De novo Powered Air-Purifying Respirator Design and Fabrication for Pandemic Response

SUMMARY The rapid spread of COVID-19 and disruption of normal supply chains resulted in severe shortages of personal protective equipment (PPE), particularly devices with few suppliers such as powered air-purifying respirators (PAPRs). A scarcity of information describing design and performance criteria represents a substantial barrier to new approaches to address these shortages. We sought to apply open-source product development to PAPRs to enable alternative sources of supply and further innovation. We describe the design, prototyping, validation, and user testing of locally manufactured, modular, PAPR components, including filter cartridges and blower units, developed by the Greater Boston Pandemic Fabrication Team (PanFab). Two designs, one with a fully custom-made filter and blower unit housing, and the other with commercially available variants (the “Custom” and “Commercial” designs respectively) were developed. Engineering performance of the prototypes was measured and safety validated using NIOSH-equivalent tests on apparatus available under pandemic conditions, at university laboratories. Feedback on designs was obtained from four individuals, including two clinicians working in an ambulatory clinical setting and two research technical staff for whom PAPR use is a standard part of occupational PPE. Respondents rated the PanFab Custom PAPR a 4 to 5 on a 5 Likert-scale 1) as compared to current PPE options, 2) for the sense of security with use in a clinical setting, and 3) for comfort. The three other versions of the designs (with a commercial blower unit, filter, or both) performed favorably, with survey responses consisting of scores ranging from 3–5. Engineering testing and clinical feedback demonstrate that the PanFab designs represents favorable alternative PAPRs in terms of user comfort, mobility, and sense of security. A nonrestrictive license promotes innovation in respiratory protection for current and future medical emergencies.


PURPOSE
This procedure establishes the method by which a generated corn oil aerosol is used for determining if powered, air-purifying respirators (PAPRs) supplied with loose fitting respiratory inlet coverings meet the facepiece-fit requirements at 42 CFR Part 84, Subpart K, Section 84.176(b).

GENERAL
This STP describes the Determination of Respirator Fit, Quantitatively Using Corn Oil Aerosol, For Powered Air-Purifying Respirators With Loose-Fitting Respiratory Inlet Coverings test procedure in sufficient detail that a person knowledgeable in the appropriate technical field can select equipment with the necessary resolution, conduct the test, and determine whether or not the evaluated product passes the test.  Environmental test chamber -The chamber shall be designed so that the individual(s) performing fit factor testing are visible at all times while in the chamber. The chamber design must include an entry vestibule designed to allow safe entry and exit from the chamber with minimal disturbance to both the aerosol concentration and the concentration uniformity. An example of a charged corn oil chamber is illustrated in Figure 5.
3.8. Chamber Communications -Electronic audio communications (chamber loudspeaker) are employed to transmit a real-time voice signal from laboratory technicians to test subjects to ensure that test subjects can clearly hear when to start and stop the test exercise regimen or receive safety information.
3.9. Facial Size Measurement Calipers -Calibrated face sizing calipers shall be used to measure the human test subject to the requirements identified in Appendix A. Examples of calipers are sliding measurement calipers: Seritex model GPM 104, 0-200 mm length, or spreading measurement calipers: Seritex model GPM 106, 0 -300 mm width. These are shown in Figure 6 and Figure 7.
3.10. Facepiece Direct Probes. The sample probes shall be of the shape defined by Liu [AIHAJ (45);278-283, 1984] and shall not interfere with the fit or function of the respirator. Figure 8 and Figure 9 are photographs of the probes used. Each probe bulkhead fitting is sealed using two rubber washers, one metal washer and one nut.

TESTING REQUIREMENTS AND CONDITIONS
4.1. Prior to beginning any testing, confirm that all measuring equipment employed has been calibrated in accordance with the testing laboratory's calibration procedure and schedule. All measuring equipment utilized for this testing must have been calibrated using a method traceable to recognized international standards when available.

4.2.
General respiratory inlet covering fit requirements for PAPRs -4.2.1. The fit test shall be performed using a panel of test subjects of various facial sizes measured in accordance with the NIOSH Bivariate Panel (NIOSH Panel). The measured face length and face width are used to designate the subject's NIOSH Panel cell number, as illustrated in Figure 10 according to the manufacturer's user instructions, prior to entering the chamber; however, upon entry into the test chamber neither the respirator, nor the respiratory inlet covering shall not be re-adjusted.

