Inverter Lifespan

[FB.]

What would you expect to be the average useful lifespan of a solar PV inverter from respected manufacturers such as SMA, Fronius, Powerone?

Options:

5-7 years.
8-10 years.
11-14 years.
15-20 years.
Over 20 years.
Don't know.

 

[Gavin A]

it all depends on how it's been installed, how warm it is during operation, how dusty it gets, if it ever gets the heat fins and fans dusted, if the fan is replaced fairly quickly after it fails or the inverter's just left to gradually die after the fan fails etc.


Assuming it's installed in a cool clean environment, and the owner keeps a bit of an eye on it, replaces the fan quickly if it does fail etc then I'd expect them to average over 20 years.

If you're asking about what the actual average is likely to be across all installations in the country... probably somewhere in the the 15-20 year range.

 

[SolarCity]

I would say that if they last the first 12 months they will be good for 20 years - assuming that they are installed with a bit of common sense.

 

[jason121]

If they last 20 years the inverter companys wont be selling many, they will need new installations to keep them going and if they are built like the inverter drives of 10 years ago they wont. The capacitors degrade. I tell my customers that might need to change the inverter in 10 years, most are not bothered, seeing that they are mostly retired.

 

[whinmoor]

If the system is well designed I'd expect it to comfortably exceed the manufacturer's warranty period. So much depends on the actual circumstances. I'd agree that if it survives the first year, it should last many more.

 

[FB.]

if the fan is replaced fairly quickly after it fails

So how much would you - or other electrical companies - be likely to charge someone to replace an inverter's fan?

 

[jason121]

I have taken a fronius TL apart and the fan is no different from a PC fan.
Depends on distance but looking around £150.00

 

[FB.]

If you're asking about what the actual average is likely to be across all installations in the country... probably somewhere in the the 15-20 year range.

With most inverters in discreet places (such as lofts), and with most people likely to not really notice a performance problem, I guess your answer should be 15-20 years if you haven't yet voted.

.

 

[FB.]

I have taken a fronius TL apart and the fan is no different from a PC fan.

I've not pulled my Aurora apart but I'd got the distinct impression from the noise and the size of the air inlets on the indoor version, that the fan would be along the lines of a PC fan.

 

[mechelec]

If they cant make a TV last more than 5 years then you can bet your boots inverters will not last 15 years. I would say 10 years would be good going.

 

[SolarCity]

TVs don't last more than 5 years?

 

[Gavin A]

So how much would you - or other electrical companies - be likely to charge someone to replace an inverter's fan?

it's a user serviceable part on most of the SMA inverters - you don't even need a screwdriver to do it on the HF inverters.

how much we'd charge would depend on distance, but it's a 1 hour job at most (possibly a bit more as we'd probably want to retest the system while we were there just to double check everything else was fine), so I'd be surprised if it worked out at more than £100, probably more like £75 inc parts and fitting.

 

[sedgy34]

So how much would you - or other electrical companies - be likely to charge someone to replace an inverter's fan?

i would imagine if an inverter is taken apart to get to the fan it will void the warranty

 

[Gavin A]

If they cant make a TV last more than 5 years then you can bet your boots inverters will not last 15 years. I would say 10 years would be good going.

the tv I'm watching is at least 15 years old and still going strong.

In fact looking around the room I can see a pair of 40 yearold speakers, a 30 year old hifi amp, 20 year CD, tape and record player, all working fine (well, actually the amp has developed a bit of a glitch, but still works after a bit of a tap).

Good quality electrical kit can last a long time, cheap tat won't.

 

[FB.]

the tv I'm watching is at least 15 years old and still going strong.

In fact looking around the room I can see a pair of 40 yearold speakers, a 30 year old hifi amp, 20 year CD, tape and record player, all working fine (well, actually the amp has developed a bit of a glitch, but still works after a bit of a tap).

Good quality electrical kit can last a long time, cheap tat won't.

It will be interesting to see how panels cope with ageing. I expect them to last much longer than the inverter, but I do wonder how long Sanyo panels will last compared to perceived lower-quality (Chinese) alternatives.
At the moment, I see a Sanyo-panelled system as being about 50% more expensive, but only 10% more generation (and that's subjective). Whether the quality will remain long after the price is forgotten, we'll have to see.

 

[jason121]

the tv I'm watching is at least 15 years old and still going strong.

In fact looking around the room I can see a pair of 40 yearold speakers, a 30 year old hifi amp, 20 year CD, tape and record player, all working fine (well, actually the amp has developed a bit of a glitch, but still works after a bit of a tap).

Good quality electrical kit can last a long time, cheap tat won't.

Tapes CD records someone stuck in 70-80-90's

 

[edexlab]

I went to SMA a couple of years ago and they were quite confident of the inverter lasting about 20 years installed correctly this wasn't the salesmen talking but some of the Engineers, of course there will be some which will fail earlier, there's always a percentage of any manufactured item which will have shorter lifespans than others.

