Keeping Filament Dry — What a Dry Box Actually Does, and How to Choose One

2026年8月22日

Some days a print just will not come out right. The nozzle crackles, the surface comes out rough, thin strings are left hanging across the part. Same settings that worked last time.

The classic case is PLA (polylactic acid, the most widely used filament material) left open and soaking up humidity. Look through user reports and you find plenty of cases where drying it in a dry box (a storage case that keeps filament dry) made the same filament print unrecognisably better.

I own a few things for dealing with this and use them, but what exactly moisture does to the material was something I had never properly looked up. And how much do you actually have to spend to fix it? This article covers where the moisture comes from and compares three approaches. (Prices are Japanese retail as of April 2026; ¥1,000 is roughly $6.70.)

What I looked at, and what I use myself

Two kinds of thing are mixed together here. The mechanism, the absorption figures and the cost estimates are gathered from published information and user reports. The vacuum bags and the dryer are things I bought and use. Each hands-on part is marked as such, and a later section separates the two explicitly.

What moisture does to filament

Filament (the long thin strand of plastic a 3D printer prints with) is hygroscopic: it readily draws water out of the air. When it does, that water boils off in the heat of the nozzle (the tip that melts and extrudes the filament) during printing, and that is what wrecks print quality.

The symptoms

PLA has a reputation for not absorbing much, but leave it open for a few weeks and the effect is plenty visible. PETG and nylon are far more absorbent; nylon starts degrading after 24 hours in the open.

Why moisture breaks it: the mechanism, by material

So water is bad. But why? Understanding this lets you decide how much care each material needs in storage and drying.

PLA: hydrolysis of the ester bonds

PLA is polylactic acid, a polymer of lactic acid units linked in a chain. What links them is a chemical joint called an ester bond.

Hydrolysis is the reaction where a water molecule (H2O) wedges into that joint and cuts the chain. Old PLA filament turning brittle and snapping is those chains getting shorter through hydrolysis.

Hydrolysis in PLA is the slow-but-certain kind: leave it in a humid environment for a few weeks and print quality starts to suffer. Nozzle temperatures (around 200°C) accelerate hydrolysis further, so printing wet PLA brings the degradation out all at once.

Where the number comes from: PLA’s 0.5% moisture limit (by weight) is the recommended maximum moisture content in NatureWorks’ technical documentation, “Ingeo Biopolymer 4043D Technical Data Sheet." Above that, the drop in molecular weight during extrusion becomes pronounced and you get bubbles, stringing and rough surfaces.

Nylon (PA): amide bonds and hydrogen bonds, both problems

Nylon chains are linked by amide bonds. The amide bond itself resists hydrolysis better than PLA’s ester bond, but nylon has a different weakness.

Nylon’s polymer chains are held together by hydrogen bonds, a weak attraction. Water molecules interpose themselves in those hydrogen bonds, and once between the chains they cause the nylon to swell. That throws off dimensional accuracy and reduces strength.

Nylon also absorbs extremely quickly: at 60% relative humidity it takes up over 2% of its weight in water within 24 hours. The acceptable moisture content for printing is 0.2% or less, so you either print it immediately after opening or keep it in a dry box permanently.

Where the number comes from: the equilibrium moisture content of nylon (PA6) is about 2.5% at 23°C and 50% relative humidity (as documented in DuPont’s “Zytel Nylon Resin Molding Guide" among others). The 0.2% recommendation for 3D printing filament is from Polymaker’s and eSUN’s product specifications.

PETG, TPU and ABS: each with its own story

ABS’s higher 0.8% limit comes from being chemically hard to break. Its problem is not hydrolysis but the purely physical one of surface moisture boiling at nozzle temperature. That also means ABS recovers completely once dried.

How fast each material takes on moisture

Here is a visual comparison of how fast each material takes up water: absorption (weight %) after 24 hours at 60% relative humidity and 23°C.

Nylon’s rate is in a different class, taking on water five times as fast as PLA. This chart is why “nylon goes straight into the dry box on opening" is standard advice.

Relative humidity against equilibrium moisture

How much water filament ends up holding is determined by the relative humidity where it is stored. That endpoint is called the equilibrium moisture content.

How to read this: compare the “printing limit" column against the column for your own storage humidity. Nylon at 20% RH already equilibrates at 0.50%, over its 0.20% limit. Silica gel alone is not enough for nylon; heated drying is mandatory.

How vulnerable each material is to moisture

Here is how much moisture each material tolerates, what temperature and duration to dry it at, and what happens when it gets wet. Use it when setting dryer temperatures and deciding storage priorities.

