Pushing the Residential Energy Boundaries (and my own buttons)

Bryan J. Rollins
6 min readMay 20, 2024

When I finally bought a home, I was excited to be able to “electrify” my home and power it entirely from renewable energy generated on the property.

I decided to see what asking for “everything” meant in the world of home electrification.

There were a few exciting things about the property I had purchased:

  1. The property was already mostly “off-grid.” It didn’t have city gas, city water, and could not get a good NBN (hi-speed internet connection). So I had gas bottles (soon to be removed once I get an induction range), water tanks, and Starlink.
  2. The shed on the property could fit four buses in it, and had a North facing roof, and could fit 30kW of solar panels, the maximum allowed in Australia for residences. I could turn the shed into the “power plant” for the property.
Unloading the gear begins at the shed. Pretty cool that the entire system fits neatly on the trailer.

My list of requirements:

  1. 30 kW of solar panels and 40 kwH of battery storage
  2. Use the shed as the power plant. All solar on the roof, all batteries, inverter, etc inside.
  3. Three-phase power — I wanted to use three-phase power on the EV charger and install a three-phase heat pump (usually they are lower overall energy consumption). During the day if there is enough solar three-phase appliances should never need to touch the grid.
  4. 100% seamless backup (i.e. power when off-grid). Specifically:
    - backup of all circuits on the property (most systems approach it by being only single phase backup, and recommending that you only back up certain circuits to maximise your battery usage).
    - any three-phase appliance will function, specifically the EV charger.
    - solar panels will be able to deliver energy to the house when off-grid
    - solar panels will be able to charge the batteries when off-grid (this was a tricky one that most systems could not handle)
  5. Seamless failover to the backup system. When grid power is lost or turned off, I don’t want a flicker.
  6. Be able to sell energy back to the grid from excess solar.
  7. As simple as a design as possible (avoid split systems, more than one inverter)
  8. All equipment should be 1 meter off the shed floor (it’s a flood zone out here)
  9. Optional: Be able to sell energy back from the batteries, ideally through a simple policy configuration or VPP program.
Installing the modules (panels) on the rails.

The System

After all that, the system I ended up with was:

  • 66 Trina 430 Watt panels for a total of 28.38kW, so not quite 30kW
  • 3 stacks of 4 Goodwe Linx batteries LX F3.3-H for 39.6kWh of storage
  • Goodwe GW29.9KW inverter
  • Tesla Wall Charger (first generation in Australia, I had a free one either come with the car)
  • Over 100 meters of three-phase cabling

The inverter is basically a small commercial inverter (or extra-large residential). Despite the amount of cable deployed, it makes the system a lot like what I wanted: a power plant in my shed.

What I learned about why this is hard

  • Almost all backup systems are single phase. For example, Telsa Powerwalls only supported backing up one phase. Fine for most use cases, but again I’m trying to build something cutting edge.
  • Most backup approaches won’t charge the batteries during an outage.
  • The list of batteries supported by VPPs is not long in Australia, and none of them worked for my requirements.
  • This system is right in between large residential and small commercial. As a result, it’s not cheap.
  • Most solutions presented to me were complex. Panels on the house and the shed, different backup circuits, i.e. complex and still didn’t meet all the requirements.
  • No one I talked to had done a system like this. So a few first proposals were just wrong about the capabilities of the design and I had to do a lot of homework, talk to friends, read articles on SolarQuotes.

What I learned about pricing:

  1. When I said, “Price is no object, just design a system that meets the requirements,” most did not believe me, and designed systems that didn’t meet the requirements but were cheaper.
  2. Three-phase power cables which can handle this much solar are crazy expensive. To run cables from the mains box to the shed (and back) was about $10,000 in cabling cost. On a standard home, this cost just doesn’t exist because the inverter is close to the mains circuit breaker.
  3. Solar panels are largely commodity priced: a little variation (10%) across vendors but not that much.
  4. Inverters for this size of system aren’t cheap.
  5. Battery pricing was similar across different systems.

Biggest lessons:

  1. SolarQuotes.com.au is an amazing resource. Watch their videos, read the blogs, and ask them questions! And definitely use their service to get quotes if you’re thinking about getting solar.
  2. Generally the installers (Nickel Energy) were good people, friendly, and I didn’t experience any shady sales tactics. Another reason to only go with vendors who are listed and reviewed on SolarQuotes.
  3. Avoid GoodWe inverters and batteries completely. I can’t stress this enough. There is a whole other blog on this coming soon.
  4. User Experience on most of the apps for management and reporting is pretty bad.
  5. Vendors don’t install. Almost all subcontract with installers. Lewis runs Static Solar & Electrical and he and his crew did a great job.
  6. Vendors are great at sales but don’t have the staff to really deliver customer support for complex installations.
  7. ChargeHQ is a great solution for EV charging, to make sure I’m only using solar generated energy and not grid power when I’m charging the car.

Overall, it worked…

Aside from some very serious problems around resilience of GoodWe equipment and the inability of my vendor to get an answer out of them, I have all of the requirements met except for the VPP, though the lack of resilience in the system does make the whole “backup” requirement a failure. Again, blog coming soon on GoodWe and how you should avoid them at all costs.

  • I am largely self-reliant except during very cloudy weeks, and sell almost all the excess power to the grid. To date, I’ve generated just under 40 MWh, which is enough to power ~3 standard homes for 1 year.
  • For my energy retailer, I tried Amber Electric for a while. I love the real-time pricing concept, but given the fact that I couldn’t sell energy at peak prices (given the limitations of the GoodWe software), it didn’t make sense because the prices for solar go negative during the middle of the day. For delivering 3MWh of energy back to the grid, my total net was around $1.00. I’ve switched to Energy Locals and get just 5 cents per kWh but that’s much better than with Amber.

And now, some energy porn

A good day:

The energy graphs for March 1st. The solar energy (dark blue) curve follows sunrise and sunset, which charges the batteries (purple), and then when full, begins selling excess energy to the grid (orange). Then once the sunsets, the load (yellow) is delivered from the battery storage (light blue) which then decreases over time.

The worst day:

The worst day in May. Total generation was just 20kWh compared to the all time best day (in February) of 192kWh generated. Weather definitely matters.

The seasonality of solar

Total generation by month. Summer is in the middle of this graph, so you see the seasonal effect of solar. Ignore the income, that’s not calculated correctly here. And the ends of the graph are shaper than they should be: July was not a full month, and neither is May at the time of this blog.

I consume just over half of the energy I produce:

I consume just over half of the energy generated — which is more than I expected. I blame this mostly on the commercial grade aircon system, but charging the EV does contribute. And even with the batteries, I still have to purchase about 7% of my energy from the grid.