Category: The house project

Für Handwerker, Renovierer und Technik-Interessierte.

Posted on

Insulating Brick Walls in Old Buildings: Field-Fired Bricks, Moisture and My Energy Concept

Ziegelwand dämmen im Altbau: freigelegte Feldbrandziegel und Kalkputz in der tiefen Fensterlaibung eines sächsischen Bauernhauses in Siebenbürgen

Should you be insulating brick walls in an old building, and if so, how? That was the question in a recent open discussion in a forum, where I was corrected quite clearly. I had written that walls need to breathe. The reply came promptly, appreciative in tone and clear in substance: that statement was simply wrong, and external insulation was the best thing I could do for my wall. In this article, I sum up what I learned from the exchange, where I got it wrong and why I still will not insulate the exterior walls of my house.

If you are new here: I am renovating a Saxon farmhouse in Transylvania, in the village of Cincu (Großschenk), built in 1888. The exterior walls are around 60 centimetres thick and made of field-fired bricks. Around 40 years ago, I trained as a building technician, even though my career later moved towards electronics and IT. Before starting the restoration, I therefore looked very closely at how to approach a house like this properly. That matters more than it might seem at first glance, because field-fired bricks behave quite differently from the solid bricks most people know from tables and calculators.

Key points at a glance:

  • Walls do not breathe. Air and indoor moisture are removed almost entirely by ventilation, not through the wall.
  • What matters in old masonry is moisture transport: the wall must stay vapour-open towards the inside and the outside, so that moisture cannot build up anywhere.
  • Field-fired bricks insulate better than modern solid bricks, roughly like light clay blocks. My 60 cm wall is equivalent to about 4 to 7 cm of polystyrene, and its U-value is roughly 0.45 to 0.7 W/(m²K).
  • On the other hand, field-fired bricks are highly hygroscopic and absorbent. Insulation that traps moisture behind it can damage them permanently.
  • For historic building fabric like mine, wall insulation is therefore often advised against altogether. I would rather save energy at the roof, ceiling, windows and floor.
  • My energy concept makes up for a good part of the higher heat demand of the uninsulated wall: wall heating with an air-to-water heat pump, a split air conditioner for dehumidifying, a heat pump water heater, a rooftop plug-in solar system with battery storage and a wood stove.

What the discussion was about

So that you can follow what I am referring to, here is the core of the post a forum member wrote in reply to my statement (translated from German, slightly adapted and shortened):

“I really admire you for your skills and your knowledge. Maybe you have the time and inclination to read up on thermal insulation and play around with this calculator. […] Your statement that walls need to breathe is absolutely wrong. External insulation, if done properly, is the best thing you can do for your wall and your wallet. The insulating effect of your 60 cm solid brick wall is equivalent to a 3 cm EPS/polystyrene board. In other words, 3 cm of polystyrene halves the heat loss. There are many insulation materials, and you can achieve a lot with little money, especially if you are as good a craftsman as you are.”

I was pleased by the first sentence. I already knew the U-value calculator by ubakus.de (in German), as I had worked with it extensively during the planning phase of my restoration. Still, I took the post as an opportunity to go through my reasoning again carefully and recalculate my wall. My conclusion: on the term “breathing”, the other person is right. But the calculation assumes an ordinary solid brick, and the recommendation to insulate externally does not suit a house like mine.

Do walls need to breathe? The truth behind the myth

The short answer: No, walls do not breathe. There is practically no significant air exchange through a plastered brick wall. Only a very small share of the moisture we release into the indoor air through cooking, showering, sleeping and plants escapes through the wall to the outside. By far the largest part is removed by ventilation, that is, through open windows, leaky joints or a ventilation system.

I found it interesting where the myth comes from. It goes back to an experiment by the Munich hygienist Max von Pettenkofer in 1858. Pettenkofer sealed the windows and doors of a room and found that the air exchange hardly changed. He concluded that the air must be flowing through the brick walls. Among other things, he apparently overlooked the stove, which was connected to the outside air via its flue. Later measurements disproved this, but the term has stuck to this day. You can find a short summary in the article “Vom Mythos der atmenden Wände” on Haustechnikdialog (in German).

In this respect, the person in the forum was right, and I had used the term too casually.

What I actually meant: vapour-open in both directions

Behind my poorly worded statement, however, lies an idea I stand by: what actually happens in old masonry is not breathing but moisture transport. Three processes play a role:

  • Vapour diffusion: Water vapour migrates through the building component, in winter mostly from the inside to the outside, in summer and with a damp façade also the other way round. How easily this happens is described by the sd-value of a material.
  • Sorption: Hygroscopic building materials absorb water vapour from the air, store it in their pores and release it again when the air is dry. In this way, they buffer moisture peaks in the room.
  • Capillary transport: Liquid water is conducted through the fine pores of brick, lime mortar and lime plaster and can evaporate at the surface.

A house from the 19th century usually has no damp-proof course against rising damp, and the foundations sit directly in the soil. Moisture therefore comes from below, from outside with driving rain and from the indoor air inside. Such walls have worked for well over a hundred years because they can absorb water and release it again to both sides. This ability to dry out is what I really meant by “breathing”.

In short: Walls do not breathe in the sense of air exchange. But old masonry must remain vapour-open towards the inside and the outside. Any layer that slows drying in one direction, whether an insulation board, a vapour retarder, a cement render or a dense paint, can cause moisture to build up behind it.

Field-fired bricks: why my wall is not an ordinary solid brick wall

The forum post spoke of a “solid brick wall”. That is an understandable misconception, because from the outside the bricks look similar. Strictly speaking, though, my house is built of field-fired bricks. These are bricks that were not fired in a brickworks with an industrial kiln but in simple field kilns or clamps, often close to the building site.

This gives them different properties from a modern, industrially fired solid brick:

  • Lower and uneven firing temperature: The bricks are less sintered, more porous and lighter. One and the same wall can contain bricks of different quality.
  • Better thermal protection: Because of the high proportion of pores, their thermal conductivity is roughly in the range of a light clay block and thus well below that of a modern solid brick.
  • Highly hygroscopic and absorbent: Field-fired bricks take up moisture from the air and liquid water much faster and in greater quantities than a modern solid brick. But they also release it again if you let them.
  • More sensitive: Softer, less strongly fired bricks are sensitive to frost and salts if they stay wet permanently. A capillary-active lime plaster protects them without sealing them in.

