Category Archives: BACKGROUND

05 . Modelling of Earth-Heat-Exchanger

There exist as analytical as well as numerical methods of modelling to calculate such earth heat exchangers, but usually without the heat pump circuit.

Fundamental for all models is the Fourier differential equation

dt

c • P

(4.1)

with a :

Temperature Conductivity [m2/s]

c :

Specific Heat Capacity [J/(kgK)]

T :

Temperature [K]

W :

Heat Power Density [W/m3]

A :

Laplace-Operator

P :

Media Density [kg/m3]

A useful analytical model is the geometry shape factor model described by K. J. Albers 1991 [2]. This method neglects the Fick’s diffusion and Darcy streams, uses a simplified heat transfer equation and makes a superposition of temperature fields of the undisturbed earth (nonsteady) and the disturbed earth with the concrete pipes (steady).

The temperature curvature over the period of one year is given by a cosine function, which leads in the undisturbed earth to an exponentially damped temperature dependence on the deepness.

with x : Position under the Earth Surface [m]

t : Time [s]

a : Temperature Conductivity [m2/s]

Air temperature:

Tl = Tm + (Tmal — Tm) cos(2n—-p)

t0

Tm : Annually Averaged Air Temperature [°C]

Tmax : Maximum Average Month Temperature [°C] t0 : Time Period, here one year [s]

t : Time [s]

p0 : Phase Shift

(4.3)

with

For the temperature filed of the undisturbed earth, the following simplified equation is used

d 2T d2T

V2T = — + — = 0

TOC o "1-5" h z

(4.4)

Earth temperature:

Eyndisturbed

(x, t) = Tm + (T

2-n • ( )-x•

(4.2)

dx2 dy2

which lead to an isothermal solution with a shape (geometry) factor S

Q = A — S — (Ti — T2) (4.5)

The final result is the superposition of an unsteady and a steady temperature field

T(X y, t) = TE, stat (x> У) + TEinstaAX t) (4.6)

In order to integrate inner heat sources like the heat pump pipes, Mr. Jorn Herz extended this shape factor model. For this purpose he made the assumption, with respect to an analytical solution, that the heat power density is alternatively independent or linearly dependent from the temperature.

This would be possible, because of the dependence of the enthalpy flux on the earth temperature, which is at the inlet higher than at the outlet, with respect to the higher temperature difference. For simplification, the temperature decrease between in — and outlet is considered to be linear with the pipe length.

Figure 5 shows the geometric conditions of the temperature and a comparison (upper picture) to the heating circuit in the earth with constant difference.

In a next step, the linear curvature (red) will be segmented in constant average values (green), in order to use the shape factor model mathematically in the usual way.

Figure 5: Linear Temperature Dependence between In — and Outlet of the Single Heat Circuits

Further, Mr. Herz dealt with the heating circuits a) as line shaped (one dimensional) heat sources and b) cylinder like heat sources (heat source and drain method), using the formulas

T (r, t) = —

О? 0

4 n-A-L

Ei

f — r2 ^

4 — a-1

V У

(4.7)

a)

u

with the exponential integral Ei(-Z) = J——————— du

Z u

T (r) = To +

Q o

2-n-A-L

-ln

r

r0

У r )

(4.8)

b)

The total temperature field of an earth pipe with the heat sources of the heat pump results as a superposition of all single calculated heat circuits with different segmented heat fluxes Q and the temperature field of the heat drains.

For further details and explanations and insight see the diploma thesis of Jorn Herz [1] and the dissertation thesis of J. K. Albers [2].

1. Measurements

All Measurements were done manually, because the automatic measurement system is already not completed.

Manual temperature measurements were done on

• the heated earth area

• the surfaces of the concrete pipes

• inner temperature of the concrete pipe (air flow)

with respect to the working status of the heat pump.

