Why an Oxygen Absorber Is Sized to the Void, Not the Bag

This page contains affiliate links. If you buy through one we may earn a commission, at no cost to you.

Larvae in a seed tub almost always means the storage was airtight and the oxygen was not removed, and the reason the oxygen was not removed is usually an absorber matched to the wrong number. There is one identity behind this and it is worth carrying around: the capacity an absorber needs is 0.2095 multiplied by the void fraction of the packed food multiplied by the container volume. Oxygen is 20.95 percent of air by volume, the air sits only in the spaces between the pieces, and the void fraction is the term nobody publishes and nobody estimates.

Which turns a shopping question into an arithmetic one with a surprise in it. Put the same 300 cubic centimetre absorber in the same published 3.6 litre jar and it leaves 0.81 percent oxygen under rice and 8.67 percent under penne. A factor of 10.7, and the only thing that changed is the shape of the food. Insects do not notice the second case at all.

The rating is a volume of oxygen, not a volume of bag

Start with the pouch, because it is where the two published numbers are. A Wallaby bundle of 100 oxygen absorbers rated at 300cc with ten MRE Mylar bags publishes bags of 6.5 by 8 inches at 7.5 mil, absorbers that ship in heat sealed packs of ten and are never repackaged, and a stated fit for powders, spices, grains, beans, pasta, cereal and freeze dried foods.

A flat pouch of width W fills out to roughly a cylinder of circumference twice W, so the radius is W divided by pi, which is 2.069 inches, and the cross section is 13.45 square inches. Allow an inch and a half at the top for the heat seal above the zipper and the usable height is 6.5 inches: 87.42 cubic inches, or 1,432 cubic centimetres.

Now take 20.95 percent of that, which is the oxygen in the bag when it is completely empty. 300.1 cubic centimetres. Which is the rating of the absorber that ships with it, to one part in three thousand.

That is not luck and it is not generosity. The absorber is sized to an empty bag of its own size, which means the void fraction of whatever you put in it is the entire safety margin. At that size the margin is enormous and the product never fails, which is exactly why nobody learns the rule before they need it. Full record in our review of the Wallaby 300cc bundle, whose own drawbacks name the limit in five words: "300cc may be light for very large containers."

The seal allowance is ours rather than published, and it is the one soft input in that paragraph, so here is what moving it does. At one inch the empty bag carries 323.2 cubic centimetres of oxygen and the absorber is 7 percent short. At two inches it carries 277.1 and the absorber is 8 percent over. The rating sits inside the allowance either way, which is the honest version of the result.

The void fractions, which are the whole of it

A void fraction is one minus the bulk density over the true density, and three representative pairs cover most of what goes into a bag like this:

White rice, 850 against 1,450 kilograms per cubic metre, gives 0.414. Penne pasta, 450 against 1,400, gives 0.679, because a tube is mostly a hole. A parrot seed mix heavy in unhulled sunflower, 400 against 1,100, gives 0.636.

Run the identity on the 6.5 by 8 inch pouch and the oxygen comes out at 124 cubic centimetres under rice, 191 under the seed mix and 204 under penne. One 300cc absorber covers all three with between 47 and 142 percent to spare. At pouch size the question genuinely does not arise.

Where it breaks, on a volume somebody published

Step up one container size and the margin is gone. A DWELLZA KITCHEN set of four flour and sugar canisters publishes a uniform 3.6 litre capacity, which is 3.8 quarts, with side locking lids on a black silicone gasket, food grade BPA free bodies that nest when empty, and reusable labels with a marker so the contents can be switched. Its own drawbacks are candid about the consequence of one size for everything: "Tiny ingredients still fill a large canister."

So put the same 300cc absorber in that jar.

Rice. Void 0.414 on 3,600 cubic centimetres is 1,490 of air, carrying 312 of oxygen. The absorber is 3.9 percent short. The residue is 12 cubic centimetres in 1,490, which is 0.81 percent oxygen.

Penne. Void 0.679 is 2,443 of air, carrying 512 of oxygen. The absorber is 41.4 percent short. The residue is 212 cubic centimetres in 2,443, which is 8.67 percent.

