Development of a parametric PET assessment model for the early design stage and evaluation of building-related heat mitigation measures for extreme heat events in the Netherlands

Abstract

In 2018, the ‘Deltaplan Ruimtelijke Adaptatie’, has been set up by the National Institute of Health and Environment (RIVM) for the purpose of facilitating the transformation to a climate adaptive living environment. As part of the Deltaplan, municipalities have been asked to perform a heat risk assessment through standardised outdoor thermal comfort maps, formulated in terms of the Physiological Equivalent Temperature (PET). The heat risk assessment has increased awareness of urban heat related issues amongst municipalities, and sparked interest in heat-proof (re)development of urban areas. However, appropriate tools for outdoor thermal comfort assessment in the early stage of urban design are currently not available, as established PET simulation tools come at large computational cost. In response to the absence of appropriate outdoor thermal comfort design tools, this thesis proposes a Grasshopper-based PET simulation tool with an adequate balance between time-efficiency and sufficient accuracy for the early design stage. Additionally, a study into the heat mitigation efficiency of varying Height-to-Width ratio (H/W ratio), street orientation and facade albedo for extreme heat events in the Netherlands provides global rules of thumb for heat-proof urban design.

Through literature review, conditions for appropriate determination of four meteorological input parameters for PET calculation (urban air temperature, mean radiant temperature, urban relative humidity and urban wind speed) have been determined, which have been used to construct the PET simulation model. The PET model has been validated to be sufficiently accurate through both literature and sense-checks and shows a considerable improved time-efficiency in comparison with established simulation tools such as ENVI-met. The model is thus considered suitable for application in the early design stage. Application of the model is limited to (1) cities in Western-Europe, (2) situations of low wind speed and (3) the months of April to September.

Because of its rather quick computation time, the Grasshopper PET simulation model has been used to formulate basic rules-of-thumb for heat proof design through a study into the effects of varying H/W ratio, street orientation and facade albedo. The study has been performed for a representative urban canyon in the Netherlands for an analysis period from 12.00 – 18.00 on an above average warm summer day. Study results show decreased spatially and temporally averaged PET in the urban canyon for increasing H/W ratio. Considering street orientation, highest average PET occurs for streets oriented towards the South-East (SE) and lowest average PET occurs for streets oriented towards the North-East (NE). Varying H/W appears to be the most effective strategy for heat mitigation with a heat mitigation potential of up to 5.6 \degree C, closely followed by varying street orientation with a heat mitigation potential of up to 4.7 \degree C. Default settings for street orientation affect the effectiveness of varying H/W ratio and vice versa: Varying H/W ratio is considered most effective for SE street orientations (heat mitigation potential of up to 5.6 \degree C) and least effective for NE street orientations (heat mitigation potential of up to 3.9 \degree C). Varying street orientation is considered most effective for larger H/W ratios (heat mitigation potential of up to 4.7 \degree C for H/W ratio 1.0) and least effective for smaller H/W ratios (heat mitigation potential of up to 3.0 \degree C for H/W ratio 0.5). From the study results, no firm conclusions can be drawn with regards to the effectiveness of varying between low albedo facades (albedo = 0.3, untreated facades) and high albedo facades (albedo = 0.8, white-painted facades): The results appear to be highly dependent on the number of ambient bounces (ab) of reflected shortwave radiation considered in mean radiant temperature calculation. Depending on the considered number of bounces, either low albedo facades (ab = 2) or high albedo facades (ab = 4) result in lower average PET in the urban canyon. At the time of writing, it is uncertain which number of ambient bounces should be considered realistic for calculation. For both considered number of ambient bounces, however, study results show that the heat mitigation potential of facade albedo is significantly lower than that of H/W ratio and street orientation.

For the considered urban canyon, a combination of H/W ratio 1.0 and SE street orientation results in the lowest average PET (approximately 38 \degree C), whereas a combination of H/W ratio 0.5 and NE street orientation results in highest average PET (approximately 47 \degree C). Dependent on the number of ambient bounces considered, either low- or high albedo facades result in highest average PET. However, the contribution of façade albedo on the mentioned PET values is limited (up to $\pm$ 1 \degree C). An exploration into the effects of ground- and façade material on the obtained average PET results suggests that varying ground- and façade material moderately affects average PET results. Further research is needed to quantify the exact effect of varying ground- and facade material on PET.

For future research, it is additionally recommended to perform a more elaborated validation with field measurements. The focus of a validation study should be on calculation of mean radiant temperature, as the calculation method implemented in the PET model is currently a draft version. Other interesting topics for future research are the implementation of vegetation in the PET model, and improvement of the wind speed calculation for more accurate wind speed modelling.