thesis - Pavel Kopecký
Transkript
thesis - Pavel Kopecký
ČESKÉ VYSOKÉ UČENÍ TECHNICKÉ V PRAZE Fakulta stavební Doktorský studijní program: STAVEBNÍ INŽENÝRSTVÍ Studijní obor: Pozemní stavby Ing. Pavel Kopecký HYGRO-THERMAL PERFORMANCE OF EARTH-TO-AIR HEAT EXCHANGERS numerical model, analytical and experimental validation, measurements in-situ, design TEPELNĚ VLHKOSTNÍ CHOVÁNÍ ZEMNÍCH VÝMĚNÍKŮ TEPLA numerický model, analytická a experimentální validace, měření in-situ, navrhování DISERTAČNÍ PRÁCE K ZÍSKÁNÍ AKADEMICKÉHO TITULU Ph.D. Školitel: Prof. Ing. Jan Tywoniak, CSc. Praha, březen 2008 Preface This report documents the work conducted in my study for PhD degree at the Czech Technical University in Prague. My involvement in earth-to-air heat exchangers began in 2003, when my supervisor Jan Tywoniak gave me a chance to participate in the design of ecology centre Sluňákov. The large earth-to-air heat exchanger plays an important role in this fascinating building. The PhD study was financed by several grants. I acknowledge the support of doctoral grant 103/03/H089 and participation within the CIDEAS research centre 1M0579. Without this support, the PhD would have never been conducted. During autumn 2004, I carried out a stay at the Danish Technical University in Lyngby under my supervisor Prof. Stephen Emmitt. This stay was made possible by Socrates Erasmus scholarship and Prof. Emmitt himself. Two courses were carried out at DTU. Mr. Carsten Rode introduced me the mysterious field of numerical methods (Numerical Methods for Building Energy Technology). His personality and the way of thinking was inspiration for me. The course, Ventilation and Climatic Systems, at famous International Centre for Indoor Environment and Energy with course manager Mr. Arsen Melikov gave me a good opportunity to spend nice weeks within the team of international students. I thank my supervisor Prof. Jan Tywoniak for all discussions and an encouragement. I also thank Prof. Petr Hájek who ensured the support of doctoral grant 103/03/H089 and CIDEAS research centre. I thank my colleagues at the department, PhD students from the group of building physics, for their every-day support and discussions. I thank Mr. Martin Jindrák for support and monitoring collaboration. The monitoring on Rychnov family house was made possible by research cooperation between CTU Prague and a company ATREA (http:/www.atrea.cz). I thank Tereza Nováková for some necessary language and text correction. -i- Last but not least, I want to thank my wife Anežka for all the support she has provided me. No doubt, I want to thank her for great help with German language course. Pavel March 2008 - ii - Abstract The thermal comfort requirements and cooling energy demand have significantly increased during the last twenty years. Air heating and/or cooling in earth-to-air heat exchanger is a possible approach for reduction of ventilation heat loss and for improvement of thermal comfort in a building. The earth-to-air heat exchanger is a pipe buried in the ground through which the air is sucked into a building. The general aim of this work is to contribute to the evaluation of hygro-thermal performance of earth-to-air heat exchangers. The overall performance of the earth-to-air heat exchanger may be understood as a bundle of particular components which interact with each other. For instance, the local thermal process (air-to-ground heat transfer near the pipe), the influence of neighbouring pipes, latent heat transfer inside pipes (condensation or evaporation), and thermal coupling with the ground surface can be mentioned. This research on earth-to-air heat exchangers was the combination of a theoretical study and experimental measurements. The theoretical predictions of the hygro-thermal model developed for simulation of the earth-to-air heat exchangers were compared with a) analytical solution for cylindrical heat exchanger and b) in-situ measurements on real-size earth-to-air heat exchangers connected with mechanical ventilation in two low-energy family houses. The model showed perfect agreement with the analytical solution and satisfactory agreement with experimental data. The validated model was then submitted to series of simulations resulting in the straightforward parametric analysis. Moreover, the evaluation of some measured data on real-size earth-to-air heat exchanger collected during 2006 and 2007 was performed. A deeper understanding of heat and moisture dynamics in the earth-to-air heat exchanger is required for the proper design of the earth-to-air heat exchanger. A design methodology was developed in order to facilitate the design of earth-to-air heat exchangers using knowledge gained through both theoretical and experimental work. - iii - The model and long-term monitoring provides information about processes which take place during earth-to-air heat exchanger operation and reveals limitations and energy saving potential of the earth-to-air heat exchanger with respect to building ventilation. Keywords: passive cooling, earth-to-air heat exchanger, hygro-thermal simulation, analytical validation, experimental validation, long-term monitoring, design - iv - Abstrakt V poslední době je snahou snižovat spotřebu energie na provozování budov. Požadavky na tepelný komfort vnitřního prostředí budov naopak stále rostou. Pro předehřev a chlazení čerstvého vzduchu přiváděného do budovy může být v domech větraných pomocí systému nuceného větrání využito zemních výměníků tepla (ZVT). ZVT je potrubí, zakopané v určité hloubce pod povrchem země, přes které je do budovy nasáván čerstvý vzduch. Tato práce je kombinací teoretického zkoumání s měřením na reálných ZVT. Teoretické předpovědi vyvinutého modelu ZVT byly porovnávány s a) analytickým řešením a b) měřením in-situ na dvou skutečných ZVT. Takto validovaný model byl potom podroben parametrické analýze. Částí této práce je i vyhodnocení měřených dat z reálného ZVT (zejména měřená data z roku 2006) a vývoj metodiky pro návrh dimenzí ZVT. Porovnání mezi měřením a simulací ZVT sice neukázalo absolutní shodu, avšak matematický model a později i dlouhodobé sledování reálných ZVT poskytují jasnou informaci o procesech vyskytujících se během provozování ZVT a odkrývají tak potenciál využití ZVT ve spojení s větráním budov. Klíčová slova: zemní výměník tepla, tepelně vlhkostní simulace, analytická validace, experimentální validace, monitoring -v- Table of contents Preface ...............................................................................................................................i Abstract .......................................................................................................................... iii Abstrakt............................................................................................................................v Table of contents.............................................................................................................vi Nomenclature..................................................................................................................xi 1 2 Introduction ............................................................................................................1 1.1 General context ...............................................................................................1 1.2 Earth-to-air heat exchanger .............................................................................3 1.3 Literature review .............................................................................................5 1.4 Objectives........................................................................................................7 1.5 Scientific methods...........................................................................................8 1.6 Outline of the work .........................................................................................9 Model for simultaneous heat and moisture transfer in EAHX........................11 2.1 Fundamentals ................................................................................................12 2.1.1 Heat transfer ......................................................................................12 Conduction ......................................................................................................... 12 Convection.......................................................................................................... 13 2.1.2 Moisture transfer ...............................................................................14 Diffusion............................................................................................................. 15 Convection.......................................................................................................... 16 2.1.3 Direct link between heat and moisture transfer.................................16 Generation of latent heat..................................................................................... 17 2.1.4 Analytical solutions for semi-infinite body.......................................17 Harmonic oscillation of surface temperature...................................................... 17 Approximate calculation of soil thermal diffusivity........................................... 18 2.1.5 2.2 Natural thermal regime of shallow subsurface..................................19 Theoretical analysis.......................................................................................20 2.2.1 Problem definition.............................................................................20 - vi - 2.2.2 Heat transfer...................................................................................... 21 Convective heat transfer coefficient................................................................... 22 2.3 2.2.3 Moisture transfer............................................................................... 24 2.2.4 Coupled heat and moisture transfer .................................................. 26 Numerical solution........................................................................................ 26 2.3.1 Stage of operation ............................................................................. 26 Calculation of new air temperature (step 1a)...................................................... 27 Calculation of new water vapour concentration and moisture flux (step 1b) ..... 27 Calculation of new temperature in the soil (step 2)............................................ 28 Calculation of new pipe surface temperature (step 3) ........................................ 31 2.4 3 2.3.2 Stage of natural soil recovery ........................................................... 31 2.3.3 Flow diagram .................................................................................... 32 2.3.4 Stability and accuracy....................................................................... 33 Summary....................................................................................................... 35 Model validation .................................................................................................. 37 3.1 3.2 Flat-plate heat exchanger.............................................................................. 37 3.1.1 Problem definition ............................................................................ 37 3.1.2 Numerical solution............................................................................ 38 3.1.3 Test of longitudinal heat conduction ................................................ 40 3.1.4 Phase-shifting phenomenon.............................................................. 40 3.1.5 Summary........................................................................................... 45 Verification exercises ................................................................................... 45 3.2.1 Test of undisturbed soil temperature calculation.............................. 45 3.2.2 Test of hygrothermal calculations..................................................... 49 Step change of inlet relative humidity................................................................ 50 Periodic inlet signals .......................................................................................... 51 3.3 3.4 Analytical validation..................................................................................... 55 3.3.1 Analytical solution............................................................................ 55 3.3.2 Simulation......................................................................................... 55 3.3.3 Comparison analytical vs. numerical................................................ 58 3.3.4 Summary........................................................................................... 60 Experimental validation................................................................................ 60 3.4.1 Long-term thermal simulation .......................................................... 60 Measurement in-situ........................................................................................... 60 Simulation .......................................................................................................... 61 Sensitivity analysis............................................................................................. 62 Comparison simulation vs. measurement........................................................... 65 - vii - Continuous mode................................................................................................ 67 Intermittent mode ............................................................................................... 67 Conclusions ........................................................................................................ 68 Summary ............................................................................................................ 69 3.4.2 Short-term hygro-thermal simulation................................................70 Measurement in-situ ........................................................................................... 70 Simulation .......................................................................................................... 70 Comparison simulation vs. measurement ........................................................... 70 Conclusions ........................................................................................................ 73 Summary ............................................................................................................ 73 3.5 3.6 4 Parametric analysis........................................................................................73 3.5.1 Simulation .........................................................................................73 3.5.2 Input parameters................................................................................75 3.5.3 Monitored outputs .............................................................................75 3.5.4 Results ...............................................................................................76 3.5.5 Conclusions .......................................................................................78 Summary .......................................................................................................78 Measurements in situ ...........................................................................................81 4.1 Introduction ...................................................................................................81 4.2 Soil ................................................................................................................83 4.3 4.4 4.2.1 Thermal properties ............................................................................83 4.2.2 Soil temperature ................................................................................84 4.2.3 Approximate calculation of soil thermal diffusivity .........................84 External environment ....................................................................................85 4.3.1 Ambient air temperature....................................................................85 4.3.2 Ambient air water vapour concentration...........................................86 4.3.3 Tendency of EAHX to condensation ................................................87 Earth-to-air heat exchanger ...........................................................................88 4.4.1 Example of measured data ................................................................88 4.4.2 EAHX operation................................................................................90 Pre-heating mode................................................................................................ 90 Cooling mode ..................................................................................................... 91 Annual statistic ................................................................................................... 91 4.5 4.4.3 Outlet air temperature........................................................................92 4.4.4 Moisture balance ...............................................................................93 4.4.5 Energy performance ..........................................................................94 Conclusions ...................................................................................................97 - viii - 4.6 5 Dimensioning of EAHXs ..................................................................................... 99 5.1 5.2 6 Summary....................................................................................................... 97 Theory........................................................................................................... 99 5.1.1 Outlet air temperature ....................................................................... 99 5.1.2 Temperature efficiency ................................................................... 100 5.1.3 Pressure loss.................................................................................... 101 5.1.4 Cooling power................................................................................. 102 5.1.5 Hollmuller design rules................................................................... 103 Simulations ................................................................................................. 103 5.2.1 Hollmuller design rules................................................................... 104 5.2.2 Constant NTU ................................................................................. 104 5.2.3 Constant Va/A.................................................................................. 105 5.2.4 Distance between pipes................................................................... 106 5.2.5 Role of pipe material....................................................................... 106 5.3 Design methodology................................................................................... 107 5.4 Examples..................................................................................................... 109 5.4.1 EAHX for family house.................................................................. 109 5.4.2 EAHX for larger building ............................................................... 110 5.5 Conclusions................................................................................................. 110 5.6 Summary..................................................................................................... 111 Conclusions......................................................................................................... 113 6.1 Results......................................................................................................... 113 6.2 Final remarks and recommendations for further research .......................... 114 Appendices................................................................................................................... 117 A1 Derivations and solutions ........................................................................... 117 Equation 2.30 ................................................................................................... 117 Equation 2.39 ................................................................................................... 118 FPHX with adiabatic boundary – analytical solution ....................................... 118 A2 Explicit finite difference method for transient heat conduction ................. 120 A3 House W ..................................................................................................... 122 Basic information ............................................................................................. 122 List of sensors .................................................................................................. 123 A4 Passive family house in Rychnov ............................................................... 124 Basic information ............................................................................................. 124 - ix - List of sensors................................................................................................... 125 Ventilation modes............................................................................................. 126 Air flow rates.................................................................................................... 127 Auxiliary energy............................................................................................... 127 A5 Thermal properties of soils and rocks .........................................................128 References ....................................................................................................................130 Publications written in context with the thesis .........................................................135 Other publications.......................................................................................................137 -x- Nomenclature The following list of symbols is not complete. However, all the symbols used within the thesis are explained in the place where they appear so that their meaning should be already clear from the text. Abbreviation EAHX, appearing throughout the following text, denotes earth-to-air heat exchanger. Similarly, FPHX denotes flat-plate heat exchanger. symbol latin letters as bs cp deq dp eAeahx eVaAeahx gv ha l ne va q rh t tp x, y, z explanation unit thermal diffusivity of soil thermal effusivity of soil specific isobaric thermal capacity equivalent diameter periodic penetration depth specific energy injected to surrounding soil related to exchange surface A specific energy injected/extracted to/from surrounding soil (related to exchange surface A and air flow rate Va) vapor flux convective heat transfer coefficient latent heat of condensation/evaporation air change rate air velocity heat flux relative humidity time period spatial coordinates [m2/s] [Ws0.5/(m2.K)] [J/(kg.K)] [m] [m/period] [Wh/(m2.period)] - xi - [(Wh/period)/(m2.m3/period) = Wh/m5] [kg/(m2.s)] [W/(m2.K)] [J/kg] [1/h] [m/s] [W/m2] [%] [s] [s] [m] r z A B Dchar E L Nu NTU Pr Q QEAHX Qcooling G Va W radial coordinate depth area, exchange surface thickness of the plate characteristic dimension energy length of EAHX Nusselt number number of transfer units Prandtl number heat flow heat flow injected/extracted to/from soil by EAHX cooling power of EAHX convective moisture flow volumetric air flow rate width of the plate for FPHX [m] [m] [m2] [m] [m] [kWh/period] [m] [-] [-] [-] [W] [W] thermal conductivity thermal conductivity of pipe material water vapour transfer coefficient (vapor concentration as driving potential) temperature density partial density of water vapour (vapor concentration) temperature efficiency of EAHX friction pressure loss in a straight pipe [W/(m.K)] [W/(m.K)] [W] [kg/s] [m3/s] [m] greek letters λ λp βρ θ ρ ρv ηEAHX ∆pfric subscripts a conv in out rec s air, ambient convection inlet outlet recovery soil or surface - xii - [m/s] [°C] [kg/m3] [kg/m3] [-] [Pa] sat v i i, j, k superscripts a dry moist ol cl saturation vapour internal indexes (counters) amplitude simulation with absolutely dry air simulation with moist air open loop of EAHX (direct suction of ambient air) closed loop of EAHX (circulation between internal environment and EAHX) - xiii - Introduction 1 Introduction The chapter will focus on the background of earth-to-air heat exchangers and a basic literature review. Furthermore, the emphasis is put on the introduction of thesis objectives and the scientific methods used in the thesis. 