Thesis Evaporative Cooling Towers

THESIS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

Evaporative Cooling Tower and Chilled Beams Design Aspects for Cooling in Office Buildings in Northern Europe BENGT BERGSTEN

Building Services Engineering Department of Energy and Environment CHALMERS UNIVERSITY OF TECHNOLOGY Göteborg, Sweden 2009

Evaporative Cooling Tower and Chilled Beams Design Aspects for Cooling in Office Buildings in Northern Europe Bengt Bergsten ISBN : 978-91-7385-348-4

© Bengt Bergsten Doktorsavhandlingar vid Chalmers tekniska högskola Ny serie nr 3029 ISSN 0346-718X Document D2009:05 Building Services Engineering Department of Energy and Environment Chalmers University of Technology SE-412 96 Göteborg, Sweden 2009 Telephone: + 46 (0)31-772 10 00 http://www.chalmers.se

Printed by Chalmers Reproservice, Göteborg, Sweden 2009

II

Evaporative Cooling Tower and Chilled Beams Design Aspects for Cooling in Office Buildings in Northern Europe Bengt Bergsten Building Services Engineering Department of Energy and Environment Chalmers University of Technology

Abstract The purpose of this thesis is to examine a comfort cooling system consisting of a hydronic cooling system with an evaporative cooling tower and chilled beams. The system has no conventional chiller. The analysis of the comfort cooling system is made through simulations in a building simulation tool, IDA Indoor Climate and Energy (IDA ICE) and monitoring of a pilot plant equipped with this system. A mathematical model of a cooling tower is developed and validated in this project and added to IDA ICE. The base case condition comprises a normal office building with a normally sized cooling system. The total internal heat gain, including solar radiation, is between 50 – 70 W/m2 and the climate conditions is equal to those in northern Europe, i.e. north of latitude 48 – 49°N. The pilot plant consists of an office space of 450 m2 which is cooled by a free cooling system. The free cooling system has an outdoor evaporative cooler which is connected to a chilled beam system in the office building. The monitoring was made during May – August 2007. The results form the simulations indicate that the comfort cooling system can maintain a thermal climate where the annual maximum indoor air temperature is between 24 – 26°C in a Nordic climate, and between 25 – 27°C in the rest of the northern Europe. The annual duration of indoor air temperature during working hours exceeding 24°C is between 1 – 5% in Nordic climates and between 3 – 8% in the rest of the northern Europe. The annual COP of the cooling tower and the cooling system is about 7 at base case conditions. Thus, the use of electric energy is about a third of the energy used in a conventional cooling system. The outcome from the pilot plant basically confirms the results regarding the indoor climate. However, the COP of the evaporative cooler during the measured period is lower compared to the results from simulations and findings from literature. A hydronic cooling system with chilled ceilings or beams and an evaporative cooling tower can be applied to both new and refurbished buildings. This comfort cooling system represents well-established techniques and no parts of the system are new or unproven on the market. Preliminary data from other sources indicate approximately equal total investment costs for a hydronic cooling system with a cooling tower compared to a conventional hydronic cooling system with a mechanical chiller. Keywords: free cooling, low energy cooling, evaporative cooling, commercial buildings, cooling tower, cooling tower model, simulation, hydronic cooling system, chilled beams, pilot plant.

III

This doctoral thesis has been financed by the Swedish National Energy Agency, BELOK, Swedish Building Research Programme Competitive Building, IMI Indoor Climate and CIT Energy Management

IV

Preface The idea behind this research project originates from the early 90’s when working as an HVAC engineer, designing HVAC systems in buildings. The question came up when learning about evaporative cooling of air; is it possible to chill a hydronic cooling system (with e.g. chilled beams which is very common in Sweden) by evaporative cooling only and still get an acceptable indoor temperature? At that time, no written sources or nobody I knew could give an answer. Many years later, I got the opportunity to work at CIT Energy Management. Here the doors to the research world were opened for me and I was encouraged to conduct research. I am truly grateful for that. Research is seldom performed in solitary and for me that was certainly not the case. Obtaining the results in this thesis, and write it, has been possible only by support from several people and organisations. I would first like to express my sincere gratitude to all my colleagues at CIT Energy Management (CIT EM) for contributing to a work environment impregnated with sincerity, helpfulness and joy. Professor Emeritus Enno Abel deserves my gratitude for initiating the project of building a pilot plant. My supervisors have been Per Erik Nilsson (CIT EM) and Jan-Olof Dalenbäck (CIT EM and Department of Energy and Environment, Building Services Engineering, Chalmers). Some people outside CIT Energy Management deserve special appreciation. Leif Nilsson at Monitoring Centre, Chalmers, who both helped me a lot in the deep waters of multi variable regression analysis, developing a curve fit equation for a numerically tricky equation in the cooling tower model. Leif also supplied and rigged the monitoring system at Kvarnberget, Gothenburg, with the highest level of professionalism. Lars Eriksson and Mika Voulle at Equa, who did a great job weeding my first primitive versions of the cooling tower model (in NMF language), and patiently answering all my NMF-related questions. Per Sahlin at Equa, for lending me the tool IDA SE before CIT EM purchased it. The real estate company Vasakronan who granted access to their building at Kvarnberget in Gothenburg for the pilot plant and financed it together with BELOK. Thank you all! The financial recourses have been supplied from the Swedish Energy Agency, Swedish Building Research Programme Competitive Building, IMI Indoor Climate and CIT Energy Management. Last but not least, I would like to express my very deepest gratitude to my beloved family for their patience, understanding and support. Without that, dear reader, this thesis would never have seen daylight. Göteborg, November 2009 Bengt Bergsten

V

A better burden can no man bear on the way than his mother wit (Ej bättre börda man på vägen bär än kunskap mycken) Havamal

VI

Table of contents Abstract Preface Nomenclature Glossary

III V XI XV

1

Introduction …………………………………………………

1

1.1 1.2 1.3 1.4

Motivation of study Scope and limitations Previous work Outline of this work

1 2 3 6

2

Cooling towers – introduction and modelling …………......

9

2.1 Introduction 2.2 Modelling heat and mass transfer in evaporative cooling towers 2.2.1 The effectiveness – NTU method 2.3 Visualizing heat and mass transfer in an evaporative cooling tower 2.4 Legionella in evaporative cooling towers

9 15 18 23 25

3

27

Cooling system with evaporative cooling tower ………......

3.1 Introduction 3.2 System configurations 3.1.1 Parallel coupling 3.1.2 Series coupling

4

27 30 30 33

Simulation model development …………………………...... 37

4.1 Prerequisites 4.2 The cooling tower model (CTM) 4.2.1 General description 4.2.2 Modes of operation 4.2.3 Control of supply temperature of cooling tower liquid 4.2.4 User input data 4.2.5 Limitations 4.2.6 Equations 4.2.7 Determining the UA-value 4.3 Validation of the model 4.3.1 Validation by published data 4.3.2 Validation by data from pilot plant

VII

37 38 39 40 40 41 42 42 57 61 61 66

5

Simulation of a free cooling system with evaporative cooling tower …………………………………..

5.1 Simulation tool 5.2 Prerequisites at base case 5.2.1 Model building 5.2.2 Analysed room 5.2.3 Internal heat and moisture generation 5.2.4 Cooling tower 5.2.5 HVAC system 5.3 Simulation methodology 5.3.1 Heat wave simulation 5.3.2 Full year simulation 5.4 Parameter analysis methodology 5.4.1 List of parameter alterations 5.4.2 Discussion and motivation of parameter alterations

6

95 95 98

Results and analysis of simulations……………………….... 99

7.1 Cooling tower performance 7.2 Indoor thermal climate 7.2.1 Base case 7.2.2 Single parameter variations 7.2.3 Multi parameter variations 7.3 Energy use 7.3.1 Base case 7.3.2 Single parameter variations 7.4 Findings from other sources

8

69 71 72 74 74 76 78 81 81 81 82 82 84

Monitoring of free cooling system with evaporative cooling tower ………………………………….... 95

6.1 Introduction 6.2 Pilot plant lay-out 6.3 Monitoring system

7

69

100 105 106 111 115 123 123 125 128

Results and analysis of measurements…………………….. 131

8.1 Ambient conditions 8.2 Indoor thermal climate 8.3 Performance and energy use

131 133 136

VIII

9

Conclusions and discussion......……………………………... 139

9.1 Conclusions 9.2 Discussion 9.2.1 Design 9.2.2 Indoor climate 9.2.3 Investments and annual costs 9.2.4 Applicability 9.2.5 Miscellaneous 9.3 Further research and development

139 142 142 147 148 150 150 152

References

153

Appendix A: NMF model

161

Appendix B: Specifications pilot plant

175

Appendix C: Specifications monitoring system

177

IX

X

Nomenclature

Latin letters Area of control volume, Area of wet surface in cooling tower [m2] A= a= Fill interface area per fill volume [m2/m3] Ca = Heat capacity flow of air [W/°C] Design heat capacity flow of air [W/°C] C a ,d = C liq1 =

Heat capacity flow of Liquid 1 [W/°C]

C liq 2 =

Heat capacity flow of Liquid 2 [W/°C]

c p ,a =

Specific heat of dry air [J/kg, °C]

c p ,a ,e =

Effective specific heat of air [J/kg, °C]

_

c p , fic = c p ,liq1 =

Fictitious mean specific heat [J/kg, °C] Specific heat of Liquid 1 [J/kg, °C]

c p ,liq 2 =

Specific heat of Liquid 2 [J/kg, °C]

c p ,w =

Specific heat of cooling liquid [J/kg, °C]

Ctrl =

∆h = ha = ha ,in =

Signal from controller, 0< Ctrl <1 [-] Enthalpy difference of moist air in the cooling tower [J/kg] Enthalpy of the surrounding air [J/kg] Enthalpy of entering ambient air [J/kg]

ha ,out =

Enthalpy of leaving air [J/kg]

dha =

Enthalpy change of air in a control volume [J/kg]

hs = Km = K= M& a = M& a =

Enthalpy of saturated air at the same temperature as the water film [J/kg] Mass transfer coefficient [kg/s, m2] Mass transfer coefficient [kg/s, m2] Mass flow of air [kg/s] Air mass flow through cooling tower [kg/s]

M& a , d =

Design air mass flow through cooling tower [kg/s]

Min _ M& a ,d = Minimum fraction of air mass flow through cooling tower for fan with variable speed drive [-] & M liq1 = Mass flow of Liquid 1 [kg/s] M& = Design mass flow of Liquid 1 [kg/s] liq1, d

M& liq 2 = Mass flow of Liquid 2 [kg/s]  M& liq    = Design mass flow relation of Liquid 1 and air [-]  M&  a  d Mass flow of cooling liquid [kg/s] m& w =

XI

M& w = n speed =

Mass flow of cooling liquid [kg/s] Speed of fan [s-1]

n speed ,d =

Design speed of fan [s-1]

∆p F = ∆p F , d =

Pressure difference in cooling tower fan [Pa] Design pressure difference in cooling tower fan [Pa]

PF = Pliq1 =

Fan electrical power [W] Electrical power for the pump in Circuit 1 [W]

∆p liq1 =

Pressure difference in pump in Circuit 1 [Pa]

Pliq 2 =

Electrical power for the pump in Circuit 2 [W]

∆pliq 2 =

Pressure difference in pump in Circuit 2 [Pa]

Pspray =

Electrical power for the spray water pump [W]

∆pspray =

Pressure difference over spray water pump [Pa]

Ptotal =

Total electric power ( PF + Pliq1 + Pliq 2 + Pspray )

Q= dQ = Q& = Ta ,in , db =

Heat flux through a control volume [W] Heat flux through a control volume [W] Cooling capacity [W] Design cooling capacity [W] Dry bulb temperature of ambient air [°C]

Ta ,in ,db ,d =

Design dry bulb temperature of ambient air [°C]

Ta , out , db =

Dry bulb temperature of ambient air [°C]

Ta , in , wb =

Wet bulb temperature of entering ambient air [°C]

Q& d =

[W]

Ta ,in ,wb ,d = Design wet bulb temperature of ambient air [°C] Ta ,out ,wb =

Wet bulb temperature of leaving air [°C

∆Tln = Tliq1, w =

The log-mean temperature difference Temperature of the warmer side of Liquid 1 [°C]

Tliq1,c =

Temperature of cold side Liquid 1 [°C]

Tliq1,c ,d =

Design temperature of cold side in Liquid 1 [°C]

Tliq1, w,d =

Design temperature of warm side in Liquid 1 [°C]

Tliq 2,in =

Temperature of incoming (return) Liquid 2 [°C]

Tliq 2,out =

Temperature of outgoing (supply) Liquid 2 [°C]

TA , d =

Design temperature approach [°C]

TR ,d =

Design temperature range [°C] Design temperature difference between chilled water and room air for a room cooling device, e.g. a cooling beam [°C] (average water temp. and air temp. at 1,1m above floor level according to NordTest method NT VVS 078 ed. 2) Water temperature in control volume [°C]

∆Tm =

Tw =

XII

dTw = dTw =

Temperature of the water film in a control volume [°C] Temperature change of cooling liquid in the control volume [°C]

∆Twb = Twb = U= U fic =

Wet bulb temperature difference of moist air in the cooling tower [°C] Wet bulb temperature of moist air in control volume [°C] Overall heat transfer coefficient (U-value) [W/ m2, °C] Fictitious U-value [W/ m2, °C]

UAe = UA = V= V&spray =

Effective heat transfer coefficient-area product [W/°C] Heat transfer coefficient-area product [W/°C] The volume of the cooling tower fill [m3] Volume flow through spray water pump [m3/s]

w=

Humidity ratio of ambient air [kg water/kg dry air]

Greek letters Total pump efficiency in Circuit 1 [-] η liq1 =

ηliq 2 = η spray =

ηF = η F ,d = η hex = ρ liq1 = ρ liq 2 = ρa =

Total pump efficiency in Circuit 2 [-] Total efficiency in spray water pump [-] Total fan efficiency [-] Design total fan efficiency [-] Efficiency of heat exchanger [-] Density of Liquid 1 [kg/m3] Density of Liquid 2 [kg/m3] Air density [kg/m3]

XIII

XIV

Glossary Approach

The different approach temperatures are illustrated below in a heat exchanger diagram. Sometimes approach is called grädigkeit in heatexchanger literature. Temperature

Return side Warm side

Wet bulb temp. out

Secondary range Liquid in secondary circuit (water)

Primary range

Supply side

Secondary Liquid in primary circuit Cold approach side Primary approach Moist air

Total approach

Wet bulb (Ambient air) temp. in Area

ASHRAE CAV CIBSE COP CTM DOE 2 Energy Plus NMF Range SPF TMY-year TRY-year WBT dWBT

American Society of Heating, Refrigeration and Air-Conditioning Engineers Constant Air Volume, Ventilation system with constant air flow The Chartered Institution of Building Services Engineers, UK Coefficient of Performance The Cooling Tower Model developed in this thesis Building simulation tool issued by US Department of Energy, USA Building simulation tool issued by US Department of Energy, USA (successor of DOE 2) Neutral Model Format, A model language which the CTM are written in See Approach and diagram above Seasonal Performance Factor (sometimes denoted SEER, Seasonal Energy Efficient Ratio) Test Meteorological Year Test Reference Year Wet Bulb Temperature Design Wet Bulb Temperature

XV

XVI

1

Introduction This chapter gives an introduction to this thesis by discussing the motivation, scope and limitation of this project. Previous work made by researchers is furthermore discussed as an introduction to the field of study. Finally, the outline of the thesis is presented with recommendations of reading for different categories of readers.

1.1 Motivation of study The use of electricity in commercial buildings has increased over the last decades (CIBSE, 2004). The main causes are the increased amount of office equipment in the commercial sector together with a rapid escalation of air conditioning equipment for handling the high internal heat gains. Adnot (2002), reports an increase in building floor area supplied with central air-conditioning in the EU, by 200% from 1990 to 2002. The growth in air-conditioned building floor area from present to 2020 is predicted to be an additional 60%. Increased use of air-conditioning leads to higher demands of electric energy, which will inevitably lead to higher emissions of carbon dioxide in the EU. This is quite contrary to the intentions laid down in the Kyoto protocol. To subdue further increase and reduce the energy demand in the building sector in the EU, a directive concerning the energy performance of buildings was launched in December 2002 (EU, 2002) and legislated in 2003. The intention with the directive is to reduce the energy use in buildings in the EU, thus reducing the import of fossil fuels and increasing the prospect of reaching the goals set in the Kyoto protocol. The use of electricity in commercial buildings for air-conditioning and comfort cooling plays a non-negligible role. According to CIBSE (2004), the share of total electric energy use for air-conditioning for offices, industrial buildings and other buildings with mixed use in the UK is between 2 – 20%. Figures in Sweden for office buildings show similar values, 10 – 30% (Energimyndigheten, 2007). Together with other major categories of electricity end-users, such as lighting, fans and office equipment, air-conditioning is a category which definitely plays an important role in the demand for electricity in most commercial buildings. Conventional air-conditioning and cooling systems not only use electric energy, which has an environmental impact, they are also a source of emission of environmentally hazardous gases from refrigerants. Parallel with the discussion about the increase in the use of electric energy in buildings, an ongoing process of phasing out the use of environmentally hazardous refrigerants is carried out on an international basis, based on the Vienna convention and the Montreal protocol. To explore new and environmental friendly ways to control excess temperatures in buildings therefore serves dual purposes, i.e. reduction of the use of electricity and environmentally hazardous refrigerants.

1

In this thesis, a new application of already established cooling techniques is investigated. The application comprises the use of a cooling tower, as the sole provider of chilled water, in conjunction with a hydronic cooling system with chilled ceilings or chilled beams. Cooling towers is a technique which originates from the late 19th century and is mainly used as heat rejection in industrial applications or in chiller systems. Hydronic cooling systems with chilled ceilings or chilled beams was introduced as a technique a few decades ago. These systems became increasingly popular during the 1990’s and are today the predominant system in Swedish commercial buildings. In other northern European countries this type of system has a growing market share. Hydronic cooling systems with chilled ceilings or chilled beams are normally operated with a chiller based on vapour compression technique. Although the common use of a conventional chiller, these systems are sometimes classified as a low energy cooling system per se, compared to a CAV-system (Liddament, 2000; Behne, 1997). A cooling system comprising a cooling tower, as the sole provider of chilled water, and a hydronic cooling system with chilled ceilings or chilled beams, is a low energy alternative and completely free of refrigerants, thus interesting to further explore. Apart from the environmental advantages, one key feature with this system is its high level of applicability in both existing and new buildings. Both parts in the system, the cooling tower and the system with chilled ceilings or beams, are well-established techniques. When refurbishing commercial buildings in Sweden and adding comfort cooling, systems with chilled ceiling or chilled beams are the predominant choice. As discussed in section 1.3 Previous work, there is a limited knowledge of what indoor climate such cooling system can provide in commercial buildings together with the annual use of energy. There is however indications of a satisfying indoor climate and a low energy use in the published literature. Further study of the possibilities and limitations of this system is therefore motivated.

1.2 Scope and limitations This thesis explore the possibilities and limitations of a comfort cooling system consisting of a cooling tower, as the sole provider of chilled water, in conjunction with a hydronic cooling system with chilled beams. The basic hypothesis is that such a system in many cases can be a serious alternative to conventional comfort cooling systems with a mechanical chiller, especially in climates similar to conditions in the northern Europe, i.e. north of latitude 48 – 49° N in Europe. The objective is to investigate the resulting indoor thermal climate and annual energy use in a commercial building with such system applied. The investigation is carried out by using an advanced building simulation program. Monitoring of temperatures and other variables on a free cooling system as described in an existing office building has also been made.

2

1 Introduction

The building type, which is used in the simulations, is a typical office building. The use of the cooling system is however not restricted to offices only but can also be applied in other types of commercial buildings. The investigation is limited to climates similar to conditions in the northern Europe, i.e. north of latitude 48 – 49° N in Europe, and is only discussing comfort cooling systems. Cooling of products or processes as well as the use of water for evaporation is excluded.

1.3 Previous work Many contributions have been made in the research for environmental-friendly and CFC-free alternative cooling techniques and systems. Among many others the IEA ECBCS annex 28 (Liddament, 2000) and the Pascool project (Santamouris, 1995) have pointed out the variety of alternative cooling principles and cooling systems that can be used in buildings. The use of a cooling tower as a free cooling source, however in conjunction with a vapour compression chiller, in all-air systems has been discussed by many, e.g. Goswami and Reveliotty (1987), Mumma et al. (1990), Murphy (1991), Hipskind et al. (1991) and Hensley (1994). Research dealing with the application of cooling towers as the sole free cooling source in air-water systems, i.e. systems comprising chilled ceilings or chilled beams for cooling, is however limited. The earliest found is by Niu and Kooi (1993) and Niu et al. (1995). Their work consisted of measurements of chilled ceilings in a laboratory and development of a simulation program, ACCURACY, with the aim of studying the performance of chilled ceilings. In their work, they also studied the performance of chilled ceilings in connection with an evaporative cooling tower. The modelling of the cooling tower was however very simplified; the supply cooling water temperature was defined as the sum of the ambient wet bulb temperature and a constant approach temperature. They found that with an internal heat gain of 50 W/m2 and a constant approach temperature of 4°C or 60 W/m 2 and 2,5°C, the indoor thermal climate on a yearly basis was acceptable for a Dutch climate, i.e. the indoor operative temperature is not exceeding 25°C with more than 50 – 100 hours a year. Almadari et al (1998) report about a test performed by BRE (Building Research Establishment, London UK) where an office room was equipped with a chilled ceiling connected to a cooling tower. At a sensible heat gain of 60 W/m2, a maximum indoor air temperature of 25,2°C was registered. Between 1997 - 2000 a partly EU financed research project under the Joule IV program, called EcoCool, was conducted with Portugal, Switzerland, Germany, United Kingdom and Finland as participating countries. The objectives of the project were: 1. To improve and optimize an existing closed wet cooling tower technology to suit a system with chilled ceilings. 2. To develop a simple and cheap control strategy for the system considering the building mass.

3

3. To improve the system with regard to minimizing investment and running costs as well as energy consumption. 4. To gain knowledge of the applicability of the system, depending on climate and use of the building. 5. To give a simple guideline, which show how to install and run the system, to building services, consultants and HVAC installers. The EcoCool project generated a number of published articles; Gan and Riffat (1999), Facão and Oliveira (2000), Sprecher et al (2000), Oliveira (2000), Gan et al. (2001), Hasan and Sirén (2002) and Hasan and Gan (2002). Gan and Riffat (1999) use computational fluid dynamic (CFD) calculations to predict the thermal performance of a closed wet cooling tower for chilled ceiling system. Facão and Oliveira (2000) carry out measurements of a closed wet cooling tower for chilled ceiling systems and obtain correlation factors which were used in existing thermal models for such cooling towers. The article of Sprecher et al (2000) is more of a description of the cooling system used in the EcoCool project. However, they report of measurements of indoor temperature in an office building equipped with a chilled ceiling system connected to a cooling tower. During the hottest days of the measurement period from August to October 1999 with ambient air temperatures up to 32°C, the maximum indoor temperature was 27°C. Indoor temperatures over 26°C occurred, but only for a limited time. The report by Oliveira (2000) is the final and official report from the Ecocool project. Gan et al. (2001) use CFD calculations to predict the thermal performance of two closed wet cooling towers for a chilled ceiling system, and compare with experimental measurements. This work also includes optimization of the performance of one of the cooling towers. Hasan and Sirén (2002) use measurements from Facão and Oliveira (2000) to validate a thermal model for a closed wet cooling tower. The model is then used to optimize the cooling tower for the required cooling load to achieve a high coefficient of performance (COP) for the cooling tower. The performance of the cooling system to cool an office building is also investigated using the TRNSYS simulation program. The outcome of the simulations is however described very briefly, as well as the prerequisite conditions. Hasan and Gan (2002) compare analytical models with the use of CFD for a closed wet cooling tower for chilled ceiling system, and find good agreement with the two calculation approaches. All articles with connection to the EcoCool project, with the exception of Hasan and Sirén (2002) and Sprecher et al (2000), mainly discuss the potential for a cooling tower to deliver chilled water in the range of 18-20°C, with a focus on detailed analysis of heat transfer characteristics of a cooling tower. Another research project with the objective to examine the ability of a cooling tower to deliver chilled water in the range of 18-20°C and other properties of a tower has been carried out and published by Costelloe and Finn (2000), (2001), (2003), (2007) and (2009). Costelloe and Finn (2000) describe findings from measurements on a laboratory test rig. The rig consists of an open counter flow cooling tower, a primary and secondary circuit with an intermediate heat exchanger. The aim of the test facility is to deliver cooling water at low approach temperatures for free cooling opportunities with

4

1 Introduction

chilled ceiling panels and chilled beams. Costelloe and Finn (2001) are focusing more on the energy performance of the cooling tower test rig and publish outcome from measurements concerning Primary energy rate, which is the inverse of the COP of the cooling tower. Costelloe and Finn (2003) use the measurements from the test rig and climate data of Dublin, Ireland and Milan, Italy, to calculate the availability of maintaining a certain cooling temperature in a coolant at these climate locations and at a total approach temperature of 3 K. The paper Costelloe and Finn (2007) outlines how the thermal effectiveness of a cooling tower are affected by five key operating parameters; load, ambient wet bulb temperature, primary and secondary water-flow rate and air-flow rate. In Costelloe and Finn (2009) they discuss heat transfer correlations from experimental results on their test rig. They also publish correlation coefficients for calculating the total heat transfer coefficient in a cooling tower. The most comprehensive published material concerning the resulting indoor climate is published by Bohler et al. (2002). They investigated the performance of a cooling system comprising an open evaporative cooling tower in conjunction with chilled ceilings, however with an intermediate heat exchanger to avoid fouling in the chilled water circuit. The investigation was made using the simulation tool ConsoClim. The research work included analysis of the operative temperature in an office at a number of different prerequisites, e.g. three design cooling capacities of the chilled ceilings, two building inertia, two solar gains, two internal heat gains, two orientations (east and west) and three different climates (locations of Trappes, Nice and Carpentras, all in France). The results from all the simulation runs are quite extensive. The resulting maximum indoor operative temperature in Trappes, located close to Paris, is in the range of 25 – 28°C when maximum cooling loads were 40 – 70 W/m2. For Nice, with a seaside Mediterranean climate, and Carpentras, with an inland Mediterranean climate, the maximum cooling loads were in many cases above 70 W/m2, hence the operative temperatures were in the range 27.5 – 30°C. However, when the maximum cooling loads for these locations were between 50 – 70 W/m2, the operative temperatures were in the range 25,5 – 27,5°C with only a few exceptions. Bohler et al. (2002) also presented the operative temperature for each case when there was no cooling of the room. Sprecher and Tillenkamp (2003) published an investigation on a system with a closed evaporative cooling tower and a water circuit embedded in the concrete slab. They used the simulation program TRNSYS for investigating such a system. Sprecher and Tillenkamp processed dynamic simulations of a large shopping centre, 10 000 m2, in Lucerne, Switzerland, with internal heat gain from people and lighting at 42 W/m2. During the warmest weak of the year, the indoor air temperature was kept below 26°C. Hasan et al. (2007) uses a simulation tool (IDA-ICE) to find out the best performance of a cooling tower combined with chilled ceiling, serving a four storey residential building. The highest yearly mean COP achieved was 8,3. It can be concluded that there are few published data concerning the resulting indoor climate in buildings equipped with chilled ceilings or chilled beams connected to an evaporative cooling tower, especially where the cooling tower is the sole provider of chilled water. The available data indicate the possibility to achieve an indoor thermal

5

climate during hot periods that in most cases can be acceptable. The sparse data and the limited experience of this type of cooling system is one, of many, reasons why conventional vapour compression cooling systems still is dominating in the building stock. In light of the sparse but promising results from published works on the above mentioned cooling system, it is therefore important to further investigate the resulting indoor thermal climate and annual use of energy under different conditions.

1.4 Outline of this thesis This thesis comprises nine main chapters and one appendix. In this section, each chapter is briefly described to give an overview of the thesis and its content. Chapter 1: Introduction This chapter gives an introduction to the thesis by discussing the motivation, scope and limitation of this project. Previous work made by researchers is furthermore discussed as an introduction to the field of study. Finally, the outline of the thesis is presented. Chapter 2: Cooling towers – introduction and modelling The chapter introduces cooling towers and air-cooled heat exchangers. It also includes a brief introduction to the physical and mathematical modelling of cooling towers. The focus is on the effectiveness – NTU method as a theoretical base for the development of the cooling tower model in chapter 4. Furthermore, the evaporative cooling tower process is visualized to facilitate a better understanding of the complex nature of this process. The chapter is finished with a short discussion about the risk of growth of legionella bacteria in a cooling tower applied as a sole free cooling source in a hydronic comfort cooling system. Chapter 3: Cooling systems with evaporative cooling tower The chapter discuss different system configurations of cooling systems with evaporative cooling towers for free cooling in combination with a conventional chiller. Chapter 4: Simulation model development In this chapter a cooling tower model is developed. The development is introduced with a discussion about the prerequisites followed by a discussion about the features of the model and a description of its equations. The chapter is finished with a presentation of a validation of the cooling tower model. The model is validated against both published data and measured data by the author. Chapter 5: Simulation of a free cooling system with evaporative cooling tower Here the basis and the prerequisites for the simulation of a cooling system with an evaporative cooling tower are laid out. The simulation methodology and the parameter analysis methodology are described including a discussion and motivation of the considered parameter alterations.

6

1 Introduction

Chapter 6: Monitoring of a free cooling system with evaporative cooling tower This chapter describes the monitoring of a free cooling system with an evaporative cooler in a pilot plant. The system lay-out and the monitoring system are introduced. Chapter 7: Results and analysis of simulations The results from the simulations described in chapter 5 “Simulation of a free cooling system with evaporative cooling tower” are presented and analyzed in this chapter. The results are separated into sections including Cooling tower performance, Indoor thermal climate and Energy use. The chapter also contains relevant findings from other researchers. Chapter 8: Results and analysis of measurements The measurement results from the pilot plant described in chapter 6 “Monitoring of a free cooling system with evaporative cooling tower” are presented and analyzed in this chapter. The results are separated into sections including Ambient conditions, Indoor thermal climate and Free cooling system. Chapter 9: Conclusions and discussion In this chapter, conclusions from previous research and the results in this thesis are presented. Conclusions are followed by a discussion concerning various aspects of comfort cooling in general and the comfort cooling system examined in thesis in particular.

7

8

2 Cooling towers – introduction and modelling Although the focus is on cooling towers the title also cover air-cooled heat exchangers. The chapter gives a short introduction to cooling towers and aircooled heat exchangers. It also includes a brief introduction to the physical and mathematical modelling of cooling towers with focus on the effectiveness – NTU method. This method forms a theoretical base for the development of the cooling tower model in chapter 4. Furthermore, the evaporative cooling tower process is visualized to facilitate a better understanding of the complex nature of this process. The chapter is finished with a short discussion about the negligible risk of growth of legionella bacteria in a cooling tower applied as a sole source in a comfort cooling system. For more in depth studies of cooling towers, Kröger (1999) and Hill et al. (1990) are recommended. A more detailed description of the effectiveness – NTU method can be found in most engineering handbooks, e.g. ASHRAE (2001a)

2.1 Introduction Air-cooled heat exchangers and cooling towers have many applications in a broad range of industrial and commercial areas. In industrial processes like power plants, chemical and process plants and computer centrals as well as in buildings equipped with a refrigeration process, excess heat has to be discharged. The excess heat can be discharged to different type of heat sinks, e.g. to a water mass like a river or lake, to the ground water or an aquifer or to the ambient air. When excess heat is discharged to the ambient air an air-cooled heat exchanger or a cooling tower is used. The difference between an air-cooled heat exchanger and a cooling tower is not exactly defined in the literature. A simple but not exact characterisation between the two is that an air cooled heat exchanger is based on exchange of sensible heat while a cooling tower exchange heat via both sensible and latent heat. In practical terms an air-cooled heat exchanger is based on a coolant in a closed loop typically connected to a finned tube heat exchanger, see figure 2.1. The coolant, usually water, is never in direct contact with the ambient air. The coolant is chilled only through heat transfer driven by the difference of the temperature of the coolant and the dry bulb temperature of ambient air. The air movement through the heat exchanger is often accomplished with one or several fans. An air cooled heat exchanger is sometimes called a fluid cooler. This type of cooler is occasionally used in Swedish commercial buildings as a device for free cooling. Normally it is used in connection with a conventional chiller which is activated when the outdoor temperature is above 10 - 12°C. At that temperature the air cooled heat exchanger cannot produce chilled water at the normal set-point temperature, i.e. 13 - 14°C.

9

Fan Finned tube air-coil Coolant in a closed loop system

Ambient air

Figure 2.1

Schematic illustration of an air-cooled heat exchanger

A cooling tower is a device, which typically uses a combination of heat and mass transfer to cool water, see figure 2.2. The water to be cooled is distributed in the tower by spray nozzles, splash bars or film fills in a manner that exposes a very large water surface to the ambient air. The cooling tower fill is sometimes also called packing. The movement of the air is achieved by fans (mechanical draft), natural draft or the induction effect from water sprays. A portion of the water is evaporated because the moisture content of the air is not saturated at the temperature of the water. Since this process of evaporation requires energy to change the water from liquid to vapour, heat is taken from the water, i.e. the water is cooled.

Spray nozzles

Fan

Coolant in an open loop system

Ambient air

Cold water basin

Figure 2.2

Schematic illustration of a cooling tower

Cooling towers can be categorized according to different criteria: By the air transport system

10

•

naturally ventilated towers (natural draught cooling towers)

•

mechanically ventilated towers (fan draught towers). Ventilation can be of either induced type or forced draught.

2 Cooling towers – introduction and modelling

By the air channelling system •

counterflow cooling towers

•

cross flow cooling towers

•

a combination of these two designs.

By the design of the heat exchanging surface •

open circuit systems, where water is cooled by direct contact with the surrounding air (wet type cooling tower). - water film fill - splash fill - cooling towers without fill for special applications

•

closed circuit systems, where the medium to be cooled is not in direct contact with the surrounding air - steel tube bundles with smooth external wall tubes suitable for external water spraying - finned tube bundles, generally used for dry towers only

•

hybrid cooling towers, a combination of wet and dry tower, with different fills and possibilities to switch between different modes of operation, i.e. wet only, dry only and a combination of wet and dry mode.

In this thesis the application of cooling towers is focused on cooling systems which are normally used together with building HVAC (Heating, Ventilation and Air Conditioning) and refrigerating systems. The following text is therefore narrowed to deal only with the types of cooling towers normally occurring in this context. Cooling towers in conjunction with building cooling systems can be divided in to open or closed towers. A schematic illustration of an open type of cooling tower is shown in figure 2.3. The water to be cooled is directly exposed to ambient air by either a falling water film in a fill arrangement or by droplets caused by an arrangement of splash bars. The air stream passing by catches small droplets, which are captured in a drift eliminator. In the drift eliminator, the small droplets accumulate to form larger drops that, by gravity, return to the fill or splash bars. The cooled water is collected in a basin in the base of the tower from where it is returned to the heat source.

11

Spray nozzles

Fan Drift eliminator

Fill arrangement Coolant in open loop system Ambient air

Cold water basin

Figure 2.3

Schematic illustration of an open type cooling tower

In the closed type cooling tower, figure 2.4, the water or coolant is kept in a closed loop and is not directly exposed to ambient air. The heat and mass transfer is taking place in a heat exchanger, usually a smooth tube heat exchanger, which is kept wet through spraying of water over the exchanger. This type of closed loop cooling tower is sometimes also referred to as an evaporative cooler.

Spray nozzles


Coolant in a closed loop system Ambient air

Spray water basin

Figure 2.4

Schematic illustration of a closed type cooling tower

The air movement is always facilitated by one or several fans regardless if it is an open or closed type cooling tower. The air can be either pushed or pulled through the tower. When pushed (by overpressure) the type is called forced draft and the fan(s) is placed where the ambient air stream enters the tower, see figure 2.5. The fan(s) forces the air

12


through the fill or the tube heat exchanger. With a forced type cooling tower, the heat exchange configuration is typically of counter flow type. Forced draft tower are characterized by relatively high air inlet velocities and low exit velocities and are thus susceptible to recirculation of the hot and moist plume air. A radial fan is the most common type of fan use in forced draft towers.

Spray nozzles Drift eliminator

Fill arrangement Coolant in an open loop system Ambient air

Fan Cold water basin

Figure 2.5

Schematic illustration of a forced draft cooling tower (illustrated with an open type cooling tower)

When the air is pulled through the cooling tower (by underpressure) the type is called induced draft, see figure 2.6. Induced draft towers may be of the counterflow or crossflow type. An axial fan is the most common type of fan used with induced draft towers and the fan is placed on top of the tower. Plume recirculation is a smaller problem in induced draft towers than it is in forced draft towers, since exit air velocity is higher. However, noise reduction is a more difficult task with this type.

13

Spray nozzles


Fill arrangement Coolant in an open loop system Ambient air

Cold water basin

Figure 2.6

Schematic illustration of an induced draft cooling tower (illustrated with an open type cooling tower).

As mentioned above one or several fans cause the air movement through the cooling tower. The fan(s) can be either single speed, dual speed or equipped with variable speed control (with a variable frequency drive, VFD). Adjusting the airflow through the tower is the normal and primary way to control the temperature of the leaving fluid. If a cooling tower has only one fan it is normally cycling on/off or cycling between two or three steps of speed to control the heat rejected. If a tower has several fans the heat rejected from the unit can be controlled either by cycling each fan on/off or between two or three steps of speed and operating the fans in several steps. The lowest operating cost is achieved with fans equipped with variable speed drives (ASHRAE/IESNA, 1999 and Stout & Leach, 2002). VFD-control of fan(s) has become more common in cooling tower applications in later years. In cold climates operation of cooling towers, a fog plume can occur which can be of significant size, especially at low temperatures and high humidity. For this reason and for water saving causes, cooling towers with basically two operation modes, dry or wet operation, has been developed. Cooling towers with two or more operational modes is often called hybrid cooling towers. An example of a principal hybrid cooling tower can be seen in figure 2.7. Wet mode is activated, i.e. the spray water pump is started, when the dry bulb temperature of the ambient air is too high for sufficient heat rejection. During wet mode the coolant can pass through the dry finned coil only, or continue through the wet surface coil using a modulating three-way valve. When the dry bulb temperature is low enough the operation can be switched to dry mode and the spray water pump is stopped.

14


Fan Dry finned coil Drift eliminator Coolant in closed loop system Spray nozzles Modulating threeway valve

Wet surface coil Ambient air

Wet deck surface Spray water basin Spray water pump

Figure 2.7

Schematic illustration of a hybrid type cooling tower. Note that hybrid cooling tower can have many different designs. This figure shows only the principle.

As mentioned earlier a cooling tower is a device which typically uses a combination of heat and mass transfer to cool water. This fact makes the physical and mathematical description more complex when engineers want to make calculations on a cooling tower. The following chapter gives a short overview of modelling heat and mass transfer in cooling towers, with focus on modelling based on the effectiveness – NTU method.

