IED Lifetime and Maintenance Strategies

What is the lifetime of an IED and what should we do about it after that period?

Now there is a question with a million answers!!
We have the same question for electromechanical relays too. We would generally have said that they would have been made with a life expectancy of >20 years but there are thousands of substation worldwide where the protection is as old as the switchgear which is 50-60 years old and everything is still working fine.
Similarly there are first generation electronic relays (diodes and transistors) from the mid 1970's that are still in service today.
and the first generation microprocessor relays from the early 1980's also still in service.

That all kind of proves that "lifetime" numbers are not necessarily physical "drop dead" estimates.

The first thing to consider, which thankfully I don't think has had to be tested in court –is what would the insurance company lawyers say in response to a blackout if there is a device that caused the blackout and it was beyond its published "lifetime" - they may well argue you have been irresponsible in leaving it in service and that is why they won't pay out the insurance. !! (there is a famous 660 MW generator failure in Australia where insurance didn’t pay out due to lack of service in accordance with the manufacturers guidelines on the machine)

But let’s keep lawyers in mind, but not the focus of this discussion.

CIGRE Technical Brochure 246 "The automation of new and existing substations: why and how" says in its second paragraph - yes literally the second paragraph setting the whole basis of all that follows:
"Primary equipment has an average lifetime of approximately 40 years and secondary equipment such as protection, control or communication equipment approximately 20 years. Consequently, the secondary equipment has to be refurbished once during the lifetime of the substation."
So good industry practice guidelines would say that is what you should be doing (and is fuel for the lawyer argument above).

In reality most utilities operating under a regulatory regime will use 15 year life estimates instead of 20 years in their accounting write off models.
So from an accounting perspective you are expected to be replacing things on a 15 year cycle. Equally arguably it is an expectation of the shareholders (or public stakeholders if government owned) to keep their assets that they have invested in are up to date, serviceable and valuable as a capital value.

What that means is for a utility with 90 substations they must be replacing their secondary systems at a rate of 6 substations per year to complete the 90 every 15 years.
Or a utility with 300 substations, completely replacing 20 per year!

Few utilities are able to achieve even close to these numbers - engineering perspective and/or and money perspective.
This is really the reason we need to adopt IEC 61850 to achieve Reliable Reusable Engineering to reduce the effort and cost of SAS refurbishment - a paper by ABB at "Distribution 2001" in Brisbane "“The Impact of the coming Standard IEC 61850 on the Life-cycle of Open Communication Systems” indicated that IEC 61850 would reduce the costs of refurbishments from 75% of the cost of a new substation down to just 25%!!

CIGRE also produced Technical Brochure 448 "Refurbishment Strategies based on Life Cycle Cost and Technical Constraints" which directly deals with these sorts of action plan strategies and justifications.
That should be read in conjunction with Technical Brochure 464 "Maintenance Strategies for Digital Substation Automation Systems"

No manufacturer will guarantee a life.

They can give estimates of MTBF which statistically they can calculate those - either from the individual component MTBF or more reliably from the in-service failures recorded - and may give results of over 100 years but that is not really the expected life and certainly not what they guarantee. It just indicates that you could reasonably expect a good 15 - 20 years out of the IEDs in general but individual failures may occur much less than that.

In some cases life may be much more definitely defined as we know things like electrolytic capacitors - often used in power supplies - will loose the electrolyte over time - perhaps 10-15 years - so some specifications preclude them from being used ... but checking on that (or other time dependant devices) and enforcing it can be difficult.

The question is not so much about individual failures but more as when should you pre-empt massive failures or unsuitability and replace the IEDs just for the sake of it. As I said, that may involve some insurance/legal risk assessments.

There is an argument that justifies longer intervals between routine testing of IEDs that have better forms of intelligence that self-monitors operation.

That doesn't necessarily mean any one IED, or a whole set of IEDs will last longer - just that you will likely know when failures are occurring.

The risk assessment still comes into play.

If a manufacturer has said something along the lines that the life is typically 20 years but not guaranteed, it doesn't really help at all other than a possible lawyers view as mentioned above - if the manufacturer said 20 and it was still in service after 40 they would carve you up for meat .... so be careful what you ask the manufacturer to state in the first place - it places implied obligations on your replacement programs, resources and costs in 20 years time. Maybe its better not to ask??  No, that would be worse as an "ostrich with its head in the sand" and even more irresponsible!  We might not like the answer or the implications but it would be like buying a car and not planning for putting petrol in it, and occasionally replacing the oil, brakes and tyres.  Eventually there comes a point where whilst the car is working OK, there will be time for a decision to just replace the whole car, but defining what triggers that decision is very hard to do 5, 10 or 15 years in advance .

Remember that a single IED failure could result in massive blackout and potentially loss of life.
Protection is no game to trifle with reliability.
Factor that into your monitoring strategies, testing cycles and replacement programs!

On the other hand, what we are obligated to do even by definitions of an engineer that you can look up on the web, as engineers we must find solutions to real issues with practical and reasonable means given specified constraints. We are not required to build walls that will survive a monster tidal wave higher than Mt Everest or another great flood of Noah proportions.

CIGRE TBs are available from
www.e-cigre.org
staff of CIGRE member organisations or people who are individual members of CIGRE can download the PDF for free under strict copyright distribution restrictions (so please don't ask me to send you a copy) - in any case they are generally less than 150 Euro or so to buy as a non-Member.

Extra Notes: