Y. Ikeda, C. M. Herr, D. Holzer, S. Kaijima, M. J. Kim. M, A, Schnabel (eds.), Emerging Experience in Past, Present and Future of Digital Architecture, Proceedings of the 20th International Conference of the Association for Computer-Aided Architectural Design Research in Asia CAADRIA 2015, 000–000. © 2015, The Association for Computer-Aided Architectural Design Research in Asia (CAADRIA), Hong Kong

ADAPTIVE POLYMER BASED BIPV SKIN

GALIT SHIFF1, YAEL GILAD2 and AMOS OPHIR3 1,2 Department of Interior, Building and Environment Design, Shenkar college of Engineering & design, Ramat Gan, Israel 3 Department of Plastic Engineering, Shenkar college of Engineering & design, Ramat Gan, Israel 1,2 {galit, yael}@gilad-shiff.com 3 amosophir@shenkar.ac.il

Abstract. This study focuses on developing three-dimensional solar panels, as an alternative to traditional flat Photovoltaic (PV) surfaces in Building Integrated Photovoltaic (BIPV). We propose to increase the energy efficiency of buildings by using the entire envelope for en-ergy production as well as by increasing the efficiency of solar energy output in orientations which were traditionally considered as non-ideal. The panels are constructed from Polycarbonate with integrated flexi-ble photovoltaic film, solar paint or dye. The methodology included digital algorithm-based tools for achieving optimized variable three-dimensional surfaces according to local orientation and location, com-putational climatic simulations and comparative field tests. In addition, the structural, mechanical and thermal properties of the integration be-tween flexible PV sheets and hard plastic curved panels were studied. Interim results demonstrate a potential improvement of 50-80% in en-ergy production per building unit resulting from geometric variations per-se. The dependence of energy production by surface geometry was revealed and an optimized method for solar material distribution on the surface was proposed. A parametric digital tool for automatic generation of optimized three-dimensional panels was developed to-gether with a database and material models of the optimized panels system.

Keywords. Building Integrated Photovoltaics; digital algorithm; cli-matic simulations; building envelope

Y. Ikeda, C. M. Herr, D. Holzer, S. Kaijima, M. J. Kim. M, A, Schnabel (eds.), Emerging Experience in Past, Present and Future of Digital Architecture, Proceedings of the 20th International Conference of the Association for Computer-Aided Architectural Design Research in Asia CAADRIA 2015, 345–354. © 2015, The Association for Computer-Aided Architectural Design Research in Asia (CAADRIA), Hong Kong

2 G. SHIFF, Y. GILAD AND A. OPHIR

1. Introduction

The project deals with the development of three-dimensional curved solar panels to generate energy, as an alternative to traditional flat photovoltaic (PV) surfaces in Building Integrated Photovoltaic (BIPV).

1.1. OVERVIEW

The Field of BIPV refers to the construction of products containing solar en-ergy production mechanisms as an integral component. Previous studies have shown that a successful combination between architecture and energy production structure design, is essential for meeting the targets of nearly zero energy buildings (Soteris, 2013), as specified for instance by European par-liament (2010).

The advantages of building integration of PV systems are that more sur-faces become available on the building for the installation of PV systems and traditional building components are replaced by the PV one, increasing their economic viability. The conflicts between architectural aesthetics and PV systems have few consequences. Architects might often find that while de-signing for optimal energy production, the integrity of the design is lost, while engineers may realize that while maintaining the aesthetic values the efficiency of the system is compromised. Successful Integration of architec-tural elements with PV technology is crucial for both performative and aes-thetical aspects, so the challenge resides in both lines of work.

The position of a PV array towards the solar radiation determines the electrical output and efficiency of the system. Therefore, optimizing the an-gle-setting of a BIPV panel should involve both orientation (azimuth angle) and inclination (altitude angle). Very often, the location and shape of build-ings do not enable optimal angle setting, which puts the entire economical viability of installing the PV system in question.

Most BIPV Integration attempts so far were within glass building materi-als, which are relatively expensive and allow only flat installations. BIPV glass costs per sqm could be twice as high as standard glass (Hagemann, 2008). In this study, Polycarbonate is used as it is easily formed, relatively cheap and is adequate for use in various disciplines such as industrial and agricultural constructions as well as sports halls. The latter are often large energy consumers and are characterized with wide apertures.

