EBME Forums Forums Maintenance Technical Assistance When do you decide a unit must be switched out?

Obviously (and normally), the residual value of a given piece of kit falls over time, just as accumulated maintenance expenditure increases.*

But we need to be clear (careful) what we are talking about. For instance:-

1) "Book" value (annual mark-down due to depreciation)
2) Current market value (estimated)
3) Actual market value (what you got when you sold it) ... etc.

Meanwhile, depreciation of 7% per annum brings the residual (book) value of an item to 70% of its original cost in about five years.

And (as I may have mentioned before) I know of places where they use the annual write-down figure (of 7%) as a judgement about whether it is worth working on kit when it is brought in for repair. That is, has the residual value fallen to less than 70% of the cost of buying a new one? But note:- a new one, not the original cost, but usually - invariably - higher; bringing our "expected life-time" to (a little) more than five years, most likely.

But if we are saying that our budget for supporting a piece of kit over its complete life-time is 70% of its original cost, then if we actually spend nothing at all on it, but factor-in depreciation only (again at 7% p.a.) then we reach that stage just beyond sixteen years. I hope my maths is correct here! LOL.

But again, should we be talking in terms of original cost ... or current cost of replacement? And so it goes on ...

Meanwhile, it has been clear (?) to me for some time that not only are equipment "costs" and equipment "dates" (generally, the onward march of time) related, but there must be some "universal formula" connecting them. Including, of course, my own favourites:- PM dates. Date of last PM, next PM and so on.

I haven't worked out (stumbled across, more likely) that indisputable piece of logic (algorithm, function ... whatever) yet - but maybe I will one day.

* Does the "cross-over" point have any significance?

OK, I realise that all of the above is a bit ... er, academic.

But stuff like:-

1) It's been lost (or stolen)
2) The surgeon refuses to use it
3) The new lab tech can't get consistent results
4) The obstetrician prefers another manufacturer
5) The nurses don't trust the results
6) They've bought a new one

... and all the rest doesn't really translate that well into computer logic.

Last night I sat down and knocked up a spreadsheet (OK, sad I know); I wanted to punch in some "real world" figures.

It seems to me that the question of "when to retire equipment" (in planning terms, at least) can be answered by keeping tabs on three inputs (regardless of purchase price):-

1) Depreciation %pa
2) Price inflation %pa
3) Maintenance budget %pa

1) Is whatever the Bean Counters reckon it to be. 7% year-on-year? That is, the "book value" of the equipment reduces by 7% of last year's value (the computer can render this as a day-on-day reduction in value).

2) The expected price increases over time for a like-for-like replacement. We could could use government figures; but let's try 3% each year (again, the computer easily gives us a daily amount).

3) The anticipated annual maintenance budget, set during commissioning as a percentage of the purchase price. To my mind this is where things get interesting ...

24 years ago I wrote (in a "paper"):-


Is 10% a valid figure ... or should it be 5%? Either way, I reckon it should fall somewhere within those limits. So, for the time being let's go for 7.5% (of the initial purchase cost) per annum as our maintenance budget.

Before going for the bottom line we need to ponder our "replacement trigger point". OK, a forth factor then:-

4) Trigger %

Previously we were talking about 70%. To my mind that means the point at which accumulated maintenance expenditure (or perhaps we should include any or all expenditure) on the kit concerned reaches 70% of the current cost of replacement.

After punching in:-

1) 7.0 %
2) 3.0 %
3) 7.5 %
4) 70 %

... we arrive at planned life-times in the order of 14 years. That's according to my spreadsheet, anyway.

If we had gone for an annual maintenance of 10% we would have reached our 70% trigger point in 9 years. But if we were confident that only 5% per year was enough, then 70% would never be reached within thirty years!

Which is probably why back in Ye Olde Days we always assumed 60% as our "BER level"; giving, in the 5% scenario mentioned, 24 years - but only 7 years at 10%, and 10 or 11 at 7.5%. So I'm pleased to say that my "traditional" figures (10%, 60%) still appear to arrive at a sensible result, even though, to be honest, we ignored depreciation at the time (and did not account for it - add it in - as an annual expenditure).

So hopefully we can agree that the planned annual maintenance budget (followed, of course, by its successful implementation) is the critical element (and the only one over which we can claim to have any influence). But what should it be?:-

1) 5%
2) 7.5%
3) 10%
4) Some other figure?

OK, when considering (planning for) depreciation and inflationary price increases we can apply an appropriate factor to decrease the residual ("book") value of equipment at any time, and also keep tabs on its like-for-like replacement cost.

But here's another thought:- should we be also applying similar calculations to spare parts (and other types of) stock ... and tools and test equipment etc.?

As those who have tried it will know, what you actually get when selling off tools and test equipment, not to mention "job lots" of parts, is often somewhat, shall we say, disappointing.

The situation with stocked parts is interesting, as if they have been sitting on the shelves for a while, even though still "new" (unused) it could be argued that their book value is depreciating. But on the other hand they should normally end up being "sold" to a job.

