1 The solar physics

The Sun is a sphere made up of gaseous elements consisting of 80% hydrogen, 19% helium and 1% of all the well-known substances. It has a diameter of 1.39 ■ 109 m and it is located at a distance of 1.495 ■ 1011 m from the Earth. However, this dis­tance may vary by ±1.7% during the year because of the orbit’s eccentricity.

The Sun is characterized by two motions: a motion of revolution around the centre of the galaxy, which has a linear speed of 300 km/s and takes 200 million years to complete, and a motion of rotation around the axis, which lasts about 4 weeks.

Inside the Sun, numerous fusion reactions take place. The heat produced by these reactions spreads from the inner layers to the outer layers by convection, conduction and radiation. From the outer layers, the heat is transmitted to the surrounding space by radiation. Among the nuclear reactions that occur in the Sun, the most important is the one which converts hydrogen into helium; the mass of a helium nucleus is smaller than that of the four original protons and this mass defect is converted into energy.

The mass of the Sun is roughly 2 ■ 1030 kg. The areas at the centre of the Sun reach temperatures of about 8-40 million kelvin and a density 100 times greater than that of water. However, the density is extremely lower in the outer layers.

It is believed that the region between 0 and 0.23R (R = solar ray), which con­stitutes 40% of the solar mass, produces 90% of the solar energy. The area between 0.7 and 1R is called the convective envelope (temperature 5000 K, den­sity 10-5 kg/m3), because of the importance of convective processes in this layer. The photosphere, the outer layer from the convective envelope, is composed of strongly ionized gases, which are capable of absorbing and emitting through a continuous spectrum of radiation. Over the photosphere, there is the inversion layer, which is hundreds of kilometres wide and is made up of cold gases. Out­side the inversion layer, there is the chromosphere, which is 10,000 km wide, and the corona, characterized by a very low density and high temperatures (106 K) (Fig. 1) [1].

The electromagnetic radiation emitted by the Sun extends over a wide wave­length interval: from 0.1 nm to 104 m; however, the greatest part of that energy falls in the interval between 0.2 and 4 |jm. In particular, 95% of the energy which reaches the Earth is included between 0.3 and 2.4 |jm. The spectrum of the solar radiation is similar to a black body’s spectrum at a temperature of 5780 K, since temperatures at the surface of a star fluctuate between 4000 and 6000 K. Therefore, it is right to assume that the behaviour of the Sun with regard to radiation is similar to the behaviour of a black body at a uniform/regular temperature (Fig. 2). This temperature of 5780 K is calculated using the Stefan-Boltzmann law [1, 3].

Analysing the spectrum more carefully, one can notice that the greatest part of the radiation falls in (1) the ultraviolet band, which extends from 0.20 to 0.38 pm; (2) the visible light band, from 0.38 to 0.78 pm; and (3) the near infrared band until about 4 pm. Only 8-9% of all the solar energy which reaches the Earth falls in the ultraviolet band; 46-47% falls in the visible band while the remaining 45% falls in the infrared band [3].

Currently, the most suitable technology for electrical energy production by ther­modynamic systems is one which uses linear parabolic collectors [45, 50, 53].

These collectors are composed of a linear concentrator with a parabolic profile whose reflecting surface follows the Sun rotating on a single axis. The concentra­tor is fixed on a support structure which guarantees the correct processing during windy conditions and the action of other atmospheric agents.

The reflecting panel is normally composed of a common glass mirror with an appropriate thickness. Solar radiation is focused towards a receiving tube that is placed along the parabolic concentrator’s fire. The energy absorbed by the receiv­ing tube is then transferred to a processing fluid (thermal vector fluid) generally made of synthetic oil which is drawn up inside. The heat gathered is normally used as shown in the Fig. 86 (i. e. for the electrical energy production).

