- Email: [email protected]
Performance of unglazed solar ventilation air pre-heaters for broiler barns
Available online at www.sciencedirect.com
Solar Energy 85 (2011) 1418–1429 www.elsevier.com/locate/solener
Performance of unglazed solar ventilation air pre-heaters for broiler barns Se´bastien Cordeau, Suzelle Barrington ⇑ Department of Bioresource Engineering, Macdonald Campus of McGill University, 21 111 Lakeshore, Ste Anne de Bellevue, Quebec, Canada H9X 3V9 Received 24 February 2010; received in revised form 24 December 2010; accepted 28 March 2011 Available online 5 May 2011 Communicated by: Associate Editor C. Estrada-Gasca
Abstract Solar radiation is an interesting heat source for applications requiring a limited amount of energy, such as pre-heating cold fresh air used in venting livestock barns. The objective of this study was to evaluate the energy recovery efficiency of a solar air pre-heater consisting of an unglazed perforated black corrugated siding where the incoming fresh ventilation air picks up heat from its face and back. Installed on the southeast wall of two broiler barns located 40 km east of Montreal, Canada, the performance of solar air pre-heaters was monitored over 2 years. Sensors inside the barns monitored the temperature of the ambient air, that pre-heated by the solar collector and that exhausted by one of the three operating fans. An on-site weather station measured ambient air temperature, wind direction and velocity and radiation energy absorbed on a vertical plane parallel to the unglazed solar air pre-heaters. The measured vertical solar radiation value was used to evaluate the heat recovery efficiency of the unglazed solar air pre-heaters. Using data from the Varennes Environment Canada weather station located 30 km northwest, the solar sensors were found to measure the absorbed solar radiation with a maximum error of 7%, including differences in exterior air moisture. Unglazed, the efficiency of the solar air pre-heaters reached 65% for wind velocities under 2 m/s, but dropped below 25% for wind velocities exceeding 7 m/s. Nevertheless, the unglazed solar air pre-heaters were able to reduce the heating load especially in March of both years. Over a period starting in November and ending in March, the solar air heaters recovered an energy value equivalent to an annual return on investment of 4.7%. Ó 2011 Elsevier Ltd. All rights reserved. Keywords: Solar energy; Recovery efficiency; Ventilation; Heating
1. Introduction Within a concept of sustainability, solar radiation is an interesting heat source for applications requiring a limited amount of energy. Despite the lack of research during the 1990s, the twenty-first century brought some renewed interest in solar energy to reduce the use of petro-fuels and their greenhouse gas generation. In agriculture, solar energy can provide heat for a wide range of applications (Arinze et al., 1993; Beshada et al., 2006; Fuller, 2007), such as the ventilation system of livestock buildings requiring heating to maintain a high ambient temperature with young animals. ⇑ Corresponding author. Tel.: +1 514 398 7776; fax: +1 514 398 8387.
E-mail address: [email protected] (S. Barrington). 0038-092X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.solener.2011.03.026
To reduce the heat load of buildings, solar systems have evolved. Glazed solar plates are the oldest and simplest form of solar heaters consisting of a black surface covered by a transparent glazing, and installed on a vertical wall or roof exposed to sunlight. A south wall covered by a black metal sheet protected by a fiberglass glazing reduced by 50% the heat load of a swine gestation barn in Saskatoon, Canada (Sokhansanj and Schoenau, 1991). While the solar collector had a short payback period of 5 years, adding a solar heat storage system was not found economical. In the late 1980s, the concept of unglazed transpired solar plate was developed to further shorten the payback period of solar heating systems. Such solar plates consist of an unprotected or unglazed but perforated black metallic surface placed over a plenum on a southern surface. Eliminating
S. Cordeau, S. Barrington / Solar Energy 85 (2011) 1418–1429
1419
Nomenclature A, B and C are coefficients changing with n and used to calculate the absorbed radiation, dimensionless Aw area of solar air pre-heaters of 73.65 m2 /floor Cp heat capacity of dry air or 1.006 kJ/kg dry air/°C CV coefficient of variation Ed diffuse solar irradiance (W/m2) Er ground reflected solar irradiance (W/m2) EDN absorbed direct solar irradiance (W/m2) ET equation of time in decimal minutes H solar time before or after noon divided by 15° of arc/h, where for example H is 0° of arc at noon Hr sunrise in hour angle about noon Hs sunset in hour angle about noon I incident radiation (W/m2) L latitude of the location in degree of arc LON longitude of the position in decimal degree of arc LSM local meridian time in decimal degree of arc LST local standard time in decimal hour n day f the year, where January 1st is n = 1 and December 31st is n = 365 Q ventilation rate offered by the ventilation system in m3/s R earth’s rotational speed of 15° of arc/h Ra absorbed solar radiation in (W/m2)
the transparent glazing reduces the cost of the collector (Gunnewiek et al., 1996). Creating an interior negative pressure, the building ventilation system is used to pull exterior air through the perforations of the black metallic sheeting, and then over the back surface of the solar plate before entering the building (Carpenter and Kokko, 1991). Although the solar collector is unprotected, its perforations allow the ventilation air to collect heat lost by convection at its surface (Kutscher et al., 1993). Air flow at the back of the solar plate is also important in recovering solar energy (Gunnewiek et al., 1996). Nevertheless, the efficiency of such a solar system depends heavily on its design influencing surface air flow distribution. Flow reversal is one of the main issues, where heated air circulating behind the solar plate will exit through the perforations (Gunnewiek et al., 1996). Wind can have a substantial effect on solar collector efficiency by reducing convective heat loss recovery and increasing flow reversal. Computational Fluid Dynamics models (CFD) were used to establish unglazed solar collector configurations to optimize heat recovery and minimize flow reversal (Gunnewiek et al., 1996; Arulanandam et al., 2000; Augustus Leon and Kumar, 2007; Stojanovic et al., 2011). To counter balance wind effect, the ventilation system must create a minimum air flow rate on the collector surface which varies with building and solar collector configuration, and wind direction (Gunnewiek et al., 2002). Nevertheless, this minimum air
Rs
Tt To WS y b U d 1 e h qg qa W
absorbed radiation measured by the vertical sensor representing the potential heat which can be transferred to the cold fresh air by the solar air pre-heaters with an area of 73.65 m2/floor air temperature at the ventilation inlet (°C) outside air temperature in °C wind velocity (m/s) coefficient used to describe the diffused absorbed radiation, dimensionless sun’s altitude in degree of arc sun’s azimuth measured from the south in degree of arc declination of the earth with respect to the sun in degree of arc energy recovery by the cold fresh air circulating over the solar air pre-heaters (kW) solar air pre-heater efficiency (%) angle of incidence of the solar radiation on the absorbing surface degree coefficient of ground reflection generally equal to 0.2 density of dry air for the temperature measured at the mouth of the fan (kg/m3) is the orientation of the vertical wall receiving the radiation and with respect to the south arc degree
flow rate reduces the surface of the solar collector and accordingly, the amount of solar energy recovered. Recently, a modified version of the unglazed solar plate collector was proposed, namely the jet impinged unglazed solar collector. This solar system consists of a non perforated and unglazed metallic solar plate installed over a plenum enclosed by a perforated back wall allowing air jets to hit the back of the solar collector at a perpendicular angle, thus increasing heat recovery (Belusko et al., 2008). Although several design monograms were produced to increase the efficiency of unglazed solar collectors, field applications can be quite challenging where building air ventilation varies with seasons. A 15 to 45% energy recovery was measured for unglazed solar air pre-heaters installed over the ventilating inlets of a swine nursery in the Quebec City area, Canada (Godbout et al., 2004). To further evaluate the agricultural potential of such unglazed solar air pre-heaters, the present paper measured their efficiency in recovering solar radiation and reducing the heating cost of two broiler barns located 40 km east of Montreal, Canada. Each barn had three floors offering an area of 535 m2/floor. Solar air pre-heaters covered a vertical surface of 73.65 m2/floor on the southeast wall for a total of 221 m2/barn. Besides a horizontal radiation sensor, a vertical radiation sensor was positioned parallel to the solar collectors to measure the absorbed solar
1420
S. Cordeau, S. Barrington / Solar Energy 85 (2011) 1418–1429
radiation. Radiation measurements were validated against the readings of the Varennes Environment Canada weather station located 30 km northwest and theoretical values. For the winters of 2007–08 and 2008–09, the efficiency of the solar air pre-heaters was measured by comparing the heat recovered by the incoming ventilation air and the solar radiation absorbed by the vertical sensor parallel to the collectors. The savings in heating was also observed by monitoring the operation time of natural gas heaters on each floor. 2. Experimental set-up and methodology 2.1. System description The unglazed solar air pre-heaters consisted of a wooden box frame built over the fresh air inlets of all 3 floors of both experimental broiler barns (Fig. 1a and b). Its unprotected facing of black corrugated metal sheeting had small perforations covering 1% of its surface through which fresh ventilation air entered the barn. The unglazed solar air pre-heaters could be bypassed by opening the bottom portion of the wooden box frame. The unglazed solar air heaters were installed on the southeast face of the experimental buildings at an angle of 50° from the south. Each 3 floors of the two experimental broiler barns had a solar air pre-heaters surface measuring 1.47 m in height and 50.1 m in length, for 73.6 m2 /floor or 221 m2/barn.
