Oddbean new post about | logout
▲ ▼
TimToolMan (TimToolManOG - TS) | 1 years ago (raw) | root | parent | reply | flag 
 @48bfa577 
Reliability as a primary goal is not focused on cost effectiveness. Cost effectiveness often happens when you get it right, but the measurement of it is difficult. If you have no unplanned outages, how do you prove cost-benifit on your maintenance cost? All of your cost of unplanned outages is theoretical- so the engineers don't apply much weight. So, maintenance budgets are cut, and profits go up..... for a while. Then things start to break.
▲ ▼
The Felon Pope :popephil: | 1 years ago (raw) | root | parent | reply | flag 
 @602e6fbe we use some hocus pocus to measure this stuff. To go from 90% reliability after 10 years to 99% costs more than going from 60% to 90%. So we evaluate that cost compared to the cost of down time and replacement for each item. If it's safety equipment no cost is considered, we go for maximum reliability (parts with failure rates of parts per million, but if it's something cheep and just used for advertising or of low consequence in the event of failure, we do parts per thousand.