Solar control mechanisms for enhanced thermal comfort and daylight control of a fully glazed space

Harris Poirazis1, Amanda O´Donnell2, Magdalena Stefanowicz2, Henrik Davidsson2

1ACC Glasrådgivare, 2Dept. of Energy and Building Design, LTH, Lund University

Malmö, Sweden, harris.poirazis@acc.glas.se

Abstract

This paper deals with the need of developing a method that can assist designers to select appropriate solar control mechanisms in highly glazed spaces, in order to provide enhanced daylight and views while avoiding overheating. The development of the method is based on real life needs and was initiated by an actual project that we, as building physicists / façade consultants were called in to assist. The paper is divided in three parts: (a) brief description of the case study – introduction to the project, (b) description of the method developed to inform façade design and (c) further development of the method. The work described is a combination of project work and academic development carried out in Lund University. The aim of this effort was to develop advanced methods and bring them to the “doorstep of design” by introducing a solar control selection process for maximizing the benefits of glass (daylight and views) while avoiding overheating.

Keywords: highly glazed spaces, atrium, thermal comfort, daylight, solar control mechanisms, frit.

1. Introduction

The aim of this paper is to describe an ongoing effort of a method development towards appropriate selection of solar control mechanisms. The method does not aim to serve optimization purposes nor to interact with Dynamic Thermal Modelling tools assessing building performance. On the contrary, the tool aims to help the designer better understand the potential improvements that can be realised by implementing different solar control mechanisms such as: (a) solar control glass (constant performance, evenly distributed and angular independent), (b) frit (constant performance, unevenly distributed and angular independent) and (c) shading devices (variable performance, evenly or unevenly distributed and quite often angular dependant).

This development project was initiated within the needs of a real life project, the NBS (Nya Barnsjukhuset, Östra Sjukhuset), a hospital located in Gothenburg, Sweden. Then the concept was realised by Harris Poirazis, a Building Physics and façade consultant at ACC Glasrådgivare and it began materializing within the context of the Master Programme in Energy-efficient and Environmental Building Design of LTH, Lund University. The development of the method started with the work of two MSc students Amanda O´Donnell and Magdalena Stefanowicz, under the supervision of Henrik Davidsson and the mentorship of Harris Poirazis. The project is an ongoing process.

2. The case study

2.1 The Building

The NBS is one of the world's preeminent Children's Hospitals. The building contains facilities for rehabilitation, surgery, intensive care, hospital wards and administration. The building has a large, fully glazed, south facing atrium as displayed in Figure 1. This atrium is meant to be used by visitors and patients all year round, therefore providing a sound indoor environment is essential for the building’s quality.

Figure 1: NBS hospital (photo extracted from the White Architects website) 1

2.2 The climate

Gothenburg experiences a rather reasonable (for Scandinavian standards) range of outdoor dry bulb temperatures throughout the year.

Looking a bit closer into an .epw weather file2 (of the closest weather station of Landvetter), one can notice that the outdoor air temperatures hardly falls below the - 10 °C during winter and stays below the 26 °C during summer. In Figure 2 a distribution of outdoor air temperatures is presented. A distinction was made between the total amount of hours within a year (black line) and the hours when the atrium is expected to be most used (orange and green lines).

Figure 2: Distribution of outdoor air temperatures of Gothenburg

2.3 The building model

A view of the thermal model for the NBS is demonstrated in Figure 3. The atrium is facing south and therefore is expected to be fully exposed to direct sun during the warmest times of the day.

Figure 3: View of the initial NBS thermal model

3. The generation of the idea

3.1 The problem

As briefly described above this method/tool development project was initiated through our efforts to answer a real life question as to how we can provide an enhanced visitors’ experience in a south facing atrium of a hospital building. In order to deploy the thoughts that led to this project a description of the string of thoughts and questions that were generated during the brainstorming is necessary.