For the question of whether I should insulate my brick wall, this has two consequences: the wall insulates better on its own than the forum calculation assumes. And its biggest risk is not the cold, but moisture that can no longer dry out.

The numbers: what does my 60 cm field-fired brick wall really achieve?

The claim in the forum was that my 60 centimetres of brick insulate as well as 3 centimetres of polystyrene. I did the maths, once for a typical solid brick and once for my field-fired brick.

How the U-value works

The U-value (thermal transmittance) indicates how much heat is lost per square metre of wall area and per degree of temperature difference between inside and outside. The unit is W/(m²K). The lower the U-value, the better the wall insulates.

It is calculated from the thermal resistance R of each layer: thickness divided by thermal conductivity (λ, lambda). The resistances of all layers are added up, plus small surface resistances. The U-value is then the reciprocal of the sum.

Solid brick, field-fired brick and polystyrene compared

A typical solid brick has a thermal conductivity of about 0.68 W/(mK). For my field-fired brick, I calculate with around 0.4 W/(mK), based on light clay blocks. Because the bricks vary so much, I also looked at a range of 0.3 to 0.5 W/(mK). EPS (polystyrene) is around 0.035 W/(mK).

  • 60 cm solid brick: 0.60 m ÷ 0.68 W/(mK) = R ≈ 0.88 m²K/W, equivalent to about 3 cm of EPS
  • 60 cm field-fired brick: 0.60 m ÷ 0.4 W/(mK) = R ≈ 1.5 m²K/W, equivalent to about 5 cm of EPS
  • 3 cm EPS: 0.03 m ÷ 0.035 W/(mK) = R ≈ 0.86 m²K/W

For an ordinary solid brick, the statement from the forum is therefore correct. My 60 centimetres of field-fired brick, on the other hand, are equivalent to about 4 to 7 centimetres of polystyrene, depending on the brick. Including plaster and surface resistances, I arrive at a U-value of around 0.58 W/(m²K), within a range of about 0.45 to 0.7.

For comparison: in Germany, the Building Energy Act (GEG) requires a U-value of no more than 0.24 W/(m²K) when an exterior wall is renovated, but provides exemptions for listed buildings and historic building fabric worth preserving. Romania has its own requirements, currently the Mc 001/2022 standard. I use the German value for orientation only.

Damp field-fired bricks insulate worse

The lambda value from tables applies to dry material. Water conducts heat many times better than the air in the pores. Precisely because field-fired bricks have so many pores and absorb so much water, their actual thermal protection depends heavily on their moisture content. For me, this is the most important idea in this article: a dry field-fired brick wall is itself a useful thermal barrier. If you keep it dry, you improve its insulating effect without fixing a single board to it. If, on the other hand, you add insulation that traps moisture behind it, you worsen precisely this property and also risk damaging the brick.

What insulation would achieve on paper

For the sake of completeness, I calculated several variants with λ = 0.4 W/(mK) for the field-fired brick. The values are approximations and apply only to dry masonry. Such a calculation does not reflect possible moisture damage:

Build-up (on 60 cm field-fired brick, plastered) U-value approx. in W/(m²K) Heat loss compared to existing wall
Existing wall without insulation 0.58 100 %
+ 5 cm insulating lime plaster (λ ≈ 0.07) 0.41 approx. 71 %
+ 3 cm EPS (λ 0.035) 0.39 approx. 67 %
+ 6 cm wood fibre (λ 0.040) 0.31 approx. 53 %
+ 10 cm EPS or mineral wool (λ 0.035) 0.22 approx. 38 %

This clearly puts the forum statement into perspective: on my field-fired brick wall, 3 centimetres of polystyrene do not halve the losses, they reduce them by around a third. A noticeable improvement would require 6 to 10 centimetres, which are exactly the thicknesses at which the intervention in the wall’s moisture behaviour is greatest. For my house, benefit and risk are therefore in a different relationship than for an ordinary solid brick.

My wall also performs better than expected in terms of comfort. With minus 10 degrees outside and 20 degrees inside, its inner surface reaches just under 18 degrees by calculation. With 10 centimetres of insulation, it would be around 19 degrees. One degree more on the wall surface is pleasant, but for my house it does not justify the risk to the building fabric. On top of that, my wall heating warms the wall surfaces directly anyway. I describe how this works further down.

My tip: Calculate your own wall yourself, but with the right material. In the U-value calculator by ubakus.de, you can enter your wall build-up layer by layer and adjust the thermal conductivity if your material is not in the database. Whether you calculate with 0.4 or 0.68 W/(mK) changes the result considerably. But bear in mind: the U-value says nothing about how moisture behaves in an old, absorbent wall over the years. Only a hygrothermal simulation, for example with WUFI, or the experience of specialists in historic buildings can tell you that.

Insulating brick walls in historic buildings: why I decided against it

For an ordinary old building made of solid bricks with dry masonry, the person in the forum has a point in terms of building physics: if you insulate at all, outside is usually better than inside, because the wall then lies on the warm side. For historic building fabric like mine, however, experts in heritage conservation and old building renovation often advise against insulating the exterior walls altogether. Based on everything I worked out before starting the restoration, I share this view. The reasons:

  1. Moisture builds up behind the insulation: With external insulation, the wall dries out less well towards the outside. With internal insulation, the wall becomes colder, and condensation can form at the boundary between insulation and masonry. In both cases, moisture collects exactly where you cannot see it.
  2. Field-fired bricks are particularly sensitive: Because they absorb so much water, salts accumulate in a permanently damp wall. Frost can then spall the soft brick from the outside, and plaster comes loose.
  3. Rising damp cannot be insulated away: Without a damp-proof course, the wall will always draw moisture from the ground. It must be able to release this moisture to both sides.
  4. Lime plaster and lime mortar: Historic building fabric is designed for mineral, capillary-active materials. Even a cement render or an emulsion paint can disturb this system, let alone an insulation layer.
  5. The village character: The street façades of the Saxon villages with their gables, cornices and gateways are part of a cultural heritage. That is why a façade would have to be rebuilt after insulating. I have found no information on whether the thick layers of plaster this requires would hold on the insulation or affect it. Depending on the location, heritage protection requirements also apply in Transylvania, which should be clarified with the responsible authority beforehand.
Important: If you are still considering wall insulation on an old house, only use materials that are vapour-open and capillary-active in both directions, and only after professional advice from someone experienced with historic buildings. Check the moisture situation first: is the wall wet at the base, are salts blooming, is the brick crumbling? Then insulation has no place on this wall until the cause has been fixed.

