Simulating HVAC Defense against COVID-19 Infection

/in

by Craig Phillips, Commissioning Director at Bernhard

_______________

In fall of 2020, The University of Alabama at Birmingham Healthcare (UAB) was faced with an increasing influx of COVID-19 patients and the question of how to best maintain medical staff safety as a top priority was at the forefront of UAB’s concerns. Changes in operation of heating, ventilation and air conditioning (HVAC) systems’ configuration to increase the systems effectiveness against the spread of COVID-19 was given careful consideration. Specifically, three questions were asked of the team.  Does an air change rate of six provide the quickest and most orderly removal of particles from the patient room?  Does maintaining a room differential pressure relationship of -0.01 inches water column relative to the adjacent corridor achieve near 100% particle containment?  Does the addition of a HEPA air scrubber have a positive effect on the time to remove particles from the room and does the location of the HEPA air scrubber play a vital role in containment of the particles? In order to better answer these questions, UAB organized a small team of healthcare engineers, an environmental health and safety manager, and the support of the hospitals C-Level stakeholders. At the inception of the efforts, UAB had just opened a new bone marrow transplant unit which provided the team with the unique opportunity of having an empty bone marrow transplant unit to utilize as a test site before the unit was to be transformed into a dedicated COVID unit.  And so began the task of creating the framework for a field testing lab to serve as the epicenter of the experiments.

The approach to achieving valid answers was to utilize one of the bone marrow transplant patient rooms with a high-supply, low-return diffuser configuration (referred to as patient room “A”) as one of the two test sites. Specifically, the configuration consisted of two perforated supply diffusers with an integrated HEPA filter, one floor-mounted return grille, and a typical ceiling exhaust in the adjacent dedicated restroom. The second test configuration (referred to as patient room “B”) consisted of a modified diffuser layout with high-supply high-return layout. Specifically, the configuration consisted of two smaller perforated supply diffusers without an integrated HEPA filter, one ceiling mounted return grille, and a typical ceiling exhaust in the adjacent dedicated restroom. The furniture layout was kept consistent and simple with a patient bed, an adjustable over-the-bed table and a chair in the corner of the room.

The method used to create measurable results started with procuring two BS 5295 rated fog generating machines that utilize pharmaceutical grade fog fluid producing the same Atomic Energy Authority (AEA) certified 0.2-micron particle sized fog. The size of the particles being generated was important because the particles should be in the same range of typical aerosolized COVID-19 particles (0.06-1.4 micron).  A high output machine, capable of 6,356 CFM, was chosen to be used in a “saturated” room test where the amount of time to completely fill the room with fog was precisely calculated. The smaller machine, capable of 100 CFM, was chose to be used in testing meant to simulate the amount of aerosolized COVID-19 particles generated from a typical infected adult male patient. To measure the effectiveness of the HVAC system to remove the fog from the room, a light intensity meter and ionization smoke detector were utilized.  The time was recorded for the smoke to visually clear, for the smoke detector to clear and for the light level to return to the same level as measured at the start of each test. In addition, other important parameters were monitored and recorded for each test such as the room pressure, the use or lack of an air scrubber, the air scrubber CFM and equivalent air change rate (fixed at 6 ACH) and the supply and return airflow (measured by the Building Automation System).

The team faced many challenges and barriers from the onset to the end of the endeavor. The first obstacle was developing a means to measure the effectiveness of the tests. The team reviewed plausible methods and ultimately settled on utilizing the ionization smoke detector and light meter combo to gauge the effectiveness of the HVAC system. The second major obstacle was selecting a fog machine(s) suitable for the testing. The team identified the machines must be pharmaceutical grade and generate particles between 0.06- and 1.4 micron.  After much research and debate, the team decided on two machines, one which had to be shipped from the United Kingdom, which presented the looming concerns of international shipping delays, which were rampant at the time and still linger today. Another obstacle was the testing site had a non-typical HVAC distribution layout with HEPA filters in the supply diffusers and a low return grille. The team decided to modify the HVAC distribution layout to better match the typical patient room with two supply diffusers and one high return grille. Furthermore, the inlet conditions to one of the terminal units was modified to obtain accurate airflow control. Perhaps the biggest obstacle was time itself. The same team that was dedicated to the experimentation was also being tasked with converting numerous patient floors into “COVID wings” and the location of the testing site was a prime location to convert into a COVID wing. As soon as the testing was completed, the HVAC system was slated to be retro-commissioned and turned over for immediate use.

