- Email: [email protected]
Passive cooling in a low-energy office building
Solar Energy 79 (2005) 682–696 www.elsevier.com/locate/solener
Passive cooling in a low-energy office building H. Breesch a
a,*
, A. Bossaer b, A. Janssens
a
Buildings and Climatic Control, Department of Architecture and Urban Planning, Ghent University, J. Plateaustraat 22, B-9000 Ghent, Belgium b Cenergie cvba, B-2600 Berchem, Belgium
Received 22 October 2003; received in revised form 19 November 2004; accepted 7 December 2004 Available online 1 January 2005 Communicated by: Associate Editor Matheos Santamouris
Abstract In office buildings, the use of passive cooling techniques combined with a reduced cooling load may result in a good thermal summer comfort and therefore save cooling energy consumption. This is shown in the low-energy office building ÔSD WorxÕ in Kortrijk (Belgium), in which natural night ventilation and an earth-to-air heat exchanger are applied. In winter, the supply air is successively heated by the earth-to-air heat exchanger and the regenerative heat exchanger, which recovers the heat from the exhaust air. In summer, the earth-to-air heat exchanger cools the ventilation air by day. In addition, natural night ventilation cools down the exposed structure which has accumulated the heat of the previous day. In this article the overall thermal comfort in the office building is evaluated by means of measuring and simulation results. Measurements of summer 2002 are discussed and compared to simulations with a coupled thermal and ventilation simulation model TRNSYS-COMIS. The simulations are used to estimate the relative importance of the different techniques. The evaluation shows that passive cooling has an important impact on the thermal summer comfort in the building. Furthermore, natural night ventilation appears to be much more effective than an earth-to-air heat exchanger to improve comfort. 2004 Elsevier Ltd. All rights reserved. Keywords: Natural night ventilation; Earth-to-air heat exchanger; TRNSYS-COMIS
1. Introduction The cooling energy demand has increased in Europe. The demand for room and packaged air conditioners has grown with 20% between 1999 and 2002 (JRAIA, 2003). The annual report summary of BSRIA on the
* Corresponding author. Tel.: +32 9 264 7861; fax: +32 9 264 4185. E-mail address: [email protected] (H. Breesch).
West European room and packaged air conditioning market in 2003 (BSRIA, 2004) confirms these conclusions. On the other hand, the Kyoto protocol binds the developed countries to reduce the collective emissions of six key greenhouse gases—among which CO2—at least by 5% by 2008–2012. This protocol encourages the governments amongst others to improve energy efficiency and to promote renewable energy (EU, 2003). Therefore, counterbalancing the energy and environmental effects of air conditioning is a strong requirement for the future. Passive cooling may contribute to the Kyoto requirements by reducing the need for cooling
0038-092X/$ - see front matter 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.solener.2004.12.002
H. Breesch et al. / Solar Energy 79 (2005) 682–696
683
Nomenclature A CD Cp g h p PMV PPD WF U v
area of the opening (m2) discharge coefficient of the opening (–) wind pressure coefficient (–) solar transmission coefficient (–) height (m) vapour pressure (Pa) predicted mean vote (–) predicted percentage of dissatisfied people (–) hourly weight factor (h) thermal transmittance (W/m2 K) wind velocity (m/s)
energy while providing a good thermal comfort (Santamouris, 2004). Two interesting and promising passive cooling techniques are natural night ventilation and earth-to-air heat exchangers. Intensive natural night ventilation is driven by wind and thermally (stack) generated pressures and cools down the exposed building structure at night. As a consequence, heat may accumulate the next day and temperature peaks will be reduced and postponed consequently. Furthermore, earth-to-air heat exchangers consist of pipes buried underground through which ventilation air is drawn. The thermal inertia of the soil dampens and postpones temperature fluctuations at the surface (Mihalakakou et al., 1992). At a sufficient depth, the ground temperature is sufficiently low in summer to cool and sufficient warm in winter to heat the ventilation air in contact with the tube in the ground (Givoni, 1994; Argirou, 1996; De Paepe and Janssens, 2003). The performances of these passive cooling techniques depend on multiple building and environmental parameters. Firstly, climatic parameters (temperature difference inside-outside, average outdoor temperature range), building characteristics (thermal inertia of the building and convective heat transfer between ventilation air and thermal mass) and technical parameters (ventilation rate by night and control strategy) define the performance of natural night ventilation (Blondeau et al., 1997; Givoni, 1998; Pfafferott et al., 2003). Following performances have been noticed. Daily peak temperatures in office buildings decrease by maximum 2–4 C (Kolokotroni, 1995; Martin and Fletcher, 1996; Geros et al., 1999; Blondeau et al., 1997; Pfafferott et al., 2003). Blondeau et al. (1997) also measured that the average predicted mean vote (PMV) decreases from 1.5 (warm) to 0.75 (slightly warm) when mechanical night ventilation was applied. In addition, Givoni (1994) showed experimentally that in a hot dry climate the excess hours considerably reduce in a high mass
Greek symbols h temperature (C) l mean r standard deviation u relative humidity (%) Subscripts i internal e external a air s surface
building. Furthermore, a yearly reduction of sensible cooling up to 30% was determined in Greece (Santamouris et al., 1994) while a 30% reduction of cooling energy consumption and 40% reduction of installed cooling capacity were calculated in a low energy office building with stack ventilation flow at night of 10 ach in the United Kingdom (Kolokotroni and Aronis, 1999). In addition, Gratia et al. (2004) calculated a 40% reduction of daily cooling demand in a high mass narrow office building with natural night ventilation on a sunny Belgian summer day. Secondly, the performance of an earth-to-air heat exchanger depends on the air flow rate by day, convective heat transfer between the pipe and the ventilation air, depth, dimensions and number of the pipes, soil temperature and soil properties (Goswami and Dhaliwal, 1985; Tzaferis et al., 1992; Givoni, 1994; Mihalakakou et al., 1994, 1996a,b; Santamouris et al., 1995; Gauthier et al., 1997). The performance is generally expressed by a heating and cooling gain (yearly, monthly, peak, sorted by ambient air temperature, usable energy supply) or a peak temperature shift in summer and winter. Other efficiency criteria are discussed in Pfafferott (2003) and Goswami and Ileslamlou (1990). The seasonal energy performance in a moderate climate is estimated at 8–10 kW h/m2 of ground coupling area for cooling and at 10–15 kW h/m2 of ground coupling area for heating. Peak cooling and heating capacity at ambient air temperatures of respectively 32 C and 5 C are estimated at 45 W/m2 of ground coupling area for cooling and heating (IEA Annex 28, 1997). The feasibility of natural night ventilation and an earth-to-air heat exchanger for cooling is primarily limited by climatic conditions and heat gains. Firstly, hot and moderate climates with large diurnal temperature difference over the summer are best suited for natural night ventilation (Kolokotroni, 1995). The minimum recommended diurnal temperature difference is approximately 10–12 C (Givoni, 1994; Santamouris and
684
H. Breesch et al. / Solar Energy 79 (2005) 682–696
Asimakopoulos, 1996; Lechner, 2001). Only a moderate climate having large temperature difference between day and night as well as between summer and winter is suited for an earth-to-air heat exchanger. In a hot climate, the cooling effect of an earth-to-air heat exchanger is limited whereas in a cold climate the cooling demand is small (Zimmermann and Remund, 1999). Secondly, both passive cooling techniques may only satisfy a low or moderate cooling demand. Furthermore, natural night ventilation is only suitable in buildings with sufficient accessible thermal mass (75–100 kg/m2). As a rule of thumb, IEA Annex 28 (1997) and Van Paassen et al. (1998) recommend a maximum of 20–30 W/m2 heat gains in heavy constructions. As an earth-to-air heat exchanger has a good peak performance but a limited seasonal cooling capacity, it is more suited to provide comfort cooling in low-energy buildings which tend to overheat in the summer (Zimmermann, 1995). Furthermore, as an earth-to-air heat exchanger merely precools the air, a combination with natural or mechanical night ventilation or with slab or ceiling cooling is interesting (Zimmermann and Remund, 1999). In Belgium, during last years, passive cooling techniques have been gradually introduced in office buildings (Heijmans and Wouters, 2002a,b; Breesch et al., 2004). The Belgian moderate climate is very suitable for passive cooling because both the external outside air and the ground may be used as a heat sink in summer. A yearly average ambient temperature of 9.7 C and an average daily maximum and minimum temperature in July of respectively 21.8 C and 12.7 C are measured (RMI, 2004) (see also Table 4). Nevertheless, designers still lack confidence to apply these techniques in the absence of information about the performance of existing applications in Belgium. Therefore, this paper discusses the performance of a low-energy office building in Belgium. The paper will focus on the performances of the building in the cooling season, but will briefly examine the performances in the heating season too. The first part describes the low-energy office building and its operation. The second part discusses the measured operation and performance of the building and applied passive heating and cooling systems. The third part focuses on the simulated performances and estimates the impact of the passive cooling techniques on the thermal summer comfort.
