UNIVERSITÀ DEGLI STUDI DI PADOVA
Facoltà di Ingegneria Dipartimento di Fisica Tecnica

Tesi di laurea FEED-FORWARD CONTROL APPLIED TO THERMALLY ACTIVATED CONCRETE SLABS – A CASE STUDY

Relatore: Correlatori:

Prof. Ing. Roberto ZECCHIN Dott. Ing. Michele DE CARLI Dott. Ing. Dietrich SCHMIDT Laureando: Alessio PULLIERO

Anno accademico 2003-2004

To my parents ….

ABSTRACT
The main purpose of this thesis is to investigate different control strategies for the heating and cooling system of the ZUB (Centre for Sustainable Building), situated in Kassel, Germany. It is an experimental office building, with a very detailed monitoring system for studying low-energy and low-exergy building technologies. The conditioning system is a TABS (Thermally Activated Building System), with water pipes embedded in the centre of a structural concrete slab, thus resulting in a ceiling radiant system. The high thermal capacity of the slab offers great opportunities to store heat, to dampen temperature fluctuations or to shift the peak-load; but, on the other hand, it implies a slow response of the system, which requires an accurate regulation strategy to front the variability of several factors, and to achieve the desired indoor temperature. A TRNSYS model of an office room has been developed, thus allowing to implement several regulations in the software and to test their performance. The main conclusion, carried out from the simulations, is that the implementation of a Feed-forward controller gives appreciable advantages in the temperature control, achieving both a more precise control of thermal conditions and a reduction of the energy consumption. The parameters taken into account by the controller are more than one and, more in detail, they are the operative temperature, the variation of external temperature, the solar radiation and the effect of internal gains. Moreover, it has been highlighted how the heat storage opportunity given by the heavy concrete slab is strictly related with its temperature variation, but, if the TABS is the only heat emission system, its temperature is univocally determined by the energy loads. Therefore, to manage the heat absorption and consequent release, a certain variation of indoor temperature has to be allowed, even if within thermal comfort restraints. According to this consideration, the new strategies act directly on the trend of set point temperature, changing its value during the day in a proper way, for exploiting the great thermal capacity of the slab; at the same time, the controller acts on the hydronic circuit to compensate the disturbances and to reach the desired temperature.

v

PREFACE
This work has been carried out mainly at the Department of Architecture and Applied Physics, at the University of Kassel, Germany. This has been possible thanks to Erasmus-Socrates exchange program, and to a bilateral agreement between Professor Roberto Zecchin, at the Department of Building Physics, University of Padua, Italy, and Professor Gerd Hauser, at Kassel University. Therefore I want to convey my gratitude to both professors and to everyone has made the exchange possible. I want to give special thanks to my direct supervisors, Dietrich Schmidt and Michele De Carli, and to all the people who have given a precious contribution to this work, in particular Christoph Kempkes, Jan Kaiser and Claudio Zilio. Last but not least, I want to thank all the colleagues at the Department, for providing an enjoyable workplace and inspiring atmosphere for me.

The Author

vii

TABLE OF CONTENTS
NOMENCLATURE ...........................................................................................................xiii Chapter 1 SUSTAINABLE BUILDINGS ............................................................................................. 1 1.1. THE CONCEPT OF SUSTAINABLE BUILDING.............................................. 1 1.2. DESIGNER’S RESPONSIBILITY ....................................................................... 1 1.3. LOW ENERGY BUILDINGS .............................................................................. 2 1.4. LOW EXERGY BUILDINGS .............................................................................. 3 1.4.1. Exergy concept and definitions ..................................................................... 3 1.4.2. Thermal Exergy ............................................................................................ 4 1.4.3. Low Exergy Buildings................................................................................... 5 1.4.4. Advantages of Low Exergy Systems............................................................. 7 1.5. HUMAN COMFORT............................................................................................ 8 1.5.1. Operative Temperature .................................................................................. 9 1.5.2. Effect of radiant heating/cooling systems on human comfort ..................... 11 1.6. RADIANT SURFACE SYSTEMS, TECHNOLOGICAL SOLUTIONS .......... 12 1.6.1. Superficial thermal layers ............................................................................ 12 1.6.2. TABS (Thermally Activated Building System) .......................................... 13 1.6.3. Large surface radiators ................................................................................ 14 1.7. PASSIVE BUILDINGS ...................................................................................... 14 1.7.1. Definition of Passive Building .................................................................... 14 1.7.2. Passive Building Technical Solutions – Overview...................................... 15 1.8. REFERENCES .................................................................................................... 20 Chapter 2 THE Z.U.B. BUILDING ..................................................................................................... 23 2.1. DESCRIPTION OF THE ZUB BUILDING ....................................................... 23 2.2. RADIANT AND VENTILATION SYSTEMS................................................... 26 2.3. MEASUREMENT EQUIPMENT....................................................................... 29 2.4. REFERENCES .................................................................................................... 32 Chapter 3 TABS (THERMALLY ACTIVATED BUILDING SYSTEMS)........................................ 33 3.1. INTRODUCTION ............................................................................................... 33 3.2. TABS FEATURES.............................................................................................. 33 3.3. DYNAMIC BEHAVIOUR.................................................................................. 35 3.4. TABS ADVANTAGES....................................................................................... 38 3.5. TABS DISADVANTAGES AND LIMITATIONS............................................ 40 3.6. TABS MODELLING ......................................................................................... 41 3.7. TARGETS AND PROPOSALS .......................................................................... 42 3.8. REFERENCES .................................................................................................... 43

ix

Chapter 4 TRNSYS A TRANSIENT SYSTEM SIMULATION PROGRAM....................................45 4.1. BUILDING SIMULATION PROGRAMS..........................................................45 4.1.1. Steady State Simulations..............................................................................45 4.1.2. Transient System Simulations......................................................................46 4.2. TRNSYS ..............................................................................................................47 4.2.1. TRNSYS History .........................................................................................48 4.2.2. TRNSYS Today ...........................................................................................48 4.2.3. Modelling of TABS with TRNSYS .............................................................50 4.2.4. Limitations of TABS modelling with TRNSYS ..........................................52 4.3. REFERENCES.....................................................................................................52 Chapter 5 ZUB BUILDING MODEL ..................................................................................................53 5.1. PREVIOUS MODEL ...........................................................................................53 5.1.1. Criticism to previous model .........................................................................53 5.2. NEW TRNSYS MODEL .....................................................................................53 5.2.1. Solar radiation and shading devices.............................................................55 5.2.2. Water inlet temperature and mass flow rate.................................................55 5.3. REFERENCES.....................................................................................................57 Chapter 6 REGULATION STRATEGIES ...........................................................................................59 6.1. INTRODUCTION................................................................................................59 6.2. ACTUAL REGULATION STRATEGIES IN THE ZUB BUILDING ..............59 6.3. FEEDBACK CONTROL.....................................................................................61 6.3.1. Thermostatic Control (On/Off) ....................................................................62 6.3.2. Proportional Control (PC) ............................................................................63 6.3.3. Proportional + Derivative Control (PDC) ....................................................64 6.4. FEED-FORWARD CONTROL...........................................................................65 6.5. TEMPERATURE CONTROL SYSTEM IN THE ZUB SIMULATIONS.........67 6.6. REFERENCES.....................................................................................................68 Chapter 7 FEED FORWARD COMPENSATOR ................................................................................69 7.1. DESCRIPTION OF THE FEED-FORWARD CONTROL.................................69 7.2. PERFORMANCES OF THE FEED-FORWARD COMPENSATOR ................70 7.2.1. Anticipation of the response of the thermally activated slab .......................70 7.2.2. Comparison between PDeC and Thermostatic Control, in the case of heating ..........................................................................................................72 7.2.3. Comparison between constant and scheduled functioning of the heating system...........................................................................................................75 7.3. REFERENCES.....................................................................................................77

x

Chapter 8 REGULATION STRATEGIES FOR THE EXPLOITATION OF THE GREAT THERMAL CAPACITY OF TABS.................................................................................... 79 8.1. HEAT STORAGE CAPACITY OF THE CONCRETE SLAB .......................... 79 8.2. SUMMER - KASSEL ......................................................................................... 80 8.3. SUMMER - VENICE .......................................................................................... 86 8.4. WINTER.............................................................................................................. 89 8.5. REFERENCES .................................................................................................... 95 CONCLUSIONS ................................................................................................................. 96 TOPICS FOR FUTURE RESEARCH ................................................................................ 98 Appendix 1 DAMPENING EFFECT OF A THERMAL CAPACITY .................................................. 99 A1.1. INTRODUCTION ........................................................................................... 99 A1.2. THE MODEL .................................................................................................. 99 A1.3. OBSERVATIONS........................................................................................... 99 A1.4. REFERENCES .............................................................................................. 101 Appendix 2 THERMAL RESPONSE OF ROOM AND SLAB TO DIFFERENT IMPULSES.......... 103 A2.1. TRANSFER FUNCTIONS ........................................................................... 103 A2.2. EFFECTS OF DISTURBANCES ................................................................. 105 A2.2.1. External temperature.................................................................................. 106 A2.2.2. Slab functioning......................................................................................... 107 A2.2.3. Internal gains ............................................................................................. 110 A2.2.4. Solar gains ................................................................................................. 112 A2.3. COMPENSATION OF OUTPUTS OF DISTURBANCES ......................... 113 A2.3.1. Compensation of external temperature variation....................................... 113 A2.3.2. Compensation of internal gains ................................................................. 115 A2.3.3. Compensation of solar radiation................................................................ 116 A2.4. REFERENCES .............................................................................................. 117 Appendix 3 RESULTS OF SIMULATIONS........................................................................................ 119 A3.1. INTRODUCTION ......................................................................................... 119 A3.2. KASSEL SUMMER CLIMATE ................................................................... 120 A3.2.1. Kassel, Summer – One Occupant .............................................................. 120 A3.2.2. Kassel, Summer – Two Occupants............................................................ 122 A3.2.3. Kassel, Summer – Effect of mass ............................................................. 124 A3.3. VENICE SUMMER CLIMATE.................................................................... 126 A3.3.1. Venice, Summer – One Occupant ............................................................. 126 A3.3.2. Venice, Summer – Two Occupants ........................................................... 128

xi

A3.3.3. Venice, Summer – Effect of mass..............................................................130 A3.3.4. Venice, Summer – Combination of TABS and AHU ................................133 A3.4. KASSEL WINTER CLIMATE .....................................................................134 A3.5. KASSEL, COST OF NIGHT VENTILATION .............................................135 A3.6. DISCUSSION OF RESULTS........................................................................135 A3.6.1. Kassel, Summer..........................................................................................135 A3.6.2. Venice, Summer.........................................................................................136 A3.6.3. Kassel, Winter ............................................................................................137

xii

NOMENCLATURE
A C c clo E e G f g h hc hr K,k & m met P Q q qk R s T Ta Tmr Top Tset U V w z & ∆S irr ∆T λ µ ηel ηid ηmec ηV ρ Area of the surface [m2]; Anergy [W] Total heat capacity [J/K] Specific heat capacity [J/(kgK)] Unit of measurement of the Clothing Factor Total exergy [W] Specific exergy [W/kg] Transfer function Function Gravity acceleration [m/s2] Specific enthalpy [J/kg]; Convective and Radiant coefficient for heat exchange [W/m2K] Convective coefficient for heat exchange [W/m2K] Radiant coefficient for heat exchange [W/m2K] Proportional constant Mass flow rate [kg/s] Unit of measurement of the Metabolic rate Power [W] Heat [Wh]; Energy [J] Heat Flux [W] Heat flux exchanged by the system with a source at the temperature Tk [W]. Thermal resistance [W/K] Specific entropy [J/kg]; Thickness of a layer or of a wall [m]; Laplacetransform variable Temperature [°C], [K] Temperature of air [°C] Mean Radiant temperature [°C] Operative temperature [°C] Set point temperature [°C] U-value [W/m2K] Volume [m3]; Disturbance Speed [m/s] Height over a point of reference [m] Velocity of production of entropy due to irreversible transformations [W] Temperature difference [K] Thermal conductivity [W/mK] Decrement factor (or amplitude attenuation) Electric efficiency Hydraulic efficiency of a pump Mechanical efficiency of an electric motor Volumetric efficiency of a pump Density [kg/m3]

xiii

Σ φ Subscripts a C e ext ES F i in int out P set T 0

Sum; Name of a generic system Time lag (or phase shift) [s]

Air Feedback Controller Outlet section External Heat Emission System Feed-forward Controller Inlet section Inlet; Input Internal Outlet; Output Process Set point Sensor Ambient

Abbreviations AHU COP DC DeC E.S. FD FE HVAC MEM NMF PC PDC PDeC PMV PPD S TABS VOC W ZUB Air Handling Unit Coefficient Of Performance Derivative Control Feed-forward compensator based on external temperature Heat Emission System Finite Differences code Finite Element code Heat Ventilation Air Conditioning Macro Element Model Neutral Model Format Proportional control Proportional and Derivative Control Proportional Control with feed-forward compensator based on external temperature Predicted Mean Vote Percentage of People Dissatisfied Regulation strategy for summertime Thermally Activated Building System Volatile Organic Compounds Regulation strategy for wintertime Centre for Sustainable Building, Kassel, Germany

xiv

Chapter 1

SUSTAINABLE BUILDINGS
1.1. THE CONCEPT OF SUSTAINABLE BUILDING The concept of sustainable building involves several topics, such as compatibility of materials, environment integration, low energy demand, use of renewable energy. All these ideas must be kept in context with other equally important design targets, such as aesthetics, accessibility, cost effectiveness, flexibility, satisfaction of users’ wishes, aims of comfort conditions, security. Therefore, an environment friendly and sustainable building process is an interdisciplinary task, that involves different building related disciplines at the same time [1]. But, considering that the primary energy demand of residential and commercial buildings counts for about one third of the total world energy demand [2], energy efficiency should be recognized as one of the most important aims [3], that involves the entire building process. 1.2. DESIGNER’S RESPONSIBILITY Despite the observation of the planet’s altered conditions, it seems that almost none of the countries considers as its bigger priority the consumption restraint, which would likely determine a sudden reduction of emissions. Amidst this political and administrative emptiness, designers are called to voluntarily assume the responsibility, in the name of their role in the society, thus addressing architecture towards more efficient environmental solutions. Designers should to consider environmental and social upgrading as a major and absolute necessity in their work [4]. Several actions can be taken into account to reduce the energy consumption of the buildings, both for new projects and for existing buildings. Some possibilities are reported in the following paragraphs.

