Definition of candidate buildings

A candidate building is a theoretical construction with specific properties. The behavior of this building will be studied to examine the effect of varying those properties. After defining the basic candidate, the first step is to decide which ones are going to be those design variables.

The basic candidate building will be the reference building defined before. This build-ing has been designed takbuild-ing into account common advice for passive housbuild-ing. As a result, it tries to maximize the solar gains during winter. Therefore, it has a rectangular shape with its longer façade facing south. Moreover, the south façade implements most of the house glazing. Finally, shading devices have been installed in these southern windows to prevent excessive gains during summer. Due to the limitations of DBES model, other passive techniques had to be neglected.

After this basic design, most decisive properties affecting the energy performance of a building are those related to its envelope and systems. Therefore, selected design varia-bles are related to thermal transmittance of the envelope, airtightness, efficiency of the heat recovery system and, finally, the used heating and cooling system. These variables are also the chosen by several studies in the field.

In the case of the thermal transmittance, the optimization study performed using DBES model in [101] suggests that the quality of all envelope components should evolve in the same direction. For example, there is no sense in improving the insulation in the floor but not in the façades or ceilings. On the contrary, U-values in ceiling, floor, fa-çades and windows must keep a similar level. For this reason, these variables have been grouped in one only design variable called “envelope package”. This strategy is also defended and adopted in [100] and [102]. As it is shown in Table 4.6 and Table 4.7, four packages are studied with rising U-values in their components.

Table 4.6. Envelope packages for the U-values of the candidate buildings in Spanish

Table 4.7. Envelope packages for the U-values of the candidate buildings in Finnish location.

It is worth to mention that, although windows are represented by their U-value, their solar factor and number of glazing were also modified when improving their quality.

Finally, each one of the packages was created as a Matlab variable containing the set of elements.

The next two design variables are airtightness and heat recovery efficiency. In this case, the same values for both variables will be studied in Helsinki and Madrid, as shown in Table 4.8.

Table 4.8. Airtightness and heat recovery efficiency values for candidate buildings in Helsinki and Madrid.

Infiltration q50 (m³/hm²) 4 | 2 | 1 | 0.5 Heat recovery efficiency 45 % | 65 % | 75 %

4. Research approach and methods 73 The values of the design variables were selected with a starting point on those of the reference building. The infiltration rate is an exception. The value 4 m³/hm² was added, after analyzing some results, to properly check the tendency of the airtightness effect over the cost and energy consumption.

One of the few variables left is the one related to heating, cooling and energy generation systems. As explained before, three different systems are considered. Moreover, each of those three systems will be studied with and without solar collectors. In the case of us-ing solar thermal energy, two parameters of these systems have been optimized: water storage volume and solar collector area. The optimization was done through several DBES simulations aiming to maximize the obtained solar thermal energy. However, it must be cared not to produce more heat than necessary during summertime and not to have excessive temperatures in the storage during the year. As a result, storage volume is 300 liters and collector area 2 m² for buildings in Madrid. In Finnish buildings, opti-mal parameters depend on the heating system. For ground source heat pump systems and district heating, values are 400 liters and 5 m², while for air-to-air heat pump sys-tems they are 400 liters and 4 m².

It is worth to mention that the power capacity of heat pumps has been set near to the maximum load of the building. As a result, the air to air heat pump for Spain has 3.5 kW capacity and for Finland 3.2 kW. Ground source heat pumps have in both cases 6 kW capacity due to the limitations of the market. In addition, GSHP also heat domes-tic water so their capacity must be slightly higher.

Finally, the design variable for photovoltaic systems is the area of the panels. This value has been considered in order to achieve a specific performance level on candidate build-ings. It is worth to mention, that the photovoltaic simulation module in DBES model can be run separately from the rest of the building model. As a result, a considerable amount of time is saved. The function “pvchanger” is responsible for running this part of the model and apply the new PV results over already simulated buildings. In addi-tion, this function was developed to find the exact PV-panel area needed to make zero the annual energy balance or to reach a certain level of performance. This script can be found in Appendix A.

The default method for providing DBES program with inputs is through an Excel file.

This method can be convenient when analyzing only one building. However, in the case of analyzing more than one thousand buildings, it is unfeasible to modify one by one those Excel files. For this reason, a new input method has been developed for this study.

This new method consist on a simple modification in the main script of DBES. After this modification, DBES loads inputs from a selected Matlab file instead of an Excel file.

For creating these input Matlab files, a new function was built. This function, called

“buildingcreator”, generates multiple input files according to a specific range of values of design variables. As a result, by just entering the design data appearing on the tables above in this function, input files for the entire population of candidate buildings will be created. The complete code of the building creator function can be found in Appendix A. Lastly, all this input files will be handled to another function in order to simulate each of the buildings, as it will be shown in the next subchapter.