Table 1 Design principles

Granularity

It consists of both the quality of being granular, i.e., in the modeling field as the ability of “resolving details in time and space” (Pfenninger et al. 2014), but also the state of being composed of many individual parts or elements. In other words, to obtain results relevant to the purpose of the analysis, the model must ensure that the output meets the required granularity of detail (Lopion et al. 2018), and must be “technically explicit i.e. the different specificities of the simulated elements (equipments, buildings, etc.) must specifically impact the load curve calculations and results” (Grandjean et al. 2012)[p.6541].

Scalability

It explicits and extends the term “aggregative”, proposed by Grandjean et al. (2012) to describe model’s ability to generate results at different levels (household, city, region, etc.), emphasising on two implicit aspects: computational cost and data availability for parametrization. Indeed, as discussed in the field of computer science, a scalable model must have “the ability to handle increased workload by repeatedly applying a cost-effective strategy for extending a system’s capacity” (Weinstock and Goodenough 2006)[p.1]. In addition, to ensure consistency across levels, the model must use observable values, e.g., consumer socio-demographics and appliance penetration rates, allowing the model to be sensitive to micro-level factors that are available at macro-level (Xu et al. 2020).

Modularity

It goes beyond the modeling domain and as described in Baldwin et al. (2014)[p.1383], consists of the decomposition of a framework (e.g., a model) into modules characterized by “the interdependence of decisions within modules; the independence of decisions between modules; and the hierarchical dependence of modules on components embodying standards and design rules”. In this case, “module” is intended according to the definition of McClelland et al. (1987) adopted by Baldwin and Clark (2000)[p.63]: “[...] a unit whose structural elements are powerfully connected among themselves and relatively weakly connected to elements in other units”.

Transparency

It requires that the documentation and communication of a model provide all the necessary information to allow the recipients to understand, reproduce and possibly validate its results (Morrison 2018; Grunwald et al. 2016; Huebner et al. 2021). As the concept of transparency relates to the requirements of the recipients, it is important to note that here the authors are referring to an expert-to-expert exchange.