2. Low-energy office building The office building ÔSD WorxÕ is located in Kortrijk, Belgium and consists of two office floors on top of a limited ground floor with building services. On the south side, the floors are connected with an open vertical circulation zone. Figs. 1 and 2 show a plan of the first floor and a vertical section. The plan of the second floor is very similar. The office floors and the ground floor have
Fig. 1. Plan of the first floor of SD Worx (Kortrijk, Belgium) with open plan office (no. 1), circulation zone (no. 2), conference rooms (no. 3) and auditorium (no. 4).
a floor area of respectively 497 m2 and 198 m2. The most important rooms on the office floor are the open plan office (no. 1), circulation zone (no. 2), conference rooms (no. 3) and other offices or an auditorium (no. 4). The vertical section indicates the open plan offices (no. 1) and the stairwell (no. 2) on both floors as well as the technical rooms (no. 5) on the ground floor. The principles of this low-energy office building are the reduction of the heating and cooling load, the use of passive cooling and heating and enhanced control automation (Cenergie cvba, 2003). 2.1. Heating season First of all, the transmission heat losses are minimized. The external walls, roof, floor and windows are
H. Breesch et al. / Solar Energy 79 (2005) 682–696
685
Fig. 2. Cross section of SD Worx (Kortrijk, Belgium) with open plan offices (no. 1), circulation zone (no. 2) on first and second floor and building services on ground floor (no. 5).
very well insulated (see Table 1). Further, the ventilation heat loss is reduced by an earth-to-air and a regenerative heat exchanger which (pre)heat the ventilation supply air. In this project the earth-to-air heat exchanger includes two concrete pipes with an internal diameter of 80 cm and a length of 40 m each, buried in clay ground under the groundwater level at a depth of respectively 3 and 5 m and connected to the ventilation system by PE-pipes with an internal diameter of 40 cm (De Paepe, 2003). Thermal properties of the clay soil on site were not measured. Standard values for clay are a thermal conductivity of 1.5 W/(m K) and a volumetric thermal capacity of 3.0 · 106 J/(m3 K) (PrEN ISO 13370, 1998). A regenerative heat exchanger recovers the heat from the exhaust air as follows. Fresh and exhaust air alternately pass through two accumulators including packed aluminium plates. These plates remove the heat from the warm extracted airflow and pass it to the cold supply airflow. Further, demand controlled ventilation reduces the ventilation heat loss. The building is divided in 11 zones, distributed over the offices and meeting rooms on both office floors. When the CO2 concentration exceeds 800 ppm, fresh air is supplied into the zone through a false floor plenum and air diffusers along the external wall. Minimizing the transmission and ventilation heat losses in this office building, resulted in a low energy
consumption during the heating season 2002–2003, i.e. 62 kW h/m2 (Cenergie cvba, 2003). 2.2. Cooling season Following actions are taken to achieve a good thermal summer comfort. Firstly, the cooling load is reduced. Zoning minimizes the solar heat gains because the offices face north and the circulation zone faces south. This glazed south-facing fac¸ade is finished with controllable external sunblinds to achieve solar shading. The heat emission of lighting and office equipment is also minimized. The installed power of the lighting is limited to 9.5 W/m2, presence detection and daylight control are applied. An exposed concrete ceiling additionally delivers a high thermal capacity, which reduces and postpones the cooling load. Secondly, passive cooling is applied. By day, an earth-to-air heat exchanger precools the supply air flow (Fig. 3). By night, outside air enters the office floors through bottom hung windows, located near the ceiling in the offices on the north side. The air cools down the exposed ceiling and leaves the building at the top of the circulation zone through outlet windows along the width of the building (see Fig. 4). This natural night ventilation system is primarily driven by thermally (stack) generated pressures. Thirdly, the use of these passive cooling techniques is
686
H. Breesch et al. / Solar Energy 79 (2005) 682–696
Table 1 Building data Night ventilation building data Zone
Reference height (m)
Dimensions H/D/W (m)
Louvers From
To
h (m)
A (m2)
CD ()
Outside Office 1st Outside Office 2nd Circulation
Office 1st Circulation Office 2nd Circulation Outside
6.0 6.3 9.5 9.8 11.4
5.66 3.06 5.66 3.06 10.65
0.27 0.33 0.27 0.33 0.33
Office 1st floor
3.5
3/12.1/21.2
Office 2nd floor
7
3/12.1/21.2
Circulation zone 3.5 Thermal building data
8.4/2.9/21.2
Wall
U (W/m2 K)
g ()
Wall
U (W/m2 K)
Window Window, horizontal, circulation
1.1 0.6
0.6 0.28
External floor External floor, circulation Internal floor External wall Internal wall Window frame Roof
0.24 0.15 1.61 1.28 1.93 3.5 0.28
Wall composition External wall Internal wall Flat roof Internal and external floor
Reinforced concrete, embedded with brick panes Wooden cupboards Hollow core concrete slabs Hollow core concrete slabs, false floor
Internal heat gains per floor Present
Maximum
People PC + Screen Laserprinter Lighting
9 pers. 9 pc. 1 pc. 235 m2
24 pers. 24 pc. 1 pc. 235 m2
Total
12.9 W/m2
24.2 W/m2
Radiant (%) 80 W/pers. 130 W/pc. 320 W/pc. 9 W/m2
Convective (%)
50 25 20 25 31
Usage diversity
50 75 80 75 33
69
1 0.75 0.5 0.6 67
Fig. 3. Operation scheme of the earth-to-air heat exchanger in SD Worx (Cenergie cvba, 2003).
H. Breesch et al. / Solar Energy 79 (2005) 682–696
687
Fig. 4. Operation scheme of natural night ventilation in SD Worx (Cenergie cvba, 2003).
automatically controlled to meet requirements at different climatic loads. When the temperature in every zone exceeds 23 C during the day, the supply ventilation rate increases proportionally to the mean indoor temperature: from 5400 m3/h at 23 C to 8000 m3/h at 26 C and above. In this situation, the exhaust air leaves the building through the windows on top of the circulation zone. When the internal temperature in a zone is less than 23 C, the ventilation rate in this zone is demand controlled and the ventilation air is mechanically extracted as in the heating season. The operation of natural night ventilation depends on multiple control parameters: the maximum inside and outside air temperature of the day before, the inside air temperature, the difference between inside and outside air temperature, the internal relative humidity, the rainfall and the wind velocity at that moment. Table 2 lists the settings of the control system. The control parameters were optimised after the first summer. While the inlet windows before were opened at insides air temperatures larger than Table 2 Control system natural night ventilation Natural night ventilation is in operation if the conditions below are fulfilled Previous day hi,max > 23 C he,max > 20 C At that moment 22 h < time < 6 h hi > 20 C hi he > 2 C ui < 70% No rainfall v < 10 m/s
20 C, the opening is now controlled by ceiling surface temperatures exceeding 22 C.
3. Measurements: discussion The building operation system monitors continuously (every 15 min) the outdoor and indoor climate, the air flow rate in the mechanical ventilation system and the control parameters. Measurements from May 15 to September 30, 2002 are used to evaluate the performances of natural night ventilation and the earth-to-air heat exchanger in the cooling season. This was the first summer that the building was in operation. The operation and performance of the earth-to-air heat exchanger are also discussed in the heating season on the basis of monitoring data from November 15 until December 20, 2002. 3.1. Natural night ventilation The monitoring (De Paepe, 2003) shows that natural night ventilation was in operation during 60% of the nights from the end of June until the end of September 2002. Fig. 5 shows the outdoor temperature, the operative and surface temperatures on both floors and the operation of natural night ventilation during a warm summer period. Natural night ventilation cooled down the inside air and exposed ceiling, which had stored the heat of the previous day, from 10 p.m. till 6 a.m. As a result, excess heat accumulated in the ceiling the following day (surface temperature increased by day) and the air temperature peaks by day decreased. No humidity or condensation problems were noticed. The performance of natural night ventilation is analysed based on the achieved air and surface temperature
32
10
30
9
28
8
26
7
24
6
22
5
20
4
18
3
16
2
14
1
Night ventilation operation
H. Breesch et al. / Solar Energy 79 (2005) 682–696
Temperature (˚C)
688
0 12 14:00 22:00 6:00 14:00 22:00 6:00 14:00 22:00 6:00 14:00 22:00 6:00 14:00
Time external
1st floor air
1st floor surface
2nd floor surface
2nd floor air night ventilation
Fig. 5. Measured operation of natural night ventilation (August 12–16, 2002).
changer in summer. The difference between these two temperatures defines the sensible cooling effect of the earth-to-air heat exchanger. During the day, when the mechanical ventilation was in operation, the earthto-air heat exchanger reduced and dampened the outside air temperature fluctuations. When the ventilation system did not work, the temperature in the earth-to-air heat exchanger adopted the ground temperature. Fig. 7 relates and compares the daily maximum temperature of the air leaving the earth-to-air heat exchanger to the daily maximum outdoor air temperature. The maximum temperature of the supply air never exceeded 22 C, i.e. an important cooling effect on warm summer days (cf. the cooling effect at the last day on Fig. 6). During days with a maximum external temperature between 12 and 22 C, the cooling effect was limited. A heating effect was noticed when the maximum outside temperature was less than 12 C. Figs. 8 and 9 illustrate the performances of the earthto-air and regenerative heat exchanger in the heating season. Fig. 8 shows that on a winter day the ventilation air was preheated by the earth-to-air and further heated
2.0 0.8 0.7 0.3
1.3 0.6 0.5 0.2
drop overnight (between 10 p.m. and 6 a.m. the next day). These temperature drops were higher on the first than on the second floor because of the higher thermal stack effect (see Table 3). The outside air temperature peak was on average postponed for 5 h. As a result, the indoor air temperature peaks occurred after the office hours. 3.2. Earth-to-air heat exchangers
32 30 28 26 24 22 20 18 16 14 12
10 9 8 7 6 5 4 3 2 1 0 0:00
18:00
6:00
12:00
0:00
18:00
0:00
18:00
6:00
12:00
0:00
18:00
6:00
12:00
cooling effect
0:00
Temperature (˚C)
The measured performance of the earth-to-air heat exchanger during the summer is shown in Figs. 6 and 7. Fig. 6 compares the external air temperature to the air temperature at the outlet of the earth-to-air heat ex-
6:00
Dhs,i (C)
l r l r
Second floor
12:00
Dha,i (C)
First floor
Ventilation operation
Table 3 Natural night ventilation: temperature drop overnight
Time external
earth-to-air
1st floor air
2nd floor air
ventilation
Fig. 6. Measured operation of the earth-to-air heat exchanger in the cooling season (August 11–14, 2002).
perature of the earth-to-air heat exchanger to the daily minimum outdoor temperature during working hours. A heating effect existed on winter days when the minimum external temperature was lower than 8 C. A limited cooling effect was noticed from a daily minimum external temperature of 10 C.
36 32 28
689
cooling effect
24 20
3.3. Thermal summer comfort
16 12
Thermal summer comfort is evaluated by FangersÕ comfort theory (Fanger, 1972). Fanger defined the conditions in which more than 10% of the occupants are dissatisfied. The critical indoor operative temperature meeting this requirement depends on the metabolism and clothing resistance. Assuming seated office work (metabolism of 65 W/m2) and light working clothes (0.7 Clo), a threshold temperature of 26.0 C is found (ISSO, 1990). Fig. 10 shows the cumulative distribution of the measured temperatures on both floors. An excellent thermal summer comfort was reached: 26 C was only exceeded during 3 h (or 0.3% of summer working hours) on the
8 4 0 0
4
8
12
16
20
24
28
32
36
Maximum outside temperature (˚C) Fig. 7. Measured cooling effect of the earth-to-air heat exchanger.
