Category Archives: Sonar-Collecttors

Automation of Central Receiver Systems Plants

Figure 1. Detail of the operation of the heliostat field and CRS at CIEMAT-PSA

From the point of view of automatic control, a Central Receiver System (CRS) plant may be decomposed, almost in a first approximation, in two subsystems. The first one is the Heliostat Field (HF) and the second one is the Receiver and Power System (RPS), sited at the top of the tower. Figure 1 shows both components interacting. The HF reflects incoming solar radiation in predefined aimpoints. Following this natural decomposition the control system of the CRS is divided into two functional blocks: the HF Control System (HFCS) and the RPS Control System (RPSCS).

each heliostat of the HF in the desired

The basic objective of the HFCS is to position coordinates all the time, depending on the PS demands. The global objective of the HFCS is the generation of a uniform time-spatial distribution of the temperature onto the volumetric receiver (in the case of the PSA CRS), by controlling the timed insertion of group of heliostats associated to predefined aimpoints in the receiver, by modifying the aimpoints coordinates and by changing heliostats from one group to another during operation. This is accomplished by applying an aimpoint strategy over the HFCS [1].

From the control viewpoint, due to the requirements of the operation, the HFCS can be considered a hard real time system and thus, it requires a real time support system [2] to guarantee the adequate positioning of the heliostats at due time. Moreover, the HFCS must permit easy integration with the RPSCS, without losing the real time capabilities. The work that is being developed at present in PSA-CIEMAT is oriented towards the development of a real time distributed control system for the HFCS. The RT-CORBA [3] technology helps to integrate heterogeneous computing platforms and has been selected to implement the communications between the different data acquisition and control sytems. In this case, the RPSCS has been implemented in LabView over Windows XP OS. This control system is described in [4] in more detail. The HFCS is being developed in C++ language using the ISO POSIX C 1003.1c API interface, over LynxOS operating system. The real time logical interface to the PSCS is being implemented using The Ace ORB (TAO) RT-CORBA implementation [5-8]. The TAO CORBA distribution is widely used for industrial and scientific purposes (http://www. cs. wustl. edu/ schmidt/TAO. html).

Preliminary tests have been performed at CIEMAT-PSA using a Cluster Beowulf platform based in a switched 100 Mb/s Ethernet network and low computing resources nodes. The experiments have successfully demonstrated the low overhead and latency that TAO introduces in two-way inter-objects calls. Depending of the test performed the total cost in time of the overhead introduced by CORBA layers (at client and server objects) and network delays are lower than 5 ms in the worst cases [9].

Using RT-CORBA in the HFCS software design, a real time logical interface to the physical field (heliostats) is provided to the rest of computers and systems connected to the network domain. These computers may interact with this logical interface with independence of the operating system and programming language, which augments and promotes the heterogeneity and overcomes the constraints imposed by proprietary technologies. An example of interaction promoted by RT-CORBA is the periodical request of coordinates from each heliostat of the HF from the LabView application in the RPSCS (figure 2).

Ethernet Network RT-CORBA

Power Stage Control System He liostat Fie 1 d C otitrol System

LabView Interface IEEE POSIX 1003 Лс

Soft Rea] Time Operating System Hard Real Time Operating System

Figure 2. Distributed control system based on RT-CORBA middleware

The main objective of the RPSCS is to regulate the pressure and temperature of the steam generated with the heat collected in the solar receiver. Figure 3 shows the components of the RPS at the CESA facility at CIEMAT-PSA. Briefly it is composed of:

• Solar receiver.

• Storage tank.

• Air-water/steam heat exchanger (HEX).

• Process load (in the figure formed by a steam turbine and a condenser).

• Actuation components: K1 and K2: valves; G1 and G2: blowers; B1: water pump; V1: steam valve.

Figure 3. Diagram of the power system of the CESA facility at CIEMAT-PSA (courtesy of J. D. Alvarez)

The system operates as follows: the concentrated solar radiation heats of the air flowing through the volumetric receiver. The air temperature gradient through the receiver can be controlled via the actuation over K1 and G1, so the output temperature of the receiver is regulated controlling the air mass flow rate m1 over the K1-G1-Receiver branch. This is the first control loop. The control algorithm implemented at present is an classical PID with anti-windup controller.

In a similar way, the air mass flow rate entering the HEX is controlled by acting on K2 and G2.

The second control loop must control the energy flux at the HEX inlet, by extracting/impelling air from/to the storage tank depending among other variables on the solar radiation conditions, receiver state, etc. This control loop is also based in a classical PID with anti­windup controller.

The RPSCS is under development at present. One of the steps that is being carried out is the development of a reliable dynamic model of the process. This model is being implemented using the object oriented modelling language for physical systems Modelica [10]. Figure 4 shows the model interface with similar layout than the real system. The object oriented methodology preserves the original topology of the real plant in the model, so any component from the real plant can be described by a model representing it. Finally all the models are connected in the same way the components are connected in the real plant.