1.1 General context We live in time when global warming becomes more and more apparent (see Figure 1.1). Most of the observed increase in global average temperatures since the mid-20th century is very likely caused by observed increase in anthropogenic greenhouse gas concentrations (IPCC report, 2007). Consequently, and amongst others, longer and more intense heat wave periods and heavy precipitation events will very likely occur in the 21st century. European countries could already notice extremely hot summer in 2003 and 2006 (see Figure 1.2), and extraordinarily mild autumn 2006 (see Oldenborgh, 2006). The amount of any fossil fuel under the ground worldwide is finite. Therefore, fossil fuels will be more expensive in the future due to their successive depletion. For instance, (Cílek and Kašík, 2007) report that the worldwide maximum in oil production (peak oil) will likely occur between 2008 and 2015. Thus, problems with oil availability and related price increase should be already expected in the following decade. In developed countries, approximately 40 % of total primary energy is consumed in buildings and approximately 25 % of the green house gas emissions (CO2 in particular) are building-related. Therefore, a future strategy focuses on: a) low energy and passive buildings (energy efficiency), and b) use of renewable energy sources for heating and cooling of buildings and electricity production. The technical measures are already available on the market. But more is required than pure increase in energy efficiency of buildings. The future development will have to focus on zero-energy buildings (Voss and Kramp, 2007), and the energy-independent buildings. -1- Chapter 1 40 40 fourier (N=5) 35 35 30 30 25 25 frequency frequency fourier (N=5) 20 15 20 15 10 10 5 5 0 1800 1850 1900 year 1950 2000 0 1800 1850 1900 year 1950 2000 Figure 1.1: Left - absolute frequency of daily ambient air temperature maximums sorted by the year of their record. Right – absolute frequency of daily ambient air temperature minimums sorted by the year of their record. Data from Prague Klementinum (1775 2006) 2001 1991 2001 1991 1981 1971 1961 1951 1941 1931 1921 1911 1901 0 1981 5 1971 10 1961 15 1951 20 5th degree polynomial 1941 25 90 80 70 60 50 40 30 20 10 0 1931 30 2006 1921 2003 1911 5th degree polynomial 1901 35 Figure 1.2: Left - number of tropical days (θambient > 30 °C). Right – number of summer days (θambient > 25 °C). Data from Prague Klementinum (1900 – 2006) The thermal comfort requirements and cooling energy demand have significantly increased during the last twenty years. Production of coldness by a “traditional” compressor cooling consumes a lot of electricity. Countries with high penetration of airconditioning, such as the USA or Canada, have serious problems with very high demand during summer temperature peaks. Therefore, the substantial reduction of heat gains (prevention and protection) and short-term storage of heat gains in building components (modulation of heat gains) should be considered to be important issues of building-energy concept. Alternative (often called passive) cooling techniques could be used for heat rejection, if necessary. Ventilation by cold night air and using coldness of ground (earth-to-air or earth-to-fluid heat exchangers) are standard examples of passive cooling techniques. Some promising cooling techniques are still under scientific investigation, see e.g. (Hollmuller et al, 2006) and (Arkar and Medved, 2007). -2- Introduction 1.2 Earth-to-air heat exchanger Air heating and/or cooling in earth-to-air heat exchanger is a possible approach for reduction of ventilation heat loss and for improvement of thermal comfort in a building. The earth-to-air heat exchanger is a pipe buried in the ground through which the air is sucked into a building. Such device is considered to be both renewable energy source and a technique for reduction of building heat use and cooling energy demand. Using soil for heat dissipation is a traditional principle. Because of high soil thermal capacity, the temperature in sufficient depth under upper soil surface remains stable, oscillating roughly around annual mean ambient temperature. Hence, the ambient air sucked through the pipe is either cooled down or heated up. The principle had already been known by ancient Persian and Greeks (Santamouris et al, 1996). Nowadays, with rising energy costs and higher requirements on thermal comfort, the utilization of earth-to-air heat exchangers is slowly coming back. Its simplicity, high cooling potential, low initial, operational and maintenance costs are presented as the advantages of the earth-to-air heat exchanger. In many cases, well designed earth-to-air heat exchanger together with efficient overall concept of a building may eliminate the need for mechanical air conditioning (Pfafferot, 2004). The setup of a simple earth-to air heat exchanger is depicted in Figure 1.3. The system typically consists of an inlet shaft with filters and a pipe buried in the ground. The pipe is connected to a mechanical ventilation system using T-fitting with damper. The damper controls whether air is sucked through the earth-to-air heat exchanger or is directly sucked through a facade inlet. Ventilation system Figure 1.3: EAHX linked to a building (open loop mode) -3- Chapter 1 An alternative setup of the earth-to-air heat exchanger has been developed by the Czech company ATREA (http://www.atrea.cz). The layout (Figure 1.4) allows air circulation between building and the earth-to-air heat exchanger (closed loop system). In order to switch over operational modes the system needs a special damper placed in a chamber somewhere near an air handling unit. Ventilation system Figure 1.4: Alternative setup of EAHX (closed loop mode) What is the current state-of-the-art in the Czech Republic? The earth-to-air heat exchanger has become quite popular when small one-pipe exchangers are widely used in low-energy family houses. Even multi-pipe exchangers have been built, e.g. kindergarten in Ostrava Proskovice (Kramoliš, 2002) or Ecological Education Centre Sluňákov near Olomouc (Hofmeister et al, 2007). However, such systems are still rare and perceived as partially experimental. Figure 1.5: Left – EAHX in Ostrava Proskovice. Right – EAHX Sluňákov near Olomouc -4- Introduction 1.3 Literature review Heat conduction in the ground has traditionally been a domain of researchers from the northern European countries. Because of the cold climate, they are more attracted by the field of foundation design, soil freezing, ground heat storage, and air pre-heating in earth-to-air heat exchangers. Researchers of the southern European countries often examine cooling potential of earth-to-air heat exchangers. The most recent research perhaps comes from Switzerland and Germany. The valuable information is often found in books or papers dealing with ground heat storage (e.g. Hellstrom, 1991) or thermo-active components (e.g. floor heating, radiant cooling ceilings). The references dealing with earth-to-air heat exchangers may be further divided into: Modelling issue (analytical, numerical or mixed models) – (Claesson, 1983), (Nilsson, 1991), (Mihalakakou et al, 1994), (Stahl, 2002), (Hollmuller, 2005a), (Hollmuller, 2003), (Hollmuller, 1998) Comparative analysis between predicted and measured data – (Nilsson, 2001), (Hollmuller, 2001), (Mihalakakou et al, 1994) Evaluation, design, and operational experience of real examples – (Pfafferott, 2003), (De Paepe, 2003), (Herkel, 2002), (Fink et. al, 2002) One of the first investigations of heat extraction from the ground was launched by (Claesson, 1983). Mathematically focused publication employed analytical solutions for heat conduction equation with different boundary conditions. The analysis was based on the technique of superposition, when the complex thermal process in the ground was considered to be superposition of the elementary ones. In his doctoral thesis, (Nilsson, 1991) dealt with pre-heating of ambient air by a system of earth-tubes as primary energy source for a heat pump. His work also employed an analytical approach and special attention was paid to latent heat effects in the soil (soil freezing). The theoretical analysis was compared with in-situ measurement of pre-heating system located in northern Sweden. (Mihalakakou et al, 1994) introduced a numerical model of the EAHX. The model was compared with an experimental setup of EAHX. (Mihalakakou et al, 1996) used the model to describe the thermal influence of design variables – pipe length, pipe -5- Chapter 1 radius, air flow rate and pipe depth. In (Mihalakakou, 2003), a neural network was used for to estimate the thermal performance of a single pipe exchanger. Another neural network approach was discussed in (Kumar, 2006). (Santamouris, 1995) developed a method for estimation of the EAHX contribution to the reduction of building cooling load. (Stahl, 2002) shortly discussed the moisture consequences of EAHX operation. High relative humidity level of outlet air during summer operation was shown; it was further stated that high air humidity near the pipe outlet might result in mould growth inside the pipe. In his doctoral thesis, (Hollmuller, 2002) dealt with earth-to-air heat exchanger. He developed a numerical simulation model, and analyzed the monitored data from real scale installations. (Hollmuller, 2003) introduced an analytical solution for cylindrical air-to-soil heat-exchanger with either external adiabatic or isothermal boundary. In parallel, it was mentioned that a special setup of EAHX (special combination of geometry, exchanger operational parameters, and periodical input signal) will lead to the outlet phase-shifting almost without amplitude-damping. The guideline for the combination of inputs in order to reach this performance was also proposed. Heating as well as cooling potential of one commercial building was evaluated by (Hollmuller, 2001) by means of his numerical simulation model and the evaluation of the monitored data. (Pfafferott, 2003) evaluated three EAHXs for office buildings. A general method for comparison of EAHXs in operation was introduced. (De Paepe and Janssens, 2003) developed a simple design tool of EAHX which was based on thermo-hydraulic performance optimization. Even though many publications are available in this research field, other research topics which have not been well explored, yet, can be deduced: The moisture consequences: Except for Hollmuller’s numerical model, existing mathematical models don not take humidity of air into account. Are really latent heat effects inside pipes so important (e.g. unwanted heating due to the release of condensation heat, positive effect of the water vapour evaporation, consequent mould growth in pipes)?. -6- Introduction A comparative analysis between simulation and measured data: (Hollmuller, 2002) made careful analysis between monitored data of real scale exchangers and his mathematical model. It was stated that water infiltration due to leaky pipe wall can significantly affect the air heat balance. The model developed by (Mihalakakou, 1994) was compared to experimental short-term measurement. Still, there is a lack of measured data, especially monitoring of real size installations would be a valuable contribution. The evaluation of EAHX’s total effect with regard to the ventilated zone: (Pfafferott, 2003) evaluated the operation of real size earth-to-air heat exchangers, but the contribution of the exchanger to overall building energy balance was not discussed. The contribution to the overall building energy balance was discussed by (Hollmuller, 2001), who pointed out the fundamental difference between preheating and pre-cooling mode of the EAHX. The air pre-heating lowers building energy demand; however, there is competitive relationship of EAHX with heat recovery because the EAHX “steals” energy which could otherwise be recovered. Hence, the total effect of the EAHX in series with heat recovery is not significantly higher than the effect of heat recovery itself, especially for very high efficiency of heat recovery (assuming very tight building envelope). The air precooling may dampen the daily temperature peaks below the comfort threshold. However, the amount of heat which is removed from the building is strongly dependent on the air-flow rate and current soil temperature. The comparison between closed loop mode and open loop mode of EAHX has not been performed, yet. Is a closed loop setup more effective for cooling than open loop system? What is the difference between closed loop and open loop mode from moisture point of view? 1.4 Objectives Development of hygro-thermal model for simulation of the earth-to-air heat exchangers is the main objective of this work. It is believed that through theoretical work the main tendencies in heat and moisture transfer in the earth-to-air heat exchanger will be explored. However, reality is always much more complicated than any complex model could be. Therefore, a reasonable model complexity should be kept in mind. -7- Chapter 1 Since the model is a pure mathematical representation (i.e. simplification) of the physical reality, validation and/or verification of the model will be another objective of this work. Since experimental validation is conditioned by good quality of experimental data, long-term monitoring of real-size earth-to-air heat exchanger and evaluation of the measured data will be the next part of this work. A complex modeling is not always required to answer simple question. Therefore, a method for design of EAHX dimensions should be developed. The result of this part should make possible to design EAHX quickly and easily without loosing a physical background. 1.5 Scientific methods The research method was a combination of a theoretical study and the experimental measurements performed on two real size EAHXs. The theoretical chapters used logical methods (mainly abstraction, analysis and synthesis) which complemented each other. The analysis was used for development of the simulation model for earth-to-air heat exchanger when complex problem of simultaneous heat and moisture transfer was divided into particular components. On the other hand, the synthesis put the components together when interrelations between components were studied. The model was developed by a step-by-step procedure when the first version of the algorithm included only heat transfer and was tested on simpler case of the flat-plate heat exchanger. Subsequently, the basic routine (dealing with heat transfer only) for the earth-to-air heat exchanger was developed adjusting the model of the flat-plate heat exchanger (pipe, two-dimensional calculation of heat conduction, boundary nodes). The moisture transfer and influence of latent heat on air balance were added later. The long-term monitoring was performed on two low-energy family houses ventilated by mechanical ventilation equipped with heat recovery and real size earth-toair heat exchangers. Monitoring on real size systems has two important advantages. First, the simulation can be compared with measurement so that the relation between reality and the model can be studied. Second, the measurement provides authentic information about EAHX real operation and about link between house ventilation system and the earth-to-air heat exchanger. -8- Introduction The process of thesis development may be represented via the following flow diagram (Figure 1.6). Theoretical analysis of physical processes in EAHX The development of a mathematical model for hygro-thermal simulation of the earth-to-air heat exchangers + validation The measurement on real size exchangers, data evaluation, measurement vs. simulation Figure 1.6: The thesis - flow diagram 1.6 Outline of the work Chapter 1: The chapter focuses on the background of earth-to-air heat exchangers in the Czech context and a basic literature review (state-of-the-art). Furthermore, the emphasis is put on the introduction of thesis objectives and the scientific methods used in the thesis. Chapter 2: The hygro-thermal model of the earth-to-air heat exchanger will be described in this chapter. The emphasis is put on the introduction of physical processes which may take place in earth-to-air heat exchangers and the presentation of developed model. Chapter 3: The developed hygro-thermal model of the earth-to-air heat exchanger will be validated in this chapter. An analytical validation and experimental validation are introduced here. Finally, the hygro-thermal performance of the validated model is evaluated by a parametric analysis. Chapter 4: The results from long-term monitoring on passive family house in Rychnov will be introduced in this chapter. The presentation will focus on the EAHX performance. Chapter 5: The chapter will focus on the development of a simple method for design of optimal dimensions of EAHXs. Chapter 2, 3, 4, 5 are considered to be the body of the work. -9- Chapter 1 - 10 - Model for simultaneous heat and moisture transfer in EAHX 2 Model for simultaneous heat and moisture transfer in EAHX The hygro-thermal model of the earth-to-air heat exchanger will be described in this chapter. The emphasis is put on the introduction of physical processes which may take place in earth-to-air heat exchangers and the presentation of developed model. The overall performance of the earth-to-air heat exchanger may be understood as a bundle of particular components which interact with each other (Figure 2.1). For instance, the local thermal process (air-to-ground heat transfer near the pipe), the influence of neighbouring pipes, latent heat transfer inside pipes (condensation or evaporation), and coupling with the ground surface (long-term behaviour of the exchanger) can be mentioned. Another complication of the computational analysis is the time scale of thermal process; even very short extraction or injection pulses take place during the operation of the exchanger (see chapter 4.4.2). short-wave radiation rain wind ambient temperature long-wave radiation cond./evap. heat λs, as conduction cond./evap. in the pipe, latent heat of phase change ground water table Figure 2.1: Boundary conditions for earth-to-air heat exchanger - 11 - Chapter 2 2.1 Fundamentals This part will deal with several heat and mass transfer topics. The emphasis is placed on the description of mathematical formulae. The sub-chapter focuses on the physical processes taking place in the earth-to-air heat exchanger. The main references for this introductory theoretical chapter are (Hagentoft, 2001) and (Hens, 2007). 2.1.1 Heat transfer Heat transfer is the process by which energy is transported as a result of temperature difference. The heat can be transferred by following mechanisms: • Conduction • Convection • Radiation The heat conduction is the most important transport process in solids, radiation or convection are usually minor. On the contrary, convection and radiation are the most important transport processes in gases and liquids, e.g. air gap embedded in solid material. Conduction The heat flux q [W/m2] in homogenous and isotropic material is expressed by Fourier’s law which says that the heat flux is proportional to the gradient of temperature and is flowing in the opposite direction of gradient temperature vector: q = −λ ∂θ ∂x (2.1) where: λ is thermal conductivity of material [W/(m.K)], θ is temperature [K]. The energy conservation law says that energy can not disappear. This general law is used to describe the thermal balance of a control volume. The inflow of heat [W/m3] minus outflow [W/m3] has to be equal to heat stored [W/m3] in the control volume per time step: − ∂q ∂h =ρ ∂x ∂t (2.2) - 12 - Model for simultaneous heat and moisture transfer in EAHX where: ρ is density of material [kg/m3], h is specific enthalpy [J/kg]. By inserting transport equation (2.1) to equation (2.2) is obtained: − ∂ ⎛ ∂θ ⎞ ∂h ⎜ −λ ⎟=ρ ∂x ⎝ ∂x ⎠ ∂t (2.3) The relationship between change of specific enthalpy h (J/kg) and temperature change in a certain material is: ∂h = c p ∂θ (2.4) where: cp is specific thermal capacity [J/(kg.K)]. By inserting the equation of state (2.4) to continuity equation (2.3), transient heat conduction equation is obtained: − ∂ ⎛ ∂θ ⎞ ∂θ ⎜ −λ ⎟ = ρc p ∂x ⎝ ∂x ⎠ ∂t (2.5) In case of constant λ and ρcp equation (2.5) can be rearranged into the following form: ∂θ ∂ 2θ =λ 2 ∂t ∂x ρc p (2.6) Thermal diffusivity a (m2/s) is the ratio: a= λ ρc p (2.7) The equation for heat conduction in two dimensions can be written as: ρc p ⎛ ∂ 2θ ∂ 2θ ∂θ = λ⎜ 2 + 2 ⎜ ∂x ∂t ∂y ⎝ ⎞ ⎟⎟ ⎠ (2.8) Convection The convective heat transfer is caused by fluid flow. The convective heat flow Qconv [W] is: Qconv = ma caθ a (2.9) where: ma is air (fluid) flow rate [kg/s], ca is specific thermal capacity of air (fluid) [J/(kg.K)] and θa is air (fluid) temperature [K]. If fluid flows along a surface and the difference between fluid and surface temperature exists, the convective heat transfer will occur. The convective heat flux qconv (W/m2) is expressed as a function of the - 13 - Chapter 2 convective heat transfer coefficient ha [W/(m2.K)] and the difference between fluid and surface temperature: qconv = ha (θ a − θ s ) (2.10) Convection caused by a pump or a fan is called forced convection. Convection caused by density differences in the fluid is called natural convection. The flow in pipes can be classified as either laminar or turbulent. No mixing of the fluid occurs in the laminar flow. The flow velocity and direction of flow in a certain point remains constant over time. In the turbulent flow the fluid is completely mixed, the flow velocity and direction of flow in a certain point is not constant over time. Due to the mixing, the temperature in the channel becomes almost uniform. Reynold’s number may characterize the flow: in straight pipes, at Re ≤ 2300 the flow is considered to be laminar and at Re ≥ 10000 the flow is considered to be turbulent. Between 2300 and 10000 the flow is in transition state (neither laminar nor turbulent). 2.1.2 Moisture transfer The moisture transfer is the process by which mass (vapour or water) is transported as a result of difference in vapour concentrations (or partial vapour pressures). The moisture can be transferred in porous materials by the following mechanisms: • Diffusion • Convection • Capillary suction • Special types (thermo-diffusion) For the earth-to-air heat exchanger, all types of moisture transfer are important. The convective vapour transfer occurs between air and inner surface of pipes. Water vapour is transferred by diffusion through the wall of pipes. Convection and capillary suction may take place in the surrounding soil. - 14 - Model for simultaneous heat and moisture transfer in EAHX Diffusion The water vapour flux gv [kg/m2s] is expressed by Fick’s law which says that vapour flux is proportional to the gradient of a driving potential and flows in the opposite direction of gradient vector. Either water vapor concentration or partial pressure of water vapor can be chosen as the driving potential: g v = −δ ρ ∂ρv ∂x (2.11) g v = −δ p ∂pv ∂x (2.12) where: δρ is water vapour permeability in [m2/s] (when ρv is driving potential), δp is water vapour permeability in [kg/(m.s.Pa)], (when pv is driving potential), ρv is water vapour concentration [kg/m3], pv is partial pressure of water vapour [Pa]. The equation of state for water vapour expresses the relationship between partial pressure and water vapour concentration under isothermal conditions: pv = Rv ρvθ (2.13) where: Rv is gas constant for water vapour (Rv = 461.5 J/kg.K). Therefore, the relationship between water vapour permeability δp and δρ is: δp = δρ (2.14) RvT The mass conservation law says that mass can not disappear. This basic law is used to describe the moisture balance of a control volume. The inflow of moisture [kg/(m3.s)] minus outflow [kg/(m3.s)] has to be equal to moisture stored [kg/(m3.s)] in the control volume per time step: − ∂g v ∂u =ρ ∂x ∂t (2.15) The relationship between change of moisture content u (kg/kg) and partial vapour pressure change in a certain material can be described by using the slope of the sorption isotherm ξ and saturated water vapour pressure psat: ∂pv psat ∂u = ξ ∂x ∂x (2.16) - 15 - Chapter 2 By inserting the equation of state (2.16) to continuity equation (2.15), transient equation for moisture transport under isothermal conditions is obtained: ∂u ∂ ⎛ psatδ p ∂u ⎞ = ⎜ ⎟ ∂t ∂x ⎝ ρξ ∂x ⎠ (2.17) Convection Convective moisture transfer is caused by air flow. The convective moisture flow G [kg/s] is: G = Va ρv (2.18) where: Va is air flow rate in [m3/s]. If the air flows along a surface and there is the difference between water vapour concentration in flowing air and water vapour concentration in air close to the surface, the convective vapour transfer will occur. Analogously to heat transfer, the water vapour flux gv (kg/(m2.s)) is expressed as a function of the transfer coefficient βρ and the difference between air and the surface water vapour concentration: g v = β ρ ( ρv − ρv ,s ) (2.19) The mass transfer coefficient can be calculated from the heat transfer coefficient according to Lewis formula: βρ = ha ρ a ⋅ ca (2.20) 2.1.3 Direct link between heat and moisture transfer Heat and moisture transfer are processes which are linked together. The moisture transfer influences the heat transfer in the following ways: • Generation of latent heat • Thermal conductivity of wet materials is increased • Thermal capacity of wet materials is increased - 16 - Model for simultaneous heat and moisture transfer in EAHX Generation of latent heat For the earth-to-air heat exchanger, the generation of latent heat is considered to be important because the periods of condensation on the inner surface of the pipe and/or evaporation of the water vapour from the inner surface of the pipe can occur. If the water vapour condenses on the inner surface of the pipe, heat will be released; therefore, the air temperature will increase1. Consequently, increased air temperature increases the temperature of the pipe, and the condensation rate is reduced (a reverse link). The evaporation has an inverse effect. The generation of latent heat ql [W/m2, condensation (+), evaporation (-)] is defined as: ql = l ⋅ gv (2.21) where: l is latent heat of condensation [2.5*106 J/kg], gv is condensing (+) or evaporating (-) flux in [kg/(m2.s)]. 2.1.4 Analytical solutions for semi-infinite body An analytical solution of the general heat conduction equation (2.6) is to be found for semi-infinite medium. The harmonic oscillation of surface temperature is an important case of boundary temperature input. Harmonic oscillation of surface temperature Semi-infinite body is submitted to a harmonic oscillation of the surface temperature with period tp. The surface temperature oscillation θs may be defined by sinus function: ⎛ 2π t ⎞ ⎟ ⎜ tp ⎟ ⎝ ⎠ θ S (t ) = θ MEAN + θ A sin ⎜ (2.22) where: θMEAN is mean surface temperature, θA is the amplitude of surface temperature, and t is time. The formula for temperature oscillation inside semi-infinite slab (in the depth z [m]) is given as: 1 One might also imagine a thin film of water on the pipe surface which decreases the value of the heat transfer coefficient and reduces heat transfer between flowing air and the pipe. - 17 - Chapter 2 θ ( z, t ) = θ MEAN + θ A exp −z dp ⎛ 2π t z ⎞ − sin ⎜ ⎟ ⎜ tp dp ⎟ ⎝ ⎠ (2.23) The response is linearly shifted, and the amplitude is exponentially dampened. The heat flow rate is given as: −z ⎛ 2π t z π ⎞ Aλ 2 d − + ⎟ Q( z , t ) = θ A exp p sin ⎜ ⎜ tp dp 4 ⎟ dp ⎝ ⎠ (2.24) At the boundary (z = 0): Q(0, t ) = ⎛ 2π t π ⎞ Aλ 2 θ A sin ⎜ + ⎟ ⎜ tp dp 4 ⎠⎟ ⎝ (2.25) where: dp is periodic penetration depth [m]. dp = at p (2.26) π Approximate calculation of soil thermal diffusivity The approximate calculation of soil thermal diffusivity as can be based on measured soil temperature profiles in several depths; see e.g. (Verhoef et al., 2002). The method is based on the analytical solution for a semi-infinite slab comparing the amplitudes and/or phase shifts of signals in different depths. If the soil surface temperature oscillation was compared with temperature oscillation in depth z, the method would be expressed by the following formulas: π⎛ ⎞ z as = ⎜⎜ ⎟⎟ t p ⎝ ln ( dampening ) ⎠ as = π⎛ z ⎞ 2 (2.27) 2 (2.28) ⎜ ⎟ t p ⎝ shift ⎠ where: dampening is the amplitude dampening [-] in depth z, shift is the phase shift [rad] and tp is the period of the oscillation [s]. If the signals from two different depths were compared, z would denote the difference between the depths. - 18 - Model for simultaneous heat and moisture transfer in EAHX 2.1.5 Natural thermal regime of shallow subsurface The estimation of natural soil thermal regime may be quite complicated and dependent on many circumstances. Generally, the heat transfer in soils has conductive character. The most important factors for thermal regime of soil layer near the earth surface (depth < 15 m) are ambient air temperature, incoming solar radiation, vegetation, and snow cover (especially for northern countries with many days with snow cover). However, closer to the earth surface some other factors may play a significant role, e.g. latent heat of condensation or evaporation, subsurface freezing or thawing, evapotranspiration, subsurface moisture transfer, and heat transfer by long wave radiation between soil surface and surrounding surfaces (the sky in particular). “The interaction of all these variables over short and long time scales determines the temperature of the ground as a complex (i.e. nonlinear) and complicated series of processes”, (Beltrami, 2003). Understanding the thermal regime of soil is not important only for the earth-to-air heat exchangers, but it can provide valuable information about the Earth. “Temperature-depth profiles contain a robust signal of the long-term surface temperature history”, (Beltrami, 2001). For instance, the phenomenon of global warming has been studied with help of geothermal data. A potential time behavior of the temperature in shallow subsurface is depicted in Figure 2.2. The temperature of the ground two meters in depth under the surface is a harmonic oscillation with the maximum approximately in September and minimum approximately in March. 