- Email: [email protected]
Solar chimney and building ventilation
Applied Energy 84 (2007) 135–146
APPLIED ENERGY www.elsevier.com/locate/apenergy
Solar chimney and building ventilation D.J. Harris *, N. Helwig School of the Built Environment, Heriot-Watt University, Edinburgh EH14 4AS, United Kingdom Available online 1 September 2006
Abstract This study is concerned with the design of a solar chimney to induce ventilation in a building. CFD modelling techniques were used to assess the impacts of inclination angle, double glazing and low-emissivity finishes on the induced ventilation rate. It was found that for a south-facing chimney, an inclination angle of 67.5° from the horizontal was optimum for the location chosen, giving 11% greater efficiency than the vertical chimney, and that a 10% higher efficiency was obtained by using a low-emissivity wall surface. Ó 2006 Elsevier Ltd. All rights reserved. Keywords: Building; Energy; Solar; Chimney; Ventilation
1. Introduction The prospect of global warming has induced architects and building engineers to search for ways of heating, cooling and ventilating buildings by passive means rather than energy-consuming mechanical devices. Among these is the solar chimney, essentially a solar energy absorber with open top and bottom, which induces airflow through a building when solar radiation impinges on it. A relatively large item, it also has a function as an architectural feature and may influence the appearance of the building to which it is attached (Fig. 1). In order to minimise both costs and visual intrusion, it is important to maximise airflow for a given set of weather conditions and size of chimney. Design of the solar chimney is therefore important both in providing efficient air movement and in preserving the architectural integrity of the building. A number of factors influence *
Corresponding author. Tel.: +44 131 451 4634; fax: +44 131 451 3161. E-mail address: [email protected] (D.J. Harris).
0306-2619/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.apenergy.2006.07.001
136
D.J. Harris, N. Helwig / Applied Energy 84 (2007) 135–146 Thermal storage material glazing
insulation
insulation
glazing Black surface Solar radiation
room
Fig. 1. Solar chimney configurations.
the design of the chimney—the location, climate, orientation of the building, size of building to be ventilated, and internal heat-gains, among others. 1.1. Basic operation of a solar chimney Fig. 2 illustrates the basic operation of a solar chimney. Solar radiation passes through glazing and is absorbed at the wall surface. The air in the chimney is then heated by convection and radiation from the absorber. The decrease in density experienced by the air causes it to rise, whereupon it is replaced by air from below, i.e. from the attached room. The rate at which air is drawn through the room depends upon the buoyancy-force experienced, (i.e. dependent upon the temperature differential), the resistance to flow through the chimney, and the resistance to the entry of fresh air into the room. Solar chimneys are generally used to provide ventilation for cooling, but also sometimes for heating, when a fan can be used to direct the warmed air into the building. A significant advantage of a chimney used only for cooling is that the demand for cooling and the supply of solar radiation are in phase. There are many choices to make in the design of a solar chimney, including height, width and depth of cavity, type of glazing, type of absorber, and the inclusion of insulation or thermal mass in the solar chimney. In this work, computer simulations were carried out in order to assist in the design process. The variables studied were slope angle, emissivity of the absorber surface, and the use of single or double glazing on the cover. A solar chimney may be thought of as a special case of a Trombe wall or thermosyphoning air panel. However, a Trombe wall normally constitutes part of the wall of a building and has the disadvantage that it occupies wall area normally taken up by glazing, so causing a loss of daylight and view. Also it usually possesses significant thermal mass. A thermosyphoning air panel typically has a metal absorber and low mass, and
D.J. Harris, N. Helwig / Applied Energy 84 (2007) 135–146
Heated air in Cavity rises
137
Black absorber surface
Solar radiation
insulation Absorbed heat Emitted to Cavity
glazing Air movement
Fig. 2. Operation of solar chimney.
occupies a similar position to the Trombe wall. A solar chimney of the type described here is located above the roof line and therefore does not interfere with the fenestration. The simplest and most obvious layout is to have a vertical chimney (Fig. 1), but it may not be attractive architecturally. A cheaper, and less visually obtrusive format, is to lay the collector along the roof slope, and for greater height a combination of the two may be used. Part of this research work was aimed at assessing the impact of the degree of tilt (from the vertical) on the performance of a solar chimney. Tilt angle is important as the warm air inside a tilted cavity will have a greater convective heat-transfer to the cooler surface of the glazing and will consequently lose some of its heat, resulting in a poorer performance. On many buildings it is cheaper to slope the chimney across the roof at the roof, slope angle, as is often done with solar water-heaters. This could also be carried out as a retrofit with minimal disturbance to the existing building. The amount of solar radiation absorbed depends on the tilt angle. There are therefore two opposing phenomena at work; one degrading the performance, the other potentially
138
D.J. Harris, N. Helwig / Applied Energy 84 (2007) 135–146
improving it. The purpose of this work is to use CFD to calculate the balance of impact, if any, by changing the angle of tilt. The roof chimney incorporates the full collector area and stack height into a pitched roof; the whole of one side of a roof may become a solar chimney. It is important that the solar chimney is insulated against the room inside in order to avoid additional heatgains to the room underneath, especially if used for cooling in summer. 1.2. Advantages of aligning the chimney along the roof slope
Very large collector areas easily achieved. May be more aesthetically pleasing than a tower. No additional towers needed. Likely to be cheaper than a tower design. Easier to retrofit.