1

1.3. LOW ENERGY BUILDINGS Despite different definitions of Low Energy Buildings have been formulated, in this context are taken into account all those buildings whose primary energy demand is considerably reduced by the adoption of active or passive measures [5], like the ones listed below. Different kinds of energy are employed in a building, for heating, cooling, dehumidification, water circulation, ventilation, illumination, electrical equipments, services, etc. To make a comparison, or to evaluate the energy demand of a building, each kind of energy should be consider separately, for its different quality, or even for its different price. A way to sum up all the energy sources at the same time, is to consider the primary energy necessary to produce a certain energy flow, by multiplying it for its primary energy factor [6]. For example, the primary energy necessary to produce 1 Joule of electricity is more than the one necessary for 1 Joule of low temperature heat. Another approach, is to consider the exergy content of a certain amount of energy, as explained in the following paragraph. The reduction of the energy demand of a building can be achieved already from the design moment, considering both general aspects and particular solutions; in this way, some synergies can be exploited, leading often to unexpected better results and in cost containment. Anyway, some interventions are possible also on existing buildings or during a restoration. Some possibilities are illustrated below [2],[5],[7]: Increase of insulation (of the building envelope, pipes, burner) and air tightness; Increase of heating/cooling system efficiency; Optimisation of solar shading and ventilation systems; Optimisation of building shape and orientation, including openings, windows (glazing and frames), overhangs; Exploitation of natural phenomena, like solar radiation, wind, ground storage capability, vegetation, neighbouring structures; Optimisation of internal gains, natural and artificial lighting, electrical devices. Proper choice of building materials, and mass of the structure.

2

1.4. LOW EXERGY BUILDINGS 1.4.1. Exergy concept and definitions The concept of exergy is well known within Thermodynamics, but is important to focus the attention on the meanings that the word can have as far as Building Physics is concerned. According to this concept, it is not enough to consider the quantity of energy, but also its quality. It is possible to define a scale of the quality of certain amounts of energy, considering their transformability into other kinds of energy. According to the second principle of Thermodynamics, the transformations of energy are not reversible, for the increase of entropy of the Universe (the system and the surrounding ambient at the same time), so that a degradation of energy occurs. For example, 100W of mechanical work can be completely transformed in heat, while 100W of heat, obtained from the cooling of a hot water tank, cannot be completely transformed in work, according to Carnot’s principle. A more precise definition of Exergy is the following: “The maximum mechanical work that can be theoretically obtained with transformations that lead the system to the dead state, being this last defined as the condition of mechanical, thermal and chemical equilibrium with the environment” [8]. Exergy can be valuated just using the two Principles of Thermodynamics, as reported below, referring to the generic steady-state system illustrated in Figure 1.1, and neglecting the chemical component.

3

q0

∑

k

qk

∑

j

& mj i Σ e P

∑

j

& mj

Figure 1.1 – A general system and its energy exchanges

Equation 1.1 can be written, according to the First Principle of Thermodynamics, while the Second Principle allows to write Equation 1.2.
& & q 0 + ∑k q k = ∑ j m j (h j ,e − h j ,i ) + ∑ j m j
& ∑ m(s j j ,e

w 2,e − w 2,i j j 2

& + ∑ j m j g (z j ,e − z j ,i ) + P

(1.1) (1.2)

− s j ,i ) =

q0 q & + ∑k k + ∆S irr T0 Tk

Multiplying Equation 1.2 for T0, and adding it to Equation 1.1, the following expression can be obtained: w 2,i − w 2,e  T0  j j  & & P = ∑k q k 1 −  + ∑ j m j h j ,i − h j ,e − T0 (s j ,i − s j ,e ) + ∑ j m j +  2  Tk  & & + ∑ m j g (z j ,i − z j ,e ) − T0 ∆S irr

[

]

(1.3)

j

The maximum obtainable work, in a unit time, can be valuated with Equation 1.3, when

& ∆S irr = 0 [9].
1.4.2. Thermal Exergy

Particularly interesting within Building Physics is the exergy associated with a heat exchange. In fact, the maximum mechanical work obtainable from a source at temperature T1, exchanging the heat Q1 with the environment at temperature T0, is:

4

 T E = Q1 1 − 0  T 1 

   

(1.4)
 T  is called Carnot factor, while the therm A = Q 0 is called anergy, and  T1 

 T The therm 1 − 0  T 1 

represents the amount of heat not transformable into work [9]. The definition of the environmental temperature T0 is arbitrary, and it depends on the target to be studied. For example, T0 could be the instantaneous value of external air temperature, and this could be interesting for studying the exergy related with a cooling machine, refreshed by external air. T0 could be the temperature of the ground, and this is useful for studying the heating with a ground-coupled heating pump. If we want to study the exergy related with the heat transfer from a radiator to the indoor ambient, T0 should be defined as the internal room temperature. According to what already stated, a low quality or low exergy heat flux is an exchange of energy from sources at a temperature close to ambient temperature.
1.4.3. Low Exergy Buildings

In Low Exergy Buildings, the target of using energy at its minimum quality is aimed. Of course, high quality energy, such as electricity, is used for certain functions, but low exergy sources are used for heating or cooling (and for ventilation, when possible). This means, for example, that heating is supplied to a thermal zone at a low temperature, while cooling is achieved with relatively high temperature. From a generic point of view, in a steady-state condition, it’s possible to state that the power given by the heating system to the zone must be the same power lost by the zone to the external, due to a temperature difference, as illustrated in Figure 1.2.

5

AE.S qint TE.S.

Aext qext Tint Text

Figure 1.2 – Heat exchanges of a simple thermal zone

If, as a first approximation, radiation exchanges are neglected, the following equations can be written: q int = h ⋅ A E.S. ⋅ (TE.S. − Tint ) q ext = U ⋅ A ext ⋅ (Tint − Text )
 T  e int = q int ⋅ 1 − int   T  E.S.    T e ext = q ext ⋅ 1 − ext  T Int     

(1.5) (1.6) (1.7)

(1.8) (1.9)

q int = q ext

Low Exergy Buildings present low values of eext and eint. Tint is strictly connected with comfort conditions, so that the only way to achieve a low value for eext is to reduce qext (Equation 1.8), thus minimizing not only the exergy but also the energy of the flux. On the other hand, different things can be done as far as eint is concerned. As a matter of fact, it
 T  is possible to reduce the Carnot’s factor 1 − int  by minimizing TE.S., that is the  T  E.S.  

temperature at which the heat flux is emitted. But, to make possible that qint=qext, it is

6

necessary to adopt a bigger surface for the heat supply (Equation 1.5), if compared to a traditional high temperature heating system. Therefore, the concept of Low Exergy Building is strictly connected with low temperature heating systems and high temperature cooling systems, but also with large surface heat emission systems.
1.4.4. Advantages of Low Exergy Systems

Considering an integrated system, with different energy sources, with different exergy contents, and different kinds of utilisation, it is immediate to think to use the energy at the minimum quality available to satisfy the need. For example, in a combined and cogenerative power plant, with a gas turbine, a steam turbine and a condenser connected to an urban heating grid, the best way to connect the components is to use the total amount of energy in a cascade of exergy values; so that the high temperature products of combustion should first feed the gas turbine, then produce high pressure steam to be expanded in the steam turbine, and only at the end the remaining energy, at a lowest exergy level, can be used for district domestic heating. If all buildings in the grid would be low-temperature heated, the steam could be condensed at a lower temperature, and more work could be obtained from its further expansion, thus increasing the efficiency of the whole system. But, for a single building, connected to the high temperature district heating grid, or heated by its own fuel-feed heat generator, what are the advantages to use low exergy systems? Several advantages have been listed: Possibility to couple the system with low exergy sources, like solar collectors, ground exchangers, trombe walls, low temperature internal gains. Even if those sources are not utilized at the moment of the design, there is the potential possibility to change the source in the future, or to use alternative energies like additional gains. Low exergy systems can use both high quality and low quality energy; traditional, high exergy systems can use only valuable sources. Possibility of heat storage with low temperature differences, minimizing the storage losses; the storage could interest the same mass of the structure, the ground, or other

7

artificial capacities, like water tanks. Besides, the storage can interest both long time fluctuations (seasonal climate changes), or short time variations (daily or shorter). Utilisation of heat pump and refrigerators with an increased efficiency or coefficient of performance (COP). Elimination of local high temperature differences (as in the proximity of radiators), reducing local losses of heat. Less air stratification, allowing a full employment of the heat, and reducing losses. Better utilisation of spaces, thanks to building-integrated technical solutions. All the advantages of large surfaces radiant systems, like better thermal comfort conditions (§§1.5.2), less ventilation losses, better utilisation of spaces. The adoption of condensing heat generators is possible, for the low temperature of return water. If the building is connected to a district heating grid, a smaller heat exchanger is requested, thanks to the bigger temperature differences (but, in this circumstance, the exergy waste is located in the exchanger itself, as in the case of the ZUB building). From a generic point of view, if in a district low exergy buildings are widespread, different politics could be thought. For example, the price of energy could be different according to its quality; moreover, the district heating grid could supply different buildings in a cascade of temperatures, thus allowing a bigger transfer of energy with the same grid. Anyway, Low Exergy Buildings have also some disadvantages: Slow response of the system due to its high thermal capacity, avoiding quick regulations. Impossibility of controlling air humidity, in the cooling period. Higher electrical expenses for pumping, due to the higher water flow rates involved.
1.5. HUMAN COMFORT

Human comfort is related to the degree of satisfaction with the environment, as measured, for example, by a scale of acceptability. Many variables influence comfort, as listed below [10]:

8

Physical Environmental Variables Thermo-physical environment (a combination of air temperature, surrounding surfaces radiant temperature, relative humidity, air speed), safety and security, protection from elements, air quality, acoustic conditions, light, aesthetics, controllability, size. Physiological Variables Metabolism, age, gender, time, health, medication, acclimatisation. Behavioural Variables Clothing, activity, location, posture, the use of controls. Psychological Variables Stress, family or personal relations, relations at work or school, work satisfaction and involvement, perception of control, psychosis. Although human comfort is clearly influenced by physical environment, the perception of the physical factors is strongly influenced by physiological, behavioural and psychological variables. Anyway, there are some physical parameters that most influence thermal comfort, and that it is possible to control in an objective way; they are air temperature, surrounding surfaces temperature, air quality, relative humidity. But also other variables can be monitored and controlled, like lighting, acoustic, air speed, security, aesthetics.
1.5.1. Operative Temperature

Air temperature (Ta) and mean radiant temperature (Tmr) are the variables that mainly influence thermal comfort. They can be considered together in only one parameter, the operative temperature TO:
TO = h c Ta + h r Tmr hc + hr

(1.10)

where hc and hr are respectively the convective and the radiant coefficient, in the heat exchange between the environment and the human body. Since hc and hr has usually very

9

similar values, as an approximation, TO can be considered as the average value of Ta and Tmr [11]. According to Fanger’s studies [12], and to EN ISO 7730 [13], it’s possible to define the values of operative temperature to be guaranteed for achieving comfort condition for most of the occupants. The desired operative temperature can be expressed as a function only of the metabolic rate and of the clothing, as illustrated in Figure 1.3.

Figure 1.3 – Thermal comfort diagram, for -0.5 < PMV < +0.5

In the diagram above, in abscissa is reported the thermal resistance of clothing, measured in clo, while on the ordinate axis there is the operative temperature. The metabolic rate,

10

expressed in met, is reported as a parameter on the diagonal lines. Above the dotted lines are written the shifts from the ideal operative temperature that maintain PMV between -0.5 and +0.5, i.e. PPD < 10%. The arrows highlight the parameters adopted in the simulations of the ZUB building: a metabolic rate of 1.2 met and a clothing factor of 0.6 and 1.2, respectively for summer and winter. They correspond to an ideal operative temperature of 24 °C in summertime, but a variation of ±1.7 °C is tolerated, maintaining PPD < 10%; in wintertime, the temperature range which corresponds to a PMV between -0.5 and +0.5 is 20.5 ±2.5 °C.
1.5.2. Effect of radiant heating/cooling systems on human comfort

As already stated above, a low temperature difference between the emission system and the ambient air implies a big surface for the heat exchange. This entails that, as far as heating is concerned, the average radiant temperature of surrounding surfaces is bigger than a conventional, high temperature, emission system. For cooling, a lower main radiant temperature is aimed. This is due to the bigger view factors of surfaces in the heat exchange between human body and surrounding zone. Since a desired value of the operative temperature has to be reached, if the mean radiant temperature is higher, the air temperature has to be lower, and vice versa (Equation 1.10). This means that, in real applications, air temperature is usually lower of 1 or 2 °C, in wintertime. As an higher mean radiant temperature is reached, the wide surface emission systems are also called radiant heating systems. If compared with high temperature heating, the wide surface heating involves a bigger radiant exchange and a lower convective exchange. As far as comfort is concerned, some advantages of this kind of systems can be listed [7], [11], [14]: Better comfort while breathing fresher air in winter. Higher relative humidity in winter (but also in summer, as dehumidification is not possible), limiting the problems of sore throat or dry skin. Better radiant symmetry, and more homogeneous conditions in the room. Elimination of cold draughts in summer. Better hygienic conditions, as weaker convective air movements can drag less dust.

11

1.6. RADIANT SURFACE SYSTEMS, TECHNOLOGICAL SOLUTIONS

The most widespread technical solutions for low temperature heating systems are superficial thermal layers, TABS (Thermally Activated Building System) and large surface radiators.
1.6.1. Superficial thermal layers

A long serpentine of pipes is laid down within a superficial layer of the floor (Figure 1.4-a), the ceiling (Figure 1.4-b) or the wall (Figure 1.4-c); an insulation layer is requested to limit the heat flux towards the wrong side. This system is the most widespread amongst low temperature heating. As far as cooling is concerned, this solution is not very diffused, for the limited cooling power; in fact, the temperature of inlet water cannot stay below the dew point, because condensation can occur, but not on a controlled surface (like for conventional systems), rather directly on the cooled building surface. Also thermal comfort restrictions limit the minimum temperature of the floor, to avoid cold sensation at feet. From a constructive point of view, several solutions have been adopted, with metal or plastic pipes, with pre-built slabs or in-situ construction. The most widespread technique consists in the lay-down of plastic tubes over an insulation layer on the floor, in a serpentine (Figure 1.5) or spiral (Figure 1.6) disposal.

a
Figure 1.4 – Example of Thermal Slabs [15]

b

c

12

Figure 1.5 – Serpentine arrangement of pipes [15]

Figure 1.6 – Spiral arrangement of pipes [15]

1.6.2. TABS (Thermally Activated Building System)

The aim is to heat or cool not only a little component, but all the building or a relevant part of it [16]. Practically, the temperature of an high mass and high surface building component is controlled, by adopting pipes embedded deep in the concrete (Figure 1.7), or by air ducts within a high mass slab, such as ceilings or floors (Figure 1.8). These solutions are particularly adopted in multi-storey buildings, with the same kind of thermal zones at each floor, so that the heat flux can reach both the upper and the lower storey [17]. Anyway, also a floating floor can be adopted, for acoustic performances, and in this case the heat flux reaches the room by the ceiling. About TABSes with air ducts, the air can be used both as heat carrier and for ventilation. The high thermal capacitance of these systems allows a certain heat storage, and the relevant time-shift of loads and responses permits peak-shavings or self balancing. In this way, it is possible to reduce the maximum power of the chiller machine, by spreading their functioning on a longer period, or let they work during the night, with better coefficients of performance or cheaper electrical tariffs. Also its combination with alternative cooling sources, such as ground exchangers, is very attractive, since the limited peak power of the natural source and its continuous availability, stay in a good accordance with the features of the TABS. These systems are wider dealt in Chapter 3.