25 20 [°C] 15 10 5 0 -5 0 30 60 90 120 150 θ a 180 210 θ 1.0 240 270 300 330 360 θ 2.0 Figure 2.2: Daily means of ambient air temperature θa and soil temperature in depths 1.0 m and 2.0 m (θ1.0, θ2.0); based on measured data provided by the Geophysical Institute in Prague (http://www.ig.cas.cz/) - 19 - Chapter 2 Geothermal heat flow (steady state heat flow from the Earth’s interior towards the ground) in depths usual for horizontal pipes (several meters) can be neglected, for this heat flow is very weak (typical temperature gradient 0.3 K/100 m), compared with much stronger vertical heat flow through the surface. The temperature in ground in depth > 15 m is often estimated by the following linear dependence: θ ( z ) = θ MEAN + qg λ z (2.29) where: qg is the estimated or measured value of geothermal heat flow [W/m2], λ is the estimated mean thermal conductivity of ground [W/m.K], θMEAN is mean surface temperature [°C], and z is depth under the ground [m]. Moreover, the exchanger may be influenced by the level of underground water which might extract significant portion of heat because of the horizontal water flow. 2.2 Theoretical analysis 2.2.1 Problem definition The geometrical setup of EAHX used in the model is depicted in Figure 2.3 and Figure 2.4. Figure 2.3: Axonometric view of EAHX - 20 - Model for simultaneous heat and moisture transfer in EAHX Figure 2.4: Longitudinal and transversal section of EAHX 2.2.2 Heat transfer Generally, the convective heat flow along the pipe has to be balanced with conductive heat flow in soil. The heat transfer processes in the earth-to-air heat exchanger may be described by three differential equations. The differential equation (2.30) describes the heat balance of flowing air when latent heat generation is taken into account (Figure 2.5): ∂θ a ha 2π r0 h 2π r g l 2π r0 + θa − a 0 θs − v =0 ∂x ma ca ma ca ma ca (2.30) where: θa is temperature of air in the pipe [°C], θs is temperature of internal surface of the pipe [°C], gv is condensing (+) or evaporating (-) amount of water vapour [kg/(m2s)], these variables differ in length of the exchanger and in time; ha is air-to-pipe convective heat transfer coefficient [W/(m2.K)], ma is air flow rate [kg/s], these variables vary in time; l is latent heat of condensation [J/kg], ca is specific thermal capacity of air [J/(kg.K)]; they are assumed to be constant, and ro is internal radius of the pipe [m]. The differential equation (2.31) is the continuity (balance) equation describing heat conduction around the pipe: ρc p ⎛ ∂ 2θ ∂ 2θ ∂θ = λ⎜ 2 + 2 ⎜ ∂y ∂t ∂z ⎝ ⎞ ⎟⎟ ⎠ (2.31) where: θ is temperature of the soil [°C], ρcp is volumetric heat capacity of the soil [J/(m3.K)], and λ is thermal conductivity of the soil [W/(m.K)]. The longitudinal component of heat flow (along the pipe) can be neglected. A test on the influence of longitudinal heat conduction is performed in chapter 3.1.3. The moisture transfer in surrounding soil was neglected. Differential equations (2.30) and (2.31) are linked together by the heat balance of internal surface (si): - 21 - Chapter 2 ha (θ a − θ s ) = −λ ∂θ ∂r (2.32) si Figure 2.5: Heat balance on longitudinal control volume Convective heat transfer coefficient The earth-to-air heat exchanger usually employs longer horizontal pipes with aspect ratio length-to-diameter in the order of hundreds. The convective heat transfer coefficient depends on the velocity and the character of fluid flow, temperature and geometrical conditions. Using dimensionless Nusselt number Nu [-], the heat transfer coefficient can be calculated from the well-known formula: ha = Nu ⋅ λa Dchar (2.33) where: λa is thermal conductivity of air [W/(m.K)], Dchar is characteristic dimension, in this case diameter of the pipe [m]. Nusselt number for forced flow in a long circular pipe can be calculated according to the following formula (Ashrae, 2001): Nu = C Rem Pr n (2.34) where: constant C, exponents m and n are determined from experimental measurements, Re is Reynolds number [-], and Pr Prandtl number [-]. For common circular pipes, constant C and exponents m and n are: C = 0.023, m = 0.8, n = 0.4 for heating, n = 0.3 for cooling. Reynolds number Re represents the ratio between inertial and viscous (frictional) forces in the fluid: Re = ρ a va Dchar µa (2.35) where: ρa is the density of air [kg/m3], va is average flow velocity in the pipe [m/s], and µa is dynamic viscosity [kg/(m.s)]. Prandtl number Pr is calculated as the - 22 - Model for simultaneous heat and moisture transfer in EAHX ratio of two transport coefficients (kinematic viscosity υa [m2/s] and thermal diffusivity aa [m2/s]): Pr = νa aa = µ a ca λa (2.36) According to (Ashrae, 2001), formula (2.34) is a good approximation in turbulent range of 10 000 < Re < 120 000, 0.7 < Pr < 120, and L/D > 60. Thermophysical properties of air are shown in Table 2.1. T (°C) -10 0 10 20 30 λa ρa cpa [J/kg.K] [W/m.K] 0.023 0.024 0.025 0.026 0.026 µa 3 [kg/m ] 1.343 1.293 1.247 1.205 1.165 1005 1005 1006 1006 1007 [kg/m.s] 1.673*10-5 1.723*10-5 1.772*10-5 1.821*10-5 1.868*10-5 υa [m2/s] 1.246*10-5 1.332*10-5 1.421*10-5 1.511*10-5 1.604*10-5 Pr (-) 0.72 0.72 0.72 0.72 0.71 Table 2.1: Thermo-physical properties of air The value of Prandtl number is almost independent on temperature; therefore, the influence of temperature on the value of Nusselt number can be neglected. The values of the heat transfer coefficient are depicted in Figure 2.6 and Figure 2.7 (the air properties were chosen for temperature 10 °C). 35 ha [W/m2.K] 30 25 20 15 150mm - c 150mm - h 200mm - c 200mm - h 250mm - c 250mm - h 10 5 100 200 300 400 500 600 700 V [m3/h] 800 900 1000 Figure 2.6: Convective heat transfer coefficient ha with respect to air flow rate and pipe diameter; c denotes cooling, h denotes heating - 23 - Chapter 2 35 ha[W/m2.K] 30 25 20 150mm - c 150mm - h 200mm - c 200mm - h 250mm - c 250mm - h 15 10 5 1 2 3 4 5 6 va [m/s] 7 8 9 10 Figure 2.7: Convective heat transfer coefficient ha with respect to velocity and pipe diameter; c denotes cooling, h denotes heating As seen from Figure 2.7, the heat transfer coefficient is almost linear function of the velocity. The following approximate relation (Hollmuller, 2002) could be used: h a = 3va + 3 (2.37) The velocity in a pipe related to air flow rate and pipe diameter is depicted in Figure 2.8. va [m/s] 15 150 mm 200 mm 250 mm 10 5 100 200 300 400 500 600 700 V[m3/h] 800 900 1000 Figure 2.8: The velocity va with respect to air flow rate and pipe diameter 2.2.3 Moisture transfer The description of moisture transfer in EAHX was simplified; variations of moisture content and latent heat effects in surrounding soil were not taken into account. - 24 - Model for simultaneous heat and moisture transfer in EAHX Basically, it is assumed that water vapour concentration in flowing air is a constant value: ρv = ρv,in (2.38) where: ρv is water vapour concentration of air in EAHX [kg/m3], ρv,in is water vapour concentration of air at the inlet of the exchanger [kg/m3]. If the pipe is made of plastic material with tight joints and no condensation or evaporation from wet pipe surface occurs, the assumption (2.38) will be valid. However, if water vapour condenses or evaporates, air will be either dehumidified or humidified. The resulting moisture balance (Figure 2.9) on the longitudinal element is set up analogously to the heat balance: β 2π r0 θ s ∂ρ v β ρ 2π r0 ρv − ρ ρ + =0 v , sat Va Va ∂x (2.39)2 where: Va is air flow rate [m3/s], ρv,sat is saturated water vapour concentration (a function of internal pipe surface temperature θs) a βρ is moisture transfer coefficient [m/s]. Figure 2.9: Moisture balance on longitudinal control volume Furthermore, it is advantageous to introduce some simplifications and limitations: the pipe can be wet only as a consequence of previous condensation3 water in the pipe does not move (water stays in the same longitudinal control volume as it has condensed) and the surface of the pipe is moist uniformly4 2 If formula (2.20) was added to (2.39) the equation analogous to (2.30) would be obtained. 3 In reality, an important phenomenon might be the constant infiltration of moisture from the soil into the pipe because of leakage in the pipe wall, (Hollmuller, 2001). - 25 - Chapter 2 air moistening is limited by saturated water vapour concentration ρv ≤ ρ θa (2.40) v , sat where: ρv,sat is saturated water vapour concentration (a function of air temperature in the pipe θa). 2.2.4 Coupled heat and moisture transfer The thermal balance of flowing air in the pipe is influenced by heat connected with a change of phase. The thermal effect of condensation or evaporation is covered by the last term in the differential equation (2.30). 2.3 Numerical solution The following system of indexes is introduced: i - x axis index (along the length of the exchanger), j – y axis index, k – z axis index, and t – time step index. Index (t - 1) denotes previous values (preceding time step). Basically, the exchanger may work in the following operation modes: • ma ≠ 0; heat exchanger is in operation, air is sucked through the pipe • ma = 0; heat exchanger is not in operation, and this stage is called the natural soil recovery5 2.3.1 Stage of operation Generally, the solution procedure within one time step of the operational stage is divided into three sub-steps: • 1a) the explicit calculation of new air temperature and 1b) the explicit calculation of new water vapour concentration in air, condensing and/or evaporating amount, and accumulated water in the exchanger • 4 2) the explicit calculation of new soil temperature In reality, the pipe is often sloped and water can move. 5 On the contrary, the forced soil recovery might occur when the exchanger will be operated during a summer night (soil is cooled down by relatively cold night air). - 26 - Model for simultaneous heat and moisture transfer in EAHX • 3) the calculation of new pipe surface temperature Calculation of new air temperature (step 1a) The analytical solution of (2.30) can be written as: ( ) θ a (i ,t ) = θ s ,eqv i ,t −1 + θ a(i −1,t ) − θ s ,eqv( i ,t −1) exp ( ) − ha 2π r0 ∆xi ma ca (2.41) θ s ,eqv( i ,t −1) = θ s ,mean( i ,t −1) + ∆θ s ,lat ( i ,t −1) (2.42) where: θs,mean is mean surface temperature in the pipe [°C] and θs,eqv is equivalent surface temperature containing the thermal contribution of condensation or evaporation ∆θs,lat [°C]. ∆θ s ,lat( i ,t −1) = g v ( i ,t −1) ⋅ l (2.43) ha 1 14 2 ha [W/(m K)] 12 10 8 2 6 3 4 4 2 4 6 2 gv [g/(m s)] 5 8 67 -3 x 10 Figure 2.10: ∆θs,lat [°C] as a function of water vapor flux gv and convective heat transfer coefficient ha Calculation of new water vapour concentration and moisture flux (step 1b) The analytical solution of (2.39) can be written as: ( ) ρv (i ,t ) = ρv,sat i ,t −1 + ρv( i −1,t ) − ρv,sat(i ,t −1) exp ( ) − β ρ 2π r0 Va ∆xi The possibilities which can occur are summarized in the Table 2.2. - 27 - (2.44) Chapter 2 Moisture flux Water vapour concentration Condition Condensation ρv (i −1,t ) > ρvθ,ssat (i ,t −1) Evaporation Dry surface ρv (i −1,t ) < ρvθ,ssat (i ,t −1) ρv (i −1,t ) < ρ vθ,ssat (i ,t −1) and Gacu (i ,t −1) > 0 and Gacu (i ,t −1) = 0 * ρv (i ,t ) = according to (2.44) ρv (i,t ) = according to (2.44) g v (i ,t ) = ρv (i ,t ) ≤ ρvθ,asat ρ v (i ,t ) = ρ v (i ,t ) ρv (i ,t ) > ρvθ,asat ( ρv(i−1,t ) − ρv(i,t ) )Va 2π r0 ∆xi g v (i ,t ) = ρv (i ,t ) = ρvθ,asat ρv (i ,t ) = ρv (i −1,t ) ** ( ρv(i−1,t ) − ρv(i,t ) )Va 2π r0 ∆xi g v (i ,t ) = 0 g v (i ,t ) = 0 ** Table 2.2: Condensation and evaporation in the model * The surface of pipe is wet; Gacu is accumulated moisture [kg]. ** The relative humidity must not exceed 100 %. Even if the pipe is wet, evaporation will not occur (air is fully saturated). Calculation of new temperature in the soil (step 2) Soil temperature is calculated in perpendicular planes to the length of the exchanger as shown in Figure 2.11. As a result, the 3D temperature field is calculated using calculation in 2D. The calculation mesh is automatically generated by developed function, denser near the pipe and the ground surface with an expansive factor towards other boundaries. - 28 - Model for simultaneous heat and moisture transfer in EAHX Figure 2.11: Perpendicular sections along length of the exchanger. The explicit finite difference method is used for the solution of equation (2.31). The method is based on the calculation of new temperatures from previous temperatures (fully explicit scheme for a discretization in time), and it is briefly introduced in appendix A2 using the information gathered in (Rode, 1997). A special attention has to be aimed to set up the heat balance of boundary nodes: Pipe-to-soil boundary: mean air temperature θa,mean and the value of convective resistance is used for constructing air/pipe-to-soil boundary condition6. Mean air temperature in each longitudinal control volume is calculated as: 6 The approximation of air temperature in a longitudinal element by mean temperature need not be always accurate. If one dealt with a theoretical case of very high (e.g. closed to infinity) air-to-pipe convective heat transfer coefficient, the approximation would lead to the overestimation of heat flow - 29 - Chapter 2 θ a ,mean( i ,t ) = θ a( i −1,t ) + θ a( i ,t ) (2.45) 2 The pipe is approximated by an equivalent square with a perimeter which equals to the perimeter of the pipe. The thermal resistance of the pipe is calculated according to (2.46). Figure 2.12: Replacement of the pipe by an equivalent square. R pipe ⎛ ⎞ rpipe 2 D pipe ln ⎜ ⎜ rpipe − t pipe ⎟⎟ ⎝ ⎠ = π ⋅ λ pipe (2.46) The boundary at the ground surface: ambient air temperature, global solar radiation on a horizontal plane, convective surface thermal resistance Ra (the influence of wind), and inserted additional thermal resistance Rs representing the influence of soil cover (vegetation, snow, defined by user) are used for constructing ambient air-toground boundary condition (Figure 2.13). Undoubtedly, the soil surface boundary condition should also consider many other factors (see section 2.1.5). Perhaps, except for the heat transfer via radiation, these factors influence a few upper centimeters of soil only, for they are based on daily time scale (e.g. morning condensation followed by evaporation). Therefore, when one deals with horizontal pipes placed in depth of several meters under the ground, it should be possible to leave out short time scale factors (short-term air-soil temperature coupling) out of balancing the upper soil surface boundary condition. Their influence on the exchanger is probably minor. between air and the ground. However, the approximation is accurate for real cases of low values of convective heat transfer coefficient. For details see chapter 3.1. - 30 - Model for simultaneous heat and moisture transfer in EAHX Figure 2.13: The ground boundary condition. The remaining walls of the rectangle: the vertical sides of the rectangle are usually assumed to be adiabatic. The adiabatic boundary condition may be placed at the bottom side of the rectangle. In some cases, this boundary may be assumed to be isothermal (e.g. with annual mean ambient temperature). Calculation of new pipe surface temperature (step 3) The surface temperature of the pipe is calculated by setting up the heat balance for the inner surface of a fictitious pipe. 2.3.2 Stage of natural soil recovery For this mode, air in the pipe is assumed to be still. The calculation follows up the calculation of the operation stage. The initial values of soil temperature are taken from the previous time step (the last time step of the operation). Again, explicit finite difference method is used for the calculation of soil temperature. - 31 - Chapter 2 2.3.3 Flow diagram A flow diagram of the basic algorithm is depicted in Figure 2.14. The calculation routine has been developed within MATLAB environment. start read initial values list of inputs pipe surface temperature soil temperature time = 1,t yes ma > 0 1a, 1b 2 stage of operation 2 natural soil recovery 3 3 print outputs end Figure 2.14: Flow diagram of the algorithm - 32 - Model for simultaneous heat and moisture transfer in EAHX 2.3.4 Stability and accuracy Using previous values for the calculation of new ones brings also some stability problems. Except for possible numerical instability of used explicit finite difference method, which is controlled by using stable time step (reasonably dense mesh), there is a stability problem due to the use of old temperatures and old condensation rates (explicit coupling in (2.41) and (2.44)). Using preceding pipe surface temperatures may influence air temperatures in a few time steps after the stage of natural soil recovery. Using preceding moisture fluxes may overestimate or underestimate the latent heat of condensation with the impact on new air temperatures. These oscillations can be limited via a calculation which repeats basic algorithm several times per one time step (Figure 2.15). The inserted iteration loop involves some other computational operations in the algorithm; therefore, it influences the speed of the calculation. Generally, the iteration is appropriate to be used for calculations of several time steps after the stage of natural soil recovery only and perhaps for cooling calculations when latent heat effects play more important role. The density of the mesh is also of vital importance since it influences the speed of the calculation, accuracy, and stability of heat conduction calculation. The denser mesh is the shorter time step we need and the longer calculation is performed. The difference between the properties of air inside the pipe (insulation without thermal capacity) and soil properties (a good conductor with high thermal capacity) causes the problem with stability of heat conduction calculation during the stage of natural soil recovery. A very short time step is necessary in order to achieve stability; therefore, the calculation might be time consuming. The possible solution could be either the calculation of heat conduction using the implicit method, or replacement the inner volume of air in the pipe by soil in order to preserve the explicit calculation. The latter way is used in the MATLAB routine. - 33 - Chapter 2 time = 1,t ma > 0 θa,new 1 gv,new θnew 2 2 natural soil recovery θs,new 3 3 i=1 stage of operation old = new i=i+1 θa,new = f(θs,old, gv,old) 1 gv,new 2 θnew 3 θs,new |θa,new - θa,old| < ε v i>n yes i>n yes print mistake Figure 2.15: Alternative flow diagram – iterative procedure for stage of operation - 34 - Model for simultaneous heat and moisture transfer in EAHX 2.4 Summary Based on the theory described in this chapter, a series of computer codes written in MATLAB environment (Figure 2.16) was developed in order to deal with the simulation of EAHXs. The main features of the model can be described as follows: Heat transfer: The model is based on: a) the analytical solution of differential equation describing heat balance of the exchanger longitudinal control volume of the exchanger (sensible and latent heat is taken into account) and b) the numerical solution of two-dimensional transient heat conduction in soil around the pipe. The algorithm allows the calculation of air and pipe surface temperature profiles along the exchanger, temperatures within the soil as well as extracted/injected heat from/to soil for different inlet temperature signals and different boundary conditions. Moisture transfer: The model is based on the analytical solution of the differential equation (analogy with heat transfer) describing moisture balance of the exchanger longitudinal control volume. The moisture transfer within soil is not included. The algorithm allows the calculation of moisture flux (condensation or evaporation amount in the pipe), accumulated moisture in the pipe, and outlet air relative humidity and water vapour concentration. Coupled heat and moisture transfer: The effect of latent heat (condensation on the pipe surface or evaporation from already wet surface) on the air balance may be taken into account. Automatic grid generator Inputs Input m-file With expansive mesh from pipe and the ground towards other boundaries From hourly Weather data data interpolation Solution loop Results Control schedules Routine Operation Air flow rate Figure 2.16: Structure of the calculation routine - 35 - Chapter 2 - 36 - Model validation 3 Model validation An analytical and experimental validation of the model will be introduced in this chapter. 3.1 Flat-plate heat exchanger The subchapter will illustrate development of the EAHX model. In fact, a routine for the flat-plate heat exchanger (FPHX) was developed earlier than already thoroughly described model for a simulation of EAHXs. The model of the FPHX is simpler (but similar in principle) than more complex case of the EAHX. The study of the FPHX surprisingly provides a good physical insight into the dynamical performance of air-to-mass heat exchangers. Especially, air-to-mass convective coupling was tested with the FPHX. Besides, some other tests were performed: a) the effect of heat conduction parallel with longitudinal axis was studied; b) the phenomenon of phase-shifting almost without amplitude dampening was studied. The goal was to verify whether longitudinal heat flow component is negligible or not and to verify the phenomenon of phase-shifting. 