Shape optimisation of residential mid-rise buildings for reduction of energy demand in temperate climate

Abstract

Shape has
long been an important parameter in improving the internal comfort of buildings
and reducing energy demand. This can be seen from historical vernacular
architectural typologies, like igloos, which have minimal thermal loss surface
to provide a comfortable internal climate.  Using a shape factor to reduce the building
envelope and to minimise thermal loss is incorporated into the Dutch Building
codes for a long time, aimed at a comfortable climate and low energy demand.
This research is focussed on optimising building shape to reduce energy demand
but combining this with demands for thermal comfort and daylight entrance.  By making use of Grasshopper a parametric
design model is created. Using this model, a large variety of building designs
was generated which are analysed on their daylight entrance and energy demand
using Honeybee and Ladybug. By analysing the outcomes of these performance
analyses, the window-to-wall-ratio and shape, quantified by shape factor Lc, of
these designs were optimised using the autonomous optimisation algorithm pilOPT
in modeFRONTIER. The optimisation objective is to minimise the total energy
demand for heating and cooling. This is assessed by calculating the normalised
energy demand for heating and cooling for both a summer and winter period. To
execute the optimisation, the Erasmus Campus Student Housing project by Mecanoo
in Rotterdam was used as a reference project. 
The optimisation results show more compact buildings, with low WWRs have
lower energy demands. This can also be seen from the Pearson correlation
between the Lc [m] and the normalised energy demand [kWh/m2]. Which is found to
be -0.624 for the first optimisation and -0.632 for the second optimisation.
This confirms current building practice in which the relative building envelope
is tried is be reduced.  Low WWRs (<
0.2) can obtain a minimum daylight factor (DF50%) of 2.1%, which is lower than
current practice. However, the minimum WWR is largely affected by the presence
of neighbouring buildings. For facade orientations with adjacent buildings,
minimum WWR is significantly higher (up to 0.8). Since buildings outside the
own plot are not considered in the daylight calculation according NEN 2057 and
the new NEN-EN 17037 situations may occur where buildings will obtain legal
requirements but acquire poor daylight entrance. For future building shape
optimisation studies, it is recommended to make use of visible part of the sky
analysis (VSF) in the assessment of daylight criteria. Using VSF can save
minutes of computation time per building analysis thereby speeding up the
optimisation process. Recommendations for further research following this
thesis are to: enlarge the number of case studies; focus on more detailed WWRs in
a smaller range, by allowing smaller steps for WWR parameters, more detailed
optima may be found; include the effect of installation efficiency, as this
will affect the primary energy demand; assessing the effect of on-site energy
production such as PV-panels on the building shape and WWR, as increasing the
building envelope might not increase the net energy demand of buildings if more
energy can be produced.