From the control viewpoint, the RPSCS signals are:

Thermal Storage Tank

Figure 4: Modelica model of the Power System

Even in the case of good control of the conditions of the inlet energy into the evaporator, variations in the outlet steam conditions in the water-steam circuit are produced. To control the outlet steam temperature/pressure it is necessary to control the pump B1 and valve V1. The control loop of the pump B1 is characterised by having a long delay (the length of the tubes is about 1330 m) and thus the actual control system based on a three state on/off controller is being changed by a Smith Predictor-based control loop [4].

• Mass flow rate through the blowers. This variable could be controlled by manipulating u1, u2, u3 and u4 represented in figure 3. This assumption is nearly true due to the almost ideal mass flow generator characteristic observed in the blower-valve sets in both loops: storage-receiver and storage-HEX mesh.

• Incoming inlet radiation at the solar receiver. This is a perturbation variable and the source of energy for the whole process.

• Mass flow rate through the water pump. The characteristic of this pump is being studied, and it seems to be highly nonlinear.

• Steam valve opening fraction.

The model output signals are all the variables that will be controlled: pressure and temperature at the inlet of the air-water heat exchanger; pressure and temperature at the inlet of the process load (main control objective), and the temperature of the storage tank. These controlled variables are also shown in figure 3.

Laboratory test results

SolarChill

03-03-2004

3-03-2004 06:00

3-03-2004 10:00

3-03-2004 14:00

0

1000

900

800

700

E

600

500

400

T3

ro

300

200

100

—— Current

—— Irrad

3-03-2004 18:00

The vaccine cooler has been tested in climate chamber at DTI. The holdover time for the vaccine cooler was measured to be about four days at 32 °C ambient temperature. For the upright version, the hold-over time is one day less due to the geometry and smaller ice volume. The tests have been used to determine the necessary PV panel size for the selected locations. As the critical parameter is the minimum current for start of the compressor, it was decided to use a panel with a short circuit current of 2.5 times the start current. In this way it is ensured that the compressor will also start at most overcast days, but the economical optimum may be found at a smaller panel size.

Fig.2 Typical current consumption on a good solar day. The many spikes in the early morning represent non-succesful start attempts.

Innovative bidirectional video-goniophotometer combining transmission and reflection measurements

Marilyne Andersen, Christian Roecker, Jean-Louis Scartezzini

Solar Energy and Building Physics Laboratory (LESO-PB), Swiss Federal Institute of Tech­nology (EPFL), CH — 1015 Lausanne, Switzerland

This paper describes the design process and setting up of a novel bidirectional go — niophotometer, relying on digital imaging and allowing the combination of transmis­sion and reflection measurements. As its measurement principle is based on the projection of the emerging light flux on a rotating diffusing screen towards which a calibrated CCD camera is pointed (used as a multiple-points luminance-meter), sev­eral strong constraints appear in reflection mode due to the conflict of incident and emerging light flux: for five out of the six screen positions (unless incidence is nor­mal), the incident beam must penetrate in a way that it is restricted to the sample area only; in addition to this, when the screen obstructs the incoming light flux, a special opening in the latter is required as well to let the beam reach the sample. The practical answer to these constraints, detailed in this paper, proved to be reliable, appropriate and efficient.

Introduction

(a) Arbitrary screen position p (b) Screen position p+1

Figure 1: Detection of transmitted light flux for two consecutive screen positions p and p+1.

To allow the integration of advanced daylighting systems in buildings and benefit from their potential as energy-efficient strategies, a detailed knowledge of their directional optical prop­erties is necessary. These properties are accurately described by the Bidirectional Transmis­sion (or Reflection) Distribution Function, abbreviated BT(or R)DF, that expresses the emerg­ing light flux distribution for a given incident beam direction (Commission Internationale de l’Eclairage, 1977). An original experimental method for their assessment, illustrated in Fig­ure 1 was first developed for transmission measurements (Andersen et al., 2001): the light emerging from the sample is reflected by a diffusing triangular panel towards a CCD camera, which provides a picture of the screen in its entirety. Within six positions of the screen and camera around the sample (each separated by a 60° rotation from the next one), a complete investigation of the transmitted or reflected light is achieved.

This innovative approach brought several major advantages when compared to characteriza­tion techniques requiring a sensor to be moved from one position to the other (Papamichael et al., 1988; Bakker and van Dijk, 1995; Aydinli, 1996; Breitenbach and Rosenfeld, 1998; Apian-Bennewitz and von der Hardt, 1998): a significant reduction of the BT(R)DF data assessment time (a few minutes instead of hours per incident direction) and a continuous information about the transmission (reflection) hemisphere, whose resolution is only limited by the pixellisation of the images.

The camera is used as a multiple-points luminance-meter and calibrated accordingly. A luminance mapping of the projection screen is carried out by capturing images of it at differ­ent integration intervals, thus avoiding over and under-exposure effects, and appropriately combining the latter to extract BT(R)DF data at a pixel level resolution.

Material samples showing large range of luminances can thus be handled without any loss of accuracy, while an appreciable flexibility is allowed in the data processing (Andersen, 2004).

For BRDF measurements (reflection mode), however, additional constraints appear due to the conflict of incident and emerging light flux.

