Quilts against sleeping bags, argued from the physics

Insulation only works when it is lofted. Under your body weight the underside of a sleeping bag is crushed to a few millimetres, does almost no thermal work, and is carried but not used. A quilt deletes that panel and the hood, saves roughly 150 to 350 g at the same warmth, and hands you a different problem: draughts.

What the underside of a sleeping bag is actually doing#

Thermal resistance comes from trapped still air, not from fibres. Still air conducts at 0.026 watts per metre per kelvin, and fill exists only to hold it still. Loft is the variable: lofted fill gives roughly 0.20 to 0.28 RSI (square metre kelvin per watt) per cm.

Now put a sleeping adult on top. A 5 cm loft under the hips collapses to two or three millimetres, and the fibres press into contact so the path is no longer through air. Apply the thickness rule to 3 mm and you get an RSI near 0.07, against about 1.1 lofted, and fibre contact makes the real figure worse.

ISO 23537 specifies the mattress under the manikin because a bag cannot insulate downward alone, so every bag rating already assumes a pad, as our sleeping pad R-value guide sets out. Manufacturers half-conceded this long ago, using 60/40 or 70/30 fill splits. A quilt takes it to 100/0.

The saving, first in grams and then in degrees#

Here is the component accounting for a three-season down bag rated near 0 degrees C comfort, carrying 600 g of 650 fill power down.

Take the middle of every range: 160 plus 70 plus 65 plus 55 removed is 350 g, against 35 plus 52 added, or 87 g. Net saving 260 g.

Now convert the fill part into degrees. Published fill weights give two anchors: a bag rated near 0 degrees C comfort carries 500 to 700 g of 650 fill power down, and a winter bag near minus 10 degrees C carries 800 to 1,100 g. Midpoints of 600 g and 950 g give 350 g for 10 degrees C, or about 35 g of 650 fill power down per degree C.

That number does the work. The 160 g of underside fill would be worth 160 divided by 35, so 4 to 5 degrees C, if it sat on top of you. Where it actually sits, our model gives it under half a degree. Deleting it costs almost nothing. The caveat: 35 g per degree is a linear model fitted to two published ranges, and required insulation is only roughly linear in the temperature gap, as ISO 11079 treats it.

Draughts, the hood and the head are the whole of the trade#

A quilt is an open system. Warm air is held in by three things only: pad straps, quilt width, and a neck closure that seals.

Model the enclosed air as 50 litres, or 0.05 cubic metres, which at 1.2 kg per cubic metre is 0.06 kg. Air holds about 1,005 joules per kilogram per kelvin, so warming it 30 degrees C above ambient costs 0.06 times 1,005 times 30, roughly 1,800 joules. A sleeping adult produces 70 to 80 watts, so one full flush costs 25 seconds of resting output. One flush is nothing. The rate is the problem:

• Flushing once every ten minutes costs 1,800 divided by 600, which is 3 watts. Undetectable.
• Flushing once a minute, which is what a quilt that lifts on every movement does, costs 30 watts, near 40 percent of resting output.

To put that into degrees, take a system losing 80 watts across a 33 degree gap between skin and ambient. That is a conductance of 80 divided by 33, or 2.4 watts per degree C, so 30 watts is roughly 12 degrees C of lost rating. This is a model, not a measurement, but it explains why a badly fitted quilt is not merely a little cold.

The head is the second leak. The claim that a fixed 40 to 45 percent of heat escapes through the head does not survive scrutiny: loss is roughly proportional to surface area, and the head is under 10 percent of it. That holds when the head is dressed like everything else, which in a quilt it is not. A bare head and face, 0.10 square metres at a surface conductance near 7 watts per square metre per kelvin, loses 23 watts across 33 degrees. The rest of the body, 1.7 square metres at a lofted conductance near 1 watt per square metre per kelvin, loses 56 watts. The head is 6 percent of the area and 29 percent of the loss. A hat with an RSI of 0.15 drops its conductance to 3.4 and its loss to 11 watts, and 12 watts divided by 2.4 is about 5 degrees C.

Treat that as an upper bound, since a real head is partly shielded by pillow, pad and quilt edge. If you already carry a hooded jacket, the weight saving stands. If you buy a 250 to 400 g jacket to make the quilt work, it is gone: our insulated jacket guide covers that cost.

Sizing a quilt: a width rule you can measure at home#

Flat widths run from about 125 cm to 145 cm at the shoulder, and that range decides whether the quilt seals. The fabric must run from under one pad edge, over the top half of your torso, and back under the other:

Minimum flat width = pad width + half your chest circumference + 20 cm of tuck.

Measure the chest over the clothing you sleep in. On a 51 cm (20 inch) pad, a 100 cm chest needs 51 plus 50 plus 20, which is 121 cm. A 130 cm chest needs 136 cm, or 149 cm on a wide 64 cm (25 inch) pad. The rule errs generous because it partly double counts the run from torso to pad edge, which is deliberate: a taut quilt lifts off the pad every time you move.

Width costs 25 to 80 g, so a sleeper who needs 145 cm has already returned a third of the saving.

