Accident Investigation

Ch.8.pdf

Home >Education homework help >Accident Investigation

CHAPTERS

Change Analysis

A change of some sort is a major factor in most accidents. Although change is a necessary component of progress, it can also be catastrophic. Change can be planned, anticipated, or desired; it can also be unintentional or unwanted. Change analysis is a technique for analyzing the changes that led up to an accident. It can be combined with other techniques or used independently.

Change analysis can be used both reactively and proactively. When it is used reactively, the investigator looks back at the events that led up to the accident and determines the unintentional or unwanted changes that may have caused it. \v'hat was different about how the process was performed this time that caused the accident to happen? When it is used proactively, safety professionals develop scenarios that introduce a planned, anticipated, or desired change into a 5Ystem and use change analysis techniques to identify potential hazards or accident situations that could arise because of the change. Change analysis can also be used to review processes or to identify the potential effects of changes before implementing a new procedure or process (Stephenson 1991) (Spear 2002). Many safety professionals use a modified change analysis process 10 comp[ ·th I · t i Y w1 the Process Safety Management regu atory reqw.remen s or managem f enc o change.

Part Ill: ,-l11aJy11· I r, h . ra re mqurs

About Change Anal . ys1s The technique of change analysis was first 1981) and was perfected by Chari K used before World W U.S. Air Force in the 1950s Billjesh epner and Benjarnu,y aru(Po,.. M . o nson I rego, I ,,

anagemem Oversight and Risk T a so used this tech . o, II, D ree (MOR1) lllqo, .

epanment of Energy. When using it th . . system deveJo""d lo II, f 1 . , e lnvesngator r- fo1 ,, o events eading up to an accide .th . . compares th '"' nt w1 a sunilar . e seo., ....

find the causes of the accident non-acc1den, sequ ,~.., . ~-

Although change analysis can be used with . it is most useful for accidents that h ~ny accident investigatio

appen while routine performed. For first-aid incidents, near misses and .d tasks are~ . I , aca ems that

sunp e tasks, change analysis may be sufficient on it oceurdoriog h d . will . s own to detennin .,,1... appene ; It provide enough information to deterrnin h ' • .., the sequence from developing again. e ow to P•ev~,

For more complex accidents, change analysis must be . . . _ structured ar.a

detailed. It 1s not enough for the rnvesagator to simply ask, "What chang.i to cause the usual sequence of events to turn into an accident sequence?" Every event that changed, including management systems events, mus, I, analyzed in detail, and frequently other analysis methods must be osed i, addition to change analysis.

The Change Analysis Approach

The approach of change analysis is to compare an accident situation " sequence with a similar accident-free situation or sequence. The basic ~ analysis sequence is illustrated in Exhibit 8.1 and a summary of compll!lOO

methods is illustrated in Exhibit 8.2.

Types of Accident-Free Situations f . tion or sequence to

It is critical to find a directly related accident- ree smia . be us,d: f s1tuauons can compare with the accident situation. Three types o

Chapter 8: Change Analysis

Analyze

Procedure as It Is Usually Performed 11,eSame . . . . able to compare the accident sequence with the ideally the ,nvesogator IS th

' d it was performed accident-free-last week, last mon , same proce ure as . . . S. accident did not occur dunng that struaaon, what was orlmyear. mcean . .

different in the sequence this time that may have comnbuted to the acadent.

A Description of the Job as It ls Supposed to Be Performed Comparing the accident sequence to a written description of how a task is supposed to be performed makes change analysis simple for the investigator. This process is also called Codes, Standards, and Regulations (CSR), Procedure Design Criteria Analysis, or Procedure Adherence Analysis. The investigator compares the accident sequence to the procedure, standard, or regulation ind discovers where change has occurred. One problem with comparing the accident sequence to a task description sequence is that the worker involved in ihe accident might not actually have changed the usual procedure. It is possible tluiworkers have neverperfonned the job as it is described by the standard or Procedure If this · th th • d . · " e case, en It was not a change but a dijfmnce from the ocnbed procedure that caused the accident.