Insulation materials: what is out of the question for my house

The forum’s point that there are many insulation materials is true. For a wall made of field-fired bricks, however, this variety narrows down quickly. Here is my assessment:

Ruled out: EPS, ETICS and vapour retarders

Polystyrene is cheap and insulates very well, but it has a high vapour diffusion resistance and is not capillary-active. On an absorbent, moisture-laden wall, it is for me the opposite of what the wall needs. The same applies to external thermal insulation composite systems (ETICS) with synthetic resin render and to internal insulation with a vapour retarder or vapour barrier.

Critical: wood fibre and mineral wool

Both are vapour-open and much more compatible than EPS. However, they are not capillary-active and still change where and how the wall dries out. On dry masonry without heritage status, they can be an option. For my field-fired brick wall, I consider the risk higher than the benefit, even without heritage status.

Critical: capillary-active internal insulation

Calcium silicate or mineral insulation boards are considered the most compatible form of internal insulation in old buildings. But they, too, make the wall colder and shift its moisture behaviour. Without a simulation and professional support, I would not tackle this in my house.

Conceivable at most: lime plaster and insulating lime plaster

For me, a good, sufficiently thick lime plaster is not insulation in the true sense but part of the moisture protection. It is vapour-open, capillary-active and protects the brick. An insulating lime plaster with lightweight aggregates such as perlite improves thermal protection somewhat but remains part of the wall’s system. If I change anything on the walls at all, it will be in this direction and only after consulting specialists.

My tip: With field-fired bricks, the best lambda value is not what counts, but whether the wall can still dry out in both directions afterwards. If you do not get a clear answer to that, you are better off leaving the wall as it is and investing in the other building components.

Where I save heat instead

Precisely because my field-fired brick wall insulates better than initially assumed and I do not want to insulate it, the other building components come to the fore. Here, a lot can be achieved with little risk:

  • Top floor ceiling and roof: Warm air rises. Insulating the ceiling to the unheated attic is usually one of the cheapest measures with a big effect, and many people can do it themselves. In my house, the ceiling to the attic had already been insulated with a clay-and-straw fill.
  • Windows and doors: Single-glazed windows and leaky doors cause draughts. Often it helps just to refurbish old windows, seal them or add a secondary inner window. Here, too, the living areas of my house already had box-type double windows, which I simply upgraded, fitting the inner sashes with a seal.
  • Floor: A lot of heat is lost, especially with floorboards directly over the ground or over a cellar. My house already had a clay fill here as well. Only the former carpentry workshop had no cellar and just a rammed earth floor with boards over it. There, I dug down to just above the bottom edge of the foundation, filled it with drainage gravel and installed a 10 cm layer of rigid EPS foam before laying the new floor.
  • Air leaks: Cracks, openings and open chimney flues cause uncontrolled air exchange.
  • Dry walls: Splash-water protection, intact gutters and an exposed plinth keep the wall dry and thus improve its insulating effect.
  • Radiant heat: A heating system that warms the walls keeps the masonry dry and provides comfort even when the air is a little cooler. In my house, wall heating takes care of this.

My energy concept: wall heating, heat pumps, solar power and a wood stove

I do not want to gloss over anything: an uninsulated wall loses more heat than an insulated one, and I need more energy as a result. Instead of wrapping up the wall, I therefore started with the building services. My setup consists of several components that complement each other:

  • Wall heating with radiant heat as the main heating system
  • Air-to-water heat pump supplying the wall heating
  • Split air conditioner for cooling, heating in spring and autumn and, above all, dehumidifying
  • Heat pump water heater for hot water
  • Solar panels on the roof, set up as a plug-in solar system with battery storage, for my own electricity
  • Wood stove to use the firewood from my own property

Wall heating: radiant heat like a tiled stove

With wall heating, water-carrying pipes are embedded in the wall plaster. The wall becomes a large, evenly warm heating surface and gives off its heat mainly as radiation, much like a tiled stove. Radiant heat warms people and surfaces directly rather than primarily the air. That is why a room feels comfortable even at a slightly lower air temperature, and less dust is stirred up than with radiators.

For an old field-fired brick wall, this has another advantage: warm wall surfaces keep the masonry dry on the inside. Condensation and mould in cold corners hardly stand a chance. In this way, the wall heating supports exactly what this article is about: a wall that stays dry and can dry out in both directions.

Low flow temperature for the heat pump

Because the heating surface is so large, a low flow temperature is sufficient. A heat pump is deliberately designed and set up for the lowest possible flow temperature. In my house, it is around 10 degrees below what gas heating systems are typically run at. This is exactly the range in which an air-to-water heat pump works efficiently. As a rule of thumb, every degree less flow temperature improves the efficiency of a heat pump by about 2 to 2.5 percent. With a difference of 10 degrees, that is a noticeable effect.

To be honest, there is also this: wall heating on an exterior wall raises its temperature and thus slightly increases the heat loss to the outside. I accept this disadvantage deliberately, because in my house the benefits for comfort, efficiency and building fabric outweigh it.

Split air conditioner: cooling, heating and above all dehumidifying

An air conditioner in a farmhouse from 1888 may sound unusual at first. For me, though, it is an important component, for three reasons:

  • Dehumidifying: This is the most important point for me. On humid summer days, the outside air brings a lot of moisture into the house, which can settle in the cooler, hygroscopic walls. The air conditioner lowers the humidity in the room. This protects the field-fired bricks and the plaster and helps the wall dry out towards the inside as well.
  • Heating in spring and autumn: When it is only cool in the mornings and evenings, the split air conditioner heats quickly and efficiently as an air-to-air heat pump, without the slow-reacting wall heating having to start up.
  • Cooling: Summers are getting hotter in Transylvania, too. The thick walls stay cool for a long time, but at some point even they heat up.

My unit has no app control. So I built my own connector that communicates with the unit via the remote control interface. This lets me integrate it into my smart home and control it automatically, for example depending on outdoor and indoor humidity or on the solar power available.

Heat pump water heater for hot water

Hot water is provided by a heat pump water heater. Together with the split air conditioner, it was originally intended as an interim solution until my final setup was decided. Both have proven themselves so well, however, that they will stay in use.

