Interdisciplinary Laboratory of Performance-Integrated Design (LIPID)

It is widely agreed that the climate ischanging and has a disastrous e ect onthe diversity of animals and plants, theincrease of natural disasters and thehigher damages they are causing, andso on. Solutions are beeing searched to ght this change, as we saw it recentlyin Paris with the 2015 United NationsClimate Change Conference. One of thesolutions proposed is to use renewableenergies, as for example solar panels.That is why it is so important to knowthe potential of this kind of energy, es-pecially in an urban environment. In-deed the roofs and fa˘cades of the build-ings o er big unused surfaces and thiswould be a good way to use them. An-other advantage is that there is no ter-rain to buy as it would be for a solarfarm, what means less costs.Figure 1: Description of the LODs asexplained by Gesquiere et al. [3]Some studies already exist on the ef-fect of the level of detail (LOD) forexample about shadows as done byBiljecky et al. [5] or heat demand fore-casts studied by Strzalka et al. [6]. Theidea of this paper is to do a similarstudy, but on another important topic:the assessment of the photovoltaic po-tential (PV potential). One point willbe particularly observed: the e ects ofthe partial shading. Indeed, as somestudies show it (see for example Kha-ing et al. [4]), the partial shading hasan enormous inuence on the electricityproduction of a PV panel. That is whyit is so important to know how signi- cant the di erence of shaded areas isbetween the LODs.In the 3D-modelling universe, theLOD has a crucial role, as it is the caseto assess the PV potential in urban en-vironment using this kind of models.According to the CityGML standards,there are ve levels of detail (see Fig.1),going from the footprint (LOD0) to aprecise model with inside and outsidedetails (LOD4) as stairs and chimneys.For the subject we are studying,meaning the assessment of the PV po-tential, the inside details are totally su-peruous. That is why we are goingto consider the LOD3 model as a per-fect model. But an LOD3 model costsa lot more to do than an LOD1 orLOD2. This brings us to the main ques-tion we will try to answer in this pa-per: Is it useful to do an LOD3 model,or are LOD1 and LOD2 models pre-cise enough to assess the PV potentialof a building, or even of a city? Thesecond part of the paper will handlethe theme of the shading e ects. In-deed, the production of electricity of PVcells is a ected when a part of the cellis shaded. Therefore, what would bethe consequences on the PV potential ifall the zones unregularly lit by the sunwould be considered as unusable for PVcells? This is the second question to beanswered in this paper.