Svenska Calculator
Book a demo
Svenska rohi@kronwatt.com
  1. Home
  2. The Home Energy School
  3. Modul 05

Part 2 · Your house

Burning heat, or moving it?

A heat pump, or a wood stove? The question sounds like a choice between two kinds of heat. It is not. It is a choice between making heat and moving it — and that difference decides what every kilowatt-hour costs you.

Rohi Perlsteyn, KronwattPublished Updated

In brief

  • Burn fuel and you never get out more than you put in. A heat pump moves heat and gives more.
  • The brochure's 4 to 5 comes from laboratory tests. In Swedish houses, 2.7 was measured.
  • The price per kWh of heat is the energy price divided by annual efficiency. Anything else compares apples with pears.
  • A stove heats the room it stands in. A heat pump heats the house. They solve different problems.
On this page
  1. Two ways to get heat
  2. The heat factor, and why yours is not the brochure's
  3. When it gets properly cold
  4. What Sweden heats with today
  5. The stove: what it is good at
  6. What does a kilowatt-hour of heat cost?
  7. So: which one?
  8. Check yourself

Two ways to get heat

There are only two. Every heating system in the world is one or the other.

The first is to burn something. Wood, pellets, oil, gas — or electricity through a coil, which comes to the same thing. You have energy in a fuel and you turn it into heat. And here there is a ceiling nobody gets past: you never get out more heat than there was energy in the fuel. In practice you get less, because some of it leaves up the chimney.

The second is to move heat that already exists. That is what a heat pump does. It makes no heat. It picks up heat that is already sitting in the ground, in the bedrock or in the outdoor air, and carries it into the house.

A heat pump is not a factory. It is a lift.

And that is why it seems to break the laws of physics without doing so. The electricity does not pay for the heat. The electricity pays only for the carrying — for driving the compressor that lifts heat from cold to warm. The heat was already out there, free, even at zero degrees.

BURN boiler 1 kWh wood in 0.8 kWh of heat the rest leaves up the chimney MOVE heat pump 1 kWh electricity + 2 kWh free from the ground 3 kWh of heat electricity pays the carrying, not the heat
The same kilowatt-hour of purchased energy. Entirely different amounts of heat in the house.

The heat factor, and why yours is not the brochure's

The number that describes how well a heat pump lifts is called the coefficient of performance — COP, and across a whole year, SCOP or the annual heat factor. It says one thing only:

heat out = electricity in × heat factor

A heat factor of 3 means one purchased kilowatt-hour of electricity becomes three kilowatt-hours of heat in the house. Two of them came out of the ground and cost nothing.

And here comes the most important figure in this module — or rather, the two figures that do not agree.

Where the figure comes from Annual heat factor
Laboratory test of new ground-source pumps, as reported in trade press 4–5
Measured in 20 Swedish houses, 2012–2014 2,7–2,9
Worst and best house in that same measurement 1,9 – 3,4
Energimyndigheten's field measurement of 20 ground-source heat pumps, May 2012 to June 2014. The pumps were 7 to 14 years old. The laboratory figure of 4 to 5 we could trace only to trade press and could not confirm with the agency — treat it as an indication. Nor have we been able to find the agency's own report page — the figures agree across several sources close to the original, but treat them as an indication rather than a verdict.

The same technology. Nearly a twofold gap, larger than anything you will find between two columns of a price list. And the spread between the worst and the best house — 1.9 against 3.4 — is wider than the gap between two different brands.

When it gets properly cold

A heat pump that takes its heat from the outdoor air has a problem ground-source does not: the colder it gets outside, the less heat there is to collect, and the higher the lift becomes. Exactly when the house needs the most, the pump can manage the least.

So nearly every heat pump installation has an immersion heater — an ordinary electric coil that steps in on the coldest days. It has a heat factor of 1. It burns. And that is entirely normal: a pump sized to cover the coldest hour of the year on its own would be too big and would run badly for the rest of it.

The rule of thumb your installer works to

A heat pump covering 60 to 70 per cent of the house's peak power demand still delivers more than 90 per cent of the year's energy. That sounds impossible until you remember how few genuinely cold hours there are. The immersion heater takes the remainder.

The rule comes from a KTH and Vattenfall report in the Effsys2 research programme, 2010. Two words in a quote belong to it: the bivalent temperature, where the immersion heater starts helping, and the operating limit, where the pump stops altogether.

What Sweden heats with today

This is worth knowing before you assume you are either first or last to something. Of Sweden's 2 040 800 permanently occupied houses, 1 345 000 have a heat pump — just under 66 per cent.

Number of houses with each system, 2024

  • Ground-source443 000

  • Air-to-air heat pump432 000

  • District heating303 000

  • Air-to-water heat pump229 000

  • Exhaust-air heat pump146 000

  • Wood boiler179 000

  • Oil boiler, still in use15 000

Energimyndigheten, Energistatistik för småhus 2024, published 10 June 2025. A house can have several systems, so the bars add up to more than the number of houses. On top of that: 741 000 houses have a stove or an open fireplace — four times as many as have a wood boiler. That says something about what a stove is really used for.

