How Sally Reduced her Electricity Bill to $56 a Month and How She Could Take it to $0.
Sally in front of her all-electric home
This Gold Coast homeowner cut her power bill by 61%
Sally moved into her all-electric Gold Coast home in mid-April 2026. By August, her average daily electricity cost had fallen from $4.72 to $1.83 — reducing her annual bill from $1,700 to $670.
The interesting part isn't that she installed solar and a battery. Plenty of people do that. The interesting part is that she requested her 30-minute interval meter data, ran it through an AI assistant, and worked out exactly which of her changes were doing the heavy lifting — and which ones weren't.
Two of the changes weren't delivering what she expected.
The starting point: $4.72 a day, and 40% of it was hot water
For her first 25 days in the house, Sally's meter recorded an average of 10.17 kWh a day drawn from the grid, split across two circuits:
| kWh/day | Rate | Cost/day | |
|---|---|---|---|
| Peak (7–10am, 3–10pm) | 3.86 | 45.25c | $1.75 |
| Off-peak (all other times) | 2.31 | 29.53c | $0.68 |
| Controlled load (hot water) | 4.01 | 16.75c | $0.67 |
| Daily supply charge | - | - | $1.62 |
| Total | 10.17 | - | $4.72 |
That's about $425 a quarter, or $1,722 a year.
The number that leaps out is the controlled load. Sally's electric hot water system was on a separate off-peak circuit, heating overnight — and it was consuming 39% of every kilowatt-hour she bought. It was cheap energy at 16.75c, which is exactly why nobody ever thinks to look at it. But cheap and large still adds up to $245 a year.
Her other problem was structural: her Red Energy plan charges peak rates from 7 am to 10 am and again from 3 pm to 10 pm. That's ten hours of peak every single day, and the 3 pm start is brutal for a solar household, because generation falls away right as the expensive window opens.
The install: 6.6 kW solar and 20 kWh battery for $13,490
On 26 May 2026, Sally had a 6.6 kW solar array and a 20 kWh battery installed for $13,490.
On the same day, she did something far cheaper that turned out to be, dollar for dollar, her best move of the whole project: she had a timer fitted to her hot water system for $250, taking it off the overnight controlled load circuit and setting it to heat in the middle of the day instead.
Six weeks later, on 10 July, she added timers to the dishwasher and washing machine, scheduling both for the middle of the day.
The move most people skip: she asked for her raw meter data
Sally's power bill told her what she spent. It didn't tell her when, and without the when, none of the analysis below is possible.
So she requested her raw interval metering data.
This is your data, and in South East Queensland you go to Energex. The request form is here:
energex.com.au — Accessing your metering data
If you have a smart meter, the data comes back as a CSV by email. It's normally free, though charges can apply if you request it more than four times in a twelve-month period or ask for a format other than the standard one. You'll need your NMI — the National Metering Identifier printed on your electricity bill — plus a start and end date, and the name and address on the account so they can verify you were the account holder for the period requested.
Three things to get right when you ask:
Ask for interval data, not a summary. A quarterly or monthly total tells you what you spent. Only the interval file tells you when, and the when is the entire analysis.
Give a generous date range. Ask from the date you moved in, or from a year before any change you want to measure. You want a clean baseline period before any changes you made.
Explicitly ask for the export data, not just consumption. If you have solar panels you want the export data as well as what you bought from the grid.
You also need to know what price you are paying for electricity at different times of the day. This information should be on your power bill.
Then she pointed AI at it
An interval data file is not human-readable. Sally's arrived as a few hundred rows of comma-separated numbers. Opening it in Excel tells you almost nothing.
This turns out to be exactly the sort of thing an AI assistant handles well. Sally uploaded the file and worked through it conversationally.
Prompts worth starting with:
"Here's my electricity interval metering data file. Work out what format it's in, identify the registers, the interval length, and the date range covered, then give me total daily consumption for each day."
"I had [size] kW of solar and a [size] kWh battery installed on [date]. Split the data into before and after, and compare average and median daily grid import."
"Show me my average consumption by half-hour for each period, so I can see how the shape of my day changed."
