Demand Simulation for Buildings and Districts

Energy demand

Consistent demand profiles are one of the most important prerequisites for planning district energy systems. If demand is over- or underestimated, systems end up being over- or undersized. Oversizing increases investment costs as well as costs during later operation. Undersizing, in turn, means that demand goes unmet and results in energy supply shortfalls. Relevant energy demands for districts include heat demand (space heating, domestic hot water, and where applicable process heat), electrical demand (plug loads, e-mobility), and cooling demand (space cooling, server room cooling, refrigeration for food retail, etc.).

Calculation methods

Although demand profiles are of great importance for planning, there is a lack of suitable calculation tools and methods for generating them. Existing profiles are often simply scaled. However, this fails to adequately account for specific district requirements or given peak loads. Detailed simulation methods (dynamic building energy simulation), on the other hand, allow buildings to be modeled in great detail and their annual demand to be estimated fairly accurately. Even here, though, uncertainties remain, since, for example, user behavior can only be accounted for roughly, on a statistical basis, and deviates from actual user behavior. A further issue with detailed simulation methods is that they require a very large number of input parameters to generate demand profiles with sufficient accuracy. For instance, information on shading systems, wall constructions, insulation thicknesses, window types, or occupancy density is needed to determine a building’s cooling or heating demand. However, this information is typically only available in later planning phases. In the early planning phase of districts, simplified methods are therefore often applied, for example based on the degree-day method.

Simulation using physical building models

Thermal building models are used to determine thermal energy demand (space heating and cooling). The simulation captures all heat flows and resistances within the building. This often requires a number of assumptions to be made, for example regarding transmission and ventilation heat losses, solar gains from solar radiation, building shading, and internal heat gains from occupants and technical systems. In particular, occupant behavior (occupancy times, ventilation habits) can often only be estimated roughly. In detailed building models, the building is divided into several thermal zones, and the heat flows between the individual zones are also taken into account. Deriving a building’s heating and cooling demand over the course of a year requires very detailed models, which include, for example, all of the building’s external surfaces along with their geographic orientation, as well as wall constructions (materials used and layer thicknesses, glazing ratio, and any existing thermal bridges).

Demand profiles

District planning typically uses demand profiles with hourly resolution for a typical reference year. For the simulation of demand profiles, in turn, a typical weather year is used (test reference year). For Germany, weather data can be downloaded free of charge from the DWD for many locations; for other locations, EnergyPlus weather years can be downloaded free of charge. Using profiles with hourly resolution is particularly important when renewable energy sources, especially photovoltaics, are being considered. For electrical simulations, a 15-minute resolution is also common. In district planning, the advantage is that a large number of buildings is often considered, and load peaks tend to average out. For a single detached house, for example, individual load peaks can occur (hair dryer, stove), lasting only a few minutes. When these load peaks are averaged over an hour, the resulting average power values are much lower. If photovoltaic electricity is to be used, an hourly view would result in much greater overlap than if demand and power generation were considered on a minute-by-minute or even second-by-second resolution. This illustrates that different calculation methods are needed for individual buildings than for district-scale solutions. In districts, as the number of buildings increases, load peaks average out more and more, resulting in smoothed demand profiles for which hourly resolution is generally sufficient (standard load profiles). In sluggish, high-inertia systems such as district heating networks, load peaks are — unlike in electrical grids — less critical, because they don’t need to be covered directly by generation units; thermal inertia smooths out load peaks in any case.