EXPERIMENTAL ANALYSIS OF AIR-CONDITIONING IN HOSPITAL ROOMS BY MEANS OF LIGHT RADIANT CEILINGS

Transcription

1 EXPERIMENTAL ANALYSIS OF AIR-CONDITIONING IN HOSPITAL ROOMS BY MEANS OF LIGHT RADIANT CEILINGS Renato M. Lazzarin Francesco Castellotti Filippo Busato Department of Management and Engineering University of Padova Stradella S. Nicola, Vicenza - Italy Abstract. The new wards of the S. Bortolo Hospital in Vicenza (Italy) are air-conditioned by a mixed systems of radiant panels and primary air. The sensible load is covered by light radiant ceiling panels, whereas the fresh air controls the latent one. An ambient of the ward was suitably equipped with sensors and data logger so to evaluate the achieved level of comfort. All the room variables which influence the condition of thermal comfort (air temperature, humidity and velocity and mean radiant temperature), as well as the ceiling temperatures were logged, both in heating and cooling operation. Besides of the calculation of the achieved level of thermal comfort in terms of Predicted Mean Vote, the thermal conductances of the panels were determined and the sizing of the system was checked solving the detailed thermal balance of each surface. Keywords: Air-conditioning, Radiant ceilings, Hospital, Thermal comfort NOMENCLATURE m Mass flow [kg/s] σ Boltzmann constant = 5.67e-8 W/(m 2 K ) C Thermal conductance [W/(m 2 K)] Subscribes F View factor a Inside air K Thermal transmittance [W/(m 2 K)] add Adduction P Perimeter [m] c Specific heat [J/(kgK)] PMV Predicted Mean Vote con Convection PPD Predicted Percentage of Dissatisfied [%] g Globothermometer q Specific thermal flux [W/m 2 ] in Inside surface Q Thermal flux [W] inf Infiltration air S Area [m 2 ] int Internal T Temperature [K] m Mean t Temperature [ C] mr Mean radiant Greeks out Outside surface α Liminar heat transfer coefficient p Panel [W/(m 2 K)] rad Radiation ε Emissivity rs Remainder surfaces

2 s s-a ven Supply Sun-air Ventilation air w west win Water West faced Window 1. INTRODUCTION As regards air-conditioning systems, the specific requirements about indoor air quality (IAQ) and microclimate determine the choice of the technology to be adopted. Concerning the wards of a hospital, radiant panels coupled to a primary air ventilation limit the dust raising assuring an optimum IAQ, besides providing a high level of hygro-thermal comfort. Such a technology supplies year round air-conditioning to some pavilions of the S. Bortolo Hospital in Vicenza (Italy). It was possible to carry out an experimental survey that permitted an evaluation of the performances both in terms of thermal comfort and energy sizing. The logging of the systems took place in two different experimental sessions, in two different rooms, respectively in March and August 2003, to limit as much as possible the inconvenience caused to the patients and personnel. In both cases the ceiling panels are made of plasterboard: in the room investigated in winter steel pipes are in close contact to aluminum embedded sheets; in the room investigated in summer polybutylene pipes are simply embedded in the plasterboard. 2. THE PANEL THERMAL ANALYSIS The aim of the surveys was the evaluation of the specific thermal power exchanged by the panels, as well as their thermal conductance. To compute the thermal power the convective and radiative components were handled separately. The convection coefficient can be calculated easily, according to the following empirical relation: ( t t ) a p α con, p = (1) 0.08 Sp Pp if the panel surface is colder than the air, where t p is the panel temperature and t a is the air temperature, S p is the area of the panel and P p its perimeter. If the panel surface is warmer than the air, the relation to be used is: ( t t ) 0.25 a p α con, p = (2) Sp Pp Hence, the specific convective thermal exchange can be expressed as: q ( t t ) p, con = α con, p p a (3) As regards the radiative contribution, one can consider uniform the surface temperature of the panel and define the average temperature of the remainder surfaces of the room as:

