Section 01Three ways heat moves, and which one each system leans on
Every thermal experience is some blend of conduction, convection and radiation. Both systems use all three. They differ in the proportions, and the proportions are the whole story.
| Mode | What it is | Traditional | Infrared |
|---|---|---|---|
| Conduction | Contact transfer, solid to solid | Element to steel to stones. Bench to skin. | Minor. Panel housing to cabin wall. |
| Convection | Bulk movement of heated air | Dominant. Hot air rises, stratifies, circulates, and delivers most of the load to your skin. | Deliberately minimised. Air stays 30–45 °C / 54–81 °F cooler, which is the entire selling point. |
| Radiation | Electromagnetic emission, no medium required | Substantial. Stones at 250–300 °C / 482–572 °F, hot walls and the stove all radiate at you. | Dominant. Long-wave emission aimed directly at the occupant. |
| Latent heat (condensation) | Vapour condensing on a cooler surface, releasing stored energy | Unique to traditional. This is löyly. Nothing in an infrared cabin can reproduce it. | None available. |
The finding that reverses the usual framing
In a traditional sauna, radiation carries about 73% of the stove's total heat output. Convection carries the rest.
This is measured, not asserted. Fan, Holmberg and Heikkinen at VTT Finland (Building Research and Information, vol. 22 no. 6, 1994) modelled a 2 × 2 × 2 m sauna with a 5.5 kW stove using a thermal-network analysis and CFD. They calculated radiant emission from the stove at 4.034 kW and convective emission at 1.466 kW. An independent check from measured airflow across the stove, 4.1 dm³/s at a 245 °C / 473 °F temperature rise, gave a convective figure of about 1.2 kW, which they judged good agreement.
So the popular framing of "infrared uses radiant heat, traditional sauna uses air heat" is not merely incomplete. It is closer to backwards. A traditional sauna is predominantly a radiant environment that also gives you convection and, uniquely, latent heat from löyly. Infrared removes two of those three and increases the remaining one.
The same paper reports stone surface temperatures of 260–310 °C / 500–590 °F, with surfaces treated as grey bodies at emissivity 0.9.
Two design rules that follow directly from the 73% figure
If radiation dominates, then line of sight to the stones is what determines whether the room heats evenly. The authors state both consequences plainly:
- "If an even temperature distribution is required in the sauna, the hot stone surface on the stove must be 'seen' by each surface space." Anything that blocks sight of the stones creates a cold zone. This is a radiation-geometry problem, not an airflow problem, and it is why heater position governs the room.
- "If the stove is elevated above the floor, the temperature in the zone lower than the stove's upper surface can be very low." This is the physics behind the feet-above-the-stones rule. It is not folklore.
Their conclusion: "The configuration and position of the sauna stove becomes the dominant factor in controlling the sauna room temperature distribution." See the geometry targets →
Section 02The electron-to-skin chain
The physics of a traditional sauna has been modelled twice in the published literature, and the two studies complement each other neatly. Fan, Holmberg and Heikkinen (VTT Finland, 1994) solved the radiant heat exchange and the airflow, validated against laboratory measurement, without modelling water evaporation. Corentin Macqueron (2014) then modelled a wood-burning sauna in the NIST Fire Dynamics Simulator, adding combustion, soot and steam cloud transport, which is the part the 1994 work explicitly left out.
The two models, for anyone who wants to go to the source
- Fan, Holmberg & Heikkinen, 1994. 2 × 2 × 2 m sauna, 5.5 kW stove mounted in a corner 200 mm above the floor, stove body 250 × 512 × 550 mm. Fresh air at 20 °C / 68 °F, 48 kg/h, exhaust 1,890 mm up the opposite wall. Solved with the WISH code on a k-epsilon turbulence model, plus an analogue thermal-network solution for radiation. Validated against the Aikäs and Holmberg laboratory tests (VTT report 1431, 1992), which is the same VTT work that underlies modern Finnish ventilation guidance.
- Macqueron, 2014. A roughly 8 m³ wood-panelled insulated sauna with a small wood-burning stove, modelled in FDS on about 170,000 hexahedral cells at 4 cm resolution. Heat release rate was tuned to reproduce the 80 °C / 176 °F the room's thermometer actually reads in steady state, which required a prescribed thermal power of 8.75 kW. Outputs include temperature and velocity fields, heat flux, soot and steam cloud transport.
