Filament Storage That Actually Works: Humidity Thresholds, Dry Box Builds, and When to Re-Dry by Material
Filament storage is the quietest print-quality problem in a shop. A spool that printed clean in July can string, pop, and delaminate by October on the same printer, profile, and slicer, because the plastic has been drinking the air the whole time.
Every polymer we run is hygroscopic to some degree. Water molecules work their way between the chains, and when the filament hits a 240°C hot end, that water flashes to steam inside the melt pool. You hear it as popping. You see it as fuzz, voids, and layers that peel apart in your hand.
This guide is the storage playbook we use on the production floor: the humidity numbers that matter, the dry boxes worth building, the desiccant that works, and the re-dry schedule per material. No magic, just moisture physics and a hygrometer you can trust.
Why Filament Storage Matters (and What Moisture Actually Does)
Hygroscopic is a polite word for thirsty. PLA, PETG, ABS, TPU, nylon, PVA, and PC all pull water out of ambient air, just at different rates. Nylon is the extreme case; a dry spool left out overnight in a 70% RH garage can be noticeably degraded by morning.
Here is what moisture does inside the hot end. Water boils at 100°C. Your nozzle runs 210°C to 290°C depending on material. Any water trapped in the filament vaporizes the instant it enters the melt zone, blows out through the molten plastic, and leaves a void. That void is what you hear as a pop and see as a bubble on the surface.
The second failure mode is chemistry. PLA is a polyester, and water hydrolyzes polyesters. Prolonged exposure cleaves the polymer chains; the plastic literally gets shorter and weaker over months in a humid shop. See the NatureWorks Ingeo PLA technical data sheets for the degradation curves. A wet PLA spool is not just hard to print, it is worse plastic.
The symptoms stack in a predictable order. Stringing gets worse first, because retractions stop cleanly sealing the nozzle when steam keeps pushing plastic out. Then surfaces go hazy and rough. Then interlayer bonding drops; a part that should take 400N to break snaps at 180N. By the time you hear popping at the nozzle, the spool has been wet for weeks.
Humidity Thresholds by Material: PLA, PETG, TPU, ABS, Nylon
Relative humidity is the number that matters, not absolute moisture content. Below are the targets we hold spools at in active storage, drawn from supplier TDS numbers and what actually prints clean on our floor.
| Material | Storage RH | Red-flag RH | Notes |
|---|---|---|---|
| PLA | <30% | >50% | Forgiving short-term, degrades over months |
| PETG | <25% | >40% | Stringing is the first tell |
| ABS/ASA | <25% | >45% | Less hygroscopic than PETG but splits layers when wet |
| TPU | <15% | >30% | Pops at the nozzle within days of exposure |
| Nylon (PA6, PA12, PA-CF) | <10% | >20% | Hardest to keep dry, most punishing when it isn't |
| PVA / BVOH | <10% | >15% | Can fuse to itself on the spool |
| PC / PC-blend | <15% | >30% | Bubbles and brown tint when wet |
Shop RH in most of North America runs 35% to 60% depending on season and HVAC. That means PLA on an open shelf sits in the yellow zone most of the year, and nylon on an open shelf is wet before you finish unboxing it.
A useful rule: if a material's spec sheet calls for drying before first use, treat it as always needing active storage. Nylon, PVA, PC, and TPU all live in sealed containers in our shop, not on shelves. PLA, PETG, and ABS get the open shelf only when we expect to burn through the spool within two weeks.
The RH target is for the air immediately around the spool. A gasket tote reading 20% at the hygrometer does not mean a spool fresh from a wet shipping box is at 20% inside its plastic. Give fresh spools 48 hours in the box before trusting the equilibrium.
Measuring Humidity Correctly: Hygrometers, Placement, and Calibration
Every cheap digital hygrometer lies by some amount. The ones in gumball packaging lie by 10% or more. The $12 to $20 units from reputable sensor brands (ThermoPro, Govee, SensorPush) lie by 2% to 5%, which is tolerable if you calibrate them.
Salt-test calibration is the only honest check. Put a tablespoon of table salt in a shot glass, add just enough water to make a slurry without dissolving it, then seal the shot glass and hygrometer together in a zip-top bag for 8 to 12 hours. A saturated salt solution holds its headspace at 75% RH at room temperature. Whatever your hygrometer reads minus 75 is its offset.
Sensor placement matters more than people think. Stick the probe to the inside wall of the dry box, not on top of the desiccant and not pressed against a spool. Air near fresh desiccant reads drier than the box average; air next to a spool reads wetter. We want the number that represents the bulk atmosphere.
Then wait. A hygrometer sealed into a box with a new desiccant charge takes 18 to 24 hours to stabilize. The fast drop in the first hour is the sensor itself equilibrating, not the box. If the reading is still dropping a day later, your desiccant is working. Leave it alone.