Temperature (˚C)
to the desired inlet temperature by the regenerative heat exchanger. Fig. 9 relates the daily minimum outlet tem-
25
9000
20
7500
15
6000 heating effect regenerative
10
4500 heating effect earth-to-air
5
3000
0
1500
external
earth-to-air
extraction air
1st floor
0:00
20:00
16:00
Time
supply air
12:00
8:00
4:00
0:00
20:00
16:00
8:00
12:00
4:00
0 0:00
-5
Supply air flow (m³/h)
Maximum earth-to-air temperature (˚C)
H. Breesch et al. / Solar Energy 79 (2005) 682–696
flow rate
Minimum earth-to-air temperature (˚C)
Fig. 8. Measured operation of the earth-to-air and the regenerative heat exchanger in the heating season (December 19–20, 2002).
20 18 16 14 12 10 8 6 4 2 0
heating effect
-6 -4 -2 0
2
4
6
8 10 12 14 16 18 20 22 24 26 28 30
Minimum external temperature (˚C)
Fig. 9. Measured heating effect of the earth-to-air heat exchanger.
Cumulative temperature distribution (% of total working hours)
690
H. Breesch et al. / Solar Energy 79 (2005) 682–696 100
91.3
90
1st floor 2nd floor
80
73.0
70 60 48.3
50 40
33.1
30 20 8.2
10 0
0.0 0.0
0.0 0.3
3.0
> 27˚C
> 26˚C
> 25˚C
> 24˚C
> 23˚C
Maximum inside air temperature 2nd floor (˚C)
Fig. 10. Measured cumulative distribution of temperatures on both floors during working hours (May 15 to September 30, 2002).
28
26
24
22
20 10
14
18
22
26
30
34
Maximum outside air temperature (˚C)
Fig. 11. Measured maximum inside temperature on the second floor during working hours as a function of the measured maximum outside temperature (May 15 to September 30, 2002).
second floor and never on the first floor. Fig. 10 also proves that high temperatures were more frequently monitored on the second floor. An inside temperature of 23 C was exceeded during 73% and 91% of the working hours on the first and the second floor respectively. Fig. 11 relates the daily maximum air temperature on the second floor to the daily maximum outdoor air temperature. The maximum air temperature on the second floor averaged 24.6 C. 98% of the maximum indoor temperatures on the second floor are from 23.5 C to 26 C.
4. Simulations: impact of passive cooling 4.1. Simulation model In a naturally ventilated building, free moving internal temperatures depend on the ventilation flow rates. Because natural (night) ventilation is driven by wind and thermally (stack) generated pressures, the ventila-
tion rate on its turn is function of the indoor air temperature. Therefore, a coupled thermal and ventilation simulation model, which iterates the mass and energy balance per zone till convergence, is necessary to simulate natural night ventilation (Breesch and Janssens, 2002). The office building ÔSD WorxÕ has been simulated by linking TRNSYS (Klein et al., 2000), a transient multi-zone thermal simulation model, to COMIS (Dorer et al., 2001; Haas et al., 2002), a multi-zone infiltration and ventilation simulation model. Both simulation programs subdivide the building in various zones, mostly corresponding to the rooms, in which the air is assumed to be perfectly mixed. In TRNSYS, a zone is represented by two temperatures: the homogeneous air temperature and the so-called star temperature (Seem, 1987). The star temperature is a weighted average of the zone air temperature and the surface temperatures of the walls surrounding the zone. The air temperature is solved from the convective heat flow balance of the zone, the star temperature is solved from the combined convective and radiation heat flow balance. The star temperature concept is introduced to facilitate the calculation of conduction heat loss over a wall. This is modelled according to transfer function relationships (Klein et al., 2000). These relationships describe the heat flux at one side of the wall varying in time, depending on a change of the heat flux at the other side of the wall or on a change of the surface temperature at one of both sides of the wall. In other words, they tell the Ôthermal historyÕ of a wall. In COMIS, each zone is represented by single values for air temperature and pressure. Air flow paths like windows, louvers, cracks and ducts, connect the zones to each other and to the outdoors. The flow rates are related to the pressure difference over the flow path by nonlinear equations. Wind pressure coefficients, relating the wind pressure at a building to the wind velocity, are attributed to external nodes. The zone pressures are solved from the system of mass flow balances. Given these pressures, the flow rates in each air flow path may also be calculated (Haas et al., 2002). The simulation model simplifies the building geometry and includes two landscape office zones and the circulation zone connecting both floors. Table 1 summarizes the characteristics of the night ventilation openings, the composition and the thermal transmittance properties of the walls and the internal heat gains of each floor. The earth-to-air heat exchanger is simplified and modeled as shown in Fig. 7. Since the aim of the simulations is to assess thermal comfort in summer, the regenerative heat exchanger is not taken into account. For an adequate simulation of the control system, a small time step of 6 min is chosen. Wind pressure coefficients are shown in Fig. 12 (Orme et al., 1994). To transform the wind velocity from the meteorological to the local site, a roughness height on site of 0.25 m (i.e. countryside
H. Breesch et al. / Solar Energy 79 (2005) 682–696
Table 4 Comparison outdoor temperature on site and TRY Ukkel (May 15 till September 30)
0.4 office grills
Cp (-)
0.2
691
flat roof
0 0
45
90
135
180
Daily average temperature (C)
-0.2
Daily temperature amplitude (C) -0.4
Daily maximum temperature (C) -0.6
l r l r l r
SD Worx 2002
TRY Ukkel
17.5 3.1 8.9 3.3 22.0 4.1
16.2 3.1 9.0 3.0 20.7 3.9
angle (˚)
Fig. 12. Wind surface pressure coefficient Cp varies with the angle between the wind direction and the normal on the surface (Orme et al., 1994).
and spread habitat) is assumed (Dorer et al., 2001). The open plan offices are designed for an occupation of 24 persons on each floor. At the moment of the measurements, nine persons were working on each floor from 8 h till 17 h. The heat gains of computers, monitors, printers and lighting are multiplied by a diversity factor. This factor takes into account that not all equipment is in use all the time or is throwing out its actual peak heat gain (see Table 1) (Wilkins and Hosni, 2000). Solar irradiation and wind direction data in situ are not available. Consequently, the weather data of the Test Reference Year (TRY) of Ukkel (Belgium) are used in the simulations to estimate the impact of passive cooling on thermal comfort. Fig. 13 compares the cumulative distributions of the daily average temperatures in both weather data. Table 4 reports the summer average and standard deviation of the daily average and maxi-
mum temperature and of the daily temperature amplitude. The summer period on site is on average slightly warmer than in the TRY and contains a larger amount of high temperatures. Fig. 14 additionally compares the cumulative distributions of the daily temperature amplitude. In conclusion, the average and the distribution of the daily temperature amplitude are very similar in both climatic data (see also Table 4). 4.2. Comparison to measurements Before discussing the impact of passive cooling on thermal summer comfort in ÔSD WorxÕ, the simulation results are first compared to the measurements. Fig. 15 shows the relationship between the simulated maximum indoor operative temperatures and the outdoor air temperatures. The simulated cumulative temperature distribution is shown in Fig. 16. These figures may respectively be compared to the measured temperatures in Figs. 11 and 10. The measured and simulated relationship between maximum indoor and outdoor temperatures (see Figs.
100 SD Worx 2002 TRY Ukkel
90
cumulative frequency (%)
80 70 60 50 40 30 20 10 0 12
14
16
18
20
22
24
26
28
Daily average temperature (˚C) Fig. 13. Weather data (May 15 to September 30): comparison cumulative distribution of daily average temperatures on site—Test Reference Year Ukkel.
692
H. Breesch et al. / Solar Energy 79 (2005) 682–696 100 SD Worx 2002 TRY Ukkel
90
cumulative frequency (%)
80 70 60 50 40 30 20 10 0 4
6
8
10
12
14
16
18
Daily temperature amplitude (˚C) Fig. 14. Weather data (May 15 to September 30): comparison cumulative distribution of daily temperature amplitude on site—Test Reference Year Ukkel.
Maximum inside operative temperature 2nd floor (˚C)
28
26
24
22
20
10
14 18 22 26 30 Maximum outside air temperature (˚C)
34
Cumulative air temperature distribution (% of total working hours)
Fig. 15. Simulated maximum operative temperature on the second floor as a function of the maximum outside air temperature from May 15 to September 30 (TRY Ukkel).
100 90 80
1st floor 2nd floor
70
63.3 64.6
60 50 37.3 37.2
40 30 20 9.1 7.5
10 0.0 0.0
0.7 0.7
> 27˚C
> 26˚C
0 > 25˚C
> 24˚C
> 23˚C
Fig. 16. Simulated cumulative distribution of air temperatures on both floors during working hours from May 15 to September 30 (TRY Ukkel).
H. Breesch et al. / Solar Energy 79 (2005) 682–696
4.3. Impact of passive cooling
11 and 15) show the same trend at outside temperatures above 18 C, i.e. when natural night ventilation was in operation. It started working from a maximum outside air temperature of 20 C (see Table 2). On the other hand, at outside temperatures below 18 C, the simulations underestimate the actual temperature peaks in the offices. These temperatures typically occurred during springtime, when the night ventilation system was out of action. Clearly, the difference between simulations and measurements may be attributed to the fact that the simulations do not account for the heating exchange between supply and extract air. Regardless of the fact that the boundary conditions used in the simulations differ from the actual climate, following difference between Figs. 10 and 16 is striking. Contrary to the measurements, the simulated temperatures on both floors hardly differ. Furthermore, the simulation model respectively overestimates and underestimates the operative temperature peaks (hi > 25 C) somewhat on the first and second floor. This difference may be explained by following limitations and assumptions in the multi-zone model. Firstly, an adiabatic floor is assumed between the first and the ground floor with services in the simulation model, as the temperature is not know on this latter floor. Due to this, the transmission losses of the first floor are neglected. Secondly, the vertical circulation zone is represented by one homogeneous temperature in the multi-zone model. In reality, thermal stratification occurred: a temperature difference of 1–1.5 C was measured between the second and the first floor in the circulation zone on a sunny summer day (see Fig. 17). In conclusion, the comparison shows that the simulation model is adequate to predict the effect of passive cooling on thermal summer comfort i.e. when the daily maximum external temperature exceeds 18 C.