For five out of the six screen positions (unless incidence is normal), the detection princi­ple can be kept identical as in transmission mode (Figure 2(a)), except that light flux must penetrate the measurement space in a way that the beam is restricted to the sample area only. As there is one position (all six for normal incidence) where the screen obstructs the incoming light flux, a special opening in the latter is required to let the beam reach the sam­ple, producing a blind spot at that specific screen position (and only in reflection mode), as illustrated in Figure 2(b).

(a) Unobstructed penetration (b) Screen hole

Figure 2: Detection of reflected light flux.

The design process of the instrument combining BTDF and BRDF measurements is pre­sented in this paper, and the mechanical components specifically developed to answer to these constraints in a practical and efficient way are described.

Domestic hot water (DHW)

The water stored in the facade is warmed by solar energy and it is delivered to the do­mestic hot water network according to the consumer demand. When there is consumption, water enters to the accumulator in facade from the city network at a constant inlet tempera­ture considered equal to 15°C.

Top temperature at tank is considered to be the delivery temperature to the load. No heat losses through the pipes are considered. The demand temperature is assumed constant and equal to 43°C.

Two consumption profiles will be considered, profile 1 presents water use at noon, whereas profile 2 does not. They are described in Table 1 and Table 2.

Time

Use

Flow rate

Total [I]

From:

To:

[l/min]

[kg/s]

8:00

8:15

Shower

5

0.083

72

8:15

8:30

Face and hands washing

1.9

0.03

27.36

13:30

13:45

Food preparation

2.0

0.0331

28.80

15:00

15:15

Fland dishwashing

1.04

0.0182

14.976

19:00

19:15

Clothes washing

4.48

0.074

64.504

20:00

20:15

Face and washing hands

1.9

0.03

27.36

Total:

235

Table 1: Water consumption profile 1-With noon draw

Table 2: Water consumption profile 2 — Without noon draw

Time

Use

Flow rate

Total [I]

From:

To:

[l/min]

[kg/s]

8:00

8:15

Shower

5

0.083

72

8:15

8:30

Face and hands washing

1.9

0.03

27.36

19:00

19:30

Food + hand dishwashing

3.75

0.062

108

19:30

19:45

Face and washing hands

1.9

0.03

27.36

Total:

235

If outlet temperature from tank is called OTL and inlet temperature ITL, useful energy from facade to satisfy domestic hot water load, is calculated as:

TOC o "1-5" h z QLOAD = mcp(OTL — ITL) (3)

Domestic hot water load will be:

Load = mcp(T demand — IT L) (4)

Then, auxiliary energy necessary to complement energy delivered by the facade will be:

Qaux = mcp(Tdemand — OTL) (5)

Psychological position towards Glare Discomfort

Two paradigms about human perception of light and glare suit with the planning of lighting conditions: Engineers currently work with a stimulus-response approach relying on criteria which express the visual capacity of the lighting conditions The second approach is the model of cognitive perception which is often applied by architects. Both paradigms relate to psychological research areas.

The stimulus-response-model refers to the psychological paradigm of the "behaviourism" which was the major theory of psychology in the 1960s. It supposes that the behaviour is determined entirely by environmental conditions. According to this model the behaviour of people can be directed by "arrangements" of environmental stimuli. Representatives of this theory suppose that what we become is mostly the result of environmental influence, not of inheritance.

This theory formulated the methodical principle that only the directly observable behaviour may be the research object, while internal mental events of the human beings are not accessible.

Although only few psychologists represent the radical behaviourism which denies any inner life, its theoretical and methodical principles have influenced the thinking so lastingly that they are still noticeable.

The cognitive model supposes that the neural information processing determines how an individual will behave. The processing logically assumes the input of information from the environment, but human action is not seen as a direct reaction to this input. The active process of the cognition or information processing is in between and the person selects actively the information which he needs to react. People do not react to reality as an objective material world, but how it displays to them as a subjective reality. The individual constructs his own interpretation of the world as an internal image — called mental model or schema — which does not correspond to the objective description (Gardner, 1989).

The experiences and perceptions are stored in schemata (or scripts if they describe a course of events) and are activated in the processing. Schemata are general conceptual frames or knowledge structures and contain assumptions about certain subjects, people and situations. Schemata are knowledge parcels about the complicated generalisations of our experiences with structures in the environment. One has for example schemata of kitchens, bedrooms and exam celebrations. Schemata (and scripts) do not contain all details of our different experiences, but represent our "average experiences" which we gain through interaction with our environment (Rumelhart, Smolensky, McClelland, Hinton, 1986).

This can be illustrated by an experiment conducted by Brewer & Treyens(1981). In this experiment test persons were led to a room which was said to be the office of the investigator. After some minutes they were brought to another room where they were
instructed to note down what they remembered from the first room. The test persons recalled very well equipment which is typical for an office schema, but did not recall objects which are not typical for an office schema. They also listed objects which are quite usual in offices but had not been present in the test room.

