Second, the technicians
tell him that within the two low level performance
states, not 67% but only 60% is generated
due to bad site conditions (humidity, air
pressure, etc.). Because of these model
refinements the expected energy availability
falls to 96.3%.
Third, they tell
him that besides the unplanned repair events,
2 weeks of planned maintenance per year
(which is 3.8%) can not be neglected. Within
this period, the total system is totally
down and does not produce any energy at
all. Therefore, the expected energy availability
shrinks to (96.3%*(100%-3.8%)) = 92.6%.
It might be argued that only under the condition
of best maintenance practices, the assumption
of exponentially distributed failure-, and
repair rates is fair and might be applied
for our model. We therefore immediately
adjust slightly our Petri-net model to what
the technicians told us to be observed in
reality (even if constant rates were not
reasonable, we would not have any difficulty
in finetuning the Petri-net model). |