The Combination of Solar Thermal Collectors and. Heat-Pumps in a Compact Unit for Passive Houses

Bernd Hafner

Viessmann Werke GmbH & Co KG, 35107 Allendorf
DrHf@ Viessmann. com

Passive houses have proved a high potential of energy saving. The actual discussion is at which costs the new concept of thermal insulation of the passive houses can be realised. On the one hand, some components e. g. windows become more expensive. On the other hand, the low heating demand allows some economies in the installation of the heating system. For example an air heating working with a normal ventilation rate of 0.4 h-1 is sufficient to cover the heating demand of the house. An exhaust air heat pump saves installation costs as it needs only electricity and the exhaust air of the ventilation system.

In passive houses the heating demand is usually lower than the energy demand for hot water production. The heating system should take this in account. The solar system for domestic hot water production has a relatively high potential for energy saving. But the space for the solar installation, specially the storage, is expensive.

Approach

The base of the compact unit Vitotres 343 is a ventilation system with a heat recovery as it is required for passive houses. A focus is set on the efficiency of the ventilation system: a counter-flow heat-exchanger and fans with DC-motors are used.

An air heat-pump with a relatively low heating power of 1.5 kW takes the exhaust air coming out of the heat-exchanger as a source. A flow of ambient air is added to the exhaust air to maintain a constant flow-rate of approximately 300 m3h-1 over the evaporator independent of the ventilation rate of the house.

A 250 litre storage tank for the domestic hot water production is included in the compact unit. The heat-pump works on an internal hydraulic heating circuit filled with water-glycol mixture. The hydraulic circuit can be switched from air-heating to hot water production. The storage tank can also be heated by a solar collector.

One of the major challenges during the development of the compact unit were the required dimensions: height about 2 m, width 60 cm and depth less than 70 cm. Already a 250 l storage tank has a height of 140 cm. And still it is almost impossible to include two heat — exchangers with a sufficient surface for a solar collector and a heat-pump in this storage tank without neglecting the companies guidelines for this heat exchanges (no use of copper in drinking water, big diameters of the pipes to reduce furring).

An other solution was developed: Both, the heat pump and the solar collector work on the same heat exchanger in the storage tank (see figure 1). A new control strategy for the heat pump and solar collector avoid the simultaneous heating of the storage by both sources.

Natural Ventilation

Today, in well-insulated buildings, ventilation and cooling may account for about 50% of the energy requirements and a well controlled and energy efficient ventilation system is prerequisite for low energy consumption for the building.

The ventilation system for Albertslund Library is based upon natural driven forces caused by the temperature difference and the differential pressure between indoor and outdoor. The natural ventilation system is fan assisted for assuring a comfortable indoor thermal climate on sunny days with rain and high gusting winds.

The success of hybrid ventilation system depend on utilizing the benefits of natural ventilation and that the design of the system is integrated, from the start and in subsequent design stages. This is the case for Albertslund library where a close teamwork with integrated energy design as the main topic was initiated from the very beginning.

1.3 Design

The design of the new library is characterized by big open areas and a high roomheight, which gives a good possibility for use the building for fresh air distribution and gaining good air quality. Renovation of the facade and installation of skylights makes the building highly suitable for natural ventilation by using the facades as fresh air intakes and remove the exit air through the skylights. Thereby, the building will act as a ventilation duct for distributing the fresh air inside the library. The fresh air is then distributed by displacement.

Fresh air intakes are situated along the facade between the facade windows. A convector is integrated in the intake for preheating of the fresh air, as seen in figure

11. The fresh air intakes are controlled with a damper in each intake which all are connected and controlled by the BMS, see figure 3.

The exit air is ventilated to the outside through openable windows in the skylights, see figure 11. The openable high performance windows are regulated by wind direction, indoor temperatures, indoor CO2 level and rain by the BMS. In both sides of each skylight hidden panel convectors are installed for avoiding cold draughts and to contribute to space heating of the library.

Figure 11 Sketches of the fresh air intake on the left and the skylight on the right.