3.1.1 Problem definition The flat-plate heat exchanger is a device depicted in Figure 3.1 and Figure 3.2. It consists of two mass panels with vertical adiabatic sides and a gap which allows airflow parallel with x-axis. Figure 3.1: Axonometric view of the flat-plate heat exchanger - 37 - Chapter 3 Figure 3.2: Longitudinal and transversal section of the flat-plate heat exchanger A few changes compared to the model of the EAHX have to be made in order to describe heat transfer in the FPHX. The heat balance of the longitudinal control volume has a similar form to equation (2.30): ∂θ a ha 2 B h 2B θa − a θ s = 0 + ma ca ∂x ma ca (3.1) where: θa is temperature of air in the exchanger [°C], θs is temperature of internal surface of the exchanger [°C], ha is convective (air-to-plate) heat transfer coefficient [W/(m2.K)], ma is air flow rate [kg/s], ca is specific thermal capacity of air [J/(kg.K)], B is width of the plate [m]. Because of the adiabatic boundary on both vertical sides, the differential equation describing heat conduction in the panel does not contain the y component: ρc p ⎛ ∂ 2θ ∂ 2θ ∂θ = λ⎜ 2 + 2 ⎜ ∂x ∂t ∂z ⎝ ⎞ ⎟⎟ ⎠ (3.2) where: θ is temperature of the panel [°C]. These two differential equations are linked together by the heat balance of the internal surface (si): ha (θ a − θ s ) = −λ ∂θ ∂z (3.3) si 3.1.2 Numerical solution The algorithm is similar to the already described procedure for earth-to-air heat exchangers (see chapter 2.3); only a few changes have to be made. The temperature of air along the pipe (the analytical solution of equation (3.1)) is calculated as: θ a ( i ,t ) = θ s ( i ,t −1) h 2B − a ∆ xi ⎞ exp ma ca + ⎛⎜ θ a − θs ⎟ ( i ,t −1) ⎠ ⎝ ( i −1,t ) - 38 - (3.4) Model validation where: i is x axis index, and t – time step index. The analysis of the exponent7 in the equation (3.4) reveals two limit cases: If NTU is infinite, term (e-∞) will equal to zero. Such situation would refer either to infinite convective heat transfer coefficient ha or infinite exchange surface (infinite length or radius) or zero air flow rate ma. The solution (3.4) will be reduced to the following form: θout = θ s (3.5) If NTU is equal to zero, term (e0) is equal to one. Such situation would refer either to zero exchange surface or zero convective heat transfer coefficient ha or infinite air flow rate ma. The solution (3.4) will be reduced to the following form: θout = θin (3.6) The temperature of the panel may be calculated in two ways: either 1D calculation (heat flow component parallel with x axis is neglected) or 2D calculation can be utilized. In the first case, plane xz is divided into the stripes which are separated from each other by an adiabatic layer (Figure 3.5). Figure 3.3: FPHX – the longitudinal control volume and 1D heat conduction in the panel 7 The exponent is called number of transfer units (NTU) in standard terminology of heat exchangers. - 39 - Chapter 3 3.1.3 Test of longitudinal heat conduction The heat flow component parallel with x axis was neglected from the analysis of the EAHX. The test with a step change of inlet air temperature had been performed with the model of the FPHX in order to prove such simplification. Let no air flows through a concrete FPHX longer time. The panel has length L = 20 m, thickness W = 1.0 m, width B = 0.3 m, cavity 0.05 m between panels, and the isothermal boundary condition at the external surface of both panels with temperature 0 °C. The initial temperature of the panel is 0 °C. At time zero, there is a sudden change of the inlet air temperature; the inlet air temperature is changed to 10 °C and the air flow rate is maintained at 125 m3/h. The convective heat transfer coefficient is calculated according to equation (2.37). As shown in Figure 3.4, the influence of longitudinal heat flow component is very weak. The difference of the outlet air temperature between 1D and 2D calculation was in order of 10-4 (negligible). x 10 10 5 7.5 4 [°C] [°C] -4 5 3 2 θ 2.5 in 1 θ1D out 0 2 4 6 8 [days] 10 12 14 0 0 diff 2 4 6 8 [days] 10 12 14 Figure 3.4: Left - temperature of inlet (θin) and outlet air (θout ); right - the difference between 2D and 1D calculation of heat conduction in the panel, diff = θout2D - θout1D 3.1.4 Phase-shifting phenomenon The characteristic phenomenon of phase-shifting almost without amplitudedampening of input harmonic signal was most probably first formulated by (Hollmuller, 2003). A rather special combination of input harmonic oscillation, thin mass layer, large exchange surface, external adiabatic boundary, and high convective heat transfer lead to such dynamical balance. This phenomenon might be used as a cooling technique for buildings in a climate with strong daily oscillation of ambient air temperature (e.g. mild climate). (Hollmuller, 2006) dealt with a laterally insulated rectangle box filled with - 40 - Model validation thermal mass through which air is sucked in order to delay input harmonic oscillation almost without dampening. A developed analytical model is compared with experimentally monitored performance of the material stored in the box (e.g. ceramic balls, ceramic slabs, perforated ceramic bricks, gravel). The effort was aimed at calibration of the analytical model and development of real-size installation for a house (so called cool shifter). This chapter compares the numerical simulation performed with the numerical model of the FPHX with an analytical solution of the FPHX (Hollmuller, 2003). The goal was to make another cross comparison and understand more deeply the dynamics of air-to-mass heat exchangers. The analytical solution of the FPHX with harmonic input (see appendix A1) has similar form as the analytical solution for cylindrical heat exchanger (3.12). A small MATLAB routine was written in order to calculate the analytical output. Different setups of the FPHX (see Table 3.1) were simulated: • setups with low values of convective heat transfer • setups with high values of convective heat transfer • extreme setups with very high convective heat transfer (~ closed to infinity) Each setup (Table 3.1) was defined by a combination of dimensionless numbers with prescribed value of the convective heat transfer coefficient. Dimensionless numbers are defined as: ∆R0 = W dp (3.7) S S = Sd p (3.8) h ha = a hd p (3.9) where: W is thickness of the plate [m], dp is daily penetration depth [m/day], Sdp is associated characteristic surface [m2], ha is convective heat transfer coefficient [W/(m2.K)], hdp is conductance between inner surface of panel and penetration depth [W/(m2.K)]. Sd p = ma ca hd p (3.10) - 41 - Chapter 3 λ = 1,6 W/m2.K ρ = 2300 kg/m3 Material properties Length L [m] Width B [m] Discretization c = 840 J/kg.K ρc = 1,93 MJ/m3.K 2,0 m 0,25 m 30 control volumes 20 control volumes 60 s z axis x axis Time step ha [W/m2.K] W [m] Va [m3/h] DAM 10,6 10,6 10,6 10,6 0,15 0,15 0,15 0,15 4 8 16 32 1 1 1 1 8 4 2 1 1 1 1 1 PS PS 8 PS 4 PS 2 PS 1 10,6 10,6 10,6 10,6 0,03 0,03 0,03 0,03 4 8 16 32 0,2 0,2 0,2 0,2 8 4 2 1 1 1 1 1 DAM DAM 8 DAM 4 DAM 2 DAM 1 106 106 106 106 0,15 0,15 0,15 0,15 4 8 16 32 1 1 1 1 8 4 2 1 10 10 10 10 PS PS 8 PS 4 PS 2 PS 1 106 106 106 106 0,03 0,03 0,03 0,03 4 8 16 32 0,2 0,2 0,2 0,2 8 4 2 1 10 10 10 10 DAM DAM 8 DAM 4 DAM 2 DAM 1 1060 1060 1060 1060 0,15 0,15 0,15 0,15 4 8 16 32 1 1 1 1 8 4 2 1 100 100 100 100 PS 8 PS 4 PS 2 PS 1 1060 1060 1060 1060 0,03 0,03 0,03 0,03 4 8 16 32 0,2 0,2 0,2 0,2 8 4 2 1 100 100 100 100 low ha DAM 8 DAM 4 DAM 2 DAM 1 high ha very high ha ∆R0 [-] S [-] ha [-] PS setup Table 3.1: Simulated setups of FPHX; DAM denotes dampening setup, PS denotes phase-shifting setup As shown in Figure 3.5, the numerical simulation precisely matched the analytical results of the case with low convective heat transfer. The simulation also - 42 - Model validation ~ matched the analytical results of the case with high convective heat transfer with S = 1 ~ or S = 2 , Figure 3.6. However, the numerical simulation did not match the analytical results of the case with very high convective heat transfer. In the latter case, the performance of the exchanger seems to be flow rate independent and exchanger geometry independent (Figure 3.7). Hence, the numerical algorithm (especially the thermal link between air and mass) has to be reviewed again in order to make such discrepancy clear. The numerical algorithm first calculates the mean value of the air temperature from input and output temperature of the longitudinal control volume (see Figure 2.11). Then, the mean air temperature together with convective resistance form the boundary condition (coupling with mass) assigned to the longitudinal control volume. Using the mean air temperature to construct the thermal link between air and panel is rather a good approximation for low convective heat transfer. However, if very high convective heat transfer occurs, the heat flow injected to panel will not be calculated correctly. Instead of the mean air temperature with convective resistance, it would be better to construct the thermal air-to-mass link as a boundary with heat flux driven by temperature difference between input and output of the longitudinal control volume. In latter case, the energy balance will always be correct. 10 7.5 5 2.5 0 -2.5 -5 -7.5 -10 0 0.5 input 1 output numerical 1.5 anal#PS1 2 anal#PS2 2.5 anal#PS4 3 anal#PS8 10 7.5 5 2.5 0 -2.5 -5 -7.5 -10 0 input 0.5 1 output numerical 1.5 anal#DAM1 2 anal#DAM2 2.5 anal#DAM4 3 anal#DAM8 Figure 3.5: Low ha – comparison of the numerical simulation (grey) with analytical solution - 43 - Chapter 3 10 7.5 5 2.5 0 -2.5 -5 -7.5 -10 0 0.5 1 output numerical input 1.5 anal#DAM1 2 anal#DAM2 2.5 anal#DAM4 3 anal#DAM8 10 7.5 5 2.5 0 -2.5 -5 -7.5 -10 0 0.5 input 1 output numerical 1.5 anal#PS1 2 anal#PS2 2.5 anal#PS4 3 anal#PS8 Figure 3.6: High ha - comparison of the numerical simulation (grey) with anal. solution 10 7.5 5 2.5 0 -2.5 -5 -7.5 -10 0 input 0.5 1 output numerical 1.5 anal#DAM1 2 anal#DAM2 2.5 anal#DAM4 3 anal#DAM8 10 7.5 5 2.5 0 -2.5 -5 -7.5 -10 0 0.5 input 1 output numerical 1.5 anal#PS1 2 anal#PS2 2.5 anal#PS4 3 anal#PS8 Figure 3.7: Infinite ha - comparison of the numerical simulation (grey) with anal. solution - 44 - Model validation 3.1.5 Summary The simulation of the FPHX was presented. First, the comparison between the simulation routines of the FPHX led to the conclusion that the heat conduction parallel with longitudinal axis of air-to-mass heat exchangers is negligible. This finding was reflected in the formulation of the model of the EAHX. Next, the results from simulation of several FPHX setups were compared with the results calculated via analytical solution of the FPHX. The comparison showed the imperfection of the numerical algorithm which caused the incorrect results for theoretical cases with very high convective heat transfer. Further research perspectives: It might be interesting, particularly from theoretical point of view, to deal with a harmonic oscillation at the external boundary of the FPHX. Two driving input signals and their conductive interaction with the mass of panels somehow form the oscillation of the output It might be interesting to deal with a coupled system of a solar air collector with a storage masonry wall (FPHX). The wall is the component of a house. The interaction of the energy producer with the wall is one problem. The interaction of the storage wall with internal environment of the house is another complication in analysis. 3.2 Verification exercises 3.2.1 Test of undisturbed soil temperature calculation The calculation of natural temperature stratification in soil is partial and difficult component of the earth-to-air heat exchanger simulation. A test of the algorithm ability for calculation the temperature of soil was performed with weather data for Holzkirchen (Figure 3.8; downloaded from http://www.wufi.de). First, the soil thermal diffusivity was determined by a cross comparison between three temperature signals (see chapter 2.1.4): i) the temperature of the surface, ii) the temperature of the ground 0.5 m in depth under the surface, iii) the temperature of the ground 1.0 m in depth under the surface. These signals were decomposed into annual harmonic pulses (Figure 3.9) so that the amplitude, the phase of each pulse, yearly - 45 - Chapter 3 penetration depth, and thermal diffusivity of soil were determined (Table 3.2). The range of calculated thermal diffusivities indicates that the ground was not perfectly homogenous in the vertical direction. However, the decrease of thermal diffusivity with depth is not standard situation since the porosity of upper layer (< 0.5 m) is typically higher than pressed bottom layers. Therefore, the calculation of thermal diffusivity from annual harmonics seems inaccurate. 40 35 30 25 20 [°C] 15 10 5 0 -5 -10 -15 -20 0 30 60 90 120 150 θ 180 θ e 210 240 θ s 270 300 330 360 θ 0.5 1.0 Figure 3.8: Holzkirchen, weather data - θe is ambient air temperature, θs is the temperature of the ground surface, θ0.5 and θ1.0 are temperatures 0.5 m and 1.0 m in depth under the surface 20 [°C] 15 10 5 0 -5 0 30 60 90 120 150 θ s 180 210 θ 0.5 240 270 300 330 360 θ 1.0 θ0.5 vs. θs θ1.0 vs. θs θ1.0 vs. θ0.5 Phase shift [rad] Comparison between signals Amplitude damping [-] Figure 3.9: Holzkirchen – decomposition of weather data into harmonics 0.8629 0.1494 Penetration depth dp [m/year] from from phase damping shift 3.39 3.35 Thermal diffusivity as [m2/s] from from damping phase shift 1.15*10-6 1.11*10-6 0.7094 0.3506 2.91 2.85 0.85*10-6 0.81*10-6 0.8221 0.2012 2.55 2.49 0.65*10-6 0.62*10-6 mean value of soil thermal diffusivity as [m2/s] 0.86*10-6 Table 3.2: Amplitude dampening and phase shift of harmonics – calculated values of yearly periodic penetration depth and thermal diffusivity - 46 - Model validation Next, the simulation of undisturbed soil temperature was performed; the time step of the simulation was 1200 seconds. A calculation domain was 10 m stripe (1D heat conduction) divided into 36 control volumes (denser grid near the ground). The bottom boundary was assumed to be isothermal with temperature equal to the annual mean temperature of ambient air on site (6.4 °C). Soil properties were assumed as follows: subsurface 0.5 m thick layer with thermal conductivity 2.7 W/mK and volumetric heat capacity 2.4 MJ/m3K; remaining soil with thermal conductivity 1.5 W/mK and identical volumetric heat capacity as sub-surface layer. The balance of the upper plane surface was set up by ambient air temperature, global solar radiation on a horizontal plane (absorptivity α according to Figure 3.11), convective surface thermal resistance Ra (the influence of wind, set to be constant value 0.04 m2K/W), and inserted additional thermal resistance Rs representing the influence of soil cover (vegetation or snow, inserted between the ground surface and the first soil node). It was supposed that [m2K/W] or [-] the ground was covered by snow between 21st November and 1st March. 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0 30 60 90 120 150 α 180 210 240 Rs Ra 270 300 330 360 Figure 3.10: Holzkirchen - surface resistances Rs, Ra, and absorptivity α used in the simulation Despite many uncertainties in input parameters (soil properties, soil cover) and some bugs evident in measured data (sudden discontinuities in soil temperature time profile), the trends of all simulated cases correspond with the trends of measured data, as shown e.g. in Figure 3.11. Basic statistics of the difference (numerical - measured) are given in Table 3.3. The temperature profile in depth of 1 m was calculated with better accuracy than the profile closer to the surface. Probably, the closer to the surface, the more uncertain the calculation was. - 47 - Chapter 3 20 [°C] 15 10 5 0 0 30 60 90 120 150 180 simulated 1.0 210 240 270 measured 1.0 θ 300 330 360 θ Figure 3.11: Holzkirchen - comparison between numerical simulation and measured soil temperature in depth of 1.0 m; from hourly values 20 [°C] 15 10 5 0 0 30 60 90 120 150 180 measured 0.5 210 240 270 simulated 0.5 θ 300 330 360 θ Figure 3.12: Holzkirchen - comparison between numerical simulation and measured soil temperature in depth of 0.5 m; from hourly values 30 25 20 [°C] 15 10 5 0 -5 -10 0 30 60 90 120 150 180 θ - simulated s 210 240 270 θ - measured s 300 330 360 Figure 3.13: Holzkirchen – daily means of soil surface temperature, the comparison between numerical simulation and measured data numerical - measured max mean min std from 18.66 1.53 -15.38 4.42 hourly means θs θ0.5 θ1.0 4.86 1.46 -2.96 1.55 hourly means 3.79 1.20 -0.76 0.96 hourly means Table 3.3: Basic statistics of the difference (numerical – measured) - 48 - Model validation Two typical situations of the ground heat balance are depicted in Figure 3.14. The heat flux through the ground surface is depicted in Figure 3.15. Figure 3.14: The ground heat balance; left – typical winter situation, right – typical summer situation 40 30 [W/m2] 20 10 0 -10 -20 -30 0 30 60 90 120 150 180 210 240 270 300 330 360 Figure 3.15: Daily means of heat flux through the ground surface (q1); positive values denote the gain (soil is heated up), negative values denote the loss (soil is cooled down); the mean value of q1 is closed to zero 3.2.2 Test of hygrothermal calculations Simulations presented in this chapter are supposed to examine some moisture calculations with the model. Generally, it is better to focus on trends of output parameters (which are expected) rather than on absolute values of output parameters (which are uncertain). The case of dry air provides a basic case. If moist air is sucked through the exchanger, the basic case will be deformed due to the release of latent heat of condensation and/or consequent evaporation. Generally, the magnitude of such deformation depends on: inlet air temperature and relative humidity, pipe surface temperature, air flow rate, pipe length and depth, and soil properties. A calculation domain was a block 2 m (width) x 2 m (height) x 30 m (length of the pipe). The domain was divided into 17 x 17 x 120 control volumes. All external - 49 - Chapter 3 walls of the rectangle were assumed to be adiabatic. The pipe (external diameter 200 mm) was approximated by an equivalent square with perimeter which equals to perimeter of the pipe. The time step of all simulations was 300 seconds. The soil was assumed to have properties λ = 1.9 W/(m.K) and ρcp = 1.9 MJ/(m3.K). Step change of inlet relative humidity Let no air flows through the pipe longer time. The temperature of the block is constant (13 °C). At time zero, there is a sudden change of the inlet air temperature and relative humidity; the inlet air temperature is changed to 30 °C and relative humidity is changed to 50 %. The air flow rate is maintained at 250 m3/h. Figure 3.17 shows the expected performance. Warm and humid air passing through the pipe is cooled down. Initially, water vapour condensation prevails, but the pipe and surrounding soil are consequently warmed up. Next, evaporation prevails over condensation until condensation fully diminishes and finally all water in tubes evaporates. 30 [°C] 25 20 θ out 15 θ 10 θdry 0 in out 0.5 1 1.5 [day] 2 2.5 Figure 3.16: Temperature of inlet (θin) and outlet air (θout); index dry denotes the simulation with dry air. The reason for the instability visible near the initial step of inlet air temperature was described in section 2.3.4 - 50 - Model validation 750 Gacu [g] 500 250 0 0 0.5 1 1.5 [day] 2 2.5 Figure 3.17: Total accumulated moisture Gacu [g] in the pipe; a marker shows the point when condensation stopped dG [g/timestep] 15 10 5 0 0.5 1 1.5 [day] 2 2.5 [control volumes Figure 3.18: Moisture rate dG = Gcond + Gevap [g/timestep]; if positive, condensation rate Gcond prevails; if negative evaporation rate Gevap prevails 100 75 50 25 0 0.5 1 1.5 [day] 2 2.5 Figure 3.19: Number of control volumes with water (120 control volumes within the pipe) Periodic inlet signals Let no air flows through the pipe longer time. The temperature of the block is constant (13 °C). At time zero, the inlet air temperature and relative humidity started to pulsate with daily frequency as shown in Figure 3.20 and in Figure 3.21 respectively. The air flow rate is maintained at 250 m3/h. - 51 - Chapter 3 θ 30 θ [°C] 25 out in θdry out 20 15 10 0 0.5 1 1.5 [day] 2 2.5 Figure 3.20: Temperature of inlet (θin) and outlet air (θout); index dry denotes the simulation with dry air [%] 90 80 70 rhout 60 0 rhin 0.5 1 1.5 [day] 2 2.5 Figure 3.21: Relative humidity of inlet (rhin) and outlet air (rhout) cout [g/m3] 17.5 cin 15 12.5 10 7.5 0 0.5 1 1.5 [day] 2 2.5 Figure 3.22: Water vapour concentration in inlet (cin) and outlet air (cout) - 52 - Model validation 1500 Gacu [g] 1000 500 0 0 0.5 1 1.5 [day] 2 2.5 [g/timestep] Figure 3.23: Total accumulated moisture Gacu [g] in the pipe 20 10 0 -10 -20 0.5 1 1.5 [day] 2 dG 2.5 Gcond(+) Gevap(-) [control volumes] Figure 3.24: Moisture rate dG = Gcond + Gevap [g/timestep]; if positive, condensation rate Gcond prevails; if negative evaporation rate Gevap prevails 75 50 25 0 0.5 1 1.5 [day] 2 2.5 Figure 3.25: Number of control volumes with water (120 control volumes within the pipe) The temperature of air along the pipe for three selected steps of simulation (step 130, step 193, and step 240) is depicted in Figure 3.26, Figure 3.27, and Figure 3.28. During the simulation, the following cases occurred: a) only condensation took place (step 130), b) only evaporation took place (step 240), and c) when both condensation and evaporation took place (step 193). - 53 - Chapter 3 [g/control volume] [°C] 30 Step 130 θmoist a θdry a 25 20 15 0 0.4 5 10 15 20 25 5 10 15 20 25 30 0.3 0.2 0.1 0 0 [m] 30 Figure 3.26: Temperature of air along the pipe and condensed amount in each control volume [°C] 30 Step 240 25 θmoist a θdry a 20 [g/control volume] 15 10 0 5 10 15 20 25 5 10 15 20 25 30 -0.25 -0.5 -0.75 -1 0 [m] 30 Figure 3.27: Temperature of air along the pipe and evaporated amount in each control volume [g/control volume] [°C] 30 Step 193 θmoist a 25 θdry a 20 15 0 5 10 15 20 25 5 10 15 20 25 0.1 0 -0.1 -0.2 -0.3 -0.4 0 30 [m] 30 Figure 3.28: Temperature of air along the pipe, condensed and evaporated amount in each control volume (condensation simultaneous with evaporation) - 54 - Model validation 3.3 Analytical validation 3.3.1 Analytical solution The analytical solution for cylindrical heat exchanger with external adiabatic boundary condition (Figure 3.29) and harmonic oscillation at the input was used for a comparison with numerical calculation. The analytical output was calculated using Excel routine designed by (Hollmuller, 2005). Figure 3.29: Cylindrical air-to-mass heat exchanger with adiabatic boundary The analytical solution for harmonic input with angular frequency ω θin (t ) = θ0 cos (ωt ) (3.11) is described by formula: ⎛ ⎛ Sh ⎞ x⎞ Sk ⎞ ⎟⎟ ⎟ cos ⎜⎜ ω ⎜ t − ⎟ − ⎝ ma ca ⎠ ⎝ ⎝ va ⎠ ma ca ⎠ ⎛ θ a ( x, t ) = θ 0 exp ⎜ − (3.12) where: S is heat exchange surface [m2] from the inlet to distance x, h is total (air/pipe + soil) amplitude-dampening exchange coefficient [W/(m2.K)], k is total (air/pipe + soil) phase-shifting exchange coefficient [W/(m2.K)], t is time [s]. Term x/va is time in which air flows from the inlet to distance x from the inlet (transit time). Details to derivation of the analytical solution and the calculation of both exchange coefficients can be found in (Hollmuller, 2003). 3.3.2 Simulation The comparative analysis was performed for three different setups of the cylindrical exchanger (Table 3.4). These setups cause different kinds of outlet air temperature modulation: - 55 - Chapter 3 annual dampening daily dampening annual phase-shifting In parallel, the choice of these setups provides a cross comparison between the developed model and respected numerical model which was examined by identical validation process (see Hollmuller, 2005). Setup Annual dampening Daily dampening Annual phase-shifting r0 [m] 0.125 0.125 0.125 rad [m] 2.0 0.6 0.6 L [m] 50 50 400 External boundary adiabatic adiabatic adiabatic Table 3.4: Setups of cylindrical heat exchanger The inlet air temperature was determined by standard annual meteorological data for Geneva and was fully decomposed into a complete sum of harmonic pulses (4380 frequencies) through Fourier series. The setups were submitted to constant air flow rate 162.5 m3/h (200 kg/h) which induced the value 4.13 W/(m2.K) of the convective heat transfer coefficient. Soil properties were considered to be λs = 1,9 W/(m.K), ρcp = 1,9 MJ/(m3.K). Two numerical calculation routines were used for the numerical simulation. Both routines use the identical basic calculation procedure described in chapter 2.3. They differ in the calculation of heat conduction in the cylinder. While the first routine uses one-dimensional implicit calculation of the heat conduction within radial coordinate r, the second routine is based on two-dimensional explicit calculation within rectangular control volumes. In the latter case, the pipe was approximated by an equivalent square with perimeter 4deq which equals to perimeter of the pipe 2πr0. The equivalent square for external adiabatic boundary was derived from the condition of identical surface areas: d ad ,eq = rad π (3.13) - 56 - Model validation r0 rad deq dad,eq Figure 3.30: Replacement of the cylinder by equivalent squares The mesh for 2D calculations was generated so it was possible to simulate all setups with time step of one hour. Although such time step implies rather coarse mesh (Table 3.5), utilization of this time step is intentional for studying the influence of the coarse grid on the accuracy of calculation. 1D radial model Annual dampening Daily dampening number of radial control 30 (∆r = 0,067 volumes m) number of longitudinal 25 (∆L = 2 m) 25 (∆L = 2 m) control volumes 2D model Annual dampening y 14 z 14 number of rectangular control volumes number of longitudinal 25 (∆L = 2 m) control volumes 0.1963 deq [m] dad,eqv [m] 3.5449 Annual phaseshifting 30 (∆r = 0,02m) Daily dampening 7 7 25+15 (the first 200 m ∆L = 8 m ) Annual phaseshifting 7 7 25 (∆L = 2 m) 50 (∆L = 8 m) 0.1963 0.1963 1.0635 1.0635 Table 3.5: Specification of simulations - 57 - Chapter 3 30 [°C] Annual dampening 1D radial model 20 20 15 10 analytical 3.3.3 Comparison analytical vs. numerical 10 0 5 -10 30 50 100 150 200 250 300 350 0 numerical 0 1D radial model [°C] Daily dampening 20 20 15 10 5 10 5 10 5 10 15 20 analytical 0 10 0 5 -10 0 30 50 100 150 200 250 300 [°C] Annual phase-shifting 350 numerical 0 1D radial model 15 20 analytical 0 20 20 15 10 10 0 5 numerical -10 0 50 Tinletmax 100 Tinletmin 150 num Toutletmax 200 250 num Toutletmin 300 anal Toutletmax 350 0 0 anal Toutletmin daily min 15 20 daily max Annual dampening from daily min 200 100 50 100 150 200 250 300 350 0 -0,3-0,2-0,1 0 0,1 0,2 0,3 Daily dampening from daily min 200 [1/year] from daily max 100 from daily max 50 100 150 200 250 300 350 0 -0,3-0,2-0,1 0 0,1 0,2 0,3 Annual phase-shifting 200 [1/year] 0.3 0.2 [°C] 0.1 0 -0.1 -0.2 -0.3 0 0.3 0.2 [°C] 0.1 0 -0.1 -0.2 -0.3 0 0.3 0.2 [°C] 0.1 0 -0.1 -0.2 -0.3 0 [1/year] Figure 3.31: 1D radial model – comparison with analytical; Left - comparison between numerical simulation (grey thick) and analytical solution (black thin) for different configurations of cylindrical heat exchanger with external adiabatic boundary conditions – daily values of maximal and minimal inlet and outlet temperatures; Right – scatter plot numerical vs. analytical (from daily maximal and minimal values) 100 50 100 150 200 250 300 350 0 -0,3-0,2-0,1 0 0,1 0,2 0,3 Figure 3.32: 1D radial model - comparison with analytical; Left – error propagation during the year, error is defined as: err = (numToutlet - analToutlet); Right – histogram of error’s frequency (grey – error calculated from daily maximums; black – error calculated from daily minimums) - 58 - [°C] Annual dampening 30 2D model analytical Model validation 20 20 15 10 10 0 5 numerical -10 30 60 90 120 150 180 210 240 270 300 360 2D model [°C] Daily dampening 30 330 0 0 5 10 5 10 5 10 15 20 analytical 0 20 20 15 10 10 0 5 numerical -10 30 60 90 120 150 180 210 240 270 300 [°C] Annual phase-shifting 30 330 360 2D model 0 0 20 20 15 10 15 20 analytical 0 10 0 5 numerical -10 0 30 60 Tinletmax 90 Tinletmin 120 150 180 num Toutletmax 210 240 num Toutletmin 270 300 anal Toutletmax 330 360 0 0 anal Toutletmin 15 daily min 20 daily max [°C] 100 50 from daily min 60 90 120 150 180 210 240 270 300 330 360 Daily dampening 0 -1 150 100 -0,5 0 0,5 1 -0,5 0 0,5 1 0 0,5 1 [1/year] 0 30 0.5 0.4 [°C] 0.3 0.2 0.1 0 -0.1 -0.2 -0.3 -0.4 -0.5 0 30 0.5 0.4 [ °C] 0.3 0.2 0.1 0 -0.1 -0.2 -0.3 -0.4 -0.5 0 30 Annual dampening from daily max 50 60 90 120 150 180 210 240 270 300 330 360 Annual phase-shifting 0 -1 100 50 60 90 120 150 180 210 240 270 300 330 360 0 -1 [1/year] 0.5 0.4 0.3 0.2 0.1 0 -0.1 -0.2 -0.3 -0.4 [1/year] Figure 3.33: 2D model; Left - comparison between numerical simulation (grey thick lines) and analytical solution (black thin lines) for different configurations of cylindrical heat exchanger with external adiabatic boundary conditions – daily values of maximal and minimal inlet and outlet temperatures; Right – scatter plot numerical vs. analytical (from daily maximal and minimal values) -0,5 Figure 3.34: 2D model - comparison with analytical; Left – error propagation during the year, error is defined as: err = (numToutlet - analToutlet); Right – histogram of error’s frequency (grey – error calculated from daily maximums; black – error calculated from daily minimums - 59 - Chapter 3 3.3.4 Summary The comparison between numerical simulation and analytical solution demonstrated the perfect agreement of numerical simulation with analytical model. Even if quite a coarse mesh was used in the 2D routine, high accuracy was achieved. This confirmed that the basic part of the algorithm (local heat transfer near pipe) works properly. The third setup of the exchanger (400 m pipe length and 0.60 m soil radius) resulted into an atypical performance. The outlet air temperature was dampened similarly to daily dampening setup (50 m length and 0.60 m soil radius), but the output signal was shifted much more. Probably, such strong phase-shifting was firstly introduced in (Hollmuller, 2003). The phenomenon also has more realistic daily mode (not so long tube) and it is intended to use it for cooling of buildings (Hollmuller, 2005). 