Inside
Outside
Fixed panel
Air inlet temperature sensor
Panel automatically adjusted with fan air flow for a fresh air speed of 4 to 5 m/s
Unglazed solar air pre-heater with 1% perforation
Cold fresh air entering through perforation Barn wall Bypass panel to shut off solar air pre-heater
Fig. 1a. Configuration of the experimental unglazed solar air pre-heaters of corrugated black metal siding.
Fig. 1b. Unglazed solar air heaters over the fresh air inlets (not visible) and the bottom by-pass door.
2.2. Experimental broiler barns In this experiment, the two identically built broiler barns were located in St Jean Baptiste, 40 km east of Montreal, Canada, at 45.5° latitude where the LSM was 75.0° at normal eastern time, and the LO was 73.1°. The barns faced the southeast at an angle w of 50° from the south, and measured 9.2 m in width, 9.0 m in total wall height and 61 m in total length where 57.9 m was occupied by the broilers. All 3 floors of both experimental barns were equipped with two natural gas heaters (Fig. 2), each generating 30 kW of net sensible heat. To conserve energy and minimize cost, the heater exhaust was discharged inside the barn, a practice not recommended because of the high CO2 levels created. Such practice had no observable effect on broiler bird performance (Cordeau and Barrington, 2010). Each heater was installed at a height of 1.5 m at both ends of the floors, and on the fresh air inlet side. The heat generated was pushed towards the centre of the floor by the heater fan. Each barn floor had a capacity of 6500 birds raised in batches during 35–45 days, from 0.035 to 2.5 kg in body mass. During the cold season and with the broilers freshly introduced in the barns, only two 400 mm fans were operated/floor at a minimum speed of 55%. The speed of these exhaust fans was increased by an automated control governed by the temperature measured using three thermocouples located in the centre but distributed over the room length (Fig. 2) at a height of 0.3 m. The speed of the 400 mm fans was increased from 55 to 100% over a temperature increment of 1.4 °C. The temperature set-point started at 32 °C and dropped by 0.27 °C/day until it reached 24 °C. After one to three weeks, a third 400 mm fan was placed in operation and all three fans handled the fresh air ventilation rate until the end of the batch. At 55 and 100% of their full speed, each 400 mm fan delivered 0.43 m3/s (CV = 14.9%) and 0.94 m3/s (CV = 7.6%), measured using a Balometer (ALNOR EBT721 Balometer, Huntingdon Beech, CA, USA).
S. Cordeau, S. Barrington / Solar Energy 85 (2011) 1418–1429
1421
Fans to exhaust air
Data Logger ▓ ▪ T2
Heater ▪ T3
N T4
◄
▪ T1 Heater
▪ T6 Incoming cold fresh air
▪ T5 Solar air heaters
Fig. 2. Experimental barn instrumentation: temperature sensors T1, T2, T3 for inside air at 0.3 m off the floor; T5 and T6 – inside inlet air, and; T4 – inside outgoing air.
2.3. Instrumentation The site was instrumented to monitor exterior climatic conditions and interior air quality on all 3 floors of both experimental barns. To monitor exterior climatic conditions, a weather station was installed on a tower 3.0 m high located in a grassed area mid way between the two broiler barns. This weather station recorded the ambient climatic conditions: wind velocity and direction, ambient air temperature, barometric pressure and relative humidity, and absorbed radiation. A Hobo data logger recorded all weather data reading every 5 min and was downloaded monthly onto a portable computer. Two solar radiation sensors (Model S-LIB Hobo Silicon Pyranometer Smart Sensor, Onset Computer Corporation, Pocasset, MA, USA) were placed in a horizontal and vertical position, to respectively measure the horizontally absorbed radiation and the radiation adsorbed by the solar air pre-heaters. The vertical radiation sensor was placed parallel to the solar air pre-heaters using a compass. According to the manufacturer, the Hobo silicon pyranometer radiation sensors measure absorbed radiation with a maximum error of ±10.0 W/m2 or ±5%, whichever is greater in sunlight. An additional error of 0.38 W/m2/°C can be associated with temperatures above 25 °C. The HOBO silicon pyranometer radiation sensors use silicon photodiodes to measure solar power with a spectral range of 300–1100 nm. Nevertheless, this range introduces a negligible error when the sensors are calibrated to measure natural sunlight (Onset Computer Corporation, Pocasset, MA USA). The accuracy of the vertical radiation sensor was verified by comparing its reading to theoretical values (Eq. (15)), while that of the horizontal sensor was verified against readings recorded by the Varennes Environment Canada weather station (Latitude of 45.7°) located 30 km northwest. The air quality on all floors of both experimental barns was recorded every 5 min. The inside air temperature of each floor was measured at three positions located a height of 0.3 m (T1, T2 and T3), next to the thermocouples controlling the ventilation system (Fig. 2). The incoming fresh
air temperature was measured at two locations inside the fresh air inlets (T5 and T6). One more temperature sensor was installed at one of the 400 mm diameter exhaust fan in continuous operation (T4). All temperature sensors were Smart Set Hobo sensors (Onset Computer Corporation, Pocasset, MA, USA). The ventilation rate was monitored by recording the rpm of one 400 mm fan/floor using an electromagnetic converter (Heber et al., 2001). The air flow rate was computed from a correlation obtained using a Balometer (ALNOR EBT721 Balometer, Huntingdon Beech, CA, USA) during which the volumetric air displacement of all 400 mm fans was measured at rpms varying from 55 to 100%. The cumulated operating time of the natural gas heaters was recorded every 5 min using a 5 V AC adapter detecting gas heater operation and sending a signal to the data logger to record operating time. 2.4. Methodology The main objective of the study consisted in measuring the heat recovered by the solar air pre-heaters and thus, the savings in heating load during the two consecutive winters, namely from November 2007 to March 2008 and November 2008 to March 2009 inclusively. The efficiency of the unglazed solar air pre-heaters was monitored continuously by comparing the measured absorbed vertical radiation to the solar heat recovered by the cold fresh air at the ventilation inlets (ASHRAE, 2009; Esmay and Dixon, 1986). This heat recovery was computed from the difference in monitored exterior and fresh air inlet temperatures multiplied by the air flow rate of the ventilation system 1 ¼ ½ðT t T o ÞCp Q qa 1 e ¼ 100 ðRs AwÞ
ð1Þ ð2Þ
where 1 is the energy recovery by the cold fresh air circulating over the solar air pre-heaters in kW, e is the solar air
1422
S. Cordeau, S. Barrington / Solar Energy 85 (2011) 1418–1429
pre-heater efficiency in%, Tt is the air temperature at the ventilation inlet in °C, To is the outside air temperature in °C, Cp is the heat capacity of dry air or 1.006 kJ/kg dry air/°C, Q is the ventilation rate offered by the ventilation system in m3/s, qa is the density of dry air for the temperature measured at the exhaust fan in kg/m3, Rs is the absorbed radiation measured by the vertical sensor representing the potential heat which can be transferred to the cold fresh air by the solar air pre-heaters in kW, and Aw is the area of solar air pre-heaters of 73.65 m2/floor. The monitoring of the unglazed solar air pre-heaters started upon reception of each batch of young birds and terminated when the fully grown birds were sent for slaughter, in an all-in all-out fashion. The acquired data was downloaded onto a portable computer between bird batches. Infra-red pictures were also taken of the solar air pre-heaters from outside to visualize the uniformity in solar heat recovery. 2.5. Statistical analysis This project measured the efficiency of the solar air preheaters in recovering solar radiation and reducing the heating cost of the two experimental broiler barns. The impact on solar air heater efficiency was therefore analyzed as a function of various climatic factors such as solar radiation level, wind speed and direction and exterior temperature. To obtain a relationship between these climatic factors and solar heat recovery, regression equations were obtained using Excel (Microsoft Office 2007, Seattle, USA). Furthermore, it was observed that cloudy periods produced a wide variation in efficiency often exceeding 100%, because the unglazed solar air pre-heaters accumulated exchangeable heat despite the drop in incident radiation. Thus, regression equations were produced for low and high radiation levels corresponding to cloudy and sunny days. 3. Calculation of incident solar radiation The accuracy of the horizontal solar sensor was verified using values measured by the Varennes Environment Canada weather station, while the accuracy of the vertical sensor was verified against calculated value for a plane with the same orientation. The incident solar radiation received by a plane on the surface of the Earth can be computed as a function of its solar orientation and position. This plane will perceive direct, indirect and diffused radiation, all of which can be calculated based on the following simplified method (ASHRAE, 2009). This method was used to validate the measured absorbed radiation by a radiation sensor positioned vertically and parallel to the experimental barn walls on which the solar air pre-heaters were installed. Because of time zones, the actual time must be converted to solar time (H) in hour angle, where the sun is positioned directly south at noon
