How we work it out
The calculator turns a few sliders into a dollar figure, and between the two sits a model of your house and a list of every retail plan we can find. This page is that model, written out: where the plans come from, what we assume about a roof and a battery, what we refuse to price, and where the answer is weakest. Everything below is read from the calculator itself when this page loads, so it cannot quietly go out of date.
Where the plans come from
Every retailer in the national market has to publish its plans as open data. We read that data directly, not a comparison site's version of it.
The Consumer Data Right register at api.cdr.gov.au lists every energy retailer and
where its product data lives. Most of the large brands are served by the Australian Energy
Regulator at cdr.energymadeeasy.gov.au. We read the rates, the daily supply charge,
the time of use windows, the feed-in tariff and its volume limits, and the eligibility rules,
straight from the retailer's own published record.
10667 residential plans across 13 distribution networks, last collected 2026-09-27.
Prices in that data exclude GST, and not uniformly: a daily supply charge and a usage rate carry it, a residential feed-in credit does not. We add it where it belongs, which is why our figures match a bill rather than a rate card.
What we assume your roof makes
This is the single number the whole answer rests on, so it is the one we most owe you an explanation of.
We describe a roof by its specific yield: kilowatt hours a day for each kilowatt of panels, averaged across a whole year. A figure of 3.6 means a 10 kW system makes about 36 kWh on an average day. It is not a sunny-day number. Midwinter, cloud, dust, the efficiency panels lose when they get hot and the losses in the inverter and the cabling are all inside it already, which is why it looks low next to the number on the box.
Our figures come from PVGIS, the European Commission Joint Research Centre's photovoltaic model, on its ERA5 dataset averaged over 2005 to 2020, for a roof mounted array at the best tilt for its latitude with PVGIS's standard 14% system losses. We use a representative point in each network's area, named below, and every row can be reproduced by anyone from a public URL.
Beside it we show the Clean Energy Regulator's zone rating for the same place, from the Renewable Energy (Electricity) (Zone Ratings and Zones for Solar (Photovoltaic) Systems) Instrument 2019, converted from megawatt hours a year into the same daily unit. That is the figure Australian law deems a system will generate when it calculates your certificates. It is a coarser instrument, with only four zones for the whole country, but it is independent of ours, and the two agreeing is worth more than either alone.
| Region | Network | Modelled at | Ours (PVGIS) | Regulator's zone | Plans held |
|---|---|---|---|---|---|
| South east Queensland | Energex | Brisbane | 3.61 | 3.79 (zone 3) | 1270 |
| Regional Queensland | Ergon | Rockhampton | 3.91 | 3.79 (zone 3) | 90 |
| Sydney, Newcastle and the Central Coast | Ausgrid | Sydney | 3.54 | 3.79 (zone 3) | 1366 |
| Western Sydney, the Blue Mountains and the Illawarra | Endeavour | Penrith | 3.34 | 3.79 (zone 3) | 1416 |
| Regional New South Wales | Essential Energy | Dubbo | 3.84 | 3.79 (zone 3) | 1770 |
| Canberra and the ACT | Evoenergy | Canberra | 3.63 | 3.79 (zone 3) | 489 |
| Melbourne city and inner suburbs | CitiPower | Melbourne | 3.3 | 3.25 (zone 4) | 684 |
| North west Melbourne | Jemena | Broadmeadows | 3.33 | 3.79 (zone 3) | 679 |
| South east Melbourne and the Mornington Peninsula | United Energy | Frankston | 3.39 | 3.25 (zone 4) | 675 |
| Western Victoria and Geelong | Powercor | Ballarat | 3.3 | 3.79 (zone 3) | 687 |
| Eastern Victoria and Gippsland | AusNet | Traralgon | 3.18 | 3.25 (zone 4) | 692 |
| South Australia | SA Power Networks | Adelaide | 3.69 | 3.79 (zone 3) | 754 |
| Tasmania | TasNetworks | Hobart | 2.88 | 3.25 (zone 4) | 95 |
The two sources agree within about 6% everywhere except Tasmania, where the regulator's zones are too coarse to tell Hobart apart from Gippsland and we follow PVGIS. Where a region shows no plans, it is not offered in the calculator at all: we would rather say we do not know than price a house against an empty list.
What this replaced, and why we are telling you
Until recently the calculator used one figure, 4.2, for every visitor in the country. That number is almost exactly the regulator's zone 2 rating, and zone 2 is Darwin. Brisbane, Sydney, Canberra, Adelaide and Perth are all zone 3; Melbourne and Hobart are zone 4. Nobody on the eastern seaboard lives in zone 2. Every figure the calculator produced was assuming a roof in the Northern Territory, which overstated generation by roughly 10% in Brisbane and 30% in Hobart, and the effect was not neutral: it flattered plans with a high feed-in tariff, because it invented exports to sell into them.
