IEA48 Simulation Tools: Reference Book - · PDF fileAuthors: Stéphane ... o Sizing of...

IEA ECBCS 48 - Subtask 1.4 1

IEA – ECBCS Annex 48 Heat Pumping and Reversible Air conditioning

International energy Agency Energy Conservation in Buildings and Community Systems Programme

IEA48 Simulation Tools: Reference Book

Annex 48 of the International Energy Agency Energy conservation in buildings and Community Systems Programme

First Draft

29th March 09

Authors: Stéphane Bertagnolio*, Benjamin Soccal*, Pascal Stabat**

* Laboratoire de thermodynamique, Université de Liège, Belgique ** Ecole des Mines de Paris, France

Operating agents : Jean LEBRUN, Laboratoire de thermodynamique – Campus du Sart-Tilman B49 (P33) B-4000 Liège ([email protected]) Philippe ANDRE, Département des sciences de l’environnement – Avenue de Longwy, 185 B6700 Arlon ([email protected])

mailto:[email protected]

mailto:[email protected]


Table of contents

I. INTRODUCTION ........................................................................................................................................ 3

II. SPECIFICATIONS ...................................................................................................................................... 3

1. GENERAL REQUIREMENTS .................................................................................................................. 3

2. HEAT RECOVERY AND REVERSIBILITY POTENTIAL IDENTIFICATION ............................... 4

3. SIZING AND ECONOMIC/ENERGY ASSESSMENT TOOL ............................................................... 6

III. PROGRESS IN THE DEVELOPMENT OF THE TOOLS ................................................................ 8

1. SIMZONE (MULTI OR MONO ZONE)/AGGREGATE ........................................................................ 8

2. SIZING AND ASSESSMENT TOOL (PREVIOUSLY CALLED “GENERATION”) ......................... 9

APPENDIX 1: RECOVERY AND REVERSIBILITY POTENTIALS ......................................................... 39


I. Introduction The aim of the project is to promote efficient solutions of heat pumping in commercial buildings in order to save primary energy and reduce CO2 emissions. The project targets the retrofit of existing buildings and but also new buildings. The main goals of the first subtask of the Annex 48 project, Analysis of building heating and cooling demands and of equipment performances, are : 1. To evaluate the potential of heat recovery and reversible use of air conditioning systems, in order to save

energy and to reduce GES emissions; 2. To deliver simulation models for those systems; 3. To develop a simplified tool, which could be used by decision makers in order to evaluate the energy

and economic potential of these solutions for their buildings, while entering a limited number of easily accessible parameters.

The document presents the specifications expected from the tools to be developed in the framework of IEA Annex 48. It is proposed to develop two types of tools :

An heat recovery and reversibility potential assessment tool; This tool aims to help practitioners and decision makers in identifying the most ―interesting‖ application cases.

A sizing tool (this task is particularly linked with Task 2 on the design);

An energy and economic assessment tool of selected heat pump systems. This tool aims developed to allow quick and easy assessment of selected heat pump systems

II. Specifications

1. General requirements

The laboratory of Thermodynamics of University of Liège has developed building energy simulation tools under EES environment. It has been chosen to start the development of the IEA Annex 48 tools from these works. The tools are split in two :

A heat pumping/heat recovery potential identification;

A selection tool : o Sizing of heat pump systems; o Energy and Economic assessment.


2. Heat recovery and reversibility potential identification

Introduction The deliverable 1.2. aimed to define indexes in order to identify and quantify the reversibility and heat recovery potentials. Two possibilities have been first considered :

To try to identify the investigated building among the database of simulated buildings (deliverable 1.1);

To use indexes based on energy bills (electricity and gas/fuel) and/or installed cooling and heating power.

The main drawback of the first solution is to restrict the index tool to buildings which have their characteristics in the range of those defined to build the database. Concerning the second solution, since it is difficult to extract chiller consumption from the electricity bill, since monthly bills are not always available and since the installed powers are a poor piece of information, no simple and reliable index can be calculated. Purpose: Identify if the new building or the existing building is a good candidate for reversible heat pump or heat recovery by using a building energy simulation tool. Help the decision maker to pre-select some interesting options. Methodology: 1/ The user should enter the building and HVAC system information:

define thermal homogeneous zones in the building

describe the floor area, roof area, wall surfaces and their orientation, window surface and their orientation for each zone

describe the thermal characteristics of the walls, window and floors (among the wall compositions database)

describe the shadings and the solar protections (efficiency and scenario of use)

define the temperature and humidity setpoints (night/day, summer/winter)

define the internal load powers (lighting, electric appliances)

enter the infiltration flow rate in vol/h

describe the typical daily occupancy profiles (using tables)

define the scenarios of lighting and electric appliance use (using tables)

choose a type of ventilation system (mechanical extract ventilation, CAV or VAV), flow rates and scenario.

choose a climate among a database (Meteonorm files from few station to be defined) or use a user file

in option1 : describe Hot Domestic Water if any (system, storage, hourly water profiles) 2/ The tool will first assess the hourly loads for space heating, cooling, humidification and dehumidification corresponding to the selected weather data. The calculations include the distribution units (TU, AHU) and so they represent directly the heat and cold production loads. The Terminal units are sized according to a classical procedure. 3/ The tool calculates some ratios to define if heat pumping or heat recovery is potentially interesting. 4/The tool provides the rough data (energy saving potentials in kWh, %, and kWh/m² for solution, calculate the combination of solutions (reversibility for space heating + hot domestic water in option)

1 HDW is considered in option in all the document. It will be included in the tools after the end of the version for space

heating and if we still have time to do it.


5/ The tool provides some information about the potential according to the rough results and HVAC systems installed in case of retrofit building: proposition of solutions (reversibility, recovery or both) some recommendations on the feasibility (warning on temperature levels in terminal units, distribution temperature in heating coils should be as low as possible; modifications to foresee, limitation of air-to water HP at low outside temperature…)

Inputs

Building & HVAC System Energy loads on a typical

reference year

Calculation of REVERSIBILITY potential for(SPACE HEATING and

HUMIDIFICATION2 (see appendix 1)

Outputs: potentials

+ recommendations

(temperature level in terminal units…)

Calculation of HEAT RECOVERY potential for SPACE HEATING

AND HUMIDIFICATION (see appendix 1)

OPTION: Calculation of HEAT RECOVERY potential for HOT

DOMESTIC WATER (centralized? use of storage?, temperature level?..)

OPTION : Calculation of REVERSIBILITY potential for HOT DOMESTIC WATER (centralized?

use of storage?, temperature level?..)

Requirements :

Limit the inputs and choose accessible inputs

Provide information on default value and main assumptions

Calculate quickly the potentials

2 Replacement of steam humidification by adiabatic humidification


3. Sizing and economic/energy assessment tool

Introduction After having identified reversibility and recovery potentials with help of the identification tool presented here above, the user has still to select the best heat pump system(s). This selection can be made with help of the deliverable 1.4. After this first selection, the chosen system(s) should be sized and assessed. In this ―sizing and assessment‖ tool, the user will convert the heating and cooling demand profiles in energy consumptions. In this tool, HVAC systems are included. Most common reversible and recovery heat pump systems have been presented and discussed in the report ―Deliverable 1.4: Reversibility and heat recovery technology description‖. The methodology can be split in two steps:

- The sizing step depends on the case of a new or a retrofit building.

In case of a new building, the HVAC system should be defined from the beginning and all the HVAC system (Terminal units and heat and cold production) should be sized according to design procedures based on pre-sizing rules discussed in Task.2.

In case of an existing building to be retrofitted, the terminal units and the reference primary HVAC system are already defined and the new primary HVAC system should be chosen. The HVAC system can be sized according to design procedures (already done for secondary HVAC system in the first tool) or according to the data on the site. The user will have the possibility to adjust the simulated reference consumption to energy bills by entering the ―real‖ installed equipment, weather file of the specific year...

- The economic/energy step should include two calculations, one with the reference HVAC system and one with the heat pump system. (both simulations are essential and allow a direct comparison between the two systems).