ET H ¼ R LST þ 12 þ LSM LON 60
ð3Þ
where LST is the local standard time in decimal hours, ET is the equation of time in decimal minutes, LSM is the local meridian time in decimal degree of arc, LON is the local longitude in decimal degree of arc and R is the earth’s rotational velocity of 15° of arc degree. The hour angle time H is simply the solar time before or after noon divided by 15 arc degree/h. For any given day of the year n and time on that day H in hour angle, the incident solar radiation is affected by the declination of the earth with respect to the sun, d
360 d ¼ 23:45 Sin ð284 þ nÞ 365
ð4Þ
where d is the sun’s declination in degree of arc and n is the day of the year where January 1st is n = 0 and December 31st is n = 365. For a given day of the year, solar radiation occurs from sunrise to sunset (Hr and Hs), respectively. Considering that at Hr and Hs, the sun’s altitude b is 0° ½SinðdÞSinðLÞ CosðH r Þ ¼ CosðH s Þ ¼ ð5Þ ½CosðdÞCosðLÞ where Hr is equal to Hs, all values of Hr, Hs and H are in hour angle before or after noon, and L is the latitude of the location in degree of arc. Between Hr and Hs, the sun’s altitude b is computed as SinðbÞ ¼ ½CosðLÞCosðdÞCosðH Þ þ ½SinðLÞSinðdÞ And the sun’s azimuth U is calculated as ½SinðbÞSinðLÞ SinðdÞ CosðUÞ ¼ ½CosðbÞCosðLÞ
ð6Þ
ð7Þ
The direct normal solar irradiation I (W/m2) is computed as A I¼ ð8Þ feðB=SinðbÞÞg where A (W/m2) and B (dimensionless) are factors defined by regression according to monthly values (ASHRAE, 2009) depending on the day of the year n A ¼ 2:229 107 n4 þ 1:648 104 n3 3:418 102 n2 þ 1:387n þ 1214
ð9Þ
B ¼ 1:140 1010 n4 8:607 108 n3 þ 1:884 105 n2 9:748n þ 0:1547
ð10Þ
The value of I can vary by as much as 15% depending on the clearness of the atmosphere (ASHRAE, 2009). Accordingly, the direct solar irradiation absorbed by a vertical plane EDN, oriented at w degree from the south is defined in terms of the cosine of the incident angle h h ¼ Cos1 ½CosðbÞCosðU wÞ
ð11Þ
S. Cordeau, S. Barrington / Solar Energy 85 (2011) 1418–1429
and
EDN
A ¼ Cosh ¼ I Cosh efB=SinðbÞg
Ed ¼
A Cy ¼ ICy efB=SinðbÞg
1423
ð14Þ
ð12Þ
The ground reflected radiation Er (W/m2) is computed as 1 A Er ¼ ½C þ SinðbÞqg 2 efB=SinðbÞg 1 ¼ I½C þ SinðbÞqg ð13Þ 2 where qg is the coefficient of ground reflection generally equal to 0.2 for mix surfaces (ASHRAE, 2009). The diffused radiation is
where C (dimensionless) is also a function of the day of the year n C ¼ 2:229 107 n4 þ 1:648 104 n3 3:418 102 n2 þ 1:387n þ 1214
ð15Þ
and y ¼ 0:55 þ 0:437 CosðhÞ þ 0:313 Cos2 ðhÞ
for h > 78 ð16Þ
Fig. 3. Horizontal solar radiation at the Varennes Environment Canada station (Latitude 45.7°) versus that of the experimental sensor on (a) March 17th, 2008 and (b) July 22nd, 2007.
1424
S. Cordeau, S. Barrington / Solar Energy 85 (2011) 1418–1429
Fig. 4. Vertically absorbed solar radiation measured on site versus that from Eq. (15), with qg: (a) 0.2 for April 6 and 7, 2008; (b) 0.6 for November 19, 2007 and (c) 0.6 for February 14th, 2009.
S. Cordeau, S. Barrington / Solar Energy 85 (2011) 1418–1429
Accordingly, a vertical plane oriented at w arc degree from the south will absorb the following direct, ground reflected and diffused solar radiation, Ra, equal to I multiplied by the sum of the respective coefficients 1 Ra ¼ I Cosh þ ðC þ sinðbÞqg Þ þ Cy 2 1 ¼ Cosh þ ðC þ SinðbÞqg Þ þ Cy ð17Þ 2 Eq. (15) was used to validate the absorbed radiation measured by the vertical sensor installed on the weather station.
1425
4. Results 4.1. Validation of solar irradiance data Fig. 3 compares the absorbed radiation for a horizontal plane, measured at the Varennes weather station of Environment Canada, and by the horizontal radiation sensor at the experimental site, for two cloud free days. The values measured at the experimental site follow the same trend as those measured by Environment Canada except for the peak which was generally 5–7% higher at the experimental site. This resulted from the fact that the Varennes
Fig. 5. Unglazed solar pre-heater energy recovery on: (a) March 5th 2009 for a wind velocity of 2.6 m/s and (b) March 12th, 2009 for a wind velocity of 7.2 m/s.
1426
S. Cordeau, S. Barrington / Solar Energy 85 (2011) 1418–1429
Environment Canada weather station was only 5 km from the St. Lawrence River, an important water body, as compared to the experimental site which was 30 km away. Since air water vapour reflects, scatters and absorbs solar radiation, the experimental readings were higher than those measured in Varennes. Fig. 4 compares the absorbed solar radiation calculated using Eq. (15), for a vertical plane parallel to the solar collectors, to that measured by the vertical radiation sensor on April 6th and 7th, 2008, November 19th, 2007 and Feb-
ruary 14th, 2009, respectively. The experimental values followed the calculated trend except for a 4% variation, resulting from the clearness of the sky which can create up to 15% variation (ASHRAE, 2009). When the ground reflection coefficient qg was adjusted with seasons and ground cover, Eq. (15) produced values corresponding to that measured after 3:00 pm, in the absence of direct solar radiation. On November 19th, 2007 with a light snow cover over frozen grounds and on February 14th, 2009 with a heavy snow cover, a value of 0.3 and 0.6 for qg respectively
Fig. 6. Heat recovery efficiency (e, %) versus wind velocity (x, m/s) for a ventilation rate of 0.86–1.14 m3/s, under bypassed and active solar pre-heaters.
Fig. 7. Heating load and inlet fresh air temperature on March 4th, 2009, for a solar energy recovery of 532 MJ.
S. Cordeau, S. Barrington / Solar Energy 85 (2011) 1418–1429
provided a good data fit. On April 6th and 7th, the recommended value of 0.2 for qg also provided a good fit. Considering that the 3.0 m high weather station was installed half way between the two experimental barns, 58 m apart and 11 m in total height including the roof, the vertical sensor observed a shadow effect before 9:00 am from November to February inclusively (Fig. 4). Accordingly, the two Hobo silicon pyranometer radiation sensors were found to measure the absorbed radiation with a maximum error of 7%, including differences in air moisture content. This maximum error is considered adequate considering that depending on the clearness of the day, values can vary by as much as 15% (ASHRAE, 2009). 4.2. Solar air pre-heater energy recovery efficiency Fig. 5a and b illustrate the effect of wind velocity on the energy recovery efficiency of the unglazed solar air preheaters for March 5th and 12th, 2009, respectively,
1427
according to Eqs. (16) and (17). Although the average ventilation rate on each day was 1.18 m3/s and 1.16 m3/s, the overall efficiency in energy recovery (Eq. (17)) was 63% and 20%, under respective wind velocities of 2.6 m/s and 7.2 m/ s. Whereas both days experienced the same incident solar radiation, 30 kW of average energy was recovered on March 5th between 9:00 am and 1:00 pm, as opposed to 10 kW on March 12th because of stronger winds. Fig. 6 correlates solar air pre-heater efficiency and wind velocity irrespective of its direction when the solar air pre-heaters are either active or bypassed (Fig. 1). Days with a low solar incident radiation under 15 kW, on the average between 8:00 am and 4:00 pm, are excluded because of limited wind effect. For active unglazed solar air pre-heaters, heat recovery efficiency was correlated to wind velocity (R2 of 0.41) and dropped by 5.7% for every 1 m/s of wind velocity increment. For bypassed unglazed solar air pre-heaters, some energy was still recovered but the efficiency was under 20% and dropped
a Regular white metal siding
Solar air pre-heater
Barn B
Stud
b Regular white metal siding Solar air preheater
Fig. 8. (a) Exterior infra-red view of the unglazed solar air pre-heaters at 0 °C outside and (b) regular photo of the same unglazed solar air pre-heater section.