Figure 86 shows the processing scheme of a solar thermal electrical system with linear parabolic collectors using synthetic oil as the thermal vector fluid. In such systems, parabolic collectors are connected in series, generally in two par­allel rows which are a few hundred metres long and form a string that represents the unitary module of the system. The strings as a whole form the so-called solar field (Fig. 85). The synthetic oil pumped towards the receiving tubes comes out from the warmed up solar field at a temperature of about 390°C and then feeds a power unit (which is right at the centre of the solar field): the thermal vector fluid transfers the heat to a steam generator to start the processing of an electrical turbo-generator group. After delivering the heat, the oil (at 290°C) comes back

to the solar field to be warmed up once again. With the linear parabolic collector technology, a maximum system processing temperature of 600°C can be reached (but it also depends on the kind of thermal vector fluid used and on its tempera­ture when coming out from the solar field).

Nevertheless, the use of synthetic oil as the thermal vector fluid, which is the case in almost all solar systems with linear parabolic collectors, does not allow reaching temperatures higher than 390°C (as seen in Fig. 86) which has a nega­tive influence on the thermodynamic performance of the steam generator group. In these systems, the conversion efficiency for the conversion of solar energy directly into electrical energy is 15%. At present, solar thermal electrical sys­tems with linear parabolic collectors have typical dimensions; the capacity of these systmes can be in the range 30-80 MWe and they can also burn a certain quantity of fossil fuel (natural gas) to produce energy when there is a lack of solar energy, so these type of systems are hybrid systems, i. e. solar-fossil fuel systems [45, 50, 51, 53].

The maturity of this technology can be proved using the example of the Kramer Junction in Mojave Desert (California), where in 1984 this kind of solar thermal electrical system (SEGS I, Solar Electric Generating Systems) with a capacity of 14 MWe was realized. This system uses both linear parabolic collectors and natural gas as the fuel for overheating and sustaining the system in case of low radiation or breakdown.

An additional eight systems were constructed from 1984 to 1991, SEGS II-SEGS IX, reaching a total power of 354 MWe. In these systems, the tech­nology had been improved and costs were reduced such that the cost of elec­tricity generated was reduced from about 30 c$/kW h (in the first system) to 8 c$/kW h (in the last system). These systems have all been producing electric­ity and have added more than 13 TW h (billions of kW h) to the electrical net [45, 50, 51, 53].

sulnr

superheater

reheatcr expansion essel

Figure 88: SEGS IX scheme.

Linear parabolic collector systems have shown a few limitations which have not allowed their wider application. The principal problems are [53]:

• electrical energy production depends on the intermittence and variability of the solar source, which necessitates the use of fossil fuels to integrate the thermal energy production and therefore the need for solar-fossil fuel hybrid systems;

• the low conversion efficiency of the systems, which is due to the limited efficiency of solar energy gathering and the low processing temperature of the thermal vector fluid (<400°C);

• the high cost of the electrical energy produced, which is a consequence of the low efficiency of the systems and the high construction cost;

• the dangers posed by the use of the working fluid (synthetic oil) which is toxic and highly inflammable at the processing temperature.

Technological development projects have been planned in many countries. In 2004, the construction cost for a solar thermal electrical system with linear parabolic collectors was about 2500-3500 €/kWe, with a predicted 30% reduction in the medium term [51].

4.4.1 Tower system with a central receiver

This technology has overcome the demonstrative phase as an industrial prototype, but it has not reached the phase of trade maturity yet. The central tower system (see Fig. 89) makes use of flat reflecting panels (which as a whole form the solar field) called heliostats (Fig. 90). These panels track the Sun by rotating on two axes and concentrate the sunlight towards a sole receiver. The receiver is installed at the top of a tower (which is at the centre of the system) and inside a fluid (thermal vector fluid) flows to absorb the solar heat.

The thermal energy which is made available by this process can be used in dif­ferent processes, especially for the production of electrical energy.

In this kind of system, the thermal vector fluid can reach high processing temperatures (>500°C), which allows achieving high efficiencies in the conver­sion of solar energy into electrical energy. Generally, the transformation happens by exploiting the heat in a traditional water-steam thermodynamic cycle (see Fig. 91).