Solar power from the roof reduces costs

On my roof there is a solar system with 2,000 Wp of module capacity and an inverter with 800 W output, set up as a plug-in solar system (in German: “Balkonkraftwerk”, balcony power plant). An EcoFlow Delta 3 Plus is currently connected to it as battery storage. It stores surplus solar power during the day and releases it in the evening and at night, when the modules no longer deliver anything. Together, they supply part of the electricity for the heat pumps, the air conditioner and the household and reduce running costs. This way, I can offset a good part of the extra consumption caused by the uninsulated wall without interfering with the building fabric.

To be honest, it is not quite perfect: the heat pump for the wall heating needs the most electricity in winter, when the solar system delivers the least. Other consumers, on the other hand, match the sun all the better. The air conditioner cools and dehumidifies exactly when the modules deliver the most power. The heat pump water heater can heat its water with solar power during the day. And in spring and autumn, when heating is still needed and the sun is already or still shining strongly, my own electricity makes a noticeable contribution to lowering costs. A contribution to environmental protection and sustainability comes on top.

By the way, a plug-in solar system is a good way to start if a large rooftop installation is not (yet) an option. It mainly covers the household’s base load during the day and, with a small storage unit such as a portable power station, part of the consumption in the evening. Across the EU, such plug-in solar devices may feed a maximum of 800 W into the grid via the micro-inverter. In Germany, the module capacity is additionally limited to 2,000 Wp. How the system has to be registered differs from country to country: in Germany, it only needs to be entered online in the Marktstammdatenregister (the national registry), which takes about 10 minutes. In Romania and other countries, you should check the local rules beforehand.

A wood stove for my own firewood

The last component is a wood stove. Firewood accumulates regularly on the property, and I would rather use it in my own stove than leave it lying around. The stove also provides pleasant radiant heat and steps in exactly when it is coldest outside and the solar system delivers the least power.

The charm of the old farmhouse remains

One point matters to me almost as much as the technology: there are no radiators on the walls in my house. The rooms keep their original look and shape, nothing obstructs walls or window recesses, and the charm of the old farmhouse is preserved inside. The main heating is there, but you cannot see it.

My tip: If you have an old house with solid, uninsulated walls, it is worth taking a look at surface heating such as wall or underfloor heating. Large heating surfaces allow low flow temperatures, and that is exactly what you need for a heat pump to work economically in an old building, too. Make sure the plaster around the heating pipes suits the masonry and stays vapour-open.

How I am approaching my house in Cincu

The discussion in the forum sharpened my terminology and confirmed my position. This is how to approach the matter systematically:

  1. Survey of the existing building: Check where the walls are damp, where plaster is flaking off, where bricks are crumbling and where salts are blooming. Measuring the moisture in the masonry is part of this.
  2. Eliminate the causes: Keep rainwater away from the plinth, check gutters and downpipes, lower the soil at the plinth where it is too high.
  3. Remove disruptive layers: Where cement render or dense paints have been applied, they must be removed and replaced with lime plaster and lime paint.
  4. Start at the top: Insulate the ceiling to the attic if this has not been done yet.
  5. Windows and doors: Refurbish them and seal them properly without losing their historic character.
  6. Leave the walls alone: The exterior walls remain uninsulated and vapour-open in both directions.

I report on the individual steps here on the blog, including the ones where something does not work out.

Contacts and further information

If you have an old house with solid brick walls yourself, these sources will help you further. Most of them are in German:

  • ubakus.de – U-value calculator: free online calculator for wall build-ups with condensation and temperature calculation.
  • Interessengemeinschaft Bauernhaus: association with a great deal of experience in preserving old rural buildings.
  • Mihai Eminescu Trust: foundation that has been restoring houses in the Saxon villages of Transylvania with traditional techniques for many years.
  • Deutsche Stiftung Denkmalschutz (German Foundation for Monument Protection): the largest private initiative for the preservation of historic monuments in Germany, with information on the careful renovation of old buildings.
  • WTA guideline 6-4 “Internal insulation according to WTA I”: the planning guide of the International Association for Science and Technology of Building Maintenance and Monuments Preservation (WTA), helpful for understanding the risks of internal insulation.
  • WUFI by Fraunhofer IBP: software for hygrothermal simulation, used by specialist planners to calculate the moisture behaviour of walls over several years.
  • Zukunft Altbau: independent information programme on the energy-efficient renovation of existing buildings.

Frequently asked questions about insulating old brick walls

What is a field-fired brick?

A field-fired brick is a historic brick that was not fired in a brickworks with an industrial kiln, but in a simple field kiln or clamp, often right next to the building site. Because of the lower and uneven firing temperature, it is more porous, lighter and much more absorbent than a modern solid brick.

How well does a 60 cm thick field-fired brick wall insulate?

Better than a wall made of modern solid brick. With a thermal conductivity in the range of a light clay block, roughly 0.3 to 0.5 W/(mK), its insulating effect is equivalent to about 4 to 7 centimetres of EPS. Depending on the brick and its moisture content, the U-value is about 0.45 to 0.7 W/(m²K). A 60 cm wall made of ordinary solid brick, by contrast, is equivalent to only about 3 centimetres of EPS.

Should you insulate an old brick wall?

For historic building fabric with field-fired bricks, lime mortar and no damp-proof course, it is often advised against altogether, because moisture can build up behind the insulation and damage the masonry. If you still want to insulate, you should only use materials that are vapour-open and capillary-active in both directions and seek professional advice.

Do old walls need to breathe?

No, walls exchange practically no air. The myth goes back to a flawed experiment by Max von Pettenkofer in 1858. Old walls must, however, be vapour-open so that moisture entering from below, through driving rain or from the indoor air can dry out again towards the inside and the outside.

What does “vapour-open in both directions” mean?

It means that no layer of the wall, neither inside nor outside, significantly slows the passage of water vapour. This allows the wall to dry out towards the drier side, depending on the season and the weather. Lime plaster and lime paint meet this requirement; cement render, emulsion paint, vapour retarders and polystyrene do not.

How can I improve the energy efficiency of an old house without wall insulation?

The biggest levers are usually the top floor ceiling, the windows and doors, the floor and air leaks. In addition, the walls should be kept dry, because dry masonry insulates better than wet masonry. On the building services side, surface heating such as wall heating, combined with a heat pump, your own solar power and possibly a wood stove, helps to offset the higher heat demand of an uninsulated wall. An air conditioner with a dehumidifying function helps keep the indoor air, and thus the walls, dry in summer.