The impact of natural night ventilation and the earthto-air heat exchanger on thermal summer comfort in the office building ÔSD WorxÕ is estimated by comparing the reached thermal summer comfort in following cases: the low-energy office building without passive cooling, with only an earth-to-air heat exchanger, with only natural night ventilation and with the combination of both techniques. Thermal summer comfort is characterised by the weighted excess hours during occupation time. A number of weighted working hours in which more than 10% of the occupants are dissatisfied (predicted mean vote or PMV > 0.5) less than 150 h, means a good thermal summer comfort (Van der Linden et al., 2002). Determination of the weighted excess hours is based on the comfort theory of Fanger (1972), which takes account of indoor environmental parameters (air and radiant temperature, air velocity and relative humidity) and personal properties (metabolism, activity level and clothing). The method of weighted excess hours takes account of the degree of discomfort by means of an hourly weight factor (WF). WF is directly proportional to the increase of the predicted percentage of dissatisfied people or PPD: 1 h with 20% dissatisfied people counts for 2 h with 10% dissatisfied. A PMV of 0.5 corresponds to a WF of 1. if PMV < 0:5 then WF ¼ 0 else WF ¼ 10 9:5 exp½ð0:03353PMV4 þ 0:2179PMV2 Þ:
The threshold value of 150 weighted excess hours corresponds to an average of 15% dissatisfied people during 5% of the annual occupation time. This method has been
32
200000
30
180000
28
160000
26
140000
24
120000
22
100000
20
80000
18
60000
16
40000
14
20000
12
Illuminance south (lx)
Air temperature (˚C)
693
0 0:00 8:00 16:00 0:00 8:00 16:00 0:00 8:00 16:00 0:00 8:00 16:00 0:00
Time external
1st floor circulation
2nd floor circulation
illuminance south
Fig. 17. Measured thermal stratification in the circulation zone (August 12–16, 2002).
694
H. Breesch et al. / Solar Energy 79 (2005) 682–696
used to assess thermal comfort in directives by the Dutch Government Buildings Agency (Rijksgebouwendienst, 1999). The hourly PMVÕs are defined from the simulated air and radiant temperatures, an air velocity of 0.1 m/s, a metabolism of 70 W/m2 and a clothing resistance of 0.7 Clo (Rijksgebouwendienst, 1999). The following equation calculates the internal vapour pressure in Ukkel (Belgium) (Hens, 1993): pi ¼ pi þ ^pi cosð2pðt 210Þ=365:25Þ; pi ¼ 1370 Pa, ^pi ¼ 220 Pa and t = day of year. Figs. 18 and 19 compare the predicted thermal summer comfort on the first and second floor in the four different cases in actual and maximum design occupation respectively. It is shown that in present occupation, natural night ventilation alone or combined with an earth-to-air heat exchanger delivers an excellent thermal comfort. The weighted excess hours equal respectively 6 h and 61 h on the first floor. Maximum occupied, none of the applied passive cooling techniques or combination performs really well. Only the combination of natural night ventilation and an earth-to-air heat exchanger pro-
weighted excess hours (h)
600 500
nightvent + earth-to-air nightvent
400
earth-to-air
300
no passive cooling
5. Conclusions Natural night ventilation and an earth-to-air heat exchanger are applied in the low-energy office building ÔSD WorxÕ in Kortrijk (Belgium). Measurements during summer 2002 show that these passive cooling techniques perform well and create a good thermal summer comfort. Simulations with the coupled thermal and ventilation simulation model TRNSYS-COMIS demonstrate that natural night ventilation is more efficient to improve thermal summer comfort than an earth-to-air heat exchanger in this project.
200 100
References
0 1st floor
2nd floor
Fig. 18. Weighted excess hours from May 15 to September 30 (TRY Ukkel) at present occupation (9 persons).
1600 weighted excess hour s (h)
vides an acceptable thermal comfort with an amount of weighted excess hours somewhat larger than 150 h. An earth-to-air heat exchanger alone and absence of passive cooling techniques perform poorly in both occupancies. The same conclusion may be drawn for the second floor. The impact of a passive cooling technique on thermal summer comfort is defined as the difference between the number of weighted excess hours in absence of passive cooling and with this technique. The impact of natural night ventilation on both floors is in present and maximum occupation respectively 3 and 5 times the impact of the earth-to-air heat exchanger. Two conclusions may be drawn from the simulation results. Firstly, passive cooling has an important impact on thermal summer comfort in the office building ÔSD WorxÕ. Secondly, natural night ventilation improves, on present modeling assumptions, thermal summer comfort much more than an earth-to-air heat exchanger.
1400 nightvent + earth-to-air
1200
nightvent
1000
earth-to-air
800
no passive cooling
600 400 200 0 1st floor
2nd floor
Fig. 19. Weighted excess hours from May 15 to September 30 (TRY Ukkel) at maximum occupation (24 persons).
Argirou, A., 1996. Ground cooling. In: Santamouris, M., Asimakopoulos, D. (Eds.), Passive Cooling of Buildings. James and James, London, pp. 360–403. Blondeau, P., Sperandio, M., Allard, F., 1997. Night ventilation for building cooling in summer. Solar Energy 61, 327– 335. Breesch, H., Janssens, A., 2002. Simulating natural ventilation: coupling thermal and ventilation model. In: Guarracino, G. (Ed.), Proceedings of 3rd European Conference on Energy Performance & Indoor Climate in Buildings and 23rd Conference of the Air Infiltration and Ventilation Centre. Lyon, France, pp. 741–746. Breesch, H., Descheemaeker, K., Janssens, A., Willems, L., 2004. Evaluation of natural ventilation systems in a landscaped office. In: de Wit, M.H. (Ed.), Proceedings of 21st Conference on Passive and Low Energy Architecture (PLEA). Eindhoven, The Netherlands, pp. 157–162. BSRIA, 2004. Report summary: West European market for air conditioning 2003. Available from:.
H. Breesch et al. / Solar Energy 79 (2005) 682–696 Cenergie cvba, 2003. A sustainable office building for SD Worx in Kortrijk (in Dutch). Available from:. De Paepe, K., 2003. Low-energy techniques in office buildings: discussion of the measurements. MSc Thesis, Ghent University, Ghent, Belgium (in Dutch). De Paepe, M., Janssens, A., 2003. Thermo-hydraulic design of earth–air heat exchangers. Energy and Buildings 35, 389– 397. Dorer, V., Haas, A., Weber, A., Feustel, H.E., Smith, B.V. (Eds.), 2001. COMIS 3.1—UserÕs Guide. EMPA, Du¨bendorf, Switzerland, pp. 119–120. EU, 2003. The Kyoto protocol—A brief summary. Available from: . Fanger, P.O., 1972. Thermal Comfort: Analysis and Applications in Environmental Engineering. McGraw-Hill, London, New York. Gauthier, C., Lacroix, M., Bernier, H., 1997. Numerical simulation of soil heat exchanger-storage systems for greenhouses. Solar Energy 60, 333–346. Geros, V., Santamouris, M., Tsangrasoulis, A., Guarracino, G., 1999. Experimental evaluation of night ventilation phenomena. Energy and Buildings 29, 141–154. Givoni, B., 1994. Passive and Low Energy Cooling of Buildings. Van Nostrand Reinhold, New York. Givoni, B., 1998. Effectiveness of mass and night ventilation in lowering the indoor daytime temperatures. Part I: 1993 experimental periods. Energy and Buildings 28, 25–32. Goswami, D.Y., Dhaliwal, A.S., 1985. Heat transfer analysis in environmental control using an underground air tunnel. ASME Journal of Solar Energy Engineering 107, 141–145. Goswami, D.Y., Ileslamlou, S., 1990. Performance analysis of a closed-loop climate control system using underground air tunnel. ASME Journal of Solar Energy Engineering 112, 76–81. Gratia, E., Bruye`re, I., De Herde, A., 2004. How to use natural ventilation to cool narrow office buildings. Energy and Buildings 39, 1157–1170. Haas, A., Weber, A., Dorer, V., Keilholz, W., Pelletret, R., 2002. COMIS v3.1 simulation environment for multizone air flow and pollutant transport modeling. Energy and Buildings 34, 873–882. Heijmans, N., Wouters, P., 2002a. Pilot study report: Probe, Limelette, Belgium. In: Heiselberg, P. (Ed.) Principles of Hybrid Ventilation, Hybrid Ventilation Centre, Aalborg University, Denmark. Available from: . Heijmans, N., Wouters, P., 2002b. Pilot study report: IVEG, Hoboken, Belgium. In: Heiselberg, P. (Ed.) Principles of Hybrid Ventilation, Hybrid Ventilation Centre, Aalborg University, Denmark. Available from: . Hens, H., 1993. Applied Building Physics 1: boundary conditions and performances. Acco, Leuven, Belgium, pp. 28–29 (in Dutch). IEA Annex 28, 1997. Selection guidance for low energy cooling technologies. International Energy Agency, Energy Conservation in Buildings and Community Systems Programme, Annex 28 Low Energy Cooling, Subtask 2 Report 1, United Kingdom.
695
ISSO, 1990. Design of indoor conditions and good thermal comfort in buildings. ISSO Research rapport 5, Rotterdam, The Netherlands (in Dutch). JRAIA, 2003. Estimates of world demand for air conditioners (1999–2007). Available from:. Klein, S.A., Duffle, J.A., Mitchel, J.C., Kummer, J.P., Thornton, J.W., 2000. Trnsys: a Transient System Simulation Program. Solar Energy Laboratory, University of Wisconsin, Madison, USA. Kolokotroni, M., 1995. Night ventilation in commercial buildings. In: Review of Low Energy Cooling Technologies, International Energy Agency, Energy Conservation in Buildings and Community Systems Programme. Annex 28 Low Energy Cooling, Subtask 1, United Kingdom, pp. 7– 11. Kolokotroni, M., Aronis, A., 1999. Cooling-energy reduction in air-conditioned offices by using night ventilation. Applied Energy 63, 241–253. Lechner, N., 2001. Heating, cooling, lighting: design methods for architects. J. Wiley and Sons, New York, pp. 245–278. Martin, A., Fletcher, J., 1996. Night cooling strategies: Final report 11621/4. BSRIA, Berkshire, United Kingdom. Mihalakakou, G., Santamouris, M., Asimakopoulos, D., 1992. Modelling the earth temperature using multiyear measurements. Energy and Buildings 19, 1–9. Mihalakakou, G., Santamouris, M., Asimakopoulos, D., 1994. On the use of ground for heat dissipation. Energy 19, 17– 25. Mihalakakou, G., Lewis, J.O., Santamouris, M., 1996a. The heating potential of buried pipes techniques—application in Ireland. Energy and Buildings 24, 19–26. Mihalakakou, G., Lewis, J.O., Santamouris, M., 1996b. The influence of different ground covers on the heating potential of earth-to-air heat exchangers. Renewable Energy 7, 33– 46. Orme, M., Liddament, M.W., Wilson, A., 1994. Numerical data for air infiltration and natural ventilation calculations. Air Infiltration and Ventilation Centre, Technical note 44, Coventry, United Kingdom. Pfafferott, J., 2003. Evaluation of earth-to-air heat exchangers with a standardised method to calculate energy efficiency. Energy and Buildings 35, 971–983. Pfafferott, J., Herkel, S., Ja¨schke, M., 2003. Design of passive cooling by night ventilation: evaluation of a parametric model and building simulation with measurements. Energy and Buildings 35, 1129–1143. PrEN ISO 13370, 1998. Thermal performance of buildings— Heat transfer via the ground—Calculation methods. CEN, Brussels, Belgium. Rijksgebouwendienst, 1999. Building physical quality of state housing (in Dutch). Available from: . RMI, 2004. Normals for Uccle 1961–1990. Royal Meteorological Institute of Belgium (in Dutch). Available from: . Santamouris, M., 2004. Night ventilation strategies. AIVC Ventilation information paper no. 4, INIVE eeig, Brussels, Belgium. Santamouris, M., Asimakopoulos, D. (Eds.), 1996. Passive Cooling of Buildings. James and James, London, UK.