Macqueron notes that apart from the 1994 paper, "few detailed thermal studies have been performed on the matter." Two studies, thirty years apart, are the entire published CFD literature on saunas.
The chain from energy source to your skin runs through nearly all of heat-transfer physics in a span of a few metres.
Electric sauna: it starts with electrons
An electric field drives electrons through the heating elements. Their energy relaxes into the atomic lattice through electron-phonon interactions. Electrical energy becomes heat, which is Joule heating. Heating elements reach roughly 450 °C / 842 °F.
Wood-burning sauna: it starts with chemistry
Heat decomposes the wood. Cellulose, hemicellulose and lignin undergo pyrolysis, and combustion oxidises the products. Chemical energy becomes heat. Flame reaches roughly 1,000 °C / 1832 °F and the flue roughly 600 °C / 1112 °F.
From that point both systems speak the same language.
Conduction carries heat through steel and into the stones. Convection heats the air. Thermal radiation transfers energy directly between stove, stones, walls and occupants. Hot air becomes less dense and rises, and buoyancy builds the familiar stratification of roughly 100 °C / 212 °F above and 40 °C / 104 °F below. Heat has become motion.
Some of it escapes, through walls and ceiling, through ventilation and leakage, and with a wood stove through the chimney. Stove power is never entirely useful heat delivered to the bather. That single sentence is why heater sizing and ventilation design matter as much as they do, and why they get their own section in Design.
Why the stratification gradient decides your bench height
A sixty-degree difference between ceiling and floor exists inside a room roughly two metres tall. Where you place a bench decides which sauna you are actually in. Feet below the level of the stones sit in a fundamentally different room from a head near the ceiling. This is the physical reason Finnish design insists on 40 to 48 inches between the top bench and the ceiling and on getting your feet up. See bench geometry →
Section 03Löyly, precisely
Löyly is the Finnish word for the burst of steam and heat that follows water thrown on hot stones. It has no clean English translation, and it is not simply "more humid air."
Water hits stones at several hundred degrees Celsius. It heats, boils, and absorbs about 2.3 megajoules per kilogram as latent heat of vaporisation. Energy is now carried by water vapour rather than by hot air.
That vapour is swept through a buoyant, turbulent three-dimensional flow until it reaches cooler skin. Part of it condenses there, releasing that stored latent heat directly onto you, while convection and radiation continue transferring energy in parallel. The perceived heat spike is a genuine energy delivery event, not a sensation of humidity.
The full chain, as modelled
electricity or chemical energy → electrons or molecular bonds → heat → conduction → radiation → convection → buoyancy → stones → vaporisation → turbulent transport → condensation → skin
Only a few metres. Almost all of heat-transfer physics along the way. Validated computational fluid dynamics can show the full transient three-dimensional journey, including plumes, recirculation zones, gradients and the path of the löyly cloud, which experimental probes at a few fixed points never capture.
Three consequences that matter to a buyer
- Stone mass is the real specification. Löyly quality depends on having enough stone at high enough temperature to flash water instantly. A heater that warms the air quickly while leaving the stones cold produces a hot room with poor löyly. This is the failure mode of an oversized heater in an undersized room.
- Löyly is why humidity numbers mislead. A traditional sauna sits at 10–20% relative humidity most of the time and spikes hard and briefly during löyly. Quoting a single humidity figure describes neither state.
- An infrared cabin cannot do this at all. There is no stone mass and no surface you can safely throw water at. Cabins that add a small steam generator produce humidity without the latent-heat delivery event, which is a different thing.