3 T rs K i i = = N T Si 1 () i = 1 S i Taking into account different configurations, the view factor between the panel and the other surfaces keeps closely around 0.87 for parallelepiped rooms [Lazzarin, Crose, 2000]. Hence, the radiative thermal exchange can be expressed as: 8 ( T T ) = 5 ( T T ) K 1 q p, rad = σ 10 (5) p j p rs j = Sp F p, j p S ε j ε j Now it is possible to calculate the specific thermal power and to define the adduction coefficient α add,p and the panel conductance C p, according to the following relation: q ( t t ) = C ( t t ) p = qp, con + qp, rad = α add, p p a p w, m p (6) where t w,m is the water mean temperature of the panel piping: t w, m tw = tw, s + (7) 2 t w is the water temperature difference between the supply, t w,s, and the return sections that can be calculated knowing the water mass flow ṁ w : S q p p t w =. c w ( 1+ δ ) m w (8) δ is a coefficient that takes into account thermal losses through the ceiling. As just seen, the method permits the calculation of the specific thermal power exchanged by the panel, as well as the correlation between the conductance C p and the adduction coefficient α add,p that is suitable only for this specific kind of panel, of course. That piece of information is useful in checking the panel sizing or in the design procedure (see Table 2). In next paragraphs the method will be applied both for winter and summer session and the sizing will be checked in the design conditions as well. It is worth stressing the importance of the calculation of the remainder surfaces temperature for a correct evaluation on the radiative contribution in the thermal exchange. On this subject if one knows the information about the mean radiant temperature, the T rs evaluation can be more reliable and the application will be discussed just after having described the experimental campaigns settings. 3. THE EXPERIMENTAL SESSIONS The experimental surveys were carried out measuring a great number of variables. The following variables concern with the indoor environment:

4 the surface panel temperature in several points; the water supply temperature; the fresh air supply temperature; the microclimate variables, i.e. air temperature, mean radiant temperature, air relative humidity and air velocity. The instruments setting is described in Table 1. The microclimate logging station is conformed to ISO 7726 [ISO 7726, 1998]. All the values were logged every 15 minutes. The water mass flow was not measured, but calculated starting from the design condition of a piping pressure drop of 98 Pa/m. Table 1. Instruments used in the campaigns. Variable Sensor Surface panel temperature Copper-constantan thermocouple Water supply temperature Copper-constantan thermocouple Fresh-air supply temperature Copper-constantan thermocouple Air temperature Pt100 resistive sensor Mean radiant temperature Globe thermometer Air relative humidity Capacitive hygrometer Air velocity Hot wire anenometer Figure 1. Panel thermocouples layout and microclimate station position: (a) winter session; (b) summer session. The meteorological variables, i.e. solar irradiance, outside air temperature and relative humidity, were measured by a station placed in the center of Vicenza. 3.1 Winter session The winter campaign lasted from the 13 th to the 27 th of March of The room ( m 3 ) has only one external wall turned to west with a 2. m 2 window. The wall thermal conductance is about 0. W/m 2 K and the window one is about 3 W/m 2 K.

5 The radiant panel (13.0 m 2 ), that covers almost all the ceiling (8%) except for a perimeter belt, is made of diffusing aluminum sheets, fastened to the steel pipes by means of steel clips that assure a close thermal contact. The sheets are tied to a wire mesh embedded in a plasterboard layer. The pipes pitch is 33 cm and the whole piping length is 39 m. The inner pipe diameter is 1 cm and the thickness is 2 mm. The panel thermocouples layout as well as the microclimate station position is sketched in Figure 1a. The perimetric position of the latter was chosen to cause no inconvenience to the patients and the personnel. Thermal comfort evaluation The evaluation of thermal comfort follows the Fanger s theory [Fanger, 1970]: the time history of the microclimate variables is reported in Figure 2, save for the air velocity that was always lower than 0.15 m/s. Figure 2. Winter session. Microclimate variables. One can note that: the air temperature is almost always in the range C and the mean radiant temperature is not much higher than the air temperature; the difference is quite clear in the afternoons when the external wall and above all the window got warm due to the solar irradiance onto them and hence their temperatures affected the measure of the globe thermometer that was close to the external wall; the air relative humidity is in the range 27 3 % and so quite low with respect to the comfort as complained by the personnel; however, it is worth pointing out that the weight of air relative humidity is quite small in the thermal comfort evaluation, according to the Fanger s theory. The predicted mean vote (PMV) evaluation assumes a metabolic rate of 1 met and a clothing resistance of 0.75 clo, typical values for a patient. The PMV and the predicted percentage of dissatisfied (PPD) are shown in Figure 3. If the accepted PMV is in the range (and the consequent PPD lower than 10%), the hygro-thermal comfort was almost always reached [ISO 7730, 198]. Yet, the mean PMV is 0.3 and so a light sensation of cold can be expected, also justified by the closeness to the external wall. However, that conclusion changes if one considers the higher metabolic rate of the personnel (1.5 met) or the higher

