FMEA, Failure Modes and Effects Analysis
FMEA is a hazard identification technique sometimes called Failure Mode and Effects Consequence Analysis (FMECA). In this description an FMEA is considered to be equivalent to an FMECA. Evaluation of identified hazards must be followed by subjective or objective form of consequence determination in order to adequately address those hazards that are the greatest risk. A corollary to this reasoning is that no company (client) is so well endowed with funds that they can spend lavishly on correction of every hazard regardless of the consequence. Therefore the use of these study names is interchangeable since consequence determination is needed to determine the relative risk of a failure or group of failures.
APPLICABILITY OF FMEA
The FMEA method examines the way equipment can fail or be improperly operated, thus producing a hazard or operability issue. Every line, piece of equipment, and operation is analyzed through the full scope of operations: Run, Idle, Startup, Shutdown and Emergency Shutdown. The level of the failure modes examined depends somewhat on the quality of the failure rate data supplied or generally available for the equipment or systems in the study. The FMEA seeks to discover the effects of inadequate control of any type energy, as do other hazard identification methods. The difference in FMEA is that failures are generally limited to a specific system or unit operation. Application of FMEA over an entire process or operation gets very time consuming and tends to lead the team toward becoming inefficient because there is so much to simultaneously consider.
IDENTIFICATION OF FAILURE MODES
All failure modes of each item are considered individually and in concert where necessary. In each instance the potential effects that can result from the type of failure examined are determined. Results are recorded in a tabular format with assignment of subjective frequency of failure and potential consequence to define the classic risk equation, which is the product of frequency and consequence. These products are then accumulated into a matrix, generally with up to five levels of frequency and five levels of consequence, with acceptable and unacceptable assignments for the higher risk tabulations. The team decides at the outset of the study what to consider an acceptable or unacceptable risk.
Each failure mode item examined is given a unique identifier for later reference. The level of resolution, as mentioned, is defined by the available description of the equipment and its potential failures. The team leader will have a compendium of failure modes for most types of equipment to a varying level of resolution, should the team come short of addressing the full scope of failure modes.
It is feasible to dissect systems down to the individual part level if data are available, but most often the team defines the failure mode in terms of loss of intended function of the system or subsystem involved. In general, effects are traced at least two operations upstream and downstream of the failed device or system to assure that at least second-step deduction has been pursued in identifying effects. In some cases the team goes even further up and downstream of the process or operation being addressed looking for hazards or operation error potentials.
EVALUATION OF RISK
The FMEA examines each item systematically line by line, piece by piece, to the desired level of resolution based upon the client’s direction. Identified risks can then be subjectively or objectively evaluated further than the initial assignment as desired. With subjective risk ranking there may be a tendency to overshoot the consequence and frequency estimate and skew the risk interpretation such that major risks may be ignored as acceptable or minor risks may be unnecessarily elevated to greater importance than is actual, leading to wasting funds correcting an acceptable risk. When major risks are relegated to minor risk status, willful violations of regulatory codes and unnecessary endangerment of employees and possibly the public can result. Given sufficient lead time, the facilitator can prepare what is commonly called a trail for the FMEA. A trail establishes a tentative order in which all failure modes on all systems and subsystems will be examined.
Outputs are as described and reported as mentioned in the PHA protocol.