LMRA: Useful or useless?- update september 2022
sven naessens
Call me mister Grey, the practician. Reflections belong outside the box, solutions are best served as a compromise.
Update in Italic
Why does the industry continue to implement (new) safety activities?
To answer this question, one must understand the fundamentals of industrial safety that have been developed over the last 150 years.
Early pioneers (e.g. Heinrich) taught us that the outcome of an accident (e.g. fatality, loss of time, …) is not always predictable. In many cases, the outcome can be considered a question of bad luck. The accident itself (e.g. fall from height) is causal but not the outcome. Due to studies like “why accidents happen and how to avoid them?” (Calder, 1898; Eastman, 1910) safety thinking evolved.
H?m?l?inen et al. (2006) estimated that the number of workers worldwide (total employment equivalent) in 1998 was about 2,16 billion. The number of occupational fatalities in that period was 245,719. Meaning that 99.99% of the workers didn’t experience an accident with fatality. In the 19th century, some people argued that accidents were inevitable and that it was the price that society had to accept as part of industrialization. In industry, there is nowadays a real belief that we can prevent ALL accidents with fatalities. Many companies have therefore set up a “zero fatality program”.
Hoorelbeke (2021) describes the evolution of occupational health and safety since the industrial revolution. Accidents have always been rare events. This means that the number of people that do not have an accident is always several order of magnitudes higher than the number that will have an accident irrespective of the safety measures in place.
The safety journey is a never-ending story because of the ambition of the industry to reach zero (fatal) accidents. New ideas and methods are studied and when useful implemented. It is common practice in the industry to try to find some evidence of the effectiveness of a safety activity before deploying it throughout the organization.
The concept LMRA
The concept of LMRA is quite straightforward: “an ultimate check just before starting the work”. A typical example of an LMRA is “looking left and right before crossing the road. However, in industry, the concept has evolved and is deployed in many different ways. To enable the verification of the effectiveness of LMRA administrative support is often used. The workers can fill in a paper to answer some questions. Afterward, this paperwork is often used to continuously improve the LMRA activity. Or to return to the basics.
Management must be aware that completing the paperwork does not miss the mark.
LMRA analisys and implementation
It is impossible to use statistical analysis to demonstrate ab initio that LMRA is useful before putting it in place. The reason is that it is not possible to perform an equal probability sampling in the population. A typical number of works performed in a large refinery is about 100,000 to 200,000 per year. LMRA will be implemented to further decrease the number of accidents. But accidents are rare events; for a large refinery, the number of accidents can easily be less than 10 per year. This also means that the number of works for which LMRA will be effective is very small to the number of works for which it will have no added value.
Most of the time LMRA will first be deployed for a certain time on a pilot project. The results (i.e. decrease in accidents) will be measured. The second step will be deployed on a larger number of sites and when the second phase is convincing it will be implemented throughout the organization.
TotalEnergies: Evidence demonstration 1
TotalEnergies is present in 130 countries in the world and produces about 400 million manhours per year. Within the company, I am in charge of one of the reporting tools, meaning there is a central database in which people (employees and contractors) can enter information about accidents, near-miss, anomalies, observations, etc.).
The worldwide database contains at the current moment 1,474,7252 entries. Most of the entries are near-miss reports (556,455) and anomalies (327,098). I performed a data mining with the following keywords/phrases “LMRA, Last Minute Risk Analysis, and last check”. The search gave 4,714 results. This means that people from all over the world entered 4,714 times information about LMRA.
Important to understand:
??????????Fields that work with code tables: these entries are called data
??????????Fields that use free text: these entries are called information
Within TotalEnergies we don’t have any fields with a code table refering to LMRA. All results below needs therefore be considered as information written in statements made by our workers and/or contractors.
I started looking at the most recent reports. Hereunder are some examples of quotes in reports created this year. To be clear I have only read the first 20 or so:
1.??????“During the LMRA the scaffolder observed that it was impossible to determine the wind direction. The works were stopped.”
o??Actions undertaken: stopped work replaced the wind-direction flag.
