diy solar

diy solar

Rooftop Solar on the Australian Grid

Also these inverters have an inner bypass (AC in to AC out). That can transfer 40-50A ... per phase.
If not enough you can parallel 6 of them.
So there is really no limitation in residential usage.
The limitation is cost. Why over capitalise with multiple inverters when one will do? Why have extra inverters just to pass enough grid power through when one normal grid tied inverter can do the job anyway?

In our cities the grid is very reliable, so most are not that concerned with an expensive backup option. But grid tied solar - the financial benefit is a no brainer in our market. That's why nearly 30% of homes have one.

Where I am though we get fairly regular outages, hence building my off-grid system which can supply power to our 3 buildings via a transfer switch when needed.

If I really want I could leave the transfer switch in the backup position and the off-grid system can operate in utility pass through mode with UPS cut over if the grid supply is cut. Meanwhile non-critical loads are not supplied via the backup, and if the grid comes back they will too, including the grid tied inverter.

the most common question that comes up weekly is "I have a solar system and still sitting in the dark in a blackout ... what can I do against it"
Yes it's an often misunderstood concept.

I have no illusions that my grid tied inverter will go off-line when the grid does. But rather than spend $15,000 to have a grid tied battery system with grid isolation and backup capability which will never payback in my lifetime, I built an off-grid system to provide backup for a fraction of the cost. As a bonus I've taken some load off-grid (pool pump) to use the otherwise spare PV capacity.
 
The limitation is cost. Why over capitalise with multiple inverters when one will do? Why have extra inverters just to pass enough grid power through when one normal grid tied inverter can do the job anyway?
I think 50A / phase bypass is more than enough for residential use.
That is 150A for this 3 phase unit. I do not think you use 33kW in any given time in a home :)
Also this is only limits the protected load side. The green unprotected load side you can have as much as your main can handle.


In our cities the grid is very reliable, so most are not that concerned with an expensive backup option. But grid tied solar - the financial benefit is a no brainer in our market. That's why nearly 30% of homes have one.

Where I am though we get fairly regular outages, hence building my off-grid system which can supply power to our 3 buildings via a transfer switch when needed.
Or you could build a hybrid system.
One other hybrid solution is the grid-tie + UPS solution. I am doing this myself. I have to because the EVE cells I can not put outside ... and the solar array is far from the house. Also the off-grid inverter is an 5048MK and that is a double conversion Online UPS (own clear sinus wave)

grid_tie_UPS.jpg

As you can see here on the exactly same spots the green unprotected load connects, and the purple protected load connects.
The only difference is that the grid-tie is a separate unit. This system has cons and pros against the one unit hybrid.

If I really want I could leave the transfer switch in the backup position and the off-grid system can operate in utility pass through mode with UPS cut over if the grid supply is cut. Meanwhile non-critical loads are not supplied via the backup, and if the grid comes back they will too, including the grid tied inverter.
Even more the MPP Solar inverters not only do bypass ... they doing it with power production on that line.
Have inner limiter. So power comes in into the inverter from grid, PV power comes in ... and out goes to the house all the PV power and a little bit from the utility power.

Yes it's an often misunderstood concept.

I have no illusions that my grid tied inverter will go off-line when the grid does. But rather than spend $15,000 to have a grid tied battery system with grid isolation and backup capability which will never payback in my lifetime, I built an off-grid system to provide backup for a fraction of the cost. As a bonus I've taken some load off-grid (pool pump) to use the otherwise spare PV capacity.

Why $15.000 battery?
These one unit Hybrid inverters use the same 48V battery.

hibrid_panel.jpg
You buy 16 cells of EVE/Lishen/CATL and you have your battery for it.
Like $3000 for 16pcs of 280-300Ah cells with BMS.

The inverter is not cheap but compared to Sol-Ark it is

A one phase 5kW inverter is $1400

A 3 phase 10kW inverter is $2390

And you connect your EVE cells to it and your Solar panels and the main and protected load panel ... and your system is ready.
(of course you will need the contract with the grid utility company ... but it is exactly like with a grid-tie inverter ... since it is a grid-tie inverter for them)
 
Last edited:
Like $3000 for 16pcs of 280-300Ah cells with BMS.
$15k is the minimum cost of decent sized grid tied 3-phase battery system with backup capability, professionally installed and meeting all battery connection standards.