Purpose: Assess the energy savings, CO2 emission reductions and the economic interest. Help the decision-maker to select the best solution in terms of costs/environmental impact. The tool should allow to study the most common solutions and to select the heat pump according the models sold (according its performance at full and part load). Methodology: 1/ The hourly (global) heating and cooling demands generated by means of SIMZONE/AGGREGATE, or by other simulation softwares (Trnsys, Energyplus,…), can be used as input of the software. 2/ Make a first selection of ―interesting‖ and ―feasible‖ heat pump systems (see information given on pre-design in ST2) based on the analysis of H/C demand profiles. The main parameters consist in the characteristics of the chosen system, of the coupled heat source(s)/sink(s) and the related investment cost. 3/ first calculation: simple sizing calculation based on standards and procedures (see French, Italian and German rules) pre-sizing and pre-design based on the results of D.1.4. and Task 2. design handbook 4/ Simulation of the selected and sized equipment and simulation with the reference system (condensing boiler + chiller for a new project or the initial system in case of retrofit). 5/ The tool provides as main outputs for the reference case (boiler + chiller) and the selected solution:

- the seasonal and yearly performances of the system (COP, EER, SEER, …) - Operating costs and pay back time (€) - the CO2 emissions (in kgCO2 and in kCO2/m²)


- primary energy consumption (kWh and in kWh/m²)

Parameters

- type of heat pump- type of heat source(s)/sink(s)- type of back-up system- system components performances

Inputs

- H/C demands profiles- weather data- Energy cost (€/kWh)- CO2 emissions (kgCO2/kWh)- HP system cost (€)

Outputs

- Performance factors (COP,…)- yearly functioning costs (€)- yearly CO2 emissions (kgCO2)- comparison with classical system (air-cooled chiller and gas boiler)

GENERATION

Systems to be considered:

- Classical separated heat and cold production: fuel/gas boiler + air/water cooled chiller

- Air-to-water Reversible heat pump (with backup heating system) - Water-to-water heat pump

Ground coupled (w or w/o storage): simple design ( reversibility only) or complex design with heat recovery

Exhaust ventilation air HP systems - Dual condenser (series or parallel) (heat recovery) - Closed water loop system and VRF: Simple assessment based on the comparison

between heating and cooling demands of different zones and on basic modeling (see Consoclim model for VRF)

terminal units considered (same as in the first tool): FCU VAV, CAV, mechanical extract ventilation TU and AHU models are already included in ―Simzone‖ (= Building model + complete HVAC system) Limitation:

We propose to exclude Hot domestic water production in this second tool Requirements :

Limit the inputs and choose accessible inputs

Provide information on default value and main assumptions Calculate quickly the potentials


III. Progress in the development of the tools

1. Simzone (multi or mono zone)/Aggregate

Given the limitations of an estimation based upon global values of heating and cooling demands, it is proposed here to use another approach, namely to generate by dynamic simulations hourly values of heating and cooling demands3 and to calculate the reversibility and recovery potentials for space heating as developed above from the results of the calculation. The connection with the measurements is done through tuning of the simulation model in order to reproduce not too bad the global values. Hourly simulations are carried out using models developed using EES (SimZone and Aggregate). The tool consists in a package of directly executable simulation files together with an explanation note including the definition of an evaluation index about reversibility and recovery potentials Identification of the savings is carried out in 3 steps:

1) Tuning of the parameters of the model using information effectively available (main dimensions of the building, types of facades, windows, orientations, occupancy schedule and rate, internal gains, effective control of the air renewal rate) and calibration to reproduce recorded energy consumptions;

2) Simulation of the system on a typical year, calculation of hourly values of heating and cooling loads;

3) Aggregation of the demands for the analyzed zones. Computation of the reversibility and recovery potentials, of the global index and system selection in view of an energy refurbishment

The three steps are carried out using the two following tools :

- Tool n°1: Building simulation and calculation of demands - Tool n°2 : Aggregation of the demands and calculation of the recovery and reversibility

potentials

A short description of the tools is given below. TOOL n°1 (“SimZone”) : monozone or multizone building simulation The behaviour of the building can be simulated by different methods:

- the building is considered as a single zone to which a secondary HVAC system is connected. This HVAC system includes an Air Handling Unit and terminal units for both heating and cooling;

- the building is analyzed floor by floor. It is possible to distinguish between intermediate floors and floors below the roof

- the building is divided in sub-zones which can represent either a single room, a part (wing) of the building or a set of peripheral spaces (in contact with an external frontage) and central spaces; each zone is connected to a secondary HVAC system (including an AHU and some TUs)

The selection of the method has to be done by the user who has to judge of the necessity to simulate a small part of the building or a larger area. The main selection factors are:

- the main orientation of the zone; - the levels and schedules of internal gains and occupancy; - the type (CAV, VAV, type of TUs) and the operation (schedule, setpoints, ventilation flow

rate) of the HVAC systems.

3 Instead of giving only sensible heating and cooling demands of the zone, the tools are developed to generate the

HVAC system heating and cooling loads for space heating/cooling and (de)humidification and take into account the

behavior and the performance of all the secondary HVAC components (coils, fans, pumps,…).


TOOL n° 2 (“Aggregate”): aggregation of the demands and calculation of the recovery and reversibility potentials Once the independent monozone simulations have been completed, the demands from the different zones (or groups of zones; 8760 hourly values of heating and cooling demands) are introduced in the aggregation tool. The number of zones of a given type is specified (number of south orientated zones, number of central zones…) and the loads are aggregated in order to determine the global demand profile. As mentioned above, a multizone release of « Simzone » is under development. This second version is based on similar modeling and implementation. Using this multi-zone tool simultaneous heating and cooling demands of up to 5 zones could be directly identifiable and the global H/C demands profile could be directly generated. The global heating and cooling demands are then compared and used in order to compute the recovery and reversibility potentials (defined in Appendix 1).

2. Sizing and Assessment tool (previously called “GENERATION”)

2.1. Component modeling

2.1.1. Heat pumps

The chiller model of a heat pump is the basis of each system being studied. The amount of energy consumed by the heat pumps can represent up to 45% of the total consumption of the HVAC (Heating ventilation air-conditioning) system in the case of a water loop system (Chapont & Delandre). In this context having an accurate model of the heat pump in order to determine the consumption is essential. The model of the chiller (and obviously the model of the heat pump when operating in a reversible mode) is based on correlations thanks to data collected from heat pump manufacturers. The aim of the model is to accurately simulate the operation and the electricity consumption of each type of heat pump especially when it operates at part load (PLR < 1)4. In this study five types of heat pump / chiller models are used:

1. The air cooled chiller

2. The reversible air-to-water heat pump

3. The simple water-to-water heat pump

4. The air-to-water with recovery heat pump

5. The reversible air-to-water heat pump used in water loop systems.

Calculating the electricity consumption of the heat pump/chiller has been done using the DOE-2 method proposed by Hydeman et al. (2002). This method aims to develop a set of curves depending on:

Entering or leaving evaporator water temperature;

Entering or leaving condenser water temperature;

The type of compressor used.

The complete model is described in Appendix A.2.1. With these curves it is possible to determine the entire electricity consumption of the device. Considering the operating mode, the temperature limit mainly depends on the type of heat pump selected. Generally, at evaporator exhaust it is located in the range 7 to 12 °C, in chiller mode. The condenser generally works between 35°C and 45 °C. Exceeding these limits is not allowed as it would damage the heat pumps.

4 PLR= Part Load Ratio. It quantifies the operation of the heat pump with respect to a full load operation.


Concerning the water loop heat pump, its operation is defined with an entering air temperature of 23°C in heating mode and 27°C in cooling mode. As far as the water loop temperature is concerned, the upper limit is defined at 40 °C while the minimum temperature is defined at -5°C. Operating outside of this range would result in damages for the heat pump so that in reality the temperature range will be narrower.

Figure 1: Schematic view of a heat pump

Several references have been utilized to develop the heat pump model:

1. DOE (Department of Energy). 1980. DOE 2 Reference Manual Part 1, Version 2.1. Berkeley, Calif.:

Lawrence Berkeley National Laboratories

2. Hydeman, M., and K. Gillespie. 2002. Tools and techniques to calibrate electric chiller component

models. ASHRAE Transactions 108(1): 733-741.

3. Hydeman, M., Taylor, S.T., Winiarski, D. 2002. Application of component models for standard

development. ASHRAE Transactions 108(1).