Part I I I: Ana/ytiral Tuhniqun

Exhibit 8.2

1. ~n "'accident-free" Ho . : situation w the Job was perfor last year with no accid med last week, last rno

2. Plant procedures ents. ntti.or How the task was supposed t according to company policy~ be~

3• An "'ideal" situation How the job would be rf circumstances. pe ormed unde~

A Description of the Job as It Should Be P rli This sounds very much like the co . . e ormed . mpanson 111 the • ts s~btly different. In this kind of comparison, the ~r::~us Patagnph, bu111 accident sequence events not with written standards :ugat.or compucs die of the way the task would be performed und .d I u_t with a description hi h er I ea c1rcumstao With t s met od. the investigator is not comparing changes but di;t ! between the accident sequence and the ideal sequence, since the:;: probably never been performed in the ideal way. lf the investigator finds 1 be~er way to perform the procedure, it can be recommended as a correCU\·r acaon.

Making Thorough Comparisons J\s you perform change analysis, you must compare all events and conditioos involved in the accident with corresponding events and conditions in tht accident- free scenario. Ask the following guestions about each e\·ent u:a.t involved? lf:i"hat happened? IV'ht11 did ii happm? When did it buppui? H,, ,,; it happen? Analyze each set of events and conditions and ask what changes may have occurred: l s the time or place different? Are the people UJ\'oh-c-d different? On a managerial level, was it managed, controlled, reviewed, ci implemented differently? (DOE 1999.) After you have made' tho...,gh comparison find the differences between the two sequences. In most situations, they are guite obvious. Finally, determine how th:u differtllct

caused the accident.

Ch(Jpter 8: C/J(Jflge A11(J!Jf1S

. I w demonstrates the chwge analys is approach. 1be short e,xamplc be o

fi h'ng every day for cwenty years, and every day he

b had gone is l . Bo ' 1

aich fish for his lunch and dinner. One day, he dbeenabetoC had

001 catch any fish, and he wondered why. He compared tl1e

ful d , 10

all of the previous successful days and found unsuccess 3} . • th:it he had used the same fishing hole and the s~e hook size, and had fished

31 the same time of day. TI1e o nly difference was chat

he h:.1d used different bait on the unsuccess ful day, so he reasoned that using different bait was the cause of his failure to catch fish.

:\cadent in,·escig.itions will never be this easy to analyze, o f course, and th~ uu]ysis does not stop with finding that different bait was the reason for fishing f:ulu re. The next steps are lO learn why Bob decided to change his b 1 nand why the new bait didn't seem to attract fi sh. ln change analysis, you

must find out al/of the changes that occurred and analyze how they affected

the outcome.

Change Analysis Procedures

(lunge analysis is performed r I on a iour-co umn worksheet (see Exhibit 8.3) . • Column 1: Write the accident sequence.

: ~:umn 2: \Xhite the events of a comparable accident-free se9uencc o umn 3: Note the differences or cha b . l and column 2. , nges etween the events in column

, Column 4: Analyze the djfferenc . how they affected the outcome. es or changes In column 3 and detem1ine

Example Scenario \\tltn perfomun ch <ornpa g ange analysis , re to the accid , •) ou must find an accid f is a non- . ent se9uencc. Alth I ent- ree SC<...juence co

recurring task that has no sp~u~f,1 o u.r _forklift -and-ladder scenario ec1 ic wri tte n des . . c n ptton, there are

Part Ill · A · na!Jtica/ T. h . ec. 111q11es

Exhibit 8.3

ACCIDENT SEQUENCE MP

~---.c.:c::.::. __ ___:S~E~Q~U~E~NC~E~ 1. 1. DIFFERENCE

ANALYSIS 2.----i:__ __ _!.:__1.