13

Figure 1.7 – TABS (Thermally Activated Slab) [18]

Figure 1.8 – TermoDeck® [19]

1.6.3. Large surface radiators

There are specimens of low exergy buildings where common radiators are used as emission systems, but their surface is extremely increased, thus allowing the use of low temperature water for heating, and high temperatures for cooling. In very wet climates, a modest air dehumidification is also possible, if the radiators are provided with a condensate collecting system [14]. This solution is not very widespread, and it is technologically closer to a traditional high temperature heating system. Moreover, a greater investment cost has to be faced, for the increased number of radiator units.
1.7. PASSIVE BUILDINGS 1.7.1. Definition of Passive Building

Many examples of ancient architecture have special design features that provide comfortable living conditions, without the expenditure of conventional energy sources. The shape and orientation of buildings, the arrangement of openings, shading devices, and numerous other design elements were often given very careful consideration, in order to take maximum advantage of the climate. The role of trees, water and vegetation around the building in determining thermal comfort was well appreciated. Massive walls and clustering of houses were also common means of balancing severe temperature

14

fluctuations. A large number of those proven concepts of natural climatic control have been forgotten in the design of buildings today, and a big amount of non-renewable energy has to be supplied to achieve comfort conditions for occupants. Passive Buildings are designed in the way to achieve comfort exploiting only natural phenomena and internal gains. Also little fans, pumps or ventilators can be used for the effective distribution of heating and cooling; these can still be called passive systems, if the total energy consumed by mechanical devices is negligible in comparison to the total energy needed for heating or cooling the building [7].
1.7.2. Passive Building Technical Solutions – Overview

A singular technical device is usually not enough to satisfy the energy needs of a building. The right approach, for the design of a passive building, is an in-depth study of all the local available sources, environment conditions, occupants needs. The interaction of all the adopted solutions, and the effect of several factors must be considered, and adapted to the particular situation, in a very integrated approach. Anyway, different technical devices, for exploiting one ore more natural sources, have been thought, designed or realized. A very concise list is presented below [7]: Solar devices Movable shutters; mirrors; reflectors; translucent layers; coloured surfaces; solar collectors; overhangs; louvers; glazed surfaces; shading by vegetation; shading and reflection by adjacent environment. Heat storage systems Water tanks; rock bed storage; ground storage; water walls; thermally activated building components; additional masses; walls with air cavities. Ventilation systems Stack effect ventilation systems; ventilation by wind pressure; courtyard effect systems (Figure 1.9); wind towers (Figure 1.10); Trombe walls; Natural Cooling systems Surface evaporation; ventilated roofs; fresh external air ventilation; night ventilation; underground or fluent water exchangers; ground coupling; evaporative cooling by

15

vegetation (Figure 1.11); increased radiation by movable elements; Vary Therms Walls. Convective loops Trombe walls (Figure 1.12); stack effect convective systems; solar chimneys (Figure 1.13); air cavities and ducts; appropriate air flux connections amongst rooms, spaces and cavities. Hydronic circuits Pipes embedded in the concrete or in the ground; heat exchangers; water walls; rain collectors. Exploitation of internal gains Rational utilisation strategies; natural lighting optimisation; insulating materials; additional masses for heat storage and dampening of temperature fluctuations. Exploitation of surrounding environment Rational positioning and orientation, considering sun beam directions, cast shadows, wind obstacles, ground conditions; vegetation; modification of external environment; appropriate external shape for the exploitation of wind pressures. Minimization of energy losses Insulating materials; multi-pane glazing, with low-conductivity gas gaps, lowemittance, reflecting or selecting films; minimization of cold bridges; reduction of air infiltrations; heat recovery systems.

Figure 1.9 – Courtyard effect [7]

16

Figure 1.10 – Wind tower [7]

Figure 1.11 – Combination of sensible and evaporative cooling [7]

17

Figure 1.12 – Trombe wall [7]

18

Figure 1.13 – Solar chimney [7]

19

1.8. REFERENCES

[1] [2] [3] [4] [5]

SBIC, Sustainable Buildings Industry Council; internet website: www.psic.org International Energy Agency; internet website: www.iea.org Renewables 2004, International Conference for Renewable Energies, 1-4 June 2004, Bonn, Germany. Internet website: www.renewable2004.de PAOLELLA A., “Design for Housing – Technology, Environment, Society”, Design for Housing, Falzea Editore, Reggio Calabria (I), 1, February 2004. WEBER T., Energy Performance of Buildings / Methodologies for experimental verification, The Royal Institute of Technology, Department of Civil and Architectural Engineering, Division of Building Technology, Stockholm (SE) 2004, Doctoral Thesis. SCHMIDT D., Methodology for the Modelling of Thermally Activated Building Components in Low Exergy Design, , The Royal Institute of Technology, Department of Civil and Architectural Engineering, Division of Building Technology, Stockholm (SE) 2004, Doctoral Thesis. BANSAL N.K., HAUSER G., MINKE G., Passive Building Design: a handbook of natural climatic control, Elsevier Science B.V., Amsterdam (ND), 1994. CAVALLINI A., Lectures notes, “Energetica”, University of Padova, academic year 2002-2003 (in Italian). CAVALLINI A., MATTAROLO L., Termodinamica Applicata, cleup, Padova (I), 1992 (in Italian).

[6]

[7] [8] [9]

[10] NILSSON P.E. ET AL., Achieving the Desired Indoor Climate, Energy Efficiency Aspects of System Design, Studentlitteratur, Lund (SE), 2003. [11] BETTANINI E., BRUNELLO P.F., Lezioni di Impianti Tecnici – volume primo, CLEUP, Padova (I), September 1993 (in Italian). [12] FANGER P.O., Thermal comfort: analysis and applications in the environmental engineering, McGraw Hill, New York, (N.Y.), 1972. [13] ISO, “ISO 7730-1994, Moderate thermal environments – determination of the PMV and PPD indices and specification of the conditions for thermal comfort”, International Organization for Standardization, Geneva (CH), 1994.

20

[14] Annex 37, International Energy Agency, “Low Exergy Heating and Cooling of Buildings”, Annex 37; internet website: www.vtt.fi/rte/projects/annex37 [15] ZECCHIN P., Dynamic Simulations for Optimizing Different Control Strategies of the Integrated HVAC System of the ZUB Office Building, University of Padova (I), 2004, Master Degree thesis. [16] DE CARLI M., HAUSER G., SCHMIDT D., ZECCHIN P., ZECCHIN R., “An Innovative Building Based on Active Thermal Slab Systems”, proceedings to the 58th ATI National Conference, 9-12 September 2003, San Martino di Castrozza, Italy. Available in internet: http://141.51.43.66:9191/SolarOpt/dokumente/veroeffentlichungen/S5-14_finale.pdf [17] HAUSER G., KEMPKES C., OLESEN B.W., “Computer Simulation of Hydronic Heating/Cooling with Embedded Pipes”, ZUB Jahresbericht 2000, ZUB Zentrum für Umweltbewusstes Bauen, Kassel (D). [18] EMPA, Swiss Federal Laboratories for Materials Testing and Research; internet website: www.empa.ch, Organization > Materials and Systems for Civil Engineering > BuildingTechnologies > Energy Systems / Building Equipment > Thermally activated building systems [19] TermoDeck®, internet website: www.termodeck.com

21

Chapter 2 THE Z.U.B. BUILDING
2.1. DESCRIPTION OF THE ZUB BUILDING

The ZUB (Centre for Sustainable Building), at the University of Kassel (Germany) (Figure 2.1), is a pilot project of the employment of thermally activated building systems. It has been built in 2001 thanks to a collaboration between Department of Building Physics, Department of Building Services and Department of Experimental Building. The ZUB is an innovative demonstrative building, in the frame of the IEA ECBCS Annex 37 “Low Exergy Systems for Heating and Cooling of Buildings” [1], designed for an annual heating demand less than 20 kWh/m2. It is based on the implementation of low energy heating/cooling systems, control strategies and new building materials, whose aim is to state the possibilities of active thermal slab systems to achieve comfort and energy saving. The building is close to the Faculty of Architecture, in an old urban neighbourhood, [2].

Figure 2.1 - The office building of the ZUB (Centre of Sustainable Building) [Meyer]

23

As far as the construction is concerned, it is a 3-storey building with an atrium, used as a light gap, and a basement (Figure 2.2). It consists mainly of three different parts: the first one for exhibitions, one part for offices and the last one for experiments and researches. The overall volume is 6882 m3, with a net floor area of 1732 m2 and a main floor space of 892 m2. The height of each floor is 3.4 m, except for the basement (2.8 m) and the ground floor (3.7 m), and the main experimental room (6.7 m) (Table 2.1). On the ground floor, most of space is occupied by the lecture hall and by the main experimental room, that is situated also in the basement. Heating and cooling equipment and ventilation systems are installed in the basement.

Figure 2.2 - Sections of the ZUB building [Sadding]

Table 2.1 - Features of the rooms in the ZUB [3] Area [m ] Height [m]
2

Offices 24 3.4

Lecture hall 174 3.7

Experimental room 94 6.7

Corridor 60 3.4

24

External walls (120 mm concrete + 300 mm insulation) are insulated with polystyrene and the large glazing surface (Figure 2.3) is south oriented with a g-value of 0.42. To optimize solar gains, the windows of the south façade and the atrium have been realised with the minimal frame-fraction of the construction (Figure 2.4) and external rolling shutters have been installed outside (Figure 2.5). The U-values of all the external surfaces are reported in Table 2.2.

Figure 2.3 – South Façade [Sadding]

[Meyer]

Figure 2.4 – Internal view of the large glazing façade [Meyer]

Figure 2.5 – External shading device [Meyer]

Table2.2 - U-values of the building structures [3] Building part Exterior walls Roof Windows Wall/floor against ground U-value [W/(m2 K)] 0.11 0.16 0.80 0.26

25

As for internal walls, the floors and the ceilings consist of concrete slabs whereas the dividing walls of offices consist of hollow bricks. A particular wall, however, has been realized to separate offices and other rooms from the atrium. It is a 635 mm thick clay wall, with an inner air cavity of 365 mm, made by massive unbaked clay bricks, that has a great heat retention capacity and the capability of dampening fluctuations of moisture [2],[3] (Figure 2.6).

1: Clay bricks, facing the atrium 2: Air gap 3: Clay bricks, facing the office

Figure 2.6 – Thick clay wall

2.2. RADIANT AND VENTILATION SYSTEMS

In the ZUB building, radiant systems (§1.6) for heating and cooling have been installed. In order to investigate different solutions, some particular innovations have been introduced. The pipes are embedded both in the upper concrete layer on the floor and in the centre of the slab, as can be seen in Figure 2.7, so that a traditional floor heating system is combined with a thermally activated slab system. Pipes are made in polyethylene with a diameter of 20 mm and a distance of 150 mm, except in the basement, where the diameter is 25 mm. The disposal of the pipes has a coil shape (Figure 1.6) with one loop for each office. In this way, each room has its own control system, and a local knob allows to modify the set temperature of the room of ±2 °C. Each circuit of the floor radiant system and the active thermal slab system is supplied by about 600 kg/h water mass flow rate, thus allowing to keep the difference between supply and return temperature lower than 4-5 °C.

26

Roof

Second floor

First floor

Ground floor

Basement

Fig. 2.7 - Position of different pipe layers in the concrete slabs of the ZUB building [2]

In the case of heating, the radiant system is connected with the district heating supply system and it is divided in two different circuits to supply the traditional floor system and the thermally activated slab systems; the two different circuits can be activated individually or at the same time, for research purposes. As for the cooling system, the hydronic pipes circuits employed are the same as the heating system but, for investigating the possible use of renewable energy sources, an additional circuit of pipes in the slab construction of the basement has been installed, to exploit the ground coolness to cool the water. Thus, the ground heat exchanger replaces the installation of a mechanical cooling machine. A particular type of ventilation system, that allows fulfilling the requirements for airchange and air-quality has been installed. According to the standard DIN 1946 [4], the required air flow rate is 7100 m3/h. In the ZUB building, however, particular studies about the ventilation system have been carried out to reduce its size and heat losses. Therefore, a mechanically balanced ventilation system, using heat recovery with two cross flow heat exchangers in a series has been installed. The designed air flow is 4000 m3/h and the inlet temperature of the supply air flow is 17.7° C. Such system is not sufficient to supply the offices and the fully occupied lecture hall at the same time. Therefore, during the maximum occupation, the air handling unit can achieve a sufficient air change in the lecture hall, while the offices can be ventilated by natural means. A VOC sensor in each room gives information to the ventilation system, thus varying the air flow in all the offices at the same time, according to the maximum requested value, maintaining the concentration of organic compounds under 10 PPM.

27

In the normal operation mode (Figure 2.4.A), fresh air is supplied directly to the office rooms and exhaust air is extracted from the atrium, and then transferred to the heat recovery unit; for research purposes, the fresh air can be supplied to the central atrium and extracted from the offices (Figure 2.4.B). A certain natural ventilation can be achieved, without the adoption of fans, just exploiting the stacking effect of the entrance hall. In this way, fresh air is supplied through the open windows and the exhaust air is aspirated into the atrium through air intakes, which are set in the clay wall, near the doors (Figure 2.5). The gradient of air density in the building, due to the temperature difference between indoor and outdoor air, allows a natural air circulation, from the windows up to the openings at the top of the entrance hall (Figure 2.4.C).