4. Hydeman, M., Webb, N., Sreedharan, P., Blanc, S. 2002. Development and testing of a reformulated

regression-based electric chiller model. ASHRAE Transactions 108(2).

2.1.2. Water Pump

Pumping water represents a significant part of the energy consumption in every HVAC system. In the case of a water loop system it could represent more than half of the entire HVAC consumption (Chapont, et al.). For the sake of simplicity we will only consider constant flow pumps and no variable flow pumps. In this work, no rigorous pump models were developed as the entire hydraulic network pressure drop is difficult to estimate. In fact, the pump model is based on correlations established between the water flow rate and the electricity consumption. These correlations are different depending on the type of hydraulic network considered. The complete pump model is described in Appendix A.2.2.

Figure 2: Schematic view of a pump

Several references have been utilized to develop the pump model:

1. Sfeir, A., Bernier, M., Million, T., Joly, A. 2005. A methodology pumping energy consumption in

GCHP systems. ASHRAE Transactions 111(2): 714-729

2. Bernier, M. And Lemire, N. 1999. Non-dimensional pumping power curves for water loop heat

pump systems. ASHRAE Transactions 105(2): 1226-1232.

3. Bernier, M. And Bourret, B. 1999. Pumping energy and variable frequency drives. ASHRAE Journal

41(12): 37-40.

4. Chapont, C., & Delandre, M. Air Conditioning by heat pumps on a water loop. ASHRAE

Transactions: Research.


2.1.3. Boiler

In non-residential buildings, many types of boilers could be found such as electric, oil or gas-fired boiler. According to Cooper (1994), the latter type is the best one. Concerning this boiler model some points have to be highlighted. Firstly a constant efficiency boiler model will not been considered; the efficiency of the boiler will be determined through the model. Secondly, according to Cooper (1994), the supply water temperature should not be below a certain point defined close to 60°C for a usual boiler. This problem will been addressed with the assumption that a mixing hydraulic network is placed before entering the boiler and can perfectly control the valves in order to preheat the entering water. Considering the boiler model it has been developed in two steps. The first boiler model is essentially based on the model described in ASHRAE Toolkit (Lebrun et al., 1999). According to Lebrun et al., this model of the boiler mainly uses an adiabatic burner, a water-gases heat exchanger and a water-ambient air heat exchanger. The aim is to simulate each heat exchanger as a counter

flow heat exchanger and to determine the exhaust water temperature through the method. The complete model is described in the Appendix A.2.3. In reality, even with this efficient model, it was not possible to introduce it in the entire program (especially in case of water loop system), as the number of iterations and the running time were too large. The problem created by the previous model in the water loop systems has been tackle in a very simple way. The principle was to create an efficiency table generated by the previous model based on the part load ratio and the entering boiler temperature, and to finally interpolate it with the aim to determine the efficiency and the fuel consumption.

Figure 3: Schematic view of a boiler

The main reference used to develop the boiler model is:

1. Lebrun, J., Bourdouxhe, J.-P., Grodent, M. 1999. HVAC1KIT: A toolkit for primary HVAC system

energy calculation. Atlanta, GA, American Society of Heating, Refrigerating and Air-Conditioning

Engineers.

2.1.4. Cooling tower

There are mainly two types of cooling tower which could be used, the direct and the indirect contact cooling tower. In this work only indirect contact cooling tower will be considered. It does not use any spray nozzles for the main water. The fluid flows through pipes in the cooling tower. An auxiliary network sprays water on the tubes in order to contribute to cooling the main water by heat absorption of the sprayed fluid. Air is flowing through the cooling tower at counter flow. The model of the cooling tower is mainly based on Stabat et al. (2004) and Marchio et al. (1999) models. The principle is based on considering the cooling tower as a counter flow heat exchanger using water and air loaded with water. The model (see Appendix A.2.4) has been developed in three steps:


1. a sizing of the cooling tower in order to determine the nominal heat transfer coefficient. Marchio

(1999) proposed a repartition key of the thermal resistance between the air and the water;

2. the simulation of the cooling tower using the previously calculated nominal heat transfer coefficient.

The aim of this part is to determine the exhaust water temperature through the method

considering the cooling tower as a counter flow heat exchanger;

3. the operating mode is determined comparing the exhaust water temperature with the set

temperature defined by the user. This part also allows calculating the electricity consumption of the

cooling tower.

Figure 4: Schematic view of the indirect cooling tower

Several references have been utilized to develop the cooling tower model:

1. Stabat, P. And Marchio, D. 2004. Simplified model for indirect-contact evaporative cooling-tower

behavior. Applied Energy 78: 433-451.

2. Bourdouxhe P, Grodent M, Lebrun J. Cooling-tower model developed in a toolkit for primary HVAC

system energy calculation—part 1: model description and validation using catalog data, Proceedings

of the fourth international conference on system simulation in buildings, Liège, December 1994.

3. Braun JE, Klein SA, Mitchell JW. Effectiveness models for cooling towers and cooling coils. ASHRAE

Transactions 92(2):164–174.

4. Lebrun J, Aparecida Silva C, Trebilcock F, Winandy E. Simplified models for direct and indirect

contact cooling-towers and evaporative condensers. 6th International conference on System

Simulation in Buildings, Liège, December 2002.

5. Bolher, A., Casari, R., Fleury, E., Marchio, D., Millet J.R., Morisot O. 1999. Méthode de calcul des

consommations d’énergie des bâtiments climatisés « ConsoClim » Présentation de la

méthode,ENEA/CVA-n°98.162R, CSTB.

6. Marchio, D., & Morisot, O. (1999). Modelisation simplifiée d'un aéroréfrigerant dans l'optique d'un

calcul des consommations d'énergie d'une installation de climatisation dans un bâtiment tertiaire.

ENEA/CVA-n°98.162R, CSTB.

2.1.5. Dry cooler

The principle of the dry cooler is based on the indirect cooling tower one avoiding spraying water on the tubes. The dry cooler can be considered as a heating coil; the air flow passes through many tubes which are travelled by the water. Its main advantage is that no secondary water pump is needed (comparing to an indirect cooling tower) and therefore the cost of maintenance is reduced (See section Erreur ! Source du

renvoi introuvable.). However some economic trade off related to the sizing of the equipment or to the consumption results in water temperature always being more than 10°C above the dry bulb air temperature. This level is often considered too high for a refrigeration facility or any industrial process (Winandy, 2008).


The dry cooler model (see Appendix A.2.5) is roughly the same as the cooling tower model and was mainly developed by Marchio et al. (1999). It is divided in the same steps of which are: the sizing, the simulation and determining the operating mode. However as no water spraying is present, only the dry air temperature is considered and no iterations caused by the web bulb temperature are present.

Figure 5: Schematic view of a dry cooler

The main reference used to develop the dry cooler model is:

1. Marchio, D., & Morisot, O. (1999). Modelisation simplifiée d'un aéroréfrigerant dans l'optique d'un

calcul des consommations d'énergie d'une installation de climatisation dans un bâtiment tertiaire.

ENEA/CVA-n°98.162R, CSTB.

2.1.6. Cooling coil

A cooling coil is a coil crossed on one side by water and on the other side by air. This device is only used in the exhaust heat pump system (Model 2) in order to absorb as much energy as possible from the exhaust ventilation air which energy will be used in a heat pump. Like in some other models the sizing and the simulation part are present. The sizing part aims to determine the nominal heat transfer coefficient thanks to nominal data and using the model described by Bolher et al. (1999). Manufacturer’s data were used to determine correlations between the air and water heat transfer coefficient and the exhaust air flow rate. It has also been possible to determine the nominal water flow rate passing through the cooling coil in function of the exhaust air flow rate. The second part of the model simulates the operation of the cooling coil in order to determine the exhaust water temperature. The model is fully explained in Appendix A.2.6

Figure 6: Schematic view of a cooling coil

Several references have been utilized to develop the cooling tower model:

1. Bolher, A., Casari, R., Fleury, E., Marchio, D., Millet J.R., Morisot O. 1999. Méthode de calcul des

consommations d’énergie des bâtiments climatisés « ConsoClim » Présentation de la

méthode,ENEA/CVA-n°98.162R, CSTB.

