Pv solar install the inverter in loft or garage

don0301

grahamc2003 wrote: »

Plans are to have a board in the loft to take the inverter plus other bits. Think I'll try to change that and have it above the fuse box in the hall. Not sure what inverter I'm getting, but I think I should really be going for the solaredge system due to the shading I'll get - too late now, but may have it retrofitted with some more panels. Mine must be the cheapest system ever, but had to take what was offered and in their stock, so couldn't have a larger system due to them not having a suitable inverter to take a 1.75kw string and a 1.25kw string (is there a problem with that needing a special inverter?). Scaffolding up - not suitable for painting either the wall or eaves, which I hoped to get done.

don't forget to get them to make the dc cable extra thick

grahamc2003 wrote: »

But the dc in will be pretty thick to carry 4kW at 20 or 30V

Dave_Fowler

I'm not sure why there is all this emphasis on making the DC cables as thick as possible. It is more important to make the AC cables thicker.

Take a typical 4kWp PV installation with 2 strings of 2kWp. The DC voltage rating of most inverters running at 4kW is around 400V - 500V DC. i.e. each string has a maximum DC current of 5 Amps. The AC output of 4kW requires a current of 16 Amps. Now as the losses in a cable are due to the square of the current through it, surely we should emphasize that the AC cable is as thick as possible.

Dave F

don0301

Dave_Fowler wrote: »

I'm not sure why there is all this emphasis on making the DC cables as thick as possible. It is more important to make the AC cables thicker.

Take a typical 4kWp PV installation with 2 strings of 2kWp. The DC voltage rating of most inverters running at 4kW is around 400V - 500V DC. i.e. each string has a maximum DC current of 5 Amps. The AC output of 4kW requires a current of 16 Amps. Now as the losses in a cable are due to the square of the current through it, surely we should emphasize that the AC cable is as thick as possible.

Dave F

rofl rofl rofl

you been paying attention to grahamc2003?

don0301

Dave_Fowler wrote: »

The silicon used in the manufacture of semiconductors is relatively stable up to a temperature of 150C. Above that temperature several failure mechanisms occur(mainly due to changes within the semiconductor junction). At temperatures up to 150C, failures due to temperature in semiconductors are mainly due temperature cycling where fractures in bonding wires and substrates can occur. One should note however that the temperature of the silicon in a device under load is not that of the ambient, but is higher depending on the power dissipated. A typical power MOSFET, as used in an inverter, will have a maximum junction temperature of 150C. It may be rated at 250W with a case temperature of 25C, but there is a 0.5C/W temperature gradient between case and junction. So an inverter working in an ambient temperature of 50C with an internal temperture perhaps 20C higher and heatsinks another 25C higher would have the devices running with a case temperature of 95C. This would allow a maximum junction-to-case temperature of 55C. This equates to a maximum power dissipation of 110W.

This reduced power due to higher ambient temperatures is why many inverters have to run at a lower power when they are hot. It is ironic that just when the inverter has a lot of DC power to convert to AC that the ambient is likely to be hot and they are unable to work at full power.

For other components the effects of temperature are different. The most affected components are the capacitors, particularly the electrolytics. Plastic film capacitors suffer from film degradation. A good film capacitor will have a maximum rated temperature of 120C but at this temperature its MTBF will be reduced. Electrolytic capacitors contain a liquid (or gell) electrolyte. Over time, even at relatively low temperatures the electrolyte will degrade. The oxide coating of the plates slowly disolves into the electrolyte and also, because for safety's sake they are usually vented, this degradation increases with temperature.

Resistors are relatively stable up to high temperatures, but their life is again shortened by high temperatures. This is particularly the case with high power wire-wound resistors where the wire deteriorates at a higher rate as they get hotter. Run a house-hold incandescent lamp on 100V and it will last 'forever' but run it hotter at 250V and it will usually fail within 1000 hours.

Whilst it is true that the lower the temperature, the longer the electronics will last, the high temperatures often found in a loft are usually associated with a time when there is maximum power available from the sunlight. This causes the maximum temperature rise within the inverter (usually an increased fan speed helps reduce the internal rise). So if only to allow the inverter to work at full output when the sun is at its strongest then it is surely true that an inverter situated in a lower (and more stable) ambient temperature is best.

Dave F

well googled

that'll be an argument for a higher ambient temperature? rofl

zeupater

Dave_Fowler wrote: »

I'm not sure why there is all this emphasis on making the DC cables as thick as possible. It is more important to make the AC cables thicker.

Take a typical 4kWp PV installation with 2 strings of 2kWp. The DC voltage rating of most inverters running at 4kW is around 400V - 500V DC. i.e. each string has a maximum DC current of 5 Amps. The AC output of 4kW requires a current of 16 Amps. Now as the losses in a cable are due to the square of the current through it, surely we should emphasize that the AC cable is as thick as possible.

Dave F

Hi Dave

The only reason to have thicker cables in our installation was to compensate for the longer runs from the first/last panels on each string on the roof to the garage without increasing the resistance losses in the DC cables ..... from memory, using 6mm cable on our system allows a 50% longer run whilst maintaining the same losses as 4mm ....

HTH
Z

Dave_Fowler

zeupater wrote: »

Hi Dave

The only reason to have thicker cables in our installation was to compensate for the longer runs from the first/last panels on each string on the roof to the garage without increasing the resistance losses in the DC cables ..... from memory, using 6mm cable on our system allows a 50% longer run whilst maintaining the same losses as 4mm ....

HTH
Z

Hi Z,

A very good reason. If you have a long cable run - AC or DC the losses are proportional to the length of the cable and to the inverse square of the cable diameter.

Dave F

don0301

zeupater wrote: »

Hi Dave

The only reason to have thicker cables in our installation was to compensate for the longer runs from the first/last panels on each string on the roof to the garage without increasing the resistance losses in the DC cables ..... from memory, using 6mm cable on our system allows a 50% longer run whilst maintaining the same losses as 4mm ....

HTH
Z

still waiting for your cable length measurements, so we can all calculate your losses

don0301

Dave_Fowler wrote: »

Hi Z,

A very good reason. If you have a long cable run - AC or DC the losses are proportional to the length of the cable and to the inverse square of the cable diameter.

Dave F

better tell the National Grid that

Dave_Fowler

don0301 wrote: »

well googled

that'll be an argument for a higher ambient temperature? rofl

Google has nothing to do with it. Don't insult me.

I have been working in defense and industrial electronics as a design engineer, project manager and latterly (for the last 26 years) as a self-employed electronics circuit designer. Reliability of components and circuits has been of major concern to me since the mid 1960's.

I'm not sure if you think the temperature will not cycle if the temperature is above 150C or if you think the failure of wire bonds or substrates will not occur above 150C. My statement was to say that failures in semiconductors at junction temperatures below 150C are mainly due to temperature cycling.

Dave F

don0301

Dave_Fowler wrote: »

Google has nothing to do with it. Don't insult me.

I have been working in defense and industrial electronics as a design engineer, project manager and latterly (for the last 26 years) as a self-employed electronics circuit designer. Reliability of components and circuits has been of major concern to me since the mid 1960's.

I'm not sure if you think the temperature will not cycle if the temperature is above 150C or if you think the failure of wire bonds or substrates will not occur above 150C. My statement was to say that failures in semiconductors at junction temperatures below 150C are mainly due to temperature cycling.

Dave F

Grats

so your saying an inverter internal temperature is sometimes/often/always above 150C?