1428
S. Cordeau, S. Barrington / Solar Energy 85 (2011) 1418–1429
Table 1 Unglazed solar air pre-heater performance. Month January–February Vertical incident radiation (GJ/(MJ/m2/day)) Average increase in air inlet temperature (°C) Solar air pre-heater efficiency (%) Average day time ventilation rate (m3/s) Average day time wind velocity and direction (m/s and °)a Solar energy recovered at the inlet (Can. $/barn/day) March Vertical incident radiation (GJ/MJ/m2/day) Average increase in air inlet temperature (°C) Solar air pre-heater efficiency (%) Average day time ventilation rate (m3/s) -Average day time wind velocity and direction (m/s and °)a Solar energy recovered at the inlet (Can. $/barn/day) November–December Vertical incident radiation (GJ/MJ/m2/day) Average increase in air inlet temperature (°C) Solar air pre-heater efficiency (%) Average day time ventilation rate (m3/s) Average day time wind velocity and direction (m/s and °)a Solar energy recovered at the inlet (Can. $/barn/day) Total solar energy at the inlet (Can. $/barn/winter season)
Expectedb
2008
2009
21.5/9.4
20.3/8.8 3.1 36.6 1.16 4.0–160 $15.80
23.2/11.2 6.2 42.5 1.28 4.7–100 $23.40
27.0/11.8
37.4/16.3 8.55 53.6 1.28 4.4–190 $42.90
28.1/12.2 4.9 35.3 1.17 5.2–105 $21.15
14.8
2007
13.8/6.0 4.3 64.1 1.17 3.0–160 $18.90
15.1/6.59 3.8 54.8 1.16 3.1–160 $17.70 $3415
$3 115
a
Environment Canada. Average solar radiation for the Montreal area according to Natural Resource Canada. The solar energy recovered at the inlet is based on a value of $0.08 Can./kW h; one season covers the period of November to March inclusively. b
by 4% for by every 1 m/s of wind velocity increment. A high wind velocity above 6.8 m/s practically eliminated all solar heat recovery when the solar air pre-heaters were bypassed. During the summer and despite bypassing, the solar air pre-heaters increased the daytime temperature of incoming fresh air by 1–2 °C with ventilation rates exceeding those of the winter. Finally, the outside temperature and the barn ventilation rate had no apparent impact on the energy recovery efficiency of the solar air pre-heaters (R2 of 0.05). In this experiment, the ventilation rate remained between 0.86 and 1.17 m3/s, with two and three fans in operation, for an unglazed solar pre-heater surface air velocity of 0.012 or 0.016 m/s. This speed respects the lower recommended values of 0.017 to 0.039 m/s for long buildings such as those used in this experiment, to minimize prevailing wind effect at 45° (Gunnewiek et al., 2002). Using a ventilation air flow rate of 0.01 m/s against unglazed solar air pre-heater, average efficiencies of 40– 50% were observed for wind velocities under 2 m/s and of 25% for wind velocities above 6 m/s (Fleck et al., 2002). Furthermore, efficiency decreased with increasing levels of incident solar radiation but remained relatively stable irrespective of wind direction. Despite the observed wind effect and during sunny days, the solar air pre-heaters substantially increased the temperature of the cold incoming ventilation air and dropped the heating requirements for the broiler barns. Fig. 7 shows a heating load dropping by an average of 18.5 kW during
8 h on March 4th 2009, for a total energy saving of 532 MJ/day, respectively. The solar air pre-heater energy exchange process was viewed on March 20th, 2009 using an infra-red camera (Fig. 8a and b). Exposed to an exterior temperature of 0 °C, the barn siding exhibited a temperature of 0–7 °C while the unglazed solar air pre-heater exhibited a temperature of 10–18 °C, with the stud support structure at 10– 12 °C. Over the air inlets, the solar air pre-heater exhibited a temperature of 22–25 °C while the barn air temperature was at 32 °C. During cold winter nights in the absence of solar energy, the air temperature inside the inlet was observed to be 1–2 °C warmer that that outside. The infra-red picture therefore indicates cold fresh air progressively picking up heat over the full surface of the solar air pre-heater with no flow reversal or preferential path wasting solar energy. For the experimental period, the overall heat recovery performance of the solar air pre-heaters is summarized in Table 1. The highest average solar energy recovery efficiencies of 64.1 and 54.8% were reached in November/December of 2007 and 2008, respectively, when the incoming solar radiation was at its lowest level of 6.00 and 6.59 MJ/m2/ day, as also observed earlier (Fleck et al., 2002). During January/February and March of each year, the energy recovery efficiency ranged between 35 and 55%. Between years, the efficiency varied with sky cloudiness and wind velocity. Nevertheless, higher levels of incident radiation above 8 MJ/m2/day lead to more incoming fresh air heat-
S. Cordeau, S. Barrington / Solar Energy 85 (2011) 1418–1429
ing and heating load reduction. Assuming an energy cost of 8¢ Can./kW h, the solar air pre-heaters were able to recover on the average, and from November to March of both years, an energy value of $14.80 Can./m2, corresponding to a 4.7% annual return on investment, defined as the ratio of the annual average energy recovered and the initial investment cost. 5. Conclusion The objective of the present study was to evaluate the heat recovery efficiency and potential of an unglazed solar air pre-heater consisting of an unglazed black corrugated metal sheeting with 1% perforation through which cold fresh air was pulled and used to ventilate broiler chicken barns. The experimental site was located 40 km east of Montreal, Canada. The study leads to the following conclusions: (1) The radiation sensors used on the experimental site measured the absorbed radiation with a maximum error of 7%, including the effect of variable exterior air moisture content; (2) The theoretical calculation method predicting absorbed radiation on a vertical surface (ASHRAE, 2009) was adequate for a ground reflection coefficient qg of 0.2 during the summer for a grass cover, and 0.3 and 0.6 during the winter for a light and heavy snow cover, respectively; (3) Besides the level of incoming solar radiation, wind velocity was found to be the main factor affecting the energy recovery efficiency of the unglazed solar air pre-heaters. The average efficiency of the solar air pre-heaters was 65% for wind velocities under 2 m/s, but dropped below 25% for wind velocities exceeding 7 m/s; (4) Overall, the solar air pre-heaters were able to recover $14.80 Can./m2/cold season, representing an annual return on investment of 4.7%.
Acknowledgements The financial contribution of the following groups is acknowledged: Natural Resources Canada, Natural Science and Engineering Council of Canada, Conseil pour le De´velopement Agricole du Que´bec and la Coop Fe´de´re´e.
1429
References Arinze, E.A., Sokhansan, J.S., Schoenau, G.J., 1993. Simulation of natural and solar heated air hay drying system. Computers and Electronics in Agriculture 8, 325–345. Arulanandam, S.J., Terry Hollands, K.G., Brundette, E., 2000. A CFD heat transfer analysis of the transpired solar collector under no-wind conditions. Solar Energy 67 (1–3), 93–100. ASHRAE 2009. Handbook of fundamentals. American Society of heating, refrigeration and air conditioning, Atlanta, Georgia, USA. Augustus Leon, M., Kumar, S., 2007. Mathematical modeling and thermal performance analysis of unglazed transpired solar collectors. Solar Energy 81, 62–75. Belusko, M., Saman, W., Bruno, F., 2008. Performance of jet impingement in unglazed air collectors. Solar Energy 82, 389–398. Beshada, E., Zhang, Q., Boris, R., 2006. Winter performance of a solar energy greenhouse in southern Manitoba. Canadian Biosystems Engineering 48, 5.1–5.8. Carpenter, S.C., Kokko, J.P., 1991. Performance of solar preheated ventilation air systems. In: Proceedings of the Annual Conference of Solar Energy Society of Canada. June 21–26, Toronto, Canada. solar Energy Society of Canada inc., Ottawa, Canada, pp 261–265. Cordeau, S., Barrington, S., 2010. Heat balance for two commercial broiler barns with ventilation air solar heaters. Journal of Biosystems Engineering 17 (3), 232–241. Esmay, M.L., Dixon, J.E. 1986. Environmental control for agricultural buildings. AVI Publishing Company Inc., Westport, Connecticut, USA. Fleck, B.A., Meier, R.M., Matovic, M.D., 2002. A field study of the wind effects on the performance of an unglazed transpired solar collector. Solar Energy 73 (3), 209–216. Fuller, R.J., 2007. Solar heating systems for recirculation aquaculture. Aquaculture engineering 36, 250–260. Godbout, S., Pouliot, F., Lachace, I., Guimont, H., Leblanc, R., Pelletier, F., Hamelin, L. 2004. Feasibility of energy recovery of a solar wall in pig nursery. ASAEB paper 044140. American Society of Agricultural, food and biological systems, St Joseph, Michigan, USA. Gunnewiek, L.H., Brundrett, E., Hollands, K.G.T., 1996. Flow distribution in unglazed transpired plate solar air heaters of large area. Solar Energy 58 (4–6), 227–237. Gunnewiek, L.H., Hollands, K.G.T., Brundrett, E., 2002. Effect of wind on flow distribution in unglazed transpired-plate collectors. Solar Energy 72 (4), 317–325. Heber, A.J., Ni, J.-Q., Haymore, B.L., Duggirala, R.K., Keener, K.M., 2001. Air quality and emission measurement methodology at a swine finishing buildings. Transactions of the American Society of Agricultural Engineers 44 (6), 1765–1778. Kutscher, C.F., Christensen, C.B., Barker, G.M., 1993. Unglazed transpired solar collectors; heat loss theory. Journal of Solar Energy Engineering 115, 182–188. Sokhansanj, S., Schoenau, G.J., 1991. Evaluation of a solar collector system with thermal storage for preheating ventilation air in farm buildings. Energy Conversion Management 32 (2), 183–189. Stojanovic, B., Hallberg, D., Akander, J., 2011. A steady state thermal duct model derived by fin-theory approach and applied on an unglazed solar collector. Solar Energy 84 (10), 1838–1851.
