Filament storage and drying for a print farm

Store filament sealed with desiccant, ideally below 20% relative humidity, and dry it when you can't vouch for how it was stored, at the temperature published by the maker of that spool. In a farm, use a separate dryer, because drying on a printer's bed takes that printer out of production for 12 hours.

We keep spools on open wall racks and shelves in our print room, and we have a dryer. That is looser than what Bambu asks for below. We have no humidity log or moisture failure figure to publish, so we can't say what that costs us.

How sensitive each material is

Bambu's wiki sets out when it wants drying. Its filament guide gives saturated water absorption at 25 °C and 55% RH. Both describe Bambu's own filaments.

Absorption alone doesn't predict the rule. PC absorbs less than PLA in this table, but Bambu requires drying for PC and only recommends it for PLA. Prusa names polyamide, polypropylene, PVA and BVOH as the highly hygroscopic group, to store dry or dry before every print.

Wet filament shows up as stringing, oozing, holes, rough surfaces and reduced strength, says Bambu. Prusa adds blobs, plus bubbling or smoke during extrusion.

Drying temperatures: makers disagree

These are air temperatures for a dryer or forced-air oven, in °C and hours.

The makers are not describing the same filament, so a gap here is not an error. Use the figure for your spool. Polymaker's pages add that drying applies "only if the material has absorbed moisture". We could read its PETG figure only through a search excerpt, because the page blocked us. Bambu's PDF guide also differs from its wiki, with 55 °C for PLA and 80 °C for PC, so don't treat any single figure as exact.

Times depend on how wet the spool is. Polymaker's PolyDryer manual lists 6 hours for PLA, PETG, ABS, ASA and PC, 12 for TPU, 18 for PA and 24 for PVA. Its basis is samples left for 5 days at 25 °C and 70% RH.

Three limits to respect:

• Too hot. Prusa warns that going over its temperatures makes filament "soften and stick together", and says to keep to its times to avoid overdrying.
• The spool. Bambu warns that some third-party spools may deform. Prusa says its old black spools take only 45 °C without a screw fitted.
• Kitchen appliances. Bambu says not to use a microwave or a kitchen oven, because the heat is uneven. Prusa says home ovens measure temperature poorly, so use an external thermometer if you do.

Drying on the bed or in an AMS

All from Bambu's wiki. Bed values are plate settings, so never put them into a dryer. Open-frame machines such as the A1 can't dry this way.

In a farm the bed method has a clear cost. Twelve hours is longer than our 9-hour plate in cost per plate, so each spool costs that printer more than one plate. A P1S also has no built-in drying function, so you set the bed and track the time by hand.

Humidity targets and dry boxes

Of the three makers here, only Bambu gave us numbers this direct, and it also sells the AMS and the desiccant:

• Typical indoor air is 45% to 65% RH.
• Most filament stays dry only for a limited period below 20% RH.
• At 20% RH, filament generally stays dry for 2 to 7 days, depending on the type.
• At about 55% RH, freshly dried filament absorbs enough to affect print quality within 2 to 12 hours.

If that last figure holds in your room, drying a spool and then putting it back on an open rack undoes the work within a day. So dry a spool shortly before it prints, or dry it into a sealed box. Bambu suggests a hygrometer inside the box so you see the real value.

Other makers say less. Prusa's article gives no humidity target. It says to use a closed box with desiccant or a vacuum bag with desiccant, and if space is short, to protect the most hygroscopic materials first. Polymaker's PolyDryer page reports an internal test, for reference only: starting from filament left open for 2 weeks at 20 °C and 60% RH, its box ended between 10% and 30% RH, depending on power level.

Two material rules from Bambu are worth writing on the shelf:

• Nylon. PA-based filament absorbs a lot after about 3 months in open air, and is then very hard to dry at 80-90 °C. Seal what you won't use soon.
• PC. Repeated heating and cooling can make it brittle. Keep it sealed after drying, so it needs fewer cycles.

Desiccant runs out too. Bambu's silica gel beads change colour, can be regenerated at 80-90 °C and never above 100 °C, and last 5 to 15 cycles.

What we would do first

This is opinion, and it follows the sources above, not our own measurements.

1. Seal everything marked Required that isn't on a printer.
2. Put a hygrometer in the room for a few weeks, so you know which line above applies to you.
3. Buy one dryer, so drying doesn't use up printer hours.
4. Note the spool in your plate log when a print fails. A failed plate costs up to about CHF 14 in our example, depending on how far it got. Only a log tells you whether moisture is part of that.

What we can't state

• How much moisture costs a farm. We found no public failure data tying room humidity to failed plates, and we have none of our own.
• How Bambu reached its hours and days. The wiki gives no method.
• What lab results mean for a spool on a shelf. A 2026 study soaked printed parts in water for up to 720 hours, and PETG held up best. We read only its abstract. A 2022 study reported by Hackaday also submerged samples. That is printed parts in water, not filament in room air.

For tracking spools in software, the manual is filament in Rackrew.