The results of the study showed that six air changes provided subjectively better results (27.5 Minutes) when compared to CDC Guidelines (46 minutes at 99% efficiency) for Environmental Infection Control in Health-Care Facilities (2003). Typical COVID patient particle generation testing shows that the infectious risk posed by a patient infected with COVID-19 is relatively low because of the low viral load typical in infected patients unless that viral load is increased by severe acute coughing. For this scenario, with a neutral room pressure, a patient who is infected and is coughing 10 times per minute poses the least risk to medical staff with an air change rate of 6 air changes per hour with a HEPA scrubber in the room close to the patient bed between the supply and return grilles in the room. The chance of medical staff inhaling the viral particles is decreased the further the distance from the patient. Increasing the air change rate decreases the time to remove the virus particles from the room, but with increased air-change rate turbulence is created which causes the airborne particles to be distributed around the room. This phenomenon at eight and ten air changes increases the probability of the virus infecting medical staff compared to the tests conducted at air change rates of six. No visual difference was witnessed, related to undesired spread of aerosolized smoke, between four and six air changes per hour. However, at four air changes per hour, the time it took for the aerosolized smoke to clear was much greater than at six air changes per hour.

The smoke saturation room testing answered three important questions:

  • The testing showed six air changes per hour provided comparable particle removal times compared to higher air changes rates of eight and ten air changes which are not feasible in most patient rooms due to existing HVAC infrastructure limitations.
  • Scrubbing the air in the room with a HEPA scrubber significantly improves the particulate removal times.
  • Orienting the scrubber so that the intake is at the door can better protect staff in the corridor areas by capturing the particles before they can exfiltrate the room.

Ideally, the scrubber would be pointed at a low return so the majority of particles are quickly removed from the room.  In addition, the testing proved that operating a room at negative 0.01 inches protects the staff in the corridor by not allowing the particles to escape the room. However, it is impractical to configure the majority of patient care spaces in such a manner due to the limitations of the existing HVAC infrastructure.  In such cases, where a patient exhibits severe coughing, it is advisable to configure a room to be at six supply air changes, ten return air changes and to install a HEPA scrubber at the door with the discharge directed at a return grille or restroom door with exhaust of ten air changes per hour.

Based on the data collected from the testing, UAB made strategic decisions to set all patient room air changes to six air changes and to include HEPA scrubbers in each COVID-19 patient room. At the height of the COVID pandemic, UAB Healthcare had 300 COVID patients admitted to the hospital. Despite the high COVID-19 patient census rates, UAB medical staff contraction rates were relatively low. Based on these promising results it is recommended healthcare institutions implement self-deployed air change compliancy verifications and when pandemic risks are great, implement additional filtration and air changes in the form of HEPA air scrubbers. If feasible, operate the patient room at a negative pressure equal to negative 0.01 in. w.c.

Craig Phillips has 16 years of industry experience and is the Director of Commissioning at Bernhard. He has extensive experience commissioning healthcare facilities and campus expansions around the country. He helped to build Bernhard’s commissioning services group, providing project management and commissioning services for projects including a 1-million-square-foot healthcare campus expansion, new data centers, commercial facility renovations, and more. Craig is a registered Professional Engineer, Certified Commissioning Authority, Certified Energy Manager, Healthcare Facility Design Professional, a NEBB Building Systems Commissioning Certified Professional, and LEED AP BD+C. Craig was the co-author of the US Military Health Care Commissioning Guidelines and co-author of Ascension Health’s Commissioning Guidelines. He holds a Bachelor of Science degree in Mechanical Engineering from Arkansas Tech University.