Improved Heat-Stress Algorithm

Dryden Flight Research Center, Edwards, California

Wednesday, August 01 2007

Corrections for solar radiation and wind increase the accuracy of determining dangerous outdoor work environments.

NASA Dryden presents an improved and automated site-specific algorithm for heat-stress approximation using standard atmospheric measurements routinely obtained from the Edwards Air Force Base weather detachment. Heat stress, which is the net heat load a worker may be exposed to, is officially measured using a thermal-environment monitoring system to calculate the wet-bulb globe temperature (WBGT). This instrument uses three independent thermometers to measure wet-bulb, dry-bulb, and the black-globe temperatures. Two reasons for this project were limited access to the Dryden monitoring system and delays of the required manual issuances for heat-stress warnings.

Existing algorithms to estimate the WBGT are less applicable in dissimilar environments. The American College of Sports Medicine (ACSM) was the algorithm selected after being used extensively and successfully during the 2000 Summer Olympics in Sydney, Australia. The ACSM algorithm only uses dry ambient temperature and vapor pressure to estimate the WBGT and it assumes moderate solar radiation and low wind speeds. However, when applied to extreme environments such as the Southwest desert region, this algorithm produces less than acceptable WBGT values. This situation increases the likelihood of overlooking unsafe conditions for employees working outdoors.

Dryden provided corrections to the environmental assumptions of moderate solar radiation and low wind built into the algorithm. Desert environments, such as the high Mojave in Southern California, have both high solar radiation (1,000 W/m²) and periods of significant winds during the summer afternoons in excess of 20 kn. When comparing the difference between the thermal environment monitor WBGT (actual) and the ACSM approximated WBGT against wind speed, it was found that a direct relationship existed with a positive slope (i.e., a large difference between the actual and estimated WBGT as the wind speed increases). A correction curve based on 40 days of comparisons was applied, and significant improvements to the WBGT approximations were observed. The daily difference between actual and calculated WBGT was reduced from approximately 2.0° to about 0.5° on the average.