Then bring in your actual tariff — this is where it stops being interesting and starts being useful:
"My peak rate is [price]/kWh applying [times of day], off-peak is [price], controlled load is [price]c, and the supply charge is [price]/day. Calculate my average daily cost for each period."
"My feed-in tariff has these bands: [list them]. How much am I actually earning from exports, and how much would I earn if I shifted X kWh a day into the highest-paying window?"
"Given those savings, what's the payback period on a [quoted price] [size] kW solar system and a [size] kWh battery?"
Two things to watch:
Ask it to report medians as well as averages. A handful of cloudy days will drag an average badly out of shape — Sally's post-install average was $2.09/day while her typical day was $1.71. Both are true; they answer different questions.
And tell it about anything unusual. Sally's data contained a 56 kWh spike starting on the evening of 9 July. That wasn't a system fault or a cold snap — it was friends charging their EV while staying over. Left in, it would have made her results look 15% worse than reality. It's excluded from every figure in this case study.
Win #1: The hot water timer — a $250 part that paid for itself in twelve months
Controlled load consumption went from 4.01 kWh/day to zero, and stayed there.
Not "reduced." Zero. From 27 May onwards, that circuit records nothing at all.
The hot water didn't stop being heated — it moved. Instead of drawing 4 kWh from the grid at 3am, the system now heats in the middle of the day, straight off the roof. The load is identical. The source changed.
Cost to install: $250. Value: $0.67 a day, or $245 a year.
That's a payback of just over twelve months, and it keeps paying every year after that. In percentage terms it's the best-performing dollar Sally spent on the entire project by an enormous margin — the solar and battery take more than a decade to return their cost, the timer took a year.
One honest qualification: the timer only works because the solar is there. Moving hot water off a 16.75c controlled load circuit onto a 29.53c off-peak rate would have made her worse off. The saving comes from the pairing, not the timer alone.
But that's exactly why it's worth flagging. If you have electric storage hot water on a controlled load tariff and you have solar, you are buying overnight electricity to heat water that your roof could heat for free eight hours later — and you are almost certainly still doing it, because nobody has told you to stop.
It's the change that's easiest to miss, precisely because controlled load looks cheap on the bill.
Win #2: The battery ate the evening peak
This is where the $13,490 went, and the data is unambiguous.
Peak-window grid import fell from 3.86 kWh/day to 0.34 kWh/day — a 91% reduction.
The clearest single indicator sits in the overnight numbers. Before the install, Sally imported electricity between midnight and 6 am on every single one of her 25 days. After the install, she imported essentially nothing overnight on 51 of 75 nights.
Solar alone cannot do that. That's a 20 kWh battery carrying an entire household through the night, most nights.
Her peak-rate cost dropped from $1.75/day to $0.15/day.
Win #3: The appliance timers — real, but smaller than expected
Here's where honest analysis earns its keep.
On the surface, the appliance timers look spectacular: daily net cost fell from $2.64 to $1.83 after 10 July. But the weather also improved over that period, so the two effects are tangled together.
Separating them out:
| Before appliance timers | After | |
|---|---|---|
| Import on good-solar days | 0.26 kWh/day | 0.33 kWh/day |
| Import on poor-solar days | 4.65 kWh/day | 2.46 kWh/day |
On sunny days, the timers made no measurable difference at all — the battery was already covering the dishwasher and the washing machine regardless of when they ran.
On cloudy days, they nearly halved her import. That's the real mechanism: running those appliances at midday means marginal solar covers them directly, instead of draining battery capacity that's needed after dark. The benefit only shows up when the battery is under pressure.
Realistic value: $100–150 a year, concentrated entirely in poor-weather stretches. Worth doing for the $30 each the timers cost. Not the biggest impact.
Where Sally's money actually goes now
| Before | After | |
|---|---|---|
| Grid import | 10.17 kWh/day | 0.97 kWh/day |
| Solar export | 0 | 5.34 kWh/day |
| Energy cost | $3.10/day | $0.34/day |
| Supply charge | $1.62/day | $1.62/day |
| Feed-in credit | 0 | −$0.13/day |
| Net | $4.72/day | $1.83/day |
| Annualised | $1,722 | $670 |
Her electricity now costs less than her connection does. On a typical day Sally spends about 15c on energy and $1.62 for the privilege of being attached to the grid. The supply charge is 88% of her bill.