The central tower technology has shown its technological practicability in the production of electrical energy by the realization and the running of numerous small-sized systems (0.5-10 MWe) in different countries all over the world (Spain,

Italy, Japan, France and USA). From this experience, which has come with maturity, it has been seen that the best size for these systems is in the range 50-200 MWe. Different kinds of thermal vector fluid (such as water, air, melted salts) have been experimented with for years; however, until now the most suitable fluid for this technology has been the melted salts mixture composed of 60% sodium nitrate (NaNO3) and 40% potassium nitrate (KNO3).

Compared to synthetic oil, which is used as the thermal vector fluid in most solar thermal electric systems with linear parabolic collectors, the melted salts mixture, used in tower systems with a central receiver, has two important advan­tages: the fluid can reach a higher processing temperature (565°C) and it is possi­ble to install a thermal energy accumulation system, which can be created by piping the mixture heated up in the receiver towards an appropriately insulated storage tank (see Fig. 91). The sodium and potassium nitrate mixture can be heated until a maximum temperature of 565°C is reached (when the temperature is higher than 565°C, nitrates decompose into nitrites causing potential corrosion problems), which is much higher than the temperature of 390°C allowed by synthetic oil; this higher temperature allows achieving a better performance in the thermodynamic cycle for the production of electrical energy as well.

The elevated cost, the environmental risks and the inflammability which character­ize synthetic oil do not allow the storage of this warm liquid in such a volume needed to achieve an efficient thermal accumulation (actually, there is no thermal accumula­tion in the solar systems with linear parabolic collectors). Instead, the cheapness, non-toxicity and low environmental risks typical of the melted salts mixture make this fluid the most suitable for use in a thermal energy accumulation system, which solves the problem of the solar source variability and allows the production of electrical energy on demand so as to make the system more flexible [45, 50, 51, 53].

Figure 91 shows the processing scheme for a tower solar thermal electric sys­tem with a central receiver; it uses the melted salts mixture (described above) as the thermal vector fluid. Heliostats concentrate the sunlight towards the receiver

inside which the melted salts mixture flows; this mixture absorbs the heat and reaches a temperature of 565°C. The warmed thermal vector fluid is then directed to an insulated storage tank where thermal energy accumulation takes place before being picked up for the production of electrical energy. When it comes out from the warm storage tank (565°C), the melted salts mixture delivers heat to the steam generator which feeds an electrical turbo-generator. After delivering the heat, the thermal vector fluid cools down (290°C); at this point, it is accumulated inside another storage tank, waiting to be once again directed to the receiver. Eventually, the most important improvements introduced by this kind of system compared to those brought about by the system with linear parabolic collectors are as follows [53]:

• The thermal vector fluid is more safe since sodium and potassium nitrate (well-known compost) is neither flammable nor toxic (compared to the syn­thetic oil).

• The increase in the processing temperature of the thermal vector fluid from 390 to 565°C improves the performance of the thermodynamic cycle.

• The introduction of thermal accumulation solves the problem of the daily vari­ation in solar intensity. This provides clear advantages as regards the processing continuity of the turbine-alternator group and avoids resorting to the fossil fuel integration. Therefore, these systems are not hybrids but exclusively feed by renewal sources.

• The mirrors are made of compound materials (honeycomb) which are lighter, stronger and cheaper than the glass used in SEGS systems.

Although tower systems with energy accumulation are more efficient as regards conversion and require cheaper initial investment than the systems with linear parabolic collectors, a few disadvantages make their installation difficult on a wide scale and for high power requirements [53]:

• the conspicuous dimensions of the central tower whose height is proportional to the mirrors’ field extension and the system’s power;

• It is very difficult to concentrate solar radiation towards the receiver which is installed at a height of 100 m. By contrast, the focal length of systems with linear parabolic collectors is lesser than 2 m.