2.2 Modelling heat and mass transfer in evaporative cooling towers In the literature, many different mathematical models have been presented to predict the heat and mass transfer in an evaporative cooling tower. These models can be separated into two main categories; analytically based and empirically based. The analytically based can further be separated into two sub categories; those requiring numerical integration of one or more differential equations and those requiring evaluation of more simple equations. The empirically based models use vendor-supplied cooling tower performance data as the base for performing calculations. When calculating cooling tower performance, different kinds of regression-techniques are used. This approach makes this type of modelling and calculating very fast compared to the analytically based models. One

15

example of using an empirically based model is found in DOE 2 (Benton et. al., 2002) which is a well known and widely spread building simulation tool. Analytically based models including differential equations, on the other hand, require substantially more computer time to be solved (Benton et. al., 2002). Many of these models are very complex and require specific and comprehensive data to calculate the heat and mass transfer. The analytically based models with less complexity regarding model equations lies somewhere between the two other groups mentioned when it comes to computational speed. Two of the most known in this category is the Merkel model and the effectiveness – NTU method. The Merkel model developed by Fredrich Merkel in 1925 (Merkel, 1925) has been widely used when developing cooling tower models. Cooling tower performance curves and the CTI Toolkit, released by the Cooling Technology Institute (www.cti.org), is one example where applications are based on the Merkel model. The effectiveness – NTU method is a well-established method for calculating heat exchangers in general. This method is also applicable in combined heat and mass transfer calculations of evaporative coolers and cooling towers (Jaber and Webb 1989). The method enables the cooling tower/evaporative cooler to be treated as a “black box”, i.e. the physical properties such as size and kind of fill arrangement in an open type of cooling tower, or size and dimensions of the tube heat exchanger in a closed type cooling tower, may not be known when calculating the heat transferred. A cooling tower model to be used at practical HVAC design and engineering must have a limited complexity with a manageable amount of data, yet accurate enough for applications in building simulation. This is especially important in the early stages of the building design phase when often very little information is available concerning details of the building and its technical installations, including a cooling tower. A model based on the effectiveness – NTU method has several advantages compared to models that are more complex: - It requires relatively few input data prior to calculation - The same model can be used for calculating both wet cooling towers, with open or closed configuration, and dry cooling towers, i.e. air cooled heat exchangers. - The accuracy of a model based on the effectiveness – NTU method is sufficient for applications in a building simulation program. The accuracy of the model has been examined by Benton et al. (2002), see table 2.1. The error presented in table 2.1 is the difference between the supply cold water temperature predicted by the effectiveness – NTU method and that predicted by a cooling tower vendor. The figures in table 2.1 are based on 669 data points for counterflow towers and 347 data points for crossflow towers. For counterflow arrangement 95% of the errors is in the interval of –0,8°C to 1,2°C. For crossflow ditto 95% of the errors are in the interval of –1,0°C to 0,6°C. The accuracy of the model developed in this thesis, named CTM (Cooling Tower Model), which is also based on the effectiveness – NTU method, is examined in chapter 4.3 Validation of the model.

16


Table 2.1

Accuracy of the effectiveness – NTU method compared to vendor supplied data, (Benton et al., 2002)

Error [°C]

Counterflow cooling tower

Crossflow cooling tower

Average error

0,2

-0,4

Maximum error

3,9

2,3

Standard deviation

1,0

0,6

95% Confidence Interval

±2,0

±1,2

Bourdouxhe et al., (1994), who published the version of the effectiveness – NTU method which is used in this thesis, validated the model with catalogue data. The model was validated against a counter flow cooling tower manufactured by Baltimore Air Coil with nominal cooling capacity of 254 kW. The error presented in table 2.2 is the difference between the supply cold water temperature predicted by the effectiveness – NTU method and that predicted by Baltimore Air Coil in the same way as in table 2.1. The error figures are calculated by the author from 13 values of cold water supply temperature presented by Bourdouxhe (1994). Table 2.2

Accuracy of the effectiveness – NTU method compared to catalogue data, (Bourdouxhe et al., 1994)

Error [°C]

Counterflow cooling tower

Average error

0,09

Maximum error

0,44

Standard deviation

0,15

95% Confidence Interval

±0,16

Hernandez et al., (1994) validated the model published by Bourdouxhe et al., (1994), with experimental data. The maximum relative difference between calculated and measured exiting cold water temperature was 0,3 % using the same cooling tower as Bourdouxhe et al. (1994). The relative difference is defined as the difference between calculated and measured supply cold water temperature divided with the difference between entering (warm) water temperature and entering (ambient) air wet bulb temperature. The effectiveness – NTU method is used in cooling tower models in well known and wide spread building simulation programs such as TRNSYS (www.trnsys.com) and EnergyPlus (Energy Plus, 2004). The cooling tower model in ASHRAE Primary HVAC Toolkit package (ASHRAE, 1999) is also based on the effectiveness – NTU method.

17

2.2.1 The effectiveness – NTU method The effectiveness – NTU method applied on cooling towers and evaporative coolers is based on the theories of Merkel (1925). His equation states that the total heat transfer taking place at any position, e.g. a control volume, in the tower is proportional to the difference between the enthalpy of air saturated at the same temperature as the water film at one point and the enthalpy of the surrounding air at the same point.

dQ = K m (hs − ha )dA

(2.1)

Where Q = Heat flux through a control volume [W] K m = Mass transfer coefficient [kg/s, m2]

hs = Enthalpy of saturated air at the same temperature as the water film [J/kg] ha = Enthalpy of the surrounding air [J/kg] A = Area of control volume [m2] The difference (hs – ha) is the enthalpy driving potential proportional to the heat flux. The Merkel theory is based on the following assumptions: 1. The total heat flux is written in terms of an enthalpy difference of the moist air. 2. The effects of loss of water through evaporation are neglected in the energy balance equation. Also the effects of drift and blow down water loss are typically neglected. 3. The leaving moist air enthalpy is assumed a function of the wet bulb temperature. 4. The thermal resistance of the water film is neglected. 5. The Lewis number is equal to unity. The Merkel equation integrated for the whole cooling tower yields T

w2 dTw KaV = ∫ m& w h − ha Tw1 s

where K = Mass transfer coefficient [kg/s, m2] a = Fill interface area per fill volume [m2/m3] V = The volume of the cooling tower fill [m3] m& w = Mass flow of cooling liquid [kg/s]

dTw = Temperature of the water film in a control volume [°C] hs − ha = The enthalpy difference and driving potential [J/kg]

18

(2.2)


KaV is called the cooling tower coefficient or sometimes the Merkel m& w number. The derivation of equation 2.1 and 2.2 can be found in Merkel (1925) and also in Kröger (1999).

The expression

The first one to publish a correct derivation of a general effectiveness – NTU method for an evaporative cooling tower from the works of Merkel was Jaber and Webb, (1989). Later Braun et al., (1989) and Bourdouxhe et al., (1994) also presented a cooling tower model based on the effectiveness – NTU method. The main part of the following derivation of a cooling tower model, based on the effectiveness – NTU method, is from Bourdouxhe et al., (1994). In a control volume in an evaporative cooling tower, the following relation can be stated: dQ = M& w c p ,w dTw = M& a dha

[W]

(2.3)

where

dQ = Heat flux through a control volume [W] M& w = Mass flow of cooling liquid [kg/s] c p , w = Specific heat of cooling liquid [J/kg, °C] dTw = Temperature change of cooling liquid in the control volume [°C] M& a = Mass flow of air [kg/s] dha = Enthalpy change of air in a control volume [J/kg] Based on Merkel’s theory and the assumption that Lewis number is equal to unity, the total heat transfer (sensible plus latent) between air and water in the control volume may be defined as follows:

dQ =

U ⋅ dA (hs − ha ) c p ,a

[W]

(2.4)

where

U = Overall heat transfer coefficient (U-value) [W/ m2, °C] dA = Wet surface area in the control volume [m2] c p ,a = Specific heat of dry air [J/kg, °C] It is assumed that the moist air enthalpy can be defined by the wet-bulb temperature. To enable calculations of moist air enthalpy using the wet-bulb temperature, a different definition of the specific heat of the moist air must be used. The moist air is therefore

19

treated as a fictitious ideal gas characterized by the following fictitious mean specific heat: _

c p , fic =

∆h ∆Twb

[J/kg °C]

(2.5)

where _

c p , fic = Fictitious mean specific heat [J/kg, °C] ∆h = Enthalpy difference of moist air over the cooling tower [J/kg] ∆Twb = Wet bulb temperature difference of moist air over the cooling tower [°C]

The liquid side conductance is much greater than the air side conductance. Therefore, the wetted surface temperature is also assumed equal to the water temperature. Based on these assumptions, the expression in (2.4) and (2.5) yields: dQ = U fic ⋅ dA(Tw − Twb )

[W]

(2.6)

where U fic = Fictitious U-value, defined in equation 2.7 [W/ m2, °C]

Tw = Water temperature in control volume [°C] Twb = Wet bulb temperature of moist air in control volume [°C] and

U

fic

=

U ⋅ c p , fic

[W/ m2, °C]

c p ,a

(2.7)

The energy balance equation (2.3) can now be written as follows: _

dQ = M& w c p ,w dTw = M& a c p , fic dTwb

[W]

(2.8)

Equation (2.8) can be rearranged

dTw =

dQ M& w c p ,w

[°C]

(2.9)

and dTwb =

20

dQ M& a c p , fic

[°C]

(2.10)


Now subtract (2.9) with (2.10), which gives

 1 1 − dTw − dTwb = d (Tw − Twb ) = dQ&  &  M& c  w p ,w M a c p , fic

   

[°C]

(2.11)

Combine (2.6) with (2.11) gives

 1 d (Tw − Twb ) 1 = U fic  − & &  Tw − Twb  M w c p ,w M a c p , fic

 dA  

[-]

(2.12)

This equation is applicable in an evaporative cooling tower. It also corresponds with the equation that occurs in the effectiveness-NTU development for an ordinary heat exchanger transferring sensible heat only:

 d (TH − TC ) 1 1 =U − &  M& c TH − TC  H p , H M C c p ,C

 dA  

[-]

(2.13)

where index H stands for the hot side and index C stands for the cold side of a simple heat exchanger. Consequently, the cooling tower can be modelled, in steady state regime, as a fictitious indirect contact heat exchanger, see figure 2.8. Twb,in cp,fic

Twb,out Air side AUfict

Tw,c

Figure 2.8

Liquid side

Tw,w cpw

Fictitious indirect contact heat exchanger

The two fluids can now be treated in the same way as in ordinary heat exchangers models. The first fluid is water (or any liquid) and the second fluid is a fictitious fluid entering the heat exchanger at the temperature Twb,in and characterized by the fictitious specific heat cp,fic . The heat exchanger is characterized by one parameter, its global heat transfer – area product, AUfic. The actual cooling tower heat transfer–area product, AU, is related to AUfic by the following expression.

21

AU = AU fic

c p ,a

[W/°C]

c p , fic

(2.14)

Consider a case with a counterflow cooling tower, see figure 2.9.

Temperature Tw,w

Liquid to be cooled, e.g. water Twb,out Tw,c Twb,in

Moist air 0

100%

Area

Figure 2.9 Temperature diagram of fictitious indirect contact counterflow heat exchanger

If m& w c p ,w is the smaller heat capacity rate, the effectiveness, ε, of a cooling tower can be defined by analogy with the effectiveness of a simple heat exchanger.

ε=

Tw,w − Tw,c

[-]

Tw,w − Twb ,in

(2.15)

or as a function of CR and NTU

ε=

1 − exp[− NTU (1 − C R )] 1 − C R ⋅ exp[− NTU (1 − C R )]

where C R =

22

C min C max

(counter flow) [-]

[-]

(2.16)

(2.17)

C min = min (M& w ⋅ c p ,w , M& a c p , fic )

[W/°C]

(2.18)

C max = max (M& w ⋅ c p ,w , M& a c p , fic )

[W/°C]

(2.19)


and NTU =

U fic ⋅ A C min

[-]

(2.20)

In the case with a crossflow cooling tower, Jaber and Webb (1989) recommends use of the unmixed/unmixed ε-NTU relation.

 exp(− NTU ⋅ C R ⋅ k ) − 1   CR ⋅ k  

ε = 1 − exp

(cross flow)

[-]

(2.21) where

k = NTU −0, 22

(2.22)

A big advantage with applying the effectiveness – NTU method on a cooling tower model is that it is applicable on both open circuit and closed circuit cooling towers as well as air cooled heat exchangers. This is utilized in the simulation model, called CTM (Cooling Tower Model), presented in chapter 4 Simulation model development.

2.3 Visualizing heat and mass transfer in an evaporative cooling tower To facilitate the understanding of the complex nature of the coincident heat and mass transfer in an evaporative cooling tower, the process can be visualized in a diagram. In this case, the enthalpy – humidity diagram (Mollier chart) in figure 2.10 can be illustrative. In figure 2.10, an example of the evaporative cooling tower process is shown. In this case, a counter flow configured tower is used as an example. The cooling of the liquid, e.g. water, is illustrated as a line on the saturation curve from 19°C (Tliq,in) to 15°C (Tliq,out) . The ambient air is entering the cooling tower at 15°C and 50% RH, which equals a wet bulb temperature of 9,7°C (Twb,in). The wet bulb temperature of the ambient air is the lower limit to which the coolant can be chilled in an evaporative process. The exhaust air from the cooling tower has a dry bulb temperature of 17°C and a relative humidity of about 90% RH, which equals a wet bulb temperature of about 16°C (Twb,out).

23

humidity (kg/ kg) 0,000 25

0,005

0,010

0,015

0,020

50% RH

Range

100%RF

Liquid

Saturation curve (100% RH)

Moist air

Tliq,in 20 Tliq,out 15

Twb,out

10 5

Enthalpy difference of entering and leaving moist air, equal to total heat exchange rate

Approach

0

Twb,in

-5 -10 -15 -20

Figure 2.10

Enthalpy – humidity diagram (Mollier chart) with an illustrated example of the evaporative cooling tower process

By using the concept of the effectiveness – NTU method, described in section 2.2.1 The effectiveness – NTU method, the example of an evaporative cooling tower process in figure 2.10 can be visualized in an ordinary heat exchanger diagram in figure 2.11. Temperature 20°C Tliq,in Twb,out

Range Liquid

15°C

Tliq,out

Approach Moist air 10°C

Twb,in Area

Figure 2.11 Temperature diagram of a counter flow cooling tower. Example from figure 2.10 is visualized together with definitions of Approach and Range.

24


In figure 2.11 the concept definitions of Approach and Range are visualized. The terms Approach and Range are commonly used in the cooling tower literature. Approach is simply a temperature difference showing how close the temperature of the leaving liquid is to the theoretical limit of the wet bulb temperature of the entering moist air. Range is the difference between the temperature of the entering and leaving liquid.

2.4 Legionella in evaporative cooling towers It is beyond the scope of this thesis to discuss the whole issue about the risk of spreading legionella bacteria through aerosols originating from cooling towers and related problems with cooling water treatment. However, the issue of legionella and cooling towers has been well known within the HVAC society since the first known epidemic outbreak in 1976 in Philadelphia, USA. Cooling towers in general have since then been considered as a potential source of legionella bacteria. There are many factors that influence the rate of growth, or killing, of the legionella bacterium. The presence of water, nutriment, oxygen, time for growth, still water, microbial biofilm and pH-value between 5,5 – 9,2 together with right temperature conditions are some of the most important parameters for a positive environment for the growth of the legionella bacterium (Stålbom and Kling, 2002). The factor, of the above mentioned, that is significantly different from conventional use of a cooling tower compared to a cooling tower applied to a hydronic comfort cooling system as a sole free cooling source, is the temperature conditions. The temperature range with conventional use of a cooling tower, e.g. as a condenser or in an industrial process, is typically between 30°C – 45°C. The temperature range of a cooling tower in the context in this thesis is typically considerably lower. The hydronic cooling system temperatures is normally above the indoor dew-point temperature, but low enough to provide sufficient cooling in the room cooling devices, hence the temperature range is typically 13°C – 20°C. As seen in figure 2.12 the growth rate of legionella bacteria in conventional use of cooling towers is high or near the maximum rate, whereas for the application of cooling towers in the context of this thesis, the bacteria is in resting mode. Thus, there is a considerable difference in risk for bacteria growth between these two cooling tower applications. It can therefore be concluded that the risk for growth of legionella bacteria in a cooling tower applied as a sole free cooling source in a hydronic comfort cooling system is reasonably small.

25

Cooling tower as a sole free cooling source Conventional use of cooling towers

Growth

Temperature

Growth – killing rate curve for legionella bacteria

Killing

Resting

Growth

Killing

Figure 2.12 Growth or killing rate for legionella bacteria at different temperatures (Stålbom and Kling, 2002)

Further reading in this matter can for example be found in ASHRAE (2000b), and Stålbom & Kling (2002).

26

3 Cooling system with evaporative cooling tower In this chapter, system aspects of cooling systems with evaporative cooling towers for free cooling combined with a conventional chiller are discussed.

3.1 Introduction This thesis explores the ability of an evaporative cooling tower to be the sole provider of chilled water to a hydronic cooling system with chilled beams. This application can clearly work for buildings with sufficient prerequisites, see chapter 5 Simulation of a free cooling system with evaporative cooling tower, 7 Results and analysis of simulation and 8 Results and analysis of measurements. However, when prerequisites are not favourable, e.g. high cooling loads, strict temperature control, warm and humid climate or need for dehumidification, the cooling system can be supplemented with mechanical cooling. In cases when a cooling system consists of both a conventional chiller, based on vapour compression, and an evaporative cooling tower for free cooling opportunities, the system configuration can have different layouts. The most common configurations are discussed in section 3.2 System configurations. However, in section 3.2 System configurations only water-side applications are discussed, since this thesis is focusing merely on hydronic cooling systems. A discussion concerning evaporative air-side free cooling is outside the scope of this thesis and can be found in De Saulles (1996) and Lindholm (2000). The question how big portion the free cooling part, i.e. the cooling tower, of the cooling system can provide of the total annual cooling needs is of vital importance. The answer is dependent on a number of factors. The most important are: The annual variation and duration of the wet bulb temperature for a given location; the ambient wet bulb temperature constitutes the lower natural limit for the provision of chilled water from an evaporative cooling tower and varies with climate and geographical location. System configuration; different system configurations can provide different amount of free cooling on a yearly basis, see section 3.2 System configurations. Required chilled water temperatures; the higher the required chilled water supply temperature the higher the portion of annual free cooling from the cooling tower. If dehumidification is required in the air handling unit a lower supply and return temperature for the air cooling coil will be necessary, e.g. 6 – 7°C in supply and about 10 - 12°C in return. The room cooling devices have higher temperature, normally 13 – 15°C, for chilled ceilings, chilled beams or fan coils. In this case the cooling coil will have a shorter utilization time than the room cooling devices. The set point for the chilled water temperature is normally constant throughout the year. However, there is usually no need for a constant temperature in an ordinary comfort cooling system. Hence, the temperature can be allowed to rise with for example

27

lower ambient temperature. Variable chilled water temperature prolongs the potential free cooling time. Design conditions of the cooling tower, i.e. primary approach and range at design wet bulb temperature; primary approach and range at design temperature determines the size of the cooling tower and the ability to chill the cooling water. A further discussion about this topic is given in section 5.4.2 Discussion and motivation of parameter alterations; Design cooling capacity (cooling tower). The actual approach varies with the ambient wet bulb temperature and the cooling load, see figure 3.1. Cooling load characteristics; how the cooling load varies diurnal and annual determines the required load on the cooling tower. The actual approach temperature, and hence the possible cooling water temperature, is dependent on the cooling tower load; the lower the load, the smaller the approach. Conditions in control system; consists of different control regimes and its present conditions. When the chiller starts and the system switch over from free cooling to conventional cooling operating can be determined by the ambient wet bulb temperature, the cooling system supply temperature or a representative indoor temperature. Depending on the temperature chosen as the switch over condition, it will influence the annual free cooling potential. An example of the relation between the actual approach and the ambient wet bulb temperature at two different ranges is displayed in figure 3.1. It should be noted that the curves varies with the design conditions of the cooling tower.

Approach temperature (°C)

12

Full load and a temperature range between supply and return water at 4,4 °C (8°F)

10 8

Half load and a temperature range between supply and return water at 2,2 °C (4°F)

6 4 2 0 0

5

10

15

20

25

Ambient wet bulb temperature (°C)

Figure 3.1

28

Relation between actual approach temperatures and ambient wet bulb temperature for an evaporative cooling tower. NB: the curves and conditions in the figure is an example of a possible case. (data from Marley (1982))

3 Cooling system with evaporative cooling tower

A complication with evaporative cooling towers is that the performance, i.e. the ability to chill water to a certain degree, varies with the ambient wet bulb temperature, which is shown in figure 3.1. The performance also varies with the actual cooling load as shown in figure 3.1. Furthermore, the performance is also determined by the design conditions for the cooling tower, i.e. the cooling tower design approach and design range at a certain design ambient wet bulb temperature. In figure 3.2, the duration of the ambient wet bulb temperature and associated cooling water temperatures from an evaporative cooling tower are shown. The available supply cooling water temperature from a cooling tower at full and half load is indicated in the figure and is based on the relation shown in figure 3.1. The actual cooling load is however not constant during a year in a building. In the figure, the available cooling water temperature is displayed resulting from a fictional example of an annual cooling load profile, ranging from full load at high wet bulb temperatures to a fractional load at lower wet bulb temperatures.

Available cooling water temperature from cooling tower at half load

Available cooling water temperature from cooling tower at full load

Available cooling water temperature from cooling tower at a fictional example of a cooling load profile

25 20 Temperature (°C)

15

Ambient wet bulb temperature

10 5 0 0

1000 2000 3000 4000 5000 6000 7000 8000

-5

Example of approach temperature at a wet bulb temperature of 10°C and at full load

-10 Time (h)

Figure 3.2

Duration of ambient wet bulb temperature and associated cooling water temperatures from an evaporative cooling tower. NB: the duration curves and conditions in the figure are only a fictional example.

In the next section, 3.2 System configurations, different configurations of free cooling systems with an evaporative cooling tower are presented and discussed.

29

3.2 System configurations There are a number of different system configurations where an evaporative cooling tower can be utilized for free cooling opportunities. They are published in e.g. Murphy, (1991), Hensley, (1994), De Saulles, (1996), Bahnfleth & Rehfeldt, (1996) and Lindholm, (2003). The different system lay-outs can be grouped in two categories; systems with the cooling tower coupled in parallel or in series with the cooling load. Most of the cooling towers in the figures below are open type cooling towers. In most cases, the open tower can be replaced by a closed loop cooling tower. In cases when the cooling system is operating in ambient temperatures below zero, freeze protection of the system must be considered.

3.2.1 Parallel coupling Direct connection

Filter

Filter Condenser

Condenser

Evaporator

Evaporator Chilled water

Chilled water Cooling load

Mechanical cooling

Figure 3.3

Cooling load

Free cooling

Free cooling with direct connection between cooling tower and cooling load. Dashed lines indicate closed flow paths. NB: The scheme is intentionally simplified to improve the visual impression.

In figure 3.3, a direct system for free cooling with an evaporative cooling tower is presented. Direct systems physically interconnect the chilled water and the condenser water circuits during free cooling operation, enabling heat to be rejected directly by the cooling tower.

30


The main benefit of this technique is that the difference between the chilled water and the ambient wet bulb temperature, i.e. the approach, is kept to a minimum. This in turn maximises the free cooling availability for a direct system. The drawback of a direct system is the need for a strict water treatment management to reduce the risk of corrosion and fouling in the chilled water circuit. A filter should therefore be considered an integral component in this type of system. Direct systems are sometimes denoted as a “strainer cycle”. Indirect connection

Condenser

Condenser

Evaporator

Evaporator

Heat exchanger

Heat exchanger

Cooling load Chilled water

Mechanical cooling

Figure 3.4

Cooling load Chilled water

Free cooling

Free cooling with indirect connection between cooling tower and cooling load. Dashed lines indicate closed flow paths. NB: The scheme is intentionally simplified to improve the visual impression.

In figure 3.4, an indirect system for free cooling with an evaporative cooling tower is presented. In an indirect system the chilled water circuit is maintained as a closed loop. The heat is rejected by means of a heat exchanger through which chilled water and cooling tower water flow. A plate-and-frame heat exchanger is usually used in this system configuration to minimize the temperature difference between the chilled water and the cooling tower water.

31

As an alternative, an indirect system may incorporate a closed circuit cooling tower, thus excluding the heat exchanger. In this case, the system is laid out as in figure 3.3. The indirect system has the advantage of avoiding the risk of increased corrosion and fouling since the chilled water circuit is separated from the cooling tower water. The penalty for using a heat exchanger is an increase in the chilled water temperature to wet bulb temperature approach. This will cause a corresponding reduction of the free cooling potential. In many cases, this is considered to be worthwhile, since the costs for maintenance and risk of fouling associated with direct systems may be regarded as too high (De Saulles, 1996). Thermosyphon

Condenser

Condenser Compressor by-pass valve

Compressor by-pass valve Evaporator

Evaporator

Cooling load

Cooling load

Chilled water

Chilled water

Mechanical cooling

Figure 3.5

Free cooling

Free cooling with thermosyphon, i.e. refrigerant migration between condenser and evaporator. Dashed lines indicate closed compressor or flow paths. NB: The scheme is intentionally simplified to improve the visual impression.

A thermosyphon system, figure 3.5, is based on a conventional mechanical chiller, added with a compressor bypass arrangement. With this arrangement, when the compressor shuts down, valves in bypass channels open to permit free migration of refrigerant vapour from the evaporator to the condenser and the flow of liquid refrigerant from the condenser to the evaporator. Many chiller manufacturers offer an accessory package that can enable this free cooling application to be used.

32


The heat transfer is limited to refrigerant phase-change and the load capability of thermosyphon systems rarely exceeds about 25%, although some chiller manufacturers can provide systems with an efficiency of up to 35% of full load. In addition, the chilled water from the cooling tower must usually be 7°C or colder to accomplish sufficient heat transfer (Murphy, 1991). Together, these two circumstances limit the use of this type of system to a relatively small portion of the year and also consequently require that the cooling load is lower (about 25%) during winter operation.

3.2.2 Series coupling The operating principle of the tower based systems described under section 3.2.1 Parallel coupling enables the chiller to be shut down and switch the entire cooling load to be handled by the cooling tower. Under this section, series coupling, also labelled load shaving or load sharing, the system provides an alternative by allowing simultaneous free and mechanical cooling. This has the advantage of allowing partial free cooling to be exploited at times when ambient temperatures are too high for full free cooling. Load shaving with a common cooling tower

Condenser Heat exchanger Evaporator

Evaporator

Chilled water

Chilled water

Cooling load

Mechanical cooling

Figure 3.6

Condenser Heat exchanger

Cooling load

Free cooling

Free cooling with load shaving through a common cooling tower. Dashed lines indicate closed flow paths and grey lines or objects indicate optional usage. NB: The scheme is intentionally simplified to improve the visual impression.

33

Free cooling systems with load shaving, as shown in figure 3.6, are indirect by its nature because the free cooling takes place on the return side of the chilled water circuit through a heat exchanger. The heat exchanger is usually a plate and frame heat exchanger to enable a close temperature approach between the cooling tower water and the chilled water circuit. De Saulles (1996) recommend a design approach of about 1,5 – 2 K for systems operating 12 hours a day or less and about 1 K for systems operating on a 24-hour basis.

Load shaving with a dedicated cooling tower

Condenser

Condenser Chilled water

Evaporator

Evaporator

Cooling load

Cooling load

Mechanical cooling

Figure 3.7

Chilled water

Free cooling

Free cooling with load shaving through a dedicated cooling tower. Dashed lines indicate closed flow paths and grey lines or objects indicate optional usage. NB: 1) The scheme is intentionally simplified to improve the visual impression. 2) The cooling tower on the condenser circuit can also be an air heat exchanger if desired.

In figure 3.7 free cooling with load shaving through a dedicated cooling tower is shown. This application is also indirect, i.e. cooling tower water circuit and chilled water circuit is separated. System control is rather simple since the level of pre-cooling can be

34


allowed to float with the ambient wet bulb temperature while the chiller controls the chilled water supply temperature in the normal way. Disadvantages of using a dedicated tower are the higher capital and maintenance costs together with the required extra space. However, the load shaving tower could be used to supplement the tower in the condenser loop at times when free cooling is not possible, i.e. at times with high ambient temperatures. In cases when the cooling load to a sufficient degree is dependent of the ambient weather conditions this would be a practical arrangement. When ambient temperatures are high the cooling tower rejecting heat from the chiller will be working at full load and would benefit from the additional capacity afforded by the load shaving tower which would otherwise be idle. This arrangement should enable smaller towers, e.g. air heat exchanger, to be specified for chiller heat rejection, which would partly or wholly offset the cost of the load-shaving tower (De Saulles, 1996)

When the ambient wet bulb temperature is at least 0,5 K lower than the return chilled water temperature the free cooling system can start pre-cool the chilled water circuit. During such times the level of pre-cooling will vary with the ambient wet bulb temperature. Since it is the return and not the supply chilled water temperature which determines when energy saving is possible, the system can begin operate at relatively high wet bulb temperature, enabling free cooling for a significant part of the year.

35

36

4 Simulation model development In this chapter a cooling tower model is developed. The development is introduced with a discussion about the prerequisites followed by a discussion about the features of the Cooling Tower Model, CTM, and a description of its equations. The chapter is finished with a presentation of the validation of the CTM.

4.1 Prerequisites The basic reason for developing a cooling tower model was the lack of such a model in the chosen building simulation program, IDA Indoor Climate and Energy (IDA ICE). IDA ICE is described in chapter 5.1 Simulation tool together with a discussion about the motives behind the selection of building simulation program. The prerequisites for the Cooling Tower Model, CTM, are that it should comply with the following requirements: Be complex enough for detailed studies of cooling system temperatures, both in the primary and secondary circuits, air and liquid flows and power levels of pumps and fan in cooling tower together with energy use of the pumps and fan. Be simple enough to use in early stages of the building design phase, i.e. have a limited amount of input data, since detailed information about a cooling tower is normally not available. Have sufficient accuracy for an application in a building simulation program. The prime interest is not detailed analysis of the conditions inside the cooling tower, instead the focus is on analysis of the performance of the cooling tower and the indoor thermal climate in a building. Allow for different control strategies of maintaining a constant supply coolant temperature, i.e. single speed fan with on-off control as well as fan with variable frequency drive (VFD). Allow for parameter variation of design parameters. The required qualifications of the person using the CTM should be moderate. For example HVAC engineers with some, but not extensive, experience of building simulation should be able to handle the model. Of the categories of cooling tower models described in section 2.2 Modelling heat and mass transfer in evaporative cooling towers, the one which comply best with the requirements above is based on the effectiveness – NTU method (ε - NTU method). In addition to the use of the ε - NTU method in the model in this thesis, the CTM, the ε NTU method is also used in cooling tower models in well known and wide spread building simulation programs such as TRNSYS (www.trnsys.com) and EnergyPlus (Energy Plus, 2004). The cooling tower model in ASHRAE Primary HVAC Toolkit package (ASHRAE, 1999) is also based on the ε - NTU method. By using the ε - NTU method the same basic model can be used for simulating both wet cooling towers, open and closed, and dry fluid coolers, i.e. air cooled heat exchangers, which gives the CTM extra fields of application.

37

4.2 The cooling tower model (CTM) The main purpose of the model developed in this thesis is for simulating an evaporative cooling tower, hence the model is called the Cooling Tower Model (CTM). However, the CTM actually describes a general air/liquid heat exchanger with the purpose to cool a liquid with outdoor air with the ability to be used with several combinations of arrangements; open or closed and wet or dry, see table 4.1. The CTM has therefore a wide field of application and can be used to simulate both cooling towers, open and closed circuit type, and any air/liquid heat exchanger with the purpose to cool a liquid with outdoor air. The matrix in table 4.1 shows the different arrangements the CTM has including only sensible heat exchange as well as both sensible and latent heat exchange. The CTM implies that cooling is supported by one or several fans inducing forced air movement. Table 4.1 Table showing different possible arrangements of the CTM in this thesis. Open

Closed

Cooling tower with spray nozzles distributing cooling water over fill packing Wet in an open cooling circuit. Heat rejection contains both sensible and latent heat. See figure 2.3

Cooling tower with closed circuit system. Spray nozzles distribute spray water over tube bundles containing a closed circuit coolant. Heat rejection contains both sensible and latent heat. See figure 2.4

This combination is impossible in practical terms.

Air cooled heat exchanger with coolant contained in a closed circuit system. Heat rejection contains only sensible heat. See figure 2.1

Dry

Apart from the different combinations of arrangements described in table 4.1, the CTM has two possible modes of fan speed control; single speed or variable speed drive (VFD), and two possible heat exchange configurations; counter flow or cross flow. The CTM is based on a cooling tower model published by Bourdouxhe et al. (1994), which also is described in ASHRAE Primary HVAC Toolkit (ASHRAE 1999). The ASHRAE model is based on the ε - NTU method. In this thesis, the ASHRAE model is a part of the CTM which has been supplemented with different combinations of arrangements, described in table 4.1, together with different modes for fan speed control and heat exchange configurations as described above. Equations for handling part load operation together with sequenced control strategies for maintaining a constant supply cooling water temperature at different cooling loads and ambient conditions are also added to the CTM. All these features make the CTM complex but also increase its possible fields of application. The CTM is written in the Neutral Model Format (NMF) language. NMF is a formal language used in IDA ICE to describe the equations and conditions in a model. In simple terms NMF describes what to be calculated but not how it is executed. The numerical execution is handled by a general-purpose equation solver (IDA Solver). The CTM consists of 47 equations comprising 60 variables and 59 parameters.

38

4 Simulation model development

4.2.1 General description In this section, an overview is given describing the main features of the CTM. The complete model, written in NMF language, is presented in appendix A. In the CTM, the user can choose between wet or dry operation mode. The wet mode implies the option of cooling liquid spraying either through a cooling tower fill (open mode) or a separate water circuit spraying over the closed circuit tube bundles containing the cooling liquid (closed mode). The CTM includes a freeze-protected circuit, Circuit 1 or the primary circuit, which is a closed loop between the cooling tower and the heat exchanger, which separates Circuit 1 from Circuit 2. Circuit 2, or the secondary circuit, is equal to the closed cooling water loop connected to the air cooling coil in the Air Handling Unit (AHU) and the cooling device(s) in the zone(s), e.g. chilled ceilings, chilled beams or fan coils. A scheme of the CTM and its connection with the rest of a possible comfort cooling system is shown in figure 4.1. Note that the scheme is simplified concerning fittings.

Cooling tower or aircooled heat exchanger

Circuit 1 (primary): Freeze-protected brine

Heat exchanger

To the air cooling coil in the Air Handling Unit (AHU) Hydronic room cooling devices, e.g. chilled beams or fan coils

Circuit 2 (secondary): Water based circuit System boundaries for the cooling tower model Possible chiller

Figure 4.1

The Cooling Tower Model (CTM) and its connection with the rest of a possible cooling system.

The liquid in Circuit 1 is called Liquid 1 and subsequently the liquid in Circuit 2 is called Liquid 2. If Circuit 1 has a freeze protective function, Liquid 1 is a freeze protecting brine. Circuit 2 is usually containing water. If the user wants to omit the freeze-protected circuit a few parameters in the CTM is simply altered. A detailed description for omitting the freeze-protected circuit is presented in appendix A.

39

4.2.2 Modes of operation The cooling tower model can be set to different modes by the user. Mode_WetDry; is wet or dry operation as described above in table 4.1.

Mode_Fan; decides if the fan(s) is/are single speed or equipped with variable speed drive. The single speed mode is applicable when cooling tower has only one fan (or maybe two fans) with single speed. If the cooling tower has several fans with single or two speed operation the user should choose the mode with variable speed drive. This mode simulates the case with multiple fans cycling on-off in sequence fairly well. If cooling tower is equipped with variable speed drive the choice of mode is obvious.

Mode_ClOp; decides if the cooling tower or the air cooled heat exchanger has an open circuit or closed circuit. The mode controls if spray water is on or off. This mode is added because the combination of Closed circuit mode and Wet mode use spray water. All other combinations of Open/Closed and Wet/Dry modes use no additional spray water.

Mode_Tower; lets the user choose between a counterflow configuration or a crossflow configuration. The crossflow configuration assumes both streams unmixed.

4.2.3 Control of supply temperature of cooling tower liquid The cooling tower model has two different ways to control the tower and the supply temperature of the cooling tower liquid. First is a simple on/off control of the whole tower via an external signal from a time schedule through a link (Tower_control). Secondly, the leaving Liquid 2 temperature is controlled in two sequences; first by modulating the mass flow through the pump in Circuit 1 and of air in the fan(s). This is done via an external signal from a PI-controller, see figure 4.2. In single speed mode the real life cycling on-off is simulated as a linear relation of the mass flow from a minimum flow (15% of max.) up to maximum flow (ASHRAE/ IESNA 1999). In variable speed mode, the fan laws are applied controlling the mass flow.

PI-controller

Figure 4.2

40

Control of cooling tower model through sequenced control


4.2.4 User input data The CTM gives the user the following optional design parameters as input data. In early stages of a building design phase, only some of the first category parameters are necessary to address. The rest of the parameters can normally remain default until later stages in the design phase. 1) Input parameters affecting size and cooling capacity of the cooling tower and the intermediate heat exchanger: QTower_d TApproachP TRange_P TApproachS TRange_S QCooling_d TAir_indb_d RelHum_d MLiq1MAir_d

Design cooling capacity in tower (in kW) Temperature difference between inlet air (TAir_in_d) and outlet Liquid 1 (TLiq1Cold_d) at design stage. P = Primary circuit. Temperature difference between inlet Liquid 1 (TLiq1Warm_d) and outlet Liquid 1 (TLiq1Cold_d) at design stage Design difference between TLiq1Cold_d and TLiq2Out in secondary circuit. S = Secondary circuit. Design difference between TLiq2Out and TLiq2In in secondary circuit Design cooling load in building (in kW) Design inlet dry bulb air temperature Design relative humidity of ambient air [%) Massflow relation of Liquid 1 and Air (MLiq1_d/MAir_d). Normally in the range 0,5 to 2.

2) Input parameters primarily affecting energy efficiency of the cooling tower: EtaFan_d dpFan_d dpLiq1 dpLiq2 dpSpray Mode_Fan EtaLiq1 EtaLiq EtaSpray

Fan efficiency (in tower) at design airflow rate Pressure difference over fan at design rate mass flow of air in tower Pressure difference in Circuit 1 (Liquid 1), assumed constant Pressure difference in Circuit 2 (Liquid 2 passing heat exchanger), assumed constant Pressure difference in spray water circuit (open) assumed constant Control mode; 0=Fan with variable speed control, 1=Single speed fan Total pump efficiency in circuit with Liquid 1 Total pump efficiency in circuit with Liquid 2 Total pump efficiency in (open) circuit with spray water

3) Miscellaneous input parameters: Min_MAir

cpLiq1 rhoLiq1

Min. airflow in fraction of design rate mass flow of air in tower. When using variable speed drive Min_MAir is normally in the range 0,1 to 0,2 Liquid 1 specific heat (default 3685= 40% propylene glycol at 15°C) Liquid 1 density (default 1039= 40% propylene glycol at 15°C)

41

Mode_WetDry Mode_Spray Mode_Tower TLiq2Out_set

Control mode; 0 = Wet mode, 1 = Dry mode Control mode; 0 = Closed circuit tower, 1 = Open circuit tower Control Mode; 0 = Counter flow tower, 1 = Cross flow tower Liquid 2 outlet set temperature

4.2.5 Limitations The CTM has like most other mathematical models some limitations. The main limitations are listed below: • The CTM does not calculate the consumption of makeup water for evaporation, drift losses or blow down. • Hybrid cooling towers with combination of, or alteration between, wet and dry mode during operation cannot be simulated within the CTM. • Natural draft cooling towers cannot be simulated in the CTM. • The CTM works with good accuracy between -20°C to + 40°C wet bulb temperature of ambient air. It should not be used outside this temperature range. For further information, see equation 4.11 to 4.13 and the adjacent discussion.

4.2.6 Equations The cooling tower part of the CTM complies with the cooling tower model in ASHRAE (1999). The equations presented in this chapter are the most important ones governing the CTM. For a complete presentation of all equations of the CTM, see appendix A. Figure 4.3 shows the layout and basic terms in the cooling tower model. Enthalpy of entering air stream The enthalpy of ambient air entering the cooling tower is:

(

ha ,in = Ta ,in , db ⋅ c p ,a + w 2501 ⋅ 10 3 + 1805 ⋅ Ta ,in , db

)

[J/kg, °C]

where ha ,in = Enthalpy of entering ambient air [J/kg] Ta ,in, db = Dry bulb temperature of ambient air [°C] c p , a = Specific heat of ambient dry air [J/kg, °C] w = Humidity ratio of ambient air [kg water/kg dry air]

42

(4.1)


Air out

Liquid 1, Warm

Cooling tower

Liquid 2, In

Heat exchanger

Liquid 2, Out Ambient air in

Figure 4.3

Liquid 1, Cold

Cooling tower model; schematic layout and notations

Enthalpy of leaving air stream in wet mode: In wet mode the enthalpy of the leaving air can be calculated by an approximate relation only dependent on the wet bulb temperature of the leaving air (ASHRAE, 1999);

ha ,out = 9362,5 + 1786,1 ⋅ Ta ,out ,wb + 11,35 ⋅ Ta2,out , wb + 0,98855 ⋅ Ta3,out ,wb [J/kg, °C] (4.2) where ha ,out = Enthalpy of leaving air [J/kg] Ta ,out ,wb = Wet bulb temperature of leaving air [°C]

In the cooling tower model, this equation is only used in the Parameter processing section where the design UA-value is calculated. In the part of the CTM which is calculated each timestep, there is however no need to calculate the enthalpy of the (ha,out − ha,in ) , leaving air. The original equation from ASHRAE (1999), i.e. c p ,a ,e = (Ta,out ,wb − Ta,in,wb ) in where the enthalpy of the leaving air is a part, is replaced by a curve fit equation in the CTM. The curve fit equation of the variable c p ,a ,e is a function of the mean wet bulb temperature and the wet bulb temperature difference of entering and leaving moist air. This is made to ensure a better numerical stability of the CTM. For further information, see equations 4.10 to 4.13.