Various geometric optimization studies have been carried out in the field of shading elements (Kaftan, 2001; Sargent et al, 2011) but less effort has been made in the field of PV devices. This was mainly concentrating on dy-namic sun-following devices improving energy production. These require energy for activation and are expensive and complicated to maintain.

346

ADAPTIVE POLYMER BASED BIPV SKIN 3

1.2. PROJECT OUTLINES

In this study, we attempt to arrive at optimized solar building envelope using static panels with simple non-conventional materiality, achieving better re-sults and innovative design via parametric geometrical manipulations.

The goal was to apply geometrical manipulation and arrive at a building envelope system demonstrating significant improvement in solar energy production efficiency per unit area. During this process, the energy produc-tion potential of curved various surfaces have been simulated and measured as compared to flat reference surface.

All surfaces were placed at multiple orientations towards the sun, for quantitative understanding of the advantages and disadvantages as well as auto-shading effects of curved BIPV surfaces. Based on this understanding, we have developed a design-to-production algorithm that uses data of local ecliptic and orientation to generate optimized geometry for each component. Further to the computational study, a material research is being conducted with Polycarbonate production technologies. Polycarbonate was chosen as material focus as it incorporates simple means of creation of variable custom curved panels through processes of extrusion and thermoforming.

2. Process overview

2.1. DESIGN AND SIMULATION OF SOLAR ENERGY PERFOR-MANCE OF CURVED SURFACES

This research included two parallel lines of work through which the energy production performance of curved surfaces has been analyzed. The first was through digital simulations and the second was through material field exper-iments. Both studies have been conducted under the same parameters includ-ing measurements in all wide range of orientations from south to west, and from horizontal to vertical, and within the local ecliptic of Tel Aviv, Israel. The orientations from south to east have been neglected due to symmetry, after accounting for the slight differences from symmetry which are known to occur (Chen, 2007).

The starting point was with five surfaces in different ‘wave’ curvatures extruded in one dimension (termed hereafter ‘1D curved surfaces’), com-pared to a sixth flat surface that was used as reference to the curved surfaces.

The six surfaces were modelled in Rhino and were inserted into Ecotect Analysis software for solar radiation simulation. Each surface was rotated in increments of 15 degrees between the south – west orientations and vertical-horizontal inclinations. This yielded 7 orientations for each axis and a total

347

4 G. SHIFF, Y. GILAD AND A. OPHIR

of 49 different positions (Figure 1). The annual average sum of total radia-tion was measured and normalized per square meter per day for the purpose of comparison. The simulation took into account auto-shading and enabled detailed analysis of different areas within each surface.

Figure 1. Visualizing the 49 different positions of one of the six surfaces tested between the south – west orientations and vertical-horizontal inclinations (left) and a table demonstrating

the radiation measurements over the different positions (right).

The same set of simulations were analyzed taking into account only the 50% best performing cells in each surface, in order to eliminate the effect of auto-shaded areas of the curved surfaces on the total results.

Identical material surfaces were installed using flexible ‘PowerFilm’ so-lar sheets mounted on Polycarbonate surfaces. These were placed under di-rect sun at the same orientations which were tested by digital simulations (Figure 2). Solar energy production was measured throughout the day. The measurements at each position were measured along three summer days at a time, in order to eliminate the impact of occasional clouding on the results.

Figure 2. Physical models using flexible ‘PowerFilm’ solar sheets on Polycarbonate surfaces.

348

ADAPTIVE POLYMER BASED BIPV SKIN 5

Further to the simulations and physical model of 1D curved surfaces, the same method was applied to introduce an additional curvature over the lat-eral dimension for the surface producing the highest solar radiation. The new surface (termed hereafter ‘2D curved surface’) were also tested with the same simulation procedure described above but with 16 different positions over the south – west orientations and vertical-horizontal inclinations (Figure 3).

Figure 3. From 1D curved surface to 2D curved surface (left). Radiation measurements for 2D curved surface over 16 different positions (right).