Without any doubt, one of the most important examples of this technology is represented by the experimental American system Solar Two of 10 MWe power, which was in operation from 1996 to 1999 in Dagget, California. Solar Two was the first system to use a melted salts mixture composed of 60% sodium nitrate and 40% potassium nitrate as the thermal vector fluid [8, 15, 42].

In Italy, as regards high-temperature solar thermal systems, the most relevant example was in the beginning of the 1980s with the construction of the world’s biggest solar power plant in Adrano, Sicily. This power plant, called Eurelios

(constructed within a CEE research project and thanks to the investment from an Italian-French-German society), has not been in use since an experimental phase which lasted for 6 years, from 1981 to 1987. Eurelios was able to produce a power of only 1 MW [2, 13].

Technological development projects are currently being implemented in USA, Spain (with the collaboration of a few countries), South Africa and Israel. In 2004, the construction cost for a solar thermal electrical system was about 4500 €/kWe, with a predicted reduction to about 2000 €/k in the medium term [51].

Tower systems with a central receiver have and will continue to have great importance in the field of both continental and in world energy in the near future. As proof of this statement is the PS10 system with its capacity of 11 MW power. This system, which began production in January 2007, located in Sanlucar La Mayor (Andalusia) is the biggest European solar thermal power plant. The PS10 Spanish power plant, whose overall cost is 35 million Euro, occupies 60 hectares of land and based on predictions it will produce 23 GW h/year at a cost of 0.1 €/ kW h produced. An interesting result when we consider that the best photovoltaic systems currently produce energy at a cost no cheaper than 0.23 €/kW h, which is close to cost of energy produced by fossil fuel systems (0.06 €/kW h) [25].

Observations on the thermal accumulation The introduction of a thermal accu­mulation system allows the elimination of the short transitory effects due to the irregularity of the solar radiation and also allows the release of the production diagram from the solar radiation diagram, as seen in Fig. 93 (where it is assumed that the electrical power installed is the same in both cases). The presence of the accumulation allows the use of a wider solar field, even if the electrical power is the same, to produce more energy and also a greater number of ‘equivalent annual hours’ of operation. These can go from 1500 h, typical of a system without the accumulation, to 2000-4000 h or more in a system with accumulation. A very big

storage tank would virtually allow the continuous production of energy. Actually, it is more appropriate to limit the accumulation to a storage tank which would allow 5-10 h of nominal power processing. This will allow users to plan the production of electrical energy to its best, concentrating it during periods of high require­ment (also increasing its trade value). In fact, as can be seen from the diagram in Fig. 94, the requirement of electrical energy in Italy has its peak during the evening — night hours and so it is delayed by nearly six hours from the solar radiation peak. This aspect is often more evident in developing countries. Generally, based on the weather forecast for two or three days (which is being introduced in the manage­ment system of the electrical generation park), it is even possible to optimize the energy production to make it available during the hours when energy costs more. An accumulation system also allows production on requirement, contributing to the creation of the required margin for the power stock of the net [45].

load factor: 25%

‘electricity

tidiostats

energy stored

accumulation of heat

Liu and Jordan extended their model of division, which is valid for monthly average radiation values, to the valuation of daily global radiation received on an inclined surface. If K is the daily clearness index, which is equal to H/Hex (i. e. the ratio between daily radiation and extraterrestrial radiation received on a horizontal surface on a specific day), the ratio between diffuse radiation and global radiation received on a horizontal surface on a certain day has to be cal­culated as follows:

D/H = 1.0045+ 0.04349K -3.5227K2 + 2.6313K3 se K < 0.75 (^

D/H = 0.166 seK > 0.75

The direct component is calculated by the difference:

B = H — D (44)

The daily global radiation on an arbitrary inclined surface can be calculated using the following expression:

E = Rb B + Rd D + Rr(B + D) (45)

where Rb, Rd and Rr are the inclination coefficients of direct, diffuse and reflected radiation determined, respectively, by eqns (29), (32) and (34), but for the chosen day [1].