A
Fig. 2.4 - Operation modes of the ventilation system

B

C

Figure 2.5 – Air intakes in the clay wall [2]

28

To reduce electricity consumption, an accurate study of lighting has been conducted. Each office gets light from the windows of the south façade and the atrium from its glass walls, and the presence of light sensors allows using artificial lights only when a room is occupied and the value of natural light is too low. The presence of these large glazing areas is also used to exploit solar gains to reduce heat consumption. Solar radiation and internal sources reduce the requirements of the heating system, while the presence of external rolling shutters allows to limit the incoming solar heat gain in the case of cooling. The choice of the artificial lights in the rooms (Table 2.3) has considered also the thermal loads they may produce.
Tab. 2.3 - Lights installed in the ZUB building Number of lights Power [W] Efficiency [lm/W] Illuminance [lx] Heat gain [W/m2] Offices 4 39 85 553 6.5 Corridor 6 55 91 500 5.5 Experimental room 24 55 91 1278 14.0 Lecture hall 36 55 91 1036 11.4 Atrium 22 20 25 63 2.5

The energy consumption for the entire office building has been evaluated according to the German energy code WSchVO’95. The heating demand is 5.3 kWh/m3 per year (16.5 kWh/m2 per year), that is 27% of the limiting maximal value required by German standards, and an electrical consumption of approximately 10 kWh/m2 per year. During the year 2002, the measurements have shown a heating energy demand of about 23 kWh/m2 and a total energy demand, including electrical consumption, lower than 40 kWh/m2 [3].
2.3. MEASUREMENT EQUIPMENT

In the ZUB building, one of the main issues is the monitoring of all the data that allow verifying and controlling concepts and researches. For this reason, an intensive project for “Solar Optimised Buildings” (a national research program promoted by the German Ministry of Economy and Technology) is currently being run. According to this

29

program, the planning and construction processes are being followed up over a period of four years and for at least two years measurements of main parameters are being drawn out. To collect data such as heat exchanges, temperature, thermal comfort and air quality, however, a complex measurement system and detailed programs are required. In the office building of the ZUB, the survey of these parameters has been planned to supervise all the building and to investigate in depth the behaviour of control strategies and systems (Table 2.4).

Tab. 2.4 - Monitoring plan of the ZUB
MONITORING BUILDING Energy Consumption Daily Employment Lighting Sensitivity Service Management Systems Interaction SINGLE ROOM Thermal Comfort Visual Comfort Air Quality Shading EQUIPMENT Ventilation System Natural Ventilation Heating System Cooling System

The adopted measurement system consists of 448 sensors, located in different positions of the building, that allows to collect 1172 data every minute. Moreover, some weather instruments have been set out of the building to measure outside parameters, such as air temperature, solar radiation, air moisture, speed and wind direction. To control all the equipment, then, thermocouples, hygrometers and water mass flow meters have been installed in the ventilation system and in the heating/cooling radiant systems. In this way, the working parameters and the efficiency can be analyzed and improved and the energy consumption can be assessed. In all rooms, finally, several sensors have been placed to gather information about internal conditions of thermal comfort, air quality and heat loads. Instruments, able to assess the intensity of lighting, both natural and artificial, analyse the possible produced heat gains. Furthermore, a set of temperature and air humidity probes in the rooms allows calculating the operative temperature. The most relevant aspect of this measurement system, however, is the installation of thermocouples inside the concrete

30

slabs. As a matter of fact, some researches deal with the behaviour of active thermal slab systems related to peak-shaving. In order to monitor the temperature of active layers, during the construction of the building, some sensors have been set in the concrete slabs (Figure 2.6). Through these instruments, the data concerning temperature and performances of the active thermal slab system can be collected, both upright and horizontally in the volume of the slab, and a detailed map of heat exchange can be drawn up. It is also possible to know the direction of heat exchanges and the behaviour of the used building materials. Nevertheless, the installation of these sensors is not sufficient to manage and analyse all the data collected. Because of the large number of parameters to be processed every day (about 846000) a careful method avoiding hitches is required. The employed sensors are connected, at first, to three computer centres that survey the measurements coming from the building, the rooms and the equipment. In these centres, they are recorded as ASCII data including the information about control strategies. Then, through a bus system, they are sent to the management system, where they are processed and recorded in a database that is also connected to a display station. Through this system, therefore, the collected information are displayed every minute, thus allowing to understand the efficiency of the installed systems, the thermal conditions in rooms, and to optimise the chosen control strategies [3].
Roof First and second floor Ground floor Basement

Fig. 2.6 – Sample of disposition of the thermocouples in the slabs [2]

31

2.4. REFERENCES

[1] [2] [3]

Annex 37, International Energy Agency – Low Exergy Heating and Cooling of Buildings – Annex 37; internet website: www.vtt.fi/rte/projects/annex37 SALDANHA, DE M., “Report Phase I”, SolarOpt. Available in internet: http://www.solaropt.de DE CARLI M., HAUSER G., SCHMIDT D., ZECCHIN P., ZECCHIN R., “An Innovative Building Based on Active Thermal Slab Systems”, proceedings to the 58th ATI National Conference, 9-12 September 2003, San Martino di Castrozza, Italy. Available in internet: http://141.51.43.66:9191/SolarOpt/dokumente/veroeffentlichungen/S5-14_finale.pdf DIN 1946-6, 1998-10, Raumlufttechnik - Teil 6: Lüftung von Wohnungen; Anforderungen, Ausführung, Abnahme (VDI-Lüftungsregeln); (in German).

[4]

32

Chapter 3 TABS (THERMALLY ACTIVATED BUILDING SYSTEMS)
3.1. INTRODUCTION

The aim of TABS (Thermally Activated Building Systems) is to involve in the heat exchange a big portion of the mass of the building, for allowing a certain heat storage and for taking advantage by the high thermal capacity of the system [1]. But the TABS is both a storage component and an emission system, therefore low temperatures, wide surfaces and heavy masses have to be involved. That is why the most widespread technical solution for TABS are thermally activated concrete slabs, and often the word TABS implies this component, even if some applications of pipes embedded in pillars, beams, columns can be found. The temperature of the heavy slab can be controlled by water pipings embedded in the depth of the concrete (Figure 1.7), or by air circulating within cavities (Figure 1.8). In the second solution, the air transfers or adsorbs heat from the concrete slab, and finally it is introduced into the room, satisfying ventilation requests and supplying an additional convective heating [2]. Anyway, the solution with water circulating in polymeric pipes embedded in the depth of the concrete is the most widespread.
3.2. TABS FEATURES

The concrete slab is therefore included in the heat transport mechanism and acts both as emission system and as heat accumulator, which can be charged during periods when conditions for cooling are unfavourable and then, for instance, discharged during the night, when outside temperatures are low (Appendix A1) [1]. The concrete slab, however, can only store heat if its temperature can be continuously increased during the charging time, according to the following equation: Q = C ⋅ ∆TS (3.1)

where Q is the energy stored, C the thermal capacity of the slab and ∆TS is the temperature increase of the slab. To be precise, C cannot be considered a constant, because the slab,

33

during the charge, exchanges heat by convection and radiation with other surfaces and with internal loads; so equation (3.1) should be rewrite as [3]:
Q = ∫ CdTS ,
1 2

C=

∂Q ∂TS

(3.2)

Also TS has been considered homogeneously identical in all the slab, but this is only an approximation, used in this context. Moreover, also the other masses contribute to the heat storage and transfer, even if their temperature is not directly controllable; for this reasons, equations (3.1) and (3.2) should be considered just an indication, and cannot be directly used for calculating the amount of storable heat. Anyway, equation (3.1) shows that, if a certain amount of energy has to be stored, a certain increase of the slab temperature TS has to be allowed [4]. The slab has a wide surface, so its superficial temperature strongly influence the mean radiant temperature; also a relevant convective exchange with the room takes place, modifying air temperature Tair. In this way, the superficial temperature of the slab, strongly influences operative temperature Top (Equation 1.10). Since Top is related with comfort conditions, the maximum TS increase that can be allowed is the one that let Top increase within a certain interval, according to EN7730 [5]. This criterion, therefore, limits the possible heat storage capacity of the concrete slab. According to Equation 3.1, if a relevant amount of heat has to be stored, being ∆TS limited by the previous considerations, an high value of the thermal capacity C has to be guaranteed. To reach the aim, both the mass and the specific heat capacity of the building materials have to be relevant, according to the following equation:

C = ∑ j c j ⋅ ρ jVj

(3.3)

Common building materials, such as concrete or bricks, can have a relevant heat storage capacity, if a consistent mass is involved in the process and if a certain temperature difference is allowed. Thermophysical properties of common building materials are reported in Table 3.1. It is possible to notice that, with the exception of wood, metal and plastics, the heat capacity of most building materials is approximately proportional to their mass, since they all have about the same specific heat capacity, equal to 0.24 Wh/kgK.

34

Since a necessary condition for heat storage is the temperature increase of the slab, these systems are particularly suitable for all those applications where a big variation of temperatures and loads take place, such as in office buildings, where the contribution of internal loads, during occupation, is relevant, or in summer time.
Table 3.1 – Thermophysical properties of common building materials [6] Material Density [kg/m ] Aluminium Construction steel Normal concrete Foam concrete Solid brick Lime/sand stone Glass Earth Wood Polystyrene hard foam Mineralwool Water 2700 7850 2400 400-800 1200-2000 1000-2200 2480 1450-2040 600-800 10-30 8-500 1000
3

Thermal Conductivity [W/mK] 165 60 2.1 0.14-0.27 0.50-0.96 0.5-1.3 0.8 0.5-2.6 0.13-0.20 0.03-0.04 0.03-0.05 0.60

Specific Heat Capacity [Wh/kgK] 0.25 0.13 0.24 0.24 0.24 0.24 0.19-0.26 0.24 0.66 0.42 0.24 1.2

3.3. DYNAMIC BEHAVIOUR

Considering the external envelope of the building, non-stationary boundary conditions on the surface of a wall are determined by daily changing of air temperature and solar radiation. Diurnal variations produce an approximate 24 hours cycle of increasing

35

and decreasing temperatures. The effect of this on a building is that the heat flow into the thermal zone is also cyclic (heat flows through the envelope during the hot period, some of it is stored, and during the cool night period the heat is partially dissipated to the environment and partially reaches the zone). Two quantities, characterizing this periodic change, are the time lag φ (or phase shift) and the decrement factor µ (or amplitude attenuation), as they are illustrated in Figure 3.1, in the case of harmonic boundary conditions.

Ta

Outside Wall Temperature Te

Ti

Inside Wall Temperature Ti

Day

Day

∆Te
Night

t

∆Ti
Night

t

1 day

1 day

T

Time lag φ

Decrement factor

µ=

∆Ti ∆Te

Te Ti

t

1 day

Figure 3.1 – Time lag (φ) and decrement factor(µ)

36

For a single layer wall, and for harmonic boundary conditions, φ and µ can be easily calculated from the following equations [6]:
 

µ = exp − s  ϕ=s where

ωρ ⋅ c   2λ  

(3.4)

ρ ⋅c 2λω

(3.5)

ω=

2π ( s −1 ) (24 ⋅ 3600) These simple equations consider the wall as a stand-alone entity, but the boundary

conditions do not depend only on external climate, but also on the situation of the wall itself, and on its exchanges with other building components. Anyway, a computer simulation of the behaviour of the slab has been performed and compared with the theoretical values obtained with Equations 3.4÷3.5, relatively to the heat conduction from the core of the concrete slab to its surface. In Table 3.2, the results of simulation are reported; it can be noticed a good correspondence amongst the results obtained by equations and by computer simulation.
Table 3.2 – Comparison between calculated and simulated values of time lag and decrement factor

CALCULATED thickness time lag [h] 5 cm 1.3 8 cm 2.0 decrement factor 0.72 0.59

SIMULATED thickness time lag [h] 5 cm 1.2 8 cm 2.2 decrement factor 0.77 0.62

Also internal walls continuously exchange long wave radiations with other surfaces and with occupants, exchange heat by convection with air, by conduction on the edges, and absorb energy by short wave radiation. For these reasons, internal masses are linked to external periodic variations in a more complex way. By computer simulation, it is possible to calculate typical time lags and decrement factors for different kinds of structures, even if their definition is not univocal. In this context, it is important to underline that the building materials, their mass and their

37

thermophysical properties influence not only the heat storage capability, but also the dynamic behaviour. Above all, the time lag is an important parameter to be considered, if one of the aims is to transfer heat from day to night. Moreover, it is important to characterize the response of the building not only to external variations, but also to internal gains. Particularly interesting is the behaviour of the building when thermally activated slabs are operating. The conduction of heat through the slab itself, its heat exchange by radiation and convection are not contemporaneous and not independent, so that its characterization is very complex, and steady state simulations cannot be used for modelling thermally activated slabs.
3.4. TABS ADVANTAGES

Since Thermally Activated Slabs are both emission systems and storage devices, some advantages can be found out in relation with the two functions. As far as emission is concerned, a Thermally Activated Slab can be consider a wide surface radiant emission system, with all its relative advantages (§§1.5.2). On the other hand, the possibility of storing some heat offers several opportunities, depending on climate conditions, environment features, available sources, installed heating or cooling generators, energy prices. An attempt of listing advantages and opportunities has been done below [1],[7],[8],[9], but they are often strictly related one to each other, so that a clear classification is not always possible. Peak-shaving For the high thermal capacity of the system, a certain dampening of temperature fluctuations is acted by the mass itself, so that the cooling or heating action can be acted in a longer period, thus reducing the maximum power of the plant. Possibility of free night cooling In those regions characterized by hot days and fresh nights, a certain quantity of heat, or in case all its amount, can be adsorbed during day and relayed in the night, implying a reduction of expenses for cooling or allowing, with favourable condition, a free night cooling.

38

Energy Saving for mechanical cooling The possibility of acting a mechanical cooling during night, when the temperature difference between inside and outside is lower, implies a better COP of machines, allowing a certain energy saving. Money Saving for mechanical cooling Often, during night, electrical cost is lower, and this involve a lower expense for cooling. Also the maximum installed power is lower, thus reducing both the fixed tariffs of electricity and the price of the chiller. Possibility of coupling with ground exchangers The coupling of TABS with ground exchangers is particularly suitable, for several reasons. First, the relative high temperatures of the ground can be useful only for wide surface cooling systems; then, the continuous availability of a low power source is in good accordance with the storing capacity of the slab. Moreover, the peakshaving effect, due to the high thermal capacity of TABS, involves a considerable reduction in the size of the ground exchanger, with relevant reductions of costs, especially in the case of earth drillings. Best exploitation of solar and internal gains During the heating period, could be interesting to store not only heat from the pipes, but also the heat deriving by internal and solar gains. Especially for rooms with wide surface glazing, the solar gain could be relevant also in winter, but its discontinuity limits its exploitability. The high thermal capacity of a TABS allows a certain storage, at relatively low temperatures, of natural gains. Moreover, if only a portion of the surface is hit by solar radiation, the circulation of water in the pipes involves a better distribution of heat, improving both the exploitation of the gain, and the uniformity of heat emission.