696
H. Breesch et al. / Solar Energy 79 (2005) 682–696
Santamouris, M., Argirou, A., Dascalaki, E., Balaras, C., Gaglia, A., 1994. Energy characteristics and savings potential in office buildings. Solar Energy 52, 59–66. Santamouris, M., Mihalakakou, G., Balaras, C.A., Argiriou, D., Asimakopoulos, D., Vallindras, M., 1995. Use of burried pipes for energy conservation in cooling of agricultural greenhouses. Solar Energy 55, 111–124. Seem, J.E., 1987. Modeling of heat in buildings. PhD thesis, Solar Energy Laboratory, University of Wisconsin, Madison, USA. Tzaferis, A., Liparakis, D., Santamouris, M., Argiriou, A., 1992. Analysis of the accuracy and sensitivity of 8 models to predict the performance of earth-to-air heat-exchangers. Energy and Buildings 18, 35–43. Van der Linden, K., Boerstra, A.C., Raue, A.K., Kurvers, S.R., 2002. Thermal indoor climate building performance characterized by human comfort response. Energy and Buildings 34, 737–744.
Van Paassen, A.H.C., Liem, S.H., Gro¨ninger B.P., 1998. Control of night cooling with natural ventilation: sensitivity analysis of control strategies and vent openings. In: Proceedings of 19th Annual AIVC Conference. Oslo, Norway, pp. 438–447. Wilkins, P.E., Hosni, M.H., 2000. Heat gain from office equipment. Ashrae Journal 42 (6), 33–44. Zimmermann, M., 1995. Ground cooling with air. In: Review of low energy cooling technologies, International Energy Agency, Energy Conservation in Buildings and Community Systems Programme. Annex 28 Low Energy Cooling, Subtask 1, United Kingdom, pp. 12–16. Zimmermann, M., Remund, S., 1999. Ground coupled air systems. In: Early design guidance for low energy cooling, International Energy Agency, Energy Conservation in Buildings and Community Systems Programme. Annex 28 Low Energy Cooling, Subtask 2, Report 2, United Kingdom (Chapter F).
Passive cooling in a low-energy office building H. Breesch a
a,*
, A. Bossaer b, A. Janssens
a
Buildings and Climatic Control, Department of Architecture and Urban Planning, Ghent University, J. Plateaustraat 22, B-9000 Ghent, Belgium b Cenergie cvba, B-2600 Berchem, Belgium
Received 22 October 2003; received in revised form 19 November 2004; accepted 7 December 2004 Available online 1 January 2005 Communicated by: Associate Editor Matheos Santamouris
Abstract In office buildings, the use of passive cooling techniques combined with a reduced cooling load may result in a good thermal summer comfort and therefore save cooling energy consumption. This is shown in the low-energy office building ÔSD WorxÕ in Kortrijk (Belgium), in which natural night ventilation and an earth-to-air heat exchanger are applied. In winter, the supply air is successively heated by the earth-to-air heat exchanger and the regenerative heat exchanger, which recovers the heat from the exhaust air. In summer, the earth-to-air heat exchanger cools the ventilation air by day. In addition, natural night ventilation cools down the exposed structure which has accumulated the heat of the previous day. In this article the overall thermal comfort in the office building is evaluated by means of measuring and simulation results. Measurements of summer 2002 are discussed and compared to simulations with a coupled thermal and ventilation simulation model TRNSYS-COMIS. The simulations are used to estimate the relative importance of the different techniques. The evaluation shows that passive cooling has an important impact on the thermal summer comfort in the building. Furthermore, natural night ventilation appears to be much more effective than an earth-to-air heat exchanger to improve comfort. 2004 Elsevier Ltd. All rights reserved. Keywords: Natural night ventilation; Earth-to-air heat exchanger; TRNSYS-COMIS
1. Introduction The cooling energy demand has increased in Europe. The demand for room and packaged air conditioners has grown with 20% between 1999 and 2002 (JRAIA, 2003). The annual report summary of BSRIA on the
* Corresponding author. Tel.: +32 9 264 7861; fax: +32 9 264 4185. E-mail address: [email protected] (H. Breesch).
West European room and packaged air conditioning market in 2003 (BSRIA, 2004) confirms these conclusions. On the other hand, the Kyoto protocol binds the developed countries to reduce the collective emissions of six key greenhouse gases—among which CO2—at least by 5% by 2008–2012. This protocol encourages the governments amongst others to improve energy efficiency and to promote renewable energy (EU, 2003). Therefore, counterbalancing the energy and environmental effects of air conditioning is a strong requirement for the future. Passive cooling may contribute to the Kyoto requirements by reducing the need for cooling
0038-092X/$ - see front matter 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.solener.2004.12.002
H. Breesch et al. / Solar Energy 79 (2005) 682–696
683
Nomenclature A CD Cp g h p PMV PPD WF U v
area of the opening (m2) discharge coefficient of the opening (–) wind pressure coefficient (–) solar transmission coefficient (–) height (m) vapour pressure (Pa) predicted mean vote (–) predicted percentage of dissatisfied people (–) hourly weight factor (h) thermal transmittance (W/m2 K) wind velocity (m/s)
energy while providing a good thermal comfort (Santamouris, 2004). Two interesting and promising passive cooling techniques are natural night ventilation and earth-to-air heat exchangers. Intensive natural night ventilation is driven by wind and thermally (stack) generated pressures and cools down the exposed building structure at night. As a consequence, heat may accumulate the next day and temperature peaks will be reduced and postponed consequently. Furthermore, earth-to-air heat exchangers consist of pipes buried underground through which ventilation air is drawn. The thermal inertia of the soil dampens and postpones temperature fluctuations at the surface (Mihalakakou et al., 1992). At a sufficient depth, the ground temperature is sufficiently low in summer to cool and sufficient warm in winter to heat the ventilation air in contact with the tube in the ground (Givoni, 1994; Argirou, 1996; De Paepe and Janssens, 2003). The performances of these passive cooling techniques depend on multiple building and environmental parameters. Firstly, climatic parameters (temperature difference inside-outside, average outdoor temperature range), building characteristics (thermal inertia of the building and convective heat transfer between ventilation air and thermal mass) and technical parameters (ventilation rate by night and control strategy) define the performance of natural night ventilation (Blondeau et al., 1997; Givoni, 1998; Pfafferott et al., 2003). Following performances have been noticed. Daily peak temperatures in office buildings decrease by maximum 2–4 C (Kolokotroni, 1995; Martin and Fletcher, 1996; Geros et al., 1999; Blondeau et al., 1997; Pfafferott et al., 2003). Blondeau et al. (1997) also measured that the average predicted mean vote (PMV) decreases from 1.5 (warm) to 0.75 (slightly warm) when mechanical night ventilation was applied. In addition, Givoni (1994) showed experimentally that in a hot dry climate the excess hours considerably reduce in a high mass
Greek symbols h temperature (C) l mean r standard deviation u relative humidity (%) Subscripts i internal e external a air s surface
building. Furthermore, a yearly reduction of sensible cooling up to 30% was determined in Greece (Santamouris et al., 1994) while a 30% reduction of cooling energy consumption and 40% reduction of installed cooling capacity were calculated in a low energy office building with stack ventilation flow at night of 10 ach in the United Kingdom (Kolokotroni and Aronis, 1999). In addition, Gratia et al. (2004) calculated a 40% reduction of daily cooling demand in a high mass narrow office building with natural night ventilation on a sunny Belgian summer day. Secondly, the performance of an earth-to-air heat exchanger depends on the air flow rate by day, convective heat transfer between the pipe and the ventilation air, depth, dimensions and number of the pipes, soil temperature and soil properties (Goswami and Dhaliwal, 1985; Tzaferis et al., 1992; Givoni, 1994; Mihalakakou et al., 1994, 1996a,b; Santamouris et al., 1995; Gauthier et al., 1997). The performance is generally expressed by a heating and cooling gain (yearly, monthly, peak, sorted by ambient air temperature, usable energy supply) or a peak temperature shift in summer and winter. Other efficiency criteria are discussed in Pfafferott (2003) and Goswami and Ileslamlou (1990). The seasonal energy performance in a moderate climate is estimated at 8–10 kW h/m2 of ground coupling area for cooling and at 10–15 kW h/m2 of ground coupling area for heating. Peak cooling and heating capacity at ambient air temperatures of respectively 32 C and 5 C are estimated at 45 W/m2 of ground coupling area for cooling and heating (IEA Annex 28, 1997). The feasibility of natural night ventilation and an earth-to-air heat exchanger for cooling is primarily limited by climatic conditions and heat gains. Firstly, hot and moderate climates with large diurnal temperature difference over the summer are best suited for natural night ventilation (Kolokotroni, 1995). The minimum recommended diurnal temperature difference is approximately 10–12 C (Givoni, 1994; Santamouris and
684
H. Breesch et al. / Solar Energy 79 (2005) 682–696
Asimakopoulos, 1996; Lechner, 2001). Only a moderate climate having large temperature difference between day and night as well as between summer and winter is suited for an earth-to-air heat exchanger. In a hot climate, the cooling effect of an earth-to-air heat exchanger is limited whereas in a cold climate the cooling demand is small (Zimmermann and Remund, 1999). Secondly, both passive cooling techniques may only satisfy a low or moderate cooling demand. Furthermore, natural night ventilation is only suitable in buildings with sufficient accessible thermal mass (75–100 kg/m2). As a rule of thumb, IEA Annex 28 (1997) and Van Paassen et al. (1998) recommend a maximum of 20–30 W/m2 heat gains in heavy constructions. As an earth-to-air heat exchanger has a good peak performance but a limited seasonal cooling capacity, it is more suited to provide comfort cooling in low-energy buildings which tend to overheat in the summer (Zimmermann, 1995). Furthermore, as an earth-to-air heat exchanger merely precools the air, a combination with natural or mechanical night ventilation or with slab or ceiling cooling is interesting (Zimmermann and Remund, 1999). In Belgium, during last years, passive cooling techniques have been gradually introduced in office buildings (Heijmans and Wouters, 2002a,b; Breesch et al., 2004). The Belgian moderate climate is very suitable for passive cooling because both the external outside air and the ground may be used as a heat sink in summer. A yearly average ambient temperature of 9.7 C and an average daily maximum and minimum temperature in July of respectively 21.8 C and 12.7 C are measured (RMI, 2004) (see also Table 4). Nevertheless, designers still lack confidence to apply these techniques in the absence of information about the performance of existing applications in Belgium. Therefore, this paper discusses the performance of a low-energy office building in Belgium. The paper will focus on the performances of the building in the cooling season, but will briefly examine the performances in the heating season too. The first part describes the low-energy office building and its operation. The second part discusses the measured operation and performance of the building and applied passive heating and cooling systems. The third part focuses on the simulated performances and estimates the impact of the passive cooling techniques on the thermal summer comfort.
