Presented at the 4th World Conference on Photovoltaic Energy Conversion, WCPEC-4, June 12 2018. We present a methodology that utilizes EL images to predict power loss due to cell cracking. We explored pixel intensity histogram normalization methods to generalize this approach to a wide range of measurement conditions. The final optimized EL metric exhibits a strong correlation with power loss for a range of module technologies where a 1% increase in the dark area due to cracks results in a 3% loss in performance. This approach enables field EL mapping to translate to module Pmax mapping across a system.