute/institute lasting fixes faster, thereby increasing reliability thru identification of remedial measures.  TapRooT® reveals root causes, causal factors, events, and/or conditions within VENDOR’s management control so that corrective action can be taken. Said more succinctly, root causes in TapRooT® are causes VENDOR management has control over.  The information generated during a TapRooT® investigation is essential to implementing an effective prevention program under the control of VENDOR’s management by using hindsight analysis of the incident to perform remedial measures.

TapRooT® is a system used to determine subsequent remedial measures VENDOR may take to improve future performance.  This investigation therefore is excluded from evidence under Federal Rule of Evidence 407 based on the policy of encouraging VENDOR to remedy hazardous conditions without fear that their actions will be used as evidence against them, that is, to encourage VENDOR to take, or at least not discourage them from taking steps in furtherance of added safety and quality.

Incidents, injuries, illness, and quality issues may occur as a result of third parties’ conduct.  TapRooT® may not focus on the acts and/or omissions of third parties, contractors, subcontractors, vendors, and/or clients. Errors made by third-parties in design, repair, assembly, installation, construction, etc. are not the focus of TapRooT inasmuch as VENDOR management has no control over errors made by these third parties except expected conformance with their duties owed to VENDOR.

Even though VENDOR makes every effort to determine what happened during an incident and to minimize future incidents through the VENDOR investigative team, TapRooT® is generated in hindsight and does not determine legal cause(s), “but for” causation, or proximate cause(s) of an incident. To infer this from a TapRooT® investigation would be a misuse of the TapRooT® analysis. Instead, TapRooT determines events, conditions and causal factors in the root cause analysis. Each causal factor may have one or more root causes.  Any causal factors and/or recommendations which may be generated in a TapRooT® investigation are based on the investigator/ investigation team’s own views, observation, educated opinions, experience, and qualifications. TapRooT® identifies remedial measures to reduce the probability of events such as the one being investigated from happening in the future.  This information is not intended to replace the advice or opinion of outside VENDOR retained experts who may have more specialized knowledge in an area made the basis of this investigation. Equally important, while information gathered during a TapRooT® investigation is obtained from sources deemed reliable, the accuracy, completeness, reliability, or timeliness of the information is preliminary in nature until the final report is issued. Thus, the findings and/or conclusions of a TapRooT® investigation are subject to change based on information and data gathered during subsequent investigations by experts who may be more focused on legal causation, which is outside the scope of TapRooT®.

(1) FRE Rule 407. Subsequent Remedial Measures

When measures are taken that would have made an earlier injury or harm less likely to occur, evidence of the subsequent measures is not admissible to prove:

  • negligence;
  • culpable conduct;
  • a defect in a product or its design; or
  • a need for a warning or instruction.

But the court may admit this evidence for another purpose, such as impeachment or — if disputed — proving ownership, control, or the feasibility of precautionary measures.

– – –

Of course, before adopting any advice to reduce potential liabilities in future courtroom actions, you should consult your own in-house or outside counsel. They may modify the forms provided above or have other wording that they prefer.

So consider the advice provided above and get your own protective wording added to all your standard reports. We are looking at ways to add this to the TapRooT® Software and we’ll let you know when we’ve figured out a way to do it. Until then, we suggest manually adding the wording to your official final reports.

Connect with Us

Filter News

Search News

Authors

Barb PhillipsBarb Phillips

Editorial Director

Chris ValleeChris Vallee

Six Sigma

Dan VerlindeDan Verlinde

Software Development

Dave JanneyDave Janney

Safety & Quality

Ed SkompskiEd Skompski

Medical

Gabby MillerGabby Miller

Communications Specialist

Ken ReedKen Reed

Equifactor®

Linda UngerLinda Unger

Vice President

Mark ParadiesMark Paradies

Creator of TapRooT®

Steve RaycraftSteve Raycraft

Technical Support

Success Stories

An improvement plan was developed and implemented. Elements of the improvement plan included process…

Exelon Nuclear

Alaska Airlines adopted System Safety and incorporated TapRooT® into the process to find the root causes…

Alaska Airlines
Contact Us