43

Enthalpy of leaving air stream in dry mode: The enthalpy of air leaving the cooling tower is:

(

ha , out = Ta , out , db ⋅ c p , a + w 2501⋅ 103 + 1805 ⋅ Ta , out , db

)

[J/kg, °C]

(4.3)

where ha , out = Enthalpy of leaving air [J/kg] Ta , out , db = Dry bulb temperature of ambient air [°C]

Control of supply temperature, i.e. Liquid 2 temperature, Tliq2,out The aim of controlling Tliq2,out, is to keep the supply temperature at the requested temperature (Tliq2,out,r). The UA-value, and consequently the leaving temperature of Liquid 2, Tliq2,out, is controlled through sequenced control of the airflow through the tower. When Tliq2,out +0.1 < Tliq2,out,r and rising, the first step to control Tliq2,out is by cycling fan on-off, i.e. UA-value is cycling between zero and UA=f(min of M& ) where a

M& a is the air flow trough the cooling tower. The second step is carried out by modulate

the air flow and Liquid 1 flow between minimum flow rate and maximum flow rate to achieve the desired temperature of Tliq2,out. When the fan is on, the air mass flow, M& a , is controlled in a linear relation to the control signal from the PI-controller. In figure 4.4, the relation airflow – control signal for a fan in single speed mode is shown. In single speed mode the real life control strategy by cycling the fan on-off is emulated as a linear relation of M& from the design a

air mass flow, M& a , d , to 0,15 M& a ,d (AHRAE/IESNA, 1999), see equation 4.4. M& a should is in this case be considered as a mean value of the air mass flow over a time period. For a fan with variable speed control, the relation airflow – control signal is almost the same as in figure 4.4, but in this case the minimum airflow is optional depending on type of variable speed drive, see equation 4.5. The default minimum value in CTM is 0,1 M& a , d for a fan with variable speed drive.

44


1,0

Relative air flow (-)

0,8

0,6

0,4

0,2

0,0 0,0

0,2

0,4

0,6

0,8

1,0

Control signal (-)

Figure 4.4

Relation airflow versus control signal from PI-controller

The equation for a fan with single speed mode is as follows:

M& a = M& a ,d ⋅ Ctrl + 0.15 ⋅ M& a ,d (1 − Ctrl ) [kg/s]

(4.4)

where M& a = Air mass flow through cooling tower [kg/s] M& a , d = Design air mass flow through cooling tower [kg/s] Ctrl = Signal from controller, 0< Ctrl <1 [-]

For a fan with variable speed control, equation 4.5, M& a is linear between M& a ,d and a minimum flow equal to Min _ M& a ,d ⋅ M& a ,d ( Min _ M& a ,d is typically 0.1 to 0.2 depending on type of variable speed drive).

M& a = M& a , d ⋅ Ctrl + Min _ M& a , d ⋅ M& a , d (1 − Ctrl ) [kg/s]

(4.5)

where

Min _ M& a , d = Minimum fraction of air mass flow through cooling tower for fan with variable speed drive [-]

45

The mass flow of Liquid 1, M& liq1 , is always regulated with variable speed drive. Minimum flow is set to 10% of M& liq1 .

M& liq1 = M& liq1,d ⋅ Ctrl + 0.1⋅ M& liq1,d (1 − Ctrl )

[kg/s]

(4.6)

where

M& liq1 = Mass flow of Liquid 1 [kg/s] M& liq1,d = Design mass flow of Liquid 1 [kg/s] The mass flow rates of air and Liquid 1 influence the heat transfer in the cooling tower, i.e. the UA-value. In the literature many different kinds of correlations have been presented to enable part load simulations of a cooling tower or an air-cooled heat exchanger, e.g. Baker & Shryock (1961), Parker & Treybal (1961) and Mitzushina et al. (1967), A common way of correlate the UA-value to variations in air flow and liquid flow rates is shown in equation 4.7. m UA = C ⋅ M& an ⋅ M& liq

[W/°C]

(4.7)

where C, n and m normally are determined empirically for a certain object. In the literature several values of C, n and m has been presented. If the UA-value can be determined at the design flow rates of air and liquid, UAd, the UA-value can be expressed as follows:

 M& UA = UAd ⋅  a &  M a,d

   

n

 M& ⋅  liq  M&  liq , d

   

m

[W/°C]

(4.8)

Lebrun and Silva (2002) has gathered and tabled values of the exponents n and m for cooling towers from four different sources including themselves, with a total of seven different values of n and m respectively. The mean values of n and m from those values are n = 0.65

(with min n = 0.39 and max n = 1.03 )

m = 0.43

(with min m = 0.15 and max m = 0.67 )

For air cooled heat exchangers (dry operation) Wetter (1999) has presented values for n and m, see table 4.2 and below.

46


Table 4.2 Values for exponent n in UA-correlation for air-cooled heat exchangers (Wetter, 1998) Tube arrangement

Exponent n (air side)

In-line

Reynolds number (Re) Re≤ 4·104

Staggered

1·103 ≤ Re ≤ 2·104 0.65

Staggered

2·104 ≤ Re ≤ 2·105 0.80 0.95 Re ≥ 2·105

Staggered

0.72

The reference dimension in Reynolds number is the fin pitch. The exponent m for the liquid side is only given as 0,85 and not depending on Reynolds number. The UA-value correlation in the present model, in wet mode, is as follows:

 M& UA = UAd ⋅  a &  M a ,d

   

0 , 65

 M& liq1 ⋅  M&  liq1,d

   

0 , 43

[W/°C]

(4.9)

The values of n and m in equation 4.9 are default values in the CTM but can easily be changed by the user. The UA-value correlation with default values of n and m, as in equation 4.9, is shown in figure 4.5. The correlation represents an almost strait line. 1,0

Relative UA-Value (-)

0,8

0,6

0,4

0,2

0,0 0

0,2

0,4

0,6

0,8

1

Relative flow rate (-)

Figure 4.5

UA-value correlation according to equation 4.9 with default values of n and m where n = 0.65 and m = 0.43 . Both M& a and M& liq1 are varied simultaneously.

47

Cooling tower Equations in this section are based on the cooling tower model in ASHRAE (1999) and equations presented in chapter 2 Cooling Towers – Introduction and modelling, section 2.2. To supplement the ASHRAE model a dry operation mode is added as well as a distinction between open or closed type cooling tower. The difference between open or closed mode is however only a spray water pump adding a small amount of extra energy for the closed type cooling tower. A fictitious mean specific heat of moist air was introduced in section 2.2, equation 2.5. The fictitious mean specific heat is in the CTM called the effective specific heat of moist air, c p ,a ,e , complying with the terminology in ASHRAE (1999). c p ,a ,e is only used in wet mode and it is written:

(h

− ha ,in )

a ,out , wb

− Ta ,in, wb )

a ,out

c p , a ,e =

(T

[J/kg °C]

(4.10)

where c p , a ,e = Effective specific heat of air [J/kg, °C] ha , out = Enthalpy of leaving air [J/kg] ha ,in = Enthalpy of entering ambient air [J/kg] Ta ,out , wb = Wet bulb temperature of leaving air [°C] Ta , in , wb = Wet bulb temperature of entering ambient air [°C]

For dry mode operation the specific heat of dry air, c p ,a is used. The original equation 4.10 is in the CTM replaced by a curve fit equation where c p ,a ,e = f (Tm , Td ) , see equation 4.11. This is made to ensure a better numerical stability of the CTM. The effective specific heat of air is expressed as: c p ,a ,e = 0,56473 ⋅ 10 −4 Tm4 − 0,0351Tm3 + 1,65114Tm2 + 2,09962 ⋅ 10 −11 Tm4Td4 + 1,84999 ⋅ 10 − 4 Tm2Td2 + 1196,30745

(4.11)

The curve fit equation comprises two variables; Tm and Td :

Td = Ta ,out , wb − Ta ,in, wb

Tm =

48

(T

a ,out , wb

+ Ta ,in , wb ) 2

(ambient wet bulb temperature difference, [°C])

(4.12)

+ 20 (mean ambient wet bulb temperature, [°C]) (4.13)


Equation 4.11 implies that the mean ambient wet bulb temperature, Tm , has an offset of 20 K to place the zero value of Tm on the lowest end of the allowed ambient wet bulb temperature range (-20 < Tm < 40°C). The curve fit equation has a relative error < 1% between –15 < Tm < 40°C and Td < 15°C. The relative error of the effective specific heat, c p ,a ,e , is the relative difference in percentage between the original equation (4.10) and the curve fit equation (4.11). At normal operation of a cooling tower with the application of chilling water in a comfort cooling system, Tm and Td will be well within these limits. A graphic presentation of the accuracy of the curve fit equation is presented in figure 4.6. A relative error of < 1% in the effective specific heat, c p ,a ,e , generates an error on the leaving Liquid 2 temperature which is less than 0.1°C. Increasing to the maximum range allowed; -20 < Tm < 40 °C and Td < 30°C, the relative error in the effective specific heat, c p ,a ,e < 3,4% .

Maximum range of ambient wet bulb temperature when operating in the northern parts of Europe. The lower end depends on type of cooling tower and cold weather operation

4%

Ambient wet bulb temperature difference, Td [°C]

3%

Accuracy (%)

2%

0,01

1%

5,0 10,0

0%

15,0 -1%

20,0 25,0

-2%

30,0 -3% -4% -20

-10

0

10

20

30

40

Mean wet bulb temperature, Tm (°C)

Figure 4.6

Accuracy of the curve fit equation (4.11) compared to the original equation (4.10) describing the effective specific heat, c p ,a ,e , of moist air. Accuracy is expressed as the relative difference in percentage between the original equation and the curve fit equation.

49

The name “fictitious heat transfer coefficient-area product”, which was introduced in chapter 4.2.1, is in this chapter replaced by the expression “effective heat transfer coefficient-area product”, UAe , to comply with the terminology in ASHRAE (1999). UAe =

UA ⋅ c p ,a ,e c p ,a

[W/m2, °C]

(4.14)

where UAe = Effective heat transfer coefficient-area product [W/°C] UA = Heat transfer coefficient-area product [W/°C] c p , a = Specific heat of dry air [J/kg, °C] The heat capacity flow of air, Liquid 1 and Liquid 2 is as follows:

Cliq1 = c p ,liq1 ⋅ M& liq1 Cliq 2 = c p ,liq 2 ⋅ M& liq 2

[W/°C]

(4.15)

[W/°C]

(4.16)

Ca = c p ,a ,e ⋅ M& a

[W/°C]

(4.17)

where Cliq1 = Heat capacity flow of Liquid 1 [W/°C] c p ,liq1 = Specific heat of Liquid 1 [J/kg, °C] M& liq1 = Mass flow of Liquid 1 [kg/s] C liq 2 = Heat capacity flow of Liquid 2 [W/°C] c p ,liq 2 = Specific heat of Liquid 2 [J/kg, °C] M& liq 2 = Mass flow of Liquid 2 [kg/s] C a = Heat capacity flow of air [W/°C] M& a = Mass flow of air [kg/s] The smallest and biggest heat capacity flow Cmin and Cmax is

50

C min = min (C liq1 , C a )

[W/°C]

(4.18)

C max = max (C liq1 , C a )

[W/°C]

(4.19)


The relation of Cmin / Cmax is defined as

CR =

C min C max

[-]

(4.20)

The Number of transfer units, NTU, in the cooling tower can now be written

NTU =

UAe C min

[-]

(4.21)

The effectiveness of the heat exchanger, ε, depending of configuration of the cooling tower or air cooled heat exchanger, is written: Counterflow configuration:

ε=

1 − exp[− NTU (1 − C R )] 1 − C R ⋅ exp[− NTU (1 − C R )]

[-]

(4.22)

Crossflow configuration:

 exp(− NTU ⋅ C R ⋅ k ) − 1   CR ⋅ k  

ε = 1 − exp

[-]

(4.23)

where

k = NTU −0, 22

(4.24)

The water to air heat transfer rate in the cooling tower, Q& , i.e. the cooling capacity: Wet mode: Q& = ε ⋅ C min (Tliq1, w − Ta ,in , wb )

[W]

(4.25)

Dry mode: Q& = ε ⋅ Cmin (Tliq1,w − Ta ,in ,db )

[W]

(4.26)

where

Tliq1, w = Temperature of the warmer side of Liquid 1 [°C] Leaving air dry or wet bulb temperature from cooling tower or air cooling heat exchanger, Ta ,out ,wb or Ta ,out ,db , has the following expression:

51

Wet mode:

 Q&  Ta ,out ,wb = Ta ,in , wb +    Ca 

[°C]

(4.27)

[°C]

(4.28)

Dry mode:

 Q&  Ta ,out ,db = Ta ,in,db +    Ca 

Outlet Liquid 1 temperature, Tliq1,c , i.e. the colder side of Liquid 1, from tower is calculated as follows:  Q&   Tliq1,c = Tliq1, w −  C  liq 1  

[°C]

(4.29)

Heat exchanger According to elementary heat exchanger theory the outlet temperature of Liquid 2, Tliq 2,out , from heat exchanger between Liquid 1 and Liquid 2 is written as: If Cliq2 ≤ Cliq1:

Tliq 2,out = Tliq 2,in − η hex (Tliq 2,in − Tliq1,c )

[°C]

(4.30)

If Cliq2 > Cliq1:  Cliq1  (T Tliq 2,out = Tliq 2,in − η hex  − Tliq1,c )  C  liq 2,in  liq 2 

[°C]

(4.31)

where

Tliq 2,in = Temperature of incoming Liquid 2 [°C] Tliq 2,out = Temperature of outgoing Liquid 2 [°C] Tliq1,c = Temperature of cold side Liquid 1 [°C]

η hex = Efficiency of heat exchanger [-] The warm side temperature of Liquid 1, Tliq1,w , from heat exchanger between Liquid 1 and Liquid 2 is calculated as:

52


 C liq 2 Tliq1, w = Tliq1,c +  C  liq1

 (Tliq 2,in − Tliq 2,out )  

[°C]

(4.32)

Use of energy The speed of fan with variable speed control is written:  M& n speed = n speed ,d  a  M&  a ,d

   

[s-1]

(4.33)

where

n speed = Speed of fan [s-1] n speed ,d = Design speed of fan [s-1] The pressure difference in a cooling tower fan with variable speed control can, using the fan laws, be written as:

 n speed ∆p F = ∆p F ,d  n  speed ,d

   

2

[Pa]

(4.34)

For a fan with single speed mode the pressure difference is:

∆p F = ∆p F ,d

[Pa]

(4.35)

where

∆p F = Pressure difference in cooling tower fan [Pa] ∆p F ,d = Design pressure difference in cooling tower fan [Pa] The total fan efficiency, η F , is often assumed constant in models of energy use of a fan. The total fan efficiency is however not constant, especially when a variable speed drive is applied. In this model the total fan efficiency is modelled as a function of fan motor speed. As stated in equation 4.33 the fan motor speed is direct proportional to air mass flow through the fan. The relation for η F with variable speed drive is chosen as a power law expression. This is an approximation for the fact that the efficiency is a function of air mass flow of the fan itself and the speed of the variable speed drive. The expression for the variable speed drive is similar to a power law curve.

 M& η F = η F ,d  & a  M a,d

   

0.5

[-]

(4.36)

53

where

η F = Total fan efficiency [-] η F ,d = Design total fan efficiency [-] The exponent is chosen to 0.5 as default, but can easily be changed by the user. For a single speed fan the total efficiency is constant:

η F = η F ,d

[-]

(4.37)

The electrical power for fan(s) in a cooling tower, PF , is expressed as: Fan with variable speed control:

PF =

∆p F ⋅ M& a ρ a ⋅η F

[W]

(4.38)

⋅ ∆p F ,d − 0.15 ⋅ M a ,d ⋅ ∆p F ,d ) [W]

(4.39)

Fan with single speed mode:

PF =

1 0.85 ⋅ ρ a ⋅ η F

(M

a

where

PF = Fan electrical power [W] ρ a = Air density [kg/m3] The linear relation in equation 4.39 corresponds to a relation in AHRAE/IESNA (1999) between power and mass flow of a single speed fan. This equation emulates the relation between mean power requirement and mean flow when a single speed fan is switching between on or off mode. In figure 4.7, the equations 4.38 and 4.39 are visualized. The two curves for variable speed drives are for comparing constant versus variable total fan efficiency (eta const. vs. eta variable).

54


Relative required electrical power

1,0 0,9 0,8 0,7 0,6 0,5 0,4 0,3 0,2 0,1 0,0 0,0

0,2

0,4

0,6

0,8

1,0

Relative air mass flow Single speed

Figure 4.7

Variable speed, eta variable

Variable speed, eta const.

Electrical power requirement for a fan as a function of air mass flow in the CTM.

The electrical power for the pump in the primary circuit (Circuit 1) between cooling tower and heat exchanger, i.e. Liquid 1, is expressed as follows:

Pliq1 =

∆pliq1 ⋅ M& liq1

ρ liq1 ⋅ηliq1

[W]

(4.40)

where Pliq1 = Electrical power for the pump in Circuit 1 [W] ∆p liq1 = Pressure difference in pump in Circuit 1 [Pa]

ρ liq1 = Density of Liquid 1 [kg/m3] η liq1 = Total pump efficiency in Circuit 1 [-]

The pump in the primary circuit is assumed to have a variable speed drive, see equation 4.6. The total pump efficiency is however assumed to be constant in contrary to the total fan efficiency. This simplification is done since the annual electrical use of the pump in Circuit 1 is a relatively small part, about 10%, of the total annual electrical use of the cooling tower or the air-cooled heat exchanger.

55

As for the pump in circuit 1, the electrical power for pumping Liquid 2 through heat exchanger in Circuit 2 is expressed in the same way:

Pliq 2 =

∆pliq 2 ⋅ M& liq 2

ρ liq 2 ⋅ηliq 2

[W]

(4.41)

where Pliq 2 = Electrical power for the pump in Circuit 2 [W] ∆pliq 2 = Pressure difference in pump in Circuit 2 [Pa]

ρ liq 2 = Density of Liquid 2 [kg/m3] ηliq 2 = Total pump efficiency in Circuit 2 [-]

When the user has chosen to simulate a wet (evaporative) closed cooling tower the electrical power demand for a pump in the water spray sump is calculated as follows:

Pspray =

∆pspray ⋅ V&spray

η spray

[W]

(4.42)

where Pspray = Electrical power for the spray water pump [W] ∆pspray = Pressure difference over spray water pump [Pa] V& = Volume flow through spray water pump [m3/s] spray

ηspray = Total efficiency in spray water pump [-] and 0.018 ⋅ Q& d V&spray = 1000

[m3/s]

(4.43)

where Q& d = Design cooling capacity of cooling tower

[kW]

Equation 4.43 is based on an approximate relation between the volume flow of spray water and the cooling capacity of an evaporative closed circuit cooling tower which is assumed to be 0.018 l/s per kW cooling capacity in tower (Stoecker, 1998).

56


A coefficient of performance, COP, can be defined for a cooling tower or air-cooled heat exchanger as:

COP =

Cooling capacity (W ) , Total electricity power (W )

COP =

Q& Ptotal

or [-]

(4.44)

where [W] Q& = Cooling capacity Ptotal = PF + Pliq1 + Pliq 2 + Pspray [W]

(4.45)

4.2.7 Determining the design UA-value The CTM requires a UA-value to be known at the maximum design operating point. A design UA-value is very seldom known by the user, especially in the early stages of the building design phase. In the CTM some of the parameters supplied by the user in section 4.2.4 User input data, are used to calculate a UA-value at design conditions, see table 4.3.

Table 4.3

Table showing input data for the calculation of the design UA-value of a specific cooling tower.

Parameter name

Symbol

Description

QTower_d

Q& d

Design cooling capacity in tower (in kW)

Tapproach_d

T A ,d

Design temperature difference between inlet air (TAir_in_d) and outlet Liquid 1 (TLiq1Cold_d). For wet mode, TAir_in_d is equal to the design ambient wet bulb temperature.

Trange_d

TR ,d

Design difference between inlet Liquid 1 (TLiq1Warm_d) and outlet Liquid 1 (TLiq1Cold_d).

TAir_indb_d

Ta ,in ,db ,d

Design inlet dry bulb air temperature

RelHum_d

RH d

Design relative humidity of ambient air [%]

MLiq1MAir_d

M& liq M& a

Mass flow relation of Liquid 1 and Air (MLiq1_d/MAir_d). Normally in the range of 0,5 to 2.

57

For wet mode the design wet bulb temperature, Ta ,in ,wb ,d , is calculated from Ta ,in ,db ,d and RH d by subroutines supplied with IDA Indoor Climate and Energy (IDA ICE), which is the simulation program this model is implemented in. For more information about IDA ICE, see chapter 5.1 According to the effectiveness – NTU method presented in chapter 2.2 the cooling tower, as well as an air cooled heat exchanger, can be represented as a simple heat exchanger, figure 4.8. Ta,in,wb

Air side

Ta,out,wb

UAe,d Tliq1,c

Figure 4.8

Liquid side

Tliq1,w

An evaporative cooling tower modelled as an indirect contact heat exchanger

From given and calculated parameters the entering and leaving liquid temperatures are easily calculated: Wet mode: Tliq1,c,d = Ta ,in ,wb ,d + TA,d

[°C]

(4.46)

Tliq1,c ,d = Ta ,in ,db ,d + TA,d

[°C]

(4.47)

Tliq1,w,d = Tliq1,c + TR ,d

[°C]

(4.48)

Dry mode:

and

where Tliq1,c ,d = Design temperature of cold side in Liquid 1 [°C] Tliq1,w,d = Design temperature of warm side in Liquid 1 [°C] Ta ,in ,wb ,d = Design wet bulb temperature of ambient air [°C] Ta ,in ,db ,d = Design dry bulb temperature of ambient air [°C] TA,d = Design temperature approach [°C] TR ,d = Design temperature range [°C]

58


After calculating Tliq1,c ,d and Tliq1,w,d , one temperature is still unknown, i.e. Ta ,out ,wb ,d in wet mode or Ta ,out ,db,d in dry mode. In the case of dry mode Ta ,out ,db,d can easily be calculated through the expression Ta ,out ,db ,d = Ta ,in ,db ,d +

Q& d C a ,d

(dry mode) [°C]

(4.49)

where

Ca ,d = M& a ,d ⋅ c p ,a

(dry mode) [W/°C]

(4.50)

and

M& a ,d =

M& liq1,d [kg/s]  M& liq     M&   a d

(4.51)

where Q& d = Design cooling capacity [W] Ca ,d = Design heat capacity flow of air [W/°C] M& a ,d = Design mass flow of air [kg/s] c p ,a = Specific heat capacity of air [J/kg °C] M& liq1,d = Design mass flow of Liquid 1 [kg/s]  M& liq    = Design mass flow relation of Liquid 1 and air [-]  M&   a d

& liq1,d , is calculated as follows: The mass flow of Liquid 1 at the design point, M M& liq1,d =

Q& d ⋅1000 c p ,liq1 (Tliq1, w,d − Tliq1,c ,d )

[kg/s]

(4.52)

where

c p ,liq1 = Specific heat capacity of Liquid 1 [J/kg °C]

59

If a wet mode is chosen, an iterative process must determine the value of Ta ,out ,wb ,d . By using a first guess of Ta ,out ,wb ,d , it is possible to calculate ha ,out (equation 4.2) and c p ,a ,e (equation 4.10) followed by Ca (equation 4.17). The leaving air wet bulb temperature,

Ta ,out ,wb ,d , can now be calculated by using equation 4.27. The iterative process is carried out until convergence with required accuracy is reached. The effective heat transfer coefficient-area product, UAe ,d , is given by the following relationship:

UAe , d =

Q& d ∆Tln

[W/°C]

(4.53)

where ∆Tln is the log-mean temperature difference defined as:

∆Tln =

(T

liq1, w , d

− Ta ,out ,d ) − (Tliq1,c ,d − Ta ,in,d )  Tliq1,w,d − Ta ,out ,d ln T  liq1,c ,d − Ta ,in ,d

   

[°C]

(4.54)

in wet mode operation: Ta ,out ,d = Ta ,out , wb ,d and Ta ,in ,d = Ta ,in ,wb ,d in dry mode operation: Ta ,out ,d = Ta ,out ,db ,d and Ta ,in ,d = Ta ,in ,db ,d Finally the design heat transfer coefficient-area product, UAd , can be calculated: c p,a UAd = UAe ,d [W/°C] (4.55) c p , a ,e The UAd -value is then used in equation 4.9 to calculate the UA -value.

60


4.3 Validation of the model Validation of the effectiveness – NTU model in general has been discussed in section 2.2 Modelling heat and mass transfer in evaporative cooling towers, where published material by Benton et al., (2002), Bourdouxhe et al., (1994) and Hernandez et al., (1994) was presented. The validation of the CTM developed within the work of this thesis has been carried out with emphasis on the temperature of the chilled water leaving the cooling tower, i.e. the supply cooling water. It is the same parameter which was examined by Benton et al., (2002), Bourdouxhe et al., (1994) and Hernandez et al., (1994). It is the parameter which generally is the most important when applying a cooling tower to a cooling system. The validation has been carried out by comparing calculated data from the developed model with relevant data from published sources and from measurements of the pilot plant in this thesis. The published material is from Costelloe & Finn (2003) and Poppe (1973) see Halasz (1999).

4.3.1 Validation by published data The material from Costelloe & Finn, (2003) is based on experimental measurement data from a test rig consisting of an open cooling tower, a primary and a secondary circuit with an intermediate plate heat exchanger, similar to the system in figure 4.3. The cooling tower in the test rig had a rated output of about 20 kW at design conditions. The published data consist of 13 cases with temperatures of the primary and secondary circuits respectively. The measurements was conducted at ambient wet bulb temperatures in the range of 6,4 – 16,5°C, with primary approach temperatures between 0,9°C and 2,3°C and total approach temperatures between 2,2°C – 4,3°C. The material from Poppe (1973) see Halasz (1999), which was used as a reference for validating Halasz (1999) own model, comes from numerical calculations of “rather accurate” differential equations describing an evaporative cooling tower (originally published by Poppe, 1973). In this case the cooling tower is working with only a primary circuit, hence there is only one approach and range temperature for each case. The published material encompasses 32 values with approach temperatures from 2 to 22°C and range temperatures from 4 to 30°C. The upper values of these approach and range temperatures are much higher than in ordinary comfort cooling applications,  M& liq   is in the where low approach and range temperatures are desirable. The value   M&   a d range from 0,25 to 8, which is a wider range than in usual applications. Typical values are between 0,5 to 2.

61

Costelloe & Finn, (2003) In table 4.4 the specifications of the different conditions at each case together with findings by Costelloe & Finn, (2003) compared with results from the CTM are presented. Table 4.4 Specifications of the different conditions at each case together with findings by Costelloe & Finn, (2003) compared with results from the cooling tower model in this thesis. Measured data by Costelloe and Finn, (2003)

Case Load Ambient Primary no. (kW) wet bulb supply temp. temp.

Results from Deviation between cooling tower model results from model in this thesis, i.e. the and measured data CTM

Second. Second. Primary supply return supply temp. temp. temp.

Second. supply temp.

Primary Second. supply supply temp. dev. temp. dev.

1

24

6,4

8,7

10,7

14,53

8,57

10,06

-0,13

-0,64

2

24

8,9

10,8

12,5

16,33

10,70

12,10

-0,10

-0,40

3

24

9,2

11,1

12,8

16,63

10,97

12,38

-0,13

-0,42

4

24

11,1

12,8

14,6

18,43

12,71

14,14

-0,09

-0,46

5

20

8,4

9,8

11,3

14,49

9,91

11,06

0,11

-0,24

6

20

9,2

10,6

12,1

15,29

10,66

11,81

0,06

-0,29

7

20

10,2

11,3

12,9

16,09

11,57

12,70

0,27

-0,20

8

20

12,5

13,5

15,1

18,29

13,68

14,83

0,18

-0,27

9

20

16,5

17,4

19,0

22,19

17,43

18,62

0,03

-0,38

10

15

8,7

9,7

10,9

13,29

9,84

10,70

0,14

-0,20

11

15

9,3

10,6

11,7

14,09

10,45

11,36

-0,15

-0,34

12

15

9,7

11,1

12,3

14,69

10,89

11,84

-0,21

-0,46

13

15

10,6

11,6

12,8

15,19

11,65

12,54

0,05

-0,26

The temperature deviations from comparing measured data by Costelloe & Finn (2003) with results from the cooling tower model in this thesis, the CTM, are shown in figure 4.9. The zero-level in the figure represents the data from Costelloe & Finn (2003), i.e. if a bar is on the negative scale the calculated value from the CTM is lower than the corresponding measured value by Costelloe & Finn (2003) and vice versa.

62


Temperature deviation (°C)

0,40 0,20 0,00 -0,20 -0,40 -0,60 -0,80 1

2

3

4

5

6

7

8

9

10

11

12

13

Case no. Primary temp dev.

Figure 4.9

Second. temp. dev.

Temperature deviations from comparing measured data by Costelloe & Finn (2003) with results from the cooling tower model in this thesis. The zero-level represents the data from Costelloe & Finn (2003). Description of the different cases is found in table 4.4,

The mean value, maximum value, standard deviation and the 95% confidence interval of the deviations in table 4.4 are presented in table 4.5.

Table 4.5

Accuracy of the CTM in this thesis compared with measured data from Costelloe & Finn (2003) based on values from table 4.4.

Deviation [°C] Mean value Maximum value Standard deviation 95% Confidence interval

Primary dev. Secondary dev. 0,00 -0,35 0,27 -0,64 0,14 0,12 ±0,08 ±0,07

When comparing results from the cooling tower model with data from Costelloe & Finn (2003) there is a good accordance. The mean value of the total temperature deviation is -0,35°C with a standard deviation of 0,12°C. Biggest total deviation is -0,64°C with all other deviations below -0,46°C.

63

Halasz, (1999) In table 4.6 the conditions at each case together with findings by Poppe (1973), see Halasz (1999), are compared with results from the CTM. Table 4.6

Conditions at each case together with data from Poppe (1973), see Halasz (1999), compared with results from the cooling tower model in this thesis. (WBT = Wet Bulb Temperature of ambient air)

Input data by Halasz, (1999) Case Return WBT Range Approach ma/mv no. temp. In (°C) (°C) (°C) (°C) 0.1 0.2 0.3 1.1 1.2 1.3 1.4 2.1 2.2 2.3 2.4 3.1 3.2 3.3 4.1 4.2 4.3 4.4 5.1 5.2 5.3 6.1 6.2 6.3 6.4 7.1 7.2 7.3 8.1 8.2 8.3 8.4

30 30 30 34 34 34 34 34 34 34 34 34 34 34 34 34 34 34 34 34 34 40 40 40 40 40 40 40 54 54 54 54

4 4 8 12 12 12 16 20 20 20 24 28 28 28 12 12 12 16 20 20 20 12 12 12 16 18 18 18 12 12 12 16

4 4 4 4 4 4 4 4 4 4 4 4 4 4 10 10 10 10 10 10 10 20 20 20 20 20 20 20 30 30 30 30

22 22 18 18 18 18 14 10 10 10 6 2 2 2 12 12 12 8 4 4 4 8 8 8 4 2 2 2 12 12 12 8

0,25 0,30 0,30 0,20 0,25 0,30 0,30 0,30 0,35 0,40 0,40 0,80 1,00 1,20 0,50 0,80 1,00 1,00 1,00 1,50 2,00 1,50 2,00 3,00 3,00 3,00 5,00 8,00 1,00 1,50 2,00 2,00

Output data, Results from cooling tower model Halasz, (1999) in this thesis, i.e. the CTM Supply WBT Supply WBT Supply WBT temp. Out temp Out temp out (°C) (°C) (°C) (°C) dev. dev 26 26 26 26 30 30 30 30 30 30 30 30 30 30 30 24 24 24 24 24 24 24 20 20 20 20 20 20 20 24 24 24

27,37 24,83 26,77 33,82 30,92 28,68 30,88 32,92 31,46 30,31 33,02 32,27 31,50 30,97 33,70 27,81 25,34 27,95 30,27 27,32 25,71 28,60 25,37 21,55 24,76 25,81 23,01 21,28 39,84 34,18 30,38 32,51

26,45 26,30 26,33 30,35 30,22 30,25 30,33 30,20 30,13 30,21 30,01 30,00 30,00 24,14 24,04 24,01 24,00 24,00 24,00 20,00 20,00 20,00 20,00 20,00 20,00 20,00 24,00 24,00 24,00 24,00

24,90 23,08 24,88 29,09 27,32 29,48 31,65 30,56 29,60 32,25 31,98 31,22 30,70 27,08 24,82 27,31 29,87 26,98 25,40 27,99 24,86 21,27 24,10 25,51 22,76 21,07 39,66 33,28 29,42 31,49

0,45 0,30 0,33 0,35 0,22 0,25 0,33 0,20 0,13 0,21 0,01 0,00 0,00 0,14 0,04 0,01 0,00 0,00 0,00 0,00 0,00 0,00 0,00 0,00 0,00 0,00 0,00 0,00 0,00 0,00

2,47 1,75 1,89 1,84 1,37 1,40 1,27 0,90 0,71 0,77 0,29 0,28 0,28 0,73 0,52 0,64 0,40 0,34 0,31 0,61 0,52 0,28 0,66 0,30 0,25 0,21 0,19 0,90 0,97 1,02

The temperature deviations in table 4.6 are shown in figure 4.10. The zero-level in the figure represents the data from Halasz, i.e. if a bar is on the negative scale, the calculated value from the cooling tower model is lower than the corresponding reference data and vice versa.

64


8.4 8.3 8.2 8.1 7.3 7.2 7.1 6.4 6.3 6.2 6.1 5.3 5.2 5.1

Case no.

4.4 4.3 4.2 4.1 3.3 3.2 3.1 2.4 2.3 2.2 2.1 1.4 1.3 1.2 1.1 0.3 0.2 0.1

-2,5

-2,0

-1,5

-1,0

-0,5

0,0

0,5

Temperature deviation (°C) Supply temp. dev.

Figure 4.10

Air wetbulb out dev

Temperature deviation from comparing data from Poppe (1973), see Halasz (1999), with results from the cooling tower model in this thesis. Data is from table 4.6. The zero-level represents the data from Poppe.

The mean value, maximum value, standard deviation and the 95% confidence interval of the temperature deviations in table 4.6 are presented in table 4.7.

65

Table 4.7

Accuracy of the cooling tower model in this thesis method compared with data from Poppe (1973), see Halasz (1999), based on data from table 4.6.

Deviation [°C]

Supply water WBT of temperature discharge air

Mean value

0,10

-0,80

Maximum value

0,45

-2,47

Standard deviation

0,14

0,58

±0,08

±0,32

95% Confidence interval WBT = Wet Bulb Temperature

When comparing results from the cooling tower model with data from Poppe (1973), see Halasz (1999), there is a good accordance. The mean value of the supply water temperature deviation is 0,10°C with a standard deviation of 0,14°C. Biggest total deviation is 0,45°C for the case when the cooling tower range was 4°C and the approach was 22°C. An approach of that magnitude is exceptionally unusual in comfort cooling applications.

Comparing with the data from section 2.2 Modelling heat and mass transfer in evaporative cooling towers, table 2.1 and 2.2, where published data of accuracy tests on the effectiveness – NTU model per se is presented, the results in table 4.5 and 4.7 are approximately equal or better. With the results in this section together with findings in section 2.2 Modelling heat and mass transfer in evaporative cooling towers, there can be concluded that the cooling tower model developed in this thesis, the CTM, has a good accuracy concerning the temperature of the supply water when comparing with published data.

4.3.2 Validation by data from pilot plant The CTM is also validated against measurement data from the pilot plant at Kvarnberget in Göteborg. The data is from May 1 – August 31, 2007. In the previous section the CTM is validated against a number of single operational conditions. In this case the CTM is validated against measured data from continuous operation when applied in IDA ICE. A presentation of the simulation tool IDA ICE is made in section 5.1 Simulation tool and description of the monitoring of the pilot plant at Kvarnberget in Göteborg is presented in chapter 6 Monitoring of free cooling system with evaporative cooling tower. Figure 4.11 shows an example period during July 24 – 27, 2007 with measured supply liquid temperature versus calculated temperature. Figure 4.12 shows all samples during the period May 1 – August 31 when there was wet operation active, presented as measured versus calculated supply liquid temperature.

66


Comparison between measured and calculated data Kvarnberget, Göteborg July 24 - 27, 2007 20 19 18

Temperature [°C]

17 16 15 14 13 12 11 10 00:00

12:00

00:00

12:00

00:00

12:00

Measured

00:00

12:00

00:00

Calculated

Time

Figure 4.11

Comparison between measured and calculated data during an example period between July 24 – July 27, 2007. Measured data is from Kvarnberget, Göteborg. Comparison between measured and calculated data Data during wet operation, Kvarnberget Göteborg , 2007

24 22

Calculated data

20 18 16 14 12 10 10

12

14

16

18

20

22

24

Measured data

Figure 4.12

Comparison between measured and calculated data. Measured data is from Kvarnberget, Göteborg. Total number of measurement samples during wet operation between May 1 – August 31, 2007

67

Table 4.8

Accuracy of the cooling tower model in this thesis method compared with data from measurements from Kvarnberget, Göteborg, 2007

Deviation [°C]

Supply water temperature

Mean value

-0,51

Maximum value

-0,98

Standard deviation 95% Confidence interval

0,14 ±0,04

The mean value, maximum value, standard deviation and the 95% confidence interval of the deviations indicated in figure 4.12 are presented in table 4.8. As can be seen in figure 4.11 and more so in figure 4.12 there is a deviation between calculated and measured data. The mean value of the deviation is -0,51 °C (table 4.8). Part of the deviation can be explained by the fact that the evaporative free cooler in the pilot plant at Kvarnberget, Göteborg, is a mixture of an adiabatic cooler and an ordinary cooling tower. When studying the evaporative free cooler in active operation it was clear that the flanges was only partly wetted due to the direction of the spray water, which was coming from underneath the cooler. This makes the evaporative free cooler partly an adiabatic cooler, partly an ordinary cooling tower. The CTM, on the other hand, is a model of an ordinary cooling tower. An adiabatic cooler cannot cool a circulating liquid as efficient as a evaporative cooling tower, everything else equal, because an adiabatic cooler is only using sensible cooling, although the ambient air is evaporatively cooled.

68

5 Simulation of a free cooling system with evaporative cooling tower In this chapter, the basis and the prerequisites are presented for the thermal analysis of a cooling system with an evaporative cooling tower. The analysis is made through simulations of a section of an office floor. In that section, comprising ten office rooms, one room is selected to be the space where the resulting thermal climate is logged and analysed. The primary thermal climate variable is the room dry bulb air temperature. The simulations are first made on a base case building and HVAC-system followed by a parameter variation analysis. In section 5.1, the selected simulation tool is presented. In section 5.2 is the prerequisites at base case described followed by section 5.3 where the simulation methodology with two different simulation cases is described. Section 5.4 explains the parameter analysis methodology including a discussion and motivation of the considered parameter alterations.

5.1

Simulation tool

There are a number of internationally well known building simulation tools that would be suited for the required analysis in this project, e.g. TRNSYS, DOE or Energy Plus. However, another tool was selected for the task; IDA Indoor Climate and Energy (ICE). It is an advanced tool for simulation of thermal comfort, indoor air quality and energy usage in buildings. It has more than 400 commercial licenses and about 1000 registered users on a free web based application called IDA Room (a limited version of IDA ICE). The users are mostly HVAC designers but also educators and researchers. Most of the licenses are in Sweden and in Scandinavia. The first version was released in May 1998 and the latest version, 4.0, was released in June 2009. Using IDA ICE would facilitate the dissemination of the results from this project in Scandinavia. Tools like TRNSYS, DOE, Energy Plus, or any other foreign advanced building simulation tool, have a very limited number of users within Scandinavia. IDA ICE has some unique characteristics as a building performance simulation tool, especially useful in this project. 1) All mathematical models are described in terms of equations in a formal language called Neutral Model Format (NMF). This makes it easy to replace and upgrade program modules. The end user can also view all NMF-models which makes the program open and transparent. In IDA ICE mathematical models may be reconnected arbitrarily by the end user. 2) Advanced users can use IDA SE (IDA Simulation Environment) in conjunction with IDA ICE to develop and implement new models and tailored user interfaces according to their own needs. The cooling tower developed model in this thesis, the CTM, is written in NMF-language and implemented through IDA SE into IDA ICE. A complete printout of the cooling tower model in NMF source code is shown in Appendix A. It should be noted that no changes in the source code of IDA ICE is required when adding a new model.