The size of the openable area in the skylights are dimensioned after a demand of rate of air change of 3 h"1 (24 000 m3) found from the internal heat and CO2 load. The program ContamW 2.0 has been used for dimensioning the size of the openable area in the intake and exit. Is has been found that the openable area needed in the intake is minimum 7 m2 and that 20 mi2 of openable area is needed in the skylights for exist. Here the openable area has been downwind prioritized.

1.4 Heating season

During the heating season the air change is kept at a minimum by regulating the damper in the fresh air intake and thereby the amount of fresh air after the CO2 level during opening hours. The natural driven forces will keep the CO2 at a satisfying level during the heating season. A convector integrated in the fresh air intake assures that the temperature of the incoming fresh air minimum is 18 °C so

cold draught is avoided. Fan assistance is used when the openable windows are shut because of high winds or rain. Furthermore, openable high performance windows are used in the facade with user control for removal of excess heat during summer.

1.5 Summer

Outside the heating season the natural ventilation system is controlled as during the heating season. However the dampers in the fresh air intakes and the openable windows are regulated by the indoor temperature during opening hours. When needed, the natural ventilation system will also be active during the night with a lower setpoint for indoor temperature then during the opening hours thus cooling the library. Preheating of the ventilation air is shut off outside heating season.

During very hot and sunny days where the natural ventilation cannot keep the indoor temperature at a satisfying level, the fans which are installed in the gable of every 2nd. skylight, will kick in.

Modelling of ASW and heat transfers

The ASW is schematized with two slabs, A outer and B inner, delimiting a duct into which air flows (see Fig. 2).

L is the duct length (in the air flow direction), h is the width, d the thickness, AF=hL the area, D=2dh/(d+h)s2d the equivalent hydraulic diameter. Later on the following quantities will be used:

■ G (Wm-2) mean solar radiation intensity,

■ a the absorptivity to solar radiation of the outer face of the slab A,

■ k (Wm-1K-1) air thermal conductivity,

■ Te sol-air temperature,

■ Ti indoor air temperature,

■ T1 temperature of the wall A inner face,

■ T2 temperature of the wall B inner face,

■ Ra and RB thermal resistances, respectively, of the slabs A and B,

■ Rtnv thermal resistance between the indoor air and the outdoor one in the case of non- ventilated duct;

■ Rt thermal resistance between the indoor air and the outdoor one in the case of ventilated duct;

■ Re thermal resistance between the ventilated duct and the outdoor environment;

■ Rcd closed air duct thermal resistance;

■ re and ri thermal resistances, respectively, of the wall’s outer and inner surface.

All the thermal resistances are given per surface unit (m2KW-1). We have: Te=T0+areG with T0 outdoor air temperature in the shade and Rtnv=RA+RB+re+ri+Rcd. The following dimensionless parameters are also used: z=Re/Rt and x=Rtnv/Rt.

State conditions are here considered to be steady and heat transfer schematized as one­dimensional. The case of ASW in which the duct thickness is small and the air flow inside is laminar is studied. Under laminar flow conditions the inlet phenomena, both dynamic and thermal, can be disregarded, if [12]:

Gz < 20 (1)

where Gz is the Graetz number, defined by:

Gz = — ■ Re — Pr L

with Pr Prandtl number and Re Reynols number (referring to the hydraulic diameter D) concerning the air flowing into the duct. Under these conditions, when the duct slabs are at

the same temperature Tp (uniform), the Nusselt number (Nu) is constant (independent from Pr and Re) and it results to be [12-13]:

Nu = — = 7.54 kp

with p thermal resistance between the inner faces of the duct slabs and the air flowing into the duct itself. From the previous relation it follows that either of the two slabs shows, compared to the fluid, a thermal resistance given by:

p =

and the heat flux q (Wm-2) absorbed by the fluid results: q = 2 (Tp

temperature of the fluid mixing [12].