According to discomfort glare the main notion for the stimulus-response-system is that a stimulus activates via a net of „internal switches" a response of the human being. This procedure is linked with reaction times. The elaboration of the Discomfort Glare Indices (DGI) fits largely on the concept of stimulus-response. During the visual operation luminance distribution and color rendering of the surrounding are the stimuli. The responses are behaviour, feeling and willingness to perform. Current intentions and concentration determine „internal switch" positions. The crucial category for this concept is the visual capacity which is linked with the maximal speed of perception which is reachable in difficult visual tasks. The time between stimulus and response as well as the accuracy of the response is the measure which expresses the match between the light conditions and human abilities.

Explaining discomfort glare with the cognitive approach one must consider that mental concepts of potential environment composition are available for human beings as a matter of visual experiences. Approved cognitive theories show that humans rely on existing cognitive schemata — which they elaborate in their life time — and are therefore able to reduce the necessary cognitive capacity to a minimum. Sensorial information from the eye — often only cue information of a complex perception — leads to the choice of a suitable mental concept from all stored patterns and brings this one to perception. In an objective physical discomfort glare condition the perceived information of the eye might not be the discomfort glare but the computer screen which activates an existing mental schemata that appraises work environment as negative. In the same condition the cue information might be for another individual that sun is shining outside and activates a schemata of summer holiday and leads to positive appraisal. In comparison to stimulus response systems this procedure reacts in negligible time and is able to operate as well if sensorial information is missing because it sets hypothesis about the environment composition.

Current cognitive psychology assumes that human perception and assessment of environmental conditions is directed by subjective schemata of world constructions. From this finding one has to state clearly that the stimulus-response approach for the elaboration of the assessment of Discomfort Glare is using a mechanical idea of man which is disproved by psychological research of the last 20 years.

Temperatures inside the absorber

The tests were disturbed by the clouds passing in the sky and by the drift of the heliostats during the solar tracking. The second effect was sensibly reduced after the change of the tracking algorithm in the control software of the heliostats field. For this reason, the last days of tests were less affected by instantaneous defocus to correct the aiming point, and the system behaved more stably. It yields more quasy-steady states to compare the results. Thus, only the results of the square metallic cups tests are presented, in reason of clarity. Moreover the position of the inner thermocouples is better fixed, because they were installed by Emitec during he manufacturing and they could not move, what leads to a higher data reliability of these tests.

Fig.3. Test results for the metallic square cups under a radiation intensity of 370 ± 20 kW/m2

Effect of cell size and mass flow rate on the axial temperature distributions.

For a roughly constant value of the solar radiation flux, the axial temperature distribution around the central channel is shown in fig.3 for different mass flow rates per cup. It is assumed that the mass flow rates per cup are all in the same range and then, the mass flow rate per cup is estimated by dividing the overall mass flow rate per the number of cups (35 ones for the square cups tests, since once cup of the Solair receiver was removed and closed). In this chart, the wall temperatures are plotted versus the parameter x/L, being x the distance from the irradiated entrance along the channel and L the total channel length. Inside each cup, a similar temperature profile is maintained for the three flow rates, but the mean temperature increases as the mass flow rate decreases, as it was expected.

The influence of the cell size on the axial distribution is also obtained from data depicted in the figure: for a smaller cell or channel diameter, the maximum wall temperature is reached earlier, and then the temperatures decrease with a stronger slope. On the other hand, for absorbers with a larger cell size, the maximum temperature is reached deeper in
the channel. With only 4 data we cannot assure, as it seems in the chart, that the maximum for the 500 cpsi absorber is located at a x/L value of 0.43, which corresponds to a distance of 30 mm from the entrance. It should be somewhere between 10 and 30 mm, but the positions chosen for the thermocouples do not allow to determine the location of the maximum temperature. Similarly, for the absorber with 600 cpsi, the maximum temperature should be near the thermocouple located at 10 mm on the channel axis.

These temperature profiles give information about the penetrability of the radiation into the absorber channels: for a higher cell density (600 cpsi), the cell size is smaller and the rays penetrate less than for an absorber with lower cell density (500 cpsi). This behaviour is related to the volumetric effect, since the radiation is absorbed in the first millimetres from the entrance and not only on the aperture surface. Absorbers with wider channels have more “volumetricity” since the volume irradiated is bigger, but the specific surface to absorb and transfer the heat are both smaller, resulting in lower temperature values.

Fig.4. Effect of the solar radiation intensity for an overall mass flow rate of 0.3090 ± 0.0003 kg/s, that corresponds to an estimated mass fow rate per cup of 0.0088 ± 0.0001 kg/s.

The behaviour for different solar radiation fluxes under the same mass flow rate is as expected: higher radiative fluxes lead to higher absorber temperatures, but the temperature distribution is similar in all cases. In fig. 4 only the results for the cup with 600 cpsi are presented, the profiles of the other absorber are analogue to those shown in the previous chart.

Air temperature distribution at the absorber outlet:

To obtain the temperature distribution for the air exiting the absorber, the data from the thermocouples located in several positions at a cross section 2 cm behind the absorber have been interpolated in Matlab with the function called Griddata, which fits a surface to the data in nonuniformly-spaced vectors. With the ‘v4’ method, griddata produces smooth surfaces. Once the data are interpolated for the defined grid, the level contours of air outlet temperature are plotted.

