ENERGY PERFORMANCE OF PV MODULES AS ADAPTIVE BUILDING SHADING SYSTEMS

P. Jayathissa∗, J. Schmidli, J. Hofer, A. Schlueter

Architecture and Building Systems, Institute of Technology in Architecture, Department of Architecture, ETH Zurich
John-von-Neumann Weg 9, Zurich, Switzerland
jayathissa@arch.ethz.ch

Shading systems improve building energy performance by controlling solar gains and natural lighting. Integrating
photovoltaics opens new opportunities for building integrated photovoltaics by combining the benefits of
adaptive shading with facade integrated solar tracking. Furthermore, it reduces the building energy demand and
simultaneously generates electricity onsite. This paper presents a methodology for simulating the photovoltaic
electricity production of a dynamic facade mounted PV system in combination with the energy consumption of a
building through shading. The simulation is conducted within the parametric Rhino / Grasshopper environment
using a high resolution radiation analysis for the calculation of PV electricity generation. Building energy
analysis is conducted through DIVA / EnergyPlus. From this simulation we can determine the optimum hourly
position and orientation of the PV panels, not only for optimal energy harvest, but also for the overall balance of
the room.
Keywords: Building Integrated PV (BIPV), Modelling, Tracking

1 INTRODUCTION

Buildings are at the heart of society and currently

account for 32% of global final energy consumption and
19% of energy related greenhouse gas emissions [1].
Nevertheless, the building sector has a 50-90% emission
reduction potential using existing technologies [1].
Within this strategy, building integrated photovoltaics
(BIPV) have the potential of providing a substantial
segment of a building’s energy needs [2]. Even the
photovoltaic (PV) industry has identified BIPV as one of
the four key factors for the future success of PV [3].
Dynamic building envelopes have gained interest in
recent years because they can save energy by controlling Figure 1: An example of an ASF constructed at the
direct and indirect radiation into the building, while still House of Natural Resources [11]
responding to the desires of the user [4]. This mediation
of solar insolation offers a reduction in heating / cooling
loads and an improvement of daylight distribution [5].
Interestingly, the mechanics that actuate dynamic 2 METHODOLOGY
envelopes couples seamlessly with the mechanics
required for facade integrated PV solar tracking. To study the electricity generation and building
Previous BIPV research analyses electricity energy consumption, a 3D geometry of the room and
production and building energy demand for static BIPV solar facade is built using the Rhinoceros software [12],
shading systems [6] [7] [8] [9]. This paper expands on and its parametric modelling plugin Grasshopper [13].
this work by analysing dynamic PV shading systems, The solar facade consists of 400mm CIGS square panels
while also taking into account mutual shading amongst that can rotate in two degrees of freedom. On the
modules and its effect on PV electricity generation. The horizontal axis, the panels can move from 0◦ (closed) to
approach allows us to reduce efficiency degradation due 90◦ (open) position in steps of 15◦. In the vertical axis, it
to partial shading of PV modules [10]. can move from 45◦ to -45◦ in 15◦ steps. Existing ASF
The work presented in this paper is applied in the systems [11] have independently actuated panels and a
context of the Adaptive Solar Facade (ASF) project [11]. continuous range of actuation, however for simplicity, we
The ASF is a lightweight PV shading system composed group all panels into one cluster that moves in unison.
of CIGS panels, that can be easily installed on any This leaves us with 49 possible dynamic configurations
surface of new or existing buildings. An example an ASF of the facade system. A corresponding workflow can be
can be seen in Figure 1. seen in Figure 2.
This paper will present a methodology of simulating
an ASF while simultaneously calculating the energy
demand of the office space behind the facade.
 Solar Facade Parameters

Building Parameters

Rhino/Grasshopper

Radiation Analysis Building Energy Analysis

LadyBug DIVA/EnergyPlus
Figure 3: A simulation result showing module insolation
Radiation from 11:00-12:00 on the 11 August for the used weather
Results file and a specific module orientation.
Electrical Circuit Simulation
Python
are coupled to an electrical circuit simulation of thin-film
PV modules with sub-cell level representation [18]. This
Building Energy model uses the standard equivalent circuit model to
PV Electricity
Demand
Supply calculate sub-cell I-V curves with a single diode, one
Post Processing series resistor, and one shunt resistor [19]. In addition to
Python
the irradiation dependency, the PV simulation includes
temperature dependency. The variation in efficiency due
to temperature is calculated through a linear dependence
Results of cell temperature to irradiance, where thermal imaging
Figure 2: Simulation Workflow was used to determine the correlation factor.