I'm inclined to believe them as I asked what the average age of inverters returned for repairs etc was, and the answer was 16 years for inverters installed for more than one year but the number of inverters returned in their first year was unpredictable as even though rigorous tests are carried out they don't compare to real life installed scenarios and they learn where the problems are and make changes throughout the production run of the inverter model, this seemed to me to be a fairly honest sounding reply and fits the idea of ''if it lasts a year it'll do many more'' .

In todays competitive sales spiel driven markets I would expect that if the same question was asked of some other companies I'm sure the figure given would be 20 years plus and they never get returns etc
They also said it was sensible to factor in the cost of a replacement for a system expected to last around 30 years

 

[Gavin A]

I asked the same of Power-One's technical manager for the Uk, and got the answer that the better models had a design life of 25-30 years in optimum conditions.

they have the advantage of not having external fans to get clogged up and fail, but the disadvantage of the fan unit not being user serviceable as it's internal.

 

[mechelec]

I asked the same of Power-One's technical manager for the Uk, and got the answer that the better models had a design life of 25-30 years in optimum conditions.

they have the advantage of not having external fans to get clogged up and fail, but the disadvantage of the fan unit not being user serviceable as it's internal.

Interesting! where does it get its air from then? does it circulate the air within the inverter which would be pointless or does it draw it from outside the enclosure leaving you with a fan that will still "clog up and fail" but is not user servicable.
If the cooling fan is so critical why are inverters often installed in loft spaces where the ambient temperature can easily hit 50 degrees, just when the inverter is working hardest. That has got to be a recipe for a shortened lifespan.

I am on my second TV in 7 years and this one has had to be repaired due to an inherant capacitor problem.

There will be a lot of PV systems sat on roofs with failed inverters that customers just wont have the money to repair.

 

[Ian-LS]

Interesting! where does it get its air from then? does it circulate the air within the inverter which would be pointless or does it draw it from outside the enclosure leaving you with a fan that will still "clog up and fail" but is not user servicable.
If the cooling fan is so critical why are inverters often installed in loft spaces where the ambient temperature can easily hit 50 degrees, just when the inverter is working hardest. That has got to be a recipe for a shortened lifespan.

I am on my second TV in 7 years and this one has had to be repaired due to an inherant capacitor problem.

There will be a lot of PV systems sat on roofs with failed inverters that customers just wont have the money to repair.

Sometimes it is because there is no viable alternative, and if this is the case an inverter with a derating temperatire of 50 degrees + would be needed and the loft will need adequate ventilation in the form of a ridge vent above the inverter if there isnt already one there. Sometimes though unfortunately it is because the installer doesn't know any better and simply its "we always install in the loft space" which we have heard many times even when there is a viable garage space where the consumer unit and suppliers meter is which will be much cooler or space under the stairs etc.

 

[SibertSolar]

Interesting thread. Not all inverters have/need internal forced circulation of air to cool down internal components. Using high quality components and a design focused on efficient layout of internal power-flow and passive air flow etc can avoid unwanted hot-spots internally thereby prolonging the lifetime of the inverter overall. A good quality unit that is matched adequately to the system should be good for 15+ years quite easily. Most top of the line string inverters have 10yrs warranty by default, with options to increase to 20yrs if required.

 

[FB.]

There will be a lot of PV systems sat on roofs with failed inverters that customers just wont have the money to repair.

Yes, that's a concern and that's why, since I came onto this forum, I have advocated a "depreciation" allowance as a margin of safety, since the customer can't get the money back like in a bank account, and in order to continue earning FiT payments and enjoying bill savings when the system gets old, they may need to have built up a cash reserve to pay for repairs to keep the old girl running.
Solar PV owners should use about half of the FiT payments in the first five years to build up a repairs fund just in case they're one of the unlucky ones whose inverter fails in years 5-7. Such a "rainy day" fund is desirable anyway, since it might be the car which needs repairs, or the boiler - all of which can amount to hundreds of Pounds quite quickly.

 

[MikeyT]

I would expect string inverters to last around the 10 year mark, considering the positioning of most in lofts which are generally more dusty and very very warm! It also depends on the client, as the majority will have forgotten about the inverter after 5 years and probably would never even check on it. Though by the time they do fail, the cost of a new one would be considerably less than now, I would hope anyway! The cost-benefit of just changing it to purchasing warranty makes it a no-brainer in my opinion right now.

 

[SW1970]

Not having any knowledge about typical location of inverters I was quite surprised to read above that they are frequently installed in lofts. Most lofts I've seen are a bit of state, so whether from dust, rodent, ambient temperature or fallen junk, it's a harsh environment for such an expensive piece of kit. FB - your idea is a good one.