Is wall heating suitable for an old farmhouse?

Yes, often very much so. Wall heating gives off radiant heat similar to a tiled stove, keeps the wall surfaces warm and dry and works with low flow temperatures. In my house, the air-to-water heat pump runs at a flow temperature around 10 degrees below what gas heating systems typically use, which makes it more efficient. In addition, the rooms remain free of visible radiators.

What I take away from the discussion

I am grateful for this exchange. It does not happen often that someone points out a mistake so clearly and at the same time so appreciatively. My statement that walls need to breathe was technically wrong, and I will no longer use it that way. In future, I will say more precisely what I meant: old walls must remain vapour-open in both directions.

The discussion also shows how important it is to know your own material precisely. With the values of an ordinary solid brick, a wall like mine is rated worse than it actually is. And it is easy to overlook where its real risk lies: in the high absorbency of the field-fired bricks. For a house like mine, the answer to the question of whether to insulate the brick wall is therefore: better not. I put my money and my working time into the roof, ceiling, windows, floor and dry walls. I offset a good part of the higher heat demand with wall heating, heat pumps, my own solar power and wood from my own property. The air conditioner helps me keep the walls dry in summer. And all this without a single radiator spoiling the charm of the old rooms.

How about you? Have you had any experience with field-fired bricks or old brick walls, insulated or uninsulated, perhaps even in a house in Transylvania? Feel free to write to me in the comments. I am happy about every tip, and as this story shows, I like to keep learning.

Posted on

Sonoff POWR316 Tasmota: Common Flashing Errors and How to Fix Them

Sonoff POWR316 Tasmota flashen: geöffneter Sonoff POWR316, über vier Jumperkabel mit einem ESP32-Devboard verbunden, daneben das geschlossene Gerät

Sonoff POWR316 with Tasmota: This is the combination I use to monitor the power consumption of my air conditioner. Flashing the Sonoff POWR316 with Tasmota sounds like a ten-minute job. For me, it turned into a whole afternoon. The POWR316, also known as the Sonoff POW Origin, has an ESP32 on board and behaves differently during flashing than older Sonoff devices with an ESP8266. In this article, I’ll show you which errors I ran into, what they mean, and which approach finally worked reliably: an ESP32 dev board as a USB-to-serial adapter combined with Espressif’s web flasher esptool-js.

In a nutshell: You flash the Sonoff POWR316 with Tasmota by connecting its 3V3, GND, RX and TX pads to a USB-to-serial adapter, holding down the button while powering it up, and writing the file tasmota32.factory.bin to address 0x0. The most common stumbling block is the power supply: many USB-TTL adapters cannot deliver enough current for the ESP32. What worked reliably for me was an ESP32 dev board as a programmer together with the flasher esptool-js.

Tested in October 2026 with: Sonoff POWR316 (ESP32-D0WD-V3, 4 MB flash), Tasmota 16.0.0 (German build tasmota32-DE), ESP32-WROOM-32 dev board with CP2102 chip, esptool-js in Chrome on Windows.

Why I rely on Sonoff POWR316 Tasmota

I am expanding my smart home step by step. This time, I want to measure the power consumption of my split air conditioner. I want to know how much it actually uses for cooling in summer and for heating in spring and autumn, and analyze the data in my smart home system.

For tasks like this, I like to use Sonoff. The devices are inexpensive, easy to integrate and mount nicely on a DIN rail. The POWR316 switches up to 16 A and measures voltage, current, power and energy. That makes it a good fit for an air conditioner.

With the original firmware, however, the devices depend on the eWeLink cloud. I want to control my devices locally, without a detour via a server abroad. That’s why every Sonoff device I own runs Tasmota. Tasmota is a free, open-source firmware for ESP chips. It comes with its own web interface, speaks MQTT and integrates with my OpenCCU-based setup as well as with Home Assistant, ioBroker, openHAB or FHEM.

Unfortunately, it’s almost always the same story: whenever I plan an expansion, there’s a new Sonoff generation, and flashing works a little differently than the last time. So that you don’t have to spend as much time trial-and-erroring as I did, I documented everything along the way.

What’s inside the Sonoff POWR316?

Before you get started, you should know what hardware you’re dealing with. This is what esptool-js read from my device during flashing:

Property Value
Chip ESP32-D0WD-V3 (revision 3)
Cores / clock Dual core, 240 MHz
Wireless Wi-Fi and Bluetooth
Flash memory 4 MB, DIO mode
Matching Tasmota build tasmota32 (I used the German build tasmota32-DE)

The key difference from the ESP8266: the ESP32 draws considerably more current at startup and especially when Wi-Fi is switched on, briefly 300 to 500 mA. That turned out to be my downfall, more on that in a moment.

When you first connect, check which chip the flasher reports. If it says “Dual Core”, use the regular tasmota32 build. Only a single-core chip (ESP32-SOLO1) needs tasmota32solo1. Sonoff occasionally changes chips between hardware revisions.

Safety first: never flash while connected to mains

The POWR316 is a mains-powered device. As long as the housing is open, it must never be connected to 230 V. During flashing, you power it exclusively via the 3.3 V from the adapter. If it is connected to the mains at the same time, this is life-threatening and can destroy your computer.

Also make sure that only 3.3 V is present on RX and TX. Many USB-TTL adapters have a jumper for 5 V and 3.3 V. If it is set to 5 V, the data lines also carry 5 V, and the ESP32 cannot handle that.

Caution, mains voltage: Only reconnect the Sonoff to 230 V once all cables to the PC have been removed and the housing is completely closed. Work inside the distribution board and installation on the DIN rail should be left to a qualified electrician.

What you need for flashing

  • a Sonoff POWR316 (opened, not connected to mains)
  • an ESP32 dev board (e.g. ESP32-WROOM-32 DevKit) or a USB-to-serial adapter with a strong 3.3 V supply
  • a four-pin header and five short jumper wires
  • a soldering iron (highly recommended)
  • the Chrome or Edge browser, as only these support Web Serial
  • the firmware file tasmota32.factory.bin from the official Tasmota server (or a language build such as tasmota32-DE.factory.bin)
  • optionally a multimeter to check the voltage

The errors I ran into – at a glance

Before I go into detail, here is the overview. If you’re currently staring at one of these messages yourself, this helps you quickly find the right section of the article.