Fig.5. Level contours of the air outlet temperature for 3 mass flow rates through the 600 cpsi absorber in the first row and the 500 cpsi absorber in the second row

The temperatures extrapolated near to the edges of the square absorbers must not to be considered because there are not experimental points near the edges to function as boundary conditions. For this reason the calculated temperatures are too low with respect to the real temperatures in the square corners.

The main difference between the distribution of the two absorbers is caused by the temperature at the absorber centre: in the 600 cpsi absorber the central temperature is in the same temperature range as the thermocouples next to it, while the central temperature of the 500 cpsi absorber is always lower than the temperatures at each side of the center. The colder central region in the 500 cpsi absorber suggests a non-uniform air mass flow distribution through this cup. The air flow distribution through the 600 cpsi absorber seems to be more uniform. From this, it is possible to conclude that the air flow distribution for both cups is different.

On the other hand, the warmer region that is observed at the left upper part of the 500 cpsi absorber may be due to the fact that the solar radiation was focused at the centre of the Solair, which is located in that position regarding to the 500 cpsi absorber cup (position 22).

Other important aspect is the great difference between the absorber wall temperature and the air outlet temperature. In volumetric solar receivers the air is supposed to reach the same temperature of the absorber matrix before exiting the channel. However, in these tests between the wall temperature at 20 mm from the absorber exit and the air temperature measured 20 mm after the exit, there is always a difference about 150°C for the 600 cpsi absorber and about 70°C for the 500 cpsi absorber. This means that the air does not reach the same temperature as the absorber in the first centimeters of the absorber channel, as some simulations suggest. Air temperature is lower than the absorber temperature all along the channel. Thus, air is always cooling the absorber. Nevertheless, in the first part of the
absorber channel, radiative heating up dominates over convective air cooling. From the point where radiation has no more effect over the channel walls, the absorber temperature begins to drop, more rapidly when that point is nearer to the entrance.

The extrapolation of the temperatures inside the absorber, considering only the last falling slope, gives quite a good estimation for the air temperatures.

A way of comparing the relative "heat transfer efficiency” from the absorber monolith to the air for each absorber module is to calculate the relative difference from the last absorber wall temperature (measured in the central axis at 20 mm from the exit) to the air outlet temperature in the same axis. The results show a better relative efficiency for the 500 cpsi absorber, but the absolute values are lower than for the other absorber. The variation in mass flow rates does not affect to the relative temperature decrease, while the variation in radiative fluxes has a slight effect: for higher radiation, worse heat transfer.

Conclusions

— The temperature distribution along the channel for a determined channel geometry (or absorber cell density) is the same for different mass flow rates or solar radiation fluxes. It achieves a maximum value inside the channel and then the temperature decreases due to the heat transfer to the cooling air.

— For a smaller cell or channel diameter, the maximum wall temperature is reached earlier, and then the temperatures decrease with a stronger slope. On the other hand, for absorber with a larger cell size, the maximum temperature is reached deeper in the channel.

— — The great difference between the absorber wall temperature and the air outlet temperature indicates that the studied absorbers are not sufficiently efficient to transfer all their energy to the air. The relative heat transfer efficiency is better for lower cell densities.

To sum up, the studied absorbers do not behave as it is expected for an ideal volumetric solar receiver, where the maximum matrix temperature would be at the end of the channel and the air would reach the same temperature of the absorber matrix before exiting the channel. These experimental results can be used to validate theoretical (numerical or analytical) models that reproduce the heat transfer in duct volumetric solar receivers.

Acknowledgements

The authors acknowledge the collaboration with the German enterprise Emitec, and the efforts of Mr. Arndt-Udo Rolle in the manufacturing of the samples are specially appreciated. Results discussions with Ma Jesus Marcos are also very appreciated.

ANALYSIS. Constant Thickness Model

To simplify the mathematical analysis, the following assumptions were made for the constant thickness film model: the wall is considered isothermal, i. e., the cooling water flow is not considered, the flow is laminar and fully developed for the liquid-desiccant and the air, the desiccant flow is smooth (not wavy), the physical properties are constant, there is thermodynamic equilibrium at the desiccant/air interface, air is an ideal gas, no shear forces are exerted by the air on the desiccant, the body force in the air is negligible, diffusion in the direction of the flow is negligible, species thermo-diffusion and diffusion — thermo effects are negligible, the rate of water vapour absorption is small, i. e., the velocity in the transverse direction is negligible, the solubility of air in the desiccant solution is negligible and the film thickness is constant.

for the liquid-desiccant film,

and for the air,

Fig. 2: schematic diagram of the dehumidifier channel with notation.

d u

Yd’

d

u

dy2

T

dx

dCA

— + Pdg = 0

■ = ая

d 2T

d

dy2 5 C

"d dx Dd dy2

(1)

(2)

(3)

The boundary conditions are:

at x=0,

Td=Tdi, Ta=T

at y=0,

T=T C

|d=|w,

dy

at y=Sd,

Td=Ta, ud=ua

= 0, ud=0,

a

Ca

dUj_

‘ dy

dp d2 ua

dx = A. y 2

(4)

