Inverter Temperature Derating

[scottliv]

As this was the hottest day of the year, I expected a high yield but was slightly disappointed. Having checked the inverter, it was the first time I had heard the fan on.

My system is a 3.75kWp (due south on the south coast with 15 x 250w Sanyo panels with an SMA SB 4000TL) with a loft mounted inverter. The ambient temperature in the loft was about 35 degrees C, but the cooling fins on the back of the inverter were in the region of 60 degrees C.

I understand that in the hotter weather will reduce the panel output but did not think it would be affected this much. I peaked at about 3.1kW, producing 24 kW in total. I assumed that as the loft temperature was so high and the fan was on the inverter was limiting on temperature, but I cannot find any information on this.

Will the inverter be derating on temperature and how would I know this was happening.

 

[poheZ]

if the inverter hits a critical temperature then an error code will apear. The fan is internal and for air circulation. Remember peaking is not as important as accumulation of kwh

Do you have any monitoring for your system?

 

[Gary K]

Do the inverters loose efficiency as they get warm. Even though it hasen't reached a temperature when it cuts out, would it produce more if it was running cooler?

 

[poheZ]

SMA inverters (i'm sure they are all the same) maintain their efficiency within their rated operating temperature thresholds. So, no!

 

[Gavin A]

they can derate if they go over a set operating temperature though.

 

[scottliv]

I have a sunny beam and there was no warnings. I have not checked sunny explorer yet. I know the inverters derate but cannot find a level that this happens.

 

[poheZ]

If no warnings on sunnybeam then dont panic as explorer wont say any different

why not send a picture of inverter location

 

[mdovey]

I understand that in the hotter weather will reduce the panel output but did not think it would be affected this much.

The efficiency of the Sanyo panel decreases 0.3% for each degree celsius - this is the temperature of the module itself which in full sunshine will be significantly higher than the ambient temperature.

Our outside thermostat hit about 30 in the shade today, whereas a few weeks ago it was reading about 15. The panels would therefore be a minimum of 15 degrees hotter which is a drop of 4.5%, but I'd suspect that in full sunlight they were probably closer to 20-30 degrees higher than a few weeks ago (which would be a 6-9% drop).

So, I'd suspect any reduction in output from your expectations is due to the panels being hotter, than any derating of the inverter.

 

[scottliv]

I have checked the SMA instructions and a 'thermometer' symbol will appear on the front panel of the inverter if it is derating due to temperature. This was not on so it must have just been hot panels.

 

[scottliv]

With a bit more digging, and in reply to Mdovey's post, the Sanyo HIT 250 will produce 188.9W @ 45 degrees C. This equates to around 2835W total for the 15 panels; so my output per panel, not including losses, would have been in excess of 205W giving a peak about 3100W.In conclusion it all seems to be running fine for the current environmental conditions.

 

[Earthstore]

Also you have not taken into account any cable loses, which there will be some.
The fan is on the inverter for a reason, just to help cool it when it has been working hard and the conditions may be warm, so it helps to keep the performance up, not unlike your car really.

The surface of a panel can get very hot in direct sunlight, far hotter than the air temperature, so, as you say all is well.
I feel some customers worry a little too much about the odd blip, they need to look at their annual performance, this takes into account all the changing weather conditions.

 

[SolarCity]

I've noticed that a few systems have failed to peak in our area in this recent hot weather. It often surprises people how much affect heat has on panels.

Regarding the inverter, there is a graph about somewhere showing how efficiency drops off on SMA inverters as they get hotter. You don't necessarily need to get the inverter up to a cricitical heat for it to have an effect on the efficiency.

 

[Gary K]

Regarding the inverter, there is a graph about somewhere showing how efficiency drops off on SMA inverters as they get hotter. You don't necessarily need to get the inverter up to a cricitical heat for it to have an effect on the efficiency.

so the answer is yes? -the cooler you can make run the inverter, the more efficient it is.
Has anyone got this graph please?