3.4 Experimental validation The capability of the model will be demonstrated by its application on simulation of two monitored real-size EAHXs. First, the long-term thermal simulation will be compared with basic measurement on a low-energy house. The sensitivity analysis will be performed as a part of simulation. Second, the short-term hygro-thermal simulation will be compared with measured data on passive house in Rychnov. 3.4.1 Long-term thermal simulation Measurement in-situ A low-energy family house equipped with mechanical ventilation, heat recovery, and a simple earth-to-air heat exchanger has been monitored since the end of summer 2004. The description of the house is to be found in appendix A3. The scheme of the monitored system with placement of sensors is shown in Figure 3.35. Table 3.6 provides a basic description of the EAHX. - 60 - Model validation Figure 3.35: House W – the scheme of ventilation system with placement of sensors Number of pipes 1 Air flow rate [m3/(h.pipe)] 100 – 350, higher values for summer ventilation Length of pipe [m] 21 Soil Diameter [mm] 200 Control strategy Depth [m] 1,9 Location no data According to θa and Velké Popovice, link with actual mode central Bohemia, of the ventilation near Prague unit* Table 3.6: House W - description of the EAHX * In reality, the operation of the EAHX is not only the function of the ambient air temperature, but it also depends on the actual ventilation mode. The ventilation system in house W may be operated in five basic modes, see (Tywoniak et al, 2007). Simulation A calculation domain was a block 8 m (width) x 5 m (height) x L (length of the pipe). The domain was divided on 17 x 15 x 21 control volumes. The pipe was approximated by an equivalent square with perimeter which equals to perimeter of the pipe. The material of the pipe was neglected. The upper edge of the equivalent square had distance Zp from the upper side of the rectangular section (soil surface). The simulations were performed according to a schedule as shown in Table 3.7; the time step of the simulation was 1200 seconds. The influence of the inlet shaft was not simulated; the inlet air temperature was equal to measured ambient air temperature. The inlet air was assumed to be absolutely dry. A boundary condition for soil surface was set up by ambient air temperature, global solar radiation on a horizontal plane (assumed constant 80% absorptivity for solar radiation), convective surface thermal resistance (the influence of wind, assumed to be constant value 0,04 (m.2K)/W) and inserted - 61 - Chapter 3 additional thermal resistance Rs representing the influence of soil cover (vegetation, snow, determined by user defined function). Other walls of the rectangle were assumed to be adiabatic. Ini* start 30.8. 2004 125 1.1. 125 25.2. 125 7.3. 0 25.5. 350 1.6. 0 25.7. 350 finish 23.8. 2005 0 Table 3.7: The specification of the simulation * Ini denotes the initiation of the simulation. The initiation period is the simulation of two years without the influence of the EAHX. Thus, soil temperature field is influenced only by upper soil surface boundary; geothermal heat flow was omitted, too. The main purpose of the initiation is to build up the undisturbed soil temperature field as accurately as possible. The blue fields in Table 3.7 denote the intermittent operation of the EAHX; suction of air through the EAHX was started when ambient air fell below 0 °C, and disrupted when ambient air temperature exceeded 5 °C. The orange fields denote the identical EAHX operation as monitored. The number in the frame is the air flow rate in (m3/h). Sensitivity analysis Sensitivity analysis is the important part of the model validation. The analysis should help to identify input parameters to which the model outputs are particularly sensitive and to get parameters to which the outputs are not sensitive. It is especially important to identify sensitive input parameters with high uncertainty. The method of Differential Sensitivity Analysis (Lomas, Eppel, 1992) was used. The method is based on varying just one input parameter for each simulation while remaining inputs stay fixed at their most likely base case values (BC). The range of input parameter represents input uncertainty. The change in predicted output parameter ∆pi represents an effect of uncertainty in i - th input parameter: ∆pi = pi − pBC (3.14) First, the sensitivity of the annual soil temperature amplitude θsoilA (°C) to uncertainty in input parameters (parameters 1 to 6 in Table 3.8) was studied. The soil temperature is monitored in the first soil control volume above the equivalent pipe during the second year of the initiation. Such soil temperature represents undisturbed - 62 - Model validation soil temperature during a year in depth closed to the place where the EAHX is buried. Next, the sensitivity of the total energy Eeahx (kWh/a) delivered to and extracted from soil during the simulated period (30.8.2004 – 23.8.2005) to uncertainty in all input parameters (Table 3.8) was studied. The effect of spatial and time discretization was not studied; it was assumed that the grid was generated dense enough to produce correct results. 1 λs [W/mK] 2 ρcp [MJ/m3K] b -1.0 BC 1.5 a +1.0 note range of common soils -1.5 3.0 +1.5 range of common soils +0.2 3 θin [°C] -0.2 from measurement 4 Ig [W/m2] -3 % from measurement 5 Rs [m2K/W] -50 % UDF 6 Zp [m] -0.25 2.0 7 L 8 Va [m] [m3/h] 9 ha [W/m2K] -3.0 -25 -50 -25 % 21 125 350 calculation Nu measurement uncertainty +3 % measurement uncertainty +50 % UDF is user defined function, guess +0.25 pipe slope +3.0 +25 +50 from +25 % pipe length pre-heating guess cooling guess Table 3.8: Uncertainties in input parameters λs ρ cp ρcp as*10-6 bs [W/mK] 1.5 2.5 0.5 1.5 1.5 [kg/m3] 2000 2000 2000 2250 1500 [J/kgK3] 1500 1500 1500 2000 1000 [MJ/m3K1] 3.0 3.0 3.0 4.5 1.5 [m2/s] 0.50 0.83 0.17 0.33 1.00 [Ws0,5/m2K] 2121 2739 1225 2598 1500 SC1 2.5 2250 2000 4.5 0.56 3354 SC2 0.5 1500 1000 1.5 0.33 866 BC 1a 1b 2a 2b Table 3.9: Soil properties in simulation. SC denotes superior case Soil temperature calculated during the initiation of simulations is depicted in Figure 3.36. The effect of input uncertainties monitored on the value of the annual soil amplitude (4.7 °C for BC) is illustrated in Figure 3.37, left. Uncertainties assumed in soil properties (Table 3.9) have the strongest impact; the undisturbed soil temperature - 63 - Chapter 3 strongly depends on the value of the soil thermal diffusivity and resulting upper soil surface temperature. 25 θsoil [°C] 20 15 10 5 30.8.2004 0 23.8.2005 30 60 BC 90 1a 120 1b 150 2a 180 210 t [days] 2b 6a 240 270 300 330 6b 4o 5o 6o 357 Figure 3.36: Soil temperatures θsoil (°C) calculated during initiation of the simulation; 4o is BC without solar radiation; 5o is BC with Rs assumed to be a constant value 1.0 m2KW-1; 6o is BC with pipe buried in depth 0.5 m a b +2.0 +1.5 +20 ∆Eeahx [kWh/a] +0.5 4.9 a ∆θsoil [°C] +1.0 -0.5 -1 451 -20 -40 -60 -80 -1.5 -2 a b +40 -100 λs 1 ρ cp θin Ig Rs Zp 2 3 4 5 [input parameters] 6 λ s ρ cp θin Ig Rs Zp L 1 2 3 4 5 6 7 [input parameters] Va ha 8 9 Figure 3.37: The effect of uncertainties in inputs monitored on the annual soil temperature amplitude θasoil (left) and total energy delivered to or extracted from soil (right) The effect of input parameters uncertainties monitored on the value of total energy injected and extracted from the soil (451 kWh for BC) is shown in Figure 3.37, right and in Figure 3.38. The positive and negative changes to the input parameters do not affect the predicted value Eeahx by similar amount. The model (i.e. the earth-to-air heat exchanger) behaves as a nonlinear system. The reaction of the model to change of soil properties is strongly non-linear. The reaction of the model to change of length, air flow rate, and depth is weakly non-linear. - 64 - Model validation Although the undisturbed soil temperature is the most stable in case 1b (the lowest value of the soil thermal diffusivity), this particular advantage does not overweight poor thermal transport near the pipe. The performance of cases 2a and 1a need another comment. Although the value of the soil thermal conductivity of case 2a is lower than in case 1a, case 2a shows the overall thermal performance better than case 1a. Both cases have similar soil thermal effusivity; however, case 2a has the value of the soil thermal diffusivity much lower than case 1a. Thus, the undisturbed soil temperature of case 2a is more dampened than in case 1a, and this particular advantage is probably the most decisive. Conclusively, a relatively thin cylinder surrounding the pipe having high thermal conductivity combined with semi-infinite body having low thermal diffusivity could lead to some improvement of EAHX. 500 [kWh/year] 400 300 200 100 0 BC 1a 1b 2a 2b SC1 SC2 Figure 3.38: House W - the effect of uncertainties in soil properties monitored on the value of total energy injected and extracted from the soil (451 kWh for BC) The non-linear reaction of the model to uncertainties in the inlet air temperature (cases 3a, 3b) is not caused by non-linear effects of the model. The temperature of the inlet air is a parameter which governed whether the EAHX was turned on, or turned off. As a result, the different operation patterns were executed, and consequently, the different portion of heat was extracted from and injected to soil. Comparison simulation vs. measurement Several measurement data can be used for a comparison with the model prediction. The following time intervals with characteristic pulses (intermittent operation of EAHX) were chosen in order to compare the simulation with measured data: a) 25.2. – 6.3.2005, and b) 25.7. – 31.7.2005. The first interval is the last term of - 65 - Chapter 3 air pre-heating during winter 2004/2005, and the second interval is the second hot spell of summer 2005 (air cooling). The comparison of measured outlet air temperature with simulated values is in Figure 3.39 and in Figure 3.40. The results of variants with the most significant uncertainties are depicted too. Those variants form the interval in which all simulation results were found. 8b 5 θ (°C) 0 1b -5 -10 25.2 26.2 27.2 28.2 θ in 1.3 θ 2.3 out 3.3 BC 4.3 1b 5.3 6.3 8b Figure 3.39: 25.2. – 6.3.2005; measured inlet (θin) and outlet (θout) air temperatures vs. simulated outlet air temperature (BC, 1b, 8b); no suction if θin = θout θ (°C) 30 1b 25 7a 20 15 25.7 26.7 27.7 θ in 28.7 θ out 29.7 BC 30.7 1b 31.7 7a Figure 3.40: 25.7. – 31.7.2005; measured inlet (θin) and outlet (θout) air temperatures vs. simulated outlet air temperature; no suction if θin = θout - 66 - Model validation Continuous mode The continuous mode of the operation (suction over all year) is studied with different values of air flow rate (125 and 350 m3/h). The influence of the continuous operation of the EAHX on the soil temperature may be significant, especially during winter (Figure 3.41). If the inlet air temperature is lower (cooling season) or higher (pre-heating season) than actual soil temperature, the recovery of soil induced by air flow rate (forced soil thermal recovery) takes place. The forced soil recovery is more common during summer. The soil near the pipe is cooled down by suction of colder night air or during colder spells in summer. Therefore, the soil temperature remained oscillating roughly around the undisturbed soil temperature profile in summer season. 20 θ (°C) 15 10 5 0 30.8.2004 30 23.8.2005 60 90 120 θ 150 180 (BC) undisturbed 210 θ350 soil 240 270 θ125 300 330 357 soil Figure 3.41: Continuous operation - soil temperature in the reference point; θundisturbed(BC) is temperature calculated during initiation of BC simulation, θsoil350 is soil temperature in the reference point - air flow rate 350 m3/h, θsoil125 is soil temperature in the reference point - air flow rate 125 m3/h; the reference point is located 0.5 m from the inlet in depth of 1.9 m – the first control volume above the pipe Intermittent mode The process of natural soil recovery in intermittent mode is quite rapid, especially immediately after turning the exchanger off (Figure 3.42). The effect of such recovery is surprisingly comparable with the effect of recovery induced by air flow. Roughly, majority of a thermal pulse is removed in period four times longer than the length of pulse. During spring or autumn, when no air is sucked through EAHX, the soil temperature is progressively restored to undisturbed profile, as shown in Figure 3.42 (seasonal heat storage is ineffective). - 67 - Chapter 3 20 θ (°C) 15 10 5 0 30.8.2004 30 23.8.2005 60 90 120 θ 150 180 210 θ (BC) (BC) undisturbed soil 240 270 on/off 300 330 357 Figure 3.42: Intermittent operation - soil temperature in the reference point; θundisturbed(BC) is temperature calculated during initiation of BC simulation, θsoil(BC) is soil temperature in the reference point (BC simulation, intermittent operation according to Table 3.7);the reference point is located 0.5 m from the inlet in depth of 1.9 m – the first control volume above the pipe The example with more significant thermal charge of soil (no suction within the interval of ambient air temperature between 5 °C and 19 °C) illustrates previous conclusion (Figure 3.43). 20 θ (°C) 15 10 5 30.8.2004 0 30 23.8.2005 60 90 120 θ 150 180 210 θ (BC) undisturbed soil 240 270 on/off 300 330 357 Figure 3.43: Intermittent operation, more significant thermal charge - soil temperature in the reference point; θundisturbed(BC) is temperature calculated during initiation of BC simulation Conclusions On the basis of presented results, the following conclusions were formulated: - 68 - Model validation Because of relatively low air-to-soil heat flows and the intermittent pattern of operation, the accurate simulation of the EAHX is rather difficult. The relative error of the model prediction can be high. The quality of the model prediction is dependent on the accurate estimation of the natural thermal stratification in soil during the initiation of the simulation. The natural thermal performance of shallow subsurface governs the long-term thermal performance of the EAHX. The operation of the EAHX disturbs the natural thermal stratification of soil and governs the short-term thermal performance of EAHX (heat injection/extraction, natural soil recovery). The accuracy of the numerical model was rather destroyed by uncertainties in several input parameters (primarily soil properties and the air flow rate which are almost always uncertain). Even more simplified models might be suitable for the simulation of the EAHX. Perhaps, they will not be so sensitive to input parameters uncertainties. Despite of their simplicity, they could be sufficiently accurate. The numerical model brings clear information about processes which take place during the operation of EAHX and allows rather detailed analysis. The perfect fit between measured data and simulation results is not from this point of view so important. Rather good accuracy of the numerical simulation was achieved, as shown in comparison with measured data. However, such comparison is not considered to be a sufficient experimental validation of the model because the air flow rate was not directly measured (the value in simulation of BC was tuned). Summary The long-term simulation of a simple real-size EAHX was presented in the chapter. Measured data of ambient air temperature, solar radiation and EAHX operation served as input parameters to series of simulations. First, the simple differential sensitivity analysis of the model was performed. Next, the comparison of the measured data with simulation was done. Then, the thermal charge of soil as a result of different operation modes (continuous, intermittent mode) was discussed. Finally, principal conclusions regarding to the accuracy of performed numerical simulation were presented. - 69 - Chapter 3 3.4.2 Short-term hygro-thermal simulation Measurement in-situ For details about measurement see chapter 4 and also appendix A4. Simulation The simulation performed with the developed numerical model focuses on hygro-thermal performance of the open loop mode during a) ten-day summer term from 13.6.2006 to 22.6.2006 (term 1) and b) three-day summer term from 15.7.2007 to 17.7.2007 (term 2). The terms are typical by very high ambient air temperature and intensive EAHX operation during daytime periods. The inlet air temperature and relative humidity were equal to measured values in the inlet shaft (position 2, Figure 4.1). One pipe of the exchanger only was simulated. The thermal influence of the neighbouring pipe was not included into the simulation. The time step used for the simulation was 5 minutes. Although moisture flows were calculated, generation of latent heat was not considered. Term 1: A calculation domain was a soil block 1.18 m (width) x 1.18 m (height) x 23 m (length of the pipe) with external adiabatic walls. The block was divided into 17 x 17 x 46 control volumes. The initial soil temperature was assumed 8.5 °C in all control volumes of the block (the value was tuned). Term 2: A calculation domain was a soil block 2 m (width) x 5 m (height) x 23 m (length of the pipe). The block was divided into 20 x 38 x 46 control volumes. The bottom and vertical boundaries were assumed to be adiabatic. The upper boundary was formed by measured values of soil temperature in depth of 1 m. The initial soil temperature matrix was built up by the numerical simulation (the initiation of the simulation). Comparison simulation vs. measurement The output parameters from the short-term simulation were compared with measured parameters. The correspondence of the model prediction with measurement is obvious (Figure 3.44 to Figure 3.49). - 70 - Model validation θout [°C] relative frequency [%] 17 simulated 16 15 14 13 12 11 12.5 10 7.5 5 2.5 11 12 13 14 15 16 17 measured -2 -1 0 +1 +2 numerical-measured Figure 3.44: Term 1 - outlet air temperature θout. Left - scatter plot measured vs. simulated. Right – relative frequency of difference between simulated and measured values 3 ρ v,out [g/m ] relative frequency [%] 14 simulated 13 12 11 10 9 8 8 12.5 10 7.5 5 2.5 9 10 11 12 13 14 measured -1.5 -1 -0.5 0 +0.5+1+1.5 numerical-measured Figure 3.45: Term 1 - outlet air water vapor concentration ρv,out. Left - scatter plot measured vs. simulated. Right - relative frequency of difference between simulated and measured values dG [g/(min.pipe)] relative frequency [%] 15 simulated 10 5 0 -5 -5 0 5 10 measured 12.5 10 7.5 5 2.5 -4 -3 -2 -1 0 +1+2+3+4 numerical-measured 15 Figure 3.46: Term 1 - overall moisture rate dG. Left - scatter plot measured vs. simulated. Right - relative frequency of difference between simulated and measured values - 71 - Chapter 3 20 19 18 17 16 15 14 13 12 11 relative frequency [%] simulated θout [°C] 12.5 10 7.5 5 2.5 11 12 13 14 15 16 17 18 19 20 measured -2 -1 0 +1 +2 numerical-measured Figure 3.47: Term 2 - outlet air temperature θout. Left - scatter plot measured vs. simulated. Right – relative frequency of difference between simulated and measured values 3 15 14 13 12 11 10 9 8 relative frequency [%] simulated ρ v,out [g/m ] 8 12.5 10 7.5 5 2.5 9 10 11 12 13 14 15 measured -1.5 -1 -0.5 0 +0.5+1+1.5 numerical-measured Figure 3.48: Term 2 - outlet air water vapor concentration ρv,out. Left - scatter plot measured vs. simulated. Right - relative frequency of difference between simulated and measured values dG [g/(min.pipe)] relative frequency [%] 15 simulated 10 5 0 -5 -5 0 5 10 measured 12.5 10 7.5 5 2.5 15 -4 -3 -2 -1 0 +1+2+3+4 numerical-measured Figure 3.49: Term 2 - overall moisture rate dG. Left - scatter plot measured vs. simulated. Right - relative frequency of difference between simulated and measured values - 72 - Model validation Conclusions On the basis of the presented results, the following conclusions were formulated: The accuracy of moisture calculations is very dependent on the accurate calculation of heat transfer. Although the absolutely perfect agreement was not achieved, the model shows good consistency (the trends are similar). The influence of latent heat transfer due to condensation and/or evaporation inside the pipe is still unclear. As shown by both measurement and simulation, condensation and consequent evaporation evidently occur inside the pipe. However, the measured outlet air temperature tended to be lower than simulated when latent heat was taken into account. Summary The short-term simulation of a simple real-size EAHX was presented in the chapter. Measured data of inlet air temperature, soil temperature and EAHX operation served as input parameters to series of simulations. The comparison between simulation and measurement serves as an experimental validation of the model. 3.5 Parametric analysis The hygro-thermal performance of the validated model is evaluated by a parametric analysis. The method is based on varying just one input parameter (see Table 3.10) for each simulation while remaining inputs stay fixed (at so called base case). The analysis should help to identify input parameters to which the model outputs are sensitive and demonstrate the model reaction on change of input parameters. 3.5.1 Simulation A simple (and therefore predictable) basic situation, the step-change of inlet air temperature or humidity, was simulated. At time zero, there is a sudden change of the inlet air temperature (θin) or inlet air humidity (ρv,in). A calculation domain was a block (1 + deq) m x (1 + deq) m x L m (length of the pipe). All external walls of the block were assumed to be adiabatic. The block was assumed to have thermal conductivity λs and - 73 - Chapter 3 volumetric heat capacity ρcp. The initial temperature of the block (θini) was constant. The pipe (with wall thickness tp and thermal conductivity λp) was approximated by an equivalent square with perimeter which equals to perimeter of the pipe. The time step of all simulations was 300 seconds. Two types of simulation were performed: Latent heat was not considered. Although moisture flows were calculated, generation of latent heat was not considered. Latent heat was considered. Generation of latent heat affects the thermal performance of EAHX. Because of latent heat release, condensed amount is lower (reverse link) compared to the latter type of simulation. Figure 3.50 shows the typical situation during five days (simulation period) after initial step-change (see also chapter 3.2.2). Initially, water vapour condensation prevails, but the pipe and surrounding soil are consequently warmed up. Next, evaporation prevails over condensation (“evap over cond”) until condensation fully diminishes (“cond stop”). Finally, all water in tube evaporates (“evap stop”). 3000 2000 ↑ Gacu,cond evap over cond cond stop [g] 1000 ← Gacu evap stop 0 -1000 ← Gacu,evap -2000 -3000 0 1 2 3 4 5 [days] Figure 3.50: Total accumulated water Gacu [g] in the pipe Gacu = Gacu ,cond + Gacu ,evap (3.15) where: Gacu,cond is cumulative total condensed amount of water vapour [kg], Gacu,evap is cumulative total evaporated amount of water vapour [kg] and Gacu is total accumulated water in the pipe [kg]. - 74 - Model validation 3.5.2 Input parameters The range of climate input parameters (θin, ρv,in, θini) covers likely situation in summer. For instance, the range of inlet water vapour concentration could be typical in summer for climate of middle Europe (Czech Republic, Germany, Austria, etc.). The range of initial block temperature covers summer soil temperature increase typical for depth of two meters. The range of soil input parameters (λs, ρcp) covers all soil types (lower values for dry soils, higher for moist soils, for details see appendix A5). The range of exchanger key design parameters (air flow rate Va, length L) covers some design alternatives, i.e. from undersized (L = 10 m, Va = 400 m3/h) to slightly oversized alternatives (L = 40 m, Va = 100 m3/h). The range of two remaining design parameters (diameter d0, thermal conductivity of pipe material λp) was chosen so as to cover standard diameters and materials (thermal conductivity of PVC is approximately 0,10 W/(m.K)). exchanger 2 climate inlet soil 1 3 4 5 6 7 8 9 λs ρcp θini θin ρv,in d0 Va L λp 10 tp unit [W/(m.K)] lower -67 % -1,0 BC 1,5 upper +1,0 + 67 % step 0,25 [MJ/(m3.K)] -50 % -1,0 2,0 +1,0 +50 % 0,5 [°C] -25 % -3,0 12,0 +3,0 +25 % 1,0 [°C] -20 % -5,0 25 +5,0 +20 % 2,5 [g/m3] -11 % -1,5 14,0 +1,5 +11 % 0,5 [m] [m3/h] [m] [W/(m.K)] -25 % 0,15 -60 % -150 -60 % -15 -67 % -0,1 0,20 250 25 0,15 0,25 +150 +15 +0,1 +25 % +60 % +60 % +67 % 0,05 50 5 0,05 [m] - - 0,005 +0,005 +100 % 0,005 Table 3.10: Perturbation of input parameters; BC denotes base case 3.5.3 Monitored outputs The following outputs from simulation were monitored: Energy injected through exchange surface A to surrounding soil block per period of simulation in [kWh/period] Eeahx = ∫ Qeahx dt (3.16) Specific energy injected to surrounding soil block (related to exchange surface A) per period of simulation in [kWh/(m2.period)] - 75 - Chapter 3 A = eeahx Eeahx A (3.17) Total condensed amount per period of simulation in [kg/period] Gcond ,tot = max ( Gacu ,cond ) (3.18) 3.5.4 Results The values of monitored inputs and outputs were relativized (related to BC simulation) in order to allow direct comparison. Both sensitivity and non-linearity may be evaluated. Sensitivity is defined as the first order derivative of sensitivity curve (the steeper curve, the higher sensitivity). Non-linearity is defined as curvature of sensitivity curve. The effect of perturbation of input parameters on the value of energy injected to rel. change in output parameter [%] the surrounding soil block is shown in Figure 3.51. +50 λs ρcp +25 θini θin 0 ρv,in d0 L Va -25 -50 λp -50 -25 0 +25 +50 rel. change in input parameter [%] tp BC Figure 3.51: Eeahx – parametric study. Dotted curves denote the cases when latent heat was taken into account The most sensitive input parameters are inlet air temperature and actual soil temperature (initial temperature of simulation). These parameters are influenced by climatic locality and therefore cannot be designed. Length and air flow rate are the most sensitive design parameters. The remaining design parameters (diameter, thermal conductivity of pipe wall, soil properties) are less sensitive. For instance, the effect of pipe thermal conductivity difference is rather small and almost negligible (keeping in mind low pipe wall thickness for BC). - 76 - Model validation The effect of perturbation of input parameters on the value of specific energy injected to the surrounding soil block is shown in Figure 3.52. The relation to the exchange surface was used because the value of specific energy evaluates the design effectiveness (i.e. taking into account the material spent on). As seen in Figure 3.52, rel. change in output parameter [%] length and diameter perform inversely compared to the latter case (Figure 3.51). +50 λs ρcp +25 θini θin 0 ρv,in d0 L Va -25 -50 Figure 3.52: λp -50 -25 0 +25 +50 rel. change in input parameter [%] eAeahx tp BC – parametric study. Dotted curves - latent heat taken into account The parametric analysis also identified input parameters to which condensed amount (moisture performance) of EAHX is particularly sensitive. The effect of perturbation of input parameters on the value of total condensed amount is shown in rel. change in output parameter [%] Figure 3.53. λs +700 ρcp +600 θini +500 +400 θin +300 ρv,in +200 d0 +100 L Va 0 -50 λp -50 -25 0 +25 +50 rel. change in input parameter [%] tp BC Figure 3.53: Gcond,tot – parametric study. Dotted curves - latent heat taken into account - 77 - Chapter 3 The moisture performance is much more sensitive to change of input parameter than thermal performance. The most sensitive parameters, which dominantly influence moisture performance, are instantaneous soil temperature (initial temperature of simulation) and temperature and humidity of inlet air. Therefore, risk of water vapour condensation is higher during spring when soil is cooled down after winter season. Low inlet air temperature leads to lower outlet air temperature and thus to higher risk of condensation. High air humidity is also a very important factor, especially simultaneously with low inlet air temperature. 