Predicting the soiling of modern glass in urban environments: A new physically-based model

Abstract

This study revisits the measurements of the MULTI-ASSESS and Long Term Soiling programs for understanding physically, and modeling, the processes controlling the soiling of modern glass in polluted conditions. The results show a strong correlation between the size distribution of particles and the evolution of the mass deposited at the surface of the glass. Over observation periods covering more than 2 years, the mass deposition on glass panels sheltered from the rain is observed to accelerate regularly with time at the sites closest to the sources of particulate matter (Roadside sites). At these sites the deposit is also richer in coarse (supermicron) mineral particles than at more distant (Urban Background and Suburban) sites, where the contribution of submicron particles (among which a significant fraction of particulate organic matter) is larger. This size and compositional segregation probably explains that the mass accumulation tends to slow down with time and finally saturate after an estimated duration of more than 10 years at the Suburban sites.The analysis of the correlation between the measured accumulated mass and haze shows that the haze-creating mass efficiency of the deposit decreases progressively as the density of particles increases on the glass panels. This is interpreted as being a consequence of the increasing influence of multiple scattering. A steady-state is eventually obtained when layers of closely packed particles are formed, which occurs for surface masses of the order of a few tens of μg cm−2. After this stage is reached, the haze increases linearly with further mass deposition at a pace conditioned by the size-distribution of the deposit. The parameterization of the evolution of the deposited mass with time, and of the correlation linking this mass to the haze allows proposing a new physically-based model able to predict the development of the haze on sheltered glass. Finally, a comparison of the model predictions with the independent measurements performed at the experimental sites of the AERO program shows that the model is able to simulate correctly the development of the haze at a variety of urban sites ranging from the Suburban to Roadside categories. This predictive tool should help developing conservation strategies adapted to the real environmental conditions of the historical and modern buildings.