maintenance cost x reliability improving

[rafaelfaturini]

I work in a oil industry and would like to connect reliability and cost of maintenance. How can I make a link between increasing reliability and the cost associated with maintenance (Corrective x Preventive)?
How to improve the time of preventive maintenance (PM)?

As exemple:

Failure history (25 failures / 3 ) 2p-Weibull beta = 2.22 / eta= 1600

Cost PM= 2700.00
Cost CM= 8600.00

 

[ehbadger]

From my (admittedly limited) experience, it will probably be unique to each piece of equipment/process you are looking at. I'm not sure you could just create a generic formula and plug in numbers. I tend to do a "life cycle analysis" type of thing. It is all so dependent on what the maintenance is, what it costs, and what it gains you.

Some considerations:
If it fails, does the whole process or plant go down? What is the lost $ for every hour it is broken? How long does it take to fix?
How much is PM costing? What are you actually doing in the PM?

Regarding PM frequency:
- For some things (vessels, etc) a RBI (risk based inspection) program can help set the inspection intervals.
- Other times they are driven by code (i.e. relief valves need testing every X years).
- What are your failure modes? What PM or inspections can actually be done to catch/prevent issues and what do those cost? Example: if you have pitting corrosion in piping then doing a UT (thickness check) isn't going to necessarily help, even if you do it every single day.
- Do you have historical data? What is the MTBF? Do your PM at such an interval that you don't approach the majority of the distribution curve.

 

[EmmanuelTop]

You need to run a RAM study, and plug in your maintenance costs and see how increased reliability reflects on your facility performance (in terms of profit/revenue).
See some references:

https://library.e.abb.com/public/f774fb4c67beea60c1257419005abe30/74-78 SR114_72dpi.pdf

http://www.energy.alberta.ca/CCS/RAMFinalReport.pdf

http://www.gasprocessingnews.com/fe...pical-availability-values-for-lng-plants.aspx

Dejan IVANOVIC
Process Engineer, MSChE

 

[BigInch]

I think these two items are better left independent of each other as much as possible. The required level of component reliability should be determined based on the component's criticality to mission success. Cost of whatever systems are required to meet mission success criteria should be the result of the cost of the items required. To do otherwise defines, or even worse, compromises the mission success criteria against cost. Is that what you really wanted to do? Does your mission criteria depend on cost??? If it's too expensive, do you change your success criteria? That can become dangerously subjective. If it's too expensive, do you accept possible loss of life? Damage to property? How much damage is acceptable? Where, and how do you draw that line?

 

[IRstuff]

But, it's a purely economic calculation:

cost of current repairs vs. non-recurring cost for redesign + new cost of repairs

If the former is greater by a sufficient margin, over life, then you can argue for making the change. If the latter is greater, then you stay with the current situation. Longevity also factors into the mix.

TTFN
I can do absolutely anything. I'm an expert!
homework forum: //

http://www.engineering.com/AskForum/aff/32.aspx

faq731-376 forum1529

 

[EmmanuelTop]

Big Inch is spot on. To expand my answer a bit - a properly executed RAM study with accurately defined input data (including downtime due to failure of critical components) will provide you with quantified results, in terms of how much each component contributes to overall reliability/availability of the plant.

Dejan IVANOVIC
Process Engineer, MSChE

 

[BigInch]

IR, I'm saying that if you define your mission criticality level, cost of extra components and maintenance does not become a chequebook balancing act; the cost being the direct result of meeting the mission's reliabililty criteria required for some specified probability of success. If that presents too much of an economic burden, redefine the mission success criteria. That way everyone can support the results, rather than argue about how much affect varying the degree of success probability happens to cost today verses the worth of the criteria. That cost might be different next year, or tomorrow.
I am not a fan of economic based decision maintainence criteria. You need maintenance to keep something running at an acceptable level. Define that level and you know how much maintenance is required then the cost falls out of that. I don't like arguing with the accountants. My advantage is that they can't argue design and maintenance criteria, so they are left screaming only about how much is in the bank. Plus they take the hit if, God forbid, something happens.

 

[IRstuff]

Sure, that's fine if you can get that sort of environment. But, there's always a desire to minimize cost, regardless of what was agreed upon at the start of the program, particularly when technology and components are constantly improving. In military applications, it's rare that the outyear costs have been fully calculated and allocated, so any potential reduction in maintenance costs 5 or 10 years down the line is ripe for a value-engineering change. Obsolescence costs are always a big deal, particularly in today's environment, where a processor board that was state of the art 2 yrs ago is now obsolete and the supplier wants 2x or 3x premium to keep building the board. Another scenario is where the reliability of the system as a whole just didn't live up to the predictions, either because of poor data or more severe operating environments and the maintenance costs are substantially higher than originally estimated.

TTFN
I can do absolutely anything. I'm an expert!
homework forum: //

http://www.engineering.com/AskForum/aff/32.aspx

faq731-376 forum1529

 

[IRstuff]

Another case is where the system is well past its originally specified operating life. The US Army is still using a particular sensor on a particular helicopter that was developed in the 1980s, which was supposed to have been replaced by a newer system in the late 1990s, but that program literally and figuratively crashed and burned, so a series of obsolescence upgrades were initiated that kept the system operating even now.

A similar example is the B-52 nuclear fleet, where planes that were originally built in the 1940s and 1950s are still flying, with pretty much every component having been replaced at least once. Since the electronics and guidance were extremely antiquated, by today's standards, there have also been serious upgrades for GPS, etc.

TTFN
I can do absolutely anything. I'm an expert!
homework forum: //

http://www.engineering.com/AskForum/aff/32.aspx

faq731-376 forum1529

 

[BigInch]

I don't disagree there. It's true for the oil industry as well. Actually especially in the case now at hand where it's not so much the cost of components having changed, but rather the reduced revenue of the product sold that has affected the ability to pay for them. Previous levels of economically acceptable reliability standards may no longer be economically possible. Nothing more than the cost of safety and reliability required by regulations will be the target.