2. Morisot, O., Marchio, D. 1998. CCSIMOL : A few inputs model of variable flows cooling coil.

Proceedings of the fifth International Conference on System Simulation in Buildings, Liège, Belgium.


3. Morisot, O., Marchio, D. 2002. Simplified Model for the Operation of Chilled Water Cooling Coils

Under Nonnominal Conditions. HVAC&R Research 8(2): 135-158.

4. Braun JE, Klein SA, Mitchell JW. Effectiveness models for cooling towers and cooling coils. ASHRAE


2.1.7. Ground heat exchanger

There are many types of ground heat exchangers mainly divided into two categories: the horizontal and the vertical ones. In this work only the vertical ground exchanger will be considered as in general non-residential buildings do not have enough space to implement horizontal heat exchanger to supply the entire building demand. The model used in this work is directly based on a model developed by Bernier (2006) which is completely explained in the Appendix A.2.7. Without entering into the details, this model is divided into different parts:

1. Evaluating through an algorithm the MLAA (Multiple Load Aggregation) (Bernier, et al., 2004).

2. Evaluating the penalty temperature which allows estimating the impact of thermal pollution of the

ground (a modification of the ground temperature).

3. Calculating the temperature of the inner surface of the tube via the calculation of the thermal

resistance of the ground and considering that it is the same for the whole tube.

4. Calculating the different thermal resistances thanks to the provided data and through the evaluation

of the Nusselt and Fourier numbers.

5. Calculating the exhaust water temperature using a simple model of semi-isotherm exchanger

between the fluid and the inner surface temperature.

In practice the simulation will be done for only one year, which neglects the effect of thermal pollution of the ground.

Figure 7: Schematic view of a ground exchanger

Several references have been utilized to develop the ground heat exchanger model:

1. Bernier, M., Kummert, M., Bertagnolio, S. 2007. Development and applications of test cases for

comparing vertical ground heat exchanger models. Proceeding of Building Simulation 2007, Beijing,

China.

2. Bernier, M.A. 2006. Closed-Loop Ground-Coupled Heat Pump Systems. ASHRAE Journal 48(9): 12-

18.

3. Bernier, M.A., Labib, R., Pinel, P., Paillot, R. 2004. A Multiple Load Aggregation Algorithm for

Annual Hourly Simulations of GCHP Systems. HVAC&R Research 10(4): 471-487.

4. Bernier, M. A. 2001. Ground-coupled heat pump system simulation. ASHRAE Transactions 107(1):

605-616.

5. Bernier, M. 2000. A Review of the Cylindrical Heat Source Method for the Design and Analysis of

Vertical Ground-Coupled Heat Pump Systems. 4th International Conference on Heat Pumps in Cold

Climates, p. 1-14.


6. Bernier, M., Chahla, A., Pinel, P. 2005. Long-term ground temperature changes in geo-exchange

systems. ASHRAE Transactions 114(2).

2.1.8. Thermal storage

A thermal storage is often a tank which contains a large volume of water. In this study thermal storages are used in order to model the effect of the thermal inertia of the system. This model will only be used in water loop systems as the water network is larger than in other systems. They will be implemented in two different ways:

1. As fictitious thermal storage in order to model the effect of the thermal inertia of the loop itself. In

this work it is not possible to use several small thermal storages to simulate the effect of the thermal

inertia spread along the water network. The assumption of a unique storage which concentrates the

entire mass of water will be done in order to reduce the running time of the program.

2. As real thermal storage which could represent an additional water tank installed along the loop.

Moreover, as an hypothesis, the final thermal storage volume will be considered as the sum of both so that all the thermal inertia effect is concentrated in a unique thermal storage. The model is based on the Winandy (2008) one. According to this model two parameters have to be provided to describe the entire thermal tank: the volume of the tank and the coefficient of heat losses. As the range of temperature does not significantly differ along the simulation time only fully mixed thermal storage will be considered5.

Figure 8: Schematic view of a thermal storage

The main references used to develop the thermal storage model are:

1. Lebrun, J. 2007. Introduction aux Machines et Systèmes Thermiques. Notes de cours. Université de

Liège, Belgium.

2. Winandy, E. (2008). Production de froid et chaleur basse température, Notes de cours. Université de

Liège.

5 One other solution could be to consider fully stratified thermal storage.


2.2. Systems modeling

2.2.1. Development of the program

As explained above, the models have been developed on EES (Engineering Equation Solver - ©F-Chart Software). In practice each file contains the heat pump system considered. Even if it would be better it has not been possible yet to develop a unique file including all the systems. However each file/Model has been developed in the same way. The different parts of the program are described hereafter.

Introduction of procedures

According to EES terminology, a procedure is a little subprogram which can be called from anywhere in the program. The specificity of a procedure is that all equations are developed in an explicit form, which means no iteration is required in order to solve them. Each system has been developed as a set of components procedures. Modelling components into procedures has been a special feature of the work achieved as they are not commonly used. It has been motivated by two main reasons:

Having explicit equations increases the robustness of the program as no iteration is performed;

Interchanging components is easier as each procedure is only defined by its inputs.

Procedures have also been used to do the sizing of the system and to determine the heat pump electricity consumption through the DOE-2 method. According to EES syntax, a procedure must be placed before any other equation.

Control algorithm

This is the decision tree of the system. It allows changing the operating mode of the system according to the conditions of use. In each part of the control algorithm different equations are implemented regarding the mode in which the system operates. The control algorithm has been implemented in a procedure because the utilisation of ―IF‖ conditions was easier. The main outputs of this part are the consumption, the capacity of the different devices and various temperatures.

Inputs/parameters required

This part contains inputs and parameters which determine the system and its environment. They are required by the program to run the simulation. These could be:

Heat and cold demand

Maximum heat and cold demand

Temperature set

Nominal difference of temperature in the heat pump/chiller and the boiler.

Water flow rate

Possible volume of the thermal storage

Possible heat losses in the building

In fact, many of the listed parameters are defined directly through a ―sizing procedure‖ which is called in this part. This ―sizing procedure‖ will define many of the above mentioned parameters depending only on the demand. Finally the user only has to define the temperature set and the nominal difference of temperature of the devices used by the system. As the sizing procedure highly depends on the type of demand introduced in the system.

Parameters of procedures

This part contains all the parameters used in the components procedures. These parameters are supplied in nominal conditions determining a sizing of the component considered. They can be easily modified by the user in order to better match the system considered. At the end of this part, all the thermal storage equations are defined if they need to be implemented.


Time

The time interval between two integrations can be changed but is first set at 1 hour.

Integrations

This part of the program follows with the calculation of all the integrated values, such as electrical consumption or the total amount of fuel burned.

Comparison

The last part of the program deals with the comparison of the system considered with the reference model (except for the reference model itself, of course). General outcomes of the program such as the primary energy consumption, the CO2 emission and the life cycle cost are calculated.

2.2.2. General hypothesis

Many assumptions have been made on components (already developed) and on systems which are essential, although it may seem that the accuracy could be lost in the results. Firstly, they reduce the number of inputs needed to evaluate the environmental and economic potential. Secondly, they reduce the time for running the program as a consequence of the first advantage.

General operations

The first hypothesis is related to the heat pump system which is assumed to be designed to supply the entire demand. All devices are sized to first supply the cooling demand. Also, in case of a heat pump supplying the heating demand, a backup heater is used if required. The second hypothesis is linked to the secondary system (mainly the ventilation system). In this study it will be considered that the ventilation system is:

1. able to transfer the entire demand of the heat pumping system;

2. designed to work in the temperature range provided by the heat pumping system (45-55°C in heat

pump mode, 7-12°C in chiller mode)

This allows focusing the program on the heating and cooling production system only. Thirdly no defrosting cycle will be considered. This assumption is only valid for the System/Model 1 as it requires an air-to-water heat pump. Finally in this work free cooling will not be considered. Free cooling is the possibility to directly take advantage of a cooling source without a thermodynamical machine. There are essentially three types of free cooling:

Natural ventilation: Introduction of the outside air (cooler than the inside air) through purpose-

provided openings without energy consumption;

Mechanical over-ventilation: Blowing of cool outside air into the building by means of fans.

By-passing of the heat pump: When the heat source is at a temperature level which makes it possible

to cool, a valve allows to by-pass the heat pump and directly feeds the building distribution system.