The more general problem concerning the heat transfer within a duct, inside which the fluid flow is laminar, with the two slabs delimiting the duct at different temperatures is discussed in [13]. In [10] a simple and intuitive procedure which makes allowance also for the radiative heat transfers (characterizable by a radiative resistance Г) is reported in order to calculate the flux q2 coming into the room through the ventilation duct; the following relation is obtained:

The mean heat flux Q0 coming into the room when the ventilated duct is closed is:

Q0 = (Te — Ti )/Rtnv

The study of an ASW energy performance can be carried out using a percentage saving S, defined by [4, 8]:

S = (Q0 — Q)/Q0

The meaning of S is strongly intuitive, particularly when it assumes values from 0 to 1; negative values of S clearly show that ventilation is not convenient.

If the air temperature at the duct inlet is assumed to coincide with the air temperature in the shade T0, the air temperature inside the duct can be written as follows [8]:

f

T(x) = Tm +(T0 — Tm )exp — Xx

( ) m v 0 m} p[ yL[H + z(1 — z)]

where: Tm=zT+(1-z)Te and Y=cRtnv, with c specific heat capacity rate. The quantity H is the radiative correction factor defined in [8]. The radiative factor H is due to the fact that the introduction of the surface thermal resistance heat transfer coefficients (instead of the convective ones) is not sufficient for quantifying entirely the radiative heat transfer inside the duct.

Project SOLABS

The European project SOLABS is developing a novel coloured collector for building facades.

This new absorber will be made of steel, coated with weather resistant, coloured selective layers, and operated as an unglazed collector. It will be designed from start for integration into the building envelope.

The innovative architectural approach is to consider this solar element as an integral part of the building and to design it as a cladding element, taking inspiration from the existing products proposed by the metal cladding market. These claddings can be roughly divided into three basic types, all already accepted and widely used in building fagades:

• Cassettes (rectangular panels with a ratio of 1:1 to 1:4)

• Tongue-and-groove panel planks (20 to 40 cm width, ~4m. max length) • Profiled metal sheets

Fig. 2 Cassette claddings

As profiled sheets involve a complicated technical adaptation of the heat collection system, they are not further considered, but the cassettes and the planks seem both to be reasonable formal options.

The “cassettes”(Fig. 2) are well suited for cladding large wall surfaces. As the choice of a right dimensioning of the module is the base for an appropriate cladding, it is important to define the proper dimensions, or range of dimensions, to offer to the architects. The more flexibility the more freedom you give to the designers, but probably at an increased cost for low standardisation. This point is even more crucial in the case of an element that is technically more complex than a simple cladding, as there are thermal constraints as well as water connexion issues to be addressed.

The trade-off limit between flexibility and standardization is difficult to draw on objective grounds.

Fig. 3 Panel planks cladding.

The "panel plank” (Fig. 3) shape seems to be more promising in that regard, thanks to the flexibility of use it ensures even though it is a standard element. Due to its very structure, the panel plank offers a good flexibility in the vertical dimension:

3 0mm

10mm

0 mm

Fig. 4 Variable width of the reveals

With a small plank width (20-30 cm.) and a variable width reveal (Fig. 4), this system allows covering almost any surface height.

Cut-to-length elements should be feasible in a reasonable range and would give the needed flexibility in the horizontal dimension.

Those formal characteristics make the plank element one effective answer also to the problems related to the integration of solar thermal collectors in renovation.

On the colour approach, the new palette will be the result of a complex trade-off between technical feasibility, thermal efficiency, rationalization and architect’s demands (expectations / wishes).

The R&D work underway in the project aims at minimizing the inevitable efficiency losses inherent to the back reflection of a part of the solar spectrum, needed to give the collector a "colour". Architects on the other hand would like to have the usual freedom of choice they enjoy when selecting a "normal" cladding.

The broad range of questions related to the design led us to organise a large survey, including web-based questionnaire and personal discussions with architects of diverse experiences from all Europe. This survey (Fig. 5) is still currently underway to help defining the formal preferences related to the shapes, colours, details and fagade positioning of the collectors, and the consequent possible paths the project could follow. (http://www. solabs. net)

The preliminary results of the survey and the high level of interest received during the personal interviews seems to confirm that the approach taken (cassettes, planks…) is correct and that the project is going in the right direction, opening wide perspectives for the years to come.