1.3. Disadvantages
Stack height is restricted by roof height. Heat transfer between heated air and glass is higher than for a vertical surface. Additional bends create greater pressure-losses. Incorporation of thermal mass may be more difficult.
1.4. Objective The purpose of the solar chimney is to generate air flow through a building, and the aim of this study was to investigate particular aspects of the design (tilt, glazing and emissivity) which would produce maximum air flow for a given set of conditions. Gan [1] showed that increasing the height of the chimney is effective in increasing throughput. However, a tall chimney is not necessarily desirable or convenient, and on a sloped roof, which is typical for this location, it would be more convenient to align the chimney with the slope of the roof. While sloping the chimney at an angle to the vertical increases the length of the chimney without increasing the vertical height, the additional heat-transfer between the air in the cavity and the glass generates additional thermal losses. The purpose of this work was to investigate whether this increased or decreased the overall effectiveness of the chimney, and further, to determine whether an optimum slope-angle existed for a given angle of latitude. In addition, the effect of double glazing the absorber, and of applying low-emissivity coatings to the absorber wall were investigated, for the same range of conditions. 1.5. Previous experimental work Research on solar chimneys has been carried out by a number of people, and research methods have included experimental measurements including temperature and air-movement patterns using a range of techniques, and computer simulations using Computational Fluid Dynamics (CFD). Experimental work has been carried out by Afonso [2], who used test cells, and Betts and Bokhari [3]. Gan [1] used these works to validate a CFD simulation model, which achieved a difference of only 2% between
D.J. Harris, N. Helwig / Applied Energy 84 (2007) 135–146
139
measured values and computer predictions. Data presented by Bouchair [4] have also been used for validation purposes. Sparrow and Azevedo [5] found, from experiment, that the natural convection between two plates reduces dramatically for a channel width of the same order of magnitude as the boundary-layer thickness. According to Andersen [6], the channel width for solar chimneys should be at least 4.7 cm. The airflow rate increases for larger cavity widths up to 0.2–0.3 m. For wider cavities, Bouchair [4] found that a back-flow developed in the middle of the cavity. Zhai et al. [7] found back-flow in an experimental set up with a 0.2 m gap. However, Gan [1] claims rising flow rates for cavities larger than 0.3 m where the inlet breadth is the same size as the cavity width. Bends in the system cause additional pressure losses due to friction, and may also influence the flow along the heated surfaces and affect the thermal performance. Afonso [2] demonstrated that increased thermal storage reduced the day-time flow but increased the night-time flow, as anticipated. The storage effect increases up to a thickness of 10 cm of brick, but beyond that thickness there is no significant change. Miyazadi et al. [8] claimed that there is change up to 0.3 m thickness. Experimental work by Charvat et al. [9] showed an increase in night-time air flow when thermal mass was added. They also showed that a solar chimney during the day gave, on average, 25% greater air flow than a conventional non-solar chimney. It is believed that there was a significant effect from the wind in both cases. The use of additional thermal mass was not investigated here. Afonso [2] found it essential to insulate the outside walls of the chimney in order to achieve solar efficiency: 5 cm was considered to be adequate. Gan [1] found that using double glazing on the cover increased the flow rate by 11%, where the chimney is also used in winter for heating (i.e. Trombe-wall type) then this would be most beneficial, but, for summer cooling only, the additional cost may not be worthwhile. 2. Description of the study carried out 2.1. Climate studied Clearly any design guidance is location specific, since there are geographical variations in solar radiation, external temperatures and sun angle. The conditions used here were for Edinburgh in Scotland at latitude 52°. 2.2. Calculation procedure The availability of general simulation codes solving both thermal and fluid dynamics in one model is still limited to a few applications, therefore the model used here followed others in combining a solution for the heat network and a CFD model. The heat network is solved in a conventional way using basic equations and heat-transfer correlations. The CFD model used was PHOENICS, which was also used by Bouchair [4]. Essentially, in a CFD model, the domain is divided into nodes and approximate solutions to the governing flow-equations are found by numerical methods, principally finite-difference, finite-element or finite-volume methods. The principles behind this and other CFD models are well known in the literature (e.g. [10]) and will not be described in further detail here. In using the model, conventional convergence and stability criteria were utilised. The heat network used is shown in Fig. 3.
140
D.J. Harris, N. Helwig / Applied Energy 84 (2007) 135–146
Fig. 3. The overall heat network.
The governing process is the basic stack-effect, which states that warmer air experiences an upward pressure in relation to cooler air, as a result of a decrease in density. It is defined mathematically by the general stack equation given here and found in CIBSE [11]. 1 1 Dp ¼ qin ðT in þ 273ÞgDh T in þ 273 T out þ 273 Wind blowing across the building may influence the performance of the chimney, but has not been investigated here. The effect of the wind may be detrimental to airflow if a positive pressure is created at the top of the chimney. However, with a roof angle of less than 30° from the horizontal, the outlet is always in a wind shadow, i.e. there is always a negative pressure, thus assisting airflow through the chimney in all conditions.