The energy consumption is due to the pump. This type of free cooling is often referred to as ―free

chilling‖.

Fluid properties

In the selected systems four types of fluids are used and it is obviously not possible to link a variation law to each of the fluid properties. Many of these have been established as fixed under nominal conditions. The following table summarises the main characteristics for those fluids.


Water Glycol Water (25%)

Gas fuel Air

[J/kg-K] 4190 3960 1800 1006

[kg/m³] 1000 1022 1.16

[J/kg] 43 106

Table 1: Hypothesis on fluid properties

Other characteristics have also been defined: the viscosity, thermal conductivity, etc.

Heat losses in hydraulic networks

Modelling the heat loss in all the hydraulic networks is a quite difficult task as such losses depend on many parameters such as the ambient temperature, the temperature of the fluid flowing in the pipes and the properties of the pipes used. In the context of the study the heat losses model will only be used when the heating or cooling demand equals zero while in the opposite case they are already accounted for in the global demand. In fact, the demand supplied is considered exactly as the heat and cold production system loads. Considering the hydraulic network of system 0 to 3, heat losses will correspond to 2.5% of the nominal demand according to Stabat (2001) but will only operate when no demand is supplied. In water loop system, heat losses are quite negligible as the temperature range of the water loop stays close to the ambient temperature. The main proof that heat losses in water loop system are negligible is that the hydraulic network is generally not isolated according to Philippe Struvay6. However in case of prolonged inactivity heat losses will be considered. In such case, after a long idle time the loop temperature tends to match the ambient air temperature.

2.2.3. Model 0: Chiller and boiler independent (Reference)

General description

This system will be used later as a reference for the other ones as it represents the simplest way to satisfy cooling and heating demands. The two heat and cold networks are completely separated; they are respectively supplied by the boiler and the air-to-water chiller. The components which are modelled in this system are:

Water cooled chiller

Boiler

Pump

6 Senior Engineer at Camair Air conditioning.


Figure 9: Model 0: Chiller and boiler independent

Heat source/sink

In this model only the outdoor air is considered as a heat sink as it is quite often used in heat pump systems. The main disadvantage of the outdoor air is its very variable temperature. In cooling mode high outdoor temperatures reduce the performance and the capacity of the heat pump.

Control algorithm

With this system, control algorithm is the simplest one as the two parts are independent; each network has its own control algorithm. The only test to be done is whether the demand is null or not. When a cooling or heating demand is detected by the system, the chiller or the boiler is switched on. The control algorithm is summarised in Figure 10:

Figure 10: Control algorithm of the Model 0

2.2.4. Model 1: Reversible heat pump

General description

The air-cooled chillers are the dominant technology on the European air-conditioning market, representing 85 % of chillers sold (EECCAC, 2003). The system can be reversed by means of a refrigerant change-over7 (4 ways valve) which inverses the refrigerant flow passage into the two exchangers:

7 Another possibility is the “system change-over” changing the hydraulic network but will not be considered here.


In cooling mode the air exchanger works as a condenser, rejecting heat, while the water-exchanger

works as an evaporator, transferring cooling power to the distribution system.

In heating mode the air exchanger works as an evaporator, absorbing heat from outside air, while

the water exchanger works as a condenser, transferring heating power to the distribution system.

In most cases, reversible air-cooled units are however installed in combination with backup boilers, for the following reasons:

1. To supply heating power when heating and cooling loads are simultaneous: when the chiller works

in cooling mode, it cannot provide any heating power. Because its condenser is air cooled, the

condenser heat cannot be recovered.

2. To complete heating power for low outdoor air temperatures, in case the air-to-water unit is not

sized for low outdoor temperatures (the heating power at –5°C can be 30% lower than heating

power at nominal conditions (Bertagnolio, et al., 2008));

3. To supply the entire heating power when, at very low outdoor temperatures, the boiler has better

performances8 than the air-to-water unit, or when the outdoor temperature is under the unit

working range.

Such system can reach good efficiency if the heating and cooling demands are alternate, as the system is designed to be reversible. This type of heat pump modelled the following components:

Air-to-water reversible heat pump

Boiler

Pump

Figure 11: Model 1: Reversible heat pump

8 in terms of primary energy consumption and/or CO2 emissions


Heat source/sink

During winter and in heating mode, low outdoor temperatures limit the capacity of the heat pump and decrease its performance which could be due to the presence of frost. Like in the model 0, in cooling mode, high outdoor temperatures reduce the performance and the capacity of the heat pump. Moreover, during winter defrosting strategies have to be used and decrease the performances and the heating capacity of the heat pump. The defrosting process of the heat pump will not be considered in our modelling of the system.

Control algorithm

The control algorithm explained below rigorously follows the sequential steps of the control procedure. The system is cooling driven which means that the priority is given to cold generation if there is a cooling demand. The control algorithm starts with the declaration of the variable as it is the necessary step to avoid any compilation problems. Variables defined in this part are mainly the outputs of the control procedure. DECLARATION OF VARIABLES The main temperatures which are used in the following control algorithm are the return temperatures of the network defined as:

applying the hypothesis previously explained that the set point is achieved (even with a backup boiler if necessary) and that the secondary system (ventilation) is able to supply the entire demand. The control algorithm follows with the evaluation of the heat demand.

IF THEN

IF THEN HEAT LOSSES CALCULATION Hot network HEAT LOSSES CALCULATION Cold network ELSE HEAT LOSSES CALCULATION Hot network CALL CHILLER PROCEDURE PUMP CONSUMPTION CALCULATION ENDIF

When the heat and cold demand are equal to zero no consumption is calculated but heat losses are taken into account. In order to calculate the heat losses it should be determined whether the program is at the first step of time or not as the heat losses calculation is based on the water temperature at the previous step of time. Electricity consumption of the pumps is equal to zero. Calculating the heat losses allows knowing the current temperature of the water network. In case of extended period without demand, the temperature of the network stabilises at the ambient temperature (established at 23°C but this value can be easily changed). Only in case the cold demand is not equal to zero, the heat pump in chiller mode is activated. At this step of the program the chiller mode procedure is called, which determines the electricity consumption and the EER of the component using the DOE-2 method. ELSE

IF THEN HEAT LOSSES CALCULATION Cold network

IF THEN CALL BOILER PROCEDURE PUMP CONSUMPTION CALCULATION

In case of a heating demand and no cooling demand, the heat losses in the cold network are calculated before doing anything else. Further, a simple measurement of the outdoor air temperature defines whether


the heat pump will be activated or not. The control algorithm does not allow the heat pump to operate in case of too low outside temperature to avoid operating at a poor COP. In the context of this work, the limit temperature of -5°C is considered. If the outdoor air is below such limit the heat pump is switched off and the boiler supplies the entire heat demand. Thereafter, the calculation of the pump electricity consumption is performed through the correlations.

ELSE MAXIMUM CAPACITY CALCULATION

IF THEN

CALL HEAT PUMP PROCEDURE

ELSE

CALL HEAT PUMP PROCEDURE CALL BOILER PROCEDURE

ENDIF PUMP CONSUMPTION CALCULATION

ENDIF

Considering that the ambient air temperature is high enough, the maximum capacity of the heat pump is defined in relation with the operating conditions. It is calculated through a procedure with uses the DOE-2 method with a PLR=1 (full load). When such parameter is not exceeded the system determines the electricity consumption of the heat pump assuming that the exhaust water temperature is the set one. In the opposite case, the system needs the help of the boiler to reach the set temperature. The exhaust heat pump water temperature is determined assuming that the device operates at its maximum capacity. As the system functions in heat pump mode only the heat pump set of correlations is used to determine the consumption of the system. Basically, in this case, there are two methods to add the boiler in the system: in parallel or in series.

In the parallel method, the heat pump and the boiler provide the same exhaust water temperature

but receive different water flows in relation to their respective capacity.

In the series method, the heat pump heats the entire water flow as much as it can even if it does not

reach the set temperature and the boiler supplies an additional power to heat the entire water flow

to the set temperature.

In the context of this work the series configuration will be used.