Yes I can go DIY for a lot less, but once you go grid tied = much greater compliance costs and DIY is no longer a sensible option. Even at $3k for cells and a BMS there will be quite an expense to make it compliant for grid connection. Must be installed by qualified professional. Approvals etc.

Off-grid there are still requirements but not such an issue. Let's say it's $4k all up. Once you are off-grid the capacity utilisation drops. I don't know why I'd bother - the payback on that for us is still going to be 15+ years. Battery economics sucks here.

Plus I already have a great grid tied inverter, I'm not going to replace it.

Hence I went with a low cost solution. 2nd hand solar PV and rails, low cost off-grid inverter, repurposed SLA batteries from data centre, transfer switch which creates a single phase supply for essential loads and isolates them from the grid. Costs about the same as the generator I used to use (and which now provides redundancy).
 
I've done quite in-depth modelling of battery economics for us - I have 3 years of 5-min interval data on consumption, production, import and export and built a battery simulation model. I can simulate a battery with given specifications (capacity, reserve setting, charge/discharge limits, round trip efficiency) and I'll know how it would perform, technically and financially.
 
$15k is the minimum cost of decent sized grid tied 3-phase battery system with backup capability, professionally installed and meeting all battery connection standards.

Yes I can go DIY for a lot less, but once you go grid tied = much greater compliance costs and DIY is no longer a sensible option. Even at $3k for cells and a BMS there will be quite an expense to make it compliant for grid connection. Must be installed by qualified professional. Approvals etc.
This part I do not understand.
Yes they are extremely complex and expensive battery systems like LG Chem Resu 10h (400Vdc). I wrote (mostly translated to English) a post about it and its dissection, parts, etc ...

Usually utility does not care about battery, only the inverter that is connected to the grid (at least here).
What criterias and approvals are there ?

But there are simple battery solutions like Pylontech 48Vdc battery systems. Really not more than cells and BMS in a box. Has all the UL, CE, ... certificates. Plug-and-play assembly.
And from that one 3,5 kWh 48V unit is €1600. Also can be paralleled.
Also ideal for MPP Solar off-grid and one unit Hybrid inverters. They even have the setting in them, communicate directly with Pylontech BMS.


Off-grid there are still requirements but not such an issue. Let's say it's $4k all up. Once you are off-grid the capacity utilisation drops. I don't know why I'd bother - the payback on that for us is still going to be 15+ years. Battery economics sucks here.

Plus I already have a great grid tied inverter, I'm not going to replace it.


Hence I went with a low cost solution. 2nd hand solar PV and rails, low cost off-grid inverter, repurposed SLA batteries from data centre, transfer switch which creates a single phase supply for essential loads and isolates them from the grid. Costs about the same as the generator I used to use (and which now provides redundancy).

That is the grid-tie + UPS hybrid solution.
So you already building it.



I've done quite in-depth modelling of battery economics for us - I have 3 years of 5-min interval data on consumption, production, import and export and built a battery simulation model. I can simulate a battery with given specifications (capacity, reserve setting, charge/discharge limits, round trip efficiency) and I'll know how it would perform, technically and financially.
I simply use the EU's solar calculator. Can be used almost anywhere in the globe (except most of Australia where it does not have stored radiation data so only calculates based on latitude).
https://re.jrc.ec.europa.eu/pvg_tools/en/#PVP
It has on-grid and (separated) off-grid calculators. And gives pretty precise forecasts.
 
Grid tie means it can push power to the grid. This cannot do that, only draw power from the grid.
The grid-tie part (inverter) does the selling/pushing ... the off-grid/UPS part (inverter) the battery based constant power.
It is a 2 unit hybrid system :) See previous pic.
 
