A review of control methodologies for dynamic glazing

Abstract

With adaptive building façade technologies, a building envelope can provide a comfortable indoor environment under varying external conditions with minimal additional heating or cooling. The control strategy applied to the adaptation of the façade is a key determining factor in the successful integration of these technologies into a building. The building envelope plays a key role in regulating light, heat and mass transfer from the outdoor environment to the indoor. Dynamic glazing can be used to adjust the amount of solar radiation entering a building. The control strategies that ultimately determine the success of these switchable technologies to affect a building's energy performance and occupant comfort are reviewed in this paper.

Full text

Dublin Institute of Technology
ARROW@DIT
Conference papers School of Electrical and Electronic Engineering
2017
A Review of Control Methodologies for Dynamic
Glazing
Eoin D. McLean
Dublin Institute of Technology, eoin.mclean@dit.ie
Brian Norton
Dublin Institute of technology, president@dit.ie
Derek Kearney
Dublin Institute of Technology, derek.kearney@dit.ie
Philippe Lemarchand
Dublin Institute of Technology, philippe.lemarchand@dit.ie
Follow this and additional works at: h9ps://arrow.dit.ie/engscheleart
Part of the Engineering Science and Materials Commons
8is Conference Paper is brought to you for free and open access by the
School of Electrical and Electronic Engineering at ARROW@DIT. It has
been accepted for inclusion in Conference papers by an authorized
administrator of ARROW@DIT. For more information, please contact
yvonne.desmond@dit.ie, arrow.admin@dit.ie, brian.widdis@dit.ie.
8is work is licensed under a Creative Commons A9ribution-
Noncommercial-Share Alike 3.0 License
Recommended Citation
McLean, E., Norton, B., Kearney, D. & Lemarchand, P. (2017), A review of control methodologies for dynamic glazing. Advance
Building Skins 2017Berne, Switzerland, 2 -3 October.

1
A Review of Control Methodologies for Dynamic Glazing
Eoin McLean, Brian Norton, Derek Kearney, Phillipe Lemarchand
Dublin Energy Lab, Dublin Institute of Technology, Dublin, Ireland
Abstract
With adaptive building façade technologies, a building envelope can provide a comfortable indoor
environment under varying external conditions with minimal additional heating or cooling. The control
strategy applied to the adaptation of the façade is a key determining factor in the successful
integration of these technologies into a building. The building envelope plays a key role in regulating
light, heat and mass transfer from the outdoor environment to the indoor. Dynamic glazing can be
used to adjust the amount of solar radiation entering a building. The control strategies that ultimately
determine the success of these switchable technologies to affect a building’s energy performance and
occupant comfort are reviewed in this paper.
Introduction
Directive 2010/31/EU (EPBD recast) states that from January 2019, all new public buildings in the
European Union (EU) will have to be designed to Nearly Zero Energy Building (NZEB) standards and all
other new buildings will have to comply with NZEB from January 2021. An NZEB is a building that has
a very high energy performance with a very significant amount of its energy requirement met by
renewable sources [1]. The expectations for EU member states in “Zone 4 - Oceanic climates”, which
includes Ireland, is as follows:
 Offices: 40-55 kWh/(m2.y) of net primary energy with, typically, 85-100 kWh/(m2.y) of primary
energy use covered by 45 kWh/(m2.y) of on-site renewable sources;
 New single family house: 15-30 kWh/(m2.y) of net primary energy with, typically, 50-65
kWh/(m2.y) of primary energy use covered by 35 kWh/(m2.y) of on-site renewable sources;
[2]
Specific building design requirements will vary according to function, site, climate, façade orientation
and regulatory/code specifications. The choice of glazing has a significant impact on overall building
energy performance [3]. Buildings located in heating-dominated climates will want to maximise solar
gains, thereby reducing artificial heating requirements. Those located in cooling-dominated climates
will want to reduce cooling loads by minimising solar gains. These relatively simplistic requirements
are complicated by the need to consider occupant comfort, changing occupancy, diurnal changes in
weather and heat stored in the building fabric. Introducing glare as a design consideration can
significantly reduce energy efficiency in heating dominated climates [4]. Air quality requirement and
acoustic comfort must be fulfilled when considering the building and the operation of all mechanical
and electrical services in an holistic manner. Some of these considerations are measurable; such as
daylight illuminance and temperature, however the quality of the view of the outside environment is
more difficult to quantify. Optimising windows for visual comfort can lead to high energy
consumption, whereas windows optimised for energy efficiency do not always meet general visual
acceptance criteria [5]. These multiple design requirements make choosing an appropriate control
strategy for window systems that can change their thermal and optical characteristics even more
challenging [4]. Liu et al. [6], found that the use of dynamic glazing enables an increase in Glazing Ratio
(GR), without compromising building performance; even with a GR of 100%, the dynamic glazing
outperformed a static façade with a GR of 20%.