Solar Energy 85 (2011) 1418–1429 www.elsevier.com/locate/solener
Performance of unglazed solar ventilation air pre-heaters for broiler barns Se´bastien Cordeau, Suzelle Barrington ⇑ Department of Bioresource Engineering, Macdonald Campus of McGill University, 21 111 Lakeshore, Ste Anne de Bellevue, Quebec, Canada H9X 3V9 Received 24 February 2010; received in revised form 24 December 2010; accepted 28 March 2011 Available online 5 May 2011 Communicated by: Associate Editor C. Estrada-Gasca
Abstract Solar radiation is an interesting heat source for applications requiring a limited amount of energy, such as pre-heating cold fresh air used in venting livestock barns. The objective of this study was to evaluate the energy recovery efficiency of a solar air pre-heater consisting of an unglazed perforated black corrugated siding where the incoming fresh ventilation air picks up heat from its face and back. Installed on the southeast wall of two broiler barns located 40 km east of Montreal, Canada, the performance of solar air pre-heaters was monitored over 2 years. Sensors inside the barns monitored the temperature of the ambient air, that pre-heated by the solar collector and that exhausted by one of the three operating fans. An on-site weather station measured ambient air temperature, wind direction and velocity and radiation energy absorbed on a vertical plane parallel to the unglazed solar air pre-heaters. The measured vertical solar radiation value was used to evaluate the heat recovery efficiency of the unglazed solar air pre-heaters. Using data from the Varennes Environment Canada weather station located 30 km northwest, the solar sensors were found to measure the absorbed solar radiation with a maximum error of 7%, including differences in exterior air moisture. Unglazed, the efficiency of the solar air pre-heaters reached 65% for wind velocities under 2 m/s, but dropped below 25% for wind velocities exceeding 7 m/s. Nevertheless, the unglazed solar air pre-heaters were able to reduce the heating load especially in March of both years. Over a period starting in November and ending in March, the solar air heaters recovered an energy value equivalent to an annual return on investment of 4.7%. Ó 2011 Elsevier Ltd. All rights reserved. Keywords: Solar energy; Recovery efficiency; Ventilation; Heating
1. Introduction Within a concept of sustainability, solar radiation is an interesting heat source for applications requiring a limited amount of energy. Despite the lack of research during the 1990s, the twenty-first century brought some renewed interest in solar energy to reduce the use of petro-fuels and their greenhouse gas generation. In agriculture, solar energy can provide heat for a wide range of applications (Arinze et al., 1993; Beshada et al., 2006; Fuller, 2007), such as the ventilation system of livestock buildings requiring heating to maintain a high ambient temperature with young animals. ⇑ Corresponding author. Tel.: +1 514 398 7776; fax: +1 514 398 8387.
E-mail address: [email protected] (S. Barrington). 0038-092X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.solener.2011.03.026
To reduce the heat load of buildings, solar systems have evolved. Glazed solar plates are the oldest and simplest form of solar heaters consisting of a black surface covered by a transparent glazing, and installed on a vertical wall or roof exposed to sunlight. A south wall covered by a black metal sheet protected by a fiberglass glazing reduced by 50% the heat load of a swine gestation barn in Saskatoon, Canada (Sokhansanj and Schoenau, 1991). While the solar collector had a short payback period of 5 years, adding a solar heat storage system was not found economical. In the late 1980s, the concept of unglazed transpired solar plate was developed to further shorten the payback period of solar heating systems. Such solar plates consist of an unprotected or unglazed but perforated black metallic surface placed over a plenum on a southern surface. Eliminating
S. Cordeau, S. Barrington / Solar Energy 85 (2011) 1418–1429
1419
Nomenclature A, B and C are coefficients changing with n and used to calculate the absorbed radiation, dimensionless Aw area of solar air pre-heaters of 73.65 m2 /floor Cp heat capacity of dry air or 1.006 kJ/kg dry air/°C CV coefficient of variation Ed diffuse solar irradiance (W/m2) Er ground reflected solar irradiance (W/m2) EDN absorbed direct solar irradiance (W/m2) ET equation of time in decimal minutes H solar time before or after noon divided by 15° of arc/h, where for example H is 0° of arc at noon Hr sunrise in hour angle about noon Hs sunset in hour angle about noon I incident radiation (W/m2) L latitude of the location in degree of arc LON longitude of the position in decimal degree of arc LSM local meridian time in decimal degree of arc LST local standard time in decimal hour n day f the year, where January 1st is n = 1 and December 31st is n = 365 Q ventilation rate offered by the ventilation system in m3/s R earth’s rotational speed of 15° of arc/h Ra absorbed solar radiation in (W/m2)
the transparent glazing reduces the cost of the collector (Gunnewiek et al., 1996). Creating an interior negative pressure, the building ventilation system is used to pull exterior air through the perforations of the black metallic sheeting, and then over the back surface of the solar plate before entering the building (Carpenter and Kokko, 1991). Although the solar collector is unprotected, its perforations allow the ventilation air to collect heat lost by convection at its surface (Kutscher et al., 1993). Air flow at the back of the solar plate is also important in recovering solar energy (Gunnewiek et al., 1996). Nevertheless, the efficiency of such a solar system depends heavily on its design influencing surface air flow distribution. Flow reversal is one of the main issues, where heated air circulating behind the solar plate will exit through the perforations (Gunnewiek et al., 1996). Wind can have a substantial effect on solar collector efficiency by reducing convective heat loss recovery and increasing flow reversal. Computational Fluid Dynamics models (CFD) were used to establish unglazed solar collector configurations to optimize heat recovery and minimize flow reversal (Gunnewiek et al., 1996; Arulanandam et al., 2000; Augustus Leon and Kumar, 2007; Stojanovic et al., 2011). To counter balance wind effect, the ventilation system must create a minimum air flow rate on the collector surface which varies with building and solar collector configuration, and wind direction (Gunnewiek et al., 2002). Nevertheless, this minimum air
Rs
Tt To WS y b U d 1 e h qg qa W
absorbed radiation measured by the vertical sensor representing the potential heat which can be transferred to the cold fresh air by the solar air pre-heaters with an area of 73.65 m2/floor air temperature at the ventilation inlet (°C) outside air temperature in °C wind velocity (m/s) coefficient used to describe the diffused absorbed radiation, dimensionless sun’s altitude in degree of arc sun’s azimuth measured from the south in degree of arc declination of the earth with respect to the sun in degree of arc energy recovery by the cold fresh air circulating over the solar air pre-heaters (kW) solar air pre-heater efficiency (%) angle of incidence of the solar radiation on the absorbing surface degree coefficient of ground reflection generally equal to 0.2 density of dry air for the temperature measured at the mouth of the fan (kg/m3) is the orientation of the vertical wall receiving the radiation and with respect to the south arc degree
flow rate reduces the surface of the solar collector and accordingly, the amount of solar energy recovered. Recently, a modified version of the unglazed solar plate collector was proposed, namely the jet impinged unglazed solar collector. This solar system consists of a non perforated and unglazed metallic solar plate installed over a plenum enclosed by a perforated back wall allowing air jets to hit the back of the solar collector at a perpendicular angle, thus increasing heat recovery (Belusko et al., 2008). Although several design monograms were produced to increase the efficiency of unglazed solar collectors, field applications can be quite challenging where building air ventilation varies with seasons. A 15 to 45% energy recovery was measured for unglazed solar air pre-heaters installed over the ventilating inlets of a swine nursery in the Quebec City area, Canada (Godbout et al., 2004). To further evaluate the agricultural potential of such unglazed solar air pre-heaters, the present paper measured their efficiency in recovering solar radiation and reducing the heating cost of two broiler barns located 40 km east of Montreal, Canada. Each barn had three floors offering an area of 535 m2/floor. Solar air pre-heaters covered a vertical surface of 73.65 m2/floor on the southeast wall for a total of 221 m2/barn. Besides a horizontal radiation sensor, a vertical radiation sensor was positioned parallel to the solar collectors to measure the absorbed solar
1420
S. Cordeau, S. Barrington / Solar Energy 85 (2011) 1418–1429
radiation. Radiation measurements were validated against the readings of the Varennes Environment Canada weather station located 30 km northwest and theoretical values. For the winters of 2007–08 and 2008–09, the efficiency of the solar air pre-heaters was measured by comparing the heat recovered by the incoming ventilation air and the solar radiation absorbed by the vertical sensor parallel to the collectors. The savings in heating was also observed by monitoring the operation time of natural gas heaters on each floor. 2. Experimental set-up and methodology 2.1. System description The unglazed solar air pre-heaters consisted of a wooden box frame built over the fresh air inlets of all 3 floors of both experimental broiler barns (Fig. 1a and b). Its unprotected facing of black corrugated metal sheeting had small perforations covering 1% of its surface through which fresh ventilation air entered the barn. The unglazed solar air pre-heaters could be bypassed by opening the bottom portion of the wooden box frame. The unglazed solar air heaters were installed on the southeast face of the experimental buildings at an angle of 50° from the south. Each 3 floors of the two experimental broiler barns had a solar air pre-heaters surface measuring 1.47 m in height and 50.1 m in length, for 73.6 m2 /floor or 221 m2/barn.