So if we apply our typical figures of 7% depreciation p.a., and 3% inflation p.a. to each item, what do we end up with? In this case its value would fall to half of what we paid for it within ten years. In business situations we would apply our mark-up on top of that, of course ... but it is that depreciation figure that is the "killer".

This sort of thing serves to illustrate why (these days) no-one wants to be left "holding the baby" as far as vast amounts of stock is concerned; leading to Just-in-Time parts provisioning, and all the rest. In other words, simply stocking parts (or indeed, anything else) costs money!

I provide a schedule to the end users/Divisions/Trust based on baseline lifecycles per equipment type. All configured behind the scenes in the database.

The list is provided in reverse date order so the oldest kit is at the top and the newly delivered stuff at the bottom.

Once provided to the lead planners or responsible persons at divisional or departmental level, they can they look at the 'expected replacement due date' and decide themselves what the priorities should be based on clinical judgement along with some support from EBME.

It goes a good way to providing them with a draft document which they can further tweak based on competing pressures (economic, peers, social, technological, standardisation, reliability etc etc etc).

Geoff, I agree and like your term of 'indicator'

Yes. We can provide "them" with all sorts of useful, pertinent and objective (and hopefully, accurate) data:-

1) Age of the equipment concerned
2) Current condition ("serviceable" and "well maintained", of course)
3) Anticipated equipment retirement date(s)
4) Anticipated uneconomic maintenance date(s)*
5) Frequency of breakdown(s)**
6) Current "book" (depreciated) value(s)
7) Cost of like-for-like replacement

... and all the rest; based upon dates in the equipment history, up-to-date market information, and painstaking recording of expenditure over the years.

As well as other stuff such as:-

1) Manufacturer's "End of Support" notification(s)
2) Dwindling of parts stock ... etc., etc.

But, as we all know, the "decision" related to the question posed in the first post in this thread will be taken, by and large, "based" upon subjective criteria (such as those hinted at in your last paragraph).

So yes ... by all means assist and advise as far as is helpful, but in the final analysis (or not, as the case is more likely to be), take a step back and let "them" - the users - make the relevant decisions.

Or, to put it another way:- rather than making a rod for your own back, put the monkey on their back.

* The point at which further maintenance expenditure on the item becomes uneconomic according to pre-established financial criteria.

** Remember MTBF?

How about some default answers?

1) Years elapsed since acceptance and commissioning
2) Serviceable
3) Date of acceptance plus 15 years [10 years, 20 ... whatever]
4) Sometime beyond the date at 3)*
5) Rare (if any)
6) Low
7) High

* And therefore (despite being "much debated") rarely invoked in reality - as long as maintenance budgets continue to be "cost-effective" (7.5% p.a. of initial equipment purchase price, or less) - and depending upon the "trigger level" used (50%, 60%, 70% etc.).

For what it's worth, these days I am using the term "Economic Support Limit"* (ESL); which is hopefully not only a bit more descriptive, but also jettisons possible confusion with historical terminology (which doesn't actually sit that well with the model - concept, metaphor, rationale, whatever - I like to employ now).

But either way:- yes, it is simply a trigger, an indicator (calculated with the passage of time, accumulated expenditure, and current equipment values) rather than an absolute. A "flagging up" that a certain stage in equipment life has been reached, and that consideration for replacement may then (now) be due.

In other words (and using Old English):- it is probably time to stop "throwing good money after bad"!

* "Economic Support Level also works.

Hi All

This is a problem we have been working with our hospital here in NZ. We have come up with a 'tool' that is based/developed from (very loosley) on comercially available products and has items of data collected (and updated in the case of four fields) for each asset on our database.

We call it 'relable economic life and technical criticality' by collecting and combining the following data:-
Risk rating -(according to ECRI -low, medium, high)
Avaliabilty for service delivery - is there a subitute avalaible - in the department/hospital/country
Technical obsolescence - relates to technical status of spares and service avaliability
Maintenance level - is the level of mainatining these units at what the manufactures recomend or greater/lesser.
Condition - what is the condition of the unit - this is one field that is updated year on year.
Years over replacement date - based from purchase date.

The data collection is based on the data that is already in our data base and the like of conditional data are added year on year automatically as the Biomed tech' test each item (this is achieved by a custom piece of software that the Biomed uses, they barcode scan the asset, the assets details are then displayed on the laptop, these details are verified by the Biomed, the Biomed then can change the status of the like of conditional analsys etc (which are defaulted to average condition etc)then the Biomed presses the test button and open the correct Fluke ansur test file for the equipment - once the testing is completed and the Biomed come back to the mainscreen they can print a tested label for the unit. then on to the next unit)

The advantage for us is that each item asset in the hospital, independant of use/value/location can be ranked in order of replacement importance.

Regards

Interesting!

So ... are those (six?) fields assigned weightings?

How is "it's time to consider replacement" flagged up?

What is the calculation (or basic "formula")?

What figure do you use to arrive at "replacement date" (anticipated life of equipment)? How many years (and is this a default for all equipment, or do you anticipate different lifetimes for different equipment types)?

I don't see any financial data being mentioned there.