4.3.
Test subject selection 4.3.1. For PAPRs with up to three sizes of loose-fitting respiratory inlet coverings, the test will be conducted employing 18 individual test subjects which represent the NIOSH Panel (Appendix 8.5). See Table 1 for the suggested test subject distribution in relation to the NIOSH Panel (NIOSH is allowing flexibility in the use of subjects from well-populated panel cells. NIOSH will attempt to test using a panel that contains at least one subject from each cell, but when subjects from all cells are unavailable they may be supplanted by adding subjects from more populated cells; however, no more than four subjects from any one cell may contribute to the overall test panel composition for any respirator under evaluation.)   , and 10 shall be tested wearing the larger size initially.
4.3.1.3.4. If a subject does not achieve a trial pass in the first inlet covering size evaluated, the subject will be retested in the next available size (second trial). A subject failing in the smaller size inlet covering, can try the medium size (second trial), and then larger size (third trial) inlet covering. A subject failing to achieve a trial pass in the medium size inlet covering can try the smaller and larger size inlet coverings. A subject failing in the larger size inlet covering can try the medium, then the smaller size inlet covering. A subject failing to achieve a trial pass in any of the sizes available for testing is considered a face-size failure.
4.3.1.3.5. The test administrator may determine whether or not third trials are needed for test completion. For a subject failing to achieve a trial pass in the small or medium size inlet covering, trying the large size may not be necessary since the large size may be obviously too big for the subject. For a subject failing to achieve a trial pass in the large and medium size inlet covering, trying the small size may not be necessary since the small size may be obviously too small for the subject.    software program. Test Administrator will start the software program and relay the information of time to start the test, exercise, and timing of the exercise being performed.
5.2.6. A fit factor test consists of a set of four two-minute standard exercises. During the test, each human subject will perform the following four exercises for two minutes each in the below listed sequence. Subjects should not touch any portion of the respirator during any part of the testing exercises. Test administrator will give verbal commands to stop and start each exercise.   6.2. The number of face-size failures will not exceed four.

6.3.
If an overall pass is achieved, but three subjects report the same issue about the comfort of the facepiece, the test will be considered a failure.
6.4. Fit factor for loose-fitting PAPR -6.4.1. For each face-size trial, an overall average fit factor of 500 must be achieved.

Menton-Sellion Length
Distance as measured with a sliding caliper in the midsagittal plane between the menton landmark and the sellion landmark.   3.1.6. Temperature and humidity chamber capable of maintaining 38 ±2.5 °C and 85 ± 5% relative humidity.
3.1.7. Respirator filter holder supplied for specific manufacturer type which is compatible with TSI filter tester. NIOSH will not be obligated to use these holders for actual certification testing. All manufacturer test fixtures must be correlated with the NIOSH test method.
3.1.8. Thermal printer (supplied with TSI 8130) or optional data acquisition system.
3.1.9. TSI, Green Line paper, part number 813010. Lot number must be included on each box. Each lot number must include the "Penetration vs. Resistance graph".

TESTING REQUIREMENTS AND CONDITIONS
4.1.
Prior to beginning any testing, confirm that all measuring equipment employed has been calibrated in accordance with the testing laboratory's calibration procedure and schedule. All measuring equipment utilized for this testing must have been calibrated using a method traceable to recognized international standards when available. Respirator filters will be challenged by a NaCl aerosol at 25 ±5 °C and a relative humidity of 30 ±10% that has been neutralized to the Boltzmann equilibrium state. The particle size distribution will be a count median diameter of 0.075 ± 0.020 micrometer and a geometric standard deviation not exceeding 1.86. Each respirator filter unit will be challenged with an aerosol concentration not exceeding 200 mg/m 3. 5.1.1. The NaCl aerosol concentration will be determined on the days that initial penetration testing is performed by the following gravimetric method and calculated as milligrams per cubic meter (mg/m 3 ).
5.1.2. Weigh a 102 mm filter to the nearest 0.1 mg., mount in the gravimetric filter holder, subject it to the generated aerosol at 30 Lpm for 40 minutes and reweigh the filter. Use a timer to monitor the duration of the test. Record the pre-and post-weights, time, and average flow rate on the data sheet and calculate the aerosol concentration in mg/m 3 by the following formula: The correlation factor is used by the software to express the upstream photometer signal in terms of the downstream photometer signal. 5.1.5. The NaCl particle size distribution shall be verified using "green line" filter discs supplied by TSI with a known penetration range. Graphs of penetration vs. resistance for two sheets and five sheets of stacked filter discs are supplied with each lot of the standard filters, with a central line and upper and lower lines representing the expected penetration range at a given resistance. The test data should fall within an acceptance zone having boundaries defined by the upper and lower curves on the graphs. The standard filter test using both 2 sheets and 5 sheets will be run at least once in each 8 hour test period to verify that the aerosol distribution is within the acceptance zone.