3.5.5 Conclusions Because EAHX performance is nonlinear, it is easier to deteriorate the performance (inherently) than to improve. Moisture performance of EAHX is inherently linked with thermal performance, i.e. air temperature drop (saturation limit) is related to instantaneous humidity of inlet air. Daily oscillation of ambient air temperature and humidity, instantaneous soil temperature (link to annual soil temperature oscillation and EAHX previous operation), configuration of the pipe (diameter, length, air flow rate), and thermal properties of soil near pipe plays an important role role. It seems that condensation inside EAHX might occur rather frequently in climate of middle Europe and probably can not be absolutely eliminated by EAHX design. Therefore, the inspection of real size EAHXs with a camera and some microbial investigations would be very valuable. Mould growth inside pipes should be still questionable, although the several existing studies, e. g. (Fluckiger, 1999), have not confirmed any hygienic problems. 3.6 Summary The chapter dealt with model validation. First, some verification exercises were performed. The aim of these tests was to validate two basic parts of the model: air-toground heat transfer near the pipe and coupling with the ground surface. Then, the model predictions were compared with: the analytical solution for cylindrical heat exchanger the measured data on two monitored real-size EAHXs - 78 - Model validation The model showed perfect agreement with the analytical solution and satisfactory agreement with experimental data. Finally, the validated and verified model was submitted to series of simulation leading to straightforward parametric analysis. - 79 - Chapter 3 - 80 - Measurements in situ 4 Measurements in situ The aim of this chapter is the evaluation of some measured data collected during year 2006 and 2007 and subsequent generalization. 4.1 Introduction The passive family house (see appendix A4) ventilated by mechanical ventilation equipped with heat recovery and a simple earth-to-air heat exchanger is being monitored since summer 2005. The scheme of the ventilation system is displayed in Figure 4.1. Table 4.1 provides a basic description of EAHX. The extensive monitoring is primarily aimed at operating the ventilation system linked with EAHX. Measured data are collected in the main data logger (measuring step 1 min) and three other independent data loggers (measuring step 5 min). The first one is placed directly in the inlet shaft (position 2 in Figure 1); the second one is a living room data logger collecting parameters of indoor air (temperature, relative humidity, and CO2 concentration). The third data logger collects undisturbed soil temperature from several depths under the surface (5, 30, 62, 105 cm). For the list of sensors see appendix A4. Figure 4.1: The scheme of ventilation system with position of sensors; the closed loop mode of EAHX (circulation of internal air through the EAHX) - 81 - Chapter 4 Number of pipes 2 Air flow rate [m3/h] 115 – 410, higher values for summer ventilation Length of pipe [m] 23 Soil Clay* λs = 1,39 W/(m.K) ρcp = 2,23 MJ/(m3.K) Diameter [mm] 200 Control strategy According to ambient air temperature and link with the actual mode of the ventilation unit Depth [m] 1.0 and 2.0 Place Rychnov near Jablonec nad Nisou, North Bohemia Table 4.1: The description of EAHX; *Homogenous moist clay; the thermal properties are based on samples taken from a borehole drilled to depth of one meter The mechanical ventilation system offers five ventilation modes divided into two air flow rate levels and three special modes switched on by external signals (e.g. by usage of W.C., bathroom, cooking in a kitchen). The more detailed description of the ventilation system is presented in (Tywoniak et al, 2007). EAHX may be operated in two principally different modes: Circulation of air between the ventilated zone and the EAHX (closed loop mode, experimentally used for cooling, see Figure 4.2, left). Direct suction of air through the EAHX (open loop mode, usual option used for air pre-heating and cooling, Figure 4.2, right). The closed loop mode of the exchanger was experimentally installed by placing a removable elbow (from 11.7.2006 to 10.9.2006). elbow datalogger Figure 4.2: The inlet shaft; Left – the closed loop mode with a removable elbow; Right – open loop mode The short step of measurement allows the accurate determination of the ventilation mode. Therefore, it allows the precise determination of time intervals when EAHX was in operation and the determination of corresponding value of the air flow - 82 - Measurements in situ rate. The monitoring of air relative humidity also allows the determination whether air flowing through the pipe was moistened (the moisture deficit indicates that condensation within the pipe prevailed). 4.2 Soil 4.2.1 Thermal properties The three soil samples from different depths were taken during installment of temperature sensors in the borehole (18.12.2006). Figure 4.3: Installment of soil temperature sensors in the borehole The thermal parameters of samples were measured by a portable heat transfer analyzer ISOMET. The results are shown in Table 4.2. It is recognizable that the value of thermal diffusivity slightly increases with depth. This should be caused by higher porosity of upper soil layers. sample depth 30 cm measurement 1 measurement 2 50 cm measurement 1 II. measurement 2 95 cm measurement 1 III. measurement 2 mean values I. λ [W/(m.K)] ρcp [MJ/(m3.K)] 1,26 1,43 2,29 1,23 1,61 1,41 1,39 2,45 2,26 2,18 2,06 2,33 2,06 2,23 as [m2/s] 0,51*10-6 0,63*10-6 1,05*10-6 0,59*10-6 0,69*10-6 0,68*10-6 0,62*10-6 Table 4.2: Thermal properties of soil samples; the first measurement of sample 2 was excluded from calculation of the mean values - 83 - Chapter 4 4.2.2 Soil temperature 30 25 [°C] 20 θs2.0m 15 10 5 0 -5 0 30 60 90 θa 120 150 0.05m θs 180 0.3m θs 210 240 270 0.6m θs 300 330 360 1.0m θs Figure 4.4: Soil temperatures in 2007 – depths 0.05, 0.3, 0.6, 1.0 m are measured values. Temperature in depth of 2 m was calculated using measured temperature in depth of 0.3, 0.6, 1.0 m (grey stripe represents uncertainty of calculation). 4.2.3 Approximate calculation of soil thermal diffusivity The soil thermal diffusivity was also calculated by comparison of daily amplitudes between measured temperature signals in depth of 5 cm and 30 cm. The temperatures from March 2007 to May 2007 (Figure 4.5) were used for the analysis. The soil thermal diffusivity calculated by this method is shown in Figure 4.6. θ θ [°C] 20 θ 5 cm 30 cm 15 10 5 0 1.3.07 0 1.4.07 31 [days] 1.5.07 61 Figure 4.5: Soil temperatures measured from 1.3.2007 to 1.5.2007 - 84 - 92 Measurements in situ -6 x 10 as mean as [m2/s] 1 0.75 0.5 0 0 1.3.07 1.4.07 31 1.5.07 61 [days] 92 Figure 4.6: Soil thermal diffusivity as calculated from amplitude damping 4.3 External environment 4.3.1 Ambient air temperature The daily and night ambient air temperature overlaps (Figure 4.7 and Figure 4.8) illustratively show three hot spells during summer. The overlaps clearly illustrate the potential duration of air cooling in the EAHX which might occur for up to 10 hours assuming cooling set point 24 °C. 12 10 6 - 18 h >24 °C >28 °C >32 °C [hours] 8 6 4 2 V. VI. VII. VIII. IX. Figure 4.7: Daily temperature overlaps (6 – 18 h) over the period (V. – IX.2006) The night temperature overlaps show up to 4 hours of possible air cooling assuming cooling set point 24 °C. On the other side, ambient air temperature below 16 °C is rare during hot nights so that the potential of direct night cooling (i.e. open window) is quite low in hot days. - 85 - Chapter 4 12 18 - 6 h >16 °C >20 °C >24 °C 10 [hours] 8 6 4 2 V. VI. VII. VIII. IX. Figure 4.8: Night temperature overlaps (18 – 6 h) over the term (V. – IX.2006) Assuming pre-heating set point temperature 0 °C, the occurrence of ambient air temperature lower than pre-heating set point leads to approximately 1500 hours suitable for air pre-heating. Assuming cooling set point temperature 24 °C, air cooling would be useful for approximately 500 hours (see Figure 4.9). 30 [°C] 20 Pre-heating 10 potential ~ 1500 h Cooling potential ~ 500 h 0 -10 θ sorted a -20 cooling setpoint pre-heating setpoint 10 20 30 40 50 60 70 duration [%], 100 % = 8220 h 80 90 Figure 4.9: Sorted ambient air temperature (from 2006) 4.3.2 Ambient air water vapour concentration The daily overlaps of ambient air water vapour concentration shows that the absolute humidity of air is the highest during hot spells (i.e. in time when cooling of air in EAHX might be needed). - 86 - Measurements in situ 12 6 - 18 h 3 >10 g/m 10 3 >12 g/m 3 [hours] 8 >14 g/m 6 4 2 V. VI. VII. VIII. IX. Figure 4.10: Daily ambient air water vapour concentration overlaps (6 – 18 h) over the term (V. – IX.2006) 4.3.3 Tendency of EAHX to condensation Due to damping of soil mass, soil temperature is substantially lower than ambient air temperature in summer. Simultaneously, absolute humidity of air is the highest during summer. The tendency to condensation was evaluated with measured data, see Figure 4.11 and Figure 4.12. 17.5 15 ρv,a ρv,sat(θs,depth) ↓ 1.0m 3 [g/m ] 12.5 10 7.5 5 ↑ 2.0m ← daily mean ρv,a > ρv,sat(θs,2.0m) 2.5 V. VI. VII. VIII. 07 07 [V. - IX. ] IX. Figure 4.11: Water vapour concentration of ambient air ρv,a and saturated water vapour concentration ρv,sat as a function of soil temperature θs,depth, measured data in 2007 Condition ρv,a > ρv,sat occurred roughly 1300 h (~54 days) for depth of 2 m, and 613 h (~25 days) for depth of 1 m. As seen in Figure 4.12, evaporation (negative deficit) cannot frequently occur during hot days of summer term (condition θa > 24 °C). It means that during hot spells (air is sucked through EAHX) moisture will tend to accumulate in the exchanger. - 87 - Chapter 4 relative frequency [%] 12.5 θa>24°C 10 7.5 5 2.5 -7-6-5-4-3-2-1 0 1 2 3 4 5 6 7 3 deficit [g/m ] Figure 4.12: Relative frequency of difference between water vapour concentration of ambient air and saturated water vapour concentration for soil temperature in depth of 2 m (measured data May – September 2007), deficit = ρv,a - ρv,sat(θs,2.0m) 4.4 Earth-to-air heat exchanger 4.4.1 Example of measured data The example of time daily profiles measured on EAHX is depicted in Figure 4.13. The left column shows day 20.6.2006 with open loop mode of EAHX, the right column shows day 19.7.2006 with closed loop mode of EAHX. The performance of the open loop mode (Figure 4.13, left column) shows a quite significant (and surprising) cooling effect of the inlet shaft itself; the response of the outlet air temperature on the change of air flow rate is also visible, but not so evident. Since the value of the convective heat transfer coefficient is only slightly non-linear function of the air velocity (in the expected range of air flow rates), the cooling effect in case of doubled value of air flow rate is, from the short-term perspective, only slightly reduced (see Figure 4.13, left column). The difference between the inlet air water vapour concentration and outlet air water vapour concentration indicates prevailing condensation within pipes. The amount of condensed water is in order of several kilograms per day. Because of the lower soil temperature and consequent higher temperature drop of outlet air, the pipe buried in depth of two meters tends to stronger condensation than the upper pipe. - 88 - Measurements in situ θ 25 θ θ 30 shaft,2 out,3 25 out,1 θ [°C] θ [°C] 30 θ θ ambient 20 15 θ θ θ ambient shaft,2 out,3 in,1 20 15 200 m3/(h.pipe) 111 m /(h.pipe) 401 m3/(h.pipe) 221 m3/(h.pipe) 3 0 8 12 [hours] 16 20 24 0 4 v,ambient 12 [hours] 16 20 24 8 12 [hours] 16 20 24 8 12 [hours] 16 20 24 v,ambient ρ 13 ρ 16 ρ 12 ρ 15 ρ v,shaft,2 v,out,3 v,out,1 14 13 11 v,shaft,2 v,out,3 ρ v,in,1 10 9 12 8 11 0 8 ρ ρ ρ [g/m3] v ρ [g/m3] v 17 4 4 8 12 [hours] 16 20 24 0 4 14 -2 -4 10 dG [g/min] dG [g/min] 12 8 6 -6 -8 -10 4 2 0 dG3 -12 dG1 -14 4 8 12 [hours] 16 20 24 0 dG(1+3) 4 Figure 4.13: The measured temperature θ, water vapour concentration ρv and corresponding overall moisture rate dG during two selected days; the overall moisture rate is the sum of condensation and evaporation rate within the whole pipe (positive value denotes prevailing condensation within the pipe); the left column represents day 20.6.2006 (open loop mode), the right column represents day 19.7.2006 (closed loop mode), the values measured by logger placed in the removable connection elbow (position 2) are influenced by unwanted air suction due to the leaky wall of elbow (e.g. see measured profile ρv,shaft,2) The example of closed loop mode (Figure 4.13, right column) leads to totally different moisture performance than the open loop mode. Although the outlet air temperature is similar to the case of open loop mode, there is a negative moisture - 89 - Chapter 4 balance between inlet and outlet. The difference between inlet air (the air sucked from the building) and outlet air water vapour concentration indicates prevailing evaporation within pipes. The moisture present in the pipes is likely the result of previous open loop mode operation (prevailing condensation). The movement of water towards the inlet shaft is probably quite slow due to the surface tension so that some water stays inside the pipes and only a limited amount of water is drained into the inlet shaft. The rate of water uptake (evaporation) is lower in comparison to condensation rate. Probably, this might occurs due to the active area which is limited to thin stripe of water at the bottom of the pipe. 4.4.2 EAHX operation EAHX operation pattern is important for understanding time scales of thermal process. The pattern is rather important for a correct choice of the simulation time step. The operation pattern of EAHX is driven by the following: Ambient air temperature drives the position of EAHX damper (before entering air handling unit). An algorithm of control system drives the ventilation mode (see appendix A4, Table A5). Two typical days of EAHX operation based on the monitoring of the family house in Rychnov are depicted in Figure 4.14 (14.1.2007, pre-heating mode) and in Figure 4.15 (20.6.2007, cooling mode). Pre-heating mode Generally, there is a tendency to ventilate as less as possible with respect to the desired reduction of ventilation heat loss and the acceptable level of air internal relative humidity during winter time (not to extract too much moisture from the internal air). The ventilation, however, should be sufficient respecting necessary supply of fresh air. Therefore, the typical operation of EAHX in pre-heating mode is very intermittent. The vertical stripes represent the feature of automatic cyclic ventilation (order of minutes per hour) of the house in the night. The EAHX operation visible near the noon is the ventilation induced by preparation of lunch. The evening operation represents the ventilation due to presence of people in the house. - 90 - Measurements in situ The average air change rate (over 14.1.2007) was 30 m3/h (0,12 1/h) only. The air flow rate seems rather low. However, assuming certain value of air infiltration and intermittent occupation of the house (4 persons, each 12 hours per day) the air flow rate related to one person is 20 m3/h approximately. 1 0 0 4 8 12 "14.1.2006" 16 20 24 Figure 4.14: Intermittent pattern of operation typical during heating period The intermittency of the operation could cause some problems to simulation. If one wanted to be correct, it would be necessary to use very short time step of the simulation (approximately 5 min for simulation of pre-heating mode). However, the shortest thermal pulse lasts a few minutes and the pulse is followed by soil thermal recovery which is much longer than the pulse itself. Therefore, the thermal impact on the surrounding soil of such pulses should be negligible. Cooling mode The operation in cooling mode is not as intermittent as in the pre-heating mode. It is usually concentrated on daytime during a hot spell. Generally, there is a tendency to ventilate as much as possible with respect to desired reduction of cooling load. 1 0 0 4 8 12 "20.6.2006" 16 20 24 Figure 4.15: Typical pattern of operation during cooling period Annual statistic The daily sum and monthly sum of operation hours as measured during 2006 are depicted on Figure 4.16 and Table 4.3. - 91 - Chapter 4 24 missing data [hours/day] 20 closed loop 16 12 open loop 8 preheating 4 1.12.2006 1.11.2006 1.10.2006 1.9.2006 1.8.2006 1.7.2006 1.6.2006 1.5.2006 1.4.2006 1.3.2006 1.2.2006 1.1.2006 0 Figure 4.16: Daily hours of operation (2006) month [h/month] 1 2 3 4 5 6 7 8 9 10 11 12 pre-heating open loop 148.5 140.1 164.7 64.6 5.1 2.4 0 0 0 3.9 48.1 200.5 cooling open loop 0 0 0 0.1 2.8 120.8 29.4 0 23.1 0 0 0 cooling closed loop 0 0 0 0 0 0 124 3.9 0 0 0 0 missing data 1.1 179.4 45.6 2.0 81.6 2.1 0.2 69.3 133.1 0.5 0.1 32.1 TOTAL [h/year] 778 176 128 547 Table 4.3: Monthly hours of operation (2006) 4.4.3 Outlet air temperature The temperature plot (Figure 4.17) illustrates the long-term temperature dynamics of EAHX. The profile of outlet air temperature is similar to theoretical undisturbed soil temperature profile. Because of to the link with the upper plane surface, the temperature slowly decreases during winter and slowly increases during summer. - 92 - Measurements in situ 30 10 θ [°C] θout [°C] 20 ol ol cl 0 30 θa 20 θs 10 0 -10 -10 -20 -20 -20 -10 0 10 20 30 θin [°C] 2.0m 0 50 100 150 200 250 300 350 t [days] Figure 4.17: Left - scatter plot inlet air temperature θin vs. outlet air temperature θout; right – outlet air temperature compared to ambient air temperature θa (hourly means) and undisturbed soil temperature in depth of 2 m θs2.0m (see chapter 4.2.2); the values of outlet air temperature are daily means over the EAHX operation (ol – open loop (bottom pipe, see Figure 4.1), cl - closed loop); based on measured data in 2006 When evaluating outlet air temperature (EAHX cannot reach efficiency over 100 %), one has to consider the inaccuracy of daily mean values. One reason for the inaccuracy is the following. Because of very short ventilation pulses taking place during winter (see Figure 4.14) and the inertia of temperature/humidity sensors, daily mean values of outlet air temperature are overestimated during winter. 4.4.4 Moisture balance The outlet air water wapour concentration illustrates the long-term moisture dynamics of EAHX. As seen in Figure 4.18 and Figure 4.19, no condensation or evaporation takes place during winter period of air pre-heating; the pipe is perfectly dry. During summer the outlet air water vapour concentration (Figure 4.18) is often lower than inlet air water vapour concentration since condensation occurs. Since the closed loop mode was used in the second half of summer, the outlet air temperatures were higher than during open loop mode operation as the outlet air temperature follows the undisturbed soil temperature profile (link with the upper plane surface). The tendency to condensation was the strongest at the beginning of summer when the soil is cold. - 93 - Chapter 4 20 ol ol cl ρ v,a ρ v,sat(θs,2.0m) 15 3 ρ v [g/m ] 15 3 ρ v,out[g/m ] 20 10 10 5 5 0 0 5 10 15 3 ρ v,in [g/m ] 20 0 0 50 100 150 200 250 300 350 t [days] Figure 4.18: Left – scatter plot inlet air water vapour concentration ρv,in vs. outlet air water vapour concentration ρv,out; right – outlet air water vapour concentration compared to ambient air water vapour concentration ρv,a (hourly means) and saturated water vapour concentration as a function of soil temperature in depth of 2 m ρv,sat(θs,2.0m) 2 10 ol ol cl 1 0 -1 dG(1+3) [kg/day] 3 3 ρ v,in - ρ v,out [g/m ] 4 7.5 ol ol cl 5 2.5 -2 0 2 4 6 8 10 12 14 16 3 ρ v,in [g/m ] 0 0 2 4 6 8 10 12 14 16 3 ρ v,in [g/m ] Figure 4.19: Left – water vapour concentration in the inlet air ρv,in vs. the difference between inlet and outlet water vapour concentration (ρv,in - ρv,out), the positive values denote prevailing condensation within the pipe; right – water vapour concentration in the inlet air ρv,in vs. overall moisture rate dG (if positive, condensation prevails within the pipe); all depicted values are daily means over the EAHX operation 4.4.5 Energy performance Energy injected to/extracted from soil Eeahx is depicted in Figure 4.20. Moreover, for terms I. – IV., and X. – XII., there is depicted the energy recovered by heat recovery Erec (calculated for the assumption of efficiency 85 % and balanced air flows), and energy of additional heating Eheat in order to heat up air to 23 °C (assumed interior temperature). The control system of ventilation system restrains air flow rate during winter in order to overcome too dry air inside. As mean ambient air absolute humidity rises - 94 - Measurements in situ during spring the control system enhances ventilation. Thus, as seen in Figure 4.20, ventilation heat loss was higher during spring than during winter. cooling (+) ol 15 Eeahx 10 Eeahx E [kWh/day] cl Erec 5 Eheat 0 no data -5 -10 -15 heating (-) 0 30 60 90 120 150 180 210 240 270 300 330 360 t [days] Figure 4.20: Daily energy balance (2006); energy to cover ventilation heat loss Ev = Eeahx + Erec + Eheat, Eeahx – contribution of the EAHX (energy injecteded to/extracted from soil), Erec – contribution of heat recovery, Eheat – necessary contribution of additional heating; superscript ol an cl denotes open loop and closed loop 5 2 1.75 1 0.75 0 -1 0.5 -5 2 3 1.25 20 15 1.5 ol ol cl surface fit 10 VaA [(m /day)x m ] Eeahx [kWh/day] x 10 5 0.25 -20 -15 -10 -5 0 5 (θin - θout) [°C] 10 15 20 Figure 4.21: Daily energy Eeahx injected to (+)/extracted from (-) soil sorted by difference between inlet and outlet air temperature (θin - θout) and parameter VaA (daily sum of air flow times exchange surface), points denote measured values, contour lines estimate a surface on a 2d grid (based on scattered data, smoothed) The thermal performance of the EAHX could be further evaluated by specific energy injected/extracted to/from surrounding soil (related to exchange surface A (here 34 m2 including the inlet shaft) and air flow rate Va). Unit may be defined in the following way [(Wh/period)/(m2.m3/period) = Wh/m5], or in [J/m5]. The quantity includes information about exchanger design (exchange surface consists of pipe length - 95 - Chapter 4 and diameter) and exchanger operation (daily sum of air flow) as well. It is the measure of how large exchange surface and high air flow rate is paid for release of energy per period of time. The parameter might allow a comparison between different EAHXs cooling 5 pre-heating -20 -10 0 10 (θin - θout) [°C] VaA 0.2 0.15 0.1 0.05 0 -0.05 -0.1 -0.15 -0.2 eeahx [Wh/m ] VaA 5 eeahx [Wh/m ] (based on monitored data or simulation). ol ol cl 0.2 0.15 0.1 0.05 0 -0.05 -0.1 -0.15 -0.2 cooling pre-heating -20 -10 0 10 (θin - θsoil) [°C] 20 ol ol cl 20 Figure 4.22: Specific energy sorted by difference between inlet and outlet air temperature (left) and by difference between inlet air and undisturbed soil temperature (right). Outlet air temperature from autumn 2006 was excluded from this analysis due to obvious inconsistency (see Figure 4.17, right) The value of cooling power is of particular interest (building cooling load is 3100 W). Cooling power typically varied between 700 W and 1500 W depending strongly on the value of air flow rate (see Figure 4.23 and Figure 4.24). Cooling energy should not be confused with energy injected to soil. Although the air may be cooled down in EAHX, the instantaneous temperature in a building θi could be even lower than the outlet air temperature and therefore cooling of building would not occur. 1500 Qcooling [W] 1250 1000 27.6. - 221m3/h 26.6. - 401 m3/h 750 500 250 0 25 26 [°C] θTi i[°C] 27 Figure 4.23: Cooling power of EAHX sorted by internal air temperature θi in two days of summer 2006 (26.6 and 27.6) for two air flow rate levels - 96 - Measurements in situ 1500 Qcooling [W] ol cl no data 1000 500 V. VI. VII. VIII. IX. Figure 4.24: Cooling power of EAHX during summer 2006; calculated for θi = 26 °C 4.5 Conclusions The impact of air pre-heating in EAHX on ventilation heat loss is not high, especially compared to energy which may be recovered by heat recovery. However, the EAHX serially connected with mechanical ventilation system equipped by heat recovery will protect heat recovery from ice formations occurring in colder climates (Heidt et al, 2002). The impact of air cooling on internal environment may be substantial. Cooling power of EAHX is the matter of dimensioning. However, in this small family house, the value of air flow rate and thus cooling power is very limited by available fun power (400 m3/h). Therefore, the value of cooling load should be primarily reduced by building-energy concept. Monitoring on real-size EAHX in Rychnov is still ongoing. There is still some work to be done. Since only some measured data were evaluated, generalization of measured data was not performed perfectly and knowledge is still rather dispersed. 