The seed mix. Void 0.636 is 2,291 of air, carrying 480 of oxygen. Residue 180 in 2,291, which is 7.85 percent.

A factor of 10.7 between the first and the second, from one absorber in one jar. And the threshold matters here in a way it did not in the pouch. Controlled atmosphere work on stored product insects puts complete mortality at oxygen below roughly one percent held for one to three weeks at room temperature, with two percent slow and five percent barely an inconvenience. That band is a representative figure from the entomological literature rather than anything either listing publishes, and it is the single input this page leans on hardest. On it, the rice jar works slowly and the seed jar does nothing.

The fix is arithmetic rather than a different product. Two 300cc absorbers carry 600 cubic centimetres against the seed mix's 480 and the penne's 512, and the bundle above ships a hundred of them.

Inverting it gives the number you actually want

Divide the rating by 0.2095 times the void fraction and you get the largest container one absorber can clear:

A 300cc unit covers 3,461 cubic centimetres of rice, 2,250 of seed mix and 2,110 of penne.

A 400cc unit covers 4,615, 3,000 and 2,814.

Two things fall out of those six numbers. The step from 300 to 400 buys exactly 1.333 times the container, which is just the ratio of the ratings. The step from rice to penne costs 1.640. The food moves the answer further than the product does, which is the reason a rating on its own cannot be matched to a jar, and the reason both of the obvious upgrades, a bigger absorber or a smaller jar, are weaker levers than simply noticing what you are storing.

The 400cc end of that comes from a Wallaby 30 count set of MRE Mylar bags supplied with 400cc oxygen absorbers, which publishes 15 mil total thickness with an internal aluminium layer, a width of about 6.5 inches, BPA free food grade Mylar that is heat sealable above the zipper, and tolerance of boiling water so the pouch doubles as the cooking vessel. Its own drawbacks end on the sentence this entire page is an expansion of: "Absorber capacity must match bag contents." That is exactly right, and it is the only listing in this catalogue that says it. More in our review of the Wallaby 400cc Mylar set.

Three gaps in our own records, one of which cannot be settled

Worth stating plainly, because they are the numbers a buyer would need.

The 400cc listing publishes a bag count and an absorber rating and no absorber count. Thirty bags, 400cc each, and no statement of how many absorbers are in the box. On a product sold as the two halves of one job, that is the figure that decides whether the set is complete.

It also publishes a width and no height, about 6.5 inches, while the 300cc bundle publishes 6.5 by 8. So the bag that ships with the larger absorber is the one whose size is less specified, and the interior volume calculation above can only be done on the smaller one.

And the film thickness cannot be compared at all. One record says 15 mil total and the other says 7.5 mil, and 15 is exactly twice 7.5. A pouch has two walls. So these may be a single film described by two conventions, a two wall sum against a one wall figure, or they may be genuinely different films. Neither record says which, so we are not arbitrating it, and it is the one measurement on these two listings that a buyer cannot use. The same two conventions problem, worked properly in a different category, is the subject of our page on why two 5 micron filters are not the same filter.

A smaller one in the right direction. A Tafura 10 pack of 32 ounce leakproof freezer containers describes 32 ounces as "one quart, or one liter", and those differ by 5.7 percent. On this arithmetic 5.7 percent of the absorber's duty is well inside every other uncertainty on the page, so it is a rounding rather than a contradiction. It is worth naming only because the same record publishes the figure that matters more and gets it right: 32 ounces is 946 cubic centimetres, which under the seed mix carries 126 of oxygen, so one 300cc absorber covers that tub two and a half times over. Our review of the Tafura containers has the rest, including the gap that the listing does not state exact dimensions.

The other route, and why it stops somewhere else

A vacuum is the obvious alternative and it does not land in the same place. A Gyreuni 95kPa vacuum sealer machine publishes 95 kilopascals of suction across a 12 in 1 function set. Our page on what kPa means on a vacuum sealer establishes the conversion that makes the figure readable, that a suction number is a fraction of one fixed atmosphere of about 101 kilopascals, so this page takes it rather than redoing it.