 

[overland]

Yesterday (25th) was by far the hottest day here since we've had 16 Trina panels installed in a 3.68kw system with a loft installed inverter. The peak power they produce is normally 3.62, but today they peaked at around 2.90kw which I assumed was down to the higher temps. However, overall for May they have achieved their expected generation for the month already.

Additionally I noticed that the expected generation figure for the month of June is lower than May which I assumed was partly down to one less day and also higher temps reducing the system's performance.

 

[mdovey]

so the answer is yes? -the cooler you can make run the inverter, the more efficient it is.
Has anyone got this graph please?

You can find the temperature derating graphs for the various SMA inverters here - http://files.sma.de/dl/1348/Wirkungsgr_Derat-TEN112110.pdf

In general below an ambient temperature of 40 degrees the inverter will be running at its maximum efficiency for the input of the panels. Above 40 degrees ambient temperature you may see derating (exactly how much above depends on the inverter model) but you should get a warning (on the inverter display, on a sunny beam etc) when this happens. If you aren't getting this warning there isn't much point cooling the inverter from a performance perspective. Cooling may help lengthen the life of the inverter, however.

 

[FB.]

so the answer is yes? -the cooler you can make run the inverter, the more efficient it is.
Has anyone got this graph please?

The Power One Aurora family of inverters are roughly as follows:

Air temperature / output

35-40'C: no loss of output (cooling fan gradually increases its speed to cool the internals).
40-45'C: 5-10% loss of output
45-50'C: 15-20% loss of output
50-55'C: 30-40% loss of output
>55'C: major power reduction and risk of shutdown due to overheating

It is also recommended to avoid mounting in direct sunlight as the inverter casing will absorb sun/heat and get hotter than if the inverter is in shade or in a loft/garage/utility room etc.

 

[FB.]

so the answer is yes? -the cooler you can make run the inverter, the more efficient it is.
Has anyone got this graph please?

And the SMA/SB family of inverters state an operating range up to 60'C, so probably fairly similar to the Power One family.
Unfortunately I couldn't find more specific temperature-derating data.

 

[mdovey]

And the SMA/SB family of inverters state an operating range up to 60'C, so probably fairly similar to the Power One family.
Unfortunately I couldn't find more specific temperature-derating data.

Err, you could look at the link I posted a few posts back!

 

[sambotc]

I've seen the cooling fans running in January when the sun was out, but still cold in the loft (2.4-2.7kw production). It's the process itself rather than the ambient temp in the inverter location (although this will no doubt have an effect)

Presumably the fans will use a small amount running full whack as well, so there will also be a slight loss because of this?

 

[Gavin A]

You can find the temperature derating graphs for the various SMA inverters here - http://files.sma.de/dl/1348/Wirkungsgr_Derat-TEN112110.pdf

In general below an ambient temperature of 40 degrees the inverter will be running at its maximum efficiency for the input of the panels. Above 40 degrees ambient temperature you may see derating (exactly how much above depends on the inverter model) but you should get a warning (on the inverter display, on a sunny beam etc) when this happens. If you aren't getting this warning there isn't much point cooling the inverter from a performance perspective. Cooling may help lengthen the life of the inverter, however.

So HF and TL inverters generally shouldn't derate even in warm UK lofts (as long as the temperature stays below about 50), but the SB series aren't so good, with some derating from as low as 30 degrees.

I think somewhere in the mists of time I'd seen those graphs before and reached the same conclusion, which is partly why we've mostly been installing HF and TL inverters, but I'd then no found the document again so I couldn't remember the specifics.