Section 04The infrared spectrum, and why vendors disagree about it
Infrared is simply light too long in wavelength to see. Where the bands begin and end depends on which standard you use, and the sauna industry uses a third convention of its own. This is a large part of why full-spectrum marketing is so confusing.
| Convention | Near | Mid | Far |
|---|---|---|---|
| CIE (IR-A / B / C)Photobiology standard | 0.78–1.4 µm | 1.4–3 µm | 3–1000 µm |
| ISO 20473Optics standard | 0.78–3 µm | 3–50 µm | 50–1000 µm |
| Sauna industry usageNot a standard | 0.75–1.5 µm | 1.5–5.6 µm | 5.6–1000 µm |
The one number with real physics behind it
A body at 37 °C / 98.6 °F radiates most strongly at a wavelength given by Wien's displacement law: 2,898 divided by absolute temperature. At 310 K that is 9.35 µm. Human skin's peak infrared absorption sits in the same neighbourhood, commonly quoted as 9.4 µm. This is why carbon panels are engineered to concentrate output around 9 to 10 µm. The claim that far infrared is "matched to the body" is genuinely grounded, and it is a claim about efficient absorption at the skin rather than about depth.
Carbon against ceramic
| Property | Carbon panel | Ceramic rod |
|---|---|---|
| Output concentration | 6–12 µm, peaking near skin absorption | Wider spread, including 5–7 µm mid infrared |
| Surface temperature | Lower, larger emitting area | Higher, smaller emitting area |
| Perceived heat | Gentler and more even | More intense, with hot spots |
| Reported EMF | 0.5–3 mG | 5–15 mG |
| Best suited to | Long, comfortable whole-body sessions | Targeted intensity, faster surface heating |
On EMF specifically
"Low EMF" is one of the loudest claims in infrared marketing and one of the least examined. Two things are true at once. Carbon panels do measure lower than ceramic rods, so the comparative claim is fair. And there is no body of evidence showing harm from the extremely-low-frequency magnetic fields produced by resistive heating elements at these magnitudes, which are comparable to ordinary household appliances. If low EMF construction costs you nothing, take it. Paying a large premium for it is buying reassurance rather than a demonstrated health outcome. Note also that the shielding conversation belongs to infrared alone, since a traditional heater is a resistive coil or a fire and nobody markets EMF numbers for it.
Near infrared and red light are a separate therapy
Red and near-infrared light in the 630–850 nm range act through photobiomodulation, absorbed by cytochrome c oxidase in mitochondria. That is a photochemical mechanism, unrelated to heating. Bundling it into a sauna is convenient, and it introduces a real dosing problem: photobiomodulation studies use dedicated panels at specified irradiance and distance, whereas a few emitters mounted in a cabin wall deliver an uncontrolled dose that varies with how you happen to be sitting. Treat integrated red light as a bonus feature rather than as delivered therapy, and read the irradiance specification and the working distance before paying for it.
Section 05Penetration depth, settled
This is the single most distorted number in the category, and correcting it changes how you should think about the whole modality.
Far infrared penetrates up to 1.5 inches, reaching muscle, fat, joints and the lymphatic system directly.
Infrared across 2.5 to 50 µm penetrates the human body to roughly 200 to 300 micrometres. At 9 to 10 µm, where carbon panels put most of their energy, published figures cluster at 1 to 2 millimetres. 1.5 inches is 38 millimetres.
Why the physics forbids it
Tissue is roughly 70% water, and water absorbs strongly across the far-infrared band. Protein absorbs there too. An absorbing medium extinguishes radiation exponentially with depth, so the energy is deposited almost entirely in the outermost layer of skin. No far-infrared wavelength meaningfully passes the dermis. Shorter near-infrared wavelengths penetrate further, which is precisely why photobiomodulation research uses them, and near infrared is a minority of a far-infrared cabin's output.
What this actually means, and why it is not fatal to infrared
Far infrared heats your skin. Cutaneous blood flow then carries that heat into the core. That is the same final pathway a hot room uses. The mechanism is skin heating followed by convective transport in blood, in both systems.
This does not make infrared useless. It makes it a competent, comfortable, efficient way to raise skin and then core temperature, which is exactly what the heat-adaptation literature says produces the adaptation. What it removes is the story that infrared reaches tissue that hot air cannot. Depth was never the mechanism.
