My thinking was the same as yours - keep the battery on the DC side as there's one fewer AC/DC conversion loss.

In my experience, the one-way loss is around 5% and round trip loss of 10%. So you can add 10% to the cost of the DC kit for comparison I suppose. When I bought my system, I wasn't too fussed about the cost and there wasn't really much difference anyway. I just wanted the better, more efficient system, so got the DC version.

There is an additional consideration if the DNO has to be consulted. An AC battery can discharge to the grid independently of the solar panels which could also be discharging to the grid. The combined export might be more than allowed by the DNO. This is not the case with DC batteries as I understand it and the DNO was happy for me to have DC battery but not AC with my second solar array.

As well as being more efficient, DC batteries have advantages when solar export is being clipped by the inverter, see below:

https://www.batterypowertips.com/when-does-dc-coupling-maximize-the-performance-battery-storage-plus-solar-faq/

@bontwoody thank you - very much food for thought. 

I had just about clocked onto the batteries charging in clipping situations but hadn't stumbled across the early and late tails part of the equation. Do you know if the presence of optimisers in a PV system materially changes this?

My line of thought over the last few days was starting to move towards microinverters and AC coupled battery storage (squaring of that in winter and/or if playing games with the likes of Agile) a good proportion of my battery charging would likely be from grid regardless. 

I have a corner of the roof that will be regularly shaded by a tree until lunch / early afternoon in early/late summer and had read horror stories that, while modern solar panels are getting increasingly good at reacting to shading using bypass diodes, that regular shading increases the risk of diode and panel failure. In some cases manufacturers determining that known shading issues counts as an installation issue, therefore no warranty support, unless otherwise protected by microinverter/optimiser.

No one looks to be questioning the reliability of the Enphase microinverters. However, the same cannot be said for most of the optimiser manufacturers out there. Albeit the system I am otherwise considering is Huawei based and failure rates seem pretty low on their optimisers.

I do have concerns that microinverters would introduce an element of clipping (that an AC coupled battery obviously can't use) and (certainly US) installers seem to say that they size to ensure good generation at low light conditions (so I would worry that by upping the top end limit e.g. using IQ8H, I might be giving up some low end production - but I've struggled to find data I can interpret on this).

Another driver for this train of thought was microinverter lifespan closely matches panel lifespan and the AC battery choice is completely separate. The worrier in me has concerns that (on a DC coupled, hybrid inverter, setup) when the likes of the battery, BMS, or the string inverter requires replacement that you could end up stuck in a loop of potentially struggling to match parts into your existing ecosystem, or missing out on being able to make use of the latest innovations, or even having to commit to replacement of a whole suite of components once the first of the bundle fails.

Is there a general consensus that as long as sticking to well known/adopted brands that compatible kit will still be available 10-15 years down the line?  

Also still looking for an answer to the question what happens when an optimiser fails? (assuming like Tigo or Huawei that the optimiser in question is one that doesn't need to be on all panels of an array) would this generally just be a loss in optimiser function rather than a complete loss of generation from the associated panel?

Happily I do keep swinging back to a position that the Huawei setup I've been quoted has been sensibly pieced together.

Posted by: @bontwoody

↑

There is an additional consideration if the DNO has to be consulted. An AC battery can discharge to the grid independently of the solar panels which could also be discharging to the grid. The combined export might be more than allowed by the DNO. This is not the case with DC batteries as I understand it and the DNO was happy for me to have DC battery but not AC with my second solar array.

As well as being more efficient, DC batteries have advantages when solar export is being clipped by the inverter, see below:

https://www.batterypowertips.com/when-does-dc-coupling-maximize-the-performance-battery-storage-plus-solar-faq/

Please don't talk about AC Batteries - there is no such thing.

The article included concerns a grid size solar PV and Battery Storage system, in a domestic system you would probably be talking about a few watts of unharvested solar output, which is of little concern. Experience from my own 4kWp solar PV system would indicate that actual generation starts at only a few watts.