2.2. COMPUTATIONAL MODELS AND ALGORITHMS FOR OPTI-MIZED BUILDING ENVELOPE

Cheng (2009) has developed an iterative algorithm for determining the opti-mal curvature of a surface, considered to be made of 100% solar material, and neglecting auto-shading. Cheng’s optimal curved surface shows a slight improvement of about 1.5% over a flat surface, occurring at nearly vertical positions. In this study, we have introduced the effect of auto-shading and enabled the identification of the best performing areas over each surface.

After analyzing and understanding the effect of orientation on potential energy production of generic surfaces and the auto-shading effect, a prelimi-nary algorithm was developed using Grasshopper plug-in for Rhino. Based on our findings, the algorithm attempts to best orientate 50% of the surface cells towards the sun using the following procedure: first, the 2D curved surface was simplified into an array of square pyramidal elements. Second, for each element in a specific location, orientation and inclination, one of its upper faces was located as close as possible to a perpendicular orientation towards the sun during noon time on the Equinox day and its area was max-imized. For limiting the auto shading effect, the maximal height of each py-ramidal element was limited to half of the base edge length (Figure 4). The resulting optimal surfaces were rapidly inserted into Ecotect Analysis soft-ware for performance test simulations allowing to check the efficiency of the

349

6 G. SHIFF, Y. GILAD AND A. OPHIR

algorithm and to locate the best 50% best performing cells on each surface. An improved algorithm accounting for element continuity across the surface is still under development.

A 3D printed model was produced to demonstrate the use of variable panel components on a sample envelope with differing orientations ranging between south to west over a quarter of a sphere (Figure 4).

Figure 4. An algorithm in Grasshopper plug-in, generating optimal 2D curved surfaces(left and middle). A 3D model demonstrating the use of optimal panels in relation to local orienta-

tion and inclination (right).

3. Results and discussion

The initial digital simulations demonstrated that in all surfaces, the impact of altitude angle (between vertical to horizontal positions) was significantly greater than the impact of azimuth angle. The best 1D curved surface (with a symmetrical 30 degrees angle of curvature in one section) was inferior to the flat surface by up to 11% for ideal orientations, and gave similar or slightly better results to the flat surface in non-ideal orientations. This is consistent with the results of Cheng (2009).

In the comparative analysis of the 50% best performing cells between the best 1D curved surface and the flat surface we have observed a significant advantage for the 1D curved surfaces over the flat surface, with up to 170% better performance in non-ideal orientations. This implies that distributing the solar material only in areas that are best orientated towards the sun could overcome the auto-shading problem and achieve higher efficiencies per solar material unit. For the architect, this enables usage of the remaining (e.g. 50%) of the envelope for other needs such as natural light penetration and aesthetic values. Essentially, for the most ideal orientations, the 50% analysis still demonstrated advantage of the flat surface over the curved surfaces.

The comparative field experiments, which measured the actual perfor-mance of 1D curved surfaces with 100% solar material coverage, has yielded

350

ADAPTIVE POLYMER BASED BIPV SKIN 7

better results than the equivalent digital simulation and demonstrated an ad-vantage to the best 1D curved surface over the flat surface in all orientations (Figure 5). This advantage ranged between 2.5% in ideal orientations to 7% in non-ideal orientations. Therefore, the anticipation is that in future experi-ments with only 50% solar material surface coverage in best performing are-as will also yield better results than those demonstrated in the equivalent dig-ital simulation.

Figure 5. Physical models voltage measurements over 24 hours for flat surface (blue), 1D curved surface with large (green) and small (red) wave. The subplots are for three vertical

inclinations and two horizontal orientations (south and south-west).

Solar radiation simulations for the 2D curved surfaces (with a symmetrical 30 degrees angle of curvature in both sections) on 50% best performing cells have shown an improvement of up to 116% over the best 1D curved surface in non-ideal azimuth values (such as perfect and nearly perfect west orienta-tion). The results of the best 1D curved surface, best 2D curved surface and

351

8 G. SHIFF, Y. GILAD AND A. OPHIR

flat surface are compared and summarized in Figure 6..The energy produc-tion potential in each orientation is shown starting from south to west, and progressing from horizontal to vertical altitude for each azimuth angle.