Unglazed solar panels in renovation.

B

21. Which solar cladding shape type will be

more appropriate for the renovation of the facades of the building beside? :

a. Cassette modules.

b. Panel planks.

22 . Would you cover with solar cladding all the available South (S-E, S-W) exposed surface?

b. I will cover only fagade A.

c. I will cover only fagade B.

23. Evaluate the importance of having dummy elements for the covering of non exposed facades:

— — +- + ++ I

A

Fig. 5 Example of a survey’s webpage

Simulation

To identify promising fields of application, a parameter study with a building simulation pro­gram was carried out. For this reason we implemented the possibility to calculate nonlinear thermal properties of construction materials in the simulation environment esp-r. We fo­cused on applications with PCMs in the interior wall materials to prevent overheating and reduce the cooling load in summer and heating energy in winter.

Especially the melting temperature was varied with respect of the application. Whether you want to prevent overheating or save heating energy, a different choice of the melting tem­perature is necessary.

As an example we simulated the thermal performance of a typical lightweight office. Important is the possibility of discharging the storage at night, therefore a night ventila­tion was modelled (ac/h=4). During daytime, only the minimum needed air-change-rate of ac/h=1 was assumed and no active cooling device is installed.

29

180 181 182 183 184 185 186 187 188 189 190

time [day of the year]

Figure 3: profile of surface temperature — building simulation

without PCM————————

The PCM was modelled with a melting temperature of 25°C with a melting range of 2 K, mixed to interior plaster. Fig. 3 shows clearly, how the temperature of the PCM-walls starts

04

17

180 181 182 183 184 185 186 187 188 189 190

time[day of the year]

Figure 4: profile of resultant room temperature — building simulation

rising slower at the beginning of the melting range (24°C-26°C). The stored latent heat is released during night, then leading to higher wall temperatures. In addition to the energy savings by the reduced cooling load, the lower surface temperatures of the walls result in an increased room comfort (Fig. 4). Under certain conditions, an active cooling device may be even unnecessary.

For the heating case, this effect of smoothing the peaks in wall and room temperatures is reachable too. But only for the transitional time in spring and fall there is enough surplus of heat to store in the PCM to expect, so that the overall energy saving for heating isn’t to big.

Daylight simulations with advanced software tools: the case study of a university library in Pisa

B. Angeli, F. Leccese and G. Tuoni

Department of Energetica “Lorenzo Poggi” — University of Pisa

Faculty of Engineering — Via Diotisalvi, 2 — 56126 Pisa (Italy)

e-mail: barbarangie@lcheapnet. it; f. leccese@ing. unipi. it; g. tuoni@ing. unpi. it

The results of a case study on the use of daylight for the lighting of reading and consultation rooms of a new library, being part of the Scientific Pole of the University of Pisa, are shown.

Introduction

In the last few years the growing electric energy consumption in non-residential buildings has required an accurate daylight analysis to reduce the demand for artificial light and improve the visual comfort of the indoors [1-2]. It is also well known that daylight conditions are able to improve the human life quality, the indoor comfort level and to raise the productivity at work places [3-4]. In order to understand the importance of well designed glazings (i. e. dimensions, exposure, glass properties), it will be enough to think that their subsizing involves an increase in the demand for electric energy due to the integrative artificial lighting, while an oversizing weighs on the energy consumption for cooling and heating.

In a recent paper [5] the authors have carried out an extensive daylighting investigation of a new covered market in La Spezia (Italy) using ADELINE Software Package and have achieved some useful directions since the beginnings of the preliminary design process.

In our country, the inadequacy of the library system and the need of its growth are often related to the theme concerning the recovery and the reuse of old buildings, localized in town central areas generally fully edified and realised at first for performing other functions. The Italian library system turns out to be, therefore, involved more and more often in this dialectic "old building — need for reuse”, in spite of the numerous reticences on this topic.