2.??????“During the LMRA the executant did see that there was a requirement for a full-face shield (benzene contact)”
o??Actions undertook: stopped work and went to get a full-face shield
3.??????“During LMRA. Found precaution of the main permit for S/I and depressurized well was not described. Protection frame was not available on this platform.”
o??Actions undertook: stopped work, rescheduled until the issue was solved
4.??????“During the LMRA before executing the job found the spreader of the stepladder be broken due to corrosion”
o??Action undertook: stopped work
5.??????“During LMRA, before starting working overboard, when checking chuteline, noticed that chuteline was not working’
o??Action undertook: stopped work
It is obvious that all issues were found:
·????????during LMRA execution, no before or after.
·????????Because of LMRA execution, not because of anything else.
·????????Resulteted in workers stopping their activity.
The search, performed on 3 keywords/phrases, gave 4,714 results which are about 0.3% of all entries. Out of these 0.3% came five examples for this year, after reading only about 20 reports.
领英推荐
For me, these 5 examples are enough to highlight that we prevented five possible accidents because we used the concept LMRA.
TotalEnergies: Evidence demonstration 2
Survey basics
TotalEnergies held another survey and this was the basics:
Survey Results
Permit To Work: 78% of the respondents declare that more than 75% of the Permits To Work are accompanied by a Safety Green Light (LMRA within TotalEnergies).
Stop Work: 59% of the respondents declare that more than once, performing a Safety Green Light has resulted in not starting the work.
Red Light Situations: 42% of the ‘yes’ respondents have given examples of “Red Lights”: identification of a nearby hazard (electrical cable, pipe, product, other), co-activity, change management, knowledge of operating procedures, means of access, PPE, harnesses, tooling compliance, weather conditions
Risk Awareness: 95% of the respondents declare that the Safety Green Light improves Risk Awareness.
Survey Analysis
Strong points
Work points
Field Interviews Analysis
Strong points
Work points
Conclusion
o??Yet seen as an administrative routine process.
o??Examples collected mostly show generic/not relevant answers
o??Border?between?Safety Talk and Safety Green Light is not clear
o??Sporadically mentioned in the description of the Permits To Work process
o??Many different terminologies are used to describe the Safety Green Light (different content?)
o??Some LBU have customized the Safety Green Light
o??Quality of the Safety Green Light is not really addressed (depends on the experience of the participants)
References
Calder John, The Prevention of Factory Accidents: Being an Account of Manufacturing Industry and Accident, 1899
Eastman C, Work-Accident and the Law, New York, Volume 2 of the Pittsburgh Survey, Russell Sage Foundation Publications, 1910.
Hoorelbeke, Process Safety, an Engineering Discipline, De Gruyter, 2021
H?m?l?inen P, Takala J, Saarela K, Global estimates of occupational accidents. Safety Science, 44, 137–156, 2006.
Call me mister Grey, the practician. Reflections belong outside the box, solutions are best served as a compromise.
2 年Update with second evidence facts added in Italic after a survey conducted by TotalEnergies
Interpreter of practice
2 年Great to see that our research is sparking further examination of LMRA. No doubt there is a lot left to learn about this practice. You make some very relevant points, but one of your main conclusions seems to be that early detection of issues happens during LMRA, and therefore that LMRA help prevent accidents. Our research showed the workers were equally heedful and hazard oriented, regardless of whether there was a organisational requirement to complete a LMRA. This raises the question regarding your conclusions, whether workers would have also noticed the issues if the company had no LMRA requirement, and would have just described it differently in the reporting system without LMRA policies. This does not challenge whether it is useful for a person to do checks before starting work, but it does challenge whether it is useful to spend organisational resources on implementing LMRA through artefects and compliance monitoring. Resources that might be better spend elsewhere to improve safety.?
Maintenance Manager NLMK
2 年Regarding the analyse of our database from our digital LMRA-first platform following reaction: The remarks that we see online have a high quality (depend on the compagnies). We see even in the remarks: team behaviour, interconnections of corporation between subcontractors. The question in the LMRA form are also adapted of the area or situation. General result: fast and in dept safety follow up. additional al lot of operational issues are also reported, wat is an input for continue improvement programs. for info [email protected]