Nayab Bushra, Timo Hartmann
Design optimization method for roof-integrated
TSSCs
Open Access via institutional repository of Technische Universität Berlin
Document type
Preprint
Date of this version
11th March 2022
This version is available at
https://doi.org/10.14279/depositonce-15328
Citation details
Bushra, Nayab; Hartmann, Timo (2022). Design optimization method for roof-integrated TSSCs. Technische
Universität Berlin, Preprint, http://dx.doi.org/10.14279/depositonce-15328.
Terms of use
cb This work is licensed under a Creative Commons Attribution 4.0 International license:
https://creativecommons.org/licenses/by/4.0/
1
Design optimization method for roof-integrated TSSCs
1
Nayab Bushra a, *, Timo Hartmann a
2
a Civil Systems Engineering, Technical University of Berlin, Gustav-Meyer-Allee 25, 13355 Berlin, Germany 3
Email addresses: [email protected] (Nayab Bushra); timo.hartmann@tu-berlin.de (Timo Hartmann) 4
* Corresponding author: Email address: [email protected] 5
Abstract 6
This paper proposes a two-step design optimization method for roof-integrated two-stage solar 7
concentrators (TSSCs) as energy supply systems. The integration process of these systems with 8
buildings is complex as several conflicting and multi-disciplinary concerns need to be addressed. Thus, 9
the proposed approach is intended to be adaptable to informed decision-making processes in early 10
design stages, and yet to be collaborative where several key stakeholders are involved. The method 11
is an extension to our previously developed approach where the performance of roof-integrated TSSCs 12
in several design scenarios is accessed along with multiple performance indicators by developing a 13
parametric model and controlling a set of design inputs. In the current study, the proposed method is 14
combined with a multi-objective design optimization method, aiming to optimize, building and TSSCs 15
geometry. The method was validated in an illustrative case study of a single-family house (California) 16
for a number of conflicting objectives e.g., maximization of direct normal solar irradiance (DNI) and 17
annual average load match index (av.LMI), and minimization of covered roof area. The validation of 18
the method shows a number of interesting results. The method enables the generation of performance-19
driven designs and searches for the most appropriate solutions, that can help to meaningfully support 20
the decision-making process. 21
Keywords: two-stage solar concentrator; roof-integrated; parametric; multi-objective; design 22
optimization; decision-making. 23
24
25
26
27
2
1 Introduction 28
The push for a less carbon-intensive built environment has led to several questions about how 29
self-sufficient buildings should be designed. Sustainability-related issues can be addressed by working 30
on building envelopes to maximize solar gains [1,2,3], and integrating advanced solar technologies 31
[1,4,5,6,7]. In recent years, building-integrated photovoltaics (PVs) have gained significant interest in 32
building energy research [2,3,8,9,10,11,12,13]. However, PVs are still behind the solar concentrators 33
(using mirrors and a receiver to convert sunlight into usable energy) in many aspects. For instance, a 34
PV requires two times more space than a concentrator to produce the same amount of energy (~ 550 35
kW), where space can be a critical factor, especially in urban areas [14]. Further, PV efficiency is very 36
low (16–22%) [15] compared with concentrator efficiency (40%) [16]. Concentrators with high 37
concentration ratios have high efficiency and energy yield and need a small-sized receiver 38
[17,18,19,20,21]. Among several designs, two-stage solar concentrators (TSSCs) show 50% to 200% 39
[18,19,20] more concentration ratio and require lesser (i.e., 77%) solar cells [21] compared with 40
traditional concentrators. TSSCs are prominent for high energy yield, efficient power delivery, and 41
deployment modularity [22,23,24]. In TSSCs, light is reflected from a primary mirror to a secondary 42
mirror, which is focused on the receiver [22]. Despite growing interest in building-integrated 43
concentrators [25], there exists less research [24,26,27,28,29,30,31,32,33] on integrating TSSCs with 44
buildings. Further, the integration of TSSCs with buildings reflects a complex decision-making process, 45
involving stakeholders from different domains e.g., building architects, civil engineers, and energy 46
specialists having multiple and conflicting objectives e.g., energy demand vs. energy yield vs. energy 47
cost [2,34]. To address this, design optimization can help to find trade-offs and support the quick 48
decision-making process. This facilitates the setup of design parameters (decision variables) and 49
fitness functions (design objectives) for generating, evaluating, and optimizing multiple designs. Design 50
optimization can be achieved by applying optimal combinations of different design strategies and 51
ranking design options according to a set of objectives [35]. Nevertheless, design optimization for 52
building integrated TSSCs requires optimization at the building level and system design level. 53
On the design level, TSSCs have several limitations e.g., complex architecture, the requirement 54
of efficient trackers, poor performance in case of misaligned mirrors, and the requirement of high-55
3
manufacturing skills [22]. Some studies [26,27,29] optimized TSSCs for minimum size and cost and 56