ELSE CALL CHILLER PROCEDURE CALL BOILER PROCEDURE PUMP CONSUMPTION CALCULATION ENDIF

ENDIF When heating and cooling demands are simultaneous both networks are considered as separated as it was the case in the Model 0. The control algorithm can be summarised in the Figure 12:


Figure 12: Control algorithm of Model 1

2.2.5. Model 2: Exhaust air heat pump

General description

This type of system takes advantage of the air which is exhausted from the building. Exhaust air heat pump systems use water-to-water units. Extracted air could be used as heat source in heating mode and as heat sink in cooling mode. In this model, we will only use the extracted air as a heat source assuming that it could be a fair heat sink (Bertagnolio, et al., 2008). Given that the non residential building demand could be quite large, the exhaust air flow could be insufficient to supply heat. In that case an additional heat producer (boiler) is required. The heat pump is not reversible and each water network is independent. This type of heat pump can achieve a good performance level as the temperature of the source is constant. For this system and although it could be possible, the condenser heat is not recovered. The recovery possibilities will be studied in the next system. This type of heat pump requires the following components:

Water-to-water heat pump model

Boiler model

Pump model

Cooling tower model

Air-water cooling coil model


Figure 13: Model 2: Exhaust air heat pump

Heat source/sink

Exhaust air represents a very interesting heat source because of availability, good coincidence with needs and its very constant temperature. Condensing the water contained in extracted air allows recovering part of the latent energy and increases the heating capacity. However, exhaust ventilation rates are restrictive in terms of capacity so that an additional heat source may be required. Even if this case is not considered in the present system, in cooling mode, condenser heat can also be evacuated in the exhaust ventilation air flow which could be at a lower temperature than the outdoor air.

Control algorithm

Like in the previous model, it is cooling driven. The control algorithm starts with the declaration of variable followed by the evaluation of the demand. DECLARATION OF VARIABLES

IF THEN

IF THEN HEAT LOSSES CALCULATION Hot network HEAT LOSSES CALCULATION Cold network ELSE HEAT LOSSES CALCULATION Hot network CALL CHILLER PROCEDURE

CALL COOLING TOWER PROCEDURES PUMP CONSUMPTION CALCULATION ENDIF

Like in other models each device is switched off when there is no demand and the heat losses are calculated. When only a cooling demand is present, the chiller procedure allows calculating the electricity consumption of the device (and the EER) knowing the power of the evaporator. The power of the condenser is determined


through the EER which is used to calculate the theoretical exhaust water temperature of the cooling tower. Finally the cooling tower procedure is called to determine the electricity consumption of the component. In fact two procedures are successively called for the cooling tower: the first one is the sizing procedure of the cooling tower which determines heat transfer coefficients according to the nominal point; the second one simulates the cooling tower and determines the electricity consumption. ELSE

IF THEN HEAT LOSSES CALCULATION Cold network CALL COOLING COIL PROCEDURE CALL HEAT PUMP PROCEDURE

If only a heating demand is present, the cooling coil water loop is switched on. The procedure called allows knowing the capacity potentially supplied by the exhaust air which determines the capacity of the evaporator. The heat pump procedure determines the COP which is used to calculate the corresponding condenser capacity.

IF THEN

CALL BOILER PROCEDURE PUMP CONSUMPTION CALCULATION ELSE

CALL HEAT PUMP PROCEDURE PUMP CONSUMPTION CALCULATION ENDIF

At this stage of the control, we have to differentiate when the heat pump is able to supply the entire demand or when it is not. In the first case, the heating demand is larger than what the condenser can provide. In this context, the exhaust condenser water temperature is below the set temperature and the PLR is equal to 1; the boiler will provide the additional heat needed to reach the temperature set. In the second case the set temperature is reached and the PLR of the heat pump is below or equal to 1.

ELSE CALL CHILLER PROCEDURE

CALL COOLING TOWER PROCEDURES CALL BOILER PROCEDURE PUMP CONSUMPTION CALCULATION ENDIF

ENDIF In the case of simultaneous demands there are 2 possibilities:

It can be assumed that both networks are separated and are operating independently;

or part of the heat rejected is considered to be collected in the water exchanger and the exceeding

heat is rejected at the cooling tower.


At this stage of the research we will focus on the first option as the second one will be considered in the next model (Model 3: Dual condenser). In particular, in case of simultaneous demands, the chiller and the cooling tower procedures are called for the cooling demand while the boiler procedure is called for the heating demand operating completely separately. The control algorithm can be summarised in Figure 14:


2.2.6. Model 3: Dual condenser heat pump

General description

The dual-condenser chiller is made of a chiller equipped with two condensers, one air-cooled and the other one water-cooled. The main advantage of this type of heat pump is that it allows the heat recovery, so heating and cooling demands can be simultaneous. There are a lot of different possibilities to integrate both condensers in the system. For the remainder of the study the double-bundle condenser system will be the only one considered. In this configuration both condensers are installed in parallel. Double-bundle condenser system has been selected because historically it has been the most often used system for chiller heat recovery (Dorgan, et al., 1999). Refrigerant gas from the compressor flows into condenser shells, allowing heat rejection to one or both condensers. Usually the each condenser is sized for full heat output of the chiller. According to Dorgan et al. (1999) the main advantages of this type of double-bundle condenser in relation to others are:

Standard construction can be used;

Efficiency is maintained during cooling only operation;

It is assumed that a back up boiler can operate to:

cover the whole heating demand when there is no cooling demand;

cover the peak heating demand.

This type of heat pump system requires the following elements:

Water-to-water dual condenser heat pump model

Boiler model


Pump model

Figure 15: Model 3: Dual condenser heat pump

Heat source/sink

As it is the case in some previous systems, the air is the heat sink when the heat pump operates in chiller mode. On the contrary, the heat source is either the boiler or the heat pump condenser. The availability of the second option clearly depends on the cold demand as there is a relation between the evaporator power and the condenser power.

Control algorithm

DECLARATION OF VARIABLES

IF THEN

IF THEN HEAT LOSSES CALCULATION Hot network HEAT LOSSES CALCULATION Cold network

ELSE HEAT LOSSES CALCULATION Hot network CALL CHILLER PROCEDURE PUMP CONSUMPTION CALCULATION ENDIF

In case of a cooling demand the heat pump only functions as a chiller and the air condenser is the only one to operate. Only the correlations of the air condenser are used. ELSE

IF THEN HEAT LOSSES CALCULATION Cold network CALL BOILER PROCEDURE PUMP CONSUMPTION CALCULATION

If there is a heat demand, only the boiler functions as it is not possible to recover heat from the chiller. So far the control algorithm is exactly the same as for the reference case.


ELSE CALL CHILLER WITH RECOVERY PROCEDURE

IF THEN

CALL BOILER PROCEDURE PUMP CONSUMPTION CALCULATION

If simultaneous demands are observed, the capacity which can be recovered is calculated. If the heat demand is greater than what the condenser can provide, only the water condenser works and the correlation for the water-to-water heat pump is used. As the set temperature cannot be achieved the boiler is switched on.

ELSE

CALL CHILLER WITH RECOVERY + AIR CONDENSER PROCEDURE PUMP CONSUMPTION CALCULATION ENDIF

ENDIF ENDIF If the recovery is greater than the heat demand, the boiler is no longer needed and the excess heat is rejected by the air condenser. The water-to-water correlation is used adding the consumption of the fan of the air condenser to the general consumption of the heat pump. The control algorithm can be summarised in the Figure 16:



2.2.7. Model 4: Water loop heat pump

General description

A water loop heat pump system is made of several reversible air-to-water heat pump units, each serving a zone, connected to a closed water loop circulating throughout the building. Each heat pump unit uses the water loop as heat source/sink rejecting or absorbing heat to/from it. The main advantage of this system is that the demands can be simultaneous as rejecting and absorbing heat at the same time in the loop keeps the temperature of the loop in a certain range (usually between 16°C and 32°C, according to Pietsch (1990)). If the temperature exceeds the upper limit a heat rejector must be switched on (generally a cooling tower). In the opposite case a heat source is used (generally a gas condensing boiler) even if it results in production of heat at low temperature. Hybrid configurations combining ground heat exchangers with a standard heat source (boiler) or a heat rejector (cooling tower or dry fluid cooler) are also considered. It allows optimising the sizing of the ground heat exchangers as they can be used as a heat source or sink. In this case, systems often use vertical ground exchangers as horizontal ground exchangers require a larger area (Bertagnolio, et al., 2008). However using a combination of the three elements (heat rejector – ground exchanger – heat source) is not common and generally only two of the three elements are combined. It is important to keep in mind that in the case of the utilisation of a ground exchanger all the water loop has to be loaded with glycol (to avoid freezing). Moreover, even if it may appear more efficient to have an independent ground loop it is rarely implemented. In some cases the main loop is separated from the heat rejector and heat source by heat exchangers. In this study only a simplified water loop system without any secondary circuit will be considered.