3.1.1 Clear glass for increased daylight

During early project stages, our client expressed the wish to provide a high level of daylight at the atrium and at the neighbouring façades of the occupied building in order to ensure an enhanced experience for the visitors. Therefore a glazing build-up with high light transmittance glazing was, initially, an obvious choice as it would increase the amount of light entering the space. This, however, would most likely lead to overheating problems due to two main factors: (a) the high solar transmittance would lead to high Tair (indoor air temperatures – as the space is not mechanically cooled during warm periods) and (b) the direct solar component falling onto the visitors would increase the perceived temperatures (as further explained in Section 3.2.2). Therefore, another, more realistic solution needed to be provided, that would reduce the overheating risk and increase the atrium’s usability.

3.1.2 Application of internal movable fabrics

The issue of high air temperatures was initially dealt with by suggesting natural ventilation openings at the vertical façade and the roof of the atrium, while the discomfort caused by the direct solar component was immediately acknowledged and the integration of internally movable shading elements was examined. This is illustrated in Figure 4. Nevertheless, this solution led us to raise two sets of questions: (a) how often will the shading elements be applied due to overheating issues? If shading is used too often, there is less benefit with a fully glazed façade than probably expected and (b) how will the application of fabrics affect the airflows (natural ventilation used to avoid overheating)? If fabrics with very low openness factor significantly block the natural ventilation it might not be possible to provide ventilation and shading simultaneously. This is illustrated in the right illustration in Figure 4.

Figure 4: Effect of movable fabrics on daylight and natural ventilation

3.1.3 Application of solar control glass

Then, the idea of placing a solar control mechanism that will provide adequate solar control while allowing for natural ventilation and maximizing the number of hours with undisturbed views was brought to the table. The possibility to integrate a solar control glass was suggested. In order to provide an acceptable indoor thermal environment and avoid excessive solar radiation, the selection of the required solar control properties of the glass would be decided in relation to the warmest spots of the atrium. This would evidently lead us to the selection of a rather unnecessary “dark” solar control glass. This is illustrated in Figure 5.

Figure 5: Even performance of a solar control glass

3.1.4 Frit

Once we looked closer into the building model, we drew two conclusions: (a) severe overheating might occur only in some areas of the atrium, as the building shades the atrium for different periods of the year and hours of the day, as displayed in Figure 6; therefore, why not apply solar protection only when and where it is required and use a more clear glass where possible? and (b) as there will be some areas where natural ventilation will not improve the occupant experience (by cooling down the occupants - as shown in Figure 7), could we improve the sensation of thermal comfort if (instead of reducing the air temperature) we reduce the direct solar component falling on them?

Figure 6: Self-shading effect of the atrium space

The idea of placing an angular independent, selectively placed solar control mechanism was put forward in order to protect specific areas of the atrium. In this way the hottest areas will not decide for the performance requirement of the vertical and horizontal glass roofs in total.

Figure 7: Effect of natural ventilation in the atrium & unevenly distributed frit on the glass roof

3.2 The method

3.2.1 Indicating the need for proper thermal comfort assessments

Typically, the calculation of Operative Temperatures (OT) takes into account the Mean Radiant Temperatures (MRT) and a well-mixed Air Temperature (according to the standard ISO 7730:2005 Ergonomics of the thermal environment3). For specific air velocities the Operative Temperatures are calculated as an average value of the two above factors. However, in real life scenarios there are two more parameters that may affect comfort: (a) the direct solar radiation that falls onto occupants (increasing the perceived temperatures) and (b) the temperature stratification. This is illustrated in Figure 8.

Figure 8: Typical, left illustration vs. proper thermal comfort assessments, right illustration.

3.2.2 Dynamic Thermal Modelling software tool

An appropriate Dynamic Thermal Modelling tool needed to be used in order to capture the physical parameters as described above. The selected tool was ROOM (BEANS suite). ROOM is a single cell dynamic thermal model based on an explicit finite difference formulation for unsteady heat flows within the building fabric.

ROOM is particularly suitable for determining the environmental conditions in spaces with a low level of servicing, such as atria, naturally ventilated buildings or shopping malls. It is also intended to calculate space heating and cooling loads. ROOM differs from programs based on the CIBSE admittance method in that it is not limited to cyclic loading conditions and the treatment of radiation within the space.