D.J. Harris, N. Helwig / Applied Energy 84 (2007) 135–146
141
2.3. Turbulence model Natural convection in enclosures has been extensively studied experimentally and numerically. In numerical studies, natural convection in enclosures represents one of the simplest multiple-scale, coupled non-linear fluid-flow problems and provides a convenient vehicle for the development of new analyses and new numerical algorithms. The rectangular cavity is the most extensively studied because many engineering applications can be simplified to this geometry. Several models describing the turbulence regime have been devised, applicable to different circumstances. The RNG turbulence model, which is a j–e model has been found by other users to give good correlations and was the one chosen here. It was validated against the experimental results of Betts [3] and Bouchair [4], giving results within 5% of both of these. For the chimney designed by Bouchair, Gan [1] established the following equation: m ¼ 0:0197ðT surface T air; inlet Þ
0:4015
:
The first estimate of flow rate is made using the standard stack-effect equation to calculate the pressure difference using indoor–outdoor temperature difference. This has to overcome all losses due to friction and separation. All property values were selected for room temperature at this stage. As the friction factor depends on Reynolds number and thus on velocity, an iterative procedure was carried out. 2.4. Solving the heat network The thermal model was run first to predict surface and air conditions and mean flow, then the results from this were entered into the CFD program to evaluate the actual flow characteristics. The heat-transfer coefficients for wall-to-air and air-to-glazing have to be established. Except for temperatures, all other values can be computed from given values. Heat-transfer coefficients were calculated using the Nusselt number, conductivity of air and hydraulic diameter. h ¼ Nuk=Dh where Nu ¼ f ðReD ; PrÞ The model incorporates a number of correlations for various conditions with respect to Reynolds and Prandtl number, selected by ‘IF’ algorithms. The correlation which proved to best meet the conditions in most cases was Gnielinski’s modification of the common Petukhov correlation, incorporating aspects of surface friction. NuD ¼
ðf =8ÞðReD 1000ÞPr 0:5
1 þ 12:7ðf =8Þ ðPr0:66 1Þ
The Nusselt number can be seen as a function of Reynolds and Prandtl numbers. For the sloped chimney, there is a vertical component of free convection which causes an additional heat-transfer between the wall and the glazing, and which can have a significant influence on the heat loss and overall performance.
142
D.J. Harris, N. Helwig / Applied Energy 84 (2007) 135–146
2.5. Main iteration of flow rate The overall main iteration takes the mean air temperature of the last cycle in order to calculate the new stack flow, giving the new flow-rate. The system is of oscillating character. A low flow rate produces high temperatures, with resulting high flow rate in the next cycle cooling the surface down again. Therefore, a double boundary iteration was successful in giving very accurate results in a few cycles, making it possible to use standard spreadsheet programs. For all conditions tested, the error between the last two cycles was below 0.1%. In order to validate the model, it was compared with the parametric study by Gan [1]. Our model correlates well with Gan’s predictions, with deviations between 0.1 and 11%, the greater deviation being at greater stack heights, where our model under-predicts the flow rate. The change in transmittance of the glass with differing angles of incidence was also accounted for. The ratio of transmittance at angle a to normal incidence is virtually the same for each glazing type. For the location chosen, the maximum solar energy would occur on a July day (latitude 52°), when there would be a simultaneous demand for maximum convective cooling: an inclination angle of 22.5° from the horizontal would give the optimum availability of solar energy. 3. Results and discussion A worst-case scenario was investigated, where the outside temperature is equal to or higher than the indoor temperature. The base temperatures were set at 22 °C for the room and outside air. Keeping other variables constant, the wider cavity (0.25 m) gives better results than 0.1 m; 0.25 m being the optimum. With a narrow cavity, significant conduction/convection effects may occur from the air in the cavity to the glass, so diminishing the performance. Once the model was established and subjected to basic validation, selected variables were changed to assess the effects on airflow of different forms of construction. Fig. 6 shows the effect of using double glazing and a low-emissivity coating on the absorber. Low-emissivity coatings allow radiation to be absorbed but limit the emission of longwave infra-red radiation back to the surroundings. The absorber and the adjacent air thus retain more of the heat which enters. The cavity size was also varied. Maximum air flow is given with the low-emissivity surface and cavity width of 0.25 m. Minimum flow is given by double glazing, high emissivity and cavity width of 0.1 m. It may be observed from the figures that double glazing gives some improvement, but it is not significant. For either thickness, low-emissivity gives a greater improvement in performance than double glazing (Fig. 4). The chimney giving the best performance (0.25 m, single glazing, low-emissivity) was studied in more detail regarding the angle. Because of the changes in the position of the sun, the performance varies throughout the day for the differing chimney angles. Fig. 5 shows some typical results at 12.00 and 16.00 h with single glazing and high emissivity. The change in mean cavity-temperature with angle of inclination from the horizontal is shown in Fig. 6. The flow rate increases for up to 67.5°, and for angles lower then 45° the flow reduces. At 22° from the vertical, the availability of solar radiation is at a maximum (at this latitude): however the flow rate is still lower than at 67.5°, due to greater heat-
D.J. Harris, N. Helwig / Applied Energy 84 (2007) 135–146
143
90 80
Flow Rate l/s
70 60 50 40 30
Low emissivity, w = 0.25 Double Glazing, w = 0.25
20
Low emissivity, w = 0.1
10
Double Glazing, w = 0.1
0 0
10
20
30
40
50
60
70
80
90
80
90
Inclination Angle Fig. 4. Effect of double glazing and low-emissivity.