How to read it: the moisture limit is a guide to “above this, print quality clearly suffers." The smaller the number, the more sensitive the material and the more attention storage and drying need. Nylon’s 0.2% means even a little moisture is too much.

Three ways to deal with moisture, compared on price and effort

There are broadly three ways to deal with moisture. Here is how they balance cost against effect.

1. Seal it in a bag (–)

The simplest thing is to put it in a bag and close it. A large zip freezer bag with the filament and some desiccant inside. Discount-store desiccant is fine.

One step up, you pull the air out as well: a vacuum bag. You seal the mouth, then draw the air out with a pump. Ten bags and a pump start at around $13.

Good:

• Cheap. Discount-store silica gel is fine
• You can start this minute

Not so good:

• It cannot dry filament that has already absorbed moisture
• It is storage only. All it does is prevent further uptake
• A torn bag or a poor seal makes it pointless

Verdict: a stopgap and a storage method. Effective for keeping new filament dry from the moment you open it, not enough for filament that is already wet.

The science of silica gel: how much can it actually take?

Silica gel is porous silicon dioxide (SiO2) that physically adsorbs water molecules in the countless fine pores on its surface. Picture a sponge catching water vapour in surface pores.

Its capacity is described by an adsorption isotherm: how many grams of water one gram of silica gel can take up at a given relative humidity.

A practical calculation: for a sealed container holding one 1kg spool (about 10 litres), the air inside is about 10 litres. Air at 25°C and 60% RH holds about 0.14g of water vapour in 10 litres. In theory about 0.5g of silica gel absorbs that. In practice the filament itself releases moisture and humidity enters when you open the lid, so 20–50g is a comfortable amount. With the 5g food-grade packets, that is four to ten of them.

When to regenerate: when indicating silica gel turns from blue to pink, it is saturated. Three to five minutes in a 500W microwave, or one to two hours in a 120°C oven, regenerates it. Sunlight is not hot enough to drive the water fully back out.

2. A filament dryer (–47)

A dedicated appliance that heats the filament to drive moisture off. You put the spool in, set a temperature and timer, and it dries over several hours.

Good:

• It brings already-wet filament back
• Adjustable temperature, so you can match the material
• Available from around $23

Not so good:

• On units without a feed port you cannot print while it dries (dry, remove, load into the printer). With a feed port, it feeds straight out while drying
• Leave it out afterwards and it takes moisture back on
• Cheap units can have uneven temperature

Thinking about drying efficiency

One way to compare “drying power" is drying rate per watt.

The energy side can be worked backwards from what it takes to evaporate one gram of water. The latent heat of vaporisation of water is about 2,260 J/g (0.628 Wh/g).

Worked example (a SUNLU S2):

• Power draw: about 48W
• Target: 1kg of PLA (0.5% moisture = 5g of water)
• Theoretical energy required: 5g x 0.628 Wh/g = 3.14 Wh
• Time to supply 3.14 Wh at 48W: about 4 minutes

Four minutes in theory, four hours in practice. The gap is the bottleneck of moisture migrating from inside the filament to the surface. The water is in the interior, so after the surface evaporates you have to wait for more to seep out. Raising the wattage barely shortens the drying time. Low and slow is the right answer.

Representative products (April 2026):

As an Amazon Associate we earn from qualifying purchases.

3. An all-in-one dry box (–67)

These dry the filament and feed it to the printer at the same time, so the spool stays in a low-humidity environment while you print.

Good:

• Drying and printing happen together, so the workflow is easy
• The dry environment is maintained throughout the print
• Set it once and leave it

Not so good:

• A little more expensive than a dryer alone
• Check filament diameter compatibility (1.75mm / 2.85mm) on some models
• They are bulky and take up space

Representative products: eSUN eBOX Lite, SUNLU S2 Plus, EIBOS Cyclopes

Dry boxes compared, product by product

How to read it: “dry while printing" is the biggest practical divide. Without it, every session means dry, remove, load. An all-in-one lets you print with the spool where it sits, which removes a lot of handling.

The food dehydrator has a wide temperature range but draws about 250W, five times a filament dryer’s 48W. That is a difference of a couple of dollars a month on the electricity bill.

What does not drying actually cost?

You might think you can manage without a dryer. But repeatedly failing prints with wet filament costs real money.

Monthly wasted filament

Here is an estimate of waste from carrying on with wet filament and no dryer. The failure rates are assumptions based on the range commonly reported by users.