Ninety five out of 101.3 leaves 6.3 kilopascals absolute, which is 6.22 percent of the gas that was there. The composition does not change, so the oxygen partial pressure is 0.2095 times 6.3, which is 1.320 kilopascals. Expressed the way the kill band is expressed, as a sea level equivalent, that is 1.30 percent.

So a 95 kilopascal vacuum arrives at the top of the band rather than under it, and a correctly sized absorber goes below it. More importantly the two residues are not the same kind of thing. A pressure deficit is restored by any leak, and the atmosphere is pressing on the bag the whole time looking for one. A chemical sink keeps absorbing until it is spent, so a small leak is consumed rather than admitted. That is why the two are not interchangeable even where the arithmetic puts them close, and it is also why the sealer is the better tool for a week in a freezer and the absorber the better tool for a year on a shelf.

Our guide to choosing oxygen absorbers for food storage gets the direction right and stops short of the line: "Capacity is rated in cc. A 300cc unit suits meal-size pouches, while larger buckets need higher totals." That is correct. The line is at 2,110 to 3,461 cubic centimetres for a 300cc unit, and which end you land on is decided by the food. Our page on storing long term food in Mylar bags safely says the same thing from the other side, "A 400cc absorber suits many meal-size pouches, while larger buckets need more", and adds the sentence that is the best single check on all of this: "About four-fifths of the air volume simply remains inside the pouch." That is 79.05 percent, and it is right, and it is the identity at the top of this page read backwards.

The honest limit, and what it leaves undone

What is exact here is the identity and every ratio built from it. Required capacity is 0.2095 times the void fraction times the volume, so the 1.333 between the two ratings and the 1.640 between rice and penne are exact whatever the densities are, because they are ratios of the same expression. The 1,432 cubic centimetre pouch interior follows from two published dimensions and one stated assumption. The 3.6 litre jar and the 300 and 400 cubic centimetre ratings are published. The 6.22 percent and the 1.320 kilopascals follow from one published suction figure and one constant.

What is representative rather than published is the rest, and the list is short but load bearing. The bulk and true densities of rice, penne and seed mix are typical values, not measurements, and the void fraction is as good as they are. The inch and a half seal allowance is ours. The filled pouch is treated as a cylinder, which overstates a pouch that is still slack and understates one that is packed tight. And the one to two percent oxygen mortality band is from the entomological literature rather than from any listing, which makes it the real limit on the central result: move that threshold to three percent and the rice jar succeeds comfortably while the seed jar still fails, and move it to five and both the single absorber cases that matter here are rescued. The direction of the finding survives any plausible threshold, because the 10.7 between the two residues does not depend on where the line is drawn. Which case falls on which side of it does.

Two things this page deliberately does not compute. It does not give a shelf life in years for anything, because that needs the film's oxygen transmission rate, and neither Mylar listing publishes one, so the oxygen that gets back in over a year is simply unknown. And it does not compute a time to mortality at any oxygen level, because that needs a named species and a temperature, and the vault question names neither. The absorber's own activation is a third unstated input worth flagging in the same breath: iron based absorbers need moisture to start working, and nothing on either record says how much, so a very dry freeze dried product may reach its rating slowly or not at all. The related mould argument, which runs on water activity rather than on oxygen, is worked properly in our page on why an airtight box does not stop bread going stale, and the trick of turning a solid food into a bed volume through a packing fraction is borrowed from why 5 kilograms of meat needs a 12 litre pressure cooker, which does the same algebra pointed at liquid instead of air.

Three numbers to carry. The rating is a volume of oxygen, so multiply 0.2095 by the void fraction by the container volume and compare. A 300cc absorber is sized to an empty bag of 6.5 by 8 inches and runs out somewhere between two and three and a half litres of packed food depending entirely on what the food is. And if the honest answer to which food it is happens to be unhulled sunflower seed, which is two thirds air, the jar that looked full is mostly the problem, and the correct response is to use two absorbers instead of one.