========~2. ======2~~-====-+-- Describe

the accident sequence.

Describe a comparable

sequence from an accident-free

situation.

Identify the differences between

the accident sequence and

the comparison sequence for

each step.

Analyzet;;- differences and describe how they affected the accident

procedures for using ladders and working in warehouse aisles, and changing light bulbs is similar to installing signs, so that procedure was used as the comparison sequence. Exhibit 8.4 lists only the events in the accident sequence that are different from those in the comparison sequence. After the firstthrt< columns are filled in, analyze differences between the accident sequence and the comparison sequence that could be significant to the accident. The fourth . . fi . t on theacadenc column lists several differences that had a s1gru 1cant unpac

1s0F EXAMPL

ACCIDENT sEQUENCE

t 1. All jobs go 1. n,isjobdid no through a safety

obiain a safety and scheduling revieW, review. -- . 2_ The warehouse 2. Maintenance supervisor per· workers perform formed the task. maintenance

tasks.

DIFFERENCE

1. No safety or scheduling review was performed.

2. The proper personnel did not perform the job.

3. No communi- cation was made to

3. Communication 3. The work was

the night warehouse supervisor.

4. No rows were barricaded.

about work will not communi- go through the cated to supervisors. the night

supervisor.

4. All aisles and ends of rows are barricaded off.

4. Rows were not barricaded.

Chapter 8: Change A11afys1s

ANALYSIS

1. A safety review was not conducted; it would have initiated the proper procedures.

2. The workers who are familiar with the procedures did not perform the job.

3. The night supervisor did not know about the job and thus did not alert the forklift drivers to potential closed aisles. Lack of communication between supervisors seems common.

4. The forklift driver did not realize that the row was closed.

-------------C-ontinued on nexr page

;:. rrn<£o Pa,t 171 · . A11ajytica/ Tech . 111q11es

CHANG E ANALYSIS OF ACCIDENT EXAMPLE SC SEQUENCE COMPARISON

5. No cones were placed in the aisles.

6· The forklift driver could ~ot clearly see in front of the forklift.

? . The forklift driver was driving faster than the posted speed.

8. The supervisor was new to the job.

SEQUENCE

5. Cones are placed before the barricades to_alert forklift drivers that aisles are closed.

6. Forklift drivers are able to clearly see in front of the forklift.

7. Forklift drivers do not exceed posted speed.

8- Supervisors are experienced.

DIFFERENCE

6. The forklift was ~verloaded and it was difficult for the driver to see.

7· The forklift was going too fast.

8- The supervisor was inexper- ienced.

ANALYSIS

S. ~e forklift dnver Wasu to seein Sed pla~ed ~~:es :amtenance

partment when aisles were closed.

6. The forklift- driver was Jn a hurry and overloaded the forklift.

7· The forklift driver was rushing to get through the shift.

8. The supervisor was new to supervision and was used to getting things done instead of using the proper procedure.

Chapter 8: Cha11ge A11alysis

summary

,,,rysis is, simple technique for analyzing differences between the (l,Jl1l' ,cident sequence and the events 10 an accident-free comparison tfllSJJl~a . · . Changes are usually important factors 10 an accident sequence.

~ence. . . . · ,,

0 , change anal)~" companng the acodent sequence to a co mparable

;~dent-fr« sequence is valuable in finding abstract causes o f accidents. lnfo!111'cion discovered during change analysis will help prevent accidents "" ,ecurring-(Please see the Appendix for a sample Change Analysis form.)

REVIEW QUESTIONS

1.fof what types of accidents is change analysis most useful?

2. \\'1cltare the three types of accident-free situat io ns t hat ma b 1oanaccident situation? Y e compared

c ange analysis? ._1_wi._,_~_•P_,._,._,_•q;_u_ired:__;t.:_o~co:;;m:;:p;::l•:;t,:e,::a_::th'.:o:r~ou~g~h~~

9 1