How my solar battery is battling coal on the open market – and the score so far

It took several years to reach full electrification of our place in Melbourne. Finally, just before 6pm on Friday 23 April 2021 the gas was cut off at the main, and the workers began filling the hole in the road.

The next move was the battery (here I would like to send a Cheerio to all Victorian taxpayers to thank you for your assistance).

The battery was installed in December. Unfortunately, that unit didn’t work and by then it was Christmas and Santa’s supply chain was all over the place. This meant our battery’s operational birthday was March 2022.

Some like to bottle up the power in their battery, reserving it for their private use. But I was keen to link the battery to the wider network.

I knew I would not be using all the power all the time – and just as you would lend a neighbour your ladder – I felt that the surplus power in the battery should be available to others. I also felt under some obligation to return the favour of the subsidy.

Most thrilling was the thought of engaging the coal-fired generators in battle on the open market. I could make my fortune with the spoils – as Don Quixote said to Sancho Panza, “this is righteous warfare, and it is God’s good service to sweep so evil a breed from off the face of the earth.”

As soon as the battery was operational, I joined the Tesla Energy Plan retailed through Energy Locals (TEP).

TEP promised not to use the power in the battery more than once a week (on average) nor leave it less than 20% full. In return, I received an extended battery warranty and $30 a month off my bill (which balanced out the supply charge of the same amount).

But by the end of 2022 no “controlled load usage” had taken place. This was disappointing. No neighbours had benefited from my clean power, and no monstrous giants had been punctured.

So, I said goodbye to the extended battery warranty (which I judged to be of little value) and switched to Amber.

Amber does not charge you for power, rather the Amber software enables household power producers and power accumulators like me to sell power into the wholesale electricity market.

Now I could joust with the monstrous giants in the Great Game of Electricity Demand and Supply.

On days when power was in short supply across the network, I could land a blow by taking sales from the coal-fired generators. On other days when the grid was awash with power and the buy price was negative, I would take a hit (as do the giant coal generators). Win or lose for me, but lose and lose for coal.

To prevent a total loss of home power, Amber lets you choose how much of your battery is reserved for private use. I chose 20% which, to keep things running overnight (for us), is plenty.

Now, each day the robots go to work checking the renewable energy weather forecast, the forward market price estimates, the space left in the battery and the quantity of power available for sale.

On a windy night when the grid price is low (or negative), the robots pump up the battery. Then, when the sun goes down, they run around the trading floor with their palms out yelling sell, sell, sell.

Each morning you wake to yesterday’s score in the app. I like to check Power Bought, Total Sales and Net Cost. Sometimes you win and sometimes you lose.

In April 2023, I started recording the daily scores – much like a cricket scorer watching a particularly slow game. (Unfortunately, this data is not downloadable.)

Then in May 2023, the Mark 1 famous name inverter failed (for the second time). Now I had a battery but no homebrew power.

This was my chance to become a day trader in the electricity market. When I remembered I pressed the button on the app to buy power in the middle of the day. Then I would sell it at night when the price was higher.

That May was one of my highest months for consumption (as you would expect), one of my better months for sales ($47.14) and one of my highest months for power costs ($36.81). Although I came out marginally in front, the process was onerous, and I often missed the best buy and sell prices. I decided that from then on, I would let the robots do the trading.

By 1 June 2023 the Mark 2 inverter was running, the power was flowing down from the roof and I and my team of robots were back in the Game.

Today after 18 months, I have some interesting numbers to report. The figures I will quote are export/import figures. Self-consumption is excluded. If the house makes 10 kWh, stores 5 kWh and uses the other 5 kWh, and there are no trades in or out, then the export score is 0 kWh of power, $0 sales, and $0 net cost.

I will note that Amber and the distribution company both charge a flat monthly fee which adds up to around $40 a month. To come up with a true $zero month, you must come out of the buy/sell Game with a surplus of $40.