2. Low-energy office building The office building ÔSD WorxÕ is located in Kortrijk, Belgium and consists of two office floors on top of a limited ground floor with building services. On the south side, the floors are connected with an open vertical circulation zone. Figs. 1 and 2 show a plan of the first floor and a vertical section. The plan of the second floor is very similar. The office floors and the ground floor have
Fig. 1. Plan of the first floor of SD Worx (Kortrijk, Belgium) with open plan office (no. 1), circulation zone (no. 2), conference rooms (no. 3) and auditorium (no. 4).
a floor area of respectively 497 m2 and 198 m2. The most important rooms on the office floor are the open plan office (no. 1), circulation zone (no. 2), conference rooms (no. 3) and other offices or an auditorium (no. 4). The vertical section indicates the open plan offices (no. 1) and the stairwell (no. 2) on both floors as well as the technical rooms (no. 5) on the ground floor. The principles of this low-energy office building are the reduction of the heating and cooling load, the use of passive cooling and heating and enhanced control automation (Cenergie cvba, 2003). 2.1. Heating season First of all, the transmission heat losses are minimized. The external walls, roof, floor and windows are
H. Breesch et al. / Solar Energy 79 (2005) 682–696
685
Fig. 2. Cross section of SD Worx (Kortrijk, Belgium) with open plan offices (no. 1), circulation zone (no. 2) on first and second floor and building services on ground floor (no. 5).
very well insulated (see Table 1). Further, the ventilation heat loss is reduced by an earth-to-air and a regenerative heat exchanger which (pre)heat the ventilation supply air. In this project the earth-to-air heat exchanger includes two concrete pipes with an internal diameter of 80 cm and a length of 40 m each, buried in clay ground under the groundwater level at a depth of respectively 3 and 5 m and connected to the ventilation system by PE-pipes with an internal diameter of 40 cm (De Paepe, 2003). Thermal properties of the clay soil on site were not measured. Standard values for clay are a thermal conductivity of 1.5 W/(m K) and a volumetric thermal capacity of 3.0 · 106 J/(m3 K) (PrEN ISO 13370, 1998). A regenerative heat exchanger recovers the heat from the exhaust air as follows. Fresh and exhaust air alternately pass through two accumulators including packed aluminium plates. These plates remove the heat from the warm extracted airflow and pass it to the cold supply airflow. Further, demand controlled ventilation reduces the ventilation heat loss. The building is divided in 11 zones, distributed over the offices and meeting rooms on both office floors. When the CO2 concentration exceeds 800 ppm, fresh air is supplied into the zone through a false floor plenum and air diffusers along the external wall. Minimizing the transmission and ventilation heat losses in this office building, resulted in a low energy
consumption during the heating season 2002–2003, i.e. 62 kW h/m2 (Cenergie cvba, 2003). 2.2. Cooling season Following actions are taken to achieve a good thermal summer comfort. Firstly, the cooling load is reduced. Zoning minimizes the solar heat gains because the offices face north and the circulation zone faces south. This glazed south-facing fac¸ade is finished with controllable external sunblinds to achieve solar shading. The heat emission of lighting and office equipment is also minimized. The installed power of the lighting is limited to 9.5 W/m2, presence detection and daylight control are applied. An exposed concrete ceiling additionally delivers a high thermal capacity, which reduces and postpones the cooling load. Secondly, passive cooling is applied. By day, an earth-to-air heat exchanger precools the supply air flow (Fig. 3). By night, outside air enters the office floors through bottom hung windows, located near the ceiling in the offices on the north side. The air cools down the exposed ceiling and leaves the building at the top of the circulation zone through outlet windows along the width of the building (see Fig. 4). This natural night ventilation system is primarily driven by thermally (stack) generated pressures. Thirdly, the use of these passive cooling techniques is
686
H. Breesch et al. / Solar Energy 79 (2005) 682–696
Table 1 Building data Night ventilation building data Zone
Reference height (m)
Dimensions H/D/W (m)
Louvers From
To
h (m)
A (m2)
CD ()
Outside Office 1st Outside Office 2nd Circulation
Office 1st Circulation Office 2nd Circulation Outside
6.0 6.3 9.5 9.8 11.4
5.66 3.06 5.66 3.06 10.65
0.27 0.33 0.27 0.33 0.33
Office 1st floor
3.5
3/12.1/21.2
Office 2nd floor
7
3/12.1/21.2
Circulation zone 3.5 Thermal building data
8.4/2.9/21.2
Wall
U (W/m2 K)
g ()
Wall
U (W/m2 K)
Window Window, horizontal, circulation
1.1 0.6
0.6 0.28
External floor External floor, circulation Internal floor External wall Internal wall Window frame Roof
0.24 0.15 1.61 1.28 1.93 3.5 0.28
Wall composition External wall Internal wall Flat roof Internal and external floor
Reinforced concrete, embedded with brick panes Wooden cupboards Hollow core concrete slabs Hollow core concrete slabs, false floor
Internal heat gains per floor Present
Maximum
People PC + Screen Laserprinter Lighting
9 pers. 9 pc. 1 pc. 235 m2
24 pers. 24 pc. 1 pc. 235 m2
Total
12.9 W/m2
24.2 W/m2
Radiant (%) 80 W/pers. 130 W/pc. 320 W/pc. 9 W/m2
Convective (%)
50 25 20 25 31
Usage diversity
50 75 80 75 33
69
1 0.75 0.5 0.6 67
Fig. 3. Operation scheme of the earth-to-air heat exchanger in SD Worx (Cenergie cvba, 2003).
H. Breesch et al. / Solar Energy 79 (2005) 682–696
687
Fig. 4. Operation scheme of natural night ventilation in SD Worx (Cenergie cvba, 2003).
automatically controlled to meet requirements at different climatic loads. When the temperature in every zone exceeds 23 C during the day, the supply ventilation rate increases proportionally to the mean indoor temperature: from 5400 m3/h at 23 C to 8000 m3/h at 26 C and above. In this situation, the exhaust air leaves the building through the windows on top of the circulation zone. When the internal temperature in a zone is less than 23 C, the ventilation rate in this zone is demand controlled and the ventilation air is mechanically extracted as in the heating season. The operation of natural night ventilation depends on multiple control parameters: the maximum inside and outside air temperature of the day before, the inside air temperature, the difference between inside and outside air temperature, the internal relative humidity, the rainfall and the wind velocity at that moment. Table 2 lists the settings of the control system. The control parameters were optimised after the first summer. While the inlet windows before were opened at insides air temperatures larger than Table 2 Control system natural night ventilation Natural night ventilation is in operation if the conditions below are fulfilled Previous day hi,max > 23 C he,max > 20 C At that moment 22 h < time < 6 h hi > 20 C hi he > 2 C ui < 70% No rainfall v < 10 m/s
20 C, the opening is now controlled by ceiling surface temperatures exceeding 22 C.
3. Measurements: discussion The building operation system monitors continuously (every 15 min) the outdoor and indoor climate, the air flow rate in the mechanical ventilation system and the control parameters. Measurements from May 15 to September 30, 2002 are used to evaluate the performances of natural night ventilation and the earth-to-air heat exchanger in the cooling season. This was the first summer that the building was in operation. The operation and performance of the earth-to-air heat exchanger are also discussed in the heating season on the basis of monitoring data from November 15 until December 20, 2002. 3.1. Natural night ventilation The monitoring (De Paepe, 2003) shows that natural night ventilation was in operation during 60% of the nights from the end of June until the end of September 2002. Fig. 5 shows the outdoor temperature, the operative and surface temperatures on both floors and the operation of natural night ventilation during a warm summer period. Natural night ventilation cooled down the inside air and exposed ceiling, which had stored the heat of the previous day, from 10 p.m. till 6 a.m. As a result, excess heat accumulated in the ceiling the following day (surface temperature increased by day) and the air temperature peaks by day decreased. No humidity or condensation problems were noticed. The performance of natural night ventilation is analysed based on the achieved air and surface temperature
32
10
30
9
28
8
26
7
24
6
22
5
20
4
18
3
16
2
14
1
Night ventilation operation
H. Breesch et al. / Solar Energy 79 (2005) 682–696
Temperature (˚C)
688
0 12 14:00 22:00 6:00 14:00 22:00 6:00 14:00 22:00 6:00 14:00 22:00 6:00 14:00
Time external
1st floor air
1st floor surface
2nd floor surface
2nd floor air night ventilation
Fig. 5. Measured operation of natural night ventilation (August 12–16, 2002).