maximum yield by considering decision variables e.g., the number of modules and arrangements, and 57
receiver properties. However, there is still more research needed to include several other decision 58
variables in the optimization process that include but are not limited to geometric concentration ratio 59
[24], mirrors’ size [31,33], the distance between mirrors [24,33], or the mirrors’ shape [30,36] to 60
generate optimal TSSC solutions. Additionally, the research on optimization of building-integrated 61
TSSCs is scant [26,27,29]. Unlike PVs, TSSCs have the leverage of design flexibility (e.g., by varying 62
mirror shapes, system dimensions, etc.) since the technology is still far from maturity [22]. Because 63
concentrators only work under direct normal irradiance (DNI) unlike PVs. Thus, successful integration 64
of TSSCs with buildings requires exploration of building surfaces that receive most of DNI, to ensure 65
optimal performance of these systems. However, in existing research [26,27,29], building-integrated 66
TSSCs are optimized as stand-alone designs before installation on buildings and are limited to existing 67
buildings. Thus, building-related parameters are ignored for optimization of TSSCs while buildings 68
control a significant proportion of incoming sunlight [2,3,8,9,10,11,12,13]. Hence, optimization of 69
building surfaces, especially roofs that are more optimal locations in urban areas, is still missing for 70
maximization of DNI, before installation of TSSCs. 71
This research envisions that design optimization of both; TSSCs and building can enable 72
informed decision making, by generating several solutions and evaluating across different objectives. 73
In this sense, the building roof can be optimized to maximize DNI, and TSSC geometry can be 74
optimized for optimal performance. One possibility is to develop parametric models by mimicking 75
design parameters [37] and applying multi-objective optimization [26,38,39] where genetic algorithm 76
(GA) based methods are widely adopted in the buildings and energy research [26,39]. In the current 77
research, there exist several parametric models combined with multi-objective optimization to 78
maximize solar gains on building envelopes, ultimately energy yield [2,3,8,9,10,11,12,13]. To the 79
authors’ best knowledge, no study proposed any model to maximize DNI by applying parametric 80
modeling and multi-objective optimization approaches. Existing models [2,3,8,9,10,11,12,13] enabled 81
building design optimization, by mimicking building-related decision variables e.g., roof design and 82
slope, or orientation. However, these models are limited to integrating PVs with buildings. Further, 83
4
these models are limited to using fixed, and commercially available PVs, and do not include PV-related 84
decision variables. 85
To the authors’ best knowledge, there exists no parametric modeling approach combined with 86
multi-objective optimization for building-integrated TSSs considering building- and TSSCs-related 87
decision variables, thus a collaborative design optimization of both is still missing. This motivates to 88
development of an integrated design optimization approach in subsequent steps: optimization of roof 89
design for maximizing DNI followed by optimization of TSSC design for improved performance. To 90
begin to address this knowledge gap, this paper introduces a two-step design optimization method that 91
allows for automatically integrating TSSCs with building roofs. The proposed method is based on our 92
previously developed performance assessment method [40] of roof-integrated TSSCs, where several 93
design alternatives were developed by applying a parametric modeling approach. Previously [40], we 94
accessed the performance of roof-integrated TSSCs by manipulation of the building- and TSSCs-95
related design parameters i.e., roof shape, roof slope, building orientation, TSSC type, geometric ratio, 96
and separation distance between mirrors. However, we did not optimize these designs through 97
optimization algorithm(s), and manually filtered designs according to performance criteria. The method 98
presented in this study enables a two-step optimization: (1) optimizing the roof design to maximize DNI 99
(single-objective), and (2) optimizing TSSC configuration for two performance objectives (multi-100
objective): maximize energy reflected by annual average load match index (av.LMI) [41,42] and 101
minimize covered roof area by TSSC modules. The proposed method applies NSGA-II (GA algorithm) 102
due to its wide applications in the design optimization of buildings and energy systems [43,44,45,46]. 103
In the proposed method, we consider decision variables that are related to both, building design (i.e., 104
roof shape, slope, and orientation) and TSSC design (type, solar cell size, number of modules, 105
geometric ratio, and separation distance between mirrors). Thus, our method helps us to find the best 106
roof design achieving maximum DNI, and best TSSC designs that can be integrated well with buildings 107
achieving maximum energy gain and requiring minimum installation space on the building’s roof. The 108
main hypothesis is that the performance of building-integrated TSSCs can be improved by applying 109
parametric modeling and multi-objective optimization approaches to building scale, and TSSC scale 110
Loading more pages...