2.1 Building Energy Demand 2.3 Optimisation

The building energy simulation is conducted using Results of the PV Electricity Supply, and Building
EnergyPlus [14] through the DIVA [15] interface. A Energy Demand for each possible configuration of the
summary of the energy simulation parameters can be ASF is post processed in Python [20]. The optimum
found in Table 1. The geometric solar facade is configuration that minimises building energy
interpreted in EnergyPlus as an external shading system. consumption, and maximises PV electricity production is
A simulation of each possible dynamic configuration of calculated for each hourly timestep.
the facade is run for each hourly timestep of the year
using a weather file for Zurich, Switzerland.
3 RESULTS
Table I: Summary of main assumptions for the
calculation of operational emissions 3.1 Determinations of Optimum Configurations
The optimal configurations of the ASF can be
Building Settings visualised using carpet-plots. Figure 4 and 5 details the
Office Envelope Roof: Adiabatic optimum angles of the panels to maximise PV generation,
Floor: Adiabatic and minimise heating, cooling and lighting demands
Walls: Adiabatic independently. For the simplicity of visualisation we
Window: Double Glazed have split the representation of the data to altitude and
(e=0.2) 3mm/13mm air azimuth angles. We can see in Figure 4(a) how open
Thermal Set Points Heating 22°C configurations (light coloured) are chosen to minimise
Cooling: 26C° the building heating demands during the winter months
Building System Hydronic Heating: COP=4 and early mornings of spring and autumn. Likewise
Hydronic Cooling: COP=3 closed configurations (dark colours) are the preferred
Lighting Load: 11.8W/m2 solutions to minimise the cooling demand during the
Lighting Control: 300lx summer months.
Occupancy Office: Weekdays from 8:00 – The PV optimisation, best seen in Figure 5(d),
18:00 follows a path similar to a classic solar tracking model.
People Set Point: 0.1persons/m2 The variation from a solar tracking model is due to the
Ventilation: 0.0094m3/s/person effects of self shading. This disparity will be further
Local Assumptions analysed in Section 3.3.
Weather File Zurich-Kloten, Switzerland When the four optimisation cases are combined to
achieve the configurations for total energy minimisation
we get some interesting results seen in Figure 4(e). We
2.2 PV Electricity Supply can see that during the twilight hours of the morning and
A solar radiance simulation is run in parallel using evening the open positions are preferred to minimise the
Ladybug [16], which uses Radiance [17] to determine the use of artificial lighting. During the summer afternoons,
incident insolation on the solar facade. The approach closed positions to minimise cooling demands are
enables us to calculate solar irradiance on the modules preferred, however enough light is allowed to pass to
with high spatial resolution including the effect of sufficiently illuminate the space. When we also include
module mutual shading as seen in Figure 3. The results the PV electricty optimisation we notice a strong
 (open)

(closed)

Figure 4: Carpet plots detailing the optimal altitude angles to minimise the (a) heating demand, (b) cooling demand, (c)
lighting demand, and (d) maximise PV electricity production. Figure (e) details the combinations for optimum building
thermal management without PV production, (f) also includes the PV production. Small angles correspond to closed
positions, whereas large angles represent open positions. The corresponding azimuth angles for each hour can be seen in the
following Figure (5).

(SE)

(S)

(SW)

Figure 5: Carpet plots detailing the optimal azimuth angles to minimise the (a) heating demand, (b) cooling demand, (c)
lighting demand, and (d) maximise PV electricity production. Figure (e) details the combinations for optimum building
thermal management without PV production, (f) also includes the PV production. Negative angles correspond to the panels
facing west, whereas positive angles represent east-facing panels. The corresponding altitude angles for each hour can be
seen in the previous Figure (4).

how the combination of electricity generation and

tendency of the ASF to follow an optimal PV production adaptive shading can compensate for the energy
pattern. This, however changes if the building system consumption of the building during most sunlit hours.
becomes more inefficient. Less efficient heating, for Overall the PV electricity compensates for 41% of the
example, would result in configurations optimised for energy demand of the office behind the facade during the
heating overpowering those of PV electricity generation. course of the year.

3.2 Net Energy Consumption 3.3 Solar Tracking vs. Optimal PV Harvest
Figure 6 shows the net energy use at the optimal Our model which optimises the photovoltaic electricity
angles described in Section 3.1. Red colours detail the harvest was compared to an optimal solar tracking model.
energy consumption intensity, and blue colours detail the Figure 7(a) shows radiation incident on the PV panels for
PV electricity supply. It is interesting to see in Figure 6(f) the two models including the effects of self shading. The
Figure 6: Carpet plots detailing the net energy consumption. Each square represents the total energy consumption for that
specific hour of the entire month. Red colours detail the energy demand, while blue colours detail the energy supply.