That single fact reframes everything she does next.
Four things Sally should do next
1. Shop on the supply charge, not the usage rate
At 0.97 kWh/day of import, a plan offering 5c/kWh cheaper energy saves her $18 a year. A plan with a supply charge 10c/day lower saves her $37. For a household this efficient, the fixed charge is the only number on the offer that really matters — and most comparison tools weight it the other way around.
2. Offset the daily supply charge by exporting from the battery when the feed-in tariff is high
Sally exports 5.34 kWh a day and earns 12.5c for it — about $46 a year. That's because almost all of it lands in the 10 am–3 pm window at 2.5c, dropping to 1.6c once she passes the 7.5 kWh daily threshold, which she did on 11 of 31 days.
Meanwhile her plan pays 15.4c/kWh for exports between 5pm and 8pm, and she is collecting essentially nothing there.
If her battery can hold back surplus and discharge to grid in that window:
1 kWh/day → +$32/year
2 kWh/day → +$63/year
3 kWh/day → +$89/year
The trap: she exports at 15.4c but buys back at 29.53c. If discharging into the evening window empties the battery before dawn, she loses 14c on every kilowatt-hour. This only makes money using genuine surplus — energy that isn’t needed overnight. With 20 kWh of storage and only ~3 kWh of overnight household load, she has ample headroom, but the inverter needs to be configured deliberately rather than left on plain self-consumption.
3. Check for a phantom controlled load charge
Her controlled load circuit has consumed nothing since 27 May. If there's still a network access or service charge attached to it, that's pure waste — worth one phone call.
4. Send any new load into the 10am–3pm window
This is the principle her whole result rests on: self-consumption is worth 30–45c/kWh. Export is worth 1.6–2.5c/kWh. That's a factor of fifteen to eighteen.
Every kilowatt-hour Sally moved into daylight — hot water in May, appliances in July — was worth roughly twenty times more than exporting it. If she ever adds an EV, a pool pump, or reverse-cycle air conditioning, scheduling it midday is where the next real money sits.
What she should not do is chase a better feed-in tariff. At these rates it's rounding error.
The return: roughly 13 years on paper, likely better in practice
Solar array and battery: $13,490 Hot water timer: $250 Total outlay: $13,740 Measured saving: $2.88/day, or $1,052/year Simple payback: 13.1 years
That headline number hides two very different investments, and it's worth splitting them:
| Cost | Savings/year | Payback | |
|---|---|---|---|
| Hot water timer | $250 | $245 | 1.0 years |
| Solar + battery | $13,490 | $807 | 16.7 years |
| Combined | $13,740 | $1,052 | 13.1 years |
The split is a little artificial — the timer only saves money because the solar exists, so the $245 is genuinely joint. But it makes the point that the cheapest intervention returned nearly 100% in year one, while the capital-intensive one is a long game.
Three caveats on the long game, and they all point the same direction.
This is winter data. Sally's measurements cover May to August. A South East Queensland summer brings substantially more generation from the same array — and her pre-solar baseline would have been higher too, once air conditioning entered the picture. Both effects widen the gap. A full-year figure would very plausibly land closer to 10 or 11 years.
The baseline is thin. Twenty-five days of autumn is a small sample to extrapolate a year from. It's the honest best estimate available, not a guarantee.
Rates move. Every projection here assumes today's tariffs hold. They won't — and historically, rising retail prices shorten solar payback rather than lengthen it.
Worth noting too: $13,490 for 6.6 kW plus 20 kWh of storage is sharp pricing. A battery that size would have been well north of that not long ago. The economics that make Sally's numbers work are considerably better than they were even two years ago.
What Sally could have done before buying solar and a battery
If saving money was Sally’s only reason for installing solar and a battery (it wasn’t), she could have tried something else first.
Using her interval data, as supplied by Energex, she could have tried finding an electricity plan that better suited her time-of-day power usage. By moving the hot water system off the controlled load circuit and having the timer fitted, she could have shifted a lot of her power usage to the middle of the day.
There are many electricity plans that offer free hours of power from 11 am - 2pm. By getting the hot water heating for free and using timers to other appliances in the same time slot, she could have reduced her power bill considerably without paying for solar and a battery.