2
As it is difficult to optimise windows for visual or thermal comfort and at the same time minimise
building energy use [5], a compromise is required when choosing the size, type and location of the
glazing. A significant difficulty in determining the best control strategy is the need for an adaptive
façade to address multiple conflicting performance requirements often across differing physical
domains such as visual comfort, thermal comfort and energy efficiency [4]. A simple example is the
conflict between minimising glare risk while maximising solar gains. Control strategies have attempted
to optimise visual and thermal comfort while simultaneously achieving low energy consumption
targets. [7]. To truly maximise the use of dynamic glazing, controlled external shading may be
required [5], [6],[8]. The need to control such shading adds a further level of complexity to any
potential control strategy.
Types of Control Strategies
Self-triggered passive/dynamic glazings include Thermochromic (TC), Thermotropic (TT) and
Photochromic (PC). Glazings that can be triggered by an external stimulus are categorised as
active/intelligent. They include Electrochromic (EC), Suspended Particle Devices (SPD) and Liquid
Crystal Devices (LCD) [4]. Their characteristics are included in Table 1.
Table 1: Characteristics of Dynamic Glazing
Power
Supply
Voltage
Req.
Power Req.
Change
Time
No. of
States
VLT
VLT
SHG
C
SHG
C
U-
Value
W/m2/
K-1
Type
Max
Min
Max
Min
EC
5VDC
To Switch
0.5W/m2
3-7 min
5
.409
.006
.309
.108
1.834
SPD
35-100
VAC
Constant
3.5-15.5
W/m2
1-3s
∞
.650
.040
.570
.050
-
PDL
C
75 VAC
Constant
<10 W/m2
0.1s
2
.800
.620
-
-
-
TC
N/A
N/A
N/A
20-30
mins
5
.493
.094
.337
.196
2.666
TT
N/A
N/A
N/A
-
2
.690
.350
-
-
5.740
PC
N/A
N/A
N/A
2 min
2
.640
.230
-
-
-
Advanced control strategies can lead to significant improvements in building energy performance
without compromising visual comfort [7],[9]. Current control strategies for glazing’s are either (i) Rule-
Based Control (RBC), (ii) Model-Predictive Control (MPC) also referred to as Receding-Horizon Control
(RHC), or (iii) Genetic Algorithms (GA). The majority of studies have examined a relatively simple rule
based control [4],[9],[10]. These strategies are generally unable to optimise contrasting requirements,
such as the optimisation of solar gain contrasted with the desire to reduce summer cooling loads.
Notwithstanding this, they frequently outperform some of the more complex alternatives [4].
(i) Rule Based Control
A RBC strategy is defined by a set of rules that rely on measurements of the current or past states of
the building (i.e. lighting levels, temperature, building energy demand). It uses an external decision
making system of sensors, control algorithms and actuators [4]. A number of different RBC’s have
been tested for control of dynamic glazing [11]. RBC control strategies use one or more pre-
determined instructions acting on measured or pre-set data values [11]. Dussaullt et al. [9] used two
RBC control strategies in their study, RBC1 and RBC2. Both were designed to maximise daylight