120
Flow Rate l/s
100 80 60 40
12.00hr, room at 24C 12.00 hr, room at 22C
20
16.00 hr, room at 24C 16.00hr, room at 22C
0 0
10
20
30
40
50
60
70
Inclination Angle Fig. 5. Single glazing, low-emissivity at 12.00 and 16.00 h.
losses. Twenty-four hour performance is shown in Fig. 7. The slope of roofs on houses in the location studied is either 45° (older houses) or 60° from the vertical (more modern houses) which means that sloping the solar chimney along the roof line gives less than optimum conditions. On average, the performance at 45° is approximately the same as at 90°, but 10% lower than at 67.5°. At 60°, the flow is 25% lower than at 90°, and 30% lower than
144
D.J. Harris, N. Helwig / Applied Energy 84 (2007) 135–146
Mean Cavity Temperature (ºC)
45 40 35 30 25 20 15 10 5 0 0
10
20
30
40
50
60
70
80
90
15
16
Inclination Angle Fig. 6. Effect of inclination angle on cavity temperature.
140 120
Flow Rate l/s
100 80 60
Angle 67.5 Angle 90
40
Angle 45 Angle 22.5
20 0 8
9
10
11
12
13
14
Sun Time (hrs) Fig. 7. Twenty-four hour performance.
at 67.5°. The increase in heat loss from the surface more than counterbalances the increase in solar radiation received, (at least during the summer months, when ventilation cooling is required most.) When the reductions in construction costs are taken into account, however, it may be more beneficial to simply construct the chimney along the slope of the roof. This study demonstrates that the reduction in flow as a result of so doing is relatively insignificant when the roof is at 45°, but considerably less at 30°.
D.J. Harris, N. Helwig / Applied Energy 84 (2007) 135–146
145
Further studies planned will include an experimental confirmation of this, and a lifecycle costing exercise to determine whether using the roof slope, even with poorer performance, can deliver a life-time improvement. Year-round performance should also be investigated. Fig. 7 shows results through the day the for 24 °C room-air temperature. It may be observed that the performance at 90° is similar in the morning hours, but falls off in the afternoon in comparison with that for the optimum angle. 4. Conclusions CFD was used to investigate the performance of a solar chimney. It was found that varying the slope of the chimney resulted in variations in performance, as measured by the airflow rate through the chimney. The optimum slope-angle for maximum flow is 67.5° from the horizontal, giving an average benefit of 11% increase in flowrate in comparison with that for a vertical chimney. This gives an improved performance in cooling and ventilating the building, and reduces the risk of overheating. Application of low-emissivity finishes to the wall offers an additional way of improving performance, giving approximately a further 10% improvement at that angle. The addition of double glazing gave a slight improvement in performance, but it was not significant enough to be cost effective. Although the effect of wind on the flow rates has not been investigated here, it would be an interesting avenue for future research. With roof angles less than 23° from the horizontal, the effect of wind always is to increase the stack suction pressure. With roof angles greater than this, wind direction plays a part in determining whether the stack pressure is increased or decreased (Sharples et al., [12]). For low-angle roofs, therefore, wind would always increase the air flow. Alternatively, the air-flow direction could be reversed by adding a fan and ductwork, when the system could be used for heating. Further work, at present under way, includes physical modelling to confirm these results, and life-cycle costing studies. Appendix A. Data South facing building, latitude 52°, 15th July. Height 3 m. Cavity width 0.1–0.3 m, best performance at 0.25 m. Base case 12.00 h suntime, giving maximum solar gains. Cavity breadth greater than 1 m. References [1] Gan G. A parametric study of Trombe walls for passive cooling of buildings. Energ Buildings 1998;27:37–43. [2] Afonso C. Solar chimneys: simulation and experiment. Energ Buildings 2000;32:71–9. [3] Betts PI, Bokhari IH. Experiments on natural convection of air in a tall cavity. In: IAHR workshop on flow modelling, Paris, vol. V; 1996. p. 25–6. [4] Bouchair A. Solar chimney for promoting cooling ventilation in southern Algeria. Build Services Eng Res Technol 1994;15(1):81. [5] Sparrow EM, Azevedo LFA. Vertical channel natural-convection spanning between the fully-developed limit and the single-plate boundary-layer limit. Int J Heat Mass Transf 1985;28(10):1847–57.
146
D.J. Harris, N. Helwig / Applied Energy 84 (2007) 135–146
[6] Anderson KT. Theoretical considerations on natural ventilation by thermal buoyancy. Trans. ASHRAE 1995;101(2):1103–17. [7] Zhai XQ, Dai YJ, Wang RZ. Experimental investigation on air heating and natural ventilation of a roof solar-collector. Denver (USA): World Renewable Energy Congress; 2004. [8] Miyazadi T, Akisawa A, Kashiwagi T, Gan G. A computational fluid-dynamics analysis of solar chimneys integrated with photovoltaics. Denver (USA): World Renewable Energy Congress; 2004. [9] Charvat P, Jicah M, Stetina J. Solar chimneys for ventilation and passive cooling. Denver (USA): World Renewable Energy Congress; 2004. [10] Anderson JD. Computational fluid-dynamics. New York: McGraw-Hill; 1995. [11] CIBSE Guide Part A, Environmental Design. CIBSE, 7th ed. 2006, p. 4–10. [12] Sharples S, Pursglove SD, Barrell DJW. The effect on pressure coefficients of adding a conservatory to a building for solar ventilation pre-heat. In: European conference on architecture Munich, Germany; 1987.