 

[Vegelen]

The inhernet inverter weak spot IMHO is the quality and type of capacitors installed inside them, aside from the human factor and innapropriate system design /installation.
I expect to see 1 or 2 Inverter failures in the first 20 years of service.

 

[SolarCity]

I don't think inverters failing outside of their warranty will be a big issue - judging by pricing trends I would say that in 5-10 years they will cost significantly less than they do now, and probably be a little more efficient.

 

[Vegelen]

Reductions in cost always tend to come at a price of long term reliability, as we can see in modern electrical appliances all around us.

 

[njcuk]

Interesting! where does it get its air from then? does it circulate the air within the inverter which would be pointless or does it draw it from outside the enclosure leaving you with a fan that will still "clog up and fail" but is not user servicable.
If the cooling fan is so critical why are inverters often installed in loft spaces where the ambient temperature can easily hit 50 degrees, just when the inverter is working hardest. That has got to be a recipe for a shortened lifespan.

I am on my second TV in 7 years and this one has had to be repaired due to an inherant capacitor problem.

There will be a lot of PV systems sat on roofs with failed inverters that customers just wont have the money to repair.

I think most customers could afford it, considering the initial outlays and returns they will be getting. Repairing one for £100-£150 is not going to break the bank.

Regards TV's, you've been unlucky. If I were you, i'd take out extended warranties.

 

[Gavin A]

Interesting! where does it get its air from then? does it circulate the air within the inverter which would be pointless or does it draw it from outside the enclosure leaving you with a fan that will still "clog up and fail" but is not user servicable.

it's designed to blow heat off the components and onto the reverse side of the heat sink on the front of the inverter, so doesn't draw air from anywhere it's just recirculating the internal air.

If the cooling fan is so critical why are inverters often installed in loft spaces where the ambient temperature can easily hit 50 degrees, just when the inverter is working hardest. That has got to be a recipe for a shortened lifespan.

yes, it will virtually always shorten the lifespan to some extent, but using inverters that are designed for higher temperature operation, fitted with higher quality capacitors, and either fitted with internal fans protected from dust, or easily replaceable fans, or even passive cooled reduces this impact, and I'm relatively confident should result in a 25-30 year optimal lifespan turning into a 15-20 year lifespan on average.

Depending on the airflow, temperature and dustyness of the loft of course, but the panels covering the south facing roof aspect does significantly reduce the peak temperatures, and we try to install an additional tile vent above the inverter if ventillation isn't too good, to reduce the worst of it.

I am on my second TV in 7 years and this one has had to be repaired due to an inherant capacitor problem.

There will be a lot of PV systems sat on roofs with failed inverters that customers just wont have the money to repair.

that's why cheap and nasty inverters with cheap capacitors are to be avoided like the plague, though it's not always obvious which makes this applies to.

 

[The Solar King]

A good indication of expected life is the price of extended warranties. Anyone who is good on the horses should be able to work out what the odds are of an inverter failing and needing to be replaced based on current cost. Just like bookies the warranty companies are in it to make a profit.

 

[Gavin A]

A good indication of expected life is the price of extended warranties. Anyone who is good on the horses should be able to work out what the odds are of an inverter failing and needing to be replaced based on current cost. Just like bookies the warranty companies are in it to make a profit.

to some extent, but as detailed above, the lifespan is dependent to a large extent on it's location and if it's ever dusted / fan replaced etc, so the inverter warranties will be based on pretty much worst case scenarios IMO.

 

[The Solar King]

This is almost like one of those answers on QI. Nobody knows. I always tell prospective customers that it is likely that the inverter will need to be replaced during the life of the system and build an appropriate cost plus inflation in to the PVSol economic model to cover.

 

[Dan]

Don't mind me while I bump some threads in the solar PV forum. If they're not current or applicable topics, just ignore the threads; they'll drop off the page. If they are, feel free to chip in and chat solar stuff in the solar panels forum.

 

[sgt_woulds]

From experience I would say that inverters with a transformer built pre 2008 can last well over 20 years - I'm talking about SMA SWR 750 - 2500 type inverters. Some that I installed 17 years ago are still working fine - quite a few fitted into lofts - and I've worked on a few that were installed before i joined the industry too.

The main reason is that these old style (G77 era) inverters used good solid components that were properly soldered.

Modern TL inverters are built down to a price with tiny, flow soldered components components, that are far more susceptible to heat and (more importantly) transient voltages. They are inherently less robust.

We used to tell our customers that a good inverter will last 12 - 14 years. Now we tell them 5 - 10 years.

You cannot choose according to brand name either. Fronius are crap. Every single one we fitted after 2010 has failed within 4 years. Mastervolts are almost as bad.

