Staffing Needs and Time - DOE-HDBK-1100-20040072

DOE-HDBK-1100-2004

events or basic events. The analysis starts with a review of system requirements,

function, design, environment, and other factors to determine the conditions, events, and

failures that could contribute to an occurrence of the undesired top event. The top event

is then defined in terms of sub-top events, i.e., events that describe the specific "whens

and wheres" of the hazard in the top event. Next, the analysts examine the sub-top events

and determine the immediate, necessary, and sufficient causes that result in each of these

events. Normally, these are not basic causes, but are intermediate faults that require

further development. For each intermediate fault, the causes are determined and shown

on the fault tree with the appropriate logic gate. The analysts follow this process until all

intermediate faults have been developed to their fault causes. The fault causes, or basic

events, include equipment failures, human response errors, and initiating events.

EVALUATING THE FAULT TREE.

After a fault tree is constructed, it can be input to a fault

tree analysis computer program, such as FTAP, IRRAS, or WAM. The output from the

computer program is a list of MCSs which cause the top event to occur. For each of the

MCSs, the analysts describe the consequences associated with that cut set. Table 4.25

shows a typical worksheet used to document the consequences associated with MCSs.

DOCUMENTING THE RESULTS.

A ranked list of MCSs for a system, along with the

consequence of each cut set, is the ultimate product of a qualitative FTA. Based on the

number and type of failures in the MCSs, the PrHA team may recommend improvements

to make the top event less likely. The fault tree model itself is often used as a

communication tool with both technical and nontechnical decision makers.

4.6.3 Staffing Needs and Time

Although the construction of fault trees is not typically done by team approach, to meet

the PSM Rule requirement, all members of a PrHA team should provide input during the

construction of fault trees. The PrHA team can meet in a room with a large chalkboard or

roll of paper and assign one person to draw the fault trees. The team can come to a

consensus on the type (AND, OR) and inputs for each fault-tree gate, and the gates can

then be added to the fault tree drawing. However, because FTA develops a model of a

system, it is fundamentally not a consensus method. If there is disagreement in the tree

construction, then it is likely that the process is not well understood.

Using FTA requires a detailed understanding of how a process or system functions,

detailed drawings and procedures, and knowledge of component failure modes and

effects. The team leader should be well trained and experienced in constructing fault

trees.