Additive manufacturing, also called three-dimensional (3D) printing, is an emerging industry growing 19.5% in 2021 [1]. Due to decreasing costs, increasing accessibility, and improving technologies of small-scale 3D printers, non-industrial space usage such as schools, libraries, and makerspaces is leading to potential exposures in public populations and their use is anticipated to increase.

3D printer feedstocks may contain substances that have a history of causing adverse health outcomes. Styrene, benzene, formaldehyde, toluene, other volatile organic compounds (VOCs), polycyclic aromatic hydrocarbons, metals, and ultrafine particles (diameter<100nm) have been measured from 3D printers in office and university settings [2–6]. Chronic adverse health outcomes from exposures to 3D printer emissions are difficult to assess because of their relatively short history of use. A health survey of 46 employees who regularly used 3D printers in various settings found that 59% had respiratory symptoms, including nasal congestion (12, 26%), rhinorrhea (9, 20%), itchy nose/throat/eyes (9, 20%), and cough (9, 20%), at least once per week in the past year, and that working more than 40h per week with 3D printers was significantly associated with a respiratory-related diagnosis [7]. The previous history of adverse health outcomes and documented potential exposures necessitates improvements to 3D printing and preliminary guidelines with health and safety approaches to protect employee and student health. Some printer companies have addressed health and safety concerns in schools [8], but few studies have assessed emissions and the potential for occupational exposures to teachers, librarians, and other employees in these settings. There is also limited knowledge on cost-effective engineering controls to reduce 3D printing emissions in these spaces. Most participants did not use any personal protective equipment (PPE).

The purpose of this project was to evaluate cases of 3D printer emissions in non-industrial settings such as schools, libraries, and makerspaces, to characterize the potential for user exposures. To accomplish the goal, we observed work processes, practices, and conditions, including PPE use, measured 3D printing process emissions, and evaluated workspace ventilation and controls. Personal air sampling was not collected as part of these evaluations.

Air sampling results showed that printer start-ups and failures were likely sources of potential exposures to ultrafine particulate (UFP), as seen in PCI results from our study (Figures 1 and 2) and in agreement with other studies [2, 12–17]. Results from PCIs in location D3 supported the hypothesis that print failures and start-ups could be sources of potentially higher exposures to UFP. PCI results in the location D4 did not show increases during the start-up of four 3D printers; this could have been attributed to the use of enclosures that were exhausted outdoors for all printers.

Measured as TWA concentrations over a full shift (typically 8h), occupational exposure limits (OELs) are the highest level of exposure an average worker may be exposed to without incurring the risk of adverse health effects. While area air samples cannot be compared directly to OELs, we measured VOC concentrations orders of magnitude lower than the relevant OEL at all locations. The highest concentrations of acetone measured was 0.13ppm (ppm) compared to the lowest OEL established by NIOSH of 400ppm, a difference of three orders of magnitude; similarly, isopropanol, and ethanol were found two and five orders of magnitude below their lowest OELs, respectively. The highest concentration of acetone was measured at location C with no LEV (297min), isopropanol was at location B with no LEV (347min), and ethanol was Location D1 with enclosures that were not exhausted (234min). For each location, participants conducted 3D-printing for the longest time that was typical. We collected air samples for the entire time of 3D-printing activities, ranging from 89 to 452min (mean 289min).

Isopropanol has been identified in other filament-based printing studies [3, 18, 19]. However, other sources could have contributed to the concentrations measured: isopropanol was used to clean printer beds and used in two of the PCIs. Acetone has also been found in previous studies [18–21] and is a common solvent in household and personal care products. Decamethylpentasiloxane was found in multiple locations but is a common ingredient in personal care products [22] and has not been found in other studies on 3D-printing emissions. Ethanol was found in several locations in facility D and has been noted in another study [14]. One sample collected at location D3 found quantifiable but low levels of methylene chloride, which is categorized as an occupational carcinogen [23]. Location D3 is a university chemistry lab; methylene chloride as a common solvent and liquid for chemical analysesandisnotlikelyassociatedwiththis3D-printingactivity. Ethyl ether has not been found in other chamber or field studies and may be from the lab setting or an activity occurring nearby. The other VOCs that we measured that were detectable but not quantifiable may have come from 3D printing emissions or other VOC sources in the room, such chemical storage, new furniture and carpets, and other products used in makerspace and hobby settings [24]. While we documented other activities occurring in sampling locations, it was not definitive where these VOCs were emitted from.

For our study, we used a sampling strategy to pair PCIs evaluating UFP number and size concentrations with air samples analyzed for VOCs that are likely associated with 3D printing with filaments and resins. Using this strategy, along with worker and workplace observations, the results from this evaluation suggest a low potential for exposures significant to health in the air from 3D printer emissions. However, this and other studies show that UFP and VOCs can be emitted from filament-based printers. Research is lacking on health hazards from exposure to emissions from 3D printers. Given the results presented above, effective engineering controls and safe work practices could be used to lower potential exposures to 3D printer emissions.

Printer model, filament characteristics, and printer settings can play a role in emissions. Both PLA and ABS filaments were used at locations in this study. Studies have found that ABS filaments may emit higher concentrations of UFP compared to PLA filaments [25]. PLA filaments are preferred because of their potential as lower emitters. Although not measured in this study, ABS filaments are also more likely to emit VOCs of concern, such as benzene or styrene [2, 12, 14, 15].