Ua STa = «. SlT; dx a dy2

(5)

u C. — d ^

a dx D dy2

(6)

Cai,

(7)

(8)

= 0,

(9)

Considering the assumptions above and the geometry presented in Fig. 2, the governing momentum, energy and species equations are:

SHAPE * MERGEFORMAT

at y=Sc

(10)

dT dC du

a = 0 , ^ = 0 , ^ = 0 .

dy dy dy

The velocity profile for the liquid-desiccant is calculated integrating (1) twice across the desiccant film:

ud

2

2

(11)

and considering the continuity across the film and a channel 1 m wide:

md = £’ Pduddy (12)

Solving (12) allows the development of the expressions for film thickness and velocity profile, according to the desiccant mass flow, md:

к

(13)

3mdrd

Pdg j

and

ud

(14)

3md ( 2у _ т! л

2 Pd v^d ^d J

The velocity profile for the air stream can be calculated also through the momentum and continuity equations for the air stream:

(15)

ma = 2‘PaUady

dd

dp

dx

3 Pa

ud

fo — sc )2

(16)

(17)

m

a

2Pa {Sd ~SC )3 _

2 (y2~5l )+(A — y)

dp 1 dx pa

SHAPE * MERGEFORMAT

The energy and species balances at the interface (y=8d) give the additional matching equations:

_ k k T. П h C

d dy a dy Pa “ ff dy

PdDd

dCj

dy

(19)

PaDa

C

dy

SHAPE * MERGEFORMAT

C

*

a

(20)

18.0153 pw

28.9645p -10.949pw

The equilibrium water mass fraction in the air at the interface is given by the water vapour pressure of the water in the desiccant solution at the interface, pw(Td, Cd), and Dalton’s law applied for a water vapour/air mixture:

Equations (2),(3),(5),(6),(18) and (19) were approximated using the finite-difference method. Central-differencing was used for the diffusion terms and backward or forward — scheme for the convection terms, depending on whether the air/desiccant flow configuration was co-current or counter-current. The equations were solved simultaneously using the software package Microsoft EXCEL[6]. In the “y” direction, 7 nodes were used for the liquid-desiccant film and 40 for the air (Figure 3). The interface between the last node in the desiccant side and the first node in the air side was represented by an additional node. This interface node has no volume, but provides, through equations (18), (19) and (20), the coupling between the air and liquid streams and between the energy and species equations. The number of nodes in the flow direction varied according to the height of the channel.

representation for one step in flow direction.

INTERFACE -1

WALL

+1

The variable thickness model uses the same equations, assumptions and number of grid nodes of the constant thickness model. The difference is that the thickness of the desiccant film is allowed to vary and is recalculated using equation (13), for every step of the simulation. Once the film thickness is recalculated, it is possible to obtain a new velocity profile for the desiccant using (14), i. e., the profile is still the profile for the fully developed laminar flow, but recalculated to account for the change in

Fig. 4: schematic of grid with thickness change in the desiccant film.

Variable Thickness Model

mass flow and film thickness. This procedure is similar to the one adopted by Jernqvist and Kockum [7] for developing falling film flow. To reduce the computational time, and because the changes in film thickness are small, the change in the air flow velocity profile was neglected.

Simulations

A solar cooling system consisting of solar thermal collectors, a heat storage (optional), an absorption chiller, a cold storage (optional) and the application system (e. g. room cooling) is a very complex arrangement. The performance of some components (e. g. collector and cooling tower) depends on ambient conditions (e. g. air temperature, air humidity). Because of the high number of interrelations between the different components and between some components and the ambient conditions it is not possible to calculate the performance of the whole system and to compare different types of certain components (e. g. collectors) manually. Simulation programs have to be used.

Using the capabilities of the TRNSYS /1/ package the performance of the small capacity absorption chiller in combination with different kinds of collectors and re-cooling options was evaluated /2/. For flat plate and vacuum tube collectors a comparison of different models (manufacturers) was done using published test data. For each type a reference collector model with a good performance and an acceptable cost-performance ratio was chosen. The properties of these models were used for further calculations. For parabolic trough collectors the STEC-Library was used (DLR model of the IST trough) /3/.

At first a system with a fixed cold water inlet temperature (12 °C; 18 °C) to the absorption chiller was simulated with climate data of Huelva/Spain.

For all combinations of the three collector types and two re-cooling options (wet cooling tower and dry cooler) the delivered cooling energy from April to August was calculated for five collector areas. The results are shown in Figure 8 and Figure 9.

The results show the great influence of the re-cooling version. If a dry cooler is used in­stead of a wet cooling tower the collector area has to be increased to obtain the same quantity of cooling during the summer.

Comparing the figures 8 and 9 the influence of the cold water temperature can be evalu­ated. If the room cooling system is capable to meet the cooling load with a higher cold wa­ter temperature (greater heat exchanger area needed) the cooling energy delivered by the absorption chiller increases (or a smaller collector area is sufficient to receive the same amount of cooling).

