Since 2008, researchers from the Centers for Disease Control and Prevention/National Institute for Occupational Safety and Health (CDC/NIOSH) have been collaborating with Naval Facilities Engineering Service Center (NAVFAC ESC) engineers and Navy Medical Center San Diego (NMCSD) industrial hygienists to evaluate ventilation in a Navy aircraft painting hangar. The Navy seeks to keep worker exposures to air contaminants, including hexavalent chromium (CrVI), hexamethylene diisocyanate (HDI), methyl isobutyl ketone (MIBK), and others, at levels that meet regulatory health and safety standards, while limiting the environmental footprint, i.e. energy use, and operational costs of painting hangar ventilation. All project work, including the present study, refers to Naval Base Coronado, Building 465, Bay 6, in San Diego, California. In early 2008, computational fluid dynamic (CFD) simulations were performed to model the relationship between air velocity and worker exposure levels in a Navy aircraft painting hangar. A walk-through survey was conducted June 16-19, 2009, encompassing range-finding personal and area air sampling (for CrVI, HDI, and any other contaminants found on the material safety data sheets) and the gathering of hangar dimensions, geometric details, and ventilation boundary conditions that would be used to set-up the CFD simulations. Next, the ventilation system's ability to control air contaminants was evaluated through comprehensive personal and area air sampling of all solvent, primer, and topcoat constituents, on July 22 and August 3, 2009 and April 13, 2010. Three visits were needed in order to monitor three painting processes, which are typically spaced days or weeks apart. Accurate statistical characterization of exposures required sampling of three processes. CFD simulations were performed and validated based on the ventilation settings available at the time of the 2009-2010 field studies. An initial tracer gas study was conducted April 12 and 14, 2010 to evaluate the performance of the hangar ventilation system under a number of supply/exhaust ventilation settings. The results from the 2009 and 2010 air sampling, tracer gas and CFD simulation studies are available in a NIOSH report [NIOSH 2011], which indicated that: 1. Balancing the air supply and exhaust could improve exposure control, consistent with ventilation standard practice. 2. From tracer gas measurements, 3/4 of the normal supply and exhaust rates provided the lowest concentrations, when compared to full flow (supply = 136 fpm; exhaust = 99.0 fpm) and half-flow (supply = 73.4 fpm; exhaust = 49.0). 3/4-flow was a supply velocity of 102 fpm and an exhaust velocity of 68.9 fpm. However, the only statistically significant difference among ventilation settings was between 3/4-flow and half-flow, which had the lowest and highest concentrations, respectively. 3. CFD simulations showed a large increase in contaminant concentration at typical worker locations, when the supply rate exceeded the exhaust rate, compared to when the supply and exhaust rates were equal. "Balancing," as in item 1, means maintaining a very small negative pressure, perhaps approximately -0.05 in. water. 4. Based on personal sampling of workers during typical aircraft refinishing operations, the ventilation system did not adequately address worker exposure and required supplementing with respiratory protection, as was already being done. 5. Because all materials measured in the aircraft refinishing process were less than 1% of any LEL, explosion from chemical concentrations was not an issue. 6. Additional tracer gas and CFD simulations were needed to fill the following information gaps: a.Tracer gas studies were performed only on the system in the unbalanced state. Additional tracer studies are needed under balanced conditions. b. CFD simulations were performed under balanced ventilation boundary conditions and under a hypothetical positive pressure scenario, rather than the measured unbalanced boundary conditions. Additional CFD simulations are needed that use the measured supply and exhaust velocities. The additional work called for in item 6 is the subject of the current report. Thus, in March 2011, NIOSH researchers conducted another tracer gas evaluation of the Navy aircraft hangar, under four additional ventilation settings that each provided negative pressure conditions. There were a total of four supply air blowers and four exhaust air fans located on the roof that served supply and exhaust plenums on opposite walls of the hangar. Each ventilation setting corresponded to a supply and exhaust fan combination. For example, a setting of 3/4-supply and 4/4-exhaust indicates that three of the four supply fans were operating, while all four exhaust fans were operating. The four ventilation settings were as follows: Setting 1: 1/4 supply and 2/4 exhaust; Setting 2: 2/4 supply and 3/4 exhaust; Setting 3: 2/4 supply and 4/4 exhaust; Setting 4: 3/4 supply and 4/4 exhaust. Tracer gas experiments were conducted over two nights, while normal hangar operations continued during the daytime. Results from each night were reported separately because the source and measurement locations and exhaust filter pressure