Why a quilt rating is not a bag rating#

ISO 23537 puts a heated manikin inside a closed bag on a specified mattress. A quilt has no defined enclosure, so its performance depends on the pad, the strap settings and how still the sleeper is. The method does not transfer, and testing costs money. Quilt temperature numbers are, with very few exceptions, chosen by the seller.

That does not make them dishonest, it makes them incomparable. Most describe a temperature at which a competent user gets through the night, which is closer to the ISO limit output than to comfort. Add 5 to 8 degrees C to a manufacturer quilt figure before setting it against a bag comfort rating, as set out in sleeping bag temperature ratings, decoded and applied by the sleep system calculator.

One specification does compare directly: fill weight at a stated fill power. A quilt with 400 g of 850 fill power down and a bag with 420 g of the same carry nearly identical insulation above you. A quilt marked minus 7 degrees C is, adjusted, a minus 2 to 0 degrees C item for most sleepers.

Who should not buy a quilt, and who should#

Buy a bag instead if:

• You rotate in your sleep. Not two roll-overs, but starting on your back and finishing on your front. Every rotation is an air exchange, priced above at up to 12 degrees C of lost rating.
• Most of your nights fall below minus 10 degrees C. The seal must hold for eight hours, and the penalty for imperfection scales with the temperature gap.
• You already sleep cold. If you need 3 degrees of extra margin before anything else, do not also take on draught management. Start with why you sleep cold.
• You want one sleep system for everything. The sequence in how to choose a sleeping bag assumes that case, and a bag is the forgiving answer.
• You will not use pad straps. Unattached quilts work above 10 degrees C and nowhere else.

A quilt suits you if you sleep on your back or stay curled on one side, camp mainly above minus 5 degrees C, or already carry a hooded jacket. The strongest case is the hammock: its fabric presses against your whole underside, compressing insulation as body weight does, and worse, since it wraps. The fix is an underquilt hung outside the hammock where nothing compresses it, at which point a top quilt is the obvious partner and a bag's back panel and hood are dead weight, as tent, tarp or hammock sets out.

Check this before you buy, on the floor at home#

1. Measure the flat width. Run a tape across the quilt at shoulder and hip and compare with the rule above, since published widths are often taken at the widest point.
2. Do the rotation test. Pad on the floor, straps fitted, get in wearing what you sleep in, and roll back to side to other side five times without holding the edges. Do it in an unheated room. An edge that lifts clear of the pad is your draught.
3. Feel for the leak. Lie still a minute, then run a hand along the join between quilt edge and pad, and along the neck closure. Moving air on the back of your hand is the 30 watt problem.
4. Weigh the whole system. Quilt, straps, and the hat and jacket you now need, against the bag being replaced. Under 100 g of difference is not worth it.

Common mistakes with quilts#

Comparing a quilt number with a bag comfort rating. The most expensive error in the category, and the source of most complaints that quilts are colder.

Buying the narrowest width to save weight. The 60 g saved by dropping from 140 cm to 127 cm is worth under 2 degrees C of fill at 35 g per degree. The draughts it opens cost more.

Treating the pad as an afterthought. With a quilt the pad is the entire downward insulation, with no crushed fill acting as a backup, so a pad one season short is felt at once. Strap geometry matters too: straps set for a 6 cm pad will not hold a 10 cm one.

Cinching the neck closure against bare skin. It seals well but chafes, so people loosen it at two in the morning and get cold. A thin buff solves it.

Frequently asked questions#

Are quilts as warm as sleeping bags?#

At the same fill weight and fill power above you, yes, provided the edges seal against the pad and you wear a hat. The insulation doing useful work in a bag is the fill above and beside you, which is exactly what a quilt keeps. The difference is how much of that survives a night of movement.

How much weight does a quilt actually save?#

Roughly 150 to 350 g against a comparable bag, mid-range near 260 g: underside fill, hood, back panel fabric and full length zip removed, less straps and extra width added. If a comparison shows more, the quilt is probably also narrower or thinner.

Do you need a hat with a quilt?#

Yes, below about 10 degrees C. Once the body is under lofted insulation, an uncovered head can carry a quarter to a third of total heat loss despite being under a tenth of your surface area. A wool hat closes most of the gap.

What width quilt should I buy?#

Take your pad width, add half your chest circumference over sleep clothing, then add 20 cm of tuck. A 100 cm chest on a 51 cm pad needs about 120 cm of flat width, a 115 cm chest about 130 cm, a 130 cm chest about 140 cm. Round up, since a taut quilt lifts.

Can you use a quilt in winter?#

You can, but it stops being the easy choice below about minus 10 degrees C, because the penalty for an imperfect seal grows with the temperature gap. Winter quilt users run a wide quilt, a wide pad, a full strap set, a hooded jacket and a balaclava. That is a system, not one purchase.

Do quilts work for side sleepers?#

They work for side sleepers who settle on one side and stay there. They work badly for people who rotate through back, side and front, because each rotation drags the edges off the pad and flushes the warm air out. If you rotate, a bag is more reliable.