69

3) The full system of equations, describing the building and the HVAC system, is solved with a general purpose, variable time step solver called IDA Solver. The time resolution can vary from the maximum time step set by the user, usually 30 – 60 minutes, down to fractions of minutes. The determination of the length of each time step is based on the difference between a predicted and the actual solution at the current time step. IDA ICE is designed to handle: Multiple zone dynamic and simultaneous bulk air flow and heat balance, including specific contributions from sun, occupants, equipment, lights, ventilation, heating and cooling devices, surface transmissions, air leakage, large vertical openings, cold bridges and internal objects such as furniture. Different heating and cooling devices in each zone, e.g. hydronic or electric heating baseboard panels, floor heating, chilled ceiling panels and passive or active (ventilated) chilled beams. Solar influx through windows with full account for local shading devices as well as surrounding buildings and other objects. Control of shading devices dependent on solar influx and/or wind. Detailed 3D direct and diffuse shading calculations. Air and surface temperatures Operating temperature at multiple arbitrary occupant locations, e.g., in the proximity of hot or cold surfaces. Full non-linear Stephan-Bolzmann radiation with view factors is used to calculate radiation exchange between surfaces. Non-linear correlations for surface film coefficients Directed operative temperatures for estimation of asymmetric comfort conditions Comfort indices, PPD and PMV, at multiple arbitrary occupant locations Daylight level at an arbitrary room location Zone CO2 and moisture levels, both of which may be used for control of VAV system air flow A model for vertical air gradient calculation is available Wind and buoyancy driven airflows through leaks and large vertical openings via a fully integrated airflow network model. This enables study of, e.g., temporarily open windows or doors between rooms. Airflow, temperature, moisture, CO2 and pressure at arbitrary locations of the air handling and distribution systems Power levels and temperatures for primary and secondary system components Total energy cost based on time-dependent prices as well detailed energy accounts As a building simulation tool with the above mentioned properties, IDA ICE can be considered as one of the most advanced building simulation tools on the market. The original development of IDA ICE was requested, specified and partly financed by a group of thirty leading Scandinavian companies within the building sector. The mathematical models were originally developed at the Royal Institute of Technology in Stockholm (KTH) and at Helsinki University of Technology. The company Bris Data, which later changed the name to Equa, developed IDA ICE. All models in the simulation tool are available as NMF source code.

70

5 Simulation of a free cooling system with evaporative cooling tower

The models are not tailored to Scandinavian needs but seek to capture the international state-of-the-art in building performance modelling. Whenever appropriate, models recommended by ASHRAE have been used. IDA ICE has been validated in several ways. A large number of inter-model comparisons have been made against the BRIS program (Brown 1990), which in turn has been extensively validated against measurements of thermal performance on real buildings. An extensive empirical validation exercise based on test cells has been carried out within the IEA Solar Heating and Cooling Task 22 (Guyon et al 1999). Validation with IEA Building Energy Simulation Test (BESTEST) (Ascherman 2000) Validation according to CEN 13791 ”Thermal Performance of Buildings – Calculation of Internal Temperatures of a Room in Summer Without Mechanical Cooling – General Criteria and Validation Procedures”, (Kropf and Zweifel 2002) Validation focused on radiant heating and cooling systems with RADTEST has been carried out within the IEA Solar Heating and Cooling Task 22, (Ascherman and Zweifel 2003)

5.2

Prerequisites at base case

The intention with the base case building and its technical systems is that it will reflect typical conditions of a normal office building, the activity in it and the standard and size of the technical systems. Hence, the following prerequisites are important to notice: The base case building does not meet the current Swedish building code concerning insulation. This is the case with the bulk part of commercial buildings in Sweden together with the rest of the building stock in the tertiary sector in the northern parts of Europe. The base case building is not equipped with smart, automatic and efficient solar shadings. The lighting is not the most energy efficient on the market, and there is no automatic control of the lighting regarding daylight levels or presence of people in the rooms. The appliances in the office rooms, e.g. computer and screen, have no energy saving functions activated. The ventilation system has no automatic control regarding operating hours, night cooling or demand control of volume flow through variable frequency drive or other device. The individual temperature control device in the rooms, controlling the chilled beams, has no built-in “intelligence” with variable set point temperatures between day and night or any fuzzy-control. It is a basic PI-controller to keep the air temperature in the room constant. If all these prerequisites were actually a part of the base case building, i.e. omit all no or not above, it would most likely lead to a lower indoor temperature during hot and sunny

71

weather, and lower energy use, than presented in this thesis. On the other hand, the results would then not reflect the conditions of a common real life commercial office building. A complementary discussion and motivation of most of the prerequisites is found in section 5.4.2 Discussion and motivation of parameter alterations. 5.2.1 Model building The base case building consists of an imaginary multi storey non-residential building. The analysed room is situated in a floor somewhere in the middle section of the building. The floor contains ten rooms laid out as in a normal office building, se figure 5.1. The office building is located in a medium climate of the northern parts of Europe, in this case chosen to be London Gatwick. The analysed room, marked with dashed lines in the figure, has two adjacent rooms on each side and five rooms on the opposite side of the corridor and the middle section. In a real life building the middle section is normally containing service rooms such as copying and printer rooms, toilets and archives. 15 m

Adjacent office rooms

Middle section

Corridors

20 m

N Analysed room

Figure 5.1

Original layout of the office floor

The analysed room is placed in a normal office environment with adjacent rooms and a corridor to take in to account the variations of the coincident heat flow and airflow with the corridor as well as the outdoor environment. It is assumed that the doors between all the rooms and the corridor always are open and that the two corridors are connected so that airflow is possible due to differences in density or pressure between all the zones. The layout of the office floor in the base case building in figure 5.1 is simplified when modelled in IDA ICE. The simplification consists of the following:

72


The adjacent row of five office rooms on the opposite side of the floor is lumped into one room. The neighbouring two rooms on each side of the analysed room are lumped together respectively. This is made of mainly two reasons. The first reason is that the exact modelling of the adjacent rooms is not necessary since their task is to exchange air and heat with the corridor so that the resulting air temperature of the corridor will be close to a real life situation. The second one is because the simulation time is linearly proportional to the number of zones in IDA ICE. Since a lot of simulations are required within this work, a short simulation time is preferred. The simplified base case building model with the office floor is shown in figure 5.2. The adjacent rooms are now double the size of the analysed room and the room on the opposite side of the corridor is five times the analysed room in size. The total size and floor area, the doorway area of the rooms in the simplified layout is however exactly the same as in the original layout. The corridor is lumped into one rectangular room with the same floor area as the two original corridors. The middle section, surrounded by the corridors in the original building, is modelled as one single imaginary inner wall placed in the corridor to simulate the thermal mass of the middle section. The wall is imaginary in the sense that it does not take up any space or hinder any air movements. The area of the single inner wall is equal to the area of the walls and slabs in the original middle section between the corridors.

Lumped middle section to a single inner wall equivalent in area and thermal weight to middle section in figure 5.1.

Lumped office rooms

12 m Lumped corridor Analysed room

Figure 5.2

Simplified layout of the office floor as in IDA ICE

The physical size of the rooms and the corridor is remained constant throughout the whole analysis. The construction of the different parts in the building, such as floor, ceiling, outer wall, windows and inner walls, are always the same in all the rooms. If a change is made in one or more construction parts due to parameter variation analysis, this change is made i all the rooms in the floor.

73

The floor and ceiling slab are made of reinforced concrete. The floor has an additional layer of linoleum carpet. There is a suspended false ceiling with acoustic tiles mounted to the slab. The inner walls are made of wood framework with single gypsum board on each side with insulation between. The outer wall is made of lightweight concrete and insulation. A top layer of plaster is added on each side. This construction type is normally categorized as medium heavy concerning thermal capacity. The U-value of the outer wall is chosen to 0,4 W/m2 K which is in accordance to Swedish building code in the mid 1970’s. The windows have double pane clear glass type with a U-value of 2,5 W/m2 K. 5.2.2 Analysed room The analysed room is where the indoor dry bulb temperature and other variables are registered. The analysed room is 12 m2 with a room height of 2,4 m between floor slab and false ceiling. It has one external wall, including a window, facing south. The size of the window is adjusted to the required solar heat gain in the room; see section 5.2.3 Internal heat and moisture generation. The remaining three walls are inner walls facing adjacent rooms and the corridor. The corridor wall has a door, which is assumed to be always open. For details concerning construction, see section 5.2.1 Model building above. 5.2.3 Internal heat and moisture generation The internal heat generation comprises heat from humans, lighting, office equipment and solar radiation through the window. In the analysed room, one person is present together with lighting fitting and office equipment. The adjacent and opposite office rooms have the same specific internal heat generation as the analysed room, for humans as well as equipment and lighting. Working hours is 08 – 17 (8 am. to 5 pm.), Monday to Friday all year round. Time for holidays is excluded in the simulations to avoid possible coincidence with hot and humid weather conditions and holidays, i.e. low internal heat generation. The person has in the base case an activity level equal to 1,2 Met and a clothing equal to 0,9 Clo during 1st of May to 30th of august. The rest of the year the clothing is equal to 1,1 Clo to reflect that people have more clothes during colder weather conditions. With given values of activity and clothing the generated sensible heat from a person is dependent of the indoor air temperature, mean radiation temperature and the indoor air velocity. If these two temperatures are equal to 22°C, the sensible heat is about 89 W in the summer (1st of May to 30th of august) and about 79 W the rest of the year. The indoor air velocity is set to a constant value of 0,1 m/s. The moisture generation is calculated from the values of the activity level (1,2 Met) and the partial pressure of the water vapour in the room air. The partial pressure of the water vapour is determined from the air temperature and the relative humidity in the room. The heat and moisture generation from persons is in IDA ICE calculated according to Fanger (1972). The heat and moisture generation is therefore dependent on for example, the level of activity, the clothing and the indoor air temperature as well as the mean radiant temperature. During time other than working hours, the internal heat and moisture generation from persons is zero.

74


Neither the lighting nor the office equipment is specified in detail; instead, it is given as a specific heat gain related to the floor area. The lighting is set to a specific heat gain of 12 W/m2, which equals a total heat generation of 144 W in the office rooms. In the corridor, the heat gain from lighting is set to 6 W/m2. The office equipment has a specific heat gain of 12 W/m2, which equals a total of 144 W for the room. In the corridor, no equipment load is present. In the heat transfer process the lighting is set to 30% convection – 70% radiation and office equipment to 80% convection – 20% radiation. During time other than working hours the internal heat generation from lighting in office rooms is set to 0 W/m2. Office equipment is to some extent often on during non working hours and is therefore set to 6 W/m2. In the corridor, the lighting is set to 1 W/m2 during non working hours to simulate that some lighting always is on. The sensible heat generation from humans, lighting and equipment with the numbers above is 364 W which equals a specific heat gain of 30,3 W/m2. The maximum total sensible heat gain during daytime is set to 60 W/m2 and the remaining load, about 30 W/m2 comes from solar radiation through a window in each office room. The window in each room is adjusted in size so that the maximum total sensible heat generation will be 60 W/m2. The adjustment of window size to tune the maximum solar radiation to 30 W/m2 is made during a heat wave simulation at design day conditions, where the sky is constantly clear and the insolation is at maximum. For explanation of heat wave simulation, see section 5.3.1 Heat wave simulation. The total solar radiation at 30 W/m2 is the maximum value during a day. The contribution to the total solar radiation comes from direct beam, diffuse and ground reflected radiation. The total radiation varies throughout the day with a maximum at noon. It is stressed that the maximum total sensible heat gain of 60 W/m2 includes the contribution from the solar radiation in the room. Expressing the solar radiation to the room, and the total heat gain, in specific numbers leads to a more general approach when performing the simulations. Setting the solar design heat gain in the room to 30 W/m2 implies practically almost infinite combinations of the following parameters: Number of window panes Physical size of the window Possible solar protective coating on one or more panes Type of coating Possible indoor and/or outdoor solar shading device Type of solar shading devices

75

5.2.4 Cooling tower The cooling tower, which is connected to the air-cooling coil in the air handling unit and the local cooling beams in the rooms, figure 5.4, is selected to have the following features: Closed circuit with steel tube bundles with smooth external wall tubes suitable for external water spraying; see schematic sketch in figure 5.3. The choice of a closed circuit tower versus an open type of cooling tower has no particular affect on the analysed variable (indoor temperature). The closed circuit has an extra pump for the spray water circulation which adds to the total energy usage of the tower. Normally the closed circuit type requires a bigger volume of the tower compared to an open type. A closed circuit tower is used more often in building related systems than an open circuit cooling tower. The reason for this emanates mainly from a desire to reduce the maintenance cost for water treatment. Counter flow configuration. This configuration is the most common and it also gives the lowest leaving temperature of the cooled fluid compared to a cross flow configuration with all other parameters equal. Wet (evaporative) type of tower. This is an obvious choice since this thesis deals with evaporative cooling towers only. An intermediate closed circuit with freeze protected coolant with a heat exchanger between the primary and secondary fluids, figure 5.4. This choice origin from the possible occurrence of cooling demand in a building during wintertime when the risk of freezing of the coolant is high. The heat exchanger is placed indoors where the air temperature never reaches freezing degrees. The coolant in the primary (intermediate) circuit is a 40% propylene glycol-water solution while the liquid in the secondary circuit is merely water.

Spray nozzles


Coolant in closed loop system Ambient air Spray water basin

Figure 5.3

76

Schematic illustration of a closed type cooling tower used in the simulations.


Supply air to zones

Part of the AHU air heating coil

System boundary for the cooling tower model

air cooling coil supply air fan Intermediate freezeprotected circuit

Closed type cooling tower

Hydronic room cooling devices, e.g. fan coils or cooling beams

Heat exchanger

Secondary water based circuit Outdoor

Figure 5.4

Indoor

The cooling tower model and its connection with the rest of a possible cooling system. Temperature

Return side Warm side

Wet bulb temp. out

Secondary range Liquid in secondary circuit (water)

Primary range

Supply side


Total approach


Figure 5.5

Temperature diagram of a counter flow cooling tower with intermediate circuit and visualisation of definition of design parameters.

The design parameters, explained in figure 5.5, which influences the performance of the cooling tower is chosen as follows:

77

Primary approach: The approach in the primary circuit is set to 4°C. The value of the approach is critical concerning economical decisions in the design phase. The economical aspect covers mainly short term aspects such as investment cost due to increasing size of the cooling tower with decreasing values of the approach. In the literature the lower limit for the design approach, based on economic reasons, is almost always 3°C (approx. 5°F). A further discussion about this topic is presented in section 5.4.2 Discussion and motivation of parameter alterations. Secondary approach: The approach in the secondary circuit is set to 1,5°C. The heat exchanger is a plate and frame type which normally offers excellent efficiency, i.e. low values on approach. Lower approach than about 1°C is in the literature, e.g. de Saulles (1996), not advisable due to too high investment costs. Primary Range: The range in the primary circuit is chosen to the same value as in the secondary circuit, i.e. 3°C. This leads to equal mass flows in the primary and secondary circuit at design conditions. Secondary Range: The secondary range is set equal to 3°C. This temperature difference is very common in comfort cooling systems with chilled ceilings or chilled beams. & M Mass flow quota of liquid and air ( l ): This relation primarily state the & M a

value of the mass flow of air since the mass flow of the liquid is given from the design cooling capacity and the range of the primary liquid. The value of the mass flow quota is normally in the range between 0,5 and 2. In the base case the mass flow quota is set to 1,0. According to Söylemez, (2003) the optimum flow quota is about 1 for mean temperature of cooling tower water at 22 – 27°C. Cooling capacity: The cooling capacity is set equal to the total design cooling capacity of the cooling beams together with design cooling capacity of the air cooler in the air handling unit. With 50 W/m2 in cooling capacity of the cooling beams and a specific airflow of 1,25 l/s m2 the total design cooling capacity of the tower is 7,69 kW. Design conditions: The ambient design conditions are set to 25°C dry bulb temperature and a relative humidity of 60% RH. This equals a wet bulb temperature of 19,5°C, which is very close to the ASHRAE design wet bulb temperature (0,4%) of London Gatwick that is 19,3°C (ASHRAE 2001a).

5.2.5 HVAC system The HVAC system consists of a ventilation system, a heating system and a cooling system. The basic configuration is a water-air system, i.e. room units on hydronic circuits provide heating and cooling and ventilation is provided by a CAV-system to meet the requirements of indoor air quality.

78


Ventilation system The ventilation system is a basic CAV-system with constant air volume flow and constant supply temperature. The system is supplied with air from an Air Handling Unit (AHU) comprising air supply and discharge fans, air-to-air heat exchanger for heat recovery, air heating coil and air-cooling coil see figure 5.6. The air-cooling coil is supplied with chilled water from the cooling tower, figure 5.4. return air fan

Air-to-air heat exchanger

return air from building Ambient air supply air to building

air cooling coil supply air fan air heating coil

Figure 5.6 Basic scheme of the Air Handling Unit (AHU) in the base case In the office floor, the office rooms are supplied with air at a rate equivalent to 1,25 l/s m2 and the supply air has a set point temperature of 17°C. The air flow rate is maintained constant throughout the year, only during workdays though. The actual supply air temperature is however not constant during warm weather conditions, due to the natural limitations of the cooling tower capacity to cool the water. In the office rooms, supply air diffusion is of mixing ventilation type. The supply air in the office rooms is transported to the corridor where the exhaust air is transported from the corridor. The ventilation system is operating constantly during workdays (Monday 06:00 - Friday 18:00) all year and is shut down during weekends. The operating scheme is not optimal concerning minimal energy usage for the ventilation system (heating and electricity). This is however not an issue in this thesis since the energy and temperature performance of the cooling system, i.e. the cooling tower, is in focus. The primary objective of the ventilation system in this case is to remove odour and pollutants, thus a moderate airflow. A secondary objective is to provide cool air to supplement the cooling from the chilled beams. About 20% of the design cooling capacity in the room is supplied from the ventilation at design conditions. The operating schedule therefore includes 24-hour operation during workdays to utilize nighttime cooling. No control function is added to switch nighttime cooling on or off based on some criteria.

79

Heating system A heating system is added to the office floor to provide a minimum indoor air temperature during colder periods. Ordinary hydronic baseboard heaters placed under the windows supply the heating to the office rooms. The heat capacity of the heaters will be sufficient to keep the set point temperature at the coldest day. The baseboard heaters are controlled by an ordinary proportional thermostat with a proportional band at 2°C and a set point temperature of 20°C. The corridor has no heating device. The primary heating source is assumed to be district heating.

Cooling system The cooling system consists of a cooling tower, as described in section 5.2.4 Cooling tower, a distribution system of pipe work, an air cooling coil in the AHU and room cooling devices, see figure 5.4. The rooms cooling devices is in this case active cooling beams. An active cooling beam combines cooling and ventilation functions into a single unit. Heating is also possible with this type of beam, but is not included in this thesis. To minimize the simulation time, every office room has only one cooling beam. In a real life situation an office room may have two or several cooling beams depending on room size, cooling loads and other design factors. To lump cooling beams into one single does not affect the simulation result since the sum of a number of cooling beams has the same cooling capacity and characteristics as one lumped beam. In the heat transfer process, a cooling beam uses nearly 100% convective heat to remove excess heat in contrast to a chilled ceiling or a cooling panel, which use approximately 50% radiation and 50% convection. For a more in depth discussion about room cooling devices in general see Nilsson (ed) (2003) and chilled ceilings and chilled beams i detail see Novoselac & Srebric (2002), Mott et al. (2001) and Sodec (1999) The sizing of the cooling beams in all office rooms is based on the following: The design room air temperature is 25°C The design mean cooling water temperature is 17,5°C with a supply temperature of 16°C and a return temperature of 19°C The design temperature difference between the room air temperature and the mean cooling water temperature is 7,5°C The design secondary temperature range is 3°C (difference between cooling water supply and return temperature), see figure 5.5. The design cooling capacity is 600 W for an office room at 12m2, equal to 50 W/m2. In a room with the size of 12 m2, as in the base case, two modern active cooling beams with the length of 1,8 m, will suffice providing the design cooling capacity. The cooling beam in each room is controlled by a PI-controller with a set point temperature at 22°C. The cooling system will be operating 24 hours a day all year to ensure cooling whenever the indoor air temperature exceeds 22°C, i.e. set point temperature.

80


5.3

Simulation methodology

One objective with this thesis is to investigate the indoor climate when a building is cooled solely by a cooling tower. The primary variables to be analysed is therefore dry bulb temperature and relative humidity of the indoor air. These variables are chosen simply because they are widely used as indicators of indoor thermal climate and they are also easy to measure. In section 5.4 Parameter analysis methodology, different sets of parameters that will be used in the sensitivity analysis are discussed. The analysis of the primary variables for each set of parameters is made in two different ways. The first way is to make simulation runs with the building and the cooling system exposed to a heat wave with a duration of five working days. This is a normal way of making simulations in the design process to either calculate the resulting indoor temperature or the necessary cooling capacity to meet the indoor temperature criterion. The other way of making the parameter analysis is to make full year simulations to examine the variation and duration of the indoor temperature during a normal year. The normal year in this case is a Typical Meteorological Year (TMY-year), issued by ASHRAE (2001b). Secondary variables such as energy usage and COP for the cooling tower, system temperatures etc. will be logged, however only with full year simulations and for a limited set of parameter variations.

5.3.1 Heat wave simulation Testing and sizing the cooling system together with the building against a simulated heat wave is a design strategy, which is normal among HVAC consultants. It is therefore obvious that this design strategy should be applied in this thesis to evaluate the performance of the cooling system and building together. The heat wave can either be chosen from real meteorological data or created by a synthetic 24 hour periodic sinus wave of ambient temperature. In this thesis the latter is chosen since it is the most commonly used during the design phase. The duration of the heat wave is chosen to five days, Monday to Friday, to coincide with the working days and consequently maximum cooling loads, and with a 24-hour periodic outdoor temperature curve. The maximum and minimum outdoor dry and wet bulb temperature is taken from ASHRAE (2001a). Design conditions are chosen based on the wet bulb temperature together with the mean coincident dry bulb temperature. In normal design of an air conditioning system the design value is the dry bulb temperature. In this case it is preferred to use the wet bulb temperature as the design value since the performance of the cooling tower normally has a greater effect on the indoor air temperature than the design dry bulb temperature has. The 0,4% value is chosen as design value. The 0,4% value represent a wet bulb temperature value that is exceeded on average by 0,4% of the total number of hours in a year (8760 h/year). 0,4% of a year equals 35 h/year.

5.3.2 Full year simulation A full year simulation, together with the heat wave simulation, gives a more complete picture of how the cooling system and the indoor parameters will perform. The full year

81

simulation is made with climate data from a Typical Meteorological Year (TMY), issued by ASHRAE (2001b). A TMY consists of twelve typical Meteorological Months’ (TMM) selected from the calendar months in a weather database comprising a period of normally 10-30 years. For example, January of 1983 may be selected as the first TMM, February of 1989 as the second TMM, and so on. All the twelve selected months will then be combined to form the TMY. The TMY’s used in this thesis are based on climate data for the years between 1982 and 1993. Selection criteria of a TMM are based on statistical analysis and evaluation of four weather parameters: Global Solar Radiation, Dry Bulb Temperature, Dew Point Temperature and Wind Speed. From these weather parameter a set of nine indices are selected and given different weightings. Of the nine indices, the ones which have the highest weighting are daily mean values of dry bulb temperature, dew point temperature and daily total of global solar radiation. Extreme values have lower weighting in the selection process. The selection process of TMM leads to a TMY, which contains a distribution of weather data reflecting a “normal” year rather than a year containing extreme values. This in turn leads to an indoor temperature distribution over a year that reflects a mean indoor condition occurring over a longer time span.

5.4

Parameter analysis methodology

When a building, a cooling tower and an HVAC-system as described in section 5.2, is coupled in a building simulation program it is obvious that the resulting temperature and humidity of the indoor air is dependent on many parameters. Some parameters have stronger and some has weaker influence on the resulting values of the primary variables described in section 5.3. To sort the parameters which have an influence on the primary variables, a single parameter sensitivity analysis will be done at first. In this case, it means that a number of parameters will, one by one, be changed from an original value stated at the base case. The change will be done in both directions for each parameter, i.e. a parameter will be changed to both a higher and a lower value. When all the parameters have been investigated, they can be sorted in strength according to their impact on the primary variables.

5.4.1 List of parameter alterations In table 5.1 a complete description of parameters, which are to be altered, is displayed. Values of both base case and the higher and lower values are presented. In section 5.4.2 Discussion and motivation of parameter alterations, each parameter is discussed and the high and low values are motivated.

82


Table 5.1 List of parameter alterations Parameter

Lower indoor temp.

Base case

Higher indoor temp.

3°C

4°C

5°C

2. Design range (both primary and secondary)

2°C

3°C

4°C

3. Design mass flow quota of liquid and air, (Ml/Ma)

2,0

1,0

0,5

4. Design cooling capacity, Q& CT

1,25 ⋅ Q& CT

Q& CT

0,75 ⋅ Q& CT

”heavier”

”medium”

”lighter”

Cooling Tower 1. Design primary Approach

Building 5. Thermal capacity (weight) 6. Location in building

Ground floor Middle floor

Top floor

7. U-value in external walls U-value in windows

0,2 W/m2 K 2,0 W/m2 K

0,4 W/m2 K 0,6 W/m2 K 2,5 W/m2 K 3,0W/m2 K

8. Internal heat gain; magnitude (design maximum value, including solar radiation)

50 W/m2

60 W/m2

70 W/m2

9. Internal heat gain; working hours

8 h (08-16)

9 h (08-17)

10 h (08-18)

10. Geographical location

Stockholm/ Arlanda

London/ Gatwick

Berlin

11. Facade direction

East

South

West

HVAC-system 12. Design temp. difference Beams: Troom − Twater

5°C

7,5°C

10°C

13. Design cooling capacity cooling beams

60 W/m2

50 W/m2

40 W/m2

14. Design ventilation rate

1,67 l/s m2

1,25 l/s m2

0,83 l/s m2

15. Supply air temperature set point

15°C

17°C

19°C

16. Design secondary approach (heat exchanger)

0°C

1,5°C

3°C

17. Temperature gradient in office room

2°C

1°C

0°C

18. Set point temperature cooling control

21°C

22°C

23°C

83

The reader should be aware that when changing the value of a parameter to a lower value, the indoor air temperature might rise and vice versa. Therefore all parameter changes causing the indoor air temperature to decrease is placed in the column “Lower indoor temperature”. Consequently, all parameter changes causing the indoor air temperature to increase are placed in the column “Higher indoor temperature”. This arrangement makes it easier to display the results of the sensitivity analysis.

5.4.2 Discussion and motivation of parameter alterations 1. Design Primary Approach The design primary approach is a vital parameter for determining the ability of the cooling tower to chill the cooling water as close as possible to the ambient air wet bulb temperature. The parameter is however influencing size and investment cost of the cooling tower. In figure 5.7 the relation between the design UA-value and Primary Approach for three different values of Primary Range, 2, 4, and 6°C, of a cooling tower. The UA-value can be considered as proportional to the size of the cooling tower and consequently the investment cost.

Relative design UA-value (-)

1,0 0,8 0,6 0,4 0,2 0,0 2,0

3,0

4,0

5,0

6,0

7,0

8,0

Primary Approach (°C) Primary Range 2,0

Figure 5.7

Primary Range 4,0

Primary Range 6,0

The relation between UA-value and Primary approach for three different Primary ranges

As can be seen the UA-value rises quickly when the Primary Approach reach under 4°C. The Primary Range also has an influence on the UA-value so that the lower the Range the higher the UA-value. In the literature, there is a unanimous agreement that the lowest practical and economical limit for the Primary Approach is 3°C (5°F). Based on this information and on figure 5.7, it is logical to set the parameter variations to 3, 4 and 5°C.

84


2. Design Primary Range The influence on the UA-value from the design Primary Range is discussed above. The values of the Primary Range are chosen as low as practically possible to make the mean temperature in the circuit as low as possible. The values of the Primary Range are also synchronized with the Secondary Range and are therefore chosen to 2, 3 and 4°C. These values are common temperature differences, Ranges, in the secondary circuit when connected to chilled ceilings or chilled beams.

& M l ) & Ma The design flow quota between the air and water mass flow, in most literature called L/G (Liguid/Gas), basically determines the air mass flow rate. Normally the liquid mass flow is known from the choice of design Range and the design cooling capacity, and the air mass flow is the given through the design mass flow quota of liquid and air (L/G). In figure 5.8, this parameters relative impact on the UA-value of a cooling tower is shown. The impact on the UA-value is relative moderate in the displayed range of L/G values. This parameter has more importance regarding the energy use since the major part of the energy use of a cooling tower is due to the fan, typically about 80% of total use of electricity.

3. Design mass flow quota of liquid and air, (


1,0 0,8 0,6 0,4 0,2 0,0 0,5

0,75

1

1,25

1,5

1,75

2

Design flow quota, L/G (-)

Figure 5.8

The relation between UA-value and the design mass flow quota of liquid and air (L/G).

In the literature, a variety of L/G values can bee found. The differences between them are not very big and most values of L/G are in the range of 0,5 to 2. In Söylemez (2004) the optimum values of L/G is in the range from 0,82 to 2,45 for a forced draft counter flow cooling tower, depending on mean water temperature and air pressure, i.e. geographical elevation. The values of the design flow quota are chosen to be 0,5, 1,0 and 2.

85

4. Design cooling capacity, Q& CT (cooling tower) The design cooling capacity of the cooling tower, Q& CT , i chosen to the same value as the total design cooling capacity of the cooling beams together with design cooling capacity of the air cooler in the air-handling unit. With 50 W/m2 in design cooling capacity of the cooling beams and a specific airflow of 1,25 l/s m2 at base case, the total design cooling capacity of the tower is 7,69 kW. The parameter range is simply chosen to ± 25% of Q& CT . The design values of the outdoor conditions at which the design cooling capacity is determined is 25°C dry bulb temperature and a relative humidity of 60%. This equals a wet bulb temperature of 19,5°C, which is very close to the ASHRAE design wet bulb temperature (0,4%) of London Gatwick that is 19,3°C (ASHRAE 2001a). The design outdoor condition has an influence on the size of the cooling tower. In figure 5.9 the relation between the UA-value and the design outdoor condition is shown.


1,0 0,8 0,6 0,4 0,2 0,0 10

12

14

16

18

20

22

Design ambient wet bulb temperature (°C)

Figure 5.9

The relation between UA-value and the design outdoor wet bulb temperature and an approach of 4°C.

5. Thermal capacity The thermal capacity of the rooms and the building structure is a well known factor for playing an important role of the resulting indoor thermal climate. In this case, no numerical values are stated since the most common way to describe the thermal capacity of a room is in terms of ‘light’, ‘medium’ and ‘heavy’. In this context, the terms are instead chosen to be ‘lighter’ ‘medium’ and ‘heavier’. This is because the term ‘light’ is normally connected to a construction merely consisting of wood joist frames, wooden floor and ceiling and all the walls made of wood framework and gypsum boards. Commercial buildings with this type of construction are extremely unusual. In this case, all the variations of thermal capacity assume a building based on a construction with concrete slabs. In figure 5.10 – 5.12 the three different variations of constructions with different thermal capacity are presented.

86


Ceiling slab Thermal capacity ’heavier’

Inner wall

Facade

Window

Floor slab

First layer on each construction is always closest to the room.

Façade

Ceiling slab

Floor slab

Inner walls

Concrete 0,15 m Insulation 0,08 m Concrete 0,08 m Plaster 0,01 m

Concrete 0,15 m Light concrete 0,02 m Plastic sheeting 0,005 m

Plastic sheeting 0,005 m Gypsum board 0,026 m Light concrete 0,02m Insulation 0,03 m (frame Concrete 0,15 m work 0,095 m) Gypsum board 0,026 m

Figure 5.10 Building construction with Thermal capacity ‘heavier’

Thermal capacity ’medium’

False ceiling


Façade

Ceiling slab

Floor slab

Inner walls

Light concrete 0,125 m Insulation 0,02 m Light concrete 0,125 m Plaster 0,01 m

False ceiling 0,05 m Concrete 0,15 m Light concrete 0,02 m Plastic sheeting 0,005 m

Plastic sheeting 0,005 m Light concrete 0,02 m Concrete 0,15 m False ceiling 0,05 m

Gypsum board 0,026 m Insulation 0,03 m (frame work 0,095 m) Gypsum board 0,026 m

Figure 5.11

Building construction with Thermal capacity ‘medium’

87

Thermal capacity ’lighter’

Wooden parquetry


Façade

Ceiling slab

Floor slab

Inner walls

Gypsum board 0,026 m Frame work and insulation 0,1 m Gypsum board 0,009 m Wood panel 0,025 m

False ceiling 0,05 m Concrete 0,15 m Light concrete 0,02 m Wooden parquetry 0,01 m

Wooden parquetry 0,01 m Light concrete 0,02 m Concrete 0,15 m False ceiling 0,05 m

Gypsum board 0,026 m Insulation 0,03 m (frame work 0,095 m) Gypsum board 0,026 m

Figure 5.12 Building construction with Thermal capacity ‘lighter’

6. Location in building This parameter examines the impact of the location of the office floor in the building. Two obvious locations are either on the ground floor, utilising contact with the colder temperature of the ground, and on the top floor, with the extra heat load from the roof. A middle floor location is normally considered as a base case location.

7. U-values in wall and windows The U-values of the base case building have been discussed in section 5.2.1 Model building. The parameter alteration “low temperature” is set to a U-value of 0,2 W/m2 °C for the outer wall and 2,0 W/m2 °C for the window. These values are equal to a modern Swedish building with well insulated walls and three pane windows. The parameter alteration “high temperature” is set to a U-value of 0,6 W/m2 °C for the outer wall and 3,0 W/m2 °C for the window. These values are equal to an older Swedish building (before 1960) with poorly insulated walls and two pane windows.

88


8. Internal heat gain; magnitude The internal heat gains from people, lighting and equipment of the base case building have been discussed in section 5.2.3 Internal heat gain and moisture generation. The basic parameters governing these internal heat sources are kept constant through the parameter variations. The parameter variations in the total heat gain come from changes in the solar radiation to the room, due to changes of the size of the window. It should be noted that the values of the specific internal heat generation in table 5.1, section 5.4.1 List of parameter alterations, includes solar radiation in the room. In the literature, there are very few published values of normal ranges in commercial buildings, e.g. office buildings, which are based on actual measurements. Based on results from a series of test calculations the “low temperature value” is set to 50 W/m2. This value is close to what can be expected in a normal office room with the heat gains from people, lighting and equipment, as discussed in section 5.2.3 Internal heat gain and moisture generation, together with solar radiation from a smaller window without solar shading or a larger window with good solar protection. The “high temperature” value of 70 W/m2 is close to what can be expected in a normal office room with a larger window with no or less efficient solar protection. Based on this discussion it is assumed that the range of 50 – 70 W/m2 covers a large part of normal office buildings.

9. Internal heat gain; working hours The internal heat gains from people, lighting and equipment of the base case building are not only a matter of the magnitude of heat gain. It is also a matter of the daily duration of the internal heat gain. The working hours has been chosen to alter with ± one hour. It can be argued that working hours can be much longer in some commercial buildings. This is true, but they are very seldom much shorter than 08 – 16 as it is in the “low temperature” parameter setting. Since the working hours of the base case is from 08-17 the “high temperature” setting is consequently 08-18. It is also quite rare that working hours start earlier than 8 o’clock in the morning.

10. Geographical location Geographical location is a difficult parameter to determine. It is not simply a matter of finding locations with evenly distributed latitudes. A geographical location parameter is more of a climate parameter. The climate part of this parameter has bigger influence than merely the latitude since the cooling tower performance is highly dependent of the ambient air wet bulb temperature. At an early stage of this project it was decided to analyse climates in the northern Europe, i.e. north of latitude 48 – 49°N, mainly to cover geographical areas where cooling system applications with chilled ceiling or chilled beams are common. The available climate data, in the format of TMY-tears, issued by ASHRAE (2001b) have been used to determine suitable locations. The data set in a TMY-year does not contain wet bulb temperature per se, but by processing all climate data in IDA ICE it was possible to get hourly wet bulb temperatures for all locations.

89

The selection started with the “low temperature” parameter in order to involve a Swedish location. The choice fell on Stockholm/Arlanda as the “low” parameter. From that climate, the duration curves of the ambient wet bulb temperature for all locations in figure 5.13 and the bar graph in figure 5.14 were used to determine the two other climates. In figure 5.14 all the locations used are identified. The base case location was selected to be London/Gatwick and the “high temperature” parameter was chosen to be Berlin. As a complement to the single parameter variation a limited multi parameter variation is performed. In this parameter variation, there are five locations used; Östersund and Stockholm/Arlanda (Sweden), London/Gatwick (UK), Berlin (Germany) and Paris/Orly (France). These locations cover the climatic interval in figure 5.14 very well.

24 Wet bulb temperature (°C)

Berlin 22

London/Gatwick

20

Stockholm/Arlanda

18 16 14 12 10 7760

7960

8160

8360

8560

8760

Last 1000 hours of one year duration (8760 h)

Figure 5.13 Duration curves of ambient wet bulb temperature (WBT) of 25 locations in northern Europe. The duration displayed is the 1000 hours with the highest values of WBT. See figure 5.14 for a listing of the locations used.

90


Kiruna Östersund Aberdeen/Dyce Göteborg/Landvetter Dublin Köpenhamn Stockholm/Arlanda Karlstad Birmingham London/Gatwick Luxenburg München Prague Hamburg Stuttgart Bremen Berlin Köln Amsterdam Düsseldorf Frankfurt Brussels Warzaw Krakow Paris/Orly 14

15

16

17

18

19

20

21

22

Design wet bulb temperature (°C)

Figure 5.14 Design values of ambient wet bulb temperature (WBT) of 25 locations in northern Europe according to ASHRAE (2001a). The design values are those were the wet bulb temperature is exceeded 0,4%, i.e. 35 h/year, during a normal year.

11. Façade direction This parameter gives three obvious parameter alterations, east as the “low” parameter, south as the base case and west as the “high” parameter.

12. Design temperature difference Troom − Twater This parameter determines the characteristics of the cooling capacity of the cooling beams. The cooling capacity of an active chilled beam is dependent of the difference of the mean cooling water temperature and the room air temperature. This relation is showed in figure 5.15. The design mean water temperatures is chosen to 15, 17,5 and 20°C and the design indoor air temperature is set to 25°C in all cases. This leads to design temperature differences of 10, 7,5 and 5°C. Design temperature differences between 10 to 7,5°C represent fairly normal design values. A design temperature of 5°C is less usual due to the increase in the size, and hence investment costs, of the room cooling devices. The

91

smallest design temperature difference represents the steepest line, i.e. cooling characteristics, in figure 5.15 and vice versa. The steeper the line, the more rapid increase in cooling capacity when the room temperature increases. A steeper line, i.e. a smaller design temperature difference, also leads to more or longer cooling beams in a room and thus higher investment costs as mentioned above.

Relative cooling capacity (%)

200 180 160 140 120 100 80

Design temp. diff. 10°C

60

Design temp. diff. 5°C

40 20 0 10

12,5

15

17,5

20

22,5

25

27,5

30

Indoor temperature (°C)

Figure 5.15 Cooling capacity characteristics of cooling beams at design mean water temperatures 15, 17,5 and 20°C and a design indoor temperature of 25°C.

13. Design cooling capacity cooling beams There are several ways to determine the required cooling capacity for a room. In this case, the base case cooling capacity is simply chosen to the “rule of thumb”-value common among Swedish HVAC engineers, namely 50 W/m2. In a room with the size of 12 m2, as in the base case, two modern active cooling beams with the length of 1,8 m, will suffice to provide the design cooling capacity. The variation is set to ± 10 W/m2 to cover realistic variations of design cooling capacities in commercial office buildings.