changer in summer. The difference between these two temperatures defines the sensible cooling effect of the earth-to-air heat exchanger. During the day, when the mechanical ventilation was in operation, the earthto-air heat exchanger reduced and dampened the outside air temperature fluctuations. When the ventilation system did not work, the temperature in the earth-to-air heat exchanger adopted the ground temperature. Fig. 7 relates and compares the daily maximum temperature of the air leaving the earth-to-air heat exchanger to the daily maximum outdoor air temperature. The maximum temperature of the supply air never exceeded 22 C, i.e. an important cooling effect on warm summer days (cf. the cooling effect at the last day on Fig. 6). During days with a maximum external temperature between 12 and 22 C, the cooling effect was limited. A heating effect was noticed when the maximum outside temperature was less than 12 C. Figs. 8 and 9 illustrate the performances of the earthto-air and regenerative heat exchanger in the heating season. Fig. 8 shows that on a winter day the ventilation air was preheated by the earth-to-air and further heated
2.0 0.8 0.7 0.3
1.3 0.6 0.5 0.2
drop overnight (between 10 p.m. and 6 a.m. the next day). These temperature drops were higher on the first than on the second floor because of the higher thermal stack effect (see Table 3). The outside air temperature peak was on average postponed for 5 h. As a result, the indoor air temperature peaks occurred after the office hours. 3.2. Earth-to-air heat exchangers
32 30 28 26 24 22 20 18 16 14 12
10 9 8 7 6 5 4 3 2 1 0 0:00
18:00
6:00
12:00
0:00
18:00
0:00
18:00
6:00
12:00
0:00
18:00
6:00
12:00
cooling effect
0:00
Temperature (˚C)
The measured performance of the earth-to-air heat exchanger during the summer is shown in Figs. 6 and 7. Fig. 6 compares the external air temperature to the air temperature at the outlet of the earth-to-air heat ex-
6:00
Dhs,i (C)
l r l r
Second floor
12:00
Dha,i (C)
First floor
Ventilation operation
Table 3 Natural night ventilation: temperature drop overnight
Time external
earth-to-air
1st floor air
2nd floor air
ventilation
Fig. 6. Measured operation of the earth-to-air heat exchanger in the cooling season (August 11–14, 2002).
perature of the earth-to-air heat exchanger to the daily minimum outdoor temperature during working hours. A heating effect existed on winter days when the minimum external temperature was lower than 8 C. A limited cooling effect was noticed from a daily minimum external temperature of 10 C.
36 32 28
689
cooling effect
24 20
3.3. Thermal summer comfort
16 12
Thermal summer comfort is evaluated by FangersÕ comfort theory (Fanger, 1972). Fanger defined the conditions in which more than 10% of the occupants are dissatisfied. The critical indoor operative temperature meeting this requirement depends on the metabolism and clothing resistance. Assuming seated office work (metabolism of 65 W/m2) and light working clothes (0.7 Clo), a threshold temperature of 26.0 C is found (ISSO, 1990). Fig. 10 shows the cumulative distribution of the measured temperatures on both floors. An excellent thermal summer comfort was reached: 26 C was only exceeded during 3 h (or 0.3% of summer working hours) on the
8 4 0 0
4
8
12
16
20
24
28
32
36
Maximum outside temperature (˚C) Fig. 7. Measured cooling effect of the earth-to-air heat exchanger.
Temperature (˚C)
to the desired inlet temperature by the regenerative heat exchanger. Fig. 9 relates the daily minimum outlet tem-
25
9000
20
7500
15
6000 heating effect regenerative
10
4500 heating effect earth-to-air
5
3000
0
1500
external
earth-to-air
extraction air
1st floor
0:00
20:00
16:00
Time
supply air
12:00
8:00
4:00
0:00
20:00
16:00
8:00
12:00
4:00
0 0:00
-5
Supply air flow (m³/h)
Maximum earth-to-air temperature (˚C)
H. Breesch et al. / Solar Energy 79 (2005) 682–696
flow rate
Minimum earth-to-air temperature (˚C)
Fig. 8. Measured operation of the earth-to-air and the regenerative heat exchanger in the heating season (December 19–20, 2002).
20 18 16 14 12 10 8 6 4 2 0
heating effect
-6 -4 -2 0
2
4
6
8 10 12 14 16 18 20 22 24 26 28 30
Minimum external temperature (˚C)
Fig. 9. Measured heating effect of the earth-to-air heat exchanger.
Cumulative temperature distribution (% of total working hours)
690
H. Breesch et al. / Solar Energy 79 (2005) 682–696 100
91.3
90
1st floor 2nd floor
80
73.0
70 60 48.3
50 40
33.1
30 20 8.2
10 0
0.0 0.0
0.0 0.3
3.0
> 27˚C
> 26˚C
> 25˚C
> 24˚C
> 23˚C
Maximum inside air temperature 2nd floor (˚C)
Fig. 10. Measured cumulative distribution of temperatures on both floors during working hours (May 15 to September 30, 2002).
28
26
24
22
20 10
14
18
22
26
30
34
Maximum outside air temperature (˚C)
Fig. 11. Measured maximum inside temperature on the second floor during working hours as a function of the measured maximum outside temperature (May 15 to September 30, 2002).
second floor and never on the first floor. Fig. 10 also proves that high temperatures were more frequently monitored on the second floor. An inside temperature of 23 C was exceeded during 73% and 91% of the working hours on the first and the second floor respectively. Fig. 11 relates the daily maximum air temperature on the second floor to the daily maximum outdoor air temperature. The maximum air temperature on the second floor averaged 24.6 C. 98% of the maximum indoor temperatures on the second floor are from 23.5 C to 26 C.
4. Simulations: impact of passive cooling 4.1. Simulation model In a naturally ventilated building, free moving internal temperatures depend on the ventilation flow rates. Because natural (night) ventilation is driven by wind and thermally (stack) generated pressures, the ventila-
tion rate on its turn is function of the indoor air temperature. Therefore, a coupled thermal and ventilation simulation model, which iterates the mass and energy balance per zone till convergence, is necessary to simulate natural night ventilation (Breesch and Janssens, 2002). The office building ÔSD WorxÕ has been simulated by linking TRNSYS (Klein et al., 2000), a transient multi-zone thermal simulation model, to COMIS (Dorer et al., 2001; Haas et al., 2002), a multi-zone infiltration and ventilation simulation model. Both simulation programs subdivide the building in various zones, mostly corresponding to the rooms, in which the air is assumed to be perfectly mixed. In TRNSYS, a zone is represented by two temperatures: the homogeneous air temperature and the so-called star temperature (Seem, 1987). The star temperature is a weighted average of the zone air temperature and the surface temperatures of the walls surrounding the zone. The air temperature is solved from the convective heat flow balance of the zone, the star temperature is solved from the combined convective and radiation heat flow balance. The star temperature concept is introduced to facilitate the calculation of conduction heat loss over a wall. This is modelled according to transfer function relationships (Klein et al., 2000). These relationships describe the heat flux at one side of the wall varying in time, depending on a change of the heat flux at the other side of the wall or on a change of the surface temperature at one of both sides of the wall. In other words, they tell the Ôthermal historyÕ of a wall. In COMIS, each zone is represented by single values for air temperature and pressure. Air flow paths like windows, louvers, cracks and ducts, connect the zones to each other and to the outdoors. The flow rates are related to the pressure difference over the flow path by nonlinear equations. Wind pressure coefficients, relating the wind pressure at a building to the wind velocity, are attributed to external nodes. The zone pressures are solved from the system of mass flow balances. Given these pressures, the flow rates in each air flow path may also be calculated (Haas et al., 2002). The simulation model simplifies the building geometry and includes two landscape office zones and the circulation zone connecting both floors. Table 1 summarizes the characteristics of the night ventilation openings, the composition and the thermal transmittance properties of the walls and the internal heat gains of each floor. The earth-to-air heat exchanger is simplified and modeled as shown in Fig. 7. Since the aim of the simulations is to assess thermal comfort in summer, the regenerative heat exchanger is not taken into account. For an adequate simulation of the control system, a small time step of 6 min is chosen. Wind pressure coefficients are shown in Fig. 12 (Orme et al., 1994). To transform the wind velocity from the meteorological to the local site, a roughness height on site of 0.25 m (i.e. countryside
H. Breesch et al. / Solar Energy 79 (2005) 682–696
Table 4 Comparison outdoor temperature on site and TRY Ukkel (May 15 till September 30)
0.4 office grills
Cp (-)
0.2
691
flat roof
0 0
45
90
135
180
Daily average temperature (C)
-0.2
Daily temperature amplitude (C) -0.4
Daily maximum temperature (C) -0.6
l r l r l r
SD Worx 2002
TRY Ukkel
17.5 3.1 8.9 3.3 22.0 4.1
16.2 3.1 9.0 3.0 20.7 3.9
angle (˚)
Fig. 12. Wind surface pressure coefficient Cp varies with the angle between the wind direction and the normal on the surface (Orme et al., 1994).
and spread habitat) is assumed (Dorer et al., 2001). The open plan offices are designed for an occupation of 24 persons on each floor. At the moment of the measurements, nine persons were working on each floor from 8 h till 17 h. The heat gains of computers, monitors, printers and lighting are multiplied by a diversity factor. This factor takes into account that not all equipment is in use all the time or is throwing out its actual peak heat gain (see Table 1) (Wilkins and Hosni, 2000). Solar irradiation and wind direction data in situ are not available. Consequently, the weather data of the Test Reference Year (TRY) of Ukkel (Belgium) are used in the simulations to estimate the impact of passive cooling on thermal comfort. Fig. 13 compares the cumulative distributions of the daily average temperatures in both weather data. Table 4 reports the summer average and standard deviation of the daily average and maxi-
mum temperature and of the daily temperature amplitude. The summer period on site is on average slightly warmer than in the TRY and contains a larger amount of high temperatures. Fig. 14 additionally compares the cumulative distributions of the daily temperature amplitude. In conclusion, the average and the distribution of the daily temperature amplitude are very similar in both climatic data (see also Table 4). 4.2. Comparison to measurements Before discussing the impact of passive cooling on thermal summer comfort in ÔSD WorxÕ, the simulation results are first compared to the measurements. Fig. 15 shows the relationship between the simulated maximum indoor operative temperatures and the outdoor air temperatures. The simulated cumulative temperature distribution is shown in Fig. 16. These figures may respectively be compared to the measured temperatures in Figs. 11 and 10. The measured and simulated relationship between maximum indoor and outdoor temperatures (see Figs.
100 SD Worx 2002 TRY Ukkel
90
cumulative frequency (%)
80 70 60 50 40 30 20 10 0 12
14
16
18
20
22
24
26
28
Daily average temperature (˚C) Fig. 13. Weather data (May 15 to September 30): comparison cumulative distribution of daily average temperatures on site—Test Reference Year Ukkel.
692
H. Breesch et al. / Solar Energy 79 (2005) 682–696 100 SD Worx 2002 TRY Ukkel
90
cumulative frequency (%)
80 70 60 50 40 30 20 10 0 4
6
8
10
12
14
16
18
Daily temperature amplitude (˚C) Fig. 14. Weather data (May 15 to September 30): comparison cumulative distribution of daily temperature amplitude on site—Test Reference Year Ukkel.