 

[Gavin A]

The Power One Aurora family of inverters are roughly as follows:

Air temperature / output

35-40'C: no loss of output (cooling fan gradually increases its speed to cool the internals).
40-45'C: 5-10% loss of output
45-50'C: 15-20% loss of output
50-55'C: 30-40% loss of output
>55'C: major power reduction and risk of shutdown due to overheating

It is also recommended to avoid mounting in direct sunlight as the inverter casing will absorb sun/heat and get hotter than if the inverter is in shade or in a loft/garage/utility room etc.

do you know if that's true for both the indoor and outdoor versions?

useful to know that they start derating 10 degrees lower than the TL series sunny boy inverters, and so aren't as good for loft mounting.

 

[rgf1]

Gavin, can you tell me more about the Power One Aurora family of inverters. I am also finding the maximum power is far less than I expected. I now have learned from the forum that this is inevitable in very how ambient temperatures.

I can find nothing on the 3.6 Kw Aurora about cooling fans - I have also found an advert by them saying there is not fan. Who is right?

 

[FB.]

do you know if that's true for both the indoor and outdoor versions?

useful to know that they start derating 10 degrees lower than the TL series sunny boy inverters, and so aren't as good for loft mounting.

Yes, the indoor (fan cooled) and outdoor (heat-sink/fin-cooled) versions may see a reduction in output from 40'C air temperature.

But I suspect that it would be somewhat modified by the amount of DC power being processed at any given time, since that is also heat-generative within the inverter.

Knowing the manufacturers and the need to put a positive sales spin on their products, I would suspect that the >40'C temperature-induced power derating occurs at relatively low power throughput, and probably the inverter could de-rate at 35'C ambient temperature if it was running close to full capacity and therefore generating a lot of heat within itself.

 

[FB.]

Gavin, can you tell me more about the Power One Aurora family of inverters. I am also finding the maximum power is far less than I expected. I now have learned from the forum that this is inevitable in very how ambient temperatures.

I can find nothing on the 3.6 Kw Aurora about cooling fans - I have also found an advert by them saying there is not fan. Who is right?

The PVI-3600 and PVI-2000 indoor models use a cooling fan.

The PVI-3600-OUTD (outdoor) models use cooling fins/heat sinks; more suitable for dusty places like lofts, or, apparently, mounting in shady locations on the exterior of a building, since without a cooling fan they don't get clogged-up with dust (although the cooling fins may need occasional wipe with a duster).

 

[FB.]

do you know if that's true for both the indoor and outdoor versions?

useful to know that they start derating 10 degrees lower than the TL series sunny boy inverters, and so aren't as good for loft mounting.

For info, though, my PVI-3600 indoor model had the following performance efficiency at different ambient temperatures:

Indoor utility room temperature about 15'C (early April, mid-morning at peak generation)
Power in: 3.90kW
Power out: 3.66kW
Efficiency: 93.8%

Indoor utility room temperature about 25'C (in recent days, early afternoon, after peak generation but as temperatures are rising after mid-day)
Power coming into inverter: 2.90kW
Power going out of inverter: 2.71kW
Inverter efficiency: 93.4%

So no noticeable de-rating at 25'C air temperature (in the room where the inverter is located) when running at 80% capacity.

The 0.4% difference is probably merely down to inverters tending to run more efficiently the closer they get to their maximum rating.

 

[rgf1]

Thanks to FB for his quick and cogent reply.

My Aurora 3.6 OUTD is clearly is an outdoor model, installed in a completely dry, fairly dustproof stables - in the tackroom where the saddles, horse tack is installed. There is little evidence of any of the fins being clogged be contaminated by dust. My question is, how do I clean the finned heatsink if it needs it. Can I remove the outer case? or is there a simple method using a nozzle on a vacuum cleaner. (Has anyone fitted the fan off the indoor model and, can it actually be fitted, or has anyone fitted a simple external fan - like a computer fan and does this work enough to compensate its losses?)

The temperature of the heatsink does run about 60 degrees C - this is the temperature displayed on the screen. Is this the temperature used to measure and therefore control the derating or is the derating temperature at the measured at the internals of the unit?. In my manual it says - page 83 - full power is guaranteed up to ambient temperature of 55 degrees C.

