RELIABILITY - The probability that a device will perform as designed.
For example if 90% of a certain radio last their design life of say 5 years. Then their reliabiltiy is 90%
AVAILABILITY - The number of items that will function as designed for
For example - Suppose you and nine of your neighbors start out for work on monday, but only 8 of you make it because of mechanical problems, accidents, speeding tickets/etc. Then the availability for work is 80%.
Actually it is a little more complex than this and there are equatiosn to define each one, but these are the basic basics.
Is there such relationship between reliability and availability? or is it totally 2 different subject?
Which is more reliable? To have 2 X 100% pumps (1 running & 1 spare) or 3 X 50% pumps (2 running & 1 spare)? Is the question relevant?
In terms of availability which one has higher availability i.e. 2 X 100% pumps or 3 X 50% pumps?
I used to hear that, in order to increase the reliability of a particular pump system we should diversify its source of power. For example having 2 motor-driven pumps + 1 steam turbine driven pump is more reliable compared to having 3 motor-driven pumps...? Is my logic right?
Below is the definition that I've extracted from our company's technical standards:
The reliability of a gas turbine is defined as the sum of the service hours (i.e. the period in which the unit is producing useful energy)divided by the sum of service hours plus forced outage hours, and is expressed as a percentage.
Service Hours : SH; Plan Outage Hours : POH; Forced Outage Hours : FOH
Total Hours per year = SH + POH + FOH = 8760 hrs
Reliability = SH / (SH+FOH)
The availability of a gas turbine is defined as the sum of the service hours (i.e. the period in which the unit is producing useful energy)divided by the sum of service hours plus total outage hours, and is expressed as a percentage.
Availability = SH / (SH+FOH+POH)
Based on the above definition, Availability will always be lower than Reliability...confused???
NORZUL: Reliability and availability are two different animals, evin if they use some of the same information.
Your pump question is frequently posed. Let me pose it in a different form as in the case of a twin light engine airplane. Many people say a twin is more reliable than a single engine because there are two engines. In fact the additional engine decreases reliability because it adds more parts and two the twin engine airplane is supposed to fly on two engines. No pilot in his right mind would try to fly a twin on one engine to save fuel or other inane reason. There are times that twins have an engine failure and the pilot has to fly on one engine, but it takes skill and some practice.
The NUMBERS of availability are lower than reliability in your case, but Reliability and Availability are two different concepts. Your availability definition is a little restricted comapred to what we use. Our equation is:
Sum of operating and standby hours divided by the sum of operating and standby hours plus sccheduled and unscheduled maintenance hours plus administrative and logistics delay hours.
To continue your pump question, you have to look at the system and the reliability and what failure means to you. Let us say that each pump has a reliability of 90%. If one pump is needed and there is a spare then you have 100% redundency, but 90% reliability, since only one is needed. Assuming the second pump comes on when needed.
For the three pumps the reliability of the system is 90% times 90% or 81% since both pumps are required. reliable. The third pump increases the reliability from 81% to 90%, but it really gets tricky because if you have a pump failure and the standby pump comes online then you should immediately replace the broken pump to retore the system relaibilty. If both pumps fail togethethe standby pump cannot supply the needed flow.
Regards
Dave
halldp@efv.usmc.mil
If you do not mind send me your phone number and we can talk.
IRSTUFF: Our equations are similar, and certainly different organizatins use different methodologies. It is the rational and the limitations that must be understood by all concerned. We crank in ADMIN dleay time which is usually small compared to other times.
Even time is an issue: Is it clock time, or tachometer time, or engine run time, or master switch time, or time based on 12 hour days or what. Different organizations have different basis that must be understood.
MTBF time, at least for military systems, is usually power-on time. A typical system might have 1000 hr MTBF, which is about 1.4 months, so if it's on 24/7, be prepared to make lots of repairs.
MTTR usually includes BIT and initialization time to verify failure and the successful repair, as well as any other associated troubleshooting time.
MLDT would be measured from the time you figure what part is needed to when you get a new one, ready to install.
From the equations, availability is derived from the reliability, but includes non-equipment factors, while the reliability is strictly a function of the equipment failure rate.