What criterias and approvals are there ?
For a start there is the Australian Standards documentation for battery installations. It's a long and detailed document behind an expensive paywall which I have no intention of purchasing for you. Also, you can't just add inverters to grid tied connections. Adding inverter capacity is not always permitted. To add a battery in such situations mean replacing perfectly functioning equipment with a even more expensive hybrid.

In lieu of the actual standards documentation here's an article however which might shed some light on the standards in place:

It's not a particularly friendly regulatory environment for DIY systems with enough power and capacity for home use. Small hobby kits are no big deal.
 
For a start there is the Australian Standards documentation for battery installations. It's a long and detailed document behind an expensive paywall which I have no intention of purchasing for you.
No need. If I want to read it I ask a colleague from there to send me. I do not care enough to read 500+ pages of raw material. The summary was enough ;)
It seems to be over protective. You can not have a DiY 48V system, because for that you need at last 60V MPPT and solar panels. And that has to be done by electrician.



Also, you can't just add inverters to grid tied connections. Adding inverter capacity is not always permitted. To add a battery in such situations mean replacing perfectly functioning equipment with a even more expensive hybrid.
You do not add the off-grid inverter to the grid-tie inverter ....
They are SEPARATE units working separately. They are not AC Coupled.
The off-grid can not generate power to the grid.


In lieu of the actual standards documentation here's an article however which might shed some light on the standards in place:

It's not a particularly friendly regulatory environment for DIY systems with enough power and capacity for home use. Small hobby kits are no big deal.

It is a really DiY hostile regulation. With only two 36V solar panels in series you already need an electrician.
Does not differentiate between NMC, LiPo and LFP batteries.
Who wrote this? The coal industry ?
 
The grid-tie part (inverter) does the selling/pushing ... the off-grid/UPS part (inverter) the battery based constant power.
It is a 2 unit hybrid system :) See previous pic.
I get that but it would mean the battery can only supply the circuits on the "essential" loads board.

Apart from the grid backup scenario, I see little point in buying an expensive lithium chemistry battery and control system for it to only be able to supply a portion of our loads. Battery economics suck enough as it is without dropping capacity utilisation down the toilet.

I have 3 buildings to supply. For a "permanent" installation with an off-grid UPS inverter feeding essential loads we could probably squeeze by with a 3-phase 8kW hybrid UPS inverter to make sure it's not a supply choke point, but 10kW might be safer.

Plus a complete rewire of main circuit board to split the loads. Plus a lithium battery system now connected in a manner which will take decades to payback (since it is not able to supplying energy to all loads, especially the power hungry ones). And then I'd also need to add extra off-grid PV capacity if I want to cycle such a battery daily instead of having it idle as standby backup with just enough PV for a top up charge.

Don't get me wrong, I've thought quite a bit about exactly what you propose but the numbers just don't stack up. At least not yet.

For now we'll live with a (cheap) 4kW off-grid hybrid inverter with 18.2kWh of (2nd hand data centre) SLA batteries and 2.2kW of (pre-loved) PV to cover us during outages and the occasional/special circumstances where we might choose to operate it in UPS mode (e.g. planned grid outages or when the grid is flakey due to storms). The small PV system keeps the pool pump running and batteries in a state of readiness for the inevitable outage. Summer is coming, it's only a matter of time before storms take out a local power line somewhere.

I have the sparky coming tomorrow to complete some of the wiring set up so the UPS mode will operate correctly (should be an easy fix but it has to be a sparky to sort it, just like I had him do the AC output wiring). Only other thing I am doing after that is upgrading the fire protection of the battery/inverter housing and to also upgrade the battery circuit protection with better quality fusing and disconnects than I currently have in place.
 
I get that but it would mean the battery can only supply the circuits on the "essential" loads board.
Exactly. And not :)
You can separate (L+N) your house circuits and choose with 1-0-2 switch who should give power to that load.
eaton_1_0_2.jpg

Like I separated the cooktop and usually use it from grid. But if there is a big blackout I can switch it manually to UPS.
The coffee machine is constantly on UPS :LOL:
 
LOL, I wouldn't put it past them.

Pretty much anyone putting together a 48V system here is walking a fine regulatory line. I had a qualified electrician install key parts of the wiring.
And I think electricians do not grow on trees there also ... so the already overused guys have now a lot of unnecessary low V work to do.
 