6 clothing resistance of a visitor (1 clo). In the former case the mean PMV is 0.5 and in the latter it is 0.1. Figure 3. Winter session. Thermal comfort evaluation. The radiant panel performance We proceeded applying the equations above mentioned and we calculated the remainder surfaces temperature starting from the mean radiant temperature definition, that is the temperature of the fictive surface that exchange the same radiative thermal flux with the room surfaces, with the assumption of black body emissivities: T mr = T win F win g + T west F west g + T int F int g + T p F p g (9) where T win is the inside surface temperature of the window, T west the inside surface temperature of the west wall and T int the inside temperature of the remainder internal surfaces, of course except for the radiant panel. With F we indicate the view factors between two surfaces described by the two subscribes; g subscribe refers to the globe thermometer. The inside temperatures of window and west wall can be estimated according to K ( t t ) = ( t t ) win s a a add, win α (10) win a K west ( t t ) = ( t t ) s a a add, west α (11) west a where K is the thermal transmittance and t s-a the sun-air temperature that takes into account of the solar irradiance gain. Otherwise, one can calculate the internal temperature surface of window and west wall by means of a software that simulates the thermal behavior of the room with the measured boundary conditions [AA. VV., 1997]. In both cases it is necessary to assume an outside adduction coefficient: here it was fixed to 25 W/(m 2 K). An other assumption is about the inside adduction coefficient whose radiative contribution is unknown: here the adduction

7 coefficient of the window and the wall was fixed to 8 W/(m 2 K). Once the view factors have been calculated, the only unknown quantity in eq. (9) is t int : hence, it is possible to calculate T rs according the eq. () with a good accuracy. Figure. Winter session. Panel specific power, water mean temperature and remainder surfaces temperature (calculated values), and surface panel temperature (average of the 7 measured points). Now the convective and radiative thermal exchanges of the panel can be easily calculated, and so the specific power that is reported in the graph of Figure, together with the water mean temperature, the remainder surfaces temperature and the panel surface temperature. Of course, the obtained panel specific power was quite low, due to the low thermal load of the room at the end of March. Sometimes it got negative: that is due to a T rs even higher than the panel temperature. It is worth stressing that the radiative contribution on the power exchanged by the panel is about 95%, a value well justified in winter operation when the ceiling panel temperature is higher than the air temperature and the convective liminar coefficient is very low. 3.2 Summer session In the summer survey the logging was carried out from the 6 th to the 25 th of August 2003 in a room with a volume of m 3. The only external wall, that is faced to south/south-west, has a thermal conductance of about 0. W/m 2 K; the double-glaze window has a thermal conductance of about 3 W/m 2 K and a total solar transmittance of 78%. The radiant panel (15. m 2 ), that is the 68% of the ceiling, is a built-in plasterboard sandwich, 15 mm thick, stuck on a layer of expanded polystyrene, 27 mm thick, that limits the back thermal losses. The polybutylene pipes, that are embedded in the plasterboard layer, have an outer diameter of 6 mm and a thickness of 1 mm. The panel thermocouples layout as well as the microclimate station position is sketched in Figure 1b. No one was present in the room during the survey and so the microclimate station was placed in the center of the room.