Section 06What your body does about it
The acute response is a thermoregulatory emergency handled gracefully. It looks a great deal like moderate cardiovascular exercise, which is the basis for the comparison to aerobic training.
| System | Response | Magnitude |
|---|---|---|
| Skin circulation | Peripheral vasodilation redirects cardiac output to the skin to dump heat | Cutaneous blood flow up ~70% |
| Heart rate | Rises to sustain output against reduced peripheral resistance | To ~100 bpm at moderate temperature; up to ~150 bpm in hotter rooms |
| Core temperature | Mild hyperthermia, the actual active ingredient | +1.0 to 1.2 °C in the Waon protocol; higher and faster in a hot room |
| Blood pressure | Falls during and after, through vasodilation | Systolic down 4–11 mmHg depending on protocol |
| Sweat and plasma | Evaporative cooling. Plasma volume falls acutely, then overshoots upward across days | ~0.5 kg lost in 30 min at 55 °C / 131 °F in healthy adults |
| Autonomic tone | Sympathetic activation during, parasympathetic rebound after | The source of the post-sauna calm |
| Endocrine | Beta-endorphin release, growth hormone response, catecholamine shifts | Protocol dependent |
This load is real, and it is why contraindications exist
Transient myocardial ischemia was recorded in 93% of stable coronary artery disease patients during sauna use. Sauna is a genuine cardiac stress test that most healthy people pass without noticing. Absolute contraindications include unstable angina, myocardial infarction within three to six months, and severe aortic stenosis. Full safety section →
Section 07The molecular layer, where the durable benefits probably live
Acute vasodilation explains a session. It does not explain a twenty-year mortality curve. The candidate mechanisms for durable adaptation are hormetic: a mild, repeated, survivable stress that upgrades the machinery.
HSP70, HSP90 and HSP110 families are molecular chaperones that prevent proteins from misfolding and aggregating. Whole-blood RNA sequencing thirty minutes after a heat session found 17 significantly upregulated genes, prominently including these families. Animal work found HSP27, HSP32 and manganese superoxide dismutase upregulated after four weeks of thermal therapy.
This is the leading proposed route to the neuroprotective findings, since protein aggregation is central to Alzheimer's and Parkinson's pathology.
Repeated heating increases expression of endothelial nitric oxide synthase, improving flow-mediated dilation. The cleanest demonstration is genetic: in mice lacking eNOS, repeated thermal therapy failed to increase blood flow or capillary density in an ischemic hindlimb. eNOS is the critical regulator of the angiogenesis this therapy induces.
In human trials, endothelium-dependent dilation improved while nitroglycerin-induced dilation did not, locating the effect in the endothelium rather than in smooth muscle.
Heat acutely raises IL-6. The size of that rise predicted the size of the antidepressant response over six weeks, and the effect tracked the ratio of IL-6 to its soluble receptor, which is the classical anti-inflammatory arm of IL-6 signalling rather than the pro-inflammatory arm.
A rare case of a proposed mechanism being tested as a moderator inside the same randomised trial.
Four weeks of thermal therapy reduced serum hydroperoxide and raised nitric oxide metabolites in heart failure patients. Heat exposure reduces TNF-alpha, CRP, prostaglandin E2 and leukotriene B4 while promoting interleukin-10. Local infrared treatment in spondyloarthritis significantly lowered TNF-alpha against control.
Consistent with the finding that frequent sauna appears to offset the mortality risk carried by elevated hsCRP.
Repeated heat conserves sodium and albumin, expanding plasma volume by roughly 5.6% across heat-acclimation protocols generally, and reported as high as 17.8% after four sauna exposures in trained cyclists. About two thirds of the expansion occurs within the first 24 hours.
Lower resting heart rate, greater stroke volume and better heat tolerance follow from it.
Sauna-like conditions lowered tau phosphorylation in wild-type mice, in tau-transgenic mice and in neuron-like cells. The effect correlated with phosphatase and kinase activity changes and, notably, not with the heat shock response.
The same paper produced the same effect using topical menthol to raise body temperature, which argues the active ingredient is mild hyperthermia rather than sauna as such.
Section 08Dose is thermal load, not minutes on a clock
The variable that drives adaptation is how much your core temperature rises and for how long it stays elevated. Room temperature, humidity, session length, air movement and your own acclimation state all feed into it.