Low-temperature thermal solar systems, which have been described until now and in which the energy transfer from the storage place to its utilization is realized by fluids moved by pumps and ventilators, are also called active systems. By the
expression ‘passive heating systems’ we generally mean all applications where the thermal hygrometric well-being conditions are obtained only by solar energy which is used without employing any conventional heating systems requiring elec­tricity or fuel. In merely passive systems, even the heat distribution and removal are realized by natural the phenomenon of conduction, convection or radiation, rather than using forced systems. Passive heating systems require the installation of wide glazed surfaces for solar energy interception and also structures with high thermal capacity storage function.

The efficiency of these systems is limited to the width of the glazed area which has to be correctly oriented, to the efficiency of the thermal storage realized by the walls and inner floors and eventually to the stored heat distribution towards the building parts characterized by scarce solar radiation. Currently, the realization of passive solar heating systems capable of guaranteeing the comfort conditions required in every inner room of a building seems to be an impossible goal both in cold climate zones and in mild climate zones such as Italy. However, its contribu­tion to the reduction of the yearly heating requirement could be relevant. Passive solar heating systems can be classified on the basis of the mechanism of energy transfer towards the heated room as follows [1, 3, 4]:

• Direct gain systems

• Indirect gain systems

• Isolated gain systems

The solar constant Ics is the average energy radiated by the Sun per time unit on a unitary surface situated outside the Earth’s atmosphere and perpendicular to the Sun’s rays. It measures 1367 W/m2 . Considering the atmospheric phenomenon of absorption and diffusion and the Sun’s inclination above the horizon, on the Earth’s soil the solar constant reaches a maximum of 1000 W/m2 (radiation on land, at midday, during a clear sky day) [1, 4].

The total power radiated by the Sun can be calculated as follows:

P = 4kRJIcs = 4n(150 • 109)21367 = 3.8 • 1026 W (1)

where Rm is the average distance between the Earth and the Sun [2].

The Earth intercepts only 1.73 ■ 1017 W of that power. Owing to nuclear reac­tions, a mass of 4.27 ■ 109 kg can be destroyed in a second; thus, nearly 0.0067% of the solar mass will be lost in a billion years [1].

Once we know the power emitted by the Sun, it is easy to calculate the heat produced internally per unit of solar volume:

q = P/(4/ 3)nR3 = 3.8 • 1026 /(4/ 3)n(7.25 • 108 )3 = 0.24W/ m3 (2)

where R (= 7.25 ■ 108 m) is the solar ray.

The quantity calculated above is a particularly low value considering that, for example, the human body’s heat production per unit volume is roughly 1400 W/m3 [2].

The hourly values of global radiation received on a horizontal surface Hh can be divided into diffuse components Dh and direct components Bh by Liu and Jordan’s method using the following expressions:

Dh/Hh = 1 — 0.09k if k < 0.22

Dh/Hh = 0.9511 — 0.1604k + 4.388k2

— 16.638k3 + 12.336k4 if 0.22 < k < 0.8 (46)

Dh/Hh= 0.165 ifk > 0.8

where k is the hourly clearness index defined by the ratio between the hourly global energy Hh received on a horizontal plane and the hourly energy received on a horizontal plane Hh, ex situated outside the atmosphere.

k=Hh/Hh>ex (47)

Hh, ex can be calculated using the equation:

Hh, ex = /J1 + 0.033cos(2rc«/365)]

• (cos L cos d cos h + senL send) (48)

The hour angle h is calculated at the centre of the considered hour or using the exact equation:

Hhex = 12/n/cs[1 + 0.033cos(2n«/365)]

• {cos L cos d(sen h1 — sen h2) + (h1 — h2)sen L sen d} (49)

h1 and h2 are the hour angles at the extremities of the said hour; in eqns (48) and (49) Hh, ex is expressed in W-h/m2.