APPLIED ENERGY www.elsevier.com/locate/apenergy
Solar chimney and building ventilation D.J. Harris *, N. Helwig School of the Built Environment, Heriot-Watt University, Edinburgh EH14 4AS, United Kingdom Available online 1 September 2006
Abstract This study is concerned with the design of a solar chimney to induce ventilation in a building. CFD modelling techniques were used to assess the impacts of inclination angle, double glazing and low-emissivity finishes on the induced ventilation rate. It was found that for a south-facing chimney, an inclination angle of 67.5° from the horizontal was optimum for the location chosen, giving 11% greater efficiency than the vertical chimney, and that a 10% higher efficiency was obtained by using a low-emissivity wall surface. Ó 2006 Elsevier Ltd. All rights reserved. Keywords: Building; Energy; Solar; Chimney; Ventilation
1. Introduction The prospect of global warming has induced architects and building engineers to search for ways of heating, cooling and ventilating buildings by passive means rather than energy-consuming mechanical devices. Among these is the solar chimney, essentially a solar energy absorber with open top and bottom, which induces airflow through a building when solar radiation impinges on it. A relatively large item, it also has a function as an architectural feature and may influence the appearance of the building to which it is attached (Fig. 1). In order to minimise both costs and visual intrusion, it is important to maximise airflow for a given set of weather conditions and size of chimney. Design of the solar chimney is therefore important both in providing efficient air movement and in preserving the architectural integrity of the building. A number of factors influence *
Corresponding author. Tel.: +44 131 451 4634; fax: +44 131 451 3161. E-mail address: [email protected] (D.J. Harris).
0306-2619/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.apenergy.2006.07.001
136
D.J. Harris, N. Helwig / Applied Energy 84 (2007) 135–146 Thermal storage material glazing
insulation
insulation
glazing Black surface Solar radiation
room
Fig. 1. Solar chimney configurations.
the design of the chimney—the location, climate, orientation of the building, size of building to be ventilated, and internal heat-gains, among others. 1.1. Basic operation of a solar chimney Fig. 2 illustrates the basic operation of a solar chimney. Solar radiation passes through glazing and is absorbed at the wall surface. The air in the chimney is then heated by convection and radiation from the absorber. The decrease in density experienced by the air causes it to rise, whereupon it is replaced by air from below, i.e. from the attached room. The rate at which air is drawn through the room depends upon the buoyancy-force experienced, (i.e. dependent upon the temperature differential), the resistance to flow through the chimney, and the resistance to the entry of fresh air into the room. Solar chimneys are generally used to provide ventilation for cooling, but also sometimes for heating, when a fan can be used to direct the warmed air into the building. A significant advantage of a chimney used only for cooling is that the demand for cooling and the supply of solar radiation are in phase. There are many choices to make in the design of a solar chimney, including height, width and depth of cavity, type of glazing, type of absorber, and the inclusion of insulation or thermal mass in the solar chimney. In this work, computer simulations were carried out in order to assist in the design process. The variables studied were slope angle, emissivity of the absorber surface, and the use of single or double glazing on the cover. A solar chimney may be thought of as a special case of a Trombe wall or thermosyphoning air panel. However, a Trombe wall normally constitutes part of the wall of a building and has the disadvantage that it occupies wall area normally taken up by glazing, so causing a loss of daylight and view. Also it usually possesses significant thermal mass. A thermosyphoning air panel typically has a metal absorber and low mass, and
D.J. Harris, N. Helwig / Applied Energy 84 (2007) 135–146
Heated air in Cavity rises
137
Black absorber surface
Solar radiation
insulation Absorbed heat Emitted to Cavity
glazing Air movement
Fig. 2. Operation of solar chimney.
occupies a similar position to the Trombe wall. A solar chimney of the type described here is located above the roof line and therefore does not interfere with the fenestration. The simplest and most obvious layout is to have a vertical chimney (Fig. 1), but it may not be attractive architecturally. A cheaper, and less visually obtrusive format, is to lay the collector along the roof slope, and for greater height a combination of the two may be used. Part of this research work was aimed at assessing the impact of the degree of tilt (from the vertical) on the performance of a solar chimney. Tilt angle is important as the warm air inside a tilted cavity will have a greater convective heat-transfer to the cooler surface of the glazing and will consequently lose some of its heat, resulting in a poorer performance. On many buildings it is cheaper to slope the chimney across the roof at the roof, slope angle, as is often done with solar water-heaters. This could also be carried out as a retrofit with minimal disturbance to the existing building. The amount of solar radiation absorbed depends on the tilt angle. There are therefore two opposing phenomena at work; one degrading the performance, the other potentially
138
D.J. Harris, N. Helwig / Applied Energy 84 (2007) 135–146
improving it. The purpose of this work is to use CFD to calculate the balance of impact, if any, by changing the angle of tilt. The roof chimney incorporates the full collector area and stack height into a pitched roof; the whole of one side of a roof may become a solar chimney. It is important that the solar chimney is insulated against the room inside in order to avoid additional heatgains to the room underneath, especially if used for cooling in summer. 1.2. Advantages of aligning the chimney along the roof slope
Very large collector areas easily achieved. May be more aesthetically pleasing than a tower. No additional towers needed. Likely to be cheaper than a tower design. Easier to retrofit.