ABB / Aurora inverters had a bad batch from 2008-ish that all failed (E031 - due to soldered components failing in 'unpredicatble' UK temperature fluctuations...) after about 4 - 5 years. And it takes on average 4 - 8 weeks to get a replacement from Italy.

SMA's TL's now last on average 5 years and 6 months, but we had a period of one year where we stopped fitting the SB2500 because of a 99% failure rate. So even amongst one manufacturers range there can be problems. This issue was compounded by SMA denying any issues - even when presented with our installation / failure figures.

The Solarmax S range was bullet proof - the P range all failed within 4 years (probably why Sputnik went bust).

We've only had 2 failures out of all the Samils we've installed some of which are coming up to 10 years old now.

We recommend the fitment of SPD's for anyone with high average voltages (so next to railways, sub-stations - or anyone who lives in Cambridge where we regularly see 260 V mains) - certainly for micro inverters but increasingly for string inverters too. Not that every customer takes that recommendation.

Don't get me started about micro-inverter failures... !

 

[FB.]

Woken from hibernation by an email.......

It's now almost five years since my panels were installed.

Power One Aurora PVI-3600 (indoor version), in downstairs utility room.
8+7 (two strings) Kinve KV250-60M panels (3.75kWp) facing ESE at 40 degrees.

FiT payments so far totalling around £8k. Bill savings hard to quantify. Almost certainly reached 'payback' point with 20years of FiT licence remaining (subject to political interference).

Generation averaging about 3300kWh per year; worst year around 3100kWh, best year around 3500kWh.
Generation tends to be mean-reverting, so if we have an unusually dull spell (small FiT claim in one quarter) it's very often followed by an unusually bright spell (high FiT claim the next quarter).

No noticeable loss of generation as the system ages although variable weather patterns could mask a gradual decline for a few years.
Winter tends to see most days generation below average, with a few bright days pulling up the averages.
Summer tends to see most days above average, with a few very dull days dragging down the averages.
Spring and autumn tend to have a roughly equal mix of above-average and below-average generation.

The only 'problem' is that a couple of years ago the inverter lost its date and time settings, and I know someone else with an identical inverter of similar age that has the same issue.
I haven't investigated because everything is still working and I've been short on time, but I wonder if it might be as simple as a small button-battery needs replacing as is commonly found in electrical devices to retain date/time when they're switched off.

Cheers,
F

 

[Electric Yo]

I've not pulled my Aurora apart but I'd got the distinct impression from the noise and the size of the air inlets on the indoor version, that the fan would be along the lines of a PC fan.

ABB / Power One UNO's and TRIO's - 3.0, 3.6, 10.0, 12.5, 20.0, 27.6 use passive cooling - therefore no fans. The IGBT s are strapped against the heat sink and use air convection for cooling. The newer TRIO's - 20.0 and 27.6 have a small fan inside to cool a capacitor set on the A/C filter boards, but that's it.

Fronius use active cooling method. So a smaller heatsink and therefore unit. But a fan helps the heatsink to convect heat.

I'd say an inverter should last around the 10-15 year mark.

 

[FB.]

ABB / Power One UNO's and TRIO's - 3.0, 3.6, 10.0, 12.5, 20.0, 27.6 use passive cooling - therefore no fans. The IGBT s are strapped against the heat sink and use air convection for cooling. The newer TRIO's - 20.0 and 27.6 have a small fan inside to cool a capacitor set on the A/C filter boards, but that's it.

Fronius use active cooling method. So a smaller heatsink and therefore unit. But a fan helps the heatsink to convect heat.

I'd say an inverter should last around the 10-15 year mark.

The 2011-model Power One Aurora PVI-3600 (indoor model) that I have definitely sucks in air (top left) and blows out air (middle right). The groan/whine of the fan starts when internal temperature rises to 50'C and stops when it drops to 40'C.

 

[Electric Yo]

The 2011-model Power One Aurora PVI-3600 (indoor model) that I have definitely sucks in air (top left) and blows out air (middle right). The groan/whine of the fan starts when internal temperature rises to 50'C and stops when it drops to 40'C.

That is a PVI-2000 I believe? I was unaware they had a 3.6KW version?

But yes your right those units have active cooling and hence are a lot smaller.

 

[FB.]

That is a PVI-2000 I believe? I was unaware they had a 3.6KW version?

But yes your right those units have active cooling and hence are a lot smaller.

It's definitely a 3600.

 

[FB.]

Link to spec sheet for PVI-2000 and PVI-3600.
https://library.e.abb.com/public/24111ae72d47765585257cfa0069b954/pvi-2000_3600_en.pdf

 

[mydigitalhome]

Yep, but if kit did not have a built in die date we would all be out of work. If it hasnt broken its an indication your stuff does not have enough features and flashing lights and buzzers. Time to invest in some new kit and have the joy of taking it out of the wrapper, the box the next box the foam and multiple plastic bags. What are you waiting for?