3
without exceeding 500 lx. If this threshold was exceeded the glazing switched to the next darkest state
that would keep the daylight level below 500 lx. The difference between the two strategies was the
operation of the glazing during the hours when the building was said to be unoccupied. RBC1 switched
to its clearest state, thereby maximising solar gain and RBC2 switched to its darkest state, minimising
cooling requirement. Of the two strategies, RBC2 performed better even outperforming some of the
more complex GA and MPC strategies. It was noted, that the use of energy efficient artificial lighting
systems has a significant impact on the effectiveness of the control strategies. Using on–off switches
where the switch is triggered by the level of indoor illuminance or global solar radiation,
Assimakopoulos et al. [7] achieved similar results within ≈ 2%, with their RBC to those achieved using
a more complex fuzzy logic control. This study however, (i) did not compare results with a standard
glazing (ii) only relates to lighting, heating and cooling energy consumption and (iii) does not present
data on daylighting or glare comfort. In a simulation study, Fernandes et al. [8] used target indoor
illuminance and luminance levels as the design parameters for determining the performance of a split
pane EC window used in conjunction with automatic roller blinds to reduce lighting energy
consumption. Their control strategy used a least-squares algorithm with linear inequality constraints.
This work did not attempt to compare control strategies but rather utilised this particular method of
control to compare the performance of an EC window with a standard reference glazing. They found
that if the blinds are operated once per day at the first instance of visual discomfort, the annual
lighting energy consumption was reduced by 37% – 48%. Favoino et al. [4] found that although RBC
strategies could outperform more complex strategies for a single performance requirement, they
were generally unable to optimise multiple performance objectives. Importantly, this study did show
that a RBC strategy could outperform the best static glazing option. A simulation carried out by Tavares
et al. [12], used a simple control strategy based only on incident solar radiation applied to south, east
and west facades in a Mediterranean climate but did not consider the effect of glare on occupant
comfort. They concluded that this type of strategy resulted in energy savings compared to a standard
single or double glazing.
(ii) Model Predictive Control
Model Predictive Control algorithms use a defined and specific system model to predict the future
response of that system over a pre-determined time horizon [13]. The main premise is that there is
useful information contained in the future of that system which can be used to improve the system
control and performance [14]. Though first developed to control power plants and petroleum
refineries, their use is now widespread. At each time step, an MPC algorithm optimizes the sequence
of control values, over the prediction horizon based on the predictions of the model [9]. The control
predictions of the model are then applied to the model in real time. An MPC model has three distinct
parts, the observer, the optimizer and the predictor. Dussault et al. [9], used an MPC control strategy
with the objective of minimising the total energy consumption of the building. While the results of this
strategy were promising, it was still outperformed by the RBC2 and GA strategies. A possible reason
for this was that the simplicity of the building model did not allow the increased intelligence of the
MPC controller to be fully utilised. A study by Favoino et al [4], found that MPC control strategies have
a better energy performance than any of the reactive RBC strategies tested. This is because MPC
strategies are able to minimise total building energy use, while the RBC strategies can only minimise
total building loads. As the results of an MPC control strategy are only as good as the predictions of
the system it is essential to identify the optimal predictors for any given system.
(iii) Genetic Algorithms
GA can be used either to find a single set of input variables that will optimise one or many performance
requirements into a single solution or a set of optimal solutions that recognises the lack of any one

4
perfect solution [15]. A GA would be recognised as easy to use and robust but can be slow compared
to other optimisation methods. Due to being probabilistic, they can produce different results with the
same inputs [9]. Dussault et al. [9], used a GA with the objective of minimising overall energy
consumption, due to the computational expense and time associated with optimal GA solutions, a
quasi–optimal solution was used [9]. It was found that with a traditional T8 fluorescent lighting system,
the GA offered the lowest energy consumption of all control strategies, but with more efficient LED
lighting, the simple RBC controllers performed as well as the GA.
Discussion
The difference in performance between their best and worst performing strategies has been found to
be less than 10% [7] due to the small dynamic range of the Solar Heat Gain Coefficient (SHGC) (0.36 –
0.18, bleached and coloured respectively) for many EC windows. The multiple design constraints are
bounded by limits set by the desire for large glazed areas to maximise daylighting and solar gains to
reduce the need for artificial lighting/heating systems or smaller glazed areas to reduce cooling
demand caused by solar gains which increases the need for artificial lighting [5]. Occupant comfort
must be considered as part of any design or control strategy. Visual comfort plays such an important
role in overall occupant comfort, that it requires very thorough consideration[16]. In cooling-
dominated climates, the energy consumption of a building is very sensitive to the chosen control
strategy and reactive control has been shown to be as effective as predictive control for dynamic
glazing, whereas in heating-dominated climates, predictive control has yielded better results [4].
Conclusion
Simple control strategies work well on simple building models. Many authors have noted that the lack
of modelling complexity has reduced the performance benefits of intelligent control strategies such
as MPC and particularly GA. RBC strategies offer a simple means of control, that may yield building
energy savings but it is generally accepted that they are unable to meet more than a single
performance objective. There is enough research to suggest that current dynamic glazing alone may
not provide sufficient flexibility to produce the desired combination of energy savings and visual
comfort. A possible solution to this is presented in [17] through the use of a hybrid window using an
infrared Chiral Liquid Crystal (CLC) mirror and SPD window to independently control solar radiant heat
transmission, visible transmission and glare through the window. Studies combining the use of shading
with dynamic glazing have suggested that when considering the application of smart facades, it may
be necessary to consider the entire façade and not simply a single part.
Research conducted to date has used building simulations and virtual modelling environments. While
these studies can clearly demonstrate the ways in which dynamic glazing may be controlled, it is
necessary to conduct physical field trials and record the results of dynamic glazing being controlled in
a variety of climates and with a variety of control methodologies.