Dry Box Builds: Gasket Tote, Vacuum Bag, and Active Heated Box
Three tiers cover a shop. Match the tier to the material and how often you open the container.
Tier 1: Gasket tote with desiccant. A 20 to 27 gallon Sterilite or IRIS gasket tote, two 500g silica gel canisters, and a calibrated hygrometer. Cost: around $35 all in. Holds 8 to 12 one-kilogram spools of PLA or PETG at 20-30% RH indefinitely, as long as you refresh the desiccant every two to four months. This is our workhorse for active-rotation PLA, PETG, and ABS.
Tier 2: Mylar vacuum bag archive. 10"x14" mylar bags, a 100-count box of 50cc oxygen/moisture absorbers, and a $25 impulse sealer. Vacuum-seal each spool individually with one absorber. Cost per spool: about $0.60. This is the correct answer for spools you will not touch for six months or more: new old stock, backup rolls, specialty colors.
Tier 3: Active heated dry box. An insulated tote with a PTC heater, fan, and controller holding 45-55°C with a humidity sensor loop. We run a modified Sterilite with a 100W PTC element, a 60mm fan, and an ESP32 controller; parts around $60. This is the only acceptable long-term home for nylon, PVA, PC, and specialty TPU. A passive box with fresh desiccant can get nylon to 15% RH, but it cannot hold it through the winter.
For teams that do not want to build Tier 3, a commercial unit like the Sunlu Filadryer E2 handles two spools at once with a controlled heater and a decent fan, and it can print-through directly from its outlet. We run three of them for nylon and TPU in the shop and they pay for themselves in saved reprints inside a month.
One thing to know about every dry box: opening it is the single biggest moisture event in its life. Room air floods in, desiccant has to absorb all of it, and the RH spikes for hours. Design your workflow so you open each box once or twice a day, grab what you need, and close it. Leaving the lid off for a 20-minute swap undoes a week of desiccant work.
Desiccant: Silica Gel vs Molecular Sieve, and When to Regenerate
Indicator silica gel, the orange-to-green beads in little packets, is fine for PLA and PETG. It adsorbs water down to about 20% RH equilibrium, which is enough for the forgiving polymers. The color change is roughly honest; when the beads go uniformly green or dark, they are full.
Silica gel is the wrong desiccant for nylon. Nylon wants under 10% RH, and silica gel physically cannot pull air that dry. This is where 3A or 4A molecular sieve earns its place. Molecular sieves are aluminosilicate beads with calibrated pore sizes that adsorb water to a much lower partial pressure, holding a box under 5% RH as long as they have capacity left.
Molecular sieve has two downsides: it is not color-indicating, and it is more aggressive. You need a hygrometer to know when to regenerate, and you do not want it in contact with wet filament directly. It will pull moisture from the plastic surface faster than it diffuses from the core and cause the outer wrap to get brittle. Use a mesh bag, separate tray, or a perforated canister.
Regeneration is where most people cheat and lose. Silica gel wants 2 to 3 hours at 120°C in a convection oven; indicator beads go back to deep orange when they are truly dry. Molecular sieve wants 2 to 4 hours at 230-250°C. A microwave or a dehydrator cannot reach that, and a kitchen oven only can if you verify with an oven thermometer. Underheating the sieve just drives surface water deeper into the bead; it tests dry at the surface and is still wet in the core.
A simpler rule of thumb: if the hygrometer in your box drifts upward three days running after a refresh, the desiccant is spent. Rotate a second, already-regenerated charge into service and bake the spent one.
When to Re-Dry: Shop Symptoms, Timers by Material, and Settings
Re-drying is for spools that got wet. Storage keeps spools dry; re-drying brings a wet one back. The giveaways are consistent across our shop: popping or hissing audible at the nozzle, fuzzy stringing that was not there last week, a dull or hazy surface finish on walls that used to be glossy, and layer adhesion that fails at stress points it did not before.
Settings matter and the glass-transition temperature is the hard ceiling. Dry too hot and the spool deforms; the plastic layers fuse into a disc you cannot unwind. Below is our shop table, cross-checked against the Prusa filament drying guide and manufacturer data sheets.
| Material | Dry temperature | Hours per 1kg spool | Notes |
|---|---|---|---|
| PLA | 50-55°C | 4-6 | 65°C warps PLA on a standard spool; stay under |
| PETG | 65°C | 6-8 | Forgiving; 70°C acceptable if you watch it |
| TPU (85A-95A) | 50°C | 8-12 | Softer durometers warp faster; stay low, dry long |
| ABS/ASA | 70°C | 4-6 | Tolerates higher but not needed |
| Nylon (PA6, PA12) | 80°C | 10-16 | PA6 is the thirstiest; a fresh-from-vacuum spool still benefits from 4h |
| PC / PC-blend | 80°C | 6-8 | Below 90°C to avoid optical fogging on clear grades |
PLA at 55°C versus 65°C is where we see the most damage. A dedicated filament dryer at 65°C for 6 hours leaves a cardboard-core spool slightly oval, and that oval causes extrusion inconsistency for the rest of the spool's life. 55°C for 6 hours gets PLA dry enough to print clean and the spool stays round.