drop could not be held precisely constant between nights. On night one, only settings 1 and 4 were tested. Results from night one indicated that setting 1 had statistically significantly higher mean tracer gas concentrations than setting 4 (1742 vs. 249.7 ppb). On night two, tracer gas testing was conducted for settings 2, 3, and 4. Results from night two indicate that mean tracer gas concentrations were statistically significantly higher for setting 2 than for settings 3 and 4 (1526 vs. 353.7 and 1193 ppb, respectively). There were no statistically significant differences between mean tracer gas concentrations of settings 3 and 4. The studies occurred on two consecutive nights, because the process of setting up equipment, altering system configurations, repeating trials (with time between trials to reach a stationary condition), and taking down equipment (to make the bay ready for the next day's painting operation) took several hours, even for testing just two or three air velocities. Also, some system configurations required additional consult with the HVAC technicians, who were not available during the second shift. Care was taken to not make system changes that risked interference with normal operations, which would begin at 0600 hrs. While the source and measurement locations and settings were duplicated as closely as possible on the second night, some variability was expected. Thus, the data from each night was analyzed separately. Still, sufficient data was collected to make comparisons between the velocity at Setting 4 (3/4 supply and 4/4 exhaust) and the velocities at Settings 1, 2, and 3. Based on these additional tracer gas tests and CFD simulations, along with the results of the original study [NIOSH 2011], the following conclusions and recommendations can be made: Conclusions 1. The first round of tracer gas experiments (reported in NIOSH  and referred to in the current report as Tracer Experiments I) and the CFD simulations of those conditions both indicated that the 3/4-flow resulted in lower exposures than either the half- or full-flows. 2. The existing equipment that serves Bay 6 cannot deliver a flow that is balanced. It should deliver a flow where the supply rate and exhaust rate are nearly equal, with the exhaust rate slightly higher to maintain a small negative gauge pressure, for the purpose of containment. With only four supply fans and four exhaust fans, along with the VFD controller on the exhaust fans that seemed unresponsive to supply changes, the system could not be adjusted with enough precision to achieve a balanced state. In other words, while operating 4 supply fans and 4 exhaust fans resulted in a positive pressure imbalance, turning off one of the supply fans resulted in a negative pressure imbalance (too much exhaust). 3. Increasing the average air velocity in the hangar from 43.3 to 85.3 fpm lowered exposures (from 1742 to 249.7 ppb). Increasing the average velocity from 66.1 to 75.3 fpm lowered exposures (from 1526 to 353.7 ppb), while increasing the average velocity from 75.3 to 85.3 fpm increased exposures (from 353.7 to 1193 ppb). Recommendations 1. Achieving balanced flow (perhaps -0.05 in. water gauge, if prevention of fugitive emissions to the environment is desired) through capital improvements at the site should be considered, based on ventilation standard practice. 2. After balancing or any other system modifications, follow-up tracer gas testing, process air sampling, and velocity sampling should be done to verify ventilation improvements. 3. Correcting the pressure imbalance should include replacing appropriate exhaust filters, pre-filters, or pre-layers during moderate or high filter loading to reduce pressure drop and save energy. The filter pressure drop value at which filters will be replaced should be recommended by NAVFAC ESC and the filter manufacturer. Balancing the system and improving system maintenance will improve operational efficiency. 4. Measurements of the concentration of flammable or explosive materials in air should be made directly in the exhaust stream to demonstrate compliance with NFPA 33: "Standard for Spray Application Using Flammable or Combustible Materials 2011," if any significant changes are made to the existing ventilation system or settings. The current study did not include this specific measurement, because no flammable materials were used in the tracer studies and because previous area air sampling during aircraft painting under the existing ventilation indicated that an explosion hazard was not present. 5. In addition to correcting existing paint finishing hangar ventilation systems, innovative design should be explored using CFD. Reducing the hangar cross-sectional area to more closely fit each aircraft size and maintain a desired velocity at a lower flow rate, directing supply air to the work zones more precisely, and bringing exhaust terminals closer to contaminant sources are examples of possible paths to consider that may reduce worker exposures, while also reducing associated energy costs. 6. Any changes in ventilation operation should include provisions to prevent possible safety hazards (doors blowing open or closed) created by changes in hangar pressure.