Hi,

I caught a bit of a discussion the on the radio the other day which said that Which magazine has said that people having PV installed are not being told that inverters are expected to need replacing every 10 years or so.

Is this true? My inverter was a large proportion of the cost of my installation, and if I need a new one every 10 years I really would have wanted t know that in advance!

Thanks.

Seems that you can get Sunny Boys from between £1000 to £1,800 depending on size.
http://www.simplypowersupply.com/Grid-Connect-Power-Inverter/default.aspx

185Wp panels from £308, so the inverter is somewhere between a half of the price (1kWp system) and a quarter (4kWp system) of the price. This will change depending on the instillation labour cost. Oh and I am sure someone can find cheaper kit somewhere else on the interwed.
I doubt if inverters can come down in price as much as say LCD TVs have over the last 10 years as it is a very different marketplace.

Will extend the payback time quite a bit though.

Posted By: tedIf the inverter isn't stressed then there is no reason why it shouldn't last longer

But don't you loose efficiency at lower load factors. I don't know that much about inverters, but in power supplies isn't it the large capacitors that tend to go first, they could be repairable/upgradable. Would need to ask someone that has a soldering iron and good eyesight.
The conversation about these inverters built into individual panels can up the other day. Sounds great in practice but a very high cost to replace when roof mounted even when assuming that you can get a direct replacement.

As ST says it is the elctrolytic capacitors that have the shortest life.
Full explanation here: http://jianghai-europe.com/wp-content/uploads/JIANGHAI_Elcap_Lifetime_-_Estimation_AAL.pdf

Basically, it is down to temperature, mainly ambient, but also the self heating of the capacitor from the current it is passing. The good news is that for every 10 degrees C below the maximum operating temperature of the capacitor, you double its life. So, for an 85 degree C capacitor with a life of 2000 hours (fairly typical), the life at 35 degrees C is 64000 hours, or about 7.3 years. This for a percentage of devices to fall outside the manufacturer's specification. This gives another factor which is more difficult to estimate since we can't tell what capacitor characteristics a given inverter can "get away with" since we don't know how much safety margin the manufacturer built in.

Having said all that, 10 years seems reasonable.

Brian.

Oddly, our PVR was acting weird a couple of years ago, so I got my daughter to replace all the power supply electrolytics (her eyesight is better than mine). Only one was out of spec and it was obvious why - it was surrounded on three side by a heatsink, so would have run quite hot. Its measured value was half the value marked on the casing before anything became really apparent with the PVR's operation, and that took 5 years of constant use to get there.

"If the inverter isn't stressed then there is no reason why it shouldn't last longer. "

For PV, by sizing the inverter in the UK at 0.8 of installed kWp of the array , is this possible potential
for additonal stress on the inverter .
I noticed the manifacturers SMA suggest 90-110% when sizing inverter against installed kWp

Re. inverter life expectancy, i thought this was common knowledge and installers should be informing client when question come up about warranties etc.

As Damon mentioned in 10 years time an inverter will not be the cost it currently is.
( so dont bother buying the extend warranty )

The on-panel micro-inverters have longevity engineered in (no electolytics, for a start). I believe they quote 20yrs. Remains to be seen if they've succeeded in this, but they've definately spent engineering effort, and some budget, on reliability engineering (because it is indeed much less conveneient/more expensive to replace on the roof).

"For PV, by sizing the inverter in the UK at 0.8 of installed kWp of the array , is this possible potential
for additonal stress on the inverter .
I noticed the manifacturers SMA suggest 90-110% when sizing inverter against installed kWp"
MSC expects sizing at 80% ,
Yes Uk irradiance average 750-850W/m2 , but recently i've measured peaks of 1200W/m2 in full sun
Inverters may be cutting out due to going over the maxium load , thus lossing benefits of these peaks

any thoughts on this Gavin ?

( great site by the way , clear info and open pricing)

mcs doesn't mention it as far as I can see, merely referring to the DTI guide, which states

"PV array: inverter ratios from 1:1 to 1:0.8 are commonly applied in the UK, though in certain circumstances and depending on the inverter used, ratios outside this are sometimes utilised (NB: Inverter power is taken to be maximum steady state a.c. power output)."

but more importantly...