4.6 Summary The chapter dealt with measurements in-situ performed on real-size EAHX. First, measured soil temperatures and thermal properties of soil were analyzed. Then, the tendency to condensation in EAHX was shown using measured soil temperature and ambient air temperature and humidity. Finally, the measured data on earth-to-air heat exchanger were evaluated. - 97 - Chapter 4 - 98 - Dimensioning of EAHXs 5 Dimensioning of EAHXs The following chapter focuses on development of a simple method for design of optimal dimensions of EAHXs. The characteristic problem of the EAHX design is whether the extension of the pipe and/or addition of another pipe (decrease of air flow rate per one pipe) will lead to the thermal improvement which will balance the increase of investment costs. Therefore, a design methodology will be developed in order to facilitate the design of EAHX. The length of the pipe, diameter of the pipe, and the number of pipes are the main design parameters. The air flow rate is not typical design parameter because the maximal value is usually known before. The highest value of air flow rate is usually needed during summer in order to release sufficient cooling power of EAHX. The design value of air flow rate can be derived from standard calculation of building cooling load or it is the value which can be realized by the ventilation system. 5.1 Theory 5.1.1 Outlet air temperature Assuming constant surface temperature along the length of exchanger, the outlet air temperature may be calculated as: θout = θ s + (θin − θ s ) exp − NTU (5.1) where: NTU (number of transfer units) is dimensionless parameter defined as: NTU = ha 2π r0 L ma c a (5.2) where: ha is air-to-pipe convective heat transfer coefficient [W/(m2K)], see chapter 2.2.2, ma is air flow rate [kg/s], ca is specific thermal capacity of air [J/(kg.K)], ro is internal radius of the pipe [m], and L is length of pipe [m]. - 99 - Chapter 5 The thinner pipe leads to moderate increase of NTU (i.e. increase of thermal efficiency)8, for the enhanced convective heat transfer coefficient prevails over the reduction of diameter. The statement may be proved by analysis of formula (5.2). Increase in air flow rate does not decrease the value of NTU considerably (see Figure 5.1). An increase in air flow rate and/or reduction of diameter increase the value of convective heat transfer coefficient, which compensates, to certain extent, the influence of air flow rate increase on NTU parameter. However, an increase in air flow rate and/or reduction of diameter increases pressure loss extensively. Therefore, as a general rule, it is convenient to split air flow rate into more pipes with minimal diameter and length leading to the sufficient thermal efficiency, reasonable pressure loss, and desired value of cooling power. On the contrary, the higher number of pipes is in conflict with economical restrictions and often with space restrictions of the building site. 3 100 m /h 3 5 5 500 m /h 3 1000 m /h 4 NTU [-] NTU [-] 4 3 3 2 2 1 1 5 10 15 20 L = 5m L = 25m L = 50m 25 30 L[m] 35 40 45 50 100 200 300 400 500 600 700 800 900 1000 3 Va [m /h] Figure 5.1: Parameter NTU as a function of length L and air flow rate Va 5.1.2 Temperature efficiency The temperature efficiency of EAHX ηEAHX shows how far the temperature of the outlet air came close to the pipe surface temperature θs: η EAHX = θin − θ out θin − θ s (5.3) Adding formula (5.1) to formula (5.3) the following formula is obtained: 8 This conclusion is valid as far as semi-empirical calculation of convective heat transfer coefficient is accurate. - 100 - Dimensioning of EAHXs η EAHX = 1 − exp − NTU (5.4) 1,0 0,9 0,8 ηEAHX [-] 0,7 0,6 0,5 0,4 0,3 0,2 0,1 0 0 0,5 1 1,5 2 2,5 3 3,5 4 4,5 NTU [-] 5 Figure 5.2: Thermal efficiency of EAHX ηEAHX as a function of NTU As seen in Figure 5.2, the higher NTU, the higher efficiency will be achieved. The combination of low air flow rate with a very long pipe maximizes the value of NTU. However, from certain value of NTU (approximately 2.0 – 2.5), the efficiency curve rapidly flattens. Therefore, the parameter NTU should not exceed 2.5. However, NTU is a “static” parameter which does not reflect the transient nature of EAHX - the thermal saturation of surrounding soil due to the operation of EAHX and consequent decrease of efficiency. Especially, high air flow rates combined with too short pipes could lead to quick thermal saturation of the pipe surrounding and consequent loss of 1,0 1,0 0,9 0,9 0,8 0,8 0,7 0,7 0,6 0,6 ηEAHX [-] ηEAHX [-] thermal efficiency. 0,5 0,4 0,3 0,3 3 0,2 500 m /h 0,1 3 1000 m /h 5 0,4 3 100 m /h 0,2 0,5 10 15 20 25 30 35 40 45 50 L [m] 0,1 L = 5m L = 25m L = 50m 100 200 300 400 500 600 700 800 900 1000 3 Va [m /h] Figure 5.3: Efficiency as a function of length L and air flow rate Va 5.1.3 Pressure loss The pressure loss in a straight pipe may be calculated as: - 101 - Chapter 5 ∆p fric = ξ v2 L ρa a 2r0 2 (5.5) where: va is mean air velocity [m/s], ρa is density of air [kg/m3], ξ and is friction factor in a hydraulic smooth pipe which may be calculated as (De Paeppe, Janssens, 2003): −2 ξ = (1.82log Re− 1.64 ) for Re > 2300 (5.6) 5.1.4 Cooling power Although the air may be cooled down in EAHX, the instantaneous temperature in a building θi could be even lower than the outlet air temperature and therefore cooling of building would not occur. Thus, heat flow QEAHX injected to soil can not substitute the cooling power Qcooling. The heat flow QEAHX is defined as: QEAHX = ma ca (θin − θ out ) (5.7) Adding formula (5.1) to formula (5.7) the following formula is obtained: QEAHX = ma ca (θin − θ s )η EAHX (5.8) The cooling power of EAHX (having θi temperature of internal air) is defined as: Qcooling = ma ca (θi − θ out ) (5.9) 3 Qcooling [W] 2000 100 m /h 2500 3 500 m /h 3 2000 1000 m /h Qcooling [W] 2500 1500 1000 500 1500 1000 500 0 0 -500 -500 5 10 15 20 25 30 35 40 45 50 L [m] L = 5m L = 25m L = 50m 100 200 300 400 500 600 700 800 900 1000 3 Va [m /h] Figure 5.4: Cooling power as a function of length L and air flow rate Va. Cooling power was calculated for the following boundary conditions: θin = 32 °C, θs = 16 °C (constant over the length of exchanger), θi = 26 °C and r0 = 0,1 m. The negative values of θcooling denote heating of the zone (θi < θout) - 102 - Dimensioning of EAHXs 5.1.5 Hollmuller design rules In his doctoral thesis, (Hollmuller, 2002) recommends to meet the following conditions (Table 5.1). va [m/s] daily dampening annual dampening 2 3 2 1,0 1m for 10 m /h (Va/A = 10) 1m for 5 m3/h (Va/A = 5) 2,0 1m2 for 15 m3/h (Va/A = 15) 1m2 for 7 m3/h (Va/A = 7) 4,0 1m2 for 20 m3/h (Va/A = 20) 1m2 for 10 m3/h (Va/A = 10) Table 5.1: Hollmuller’s design rules (Hollmuller, 2002) Daily dampening denotes dampening of daily ambient air temperature oscillation which occurs naturally during summer. Annual dampening denotes dampening of annual ambient air temperature oscillation. If a design for long-term air heating (e.g. seasonal pre-heating of air for an air-to-fluid heat pump) is intended, the rules for annual dampening should be applied. The rules are also depicted as the bottom line in Figure 5.11. 5.2 Simulations Some simulations were performed in order to study several special setups of EAHX, i.e. simulations testing a) Hollmuller design rules, b) dimensioning according to curve of constant NTU, c) dimensioning according to line of constant Va/A. Moreover, the distance between pipes for case of standard summer EAHX operation and the influence of pipe wall material on thermal performance of EAHX was studied. A calculation domain was a block 1 m (width) x 1 m (height) x L m (length of the pipe). The domain was divided into 19 x 19 x 25 control volumes. All external walls of the rectangle were assumed to be adiabatic. The pipe (external diameter 200 mm) was approximated by an equivalent square with perimeter which equals to perimeter of the pipe. The time step of simulation was 600 seconds. The soil was assumed to have properties λ = 1.5 W/(m.K) and ρcp = 2.0 MJ/(m3.K). The initial temperature of the block was constant (16 °C). At time zero, there was a sudden change of the inlet air temperature; the inlet air temperature was changed to 24 °C and started to oscillate periodically with amplitude 8 °C and period of 1 day. An intermittent operation of EAHX was assumed when the air was sucked through the exchanger if θin > 24 °C (12/12 h). - 103 - Chapter 5 5.2.1 Hollmuller design rules alternative 1 2 3 3 Va [m /h] 100 200 400 va [m/s] 1.0 2.0 4.0 L [m] 17.0 22.6 33.9 3 2 Va/A [m /h/m ] 10 15 20 Table 5.2: Hollmuller design rules – simulation alternatives (for internal diameter do = 188 mm, i.e. DN 200) θin θ [°C] 30 θout,1 θout,2 25 θout,3 20 15 1 2 3 4 5 t [days] Figure 5.5: Hollmuller design rules – the inlet θin and outlet θout air temperature 5.2.2 Constant NTU alternative 1 2 3 4 Va [m /h] 100 200 300 400 L [m] for NTU = 2.0 22.2 25.5 27.7 29.3 3 Table 5.3: Constant NTU – simulation alternatives θin θ [°C] 30 θout,1 θout,2 25 θout,3 θout,4 20 15 1 2 3 4 t [days] Figure 5.6: Constant NTU – the inlet θin and outlet θout air temperature - 104 - 5 Dimensioning of EAHXs 5.2.3 Constant Va/A alternative 1 2 3 4 Va [m /h] 100 200 300 400 L [m] for Va/A = 15 11.3 22.6 33.9 45.2 3 Table 5.4: Constant Va/A – simulation alternatives θin θ [°C] 30 θout,1 θout,2 25 θout,3 θout,4 20 15 1 2 3 4 5 t [days] Figure 5.7: Constant Va/A – the inlet θin and outlet θout air temperature Following the constant NTU curve (see Figure 5.8) leads to the deterioration of thermal performance. The constant NTU curve is insufficiently steep. Following the constant Va/A = 15 line (see Figure 5.8) leads to the inverse performance regarding the previous case of constant NTU. The line is steep enough to dampen the outlet air temperature more in cases with higher air flow rates. Conclusively, the design should be a compromise between too steep line of constant Va/A = 15 and too flat curve of constant NTU = 2.0. 40 DN 200 Va/A = 10 35 L [m] 30 NTU = 2.5 NTU = 2.0 25 NTU = 1.5 20 15 100 Va/A = 15 Va/A = 20 150 200 Hollmuller 250 300 Va [m3/h] 350 400 Figure 5.8: Display of NTU curves, lines of constant Va/A and Hollmuller design rules for DN 200 (points according to Table 5.2) - 105 - Chapter 5 5.2.4 Distance between pipes Three distances between pipes (0.3 m, 0.55 m, 0.8 m) are simulated. The pipe material is neglected. The length of EAHX is assumed L = 25 m; the air flow rate is maintained at 250 m3/h. θ [°C] 30 25 θin θout(Yp = 0,3 m) θout(Yp = 0,55 m) θout(Yp = 0,8 m) 20 15 1 2 3 4 5 t [days] Figure 5.9: Different distances between pipes – the inlet θin and outlet θout air temperature; Yp denotes horizontal distance between pipes As seen in Figure 5.9, the minimal horizontal distance between pipes is in order of several tens of centimeters, because of the intermittent character of summer cooling (12 h pulse/12 h soil recovery). Three times the value of daily penetration depth 3dp (Yp > (0.4 – 0.55) m depending on the soil type) should be a safe value. 5.2.5 Role of pipe material Three real materials are studied: PVC with λ = 0,15 W/(m.K), PP with λ = 0,22 W/(m.K), and PP AWADUCT Thermo with λ = 0,27 W/(m.K). Thickness of pipe wall is assumed tp = 5 mm; the length of EAHX is assumed L = 25 m. The air flow rate is maintained at 250 m3/h. - 106 - Dimensioning of EAHXs θin 30 insul θ [°C] PVC θout 25 PP θout AWA θout 20 15 Rpipe = 0 1 2 3 4 5 t [days] Figure 5.10: Different thermal conductivities of pipe material - the inlet θin and outlet θout air temperature; insul denotes 5 mm thick material with thermal conductivity 0.04 W/(m.K); Rpipe = 0 denotes infinite thermal conductivity of pipe As seen in Figure 5.10, the difference between different materials (PVC, PP, and AWADUCT Thermo) is rather small (order of tenths °C) and almost negligible. The increase of material costs will not be balanced by the increase of thermal efficiency. However, the foam wall pipe (marked as a case “insulation”) deteriorates the thermal performance considerably and therefore should be avoided. Moreover, a poor thermal contact of soil with the pipe (i.e. low quality of work) has similar effect as a thin layer of insulation. Thermal resistance of 5 mm thick air gap is the same magnitude as 10 mm thick foam core. 5.3 Design methodology Using already known (maximal) value of the air flow rate, a combination of design parameters (length of EAHX L, internal diameter of the pipe 2r0, and number of pipes np) should be identified so that the following aspects will be considered: Sufficient efficiency of EAHX. The setup of one pipe, i.e. the relationship between air flow rate per one pipe, internal diameter, and length, manages thermohydraulic efficiency. A range in which EAHX reaches sufficient efficiency is depicted in Figure 5.11 (design for air cooling). The lower bounds are lines originating from Hollmuller design rules for daily amplitude dampening. The upper bounds are NTU curves (NTU = 2.5, i.e. theoretical efficiency 92 %). Generally, a thinner pipe is advantageous because of both thermal efficiency and material costs. In reality, three pipe diameters are applicable. Diameter DN 160 is limited to air flow rate 250 m3/(h.pipe), DN 200 up to 400 m3/(h.pipe), and DN - 107 - Chapter 5 250 up to 600 m3/(h.pipe) which is va ≈ 4 m/s. Higher air flow rates should be split into more pipes in order to ensure reasonable pipe pressure loss. Figure 5.11: A range in which EAHX reaches sufficient efficiency (design for air cooling) Total pressure loss of EAHX (including filters, the inlet drill, elbows and other additional resistances) should not be too high. The limit value has to be determined by a designer of the ventilation system. The value depends on the type of ventilation unit (available fan power) and the pressure loss of the building ductwork. Sufficient cooling power of EAHX with respect to the value of cooling load. The number of pipes manages the total cooling power of EAHX. However, the design of EAHX for removal of total cooling load need not be always effective. It could be a good practice to design EAHX so that it removes a significant part of total cooling load (e.g. one half). The value of cooling load should be primarily reduced by building-energy concept, of course. A good thermal stability of a building is crucial for design of realistic EAHX and effective cooling. For instance, cooling power of EAHX (usual size for family house) is so limited that one south oriented roof window may affect significantly the ratio cooling power of EAHX vs. cooling load of building. - 108 - Dimensioning of EAHXs 40 35 35 30 30 25 20 DN 200 Qcooling L [m] L [m] 40 DN 160 Qcooling 1000 W 25 20 500 W 500 W 15 15 200 W 200 W 50 100 150 3 Va [m /h] 200 250 100 150 200 250 300 3 Va [m /h] 350 400 Figure 5.12: Approximate cooling power of EAHX; calculated for the following boundary conditions: θin = 32 °C, θs = 16 °C (constant over the length of exchanger), and θi = 26 °C Reasonable investment costs. The limit should be agreed by a designer of the ventilation system together with investor. The important part of EAHX budget is cost of excavation cut. Therefore, EAHX should be designed together with building design; important savings may be achieved by integration of work on excavation cuts. Fulfillment of other boundary conditions. The space restrictions of building site and other conditions should be considered (e.g. length limit, limit on number of pipes, radon risk at the site, or the type of soil). The outlet air temperature should be higher than minimal value (thermal comfort criterion). 5.4 Examples 5.4.1 EAHX for family house The first example shows dimensioning of EAHX for cooling of a passive family house with theoretical cooling load 2900 W (960 W are internal heat gains) and design air flow rate 400 m3/h. Some variants are shown in Table 5.5. diameter number air flow length approx. of rate per of cooling pipes one pipe pipe power DN 160 DN 160 2 3 200 m3/h 133 m3/h 23 m 18 m > 1 kW - 109 - pressure costs for drop one meter of pipe (approx.) 22 Pa 175,-/m 8 Pa 175,-/m total costs (PVC) 8050,9450,- Chapter 5 DN 200 DN 200 DN 250 1 2 1 400 m3/h 200 m3/h 400 m3/h 34 m 23 m 33 m 36 Pa 7 Pa 12 Pa 275,-/m 275,-/m 500,-/m 9350,12650,16500,- Table 5.5: Family house - design alternatives 5.4.2 EAHX for larger building The second example shows dimensioning of EAHX for cooling of larger building with theoretical cooling load 15000 W and design air flow rate 3000 m3/h. Some variants are shown in Table 5.6. All design variants lead to similar cooling power. However, the space necessary for pipe placement, pipe costs and pressure drop is different. diameter number air flow length of rate per of pipes one pipe pipe DN 160 DN 160 DN 200 DN 200 DN 250 12 18 8 12 5 250 m3/h 167 m3/h 375 m3/h 250 m3/h 600 m3/h 27 m 21 m 35 m 26 m 41 m approx. cooling power > 8 kW pressu re loss 38 Pa 15 Pa 33 Pa 12 Pa 31 Pa costs per one meter of pipe (approx.) 175,-/m 175,-/m 275,-/m 275,-/m 500,-/m total costs of pipe material (PVC)* 56700,66150,77000,85800,102500,- Table 5.6: Larger building - design alternatives * Not only material costs play role, but the cost of excavations and human work should be considered. 5.5 Conclusions During design of EAHX dimensions, a combination of design parameters (length of EAHX, internal diameter of the pipe, and number of pipes) should be identified for design air flow rate. As a general rule, it is convenient to split design air flow rate into more pipes with minimal diameter and length leading to the sufficient temperature efficiency, reasonable pressure loss, and desired value of cooling power. The placement of more pipes into one excavation cut is suitable as an economical possibility (keeping in mind the minimal distance between pipes, see 5.2.4). Dimensioning should not be confused with the performance prediction (i.e. simulation). The cooling power depicted in Figure 5.12 should be understood as an - 110 - Dimensioning of EAHXs approximate maximal achievable value. The designer should keep in mind the boundary conditions for which the cooling power was calculated. The effectiveness of the design and/or operation of EAHX could be evaluated by a parameter called COP (coefficient of performance), i.e. the ratio between the expected cooling power and the electric power needed for transport of air through the exchanger. For instance, the cooling power of EAHX monitored on the passive house in Rychnov (see chapter 4.4.5) was approximately 1500 W with necessary electric power of 150 W, i.e. COP equal to 10. In this way, EAHX could be assumed to be more effective than standard air conditioning by SPLIT units, which reach mean COP roughly equal to 3. 5.6 Summary The chapter dealt with dimensioning of EAHXs. First, the relationship between key design parameters (air flow rate, length, diameter, and number of pipes) was studied. Then, a set of simulations was performed in order to study several special setups of the pipe. Finally, a simple design methodology was developed in order to facilitate the design of EAHX. - 111 - Chapter 5 - 112 - Conclusions 6 Conclusions In this final chapter, the review of results and final remarks and recommendations for further research will be stated. Conclusions are presented in each chapter at the end. 6.1 Results The earth-to-air heat exchanger was studied by means of numerical simulation and long-term monitoring on real-size systems. The work presents a contribution to understanding the hygro-thermal performance of EAHXs. The main steps and achievements are summarized below. 1) Hygro-thermal model for simulation of the earth-to-air heat exchangers. A numerical model was developed for simulation of EAHXs. The model utilizes a) the analytical solution of differential equation describing heat and moisture balance of the exchanger longitudinal control volume (sensible and latent heat is considered) and b) the numerical solution of two-dimensional transient heat conduction in soil surrounding the pipe (explicit finite difference method), see chapter 2.3. The solution procedure (algorithm) utilizes fully explicit scheme for a discretization in time, see chapter 2.3.3. Since this approach suffers from stability problems, an alternative iterative procedure suppressing the instability was proposed (chapter 2.3.4). 2) Validation of the model. The model was validated against both the analytical solution and measured data on real size systems. The model showed perfect agreement with the analytical solution for an externally insulated exchanger with harmonic oscillation at the input (chapter 3.3). Furthermore, the model was compared with detailed measurement on real-size EAHX (chapter 3.4). The simulated outlet air temperature, outlet air water vapour concentration and overall moisture rate were rather close to measured values. The validated model was finally submitted to series of simulations leading to straightforward parametric analysis (chapter 3.5). The sensitivity of all input parameters was identified. - 113 - Chapter 6 3) Long-term monitoring of real-size earth-to-air heat exchanger and evaluation of the measured data. The results of measurement showed the hygro- thermal performance of real-size EAHX. Plots of outlet air temperature, outlet air water vapour concentration, energy gain, and cooling power were presented (chapter 4.4). The general tendency to condensation in EAHX (the attribute of climatic locality) was shown using measured soil temperature and ambient air temperature and humidity (chapter 4.3.3). 4) Simple method for design of EAHX dimensions was developed using a) design rules founded in the literature and b) theoretical relations between design parameters (length and air flow rate in particular) and performance indicators (e.g. temperature efficiency and cooling power). The method was presented as a stepby-step procedure stressing all aspects which should be considered during design process (chapter 5.3). 6.2 Final remarks and recommendations for further research Based on some research results, recommendations for further research can be made: 1) The EAHX model could be further extended, e.g. considering the neglected physical phenomena. Instead of air, heat carrier could be water or anti-freeze fluid (EWHX, earth-towater heat exchanger). It might be interesting to compare thermal performance of EAHX with EWHX. Water as heat carrier fluid will achieve much higher values of convective heat transfer coefficient than air systems. On the contrary, higher thermal capacity of water will cause slower tendency to heat up/cool down the fluid (compared to air as heat carrier). Generally, the principles of the EWHX model will be the same as the principles of the EAHX model. Only temperature dependency of fluid thermo-physical properties (impact on viscosity and thus on convective heat transfer) is not probably negligible. Based on monitored data, the thermal influence of the inlet shaft was considered to be significant. This surprising observation could lead to conclusion that air cooling due to the suction through a vertical short shaft or water well (e.g. a 2 m deep - 114 - Conclusions shaft connected to a building by rather short pipe) will be sufficient in some cases. Therefore, the inlet shaft should be also modeled (in case there is any inlet shaft). A CFD analysis would be valuable. The EAHX model neglected all kinds of moisture transfer through soil. However, the flow of underground water table may have significant influence on soil temperature. Therefore, the bottom boundary condition should be formulated with respect to defined underground water flow. Moreover, the shallow subsurface is influenced by precipitation. 2) As shown by parametric analysis of the model, soil temperature is one of the key factors for hygro-thermal performance of EAHX. Soil temperature could be intentionally influenced by the specific arrangement of soil surface (see Figure 6.1) or the specific arrangement of soil layers. The thermal influence of a building on soil temperature is another problem. 30 [°C] 25 20 asphalt - 0,5m bare soil - 0,5 m 31.07.06 26.07.06 21.07.06 16.07.06 11.07.06 06.07.06 01.07.06 15 Figure 6.1: Measured soil temperature on the site of Geophysical Institute in Prague (http://www.ig.cas.cz/) in depth of 0.5 m (two kinds of soil surface) 3) The emphasis on component modeling should be shifted to overall cooling efficiency. Therefore, this work should be extended by simulations of a building/room model coupled with the EAHX model. The general idea is that the EAHX model could become a part of the International Building Physics Toolbox, see (Kalagasidis et al, 2007). The Simulink block could be developed in the same manner like floor heating module (Weitzmann, 2002), using so-called S-function (Mathworks, 2005). - 115 - Chapter 6 4) Due to the presence of water in pipes, mould growth is still questionable. Therefore, the inspection of real size installations with a camera and some microbial investigations would be also very valuable. 5) Useful information could be gained by a simple computational experimentation with the model of flat-plate heat exchanger (for geometry of FPHX see Figure 3.1 and Figure 3.2). One case, which could be modeled, using an adjusted model for FPHX, is the phase change material (PCM) heat exchanger. The plate of the exchanger is not made of ordinary storage material, but is made of PCM. Such exchanger could be used for cooling of ambient or internal air (melting process of PCM). During night, when cold ambient air is sucked through the exchanger, the reverse process (solidification) will maintain thermal balance of PCM. The problem of phase change may be modeled by standard heat conduction equation assuming temperature dependent thermal capacity and thermal conductivity. 