5
References
[1] F. Madonna and F. Ravasio, “Definition of nearly zero energy building and cost-optimal energy
performance in 2020.”
[2] “COMMISSION RECOMMENDATION (EU) 2016/ 1318 - of 29 July 2016 - on guidelines
for the promotion of nearly zero-energy buildings and best practices to ensure that, by
2020, all new buildings are nearly zero-energy buildings,” Comm. Recomm., 2016.
[3] Julijana Velevskaa, M. P. Gjorgjevichb, and Nace Stojanovc, “Electrochromic Nickel Oxide Thin
Films for Solar Light Modulation,” Int. J. Sci. Basic Appl. Res., vol. 31, no. 3, pp. 94–104, 2017.
[4] F. Favoino, F. Fiorito, A. Cannavale, G. Ranzi, and M. Overend, “Optimal control and
performance of photovoltachromic switchable glazing for building integration in temperate
climates,” Appl. Energy, 2016.
[5] C. E. Ochoa, M. B. C. Aries, E. J. van Loenen, and J. L. M. Hensen, “Considerations on design
optimization criteria for windows providing low energy consumption and high visual comfort,”
Appl. Energy, vol. 95, pp. 238–245, 2012.
[6] M. Liu, K. B. Wittchen, and P. K. Heiselberg, “Control strategies for intelligent glazed façade and
their influence on energy and comfort performance of office buildings in Denmark,” Appl.
Energy, vol. 145, pp. 43–51, 2015.
[7] M. N. Assimakopoulos, A. Tsangrassoulis, M. Santamouris, and G. Guarracino, “Comparing the
energy performance of an electrochromic window under various control strategies,” Build.
Environ., vol. 42, no. 8, pp. 2829–2834, 2007.
[8] L. L. Fernandes, E. S. Lee, and G. Ward, “Lighting energy savings potential of split-pane
electrochromic windows controlled for daylighting with visual comfort,” Energy Build., vol. 61,
pp. 8–20, 2013.
[9] J. M. Dussault, M. Sourbron, and L. Gosselin, “Reduced energy consumption and enhanced
comfort with smart windows: Comparison between quasi-optimal, predictive and rule-based
control strategies,” Energy Build., vol. 127, pp. 680–691, 2016.
[10] F. Gugliermetti and F. Bisegna, “Visual and energy management of electrochromic windows in
Mediterranean climate,” Build. Environ., vol. 38, no. 3, pp. 479–492, Mar. 2003.
[11] E. S. Lee et al., “Advancement of Electrochromic Windows,” Lawrence Berkeley Natl. Lab.,
2006.
[12] P. Tavares, H. Bernardo, A. Gaspar, and A. Martins, “Control criteria of electrochromic glasses
for energy savings in mediterranean buildings refurbishment,” Sol. Energy, vol. 134, 2016.
[13] S. J. Qin and T. A. Badgwell, “A survey of industrial model predictive control technology,”
Control Eng. Pract., vol. 11, no. 7, pp. 733–764, 2003.
[14] H. B. Gunay, J. Bursill, B. Huchuk, W. O’Brien, and I. Beausoleil-Morrison, “Shortest-prediction-
horizon model-based predictive control for individual offices,” Build. Environ., vol. 82, pp. 408–
419, 2014.
[15] L. Gosselin, M. Tye-Gingras, and F. Mathieu-Potvin, “Review of utilization of genetic algorithms
in heat transfer problems,” Int. J. Heat Mass Transf., vol. 52, no. 9, pp. 2169–2188, 2009.
[16] Y. Al horr, M. Arif, M. Katafygiotou, A. Mazroei, A. Kaushik, and E. Elsarrag, “Impact of indoor
environmental quality on occupant well-being and comfort: A review of the literature,” Int. J.
Sustain. Built Environ., vol. 5, no. 1, pp. 1–11, 2016.

6
[17] P. Lemarchand, J. Doran, and B. Norton, “Smart Switchable Technologies for Glazing and
Photovoltaic Applications,” Energy Procedia, vol. 57, pp. 1878–1887, 2014.