"Guidance on inverter : array sizing can be obtained from the inverter
manufacturers – typically from system sizing software."

so if the inverter manufacturer's system sizing software says yes, the DTI & MCS guidance also says yes.

The thing to remember with all this is that when sunlight is brightest, the panels will also be at their hottest, so can easily be operating at 20-25% below their rated output. This could be partially offset by above 1000Wm2 sunlight levels as you suggest, it's relatively rare for it to be much above 1100Wm2 for long IME, as air pollution tends to go up during long hot sunny spells.

It's also worth bearing in mind that the panels will degrade over time, so after 10 years the peak array output is likely to have reduced by 5-10%. Also, the inverters will derate on the odd occasion that they do get too hot by simply adjusting the MPPT point slightly, so while larger systems will mean the inverter will be operating at warmer average temperatures and needing to use their fans etc more, they shouldn't actually be exceeding the design temperatures. I'd think the fan and varistors would wear out faster though through having more use, so they would need checking and replacing more often, and the potential for greater damage if they're not replaced would consequently be higher.

Yes I always thought you refer to manifacturer guidance above any other guidance. etc.
MCS guy made a fuss on our assessment job, saying we over sized the inverter ,
we used sma software to size , so I couldn't see what his problem was.

good info thanks, make sense re. panels degrading so inverter sized 80-100% systems size will suit that.

lots of people suggesting the best place for the inverter is in the loft.
I wonder if this is just because its easier for the installers.
I'm thinking inside the insulated envelope has got to be better , much more stable temperature range unlike most lofts, so cooler in the summer and heat from the inverter will add (if only a little ) to the house in the winter. this might reduce stress on the inverter .

You have to run the DC to the inverter and the AC to the board so anywhere on that route makes little difference, You've got to run one or the other
DC cable runs are suppose to be kept as short as possible to reduce Volt drop . but as most oversize with 4mm2
and domestic runs are probably never that long anyway, they should be within specified limits
The other possible negative of DC runs in the home is DC cable do look a bit like coax , so running that in trunking through a bedroom cupboard to say a utility room below , may create a problem if an uninform TV aerial guy come along and cuts them ( 400V, 6A, DC ) though they should be clearly labelled.

For the record, I just went and inspected the heatsinks of my SB1100s and SB1200s running very roughly 50% capacity at the moment: barely warm to the touch and very little dust to dislodge even after three years in the case of the oldest one (though the installers possibly did some cleaning themselves on subsequent visits/installs).

Rgds

Damon

Posted By: Gavin_A... mounted in a fully dust and vapour proof housing, with an internal fan...

I've had several PC fail because the fan failed or the heatsink clogged with dust but I guess there are ways to design out that problem.

Posted By: MarkBennettHowever good they are, I can't see anything with a fan in lasting for 30 yrs even in a controlled environment.

Th emodel we have is a Aurora PVI-3.6-OUTD-S-UK. I was under the impression that they were fanless, but looking through the manual I see that there is mention of a "Fan Fail" status message (classed as a "warning" rather than an "error"). If there really is a fan in there then it runs very quietly and I'd certainly never guessed there was one.

The big heatsink has its fins on the outside of the case.

Posted By: MegacyclesThe same inverter here, and the fan comes on under high output conditions, I think it switches on when the heatsink is>50'C.

The highest temp I've seen reported on the LCD here was a touch under that (around 48C, I think).

Any air movement it generates appears to confined to inside the cabinet.

Makes sense, given the IP rating.

PS. Since it's an "outdoor" inverter installed in a loft space[1], I've given it a bit of artificial air movement in the form of an old PC fan slung below the heat sink, blowing up, and powered by a Maplin cheapo PV designed to trickle charge a battery. That seems to knock a small number of degrees off the temp when output is high.

[1] Pretty well insulated. Even in summer it's often cooler up there than down in the living area. It has only rarely exceeded 23C up there so far this year.