Figure 6. Digital simulations results: annual average sum of total radiation per square meter per day for 50% best performing cells on a flat surface, 1D curved surface and 2D curved

surface. The pairs of values over the x-axis represent the vertical inclination and horizontal orientation respectively.

These results were therefore incorporated in the form generation algo-

rithm, and comparative efficiency diagrams have been drawn as a basis for design choices between flat, 1D and 2D curved surfaces for different orienta-tions (Figure 7).

Figure 7. Performance diagram in relation to local orientation and inclination for flat surface (left) and 2D curved surface (right). The values are normalized to 100% obtained for the flat surface in ideal conditions. The colours also match these values to emphasize the differences.

352

ADAPTIVE POLYMER BASED BIPV SKIN 9

4. Conclusions and future work

Our encouraging results suggest that BIPV is an emerging opportunity for architects to design a continuous building envelope, without compromising an increased solar gain, thanks to use of previously ‘useless’ facades’ orien-tations. Using curved geometries that can be easily generated in standard ex-trusion and thermoforming production processes today imply a potential for a commercial product.

Future work will include improving the algorithm according to the simu-lation and material experiments, testing the algorithm in specific architectur-al projects and contexts and establishing of CAD-CAM production proce-dure.

In the material testing field, future work will include the development of 1D and 2D curved surfaces with advanced solar material such as solar dye or paint, for better embedding over the surface. The overall goal is to achieve an efficient, technically sound and aesthetically pleasing prototype.

Acknowledgements This project is an R&D collaboration between the Architecture and the Plastics Engineering departments at Shenkar College together with an industrial partner from the Plastics building products industry. The authors wish to thank the Accelerating Creative Talents (ACT) Shen-kar program, Shenkar College of Engineering and Design, and PALRAM Inc. for supporting the project. They also wish to thank Tali Shabtai and Yuval Kahlon for assisting with compu-tational models and simulation and Gilad Otorgust, Amit Ron and Ma’ayan Ben Ari for as-sisting with material development and testing.

References Chen, W.; Shen, H., and Chu, Y.: 2007, Performance evaluation of PV arrays at different tilt

Angles and orientations in BIPV, in D. Y. Goswami and Y. Zhao (eds.), Proceedings of ISES World Congress 2007 (Vol. I – Vol. V), Springer, erlin Heidelberg, 315-317.

Cheng, S.: 2009, Curved photovoltaic surface optimization for BIPV: an evolutionary ap-proach based on solar radiation simulation. Masters thesis, University College London (UCL). Available from: <http://eprints.ucl.ac.uk/18982/1/18982.pdf> (accessed 20 De-cember 2014).

European parliament: 2010, Directive 2010/31/EU of the European parliament and of the council of 19 May 2010 on the energy performance of buildings, Official Journal of the European Union, L153/13. Available from: <http://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32010L0031> (accessed 20 December 2014).Hagemann, B.: 2008, Building integrated Photovoltaics, A new design opportunity for Architects, EU Sunrise Project. Available from: < http://www.aie.eu/files/RES%20TF/BIPV_web.pdf> (accessed 25 March 2015).

Kaftan, E.: 2001, The cellular method to design energy efficient shading form to accommo-date the dynamic characteristics of climate, Proceedings of PLEA 2001 - the 18th Confer-ence on Passive and Low Energy Architecture, 7-9 November 2001, Florianopolis, Brazil, 2, 829-833.

Sargent, J.; Niemasz, J., and Reinhart, C.F.: 2011, SHADERADE: combining rhinoceros and energyplus for the design of static exterior shading devices, Proceedings of Building Sim-

353

10 G. SHIFF, Y. GILAD AND A. OPHIR

ulation 2011: 12th Conference of International Building Performance Simulation Associ-ation, November 14-16, Sydney, Australia, 310-317.

Soteris, A.K.: 2013, Building integration of solar renewable energy systems towards zero or nearly zero energy building, International journal of low carbon technologies. 0, 1-7, Available from: < http://ijlct.oxfordjournals.org/content/early/2013/10/09/ijlct.ctt071.full> (accessed 20 De-cember 2014).

354