In this paper some of the results of a case study on the use of daylight for the lighting of reading and consultation rooms of a new library, being part of the Scientific Pole of the University of Pisa, are illustrated [6]. This library is now being built reusing the pavilions of the ex- “Marzotto’s Textile Industry” close to the town medieval walls and just a few minutes far from the town centre and the leaning tower. The use of Adeline has allowed studying the illuminance on the work planes both under overcast sky and under clear sky conditions and, then, choosing the best distribution of desks and furnishings (i. e. shelves for books), avoiding dazzlement phenomena and reducing to the minimum the use of artificial lighting during the library working hours.

Calculation of the view-through index

The potential to view an object in the surrounding environment through a window can be hindered by:

• low luminous transmittance of the window;

• significant amount of diffuse transmittance component;

• back reflection from the window;

• obstructing elements in the window.

1view—thr

— f (?

prim—reg ’ ^scat—diff , Pback )

Since these individual influences are combined into a single complete equation, the total equation for the view through index looks complicated [4]. The complexity of the equation, however, has no negative effect on the simplicity of the input and output, and the total equation can be summarised as follows:

lum

The following parameters are involved:

Tprim-гед the regular portion of transmittance;

Tscat-diif the scatter or diffuse portions of transmission;

pretro the back reflectance (from indoor to indoor) of the daylight product;

Bobstr the disturbance factor for the blockage;

clum the correction factor for the accommodation to the level of luminance from

the light through/from the daylight product due to transmission and back reflection.

The disturbance factor, B0bstr, is calculated according to:

f por obstr

X ref

Bobstr Є

In which:

fpor is the porosity factor of the obstruction (i. e. the transparent part of the product’s

surface);

Xobstr is the characteristic length of the obstruction in the daylight product;

Xref is the reference length for obstructions in the daylight product.

The view through index value varies between 0 and 1, corresponding to vision denied and perfect view, respectively.

The view through index calculation requires goniophotometric measurements, therefore the index was calculated only for samples 12 and 13.

Table I Calculation of view through index

Parameter

Sample 12

Sample 13

Xostr (mm)

1.05

0.77

fpor

0.669

0.738

Bostr

0.933

0.960

Pretro

0.049

0.112

clum

0.623

0.767

^vrim-reg

0.542

0.676

A@prim^scat-diff

@width-scat-dif

2.6°

2.5°

scat-dif

0.079

0.088

fiew-thr

0.827

0.862

The next table I shows the value of the relevant parameters and the final view through index for sample 12 and sample 13.

The results show that the presence of the decor on the samples tested has a small influence on the capability of view through the window (index values not far from unity), and that sample 13 should allow a better vision of the landscape than sample 12 (index values respectively of 0.862 and 0.827).

Tests of the Adsorption Process

Evaluation of tests of adsorption cycles showed an energy output of exactly 115 kWh/(m3 silica gel) in the adsorber 1 and 123 kWh/(m3 silica gel) in the adsorber 2 with a flow temperature out of the adsorber of 32°C minimum. The power values and energies during an adsorption cycle are presented as an example in figure 4. The high power peaks in the evaporator heat exchanger result from the high flow temperature in the collector loop. The average power of the adsorber heat exchanger during the test, which could be supplied to a load in a real application, equaled 2.87 kW with a maximum value of 3.9 kW. In this test, the energy input (evaporator) was

24.3 kWh and the energy output (adsorber 2) 27.4 kWh.

Simulations have shown that a storage tank density with a maximum of about 150 kWh/(m3 silica gel) can be expected with the type of silica gel used. With a density of 790 kg/m3 and a silica gel mass of 808 kg (dry substance) in adsorber 2, a storage tank density of 123 kWh/m3 was achieved experimentally. The theoretically calculated value of around 150 kWh/(m3 silica gel) could be reached with an optimized operating strategy.