3.3 The Challenge

3.3.1 Thermal analysis input

Using the ROOM Dynamic Thermal Modelling tool we assessed the Mean Radiant Temperatures (MRTs) for every hour of the year. Those temperatures include the direct solar component and they are calculated for several points at each floor as demonstrated in the Figure 9.

Figure 9: View of the Mean Radiant Temperatures (MRT) for each of the atrium floors

3.3.2 Tracing the solar control requirements on the envelope

Once the Mean Radiant Temperatures are calculated for each hour of the year and the dry bulb temperatures at different heights are assessed, the Operative Temperatures can be calculated and therefore qualified according to the set thermal comfort performance requirements set by the designer. In other words, this output data can provide us with a clear understanding as to when and where we need to protect the occupied space most from the sun (solar radiation). This leads to the following challenge: can we trace this information on the envelope (instead of displaying it on a floor level) and distribute frit in such densities that we are able to provide enhanced solar control performance only where we need it more without significantly compromising the overall daylight availability and views of the atrium space?

4. The tool

4.1 The context

The tool was developed within the context of the Master Program in Energy-efficient and Environmental Building Design in LTH, Lund University. The description of this tool is based on the work carried out by the former MSc students: Amanda O´Donnell and Magdalena Stefanowicz4.

4.2 Description of the tool

4.2.1 General

The aim of the tool is to correlate indoor thermal comfort with the need for solar control protection on the façade, in this case an unevenly distributed frit (screen print).

During this work, it was chosen to work with parametric-based modelling in Grasshopper, translating the method into a design tool. Since Rhino/Grasshopper are commonly used 3D modelling tools, the tool can easily be integrated into design processes and potentially influence design. The tool visually and numerically informs designers of when, where and to which extent frit should be applied, in order to satisfy indoor thermal comfort criteria to the extent possible with passive design of shading. The need for shading is quantified according to the occurrence of problematic solar gains and qualified according to the relative influence of these gains on the indoor climate.

Figure 10: Plotting the potentially overheating hours on a sun path

The required input is a .3dm geometry and annual solar adjusted operative temperatures for coordinates of interest to the study. The input can be generated by the tool, as well. The processing of the tool is comprised of several consecutive steps resulting in various output, which can be viewed directly in the Rhino/Grasshopper interface or exported to excel. Below are a few examples of steps included in the tool.

First, the annual operative temperature (that includes the direct solar component) profiles for each grid over the floor plan are imported from Excel. The range of overheating problems are represented by shading need factors and corresponding colours. Then, for each hour with thermal discomfort the solar position on the glazed façade is raytraced, for which intersection points on the facade can be identified, second illustration in Figure 11.

Figure 11: Raytracing the grids to the sky & deriving frit % on the facade

Based on desired indoor temperature and thermal properties of the facade, specific suggestions to the g-value of the glass and the distribution of frit are indicated, third and fourth illustrations in Figure 11.

4.3 Description of the method

The ray tracing module includes four stages of processing data: ray tracing intersection points, calculating shading area, re-distributing g-values and rearranging frit pattern on the façade. All four stages of the processing were established as definitions in Grasshopper (GH).

4.3.1 Ray tracing definition

The ray tracing definition was created for locating and visualizing the intersection points on the façade. The definition was based on backward ray tracing method, utilizing the Ladybug component “Bounce_From_Surface” in GH. The component generates the solar vectors directed from the sensor points to the sun.

Figure: 12 Solar vectors, generated with backward ray tracing. Intersections with the glazed façade are

illustrated by the crosses

4.3.2 Shading area definition

Once the solar position during hours of potential overheating have been traced in the façade, the total required frit area is calculated. This is done with respect to the properties of the frit specified by the user. The main principle of the frit area needed is illustrated in Figure 13.