Message Cause Solution
Couldn’t sync to ESP ESP32 doesn’t respond: not in flash mode, RX/TX wrong or not enough power stronger power supply, hold the button while powering up
Invalid head of packet (0x65) ESP32 sends normal boot text instead of a bootloader response force flash mode, keep the button pressed
Endless POWERON_RESET voltage drops dev board or external 3.3 V source, solder in a pin header
invalid header: 0x20202020 no valid bootloader in flash erase and reflash with esptool-js, check the checksum
Reboot right after “WifiManager active” Wi-Fi startup draws more current than USB can supply reassemble the Sonoff and start it with its own power supply

My first attempt with the USB-TTL adapter

I first tried the classic way: with my USB-TTL adapter with a CH340 chip. I had connected 3V3, GND and RX/TX crossed, i.e. TX of the adapter to RX of the Sonoff and vice versa. Then I opened the Tasmota Web Installer, selected my firmware file, pressed the button on the Sonoff, plugged in USB, selected the port and clicked Install.

Result: an error. Same with a second adapter with an FTDI chip. Swapping RX and TX didn’t change anything either.

Error 1: “Couldn’t sync to ESP”

The message read:

Failed to enter bootloader mode: Couldn't sync to ESP.
Try resetting manually. Last error: Sync timeout or abandoned

At first, I thought the connection was established, because the installer showed a Device Dashboard after clicking “Connect”. However, that only means the adapter’s COM port was opened. It says nothing about whether the ESP32 behind it is responding.

“Couldn’t sync to ESP” means that the ESP32 isn’t responding at all. The usual causes are:

  • The ESP32 is not in flash mode because the button wasn’t pressed during power-up.
  • RX and TX are connected incorrectly.
  • The power supply is insufficient.

In my case, it was the power supply. According to its datasheet, an FTDI adapter with the FT232RL chip delivers only about 50 mA on its 3.3 V pin. Many CH340 adapters don’t manage much more. That’s not enough for an ESP32. The voltage drops, and the chip doesn’t start properly.

The solution: an ESP32 dev board as a programmer

An ESP32 or ESP8266 dev board has two things on board that I needed: a USB-to-serial chip (a CP2102 in my case) and a strong 3.3 V voltage regulator that delivers much more current than a cheap adapter.

The trick: you tie the dev board’s EN pin firmly to GND. This keeps the ESP32 on the dev board permanently in reset so it doesn’t interfere. What remains is the USB-to-serial chip, which now talks directly to the Sonoff. On an ESP8266 board such as the NodeMCU, the pin is called RST.

Wiring

ESP32 dev board Sonoff POWR316
EN connect to GND on the dev board
3V3 3V3
GND GND
TX0 (GPIO1) TX
RX0 (GPIO3) RX

Yes, you read that right: here the lines are not crossed; TX goes to TX and RX to RX. The labels on the dev board refer to its own ESP32. The USB chip transmits on that ESP32’s RX line. With the board’s own ESP32 disabled, this results in a 1:1 connection. If it doesn’t work this way, swap the two lines as a test. Nothing can break.

On the common ESP32-WROOM-32 dev boards (DOIT DevKit V1 or ESP32-DevKitC), you’ll find TX0 and RX0 next to each other between pins D22 and D21. EN is on the opposite side at the very top, next to 3V3.

I always solder a pin header into my Sonoff devices. Many people just plug the wires in loosely, but that makes things much more error-prone. A loose contact on 3V3 or GND leads to exactly the errors I describe here and wastes a lot of time.

Error 2: “Invalid head of packet (0x65)”

With the dev board, a new message appeared:

Failed to enter bootloader mode: Invalid head of packet (0x65)

That was progress, because data was now coming through. 0x65 is the character “e”. So the installer received normal text instead of a bootloader response. That’s typical of the beginning of the ESP32 boot message ets Jul 29 2019 …. The ESP32 was running, but not in flash mode.

The console helps you understand

I only really made progress once I looked at the Sonoff’s output in the console. In the Web Installer, you’ll find it under “Logs & Console”. This line appeared there, over and over again:

rst:0x1 (POWERON_RESET),boot:0x13 (SPI_FAST_FLASH_BOOT)

This line tells you two things:

  • boot:0x13 means: the ESP32 boots normally from flash. For flash mode, it should say boot:0x3 (DOWNLOAD_BOOT) followed by “waiting for download”.
  • POWERON_RESET in an endless loop means: the voltage drops so far that the chip restarts completely.
Message in the console Meaning
no output no connection: check wiring or power
boot:0x13 (SPI_FAST_FLASH_BOOT) normal boot, not in flash mode
boot:0x3 (DOWNLOAD_BOOT) / waiting for download flash mode active, you can flash
constant POWERON_RESET or truncated lines power supply drops
invalid header: 0x… no valid bootloader in flash

The decisive trick: don’t let go of the button

In the end, the solution for flash mode was simple. I kept the button on the Sonoff pressed during the entire flashing process, from plugging in the USB cable until the message that the installation was complete.

The background: the button pulls GPIO0 to ground. Only if this happens during power-up does the ESP32 start in flash mode. In addition, my adapter has no DTR and RTS lines to the Sonoff. The flasher therefore can’t restart the ESP32 and put it into the bootloader by itself. If the chip restarts in between, for example due to a brief voltage drop, it ends up in normal mode again without the button pressed, and the connection breaks.

My tip for flash mode: Press the button, only then apply power, and keep holding the button until “Installation complete” or “Hash of data verified” appears. Only then let go.

Error 3: “invalid header: 0x20202020”

The Web Installer now reported “done”. I restarted the Sonoff without the button pressed and only saw this in the console:

invalid header: 0x20202020
invalid header: 0x20202020
ets Jul 29 2019 12:21:46
rst:0x10 (RTCWDT_RTC_RESET),boot:0x13 (SPI_FAST_FLASH_BOOT)

“Invalid header” means that the ESP32 can’t find valid code at the location in flash where its bootloader should be. 0x20 is the space character. So there was text instead of firmware. There are two possible causes:

  • The firmware file is corrupt. This often happens when you use “Save link as” on GitHub and download the HTML page instead of the actual file. A genuine tasmota32.factory.bin is about 3 MB in size. My file had the correct size.
  • The write process didn’t complete cleanly without the installer noticing.