Also, another question for the gurus of the forum to get their teeth into. Maybe this is another thread, or there already is a thread on this - please redirect me if so.

Page 81 of my manual - which is PVI-3.0/3.6/4.2-OUTD-UK Rev 1.0 states that the maximum recommended output is 4150 W - the absolute maximum I have ever recorded is a consistent 3984 W - this must be a clipped output as in the winter I get this value every sunny cold day.

Questions:

1. Why can I not get 4150 W
2. Is the inverter clipped to this to meet the UK regs.
3. If so how can the clipping be removed.

Many thanks to all replies

rgf1

 

[FB.]

Curious.

On the copy of the manual I'm looking at, page 83 has no temperature data.

But on page 90, the OUTD manual says:

"....

AURORA ensures top power up to 40°C ambient temperature, as faras it is not directly exposed to the sun......"

But on page 25 it says:

Temperatures exceeding 45°C for the PVI-4.2, 55°C for PVI-3.6
and 55°C per PVI-3.0, may result in power derating.
Output power derating is more likely to occur in case of high output
power and high ambient temperature.

 

[FB.]

Thanks to FB for his quick and cogent reply.

My Aurora 3.6 OUTD is clearly is an outdoor model, installed in a completely dry, fairly dustproof stables - in the tackroom where the saddles, horse tack is installed. There is little evidence of any of the fins being clogged be contaminated by dust. My question is, how do I clean the finned heatsink if it needs it. Can I remove the outer case? or is there a simple method using a nozzle on a vacuum cleaner. (Has anyone fitted the fan off the indoor model and, can it actually be fitted, or has anyone fitted a simple external fan - like a computer fan and does this work enough to compensate its losses?)

The temperature of the heatsink does run about 60 degrees C - this is the temperature displayed on the screen. Is this the temperature used to measure and therefore control the derating or is the derating temperature at the measured at the internals of the unit?. In my manual it says - page 83 - full power is guaranteed up to ambient temperature of 55 degrees C.

Also, another question for the gurus of the forum to get their teeth into. Maybe this is another thread, or there already is a thread on this - please redirect me if so.

Page 81 of my manual - which is PVI-3.0/3.6/4.2-OUTD-UK Rev 1.0 states that the maximum recommended output is 4150 W - the absolute maximum I have ever recorded is a consistent 3984 W - this must be a clipped output as in the winter I get this value every sunny cold day.

Questions:

1. Why can I not get 4150 W
2. Is the inverter clipped to this to meet the UK regs.
3. If so how can the clipping be removed.

Many thanks to all replies

rgf1

My 3.6kW Aurora indoor model is reporting internal temperature of 45'C with the air temperature around the inverter at 25'C and with 3kW of power going through it.
It is not de-rating at those temperatures and outputs - proven by the efficiency remaining in the 93-94% range just as it was a couple of months ago with the air around the inverter at about 15'C.

 

[moggy1968]

both the SB series and power one indoor inverter are less efficient than the TL, HF or outdoor versions.
They are older designs of inverter, some might say obsolete

 

[Gary K]

both the SB series and power one indoor inverter are less efficient than the TL, HF or outdoor versions.
They are older designs of inverter, some might say obsolete

I wish I had specced a 4000tl rather than the 3800 I have got, When I checked the info linked above ,I was surprised to see not only is it more efficient in general as to be expected but the 3800 drops off even more approaching max output and is most efficient at 30-50% of max knowing what I know now, and after checking the max outputs on PVoutput,I now believe the inverter could be equally if not more important than the panel choice

 

[moggy1968]

oh yes, to have a good quality system, you need a good quality system, not just decent panels. The components need to be properly matched.

You wouldn't put bicycle tyres on a ferrari!

 

[FB.]

I've had a max/min thermometer in the loft for the last few days, to see what the temperature variation has been.