39

3.5. TABS DISADVANTAGES AND LIMITATIONS

Besides advantages, Thermally Activated Building Systems have some limitations, some of them related to the same features that are peculiar advantages of these systems. For example, their high thermal capacity gives several benefits, as illustrated above, but, on the other hand, makes the system very slow, and a quick variation of temperature is impossible. An attempt of listing TABS disadvantages and limitations has been done below [1],[7],[8]: Slowness of response The high thermal capacity makes impossible quick variations of temperature and quick starts of the plant. Therefore, in buildings with random or occasional occupation, this kind of systems is not very suited. The time lag has the disadvantage of shifting also the effect of a desired regulation; this implies that every action should be acted considering its future effect. Or, with other words, an actual desired variation of internal temperature should have been achieved by a previous action. Higher costs The heavier masses involved, imply an higher cost of building materials, even if a containment of the expenses can derive from the simplicity of construction. Emission limitations A thermally activated slab is more simply employed on multi-storey buildings, so that the heat flux from both surfaces of the slab can reach two zones on two adjacent floors, and insulation layers are not required, containing the costs. The correct positioning of the pipes can guarantee that the right portion of heat reaches each floor. This solution is not possible when, for example, there are different utilisations or different ownerships of the adjacent thermal zones, or when a floating or raised access floor is adopted. In these cases, the heat emission has to be acted only by the ceiling (or, in case, by the floor), and an insulation layer has to be adopted, to contain heat flux on the wrong direction.

40

Limited cooling power The temperature of the slab must be always over the dew point, for avoiding condensation directly on the floor or on the ceiling. This implies that, with high relative humidity, the cooling power is very limited, and a dehumidification is not possible. If the control of humidity is necessary, another system has to be added, stand-alone or integrated within the ventilation plant. Also thermal comfort restrictions limit the minimum superficial temperature, especially in the case of floor cooling. Difficult modelling and design The non-steady state modelling and the non-mature knowledge of this relatively new technology limits its adoption and its diffusion.
3.6. TABS MODELLING

As settled above, only non-steady state simulations can be useful for modelling Thermally Activated Building Systems. The heat transfer from pipes embedded in the concrete, or from air cavities, to the mass and eventually to the surface takes place thanks to temperature gradients that promote conduction. Therefore, the temperature distribution within the mass of the component has to be considered in detail. Finite Element (FE) or Finite Difference (FD) codes can reproduce such distributions as long as the boundary conditions are known. But, in the case of TABS design, boundary temperatures depend on the interaction between the considered slab with the room air, the adjacent surfaces and the heat sources. Therefore, the FE or FD model should be integrated into the thermal room or building model. But the two computational approaches, FE and FD on one hand, and transient building simulations on the other hand, are very different, and their integration is not very simple, and would lead to heavy computational efforts. Transient building models consider each component from a less detailed point of view; the response functions of each object are known, so that, to a certain input corresponds a determined output. In the building simulation all inputs and outputs of each entity have to be properly linked, so that the developing of thermal conditions depends on the interaction of all factors. For the implementation of TABS in this kind of simulations, a non-detailed point of view of the

41

component has to be kept. The aim is to know the outputs of the object to certain inputs, in a dynamic situation, without entering in the detail of the local heat transfer. The MEM (Macro Element Model) method are very suitable for the purpose, since the building component is divided in a limited number of elements. According to this method, a TABS can be modelled by an optimised network of thermal resistances and capacities. The resulting simplified model can be described in the Neutral Model Format (NMF), allowing it to be implemented in dynamic simulation programs, such as TRNSYS or IDA [7]. A similar model of thermal activated slabs is already implemented in TRNSYS 15, and it has been wider described in §§4.2.3.
3.7. TARGETS AND PROPOSALS

The main proposal of this work is to improve the regulation strategy of the thermally activated slab actually adopted in the ZUB building. To reach the aim, a TRNSYS model of an office room allows to simulate the performances of the heating and cooling system, varying several parameters and monitoring many outputs. The first task is to handle the slow response of the concrete slab, due to its big thermal capacity, observing the relations between inputs and outputs, disturbances and temperature variations. The implementation of fictitious climate data, gives the possibility to act a control taking into account also the conditions of the weather in the following time. The second aspect to consider, is the exploiting of the big thermal capacity of the heavy concrete slab, and its heat charge and discharge. The heat storage should occur in the proper time, to achieve a certain energy saving and, in the case, a reduction of the maximum power involved. Being the TABS both an emission system and an heat storage device, its two functions have to work in good accordance, without any interference. The regulation strategies are tuned on the model of a room of the ZUB building, but it is interesting to verify their performances also in other conditions and with other climates, to find out more generic results.

42

3.8. REFERENCES

[1]

DE CARLI M.,KOSCHENZ M., SCARPA M., ZECCHIN R., “Metodologia Semplificata per il Dimensionamento di Sistemi Radianti ad Alta Inerzia Termica”, proceedings to the 58th ATI National Conference, 9-12 September 2003, San Martino di Castrozza, Italy, (in Italian). Termodeck®, internet website: www.termodeck.com CAVALLINI A., MATTAROLO L., Termodinamica Applicata, cleup, Padova (I), 1992 (in Italian). KOSCHENZ M., DORER V., “Interaction of an air system with concrete core conditioning”, Energy and Buildings, 30, Elsevier Science (1999), pg. 139-145. ISO, “ISO EN 7730-1994, Moderate thermal environments – determination of the PMV and PPD indices and specification of the conditions for thermal comfort”. International Organization for Standardization, Geneva (CH), 1994. BANSAL N.K., HAUSER G., MINKE G., Passive Building Design: a handbook of natural climatic control, Elsevier Science B.V., Amsterdam (ND), 1994. SCHMIDT D., Methodology for the Modelling of Thermally Activated Building Components in Low Exergy Design, , The Royal Institute of Technology, Department of Civil and Architectural Engineering, Division of Building Technology, Stockholm, Sweden 2004, Doctoral Thesis. NILSSON P.E. ET AL., Achieving the Desired Indoor Climate, Energy Efficiency Aspects of System Design, Studentlitteratur, Lund (SE), 2003. EMPA, Swiss Federal Laboratories for Materials Testing and Research; internet website: www.empa.ch, Organization > Materials and Systems for Civil Engineering > BuildingTechnologies > Energy Systems / Building Equipment > Thermally activated building systems

[2] [3] [4] [5]

[6] [7]

[8] [9]

43

Chapter 4

TRNSYS A TRANSIENT SYSTEM SIMULATION PROGRAM
4.1. BUILDING SIMULATION PROGRAMS

With the nomenclature of Building Simulation Programs are here mentioned those software that try to model all a building or a thermal zone, considering the interaction of several components. Other kinds of software, as Finite Element, Finite Differences, or Fluid-dynamics models, are intended to investigate only a particular aspect of the building, rather than the whole construction at the same time. Amongst Building Simulation Programs, a further distinction can be done between Steady State Simulations, and Transient System Simulations. Their distinction is illustrated in the following paragraphs.
4.1.1. Steady State Simulations

The most widespread building simulation programs analyse steady state conditions, calculating heat losses through the external envelope of the thermal zone, in the most critical winter conditions, when the temperature difference between internal and external is the maximum. The most important parameter to be defined for this kind of simulations are the design minimum external temperature, internal temperature, U-value of external walls and windows, simple geometrical information, thermal bridges values. Some software allow also the design of the heating plant, so that further information have to be given, as type and positioning of emission system, lay-out of pipes, inlet water temperature, heat generator features, etc. Steady state simulation programs could be useful for calculating the maximum heat demand of a building, and therefore for dimensioning the heating plant. Since the most critical conditions are considered, internal and solar gains are neglected, as well as the dampening effect of the mass. This simplified approach can lead to the over-sizing of the

45

heating system, and to its functioning at a partial power, with reduced efficiency. Anyway, it can be an acceptable approximation if external conditions are not too variable and if the contribution of internal gains are not relevant, like in wintertime, in residential buildings. For the same reason, steady state simulations cannot be used for summer conditions, when external temperature changes considerably from day to night, or for office buildings, where internal gains are relevant. Also for heavy structures or wide glazed rooms, steady state simulations are not the best choice, for the relevant peak-shaving effect due to the mass, and for the appreciable contribution of solar gains, in surfaces.
4.1.2. Transient System Simulations

the case of wide transparent

Transient system simulations consider the behaviour of the system in the time, with variable conditions, both of external parameters and internal factors. Different periods of time can be investigated, from few hours to several years, for highlighting different features of the system. Not only the heat transfer characteristics are taken into account, but also the heat storage capacity, and the complex interaction of each surface with each other component, both external and internal. A more detailed description of the system has to be given as input, from geometrical parameters to thermo-physical properties of materials, from internal gains to solar radiation, from infiltration and ventilation parameters to heat exchange coefficients. Above all, a detailed list of external weather data, as air temperature, solar direct and diffuse radiation, relative humidity, has to be input for each time step. The program can elaborate all inputs and parameters, knowing their relations, and calculate the actual condition of the system, considering its situation at the previous time step. The model has to be re-elaborated by the software for each time step, thus simulating its behaviour upon all the considered period. As far as the heat exchange of a wall is concerned, two approaches have been adopted by two kinds of transient simulation programs: based on geometrical parameters on one hand, and based on transfer functions on the other hand. Geometrically-based software, need precise information about the relative position of each surface, thus view factors can be calculated, for determining long-wave radiation exchange. As far as the

46

other approach is concerned, heat conduction through the wall is determined by transfer functions, while long-wave radiation exchange between the surfaces within the zone and the convective heat flux from the inside surfaces to the zone air are approximated using a star network of thermal resistances. This method uses an artificial temperature node to consider the parallel energy flow from a wall surface by convection to the air node and by radiation to other wall and window elements [1]. Finally, amongst very approximated models, the CARRIER Method [2] should be mentioned, even if it lies between steady state and transient simulations; it can be implemented in a computer or can be used for manual calculations. According to this method, a steady state approach has to be used, but fictitious temperature differences have to be adopted, depending on several factors, such as mass and orientation of the wall, maximum external temperature, daily temperature excursion, latitude, month.
4.2. TRNSYS

TRNSYS is a transient system simulation program with a modular structure, developed under grants from the RANN program of the National Science Foundation, the Energy Research and Development Administration and the U.S. Department of Energy. It recognizes a system description language in which the user specifies the components that constitute the system and the manner in which they are connected. The TRNSYS library includes many of the components commonly found in thermal and electrical energy systems, as well as component routines to handle input of weather data or other timedependent forcing functions and output of simulation results. The modular nature of TRNSYS gives the program tremendous flexibility, and facilitates the addition to the program of mathematical models not included in the standard TRNSYS library. TRNSYS is not only a building simulation software, but it is well suited to detailed analyses of any system whose behaviour is dependent on the passage of time. Main applications include: solar systems (solar thermal and photovoltaic systems), low energy buildings and HVAC systems, renewable energy systems, cogeneration, fuel cells [3].

47

4.2.1. TRNSYS History

In 1975, a graduate student at the Solar Energy Laboratory was working on writing a FORTRAN program to model the performance of a solar house in Colorado. The building designers decided to make a duct work change to the building, a decision that translated into three months of reprogramming for the graduate student. It was decided that, instead of writing a monolithic program to model the entire system, it would have been far simpler to write a series of programs, each of which models a single component. These subroutines had a set of inputs and outputs and could be called from an input file in whatever order the system designer wanted. Changes in the system configuration would then be made by changing the input file, without further modification to the FORTRAN code [1].
4.2.2. TRNSYS Today

The input file in today’s TRNSYS looks quite the same as it did at the very beginning. The file is text based and each component of the system is modelled by a separate subroutine. These subroutines, called Types, have different functions in the system; they are often physical pieces of equipment, (pumps, pipes, thermal zones) but they can also be utility components, such as data readers or radiation processors. Each Type has a certain number of inputs and outputs, but there are two kinds of inputs, variable with time, and constant upon all the simulation; these last are called parameters. To make the use of the software very simple, a graphical interface has been developed (Figure 4.1). The user can drag icons representing the types onto a desktop, and connect them together with a linking tool. If two icons are linked together, the output of one type is considered as the input of the following component. The parameters, instead, have to be specified for each component in a dedicated window. The software then automatically takes the graphical information and writes the text of the TRNSYS input file. As far as building simulation is concerned, a special graphical front allows to simply edit a specific Type, corresponding to a multi-zone building. In this way, many information regarding physical features of the building and its utilisation can be condensed together in

48

one TRNSYS component, allowing anyway its simple modification with a further graphical interface (PreBid), like illustrated in Figure 4.2.

Multi zone Building

Solar Radiation Generator Output

Data reader Psychrometric properties Link

Convective coefficients calculator

Figure 4.1 –IISiBat: TRNSYS graphical interface

Figure 4.2 –PreBid: graphical interface for the multi-zone building Type.