14. Design ventilation rate The HVAC-system in this project is an air-water system. This type of system uses the ventilation primarily to supply the room with fresh air and extract contaminated air. The cooling or heating effect from the ventilation is always secondary. In the base case the specific ventilation rate 1,25 l/s m2 is used. This equals a ventilation rate of 15 l/s in an office room of 12 m2 with one person present. 15 l/s is a normally used figure which ensures, with sufficient marginal, that the CO2 level never exceeds 1000 ppm. Ventilation rates below 10 l/s in a one-person office room are very rare. Based on this, a parameter variation of ± 5 l/s seems reasonable.

92


15. Supply air temperature set point A supply air temperature set point near 17°C is fairly normal in air-water HVAC systems with mixing ventilation. The ventilation is not primarily used for cooling purposes, but it is quite normal to cool the air to gain some additional cooling from the ventilation, especially during warmer periods of the year. A supply temperature below 15°C may involve increased problems with thermal discomfort due to cold draft. A parameter variation of ± 2°C is hence chosen.

16. Design secondary approach The design secondary approach is a temperature difference between the primary circuit cold fluid entering temperature, Tliq1_cold, and the secondary circuit chilled fluid leaving temperature, Tliq2_out, see figure 5.16.

Primary circuit Tliq1_cold

Secondary circuit Tliq2_out

Secondary Approach = Tliq2_out - Tliq1_cold Figure 5.16 Definition of secondary approach The secondary approach, together with the design cooling capacity, defines the size, i.e. the UA-value, of the heat exchanger. In this case, a plate type heat exchanger is used. A plate type heat exchanger with a secondary approach of 1,5°C has a good performance but is not extremely costly (de Saulles, 1996). The “low” parameter alteration, 0°C, is the same as no heat exchanger between the two fluids. This is an option that, in some cases, can be feasible when the operation of a cooling tower in wet mode during freezing conditions can be handled with other solutions. The “high” parameter alteration is consequently 3°C.

17. Temperature gradient in office room A temperature gradient is almost always generated in rooms occupied by people and with other internal heat sources. The size of the temperature stratification is determined not only by internal heat sources but also by e.g. type of ventilation, room height, solar radiation and type of heating and cooling devices. The reason why the temperature gradient can influence the resulting indoor air temperature may not be obvious. The cooling system with chilled ceilings in a room consists of the beam/beams, a temperature sensor and a control device, the supply and return pipes with control valve and actuator. The sensor is normally placed on an inner

93

wall at a height of about 1,4 m above floor. The chilled beams are placed close to the ceiling, or the false ceiling whenever applicable, i.e. 2,4 – 2,5 m above the floor. The sensor, together with the control device, will strive for maintaining a temperature equal to the set point temperature. In the office room model the control device is a PIcontroller, thus there will be no temperature offset. If there is a temperature stratification in the room, the chilled beams will be in a position with higher temperature than the temperature sensor. As mentioned above and at parameter #12. Design temperature difference Troom − Twater , the design temperature difference ( Troom − Twater ) determines the available cooling capacity of a chilled beam. Troom is in this case the temperature in the close vicinity of the chilled beams. In the room model in IDA ICE the temperature sensor and the chilled beams are placed on different height in the room, hence the available cooling capacity of the chilled beams will vary depending on the magnitude of the temperature gradient but with the same set point temperature. The temperature gradient can in IDA ICE be either calculated or given. In this case, it is a given parameter and it is assumed linear from floor to ceiling. Novocelac and Srebric (2002) report a list of measured or simulated temperature gradients from six different references. The gradients were in the range 0-2 °C/m. It should be mentioned that these gradient were in cases with displacement ventilation and in this case mixing ventilation is assumed. The figures in table 5.1 represent the total temperature difference between floor and ceiling. The base case value of 1°C is hence equivalent to a gradient of 0,4°C/m. The decided parameter variations of 0 and 2°C are equivalent to a gradient of 0 and 0,8°C/m.

18. Set point cooling control The air temperature in an office room has a typical diurnal variation with lower temperature during night time and higher during daytime. When the temperature rises in the morning, it will at times with high internal heat gain reach the cooling set point temperature. The cooling devices in the room will then start to cool the room acting as a “brake” on the rising temperature, thus hindering the temperature rise. If the set point is at a lower value, the cooling will start earlier and rise of the temperature will be less. The temperature will then level in the afternoon on a lower value than with a higher set point temperature. The person working in an office room normally sets the set point temperature on a local control device. In reality the set point temperature may vary from room to room depending on the thermal climate preferences, the awareness and knowledge of the heating and cooling system of each person together with the function of the control device and cooling device in each room. This situation is almost impossible to emulate in building simulation. The decided values of this parameter should therefore be considered as estimated mean set point temperatures The base case set point temperature is set at 22°C. The “low” value is seldom lower than 21°C and hence the “high” value is consequently set to 23°C.

94

6

Monitoring of free cooling system with evaporative cooling device In this chapter a free cooling pilot plant is described. The pilot plant comprises an evaporative cooling device connected to an existing hydronic cooling system with chilled beams in a real office building. The pilot plant has been monitored during the summer period of 2007.

6.1 Introduction To test the theoretical results from the simulations, a project was initiated to build and monitor a free cooling system with a evaporative cooling connected to an existing hydronic cooling system with chilled beams in a real office building. The free cooling system was sized as a pilot plant and thus only serving a smaller part of the office building to fit the budget of the project. The pilot plant together with monitoring and evaluation was financed by BELOK, Vasakronan and CIT Energy Management. The pilot plant was designed and built during 2006 and monitoring was conducted during the summer period of 2007. The pilot plant is located adjacent to a commercial office building at Kvarnberget at the city centre of Gothenburg, Sweden. The design work was performed by an HVAC-consultant (ÅF-konsult) in cooperation with a representative of the building owner (Vasakronan) and the project manager (the author of this thesis).

6.2 Pilot plant system lay-out A sketch of the free cooling system connected to an existing hydronic comfort cooling system with chilled beams is outlined in figure 6.1. The connected system is a branch of the entire comfort cooling system of the building. The branch is serving part of the first floor of the building and the connection to the existing chiller is maintained through pipefitting with motor-operated shut-off valves. The shut-off valves are normally closed and the connection to the chiller is only opened if the indoor temperature rises above a predefined level or if the free cooling system is malfunctioning. The supply air is chilled by an existing air cooling coil in the central air conditioning unit connected to the existing chiller. The temperature in the existing supply air system is held constant. The supplemental heater in the supply air to the office space which is connected to the free cooling system has a special purpose. The heater is fitted to the existing duct system and its purpose is to emulate the somewhat poorer cooling capacity an air cooling coil would have connected to the free cooling system. The heater is controlled so that the supply air temperature is held at a temperature similar to what an evaporative free cooling system would have had at a given ambient air condition.

95

Conventional ventilation system with air conditioning connected to a chiller. (only supply part of system is shown)

Connection to conventional cooling system with a chiller. Shut-off valves is normally closed Freeze-protected circuit (primary)

Supplementary heater to emulate a slightly poorer cooling than with conventional air conditioning

Room cooling devices (chilled beams) Free cooling device

Chilled water circuit (secondary) Plate heat exchanger

Figure 6.1 Simplified sketch of the pilot plant at Kvarnberget, Gothenburg. Thin lines indicate existing systems. The evaporative cooling device in the system is not a cooling tower in a traditional meaning, but an air cooled heat exchanger equipped with a spray-water device which gives the heat exchanger a function similar to an evaporative cooling tower, see figure 6.2. The heat exchanger is wetted by fresh water, i.e. once through spray water system, and thus there is no risk for growth of legionella bacteria in the spray water. In order to avoid misinterpretation the evaporative cooler in the pilot plant is hereafter named free cooler. The reason for choosing this kind of evaporative cooler is partly cost related and partly market supply related. The investment cost for an air cooled heat exchanger is significantly lower than for a closed loop cooling tower with the same cooling capacity. The market supply of closed loop cooling towers with the cooling capacity required in this case (< 20 kW) is very limited. The dominating market is power plants and industry with typical requirements of cooling capacities from a few hundred kW up to hundreds of MW. Air cooled heat exchangers, without additional water spray device, are occasionally used as free cooling sources in conventional comfort cooling systems. For this reason it is also interesting to analyze an established product on the market, i.e. an air cooled heat

96

6 Monitoring of free cooling system with evaporative cooling device

exchanger with an evaporative utility, to investigate its potential as a cooling device in a hydronic comfort cooling system. Detailed data about the pilot plant is presented in appendix B.

Figure 6.2 The free cooler in the pilot plant. A traditional air cooled heat exchanger equipped with a spray water devise for wetting of the heat exchanger The free cooler in figure 6.2 chills a primary circuit, containing a propylene glycol solution (40%) as freeze protection. This facilitates operation in ambient temperatures below 0°C. The primary circuit transfer heat from a secondary circuit through a plate heat exchanger. The plate heat exchanger has a design approach/grädigkeit of 1°C and a design temperature difference between supply and return, a.k.a. range, of 3°C in the primary circuit as well as in the secondary ditto. In the secondary circuit there are chilled beams which have a design ∆Tm of 8°C and a design specific cooling capacity of 50 W/m2 floor area. The circulation in the primary and secondary circuits is maintained by pumps equipped with variable frequency drives (VFD). The control system strives to maintain a constant supply temperature in the secondary circuit. The constant supply temperature is maintained through sequential control where, in order and at increasing cooling demand, the pump in the primary circuit increase its speed, the fans in the free cooler increase their speed and finally the spray water is turned on. The control system also includes a function which disconnects the free

97

cooling system and connects the conventional comfort cooling system, through a set of shut-off valves, if the indoor temperature exceeds 25°C. This function was applied through demand from both the building owner Vasakronan and the tenant ÅF-konsult. The part of the building at which the pilot plant is installed is erected during the 1970’s with corresponding building standard. The office space which the free cooling system is serving has a floor area of 450 m2 with an outer wall facing south east. The people density is nominally 14 m2/person and at normal day-time operation the density is about 16 – 20 m2/person. Design power density for lighting is 15 W/m2. There are no external sun shading devices.

6.1 Monitoring system Monitoring of the pilot plant was conducted during the summer period of 2007, i.e. may – august 2007. Each measurement point has been sampled every 10 minutes and the samples have been processed into one hour average figures. The following points have been monitored: Outside - Air temperature (one point) - Relative humidity (one point) - Solar radiation (one point, measurement of both diffuse and total radiation) Inside - Air temperature (three points distributed in the office area) - Relative humidity (one point) - Air temperature gradient (three points at different heights above floor level) - Electric energy in office space (one point in the distribution box) - Liquid temperature in the secondary circuit (two points, supply and return) - Liquid flow rate in the secondary circuit (one point) Free cooling system - Liquid temperature in the secondary circuit (two points, supply and return) - Liquid temperature in the primary circuit (two points, supply and return) - Liquid flow rate in the secondary circuit (one point) - Liquid flow rate in the primary circuit (one point) - Liquid flow rate in the spray water supply (one point) - Cooling capacity in the secondary circuit (one point) - Cooling capacity in the primary circuit (one point) - Electric energy in pumps (two points, primary and secondary circuits) - Electric energy in fans (one point, fans in free cooler) Further information about the monitoring system can be read in appendix C.

98

7 Results and analysis of simulations The results from the simulations described in chapter 5 “Simulation of a cooling system with an evaporative cooling tower” are presented and analyzed in this chapter. The results are separated into sections including Cooling tower performance, Indoor thermal climate and Energy use. The chapter also contains relevant findings from other researchers.

The results in section 7.1 Cooling tower performance are solely based on the prerequisites according to the base case conditions, see chapter 5 Simulation of a cooling system with evaporative cooling. In section 7.2 Indoor thermal climate, the results are based on prerequisites in line with the base case conditions as well as all the parameter variations. In section 7.3 Energy use, the results are based on the same presumptions as in section 7.2, but with a limited number of parameter variations. The results presented in diagrams based on a full year simulation have a resolution of 15 minutes, i.e. the presented variable is logged every 15 minutes during a year. Some diagrams show one or several variables during three days in August; Monday – Wednesday the 18th – 20th of August. This period in the TMY year of London/Gatwick, which is the base case climate/location, is the one with the highest wet bulb temperature together with one of the warmest periods of the summer, i.e. the outdoor climate is at, or close at, ambient design conditions, see figure 7.1. The presented indoor temperatures during day time at this period are therefore close to, or at, their maximum. One exception is however the approach temperatures and the COP which are at their minimum during this period. Monday

Temperature (°C)

30 5496

Wednesday

Tuesday

5520

5544

5568

25

20

15

10 Time Wet bulb

Figure 7.1

Dry bulb

Ambient conditions during August 18 – August 20, showing dry bulb and wet bulb temperature.

99

7.1 Cooling tower performance The performance of the cooling tower in this section is focused on how close to the ambient wet bulb temperature (WBT) the supply water temperature in the secondary circuit can be brought, i.e. the size of the total approach. The other type of performance discussed in this chapter, namely the energy use, is presented in section 7.3 Energy use. As mentioned in chapter 5 Simulation of a cooling system with evaporative cooling tower, the primary approach is the difference between the chilled primary circuit temperature and the ambient wet bulb temperature. The total approach is the difference between the supply temperatures in the secondary circuit and the ambient wet bulb temperature, see figure 7.2. Temperature Return side Warm side

Wet bulb temp. out

Secondary range Liquid in secondary circuit (water) Supply

Primary range

side


Total approach


Figure 7.2

Temperature diagram of a counter flow cooling tower with intermediate circuit and visualisation of definition of design parameters.

The size of the total approach temperatures is one of the most important factors for the ability to provide sufficient cooling to a building. The total approach is of course only important during the warmer part of the year, when high cooling loads and high wet bulb temperatures occur simultaneously. At cooler weather conditions, the cooling tower will normally operate at part load. One important question to answer is how well the cooling tower model can calculate the two approach temperatures during the influence of varying climate conditions and variations of mass flows and temperatures in the cooling system due to varying cooling loads and temperatures in the building. This is discussed in section 4.3 Validation of the model. In this section, the question is further examined by running full year simulations at base case conditions.

100

7 Results and analysis of simulation

Variations of the ambient wet bulb temperature, the chilled primary circuit temperature and the leaving temperature in the secondary circuit during three days in August are shown in figure 7.3. The figure visualizes how these temperatures fluctuate during this period. The ambient WBT varies from between 12 – 14 °C during night time up to peaks between 18 – 20 °C during daytime. The amplitude of the WBT is approximately 6 °C all three days, which is less than the amplitude for the ambient dry bulb temperature. The maximum dry bulb temperature was 28,8°C and the amplitude for the ambient dry bulb temperature was between 13-14 °C during Monday and Tuesday and around 9 °C on Wednesday, see figure 7.1. The chilled primary circuit temperature is close to equal with the ambient WBT during night time. Due to absence of cooling load during this time of night, the mass flows in both secondary and primary circuits are close to zero. The mass flows in both circuits are never equal to zero in the model of numerical reasons. The chilled primary circuit temperature is therefore very close to the ambient WBT at night time. During daytime the chilled primary circuit temperature is between 1 – 2°C above the WBT. These numbers are the primary approach temperature, which is shown in figure 7.4. The supply water temperature in the secondary circuit is not nearing the ambient WBT during night time as the chilled primary circuit temperature does. Both the primary and secondary circuit mass flows should go down to zero during the night, since there is no cooling load in the rooms, and consequently the temperature differences should reach zero, but they do not. This is mainly caused by different minimum limits in the cooling tower model, due to numerical reasons, for the mass flows in the two circuits. The supply water temperature in the secondary circuit is fluctuating between 1 – 3 °C above the WBT during this three-day period. These numbers are the total approach temperature, which is shown in figure 7.4. The fluctuations are mainly dependent on variations in mass flow in the secondary circuit as a result of changes in the indoor temperature over the day due to the changes in cooling load. Interesting to notice is that during Tuesday, August 19th, the leaving temperature in the secondary circuit, i.e. the supply temperature to the cooling beams, is between 20 – 21°C during working hours and the indoor temperature still only peaks at slightly above 25°C (figure 7.9, section 7.2 Indoor thermal climate). Bearing in mind that the design cooling heat gain is 60 W/m2 and the design cooling capacity of the chilled beams is 50 W/m2, it is indeed remarkable that a cooling system based only on a cooling tower as a “chiller” gives a resulting indoor temperature of about 25°C. It should be noted that the design cooling capacity of the chilled beams at base case conditions is defined at a design temperature difference of 7,5°C between mean cooling water temperature and the room air temperature, ∆Tair-liq. In this case, ∆Tair-liq is merely in the range 2 -3°C!

101

Monday

Wednesday

Tuesday

22

Temperature (°C)

20 18 16 14 12 10 00:00 06:00 12:00 18:00 00:00 06:00 12:00 18:00 00:00 06:00 12:00 18:00 00:00 Time (hh:mm) Ambient WBT

Figure 7.3

Prim. circuit, cold side

Sec. circuit, supply

Ambient WBT, primary circuit cold side temperature and secondary circuit supply temperature between August 18 – August 20. Base case conditions.

Monday

Wednesday

Tuesday

4,0

Temperature (°C)

3,5 3,0 2,5 2,0 1,5 1,0 0,5 0,0 00:00 06:00 12:00 18:00 00:00 06:00 12:00 18:00 00:00 06:00 12:00 18:00 Time (hh:mm) Primary approach

Figure 7.4

102

Total approach

Primary and total approach temperatures between August18 – August 20. Base case conditions.


In figure 7.5 the primary and total approach temperatures during a full year simulation is shown plotted against the ambient WBT. A few values below -5 °C WBT are omitted in the diagram. The values of the primary and total approach temperatures are very scattered but conglomerations of data form a linear, or almost linear, relationship with the WBT. The two approach temperatures are decreasing with increasing WBT. In figure 7.6 the primary approach temperature values and relation with the WBT is zoomed in between 5 – 20 °C WBT. In figure 7.7, the same thing is made with the total approach temperature. The primary approach temperature dependency with the WBT at higher WBT is less steep as with lower values of WBT. The change in inclination takes place at an ambient wet bulb temperature of about 13°C. This is due to that the set point temperature for the chilled water supply temperature in the secondary circuit is 16°C at base case. This set point temperature is achievable with a maximum ambient wet bulb temperature of 13°C together with a total approach of about 3°C. With lower ambient wet bulb temperatures than 13°C, the cooling tower starts to decrease the fan speed and thus the cooling capacity. Most of the primary approach values are in the range of 1 – 2 °C when ambient wet bulb temperatures are higher than 13°C. In figure 7.7, the total approach temperature shows the same dependency with the WBT as in figure 7.6. Most of these values are in the range of 1,5 – 3 °C when ambient wet bulb temperatures are higher than 13°C.

20

Approach temperature (°C)

18 16 14 12 10 8 6 4 2 0 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Ambient wet bulb temperature (°C) Prim approach temp.

Figure 7.5

Sec. approach temp.

Primary and total approach temperatures in a full year simulation plotted against the ambient wet bulb temperature. Base case conditions.

103

10

Primary approach (°C)

9 8 7 6 5 4 3 2 1 0 5

6

7

8

9

10 11 12 13 14 15 16 17 18 19 20

Ambient wet bulb temperature (°C) This thesis

Figure 7.6

Costelloe & Finn (2003)

Hasan & Gan (2002)

Primary approach temperatures in a full year simulation plotted against the ambient wet bulb temperature. Base case conditions.

10 9 Total approach (°C)

8 7 6 5 4 3 2 1 0 5

6

7

8

9

10 11

12 13

14 15

16 17

18 19

20

Ambient wet bulb temperature (°C) This thesis

Figure 7.7

104

Costelloe & Finn (2003)

Total approach temperatures in a full year simulation plotted against the ambient wet bulb temperature. Base case conditions.


The presented temperatures in figures 7.5 to 7.7 apply well with published research of Costelloe and Finn (2002), Hasan and Gan (2002) and Facao and Oliveira (2000). Costelloe and Finn (2002) have made measurements on a prototype open type cooling tower test rig with a primary and secondary circuit (described in chapter 1.3 Previous work). They found that the primary approach temperatures was in the range of 0,9 – 2,3°C and total approach temperatures in the range of 2,2 – 4,3 °C, with ambient WBT between 6,4 – 16,5 °C. Their findings are plotted in figure 7.6 and 7.7. Hasan and Gan (2002) have made measurements on a closed wet cooling tower for the generation of cooling water for use with chilled ceilings. Their findings are plotted in figure 7.6. The approach temperature is calculated by the author from published data by the Hasan and Gan (2002). Facao and Oliveira (2000) have also made measurements on a closed wet cooling tower for the generation of cooling water for use with chilled ceilings. They reported an approximate 8% increase in cooling tower thermal efficiency, ε, when the ambient WBT rises from 10°C to 20°C (increased thermal efficiency, ε, leads to lower approach temperatures). The increase is linear and the influence is similar at different air and water mass flow rates. Together with the validation of the cooling tower model in chapter 4.3 Validation of the model against published data on cooling towers and the results from the full year simulations in figure 7.5 to 7.7, it seems that the cooling tower model can calculate accurate values of primary and total approach temperatures. This is an important result since the total approach gives the available cooling water supply temperature. As shown in chapter 5.4.2 Discussion and motivation of parameter alterations, the cooling water temperature is of great importance for the available cooling capacity of the chilled beams and hence, the resulting indoor temperature. In the next chapter, the indoor thermal climate with this cooling system applied to an office building will be discussed.

7.2 Indoor thermal climate In this section, results concerning the indoor thermal climate are presented. The air dry bulb temperature as well as the operative temperature is studied. The air temperature is presented as a mean value of all temperatures in the room since temperature stratification in the analysed room is set as a constant parameter value in all simulation cases (between 0 - 2°C from floor to ceiling). The operative temperature is calculated as for a seated person at 0,6 m above floor level in the middle of the analysed room. When calculating the operative temperature the air temperature and mean radiant temperature at 0,6 m above floor level is used. However, more focus will be put on the air temperature in the presentations of the results since most indoor thermal climate requirements in building projects is formulated with the air temperature in mind.

105

The relative humidity is also of interest since all cooling, of supply air as well as in the building, is sensible. No dehumidification occurs, which means that the relative humidity increases when the air temperature decreases, everything else constant. Hence, there might be concerns for unpleasantly high relative humidity at times. All results in this chapter are from the analysed room. For an explanation of the analysed room and its prerequisites, see chapter 5.2.1 Model building. The presentation of the results starts with indoor conditions at base case, followed by results of a parameter variation analysis.

7.3.1 Base case The presentation of the results of the thermal indoor climate for the base case starts with a period containing the two last days of a five day heat wave followed by three days in August, as described earlier, from the full year simulation.

30

100%

29

90%

28

80%

27

70%

26

60%

25

50%

24

40%

23

30%

22

20%

21

10%

20 00:00

06:00 12:00

18:00

00:00 06:00

12:00 18:00

Relative humidity (%)

Temperature (°C)

Variations of the indoor air temperature, the operative temperature and the indoor relative humidity during the last two days in a five-day heat wave simulation are shown in figure 7.8, and during three days in August in a full year simulation are shown in figure 7.9.

0% 00:00

Time (hh:mm) Operative temperature

Figure 7.8

106

Air temperature

Relative humidity

Indoor air temperature, operative temperature and indoor relative humidity during the last two days in a five day heat wave simulation.


Tuesday

Wednesday

30

100%

29

90%

28

80%

27

70%

26

60%

25

50%

24

40%

23

30%

22

20%

21

10%

Relative humidity (%)

Temperature (°C)

Monday

20 0% 00:00 06:00 12:00 18:00 00:00 06:00 12:00 18:00 00:00 06:00 12:00 18:00 00:00 Time (hh:mm) Operative temp

Figure 7.9

Mean air temp

Relative humidity

Indoor air temperature, operative temperature and indoor relative humidity between August 18 – August 20, full year simulation.

In figure 7.8, the indoor air temperature range is about 21,5 – 25.5 °C. The operative temperature is slightly higher except during morning hours when the two temperatures are equal. The indoor relative humidity (RH) is in the range of 65 – 75 %. Interesting to notice is a slight decrease in the RH during daytime although one person is present and producing moisture. The decrease in RH can be explained with the increase in air temperature during working hours, which compensates for the moisture emitted from the person present. The ventilation rate is constant throughout the day. Figure 7.9 show a similar temperature pattern as figure 7.8. The indoor air temperature range is however slightly smaller, around 21 – 25 °C. The difference between the air temperature and the operative temperature is about the same as in figure 7.8. The largest difference is in the afternoon and is mainly explained by increasing surface temperatures in the room, mostly due to solar radiation. The solar radiation is quite high during Monday and Tuesday but lower on Wednesday. Consequently, the air temperature is lower on Wednesday as well as the difference between the air temperature and the operative temperature. The indoor relative humidity in figure 7.9 varies a bit more and is in the range 55 – 75 %. Interesting to notice is that the same decrease, after a small increase, in the relative humidity during daytime as in figure 7.8 can be found here as well. Figure 7.10 and 7.11 show a comparison between indoor air temperature with cooling and without cooling. The latter is when the cooling tower is shut down. Without cooling, the indoor temperature would reach levels which most people find unpleasant. In the morning around 8 am, the temperature is 23 - 24 °C and quickly rises during the day to reach 28 – 29,5 °C in the afternoon. With the cooling system operating, it

107

manages to lower the indoor temperature by about 3,5 – 4°C during the afternoon, which is a significant reduction. It is significant in the sense that it makes a big difference in the thermal sensation of a person being exposed to temperatures around 24 - 25 °C compared to 28 – 29,5 °C. According to ISO 7730 (1994) the PPD index for a person is 8% at 24 °C, 13% at 25 °C and 55% at 29 °C, at 0,9 Clo and 1,2 Met as in the base case during the summer. A more detailed discussion about thermal sensation and indoor thermal climate can be found in ASHRAE (2001a) or Nilsson (ed) (2003). 30 29 Temperature (°C)

28 27 26 25 24 23 22 21 20 00:00

06:00

12:00

18:00

00:00

06:00

12:00

18:00

00:00

Time (hh:mm) Air temp - no cooling

Air temp - with cooling

Figure 7.10 Indoor air temperatures with and without cooling during the last two days in a five day heat wave simulation. Monday

Tuesday

Wednesday

30 29 Temperature (°C)

28 27 26 25 24 23 22 21 20 00:00 06:00 12:00 18:00 00:00 06:00 12:00 18:00 00:00 06:00 12:00 18:00 00:00 Time (hh:mm) Air temp - no cooling

Air temp - with cooling

Figure 7.11 Indoor air temperatures with and without cooling between August 18 – August 20, full year simulation.

108


In figure 7.12, the indoor temperatures during working hours throughout a year, with and without cooling, are shown in the form of duration curves. With cooling, the system manages to hold the indoor temperature below e.g. 24 °C during 97,9% of working hours in a year. The same figure without cooling is 60%. See also table 7.2.

30


29 28 27 26 25 24 23 22 21 20 0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Working hours during one year (%) With cooling

Without cooling

Figure 7.12 Duration of indoor air temperature with and without cooling during one year of working hours (08-17), full year simulation. This figure shows that the cooling system can keep the indoor temperature relatively constant at 22 °C at working hours during most of the year. It should be noted that the data behind the curves are not coincident in time, i.e. the temperatures on the two curves, for example at 90%, do not necessarily occur at the same day and hour of the year. The other indoor variable investigated is the indoor relative humidity. Figure 7.13 shows the indoor relative humidity plotted against indoor air temperature during one year of working hours (08-17) at a full year simulation. The data is for the case when cooling is activated. Interesting to observe is that the maximum relative humidity, 79%, occurs at indoor temperatures around 23°C. At temperatures above 23,5°C the relative humidity decreases and is about 60% at the highest indoor temperatures. The same tendency can be seen at the other two climates, which have been investigated, Stockholm and Berlin. Table 7.1 shows the maximum indoor relative humidity with coincident temperature and maximum indoor temperature with coincident relative humidity for the three climates.

109

Indoor relative humidity (%)

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 20

21

22 23 24 Indoor air temperature (°C)

25

26

Figure 7.13 Indoor relative humidity plotted against indoor air temperature with cooling during one year of working hours (08-17), full year simulation. Table 7.1

Maximum indoor relative humidity with coincident indoor temperature and maximum indoor temperature with coincident indoor relative humidity. Full year simulation.

Climate

Maximum Coincident indoor relative indoor humidity (%) temperature (°C)

Maximum indoor temperature (°C)

Coincident indoor relative humidity (%)

Stockholm/Arlanda

85

23,4

24,8

59

London/Gatwick

80

22,7

25,3

61

Berlin

78

21,4

26,1

70

In table 7.2 different durations concerning indoor temperature and relative humidity are listed. The values are taken from the figures 7.12 and 7.13. The durations are expressed as percentage of working hours (08 – 17) when a limit value of either the indoor air temperature or the indoor relative humidity is exceeded.

110


Table 7.2 Percentage of working hours (8 – 17) which a limit value is exceeded of either the indoor air temperature or the indoor relative humidity. Full year simulation.

Climate

Percentage of working hours the indoor air temperature exceeds 24°C

Percentage of working hours the indoor air temperature exceeds 25°C

Percentage of working hours the indoor air relative humidity exceeds 70%

Percentage of working hours the indoor air relative humidity exceeds 80%

Stockholm/Arlanda

1,4%

0,0%

2,7%

0,8%

London/Gatwick

2,1%

0,3%

3,8%

0,0%

Berlin

5,6%

2,1%

4,1%

0,0%

7.3.2 Single parameter variations – indoor temperature The indoor air temperature in the analysed room is a result of the performance of many parts in the cooling system, the building and the activity in it. The performance of the different parts is influenced by many parameter values. An obvious question in this context is; how “much” is each parameter contributing to the resulting indoor air temperature. To answer the question, parameter variations are made to examine the potential of each parameter regarding its influence on the indoor air temperature. The variations for each parameter are made with one higher and one lower value. To make the variations consistent the variations are called “lower indoor temperature” and “higher indoor temperature”. This is made since lowering the actual value of some parameters raises the indoor temperature and vice versa. All variations that cause a drop in the indoor temperature are gathered in the column labelled “lower indoor temperature” and all variations that cause a raise in the indoor temperature are in the column “higher indoor temperature”. A total of 18 parameters have been altered. A list of all parameter alterations can be seen in table 7.3. A more detailed discussion of the parameters can be found in chapter 5 Simulation of a cooling system with evaporative cooling tower.

111

Table 7.3

List of all parameter alterations

Parameter

Lower indoor temp.

Base case

Higher indoor temp.

Cooling Tower 1. Design primary approach

3°C

4°C

5°C

2.

Design primary range (both primary and secondary)

2°C

3°C

4°C

3.

Design mass flow quota of liquid and air, M& ( l ) M& a

2,0

1,0

0,5

4.

Design cooling capacity, Q& CT

1,25 ⋅ Q&CT

Q& CT

0,75 ⋅ Q& CT

Building 5. Thermal capacity (weight)

”heavy”

”medium”

”lighter”

6.

Location in building

Ground floor

Middle floor

Top floor

7.

U-value in external walls U-value in windows

0,2 W/m2 K 2,0 W/m2 K

0,4 W/m2 K 0,6 W/m2 K 2,5 W/m2 K 3,0W/m2 K

8.

Internal heat gain; magnitude (maximum )

50 W/m2

60 W/m2

70 W/m2

9.

Internal heat gain; working hours

8 h (08-16)

9 h (08-17)

10 h (0818)

10. Climate (Geographical location)

Stockholm/ Arlanda

London/ Gatwick

Berlin

11. Facade direction

East

South

West

HVAC-system 12. Design temp. difference Beams: Troom − Twater

5°C

7,5°C

10°C

13. Design cooling capacity cooling beams

60 W/m2

50 W/m2

40 W/m2

14. Design ventilation rate

1,67 l/s m2

1,25 l/s m2

0,83 l/s m2

15. Supply air temperature set point

15°C

17°C

19°C

16. Design secondary approach

0°C

1,5°C

3°C

17. Temperature gradient in office room

2°C

1°C

0°C

18. Set point temperature cooling control

21°C

22°C

23°C

Figure 7.15 shows results from parameter variations with heat wave simulations and figure 7.16 shows results from parameter variations with full year simulations. Both figure 7.15 and 7.16 show the differences in maximum indoor temperature compared to the results from the base case. For example, when varying the parameter climate to a value equal to “lower indoor temperature” in table 7.3, using heat wave simulations, fig 7.15, the maximum indoor air temperature is 0,76°C lower than for the base case. For heat wave simulations, figure 7.15, no single parameter can affect the maximum indoor temperature more than 0,76°C. 12 parameters out of 18 could not change the maximum indoor temperature more than 0,4°C. Interesting to notice is that all cooling tower parameters are in the range of ±0,2°C. The two most powerful parameters are climate and internal heat gain. Facade direction - “higher temperature” and location in

112


building - “higher temperature” as well as thermal capacity – “lower temperature” and design temperature difference cooling beams – “lower temperature” parameter variation also generates a relatively big change. In fig 7.16, full year simulation, the pattern is almost the same as with figure 7.14. The parameters working hours and ventilation rate give a bit more change with the full year simulation than with the heat wave simulation. For the rest of the parameters the differences are rather small compared to figure 7.15. Not surprisingly, the climate parameter is powerful. A warmer climate raises the cooling load at the same time as the cooling capacity of the cooling tower is reduced. The magnitude of the indoor temperature variation when varying climate locations from “higher temperature” to “lower temperature” is purely a result from the choice of climate range. If the variations in climate had been from Kiruna in the north of Sweden to Palermo in Sicily, Italy, the variation in indoor air temperature would naturally have been larger. The variation in internal heat gain is ±10 W/m2. This range, from 50 W/m2 to 70 W/m2 (sensible heat), probably gather a large part of existing maximum cooling loads in office buildings. Since the internal heat gain directly affects the indoor air temperature, it is expected that the variation in indoor air temperature in figure 7.15 and 7.16 is relatively high. The facade direction west gives high temperatures in the afternoon and the location on the top floor leads to a higher transmission of heat through the ceiling, which adds to the internal heat gain and raises the temperature relatively high. High thermal capacity is a well-known factor for reducing indoor temperatures. The parameter variation “lower temperature” has a bigger influence than the “higher temperature” one. This parameter variation is not a result of an equal change up or down of the value of thermal capacity. The variation is more a result of the different building constructions, especially concerning the fabrics of the inner layers on the floor or ceiling. The parameter “lower temperature” implies large areas of exposed concrete slab or wall, which gives this alternative a stronger impact than the parameter alternative “higher temperature” does. A detailed description of the different building constructions for the different parameter variations is found in chapter 5.4.1 List of parameter alterations. A smaller design temperature difference for the cooling beams directly influences the size of the cooling beam, i.e. they become bigger. A consequence of a smaller design temperature difference is that the cooling capacity rises faster when the indoor air temperature increases. This leads to lower amplitudes in the daily indoor temperature swing. A more detailed discussion about this is presented in chapter 5.4.2 Discussion and motivation of parameter alterations.

113

Prim Approach Prim Range Flow quota Cooling capacity tower Thermal capacity Location in building U-value Internal heat gain Working hours Climate Facade direction Design temp. diff. Beams Cooling capacity Beams Ventilation rate Supply air temp. set point

-0,76°C

Secondary Approach Temperature gradient Set point Cooling -1,0 -0,8 -0,6 -0,4 -0,2 0,0 0,2 0,4 0,6 0,8 1,0

temperature difference (°C) Lower temperature

Higher temperatue

Figure 7.15 Differences in maximum indoor air temperature, compared with the results of the base case, for parameter alterations in table 7.3. Heat wave simulations.

114


Prim Approach Prim Range Flow quota Cooling capacity tower Thermal capacity Location in building U-value Internal heat gain Working hours Climate Facade direction Design temp. diff. beams Cooling capacity Beams Ventilation rate Supply Air Temperature Secondary Approach Temperature gradient Set point Cooling -1,0 -0,8 -0,6 -0,4 -0,2 0,0 0,2 0,4 0,6 0,8 1,0 Temperature difference (°C) Higher temperature

Lower temperatue

Figure 7.16 Differences in maximum indoor air temperature, compared with the results of the base case, for parameter alterations in table 7.3. Full year simulations

7.3.3 Multi parameter variations Since it is not practically possible to make a full multi factorial analysis with 18 parameters (this would require 262 144 simulation runs) a limited parameter variation is made with three of the most dominant parameters (see figure 7.16); climate/location, thermal capacity and design internal heat gain (including solar radiation). In this case, there are five locations; Östersund and Stockholm (Sweden), London (UK), Berlin (Germany) and Paris (France), two thermal capacities; medium and heavy, and three design heat gains; 50, 60 and 70 W/m2, in the parameter variations. The design heat gain includes maximum solar radiation in the room. Combining all possible parameter combinations involves 30 full year simulation runs. The result is shown in figure 7.17 – 7.20

115

Maximum indoor air temperature (°C)

28

R2 = 0,7875

Medium thermal capacity

27

R2 = 0,8763

26 25 R2 = 0,8332

24 23 22 15

16 17 18 19 20 21 Design wet bulb temperature, ASHRAE 0,4%, (°C)

22

50 W/m2

60 W/m2

70 W/m2

Linjär (50 W/m2)

Linjär (60 W/m2)

Linjär (70 W/m2)

Percent of working hours room temperature exceeds 24 °C

Figure 7.17 Maximum indoor air temperature, at five different design wet bulb temperatures, i.e. climate/location and for three different internal total heat gains. Medium weight thermal capacity. Full year simulations.

10% 9%

Medium thermal capacity

8% 7%

R2 = 0,8369

6%

R2 = 0,7945

5% 4% 3%

R2 = 0,7131

2% 1% 0% 15

16

17

18

19

20

21

22

Design wet bulb temperature, (ASHRAE 0,4%) (°C) 50 W/m2

60 W/m2

70 W/m2

Linjär (50 W/m2)

Linjär (60 W/m2)

Linjär (70 W/m2)

Figure 7.18 Percentage of working hours when indoor temperature exceeds 24°C, at five different design wet bulb temperatures, i.e. climate/location and for three different internal heat gains. Medium weight thermal capacity. Full year simulations.

116


Maximum indoor air temperature (°C)

28 R2 = 0,8637

Heavy thermal capacity

27

R2 = 0,7973

26 25 24

R2 = 0,8566

23 22 15

16 17 18 19 20 21 Design wet bulb temperature, ASHRAE 0,4% (°C)

22

50 W/m2

60 W/m2

70 W/m2

Linjär (50 W/m2)

Linjär (60 W/m2)

Linjär (70 W/m2)

Percent of working hours room temperature exceed 24 °C

Figure 7.19 Maximum indoor air temperature, at five different design wet bulb temperatures, i.e. climate/location and for three different internal heat gains. Heavy weight thermal capacity. Full year simulations.

10% 9% 8%

Heavy thermal capacity

7%

R2 = 0,7816

6% 5%

R2 = 0,706

4% 3% 2%

R2 = 0,6049

1% 0% 15

16

17

18

19

20

21

22

Design wet bulb temperature, (ASHRAE 0,4%) (°C) 50 W/m2

60 W/m2

70 W/m2

Linjär (50 W/m2)

Linjär (60 W/m2)

Linjär (70 W/m2)

Figure 7.20 Percentage of working hours when indoor temperature exceeds 24°C, at five different design wet bulb temperatures, i.e. climate/location and for three different internal heat gains. Heavy weight thermal capacity. Full year simulations.