Inside
Outside
Fixed panel
Air inlet temperature sensor
Panel automatically adjusted with fan air flow for a fresh air speed of 4 to 5 m/s
Unglazed solar air pre-heater with 1% perforation
Cold fresh air entering through perforation Barn wall Bypass panel to shut off solar air pre-heater
Fig. 1a. Configuration of the experimental unglazed solar air pre-heaters of corrugated black metal siding.
Fig. 1b. Unglazed solar air heaters over the fresh air inlets (not visible) and the bottom by-pass door.
2.2. Experimental broiler barns In this experiment, the two identically built broiler barns were located in St Jean Baptiste, 40 km east of Montreal, Canada, at 45.5° latitude where the LSM was 75.0° at normal eastern time, and the LO was 73.1°. The barns faced the southeast at an angle w of 50° from the south, and measured 9.2 m in width, 9.0 m in total wall height and 61 m in total length where 57.9 m was occupied by the broilers. All 3 floors of both experimental barns were equipped with two natural gas heaters (Fig. 2), each generating 30 kW of net sensible heat. To conserve energy and minimize cost, the heater exhaust was discharged inside the barn, a practice not recommended because of the high CO2 levels created. Such practice had no observable effect on broiler bird performance (Cordeau and Barrington, 2010). Each heater was installed at a height of 1.5 m at both ends of the floors, and on the fresh air inlet side. The heat generated was pushed towards the centre of the floor by the heater fan. Each barn floor had a capacity of 6500 birds raised in batches during 35–45 days, from 0.035 to 2.5 kg in body mass. During the cold season and with the broilers freshly introduced in the barns, only two 400 mm fans were operated/floor at a minimum speed of 55%. The speed of these exhaust fans was increased by an automated control governed by the temperature measured using three thermocouples located in the centre but distributed over the room length (Fig. 2) at a height of 0.3 m. The speed of the 400 mm fans was increased from 55 to 100% over a temperature increment of 1.4 °C. The temperature set-point started at 32 °C and dropped by 0.27 °C/day until it reached 24 °C. After one to three weeks, a third 400 mm fan was placed in operation and all three fans handled the fresh air ventilation rate until the end of the batch. At 55 and 100% of their full speed, each 400 mm fan delivered 0.43 m3/s (CV = 14.9%) and 0.94 m3/s (CV = 7.6%), measured using a Balometer (ALNOR EBT721 Balometer, Huntingdon Beech, CA, USA).
S. Cordeau, S. Barrington / Solar Energy 85 (2011) 1418–1429
1421
Fans to exhaust air
Data Logger ▓ ▪ T2
Heater ▪ T3
N T4
◄
▪ T1 Heater
▪ T6 Incoming cold fresh air
▪ T5 Solar air heaters
Fig. 2. Experimental barn instrumentation: temperature sensors T1, T2, T3 for inside air at 0.3 m off the floor; T5 and T6 – inside inlet air, and; T4 – inside outgoing air.
2.3. Instrumentation The site was instrumented to monitor exterior climatic conditions and interior air quality on all 3 floors of both experimental barns. To monitor exterior climatic conditions, a weather station was installed on a tower 3.0 m high located in a grassed area mid way between the two broiler barns. This weather station recorded the ambient climatic conditions: wind velocity and direction, ambient air temperature, barometric pressure and relative humidity, and absorbed radiation. A Hobo data logger recorded all weather data reading every 5 min and was downloaded monthly onto a portable computer. Two solar radiation sensors (Model S-LIB Hobo Silicon Pyranometer Smart Sensor, Onset Computer Corporation, Pocasset, MA, USA) were placed in a horizontal and vertical position, to respectively measure the horizontally absorbed radiation and the radiation adsorbed by the solar air pre-heaters. The vertical radiation sensor was placed parallel to the solar air pre-heaters using a compass. According to the manufacturer, the Hobo silicon pyranometer radiation sensors measure absorbed radiation with a maximum error of ±10.0 W/m2 or ±5%, whichever is greater in sunlight. An additional error of 0.38 W/m2/°C can be associated with temperatures above 25 °C. The HOBO silicon pyranometer radiation sensors use silicon photodiodes to measure solar power with a spectral range of 300–1100 nm. Nevertheless, this range introduces a negligible error when the sensors are calibrated to measure natural sunlight (Onset Computer Corporation, Pocasset, MA USA). The accuracy of the vertical radiation sensor was verified by comparing its reading to theoretical values (Eq. (15)), while that of the horizontal sensor was verified against readings recorded by the Varennes Environment Canada weather station (Latitude of 45.7°) located 30 km northwest. The air quality on all floors of both experimental barns was recorded every 5 min. The inside air temperature of each floor was measured at three positions located a height of 0.3 m (T1, T2 and T3), next to the thermocouples controlling the ventilation system (Fig. 2). The incoming fresh
air temperature was measured at two locations inside the fresh air inlets (T5 and T6). One more temperature sensor was installed at one of the 400 mm diameter exhaust fan in continuous operation (T4). All temperature sensors were Smart Set Hobo sensors (Onset Computer Corporation, Pocasset, MA, USA). The ventilation rate was monitored by recording the rpm of one 400 mm fan/floor using an electromagnetic converter (Heber et al., 2001). The air flow rate was computed from a correlation obtained using a Balometer (ALNOR EBT721 Balometer, Huntingdon Beech, CA, USA) during which the volumetric air displacement of all 400 mm fans was measured at rpms varying from 55 to 100%. The cumulated operating time of the natural gas heaters was recorded every 5 min using a 5 V AC adapter detecting gas heater operation and sending a signal to the data logger to record operating time. 2.4. Methodology The main objective of the study consisted in measuring the heat recovered by the solar air pre-heaters and thus, the savings in heating load during the two consecutive winters, namely from November 2007 to March 2008 and November 2008 to March 2009 inclusively. The efficiency of the unglazed solar air pre-heaters was monitored continuously by comparing the measured absorbed vertical radiation to the solar heat recovered by the cold fresh air at the ventilation inlets (ASHRAE, 2009; Esmay and Dixon, 1986). This heat recovery was computed from the difference in monitored exterior and fresh air inlet temperatures multiplied by the air flow rate of the ventilation system 1 ¼ ½ðT t T o ÞCp Q qa 1 e ¼ 100 ðRs AwÞ
ð1Þ ð2Þ
where 1 is the energy recovery by the cold fresh air circulating over the solar air pre-heaters in kW, e is the solar air
1422
S. Cordeau, S. Barrington / Solar Energy 85 (2011) 1418–1429
pre-heater efficiency in%, Tt is the air temperature at the ventilation inlet in °C, To is the outside air temperature in °C, Cp is the heat capacity of dry air or 1.006 kJ/kg dry air/°C, Q is the ventilation rate offered by the ventilation system in m3/s, qa is the density of dry air for the temperature measured at the exhaust fan in kg/m3, Rs is the absorbed radiation measured by the vertical sensor representing the potential heat which can be transferred to the cold fresh air by the solar air pre-heaters in kW, and Aw is the area of solar air pre-heaters of 73.65 m2/floor. The monitoring of the unglazed solar air pre-heaters started upon reception of each batch of young birds and terminated when the fully grown birds were sent for slaughter, in an all-in all-out fashion. The acquired data was downloaded onto a portable computer between bird batches. Infra-red pictures were also taken of the solar air pre-heaters from outside to visualize the uniformity in solar heat recovery. 2.5. Statistical analysis This project measured the efficiency of the solar air preheaters in recovering solar radiation and reducing the heating cost of the two experimental broiler barns. The impact on solar air heater efficiency was therefore analyzed as a function of various climatic factors such as solar radiation level, wind speed and direction and exterior temperature. To obtain a relationship between these climatic factors and solar heat recovery, regression equations were obtained using Excel (Microsoft Office 2007, Seattle, USA). Furthermore, it was observed that cloudy periods produced a wide variation in efficiency often exceeding 100%, because the unglazed solar air pre-heaters accumulated exchangeable heat despite the drop in incident radiation. Thus, regression equations were produced for low and high radiation levels corresponding to cloudy and sunny days. 3. Calculation of incident solar radiation The accuracy of the horizontal solar sensor was verified using values measured by the Varennes Environment Canada weather station, while the accuracy of the vertical sensor was verified against calculated value for a plane with the same orientation. The incident solar radiation received by a plane on the surface of the Earth can be computed as a function of its solar orientation and position. This plane will perceive direct, indirect and diffused radiation, all of which can be calculated based on the following simplified method (ASHRAE, 2009). This method was used to validate the measured absorbed radiation by a radiation sensor positioned vertically and parallel to the experimental barn walls on which the solar air pre-heaters were installed. Because of time zones, the actual time must be converted to solar time (H) in hour angle, where the sun is positioned directly south at noon