49

4.2.3. Modelling of TABS with TRNSYS

In TRNSYS 15, a modelling of thermally activated slabs is implemented in the Multi-zone Building Type. The slab (Figure 4.3) is transformed in a net of thermal resistances and capacities, with a limited number of nodes. Only geometrical and thermophysical features of the TABS have to be defined as parameters, while the software can automatically calculate the appropriate values of the resistances, arranged in a triangular network (Figure 4.4). The triangular network can be transformed into an equivalent star network (Figure 4.5). Due to this transformation, the information on the pipes can be expressed by one single resistance, Rx. This means that the resistance of each construction element now depends only on its U value. Therefore, despite the multi-dimensional nature of the problem, the thermal transmittance through both halves of the construction element can be calculated by means of the one-dimensional equation for thermal conductance. With some approximations, an expression of Rx can be written:
 d  d x ⋅ ln x   π ⋅δ  Rx ≈ 2 ⋅ π ⋅ λb

(4.1)

Thus, the resistance Rx depends only on two geometric variables, i.e. the distance between pipes dx and the pipe diameter δ, and on the thermal conductivity of the material λb in the pipe plane (Figure 4.3).
T1 y d1 T3 U1 x d2 δ dx U2 h2 T2 Figure 4.3 – Structure of the thermo-active element h1

50

T1 q1 Ra T3 Rc d2 Rb T2 q2 Figure 4.4 – Triangular network of resistances d1

T1 R1 T3 Rx Tk

q1 d1

R2 T2 q2

d2

Figure 4.5 – Star network of resistances

The transformation from triangular to star-shaped network results in temperature in the pipe plane when y=0, and it is called core temperature.

the additional

temperature Tk for the central point of the star network. This temperature equals the mean Moreover, the heat transfer from the fluid within the pipe, at temperature Tw, through the pipe shell to the concrete, at temperature T3, has to be considered. This is called thermal transmittance, and can be separated in two different processes: forced convection within the pipe and thermal conduction through the pipe shell. Two resistances can be defined for the above described heat transfer, respectively Rw and Rr. Finally, must be considered that the water temperature is not constant, but changes along the pipe coil. An opportune thermal resistance Rz can describe the heat losses of the water, passing from inlet temperature Tin to the temperature Tw, in the considered element. All the resistances above described have to be considered in series arrangement; therefore, their value can be summed up in an equivalent total resistance Rt (Figure 4.6): Rt = Rz + Rw + Rr + Rx
Tin Rz Tin Figure 4.6 – Total resistance between water inlet temperature and core temperature Tw Rw Rt Tk Rt T3 Rx Tk

51

The resistances utilized in the model have to be recalculated at each time step, thus the method can be applied to dynamic (time-dependent) calculations. The method have been validated by comparison with FEM (Finite Element Method), and a good correspondence of results has been noticed [1].
4.2.4. Limitations of TABS modelling with TRNSYS

The method implemented in TRNSYS for the modelling of thermally activated slabs has some limitations. The water mass flow rate has to be bigger than 13 kg/hm2, for guarantee the validity of some approximations. Anyway, if lower values of mass flow rate have to be adopted, it is possible to split the pipe coil in n sections, linking the output of one element to the input of the following one. Another limitation in the simulation of wide surface emission systems is that both the adjacent concrete layers have to be thicker than one third of the pipes spacing, so that only massive slabs can be considered, while superficial active layers cannot be modelled. However, if known from other kinds of calculation, it is possible to input the heat emission of the layer, without specifying other parameters as water temperature, mass flow rate or pipes features.

4.3. REFERENCES

[1]

TRNSYS 15, Reference Manual, TRNSYS (a Transient System Simulation Program), Solar Energy Laboratory, University of Wisconsin-Madison, Madison WI53706 USA. BETTANINI E., BRUNELLO P.F., Lezioni di Impianti Tecnici – volume secondo, CLEUP, Padova (I), September 1993. TRNSYS primary internet website, http://sel.me.wisc.edu/trnsys/default.htm

[2] [3]

52

Chapter 5

ZUB BUILDING MODEL
5.1. PREVIOUS MODEL

Several computer models of the ZUB building have been set up, but one in particular has been adopted as starting point for the actual simulation. It is a TRNSYS model, developed by an Italian student at the University of Kassel, Germany (Zecchin P.). In this computer simulation, a portion of the building, from the basement to the second floor, have been considered, as well as a model of coupling with the ground exchanger has been performed. Since the target of the actual simulation is to develop advanced regulation strategies for the TABS, a simplification of the previous model has been necessary, for avoiding too heavy computational efforts, and for a better monitoring of the results. For these reasons, the new model considers only one office room, in an intermediate floor.
5.1.1. Criticism to previous model

A validation of the previous TRNSYS model of the ZUB building, by comparison with measured data, or by comparison with other computer simulations, is still lacking. In the opinion of the author, some building components were not properly modelled, some choices were not the best, and some values were completely wrong. For these reasons, a substantial modification of the model has been performed, and only the characteristics of walls, layers and windows have been maintained exactly as they were.
5.2. NEW TRNSYS MODEL

As already settled above, only one office room, in an intermediate floor, has been considered in the model. The considered thermal zone is supposed to be representative of a normal office in the building, being surrounded on both sides, over and above by almost identical rooms. Therefore, the office has two internal walls facing identical thermal zones (16 m2), an internal wall facing an atrium (17.7 m2), and a big glazed surface facing South

53

(17.7 m2). The dimension of the room are 5.2 m (southern side), 4.7 m, and 3.4 m height, therefore with a floor surface area of 24.44 m2. The ceiling and the floor are 25 cm thick concrete slabs, with pipes embedded at 5 or 8 cm from the lower surface, and then covered by an 8 cm concrete floating floor, separated from the slab by 2 cm of insulation. In the TRNSYS model, constant convective coefficients have been adopted for vertical walls, while for emission surfaces variable convective coefficients are calculated at each time step, as illustrated in Table 5.1.
Table 5.1 – convective coefficients [1] Vertical walls Inside Outside Floor Horizontal slabs (heating) Ceiling 2.5 17.8

h conv = 2.11(Tfloor − Tair )

0.31

W/m2K

h conv = 1.87(Tceiling − Tair )

0.25

During the occupancy of the room, from 8 a.m. to 6 p.m., the presence of one person seated, light working or typing, supplies 75 W of sensible heat and 75 W of latent heat [2], while a PC with monitor (140 W) and artificial lighting (10 W/m2) are the other internal gains. As far as thermal comfort is concerned, a metabolic rate of 1.2 met has been assumed, while the clothing factor is 1.2 clo for the heating period, and 0.6 clo in summer [3]. To this parameters corresponds an optimal value of operative temperature of 20.5 °C in wintertime, and a temperature of 24 °C in summer, as already reported in §§1.5.1. An improvement of the model could consider variable clothing factors based on outdoor air temperature in the morning [4]. A mechanical ventilation system ensures an air change of 0.5 h-1 during the occupancy, while an heat exchanger with an efficiency of 0.8 provides a certain heat recovery. Not any control of inlet air temperature is present, neither humidification in wintertime, nor dehumidification in summertime.

54

5.2.1. Solar radiation and shading devices

The ZUB building is situated in Kassel, Germany, at a latitude of 51.4° North. Direct and diffuse solar radiation on horizontal are values of the test reference year of Trier (D), a city belonging to the same climate zone of Kassel. From these data, a cloudiness factor is calculated to estimate the fictive sky temperature. An external shading device, with a factor of 0.81 (where 0 is no shading), is applied when incident direct radiation exceeds 100 W/m2 (and it is taken away when beam radiation falls beneath 50 W/m2); it is relevant to consider that the windows have a surface of 16.84 m2.
5.2.2. Water inlet temperature and mass flow rate

TRNSYS gives the possibility to define water inlet temperature and mass flow rate as constant values, scheduled parameters or time-dependent inputs (§§4.2.2). The minimum value for the water mass flow rate is limited at 13 kg/hm2 (§§4.2.4), and this means that at least 318 kg/h of water have to feed one thermally activated slab. Therefore, the possibility of regulating continuously the power by mean of the mass flow rate is very limited, even if a substantial improvement can be obtained dividing the slab in smaller portions. The best way to obtain a precise and quick regulation of the emitted power is to control inlet temperature. In the reality, this can be acted adopting a three-way valve, with the partial injection in the circuit of water at fixed temperature. But also this solution presents some problems; for example, during the heating period, the inlet water temperature could be varied between 23 °C and 33 °C, according to the required heating power, while the slab reaches a superficial temperature of 24÷28 °C. If the required power decreases quickly, inlet water temperature is rapidly decreased, and this could be lower than the slab core temperature, leading to an unwanted refrigeration of the mass. A further control has been added to prevent this event, interrupting the mass flow if the water is colder than the slab. When an heating action is required again, the mass flow is immediately activated, but the heat is relayed to the room only after the slab surface increases its temperature. This heat transfer process, from the pipes to the surface, requires a relevant time (the time lag is

1

In some simulations, an external shading device with a shading factor of 0.5 has been adopted.

55

approximately one hour, depending strongly on the position of the pipes). At the end, this results in a considerable delay in the control, and in an imprecise achievement of desired temperature. A relevant improvement has been obtained controlling the temperature difference between outlet and inlet water. According to this approach, the control signal acts on the temperature difference between outlet and inlet water, ∆T = Tin − Tout (for heating; the opposite for cooling), and the inlet water temperature is imposed by this criterion. When no heat is requested, ∆T tends to zero but, in this case, another control stops the mass flow when the temperature difference is less than 0.5 °C, corresponding to a minimum power of 232 W. Controlling continuously ∆T, the slab core temperature is always very close to the water temperature, thus diminishing the time lag for heat conduction, and improving the control of TABS conditions. With other words, this control acts on the power given to the slab (Equation 5.1), rather than on its temperature, that is free to change for compensating heat losses and gains.
& P = m ⋅ c ⋅ ∆T

(5.1)

& Being m (water mass flow rate) and c (specific heat) constant values, the control of the variable ∆T implies the control of the power P. From a theoretical point of view, no boundaries to the temperature are imposed2 but, in the simulation, the temperatures assume always realistic values. For a better simulation of reality, a maximum value of ∆T is adopted, thus limiting the maximum heating and cooling power. The mass flow rate is set to a constant value of 400 kg/h (16.4 kg/hm2), but different strategies have been adopted and, in most of them, water circulation is activated only in certain periods of time, also to limit the energy consumption for pumping (Chapter 7).

Only during the cooling period, inlet water temperature is limited by a minimum value, the dew point temperature, for avoiding moisture condensation. This is a further limit to the maximum refrigeration power.

2

56

5.3. REFERENCES

[1]

TRNSYS 15, Reference Manual, TRNSYS (a Transient System Simulation Program), Solar Energy Laboratory, University of Wisconsin-Madison, Madison WI53706 USA. ISO, “ISO EN 7730-1994, Moderate thermal environments – determination of the PMV and PPD indices and specification of the conditions for thermal comfort”. International Organization for Standardization, Geneva (CH), 1994. BETTANINI E., BRUNELLO P.F., Lezioni di Impianti Tecnici – volume primo, CLEUP, Padova (I), September 1993 (in Italian). BRUNELLO P., DE CARLI M., TONON M., ZECCHIN R. “Aspetti energetici ed economici nel condizionamento con sistemi radianti ad attivazione termica della massa”, AiCARR, Conference, Milan (I) 2002.

[2]

[3] [4]

57

Chapter 6

REGULATION STRATEGIES
6.1. INTRODUCTION

Since the model of the ZUB building is very simplified, the aim of the simulation is not to find out relevant values to be compared to measured data, rather to develop new regulation strategies for an improvement of the TABS behaviour. More in detail, two kinds of problems have to be faced. On one hand, the slowness of the system, due to its high thermal storage capacity, makes impossible a quick temperature variation of the slab, and implies a relevant time delay in the response (§§A2.2.2). For this reason, the control of the heating and cooling system should try to front external and internal changes in advance but, at the same time, it has to spread its action on a long period. On the other hand, the heat storage capacity of the high mass concrete slab has to be exploited, and this involves that temperature changes have not to be always opposed, but should be partially allowed (§3.2). In the following analysis, an attempt to divide the two aspects has been done. At the beginning, all efforts have been focussed on the achieving of perfect required temperature, facing all kinds of loads, and purchasing information on the system response. Only later, the heat storage capacity of the slab has been taken into account, and internal temperature fluctuations have been allowed, within certain limits, to exploit the heat storage possibilities offered by the TABS.
6.2. ACTUAL REGULATION STRATEGIES IN THE ZUB BUILDING

In the ZUB building, each office room has an independent control of the heating and cooling system. A dedicated loop in the hydronic circuit, controlled by a three-way valve and a thermostat in each room allow an autonomous feedback loop. An hand drive gives the possibility to shift the set point temperature (20 °C in wintertime) of ±2 °C, according to the user’s preferences. The thermostat controls the water mass flow, according to a twopoints strategy, as described below, in §§6.3.1. In the heating period, the inlet water

59

temperature depends on external temperature, according to the function illustrated in Figure 6.1. It has to be noticed that, during the heating period, the inlet water temperature is reduced of 4 °C during the night, but the thermostat and the pumps are still working.

Figure 6.1 – Temperature of inlet water, depending on external temperature

During the cooling period, the temperature of the inlet water cannot be controlled, but depends on the conditions of the ground heat exchanger. Anyway, the set point temperature is fixed at 23 °C, and the water circulation is activated when the indoor temperature increases over this value. Must be noticed that, according to thermal comfort matters, the ideal value of operative temperature is 24 °C (§§1.5.1); therefore, cooling is activated in advance, but it is also stopped later, when its action is no longer necessary. In fact, it happens often that the cooling power is not enough to achieve the set value of 23 °C, and a temperature increase has to be tolerated.

60

As far as the air change is concerned, a VOC sensor in each room gives information to the mechanical ventilation system, thus varying the air flow in all the offices at the same time, according to the maximum requested value, maintaining the concentration of organic compounds under 10 PPM (§2.2). Therefore, the air flow is not constant, but it lays usually between 0.3 and 1 air exchanges per hour.
6.3. FEEDBACK CONTROL

Before considering particular regulation techniques, the classical theory of regulations has been taken into account. A general feedback block diagram is illustrated below:
Disturbance V

GV
Reference Value (Set Point) + Sensor

Controller

GC

Process

GP

+

+

Output

GT

Figure 6.2 – Block diagram of a generic feedback control system [1].

As can be seen from the figure above, feedback is obtained by measuring the output signal of a process and comparing it with the reference value. The difference between these two signals, known as error signal, is the input signal to the controller. This last, sends the manipulated signal to affect the process in the right direction. Most building control systems today are feedback control systems, which means that the calculation of the control signal is based on measurements of the controlled output. The most common application is an heating system (Process), controlled by a thermostat (Sensor) that

61

compares the internal temperature (Output) with a set temperature (Reference Value) and gives information (Control) to the system. In this case, Disturbances could be variation of external temperature, solar radiation, internal gains, ventilation, air infiltration. Each factor affects the system with a different transfer function G3, and therefore both with different intensity and time behaviour. The feedback control system has a number of well-known advantages, which explain the extensive use of feedback systems in industrial and building services application today. Two important advantages are that these systems normally give better attenuation of lowfrequency process disturbances and better robustness (less sensitivity to parameter changes) when compared to systems without feedback [1].
6.3.1. Thermostatic Control (On/Off)

One of the most simple controllers for heating systems is a traditional thermostatic control or on/off control [2], whose output function is illustrated in Figure 6.3. This kind of control has turned out to be not very suited for TABS, for their high thermal inertia, that leads to big temperature oscillations. More in detail, the thermostat gives power to the water exactly in the moment in which the thermal zone requires an heat flux. This means that the energy reaches the room only after a considerable time delay, when the slab surface temperature increases its temperature. In this moment, the energy demand of the zone could be completely different, involving a considerable internal temperature change. Only in this last moment, the thermostat can act a new intervention, but its effect will be observed delayed in the time [1]. A sensible improvement can be obtained reducing or annulling the dead band (Figure 6.3), thus anticipating the action of the controller. A certain hysteresis in the heating process is guaranteed by the high thermal capacity of the

The transfer function is a way to describe the dynamic relationship between the input and the output of a system. It is obtained by Laplace-transforming the differential equation between the input and output signals, when all initial values are zero, and rearranging so that Y(s)=G(s)U(s), where G(s) is the transfer function, U(s) are the inputs, Y(s) are the outputs, and s is the Laplace transform variable, as found in the Laplace
∞

3

transform:

∫ f (τ )e
0

−sτ

dτ [1]. More information can be found in §A2.1.