117

The five locations are represented by their respective design wet bulb temperature in the figures 7.17 – 7.20. The design condition for the locations is the design ambient wet bulb temperature from ASHRAE (2001a). This design condition is statistically exceeded 0,4% of the year, i.e. 35 h/year. Locations in northern Europe with their respective design condition are presented in table 7.4 below. Table 7.4

Design wet bulb temperature at different locations in northern Europe

Location

118

Design wet bulb temperature (ASHRAE 0,4%)

Coincident dry bulb temperature (ASHRAE 0,4%)

Kiruna

14,2

18,9

Östersund

15,9

21,7

Aberdeen

17,5

21,0

Göteborg

17,7

23,5

Dublin

17,9

20,5

Copenhagen

18,2

23,2

Oslo/Fornebu

18,4

24,1

Stockholm

18,4

23,6

Birmingham

18,5

23,8

Helsinki

18,7

23,6

Tallin

19,0

23,0

London/Gatwick

19,3

25,0

Luxemburg

19,5

25,4

Riga

19,6

23,7

Munich

19,6

26,7

Prague

19,7

26,2

Hamburg

19,9

25,7

Stuttgart

19,9

27,3

Berlin

20,1

27,0

Amsterdam

20,3

24,8

Koln

20,3

27,1

Frankfurt

20,5

27,8

Brussels

20,7

26,5

Warzaw

21,0

27,6

Vienna

21,1

28,4

Krakow

21,2

27,9

Paris/Orly

21,4

27,9

Budapest

21,4

30,5


For a given design internal heat gain and thermal capacity, the maximum room air temperature shows a fair linear correlation with the design wet bulb temperature. The linear correlation has an R2 value between 0,78 and 0,89 The same plotting as above can be made with the variable “the percentage of working hours when the indoor temperature exceeds 24°C”, named %24°C, against the design wet bulb temperature. The plotting is made at the same given design internal heat gains and thermal capacities as above. In this case, a linear correlation can be found between %24°C and the design wet bulb temperature. Here the linear correlation is somewhat weaker with an R2 value between 0,60 and 0,84. The range of design wet bulb temperature scale in the figures 7.17 – 7.20, i.e. 15 – 22°C, cover entirely the wet bulb design conditions in the following countries in Europe (in alphabetical order); Austria, Belgium, Czech Republic, Denmark, Estonia, Finland, Germany, Iceland, Ireland, Latvia, Lithuania, Luxembourg, Netherlands, Norway, Poland, Slovakia, Sweden and United Kingdom. The range 15 – 22°C also covers the northern parts of France, parts of Switzerland and the northern parts of Hungary. Parameter variations are also made to study the upper and lower limits of the indoor air temperature when all parameters are set to either “lower indoor temperature”, i.e. best case, or “higher indoor temperature”, i.e. worst case. Examining the best and worst case gives an image of the span the indoor air temperature can vary within, when altering the parameters within the given limits in table 7.3. 36


34

Berlin

Without cooling

32

London

30

Stockholm 28

With cooling

26

Without cooling

Berlin London

24

With cooling

Stockholm 22 20 0%

10%

20%

30%

40%

50%

60%

70%

80%

90% 100%

working hours during a year (%)

Figure 7.21 Duration of the indoor air temperature on three different locations with and without cooling at best case parameter settings. Full year simulations.

119

In figure 7.21, the annual duration of the indoor air temperature on three different locations, Stockholm, London and Berlin, with and without cooling at best case parameter settings are presented. In the case with cooling, the analysed system performs as well as any conventional vapour compression cooling system. Most of the year, the indoor temperature is held constant close to the cooling set point, 21°C, at all three locations. The maximum indoor temperature in Berlin is 23,3°C when cooling is activated. Even at no cooling, the indoor conditions are not extreme. Stockholm and London has maximum indoor temperatures slightly above 26°C and Berlin above 28°C. When conditions are turned to worst case, as in figure 7.22, the indoor temperatures are of course higher. With cooling, Berlin peaks at 30,1°C, London at 29,1°C and Stockholm at 28,4°C. Without cooling the maximum temperatures are between 32°C to 35°C. Note that the set point temperature for the cooling beams is 23°C at the worst case.

36


Berlin

Without cooling

34

London

Without cooling

32

Stockholm

30

Berlin

With cooling

28

With cooling

London

26

Stockholm

24 22 20 0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

working hours during a year (%)

Figure 7.22 Duration of the indoor air temperature at three different locations with and without cooling at worst case parameter settings. Full year simulations. Interesting to notice is that the difference between maximum indoor temperatures with and without cooling seems to bee rather conform. The difference between maximum indoor temperatures with and without cooling is also called temperature depression. In table 7.5 is the temperature depression shown for different locations and different cases. For Stockholm and London the temperature differences is in the range 3,5° to 4,9°C. For Berlin it is somewhat higher 4,8°C to 5,4°C. The results in table 7.5 indicate that a

120


temperature depression of somewhere between 3,5°C and 5,5°C between the maximum indoor temperature with and without cooling can be expected using this type of cooling system. Table 7.5

Indoor temperature depression, i.e. difference in maximum indoor air temperature between cases with and without cooling for Base case, Best case and Worst case. Full year simulation

Climate

Base case

Stockholm/Arlanda London/Gatwick

4,2°C

Berlin

Best case

Worst case

4,9°C

3,6°C

3,8°C

3,5°C

5,4°C

4,8°C

In table 7.6 the maximum indoor air temperatures for best case, worst case with cooling and the difference between them are shown. The difference is very consistent ranging from 6,5°C to 6,8°C. Table 7.6

Maximum indoor air temperature for Best case, Worst case with cooling and the difference between them. Full year simulation

Climate

Best case

Worst case

Difference

Stockholm/Arlanda

21,8°C

28,4°C

6,6°C

London/Gatwick

22,6°C

29,1°C

6,5°C

Berlin

23,3°C

30,1°C

6,8°C

In table 7.7 the percentage of working hours which the indoor temperature exceeds 24, 25 and 26 °C shown, in best case and worst case, with and without cooling. The figures are calculated from the results shown in figure 7.21 and 7.22. Table 7.7

Percentage of working hours which the indoor temperature exceeds 24, 25 and 26 °C in best case and worst case, with and without cooling. With cooling Best case

Climate

Worst case

24°C 25°C 26°C 24°C 25°C 26°C 24°C

Without cooling Best case

Worst case

25°C 26°C 24°C

25°C

26°C

Stockholm/Arlanda 0,0% 0,0% 0,0% 11,5% 5,0% 1,9% 15,1% 4,1%

0,6% 31,2% 23,5% 15,9%

London/Gatwick

0,0% 0,0% 0,0% 12,9% 6,7% 2,4% 16,4% 5,1%

0,3% 33,7% 23,2% 15,9%

Berlin

0,0% 0,0% 0,0% 16,5% 9,8% 6,5% 24,3% 12,3% 5,5% 38,0% 29,9% 22,2%

121

At best case with cooling, the cooling capacity is much oversized and the indoor temperature is never exceeding 24°C in any climate location. At worst case with cooling, the numbers for Stockholm and London are relatively equal, from around 2% at 26°C up to 11-13% at 24°C. For Berlin, the result is 3-4% units above the other two locations. Bearing in mind that these are numbers for the worst case, the time each temperature limit is exceeded is close to where normal and acceptable indoor summer temperature design limits according to presented requirements in table 7.8. Table 7.8 Recommended peak summer conditions for natural ventilated or air conditioned buildings in design guides (Lawrence Race, 2003) Organisation

Requirement

Criteria (°C)

CIBSE Guide A (1999)

Not exceeded for more than 5% 25 resultant temp of annual occupied period (≈ operative temp)

BRE Environmental Design manual – Summer conditions in naturally ventilated offices (1988)

“Satisfactory” 23±2 “Intermediate” 24±2 “Min acceptable” 25±2 Acceptable design risk for these conditions not to be exceeded for more than 30 days in 10 year

MOD Defence Works Functional Standard, DMG 07:1996

Not exceeded for more than 2,5% of a month

30 resultant temp (≈ operative temp)

Energy Efficiency Office – ‘Performance specification for the energy efficient office of the future’ – Best Practice Programme Report 30 (1995). Applies to air-conditioned and naturally ventilated buildings

No more than 1% of year No more than 5% of year

>28 >25 resultant temp (≈ operative temp)

Lawrence Race (2003) has made a compilation of requirements based upon durations from published design guides. As can be seen, there are big differences in the requirements but also in the criteria. The requirements and criteria from BRE and MOD are exclusively for naturally ventilated buildings. The conclusion of the parameter variation is that climate location and internal heat gains play a relatively big role influencing the indoor temperature. Other parameters characterized as medium potent are; location in building, facade direction, thermal capacity and design temperature difference of cooling beams. The difference between best and worst case considering the maximum temperatures, in a full year simulation, is very consistent ranging from 6,5°C to 6,8°C. The percentage of working hours which the indoor temperature exceeds 24, 25 and 26 °C in is zero for the best case and for the worst case close to where normal and acceptable indoor summer temperature design limits are for many buildings. From the analysis of the results it can be concluded that the analysed cooling system is capable of cooling a building with relatively high internal heat gains, still maintaining an acceptable indoor air temperature.

122


7.3 Energy use The energy use of the cooling system is presented and analysed in this chapter. The energy use, i.e. use of electricity, is important since it generates both operating costs and contributes to pollution of the environment. The results are presented at parameter settings corresponding to the base case if not otherwise mentioned.

7.3.1 Base case At base case the COP and the cooling capacity of the cooling tower is presented during three days, 18th to 20th of August in figure 7.25. The COP is previously defined in chapter 4 Simulation model development, equation 4.44. Tuesday

Wednesday

10

10

8

8

6

6

4

4

2

2

Cooling capacity (kW)

COP

Monday

0 0 00:00 06:00 12:00 18:00 00:00 06:00 12:00 18:00 00:00 06:00 12:00 18:00 00:00 Time (hh:mm) COP

Cooling capacity

Figure 7.25 COP and cooling capacity of cooling tower between August 18 – August-20. Full year simulation. The COP is ranging between 4 – 8,5 during daytime whereas it drops down to zero during night time. The drop in COP down to zero comes from the drop of cooling capacity during night time when there is no need for cooling in the rooms and the ambient temperature is below the set point temperature of the supply air. The cooling capacity of the cooling tower, during day time, is in the range 2 – 3 kW. The COP during a full year is plotted against the ambient wet bulb temperature in figure 7.26. Very little is found in literature concerning COP of a cooling tower applied to chilled ceilings or chilled beams. Costelloe and Finn (2001) have published data from measurements of the COP on a cooling tower at different ambient air wet bulb temperature (WBT), with the purpose of cooling water for chilled ceilings or chilled

123

beams. Their findings are incorporated in figure 7.26. As can be seen, their data is well in accordance with the results from the simulations in this thesis. The simulation result gives a yearly mean COP of 7,0 based on the quota between yearly total cooling energy and yearly total electrical energy for pumps and fans in the system.

100 90

Cooling tower COP

80 70 60 50 40 30 20 10 0 0

2

4

6

8

10

12

14

16

18

20

Am bient w et bulb tem perature (°C) Simulation results

Costelloe & Finn 2001

Figure 7.26 COP cooling tower in relation to the ambient air wet bulb temperature. Full year simulation. The COP of the cooling tower increases when the WBT drops under 12-13°C. This is due to reduction of fan speed, and thus reduction of energy use, when the required cooling tower capacity is lowered to maintain a constant supply temperature in the chilled liquid. The electricity to the fan stands for the major part of the energy use of the cooling tower, hence the rise in COP. In table 7.9, the energy use on a yearly basis of the different devices in the cooling tower is presented.

124


Table 7.9

Energy use of the different devices in the cooling tower and comfort cooling system at base case conditions. Full year simulation Use of electricity (kWh/year)

Device

Use of electricity (%)

Pump for spray water

39

6,8

Pump in primary circuit

52

9,1

Pump in secondary circuit

21

3,7

Cooling tower fan

459

80,4

Total use

571

100,0

7.3.2 Single parameter variations – energy use The energy use of the cooling tower and the cooling system is a result of the performance of many parts in the cooling system, the building and the activity in it. The performance of the different parts is influenced by many parameter values. An obvious question in this context is; how “much” is each parameter contributing with to the resulting energy use. To answer this question parameter variations in this chapter are made to examine the potential of some parameters regarding its influence on the energy use. The parameters used are the examined cooling tower parameters, see chapter 7.2.2 Single parameter variations – indoor temperature, and two other parameters being presumed as giving a large affect on the cooling tower energy use. Table 7.10 shows the selected parameters. Table 7.10 List of parameter alterations Parameter Cooling Tower 4. Design primary approach

Lower indoor temperature

Base case

Higher indoor temperature

3°C

4°C

5°C

5.

Design primary range (both primary and secondary)

2°C

3°C

4°C

6.

Design mass flow quota of liquid and M& air, ( l ) M& a

2,0

1,0

0,5

5.

Design cooling capacity, Q& CT

1,25 ⋅ Q& CT

Q& CT

0,75 ⋅ Q& CT

Building 19. Internal heat gain; magnitude (maximum )

50 W/m2

60 W/m2

70 W/m2

Stockholm/ Arlanda

London/ Gatwick

Berlin

10. Geographical location/ Climate

125

The variations for each parameter are made with one higher and one lower value. To make the labelling of the variations consistent with the variations in chapter 7.2.2 Single parameter variations – indoor temperature, they are called “lower indoor temperature” and “higher indoor temperature”. They are referred to these names since lowering the actual value of some parameters raises the indoor temperature and vice versa. All variations that cause a drop in the indoor temperature are gathered in the column labelled “lower indoor temperature” and all variations that cause a raise in the indoor temperature are in the column “higher indoor temperature”. A more detailed discussion of the parameters can be found in chapter 5 Simulation of a free cooling system with evaporative cooling tower. The variables examined are the COP (Coefficient of Performance) of the cooling system and the use of electrical energy in the cooling system.

Difference of annual cooling tower energy use (%)

100% 80% 60% 40% 20% 0% -20% -40% -60% Prim approach

Prim Range

Flow quota

Higher temperature

Cooling capacity CT

Internal heat gain

Climate

Lower temperature

Figure 7.27 Differences in cooling system energy use in comparison to base case for different parameter alterations. For an explanation of the labels “higher temperature” and “lower temperature” see above. Full year simulation. In figure 7.27 the variations in cooling tower annual energy use is presented due to parameter variations. Two parameters are dominating; Primary range and the Flow quota. The latter is quite often referred to as the L/G, i.e. the mass flow relation Liquid/Gas, in cooling tower literature. These two parameters directly influence the air mass flow. The air mass flow is in turn directly influencing the energy use of the fan which is the dominating part of the contributors to the total energy use, see table 7.9. The primary range influences the mass flow of liquid at a given cooling capacity of the

126


cooling tower. The primary range, which is inverse proportional to the liquid flow, directly influences the air mass flow through the flow quota (L/G), hence the total use of energy in a cooling tower. The mass flow quota (L/G) has an inverse proportionality to the annual energy use, the higher the L/G the lower the design air mass flow; hence a lower use of energy is obtained. The flow quota has the strongest impact on the energy use, with a change of up to 75% of the annual use of energy going from a flow quota of 1,0 to 0,5. The other parameters in figure 7.27 have a more indirect and a bit weaker influence on the annual energy use. The influence is within ±20% of the energy use in the base case with the exception of one climate alteration. 100%

Difference of mean annual cooling tower COP (%)

80% 60% 40% 20% 0% -20% -40% -60% Prim approach

Prim Range

Flow quota

Higher temperature

Cooling capacity CT

Internal heat gain

Climate

Lower temperature

Figure 7.28 Differences in cooling system annual COP in comparison to base case for different parameter alterations. For an explanation of the labels “higher temperature” and “lower temperature” see above. Full year simulation. The variations of the coefficient of performance (COP) for the cooling tower and the comfort cooling system together are shown in figure 7.28. The bars in the figure are more or less a mirror image of the bars in figure 7.27. However, there are some discrepancies. The most obvious is the Primary range but also the Primary approach.

127

7.4 Findings from other sources There are very few findings from other published sources where the thermal indoor climate is analysed in a building with a cooling system which is the same as, or equal to, the one presented in this thesis. Facão (n.d.) presents outcome from the EcoCool project and Bohler et al. (2002) presents findings from simulations using the simulation program ConsoClim. In figure 7.22 duration curves from Facão (n.d.) show the outcome of a simulation for a whole cooling season in Zurich, using the simulation program TRNSYS. The simulation was performed on an office building with a typical construction and utilization pattern together with a cooling system containing a closed wet cooling tower and chilled ceilings. During working hours, indoor temperature reaches a maximum of 27.5ºC, and is above 26ºC less than 40 hours, which is about 1% of the annual working hours. The design cooling capacity for the chilled ceiling is unfortunately not mentioned. Bearing in mind that the climate in figure 7.22 is Zurich, the duration curves of Facão (n.d.) comply fairly well with the duration curve in figure 7.11, except for at low ambient temperatures (there seems to be no heater in the room).

Diagram from Facão (n.d.) with results from the EcoCool project

Room air temperature Ambient air dry bulb temperature

Ambient air wet bulb temperature

Figure 7.22 Diagram from Facão (n.d.) showing duration of indoor air temperature together with temperatures of ambient air and chilled water and cooling tower performance. Outcome from TRNSYS simulations. The diagrams 7.23 and 7.24 show findings from Bohler et al. (2002). They made simulations with the simulation tool ConsoClim (a French program) on an office room of 15 m2 equipped with chilled ceilings connected to an open cooling tower with an intermediate heat exchanger to avoid fouling in the chilled water secondary circuit. There was no conventional chiller attached. The research work included analysis of the operative temperature in an office at a number of different prerequisites, i.e. three design cooling capacities of the chilled ceilings, two building inertia, two solar gains, two internal heat gains, two orientations (east and west) and three different climates

128


(locations of Trappes, Nice and Carpentras). In figure 7.23 and 7.24, made by the author with data from Bohler et al. (2002), the figures from simulations using the location Trappes is shown. Trappes is located close to Paris. The room cooling load contains both latent and sensible loads as well as effects from thermal storage in the building fabrics. The maximum operative temperature is the mean operative temperature on the three hottest hours of occupation. The three different cooling capacities, 30, 40 and 50 W/m2 is the design cooling capacity at a temperature difference between the room air temperature and the mean cooling water temperature of 6ºC.

Maximum* operative temperature (°C)

30 Medium thermal weight 28

Cooling capacity chilled ceiling 30W/m2

26

40 W/m2 50 W/m2

24

22

20 0

Figure 7.23

20 40 60 Room cooling load (W/m2)

80

Outcome of simulations from Bohler et al. (2002). Location is Trappes near Paris and with medium thermal weight in the room. *) The maximum operative temperature is the mean operative temperature on the three hottest hours of occupation.

129

Maximum* operative temperature (°C)

30 Heavy thermal weight

28

Cooling capacity chilled ceiling

26

30 W/m2 40 W/m2 50 W/m2

24

22

20 0

20

40

60

80

100

Room cooling load (W/m2)

Figure 7.24

Outcome of simulations from Bohler et al. (2002). Location is Trappes near Paris and with heavy thermal weight in the room. *) The maximum operative temperature is the mean operative temperature on the three hottest hours of occupation.

It is difficult to translate the room cooling load, including both latent and sensible heat, as used by Bohler et al. (2002) to the total internal sensible heat gain, including solar radiation, used in this thesis. The total internal sensible heat gain in a room is however almost always higher than the sensible cooling load. The difference between them can vary considerably depending on several factors, e.g. the thermal capacity of the building fabric. Considering this, the figures from Bohler et al. (2002) apply well to the results in this thesis.

130

8

Results and analysis of measurements The results from the monitoring described in chapter 6 “Monitoring of free cooling system with evaporative cooling device” are presented and analyzed in this chapter. The results are separated into sections including Ambient conditions, Indoor thermal climate, Cooling tower performance and energy use.

8.1 Ambient conditions The ambient conditions during the summer period, i.e. May 01 – August 31 year 2007, was relative normal according to statistical data from the Swedish Meteorological and Hydrological Institute (SMHI). When comparing the monthly mean values of the measured outdoor temperature with the normal temperatures for Gothenburg from SMHI during the period of 1961 – 1990 (VVS 2000, 2003), the difference are relatively small for the whole period of May – August, see table 8.1. In table 8.1 the measured monthly mean values are between 0,3°C - 2°C higher than the normal temperatures for May, June and August. In July the measured monthly value is 1°C lower. For the whole period, May – August, the mean value of the measured temperatures is 0,7°C above the statistically normal temperature. Table 8.1

Comparison between measured monthly outdoor mean temperatures and statistically normal temperatures (SMHI; Gothenburg 1961 – 1990) Monthly mean outdoor temperature [°C]

Period

Measured values, Kvarnberget Gbg

Normal values, Gothenburg (1961-1990) SMHI

May

11,8

11,5

June

17,6

15,6

July

16,0

17,0

August

17,6

16,2

May - August

15,8

15,1

In figure 8.1 the measured outdoor temperature is shown during the period of May 1 – August 31, 2007. During about a week in the beginning of June, the only real heat wave occurred with temperatures above 30°C during several days. In the first half of August there was a shorter period of warm and sunny weather. As for the rest, the whole period is characterized by varying temperatures between 10 – 25°C.

131

Outdoor temperature 01 may - 31 aug Kvarnberget, Göteborg 2007 40 35

Outdoor temperature [°C]

30 25 20 15 10 5 0

Time

Figure 8.1 Measured outdoor temperature at Kvarnberget, Gothenburg during the period of May 1 – August 31, 2007. A duration curve of the measured outdoor temperature during May – August is shown in figure 8.2. The outdoor temperature was below 20°C during 85% of the time. Duration outdoor temperature 1 may - 31 aug Kvarnberget, Göteborg 2007 40 35 30 Temperature [°C]

25 20 15 10 5 0 0

10

20

30

40

50

60

70

80

90

100

Time [%]

Figure 8.2 A duration curve of the measured outdoor temperature at Kvarnberget, Gothenburg during the period of May 1 – August 31, 2007.

132

8 Results and analysis of measurements

In the next chapter the indoor thermal climate in the office building at Kvarnberget in Gothenburg, with the described pilot plant system applied, will be discussed.

8.2 Indoor thermal climate In this section, results concerning the indoor thermal climate in the office building at Kvarnberget are presented. The resulting indoor thermal climate is not solely dependent on the outdoor climate such as air temperature and solar radiation. It is also, to a fairly high degree, dependent of the magnitude of the indoor heat generation from people, lighting and electrical appliances as well as the ability of the building structure to store excess heat. Furthermore the indoor thermal climate is of course dependent on the ability of the comfort cooling system to remove excess heat. The indoor air temperature in relation to the corresponding outdoor air temperature is shown in figure 8.3. The dark (blue) dots are representing the indoor temperature at times when the free cooling system is running. The lighter (orange) dots represent the indoor temperature when the conventional comfort cooling system with a vapour compression chiller is running. As described in chapter 6.2 Pilot plant system lay-out there is a function in the control system which disconnects the free cooling system if the indoor air temperature exceeds 25°C. Thus there are no registered indoor air temperature readings above 25°C when the free cooling system is running. Besides that, the free cooling system has experienced malfunctions at a few occasions and then the conventional comfort cooling system has taken over the indoor climate control. However, as figure 8.3 indicates, the free cooling system seems to be able to keep the indoor air temperature below 25°C at outdoor temperatures up to 27°C.

133

Indoor temperature vs outdoor temperature Kvarnberget Gothenburg, May 1 - Aug 31 2007 30 29

IIndoor temperature [°C]

28 27 26 25 24 23 22 21 20 19 18 0

5

10

15

20

25

30

35

40

Outdoor temperature [°C] With free cooling

With ordinary chiller

Figure 8.3 Indoor air temperatures in relation to the outdoor air temperature at Kvarnberget, Gothenburg, during the period May 1 – August 31, 2007. The duration of the indoor air temperature, when the free cooling system has been active and running, is shown in figure 8.4. The free cooling system has been running active during 2585 h of the period May 1 – August 31, which is equivalent to 88% of this period. The time scale 0 – 100% in figure 8.4 is similar to 0 – 2585 h. The duration of the indoor relative humidity during the period of May 1 – August 31, 2007 is shown in figure 8.5. In this case the relative humidity for all hours throughout the period of May 1 – August 31 is shown (a total of 2952 h). This is because the supply air to the office space involved is delivered from the central air conditioning unit where the air is cooled and also dehumidified. The relative humidity has been in the interval 30 – 70% during the period in question.

134


Duration of indoor temperature when free cooling system is active Kvarnberget Göteborg, May 1 - August 31 2007 30

IIndoor temperature [°C]

28 26 24 22 20 18 0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Time [%]

Figure 8.4 Duration of the indoor air temperature when the free cooling system is active and running. Kvarnberget, Göteborg for the period May 1 – August 31, 2007

Figure 8.5 Duration of the indoor relative humidity for the period of May 1 – August 31, 2007 at Kvarnberget, Göteborg.

135

8.3 Performance and energy use The resulting indoor climate when applying evaporative cooling is of prime interest in this thesis. The energy use and the performance of the free cooling system is however also of great interest. Few, if any, building owners can neglect the running costs and the environmental impact from a cooling system. One performance factor of interest is the supply temperature in the secondary circuit, the circuit connected to the chilled beams in the building (see figure 5.5 and 6.1). Secondary supply temp. vs outdoor temp. evaporative cooling Kvarnberget, Göteborg 1 maj - 31 aug 2007

Secondary supply temperature [°C]

35 30 25

B

20 15

A

10 5 0 0

5

10

15

20

25

30

35

Outdoor drybulb temperature[°C]

Figure 8.6

Secondary supply temperature vs. outdoor temperature at evaporative cooling mode for the period of May 1 – August 31, 2007 at Kvarnberget, Göteborg

Figure 8.6 shows the secondary supply temperature (SST) in the system with chilled beams vs. the outdoor drybulb temperature. In area A the set point temperature is 12°C. During the period the samples in area B was measured, the outdoor temperature was above 12°C. There is a distribution of SST in area A around the diagonal line. The diagonal line indicate where the SST is equal to the outdoor temperature. The SST:s which are above the diagonal line, i.e. is higher than corresponding outdoor temperature, occurs mostly during night time or early morning when the outdoor relative humidity is close to 100% . The SST:s which are below the diagonal line, i.e. is lower than corresponding outdoor temperature, occurs mostly during daytime when the relative humidity is below 100%. During these circumstances the wet bulb temperature is several degrees below the dry bulb temperature. Consequently it is possible to achieve SST:s to a temperature below the outdoor dry bulb temperature.

136


In area B the set point temperature is 17°C in order to examine the energy use at a higher set point temperature.

COP vs outdoor temperature evaporative cooling Kvarnberget, Göteborg, May 1 - August 31 2007 8 7

COP [-]

6 5 4 3 2 1 0 0

5

10

15

20

25

30

35

Outdoor dry bulb temperature [°C]

Figure 8.7

COP of the free cooling system vs. outdoor dry bulb temperature at evaporative cooling mode for the period of May 1 – August 31, 2007 at Kvarnberget, Göteborg

In figure 8.7 the COP of the evaporative cooler vs. the outdoor temperature is shown. There is a fairly large variation of COP at a given outdoor temperature. A reason for this is a variation in the cooling demand in the building while the evaporative cooler is running at full speed to cool the liquid to the set-point temperature. The variation in COP is not only from differences in cooling demand. COP for an evaporative cooler is usually increases with decreasing outdoor temperature. The basic reason comes from the reduction of fan speed at lower temperatures since the set-point temperature can be reached with lower air flow through the evaporative cooler. Simultaneously the cooling demand can be fairly high at lower temperatures. This two factors together lead normally to an increasing COP at lower outdoor temperatures. At lower temperatures the COP can reach about 10 and even reach 20 – 30 and more at outdoor temperatures below 5 - 10°C. Regarding the evaporative cooler in the pilot plant, the COP is relatively low. Most of the time the COP is equal to or lower than the COP for a conventional chiller. However it is important to bear in mind that the COP varies with the outdoor wet bulb temperature and is lower at high wet bulb temperatures. The yearly mean COP, often called SPF (Seasonal Performance Factor), is therefore usually higher than the values showed in figure 8.7.

137

There are basically two main reasons for the low COP of the evaporativ cooler in the pilot plant. The first is a fairly low L/G-value (see chapter 5.4.2 Discussion and motivation of parameter alterations for explanation of L/G). In this case L/G is 0,15 at design rate. Compared to normal range of L/G of 0,5 – 2 for cooling towers it can be considered low. A low L/G implies high air flow rates, which in turn leads to a low COP. The second reason involves the design of the evaporative cooler. As mentioned earlier, the evaporative cooler in the pilot plant is a type of standard air cooled heat exchanger equipped with a spray water device for the evaporative function. It is primarily designed for cooling liquid with outdoor air without the spray water function. When the spray water function is active the water is sprayed from underneath, the whole heat exchanger area is not completely wetted due to a quite narrow spacing of the flanges of the heat exchanger. Due to the narrow spacing the water tends to adsorb to the flanges and also between them, due to the surface tension of the water. This leads to a more narrow space for the air stream to pass, which both influence the heat exchange and the pressure drop of the air in a negative way. Together, this influences the COP of the evaporative cooler in a negative way. Standard air cooled heat exchangers, sometimes labeled fluid cooler, is primarily designed for high cooling capacity at a low price, not to have the lowest life cycle cost. Among the biggest producers of air cooled heat exchangers, there are none, who declare COP for their range of products.

138

9

Conclusions and discussion In this chapter, conclusions from previous research and the results in this thesis are presented. Conclusions are followed by a discussion concerning various aspects of comfort cooling in general and the comfort cooling system examined in the thesis in particular.

9.1 Conclusions Research on hydronic comfort cooling systems with cooling towers as a sole free cooling source in conjunction with chilled ceilings or chilled beams have been in progress since the late 90’s. The amount of published research is relatively limited. However, the collected experience from the published material gives clear indications of the following: It is possible to maintain cooling water supply temperature below 18 – 20°C during a large part of the year in the northern parts of Europe. The northern parts are approximately north of latitude 48 - 49° N. For most climates north of latitude 48 - 49° N the availability is above 90%. Costelloe and Finn (2002) reports annual availability of chilled water from a cooling tower with a fixed approach of 3°C at different supply water temperatures for the locations Dublin and Milan. For cooling water supply temperature below 18 – 20°C the availability is 97 – 99% of the year for Dublin and 67 – 78% for Milan. The cooling water temperature obviously exceeds its design temperature for short periods during warm weather. Despite this, the maximum indoor temperature in most cases is in the range 25 - 27°C, based on measurements or outcome from building simulations. However, the maximum temperatures occur only during a limited time. There is limited information about the design prerequisites concerning building, internal heat gains and cooling capacity in the published material. Published findings concerning duration of indoor temperature during for example a year have not been found. The COP of a cooling tower, with the mentioned application, is significantly higher than of a conventional chiller. A measured yearly average COP of around 33 has been reported (Costelloe & Finn, 2002). Other sources, e.g. Hasan et.al. (2007), indicate however lower values, around 7 - 8. There is however limited information regarding measured COP of cooling towers in literature.

The following can conclude the results from the research presented in this thesis: 1. A cooling tower model (CTM) is developed for the building simulation program IDA ICE. The CTM has been validated against both published data, which is based on measurement data and data calculated from an accurate model, and measured data from a pilot plant. The effectiveness – NTU method forms a basis for the CTM. The effectiveness – NTU method is also used in other cooling tower models in internationally recognized building simulation programs like TRNSYS and EnergyPlus.

139

The CTM can be used to simulate air heat exchangers as well as most common types of cooling towers in conjunction with a hydronic cooling system. The CTM can be configured as an open tower or one with a closed circuit. The heat exchange configuration can be either counter flow or cross flow in the model. The cooling tower fan in the CTM can be either single speed type or equipped with variable speed drive. The main results from CTM are mass flow and temperature of supply cooling liquid, energy use of fans and pumps in the system and COP of the cooling tower during the simulated period. 2. Results from the simulations of a base case building gives a maximum indoor air temperature about 25,5°C, whether during a heat wave or a full year simulation. When varying the design total internal heat gain, including solar radiation, between 50 – 70 W/m2 the maximum indoor air temperature is in the range 25,7 – 26,7°C at a design wet bulb temperature (dWBT) of 21°C 24,9 – 26,0°C at a dWBT of 19°C 24,0 – 25,2°C at a dWBT of 17°C All figures are for a building with medium thermal capacity and prerequisites according to the base case. For a building with heavy thermal capacity the maximum temperature figures are 0,5 – 0,7°C lower than those with medium thermal capacity. Design wet bulb temperatures of 17, 19 and 21°C is in accordance with ASHRAE climate data (0,4%) and represent climates in the northern Europe above latitude 48 - 49°N. The design total internal heat gain includes design heat gains from humans, lighting, office equipment and transmitted solar radiation. The duration of the indoor air temperature during working hours has also been presented. The percentage of indoor air temperature during working hours exceeding 24°C is in the range 3 – 8% with total internal heat gains between 50 – 70 W/m2 at a dWBT of 21°C and medium thermal capacity. At heavy thermal capacity, the range is 2 – 5%. For a dWBT of 17°C the range has decreased to 0 – 2% and 0 – 1% respectively. 3. Results concerning the indoor relative humidity (RH) from the simulations of the base case building, and during a year, give a maximum value of 80% RH at coincident indoor air temperature of 22,7°C. At maximum indoor air temperature, 25,3°C, the relative humidity is 61% RH. When varying the climate parameter, the same pattern is observed, i.e. the maximum relative humidity is in the range 78 -85% RH at coincident indoor temperatures of 21,4 – 23,4°C and at maximum indoor air temperatures, 24,8 – 26,1°C, the coincident relative humidity is 59 – 70% RH. 4. The annual coefficient of performance (COP) for the cooling tower at base case is 7,0. The annual coefficient of performance comes from the quota of the annual cooling energy and the annual electric energy to pumps and fans in the system. The variation of the COP over a year is from close to zero to about 100. The annual total electrical energy use is about 2 kWh/(m2, year), where the area is the

140

9 Conclusions and discussion

total area of the simulated office floor including office rooms, corridors and middle section. The cooling tower fan uses the major part of the total energy, about 80% and is in this thesis always equipped with a variable speed drive. When varying the design parameters of the cooling tower, i.e. primary approach, primary range and the mass flow quota, the annual COP differ between 10 – 70% from the base case COP, i.e. a COP of 7. The annual electric energy use varies between a few percent up to 75%. 5. The single parameter variations gives information of the relative impact each parameter has concerning indoor air temperature. The parameters with the highest impact are “Climate/location”, “Total internal heat gain” and “Thermal capacity of the building fabric”. The parameters ”Location in building” (especially when at top floor) and ”Design temperature difference of chilled beams” also have a fairly high impact. Interesting to notice is that all cooling tower related design parameters as well as the “Secondary range”-parameter has a relatively low impact on the indoor air temperature, i.e. less than about 0,2°C. The results from simulation in this thesis concerning maximum indoor air temperature and the COP of a cooling tower largely confirm the findings in earlier published research. The results from the monitoring of the pilot plant fairly well confirm the results from the simulations regarding indoor climate. The COP of the evaporative cooler in the pilot plant is however lower than compared to findings in the literature and results from simulations in this thesis. The reason is mainly a low L/G-value together with constrained airflow through the partly wetted flanges of the heat exchanger. This thesis brings the following new information: A validated cooling tower model in NMF-language. The impact each of 18 different design parameters in the cooling tower, the building and the HVAC-system, have on the thermal indoor environment. The indoor thermal climate in a commercial building with the examined comfort cooling system over a large geographical span, i.e. climates equal to the northern parts of Europe, which is approximately north of latitude 48 - 49°N. The annual duration, i.e. the percentage of working hours when a given indoor temperature limit, e.g. 24, 25 and 26°C, is exceeded, during a whole year. The indoor relative humidity during a whole year in a normal office room with the examined comfort cooling system applied. The impact each of six different design parameters, in the cooling tower and the building, have on the use of electric energy in the cooling tower and the cooling system. Operational experience from a pilot plant; the evaporative cooler could keep a good indoor thermal climate in a normally occupied office with chilled beams based on normal design criteria. The energy efficiency of the evaporative cooler was however poor. It had a relatively low COP during the measured period (May 1 – Aug 31, 2007) ranging from close to zero up to about 4.

141

9.2 Discussion In this thesis, a hydronic comfort cooling system with a cooling tower as the sole free cooling source is discussed. Although hydronic comfort cooling systems with chilled ceilings, chilled beams or fan coils and cooling towers by themselves are wellestablished techniques, the combination of them can be considered as novel. New and unknown techniques are often treated with scepticism, especially in the building industry. Sometimes the scepticism is of good reasons but at times, it can be derived from lack of knowledge. For this reason, it is important to gain and disseminate knowledge of the comfort cooling system in this thesis, and other low energy and CFCfree systems, to give design teams more options concerning technical systems with low environmental impact to chose from when selecting solutions for new or refurbished buildings. In the following sections, various aspects of comfort cooling in general and the comfort cooling system examined in thesis in particular are discussed.

9.2.1 Design Design aspects of a cooling tower system In a normal design procedure of an arbitrary heat exchanger, the media which is to be heated or chilled, e.g. air or water, and the heating or chilling media are assumed to have constant design temperatures. The two mass flows are also normally assumed constant. These assumptions make it easy to select a proper size of an arbitrary heat exchanger from either design charts or design computer programs issued by a producer. Normal design of a cooling tower in, for example, an industrial process with high temperatures is straightforward. After determined the design wet bulb temperature (dWBT) for the actual location, the difference between the dWBT and the required supply temperature of the cooling water settles the approach temperature. The required cooling capacity determines the mass flow and the range temperature. After settling a proper L/G, the mass flow of air in the tower is determined. The matter of designing a cooling tower as the sole cooling source in a hydronic comfort cooling system is however not obvious. Figure 9.2 illustrates the dilemma.

142


Temperature [°C] 23 22 21 20 19 18 17 16 15 14 13

Figure 9.2

3

Design supply temperature from cooling tower

2

Design wet bulb temperature

1

Ordinary design supply temperature for chilled beams

Temperature 1 and 3 are normally equal!

Example of the design dilemma when designing a cooling tower as the sole cooling source in a hydronic comfort cooling system

Assuming the dWBT is 19°C and the cooling tower has a primary approach of 4°C, the supply water from a cooling tower has a design temperature of 23°C. If there is a secondary circuit, an additional 1 – 2°C must be added to the supply water design temperature to make up for the secondary approach of the intermediate heat exchanger. The chilled ceilings or beams have in this example a design supply temperature of 14°C. The design temperature can of course be raised up to 18 – 20°C, with a following increase in size of the chilled ceilings or beams, but going as far as 23°C or more is practically and economically impossible since a normal design indoor maximum temperature is in the range 24 - 26°C. A major part of the room would be filled with chilled ceilings or beams at these conditions! How can this dilemma be solved? There are principally two ways to deal with this problem: 1. The static approach: Lower the dWBT until temperature 1 and 3 meets as in normal static design of heat exchangers. Temperature 1 can possibly be raised up to 18 – 20°C to minimize the decrease of the dWBT. Lowering the dWBT will increase the size and investment cost of the cooling tower. In section 7.4.2 Discussion and motivation of parameter alterations, figure 7.9 and adjacent text, this is discussed. On the other hand, a rise in the design supply temperature of the chilled ceilings or beams will increase the size and investment cost of them. It is easy to see that there is an interesting optimization problem here; at what dWBT will the total investment cost for the cooling tower and the chilled ceiling or beams be at minimum? This has however not been investigated in this thesis. Still it is an important question which should be looked into more thoroughly. The result of a low dWBT would be an oversized cooling tower, i.e. the cooling tower can produce supply cooling water at a smaller approach than a tower

143

designed at a higher dWBT when the actual WBT is above the (low) dWBT. This is of course positive but not necessarily needed. If it is necessary depends on the indoor thermal requirements, the cooling load and the design conditions of the chilled ceilings or beams. A matter of fact is however that the “oversized” cooling tower generates higher investment costs. 2. The dynamic approach: This approach involves an acceptance of the dilemma, i.e. leave the design temperatures where they are. This is the approach which has been applied in this thesis. For the base case, the cooling tower is designed with a dWBT of 19,5°C with a primary approach of 4°C and a secondary ditto of 1,5°C. At these design conditions, a secondary supply water temperature would be 25°C. This temperature equals exactly the design indoor temperature. Designing the chilled beams at this temperature would imply an infinite size of the beams! The design supply temperature in the base case is however at a moderate level of 16°C with a secondary range of 3°C. With a design heat gain of 60 W/m2 and a design cooling capacity of 50 W/m2 the resulting maximum temperature in the analysed room is 25,3°C for the base case. How can the resulting indoor temperature be almost equal to the design indoor temperature when the design temperature condition of the cooling water system obviously is not met? The answer is found in the dynamic heat balance of the room. There is no heat flux or temperature inside or outside a normal room that is static, especially not during the warmer part of the year. An example can illustrate this: When the maximum indoor temperature at base case, 25.3°C, was registered there was a few days of heat wave with maximum outdoor dry bulb temperatures about 28.5°C. However the WBT reached the design wet bulb temperature (dWBT=19.3°C) only for a very short period during one day. The rest of the days during the warm period when the maximum indoor temperature occurred, the WBT was often several degrees below the dWBT (about 1 - 4°C). The chilled beams could however not reach the design cooling capacity of 600 W, i.e. 50 W/m2. Instead, the cooling capacity of the chilled beams varied throughout the day between 150 W and 260 W. This is mainly due to variations in the WBT, which gives variable supply temperatures, together with variations in the indoor air temperature. The indoor air temperature varied during the day from about 22°C in the morning to about 25°C in the afternoon. Another important factor, apart from the cooling system, is the heat storage in the building fabric. Its maximum was more than double the cooling capacity of the chilled beams. The lesson to learn from this is that none of the heat fluxes in a room can be regarded as constant. Not even if a conventional chiller provides constant supply cooling water temperature. The required indoor temperature can still be obtained even if all the temperatures fluctuate, including the cooling water supply temperature. It is important to realize that the design conditions in a building hardly ever occur, whereas for example the design conditions for a heat exchanger in an industrial process can be ever present. In the former case, the design conditions are merely a way to size a cooling device. The matter of greatest importance is normally how and when the different heat

144


fluxes in a room interact, i.e. the dynamic pattern of the heat fluxes. In general, the pattern itself has the greatest influence on the resulting temperature. In the latter case, the design conditions are crucial since these conditions always occur and there are no other heat fluxes influencing the process. In this discussion, it is essential to stress that the capacity of a cooling device in a room is not unimportant, it is just less important compared to the dynamic pattern of all heat fluxes. For the industrial heat exchanger example, the design heat (or cooling) capacity is however crucial. A metaphor of the dynamic pattern of heat fluxes can be two teams playing soccer. One team is the “heating” team and the other is the “cooling” team. The outcome of the game is the resulting indoor air temperature. If one or two team members in the cooling team have a minor injury or a period of tiredness, the team still can perform well (and win!). The performance will of course decrease if the team plays with reduced number of players or if they all play bad. In team sports as well as in the dynamic heat balance of a room, the result is the collected effort of each member in each team. The only way to find out if the design cooling capacity of the room cooling devices, when connected to a cooling tower, is sufficient to keep the indoor air temperature below the maximum allowed room temperature is to run simulations with a qualified building simulation tool. Only then can all the dynamic heat fluxes and fluctuating temperatures be taken into account. This design approach may feel unusual for an HVAC engineer, but the difference from traditional design is not very big. The only actual difference is that the supply cooling water temperature is varying due to the variable wet bulb temperature during the day. All other temperatures and heat fluxes varies anyway, whether the supply cooling temperature is constant or not.