Nevertheless, it must be realized that energy densities achieved with commercial silica gels under technical conditions are rather insufficient. The energy density that has been reached experimentally, is only in the range of latent heat storage with about 1.8 to 2.2 times the value of water. This means that the storage volume is cut

Figure 4: Power and energy curves of adsorber 2 during an adsorption process.

in half compared to a system with water. Modified sorption materials were tested by the Fraunhofer Institute for Solar Energy Systems, which achieve an energy density which is 5 times higher than water under technical conditions. But up to now, these materials showed insufficient stability under operating conditions. The development of stable modified sorption materials will be an important task and a great challenge.

Studies of a Heat Pump Using Dual Heat Sources. of Solar Heat and Ambient Air

Sadasuke Ito, Toshiyuki Matsubayashi and Naokatsu Miura
Department of System Design Engineering
Atsugi, 243-0292 Japan
Tel/Fax +81-462-91-3091
Email: ito@sd. kanagawa-it. ac. jp

1. Introduction

Increased use of renewable energy such as solar energy is important to prevent global warming. The aim of the demand of solar heat in 2010 is 19.1 Mm3 in equivalent oil, which is about 4 times of that in 1997, to achieve the goal proposed in the Kyoto Protocol. However, the demand of solar heat has not increased since that time. Much more effort is necessary technically and politically to increase the demand of solar heat. Solar heat for hot water supply is usually collected by a solar collector though which a heat transfer medium of air, water, or antifreeze liquid flows. Heat loss from the collector increases as the temperature of the heat transfer medium increases. Therefore, useful heat can not be obtained when the solar radiation becomes too small.

Solar collectors can be used also as evaporators of heat pumps for raising the evaporation temperature of the refrigerant by solar heat to increase the thermal performance of the heat pumps. Charter and Taylor0 reported on such a system using flat-plate collector without a cover. A heat pump can be operated efficiently even when there is no or small solar radiation if the collectors have fins to absorb heat easily from the ambient air.

Ito et al2) demonstrated that solar collectors with photovoltaic modules on the flat surfaces can be used for generating electric power and collecting solar heat at the same time. These collectors, which are called by PV/T panels or panels, can give high efficiency of conversion of solar radiation to useful energy. Heat is absorbed from the ambient air also when the refrigerant in the panels evaporates at the temperature less than the ambient air temperature. If the panels are set on the roofs of houses, it is difficult for the panels to absorb heat from the back sides. In order to operate the heat pump efficiently when there is no or small solar radiation, we propose to use an air-refrigerant heat exchanger together with PV/T panels as the evaporators.

Roll-bond type collectors without photovoltaic modules on the surfaces are
used in the present studies. Experimental and analytical works were done previously on the performance of a heat pump with these panels3). In the present study, the thermal performance of a heat pump using the panels and an air-refrigerant heat exchanger is investigated and the effectiveness of using the heat exchanger is discussed.

2. Experiment

Fig.1 Experimental apparatus.

Fig.1 shows the experimental apparatus. The heat pump system is mainly composed of panels (evaporator), an air-refrigerant heat exchanger (heat exchanger, evaporator), an electrically driven rotary compressor with rated capacity of 250 W, a cylindrical condenser made of two copper tubes soldered together, a C-charged type thermostatic expansion valve at each inlet of the evaporators. Refrigerant 22 circulates these components. The panels are an aluminum roll-bond type. Three panels with the total area of 1.91 m2 are connected in series. The panels are mounted on plywood and installed at 50° tilt angle facing south on the roof of a house. The heat exchanger is made of a

copper tube of 10 mm outside diameter with many aluminum fins with thickness of 0.1 mm. The height, width and depth of the heat exchanger is 200 mm, 230 mm and 45 mm, respectively. A fan with a capacity of 8 W is used to draw the ambient air though the heat exchanger. Either one of the evaporators or both arranged in parallel are used in an operation of the heat pump.

In an actual system, a thermal storage tank would be used. In the present study, a constant temperature bath is used instead so that the thermal performance of the heat pump can be investigated under the condition of a constant temperature of the water flow at the inlet of the condenser. The flow rate of water is set to 3.0 l/min. The temperatures and pressures at various points are measured by using thermocouples and Bourdon tube pressure gauges, respectively.

3. Analysis