We mention it because a methodology page that only lists the things we got right is marketing.
Where this is still weak
A yield figure describes a well-installed array facing the right way with nothing shading it. A roof facing east and west, or with a tree over one string, will make less, and we have no way to know that from a slider. If your own bills or your inverter app say otherwise, believe them over us, and tell us: for our own clients we price against their meter readings rather than this model.
The battery, the inverter and the meter
We run your house hour by hour for a week and then scale it up, rather than dividing a year by 365. The order is the one a real system follows: solar covers what the house is using, the battery takes what is left over, the battery gives it back after dark, and only what neither can handle meets the grid in either direction.
- Battery round trip
- 90%. Storing a kilowatt hour and taking it back out again loses some of it. A battery is not a bucket.
- Export limit
- 5 kW on a single phase supply and 15 kW on three phase. Your network caps how much you may push back into the street regardless of what the roof makes or the inverter could send. It is the reason a bigger system does not always earn more.
- Your inverter
- Caps every flow in both directions. A 10 kW roof behind a 5 kW inverter is a 5 kW roof for the purposes of the busiest hour of the day.
- Day and night
- We split your use at the daylight boundary rather than assuming a flat draw, because when you use power decides almost everything on a time of use plan.
What we refuse to price
Most ways of getting this wrong flatter the answer. A missing limit in a retailer's data looks exactly like no limit at all, and a model given no limit will happily spend one.
- Wholesale linked plans
- Shown for comparison, never recommended. What these publish is one annual average with no time of day in it, and time of day is the whole of it: a solar house exports in the middle of the day, which is when wholesale prices are lowest and regularly below zero in spring. Pricing that at an annual average has produced figures hundreds of dollars adrift. Until the rates are shaped by the hour, it is not a number anyone should be moved on.
- Demand tariffs
- Left out entirely. A demand charge is set by your single worst half hour in a month, and a model that priced one on its energy rates alone would make it look far cheaper than it is.
- Feed-in tariffs with no stated volume limit
- A high rate with a daily cap and a high rate without one are written identically in the published data, and the difference can be thousands of dollars a year. Where a limit has been confirmed we apply it and say so. Where we only have it second hand, the plan's card says the limit is reported rather than confirmed.
- Free windows with no stated allowance
- We assume 24 kWh a day rather than unlimited. The government mandated free power offers set that allowance, so an unstated limit almost always means it. Treating it as unlimited let the model fill a battery for nothing every lunchtime and sell it back all evening, which is not a business any retailer is in.
- Impossible rates
- Anything above 44c a kilowatt hour for exported power is treated as a collection error rather than an offer: a figure entered in dollars, or a decimal point in the wrong place.
You can see the effect of this yourself. The calculator can be told to price only the plans whose terms are completely recorded, and it will tell you how many of the field that leaves out.
A plan can be real, published, cheapest, and still not yours
Retailers publish who may take each plan, in their own words, and the range is wider than most people expect. Some want a membership. Some are open only to new customers, or to somebody moving house. Some need a smart meter, or refuse a controlled load. Some require an electric vehicle. And some exist only for people who bought their solar system through that retailer's own installer, which means the plan is genuinely published, genuinely the best rate in the state, and genuinely unavailable to you.
We read those rules and carry them with the plan, so a condition that decides whether you can actually sign up travels with the price instead of turning up at the end. What we deliberately do not do is flag the ones that say nothing: nearly every plan notes that you have to live in the area it is sold in, and a warning on every plan in the list is furniture rather than a warning.
We got this wrong for a while, and fixing it is why this page exists. The check was reading the discounts and the loyalty schemes and never opening the eligibility rules at all, so a plan that required buying your system through one particular retailer went out looking like a plan anybody could ring up and ask for.
What the number means, and what it does not
It is an estimate of a year on the best plan we can price for a house like the one you described, in the region you chose, on published rates, assuming the house runs the way the model runs it.
It is not a quote, and it is not a promise. Rates change, and a plan that wins today can be repriced next quarter. A figure that comes out in credit depends on a battery being charged and discharged to suit that plan every single day, which is the part that has to actually happen. And the further your roof, your shading or your habits are from the model, the further the answer will be from your bill.
Where we can do better than a model, we do: for our own clients we price against their actual meter readings, hour by hour, and the difference between the two is sometimes large enough to change which plan wins.