6) The thermal phase-shifting (Hollmuller et al, 2006 and Hollmuller et al, 2007) is a promising technique for cooling of buildings. The important problem of a phaseshifter is equalization of flow in the exchanger and precise dimensioning. The performance of the phase-shifter is very dependent on the value of convective heat transfer coefficient which is quite uncertain. The model of the flat-plate heat exchanger could be used in order to study this phenomenon thoroughly. A CFD analysis and experimental work would be also valuable. - 116 - Appendices Appendices A1 Derivations and solutions Equation 2.30 Heat flows according to Figure 2.5: Qin = ma caθ a [W] (A.1) [W] (A.2) Qs = ha (θ a − θ s ) 2π r0∂x [W] (A.3) Qlat = g vl 2π r0∂x [W] (A.4) Qout = ma ca (θ a + ∂θ a ) Heat balance according to Figure 2.5: Qin − Qout − Qs + Qlat = 0 (A.5) − ma ca ∂θ a − haθ a 2π r0∂x + haθ s 2π r0∂x + g vl 2π r0∂x = 0 ⎛ 1 ⎞ ⎛ 1 ⎞ /× ⎜ − ⎟ /× ⎜ ⎟ (A.6) ⎝ ma ca ⎠ ⎝ ∂x ⎠ ∂ θ a ha 2π r0 h 2π r0 g l 2π r0 + θa − a θs − v =0 ∂x m a ca m a ca m a ca (A.7) Differential equation (2.30) has the form: y′ + p × y = q (A.8) The solution is the superposition: y = yH + yP (A.9) where yH is the solution of the homogenous differential equation (the equation without right side) obtained by the separation of variable and yP is the particular solution of the original equation, for details see (Bartsh, 1971). - 117 - Appendices Equation 2.39 Moisture flows according to Figure 2.9: Gin = Va ρv [kg/s] (A.10) Gout = Va ( ρv + ∂ρv ) [kg/s] (A.11) Gs = g v 2π r0∂x [kg/s] (A.12) [kg/(m2.s)] (A.13) where: g v = β ρ ( ρv − ρ v,sat ) gv (+) denotes condensation, gv (-) denotes evaporation Moisture balance according to Figure 2.9: Gin − Gout − Gs = 0 (A.14) −Va ∂ρv − β ρ ρv 2π r0∂x + β ρ ρ v,sat 2π r0∂x = 0 ⎛ 1 ⎞ ⎛ 1 ⎞ /× ⎜ − ⎟ /× ⎜ ⎟ (A.15) ⎝ Va ⎠ ⎝ ∂x ⎠ β 2π r0 θ s ∂ρ v β ρ 2π r0 + =0 ρv − ρ ρ v , sat Va Va ∂x (A.16) The solution is performed by identical procedure as described for equation (2.30). FPHX with adiabatic boundary – analytical solution Based on (Hollmuller, 2003) and material sent by Mr. Hollmuller via email. Input: θin (t ) = θ 0 cos (ωt ) (A.17) Output: ⎛ ⎛ Sh ⎞ x ⎞ Sk ⎞ ⎟⎟ ⎟ cos ⎜⎜ ω ⎜ t − ⎟ − ⎝ ma ca ⎠ ⎝ ⎝ va ⎠ ma ca ⎠ ⎛ θ a ( x, t ) = θ 0 exp ⎜ − (A.18) where S is heat exchange surface [m2] from the inlet to distance x, h is total (air/pipe + soil) amplitude-dampening exchange coefficient [W/(m2.K)], k is total 2 (air/pipe + soil) phase-shifting exchange coefficient [W/(m .K)], t is time [s]. Term x/va is time for which air flows from the inlet to distance x from the inlet (transit time). - 118 - Appendices S = 2 Bx h= k= ha ( hs + ik s ) ha + ( hs + iks ) ha ( hs + ik s ) ha + ( hs + ik s ) (A.19) (A.20) Re (A.21) Im where hs and ks are values for infinite ha. ⎛ W ⎞ sinh ⎜ (1 + i ) ⎟⎟ ⎜ d λ p ⎝ ⎠ hs + ik s = (1 + i ) dp ⎛ W ⎞ cosh ⎜ (1 + i ) ⎟ ⎜ d p ⎟⎠ ⎝ - 119 - (A.22) Appendices A2 Explicit finite difference method for transient heat conduction The numerical treatment is based on discretized form of equation (2.31) in time and space. The heat balance of a control volume (Figure A1:) is expressed as: ( ρ c p ) jk ∆y j ∆zk (θ j,k new − θ old j ,k ) = ( qW − qE + qS − q N ) ∆t (A.23) where heat flow components are: qW = (θ j−1,k − θ j,k ) ∆zk ∆y j −1 2λ j −1,k qE = qN = 2λ j ,k (θ j,k − θ j+1,k ) ∆zk ∆y j 2λ j ,k qS = + (A.24) ∆y j + (A.25) ∆y j +1 2λ j +1,k (θ j,k −1 − θ j,k ) ∆y j (A.26) ∆zk −1 ∆zk + 2λ j ,k −1 2λ j ,k (θ j,k − θ j,k +1 ) ∆y j (A.27) ∆zk ∆zk +1 + 2λ j ,k 2λ j ,k +1 Figure A1: Heat balance of control volume - 120 - Appendices After some rearrangements equation (A.23) takes shape: ( ) ( ) ( ) ( ) ( HO jk θ j,k new − θ old j ,k = HY jk θ j −1,k − θ j ,k + HY j +1,k θ j +1,k − θ j,k + HZ jk θ j,k −1 − θ j,k + HZ j ,k +1 θ j ,k +1 − θ j ,k ) (A.28) where: HOjk, HYjk, HYj+1,k, HZjk and HZj,k+1 are helping functions defined as: HO jk = HY jk = ( ρ c p ) jk ∆y j ∆zk ∆zk ∆y j −1 2λ j −1,k HY j +1,k = HZ jk = (A.29) ∆t + (A.30) ∆y j 2λ j ,k ∆zk ∆y j ∆y j +1 + 2λ j ,k 2λ j +1,k (A.31) ∆y j (A.32) ∆zk −1 ∆zk + 2λ j ,k −1 2λ j ,k HZ j +1,k = ∆y j ∆zk ∆zk +1 + 2λ j ,k 2λ j ,k +1 (A.33) Finally, temperatures on the right side of equation (A.28) are considered to be old (already calculated or initial) and unknown (new) temperature is placed on the left side of the equation: ( ) old old old old (A.34) HOjkθj,knew = HYjkθold j−1,k + HYj+1,kθj+1,k + HZjkθj,k−1 + HZj,k+1θj,k+1 + HOjk − HYjk −HYj+1,k − HZjk − HZj,k+1 θj,k The advantage of explicit method is its simplicity; the disadvantage is that the method is not always numerically stable. The computational time step has to fulfill the criterion of the stability ∆t < ∆tcrit; the requirement can be derived from following condition: ( HO jk − HY jk − HY j +1,k − HZ jk − HZ j,k +1 ) ≥ 0 ( ρ c p ) ∆y j ∆zk ∆t = crit HY jk + HY j +1,k + HZ jk + HZ j ,k +1 - 121 - (A.35) (A.36) Appendices A3 House W Basic information Description Single family wooden based house (system two-by-four) with floor and attic, without cellar. Calculated heat use: 44 kWh/(m2.a) 20 km south of Prague, Velké Popovice Location Gypsum board Wall 2 U = 0.19 W/m .K Wooden grill filled by mineral wool Built in 2003 40 mm Water vapor barrier (polyethylene foil) Mineral wool between load bearing timber frame 140 mm OSB plate Wooden grill filled by mineral wool 60 mm Diffusion open membrane Ventilated air cavity Wooden cladding Gypsum board Roof 2 U = 0.17 W/m .K Wooden grill filled by mineral wool 60 mm Water vapor barrier (polyethylene foil) Mineral wool between rafters 180 mm Diffusion open membrane Ventilated air cavity + roof cladding Floor layers Ground floor 2 U = 0.25 W/m .K Concrete slab 50 mm Expanded polystyrene with ventilation ducts 130 mm Ground water insulation (bitumen membrane) Concrete slab on ground 150 mm 2 2 Wooden frame U = 1.7 W/m .K, Glazing U = 1.1 W/m .K Windows Heat storage (915 l) heated by electricity and solar system (roof Heating solar collectors 8.4 m2). Heat from the storage is used for heating up of ventilation and circulation air and for hot water preparation. Additional heat source is a wood burning stove (7 kW) in living room. Mechanical ventilation with heat recovery and EAHX Ventilation Table A1: Basic information about house W - 122 - Appendices List of sensors Nr. 1 2 3 4 5 6 Purpose Temperature of ambient air Outlet air temperature Temperature of circulation air Temperature (circulation + fresh air) Temperature of waste air (after heat recovery) Internal air temperature (in staircase) Signal Type Table A2: List of sensors on house W This monitoring had several imperfections. The interval of the measurement (20 min, later changed to 5 min) was too long, the air flow rate was not measured directly, and the relative humidity of inlet and outlet air was not measured. Therefore, the ventilation mode (current state of the ventilation unit) was often difficult to figure out from measured data. - 123 - Appendices A4 Passive family house in Rychnov Basic information Description Single family wooden based house (panel system) with floor and attic, without cellar. Calculated heat use: 14 kWh/m2.a Location Built in Wall U = 0.10 W/m2.K Rychnov near Jablonec nad Nisou 2004 Gypsum board 2x Wooden grill filled by mineral wool 120 mm Water vapor barrier (polyethylene foil) Gypsum board Thermal insulation (mineral wool) between load 240 mm bearing timber frame (I profiles) Gypsum board Thermal insulation (expanded polystyrene) 60 mm Plaster Roof U = 0.11 W/m2.K Gypsum board 2x Water vapor barrier (polyethylene foil) Wooden grill filled by mineral wool Mineral wool Mineral wool between rafters Diffusion open membrane Ventilated air cavity Roof cladding 60 mm 240 mm 170 mm Slab on ground U = 0.18 W/m2.K Windows Heating Ventilation Wooden frame, Glazing heat mirror U = 0.62 W/m2.K Heat storage (615 l) heated by electricity and solar system (roof solar collectors 5.34 m2). Heat from the storage is used for heating up of ventilation and circulation air and for hot water preparation. Mechanical ventilation with heat recovery and earth-to-air heat exchanger in closed loop mode. Table A3: Basic information about house in Rychnov - 124 - Appendices List of sensors Nr. Purpose Signal Main datalogger: COMET MS3+ 1 Temperature + relative DC (4 – 20 mA) humidity 3 Temperature + relative DC (4 – 20 mA) humidity 5 Temperature – circulation air DC (4 – 20 mA) 6 Air flow rate – circulation fan AC 0 – 400V 7 Air flow rate – exhaust fan AC 0 – 400V 8 Flap position DC 0 – 10V Data logger in the inlet shaft: COMET S3631 2 Temperature + relative humidity 4 Temperature Data logger – soil: COMET S0141 s1 Temperature – 5 cm s2 Temperature – 30 cm s3 Temperature – 62 cm s4 Temperature – 105 cm Data logger – living room: LR1 Temperature + relative humidity CO2_LR CO2 concentration – living room CO2_C1 CO2 concentration – circulation air Type NH 421.65 NH 421.65 Pt 1000 module E3 module E3 module D2 embedded in the logger Pt 1000 (external) Pt 1000 (external) Pt 1000 (external) Pt 1000 (external) Pt 1000 (external) embedded in the logger external, SIEMENS external, SIEMENS Table A4: List of sensors on house in Rychnov - 125 - Appendices err1 err2 Ventilation modes 0 1 NORM 2 NORM 3 NORM 4 NORM 5 MAX MAX no ventilation missing data e1 e1 i1 i1 ventilation with heat recovery heating + ventilation with heat recovery i2 e1 c1 i1 i1 c2 + e2 NORM MAX i2 e1 c1 e2 MAX i2 i2 trace error MAX c2 heating or cooling (circulation between building and EAHX) modes 2a, 2b, 5a e2 e2 night ventilation external signals cooling (suction through EAHX) enhanced ventilation induced by external signals Table A5: Definition of ventilation modes The ventilation system installed in the house utilizes five basic ventilation modes (Table A5). Each mode is defined by the combination between fan operation and the position of circulation damper (Table A6). Some other ventilation modes (marked 2a, 2b, 5a) are induced by external signals (cooking, use of shower or WC). Ventilation mode 0 1 2 3 4 5 fans circulation damper Mc Mv closed e1 between closed c1 off 0 0 0 0 0 ventilation 1 1 0 0 1 circulation + ventilation 1 1 0 1 0 circulation 1 0 1 0 0 under pressure 0 1 0 0 1 over pressure 1 0 0 0 1 Table A6: Ventilation modes - basic combinations (1 – yes, 0 – no), Mc denotes circulation fun, Mv denotes ventilation fan - 126 - Appendices Air flow rates The average velocity of air flow was measured at each outlet drill of the house for each ventilation mode. Thus, each of ventilation modes was assigned with the value of air flow rate through EAHX. 0 0 0 0 0 109 296 26 213 224 401 0 0 5a 4MAX 0 5MAX 4NORM 0 115 198 115 115 198 198 5NORM 3MAX 2b 2a 2MAX 1MAX 3NORM 0 2NORM mode open loop closed loop 1NORM off Figure A2: Simple velocity (air flow rate) measurement on site (September 2006) 221 401 415 0 0 0 Table A7: Air flow rates in [m3/h] assigned with ventilation modes. Mc denotes circulation fun; Mv denotes ventilation fan 1NORM 1MAX 2NORM 2MAX 2a 2b 3NORM 3MAX 4NORM 4MAX 5NORM 5MAX 5a fun Mc Mv off Auxiliary energy 0 0 29 39 68 80 71 39 162 39 71 80 162 80 71 0 162 0 0 39 0 80 68 0 158 0 158 80 Table A8: Electric inputs assigned with ventilation modes (according to provided information from manufacturer and measurement of air flow rates). Mc denotes circulation fun, Mv denotes ventilation fan - 127 - Appendices A5 Thermal properties of soils and rocks A few references related to soil properties have been found. (Sundberg, 1989) introduced typical thermal properties of some soils (CLAY, SILT, SAND, PEAT). (Fink et al, 2002) provided data for three soil types (moist soil, dry soil, clay soil). The software GAEA (Heidt et al, 1997) and software Awadukt Thermo (provided by company Rehau) contain similar databases of soil thermal properties, but the source of their information is unknown. (Janssen, 2002) comprehensively defined four main soil types (SAND, SILT, LOAM, CLAY) and three extreme soil types using textural triangle defined by United States Department of Agriculture (USDA). The thermal properties (thermal conductivity, volumetric thermal capacity) were presented graphically as a function of moisture content. The thermal conductivity of dry soils varies between 0.3 and 1.0 W/(m.K). The thermal conductivity of moist soils typically varies between 0.9 and 2.5 W/m.K. Thermal conductivity rises fast at lower moister content, and then the curve flattens. Thermal conductivity of frozen soil is substantially higher than thawed soil (thermal conductivity of ice = 2.2 W/(m.K)). Dependence of thermal conductivity on soil type is significant. Because of high quartz content, sand is the best conductor whereas clay is the worst (Janssen, 2002, page 158, Figure 7.17). Silt and loam are located in between. Volumetric heat capacity typically varies between 1.0 and 3.0 MJ/(m3.K), higher values for water saturated soils. Dependence of volumetric heat capacity on soil type is very weak; the dominant factor is moisture content (linear function of moisture content). The thermal diffusivity of soils is typically in order of 10-7 m2/s. Thermal properties of soils and rocks are listed below, based on several information sources. The measurement of several soil samples taken from the site should be preferably used for the estimation of soil thermal properties (λ, ρcp), since the guess of the thermal properties may be rather inaccurate. - 128 - Appendices ρ soil type "Rehau Awadukt Thermo" "EN ISO "AEE Intec" 13370 " "GAEA" source english coarse gravel calcareous loamy soil moist loamy soil sand dry sand sandy ground sandstone clay clay gravel homogenous rock moist soil dry soil cp ρcp german λs [kg/m3] [J/kg.K] [MJ/m3.K] [W/m.K] Grobkiesig 2000 1840 3,68 0,52 Kalkhaltige Erde 1670 2230 3,72 0,71 Lehmboden 1650 2850 4,70 2,30 lehmig feucht 1800 1340 2,41 1,49 Sand 1780 1390 2,47 0,93 Sand trocken 1500 920 1,38 0,70 Sandboden 1520 1650 2,51 1,24 Sandstein 2250 710 1,60 1,87 Tonboden 1500 880 1,32 1,28 Ton (ISO 13370) 3,00 1,50 Kiesig (ISO 13370) 2,00 2,00 Felsen (ISO 13370) 2,00 3,50 Erde, feucht 1800 1260 2,27 2,50 Erde, tonig 1500 1080 1,62 1,51 Erde, trocken 1500 840 1,26 0,35 Quarzsand, trocken (Darby 1978) 1650 1,32 0,27 Quarzsand, 8% Feuchte (Darby 1978) 1750 1,77 0,59 Sand trocken (Karl 1965) 1650 1,38 0,70 Sand, 15% Feuchte (Darby 1978) 1780 2,46 0,92 Sandboden, 9% Feuchte (Neiss 1977) 1440 2,17 0,98 Sandboden, 13% Feuchte (Neiss 1977) 1600 2,88 1,50 trockener Sand (ISO 13370) 1998 4,94 1,60 Sand feucht (Jager 1981) 1500 1,80 1,88 Sand/Kies (EN 12524) 1950 2,05 2,00 nasser Sand (ISO 13370) 2166 6,52 2,10 Erdreich, grobkiesig (VDI 1984) 2000 3,69 0,52 Erde, kalkhaltig (Darby 1978) 1670 3,72 0,71 Lehm feucht (Cube 1977) 1800 2,41 1,45 Lehmboden, 36% Feuchte (Neiss 1977) 1650 4,69 2,30 Lehm gesätigt (Jager 1981) 1800 2,57 2,90 Ton (ISO 13370) 1820 5,74 1,20 Ton boden (VDI 1984) 1500 1,32 1,28 Ton/Schlick/Schlamm (EN 12524) 1500 3,15 1,50 Schluff (ISO 13370) 1920 5,64 1,50 Sandstein (VDI 1974) 2250 1,60 1,90 Granit (EN 12524) 2600 2,60 2,80 Felsen (ISO 13370) 2500 6,25 3,50 Torf (ISO 13370) 1500 5,83 0,35 Table A9: Thermal properties of soils - 129 - as bs [m2/s] 1,4E-07 1,9E-07 4,9E-07 6,2E-07 3,8E-07 5,1E-07 4,9E-07 1,2E-06 9,7E-07 5,0E-07 1,0E-06 1,8E-06 1,1E-06 9,3E-07 2,8E-07 2,0E-07 3,3E-07 5,1E-07 3,8E-07 4,5E-07 5,2E-07 3,2E-07 1,0E-06 9,8E-07 3,2E-07 1,4E-07 1,9E-07 6,0E-07 4,9E-07 1,1E-06 2,1E-07 9,7E-07 4,8E-07 2,7E-07 1,2E-06 1,1E-06 5,6E-07 6,0E-08 [Ws0,5/m2.K] 1383 1626 3289 1896 1517 983 1763 1728 1300 2121 2000 2646 2381 1564 664 591 1019 984 1509 1458 2078 2811 1839 2023 3701 1385 1629 1869 3286 2728 2625 1300 2174 2908 1742 2698 4677 1429 References References 1) Ashrae: Ashrae Handbook of Fundamentals, American Society of Heating, Refrigerating and Air Condition Engineers, 2001. 2) Arkar, C., Medved, S.: Free cooling of a building using PCM heat storage integrated into the ventilation system, Solar Energy 81, 2007. 3) Bartsch, H., J.: Matematické vzorce, SNTL, Praha, 1971. 4) Beltrami, H.: Inference of climate change from geothermal data, Global and Planetary Change 29, 2001. 5) Beltrami, H., Kellman, L.: An examination of short- and long-term air-ground temperature coupling, Global and Planetary Change 38, 2003. 6) Benkert, St., Heidt, F., D., Schöler, D.: Calculation tool for earth-to-air heat exchangers GAEA, Building Simulation, Prague, 1997. 7) Cílek, V., Kašík, M.: Nejistý plamen. Dokořán, Praha, 2007. 8) Claesson J., Dunand A.: Heat extraction from the ground by horizontal pipes. A mathematical analysis, Swedish council for Building Research, Stockholm, 1983. 9) De Paepe, M., Janssens, A.: Thermo-hydraulic design of earth-air heat exchangers, Energy and Buildings 35, 2003. 10) Fink, C. Blümel, E., Kouba, R. Heimrath, R.: Passive Kühlkonzepte für Büro- und Verwaltungsgebäude mittels luft- bzw. Wasserdurchströmten Erdreichwärmetauschern, AEE INTEC, 2002. 11) Fluid properties calculator, available at url http://www.electrooptical.com/html/unitconv/convertcalcs/physical/fluid_properties.asp 12) Gieseler, U.,D.,J., Bier, W., Heidt, F., D.: Cost efficiency of ventilation systems for low-energy buildings with earth-to-air heat exchange and heat recovery, PLEA, Toulouse, 2002. - 130 - References 13) Hagentoft, C.A., Introduction to building physics, Studentliteratur, 2001. 14) Hellström, G.: Ground Heat Storage, Department of Mathematical Physics, University of Lund, Sweden, 1991. 15) Hens, H.: Building Physics – Heat, Air and Moisture, Ernst&Sohn, 2007. 16) Herkel, St.: Verwaltungsgebäude DB Netz AG in Hamm, Abslussbericht, Fraunhofer ISE, Freiburg, Germany, 2002. 17) Hofmeister, O., Kopecký, P., Tywoniak, J.: Sluňákov Ecological Education Centre – Basic Information, Proceedings of conference Central Europe towards Sustainable Building 2007, Prague 2007. 18) Hollmuller, P., Lachal, B.: TRNSYS compatible moist air hypocaust model: description and validation, Centre universitaire d’études des problèmes de l’energie, Genève, 1998. 19) Hollmuller, P., Lachal, B: Cooling with air-to-earth heat exchangers versus direct night cooling: a parametric study for different climates, 18th conference PLEA, Florianópolis, 2001. 20) Hollmuller, P., Lachal, B.: Cooling and pre-heating with buried pipe systems: monitoring, simulation and economic aspects, Energy and Buildings 33, 2001. 21) Hollmuller, P.: Utilisation des échangeurs air/sol pour le chauffage et le rafraîchissement des bâtiments. Mesures in situ, modélisation analytique, simulation numérique et analyse systémique, Faculté des Sciences de l'Université de Genève, 2002. 22) Hollmuller, P.: Analytical characterization of amplitude dampening and phase- shifting in air/soil heat-exchangers, Journal of Heat and Mass Transfer 46, 2003. 23) Hollmuller, P., Lachal, B., Zgraggen, J., Pampaloni, E.: Dephaseur Thermique Diffusiv – raport final, Centre universitaire d’études des problèmes de l’energie, Genève, 2004. 24) Hollmuller, P., Lachal, B.: Buried pipe systems with sensible and latent heat exchange: validation of numerical simulation against analytical solution and long-term monitoring, 9th conference of IBPSA, Montreal, 2005a. - 131 - References 25) Hollmuller, P., Lachal, B., Zgraggen, J., M.: Rafraichisement de batiments par dephasage thermique controlé, Proceedings of Cisbat, 2005b. 26) Hollmuller et al.: Potential of inertial ventilation for passive cooling in Brazilian climates, Palenc 2005, Santorini, 2005c. 27) Hollmuller, P.: excel routine fouriercyl.xls, from email communication during 2005d. 28) Hollmuller, P., Lachal, B., Zgraggen, J.,: A new ventilation and thermal storage technique for passive cooling of buildings: thermal phase-shifting, Centre universitaire d’études des problèmes de l’energie, Genève, 2006. 29) Hollmuller, P., Lachal, B., Zgraggen, J.,: A new heat exchange and storage technique for ventilation: controlled thermal phase-shifting, Proceedings of Heat-SET, Chambéry, 2007. 30) IPCC - Mezivládní panel pro změnu klimatu (kolektiv autorů): Změna klimatu 2007: Fyzikální základy. Shrnutí pro politické představitele, český překlad. Ke stažení na http://www.chmi.cz/cc/ipcc.html. 31) Janssen, H.: The influence of soil moisture transfer on building heat loss via the ground, Katholieke Universiteit Leuven, 2002. 32) Kalagasidis, A., S., Weitzmann, P., Nielsen, T., R., Peuhkuri, R., Hagentoft, C., E., Rode, C.: The International Building Physics Toolbox in Simulink, Energy and Buildings 39, 2007. 33) Kramoliš, P.: Realizovaný projekt nástavby mateřské školy v Ostravě – Proskovicích, Alternativní energie 3, 2002. 34) Kumar, R., Kaushik, S.,C., Garg, S., N.: Heating and cooling potential of an earth-to-air heat exchanger using artificial neural network, Renewable Energy 31, 2006. 35) MathWorks: Writing S-functions, 2005. http://www.mathworks.com 36) Mihalakakou, G., Santamouris, M., Asimakopoulos, D.: Modelling the thermal performance of earth-to-air heat exchangers, Solar Energy 53, 1994. 37) Mihalakakou, G., Lewis, J., O., Santamouris, M.: On the heating potential of buried pipes techniques – application in Ireland, Energy and Buildings 24, 1996. - 132 - References 38) Mihalakakou, G.: On the heating potential of a single buried pipe using deterministic and intelligent techniques, Renewable Energy 28, 2003. 39) Lomas, J., Eppel, H.: Sensitivity analysis techniques for building thermal simulation programs, Energy and Buildings 19, 1992. 40) Nilsson C.A.: Preheating of Ambient Air by a System of Earth Tubes as a Heat Source for Buildings, Chalmers University of Technology, Göteborg, Sweden, 1991. 41) Oldenborgh, G., J.: Extraordinarily mild European autumn 2006 due to global warming?, Global Change NewsLetter 67, 2006. 42) Pfafferott, J.: Evaluation of earth to air heat exchangers with a standardized method to calculate energy efficiency, Energy and Buildings 35, 2003. 43) Pfafferott, J.: Enhancing the Design and the Operation of Passive Cooling Concepts, Fraunhofer IRB Verlag, 2004. 44) Rode, C.: Series of lecture notes on numerical transient heat conduction, materials for DTU course Numerical Methods for Building Energy Technology, 1997. 45) Santamouris, M., Asimakopoulos, D.: Passive cooling of buildings, James and James, London, 1996. 46) Santamouris, M., Mihalakakou, G., Argiriou, A., Asimakopoulos D.: On the performance of buildings coupled with earth to air heat exchangers, Solar Energy 54, 1995. 47) Stahl, F.: Preheating of Supply Air through an Earth Tube System – Energy demand and moisture consequences, Building Physics 2002 – 6th Nordic Symposium, 2002. 48) Sundberg, J.: Thermal Properties of Soils and Rocks, Chalmers University of Technology, Göteborg, Sweden, 1989. 49) Tywoniak, J., Morávek, P., Kopecký, P.: Zur Wohnungslüftung in Niedrigstenenergiehäusern in Tschechien, 2. Europäisches Blower Door Symposium, Kassel, 2007. 50) Verhoef, A. et al: Thermal properties for vineyard (EFEDA-I) and savanna (HAPEX - Sahel) sites, Agricultural and Forest Meteorology 78, 1996. - 133 - References 51) Voss, K., Kramp, M.: Zero-Energy/Emission-Buildings – Terms, Definitions and Building Practice, CESB 2007, Prague, 2007. 52) Weitzmann, P.: A floor heating module using an S-function approach for the International Building Physics Toolbox, Department of Civil Engineering, Technical University of Denmark, 2002. - 134 - Publications written in context with the thesis [1] Kopecký, P.: Numerical modeling of the flat-plate heat exchanger, Sborník konference Tepelná ochrana budov, Štrbské Pleso, Vysoké Tatry, 2005. [2] Kopecký, P.: Numerické modelování zemních výměníků tepla, Sborník konference Budova a Energia, Podbanské, Vysoké Tatry, 2005. [3] Kopecký, P., Tywoniak, J.: Pre-heating and pre-cooling of fresh air in the earth-to- air heat exchanger (EAHX) – Simulation and monitoring of a simple EAHX in lowenergy family house, Proceedings of conference CISBAT 2005, Lausanne, Switzerland, 2005. [4] Kopecký, P.: Zemní výměník tepla – matematický model, In: Tepelná ochrana budov 6/2005. [5] Kopecký, P.: Zemní výměník tepla – matematický model, validace, experimentální měření, simulace, Dílčí výzkumná zpráva pro výzkumné centrum CIDEAS, 2005. [6] Kopecký, P.: Pokročilá simulace zemních výměníků tepla – srovnání simulace a experimentálních dat z měření, konference Tepelná ochrana budov, Praha, 2006. [7] Kopecký, P., Tywoniak, J.: Hygro-thermal numerical simulation model for earth-to- air heat exchangers: validation process and the example of simulation, Proceedings of conference Advance Engineering Design 2006, Praha, 2006. [8] Kopecký, P.: Zemní výměník tepla: model a validace, In: Vytápění, Větrání, Instalace 4/2006. [9] Kopecký, P.: K energetickému přínosu zemního výměníku tepla, Sborník konference Pasivní domy 2006, Brno, 2006. [10] Kopecký, P.: Matematický model pro simulaci zemních výměníků tepla – experimentální validace, Sborník konference Simulace budov a techniky prostředí 2006, Praha, 2006. - 135 - [11] Kopecký, P.: K návrhu dimenzí zemních výměníků tepla I., Dílčí výzkumná zpráva pro výzkumné centrum CIDEAS, 2006. [12] Kopecký, P., Tywoniak, J.: Advanced simulation of the earth-to-air heat exchangers – a comparison between simulation and measured data, Proceedings of conference 12. Bauklimatische Symposium, Dresden, 2007. [13] Tywoniak, J., Morávek, P., Kopecký, P.: Zur Wohnungslüftung in Niedrigstenenergiehäusern in Tschechien, 2007. [14] Kopecký, P.: Pasivní dům v Rychnově – vyhodnocení některých měřených dat za rok 2006, Sborník konference Tepelná ochrana budov, Štrbské Pleso, Vysoké Tatry, 2007. [15] Kopecký, P.: Hygro-thermal performance of earth-to-air heat exchanger: long- term data evaluation and short-term simulation, Proceedings of conference CISBAT 2007, Lausanne, Switzerland, 2007. [16] Kopecký, P.: Zemní výměník tepla – vyhodnocení měřených dat a tepelně vlhkostní simulace, Sborník konference Pasivní domy 2007, Brno, 2007. [17] Kopecký, P.: Earth-to-air heat exchanger: hygro-thermal performance, Proceedings of conference Passivhaustagung 2008, Nuremberg, 2008. [18] Kopecký, P.: Návrh dimenzí zemních výměníků tepla, In: Vytápění, Větrání, Instalace 2/2008. - 136 - Other publications [1] Tywoniak, J., Morávek, P., Kopecký, P., Kramoliš, P.: Příklad stavebně energetické koncepce budovy, Sborník konference CEERES 2003, Praha, 2003. [2] Kopecký, P.: Integrace solárních systémů do obvodových konstrukcí budov, Sborník konference Juniorstav 2004, Brno, únor 2004. [3] Kopecký, P.: Vliv zadní izolace na tepelný zisk integrovaného neprovětrávaného kolektoru, Sborník konference Tepelná ochrana budov 2004, Praha, 2004. [4] Adamovský, D., Brázda, P., Kopecký, P., Nováček, J., Tencar, J.: Ideové řešení rekonstrukce obvodového pláště budov FSv ČVUT v Praze, Sborník konference Tepelná ochrana budov 2004, Praha, 2004. [5] Hofmeister, O., Kopecký, P., Tywoniak, J.: Připravovaný projekt střediska ekologické výchovy Sluňákov, Sborník konference Tepelná ochrana budov 2004, Praha, 2004. [6] Kopecký, P., Vonka, M.: Návrh rekonstrukce obvodového pláště budovy stavební fakulty a fakulty architektury v Praze v širších environmentálních souvislostech, In: Tepelná ochrana budov 4/2004. [7] Vonka, M., Mukařovský, J., Kopecký, P.: Pasivní domy – Krems an der Donau 2004, In: Stavba 4/2004. [8] Hofmeister, O., Kopecký, P., Tywoniak, J.: Sluňákov Ecological Education Centre – Basic Information, Proceedings of conference Central Europe towards Sustainable Building 2007, Prague 2007. - 137 -