NO RECOVERY

RECOVERY


Adding a thermal storage allows a better utilisation of the water loop: the water loop temperature remains longer in the appropriate range without the use of the cooling tower or of the boiler; it smoothens the temperature curve of the loop. The volume of the loop itself can be adjusted using 14 litres of water per kW installed (Pietsch, 1990). This type of heat pump systems requires the following models:

Water-to-air heat pump

Boiler

Cooling tower

Dry cooler

Water loop pump

Ground exchanger

Three configuration examples are presented in Figure 17. In these examples, the peripheral zone uses heat pumps in heating mode while the core zones uses heat pumps in cooling mode. At the other side of the loop a combination of two components out of the three presented can be used to keep the water loop temperature in the right range.

Model 4a

Model 4b


Figure 17: Model 4: Water loop heat pump: configuration example

According to this schema, the thermal storage is placed at the exhaust of the heat pumps. When the demand is heating dominated the temperature of the thermal storage decreases while it increases in the opposite case. Locating the thermal storage this way smoothens the temperature curve and delays the use of the heat rejector or of the boiler. The development of this model has been helped by:

1. Jeffrey, J., Splitler, D., Fisher, D., Xu, .X, Simulation of Hybrid Ground Source Heat Pump Systems

and Experimental Validation . 7th International Conference on System Simulation in Buildings,

Liège, December 11-13, 2006

2. Kavanaugh, S., A Design Method for Hybrid Ground Source Heat Pump. ASHRAE

TRANSACTIONS 1998, V. 104, Pt. 2.

Heat source/sink

The soil composition has a great influence on its thermal properties and on the overall performance of the ground heat exchanger. The thermal diffusivity of the soil is the dominant factor for the heat transfer in the ground. The availability of the ground in terms of temperature is excellent as the temperature two meters below the surface is often around 10° C all year long (Winandy, 2008).

Control algorithm

The control algorithm is depending on the type of water loop considered. The present section will be developed in three parts in relation to the main types of water loop. The water loop system program requires a lot of iterations and the control algorithm was usually defined as follows:

1. Given a supply heat pump temperature the exhaust heat pumps temperature is calculated through

the evaluation of the demand. A conversion of the demand is done in terms of heat rejected or

absorbed in the loop thanks to the COP and EER of the heat pumps.

2. The exhaust heat pumps water flows in the thermal storage and the exhaust thermal storage

temperature is calculated.

3. This temperature is entered in the control algorithm which determines whether a component should

be activated to increase or decrease the water loop temperature.

4. Exiting the control algorithm a new supply heat pumps temperature is defined and this procedure is

repeated.

It is easily understandable that running a year simulation using this method is not possible as the number of iteration would be extremely high. Also a phenomenon of entering and going off a ―IF‖ condition may

Model 4c


appear. For example at a given time the water loop requires the help of a boiler to increase its temperature. Afterwards a new and different supply heat pumps temperature is calculated at the next iteration. Due to the buffer effect of the thermal storage at this iteration the action of the boiler is no more needed. However another supply heat pumps temperature is calculated which, this time, leads to the utilisation of the boiler. As described the temperature enters alternately in the ―IF‖ condition of the boiler which leads to the non convergence of the model. A way to tackle this problem has been to use the temperature of the previous set of time to calculate the actual heat pumps supply temperature. The way to run the simulation is the same as previously, except that the starting supply temperature is taken at the previous step of time. With this method the supply heat pumps temperature calculated at the end will only be used at the next time step. All parameters such as the COP or the EER must be calculated at the temperature of the previous time step to be consistent. No iteration is required so that the running time is drastically shortened and the convergence is ensured. The running procedure is summarised in Figure 18:

Figure 18: Method of running the simulation of the water loop system

Model 4a: Heat rejector and boiler

This type of water loop utilises a heat rejector and a heat source (here a gas boiler) with the aim of keeping the temperature of the loop in a certain range. The heat rejector can be either a cooling tower or a dry cooler.

Previously Currently


As explained earlier the calculation of the exhaust thermal storage temperature is performed before entering in the control algorithm which is described hereafter. DECLARATION OF VARIABLES

IF AND THEN HEAT LOSSES CALCULATION ELSE A first estimate of the demand is completed. In this case heating and cooling demands are null so everything is switched off and the circulating pump does not operate. The following part of the control will be analysed with a heating dominated demand.

PUMP CONSUMPTION CALCULATION

IF THEN

IF THEN

ELSE CALL BOILER TABLE

ENDIF

Before considering any component activation, the pump electricity consumption is calculated assuming that it is constant regardless of the type of component switched on. For the first case ( ), even if the circulating pumps function, no element is switched on as the water loop temperature is too high to activate any heat source. This system mode is called ―stand-by‖ as no device is switched on. The value of the lower limit temperature can be determined as:

keeping a constant limit at 16°C (Pietsch, 1990);

varying the limit along the year. For instance the limit could be 28°C in summer time and 5°C in

winter time.

Below such lower limit the boiler is switched on. As already explained for the water loop system no boiler model is presented but a performance table is used as the global system could lead to numerical instabilities. The next section of the algorithm considers a cooling dominated demand. ELSE IF THEN

If the exhaust thermal storage temperature is still in the acceptable range no device is switched on and only circulating pumps operate.

ELSE IF THEN HEAT REJECTOR PROCEDURES

ELSE

ENDIF

ENDIF ENDIF

ENDIF If the exhaust thermal storage temperature is above the upper limit, one must evaluate if switching on the heat rejector leads to a global benefit or not. is defined as the difference between the water loop


temperature at the exhaust of the thermal storage and the air temperature. If such difference is greater than a given control value (often defined as 10°C) switching on the heat rejector is beneficial. On the opposite using the heat rejector would not be efficient and one can assume that the temperature of the loop can increase until the previous condition is achieved. Even if it is not implemented in the program and in case of heat wave, the temperature of the loop is assumed to increase during the day and the heat rejector operates at full load during the night in order to decrease the water loop temperature to a more acceptable level. The difference of temperature control changes according to the heat rejector use: in the case of a cooling tower, the wetbulb temperature is used, while the dry air temperature is used for a dry cooler.

Figure 19: CT/DC and boiler water loop control algorithm

Model 4b: Heat rejector and GHX

This type of water loop considers that the ground heat exchanger (GHX) is used as a heat source device but also in order to reject heat when heat pumps operate in cooling mode. Like in the previous model the control algorithm starts with an estimation of the general demand. DECLARATION OF VARIABLES

IF AND THEN HEAT LOSSES CALCULATION ELSE In this case everything is switched off as no demand is supplied.



IF THEN IF THEN

ELSE CALL GHX SUBPROGRAM

NEW PUMP CONSUMPTION CALCULATION

ENDIF

This part of the control algorithm is significantly the same as in the previous one. In the first case the mode ―stand-by‖ is activated and the ground heat exchanger is bypassed; the system works on its own thermal inertia. However the ground exchanger, instead of the boiler, is considered as a heat source in this model. Moreover the minimum temperature of the loop has been changed to 10°C as the temperature of the ground is close to 10°C all year long. Indeed, the exhaust thermal storage water temperature has to be below the ground temperature to keep the system effective. It is important to note that a procedure is not used for the GHX simulation but a SUBPROGRAM. ―Subprogram‖ allows iterations and can be called inside a procedure (here the control procedure) while ―Procedure‖ cannot. The main difference with the previous model is that the pump consumption is modified as it is considered that the pressure drop is higher if the ground heat exchanger is used. As the pump model is based on a correlation it is not possible to evaluate the difference of consumption. Arbitrarily, an addition of 20% is used to simulate the increase of pump consumption due to the use of the ground heat exchanger. In the future, a possible improvement could be to consider the increase of electricity consumption of the pump due to the utilisation of the ground heat exchanger. The demand is now cooling dominated.