Maximum inside operative temperature 2nd floor (˚C)
28
26
24
22
20
10
14 18 22 26 30 Maximum outside air temperature (˚C)
34
Cumulative air temperature distribution (% of total working hours)
Fig. 15. Simulated maximum operative temperature on the second floor as a function of the maximum outside air temperature from May 15 to September 30 (TRY Ukkel).
100 90 80
1st floor 2nd floor
70
63.3 64.6
60 50 37.3 37.2
40 30 20 9.1 7.5
10 0.0 0.0
0.7 0.7
> 27˚C
> 26˚C
0 > 25˚C
> 24˚C
> 23˚C
Fig. 16. Simulated cumulative distribution of air temperatures on both floors during working hours from May 15 to September 30 (TRY Ukkel).
H. Breesch et al. / Solar Energy 79 (2005) 682–696
4.3. Impact of passive cooling
11 and 15) show the same trend at outside temperatures above 18 C, i.e. when natural night ventilation was in operation. It started working from a maximum outside air temperature of 20 C (see Table 2). On the other hand, at outside temperatures below 18 C, the simulations underestimate the actual temperature peaks in the offices. These temperatures typically occurred during springtime, when the night ventilation system was out of action. Clearly, the difference between simulations and measurements may be attributed to the fact that the simulations do not account for the heating exchange between supply and extract air. Regardless of the fact that the boundary conditions used in the simulations differ from the actual climate, following difference between Figs. 10 and 16 is striking. Contrary to the measurements, the simulated temperatures on both floors hardly differ. Furthermore, the simulation model respectively overestimates and underestimates the operative temperature peaks (hi > 25 C) somewhat on the first and second floor. This difference may be explained by following limitations and assumptions in the multi-zone model. Firstly, an adiabatic floor is assumed between the first and the ground floor with services in the simulation model, as the temperature is not know on this latter floor. Due to this, the transmission losses of the first floor are neglected. Secondly, the vertical circulation zone is represented by one homogeneous temperature in the multi-zone model. In reality, thermal stratification occurred: a temperature difference of 1–1.5 C was measured between the second and the first floor in the circulation zone on a sunny summer day (see Fig. 17). In conclusion, the comparison shows that the simulation model is adequate to predict the effect of passive cooling on thermal summer comfort i.e. when the daily maximum external temperature exceeds 18 C.
The impact of natural night ventilation and the earthto-air heat exchanger on thermal summer comfort in the office building ÔSD WorxÕ is estimated by comparing the reached thermal summer comfort in following cases: the low-energy office building without passive cooling, with only an earth-to-air heat exchanger, with only natural night ventilation and with the combination of both techniques. Thermal summer comfort is characterised by the weighted excess hours during occupation time. A number of weighted working hours in which more than 10% of the occupants are dissatisfied (predicted mean vote or PMV > 0.5) less than 150 h, means a good thermal summer comfort (Van der Linden et al., 2002). Determination of the weighted excess hours is based on the comfort theory of Fanger (1972), which takes account of indoor environmental parameters (air and radiant temperature, air velocity and relative humidity) and personal properties (metabolism, activity level and clothing). The method of weighted excess hours takes account of the degree of discomfort by means of an hourly weight factor (WF). WF is directly proportional to the increase of the predicted percentage of dissatisfied people or PPD: 1 h with 20% dissatisfied people counts for 2 h with 10% dissatisfied. A PMV of 0.5 corresponds to a WF of 1. if PMV < 0:5 then WF ¼ 0 else WF ¼ 10 9:5 exp½ð0:03353PMV4 þ 0:2179PMV2 Þ:
The threshold value of 150 weighted excess hours corresponds to an average of 15% dissatisfied people during 5% of the annual occupation time. This method has been
32
200000
30
180000
28
160000
26
140000
24
120000
22
100000
20
80000
18
60000
16
40000
14
20000
12
Illuminance south (lx)
Air temperature (˚C)
693
0 0:00 8:00 16:00 0:00 8:00 16:00 0:00 8:00 16:00 0:00 8:00 16:00 0:00
Time external
1st floor circulation
2nd floor circulation
illuminance south
Fig. 17. Measured thermal stratification in the circulation zone (August 12–16, 2002).
694
H. Breesch et al. / Solar Energy 79 (2005) 682–696
used to assess thermal comfort in directives by the Dutch Government Buildings Agency (Rijksgebouwendienst, 1999). The hourly PMVÕs are defined from the simulated air and radiant temperatures, an air velocity of 0.1 m/s, a metabolism of 70 W/m2 and a clothing resistance of 0.7 Clo (Rijksgebouwendienst, 1999). The following equation calculates the internal vapour pressure in Ukkel (Belgium) (Hens, 1993): pi ¼ pi þ ^pi cosð2pðt 210Þ=365:25Þ; pi ¼ 1370 Pa, ^pi ¼ 220 Pa and t = day of year. Figs. 18 and 19 compare the predicted thermal summer comfort on the first and second floor in the four different cases in actual and maximum design occupation respectively. It is shown that in present occupation, natural night ventilation alone or combined with an earth-to-air heat exchanger delivers an excellent thermal comfort. The weighted excess hours equal respectively 6 h and 61 h on the first floor. Maximum occupied, none of the applied passive cooling techniques or combination performs really well. Only the combination of natural night ventilation and an earth-to-air heat exchanger pro-
weighted excess hours (h)
600 500
nightvent + earth-to-air nightvent
400
earth-to-air
300
no passive cooling
5. Conclusions Natural night ventilation and an earth-to-air heat exchanger are applied in the low-energy office building ÔSD WorxÕ in Kortrijk (Belgium). Measurements during summer 2002 show that these passive cooling techniques perform well and create a good thermal summer comfort. Simulations with the coupled thermal and ventilation simulation model TRNSYS-COMIS demonstrate that natural night ventilation is more efficient to improve thermal summer comfort than an earth-to-air heat exchanger in this project.
200 100
References
0 1st floor
2nd floor
Fig. 18. Weighted excess hours from May 15 to September 30 (TRY Ukkel) at present occupation (9 persons).
1600 weighted excess hour s (h)
vides an acceptable thermal comfort with an amount of weighted excess hours somewhat larger than 150 h. An earth-to-air heat exchanger alone and absence of passive cooling techniques perform poorly in both occupancies. The same conclusion may be drawn for the second floor. The impact of a passive cooling technique on thermal summer comfort is defined as the difference between the number of weighted excess hours in absence of passive cooling and with this technique. The impact of natural night ventilation on both floors is in present and maximum occupation respectively 3 and 5 times the impact of the earth-to-air heat exchanger. Two conclusions may be drawn from the simulation results. Firstly, passive cooling has an important impact on thermal summer comfort in the office building ÔSD WorxÕ. Secondly, natural night ventilation improves, on present modeling assumptions, thermal summer comfort much more than an earth-to-air heat exchanger.
1400 nightvent + earth-to-air
1200
nightvent
1000
earth-to-air
800
no passive cooling
600 400 200 0 1st floor
2nd floor
Fig. 19. Weighted excess hours from May 15 to September 30 (TRY Ukkel) at maximum occupation (24 persons).
Argirou, A., 1996. Ground cooling. In: Santamouris, M., Asimakopoulos, D. (Eds.), Passive Cooling of Buildings. James and James, London, pp. 360–403. Blondeau, P., Sperandio, M., Allard, F., 1997. Night ventilation for building cooling in summer. Solar Energy 61, 327– 335. Breesch, H., Janssens, A., 2002. Simulating natural ventilation: coupling thermal and ventilation model. In: Guarracino, G. (Ed.), Proceedings of 3rd European Conference on Energy Performance & Indoor Climate in Buildings and 23rd Conference of the Air Infiltration and Ventilation Centre. Lyon, France, pp. 741–746. Breesch, H., Descheemaeker, K., Janssens, A., Willems, L., 2004. Evaluation of natural ventilation systems in a landscaped office. In: de Wit, M.H. (Ed.), Proceedings of 21st Conference on Passive and Low Energy Architecture (PLEA). Eindhoven, The Netherlands, pp. 157–162. BSRIA, 2004. Report summary: West European market for air conditioning 2003. Available from:
H. Breesch et al. / Solar Energy 79 (2005) 682–696 Cenergie cvba, 2003. A sustainable office building for SD Worx in Kortrijk (in Dutch). Available from:
695
ISSO, 1990. Design of indoor conditions and good thermal comfort in buildings. ISSO Research rapport 5, Rotterdam, The Netherlands (in Dutch). JRAIA, 2003. Estimates of world demand for air conditioners (1999–2007). Available from:
696
H. Breesch et al. / Solar Energy 79 (2005) 682–696
Santamouris, M., Argirou, A., Dascalaki, E., Balaras, C., Gaglia, A., 1994. Energy characteristics and savings potential in office buildings. Solar Energy 52, 59–66. Santamouris, M., Mihalakakou, G., Balaras, C.A., Argiriou, D., Asimakopoulos, D., Vallindras, M., 1995. Use of burried pipes for energy conservation in cooling of agricultural greenhouses. Solar Energy 55, 111–124. Seem, J.E., 1987. Modeling of heat in buildings. PhD thesis, Solar Energy Laboratory, University of Wisconsin, Madison, USA. Tzaferis, A., Liparakis, D., Santamouris, M., Argiriou, A., 1992. Analysis of the accuracy and sensitivity of 8 models to predict the performance of earth-to-air heat-exchangers. Energy and Buildings 18, 35–43. Van der Linden, K., Boerstra, A.C., Raue, A.K., Kurvers, S.R., 2002. Thermal indoor climate building performance characterized by human comfort response. Energy and Buildings 34, 737–744.
Van Paassen, A.H.C., Liem, S.H., Gro¨ninger B.P., 1998. Control of night cooling with natural ventilation: sensitivity analysis of control strategies and vent openings. In: Proceedings of 19th Annual AIVC Conference. Oslo, Norway, pp. 438–447. Wilkins, P.E., Hosni, M.H., 2000. Heat gain from office equipment. Ashrae Journal 42 (6), 33–44. Zimmermann, M., 1995. Ground cooling with air. In: Review of low energy cooling technologies, International Energy Agency, Energy Conservation in Buildings and Community Systems Programme. Annex 28 Low Energy Cooling, Subtask 1, United Kingdom, pp. 12–16. Zimmermann, M., Remund, S., 1999. Ground coupled air systems. In: Early design guidance for low energy cooling, International Energy Agency, Energy Conservation in Buildings and Community Systems Programme. Annex 28 Low Energy Cooling, Subtask 2, Report 2, United Kingdom (Chapter F).