Figure 13: Facade g-value calculation principle

Where,

gAVE – average g-value of the facade Ash – Total shading frit area need, m2 Afacade – Total façade area, m2 TRsh – Shading frit transparency gNON-SHADED – Non-shaded glass g-value Then, once the area with and without frit has been assessed, the frit is distributed along the façade to the places where the intersections are dense (demonstrating the frequency of the incidences that frit required), but also to the areas that for the specific hours the difference between Mean Radiant and Air Temperature of the occupied space is the highest (expressing the intensity of the problem and therefore the need to apply frit on a specific part of the façade). Each geometry of frit was assigned to location of the corresponding intersection points. As illustrated in Figure 14, resulted frit surfaces are likely to overlap, which reduces the absolute area of the shading. In order to obtain a frit pattern, which meets the need of the total shading frit area, each frit was scaled according to an assigned ∆T factor

Figure 14: Resulted non-uniformly distributed frit pattern

4.3.3 Redistributed g-value definition

Highly glazed facades, are often consisted of smaller glazing units as displayed in Figure 15. Those units can have the same or different g-value.

Figure 15: Facade with uniform g-value of glazing before and after being divided into a grid of six glazing units. The re-distributed g-value definition was created in order to distribute g-values on the glazing units of the façade corresponding to the required frit density that can reduce the overheating risk in the indoor “hot areas”. The number of divisions is decided by the user. Once the façade is divided in smaller parts, each part gets a corresponding g-value depending on the frit distribution and intensity factors as described above. The redistributed g-value output included a list of g-values between 0 and 1. The number of g-values correspond to the number of glazing units which the façade was divided into.

Figure 16: Glazed facade divided into 6 glazing units, with corresponding g-values assigned to each unit

Resulting g-values below zero suggest that the shading need was greater than what is achievable by introducing 100% dense frit. Hereby, if resulted g-value were negative, the non-fritted glass g-value needed to be lowered to compensate for the higher shading need.

4.3.4 Uniformly distributed frit pattern

Once the corresponding g-value is defined for each unit, an equivalent frit density is suggested. The redistributed g-value definition outputs either the values of calculated g-values or the uniformly distributed frit pattern by displaying them on the 3D model, as shown in Figure 17.

Figure 17: Glazed facade before and after distribution of g-values, Non-uniformly and uniformly distributed frit patterns.

5. Future development

5.1 The concept

The concept of the tool’s further development is to extend the purpose of its usage beyond the frit, to other solar control mechanisms such as angular independent shading devices (e.g. internally, or externally placed fabrics), angular dependent shading devices (e.g. louvres, fixed shading elements or overhangs) and even inform design as to the possibilities to increase glazing in areas that will not negatively affect the overheating risk. The output of the tool will aim to give the “best value” of transparent elements with minimum negative consequences, providing enhanced daylight quality, avoiding overheating and allowing for visual connection of the occupants with the outdoors for long periods over the year.

5.2 Brief description

5.2.1 Initial assessment of overheating risk

The core of the tool will follow the line of thinking as previously described with the frit. Hourly data will be exported from a Dynamic Thermal Modelling tool with regard to Mean Radiant or Operative Temperatures. Following the previous example of the atrium, the designer may select specific number of nodes (in our case nine nodes for the ground level, seven for the second and 3 for the third, as displayed in the following Figure 18).

Figure 18: Nodes on the floor plans to be examined (outputted from ROOM software tool).

5.2.2 Implementation of data in sunpath diagrams

Once the hourly data is exported, a visual output of the sunpath is created, incorporating the temperatures for every hour of the year. For each node, one sunpath is created. The coloring is set according to the performance requirements of the designer.

Figure 19: Informative sunpath for overheating risk.

Then, the tool will inform a database of products and through an intelligent processor, will interactively suggest solar control mechanisms according to the designers’ needs and recommendations. An output example of the evaluation of different performance solutions is demonstrated in Figure 20.

Figure 20: indicative tool output.

6. References

[1] Photo extracted from the White Architects website: http://www.white.se/projects/nya-barnsjukhuset-ostra-sjukhuset/.

[2] Energy Plus website, https://energyplus.net/weather-location/europe_wmo_region_6/SWE//SWE_Goteborg.Landvetter.025260_IWEC .

[3] ISO, 2005. ISO 7730:2005(E). Ergonomics of the thermal environment, s.l.: ISO copyright office.

[4] Amanda O´Donnell, Magdalena Stefanowicz, Refined distribution of frit; Method and design tool for improved thermal comfort in glazed spaces, Energy-efficient and Environmental Building Design of LTH, Lund University