Sleeping pad R-value, and the heat you lose to the ground

R-value describes how well a sleeping pad resists heat leaving your body and entering the ground. Since ASTM F3340 was adopted across the industry in 2020 it is a measured quantity from a guarded hot plate rather than a claim, so pads are finally comparable. The consequence most buyers get backwards: the first two or three R-values buy enormous amounts of warmth, and everything above about R-5 buys very little.

What ASTM F3340 changed, and why old numbers do not compare#

Before 2020, a pad's R-value was whatever the manufacturer's own method produced: some used a hot plate, some calculated from material properties. Two pads both labelled R-4 could differ by a full point, and a buyer had no way to tell.

F3340 fixed the method rather than the marketing. The pad sits between a heated plate and a cold plate, with guard heaters forcing all the measured heat straight down through the sample. Thermal resistance follows from the power needed to hold the temperature difference, the plate area and that difference. The test compresses the pad under a defined load, so the result describes a pad with a person on it rather than one on a bench.

Treat any R-value from before 2020 as an estimate, and be wary of pads sold today that still do not quote F3340. The test costs money, and a manufacturer that has paid for it says so.

The arithmetic: turning R-value into watts#

Steady conduction through a flat layer follows Fourier's law, which in SI terms is:

q = A x delta-T / R

Here q is heat flow in watts, A is contact area in square metres, delta-T is the temperature difference across the pad in kelvin, and R is thermal resistance in square metre kelvin per watt.

A US R-value becomes SI by multiplying by 0.176, so a pad labelled R-4 has an SI resistance of 0.704. You also need a contact area, because a person lying down does not touch the pad everywhere. We model contact at 0.45 square metres, and that assumption sits behind every number below.

Take a cool but not freezing night: the underside of your body near 30 degrees C, the ground at 10 degrees C, so delta-T is 20 kelvin. The numerator is fixed at 0.45 x 20 = 9, and every result below is 9 divided by the SI resistance.

Diminishing returns: what each extra R-value actually buys#

The last column is the whole story. The R-value between R-1 and R-2 is worth 25.6 watts; the one between R-7 and R-8 is worth 0.9 watts, about one twenty-eighth as much. That is not an artefact of these assumptions. Heat flow scales with 1 over R, so the curve is a hyperbola whatever area or temperature difference you choose. Change the inputs and every number moves, but the ratios do not.

For scale, a sleeping adult produces roughly 70 to 90 watts of metabolic heat in all directions combined. Losing 51 watts downward on an R-1 pad means the ground takes most of your output, which is why thin summer mats leave people cold in bags rated well below the actual temperature. At R-6 the ground is a minor term, and the next upgrade competes with drafts and damp fill rather than with conduction. Above about R-5, extra resistance is close to inert on a three season night and the grams are better spent elsewhere in the sleep system.

Warmth per gram across the three pad families#

Pads fall into three families with genuinely different failure modes. The R per 100 g column uses the midpoint of each stated range, with the arithmetic shown so you can substitute your own pad.

Insulated air pads win warmth per gram by more than a factor of two. An air pad is almost all air, and the manufacturer adds only the down, synthetic fill or reflective film needed to stop that air moving.

The bottom row explains itself once you know the mechanism. Still air conducts at about 0.026 watts per metre kelvin, better than any common solid. But a tall open chamber with a warm top and a cold bottom sets up a convection loop that carries heat across the gap far faster than conduction would. Every insulation material ever made exists to subdivide air into cells too small to circulate, and an empty air chamber skips that step.

Stacking, and the real case for foam under air#

Resistances in series add. A R-2 foam mat under a R-4.5 air pad gives R-6.5, dropping modelled ground loss from 11.4 watts to 7.9 watts. There is no efficiency penalty, because heat simply crosses two resistances in a row.

The honest version of the standard advice: a foam pad under an air pad is not the lightest way to add R-value. Compare two upgrades from a R-4 insulated air pad.

• Add a R-2 foam mat. Result R-6, loss falls from 12.8 W to 8.5 W: 4.3 W for about 375 g, or 1.15 watts per 100 g.
• Replace it with a R-7 insulated air pad. Loss falls from 12.8 W to 7.3 W: 5.5 W for roughly 200 g, or 2.75 watts per 100 g.

Per gram the warmer air pad wins comfortably. The foam mat wins on the other three axes. It is by a wide margin the least expensive R-value available, it outlives several air pads, and it is the only redundancy in a system that otherwise has a single point of catastrophic failure.

If you carry an insulated air pad where a failure would be dangerous, the foam mat is not a warmth accessory. It is the spare tyre.

Self-inflating pads carry their own redundancy, which is their quiet advantage. Puncture one and the open cell foam inside still holds its shape under body weight and still insulates, at a reduced but usable level. You lose comfort rather than safety margin.

What the ground underneath you is doing#

R-value describes the pad. The other half of the problem is what the pad sits on, and the differences are larger than most people expect.