The simulations also provided other information such as the maximum cooling perform­ance of the absorption chiller during the operation period, distribution of the cooling energy over the month (and days), water demand for the wet cooling tower, delivered cooling en­ergy per collector area. By comparing the results for different collector areas an optimal size can be chosen.

At a second step a more realistic system with changing cold water temperatures and in­cluding the thermal behaviour of the building (cooling load) was simulated. To compare the different arrangements

— the excess temperature: ATe = TRoom — 26 °C

— the duration (tet) of the periods with inside room temperature above 26 °C and

— the integral: |ДТЕ dTET [Kh]

were calculated. Figure 10 shows the results for one configuration with different collector areas.

Simulations using sea water for re-cooling the absorption chiller were also done. The re­sults for Huelva showed that there is no higher performance of the absorption chiller than using a wet cooling tower. Sea water cooling has some other restrictions too (only near the sea, high cost, need to use suitable materials).

40 50 60 70 80 90 100

Gross Collector Area [m2]

Figure 8: Delivered cooling energy for different configurations in Huelva and a fixed cold water inlet temperature to the absorption chiller of 12°C

Figure 9: Delivered cooling energy for different configurations in Huelva and a fixed cold water inlet temperature to the absorption chiller of 18°C

All the results are valid for the specified equipment (collectors, coolers) only. For evalua­tion of a certain project simulations have to be done again with the characteristics of the equipment to be used and suitable climate data.

The experiences gained during the simulations and supervising solar cooling installations showed that great care should be taken regarding system control, storage integration and design of the room cooling system. As mentioned before the cold water temperature is important for the achievable cooling output.

—І і і p і і T ‘ I T

40 50 60 70 80 90

Gross Collector Area [m2]

Figure 10: Integral of the excess room temperature in Huelva for different configurations

Conclusions

A H2O/LiBr absorption chiller with a nominal cooling capacity of 15 kW was developed. A special heat exchanger design permits the use of low temperature heat such as solar or waste heat to drive the absorption chiller.

After prototype testing a field test at three sites was carried out in the summer of 2003. The field test showed good results. The absorption chiller is working reliable and flexible over a wide range of external conditions.

Because of the high number of cooling applications in the capacity range below 50 kW the new chiller has the potential to increase the usage of solar thermal energy for cooling pur­poses.

The waste heat of cogeneration systems can also be used to drive the absorption chiller creating new possibilities for trigeneration systems (power, heat and cold).

The performance of solar thermal cooling systems can be predicted using the capabilities of system simulation programs.

Nomenclature

COP

coefficient of performance

ATe

excess temperature

Ig

global solar insolation

n

solar collector efficiency

Ta

ambient temperature

Tet

duration of periods with temperature

Tm

temperature of the collector fluid

above 26 °C

T Room

room temperature

Literature

/1/ Solar Energy Laboratory, University of Wisconsin (2000). TRNSYS 15, A Transient System Simulation Program, Madison, WI, USA.

/2/ Heinrich, C. (2004). Modelling and Simulation of solar thermal driven single effect absorption chillers of small capacity for climatisation, Diploma Thesis (German), Dresden, Germany.

/3/ Schwarzbozl, P.; Eiden, U. e.; Pitz-Paal, R.; Jones, S. (2002). A TRNSYS Model Library for Solar Ther­mal Electric Components (STEC) — Reference Manual — Release 2.2, Cologne, Germany.

Figure 13: Radiance model of Amon Carter Museum’s south gallery on November 28 at 3:40 PM; isolux contours (top). North galleries adjacent to atrium (2nd floor)

Two of the windowless north galleries of the second floor receive direct sunlight that passes through the high southwest windows of the atrium (see yellow areas in Figure 4- right). Illuminance levels measured over the painting in one of the two galleries on January 19, 2003 (Figure 15) reached an extremely high value of 11,500 lux, which is about 57 times higher than the recommended IES standards for museums. The total illuminance-hour per year over these display areas over exceeds the total exposure limits for moderately light susceptible displayed materials. The fish-eye photo (Figure 16) shows that the painting receives direct sun for one and a half hours in the early afternoon from November to January. The high illuminance levels over the display areas are due to the high visible transmittance of the atrium’s glass (Tvis=59%). The glass of the atrium does not have any shading device to intercept the incoming direct sun to the adjacent galleries. To protect paintings from over exposure to light, curators have been rotating oil paintings frequently (see dates on Figures 15 and 16).

Figure 15: Oil painting (1848), Amon Carter Museum’s gallery adjacent to atrium, January 19, 2003, 4:30 PM.

Figure 16: Fish eye photo taken from painting’s viewpoint with sun path diagram, at gallery adjacent to atrium (March 5, 2004).

Figure 17: Direct glare on visitor’s eye at Amon Carter’s north gallery adjacent to atrium (March 5, 2004, at 4:50 PM).

Another problem observed in these galleries is the glare that visitors experience while moving through the gallery and to the atrium. Figure 17 illustrates how a visitor has to protect her eyes from direct sun while trying to see the painting and read the sign next to it. This visitor is receiving about 7,770 lux over her eyes when not covered. The extremely high variations of light levels within the field of view of the visitors to these galleries makes them uncomfortable to adjust their eyes to the low light levels over the paintings and to the bright atrium’s glass areas that are within her field of view.