Of much greater concern is the need to ensure that there is minimum loss of generation due to panel shading or full output clipping.

@derek-m apologises I was in a rush and abbreviating. I will insert 'coupled' into the relevent paces in my post 🙂 Update seems I cant actual edit the offending post now

Posted by: @declan90

↑

@bontwoody thank you - very much food for thought. 

I had just about clocked onto the batteries charging in clipping situations but hadn't stumbled across the early and late tails part of the equation. Do you know if the presence of optimisers in a PV system materially changes this?

My line of thought over the last few days was starting to move towards microinverters and AC coupled battery storage (squaring of that in winter and/or if playing games with the likes of Agile) a good proportion of my battery charging would likely be from grid regardless. 

I have a corner of the roof that will be regularly shaded by a tree until lunch / early afternoon in early/late summer and had read horror stories that, while modern solar panels are getting increasingly good at reacting to shading using bypass diodes, that regular shading increases the risk of diode and panel failure. In some cases manufacturers determining that known shading issues counts as an installation issue, therefore no warranty support, unless otherwise protected by microinverter/optimiser.

No one looks to be questioning the reliability of the Enphase microinverters. However, the same cannot be said for most of the optimiser manufacturers out there. Albeit the system I am otherwise considering is Huawei based and failure rates seem pretty low on their optimisers.

I do have concerns that microinverters would introduce an element of clipping (that an AC coupled battery obviously can't use) and (certainly US) installers seem to say that they size to ensure good generation at low light conditions (so I would worry that by upping the top end limit e.g. using IQ8H, I might be giving up some low end production - but I've struggled to find data I can interpret on this).

Another driver for this train of thought was microinverter lifespan closely matches panel lifespan and the AC battery choice is completely separate. The worrier in me has concerns that (on a DC coupled, hybrid inverter, setup) when the likes of the battery, BMS, or the string inverter requires replacement that you could end up stuck in a loop of potentially struggling to match parts into your existing ecosystem, or missing out on being able to make use of the latest innovations, or even having to commit to replacement of a whole suite of components once the first of the bundle fails.

Is there a general consensus that as long as sticking to well known/adopted brands that compatible kit will still be available 10-15 years down the line?  

Also still looking for an answer to the question what happens when an optimiser fails? (assuming like Tigo or Huawei that the optimiser in question is one that doesn't need to be on all panels of an array) would this generally just be a loss in optimiser function rather than a complete loss of generation from the associated panel?

Happily I do keep swinging back to a position that the Huawei setup I've been quoted has been sensibly pieced together.

Sorry Declan, those sorts of issues are beyond my level of expertise. I would hope that companies like Solaredge and Huawei would maintaince parts for 15 year old installations, but who knows!

@bontwoody I can also see the valid point that the DNO could understandably worry about independent discharge to grid.

Easy to accidentally start thinking the kW numbers we're talking about as 'small', i.e. comparable to a domestic EV charger or 'its a domestic array and a battery in my garage after all' but when 3.68 kWh is 16A I can see that a 6-7 kW array plus a battery with reasonable discharge rate soon adds up! 

80A (my incoming fuse rating) is just over 18kW.

Would be quite conceivable to be asking to run 50A back to grid in an AC coupled scenario. 

My gut says ask for too much and you risk getting limited back further than if you'd asked for a more reasonable amount in the first place ... 

@declan90 Too true, balancing the grid must be a tricky job. My experience was that the DNO wasnt overly keen on getting into a conversation with me, although I did get an answer. The solar installer contacted him and agreed what was acceptable, which to be fair is working well. Im trying to get my tariff switched now so I can charge my battery at cheap times and export to the grid.

Posted by: @bontwoody

↑

@derek-m apologises I was in a rush and abbreviating. I will insert 'coupled' into the relevent paces in my post 🙂 Update seems I cant actual edit the offending post now

Don't worry, I obviously knew what you meant, but I still think it is important that the correct terminology is used so that there is little room for confusion.