5.2.
Respirator filters will be pre-conditioned at 85 ± 5% relative humidity and 38 ±2.5°C for 25 ±1 hours. After conditioning, the filters shall be sealed in a gas tight container and tested within 10 hours.

5.3.
Filters will be mounted and sealed on holders to prevent leakage around the filter holder. Single air purifying respirator filters will be tested at a challenge flow rate of 85± 4 Lpm. Filters used as pairs on a respirator are tested using a single filter of the pair at 42.5 ± 2 Lpm challenge flow rate. Filters used in threes are tested using a single filter of the set at 28.3 ± 1 Lpm challenge flow rate.
5.3.1. The challenge flow rate must be checked for stability for at least 30 seconds prior to testing.

5.4.
A sample of 20 filter units will be tested against the NaCl aerosol. Three filters will be loaded until the aerosol mass loading levels as shown in the table below are reached and evaluated to determine the method for the remaining 17 filters. This is the mass amount of NaCl aerosol that has contacted the filter.
5.4.1. Type 1. If preliminary testing of all three initial test filters consistently results in a straight line (Figure 3), for the remaining 17 filters, record the initial penetration reading.
5.4.2. Type 2. If filter testing of all three initial test filters consistently results in a curve which indicates increased efficiency during the complete run ( Figure 3), for the remaining 17 filters, record the initial penetration reading. 5.4.3. Type 3. If filter testing of all three initial test filters consistently results in decreased efficiency over time (Figure 3), load the remaining 17 filters with NaCl to the level specified in the table above and record the maximum penetration reading.
5.4.4. Type 4. If filter testing of all three initial test filters consistently results in increased efficiency, then a decrease in efficiency, and then flattens out during the remainder of the complete run ( Figure 3), for the remaining 17 filters, record the maximum penetration reading after reaching and maintaining a flat line for a period of 20 minutes following the decreasing segment in efficiency.
5.4.5. For any other filter type, determine loading at which maximum penetration consistently occurs and test at that loading value for the remaining 17 filters.
5.4.6. If any one of the 20 filters have a penetration greater than 0.030%, further testing of that filter will be terminated. Any filter that exceeds the specified limit shall be remounted and retested to ensure that leakage was not caused by a mounting leak If retesting eliminates the excessive leakage and testing has gone beyond the initial penetration, that sample will be considered an invalid sample, and another tested in its place.

5.5.
The penetration of the first three filters will be measured, recorded, and printed at approximately 1-minute intervals during the test period. The highest penetration observed throughout the test of each filter will be recorded as the maximum penetration of that filter.

5.6.
Determine and record on the data sheet the maximum filter penetration for each of the 20 filters.   Prior to beginning any testing, confirm that all measuring equipment employed has been calibrated in accordance with the testing laboratory's calibration procedure and schedule. All measuring equipment utilized for this testing must have been calibrated using a method traceable to recognized international standards when available.

4.2.
Noise level test must be performed in a location that has a maximum background noise level of no more than 60 dBA.      Three complete respirator systems will be evaluated for low flow warning device activation.

4.4.
Determination of Low Flow Warning Device Visibility will be completed for any respirator with a visual warning device. The standard testing procedure is described in CVB-APR-STP-0087.

4.5.
Determination of Low Flow Warning Device Sound Level on Series PAPR100 will be completed for any respirator with an audible warning device, where the warning provided is audible only, or other warnings are not readily apparent. The standard testing procedure is described in CVB-APR-STP-0085.      contains all the frequencies in the audio spectrum with equal energy at each octave. Pink noise contains less energy at the higher audio frequencies than at the lower ones -See Figure 1.