For TPU, covered in more depth in our TPU filament temperature guide, the rule is low and slow. 50°C for 10 hours. Higher temperatures flash the surface plasticizers and leave a tacky residue that gums up the extruder gears within a few prints.
A dry cycle is not complete until the hygrometer inside the dryer settles under the material's storage target. If your TPU dryer reads 18% RH at hour 10, give it another two hours. Pulling early means the spool starts re-absorbing during the next print.
Printing Straight From the Dry Box: PTFE Routing and AMS Considerations
A dry spool exposed to shop air during a 14-hour nylon print is a wet spool by hour six. The dry box is not just storage, it is the feed source during the print.
The simplest build: cut a 1/4" hole in the end of the gasket tote, push a PC4-M10 pneumatic coupler through with a rubber grommet, and run 4mm PTFE tube from the coupler to the printer's extruder inlet. The spool rolls on two 608 bearings mounted inside the box. Keep the exit hole small and grommeted. A loose fitting leaks air both ways.
AMS-style multi-material units ship with desiccant trays and a humidity sensor, which helps but has limits. These enclosures are not airtight, and in a 50% RH shop the internal RH stabilizes around 25-30% even with fresh desiccant. That is fine for PLA and PETG. It is not enough for nylon or TPU. Vendor guidance for those materials is to feed from an external dry box through the external spool slot.
Refresh AMS desiccant every four to six weeks of regular use. The stock packets saturate fast because the AMS opens for every filament change. Keep a stash of 30g silica canisters regenerated and ready; swap, do not top off.
Every time you open the AMS lid, you reset the internal humidity to room RH. If you need to swap spools on an AMS currently printing nylon, swap fast and close faster. Better: use the external-feed slot and leave the AMS lid closed for weeks at a time.
FAQ
What humidity level is safe for filament storage?
Under 20% RH covers almost everything in your shop. PLA, PETG, and ABS are forgiving up to 30-40% for a few weeks, but TPU, nylon, PVA, and PC need to live below 15% to print clean. Your indoor shop air is probably 35% to 60% depending on season and HVAC, which means every spool that isn't in a sealed container is drifting toward wet.
Can you dry filament in a kitchen oven?
Only with a calibrated oven thermometer sitting on the rack next to the spool. Most kitchen ovens cycle 15-25°F above and below their setpoint, which means an oven set to 150°F can easily hit 175°F, enough to warp PLA into an unusable oval. A dedicated filament dryer runs around $60 and holds temperature within a couple of degrees; it pays for itself the first time it saves a spool.
How long does filament last in a sealed bag with desiccant?
A vacuum-sealed mylar bag with a fresh 50cc oxygen absorber keeps PLA and PETG printable for 12 to 18 months with no measurable degradation. Nylon and TPU drift sooner because the bag's seal is not perfect over long periods, so re-check RH on those even when the bag looks intact. If the bag has lost its vacuum pull, assume moisture ingress and re-dry before loading.
Does PLA really absorb moisture if it's stored dry?
Yes, just slowly compared to the thirsty plastics. PLA at a 50% RH shop will reach noticeable moisture content in two to four weeks, enough to cause stringing and surface haze on fine detail. The polymer chains also slowly hydrolyze over months, so a year-old open PLA spool is not the same plastic it was new, even if it prints without popping.
Is a vacuum sealer or a dry box better for long-term storage?
Vacuum bags for the archive, dry boxes for the active rotation. A gasket dry box loses RH every time you open it, so spools you touch weekly are the right residents. Spools you will not see for six months belong in individual mylar bags with a fresh absorber: zero opening events, zero desiccant refreshes, zero worry.
Closing
Pick one tier and start. Most shops jump straight to building a heated chamber and never calibrate a hygrometer; we have seen that fail more than it succeeds. Buy a $15 sensor, salt-test it this week, and measure your actual shop RH for seven days. Then size the storage to what you measured, not what you guessed.
If you're printing nylon, TPU, or PC regularly, a dedicated active dryer earns its shelf space in the first month. The Sunlu Filadryer E2 is the one we point people at for a two-spool solution that prints straight through the outlet, same unit we run three of on our production floor for the thirsty plastics.