ET H ¼ R LST þ 12 þ LSM LON 60
ð3Þ
where LST is the local standard time in decimal hours, ET is the equation of time in decimal minutes, LSM is the local meridian time in decimal degree of arc, LON is the local longitude in decimal degree of arc and R is the earth’s rotational velocity of 15° of arc degree. The hour angle time H is simply the solar time before or after noon divided by 15 arc degree/h. For any given day of the year n and time on that day H in hour angle, the incident solar radiation is affected by the declination of the earth with respect to the sun, d
360 d ¼ 23:45 Sin ð284 þ nÞ 365
ð4Þ
where d is the sun’s declination in degree of arc and n is the day of the year where January 1st is n = 0 and December 31st is n = 365. For a given day of the year, solar radiation occurs from sunrise to sunset (Hr and Hs), respectively. Considering that at Hr and Hs, the sun’s altitude b is 0° ½SinðdÞSinðLÞ CosðH r Þ ¼ CosðH s Þ ¼ ð5Þ ½CosðdÞCosðLÞ where Hr is equal to Hs, all values of Hr, Hs and H are in hour angle before or after noon, and L is the latitude of the location in degree of arc. Between Hr and Hs, the sun’s altitude b is computed as SinðbÞ ¼ ½CosðLÞCosðdÞCosðH Þ þ ½SinðLÞSinðdÞ And the sun’s azimuth U is calculated as ½SinðbÞSinðLÞ SinðdÞ CosðUÞ ¼ ½CosðbÞCosðLÞ
ð6Þ
ð7Þ
The direct normal solar irradiation I (W/m2) is computed as A I¼ ð8Þ feðB=SinðbÞÞg where A (W/m2) and B (dimensionless) are factors defined by regression according to monthly values (ASHRAE, 2009) depending on the day of the year n A ¼ 2:229 107 n4 þ 1:648 104 n3 3:418 102 n2 þ 1:387n þ 1214
ð9Þ
B ¼ 1:140 1010 n4 8:607 108 n3 þ 1:884 105 n2 9:748n þ 0:1547
ð10Þ
The value of I can vary by as much as 15% depending on the clearness of the atmosphere (ASHRAE, 2009). Accordingly, the direct solar irradiation absorbed by a vertical plane EDN, oriented at w degree from the south is defined in terms of the cosine of the incident angle h h ¼ Cos1 ½CosðbÞCosðU wÞ
ð11Þ
S. Cordeau, S. Barrington / Solar Energy 85 (2011) 1418–1429
and
EDN
A ¼ Cosh ¼ I Cosh efB=SinðbÞg
Ed ¼
A Cy ¼ ICy efB=SinðbÞg
1423
ð14Þ
ð12Þ
The ground reflected radiation Er (W/m2) is computed as 1 A Er ¼ ½C þ SinðbÞqg 2 efB=SinðbÞg 1 ¼ I½C þ SinðbÞqg ð13Þ 2 where qg is the coefficient of ground reflection generally equal to 0.2 for mix surfaces (ASHRAE, 2009). The diffused radiation is
where C (dimensionless) is also a function of the day of the year n C ¼ 2:229 107 n4 þ 1:648 104 n3 3:418 102 n2 þ 1:387n þ 1214
ð15Þ
and y ¼ 0:55 þ 0:437 CosðhÞ þ 0:313 Cos2 ðhÞ
for h > 78 ð16Þ
Fig. 3. Horizontal solar radiation at the Varennes Environment Canada station (Latitude 45.7°) versus that of the experimental sensor on (a) March 17th, 2008 and (b) July 22nd, 2007.
1424
S. Cordeau, S. Barrington / Solar Energy 85 (2011) 1418–1429
Fig. 4. Vertically absorbed solar radiation measured on site versus that from Eq. (15), with qg: (a) 0.2 for April 6 and 7, 2008; (b) 0.6 for November 19, 2007 and (c) 0.6 for February 14th, 2009.
S. Cordeau, S. Barrington / Solar Energy 85 (2011) 1418–1429
Accordingly, a vertical plane oriented at w arc degree from the south will absorb the following direct, ground reflected and diffused solar radiation, Ra, equal to I multiplied by the sum of the respective coefficients 1 Ra ¼ I Cosh þ ðC þ sinðbÞqg Þ þ Cy 2 1 ¼ Cosh þ ðC þ SinðbÞqg Þ þ Cy ð17Þ 2 Eq. (15) was used to validate the absorbed radiation measured by the vertical sensor installed on the weather station.
1425
4. Results 4.1. Validation of solar irradiance data Fig. 3 compares the absorbed radiation for a horizontal plane, measured at the Varennes weather station of Environment Canada, and by the horizontal radiation sensor at the experimental site, for two cloud free days. The values measured at the experimental site follow the same trend as those measured by Environment Canada except for the peak which was generally 5–7% higher at the experimental site. This resulted from the fact that the Varennes
Fig. 5. Unglazed solar pre-heater energy recovery on: (a) March 5th 2009 for a wind velocity of 2.6 m/s and (b) March 12th, 2009 for a wind velocity of 7.2 m/s.
1426
S. Cordeau, S. Barrington / Solar Energy 85 (2011) 1418–1429
Environment Canada weather station was only 5 km from the St. Lawrence River, an important water body, as compared to the experimental site which was 30 km away. Since air water vapour reflects, scatters and absorbs solar radiation, the experimental readings were higher than those measured in Varennes. Fig. 4 compares the absorbed solar radiation calculated using Eq. (15), for a vertical plane parallel to the solar collectors, to that measured by the vertical radiation sensor on April 6th and 7th, 2008, November 19th, 2007 and Feb-
ruary 14th, 2009, respectively. The experimental values followed the calculated trend except for a 4% variation, resulting from the clearness of the sky which can create up to 15% variation (ASHRAE, 2009). When the ground reflection coefficient qg was adjusted with seasons and ground cover, Eq. (15) produced values corresponding to that measured after 3:00 pm, in the absence of direct solar radiation. On November 19th, 2007 with a light snow cover over frozen grounds and on February 14th, 2009 with a heavy snow cover, a value of 0.3 and 0.6 for qg respectively
Fig. 6. Heat recovery efficiency (e, %) versus wind velocity (x, m/s) for a ventilation rate of 0.86–1.14 m3/s, under bypassed and active solar pre-heaters.
Fig. 7. Heating load and inlet fresh air temperature on March 4th, 2009, for a solar energy recovery of 532 MJ.
S. Cordeau, S. Barrington / Solar Energy 85 (2011) 1418–1429
provided a good data fit. On April 6th and 7th, the recommended value of 0.2 for qg also provided a good fit. Considering that the 3.0 m high weather station was installed half way between the two experimental barns, 58 m apart and 11 m in total height including the roof, the vertical sensor observed a shadow effect before 9:00 am from November to February inclusively (Fig. 4). Accordingly, the two Hobo silicon pyranometer radiation sensors were found to measure the absorbed radiation with a maximum error of 7%, including differences in air moisture content. This maximum error is considered adequate considering that depending on the clearness of the day, values can vary by as much as 15% (ASHRAE, 2009). 4.2. Solar air pre-heater energy recovery efficiency Fig. 5a and b illustrate the effect of wind velocity on the energy recovery efficiency of the unglazed solar air preheaters for March 5th and 12th, 2009, respectively,
1427
according to Eqs. (16) and (17). Although the average ventilation rate on each day was 1.18 m3/s and 1.16 m3/s, the overall efficiency in energy recovery (Eq. (17)) was 63% and 20%, under respective wind velocities of 2.6 m/s and 7.2 m/ s. Whereas both days experienced the same incident solar radiation, 30 kW of average energy was recovered on March 5th between 9:00 am and 1:00 pm, as opposed to 10 kW on March 12th because of stronger winds. Fig. 6 correlates solar air pre-heater efficiency and wind velocity irrespective of its direction when the solar air pre-heaters are either active or bypassed (Fig. 1). Days with a low solar incident radiation under 15 kW, on the average between 8:00 am and 4:00 pm, are excluded because of limited wind effect. For active unglazed solar air pre-heaters, heat recovery efficiency was correlated to wind velocity (R2 of 0.41) and dropped by 5.7% for every 1 m/s of wind velocity increment. For bypassed unglazed solar air pre-heaters, some energy was still recovered but the efficiency was under 20% and dropped
a Regular white metal siding
Solar air pre-heater
Barn B
Stud
b Regular white metal siding Solar air preheater
Fig. 8. (a) Exterior infra-red view of the unglazed solar air pre-heaters at 0 °C outside and (b) regular photo of the same unglazed solar air pre-heater section.