Risk of condensation on chilled beams One argument against chilled ceilings or beams is the risk of condensation with subsequent risk of indoor precipitation. This is a risk which is apparent at combinations of warm and humid climate, limited or no dehumidification of supply air, and low cooling system temperatures. There is however several ways of eliminating this risk in systems with normal constant supply cooling water temperatures in the range of 12 – 14°C. For example, humidity sensors can be used either in an exhaust air duct or in zones with increased risk for condensation. If the humidity rises above a given limit, a control device actuates a rise in the supply cooling water temperature above the dew point temperature to eliminate the condensation risk. There is also a special sensor on the market which detects condensation on single room cooling devices and simply closes a valve to the cooling device, thus eliminating the risk of condensation locally. For a hydronic cooling system with chilled ceiling or beams connected to a cooling tower as the sole provider of chilled water, the situation is different. In this case, the cooling water supply temperature has a natural minimum limit in the wet bulb

145

temperature. In practical terms, the cooling water supply temperature is a few degrees Celsius above the WBT. In figure 9.3, an example of possible conditions in a room is visualized. The room is chilled with a cooling system with only sensible cooling, e.g. a hydronic cooling system with chilled ceiling or beams connected to a cooling tower as the sole provider of chilled water. In the example in figure 9.3, the outdoor condition is 25°C with a humidity ratio of 0,011 kg water/kg dry air, i.e. close to 60% relative humidity. In the building, this air is chilled to 22°C. The vertical thick line illustrates a case with no humidification in the room or with very high air change rates, i.e. ventilation rates, together with a limited moisture load. At this case, the dew point temperature is about 15,5°C. The cases with dotted lines indicate the resultant condition in the room with various combinations of moisture loads and air change rates. The two cases which have a resulting relative humidity of 70% and 80% RH represent cases with normal air change rates and moisture loads in commercial buildings. In these cases, the dew point temperatures are well below the upper limit dew point temperature. In the case with a resulting relative humidity of about 95% RH, the actual dew point temperature is above the upper limit dew point temperature, hence there is a risk for condensation on the room cooling devices. This case has a very high moisture load or a combination of high moisture load and low air change rate, which is unusual in most commercial buildings. humidity (kg/ kg) 0,000 25

0,005 Outdoor air condition

0,010

0,015

0,020

70% RH 80% RH 90% RH 100% RH 100%RF

20

Possible indoor air conditions in room

15

Approach

10 5

Upper limit dew point temperature

Outdoor wet bulb temperature

0 -5

Dew point temperature at no humidification in room

-10 -15 -20

Figure 9.3

146

Enthalpy – humidity diagram (Mollier chart) with an illustrated example of temperature and humidity conditions in a room with chilled ceiling or beams in a hydronic cooling system connected to a cooling tower as the sole provider of chilled water. Sensible cooling only.


As shown in section 7.2 Indoor thermal climate, the highest values of indoor relative humidity, 78 – 85% RH, occurs at indoor air temperatures between 22 and 24°C. Above that, the indoor relative humidity decreases. Based on the results in this thesis concerning the indoor thermal climate and the discussion above, it can be considered as extremely unusual to experience condensation on room cooling devices connected to a cooling tower as the sole provider of chilled water.

Design cooling capacity of chilled beams There are several ways to determine the required cooling capacity for a room. Usually some form of simulation program is used. If it is a qualified simulation tool, it considers all important factors in a dynamic heat balance to determine the required cooling capacity in a room. The required cooling capacity is determined with a given indoor thermal requirement. In some cases, the design cooling capacity is chosen from a “rule of thumb” figure. In this thesis, the base case cooling capacity is chosen to the “rule of thumb”-value common among Swedish HVAC engineers, namely 50 W/m2. This is done simply to obtain an even number and also to apply with normal cooling capacities in Swedish buildings. In a room with base case conditions and with a base case design temperature difference between room air temperature and the mean cooling water temperature (∆Ta-l) of 7,5°C, two modern active cooling beams with the length of 1,8 m, will suffice to provide the design cooling capacity. A ∆Ta-l of 7,5°C is a bit lower than standard ∆Ta-l which is in the range 8 - 10°C. If a mean standard ∆Ta-l is about 9°C the chilled beams in the base case are about 15% bigger than with a ∆Ta-l of 9°C.

9.2.2 Indoor climate Thermal climate The results in section 7.2 Indoor thermal climate indicate that the indoor climate can be kept at conditions where most people find it comfortable at a high or very high duration of normal working hours, i.e. the indoor air temperature does not exceed 24°C or 25°C more than between fractions of a percent up to 10 percent. This conclusion is valid for normal commercial buildings in climates similar to those in the northern parts of Europe, i.e. north of latitude 48 – 49°N. A cooling system comprising a cooling tower in conjunction with chilled ceilings or beams can however provide comfortable indoor conditions in warmer climates than examined in this thesis. The design internal heat gains including solar radiation must then be kept at moderate values, i.e. below about 40 – 45 W/m2 at medium thermal capacity or below about 45 – 50 W/m2 at higher thermal capacity. The warmer the climate the lower the allowed internal heat gain. An exception from this statement may be in climates with combinations of hot and humid conditions. At climates with design wet bulb temperatures higher than about 25 - 27°C the permitted internal heat gains are probably at such low values which would be practically impossible to reach in commercial buildings with reasonable efforts. The upper values of plausible design wet bulb temperatures for this kind of comfort cooling system has however not been investigated in this thesis.

147

Parameter variations In section 7.2.3 Multi parameter variations, variations with more than one parameter at a time have been performed. The number of parameter combinations have however been kept at reasonable levels to limit the number of simulation cases. One parameter analysis includes full variation of five locations/climates, two thermal capacities and three levels of design internal heat gain. The locations are translated to design wet bulb temperatures, thus allowing the maximum room air temperature to be plotted against the design wet bulb temperature at different combinations of design internal heat gain and thermal capacity. The diagrams discussed above and presented in section 7.2.3 Multi parameter variations can be used in early stages of the design phase of a building to get a first indication of probable maximum indoor air temperatures, alternately the probable percentage of working hours when the indoor temperature exceeds 24°C at given conditions, named %24°C. Alternately, the diagrams can be used to get indications of at which conditions a required maximum indoor temperature or maximum value of %24°C can be achieved. The other diagrams in section 7.2.3 Multi parameter variations, are presenting annual duration curves of the indoor air temperature at three different locations; Stockholm/Arlanda, London/Gatwick and Berlin at the conditions “Best case” and “Worst case”. Best case is defined as when all the examined parameters are adjusted to values equal to “Lower temperature”. Worst case is defined as when all the examined parameters are adjusted to values equal to “Higher temperature”. These diagrams give indications of the span of the annual duration of the indoor air temperature between the conditions “Best case” and “Worst case”. At “Best case” the indoor air temperature are kept at the set point temperature, i.e. 21°C, for a long period of time and maximum temperatures were kept below about 23°C. This is a performance equal to a conventional cooling system with high capacity and fairly strict temperature control. At “Worst case”, the annual duration of the indoor air temperature is of course at somewhat higher level. However, the results in this case are in accordance with normal and usually acceptable design limits concerning indoor temperature duration, i.e. a duration of the indoor air temperature where the percentage of working hours when the indoor temperature exceeds 25°C is between 5 to 10%. The maximum occurring indoor temperatures where between 28 - 30°C, however occurring for a very limited time.

9.2.3 Investments and annual costs An analysis of investment costs for different types of cooling towers compared with investment costs for conventional chiller has not been made in this thesis. Sources in the literature discussing investment costs for different types of cooling towers are extremely rare to find.

148


However in one reference, Facão (n.d.), which is a (unofficial) homepage where the EcoCool project is presented, a table with investments and annual costs is presented, see table 9.1. The EcoCool project is described in section 1.3 Previous work. According to Facão (n.d.), the first two column in the table presents investment costs for cooling equipment, i.e. chiller or cooling tower, and the chilled ceiling system including distribution system and room cooling equipment. The third column is the sum of the first two. The last column presents total annual costs, which includes investment costs with a depreciation period of 5 years together with annual costs for energy and maintenance. The data is from Switzerland and from about year 2000. There is no explanation to what area the numbers in the table are related to; [Euro/m2]

Table 9.1

Comparison between investment costs for different kinds of comfort cooling systems in the EcoCool project, Facão (n.d.).

System 1) Ecocool (closed cooling tower) 2) Ecocool (open cooling tower) 3) Chilled ceil. with refrig. mach. 4) Conventional AC (all-air)

Cooling equipment [Euro/m2]

Chilled ceiling [Euro/m2]

Investment cost [Euro/m2]

Total annual costs [Euro/m2]

35

155

190

45

17

155

172

41

22

155

177

43

-

-

255

61

The total investment cost for the EcoCool system number one are 7% above that of a system with chilled ceiling and a conventional chiller (refrigerant machine), i.e. system number 3, and EcoCool system number 2 are 3% below the mentioned system. The distribution system and room cooling equipment for the first three systems in table 9.1 are equal, hence the investment costs for that part of the entire cooling system are equal. The investment cost for a conventional all-air air conditioning system is considerably higher than the first three systems, i.e. about 40% higher. When comparing the first three systems, the total annual costs show no significant differences. The figures in table 9.1 indicate that there are small differences in total investment costs for hydronic systems whether they are chilled with a conventional chiller or a cooling tower (open or closed type). The investment cost for the open type cooling tower is however about 50% of the investment cost for a closed type tower. The relative share of the investment cost for the cooling towers is about 10 – 20% of the total investment cost.

149

It should be noted that the cooling towers in the EcoCool project were prototypes, hence the investment costs are most likely on the higher end. If the towers were mass fabricated the price would have been significantly lower.

9.2.4 Applicability Each low energy cooling system has it advantages and disadvantages. This is of course true also for the cooling system investigated in this thesis. Among the advantages for hydronic cooling systems with a cooling tower there is one which stands out; namely the applicability of the system. Some of the alternative cooling systems mentioned in section 2.2 Low energy cooling alternatives can only be applied to a new building, such as hollow core slab cooling with air or slab cooling with water. A few of the alternatives require access to the ground or an ambient water resource in the vicinity. This requirement can often be difficult to fulfil in the central parts of a city where most commercial buildings are located. A hydronic cooling system with a cooling tower can be applied to both new and refurbished buildings. In Sweden, it is common to replace an old air conditioning system distributing chilled air with a hydronic cooling system equipped with chilled beams. Hydronic systems with chilled beams are now the predominant cooling system in Sweden and all actors in the building industry consider this system as well established and accepted. Using this type of system in conjunction with a cooling tower, instead of a conventional chiller, requires only access to the ambient air. There is no need for access to the ground or any ambient water resources. The space required for the location of the cooling tower, normally on roofs or adjacent to the building, equals approximately the space required for the cooling of condenser heat from a conventional chiller. All parts of this cooling system represent well-established techniques. No parts of the system are new or unproven on the market. All these advantages, together with moderate investment costs, low annual costs and the environmental advantages, makes this cooling system an attractive alternative to conventional cooling.

9.2.5 Miscellaneous Legionella The minimal risk of growth of legionella bacteria in cooling towers applied as a sole free cooling source in a hydronic cooling system is discussed in section 4.4 Legionella in evaporative cooling towers. There is however a great concern for the threat of the legionella bacteria from cooling towers among both the public and HVAC engineers. In that context, it may be daring to introduce a cooling tower in a low energy comfort cooling system where the tower would be located adjacent to or on the roof of commercial buildings. If this system is

150


introduced in several buildings in city centres, there will be several cooling towers close to where numerous people are working and living. The risk for growth of legionella bacteria is close to zero in this application of a cooling tower as shown in section 4.4 Legionella in evaporative cooling towers. It is equal to the risk of legionella growth in public outdoor swimming pools, shallow ponds or public fountains, which also are close to where people live or work. Despite this, there still might be concerns only because it is a cooling tower. To many people, cooling towers are dangerous per definition. This is a concern which must be addressed.

Cooling tower size When discussion prejudiced thinking, there might be people associating cooling towers with those in huge power plants at heights of 30 meters and more, exhausting enormous fog clouds. A small contribution against this thinking, when applying cooling towers to hydronic comfort cooling systems in commercial buildings, can be made by presenting pictures of the cooling tower used in the EcoCool project, se figure 9.4.

Figure 9.4

Picture of no. 2 prototype cooling tower in the EcoCool project, Facão (n.d.).

The cooling tower in figure 9.4 has a design cooling capacity of 10 kW. The cross section of the cooling tower measures 0,5 m × 1,23 m with a height of 1,74 m. Required cooling capacities of normal commercial buildings are of course normally higher than 10 kW, usually in the range from 10 to about 300 kW. The intention with figure 9.4 is that the reader will get an idea of the magnitude of a cooling tower with cooling capacities close to normal for commercial buildings. In most cases, a cooling tower for cooling a commercial building is however somewhat bigger than in figure 9.4, but not more than about a few times bigger.

151

9.3 Further research Although a fair amount of new knowledge and information is presented in this thesis, several topics or questions can be further investigated; There is very little information from real life buildings with hydronic comfort cooling systems in conjunction with a cooling tower. A pilot plant has been evaluated in this thesis but further measurements on a real full size object would give important information about the performance of this kind of cooling system together with the indoor climate it can provide. The described comfort cooling system could possibly benefit from using thermal storage in the system, e.g. by using water tanks to store chilled water. The chilled water can be produced during nighttime when the ambient wet bulb temperature is lower. This would most certainly reduce the maximum indoor air temperature. A thermal storage can reduce the size of the cooling tower hence reduce investment costs. The economics of a hydronic comfort cooling system with a cooling tower needs to be further investigated. The different types of cooling towers, open and closed, have different investment costs and life cycle costs. How big are the differences? How do variations in cooling tower design parameters affect investment costs and life cycle costs? Is there an optimum combination of design wet bulb temperature, total approach and design supply and return temperature of chilled water, i.e. the size of the chilled beams? Is it an important question or could we care less?

152

References Adnot, J., (2002), "Central" (Commercial) air-conditioning systems in Europe, Proceedings of 2nd International conference on Improving Electricity Efficiency in Commercial Buildings (IEECB), Nice, 27-29 may 2002, p. 143-149. Almadari F., Butler D.J.G., Grigg P. F., Shaw M. R., (1998), Chilled ceilings and displacement ventilation, Renewable Energy 15, pp 300 – 305. 1998 Aschermann M., (2000), Validation of IDA ICE, Version 2.11.06 – With IEA task 12 Envelope BESTEST, Hochschule Technik+Architektur, Luzern Switzerland, 2000-0710. Aschermann M., Zweifel G., (2003), RADTEST – Radiant heating and Cooling Test Cases, A Report of Task 22, Subtask C Building Energy Analysis Tools Comparative Evaluation Tests, International Energy Agency (IEA), April 2003. ASHRAE, (1999), HVAC1 Toolkit: A Toolkit for Primary System Energy Calculations. Atlanta: American Society for Heating, Refrigerating and Air-Conditioning Engineers, Inc. (ASHRAE), 1999. ASHRAE, (2000a), 2000 ASHRAE Handbook – HVAC Systems and Equipment, American Society of Heating, Refrigeration and Air-Conditioning Engineers, Inc. (ASHRAE), Atlanta USA, 2000. ISBN 1-883413-81-8. ASHRAE, (2000b), ASHRAE Standard – Minimizing the Risk of Legionellosis Associated with Building Water Systems. ASHRAE Guideline 12-2000, American Society for Heating, Refrigerating and Air-Conditioning Engineers, Inc. (ASHRAE), Atlanta, USA, 2000. ASHRAE, (2001a), 2001 ASHRAE Handbook – Fundamentals, American Society of Heating, Refrigeration and Air-Conditioning Engineers Inc. (ASHRAE), Atlanta USA, 2001. ISBN 1-883413-88-5. ASHRAE, (2001b), International Weather for Energy Calculations 1.1 (IWEC Weather Files), American Society of Heating, Refrigeration and Air-Conditioning Engineers Inc. (ASHRAE), Atlanta USA, 2001. Published on a CD-ROM, ISBN 1-931862-17-6. ASHRAE/IESNA, (1999), 90.1 User’s manual – ASHRAE/IESNA Standard 90.1 – 1999, American Society of Heating, Refrigeration and Air-Conditioning Engineers Inc. (ASHRAE), Atlanta USA, 2000. ISBN 1-883413-79-6. Bahnfleth, W.P., Rehfeldt, J.A., (1996), Methods for enhancing the evaporative cooling potential of central chilled water plants, Proceedings, IDEA Conference, Washington D.C., June 9 – 12 1996 Baker D.R., Shryock H.A (1961), A Comprehensive Approach to the Analysis of Cooling Tower Performance, ASME Journal of Heat Transfer, vol. 83, pp 339 – 350, 1961

153

Behne M., (1997), Alternatives to Compressive Cooling in Non-Residential Buildings to Reduce Primary Energy Consumption, Retrieved June 11, 2003, from the Energy Performance of Buildings Group of the Indoor Environment Program of the Environmental Energy Technologies Division at Lawrence Berkeley National Laboratory, Berkeley California Web site < http://epb1.lbl.gov/thermal/dfgfinal.pdf > Benton D.J., Bowman C.F.., Hydeman M., Miller P., (2002), An Improved Cooling Tower Algorithm for the CoolTools Simulation Model. ASHRAE Transactions, vol. 108, part 1, pp 760 – 768, 2002 Bohler A., Fleury E., Millet J. R., Marchino D., Stabat P., (2002), Guidance and tools for chilled ceilings combined with a wet cooling tower, Proceedings from conference; EPIC AIVC 2002 – Energy efficient & healthy buildings in sustainable cities, Lyon France, 23 – 26 October 2002. Bourdouxhe J-P., Grodent M., Lebrun J., Silva C., (1994), Cooling tower model developed in a Toolkit for primary HVAC System Energy Calculation - Part I: Model description and validation using catalog data. Proceedings of the fourth conference on system simulations in buildings, Liège, December 5-7, 1994 Braun J.E., Klein S.A., Mitchell J.W, (1989), Effectiveness Models for Cooling Towers and Cooling Coils. ASHRAE Transactions, vol. 95, part 2, pp 164 – 174, 1989 BRESCU, (1995), Avoiding or minimising the use of air-conditioning – A research report from the EnREI Programme, GIR 31 (Watford: BRESCU) Retrieved March 5, 2004, from Action Energy, a government-funded programme, Web site: < http://www.actionenergy.org.uk >. BRESCU, (2001), New ways of cooling – information for building designers, GIR 85 (Watford: BRESCU) Retrieved March 9, 2004, from Action Energy, a governmentfunded programme, Web site: < http://www.actionenergy.org.uk > Brown G., (1990), The BRIS Simulation Program for Thermal design of Buildings and Their Services, Energy and Buildings 14 (1990) pp 385 – 400. CEN, (1998), Ventilation for buildings – Design criteria for the indoor environment, CEN Report CR 1752:1998 E, European Committee for Standardization, 1998. CIBSE, (2001), CIBSE Guide B2 – Ventilation and air conditioning, CIBSE – The Chartered Institution of Building Services Engineers, London, UK. 2001. ISBN: 1 9 0 3 2 8 7 1 6 2 CIBSE, (2004), CIBSE Guide F – Energy efficiency in buildings, CIBSE – The Chartered Institution of Building Services Engineers, London, UK. 2004. ISBN: 1 9 0 3 2 8 7 3 4 0 Costelloe B., Finn D., (2000), The design and performance of an evaporative cooling test rig for a maritime climate, Proceedings from conference; Joint CIBSE/ASHRAE conference, Dublin Ireland, September 2000, pp 830 - 845. Costelloe B., Finn D., (2001), Energy reduction by enhanced evaporative cooling of buildings in maritime climates, Conference on Renewable Energy in Maritime and Island Climates (REMIC), Belfast, May 2001. Costelloe B., Finn D., (2003), Indirect evaporative cooling potential in air-water systems in temperate climates, Energy and Buildings 35 (2003) pp 573 – 591.

154

References

Costelloe B., Finn D., (2007), Thermal effectiveness characteristics of low approach indirect cooling systems in buildings, Energy and Buildings 39 (2007) pp 1235 – 1243. Costelloe B., Finn D., (2009), Heat transfer correlations for low approach evaporative systems in buildings, Applied Thermal Engineering 29 (2009) pp 105 – 115. De Saulles, T., (1996), Free cooling systems – design and application guidance, The Building Services Research and Information Association, BSRIA, Nov 1996, ISBN 0 86022 453 8 Energimyndigheten, (2007), Förbättrad energistatistik för lokaler- Stegvis STIL Rapport för år 1 – Inventeringar av kontor och förvaltningsbyggnader, ER 2007:34, ISSN 1403-1892, Statens Energimyndighet, 2007. Energy Plus, (2004), Energy Plus Engineering Document – The reference to Energy Plus Calculations, march 29, 2004. Retrieved May 19, 2004, from US Department of energy homepage < www.eere.energy.gov/buildings/energyplus/documentation.html > EU, (2002), Directive 2002/91/EC of the European Parliament and of the Council of 16 December 2002 on the energy performance of buildings, Retrieved January, 2003, from The EU site: < www.europa.eu.int/scadplus/leg/en/lvb/l27042.htm > Facão J., (n.d.), Material retrieved March 4, 2004, from the Web site titled “Ecological Cooling For Buildings Combining a Closed Wet Cooling Tower with Chilled Ceilings”, at < http://paginas.fe.up.pt/~jfacao/ecocool/ >. The web site contains unpublished material from the EU-funded research program EcoCool (1997 – 2000). Facão J., Oliveira A., (2000), Thermal behaviour of closed wet cooling tower for use with chilled ceilings, Applied Thermal Engineering 20 (2000) pp 1225 – 1236. Fanger, P.O., (1972), Thermal Comfort, McGraw-Hill, New York, USA, 1972. Gan G., Riffat S.B., (1999), Numerical simulation of closed wet cooling towers for chilled ceiling systems, Applied Thermal Engineering 19 (1999) pp 1279 – 1296. Gan G., Riffat S.B., Shao L., Doherty P., (2001), Application of CFD to closed wet cooling towers, Applied Thermal Engineering 21 (2001) pp 79 – 92. Goswami, D., Reveliotty, C.N., (1987), Free cooling by cooling tower water, ASHRAE Journal, vol. 29, no. 1, pp 32 – 37, 1987 Guyon G., Moinard S., Ramdani N., (1999), Empirical validation of building energy analysis tools by using tests carried out in small cells, Proceedings of IBPSA Building Simulation ’99 conference, Kyoto, Japan 1999. Halasz B., (1999), Application of a general non-dimensional mathematical model to cooling towers, International Journal of Thermal Science vol. 38, pp 75 – 88, 1999, with citation from: Poppe M., (1973), Wärme- und Stoffübertragung bei der Verdunstungskühlung im Gegen- und Kreutzstrom, VDI Forschungsheft 560, 1973. Hasan A., Sirén K., (2002), Theoretical and computational analysis of closed wet cooling towers and its application in cooling of buildings, Energy and Buildings 34 (2002) pp 477 – 486. Hasan A., Gan G., (2002), Simplification of analytical models and incorporation with CFD for the performance predication of closed wet cooling towers, International Journal of Energy Research 26 (2002) pp 1161 - 1174.

155

Hasan A., Vuolle M., Sirén K., Holopainen R., Tuomaala P., (2007), A cooling tower combined with chilled ceiling: system optimisation, International Journal of Low Carbon Technologies vol. 2 issue 3 pp 217 – 224 (July 2007) Hensley, J.C., (1994), The application of cooling towers for free cooling, ASHRAE Transactions, vol. 100, part 1, paper no. NO-94-7-3, pp 817 – 823, 1994 Hernandez O.S.., Hannay J., Silva C., (1994), Cooling tower model developed in a Toolkit for primary HVAC System Energy Calculation - Part II: Model validation using experimental results. Proceedings of the fourth conference on system simulations in buildings, Liège, December 5-7, 1994. Hill G.E., Pring E.J., Osborn P.D., (1990), Cooling Towers – Principle and Practice, 3rd ed. Butterworth-Heinemann, 1990. Hipskind, J.T., Hawks, K.H., Goldschmidt, V.W., Tumilty, J.E., (1991), Natural cooling cycle in cooling towers: analysis of utilization, ASHRAE Transactions, vol. 97, part 2, paper no. IN-91-21-1, pp 1197 – 1208, 1991 ISO, (1994), Moderate Thermal Environments, Determination of the PMV and PPD Indices and Specification of the Conditions for Thermal Comfort, International standard, 1st edition, 1994. Jaber H., Webb R.L., (1989), Design of Cooling Towers by the Effectiveness – NTU Method, Journal of Heat Transfer, vol. 111, pp 837 – 843, November 1989 Kropf S., Zweifel G., (2002) Validation of the Building Simulation Program IDA ICE, According to CEN 13791”Thermal Performance of Buildings – Calculation of Internal Temperatures of a Room in Summer Without Mechanical Cooling – General Criteria and Validation Procedures”, Hochschule Technik+Architektur, Luzern Switzerland, February 2002. Kröger D.E., (1998), Air-cooled heat exchangers and cooling towers – thermal-flow performance and design, (Department of Mechanical Engineering, University of Stellenbosch), 875 p. Distributed by Begell House, Inc., New York, 1998 Lawrence Race G., (2003), A ‘Just Sufficient’ approach to building services design, Final report 70176-1, BSRIA, Bracknell, UK, June 2003. Lebrun J., Silva C., (2002), Cooling Tower model and experimental validation, ASHRAE Transactions, v 108 Part 1, p 751-759. 2002 Lindholm T., (2000), Evaporative and Desiccant Cooling Techniques – Feasibility when Applied to Air Conditioning, (Dissertation) Department of Building Services Engineering, Chalmers University of Technology, Göteborg, Sweden, 2000 Lindholm T., (2003), Frikyla – Analys av teknik och systemlösningar (Free Cooling – Analysis of technology and system solutions), slutrapport från eff-Sys, Energimyndighetens utvecklingsprogram Effektivare kyl- och värmepumpssystem, 2003 (in Swedish) Liddament M., (2000), Low Energy Cooling – Technical Synthesis Report, IEA – ECBCS Annex Publications - Annex 28, Low Energy Cooling Systems, Coventry, UK, ECBCS, 2000, ISBN: 1 902177 16 3

156

References

Marley, (1982), The Application of Cooling Towers for Free Cooling, Technical report H-002, December 1982, Marley Cooling Technologies, Retrieved May 10, 2003, from Marley’s home page < http://www.marleyct.com/ > Merkel F., (1925), Verdunstungskühlung, VDI-Zeitshrift, vol. 70, no. 70, pp 123 – 128, January 1925 Mizushina T., Ito R., Miyashita H., (1967), Experimental study of an evaporative cooler, International Chemical Engineering, 7 (4), pp 727 – 732, 1967 Mott, Green & Wall, (2001), Chilled ceilings and beams, Building Services Journal, v. 23, no. 5, pp 22 - 26 (2001). Murphy, D., (1991), Cooling towers used for free cooling, ASHRAE Journal, vol. 33, no. 6, pp 16 – 26, June 1991 Nilsson P.E. (ed), (2003), Achieving the Desired Indoor Climate – Energy Efficient Aspects of System Design, The Commtech Group, Studentlitteratur 2003 ISBN: 9144032358 Niu J., Kooi J.v.d., (1993), Dynamic simulation of combination of evaporative cooling with cooled ceiling system for office room cooling, Proceedings from conference; Building Simulation ’93 – Third International Conference, Adelaide Australia, 16-18 August 1993. Niu J., Kooi J.v.d., Ree H.v.d., (1995), Energy saving possibilities with cooled-ceiling systems, Energy and Buildings 23 (1995) pp 147 – 158. Nordman B., Piette M.A., Kinney K., Webber C., (1997), User Guide to Power Management for PCs and Monitors, LBNL-39466, Environmental Energy Technologies Division, Lawrence Berkeley National Laboratory (LBNL), University of California, Berkeley USA, January 1997. Retrieved March 10, 2004, from LBNL home page < http://www.enduse.lbl.gov >. Novoselac, A., Srebric J., 2002, A critical review on the performance and design of combined cooled ceiling and displacement ventilation systems, Energy and Buildings, 34 (2002) pp 497 – 509. Oliveira A., (2000), Ecological cooling for buildings by combining a closed wet cooling tower with chilled ceilings, Faculty of Engineering, University of Porto, Publishable report, 1 December 1997 to 30 May 2000. Parker R.O., Treybal R.E., (1961), The Heat, Mass Transfer Characteristics of evaporative coolers, Chemical Engineering Progress Symposium Series, vol. 57, no. 32, pp 138 – 149, 1961 Roberson J., Homan G., Mahajan A., Nordman B., Webber C., Brown R., McWhinney M., Koomey J., (2002), Energy Use and Power Levels in New Monitors and Personal Computers, LBNL-48581, Energy Analysis Department, Environmental Energy Technologies Division, Lawrence Berkeley National Laboratory (LBNL), University of California, Berkeley USA, July 2002. Retrieved March 10, 2004, from LBNL home page < http://www.enduse.lbl.gov >.

157

Roberson J., Webber C., McWhinney M., Brown R., Pinckard M., Busch J., (2004), After-hours Power Status of Office Equipment and Inventory of Miscellaneous Plug-Load Equipment, LBNL-53729, Energy Analysis Department, Environmental Energy Technologies Division, Lawrence Berkeley National Laboratory (LBNL), University of California, Berkeley USA, January 2004. Retrieved March 10, 2004, from LBNL home page < http://www.enduse.lbl.gov >. Sahlin P., Bring A., Sowell E., (1989), A Neutral Model Format for Building Simulation Models, Proceedings of IBPSA Building Simulation ’89 conference, Vancouver, Canada 1989. Sahlin P., (1996), NMF Handbook – An Introduction to the Neutral Model Format, NMF version 3.02, ASHRAE RP – 839, report from Building Sciences, KTH, Stockholm, Sweden. 1996. Santamouris M. (coord.), (1995), Natural cooling techniques – Design methodology and application to southern Europe, Pascool Final Report, European Commission: Directorate General XII for Science Research and Development, 1995 Santamouris M., Asimakopoulos D., (ed), (1996), Passive cooling of Buildings, James & James (Science Publishers), London UK, 1996, ISBN: 1 873936 47 8 SCANVAC, (2000) , Klassindelade inneklimatsystem – Riktlinjer och specifikationer, SCANVAC, VVS-tekniska föreningen, Förlags AB VVS, Stockholm 2000. ISBN: 91973834-3-0 (in Swedish) Sodec, F. (1999), Economic viability of cooling ceiling systems, Energy and Buildings, 30 (1999) pp 195 – 201. Sprecher P., Tillenkamp F., (2003), Energy saving systems in building technology based on concrete-core-cooling, International Journal of Ambient Energy 24 (1) (2003) pp 29 – 34. Sprecher P., Borth J., Niessen R., (2000), Betriben von Kühldecken mit geschlossenem Nasskühlturm ohne Kältemaschine, Retrieved March 5, 2004, from The Energy Conservation in Buildings research, pilot and demonstration programme of the Swiss Federal Office of Energy Web site: < http://www.empa-ren.ch/InternetFiles/Programm/Aktuelles/aktualitaeten/Status-Seminar/tagungsband.htm > Energimyndigheten (2005) , Förbättrad energistatistik för lokaler – ”Stegvis STIL” Rapport för år 1, Energimyndigheten, 2005, (in Swedish) Stoecker W.F, (1998), Industrial refrigeration handbook, p. 266, McGraw-Hill 1998, ISBN 0-07-061623-X Stout M.R., Leach J.W., (2002), Cooling Tower Fan Control for Energy Efficiency, Journal of Energy Engineering, vol. 99, No. 1, pp 7 – 31, 2002 Stålbom G., Kling R., (2002), Legionella – Risker i VVS-installationer (Legionella – Risks in HVAC systems), VVS-Installatörerna, 2002. ISBN 91-631-2265-0 (in Swedish) Söylemez, M.S., (2004), On the optimum performance of forced draft counter flow cooling towers, Energy Conversion and Management 45 (2004) pp 2335 – 2341. Wang S., (2001), Handbook of air conditioning and refrigeration, Second edition, McGraw-Hill 2001 ISBN: 0-07-068167-8

158

References

Wetter M., (1998), Simulation model - Finned water-to-air coil without condensation, Simulation Research Group, Building Technologies Department, Environmental Energy Technologies Division, Lawrence Berkely National Laboratories, USA, LBNL-42355, November 1998 VVS 2000, (2003), VVS 2000 Tabeller & Diagram – Kapitel: Meteorologi och klimatologi, VVS Tekniska Föreningen, Förlags AB VVS, 2003, ISBN 91-973834-5-7 (in Swedish)

159

160

APPENDIX A

The Cooling Tower Model (CTM) in NMF code In this appendix the cooling tower model in NMF code is presented. For information about the Neutral Model Format, see Sahlin et al. (1989) and Sahlin (1996). CONTINUOUS_MODEL Cooling_Tower ABSTRACT "Cooling Tower INTRODUCTION This model describes a general device for cooling a liquid with ambient air. The model can be used for simulation of cooling towers, open and closed circuit type, wet or dry type but also any air/liquid heat exchanger with above mentioned purpose. The model implies that cooling is supported by one or several fans inducing forced draft. The model is based on the cooling tower model described in ASHRAE Primary HVAC Toolkit [1]. The ASHRAE model is based on the effectiveness-NTU method which makes it possible to simulate any heat exchanging device with the above mentioned purpose. The ASHRAE model is supplemented with different modes and different control strategies for part load control. This makes this model more general for different applications for cooling liquids with ambient air in HVAC systems. In the following text the cooling device is called 'cooling tower' for practical reasons.

GENERAL DESCRIPTION The model includes a freeze protected circuit, Circuit 1 (primary circuit), which is a closed loop between the cooling tower and a heat exchanger. The heat exchanger separates Circuit 1 from Circuit 2. Circuit 2 (secondary circuit) is the one supplying chilled liquid (water) to the air cooler in AHU and the cooling device(s) in the zone(s). The liquid in Circuit 1 is called Liquid 1 and subsequently the liquid in Circuit 2 is called Liquid 2. If the user wants to omit the freeze protected circuit the following parameters is set to: cpLiq1 = 4187 (heat capacity for water, which is normally used) rhoLiq1 = 999 (density for water, which is normally used) TApproachS = 0 (the same as no heat exchanger or one with infinite capacity)

MODES The cooling tower can be set to different modes by the user. - Mode_WetDry; The user can choose between wet or dry operation mode. The wet mode implies the liquid to be cooled (liquid 1) is sprayed over a cooling tower fill (open mode) or water spraying over the closed circuit tube bundles containing the liquid to be cooled (closed mode). - Mode_Fan; if the fan(s)is/are single speed or equipped with variable speed drive. The single speed mode is applicable when cooling tower has only one fan (or maybe two fans) with single speed. If the cooling tower has a two speed fan or several fans with single or two speed operation the user should choose the mode with variable speed drive. This mode simulates the case with multiple fans cycling on-off in sequence fairly well. If cooling tower is equipped with variable speed drive the choice of mode is obvious. - Mode_ClOp; for choice of open circuit or closed circuit cooling tower. The mode controls if spray water is on or off. This mode is added because the

161

combination of Closed circuit mode and Wet mode use spray water. All other combinations of Open/Closed and Wet/Dry modes use no additional spray water. - Mode_Tower; lets the user choose between a counterflow configuration or a cross flow configuration. The cross flow configuration assumes both streams unmixed. CONTROL The cooling tower has three different ways of control. - First way of control is a simple on/off control of the whole tower via an external signal from a time schedule through a link (Tower_control). - Secondly the leaving liquid 2 temperature is controlled in two sequences; First by modulating the massflow of air through the fan(s). This is done via an external signal from an external controller through a link (Fancontrol). In single speed mode the real life cycling on-off is simulated as a linear relation of the mass flow from a minimum flow (15% of max.) up to maximum flow. In variable speed mode the fan laws are applied controlling the mass flow. - Last step is to cycle the fan(s) on and off when the ambient air temperature is below the limit for modulating the fan(s).

USER INPUT The user must give the following design stage parameters as input: Input parameters affecting size and cooling capacity of the cooling tower: QTower_d T_ApproachP TRange_P

'Design cooling capacity in tower (in kW)' 'Temperature difference between inlet air (TAir_in_d) and outlet liquid 1 (TLiq1Cold_d) in primary circuit at design stage' 'Difference between inlet liquid 1 (TLiq1Warm_d) and outlet liquid 1 (TLiq1Cold_d) in primary circuit at design stage'

TAir_indb_d RelHum_d

'Design inlet dry bulb air temperature' 'Design relative humidity of ambient air (%)'

MLiq1MAir_d

'Massflow relation of liquid 1 and Air (MLiq1_d/MAir_d).Normally in range 0.5 to 2'

Input parameters primarily affecting energy efficiency of the cooling tower: EtaFan_d dpFan_d dpLiq1 dpLiq2 dpSpray

'Fan efficiency (in tower) at design air flow rate' 'Pressure difference over fan at design rate massflow of air in tower' Pressure difference in circuit 1 (liquid 1), assumed constant' Pressure difference in circuit 2 (liquid 2 passing heat exchanger), assumed constant' 'Pressure difference in spray water circuit (open) assumed constant'

Mode_Fan

'Control mode; 0=Fan with variable speed control, 1=Single speed fan'

EtaLiq1 EtaLiq2 EtaSpray

'Total pump efficiency in circuit with liquid 1' 'Total pump efficiency in circuit with liquid 2' 'Total pump efficiency in (open) circuit with spray water'

Miscellaneous input parameters: EtaHex Min_MAir

'Heat exchanger efficiency (between circuit 1 and 2)' 'Min. airflow in fraction of design rate massflow of air in tower. When using variable speed drive Min_MAir is normally in range 0.1 to 0.2'

cpLiq1 rhoLiq1

'Liquid 1 specific heat (default 3685= 40% propylene glycol at 15°C)' 'Liquid 1 density (default 1039= 40% propylene glycol at 15°C)'

Mode_WetDry Mode_Spray Mode_Tower

'Control mode; 0=Wet mode, 1=Dry mode' 'Control mode; 0=Closed circuit tower, 1=Open circuit tower' 'Control Mode; 0=Counterflow tower, 1=Crossflow tower'

LIMITATIONS - The model does not calculate the consumption of makeup water for evaporation, drift losses and blow down. - Hybrid cooling towers with combination of, or alteration between, wet and dry mode

162

Appendix A

can not be simulated in this model. - Natural draft cooling towers can not be simulated in this model. - The model works with good accuracy between -20 to 40 C wet bulb temperature of ambient air. It should not be used outside this temperature span. Refereces [1] ASHRAE Primary HVAC Toolkit [2] ASHRAE/IESNA Standard 90.1-1999 [3] Simulation model - Finned water-to-air coil without condensation, LBNL-42355, 1998 (model in SPARK) [4] Stoecker W.F, Industrial refrigeration handbook, p. 266, McGraw-Hill 1998 [5] Lebrun J., Silva C., Cooling Tower model and experimental validation, ASHRAE Trans. 2002 Date:

June 19, 2003

By:

Bengt Bergsten CIT Energy Management

Call:

Enthal from PSYCHRO1.NMF Humrat from PSYCHRO2.NMF Wetbulb from PSYCHRO3.NMF

Revisions: 030625 BB

Mode with both counterflow and cross flow configuration added

030826 BB

Equation for calculating hAir_out changed for better numerical stability when calculating cpAir_e

030828 BB

Original equation for cpAir_e replaced by curve fit equation. Original equation not numerically stable around 0 deg C.