The stove: what it is good at

A modern wood stove is not a bad machine. Energimyndigheten tested fifteen stoves and most reached 75 to 80 per cent efficiency, with a full range of 63 to 80. An 1980s reference stove managed 59 per cent, a 1990s one 66. The improvement is real.

The building rules long set a floor: a wood stove had to reach at least 65 per cent, a pellet stove 79 per cent and a solid-fuel boiler under 100 kW 87 per cent. Boverket replaced the building regulations on 1 July 2025 with a transition to 30 June 2026, and we could not confirm that those tables carried over — check the current text before quoting them. But notice what is missing from that list.

If you burn wood it is worth knowing what you are buying. Firewood is sold in three different measures, and they are not the same: a stacked cubic metre is about 0.65 solid cubic metres, a loose-tipped one about 0.50. Ten loose cubic metres is therefore only around 7.7 stacked.

As a rule of thumb, official Swedish statistics use 1.24 MWh per stacked cubic metre of mixed firewood. Birch is higher — around 1 830 kWh per stacked cubic metre at 20 per cent moisture. And moisture is not a detail: wet wood holds water that must be boiled away before anything becomes heat, and you never get that energy back.

What does a kilowatt-hour of heat cost?

Now we can answer the real question. There is only one formula, but it is the most important one in the whole course:

kr per kWh of heat = energy price ÷ annual efficiency

We will use 2.00 kr/kWh for electricity. That is not a claim about what you pay — it is a worked example. It builds on SCB's measured total price for an electrically heated house in the second half of 2025, 209.21 öre/kWh including grid, tax and VAT, less the energy-tax cut of 1 January 2026. Replace it with your own. It is on your invoice.

System Annual efficiency kr per kWh of heat
Direct electric heating1,02,00
Electric boiler1,02,00
District heating, national average—1,35
Air-to-water heat pump2,70,74
Ground-source, as measured in the field2,70,74
Ground-source, well installed4,00,50
 The annual efficiencies are chosen from the ranges above — they are assumptions, not measurements of your house. Our own engine uses 3.0 as its default for a heat pump, which gives 0.67 kr; we have taken the more cautious 2.7 here. The district-heating price is Energiföretagen's 2026 average for a villa, 1 351 kr/MWh including VAT, across 404 networks. The spread between municipalities is enormous: from 0.65 to 1.82 kr/kWh.

So: which one?

If the question is what should heat my house, then a heat pump and a stove are not two answers to the same question. They do different jobs.

The heat pump is the base load. It runs by itself, around the clock, in every room with a radiator or underfloor loop, including when you are not at home. It is the cheapest kilowatt-hour in the table above, and it is that every hour of the year.

The stove is something else: it heats the room it stands in, it needs somebody at home to put wood in, and it works when the power is out. That last one is a genuine value, but it is a preparedness value, not an operating one. That 741 000 houses have a stove while only 179 000 have a wood boiler says exactly that.

If you already have wood on your own land and the time to burn it, the arithmetic changes — the wood is then nearly free and the stove can carry a real part of the winter. But compare one thing at a time, and calculate across the whole year, not across the coldest week.

Two rules before anything is installed. A fireplace or stove needs a notification to the municipality and a go-ahead before work starts — the paragraph was renumbered on 1 December 2025, so check the current text. A borehole for ground-source heat needs permission or notification from the municipality, and an outdoor unit has noise rules to respect towards the neighbours.

Check yourself

Six questions. Getting one wrong is the useful part — the explanation is written for exactly that answer.

  1. Question 1Why can a heat pump give three kilowatt-hours of heat for one kilowatt-hour of electricity?

    • It sounds reasonable, but the word efficiency hides the point. An electric boiler is already close to 100 per cent — it cannot possibly get three times better. The heat pump does something else: it moves heat that already exists in the ground instead of making new heat.

    • Exactly. The electricity pays for the carrying, not for the heat. It is also why the laws of physics are untouched: no energy has been created, it has only been carried from one place to another.

    • A heat pump has no chimney — it burns nothing. Chimney loss is what separates an 80 per cent boiler from a 95 per cent one, but that is a different order of magnitude from a factor of three.

  2. Question 2Your quote promises an annual heat factor of 4.5. In Energimyndigheten's field measurement, Swedish houses came in at 2.7. What is the most likely explanation?

    • Tempting, but usually wrong — and it leads you away from the question that actually helps. The laboratory figure is correctly measured, just under different conditions: a set temperature, a set flow temperature, no hot water, a perfect installation. Ask instead which flow temperature the quote assumes.

    • Yes. And the best thing about that answer is that it points at what you can do: lower the flow temperature, look at the radiators, size it correctly. The spread in the measurement — 1.9 to 3.4 — was wider between houses than between brands.

    • Half right, and therefore dangerous. The pumps were 7 to 14 years old, so today's are better. But the measurement was made in real houses with real hot water, and that is the part that pulls the figure down — it does not disappear because the pump is new.