A second attempt with the online firmware from the Web Installer didn’t help either. Afterwards, the console showed truncated boot messages, a clear sign of an unstable power supply. Which of the two causes it was exactly could no longer be determined. But the next step solved the problem reliably.

The Sonoff won’t be damaged by failed flashing attempts. The bootloader in the ESP32’s ROM cannot be overwritten, so you can always reflash.

The approach that worked: esptool-js by Espressif

The breakthrough came with esptool-js, the web flasher from ESP32 manufacturer Espressif. Like the Tasmota Web Installer, it runs directly in the browser, but it shows a detailed log. Above all, it verifies via checksum at the end whether the data has really arrived correctly in flash.

Step by step: flashing the Sonoff POWR316 with esptool-js

  1. Wire the ESP32 dev board as described above (EN to GND).
  2. Hold down the button on the Sonoff and then plug the dev board into your computer via USB.
  3. Open espressif.github.io/esptool-js in Chrome or Edge.
  4. Set the baud rate to 115200 and click “Connect”. Select the dev board’s COM port.
  5. Check the log: it should show “Chip is ESP32-D0WD-V3” and the flash ID.
  6. Click “Erase Flash” and wait until the process is complete.
  7. Enter 0x0 as the flash address and select the file tasmota32.factory.bin.
  8. Click “Program”. Keep holding the button and don’t move anything.
  9. Wait until the log shows “Hash of data verified.”
  10. Unplug USB, release the button.

For me, the end of the log looked like this:

Wrote 3153024 bytes (1964023 compressed) at 0x0 in 181.571 seconds.
File  md5: 67622a083788b5ec91a4657ada1e0627
Flash md5: 67622a083788b5ec91a4657ada1e0627
Hash of data verified.

The two MD5 checksums are identical. This confirms that the firmware is stored correctly in memory, byte for byte. At 115200 baud, the process takes about three minutes. That’s slow, but stable.

You can ignore the line “Hard resetting via RTS pin…” at the end. Without a connected RTS line, nothing happens. You do the restart yourself by briefly unplugging USB.

The key points for flashing: A strong 3.3 V supply via an ESP32 dev board (EN to GND), keep the button pressed during the entire process, use esptool-js to first “Erase Flash” and then write the .factory.bin to address 0x0. Only when “Hash of data verified” appears in the log is the firmware guaranteed to be stored correctly.

What about the original firmware?

Once you erase the flash memory, the eWeLink firmware is gone. The only way back is a backup made beforehand. If you want to keep that option open, read out the memory before erasing it with the command-line tool esptool, for example with esptool.py read_flash 0 0x400000 backup.bin. I skipped this because I don’t intend to use the cloud firmware anymore anyway.

Tasmota starts, but no Wi-Fi appears

After flashing, I restarted the Sonoff without the button. The console finally showed Tasmota (in German, as I had flashed the German build):

00:00:00.002 HDW: ESP32-D0WD-V3 v3.1
00:00:01.403 Projekt tasmota - Tasmota Version 16.0.0(release-tasmota32)
00:00:01.813 WIF: WifiManager aktiv für 3 Minuten
ets Jul 29 2019 12:21:46
rst:0x1 (POWERON_RESET),boot:0x13 (SPI_FAST_FLASH_BOOT)

So Tasmota was running. But at exactly the moment it tried to switch on Wi-Fi, the ESP32 restarted again. The Wi-Fi radio draws the most current when it switches on, and even the dev board can’t handle that via USB. The supply is sufficient for flashing, but not for Wi-Fi operation.

The messages QPC: Count and FRC: Some settings have been reset are harmless. Tasmota counts quick restarts and resets the settings after several of them. On a freshly flashed device, nothing is lost.

The solution: the Sonoff needs its own power supply

  1. Remove all cables to the dev board.
  2. Fully reassemble the Sonoff and close the housing.
  3. Connect the Sonoff to 230 V as usual.
  4. Use your smartphone to search for the Wi-Fi network tasmota-XXXXXX-XXXX and connect to it.
  5. If the login page doesn’t open automatically, open 192.168.4.1 in your browser.
  6. Select your own Wi-Fi network, enter the password and save.

The Sonoff then connects to your home network. You’ll find its IP address in your router’s device list.

Tasmota only offers the hotspot for three minutes. If you missed it, briefly disconnect the Sonoff from power and reconnect it.

If Tasmota runs stably on your programmer, you can alternatively enter the Wi-Fi credentials directly in the console:

Backlog SSID1 YourWiFi; Password1 YourPassword

Setting up the Sonoff POWR316 Tasmota template

After the first start, Tasmota doesn’t yet know which pins the relay, button, LEDs and energy monitoring chip are connected to. For that, you need a template. This one works on my POWR316:

{"NAME":"Sonoff POWR316","GPIO":[32,0,0,0,0,576,0,0,0,224,0,0,3104,0,320,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0],"FLAG":0,"BASE":1}

The template defines the button, the relay, the two status LEDs and the receive pin for the energy monitoring chip. Here’s how to apply it:

  1. Open the Sonoff’s web interface via its IP address.
  2. Go to Tools → Console.
  3. Type Template, a space and then the complete template, and press Enter.
  4. Then enter Module 0. This activates the template, and the Sonoff restarts.
  5. After the restart, the main page shows the relay toggle button and the readings for voltage, current, power and energy.

You’ll find templates for other devices in blakadder’s Tasmota template repository.

For accurate power monitoring, it’s worth calibrating afterwards with a known load, such as an incandescent bulb or a fan heater. The Tasmota documentation on calibration explains how.

Settings for monitoring the air conditioner

With an air conditioner, my main goal is to measure, not to switch. That’s why I adjusted a few things in the Tasmota console:

  • PowerOnState 1 ensures that the relay always starts switched on after a power outage. This way, the air conditioner doesn’t stay off unintentionally.
  • TelePeriod 60 sends the readings via MQTT every minute. The default is 300 seconds.
  • Under Configuration → Configure MQTT, you enter your MQTT broker. Home Assistant then automatically detects Tasmota devices via the Tasmota integration.
My tip for the air conditioner: Don’t simply switch off a running air conditioner via the relay. The compressor should be shut down by the unit’s own controls. I only use the POWR316’s relay as a measuring point here.

By the way, you can easily install future Tasmota updates via the web interface under Firmware Upgrade. You won’t need to open the Sonoff again for that.