How would you like a bill like this?

I will also admit there are several one- or two-day holes in the data and I will put aside data from June this year as we had 4 not the usual 2 people in the house.

Here is the score so far.

Across eighteen months, Power Bought in a day has ranged from 0 to 28 kWh – one of my no-panels trading days. Sales in a day have ranged between -$3.83 (the sunny day after Australia Day when prices were negative) and +$75.56 (on the day in February when a storm knocked out the transmission system). Net cost per day has ranged from $-63.72 (that storm day) and +$9.74 on a day in July this year when we must have been buying in expensive power to keep warm.

So far Total Sales is +$408 and total net cost is -$98.22.

Counting all the Amber and network charges, my total cost for energy for the last Financial Year was $273.89. This was up a bit from last FY but down from a pre-panels gas/electricity bill of $2,102.92 in 2018-2019.

The best month for Power Bought was January this year when we reluctantly bought in 1.03kWh. The biggest buying month was that cold June this year when we bought 389.74 kWh.

The best month for Sales was February this year when sales were +$83.15. The worst sales month was January when negative sale prices cost us a total of -$6.45.

The best month for net cost was the stormy February this year when we made $64.72. June this year was the worst month for net cost caused by what the Germans call dunkelflaute (dark (solar) & calm (wind)). The month was cold, there was no sun for homebrew power and no wind in Victoria or South Australia to keep buying costs low. June cost us $109.35.

If I compare Q3 of this calendar year (July to September) to the same period last year I find:

• Our power use was up by one third (thanks to that cold July).
• Sales were up by one third. September sales were down at $6.40 when last year they were $22.24. But sales in July rose from $12.07 to $40.53.
• Net cost was up three times. This was because the increased use of power and the more costly power combined to offset the increase in sales.

The next problem to solve is preventing costly exports when the price is negative. This seems to be occurring more often, which is good news as it means there is more solar and wind power in the system. However, there is no community benefit from exacerbating an oversupply of renewable energy and I lose dollarbucks in the Game.

Recently, Amber introduced its customers to a third-party supplier Village Energy who sell a $160 Voltello box that watches the price on Amber and can speak to the inverter linked to my solar panels.

The set-up required persistence, especially to link the box to the software settings on the famous name inverter. But with assistance, that was accomplished.

The performance of the box is limited by our not-so-smart meter. Some time ago, Victoria replaced all electricity meters with smart meters, but that nation-leading program began when half the mobile phones were Nokias. Back then a meter check-in every half hour was revolutionary. Today 5-minute check-ins are the new standard and that is what the box is looking for.

Nonetheless, there is a workaround (which I will spare you) that seems to be effective, not at preventing all negative sales, but at reducing the scale.

Here is a recent day, 2 October – the battery app tells the story; the Voltello box tells the inverter to switch off and the Amber app gives the score.

The battery shows the fusion reactor rising in the sky and the panels generating power (yellow) which filled the battery (green mountain). After the battery was full, the panels exported for a period and then the Voltello box saw that prices were negative and switched off the inverter, so no further storage or exports took place.

Then, in the evening, we cooked tea (blue squiggles) while the robots sold down the power in the battery (grey tower). The sales straight from the panels and later from the battery are consolidated and reported in the Amber app as Solar exports.

Score that day was a net of +$1.07 – a good home game win. (We are just taking it one day at a time.)

Across the last month the news is less good. As you can see, looking back over the last 30 days, I am in the hole for 29¢. (Think of what you could be buying instead!)

Nonetheless, I am looking forward to the test over summer, confident we can beat last summer’s score.

McGilvray is in commentary reading the cable messages and tapping his pencil on the desk. “It’s a short ball, Bradman moves back and pulls fiercely past square leg, Hutton running round from deep fine leg has no chance and the ball goes under the ropes for another four.”

Harry is a transport consultant based in Melbourne. He thinks electricity is fun too.