1428
S. Cordeau, S. Barrington / Solar Energy 85 (2011) 1418–1429
Table 1 Unglazed solar air pre-heater performance. Month January–February Vertical incident radiation (GJ/(MJ/m2/day)) Average increase in air inlet temperature (°C) Solar air pre-heater efficiency (%) Average day time ventilation rate (m3/s) Average day time wind velocity and direction (m/s and °)a Solar energy recovered at the inlet (Can. $/barn/day) March Vertical incident radiation (GJ/MJ/m2/day) Average increase in air inlet temperature (°C) Solar air pre-heater efficiency (%) Average day time ventilation rate (m3/s) -Average day time wind velocity and direction (m/s and °)a Solar energy recovered at the inlet (Can. $/barn/day) November–December Vertical incident radiation (GJ/MJ/m2/day) Average increase in air inlet temperature (°C) Solar air pre-heater efficiency (%) Average day time ventilation rate (m3/s) Average day time wind velocity and direction (m/s and °)a Solar energy recovered at the inlet (Can. $/barn/day) Total solar energy at the inlet (Can. $/barn/winter season)
Expectedb
2008
2009
21.5/9.4
20.3/8.8 3.1 36.6 1.16 4.0–160 $15.80
23.2/11.2 6.2 42.5 1.28 4.7–100 $23.40
27.0/11.8
37.4/16.3 8.55 53.6 1.28 4.4–190 $42.90
28.1/12.2 4.9 35.3 1.17 5.2–105 $21.15
14.8
2007
13.8/6.0 4.3 64.1 1.17 3.0–160 $18.90
15.1/6.59 3.8 54.8 1.16 3.1–160 $17.70 $3415
$3 115
a
Environment Canada. Average solar radiation for the Montreal area according to Natural Resource Canada. The solar energy recovered at the inlet is based on a value of $0.08 Can./kW h; one season covers the period of November to March inclusively. b
by 4% for by every 1 m/s of wind velocity increment. A high wind velocity above 6.8 m/s practically eliminated all solar heat recovery when the solar air pre-heaters were bypassed. During the summer and despite bypassing, the solar air pre-heaters increased the daytime temperature of incoming fresh air by 1–2 °C with ventilation rates exceeding those of the winter. Finally, the outside temperature and the barn ventilation rate had no apparent impact on the energy recovery efficiency of the solar air pre-heaters (R2 of 0.05). In this experiment, the ventilation rate remained between 0.86 and 1.17 m3/s, with two and three fans in operation, for an unglazed solar pre-heater surface air velocity of 0.012 or 0.016 m/s. This speed respects the lower recommended values of 0.017 to 0.039 m/s for long buildings such as those used in this experiment, to minimize prevailing wind effect at 45° (Gunnewiek et al., 2002). Using a ventilation air flow rate of 0.01 m/s against unglazed solar air pre-heater, average efficiencies of 40– 50% were observed for wind velocities under 2 m/s and of 25% for wind velocities above 6 m/s (Fleck et al., 2002). Furthermore, efficiency decreased with increasing levels of incident solar radiation but remained relatively stable irrespective of wind direction. Despite the observed wind effect and during sunny days, the solar air pre-heaters substantially increased the temperature of the cold incoming ventilation air and dropped the heating requirements for the broiler barns. Fig. 7 shows a heating load dropping by an average of 18.5 kW during
8 h on March 4th 2009, for a total energy saving of 532 MJ/day, respectively. The solar air pre-heater energy exchange process was viewed on March 20th, 2009 using an infra-red camera (Fig. 8a and b). Exposed to an exterior temperature of 0 °C, the barn siding exhibited a temperature of 0–7 °C while the unglazed solar air pre-heater exhibited a temperature of 10–18 °C, with the stud support structure at 10– 12 °C. Over the air inlets, the solar air pre-heater exhibited a temperature of 22–25 °C while the barn air temperature was at 32 °C. During cold winter nights in the absence of solar energy, the air temperature inside the inlet was observed to be 1–2 °C warmer that that outside. The infra-red picture therefore indicates cold fresh air progressively picking up heat over the full surface of the solar air pre-heater with no flow reversal or preferential path wasting solar energy. For the experimental period, the overall heat recovery performance of the solar air pre-heaters is summarized in Table 1. The highest average solar energy recovery efficiencies of 64.1 and 54.8% were reached in November/December of 2007 and 2008, respectively, when the incoming solar radiation was at its lowest level of 6.00 and 6.59 MJ/m2/ day, as also observed earlier (Fleck et al., 2002). During January/February and March of each year, the energy recovery efficiency ranged between 35 and 55%. Between years, the efficiency varied with sky cloudiness and wind velocity. Nevertheless, higher levels of incident radiation above 8 MJ/m2/day lead to more incoming fresh air heat-
S. Cordeau, S. Barrington / Solar Energy 85 (2011) 1418–1429
ing and heating load reduction. Assuming an energy cost of 8¢ Can./kW h, the solar air pre-heaters were able to recover on the average, and from November to March of both years, an energy value of $14.80 Can./m2, corresponding to a 4.7% annual return on investment, defined as the ratio of the annual average energy recovered and the initial investment cost. 5. Conclusion The objective of the present study was to evaluate the heat recovery efficiency and potential of an unglazed solar air pre-heater consisting of an unglazed black corrugated metal sheeting with 1% perforation through which cold fresh air was pulled and used to ventilate broiler chicken barns. The experimental site was located 40 km east of Montreal, Canada. The study leads to the following conclusions: (1) The radiation sensors used on the experimental site measured the absorbed radiation with a maximum error of 7%, including the effect of variable exterior air moisture content; (2) The theoretical calculation method predicting absorbed radiation on a vertical surface (ASHRAE, 2009) was adequate for a ground reflection coefficient qg of 0.2 during the summer for a grass cover, and 0.3 and 0.6 during the winter for a light and heavy snow cover, respectively; (3) Besides the level of incoming solar radiation, wind velocity was found to be the main factor affecting the energy recovery efficiency of the unglazed solar air pre-heaters. The average efficiency of the solar air pre-heaters was 65% for wind velocities under 2 m/s, but dropped below 25% for wind velocities exceeding 7 m/s; (4) Overall, the solar air pre-heaters were able to recover $14.80 Can./m2/cold season, representing an annual return on investment of 4.7%.
Acknowledgements The financial contribution of the following groups is acknowledged: Natural Resources Canada, Natural Science and Engineering Council of Canada, Conseil pour le De´velopement Agricole du Que´bec and la Coop Fe´de´re´e.
1429
References Arinze, E.A., Sokhansan, J.S., Schoenau, G.J., 1993. Simulation of natural and solar heated air hay drying system. Computers and Electronics in Agriculture 8, 325–345. Arulanandam, S.J., Terry Hollands, K.G., Brundette, E., 2000. A CFD heat transfer analysis of the transpired solar collector under no-wind conditions. Solar Energy 67 (1–3), 93–100. ASHRAE 2009. Handbook of fundamentals. American Society of heating, refrigeration and air conditioning, Atlanta, Georgia, USA. Augustus Leon, M., Kumar, S., 2007. Mathematical modeling and thermal performance analysis of unglazed transpired solar collectors. Solar Energy 81, 62–75. Belusko, M., Saman, W., Bruno, F., 2008. Performance of jet impingement in unglazed air collectors. Solar Energy 82, 389–398. Beshada, E., Zhang, Q., Boris, R., 2006. Winter performance of a solar energy greenhouse in southern Manitoba. Canadian Biosystems Engineering 48, 5.1–5.8. Carpenter, S.C., Kokko, J.P., 1991. Performance of solar preheated ventilation air systems. In: Proceedings of the Annual Conference of Solar Energy Society of Canada. June 21–26, Toronto, Canada. solar Energy Society of Canada inc., Ottawa, Canada, pp 261–265. Cordeau, S., Barrington, S., 2010. Heat balance for two commercial broiler barns with ventilation air solar heaters. Journal of Biosystems Engineering 17 (3), 232–241. Esmay, M.L., Dixon, J.E. 1986. Environmental control for agricultural buildings. AVI Publishing Company Inc., Westport, Connecticut, USA. Fleck, B.A., Meier, R.M., Matovic, M.D., 2002. A field study of the wind effects on the performance of an unglazed transpired solar collector. Solar Energy 73 (3), 209–216. Fuller, R.J., 2007. Solar heating systems for recirculation aquaculture. Aquaculture engineering 36, 250–260. Godbout, S., Pouliot, F., Lachace, I., Guimont, H., Leblanc, R., Pelletier, F., Hamelin, L. 2004. Feasibility of energy recovery of a solar wall in pig nursery. ASAEB paper 044140. American Society of Agricultural, food and biological systems, St Joseph, Michigan, USA. Gunnewiek, L.H., Brundrett, E., Hollands, K.G.T., 1996. Flow distribution in unglazed transpired plate solar air heaters of large area. Solar Energy 58 (4–6), 227–237. Gunnewiek, L.H., Hollands, K.G.T., Brundrett, E., 2002. Effect of wind on flow distribution in unglazed transpired-plate collectors. Solar Energy 72 (4), 317–325. Heber, A.J., Ni, J.-Q., Haymore, B.L., Duggirala, R.K., Keener, K.M., 2001. Air quality and emission measurement methodology at a swine finishing buildings. Transactions of the American Society of Agricultural Engineers 44 (6), 1765–1778. Kutscher, C.F., Christensen, C.B., Barker, G.M., 1993. Unglazed transpired solar collectors; heat loss theory. Journal of Solar Energy Engineering 115, 182–188. Sokhansanj, S., Schoenau, G.J., 1991. Evaluation of a solar collector system with thermal storage for preheating ventilation air in farm buildings. Energy Conversion Management 32 (2), 183–189. Stojanovic, B., Hallberg, D., Akander, J., 2011. A steady state thermal duct model derived by fin-theory approach and applied on an unglazed solar collector. Solar Energy 84 (10), 1838–1851.