Like I separated the cooktop and usually use it from grid. But if there is a big blackout I can switch it manually to UPS.
Nice thought but no real need. The only circuits deemed "non-essential" and which remain on the grid side only are those which have a dedicated circuit with too high a power draw for the off-grid system to supply. Off-grid supply limit is 4kW. Ample power for outage coverage, but not sufficient for ongoing regular household operation.

Non essential circuits includes oven, induction stove, ducted aircon, another large aircon system, plus of course the 3-phase grid tied inverter and the circuit providing power to the off-grid inverter (can't have that feeding power to itself). Resistive element hot water storage system as well - it's a utility switched controlled load and a separately metered circuit already.

Else every GPO and light has backup. The backup also extends to supplying power via 3-phase sub-mains to two other buildings, and the high power draw appliances in those buildings will have backup power available, so they need to be manually managed. Not difficult but it would be problematic if we were operating in UPS mode and there was an outage while these high power draw appliances were on. Hence use of the UPS mode will be for specific circumstances.

It's possible to change the 3-phase sub-mains over to individual phase switching to enable isolation of one or two phases in each outbuilding. But for now I have a functioning backup which is simple to operate and hasn't cost the Earth to build.

Enough juice in the batteries to cover ~15 hours of typical essential loads to 50% DOD, plus the solar PV supplemental supply to extend that and the generator can be fired up to recharge batteries / pass through power if we need for a longer outage. Most outages are a few hours. Occasionally much longer.

One of the reasons for having just one transfer switch is the simplicity of operation for others. If I am not here, they know all that's required is to flip one switch and backup will be operating immediately. All the circuits which needs to be isolated from backup have been taken care of, with exception of a few things in the outbuildings (and they know that) they know to ensure are off.

And I think electricians do not grow on trees there also ... so the already overused guys have now a lot of unnecessary low V work to do.
Yeah, they are stupid busy. Unless its an emergency, it takes months to get a service appointment. Trades of all kinds here are flat out. We have a 1+ year wait for a builder.
 
Well the records keep coming. Briefly today the Australian national grid hit >60% renewable energy supply, and on a Monday no less.

Screen Shot 2021-09-20 at 8.54.54 pm.png
 
Solar contributed 36%, including 30% from rooftop.
Does the utility network have a way to command curtailment by rooftop PV? Or something like pumped hydro to store excess?
Or, are they able to accept whatever solar produces and adjust coal/gas power plant output accordingly?

I think California once reached 50% solar, one sunny spring day with mild weather.

You have 19% wind. Some US regions (Oregon) have had to idle wind turbines when grid didn't need the power.
 
Wind and solar farms are curtailed all the time. Coal ramps down/up as it can but that has limits. At the time of that 60% mark there would have been a lot of solar and wind farm curtailment.

Rooftop is basically a free for all, with the main control being limits on system sizes and how much power a system may export. A typical per phase limit for export is 5kW but it varies by distribution zone. Increasingly some centralised control of rooftop solar PV is being introduced and new inverter standards now apply in the state of SA where their grid for the whole year averages 60% renewable (no hydro). It will give the ability for export/production to be dialled back during times of extremely low grid demand. The quid pro quo is they are permitting those homes to have a much larger system and higher export limit which will be more useful during most of the year when solar production is lower. It's how they are going to get even more renewables into the grid - by enabling rooftop curtailment when needed, as well as grid scale management.
 
Another new record in South Australia:

Rooftop solar helps send South Australia grid to zero demand in world first

South Australia on Sunday became the first gigawatt scale grid in the world to reach zero demand when the combined output of rooftop solar and other small non-scheduled generators exceeded all the local customer load requirements.

The fact that South Australia’s demand could fall to zero, and even into negative territory, was flagged as a possibility this spring by the Australian Energy Market Operator, who noted it would be the first gigawatt scale grid to do this and said rooftop solar alone could exceed all local demand.

aemo-negative-demand-south-australia-scaled.jpg
 
Back
Top