semi合集-English.pdf - 第6983页
SEMI S10-1103 © SEMI 1996, 2003 13 R2-2.16 Work Space Analysi s — In this technique, the position of t he person performi ng a task is evaluated. All external and task related circumstances are evaluated and their risks …
SEMI S10-1103 © SEMI 1996, 2003 12
R2-2 Summary of Risk Assessment
Techniques
R2-2.1 Work Safety Analysis — A checklist is used to
check the working place on criteria such as
environment, ergonomics, and organization. On all the
criteria of the checklist a rating is given. The overview
of the ratings gives the “comfort” of the working place.
R2-2.2 MORT: Modified Fault Tree Analysis — Used
for detection of organization and policy errors being the
root causes of incidents. MORT uses a systematic
control list. MORT was developed as a safety
management program.
R2-2.3 Checklists — List of questions which is used as
guidance for the assessment. Almost every technique
uses a checklist. Checklists can be developed for
generic or specific situations (e.g., EN 1050 safety
checklist).
R2-2.4 Action Error Analysis — The consequences of
correct and incorrect actions are evaluated. Very much
based on handling of the person and the controls used.
R2-2.5 HAZOP — Hazard and Operability study:
Small team of experts of the machine/process which,
using a systematic checklist, checks for realistic
failures.
R2-2.6 FMEA — Failure Modes, Effects Analysis
(FMEA) is a tabulation of the system/plant equipment,
failure modes, and each failure mode' s effect on the
system/plant equipment. The failure mode is a
description of how equipment fails (e.g., open, closed,
on, off, leaks). The effect of the failure mode is the
system response or accident resulting from the
equipment failure. FMEA identifies single failure
modes that either directly result in or contribute
significantly to an important accident. Human/operator
errors are generally not examined in FMEA, although
the effects of a mis-operation are usually described by
an equipment failure mode. FMEA is not efficient for
identifying combinations of equipment failures that lead
to accidents. The FMEA can be performed by two
analysts or a multi-disciplinary team of professionals. A
Failure Modes, Effects and Criticality Analysis
(FMECA) is an FMEA with criticality rankings (Based
on MIL-STD-1629A).
R2-2.7 What-if — Possible failures are checked on
possible consequences. A checklist, suited for the
equipment to evaluate, may be used.
R2-2.8 Event Tree Analysis — Event tree analysis is a
technique for evaluating potential accident outcomes
resulting from a specific equipment failure or human
error known as an initiating event. Event tree analysis
considers operator response or safety system response
to the initiating event in determining the potential
accident outcomes. The results of the event tree
analysis are accident sequences, a chronological set of
failures or errors that define an accident. These results
describe the possible accident outcomes in terms of the
sequence of events (successes or failures of safety
functions) that follow an initiating event. Event tree
analysis is well suited for operations that have safety
systems or emergency procedures in place to respond to
specific initiating events.
NOTE R2-1: This technique is also known as Fault Hazard
Analysis.
R2-2.9 Fault Tree Analysis — Fault Tree Analysis
(FTA) focuses on one particular accident event and
provides a method for determining causes of that
accident event. The fault tree itself is a graphic model
that displays the various combinations of equipment
faults and failures that can result in the accident event.
The solution of the fault tree is a list of the sets of
equipment failures that are sufficient to result in the
accident event of interest. The strength of FTA as a
qualitative tool is its ability to break down an accident
into basic failures. This allows the safety analyst to
focus preventive measures on these basic causes to
reduce the probability of an accident.
R2-2.10 Circuit Logic Analysis — Logic of an
electrical circuit is checked by making a logic diagram.
Possible errors are reviewed for their consequences.
R2-2.11 Interface Analysis — Check how process,
equipment and systems are connected to each other.
Identify the risks of each connection.
R2-2.12 Mapping — The first step in determining a
risk profile is to develop a checklist of the areas of risk.
These factors can be evaluated quantitatively (as
accurate measurements) or qualitatively (as levels or
descriptive states). The second step is carrying out
sensitivity analyses for the quantitatively assessed
factors.
R2-2.13 Procedure Analysis — Procedures are
reviewed and checked if hazards occur in the steps or
sequence of steps.
R2-2.14 Contingency Analysis — Contingency
analysis is a method of treating uncertainty that
explores the effect on the alternatives of change in the
environment in which the alternatives are to function.
This is a “what-if” type of analysis, with the what-ifs
being external to the alternative.
R2-2.15 Mathematical Malfunction Analysis —
Identification of malfunctions, deviations in design are
made and the effect calculated. This technique is widely
used for rotating equipment.
SEMI S10-1103 © SEMI 1996, 2003 13
R2-2.16 Work Space Analysis — In this technique, the
position of the person performing a task is evaluated.
All external and task related circumstances are
evaluated and their risks are identified.
R2-2.17 Task Analysis — Task analysis is the
systematic examination of a task to identify all loss
exposures related to the task. A task will be reviewed in
the step-by step manner. Task Analysis should be done
during design and on final design. Task analysis is also
known as Job Safety Analysis (JSA). Verification of the
task analysis should be done by a Task Observation or
Job Safety Observation (JSO).
SEMI S10-1103 © SEMI 1996, 2003 14
RELATED INFORMATION 3
EXAMPLE OF RISK RANKING METHOD
NOTICE: This related information is not an official part of SEMI S10 and is not intended to modify or supersede
the official guideline. It has been derived from practical application by the task force members. Publication is
authorized by the vote of the responsible committee April 2, 2003.
R3-1 Example of Risk Ranking Method
R3-1.1 One method for risk analysis is the risk ranking
number (RRN) system. Numerical values are assigned
to descriptive phrases. A RRN is calculated, thus
indicating the risk.
R3-1.2 The RRN is a function of severity (S) and
probability of occurrence (PO).
R3-1.3 To get the RRN, use function (1) or
(2),
depending on which data and experiences are available.
RRN = S × PO (1)
S × FE × POH × PA (2)
Where: PO = FE × POH × PA
R3-1.4 Severity (S) — when assessing the severity of
harm, the following factors should be taken into
account:
• degree of possible harm/injury
• number of people at risk
• amount of property loss
• harm to environment
R3-1.5 Probability of Occurrence of Harm (PO) —
The probability of occurrence taken into account the
factors described in Sections R3-1.6-R3-1.8.
R3-1.6 Frequency and Duration of Exposure (FE) —
need for access to the danger zone, e.g., operator,
maintenance, service personnel
• number of persons exposed to hazard
• time spent of exposure to the danger zone, e.g.,
seconds, minutes
• frequency of access, e.g., annually, monthly, daily
R3-1.7 Probability of Occurrence of Hazardous
Situations (POH) —
• accident history
• history of damage to heath, property and/or
environment
• risk comparisons to other industries with
similarities, e.g., chemical industry, petroleum
industry
• statistical data
R3-1.8 Possibility to Avoid the Harm (PA)
R3-1.8.1 This is a number between 1 (not possible) to
0 (avoiding always possible)
• nature of persons, e.g., skilled or unskilled persons,
unmanned operation
• the speed of appearance of the hazards, e.g., slow,
fast
• awareness of critical situations, e.g., warning signs,
direct observation, information
• possibility to avoid the hazardous situations, e.g.,
reflex, possibility of escape
• knowledge and experience about the equipment
R3-1.9 Risk Ranking Number
R3-1.9.1 The risk-ranking number is a function of
severity (S) and probability of occurrence (PO).
R3-1.9.2 For these criteria, specific numbers need to be
defined. Based on the risk ranking appropriate actions
should be carried out according to the control strategy.