How to Efficiently Create a Reliable System
Given enough time and money, space and power, it would be pretty easy to build an almost 100% reliable system. Just have full redundancy and ample spare parts. In most case this is impractical due to money and in some cases impossible due to size or weight constraints. This article will define a means of analysis that will help define a reliable system efficiently. It consist of a few basic steps but is by no means simple to execute.
We’re going to combine two types of analysis that are used commonly by engineers. FMEA and Risk Management. For a full treatment of this technique see chapter 3 of “The Reliable Boat”
FMEA or failure modes and effects analysis was developed by Society of Automobile Engineers (SAE) Reliability engineers to allow them to analyze the effects caused by system element failures.
Risk Management is a project management technique which tries to see forward in time to determine what could happen to prevent a project from meeting its goals. It is used by many different project management teams from industry as well as the military.
Analyze the functionality of your system
What are the functions: functionality is usually defined in a hierarchy. Mobility is a function, loss of cooling to the engine is under mobility. It is therefore a fairly critical function.
Rate criticality are those functions -The following table will help you define criticality.
Table 1: Criticality Table
|Loss of redundancy||You might have 2 bilge pumps providing adequate functionality if one of them fails. Loss of redundancy is when one of them fails, leaving the system open to catastrophic failure if the last one fails|
|Degradation of operation||You can still perform the function but not as well as before the failure. For example, maybe your engine stays running and turns your props but it will stall when the RPM Is increased.|
|Loss of capability||If the function is no longer available, you have loss of capability. A capability might be navigation lighting. If the lights don’t work you have lost capability.|
|Loss of mobility||Mobility is a special function. Without it you are at the mercy of the wind and waves. Mobility is the main goal of your boat.|
|Loss of life||Self explanatory. For example, if you lose a person overboard there is a potential for loss of life… any risk of fire or flood is a potential for loss of life.|
How can the function degrade or fail
A failure might be that you bilge pump becomes stuck. If this is the only bilge pump then the failure is critical and could result in sinking and loss of life. If you have several bilge pumps then the issue is a much less critical loss of redundancy.
How likely is the degradation or failure
A bilge pump becoming stuck is fairly likely. I’ve had several bilge pumps fail. I would say highly likely.
Calculate risk factor
If a number can be assigned to the consequence and the likelihood, then an calculation can be made to determine the relative risk of a particular failure. The following tables depict the consequence and the likelihood with an associated number.
Table 2 Consequence of a Fault
|Loss of redundancy|
|Degradation of operation|
|Loss of capability|
|Loss of mobility|
|Loss of life|
After you assign a likelihood that the fault will occur and assign a consequence of the fault happening you can calculate the risk factor. Then you apply a threshold to define if the risk level is Low, Medium or High. Figure 1 shows the risk factor and risk level calculations. The calculation is as follows:
Risk Factor Calculation (places more emphasis on consequence and less on likelihood):
RF = 1-(1-C)L
Alternate Risk Factor Calculation (treats likelihood and consequence the same)
RF = .5 *(C+L)
C is the Consequence
L is the Likelihood
A Risk Level (RL) is then assigned based on the Risk Factor as follows:
Risk level is
Low if RF < .2
Medium if 0.4 > RF >= 0.2
High if RF >= 0.4
Risk level colors can be assigned as follows:
Red = High, Yellow = Medium (Moderate), and Green = Low
A depiction of the risk factor and risk level calculation can be seen in Figure 1. Each risk can be thought of as being in one of the colored squares.
Don’t get too hung up on exact values or precise meanings of the likelihood, consequence and risk factor. Remember this is a pseudo objective approach to analyzing the reliability of your boat. It allows you to determine what the greatest risks to your objectives are so you can properly apply resources. Keep applying commons sense. It should point you to the systems and components that need to be addressed and the relative priority that they should be addressed.
To see an example of this being done in a spreadsheet see failure_modes
For more reading on the subject please see “The Reliable Boat”.