How long a dryer takes to pay for itself

A note: that is calculated for PLA (about ¥2,500 per kg). With more expensive materials like PETG and nylon (¥3,000–5,000 per kg) the waste costs more and the payback is shorter. For a heavy user printing 20 or more times a month, any of these pays back within six months.

Storage methods compared: three ways to keep filament dry

How you store dried filament matters just as much. Dry it and store it badly and you are back where you started within days.

How to read it: aim for under 30% for PLA, under 20% for PETG, under 15% for nylon. Pick on the balance of cost and effort. A sealed container with silica gel is good value and suits most people. For storing nylon safely without vacuum bagging, an all-in-one dry box running continuously is the realistic option.

Keeping it dry once it is dry

Dried filament gets wet again if you leave it out. Pulling together experienced users’ explanations and manufacturers’ recommendations, these rules keep filament usable.

The basics:

1. Store unused filament in a sealed container with silica gel. A large sealed container with silica gel, spools standing upright, is the ideal. The silica gel can be regenerated in a microwave periodically.
2. Use it within 48 hours of opening, or move it to dry storage. PETG and nylon in particular start absorbing immediately. “I will use it later" is the most wasteful pattern there is.
3. Buy a hygrometer. A cheap digital hygrometer inside the container shows the state at a glance. Aim to keep storage below 20% RH.
4. Silica gel is reusable. Blue turning pink is the saturation signal. Three to five minutes at 500W in a microwave turns it blue again.

Storage kit (April 2026)

Here is the combination that comes up most often in reviews.

About ¥2,600–3,100 ($17–21) in total. Combine that with a dryer and the dry-then-store loop is complete.

Common filaments and prices (April 2026)

Which filament you buy in the first place matters too. Here are the staples.

What it costs to run for a year (April 2026)

Keeping 3D printing as a hobby costs more than filament. Knowing the annual figure makes budgeting easier.

Calculated at April 2026 prices. Excludes the printer itself and electricity.

How to read it: 2kg a month is roughly “printing something medium-sized once or twice a week." A light user might manage on 1kg; heavy use can pass 3kg. Nozzles are ¥200–300 in brass and around ¥500 in hardened steel. Unless you use carbon-fibre-filled filament, brass is fine.

$370–480 a year may sound like a lot, but weighed against being able to make the parts and cases you want, most people seem to find it good value.

What is hands-on here, and what this article cannot tell you

Two kinds of thing are mixed together in this article. Here is the split.

Things I use myself:

• Two kinds of vacuum bag (Vacbird, eSUN) and a USB electric pump. That the pump works on both brands of bag, and that the vacuum holds for months when the mouth is locked properly, are things I found here
• One filament dryer (Space Pi X4). Stringing dropped and prints settled down in PLA, PETG and TPU after I started drying — but that is judged by eye, not measured

Things I only looked up:

• Absorption rates and equilibrium moisture content are manufacturer figures and representative values from the literature. They change with room humidity and how you store it
• The “fewer failures" reports come from what users have written. They are not measured comparisons under matched conditions
• The cost-of-not-drying figures are calculations from published numbers
• Nylon I hardly print. Everything here about nylon comes from published sources
• Prices and ratings for the other dryers and dry boxes were checked in April 2026 and may have moved

Things I have not measured:

• How much water actually comes out. I could see the stringing drop, but I have not weighed the moisture itself. Weighing a spool before and after drying would give it in grams
• Whether a vacuum bag or a dry box holds humidity lower. That needs a hygrometer inside each and a log over days

I still lined the approaches up, because the reports of moisture causing failures do not contradict the mechanism at the material level. That much can be checked without printing a comparison.

In summary: a few tens of dollars changes your print quality

Moisture in filament quietly eats away at 3D print quality, and the less often you use the printer, the more it affects you.

Water cutting PLA’s ester bonds through hydrolysis; water molecules wedging into nylon’s hydrogen bonds and swelling it. The mechanism differs by material, but “water gets in and it breaks" is common to all of them. And because humidity is invisible, managing it by numbers matters: check the hygrometer reads under 20%, watch the silica gel for colour change.

An investment of $20–67 covers it. From the reviews and reports I compared, all-in-one dry boxes were the better-regarded option, largely because printing while drying means less handling.

Start by printing with what you have. Crackling or stringing is the sign. There are many reports of the same filament with the same settings coming out unrecognisably better once the storage environment was fixed.

Given that one spool costs $10–20, a dryer or a dry box pays for itself fairly quickly. Counted against the filament thrown away on failed prints, you could call it a saving.

Deciding what to buy

Products mentioned