Waiting for things to fail.

[Mattb4]

Watching paint dry is a walk in the park compared to monitoring some of this solar equipment for when it will fail. Hard to establish an item's worth until you have some data on how long it lasts. Then there is the fact something new or better is out there but you can not justify ripping out a working device simply to replace it. No, you wait and wait as items just ugly away with age.

I recently replaced a perfectly good working smart phone of 5+ years with a new. The reason being the new was a good buy and was up to date on program features and protections. Though in reality I just coveted the new phone more than my old one. At least with computers/smart phones it is mostly accepted that you can seek an early upgrade path. Though I am not one for the latest and greatest. I like value.

So waiting on panels, batteries and AIO's to fail is a slow and tedious process. Do I have to wait 2,3,5 or more years?

How long should you expect to have to use something before you can replace it even though it still works? Consider the Pyramids, built thousands of years ago, in serious need of renovation and update but because they are still standing the wait goes on.

 

[Texican]

2012 SMA gridtie inverter just wont die, wish it would

 

[Johneb]

I read on the outback forum of a guy who has had 12 trouble free years. I would say find something else to do. I hear solitaire is fun.

 

[wiseacre]

If you don't want to wait to see how long it takes to fail have me install it

 

[G00SE]

I feel like that accurately describes some of the inverter landscape. I sure hope I don’t have to replace batteries or panels this decade

 

[grebaba]

Take up hot rodding. Racing cars will show you how long your parts will last.
The upgrade path is fairly quick.
Greg

 

[Hedges]

If you don't want to wait to see how long it takes to fail have me install it

Did you install the SB 7.7 -41 I just picked up?

I had several 2500SWR operating for about 17 years starting in 2003. Some 32 year MTBF calculated from my failure results.

I've picked up some -41 series to meet latest requirements for new installs. 2 out of 3 were good.
The 7.7 didn't light anything up. I opened the cover and found droplets all over. Bunch of swelled and split caps.

I think I'll check resistance of each MPPT input A/B/C. If some are shorted but not all, I'll wash off the conductive goo and try powering it from the surviving inputs. Then consider replacing caps.

(When commissioning a new system, leave DC switch open. Check PV voltage and polarity before closing.)

 

[Danke]

Well, if you’re bored, wash the panels, dust off the batteries, and wax the inverters. Maybe a new paint job too?

 

[Steve777]

We live in a culture which promotes buy the latest and greatest, whether you need it or not. I suppose it keeps consumption high, but really does not serve most consumers all that well (IMO)...

My general rule is use something until it breaks or no longer will do what you want from it. Planning my PV expansion is a good example (I think). Still using an original Trace SW4024 from like 25 years ago. Still works perfectly, no complaints whatsoever. But I would like to add ~10KW more PV and do another net metering agreement with my PoCo. The easiest and cheapest solution would be to add a second SW4024 and another SCC. But neither the PoCo nor the local Bldg Dept will accept a SW4024 anymore (not UL1741SB). So I am forced to replace the old Trace and go with a newer AIO (most likely) with all the right certs.

While a newer inverter will have more bells and whistles (data monitoring, multiple level load controls, battery comm, etc), there is no real "need" as the old Traces did all the basics and are VERY reliable. But since the PTB who need to approve the change won't have them anymore, not much I can do if I want to expand.

I always wonder some, when I read here about folks that are going to put in XYZ inverter now and plan on replacing it in a few years with something newer/better. Why wouldn't you just put in the best unit you can that does what you need now and use it till it breaks? Guess I am just not the ideal consumer.

To answer the OP's original question, keep using it until it breaks (or is making signs that might be imminent), or until it can't do something that you need.

 

[wdwtx2.0]

To answer the OP's original question, keep using it until it breaks (or is making signs that might be imminent), or until it can't do something that you need.

+1 with exceptions.

 

[Mattb4]

Well, if you’re bored, wash the panels, dust off the batteries, and wax the inverters. Maybe a new paint job too?

Longevity experimenting, or finding how long something lasts, is fraught with moments of sure tedium punctuated with periods of intense boredom. But how else can you tell if it will work for X number of years until someone actually runs it for X numbers of years?

Reminds me of the old Maytag Repairman TV commercials back many years ago.

 

[Mattb4]

We live in a culture which promotes buy the latest and greatest, whether you need it or not. I suppose it keeps consumption high, but really does not serve most consumers all that well (IMO)...

My general rule is use something until it breaks or no longer will do what you want from it. ...

One advantage of the promotion of new things is it gives you new features. Only problem with some of the new features is they may or may not be those I want or need. Sometimes after longevity testing an item I find it is good enough at failure that I simply want to replace it with the same. If the period before failure is quite long (or short for some stuff) the item is no longer available.

So if it lasts really long proving its worth it becomes obsolete.

 

[50ShadesOfDirt]

I think the answer is "warranties" ... no waiting around needed if you've done the research, and avoided the:
- "90-day" to "6-month" garbage
- "we love our product, and think you will too (but it still only has a 1-year warranty)"
- "10 generations later, we think we've got it right, so we're raising the warranty to ... 2-years!"

In most cases, warranties have gone into a death spiral, sinking lower and lower ...

It's one of the reasons I was happy about my Magnum LF 4024 inverter purchase, w/ 5-yr warranty, and it's still going years later, beyond that warranty period ...

It's one of the reasons why I avoid most AIO's these days, because the warranty, the support mechanisms (or lack thereof), website shenanigans, and other indicators scream "run, don't walk away from me" ... but only if you do the research.

Something like a Midnite Solar AIO, with much longer warranties, starts to speak to me that quality is there, plus their service record seems to indicate they'll actually help you if something happens ... this might be the first AIO where you don't have to buy two at once, and keep one of them on the shelf.

But, who does the research these days ...

 

[SparkyJJO]

If I never have to replace my inverter I'd be happy. I'd just want two of them. Maybe three.... More power!
Dang it, I'm still a bit salty about AltE canceling my GS8048A order... Meh.

 

[Hedges]

But how else can you tell if it will work for X number of years until someone actually runs it for X numbers of years?

HAST.

Highly accelerated stress test - Wikipedia

en.wikipedia.org

Put your inverter in an oven. Add water to create humidity. Cycle the oven on and off.
The temperature can be selected to achieve a particular acceleration factor, e.g. 1 month equivalent to 10 years.
Capacitors, semiconductors, etc. all have an acceleration factor. Whichever achieves acceleration at lowest temperature will be limiting. So won't necessarily find what failures will actually occur first, nor when.

Each individual part might have be separately tested, and a lifespan vs. temperature curve or equation created.

The approach of calculating life of each part is taken, based on calculated/simulated thermal and electrical stress.
I think this tends to give an optimistic life; something always happens to shorten life in the real world.

High temperature burn-in tends not to show much (except maybe electrical capacitor life?)
Temperature cycling stresses joints and PCB vias, and may be more predictive.

 

[Mattb4]

HAST.



Put your inverter in an oven. Add water to create humidity. Cycle the oven on and off.
...

Yum. With a honey glaze it sounds delicious.

Though running in the wires for input and output might look a bit odd.

 

[SparkyJJO]

At work we have thermal chambers that have routes available for wires/etc to connect to the tested equipment.