3.3.
A stereo amplifier used to transmit and amplify the signal for the pink noise background.
3.4. The two loudspeakers positioned midway between the test speaker and listeners. Details of the exact positioning of the speaker equipment are further defined in Section 5.3.2.
3.5. Two Type 2 digital sound level meters (Sper Scientific, LTD, Model 840029 -See Figure 3.) with an "A" weighting decibel scale of 30 to 130 dB or equivalent. One sound level meter will be positioned in front of the Test Speaker and the second sound level meter will be positioned at head level beside listener L2-see Figure 1. Prior to beginning any testing, confirm that all measuring equipment being used has been calibrated in accordance with the testing laboratory's calibration procedure and schedule. All measuring equipment utilized for this testing must have been calibrated using a method traceable to recognized international standards when available.

4.2.
Administrator 1 will monitor the test speakers' speech and the actions of the listener group to make sure they are responding to the test speaker.

4.4.
The eight test subjects will be divided into two groups, a listener group comprised of three subjects (at least one female and one male) and a speaker group comprised of five subjects (at least one female and one male).

4.5.
All subjects shall be fluent in English. In addition, these subjects shall have no obvious or strong regional or foreign accents.

4.6.
The eight test subjects shall be trained in the donning and use of the respirator per manufacturer's instructions by the test administrator.

4.7.
Each subject shall be sized and fitted for the respirator per manufacturer's instructions.
Where an individual is qualified to wear multiple sizes of the respirator, the subject shall select the respirator size that provides the most comfortable fit.

5.2.
Record the background noise of the room, assuring no external noise, at the center listener's head position.

5.3.
Training of the Test Speakers.
5.3.1. Without the listeners in the room, one subject at a time will be seated in the test speaker's position and given an MRT word list -See Appendix A. A sound level meter will be positioned in front of the test speaker to assess the volume of their voice as the word list is read. The sound level meter shall be set to display "A"weighted sound levels. 5.3.2. Test administrator 1 shall be seated beside the test speaker to monitor the reading of the word list. The administrator shall instruct the test speaker to read, without placing any unusual emphasis on any stimulus word, at a rate of approximately one phrase every 6 seconds, using the introductory phrase "The word is (list word)". In addition, the test speaker shall be directed to communicate each word without using any visual gestures, such as hand signals, and without repeating any of the list words.

U n c o n t r o l l e d C o p y F o r I n s p e c t i o n a n d U s e U n c o n t r o l l e d C o p y F o r I n s p e c t i o n a n d U s e U n c o n t r o l l e d C o p y F o r I n s p e c t i o n a n d U s e U n c o n t r o l l e d C o p y F o r I n s p e c t i o n a n d U s e
5.3.3. Test administrator 2 shall be seated in the center test listener's position and will instruct the test speaker to begin reading the word list at a voice level of 75 to 85 dBA.

5.3.4.
During the reading of the word list, test administrator 2 will monitor the "A"weighted output volume of the speaker and provide feedback as to the loudness of the speaker's performance throughout the reading of the word list. Test administrator 1 will provide feedback to the test speaker regarding the pronunciation of the list words and the rate of performance. Additional training will be done if warranted per the judgment of the test administrators. These procedures will be completed without the use of a respirator and without any background noise above ambient conditions. 5.3.5. Additionally, training with background noise of 60 ± 2 dBA (consisting of pink noise), will also be completed according to the above procedures.
5.3.6. This will be repeated for all subjects in the speaker group.

5.4.
Training of the Test Listeners.
5.4.1. The three test listeners are seated facing a single test speaker at a distance of 10 ft. The listeners are seated next to one another with approximately one foot between them. Each listener will be given a laptop with a list of multiple-choice words for recording his or her responses.

An
MRT word list will be provided to a trained test speaker. The test speaker will be instructed to communicate the word list to the test listeners as was done during speaker training. 5.4.3. As a group, listeners will be instructed to listen attentively as the test speaker reads 50 words to them each with the introductory phrase "The word is (list word)." The test listeners shall be directed to select, or make a best guess of the word, that was perceived to be spoken from the six possible response words provided to them on the laptops. Test listeners are instructed to provide a "thumbs-up" hand signal to the speaker as a cue to say the next phrase.
5.4.4. During the reading of the word list, administrator 2 will monitor the A-weighted output volume of the speaker at the listening position and provide feedback as to the loudness throughout the reading of the word list. In addition, test administrator 1 will provide the test speaker feedback regarding the rate of performance. Administrator 1 will also monitor the actions of the listener group and provide additional instructions as needed. Test administrators will determine if additional training is required. These procedures will be completed without the use of a respirator and without any added background noise (pink noise) above ambient conditions. 5.4.5. Additionally, training with added background noise of 60 ± 2 dBA (consisting of pink noise), will also be conducted according to the procedures above.