1.3. Disadvantages
Stack height is restricted by roof height. Heat transfer between heated air and glass is higher than for a vertical surface. Additional bends create greater pressure-losses. Incorporation of thermal mass may be more difficult.
1.4. Objective The purpose of the solar chimney is to generate air flow through a building, and the aim of this study was to investigate particular aspects of the design (tilt, glazing and emissivity) which would produce maximum air flow for a given set of conditions. Gan [1] showed that increasing the height of the chimney is effective in increasing throughput. However, a tall chimney is not necessarily desirable or convenient, and on a sloped roof, which is typical for this location, it would be more convenient to align the chimney with the slope of the roof. While sloping the chimney at an angle to the vertical increases the length of the chimney without increasing the vertical height, the additional heat-transfer between the air in the cavity and the glass generates additional thermal losses. The purpose of this work was to investigate whether this increased or decreased the overall effectiveness of the chimney, and further, to determine whether an optimum slope-angle existed for a given angle of latitude. In addition, the effect of double glazing the absorber, and of applying low-emissivity coatings to the absorber wall were investigated, for the same range of conditions. 1.5. Previous experimental work Research on solar chimneys has been carried out by a number of people, and research methods have included experimental measurements including temperature and air-movement patterns using a range of techniques, and computer simulations using Computational Fluid Dynamics (CFD). Experimental work has been carried out by Afonso [2], who used test cells, and Betts and Bokhari [3]. Gan [1] used these works to validate a CFD simulation model, which achieved a difference of only 2% between
D.J. Harris, N. Helwig / Applied Energy 84 (2007) 135–146
139
measured values and computer predictions. Data presented by Bouchair [4] have also been used for validation purposes. Sparrow and Azevedo [5] found, from experiment, that the natural convection between two plates reduces dramatically for a channel width of the same order of magnitude as the boundary-layer thickness. According to Andersen [6], the channel width for solar chimneys should be at least 4.7 cm. The airflow rate increases for larger cavity widths up to 0.2–0.3 m. For wider cavities, Bouchair [4] found that a back-flow developed in the middle of the cavity. Zhai et al. [7] found back-flow in an experimental set up with a 0.2 m gap. However, Gan [1] claims rising flow rates for cavities larger than 0.3 m where the inlet breadth is the same size as the cavity width. Bends in the system cause additional pressure losses due to friction, and may also influence the flow along the heated surfaces and affect the thermal performance. Afonso [2] demonstrated that increased thermal storage reduced the day-time flow but increased the night-time flow, as anticipated. The storage effect increases up to a thickness of 10 cm of brick, but beyond that thickness there is no significant change. Miyazadi et al. [8] claimed that there is change up to 0.3 m thickness. Experimental work by Charvat et al. [9] showed an increase in night-time air flow when thermal mass was added. They also showed that a solar chimney during the day gave, on average, 25% greater air flow than a conventional non-solar chimney. It is believed that there was a significant effect from the wind in both cases. The use of additional thermal mass was not investigated here. Afonso [2] found it essential to insulate the outside walls of the chimney in order to achieve solar efficiency: 5 cm was considered to be adequate. Gan [1] found that using double glazing on the cover increased the flow rate by 11%, where the chimney is also used in winter for heating (i.e. Trombe-wall type) then this would be most beneficial, but, for summer cooling only, the additional cost may not be worthwhile. 2. Description of the study carried out 2.1. Climate studied Clearly any design guidance is location specific, since there are geographical variations in solar radiation, external temperatures and sun angle. The conditions used here were for Edinburgh in Scotland at latitude 52°. 2.2. Calculation procedure The availability of general simulation codes solving both thermal and fluid dynamics in one model is still limited to a few applications, therefore the model used here followed others in combining a solution for the heat network and a CFD model. The heat network is solved in a conventional way using basic equations and heat-transfer correlations. The CFD model used was PHOENICS, which was also used by Bouchair [4]. Essentially, in a CFD model, the domain is divided into nodes and approximate solutions to the governing flow-equations are found by numerical methods, principally finite-difference, finite-element or finite-volume methods. The principles behind this and other CFD models are well known in the literature (e.g. [10]) and will not be described in further detail here. In using the model, conventional convergence and stability criteria were utilised. The heat network used is shown in Fig. 3.
140
D.J. Harris, N. Helwig / Applied Energy 84 (2007) 135–146
Fig. 3. The overall heat network.
The governing process is the basic stack-effect, which states that warmer air experiences an upward pressure in relation to cooler air, as a result of a decrease in density. It is defined mathematically by the general stack equation given here and found in CIBSE [11]. 1 1 Dp ¼ qin ðT in þ 273ÞgDh T in þ 273 T out þ 273 Wind blowing across the building may influence the performance of the chimney, but has not been investigated here. The effect of the wind may be detrimental to airflow if a positive pressure is created at the top of the chimney. However, with a roof angle of less than 30° from the horizontal, the outlet is always in a wind shadow, i.e. there is always a negative pressure, thus assisting airflow through the chimney in all conditions.