Once we have obtained the value of the hourly global radiation Hh received on a horizontal plane, we can calculate the hourly diffuse radiation Dh. The hourly direct radiation on a horizontal plane is calculated by difference:

Bh=Hh-Dh (50)

The hourly global radiation received on an inclined surface turns out to be:

Eh=Rb Bh+Rd Dh+Rr(Bh+Dh) (51)

In this case, Rb is calculated at the centre of the said hour [1].

The direct gain system is the most common and simplest solution for a passive solar heating system: solar radiation enters the room through a glazed surface and directly warms it (Fig. 69). So, the living space works as a solar collector, but it must have the means and structures that are capable of absorbing and storing the intercepted thermal energy to keep the internal air temperature constant as much as possible. In this way, the daily overheating and the excessive decrease

in night temperature can be reduced. A direct gain system needs a wide glazed surface oriented southward to allow transfer of the winter solar radiation through direct communication with the living space. The southward orientation generally allows intercepting the greatest quantity of solar energy during winter, whereas in summer since the sun is very high, the transmitted radiation is less and it can be minimized by a suitably proportioned horizontal object (overhang). The choice of window components is very important in planning the solar heating system. Windows with a high heat transmission coefficient are preferred to maximize the quantity of intercepted radiation when the radiation is very poor and also to restrict heat losses.

A wide glazed surface oriented southward can cause overheating inside the house during the day and excessive inner air temperature fluctuations when there is no direct solar radiation (during the night or when the sky is cloudy). To solve these problems, it is important to use a thermal mass that is connected with the walls and floors whose surfaces and thermal capacities are well proportioned and also well positioned to intercept solar radiation and store thermal energy. During the day, the heat produced by the intercepted radiation is not completely released into the room; it is partially stored and released later after a delay of a few hours to stabilize the air temperature of the house. The storage thermal mass is generally made of masonry. Masonry materials for thermal storage are characterized by a high thermal capacity: cement blocks, concrete, bricks, stone, etc. To restrict dis­persion of stored energy inside the thermal mass, the brickwork walls are insulated on the outside, while floors are realized with a perimeter or an extrados insulation [1, 3,4].

Since the Earth’s orbit around the Sun is elliptical, the distance between them varies during the year, causing a ±3.3% fluctuation of the extraterrestrial radiation (Fig. 3). This radiation can be roughly calculated for every day of the year using the following equation:

where n(t) is the progressive number of the day of the year [1].

As regards Italy, there are medium-height solar radiation regimes with a big varia­tion between northern and southern regions. Keeping in mind that the parameters required to determine univocally the position of an intercepting surface are the surface’s inclination and its azimuth orientation, we now list the principal sources for the retrieval of radiation data.

‘La radiazione globale al suolo in Italia’ [10], a paper edited by ENEA, is with­out any doubt the bibliographic source which supplies, on a national level, the most detailed information about the average global radiation (i. e. the one which includes direct, diffuse and reflected components) received on a square metre of a horizontal surface per month and year. The same kind of information, which refers to a smaller number of Italian places, is also provided by UNI ISO 10349 international norms.

‘L’atlante europeo della Radiazione Solare’ [11] is without any doubt one of the most authoritative sources for the valuation of solar radiation received in a certain period of time on a surface which is exposed in any manner. This atlas gathers all the data supplied by national metrological offices. These data, gathered in maps and tables, are the result of a 10-year study. The atlas is divided into two volumes: the first takes into consideration the horizontal surfaces and the second the inclined sur­faces. The first volume reports, for every Italian place we consider, the values of the daily average radiation expressed in W-h/m2 or in kW-h/m2. Every place is charac­terized by its latitude, longitude and height above sea level.

As regards the design of solar panels, the results of the second volume appear to be much more interesting than the first. As a matter of fact, solar panels are usually arranged with a certain inclination on a horizontal plane, and the second volume reports the values of the daily average radiation (global and diffuse) per month and year for different positions of the intercepting surface. However, in the European Atlas of Solar Radiation, only the principal cities of each country are mapped.