62

system. However, this solution is suitable for its simplicity and, in the heating period, leads to acceptable results (See Chapter 7). WINTER
ON Dead Band = 0

OFF

20

Tset

22

Tint [°C]

Dead Band

Figure 6.3 – Thermostatic Control (Winter functioning)

6.3.2. Proportional Control (PC)

The aim of adopting a Proportional Control is to dampen temperature fluctuation by using the right amount of power, instead of its maximum value. A linear function defines continuously the value of the power to be supplied. The function is proportional to the difference between operative temperature and set temperature (Fig. 6.4).
PC 1 0 20 Dead Band 22 Top [°C]

WINTER

Figure 6.4 – Proportional Control (PC) (Winter functioning)

63

The choice of the dead band influences strongly the results; if its value tends to zero, the proportional control tends to a thermostatic control. A proportional regulation of ∆T turned out to be not suited for high capacity systems, because the action of the controller is very weak at the beginning, while the maximum power is given only when operative temperature is shifted from set temperature, and a further time is requested to lead the system close to the desired condition [2], [3]. This results in a considerable oscillation and non-precise control of operative temperature.
6.3.3. Proportional + Derivative Control (PDC)

An attempt of contrasting the delay in the action of proportional control has been done adding a Derivative Control (DC). This is based on the difference between the operative temperature at the current time and its value at the previous time step (Figure 6.5).

Top DC

Tset

τprevious

τactual

τ

Figure 6.5 – Derivative Control (DC)

For example, in wintertime, if the temperature is going to decrease, DC increases heating, even if operative temperature is above the set point value. This can anticipate the action, before the observed variable reaches undesired values. The DC must be combined

64

with a PC, for guaranteeing that the trend of values tends to reach the set temperature, rather then another value [3]. PDC requires a proper balancing of the proportional effect and the derivative contribution, by adjusting the weighting factors KP and KD (Equation 6.1):
PDC = K P ⋅ PC + K D ⋅ DC

(6.1)

In the ZUB Building simulations, the tuning of the PDC has been done first according to the rules of Ziegler and Nichols for a close loop [2], [3], and afterwards by a manual adjust of values. Those simulations showed a big instability of the system, probably due to the fact that internal temperature is both an input and a response for the regulation loop. For example, if it is going to be colder, the DC control increases heating, and this results in an increment of operative temperature. At the following time step, the sensor measures a temperature rise, and consequently decreases heating again. This leads to a big oscillation of the system, and to its instability. A way to solve the problem could be to calculate exactly the transfer function that links the power to the fluid (as input) with the variation of operative temperature (as output); after that, it is possible to eliminate the effect of a previous action, and valuate the error signal due only to disturbances. This approach resulted in a very difficult implementation, and it has been abandoned. Another attempt to reduce the instability of a PDC has been done implementing an adaptive control, i.e. adopting variable weighting factors in Equation 6.1. This strategy consisted in the reduction of the influence of the derivative contribution, if it was leading to big oscillation of operative temperature. For each time step, the appropriate value of KD (Equation 6.1) has been calculated, observing its effect at the previous time steps. Very heavy computational efforts and poor results suggested to abandon this approach.
6.4. FEED-FORWARD CONTROL

The contrasting of process disturbances can be improved using feedback control loops in combination with feed-forward control based on measurable disturbances (Figure 6.6). Feed-forward makes possible to eliminate a disturbance before it really affects the output signal of a process. For example, if the external temperature decreases, it takes a

65

certain time before it results in a reduction of the indoor temperature. Then, it will need a further time before the system responds to the decreased indoor temperature, hence it might take a long time before the energy balance in the building is recovered and the desired set-temperature is reached again. A feed-forward controller uses information from measurable disturbances to improve the control performance. A feed-forward block diagram (including ordinary feedback) is shown below:
Disturbance V
Sensor

GT2

Controller

GF

GV
+

Reference Value + -

Controller

GC

+ +

Process

GP

+

Output

Sensor

GT1

Figure 6.6 – Block diagram of a generic Feed-forward Control [1].

To eliminate the disturbance V, the following relationship must be fulfilled: V ⋅ GT 2 ⋅ GF ⋅ GP = −V ⋅ GV It means that the feed-forward transfer function GF must be:
GF = −GV ⋅ GT 2 ⋅ GP
−1 −1

(6.1)

(6.2)

Theoretically, it is possible completely to eliminate the disturbance using feed-forward, providing the transfer function GF, calculated by Equation 6.2. In practice, however, the disturbances are often not eliminated, partly because of the fact that the transfer function GV, from the disturbance to the output, may be not exact, and partly because there are other

66

not measurable disturbances which affect the process. Another problem is that the controller transfer function GF sometimes is difficult to implement4. A simplified approach is the adoption of static feed-forward transfer functions; it means that the link between inputs and outputs is no longer a time-dependent function, but a constant. This implies that all the information on the time behaviour of the system are lost, and their average effect is taken into account by a constant [1].
6.5. TEMPERATURE CONTROL SYSTEM IN THE ZUB SIMULATIONS

In the simulations of the ZUB building, several regulation strategies have been adopted, but they all can be resumed in the diagram below: GF3 Forecast radiation GF2 Scheduled On/Off GF1 De-Controller Tset
+ Figure 6.7 – Block diagram of the control system adopted in the simulations of the ZUB building V3 (solar radiation) V2 (internal gains) V1 (external temperature)

GV1/GV2/GV3
+

P-Controller

GC

+ +

Slab

GP

+

Top

The factors that can influence operative temperature are the solar radiation, internal gains (people, computers and lighting) and external temperature. Other disturbances, such as wind speed and direction, are not considered in the model. Other parameters can influence the operative temperature (for example ventilation, infiltration, solar shading,

One condition which must be fulfilled when using a feed-forward link is that the inverse of the transfer function Gp must be stable. By using only the static part of a feed-forward transfer function, the problem of instability of feed-forward transfer function is eliminated [1].

4

67

fictive sky temperature), but they have been defined as functions of the previous factors, or set as constant values. In Appendix 2, a detailed illustration of the effect of each disturbance has been presented; the transfer functions GV1, GV2, GV3 (Figure 6.7), that link the disturbances (as input) to the operative temperature (as output) have not been precisely calculated, but some important observation have been made. Moreover, it has been investigated the effect of heating and cooling the slab on the operative temperature, described by the transfer function GP (Figure 6.7). As far as the feedback loop is concerned, a simple proportional controller has been adopted (GC), but a feed-forward compensator (GF1), based on the variation of external temperature, can improve its functioning, as illustrated in Chapter 7. Internal gains are well known, and present every day in the same period of time; therefore, the strategies can consider this parameter, choosing the right functioning period of the heating/cooling system, for achieving thermal comfort conditions only during occupation time, allowing a certain energy saving, both for heat generation and for pumping. In Figure 6.7, the scheduled pumping activation is represented by the function GF2. Finally, in winter time, the function GF3 acts on the set temperature (Figure 6.7), after having red the forecast values of solar radiation; this is not a proper feed-forward control, because it is not based on measured data, but anyway it can act in advance, before the phenomenon occurs (§8.4). This controller leads both to a considerable energy saving and to better thermal comfort conditions (Table 8.1, Figure 8.13). A more detailed description of the regulation components and strategies is shown in the following paragraphs.
6.6. REFERENCES

[1]

SOLEIMANI-MOHSENI, M., Feed-forward Control and Dynamic Modelling in Temperature Control of Buildings, Chalmers University of Technology, Building Services Engineering, Gothemburg, Sweden 2002, Thesis for Licentiate Eng. BRUNELLI A., “Misure e Controlli”, Manuale dell’ingegnere meccanico, Hoepli, Milano, Italy, 2001, pg. 1924-1938 (in Italian). ZORZINI G., Principi di Regolazione Automatica – Vol. 1°, Cleup, Padova, Italy, 1989, (in Italian).

[2] [3]

68

Chapter 7

FEED FORWARD COMPENSATOR
7.1. DESCRIPTION OF FEED-FORWARD CONTROL

A Feed-forward compensator (§6.4) has been combined with a traditional proportional control (§§6.3.2). A common feedback close loop (§6.3) maintains the indoor temperature close to the set value. A shifting of operative temperature from the desired value is revealed by a sensor positioned inside the room, that controls the heat power emitted to the office, proportionally to the difference between Top and Tset. The proportional control can be described analytically by the following function, during the heating period:

min 1 , k PC ⋅ (Tset − Top ) if Top < Tset  PC =  if Top ≥ Tset 0 

[

]

(7.1)

The choice of the constant kPC influences the amplitude of the dead band (§§6.3.2). This regulation strategy involves a certain shifting of the temperature from the desired value, because the corrective action depends on the observed error. The coupling with a Feed-forward compensator (§6.4) can theoretically eliminate the effect of the disturbance due to the variation of external temperature, before it affects the operative temperature [1]. This approach entails the precise knowledge of the transfer functions that link the slab functioning and the external climate with the indoor temperature. The implementation of these functions resulted in computational efforts and in solving problems, therefore this approach has been abandoned. A static Feed-forward compensator has been adopted; this means that it is taken into account only the variation of external temperature, compared to the previous time step, and the function gives as output only a value, to control the heat emitted to the room. During the heating period, the Feedforward compensator can be described by the following function:

DeC = max{− 1, min 1, k DeC ⋅ (Tj−1 − Tj ) }

[

]

(7.2)

The choice of the constant kDeC required many attempts to calibrate its value, analysing its effect on the behaviour of the system; as a starting point, the real climate data have been

69

analysed, observing the maximum and most common external temperature variations (§§A2.3.1), and adapting the constant kDeC so that the function DeC could oscillate between -1 and 1. It has to be noticed that the function can assume also negative values. This makes sense only if it is considered that the Feed-forward compensator is combined with the proportional controller; therefore, the first can influence the last in both directions, reinforcing or diminishing the action of the proportional control. The final control of the power given to the hydronic circuit of the thermally activated slab, concerning only the proportional control and the Feed-forward compensator based on the variation of external temperature, can be described by the following function:
max{− 1, min[1, (K P ⋅ PC + K De ⋅ DeC)]} if PC > 0 PDeC =  if PC = 0 0

(7.3)

The annulling of the final control when the proportional control is zero is necessary to avoid that the hydronic circuit is working also when an heating or cooling action is not requested. The calibration of the control, by the proper assumption of the two constants KP and KDe required an empirical approach, adjusting their values after having observed the behaviour of the system. A more theoretical approach has been not possible, for the scarcity of literature found on the topic. From this point of view, it would have been more simple to define the control on the valuation of transfer functions, because the setting of a static Feed-forward control is not precisely defined in the traced literature.
7.2. PERFORMANCES OF THE FEED-FORWARD COMPENSATOR 7.2.1. Anticipation of the response of the thermally activated slab

The application of the Feed-forward controller resulted in an anticipation of the control, as can be noticed in Figure 7.1, in the case of summer cooling. The action of the simple proportional control is anticipated of about one hour, compensating the weak action of the slab in this initial period (§§A2.2.2). Moreover, the Feed-forward controller can diminish the action of the traditional control before the indoor temperature reaches the set point. In summertime, this means that the cooling is diminished in the late afternoon, even if internal temperature is over the desired value, because external temperature is going to

70

diminish, and therefore the heat load is going to decrease. More in detail, the anticipation at the beginning of the cooling or heating action can achieve a better compensation of disturbances, while the anticipation of the shut off of the power guarantees a certain energy saving.

PDeC - Summer
0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 -0.1 -0.2 -0.3 PDeC PC DeC Top Tset 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 24 23.5 23 22.5 25 24.5 26 25.5

Time [h]

Figure 7.1 – Anticipation of the control due to the Feed-forward compensator

Figure 7.2 demonstrates that the introduction of the Feed-forward compensator can practically eliminate the effect of the variation of external temperature before this last affects the operative temperature. As a matter of fact, the operative temperature is close to a constant value, therefore also the proportional control is approximately constant; but the power given to the hydronic circuit is strongly variable, according to the function PDeC. This means that the action of the DeC is approximately the opposite of the disturbance due to the variation of external temperature; in fact, in the considered interval of time, no internal gains are present and the solar radiation is very modest.

Top [°C]

71

PDeC - Winter
Temperature [°C]
22.5 22.3 22.1 21.9 21.7 21.5 21.3 21.1 20.9 20.7 20.5 0 Top PC 5 10 DeC 15 PDeC 20 25 30 35 40 45 50 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 -0.1 -0.2 -0.3

Time [h]

Figure 7.2 – Compensation of disturbances operated by Feed-forward control

7.2.2. Comparison between PDeC and Thermostatic Control, in the case of heating

The Feed-forward regulation strategy has been compared with a common On-Off strategy, in the heating case, with the climate data of January, in Kassel. According to the traditional strategy, a thermostat controls the water flow circulating in the hydronic circuit, while the inlet temperature has been fixed to a constant value (30 °C). As far as the PDeC regulation strategy is concerned, the control acts on the temperature difference between inlet and outlet water, thus controlling the power given to the hydronic circuit, while the water mass flow rate is always constant. The heating system is switched on 24 hours a day for both strategies. Figure 7.3 shows the results of the comparison, while in Figure 7.4 a particular is highlighted.