GENERAL
This procedure describes the Determination of Air Flow For Powered Air-Purifying Respirators test in sufficient detail that a person in the appropriate technical field can conduct the test and determine whether or not the product passes the test.
The list of necessary test equipment and materials is as follows: 3.1.1. Air tight chamber -approximately 24 inches by 24 inches by 16 inches with a hinged door, a 3 inch diameter inlet for accepting breathing tubes and adapters, a one inch diameter outlet, and a 1/4 inch outlet for a manometer probe.  Prior to beginning any testing, all measuring equipment to be used must have been calibrated in accordance with the manufacturer's calibration procedure and schedule. At a minimum, all measuring equipment utilized for this testing must have been calibrated within the preceding 12 months using a method traceable to the National Institute of Standards and Technology (NIST).

4.2.
Normal laboratory safety practices must be observed. This includes safety precautions described in the current ALOSH Facility Laboratory Safety Manual.
4.2.1. Safety glasses, lab coats, and hard-toe shoes must be worn at all times.
4.2.2. Work benches must be maintained free of clutter and non-essential test equipment.
4.2.3. When handling any glass laboratory equipment, lab technicians and personnel must wear special gloves which protect against lacerations or punctures.

PROCEDURE
Note: Reference Section 3 for equipment, model numbers and manufacturers. For calibration purposes use those described in the manufacturer's operation and maintenance manuals.

5.1.
Set up the equipment as show in Figure 1.

5.2.
Connect the respirator to the chamber. On units with breathing tubes, the blower is placed outside of the chamber and the breathing tube is attached to an adapter identical to the manufacturer's connector on the facepiece, helmet, or hood. On units where the blower is in the helmet, an adapter is attached to the helmet and inserted through the inlet of the chamber and sealed.

5.3.
Close the door of the chamber.

5.4.
Check the electric manometer for zero, adjust to zero if necessary.

5.5.
Turn on the PAPR and the vacuum pump.

5.6.
Attach the tubing to the electric manometer and adjust to zero using the valve on the vacuum pump.

5.7.
Check all connections for leaks.

5.8.
If a leak is detected, reseal and readjust the vacuum.

5.9.
Time 1 minute on a stopwatch and count the number of CFM on the dry test meter. 1 revolution equals 10 CFM.  (a) Each respirator and respirator component shall when tested by the applicant and by the Institute, meet the applicable requirements set forth in subparts H through L of this part.
(c) In addition to the minimum requirements set forth in subparts H through L of this part, the Institute reserves the right to require, as a further condition of approval, any additional requirements deemed necessary to establish the quality, effectiveness, and safety of any respirator used as protection against hazardous atmospheres.
(d) Where it is determined after receipt of an application that additional requirements will be required for approval, the Institute will notify the applicant in writing of these additional requirements, and necessary examinations, inspections, or tests, stating generally the reasons for such requirements, examinations, inspections, or tests.
84.1157. Chemical cartridge respirators with particulate filters; performance requirements; general. Chemical cartridge respirators with particulate filters and the individual components of each such device shall, as appropriate, meet the following minimum requirements for performance and protection: (a) Breathing resistance test.
(1) Resistance to airflow will be measured in the facepiece, mouthpiece, hood, or helmet of a chemical cartridge respirator mounted on a test fixture with air flowing at a continuous rate of 85 liters per minute, both before and after each test conducted in accordance with paragraphs (d) through (  All videotapes and photographs of the actual test being performed, or of the tested equipment shall be maintained in the task file as part of the permanent record.
7.3. All equipment failing any portion of this test will be handled as follows: 7.3.1. If the failure occurs on a new certification application, or extension of approval application, send a test report to the RCT Leader and prepare the hardware for return to the manufacturer. 7.3.2. If the failure occurs on hardware examined under an Off-the-Shelf Audit the hardware will be examined by a technician and the RCT Leader for cause. All equipment failing any portion of this test may be sent to the manufacturer for examination and then returned to NIOSH. However, the hardware tested shall be held at the testing laboratory until authorized for release by the RCT Leader, or his designee, following the standard operating procedures outlined in Procedure for Scheduling, and Processing Post-Certification Product Audits, RB-SOP-0005-00.