D.J. Harris, N. Helwig / Applied Energy 84 (2007) 135–146
141
2.3. Turbulence model Natural convection in enclosures has been extensively studied experimentally and numerically. In numerical studies, natural convection in enclosures represents one of the simplest multiple-scale, coupled non-linear fluid-flow problems and provides a convenient vehicle for the development of new analyses and new numerical algorithms. The rectangular cavity is the most extensively studied because many engineering applications can be simplified to this geometry. Several models describing the turbulence regime have been devised, applicable to different circumstances. The RNG turbulence model, which is a j–e model has been found by other users to give good correlations and was the one chosen here. It was validated against the experimental results of Betts [3] and Bouchair [4], giving results within 5% of both of these. For the chimney designed by Bouchair, Gan [1] established the following equation: m ¼ 0:0197ðT surface T air; inlet Þ
0:4015
:
The first estimate of flow rate is made using the standard stack-effect equation to calculate the pressure difference using indoor–outdoor temperature difference. This has to overcome all losses due to friction and separation. All property values were selected for room temperature at this stage. As the friction factor depends on Reynolds number and thus on velocity, an iterative procedure was carried out. 2.4. Solving the heat network The thermal model was run first to predict surface and air conditions and mean flow, then the results from this were entered into the CFD program to evaluate the actual flow characteristics. The heat-transfer coefficients for wall-to-air and air-to-glazing have to be established. Except for temperatures, all other values can be computed from given values. Heat-transfer coefficients were calculated using the Nusselt number, conductivity of air and hydraulic diameter. h ¼ Nuk=Dh where Nu ¼ f ðReD ; PrÞ The model incorporates a number of correlations for various conditions with respect to Reynolds and Prandtl number, selected by ‘IF’ algorithms. The correlation which proved to best meet the conditions in most cases was Gnielinski’s modification of the common Petukhov correlation, incorporating aspects of surface friction. NuD ¼
ðf =8ÞðReD 1000ÞPr 0:5
1 þ 12:7ðf =8Þ ðPr0:66 1Þ
The Nusselt number can be seen as a function of Reynolds and Prandtl numbers. For the sloped chimney, there is a vertical component of free convection which causes an additional heat-transfer between the wall and the glazing, and which can have a significant influence on the heat loss and overall performance.
142
D.J. Harris, N. Helwig / Applied Energy 84 (2007) 135–146
2.5. Main iteration of flow rate The overall main iteration takes the mean air temperature of the last cycle in order to calculate the new stack flow, giving the new flow-rate. The system is of oscillating character. A low flow rate produces high temperatures, with resulting high flow rate in the next cycle cooling the surface down again. Therefore, a double boundary iteration was successful in giving very accurate results in a few cycles, making it possible to use standard spreadsheet programs. For all conditions tested, the error between the last two cycles was below 0.1%. In order to validate the model, it was compared with the parametric study by Gan [1]. Our model correlates well with Gan’s predictions, with deviations between 0.1 and 11%, the greater deviation being at greater stack heights, where our model under-predicts the flow rate. The change in transmittance of the glass with differing angles of incidence was also accounted for. The ratio of transmittance at angle a to normal incidence is virtually the same for each glazing type. For the location chosen, the maximum solar energy would occur on a July day (latitude 52°), when there would be a simultaneous demand for maximum convective cooling: an inclination angle of 22.5° from the horizontal would give the optimum availability of solar energy. 3. Results and discussion A worst-case scenario was investigated, where the outside temperature is equal to or higher than the indoor temperature. The base temperatures were set at 22 °C for the room and outside air. Keeping other variables constant, the wider cavity (0.25 m) gives better results than 0.1 m; 0.25 m being the optimum. With a narrow cavity, significant conduction/convection effects may occur from the air in the cavity to the glass, so diminishing the performance. Once the model was established and subjected to basic validation, selected variables were changed to assess the effects on airflow of different forms of construction. Fig. 6 shows the effect of using double glazing and a low-emissivity coating on the absorber. Low-emissivity coatings allow radiation to be absorbed but limit the emission of longwave infra-red radiation back to the surroundings. The absorber and the adjacent air thus retain more of the heat which enters. The cavity size was also varied. Maximum air flow is given with the low-emissivity surface and cavity width of 0.25 m. Minimum flow is given by double glazing, high emissivity and cavity width of 0.1 m. It may be observed from the figures that double glazing gives some improvement, but it is not significant. For either thickness, low-emissivity gives a greater improvement in performance than double glazing (Fig. 4). The chimney giving the best performance (0.25 m, single glazing, low-emissivity) was studied in more detail regarding the angle. Because of the changes in the position of the sun, the performance varies throughout the day for the differing chimney angles. Fig. 5 shows some typical results at 12.00 and 16.00 h with single glazing and high emissivity. The change in mean cavity-temperature with angle of inclination from the horizontal is shown in Fig. 6. The flow rate increases for up to 67.5°, and for angles lower then 45° the flow reduces. At 22° from the vertical, the availability of solar radiation is at a maximum (at this latitude): however the flow rate is still lower than at 67.5°, due to greater heat-
D.J. Harris, N. Helwig / Applied Energy 84 (2007) 135–146
143
90 80
Flow Rate l/s
70 60 50 40 30
Low emissivity, w = 0.25 Double Glazing, w = 0.25
20
Low emissivity, w = 0.1
10
Double Glazing, w = 0.1
0 0
10
20
30
40
50
60
70
80
90
80
90
Inclination Angle Fig. 4. Effect of double glazing and low-emissivity.