To obtain the values of radiation received on variously oriented and inclined sur­faces, there are a few algorithms (which can be easily found in the currently avail­able design software) among which the most well known and used is that of Liu and Jordan which is discussed in par. 11, 12 and 13. The steps to get a correct extrapo­lation of the data for a surface which is positioned in any manner, starting from the values for a horizontal surface, are outlined in the UNI 8477 norms (first part) [4, 5].

As regards indirect gain systems, solar radiation does not directly enter the room that has to be heated up, but it falls on a thermal mass which is placed between the Sun and the living space. The solar energy absorbed by that mass is converted into thermal energy and then distributed inside the room in different ways. By the position of the thermal mass, we can distinguish two kinds of indirect systems: solar walls where the thermal mass is contained inside a wall and roof-ponds where the thermal mass is put on the roof of the room which is to be heated up. Indirect gain systems need a wide glazed surface oriented southward and the ther­mal mass used for storing the absorbed energy is placed behind it at a distance of at least 10 cm.

The storage is normally made of brickwork or water (with the latter put inside metallic, plastic or concrete waterproof containers) [1, 3, 4].

2.3.2.1 Brickwork solar walls Solar radiation received by a solar wall which is painted with a dark colour is absorbed causing a superficial heating up (Fig. 70). This heat, which is like a temperature damped wave, is transferred by con­duction to the wall’s inner surface and from there it spreads all over the room by radiation and convection. The delay and the temperature wave damping depend on the storage material and thickness. Dispersion of stored heat towards the outside is resticted by the insulation created by the air space between the glazed surface and the solar wall.

Trombe’s wall (Fig. 71) is different from the solar wall owing to the pres­ence of air holes in both the lower and the upper parts of the wall. In this way, the activation of a mechanism for natural circulation of air through the heated area is favoured. The warm air volume between the glazed surface and the thermal mass can reach high temperatures (about 65°C). The air holes in the upper part of the storage mass allow warm air to move up and enter the room, while the colder air which is inside the room is recalled inside the collector through the holes in the lower part of the storage mass. The openings should be located by means of dampers to prevent the reverse movement during the night [1, 3, 4].

2.3.2.2 Water wall The processing by a water wall (Fig. 72) is based on the same principle which regulates the processing by a solar wall, excepting that heat transmission through the wall also depends on thermal convection and not only on conduction. Because of the high thermal capacity of water and the inner convec­tive currents, which make it an almost isothermal heat accumulation, the system can work with a higher efficiency compared with brickwork solar walls. One of the most important problems is where to confine the liquid. Until now, bottles, tubes, watertight tanks, barrels, drums and cement walls filled with water have been used as containers (see Fig. 72) [1, 3, 4].

2.3.2.3 Roof-pond As regards roof-pond passive systems, the thermal mass is placed horizontally on the building’s roof (Fig. 73). The storage medium is water which is enclosed in small bags similar to little mattresses. They com­pletely or partially cover the roof which works as the ceiling of the rooms that are to be heated up. Water containers have to be placed in direct contact with the ceiling which sustains them to make the heat exchange between the inner room and the storage easier. During the hot season, the storage is exposed to solar radiation during the day; the intercepted energy is then transferred by conduction through the roof structure and directly exchanged by radiation from the ceiling of the room to be heated. During the night or during cloudy days, a mobile insulation mechanism covers the hot water and restricts its heat dispersion. In contrast to the systems with solar walls, systems with water walls are not always provided with a transparent cover to put on water. The use of translucent containers or glazed surfaces put on the water mirror is an efficient solution to reduce sensitive and latent (evaporation) heat losses when climate is very cold.

In places where there are high thermal ranges between day and night and where humidity is very low, the water storage on the roof can also be also for summer refresh­ing by insulating it during the day and exposing it during the night. The utilization of the water wall also presents numerous problems: besides the extra structural costs, the system does not guarantee sufficient advantages at high latitudes because of the reduced solar radiation intercepted by the horizontal plane; moreover, the stored heat can be spread by radiation only over the floor below the roof [1, 3, 4].