72

Comparison PDeC / Thermostat

T [°C]

23

22.5

22

21.5

21

20.5

20 0
PDeC

100

200

300

400

500

600

700

800

Thermostat

h

Figure 7.3 – Comparison of the Feed-forward regulation strategy with a common thermostatic control

Particular
T [°C]
21.5 21.4 21.3 21.2 21.1 21 20.9 20.8 20.7 20.6 20.5 0 10 20 30 40 50 60

h
PDeC Thermostat

Figure 7.4 – Particular of the comparison between PDeC and On/Off control

73

The peaks that can be noticed in both pictures are due to the effect of solar radiation and internal gains; a precise control of operative temperature is possible only when no relevant gain is affecting the temperature. In Figure 7.4, the first peak is due prevalently to internal gains, while the second one is the effect of a relevant solar radiation. It is also interesting to highlight that, adopting the PDeC strategy, the values of operative temperature are very close to the desired value and aligned with the minimum values obtained with a thermostatic control. In this way, a modest energy saving is possible, only for avoiding useless increases of temperature. In the month of January, a reduction of 1.6% of heat consumption could be possible, if the strategy PDeC is adopted; on the other hand, an increase of the electrical expenses for water circulation must be accepted, for the lower heat powers involved, that imply a longer functioning of the heating system. A detailed comparison of results is reported in Table 7.1. Moreover, the absolutely precise control of the minimum values of operative temperature suggests the possibility of adopting other strategies, as illustrated in §8.4.
Table 7.1 – Comparison of energy demand and pumping expenses5 of the Thermostatic and PDeC regulation strategies; January. Heat [kWh] Thermostat PDeC Difference 368.7 362.7 -1.61% Pmax [kW] 2.4 1.3 -46% Pumping ON [h] 191 461 +150% Heat cost [€] 15.32 15.07 -1.6% Pumping cost [€] 0.10 0.25 +150% Functioning cost [€] 15.42 15.32 -0.6%

To calculate the costs of heat and pumping, in this and in the next Tables, the following parameters have been assumed: electricity cost = 0.08 €/kWh heat cost = 0.04156 €/kWh pressure lost in the pipes = 24 mmH2O/m hydraulic and volumetric efficiency of pumps: ηV ⋅ η id = 0.7 mechanical and electrical efficiency of pumps and electrical motor: η m ⋅ η el = 0.9

5

74

7.2.3. Comparison between constant and scheduled functioning of the heating system

The high thermal inertia of the thermally activated slab does not allow to decrease significantly the temperature of the room during the night, when no occupant is present. Anyway, an attempt to do it has been done, decreasing continuously the temperature of the slab from the afternoon to the following morning. According to this strategy, the water circulation is activated only from 1 a.m. to 12 a.m., but it is usually stopped in the morning, when internal gains reduce the heat demand, and therefore the functioning of the hydronic system is no more necessary. In this way, a modest heat saving is achieved (Table 7.2), for the decrease of the temperature during the night (Figure 7.5) and, moreover, a reduction of the pumping expenses (Table 7.2) derives from the shorter operation time.

Comparison Constant / Scheduled Pumping

T [°C]

23

22.5

22

21.5

21

20.5

20 0 100 200 300 Scheduled Pumping 400 500 600 700 800 Constant Pumping

h

Figure 7.5 – Comparison between constant and intermittent scheduled of the heating system

75

In Figure 7.5, a certain reduction of the operative temperature can be noticed but, if it is considered only the occupancy time, the values of temperatures are approximately the same obtained with the continuous functioning of the system (Figure 7.6).

Comparison Constant / Scheduled Pumping - Only Occupancy Time

Top [°C]

23

22.5

WEEKENDS
22

No internal gains

21.5

21

20.5

20 0 50 100 150 200 250 300 350 Constant Pumping Scheduled Pumping

h

Figure 7.6 – Comparison between constant and scheduled functioning, only in the occupancy period

A certain difference amongst the values obtained with the two strategies can be noticed only during weekends, when no internal gains are present, and in the case of weak solar radiation, as highlighted in Figure 7.6. In Table 7.2, the energy demand and the costs are compared, in the case of constant or scheduled functioning of the heating system. In Table 7.3, instead, PDeC strategy with scheduled pumping is compared with the traditional thermostatic control, working 24 hours a day. Compared with the most traditional solution, higher pumping expenses have to be tolerated, but a certain reduction of energy and costs can be noticed, even if it is not so relevant as the ones obtained with the more advanced regulation strategies illustrated in Chapter 6.

76

Table 7.2 - Energy demand and pumping expenses in the case of Constant and Scheduled pumping; January Heat [kWh] Constant Scheduled Difference 362.7 358.9 -1.0% Pmax [kW] 1.3 1.6 +23.1% Pumping ON [h] 461 282 -38.8% Heat cost [€] 15.07 14.92 -1% Pumping cost [€] 0.25 0.15 -40% Functioning cost [€] 15.32 15.07 -1.6%

Table 7.3 – Energy demand and pumping expenses for Thermostatic control with constant functioning and PDe Control with scheduled pumping; January Heat [kWh] Thermostat + constant pumping PDeC + scheduled pumping Difference 368.7 358.9 -2.7% Pmax [kW] 2.4 1.6 -33.3%% Pumping ON [h] 191 282 +47.6% Heat cost [€] 15.32 14.92 -2.6% Pumping cost [€] 0.10 0.15 +50% Functioning cost [€] 15.42 15.07 -2.3%

7.3. REFERENCES

[1]

SOLEIMANI-MOHSENI, M., Feed-forward Control and Dynamic Modelling in Temperature Control of Buildings, Chalmers University of Technology, Building Services Engineering, Gothemburg, Sweden 2002, Thesis for Licentiate Eng.

77

Chapter 8

REGULATION STRATEGIES FOR THE EXPLOITATION OF THE GREAT THERMAL CAPACITY OF TABS
8.1. HEAT STORAGE CAPACITY OF THE CONCRETE SLAB

A thermally activated slab is first of all an emission system, and the heat that can be emitted in the room depends on the temperature difference amongst the surface of the slab, internal air and other surrounding surfaces. If the heat demand of the room is modified, the superficial temperature of the slab has to change, to emit the right amount of energy necessary to achieve the desired operative temperature. From this point of view, the temperature of the slab is univocally determined, and the power given to the water as well6. But, having the slab a relevant mass, and therefore a considerable thermal capacity, its temperature variation implies a certain heat storage or discharge (§3.2). Just to have an idea of the typical values, in summertime the temperature variation of the slab, during the day, could be about 1°C. This means that the heat first stored and then relayed is:
QSTORED = ρVc∆T = 2400 Wh   175 2  m   kJ kg ⋅ 6.11m 3 ⋅1.05 ⋅1K = 15.397kJ = 4.28kWh 3 m kgK

(8.1)

For example, if this heat is adsorbed during a period of 10 hours, it results in an average cooling power of about 400 W (16.4 W/m2). According to Equation 8.1, during the charging or discharging phase, a determined temperature variation of the slab is necessary, in the right moment. Therefore, being the slab both an emission system and an heat storage device, its two functions could be in contrast. In other words, the energy balance in the room implies a certain temperature variation of the slab, and this results in an heat storage process, but this can occur in the wrong moment, when it is useless or even unbecoming. On the other hand, if the aim is to optimize the charge and discharge of the slab, its
For this reasoning, it is supposed that the temperature of the slab is homogeneously the same in all its volume; several simulations demonstrate that the temperature difference between surface and core of the slab is less than 30% of the usual temperature fluctuation of the core.
6

79

temperature has to be varied, and this results in a modification of operative temperature, that could be not acceptable. A compromise has been found allowing a certain fluctuation of operative temperature, within the limits imposed by comfort condition, so that the percentage of dissatisfied people is less than 10%, according to ISO EN 7730 [1], and as illustrated in §§1.5.1. Different regulation strategies have been adopted for summer, winter and for other climates, and a comparison of their effectiveness is reported in the following paragraphs.
8.2. SUMMER - KASSEL

For the effect of disturbances, the heat to be taken away by the slab is not constant, but has a maximum in the last hours of occupation. For this reason, the slab has to decrease its temperature from the morning to the evening, as shown in Figure 8.1.

Tset constant (24 °C)
Temperature [°C]
24.5 24 23.5 23 1

22.7
22.5

22.5
22 21.5 21 1 2 3 4 5 6 7 T slab 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Occupancy T operative

21.2
0

Time [h]

Figure 8.1 – Summer, Kassel, constant value of set temperature

The temperature decrease of the slab, during the occupation time, is about 1.2 °C, corresponding to a discharged heat of about 5.1 kWh (200 Wh/m2) (Equation 8.1).

80

Occupants

Moreover, if it is considered only the functioning period (5 hours) the temperature decrease of the slab is 1.5 °C; this means that about 6.4 kWh (260 Wh/m2) of heat must be taken away by the fluid circulating in the pipes, in only 5 hours. This implies that an additional power of about 1.3 kW (53 W/m2) has to be guaranteed, just to decrease the temperature of the slab, independently from the heat taken away from the room. From this point of view, the adoption of an heavy slab is not convenient, since it doesn’t work in the proper way, adsorbing exergy during the day, and relaying it during the night. This results in an oversize of the system, and in the functioning of machines with a worse COP7. Anyway, the cooling system is working for a reduced number of hours (Tables A3.1÷A3.24), therefore the effect of an high mass slab should be investigated in all the summer period. As a matter of fact, the high mass slab is surely profitable when the hydronic cooling is not working (Figure 8.2), that is in all the periods when the heat load is not very relevant. In this conditions, the slab takes up some heat during the occupation period, increasing its temperature (from 22.1 to 22.6 °C in Figure 8.2), and it relays it during the night, when a lower external temperature allows free cooling or better COP for the mechanical cooling machines. In this case, it is important that the slab could store as much heat as possible, without increasing too much its temperature, since it affects the operative temperature. With regard to this aspect, an heavy mass is useful to contain the temperature increase of the slab and, in some cases, it can achieve better comfort conditions without using any cooling device (Figure 8.2), [2].

In the ZUB building there are no cooling machine, but the water circulating in the pipes is cooled by a ground heat exchanger. Since with this system the values of inlet temperature depend on the condition of the ground, in the simulations it has been adopted a mechanical cooling, for better investigating the functioning of the thermally activated slab.

7

81

TABS without cooling
Temperature [°C]
25 1

Light SLAB
24.5 24 23.5 23 22.5 22 21.5 21 20.5 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 0

Heavy SLAB

Time [h]
T op - Heavy T slab - Heavy T op - Light T slab - light Occupancy

Figure 8.2 – Comparison between an heavy and a light concrete slab, when cooling is not activated.

The temperature increases of the slabs, during the day, are 0.5 °C for the heavy one, and 2.2 °C for the light one. Referring to Equation 8.1, the heavy slab can take up about 1.86 kWh (76 Wh/m2), while the light one can absorb only 1.2 kWh (49 Wh/m2). This difference can be explained if it is taken into account the bigger temperature difference between surface and room, occurring in the case of heavy slab. Moreover, the little increase of the temperature of the heavy slab, helps to contain the rise of the operative temperature not only taking away some heat from the room, but also contributing to keep low the mean radiant temperature. Of course, the values reported above have to be considered just an indication, since the temperature is not homogeneously constant in the slab but, for example, the surface has a temperature closer to the one of the room. The wrong functioning of the TABS, illustrated in Figure 8.1, suggests to adopt different regulation strategies. As already stated, a certain shift of the value of operative temperature has to be allowed, within the boundaries imposed by thermal comfort

82

Occupancy

conditions. According with ISO EN 7730 [1], the temperature range that implies a percentage of dissatisfied people less than 10% is 24 °C ± 1.7 °C (§§ 1.5.1). For limiting the diurnal decrease of the temperature of the slab, a certain increase of operative temperature has to be allowed, from a minimum value in the morning, to a maximum in the evening. Therefore, new regulation strategies have been investigated, acting directly on the value of the set point temperature. At the same time, the PDeC controller chooses the proper actions to achieve the desired temperature. Three different regulation strategies, S1, S2, S3, (Figure 8.3) have been compared below. It is important to highlight that, for each strategy, the functioning is activated from 3 a.m. to 6 p.m., but water pumping is not running if the required power is too low (§§5.2.2). Therefore, the cooling system works usually only in the afternoon with S1 and S3, while it works prevalently in the early morning with S2.

Regulation strategies
26 1

Temperature [°C]

25.5 25 24.5 24

S2

S3

S1
23.5 23 22.5 22 1 S1 2 3 S2 4 5 S3 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 0

Occupancy

Time [h]

Figure 8.3 – Different regulation strategies for summertime in Kassel.

Occupancy

83

Figure 8.4 demonstrate that the temperature decrease of the slab is limited, or even a modest increase can be noticed. From a theoretical point of view, this should lead to a certain energy saving, since the higher temperature of the slab during the night involves an higher cooling by mean of external air. In practice, instead, the energy consumption is comparable with S1; this can be explained because often the slab cannot increase enough its temperature, after the strong cooling in the morning (Figure 8.4). Or, in other words, the cooling in the morning could be not necessary, if no relevant loads occur during the day8.

Strategy S2
26 1

Temperature [°C]

25 24.5 24 23.5 23 22.5 22 21.5 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 S2 - Tset Occupancy S1 - Top S2 - Top S1 - Tslab S2 - Tslab

22.6

22.9

0

Time [h]

Figure 8.4 – Comparison of S2 with S1.

To solve the problem, strategy S3 (Figure 8.3) has been investigated. In this case, usually the cooling is activated in the last hours of the day, but only if strictly necessary. It means that, allowing a little increase of temperature in the last hours of occupation, within thermal comfort condition, a relevant energy saving is possible. S3 is particularly
8

Since the variable part of disturbances affecting operative temperature are solar radiation and external temperature, a forecasting of their values could be a good parameter for choosing the value of the minimum set temperature in the morning.

84

Occupants

25.5

convenient when little heat loads are affecting the room, because it happens often that operative temperature does not increase over the maximum limit imposed by comfort conditions (Figure 8.5).
Strategy S3
26 1

Temperature [°C]

25.5 25 24.5 24 23.5 23 22.5 22 21.5 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 S3 - Top Occupancy S1 - Top S3 - Tslab S1 - Tslab S3 - Tset 0

23.2

23.7

Time [h]

Figure 8.5 – Comparison of S3 with S1.

The three regulation strategies have been compared in Appendix 3, analyzing different situations, with different solar shading devices and with different internal gains, adopting a night ventilation or an air change only during the occupancy time. The simulations have been run only in the months of July and August, when the heat loads are higher, for better highlighting the effects of the different strategies. Each simulation has been done twice, adopting two different values of the maximum power given to the hydronic circuit. Detailed results are reported in Appendix 4, but some results are resumed below. With the climatic data of Kassel, the adoption of a constant ventilation (both day and night, 0.5 h-1) can achieve an energy saving for cooling of 20÷40%, but the electrical expenses for fans have to be taken into account (§A3.5); also the adoption of an heavier

Occupants

85

solar shading device, with a coefficient of 0.8 (1 = blind) instead of 0.5, involves an energy saving of 20÷60%. With both this conditions, if the office is occupied only by one person, a cooling system is not necessary, since the adoption of an high mass slab (without hydronic circuit) can achieve sufficient comfort condition (PPD