Days have been sunny - sometimes a little haze or a thin cloud - with outdoor temperatures peaking around 26'C.
The loft has been peaking at 33'C; 7'C above ambient.

It should be borne in mind that about three-quarters of one roof aspect has now been covered by solar panels which will be blocking the sun's heat from that direction (but that would be the case with most solar arrays; it covers part of the roof and keeps the roof cooler).

If the roof hadn't been part-covered by panels then I suspect it would have added about 11'C to the loft temperature, taking it up to about 37'C in recent days (taking into account the additional unshaded roof and its likely ability to absorb the sun's heat).

I remember having to change the ball and float on the water tank in my old house on a hot sunny day about ten years ago and it was probably 40'C in the loft.

 

[babba]

I've had a max/min thermometer in the loft for the last few days, to see what the temperature variation has been.

Days have been sunny - sometimes a little haze or a thin cloud - with outdoor temperatures peaking around 26'C.
The loft has been peaking at 33'C; 7'C above ambient.

It should be borne in mind that about three-quarters of one roof aspect has now been covered by solar panels which will be blocking the sun's heat from that direction (but that would be the case with most solar arrays; it covers part of the roof and keeps the roof cooler).

If the roof hadn't been part-covered by panels then I suspect it would have added about 11'C to the loft temperature, taking it up to about 37'C in recent days (taking into account the additional unshaded roof and its likely ability to absorb the sun's heat).

I remember having to change the ball and float on the water tank in my old house on a hot sunny day about ten years ago and it was probably 40'C in the loft.

So you should not put inverters in lofts ?

 

[Earthstore]

So you should not put inverters in lofts ?

This has been covered before, but just a little refresher; Most installs differ, and it can be a case of saftey over performance, DC cables have no mechanical protection, no trips, nothing. So if you feel it is safe to route DC cables through a house with 600V+ maybe to gain a little performance on a hot day, then that is your decision, As FB stated the temperature in his loft was 33 degrees, so he was way below the 40 advised for the SMA, the biggest loses come from the panel temperature and not the inverter.

 

[FB.]

This has been covered before, but just a little refresher; Most installs differ, and it can be a case of saftey over performance, DC cables have no mechanical protection, no trips, nothing. So if you feel it is safe to route DC cables through a house with 600V+ maybe to gain a little performance on a hot day, then that is your decision, As FB stated the temperature in his loft was 33 degrees, so he was way below the 40 advised for the SMA, the biggest loses come from the panel temperature and not the inverter.

I agree. The panel output is about 20% lower in this weather, so the inverter won't be running as hot, so may cope with a slightly higher temperature than expected.

I have a P1 Aurora indoor model in a downstairs utility, which hasn't exceeded the outdoor 26' temperature in recent days, whereas it would have been 33'C in the loft (perhaps another degree or two warmer in the loft due to the inverter's own heat generation; the inverter adds a few degrees of heat to my cool downstairs utility room), but even if it was in the loft, it should still be within its 40'C de-rating.
Mine runs two strings, running down the outside of the house in a conduit on the wall between mine and the neighbours house, so out of view. Cables come back into the house next to the inverter, into two DC isolators. The DC voltages are about the same as mains voltages (200-300V DC, 7-8A current at full power).

I find the inverter's location convenient for taking a look at what it's doing without needing a desktop monitoring device (costing £100+ and which may only last a few years like most electronics) - the screen on the indoor model is more useful than certain other inverters - pic below:

 

[babba]

DC cable can be protected, I think the inverter will be working its socks of in tempetures stated by FB and is sure to shorten its life

 

[moggy1968]

DC cable can be protected, I think the inverter will be working its socks of in tempetures stated by FB and is sure to shorten its life

All things being equal we will mount the inverter somewhere other than the loft, but sometimes it's inevitable. SMA HF and TL inverters are rated to 60C so typical loft temperatures are well within it's operating range.