# Expected service life, and replacement justifications

B

#### Bob Welker

Are there any generally accepted rule-of-thumb life expectancy numbers for common types of control equipment?

Does anyone wish to share any apocryphal rules they've developed from experience(for instance, if such-and-such a device has been installed and
running for x number of years then it's edging into the far end bathtub curve)?

I don't need precise numbers ... +/- 25% of the average life would be OK.

I know the rule-of-thumb approximation of the Arrhenius equation (reduce life expectancy by 1/2 for every 10 degrees C temperature rise).

Are there any others I should be aware of?

H

#### Harith

All control equipments are basically the sum of their components, as such if you refer to any system and check each individual component and
find the week point in the design you can estimate the failure rate to an exact number. To the best of my experience the week link in most
of today^Òs control devices is the power supply and in particularly the capacitance in them, which under normal circumstances do not have a
life cycle of more then 10 years. We have experienced an EXP failure rate after the first 9 years of operation on our power supplies, which
required us to replace all of our supplied, to avoid that failure.

B

#### Bob Peterson

> All control equipments are basically the sum of their components, as such
> if you refer to any system and check each individual component and find
> the week point in the design you can estimate the failure rate to an
> exact number. To the best of my experience the week link in most of
> today=E2=80=99s control devices is the power supply and in particularly the
> capacitance in them, which under normal circumstances do not have a life
> cycle of more then 10 years. We have experienced an EXP failure rate
> after the first 9 years of operation on our power supplies, which
> required us to replace all of our supplied, to avoid that failure.

In general, I would tend to agree with this, however.

Ten years is a long time in the life cycle of modern control components. Think about the stuff that you bought ten years ago versus whats available today to replace it.

My philosophy is anything over 7 years old is probably a candidate for replacement, rather then fixing. Why repair something that is obsolete?
Its still obsolete.

A lot of control equipment, particularly instruments, lend themselves well to this philosophy, as it is often less costly to just replace them with new devices rather then try to repair them as they fail.

This philosophy kind of falls apart with PLC equipment as its relatively easy to find a few hundred dollars in the maint/repair budget to replace a single transmitter, but not so easy to find \$10,000 to replace an obsolete PLC and
I/O. In these cases, sometimes it is better to try and replace a single I/O card that fails, or a power supply, but at some point the stuff you are trying to replace is no longer even available from the manufacturer (or is obscenely priced), so you end up having to go get used stuff off the
secondary market. At that point, its a solid candidate for replacement.

Many plants are wise to this problem and include money in their budgets to replace aging PLCs. I know of one plant that replaced dozens of PLC3s with PLC5s over about 5 years. Along the way they replaced hundreds of obsolete 1771 I/O racks, analog cards, and power supplys. They knew that somewhere along the way these things were going to fail and they felt it was better to replace them during planned shutdowns rather then waiitng for one of them to fail during production time when it really costs \$\$\$.

Another closely related issue is that after a long period of time, the people who trained on these obsolete items have moved on, or forgotten
what they used to know, so when they do fail, its difficult to troubleshoot them, and getting help from the manufacturer on these obsolete things is not all that easy, since they are in the same boat.

Bob Peterson