Checklist: flashing Sonoff POWR316 Tasmota without frustration

  • Power supply: An ESP32 needs up to 500 mA. Use an ESP32 dev board or an external 3.3 V source, not a weak USB-TTL adapter.
  • Wiring: Solder in a pin header and keep the wires short. With the dev board and EN tied to GND, connect TX to TX and RX to RX.
  • Flash mode: Press the button before powering up and keep it pressed during the entire flashing process.
  • Check: The console must show “DOWNLOAD_BOOT” or “waiting for download”.
  • Tool: Use esptool-js with “Erase Flash” and address 0x0. Look for “Hash of data verified”.
  • First start: Start the reassembled Sonoff with its own power supply and set up Wi-Fi via the hotspot.
  • Browser: Only use Chrome or Edge; Firefox and Safari don’t support Web Serial.

Frequently asked questions about Sonoff POWR316 Tasmota

Which Tasmota version do I need for the Sonoff POWR316?

The POWR316 has a dual-core ESP32 (ESP32-D0WD-V3). You therefore need tasmota32, or a language build such as tasmota32-DE for a German interface. For flashing via cable, use the file ending in .factory.bin and write it to address 0x0.

Can I flash the Sonoff POWR316 without soldering?

In principle, yes, with wires pressed against the pads or with test probes. However, I advise against it. Loose contacts quickly lead to voltage drops and aborted flashing with the ESP32. A soldered-in pin header saves you a lot of time.

What does “Couldn’t sync to ESP” mean?

The flasher receives no response from the ESP32. Usually, the chip isn’t in flash mode, RX and TX are swapped, or the power supply is insufficient.

Why does my ESP32 keep restarting?

Repeated “POWERON_RESET” messages indicate that the supply voltage is dropping. The ESP32 briefly needs up to 500 mA, especially when switching on Wi-Fi. Most USB-TTL adapters can’t deliver that.

Can I damage the Sonoff while flashing?

Not through a failed flashing attempt. The bootloader in ROM remains intact, so you can always reflash. It gets dangerous if you apply 5 V to the data lines or connect the device to 230 V with the housing open.

Can I go back to the original firmware?

Only if you made a backup of the memory before flashing. Sonoff does not offer the eWeLink firmware for download.

Does the Tasmota Web Installer work with the POWR316?

In principle, yes. For me, esptool-js worked better because it shows a detailed log and verifies the written data via checksum. That way, you can see immediately whether the firmware is stored correctly.

Further reading

My approach in one sentence

For Sonoff POWR316 Tasmota, the rule is: use an ESP32 dev board as a programmer, keep the button pressed during the entire process, flash with esptool-js, and then start the Sonoff with its own power supply. That’s how it worked for me after a long afternoon. Now the POWR316 monitors my air conditioner’s power consumption, completely without the cloud.

Have you had different experiences, or did you get stuck at a point I didn’t cover here? Feel free to let me know in the comments.

Posted on

Building a DIY Doll Bed

Puppenbett selber bauen: Das fertige, metallfreie Holzbett aus Fichte und Multiplex mit detailliertem Lattenrost steht auf einem Holzboden.

Sometimes it’s the everyday coincidences that give the decisive push for a new building project. During a bike ride through the neighborhood, Elisabeth recently noticed a small doll wardrobe that had been left outside a garden gate to be given away. The good piece was made of solid spruce wood and was completely untreated. Continue reading “Building a DIY Doll Bed”

Posted on

Building Protection: Stopping Rising Damp in Old Buildings and Restoring Historic Floors

When you renovate an old building, you know this moment: You enter a room on the ground floor that hasn’t been properly heated or ventilated for years. It smells earthy, slightly musty, and the cold literally creeps up your legs from below. During my restoration project, an old farmhouse from 1888 in Transylvania, I faced exactly this fundamental problem. Continue reading “Building Protection: Stopping Rising Damp in Old Buildings and Restoring Historic Floors”

Posted on

Applying Lime Plaster to Mixed Masonry

Eine teils verputzte historische Wand aus Bruchstein und Feldbrandziegeln, die zeigt, wie man fachgerecht Kalkputz auf Mischmauerwerk auftragen kann, mit einer Maurerkelle im Vordergrund.

When I bought my old farmhouse from 1888 in Romania in 2016, I knew there would be a lot of work ahead of me. But only when I started knocking the layers of plaster off the walls did the true structural challenge of this project reveal itself. Beneath the crumbling plaster emerged a fascinating, yet building-physics-wise highly demanding structure: a solid mixed masonry consisting of hard, cool Saxon rubble stones and relatively soft, extremely absorbent field-fired bricks. Continue reading “Applying Lime Plaster to Mixed Masonry”

Posted on

Shaping and Restoring a Warped Historic Wooden Door

Nahaufnahme einer detailliert gefertigten, unfertigen Holztür, die auf Holzböcken in einer traditionellen Schreinerei ruht. Die Tür zeigt ein komplexes geometrisches Gitterwerk mit Bögen und leeren Glasfächern im oberen Teil sowie erhöhte Paneele im unteren Teil. Im Hintergrund Werkbänke mit Handwerkzeugen wie Hobeln, Meißeln und Hämmern. Tageslicht fällt durch Fenster

Anyone who has ever opened the massive, handcrafted softwood door of an old Saxon farmhouse knows the feeling. The heavy weight, the deep sound of the falling latch, and the inimitable patina of over a hundred years tell stories that no modern industrial door leaf could ever convey. When restoring historical buildings, I realize time and again: these doors are not just functional elements; they are the soul of the house. They were built by craftsmen who understood their trade from the ground up. Continue reading “Shaping and Restoring a Warped Historic Wooden Door”

Posted on

Supporting Ceiling Beams in an Old Building: How to Safely Prop Up a Wooden Beam Ceiling

Temporäres Abfangen alter Deckenbalken aus Eiche im Altbau mit lotrecht aufgestellten Stahl-Baustützen auf lastverteilenden Bohlen.

Supporting Ceiling Beams in an Old Building: How to Safely Prop Up a Wooden Beam Ceiling

Imagine standing in a room that has survived almost unchanged for well over a hundred years. The air smells of fine dust, dry wood, and the history of generations. That is exactly how I felt when I took over my house from 1888 and grasped the full dimensions of the ground floor for the first time. Continue reading “Supporting Ceiling Beams in an Old Building: How to Safely Prop Up a Wooden Beam Ceiling”