COST PREDICTION

[jpclmsa]

I have to predict how much money will a building spend in the next 10 years with corrective (failure)tasks. I think that probably there is somewhere a table with mtbf (mean time between failure) coeficient.Does anyone now where can i find this information (e.g. how much will cost a chiller only in what corrective interventions its concerned)?

 

[IRstuff]

It's not rocket science, but the data is extremely difficult and cost to acquire. While MTBF numbers for a particular piece of hardware might be readily available, the impact of any particular failure is probably not obtainable.

In military programs, such equipment is divided into line-replaceable units, each with their individual MTBF's, so such an analysis will at least be at a reasonable level.

One possibility is if you can get a hold of a book similar to those that mechanics have for estimating repair costs on cars. This coupled with a reasonable swag at lower level MTBF's might get you there.

TTFN

 

[PacificSteve]

Means makes a great reference book related to operations and maintenance costs of a wide variety of equipment.

Trying to piece together all of your costs, repair types, precentive maintenance, etc. can be a herculean task, and can lead to widely varying results.

When in doubt, and as a rough reality check, use 7% of the installed equipment cost per year.

PacificSteve

 

[DRWeig]

Hi!

I agree with all of the posts above -- and I've also used some accounting info in the past that helped keep me out of hot water. I don't have it handy, so I'll set you to looking.

The IRS publishes a handbook for accountants to use in determining depreciation of fixed assets, and in that publication is a table that lists historical standards for maintenance and repair as a percentage of the installed cost or basis. It is apparently derived from a huge amount of data, so it's a good standard to cite as additional data to back up your estimates.

It's all crystal ball stuff anyway, but when the boss asks an engineer to predict a number and that number gets exceeded eight years later, it's good to have some solid statistics from a respectable source to help spread the blame. I'd give him a worst case scenario too, not just an average or "most probable" prediction. Maybe even draw the whole bell curve.

I don't remember exactly the different IRS figures for mechanical equipment but I think that PacificSteve is right there in the ballpark with his seven percent figure.

Good luck with your study, let us know how it turns out!

Old Dave