The heat-acclimation literature quantifies this directly. A Bayesian meta-regression across 211 papers found that adding 15 minutes per exposure lowered end-exercise core temperature a further 0.04 °C and expanded plasma volume a further 0.4%, while each additional exposure added 1.9 g of haemoglobin mass. Raising ambient temperature 5 °C / 41 °F lowered end-exercise heart rate a further 2 bpm. Dose-response is measurable and gradual.
| Environment | Air temp | RH | Core temp trajectory |
|---|---|---|---|
| Finnish sauna, with löyly | 80–100 °C / 176–212 °F | 10–20%, spiking | Fast rise. 8–15 minutes per round is typical and self-limiting. |
| Finnish sauna, dry | 80–90 °C / 176–194 °F | 10–15% | Moderately fast. 15–20 minutes tolerable. |
| Bio sauna / sanarium | 50–60 °C / 122–140 °F | 40–55% | Slower, and humidity blocks evaporative cooling, so load is higher than the temperature suggests. |
| Steam room | 40–48 °C / 104–118.4 °F | ~100% | Evaporative cooling is nearly eliminated. Perceived intensity high. |
| Waon protocol | 60 °C / 140 °F | Dry | +1.0 to 1.2 °C in 15 minutes, then held by 30 minutes of blanket rest. The rest phase is part of the dose. |
| Consumer infrared cabin | 45–65 °C / 113–149 °F | Ambient | Slow. Reaching a core temperature of 101.3 °F took a measured average of 82 minutes, range 61 to 110. |
The 82-minute finding is the most practically useful number on this page
Twenty-five healthy adults sat in a commercially available infrared device with rectal core temperature monitored. Every one of them reached the 101.3 °F target used in the depression trials. It took an average of 82.12 minutes. If you use an infrared cabin for the standard 25 or 30 minutes, you are almost certainly not reaching the core temperature that the clinical literature studied. That is not a reason to avoid infrared. It is a reason to sit longer, and to know that the comfort advantage is partly purchased with time.
Section 09The experiment that suggests the mode may not matter
If thermal load is the active ingredient, then any route to the same load should produce the same adaptation. Somebody tested exactly that.
Four modes, one clamped core temperature
Thirteen physically active participants, five of them women, completed four separate five-day regimes of sixty minutes per day in randomised order, separated by at least four weeks. Rectal temperature was held at neutrality by 36.6 °C / 97.9 °F water immersion as the control, or raised by exactly 1.5 °C by one of three routes: 40 °C / 104 °F hot water immersion, sauna at 55 °C / 131 °F, or exercise in humid heat at 40 °C / 104 °F.
- Plasma volume expansion was similar in all heated conditions. Only exercise-in-heat exceeded the thermoneutral control, and only by 4%.
- Roughly two thirds of the six-day plasma volume expansion happened in the first 24 hours, regardless of mode.
- Resting core temperature fell 0.1 to 0.3 °C in every heated condition, with no difference between them.
- Systolic blood pressure fell 4 mmHg, again with no significant difference between modes.
- The adaptation was mediated by conservation of both sodium and albumin, with no evidence these played different roles in different modes.
The authors' conclusion: individuals can expect similar adaptation to heat regardless of the mode used, and the mode chosen "should depend on the individual and accessibility to equipment."
Two independent results point the same way. Whole-body hot packs produced the same brachial artery dilation as Waon therapy in a randomised crossover. And far infrared added to convective heat at the same temperature was not superior to convective heat alone for pain in older adults.
What follows from this, honestly stated
The evidence for heat is much stronger than the evidence for any particular machine. The large epidemiology happens to have been done on Finnish saunas, which is a strong reason to trust the Finnish sauna dose and a weaker reason to believe the Finnish sauna is uniquely capable.
Held together, this argues for choosing on adherence. Buy the heat you will use most often, at a session length you will tolerate, in a place you will actually walk to. Then read Choosing, where that principle turns into a decision.
Sources for this chapter: Fan, Holmberg & Heikkinen, Building Research and Information 1994 (radiant fraction, stone temperatures, line-of-sight and inlet-position findings); Macqueron, arXiv:1404.6774, 2014 (FDS modelling, steam cloud transport); ICNIRP statement on far infrared radiation exposure; Wien's displacement law; Kissling et al., Experimental Physiology 2021; Kominami et al., International Journal of Hyperthermia 2020; Mason et al., International Journal of Hyperthermia 2021; McDonald et al., Comprehensive Physiology 2025; Miyata et al., Circulation 2010; Guisle et al. 2021; Akonom et al. 2026. Full details in the Library.