system configuration for control, and delivers a final

energy performance assessment of the system. The
methodology can be applied to evaluate different PV
system geometries, actuation control strategies, building
systems, building typologies and climates.
Our utilisation of this model to evaluate the Adaptive
Solar Facade (AFS) details the advantages of an adaptive
system to a static system. The ASF is able to orientate
itself to the most energy efficient position, thus finding
the optimum balance between PV generation, and
daylight control to minimise heating, cooling and lighting
loads. The optimum orientation however, strongly
depends on the general efficiency of the building.
Decreasing the efficiency of the heating, cooling or
lighting systems will give higher preference for
Figure 7: Variation in energy production between configurations optimised for building thermal
optimal solar tracking, and optimal electricity harvest management through adaptive shading, than for PV
electricity production.
red line represents the solar tracking model without self When the model was run, purely to optimise
shading. Even though the radiation incident on the PV photovoltaic electricity production, we see that the
panels for both models are similar, the effect of self optimal angles do not follow a classic solar tracking
shading reduces the efficiency of the PV panels as seen in model. This is because an optimal solar tracking model
Figure 7(b). As a result, the optimum angles for dynamic results in high module self shading which reduces the
BIPV modules does not follow a solar tracking model, overall efficiency of the PV panels. This methodology,
unless the panel spacing is large enough to prevent self presented in this paper, can therefore also be applied to
shading. dynamic photovoltaic systems in general where self
shading is present.
This work ultimately presents a methodology for the
4 DISCUSSION AND CONCLUSION planning and optimisation of sophisticated adaptive BIPV
systems. Future work will use this methodology to
In this paper, we present a simulation methodology to determine the environments and building typologies that
evaluate a dynamic photovoltaic shading system, could benefit from adaptive BIPV systems.
combining both electricity generation, and the energy
demand of the building. This is achieved through the use
of DIVA/EnergyPlus for the building energy calculation,
LadyBug/Radiance for the radiation simulation, and
Python for the PV electrical simulation. A post
processing algorithm then determines the optimum
REFERENCES

[1] Fifth assessment report, mitigation of cliamte change,

Intergovernmental Panel on Climate Change (2014) 674–
738.
[2] P. Defaix, W. van Sark, E. Worrell, E. de Visser,
Technical potential for photovoltaics on buildings in the
eu-27, Solar Energy 86 (9) (2012) 2644–2653.
[3] M. Raugei, P. Frankl, Life cycle impacts and costs of
photovoltaic systems: current state of the art and future
outlooks, Energy 34 (3) (2009) 392–399.
[4] R. Loonen, M. Trˇcka, D. Co ́stola, J. Hensen,
Climate adaptive building shells: State-of-the-art and
future challenges, Renewable and Sustainable Energy
Reviews 25 (2013) 483–493.
[5] D. Rossi, Z. Nagy, A. Schlueter, Adaptive distributed
robotics for environmental performance, occupant
comfort and architectural expression, International
Journal of Architectural Computing 10 (3) (2012) 341–
360.
[6] M. Mandalaki, K. Zervas, T. Tsoutsos, A. Vazakas,
Assessment of fixed shading devices with integrated pv
for efficient energy use, Solar Energy 86 (9) (2012)
2561–2575.
[7] S.-H. Yoo, H. Manz, Available remodeling simulation
for a bipv as a shading device, Solar Energy Materials
and Solar Cells 95 (1) (2011) 394–397.
[8] F. Sara, et al, Maximizing the solar photovoltaic yield
in different building facade layouts, EU PVSEC (2015).
[9] P. Jayathissa, Z. Nagy, N. Offedu, A. Schlueter,
Numerical simulation of energy performance and
construction of the adaptive solar facade, Proceedings of
the Advanced Building Skins 2 (2015) 52–62.
[10] J. Hofer, A. Groenewolt, P. Jayathissa, Z. Nagy, A.
Schlueter, Parametric analysis and systems design of
dynamic photovoltaic shading modules, EU PVSEC 2015
conference, Hamburg, Germany (paper in review).
[11] Z. Nagy, S. Bratislav, P. Jayathissa, M. Begle, J.
Hofer, G. Lydon, A. Willmann, A. Schlueter, The
adaptive solar facade: From concept to prototypes, under
review.
[12] Rhinoceros v5 (2015).
URL https://www.rhino3d.com/
[13] Grasshopper - algorithmic modeling for rhino
(2015). URL http://www.grasshopper3d.com/
[14] D. B. Crawley, L. K. Lawrie, C. O. Pedersen, F. C.
Winkelmann, Energy plus: energy simulation program,
ASHRAE journal 42 (4) (2000) 49–56.
[15] Diva for rhino. URL http://diva4rhino.com/
[16] M. Roudsari, M. Pak, A. Smith, Ladybug: A
parametric environmental plugin for grasshopper to help
designers create an environmentally-conscious design.
[17] G. J. Ward, The radiance lighting simulation and
rendering system, in: Proceedings of the 21st annual
conference on Computer graphics and interactive
techniques, ACM, 1994, pp. 459–472.
[18] H. Johannes, A. Groenewolt, P. Jayathissa, Z. Nagy,
A. Schlueter, Parametric analysis and systems design of
dynamic photovoltaic shading modules, Energy Science
and Engineering.
[19] A. Mermoud, T. Lejeune, Performance assessment
of a simulation model for PV modules of any available
technology, Proceedings of the 25th European
Photovoltaic Solar Energy
Conference, WIP, Munich, 2010, iD: unige:38547.
[20] Python. URL https://www.python.org/