120
Flow Rate l/s
100 80 60 40
12.00hr, room at 24C 12.00 hr, room at 22C
20
16.00 hr, room at 24C 16.00hr, room at 22C
0 0
10
20
30
40
50
60
70
Inclination Angle Fig. 5. Single glazing, low-emissivity at 12.00 and 16.00 h.
losses. Twenty-four hour performance is shown in Fig. 7. The slope of roofs on houses in the location studied is either 45° (older houses) or 60° from the vertical (more modern houses) which means that sloping the solar chimney along the roof line gives less than optimum conditions. On average, the performance at 45° is approximately the same as at 90°, but 10% lower than at 67.5°. At 60°, the flow is 25% lower than at 90°, and 30% lower than
144
D.J. Harris, N. Helwig / Applied Energy 84 (2007) 135–146
Mean Cavity Temperature (ºC)
45 40 35 30 25 20 15 10 5 0 0
10
20
30
40
50
60
70
80
90
15
16
Inclination Angle Fig. 6. Effect of inclination angle on cavity temperature.
140 120
Flow Rate l/s
100 80 60
Angle 67.5 Angle 90
40
Angle 45 Angle 22.5
20 0 8
9
10
11
12
13
14
Sun Time (hrs) Fig. 7. Twenty-four hour performance.
at 67.5°. The increase in heat loss from the surface more than counterbalances the increase in solar radiation received, (at least during the summer months, when ventilation cooling is required most.) When the reductions in construction costs are taken into account, however, it may be more beneficial to simply construct the chimney along the slope of the roof. This study demonstrates that the reduction in flow as a result of so doing is relatively insignificant when the roof is at 45°, but considerably less at 30°.
D.J. Harris, N. Helwig / Applied Energy 84 (2007) 135–146
145
Further studies planned will include an experimental confirmation of this, and a lifecycle costing exercise to determine whether using the roof slope, even with poorer performance, can deliver a life-time improvement. Year-round performance should also be investigated. Fig. 7 shows results through the day the for 24 °C room-air temperature. It may be observed that the performance at 90° is similar in the morning hours, but falls off in the afternoon in comparison with that for the optimum angle. 4. Conclusions CFD was used to investigate the performance of a solar chimney. It was found that varying the slope of the chimney resulted in variations in performance, as measured by the airflow rate through the chimney. The optimum slope-angle for maximum flow is 67.5° from the horizontal, giving an average benefit of 11% increase in flowrate in comparison with that for a vertical chimney. This gives an improved performance in cooling and ventilating the building, and reduces the risk of overheating. Application of low-emissivity finishes to the wall offers an additional way of improving performance, giving approximately a further 10% improvement at that angle. The addition of double glazing gave a slight improvement in performance, but it was not significant enough to be cost effective. Although the effect of wind on the flow rates has not been investigated here, it would be an interesting avenue for future research. With roof angles less than 23° from the horizontal, the effect of wind always is to increase the stack suction pressure. With roof angles greater than this, wind direction plays a part in determining whether the stack pressure is increased or decreased (Sharples et al., [12]). For low-angle roofs, therefore, wind would always increase the air flow. Alternatively, the air-flow direction could be reversed by adding a fan and ductwork, when the system could be used for heating. Further work, at present under way, includes physical modelling to confirm these results, and life-cycle costing studies. Appendix A. Data South facing building, latitude 52°, 15th July. Height 3 m. Cavity width 0.1–0.3 m, best performance at 0.25 m. Base case 12.00 h suntime, giving maximum solar gains. Cavity breadth greater than 1 m. References [1] Gan G. A parametric study of Trombe walls for passive cooling of buildings. Energ Buildings 1998;27:37–43. [2] Afonso C. Solar chimneys: simulation and experiment. Energ Buildings 2000;32:71–9. [3] Betts PI, Bokhari IH. Experiments on natural convection of air in a tall cavity. In: IAHR workshop on flow modelling, Paris, vol. V; 1996. p. 25–6. [4] Bouchair A. Solar chimney for promoting cooling ventilation in southern Algeria. Build Services Eng Res Technol 1994;15(1):81. [5] Sparrow EM, Azevedo LFA. Vertical channel natural-convection spanning between the fully-developed limit and the single-plate boundary-layer limit. Int J Heat Mass Transf 1985;28(10):1847–57.
146
D.J. Harris, N. Helwig / Applied Energy 84 (2007) 135–146
[6] Anderson KT. Theoretical considerations on natural ventilation by thermal buoyancy. Trans. ASHRAE 1995;101(2):1103–17. [7] Zhai XQ, Dai YJ, Wang RZ. Experimental investigation on air heating and natural ventilation of a roof solar-collector. Denver (USA): World Renewable Energy Congress; 2004. [8] Miyazadi T, Akisawa A, Kashiwagi T, Gan G. A computational fluid-dynamics analysis of solar chimneys integrated with photovoltaics. Denver (USA): World Renewable Energy Congress; 2004. [9] Charvat P, Jicah M, Stetina J. Solar chimneys for ventilation and passive cooling. Denver (USA): World Renewable Energy Congress; 2004. [10] Anderson JD. Computational fluid-dynamics. New York: McGraw-Hill; 1995. [11] CIBSE Guide Part A, Environmental Design. CIBSE, 7th ed. 2006, p. 4–10. [12] Sharples S, Pursglove SD, Barrell DJW. The effect on pressure coefficients of adding a conservatory to a building for solar ventilation pre-heat. In: European conference on architecture Munich, Germany; 1987.