031012 BB

A tank (thermal mass) is added. Event Fan_on removed. Better stability.

080909 BB

Minor changes in Eff-NTU equations + thermal mass in MAir. ”

EQUATIONS /**************** Psychrometics of ambient air ***********************/ /* calculate wet and dry bulb temperatures of entering air stream */ TAir_wb := Wetbulb(TAir, HumAir) ; TAir_inwb := IF LINEARIZE (1) THEN TAir - 3 ELSE /* limits due to cpAir_e regression equation */ IF TAir_wb < -20 THEN -20 ELSE_IF TAir_wb > 40 THEN 40 ELSE TAir_wb END_IF END_IF ; TAir_indb := TAir ;

/********************* Control 1 ******************************/ /* Control of leaving liquid 2 temperature, TOut, via air flow in cooling tower fan. The control is managed by an external controller in IDA ICE Primary systems */ TLiq2Out_r = TLiq2Out_set ;

163

/* Further control equations is placed under the heading "Control 2" */

164

Appendix A

/* Limiting the incoming control signal */ Ctrl := IF LINEARIZE (1) THEN 0.5 ELSE IF Ctrl_In > 1 THEN 1 ELSE_IF Ctrl_In < 0 THEN 0 ELSE Ctrl_In END_IF END_IF ; /* When Fan is on, the air flow is regulated as below depending if fan is in variable speed mode or single speed mode. In single speed mode the real life control strategy by cycling on-off is simulated as a linear relation of MAir from MAir_d to 15% of MAir_d [2]. For fan with variable speed control MAir is linear between MAir_d and a minimum flow equal to Min_MAir * MAir_d (Min_MAir is typically 0.1 to 0.2) */ MAir + 60*MAir' = IF NINT(TowerOn) == 0 /* Tower is off */ THEN MAir_0 * MAir_d /* A minimum level for numerical stability reasons */ ELSE MAir_d * Ctrl + MAir_0 * MAir_d * (1 - Ctrl) END_IF ; /* MLiq1 is regulated in the same way as MAir. Minimum flow is set to 30% of MLiq1_d. For an open cooling tower the minimum flow must not be to small so the distribution of the cooling water is sufficient on the cooling tower fill. */ MLiq1 := IF NINT(TowerOn) == 0 /* Tower is off */ THEN MLiq1_0 * MLiq1_d /* A minimum level for numerical stability reasons */ ELSE MLiq1_d * Ctrl + MLiq1_0 * MLiq1_d * (1-Ctrl) END_IF ;

/********************** Miscellaneous ***********************/

/* Pressure difference. Inlet pressure is given as reference pressure for massflow circuit. This corresponds to the grounding of an electrical circuit, or an expansion vessel for a fluid circuit. The outlet pressure is at present a S_P parameter (dpLiq2) */ pLiqOut = dpLiq2 ; pLiqIn = 0 ; /* total massflow in liquid 2 (help variable) going through the heat exchanger. */ MLiq2 := M1 + M2 ; /* Entering liquid 2 average temp. */ TLiq2In := IF MLiq2 <= 0 THEN (TIn1 + TIn2)/2 ELSE (M1 * TIn1 + M2 * TIn2) / MLiq2 END_IF ;

165

/************************** Cooling tower ******************************/ /* Equations based on model in ref. [1]. A dry operation mode is added */

/* Help variables for equation cpAir_e */ Td := ABS(TAir_outwb - TAir_inwb) ; /* difference of in- and outgoing wet bulb air temp */ Tm := ((TAir_outwb + TAir_inwb)/2) + 21 ; /* mean value of in- and outgoing wet bulb air temp */ /* The effective specific heat of air in cooling tower. */ /* The original equation from ref.[1] cpAir_e = (hAirOut - hAirIn)/(TAir_outwb - TAir_inwb) is replaced by a curve fit equation where cpAir_e = f(Tm, Td) of numerical reasons. The curve fit equation has a relative error <1% between -15= 0.5 THEN cpAir /* Cooling Tower in dry mode */ ELSE_IF LINEARIZE (1) THEN cpAir + 1810 ELSE 6.56473E-4*ABS(Tm)*Tm**3 - 0.0351*Tm**3 + 1.65114*Tm**2 + 1196.30745 + 1.84999E-4*(Td**2)*(Tm**2) + 2.09962E-11*(ABS(Td)*Td**3)*(ABS(Tm)*Tm**3) /* Cooling Tower in wet mode */ END_IF ; /* Calculating the UA value at part load */ /*The power law relation between UA and MAir/MAir_d and MLiq1 / MLiq1_d with corresponding exponents n2 and n3 is a correlation based on ref. [3] and [5] The presented values of n2 and n3 from other references in [3] and [5] shows a big variation. With n2=0.65 and n3=0.43 as default values they represent mean values of n2 and n3 from the ones presented in [3] and [5]. */ UA := IF LINEARIZE (1) THEN UA_d * (MAir / MAir_d) * (MLiq1 / MLiq1_d) ELSE_IF MAir <= 1E-3 THEN UA_d * (MAir / MAir_d) * (MLiq1 / MLiq1_d) ELSE UA_d * (MAir / MAir_d)**n2 * (MLiq1 / MLiq1_d)**n3 END_IF ; /* The effective heat transfer coefficient-area product */ UA_e := UA * cpAir_e / cpAir ; /* Heat capacity flow of air, liquid 1 and liquid 2 */ CLiq1 := cpLiq1 * MLiq1 ; CLiq2 := cpLiq2 * MLiq2 ; CAir := cpAir_e * Mair ; /* Smallest and biggest heat capacity flow */ CMin := max(1E-6, min(Cliq1, CAir)); CMax := max(CLiq1, CAir) ; /* Relation Cmin/Cmax */ CRatio := IF CMax < 1E-6 THEN 1 ELSE max(1E-4, CMin) / CMax END_IF ;

166

Appendix A

/* Number of transfer units in tower */ NTU := IF CMin <= 0 THEN 1E-4 ELSE UA_e / CMin END_IF ; /* The effectiveness of the heat exchanger */ k := NTU**(-0.22) ; /* help variable in cross flow equation */ Eff := IF LINEARIZE (1) THEN 0.5 * NTU ELSE_IF Mode_Tower < 0.5 /* Counterflow configured cooling tower: */ THEN (1 - EXP(- NTU * (1 - CRatio))) / (1 - CRatio * EXP(- NTU * (1 - CRatio))) /* Crossflow configured cooling tower: */ ELSE 1 - EXP((EXP(-NTU * CRatio * k) - 1 ) / (CRatio * k)) END_IF ; /* Water-air heat transfer rate in cooling tower */ QTower := IF NINT(TowerOn) == 0 /* Cooling Tower is off */ THEN 0 ELSE_IF Mode_WetDry < 0.5 /* Cooling Tower is on and in wet mode: */ THEN Eff * CMin * (TLiq1Warm - TAir_inwb) / 1000 /* Cooling Tower is on and in dry mode: */ ELSE Eff * CMin * (TLiq1Warm - TAir_indb) / 1000 END_IF ; /* Outlet air dry/wet bulb temperature from tower */ TAir_outdb = IF LINEARIZE (1) THEN TAir + 2 ELSE_IF Mode_WetDry < 0.5 /* Cooling Tower is in wet mode */ THEN TAir_outwb /* is good approx. when state of air is close to saturated */ ELSE TAir_indb + (QTower * 1000 / CAir) /* Cooling Tower is in dry mode */ END_IF ; TAir_outwb = IF LINEARIZE (1) THEN TAir + 2 ELSE_IF Mode_WetDry < 0.5 /* Cooling Tower is in wet mode */ THEN TAir_inwb + (QTower * 1000 / CAir) ELSE Wetbulb(TAir_outdb, HumAir) /* Cooling Tower is in dry mode */ END_IF ; /* Outlet liquid 1 (cold side) temperature from tower */ TLiq1Cold := TLiq1Warm - (QTower * 1000 / CLiq1) ; /********************* Heat exchanger and tank ******************************/ /* Outlet temperature of liquid 2 (help variable) from heat exchanger between liquid 1 and liquid 2 */ TLiq2Out = IF LINEARIZE (1) THEN TAir + 4 ELSE IF NINT(TowerOn) == 0 /* Cooling Tower is off */ THEN TLiq2In ELSE_IF CLiq2 <= CLiq1 THEN TLiq2In - EtaHex * (TLiq2In - TLiq1Cold) ELSE TLiq2In - (CLiq1 / CLiq2) * EtaHex * (TLiq2In - TLiq1Cold) END_IF END_IF ;

167

/* Outlet temperature of liquid 1 (TLiq1Warm) from heat exchanger between liquid 1 and liquid 2 */ TLiq1Warm = IF LINEARIZE (1) THEN TAir + 3 ELSE TLiq1Cold + (CLiq2 / CLiq1) * (TLiq2In - TLiq2Out) END_IF ; /* Tank after heat exchanger in secondary circuit */ T_TankIn := IF CTRL_T_TankIn < 0.5 THEN IF TLiq2Out <= TLiq2Out_set THEN TLiq2Out_set ELSE TLiq2Out END_IF ELSE IF NINT(Warm) == 0 /* check for not heating the liquid: */ THEN TLiq2In ELSE_IF NINT(Cold) == 0 /* check the liquid is not below set point temp: */ THEN min(TLiq2Out_set, TLiq2In) ELSE TLiq2Out /* TLiq2Out <= TLiq2In OR TLiq2Out >= TLiq2Out_set: OK! */ END_IF END_IF ; M_Tank * T_Tank' = MLiq2 * (T_TankIn - T_Tank) ;

/********************* Use of energy ******************************/ /* Speed of fan with variable speed control */ n_speed := IF MAir < MAir_0 * MAir_d * 1.01 THEN 0 /* Fan(s) assumed to be off */ ELSE_IF MAir < Min_MAir * MAir_d THEN n_speed_d * Min_MAir * MAir_d / Mair_d ELSE n_speed_d * MAir / Mair_d END_IF ; /* Pressure difference in cooling tower fan (through fan laws) */ dpFan := IF Mode_Fan > 0.5 /* Fan with single speed mode: */ THEN dpFan_d /* Fan with variable speed control mode: */ ELSE_IF LINEARIZE (1) OR (n_speed / n_speed_d) < 1E-3 THEN dpFan_d * (n_speed / n_speed_d) ELSE dpFan_d * (n_speed / n_speed_d)**2 /* according to fan laws */ END_IF ; /* Total fan efficiency as a function of fan speed. */ /* The relation fan efficiency - fan speed is chosen as a power law expression which is an approximate relation for the fact that total fan efficiency is not constant when a variable speed drive is used. The total energy usage in the tower is dominated by energy supplied to the fan(s). For single speed fan 'EtaFan' is constant */ EtaFan := IF Mode_Fan > 0.5 /* Fan with single speed mode: */ THEN EtaFan_d /* Fan with variable speed control mode: */ ELSE_IF LINEARIZE (1) THEN EtaFan_d * (MAir / Mair_d) ELSE_IF MAir < Min_MAir * MAir_d THEN EtaFan_d * (Min_MAir * MAir_d / Mair_d)**n1 ELSE EtaFan_d * (MAir / Mair_d)**n1 END_IF ;

168

Appendix A

/* Electrical power for fan in cooling tower */ PFan := IF NINT(TowerOn) == 0 OR MAir < MAir_0 * MAir_d * 1.01 /* Cooling Tower is off OR Fan(s) is off */ THEN 0 ELSE_IF Mode_Fan < 0.5 /* Fan with variable speed control mode: */ THEN dpFan * MAir / (EtaFan * rhoAir) /* Fan with single speed mode: */ ELSE_IF MAir < 0.15 * MAir_d THEN 0 ELSE (MAir * dpFan_d / (0.85 * EtaFan * rhoAir)) - (0.15 * Mair_d * dpFan_d / (0.85 * EtaFan * rhoAir)) /* This linear relation corresponds to relation in [2] */ END_IF ; /* Electrical power for pump in circuit 1 between cooling tower and heat exchanger, i.e. liquid 1 */ PLiq1 := IF NINT(TowerOn) == 0 OR MLiq1 < MLiq1_0 * MLiq1_d * 1.01 /* Cooling Tower is off */ THEN 0 ELSE dpLiq1 * MLiq1 / (EtaLiq1 * rhoLiq1) END_IF ; /* Electrical power for pumping liquid 2 through heat exchanger in circuit 2 */ PLiq2 := IF NINT(TowerOn) == 0 /* Cooling Tower is off */ THEN 0 ELSE dpLiq2 * MLiq2 / (EtaLiq2 * rhoLiq2) END_IF ; /* Electrical power for pump in water spray sump. */ /* Equation is based on approximate relation between MSpray and QTower [4] which is 0.018 l/s per kW cooling capacity in tower */ VSpray := IF Mode_WetDry < 0.5 AND Mode_ClOp < 0.5 /* Cooling Tower is in wet mode AND closed circuit tower mode */ THEN (0.018 / 1000) * QTower_d /* QTower_d in kW */ ELSE 0 /* All other combinations of Mode_WetDry and Mode_ClOp */ END_IF ; PSpray := IF NINT(TowerOn) == 0 /* Cooling Tower is off */ THEN 0 ELSE dpSpray * VSpray / EtaSpray END_IF ; /* COP for cooling tower */ COP := IF NINT(TowerOn) == 0 /* Cooling Tower is off */ THEN 0 ELSE QTower * 1000 / (PFan + PLiq1 + PLiq2 + PSpray) END_IF ; /*Total electrical power demand for cooling tower */ PTot := IF NINT(TowerOn) == 0 /* Cooling Tower is off */ THEN 0 ELSE (PFan + PLiq1 + PLiq2 + PSpray) END_IF ;

169

/********************* Control 2 ******************************/ /* Operating mode TowerOn */ TowerOn := IF Event(G2, TowerCtrl) > 0.5 THEN 1 ELSE 0 END_IF ; /* Control of leaving liquid 2 temperature, TOut, via air flow in cooling tower fan */ TOut = IF LINEARIZE (1) THEN TAir + 4 ELSE T_Tank END_IF ; /* check for not heating the liquid at T_TankIn : */ Warm := IF Event(G1, TLiq2Out - TLiq2In) > 0 THEN 0 /* TLiq2Out > TLiq2In : Not OK! */ ELSE 1 END_IF ; /* check for not cooling the liquid beneath set point temp. at T_TankIn: */ Cold := IF Event(G3, TLiq2Out - TLiq2Out_set) < 0 THEN 0 /* TLiq2Out < TLiq2Out_set : Not OK! */ ELSE 1 END_IF ; /* The rest of control equations are placed under the heading "Control 1" */

/****************************** End of Equations ************************************************/

LINKS /* type T W

name T_Ambient W_Ambient

variables .... */ TAir ; /* ambient outdoor air temperature */ HumAir ; /* ambient outdoor absolute humidity */

PMT PMT

Inlet1 Inlet2

PLiqIn, POS_IN M1, TIn1 ; /* from AHU */ PLiqIn, POS_IN M2, TIn2 ; /* from Zone(s) */

PMT PMT

Outlet1 Outlet2

PLiqOut, POS_OUT M1Out, TOut ; /* to AHU */ PLiqOut, POS_OUT M2Out, TOut ; /* to Zone(s) */

ControlLink ControlLink

Fan_Control Tower_Control

Ctrl_In ; TowerCtrl ;

T T

Temp_Setpoint Temp_Measure

TLiq2Out_r ; /* Liquid 2 leaving temperature set point */ TLiq2Out ; /* Liquid 2 leaving temperature */

VARIABLES /* type Pressure Pressure Pressure

name pLiqOut pLiqIn dpFan

role OUT OUT LOC

def 15000 0 75

min 0 0 SMALL

max BIG BIG BIG

description */ "Outlet liquid 1 pressure" "Inlet liquid 1 pressure" "Pressure difference over fan"

MassFlow

M1

IN

0.30

0

BIG

MassFlow

M2

IN

0.15

0

BIG

"Inlet massflow of liquid from air cooler in AHU" "Inlet massflow of liquid from room cooler(s) in zone(s)"

170

Appendix A

MassFlow

M1Out

IN

0.30

0

BIG

MassFlow

M2Out

IN

0.15

0

BIG

MassFlow MassFlow MassFlow

MLiq1 MLiq2 Mair

LOC LOC LOC

0.45 0.45 0.89

SMALL BIG SMALL BIG SMALL BIG

VolFlow

VSpray

LOC

1.4E-4

SMALL BIG

"Volume flow of spray water used in wet and closed mode"

Temp Temp Temp Temp Temp Temp Temp Temp Temp Temp Temp Temp Temp Temp Temp Temp Temp Temp

TAir TOut TIn1 TIn2 TLiq2Out TLiq2In TLiq1Cold TLiq1Warm TAir_indb TAir_outdb TAir_inwb TAir_outwb TLiq2Out_r Td Tm T_Tank T_TankIn TAir_wb

IN OUT IN IN OUT LOC LOC OUT LOC OUT LOC OUT OUT LOC LOC OUT LOC LOC

15 19 22 22 19 22 17.5 20.5 15 16 10 16 16 4 13 19 19 10

-30 -30 -30 -30 -30 -30 -30 -30 -30 -30 -20 -20 -20 SMALL -20 -30 -30 -20

50 50 50 100 100 100 100 100 50 50 40 40 100 50 60 100 100 40

"Ambient air dry bulb temperature" "Final outlet liquid 2 temperature" "Inlet temp. of liquid from air cooler" "Inlet massflow of liquid from room cooler" "Outlet 2 liquid temperature" "Average temp. of inlet liquid 2" "Cold side liquid 1 temp." "Warm side liquid 1 temp." "Inlet dry bulb temp. of amb. air (=TAir)" "Outlet dry bulb temp. from tower" "Inlet wet bulb temp of amb. air" "Outlet wet bulb temp. from tower" "Outlet liquid 2 requested temperature" "Wet bulb temperature difference" "Wet bulb mean temperature" "Temperature in Liquid 2 tank" "Temperature entering Liquid 2 tank" "Help variable: Ambient wet bulb temperature"

HumRatio

HumAir

IN

0.006

0

BIG

"Humidity ratio ambient air (kg water/kg dry air)"

Control Control Control

TowerCtrl Ctrl_In Ctrl

IN IN LOC

1 0.5 0.5

0 0 0

1 1 1

Generic Generic

Warm Cold

A_S A_S

1 1

0 0

1 1

Generic Generic Generic Generic

TowerOn G1 G2 G3

A_S A_S A_S A_S

1 1 1 1

0 -BIG -BIG -BIG

1 BIG BIG BIG

"Control signal; 0 = Tower Off 1 = Tower On" "Fan control in signal, help variable" "Fan control, 0 = Min. massflow 1 = Max. massflow " "Is liquid 2 heated? 0 = Yes, 1 = No" “Is liquid 2 cooled beneath set point? 0 = Yes, 1 = No" "Tower operating mode 0 = Off, 1 = On" "G-stop Warm" "G-stop TowerOn" "G-stop Cold"

Factor Factor

Eff UA_e

LOC LOC

0.56 925

SMALL 1 SMALL BIG

Factor Factor Factor Factor Factor Factor Factor Factor Factor Factor

UA Cratio CMin CMax CLiq1 CLiq2 CAir NTU COP n_speed

LOC LOC LOC LOC LOC LOC LOC LOC LOC LOC

350 0.55 924 1666 1666 1884 924 1 5 0.5

SMALL SMALL SMALL SMALL SMALL SMALL SMALL SMALL SMALL SMALL

Factor Factor

EtaFan k

LOC LOC

0.42 0.6

SMALL 1 SMALL BIG

BIG 1 BIG BIG BIG BIG BIG BIG BIG BIG

"Inlet massflow of liquid to air cooler in AHU" "Inlet massflow of liquid to room cooler(s) in zone(s)" "Massflow of liquid 1" "Total massflow of liquid 2, help variable" "Actual massflow of air in tower"

"Tower efficiency" "Effective heat transfer coeff.-area product in tower" "Heat transfer coeff.-area product in tower" "Relation Cmin/Cmax" "CMin = min ( CLiq1, CAir )" "CMin = max ( CLiq1, CAir )" "Heat capacity flow of liquid 1" "Heat capacity flow of liquid 2" "Heat capacity flow of air " "Number of transfer units in tower" "Overall coefficient of performance (efficiency)" "Fan speed (rpm) at massflow rate of air in tower" "Fan efficiency (in tower) as a function of MAir" "Help variable when calculating Eff for a cross flow tower"

171

HeatFlux_k

Qtower

LOC

5

SMALL BIG

"Heat transfer rate in tower (in kW)"

HeatCapM

cpAir_e

LOC

2657

SMALL BIG

"Effective specific heat of air in tower "

ElPowerCons ElPowerCons ElPowerCons ElPowerCons ElPowerCons

Pfan PLiq1 PLiq2 Pspray Ptot

LOC LOC LOC LOC LOC

52 9 3 2.2 63

SMALL SMALL SMALL SMALL SMALL

BIG BIG BIG BIG BIG

"Supplied electric power to cooling tower fan" "Supplied electric power to pump in liquid 1" "Supplied electric power to pump in liquid 2" "Supplied electric power to spray water pump" "Total supplied electrical power to cooling tower"

role S_P C_P

def 0.6 0.67

min 0.01 0.01

max 1 1 1 1 1

description */ "Fan efficiency (in tower) at design air flow rate" "Heat exchanger efficiency (between circuit 1 and 2)" "Total pump efficiency in circuit with liquid 1" "Total pump efficiency in circuit with liquid 2" "Total pump efficiency in (open) circuit with spray water"

PARAMETERS /* type name Factor EtaFan_d Factor EtaHex Factor Factor Factor

EtaLiq1 EtaLiq2 EtaSpray

S_P S_P S_P

0.5 0.5 0.5

0.01 0.01 0.01

Factor

UA_ed

C_P

1928

SMALL BIG

Factor

UA_d

C_P

531

SMALL BIG

Factor Factor Factor Factor

CAir_d CLiq1_d CLiq2_d MLiq1MAir_d

C_P C_P C_P S_P

2543 2563 2912 1.0

SMALL SMALL SMALL 0.01

Factor

Min_MAir

S_P

0.1

SMALL 1

Factor

MAir_0

S_P

0.02

SMALL 1

Factor

MLiq1_0

S_P

0.02

SMALL 1

Factor

n_speed_d

S_P

1

SMALL BIG

Factor

n1

S_P

0.5

0.01

1

Factor

n2

S_P

0.65

0.01

1

Factor

n3

S_P

0.43

0.01

1

Pressure Pressure

pAtm pw_d

S_P C_P

101325 SMALL BIG 1024 SMALL BIG

Pressure

dpFan_d

S_P

300

SMALL BIG

Pressure Pressure

dpLiq1 dpLiq2

S_P S_P

10000 15000

SMALL BIG SMALL BIG

Pressure

dpSpray

S_P

15000

SMALL BIG

HeatCapM

cpLiq1

S_P

3685

500

5000

Density

rhoLiq1

S_P

1039

500

2000

HeatCapM Density HeatCapM HeatCapM

cpLiq2 rhoLiq2 cpAir cpAir_ed

S_P S_P S_P C_P

4187 999 1005 3656

500 500 900 SMALL

5000 2000 1012 BIG

172

BIG BIG BIG 10

"Design rate of effective heat transfer coeff.-area product in tower" "Design rate of heat transfer coeff.-area product in tower" "Design rate of heat capacity flow of air " "Heat capacity flow of liquid 1 at design rate" "Heat capacity flow of liquid 2 at design rate" "Massflow relation of Liquid 1 and Air (MLiq1_d/MAir_d). Normally in range 0.5 to 2" "Min. airflow when calculating electricity use in fan(s)" "Min. airflow in fraction of design rate mass flow of air in tower" "Min. liquid 1 flow in fraction of design rate liquid 1 flow" "Reference fan speed (rpm) at design rate mass flow of air in tower" "Exponent in power law relation between EtaFan and MAir" "Exponent in power law relation between UA and MAir" "Exponent in power law relation between UA and MLiq1" "Ambient air pressure" "Design rate partial vapour pressure of ambient air" "Pressure difference over fan at design rate mass flow of air in tower" "Pressure difference in circuit 1 (liquid 1)" "Pressure difference in circuit 2 (liquid 2 passing heat exchanger)" "Pressure difference in spray water circuit (open circuit)" "Liquid 1 specific heat (3685= 40% propylene glycol at 15°C)" "Liquid 1 density (1039= 40% propylene glycol at 15°C)" "Liquid 2 specific heat" "Liquid 2 density" "Ambient air specific heat" "Design effective specific heat of air in tower "

Appendix A

Density

rhoAir

S_P

1.19

0.5

2

MassFlow MassFlow MassFlow

MAir_d MLiq1_d MLiq2_d

C_P C_P C_P

0.7 0.7 0.7

SMALL BIG SMALL BIG SMALL BIG

"Design rate massflow of air in tower" "Design rate massflow of liquid 1" "Design rate massflow of liquid 2"

Temp Temp Temp Temp Temp Temp

TLiq2Out_set TAir_indb_d TAir_inwb_d TAir_out_d TAir_outwb_d TAir_in_d

S_P S_P C_P C_P S_P C_P

16 25 19.5 22.5 22.5 19.5

-20 -20 -20 -20 -20 -20

100 40 40 40 40 40

Temp Temp

TLiq1Cold_d TLiq1Warm_d

C_P C_P

23.5 26.5

-20 -20

100 100

"Liquid 2 outlet set temperature" "Design inlet dry bulb air temperature" "Design inlet wet bulb air temperature" "Given value of TAir_out_wb " "Help: Design outlet air wet bulb temp." "Help: Design inlet air wet/dry bulb temp. to tower" "Design cold side liquid 1 temp." "Design warm side liquid 1 temp."

Temp

T_ApproachP

S_P

4

0.1

30

Temp

TRange_P

S_P

3

0.1

30

Temp

TApproachS

S_P

1.5

0.1

30

Temp

TRange_S

S_P

3

0.1

30

Temp

dT_Ln

C_P

4

0.1

BIG

HumRatio

HumAir_d

C_P

0.012

SMALL BIG

Factor

RelHum_d

S_P

60

SMALL 100

"Design humidity ratio ambient air (kg water/kg dry air)" "Design relative humidity of ambient air (%)"

Enthalpy Enthalpy

hAirIn_d hAirOut_d

C_P C_P

55446 66501

-BIG -BIG

"Design enthalpy inlet (=ambient) air" "Design enthalpy outlet air"

Mass

M_Tank

S_P

100

SMALL BIG

"Mass of liquid in tank and material in system"

HeatFlux_k HeatFlux_k

QTower_d QCooling_d

S_P S_P

7.69 7.69

SMALL BIG SMALL BIG

"Design cooling capacity in tower (in kW)" "Design cooling load (in kW)"

Generic

Mode_WetDry

S_P

0

0

1

Generic

Mode_ClOp

S_P

0

0

1

Generic

Mode_Fan

S_P

0

0

1

Generic

Mode_Tower

S_P

0

0

1

Generic

CTRL_T_TankIn S_P

0

0

1

"Control Mode; 0=Wet Mode 1=Dry Mode" "Control Mode; 0=Closed circuit tower 1=Open circuit tower" "Control Mode; 0=Fan with variable speed control 1=Single speed fan" "Control Mode; 0=Counterflow tower 1=Crossflow tower, both streams unmixed" ""

BIG BIG

"Ambient air density"

"Design difference between TAir_in_d and TLiq1Cold_d in primary circuit" "Design difference between TLiq1Warm_d and TLiq1Cold_d in primary circuit" "Design difference between TLiq1Cold_d and TLiq2Out in secondary circuit" "Design difference between TLiq2Out and TLiq2In in secondary circuit" "Log-mean temperature difference between liquid 1 and air"

173

PARAMETER_PROCESSING /**** Check for impossible comb. of Mode_WetDry (wet/dry) and Mode_ClOp (Closed/Open) *****/ IF Mode_WetDry > 0.5 AND Mode_ClOp > 0.5 THEN CALL NMF_ERROR ("Combination of dry Mode and open circuit Mode impossible!") END_IF ; /******************** Psychrometics at design stage *********************/ /* Ambient air partial vapour pressure, pw (Pa) and humidity ratio, HumAir_d (kg water/kg dry air) */ pw_d := RelHum_d * Satpres(TAir_indb_d) / 100 ; HumAir_d := Humrat(pAtm, pw_d) ; /* calculate wet bulb temperature of entering air stream at design stage */ TAir_inwb_d := Wetbulb(TAir_indb_d, HumAir_d) ; /* Design inlet air wet bulb/dry bulb temperature from tower */ TAir_in_d := IF Mode_WetDry < 0.5 THEN TAir_inwb_d /* Cooling Tower is in wet mode */ ELSE TAir_indb_d /* Cooling Tower is in dry mode */ END_IF ; /* enthalpies of entering and leaving air streams at design stage (Cooling Tower in wet mode) */ hAirIn_d := Enthal(TAir_indb_d, HumAir_d) ; hAirOut_d := 9362.5 + 1786.1 * TAir_outwb_d + 11.35 * TAir_outwb_d**2 + 0.98855* TAir_outwb_d**3 ; /* hAirOut = f (TAir_outwb_d) (ref [1]) is a good approximation as long as humidity ratio in leaving air is fairly close to saturation curve. This is normally the case when cooling tower is in wet mode. */ /************ Calculation of UA_d and MLiq1_d at design rate***************/ /* Equations based on model in ref. [1] */ /* The design effective specific heat of air i cooling tower */ cpAir_ed := IF Mode_WetDry < 0.5 /* Cooling Tower in wet mode */ THEN (hAirOut_d - hAirIn_d) / (TAir_outwb_d - TAir_inwb_d) ELSE cpAir /* Cooling Tower in dry mode */ END_IF ; /* Design outlet liquid 1 (cold side) temperature from tower in primary circuit */ TLiq1Cold_d := TAir_in_d + T_ApproachP ; /* Design inlet liquid 1 (warm side) temperature to tower in primary circuit */ TLiq1Warm_d := TLiq1Cold_d + TRange_P ; /* The design mass flow of liquid 1 and 2 */ MLiq1_d := QTower_d * 1000 / (cpLiq1 * (TLiq1Warm_d - TLiq1Cold_d)) ; MLiq2_d := MLiq1_d * QCooling_d * cpLiq1 * TRange_P / (QTower_d * cpLiq2 * TRange_S) ; /* Design volume flow of air to massflow of air */ Mair_d := MLiq1_d / MLiq1MAir_d ; /* Design heat capacity flow of air and liquid */

174

Appendix A

CAir_d := cpAir_ed * Mair_d ; CLiq1_d := cpLiq1 * MLiq1_d ; CLiq2_d := cpLiq2 * MLiq2_d ; /* Calculation of leaving air temperature. When Cooling Tower is in wet mode TAir_out_d = TAir_outwb_d which is a supplied parameter. The leaving wet bulb temperature has to be calculated iteratively elsewhere. A routine for calculating TAir_outwb_d should be added */ TAir_out_d := IF Mode_WetDry < 0.5 /* Cooling Tower is in wet mode: */ THEN TAir_outwb_d /* Cooling Tower is in dry mode: */ ELSE QTower_d + (cpAir * Mair_d * TAir_indb_d) / (cpAir * Mair_d) END_IF ; /* Log-mean temperature difference between liquid and air in primary circuit*/ dT_Ln := ((TLiq1Warm_d - TAir_out_d ) - (TLiq1Cold_d - TAir_in_d)) / LOG((TLiq1Warm_d - TAir_out_d) / (TLiq1Cold_d - TAir_in_d)) ; /* The design effective heat transfer coefficient-area product in primary circuit */ UA_ed := QTower_d * 1000 / dT_Ln ; /* The design heat transfer coefficient-area product in primary circuit */ UA_d := UA_ed * cpAir / cpAir_ed ;

/******* Calculation of efficiency of intermediate heat exchanger at design rate ********/ EtaHex := IF CLiq2_d < CLiq1_d THEN TRange_S / (TRange_S + TApproachS) ELSE (CLiq2_d / CLiq1_d) * TRange_S / (TRange_S + TApproachS) END_IF ; /*******************************************************************************************/ END_MODEL

175

APPENDIX B

Pilot plant at Kvarnberget, Gothenburg In this appendix the pilot plant at Kvarnberget, Gothenburg in Sweden is described

Heat exchanger (between freeze protected primary circuit and secondary circuit) A plate heat exchanger for free cooling, product Alfa Laval type CB76-100E or similar Design data: Heat transfer capacity 16,5 kW Warm side: Design volume flow, water, 1,3 l/s Design temperature in/out 25,6/22,6 °C Design pressure drop 10 kPa Cold side: Design volume flow, propylene glycol solution 40%, 1,45 l/s Design temperature in/out 21,6/24,6°C Design pressure drop 10 kPa

Evaporative cooler Air cooled heat exchanger with water spray function, product Asarums Industries model XP90-5 460 rpm-15, vertical direction of air. Design data: Rated cooling capacity 17 kW Air flow rated 8,61 m3/s Ambient air condition 23°C/70% RH Liquid flow 1,58 l/s prop. glycol 40 % Liquid temperature return 24,6°C Liquid temperature supply 21,6°C Pressure drop liquid 54 kPa Dri-Batic flow 1,35 l/s 3 Fans diam. 890 mm/ 450 rpm Motor spec. 3/230 V 0,34kW 1,64A supplied with variable speed drive. In the secondary circuit there are chilled beams which have a design ∆Tm of 8°C and a design specific cooling capacity of 50 W/m2 floor area. The circulation the primary and secondary circuits is maintained by pumps equipped with variable frequency drives (VFD). The control system strives to maintain a constant supply temperature in the secondary circuit. The constant supply temperature is maintained through sequential control where, in order and at increasing cooling demand, the pump in the primary circuit increase its speed, the fans in the free cooler increase their speed and finally the spray water is turned on. The control system also includes a function which disconnects the free cooling system and connects the conventional comfort cooling system, through a set of shut-off valves, if the indoor

175

APPENDIX B

temperature exceeds 25°C. This function was applied through demand from both the building owner Vasakronan and the tenant ÅF-konsult. The part of the building at which the pilot plant is installed is erected during the 1970’s with corresponding building standard. The office space which the free cooling system is serving has a floor area of 450 m2 with an outer wall facing south east. The people density is nominally 14 m2/person and at normal operation the density is about 16 – 20 m2/person. Design power density for lighting is 15 W/m2. There are no external sun shading devices.

176

APPENDIX C Monitoring system at Kvarnberget, Gothenburg In this appendix the monitoring system at Kvarnberget, Gothenburg in Sweden is described The monitoring of the pilot plant with this measurement system was carried out during the May 1 to the August 31, 2007. Specifications of the sensors used are listed at the end of this appendix. The following points where monitored: Outdoor Solar radiation: Measurement of total and global diffuse solar radiation with pyranometers. Sample interval is 30 seconds and samples are processed to one hour average figures. The solar radiation is measured on the roof of the office building at Kvarnberget Ambient air temperature: One sensor, placed on the roof. Sample interval 6 minutes and samples are processed to one hour average figures. Ambient air relative humidity: One sensor, placed on the roof. Sample interval 6 minutes and samples are processed to one hour average figures. Indoor Volume flow secondary circuit: Flow rate is measured with a volume flow meter at the office floor which is monitored. Pulses from the meter is sampled in 15 minutes interval and processed to one hour average figures of flow rate. Liquid temperature secondary circuit: Two sensors placed in the supply and return liquid coolant serving the monitored office floor. Sample interval is 15 minutes and samples are processed to one hour average figures. Indoor air temperature: Three sensors distributed to one in each “room” in the open office space. The sensors are mounted on an inner wall at 1,1 m above floor level. At one of the sensors there are two additional sensors which measure the air temperature at 10 cm below the false ceiling and 10 cm above floor level respectively to record the temperature gradient in the office space. Sample interval is 15 minutes and samples are processed to one hour average figures. Indoor air relative humidity: One sensor placed in one of the “rooms” in the open office space at 1,1 m above floor level. Sample interval is 15 minutes and samples are processed to one hour average figures. Supply air temperature: One sensor placed in the supply duct. Sample interval is 15 minutes and samples are processed to one hour average figures. Electric energy: Electric power is recorded in the distribution box serving the office space. Electric energy is integrated during 15 minutes and samples are processed to one hour average figures.

177

Free cooler Volume flow primary circuit: Flow rate is measured with a volume flow meter in the primary circuit. Pulses from the meter is sampled in 15 minutes interval and processed to one hour average figures of flow rate. The primary circuit has a freeze protected liquid (ethylene glycol at 40% concentration). Liquid temperature primary circuit: Two sensors placed in the supply and return of the freeze protected liquid close to the plate heat exchanger. Sample interval is 15 minutes and samples are processed to one hour average figures. Liquid temperature secondary circuit: Two sensors placed in the supply and return liquid coolant (water) serving the monitored office floor close to the plate heat exchanger. Sample interval is 15 minutes and samples are processed to one hour average figures. Electricity: Electric power to the fans in the free cooler and circulating pumps in primary and secondary circuits is measured. Pulses from the meter is sampled in 15 minutes interval and processed to one hour average figures of electric energy. Volume flow spray water: Flow rate is measured with a volume flow meter in the supply pipe of fresh water. Pulses from the meter is sampled in 15 minutes interval and processed to one hour average figures of flow rate.

List of sensors, office floor Sensor

Channel Description

Sensor

Analogue sensor REFRES GT01 GT02 GT03 GT04 GT05 GT06 GT07 GT08 GF09

A00 A01 A02 A03 A04 A05 A06 A07 A08 A09

Reference resistance Room temperature, south west Room temperature middle, ceiling Room temperature middle, intermediate Room temperature middle, floor Room temperature north east Temperature coolant, supply Temperature coolant, return Temperature supply air Rel. humidity, office

Pt100 Pt100 Pt100 Pt100 Pt100 Pt100 Pt100 Pt100 Vaisala

100 Ohm #1 #4 #5 #6 #2 #309046 #309050 kanal#2 A

Electric power Volume flow, coolant

Ackr. electric meter Krohne

Digital sensor EL00 FL01

D00 D01

Calculated data QE00

178

E00

Cooling power, secondary circuit

GT06, GT07, FL01

Appendix C

List of sensors, outdoor Sensor

Channel

Description

Analogue sensor REFRES A00 Reference resistance GTU01 A01 Outdoor temperature GFU02 A02 Outdoor relative humidity GST03 A03 Solar radiation, total GSD04 A04 Solar radiation, diffuse

Sensor

100 Ohm Pt100 Vaisala B Kipp&Zonen #850918 Kipp&Zonen #850870

Digital sensor None Calculated data None

List of sensors, evaporative free cooler Sensor

Channel

Description

Analogue sensor REFRES A00 Ref. resistance GTK01 A01 Temp. primary supply GTK02 A02 Temp. primary return GTK03 A03 Temp. secondary supply GTK04 A04 Temp. secondary return

Sensor 100 Ohm Pt100 Pt100 Pt100 Pt100

#309045 #309048 #309052 #309053

Digital sensor FLK01 FLK02 ELK01 ELK02 ELK03

D00 D01 D02 D03 D04

Volume flow, coolant Volume flow, spray water Electricity, fans Electricity, pump prim. circ. Electricity, pump sec. circ.

XK01

D05

Operational mode, cooling system

Krohne Krohne

Calculated data QEK00

E00

Cooling power, primary circuit

A01, A02, D00

179

180

Thesis Evaporative Cooling Towers

Recommend Documents