Snow is an insulator in its own right. Published values put its thermal conductivity between roughly 0.05 and 0.25 watts per metre kelvin depending on density, new snow at the bottom of that range and dense wind slab at the top. That is worse than still air at 0.026 but the same order of magnitude. Snow also has a floor temperature, since a melting surface cannot go below 0 degrees C, so delta-T across your pad is often smaller on snow than on bare frozen ground the same night.

Saturated soil is the opposite case. Water conducts at about 0.6 watts per metre kelvin, and wet ground is commonly quoted between 1.5 and 2.5. A puddle under the tent is a heat sink of effectively unlimited capacity. Site choice does more here than R-value can, which is the subject of pitching and site selection.

The season guide pairs with bag ratings rather than replacing them. ISO 23537 tests a bag on a specified mattress, so a comfort rating silently assumes a pad, as sleeping bag temperature ratings explains. Each full R-value you are short feels like roughly 2 degrees C of extra cold.

Check this yourself, two ways#

Neither test needs anything you cannot buy in a hardware shop, and both answer the question that matters: is your pad the weak link?

The half-on, half-off night. Lay a foam mat so it covers your torso and hips but stops at mid-thigh, with your usual pad over the whole length. Sleep a normal night on ground below about 10 degrees C. The question in the morning is whether your legs and your torso felt different. A clear difference at the foam boundary says your base pad is under-specified. No difference says the pad is adequate and your cold nights come from drafts, damp fill or food, all covered in why you sleep cold.

The infrared thermometer reading. Read the surface temperature of bare ground next to your pitch, lie on the pad twenty minutes, get up, and immediately read the ground where your torso was. Ground under a good pad has warmed very little, because little heat reached it. Ground under a poor pad warms noticeably, and every degree it gained came out of you. Infrared thermometers read surfaces, so keep material and distance identical between readings.

Common mistakes#

Buying by season name rather than ground condition. "Three season" covers a frosty October valley floor and a mild May meadow. Ground temperature sets your heat loss, and it lags air temperature by days, which is why late autumn feels colder than spring on identical forecasts.

Inflating by mouth. Every breath puts warm, saturated air inside the pad. That moisture can freeze inside the chambers in winter, and it feeds mould in a pad you cannot clean. Use a pump sack, and store the pad open and unrolled, as covered in gear storage.

Over-inflating. A rock hard pad pushes your hips and shoulders into a thin compressed layer, which makes contact worse rather than better. Inflate fully, then bleed air until your hip just fails to reach the ground.

Treating a quilt as pad-independent. A quilt has no insulation underneath by design, so the pad carries the entire downward job. Quilt users should run one full R-value above the season guide, a point developed in quilts against sleeping bags.

Ignoring width. A pad narrower than your shoulders leaves an arm on cold ground half the night, and no R-value fixes that.

Frequently asked questions#

What R-value do I need for winter camping?#

For consistently frozen ground, aim for R-5 or better. For snow, aim for R-7 or better, which most people reach by stacking a foam mat under an insulated air pad rather than buying one very thick pad. Below about minus 15 degrees C, treat the figure as a minimum and add margin, because ground contact is the one loss you cannot vent or adjust during the night.

Can you add the R-values of two sleeping pads together?#

Yes. Thermal resistances in series add directly, so a R-2 foam mat under a R-4 air pad gives R-6. There is no efficiency penalty, because heat has to cross both layers in sequence. The practical caveats are that the stack must stay together under you all night, and that the top pad dominates how the combination feels.

Why do older sleeping pad R-values not match new ones?#

Before ASTM F3340 was adopted in 2020, each manufacturer used its own method and some figures were calculated rather than measured. Two pads labelled R-4 under that regime could differ by a full point. Current numbers come from a guarded hot plate test with defined compression and temperature conditions, so they are comparable across brands. Older catalogue figures are estimates.

Does body weight change a pad's R-value?#

Somewhat. The test compresses the sample under a defined load, so the published figure already assumes a person is on it. A heavier sleeper compresses foam and displaces air from under pressure points more than a lighter one. The effect is largest on thick air pads and smallest on closed cell foam, which is nearly incompressible.

Is a foam pad worth carrying if I already have an insulated air pad?#

For warmth per gram, no: a warmer air pad is the better deal. For everything else, yes. Foam adds about R-2 for roughly 375 g, costs little, never punctures, and is the only backup if the air pad fails in the cold. That is the difference between an unpleasant night and a dangerous one.

Does an uninsulated air pad keep you warm at all?#

Barely. Air insulates well when held still, but a tall open chamber lets convection circulate warm air from the top surface down to the cold bottom and back. That loop moves heat across the gap far faster than conduction would. Uninsulated air pads are comfort devices for warm nights, and they are the usual explanation when someone is cold in an otherwise adequate bag.

How does sleeping on snow compare with sleeping on frozen ground?#

Snow is a moderate insulator, conducting at roughly 0.05 to 0.25 watts per metre kelvin depending on density, and a melting surface cannot fall below 0 degrees C. Bare frozen ground conducts far better and can be much colder. A levelled, compacted snow platform is often a warmer pitch than frozen earth on the same night, though both call for R-5 or more.