Table 1: Summary of Lighting Conditions at the three museums.

Modern Art Museum

Kimbell Art Museum

Amon Carter Museum

Sunlight penetration on galleries

■ West-facing galleries

■ No (entrance of daylight is well controlled)

■ East-facing lobby

■ South-facing gallery (2nd floor)

■ Interior galleries adjacent to atrium (2nd floor)

Daylighting systems (orientation)

■ Toplighting (north-, south-facing clerestories)

■ Sidelighting (west­facing)

■ Windowless galleries

■ Toplighting (narrow- strip skylights with reflectors)

■ Sidelight high strip windows (east-, west­facing)

■ Toplighting (high windows in atrium)

■ Sidelighting (east-, south-facing)

■ Windowless galleries

Range of illuminance levels on display areas

■ 50 lux to 4,000 lux (under direct sunlight)

■ 50 lux and under

■ 50 lux (windowless galleries)

■ Up to 11,000 lux (under direct sun)

Display objects

■ Oil paintings

■ Sculptures (wood)

■ Paper, prints

■ Photographs

■ Oil paintings

■ Watercolors

■ Oil paintings

■ Daguerrotypes

■ Photographs

■ Sculptures (metals)

Glazing

■ Blue tinted, Tvis=59% with white interior screens.

■ Skylights

(polycarbonate with UV filter)

■ Dark tinted, Tvis=5%, UV protection

■ Tinted, Tvis=12%

■ Clear glass, Tvis=60%

Conclusions

The most noticeable problem, in the galleries of two of the museums presented in this paper, has been the sunlight penetration over the displayed museum objects. This fact is harming valuable art collections, and also creating visually uncomfortable environments for visitors. Sunlight penetration occurs mainly in galleries with side lighting windows that face the sun — east, south, and west-; orientations that are the most difficult to control. Even though the time of over exposure is relatively limited, the illuminance levels reached at these times are extremely high, 10 to 57 times the maximum recommended IES standards for light susceptible objects. Dark tinted glass, screens, or small overhangs are not enough measures to block the entrance of direct sun over the exhibit areas at the Modern Museum (west-facing galleries) and at the Amon Carter Museum (east-facing galleries).

To control the light levels in these galleries more aggressive changes should be done, like modifying the fagade by adding horizontal or vertical shading devices, external louvers, trellises, or trees to filter sunlight. All these modifications could change the image of the building, which may not be acceptable by designers. There are other less intrusive solutions that include the use of miniaturized sunscreens, interior louvers or baffles; or computer controlled dynamic window systems (Reference 6), where direct sunlight over sidelight windows can be intercepted and filtered reducing interior light levels as low as 25 lux or under.

On the other hand the third museum, the Kimbell Art, presents a well thought and carefully designed toplighting system that can introduce adequate light levels throughout the year to accommodate the lighting requirements of exhibits, and at the same time provide a connection to the exterior environment by rendering the galleries with natural light. This museum is an excellent example that it is possible to successfully illuminate museum galleries with daylighting.

If we already know that sidelight windows in museum galleries create many problems to the display of light susceptible artifacts, why are we still including them as a source of illumination in museum galleries? Is it due to the lack of knowledge about solar geometry? Or it is just that architects underestimate the effects of sunlighting in museums? The answers to these questions may not be clear and simple, meanwhile we might be finding museum galleries with sidelight windows with strong sunlight over display areas, such as in two of the three museums presented in this paper (see summary in Table 1). In the meantime, curators and facility managers at these museums have to create means to protect their valuable art collection (i. e. rotating display objects, use boards to cover them on a daily basis, display least light susceptible objects, cover completely the windows with boards or black cloths, among other solutions) (Reference 7). The main goal of this paper was to make us reflect on the way that museums are still being designed, and on how could we make them better to preserve and display artifacts that are important part of human history.

The air solar system

To assure a complete self-sufficient system, the air solar system should cover the 20% of the remaining demand. With this aim, 12 sq m of Solarwall system (the same surface of the solar panels) has been added.

Figure 11 shows how the system works. A fan collects the preheat air from the Solarwall and it is introduced through a special insulated channel directly to the living room. The system keeps the house dry also when it is not occupied for long periods. In sunny days the air system will delay the switching on of the conventional solar water system.

The calculation indicates that only in December the self-sufficiency could not be reached. Nevertheless, the results depend completely on the weather. In fact, on a sunny day the house could be heated mainly by the air solar system, but on a cloudy day the risk to be insufficient is very high.

Then, it is necessary to integrate the solar system with a boiler, using renewable energy source, which means a wood heater.

The wood heater

When necessary, a wood heater integrates the solar system and covers the small daily energy need, about 5,3 kWh/day. This small requirement can be satisfied burning around 1 kg of pellets o using directly the local olives trees pruning in a stove with a 85% efficiency.

In order to cover the remaining winter heating demand not cover by solar energy of 171 kWh, only 43 kg of pellets (that means 7,74 Euro for the whole period) or 40 kg of olives trees pruning is required.