While some locations evaluated, such as locations B, C, and D2 were not traditional “laboratories,” these locations could be considered laboratories since chemicals or activities were used that can generate chemical emissions. The air change rates we calculated in most locations (A1, A2, A5, D1, D2, and D4) were below recommendations of at least 6–10 air changes per hour for laboratory spaces, as recommended by the National Research Council [26]. However, some of the 3D printing activities, such as those in A2 and D4 were in an enclosure exhausted to the outdoors.

Maintaining negative airflow pressurization for 3D printing spaces is preferred to prevent migration of VOCs, UFP, and odors from moving out of the areas. In locations where we were able to evaluate the ventilation systems, we found some labs were positively pressured compared to surrounding areas. This means that air flowed out of the location into adjacent areas. A better practice is to maintain the 3D-printing spaces as negatively pressured compared to other areas so that air flows into the location and reduces the possibility of unnecessary exposures to occupants of adjacent areas.

We observed inconsistences in work practices and the implementation of hazard communication that may contribute to unnecessary exposures. For example, user workstations were near 3D printers during printing in some locations, SDSs were not always readily available for print materials, and standard operating procedures for 3D printing activities were not available (hazard communication). Additionally, we noted instances where PPE was required to be used in a location, and both students and workers were not following PPE requirements, or there was confusion about what PPE was required. Positioning 3D printers as far away as possible from people during printing and training users on how to safely use 3D printers, available controls, and hazards from use could help reduce unnecessary exposures.

Anecdotally, university health and safety personnel at both university sites explained that 3D printing activities were widespread throughout campus and often unknown to health and safety staff. K-12 locations typically had a better understanding of 3D printer locations and activities.

The field of industrial hygiene strives to control exposures to hazards in the workplace to protect workers. The hierarchy of controls is a framework to determine which actions will best control exposures [27]. Engineering controls can reduce or prevent hazards from coming into contact with workers, and include modifying the workspace, equipment, barriers, and ventilation. Administrative controls establish work practices and processes that reduce exposures to hazards. Additionally, PPE is generally recommended when engineering and administrative controls to reduce exposures to a desirable level are not feasible or available.

While air sampling results suggest potential for low level exposures to VOCs and UFP, steps can be taken to prevent unnecessary exposures to these, control odors, and reduce emissions moving out of printing spaces. The following considerations and supporting documents can be incorporated into plans for health and safety professionals to consider when planning for the use of 3D printers. These materials provide information for all users of 3D printers, including teachers, librarians, and students. These considerations are based on general findings from our evaluations at these and other non-industrial locations and are organized based on the hierarchy of controls [27].

The NIOSH Approaches to safe 3D printing: a guide for makerspace users, schools, libraries, and small businesses [9] describes the use of an enclosure with LEV, which may include a laboratory fume hood or a bench-top ventilated enclosure, as an engineering control to capture and remove UFP and chemicals emitted by 3D printers. LEV may exhaust printer emissions directly outdoors or through a high-efficiency particulate air (HEPA) filter to remove particles prior to exhaust or recirculation. Locations A2 and D4 used enclosures that were aligned with this approach. PCIs at location D4 did not show increases in UFP following startup. At location A2, PCIs measured increases in UFP when enclosures were opened following a printer malfunction. Some researchers have shown that the use of a simple ventilated enclosure with an exhaust fan and attached HEPA filter reduced the emissions released to the room by 97% [28]. The enclosure used in location D1 is another example of an enclosure for 3D printers, which would more effectively capture emissions if ported and connected to an exhaust system. Examples of a ventilated enclosure designed to contain 3D-printer emissions from multiple printers were observed in location D4.

Another option for controlling 3D printing emissions is connecting LEV to the printer print heads [29] or effectively positioning LEV near the printers to capture emissions. If there are concerns about VOC emissions or odors, consider adding gas and vapor filters to the exhaust of these ventilation control systems. For 3D printers placed in ventilated enclosures, users should wait for printer emissions to be removed from the ventilated enclosure before opening the enclosure. The amount of time required for printer emissions to be removed is based on the size of the enclosure and how much airflow is exhausted. Users should also wait for the emissions to decrease before approaching a printer that has malfunctioned.

Other steps can be taken to adhere to best-practice and prevent potential exposures, control odors, and reduce emissions in the printing locations. For example, starting a print when there are fewer occupants in the area or leaving the area following startup. Starting the printer head heating without filament (if possible), cleaning the nozzle after printing of residual filament, and setting a lower operating temperature for the printer nozzle when possible are additional printing strategies that can help lower printer emissions [28, 30]. Research has also shown that certain filaments and print materials emit more VOCs and UFP than others. For example, ABS-based filaments have been shown to emit more UFP than PLA-based filaments [28]. Choosing lower-emitting filaments can also reduce the potential and amount of exposure.

NIOSH’s Advanced Manufacturing webpage has 3D printing guidance documents for additional information. For example, NIOSH developed printable posters including one specific to filament-based 3D printing that cite health and safety concerns with different control options and other information to reduce exposure to potential hazards, as well as a one-page guidance for plastic filament. NIOSH also recently released guidance titled Approaches to safe 3D printing: a guide for makerspace users, schools, libraries, and small businesses [9].

Quantifiable amounts of some VOCs (including isopropanol, acetone, ethanol) were measured in air samples and UFP emissions were detected by PCI during 3D printing. The PCI results showed that printer start-ups and failures were likely sources of potentially higher exposures to UFP. To decrease the potential for exposures to students and workers, we recommended separating print spaces, using LEV such as ventilated enclosures, and administrative controls, such as implementing policies on allowable filaments and training on preventing exposures during 3D printing tasks.