8 The following discussion will be condensed with respect to the winter one to contain the exposure. Thermal comfort evaluation The microclimate variables are reported in Figure 5: the thermal comfort was almost always satisfied with a mean PMV of about 0.2. It is worth noting that there were no internal gains during the survey, and so one can expect a slightly warmer condition with the gains of people, lights and other devices. Figure 5. Summer session. Microclimate variables. The radiant panel performance Of course, the panel specific power, that is reported in the graph of Figure 6, was always negative. Unlike winter operation, in summer operation the convective contribution on the panel thermal exchange is considerable (38%), due to the higher convective liminar coefficient. In summer radiant conditioning it is necessary to check that no water vapor condenses on the panel surface: to such an aim the inside dew point temperature has to be at least 1 C higher that the panel coldest zone. That checking was satisfied in the survey, but it is not here reported.

9 Figure 6. Summer session. Panel specific power, water mean temperature and remainder surfaces temperature (calculated values), and surface panel temperature (average of the 5 measured points).. THE SIZING CHECK With the information regarding the relation between the conductance and the adduction coefficient of the panel, obtained from the experimental results, we can proceed checking the sizing of the panel in the design conditions. For the city of Vicenza in winter time all that means an outside air temperature of 5 C and no internal or solar gains; in summer time the design outside temperature is 31 C and it is necessary to take into account the internal and solar gains. One can solve the detailed thermal balance of the surface i, according to the following relation: C i N ( t t ) + ( t t ) + ε σ F ( T T ) + q 0 in, i out, i con, i in, i a i j in, i in, j rad, i = j = 1 α with i j (12) where q rad,i is the radiative internal gain on the surface i. For the balance of the inside air one takes into account the ventilation and infiltration air volume flows, the thermal losses (or gains) of the external walls and the internal convective gains, q con,int : M ( ta tven ) + m inf ca ( ta test ) + Si α con i ( ta tin, i ) m ven ca, + qcon, int = 0 (13) i = 1 All the boundary conditions are summarized in Table 2, as well as the results in terms of inside surfaces temperatures, inside air temperature and power exchanged by the panel. Comparing the latter with the room thermal load (that was calculated fixing the inside air temperature to 21 C and 26 C, respectively for the winter and summer time), one can calculate the under or over-sizing of the panel in the design conditions. After all, the panel investigated in winter results slightly over-sized, whereas we found a correct sizing in the room investigated in summer.

10 Discussing the problem in the opposite side, that is the design procedure, there are two possible approaches. According to the former the panel power should be equal to the room thermal load and to obtain such a power a suitable supply water temperature should be selected. As regards the other approach the aim could be to achieve an appropriate PMV, for example PMV=0. With this procedure the supply water to the panel heating the room surveyed in winter could be lowered to 7 C, whereas the panel size for the room surveyed in summer should be enlarged to the whole ceiling, even if the panel is correctly sized for a PMV=0.3. Table 2. Boundary conditions and results of the panel sizing check. Boundary conditions Winter Summer Outside air temperature [ C] Sun-air temperature [ C] Supply ventilation air temperature [ C] Ventilation air volume flow [V/h] 6 Infiltration air volume flow [V/h] C p [W/m 2 K] 0.091exp(0.66α add,p ) 5e-8exp(0.972α add,p ) Supply water temperature [ C] Internal and solar gains [W] no 80 (con), 320 (rad) Room thermal load [W] Results Winter Summer Air temperature [ C] Panel temperature [ C] External wall temperature [ C] Window temperature [ C] Remainder surfaces temperature [ C] C p [W/m 2 K] α add,p [W/m 2 K] Panel power [W] Panel specific power [W/m 2 ] Over-sizing [%] 9% 1% PMV (centre of room) PMV (close to external wall) REFERENCES AA.VV., 1997, TRNSYS: a transient system simulation program, Solar Energy laboratory, Madison, WISC, US. Fanger, P.O., 1970, Thermal comfort, McGraw Hill, New York. ISO 7726, 1988, Ergonomics of the thermal environment - Instruments for measuring physical quantities, International Standard Organisation, Geneva. ISO 7730, 198, Moderate thermal environments, determination of the PMV and PPD indices and specification of the conditions for thermal comfort, International Standard Organisation, Geneva. Lazzarin, R., Crose, D., 2000, Il soffitto radiante nella climatizzazione ambientale, SG editoriali, Padova.