Boiler Accident Reports: To Err Is Human
The first composite evaluation of this data, published in 1997, covered a five-year period from 1992 to 1996. With a benchmark in place, the 1997 analysis provided a particularly useful perspective on incident cause and effect. Five years later, it can be stated that the overriding conclusion reached from the second five-year study reinforces the findings of the first: namely, that human error remains the foremost cause of boiler and pressure vessel incidents in North America.
OVERVIEWTragically, a total of 127 people have lost their lives as the result of boiler and pressure vessel accidents during the past 10 years. On average, that is just less than 13 fatalities per year. The annual number of deaths has seesawed over the past 10 years (see Figure 1), with the only sustained downward trend — over three years — taking place between 1999 and 2001. While this may appear to be a positive revelation, it must be considered in the context that 1999 saw the most deaths (21) over the 10-year period. The lowest number of fatalities over the reporting period, 8, was recorded in 1994.
In the category of injuries, a total of 720 were recorded between 1992 and 2001 — an average of nearly 72 per year. Again, 1999 was not only the most deadly in the boiler and pressure vessel industry, it also saw the highest number of injuries with 136. By comparison, the year 2000 experienced the lowest number of injuries at 27.
When it comes to number of accidents, there is little positive news. Each year during the 1992 to 2001 reporting period saw at least 2,000 accidents, with a total of 23,338 accidents. That averaged 2,334 accidents per year. The highest number of accidents (2,686) occurred in 2000, while the lowest number (2,011) took place in 1998 (see Figure 2).
INJURY-PER-ACCIDENT RATIOWhile numbers may climb and fall each year, the one true measure of how the industry is faring can perhaps best be found in a statistic not officially included as part of the reporting system: the injury-per-accident ratio.
Since 1992, this ratio has ranged from one injury for every 99 accidents in 2000 (the safest year) to one injury for every 16 accidents in 1999 (the most dangerous). Last year’s ratio of one injury for every 26 accidents was the third worst year for safety during the 10 year reporting period. The average ratio of injuries to accidents for the 10-year period was one injury for every 32 accidents.
HUMAN ERROROf the 23,338 accidents recorded over the past 10 years, 83 percent were a direct result of human oversight or lack of knowledge (i.e., low-water condition, improper installation, improper repair, operator error or poor maintenance). Human oversight and lack of knowledge were also responsible for 69 percent of the injuries and 60 percent of recorded deaths.
As anyone who has followed these Incident Reports knows, low-water condition and operator error or poor maintenance have stood atop the list of boiler accident causes for all 10 years (includes power boilers, steam-heating boilers, and water-heating boilers). While low-water condition has been the predominant cause during this time period, operator error or poor maintenance has surpassed its causal counterpart just three times: in 1998, 1999, and 2000. (After this three-year hiatus, low-water condition regained its position as leading cause last year.)
Other major causes of boiler accidents reflect a mixed combination of human oversight and mechanical breakdown. In five of the 10 years, burner failure was the third leading cause of incidents (1992, 1994, 1995, 1996, and 1997), followed by limit controls in 1993 and 2000. Improper installation was the third leading cause in 1998, with improper repair coming in third in 1999 (see Figure 3).
In what comes as a surprise to many, the combined third leading cause for both boiler and pressure vessel accidents last year (2001) was unknown/undetermined — a category introduced as part of the Incident Report in 1999.
With unknown/undetermined accident causes exceeding 7 percent in 2001, National Board Executive Director Donald Tanner commented, “What we don’t know can hurt us. While being able to identify and isolate a problem may not necessarily give us complete comfort, it does provide certainty — the knowledge of understanding what needs to be corrected.”
DANGEROUS EQUIPMENTWhen it comes to equipment categories, unfired pressure vessels proved by far to be the deadliest. During the 10-year reporting period, a total of 64 persons were killed by unfired pressure vessels, followed by power boilers (44 fatalities), water-heating boilers (14), and steam-heating boilers (5).
Unfired pressure vessels were also the leading cause of injuries (289), followed again by power boilers (250), water-heating boilers (92), and steam-heating boilers (89).
In a peculiar twist, however, the above listing is reversed when it comes to total number of accidents over the 10-year period, with steam-heating boilers causing the most overall with 9,588 incidents, followed by water-heating boilers (6,928), power boilers (4,311), and unfired pressure vessels (2,511).
Indeed, the yearly breakdown finds steam-heating boilers causing the most incidents in seven of the last 10 years, while unfired pressure vessels recorded the fewest incidents each year (see Figure 4).
BRIEFLY NOTEDAdditional observations from the 10-year reporting period reveal:
Commenting on the results of the 10-year Incident Report comparison, the National Board’s executive director emphasizes the need to keep all statistical information in proper perspective.
“Since establishing the National Board Violation Findings program two years ago,” Mr. Tanner observes, “we have been able to track nearly 100,000 boiler and pressure vessel inspection violations. Had these violations not been identified and corrected, our Incident Reports may have reflected numbers of a more catastrophic nature.”
Reprinted with permission of The National Board of Boiler and Pressure Vessel Inspectors, National Board Bulletin, Summer 2002. For more information, visit www.nationalboard.org (website).
Publication date: 11/11/2002