  3. Question 3District heating costs 1.35 kr/kWh. Electricity costs 2.00 kr/kWh. Which is cheaper to heat the house with?

    • This is the most common arithmetic mistake in Swedish households, and it looks entirely reasonable. The district-heating price is for heat delivered; the electricity price is for electricity purchased. Until you divide by the annual efficiency, you are not comparing the same goods.

    • Exactly. If the electricity drives direct radiators it is 2.00 kr per kilowatt-hour of heat, and district heating wins. If it drives a heat pump with an annual factor of 2.7 it becomes 2.00 / 2.7 = 0.74 kr, and the heat pump wins comfortably. The same electricity price, entirely different answers.

    • Nobody knows where the electricity price is going, and this course will never claim to. The question can be settled without guessing: divide by the annual efficiency, and you are comparing like with like.

  4. Question 4Why does nearly every heat pump installation contain an immersion heater?

    • It does serve that purpose too, but it is not why it is there. It is planned in from the start: the pump is sized for 60 to 70 per cent of the house's peak demand, and the immersion heater takes the coldest hours.

    • Yes, and the logic is elegant: a pump covering 60 to 70 per cent of the power still delivers more than 90 per cent of the year's energy, because genuinely cold hours are few. A pump that handled the coldest hour alone would run badly the rest of the year.

    • The opposite — an immersion heater has a heat factor of 1 and lowers the annual figure every hour it runs. It is there in spite of that, because the alternative is an oversized pump that runs worse all year.

  5. Question 5What does it tell you that 741 000 Swedish houses have a stove, but only 179 000 have a wood boiler?

    • If the stove were the house's heating system, the two figures would sit close together. That one is four times the other points elsewhere: the stove is a supplement and a reassurance, not the house's base heat.

    • Yes. A stove heats the room it stands in and needs somebody at home. The wood boiler is the one that heats the whole house through the water, and there are far fewer of those. That says more about the stove's role than any brochure does.

    • They are not banned. The building rules have set requirements — a solid-fuel boiler under 100 kW had to reach at least 87 per cent efficiency — but a requirement is not a ban.

  6. Question 6You burn wet wood instead of dry. What happens?

    • The wood does carry the same energy in the timber itself. But the water in it has to be boiled away first, and that energy comes out of the same fire. So you are paying to evaporate water instead of heating the house.

    • Yes. The steam leaves up the chimney and takes the heat with it. That is why firewood is sold with a stated moisture content, and why dry wood is worth more per cubic metre — you are buying energy, not volume.

    • It does burn more sluggishly — but sluggish is not the same as long and useful. A cool, incomplete burn gives less heat and more emissions, so it is worse on both counts.

Your answers live only in your browser, and vanish when you leave the page.

Sources

  1. Energimyndigheten, Energistatistik för småhus 2024, published 10 June 2025: number of houses by heating method and by system.
  2. Energimyndigheten's field measurement of 20 ground-source heat pumps, May 2012 – June 2014: annual heat factors of 2.7 and 2.9, spread 1.9–3.4. The agency's own report page can no longer be reached; the figures come from sources close to the original and should be checked against it before print.
  3. Energimyndigheten's test of 15 wood stoves, page updated 7 March 2016: efficiency of 75–80 per cent for most, full range 63–80; reference stoves at 59 and 66 per cent. The test predates the ecodesign requirements.
  4. Boverket's building regulations, sections 6:7411 and 6:7412: minimum efficiency of 65 per cent for a wood stove, 79 for a pellet stove, 87 for a solid-fuel boiler under 100 kW. Boverket replaced the regulations on 1 July 2025 with a transition period to 30 June 2026 — check the current text.
  5. Naturvårdsverket and Boverket, Elda rätt (2009): the emission requirements do not apply to open fireplaces and tiled stoves intended for the pleasure of a fire. The claim that the net can be negative comes from the trade body Svensk Ved and is not confirmed by any public authority.
  6. Liss, Dalarna University (2005), and SCB's conversion figures for household firewood: 1.24 MWh per stacked cubic metre of mixed wood; birch around 1 830 kWh per stacked cubic metre at 20 per cent moisture. The volume measures: 1 m³ stacked ≈ 0.65 m³ solid, 1 m³ loose ≈ 0.50 m³ solid.
  7. Energiföretagen, district-heating prices 2026, villa test customer at 20 MWh a year: average 1 351 kr/MWh including VAT across 404 networks, lowest 651 and highest 1 817.
  8. SCB, Priser på elenergi och på överföring av el (EN0301): a total price of 209.21 öre/kWh in the second half of 2025 for the 15 000 kWh and above category. Skatteverket: the energy tax falls from 43.9 to 36.0 öre/kWh on 1 January 2026. The 2.00 kr/kWh worked example is our own combination of the two, not a forecast.
  9. KTH and Vattenfall, the Effsys2 research programme (2010): a heat pump covering 60–70 per cent of peak demand covers more than 90 per cent of annual energy.