ELSE IF THEN

CALL GHX SUBPROGRAM NEW PUMP CONSUMPTION CALCULATION

Even if the temperature is still is the acceptable range, the water flows through the ground heat exchanger in order to delay the use of the heat rejector. The maximum temperature is the same as in the previous case (32°C) to use the heat rejector. In the performance analysis part it will be shown that the use of the GHX alone in cooling mode completely avoids the use of the heat rejector as it prevents the temperature to increase too much.

ELSE IF THEN HEAT REJECTOR PROCEDURES

CALL GHX SUBPROGRAM


ELSE

CALL GHX SUBPROGRAM NEW PUMP CONSUMPTION CALCULATION

ENDIF

ENDIF ENDIF

ENDIF When the temperature is no more in the acceptable range, the last part of the algorithm is roughly the same as in the previous system with the test on the heat rejector which can be used or not. In case of favourable conditions, the water goes first through the heat rejector and then in the ground heat exchanger in order to decrease one more time its temperature. Under unfavourable conditions, the water only goes in the GHX.


Figure 20: CT/DC and GHX water loop control algorithm

Model 4c: GHX and boiler

This last type of water loop provides an easy way to avoid a damageable decrease of the water loop temperature by using a boiler. However, in practice, this system is rarely used. The control algorithm is described hereafter. DECLARATION OF VARIABLES

IF AND THEN HEAT LOSSES CALCULATION ELSE


IF THEN IF THEN

In the first case of a heating dominated demand, the exhaust thermal storage temperature is above the limit (10°C); so the system works on its thermal inertia.



IF THEN

ELSE

CALL BOILER PROCEDURE


ENDIF

ENDIF

If the exhaust thermal storage temperature is below the temperature limit of 10°C, the ground heat exchanger is activated. The next step of the algorithm is to determine if the boiler must be switched on or not. If the exhaust ground heat exchanger water temperature is below a certain control temperature (often defined around 8°C) the boiler has to be activated. The use of the boiler is well adapted in case of highly heating dominated demand as it allows the loop temperature not to fall below a damageable point no matter how the ground heat exchanger is sized. The demand is now cooling dominated.

ELSE IF THEN


NEW PUMP CONSUMPTION CALCULATION ENDIF

ENDIF ENDIF As the temperature is still in the acceptable range the system uses its own thermal inertia; the ground heat exchanger is bypassed. In comparison with the Model 4b, the GHX is not used in the acceptable range considering that this type of system is mainly used in case of heavily heating dominated demand and as having a lower water temperature penalises the COP of heat pumps. Moreover the maximum temperature of the loop before the GHX is activated is here defined at 20°C instead of 32°C. Choosing the temperature of 20°C results in a trade off. On one side as this system is well designed for a heating dominated demand the temperature of the loop must certainly be quite high in order to keep a good COP of the heat pump so that the GHX is not used before 20°C. On the other side choosing a high maximum limit could lead to a sharp decrease of the temperature as the hot fluid is directly sent to the ground heat exchanger which has a temperature of 10°C.


Figure 21: GHX and boiler water loop control algorithm


IV. References

Andre, P., Bertagnolio, S., Caciolo, M., Franck, P.-Y., Rogiest, C., Sarrade, L., et al. (2008). Analysis of building

heating and coolign demands in the purpose of assessing the reversibility and heat recovery potentials.

Annex 48 of the International Energy Agency, Energy conservation in buildings and Community Systems

Programme.

Bernier, M. (2000). A Review of the Cylindrical Heat Source Method for the Design and Analysis of Vertical

Ground-Coupled Heat Pump Systems. 4th International Conference on Heat Pumps in Cold Climates, p. 1-

14.

Bernier, M. (2006). Closed-Loop Ground-Coupled Heat Pump Systems. ASHRAE Journal 48(9): 12-18.

Bernier, M. (2001). Ground-coupled heat pump system simulation. ASHRAE Transactions 107(1): 605-616.

Bernier, M., & Bourret, B. (1999). Pumping energy and variable frequency drives. ASHRAE Journal 41(12):

37-40.

Bernier, M., & Lemire, N. (1999). Non-dimensional pumping power curves for water loop heat pump

systems. ASHRAE Transactions 105(2): 1226-1232.

Bernier, M., Chahla, A., & Pinel, P. (2005). Long-term ground temperature changes in geo-exchange

systems. ASHRAE Transactions 114(2).

Bernier, M., Kummert, M., & Bertagnolio, S. (2007). Development and applications of test cases for

comparing vertical ground heat exchanger models. Beijing, China: Proceeding of Building Simulation 2007.

Bernier, M., Labib, R., Pinel, P., & Paillot, R. (2004). A Multiple Load Aggregation Algorithm for Annual

Hourly Simulations of GCHP Systems. HVAC&R Research 10(4): 471-487.

Bertagnolio, S., Caciolo, M., Corgier, D., & Stabat, P. (2008). Review of heat recovery and heat pumping

solutions. Annex 48 of the International Energy Agency, Energy conservation in buildings and Community

Systems Programme.

Bolher, A., Casari, R., Fleury, E., Marchio, D., & Millet, J. (1999). Méthode de calcul des consommations

d’énergie des bâtiments climatisés « ConsoClim » Présentation de la méthode. CSTB.

Bourdouxhe, P., Grodent, M., & Lebrun, J. (December 1994). Cooling-tower model developed in a toolkit for

primary HVAC system energy calculation—part 1: model description and validation using catalog data

simulation in buildings. Liège: Proceedings of the fourth international conference on system.

Braun, J., Klein, S., & Mitchell, J. (1989). Effectiveness models for cooling towers and cooling coils. ASHRAE


Cane, D., Clemes, B., & Morrison, A. (1998). Operating experiences with commercial ground-source heat

pumps- Part 1. ASHRAE Transactions.


Casari, R., Fleury, E., Marchio, D., Millet, J., & Bohler, A. (1999). Méthode de calcul des consommations

d’énergie des bâtiments climatisés « ConsoClim » Présentation de la méthode. CSTB.

Chapont, C., & Delandre, M. Air Conditioning by heat pumps on a water loop. ASHRAE Transactions:

Research.

Cooper, W. (1994). Operative experience with water loop heat pump systems. ASHRAE Transactions:NO-94-

26-1.

DOE (Department of Energy). (1980). DOE 2 Reference Manual Part 1. Berkeley, California: Version 2.1.

Dorgan, C., Linder, R., & Dorgan, C. (1999). Chiller heat recovery: Application guide. Atlanta, Georgia:

ASHRAE.

Energy Efficiency and Certification of Central Air Conditioners (EECCAC). (2003). ARMINES.

Geothermal Bore Technologies. (2007). Retrieved from http://www.geoclip.com/

Hydeman, M., & Gillespie, K. (2002). Tools and techniques to calibrate electric chiller component models.

ASHRAE Transactions 108(1): 733-741.

Hydeman, M., Taylor, S., & Winiarski, D. (2002). Application of component models for standard

development. ASHRAE Transactions 108(1).

Hydeman, M., Webb, N., Sreedharan, P., & Blanc, S. (2002). Development and testing of a reformulated

regression-based electric chiller model. ASHRAE Transaction 108 (2).

Jeffrey, J., Splitler, D., Fisher, D., & Xu, X. (2006). Simulation of Hybrid Ground Source Heat Pump Systems

and Experimental Validation. Liège: 7th International Conference on System Simulation in Buildings.

Kavanaugh, S. (1998). A Design Method for Hybrid Ground Source Heat Pump. ASHRAE TRANSACTIONS, V.

104, Pt. 2.

Kavanaugh, S., & Rafferty, K. (1997). Ground-Source Heat Pumps: Design of Geothermal Systems for

Commercial and Institutional Buildings. ASHRAE.

Lebrun, J. (2007). Introduction aux Machines et Systèmes Thermiques. Notes de cours. Université de Liège,

Belgium.

Lebrun, J. (2004). Thermodynamique appliquée et Intrduction aux machines thermiques. Liège: Université

de Liège, Département de thermodynamique.

Lebrun, J., Aparecida Silva, C., Trebilcock, F., & Winandy, E. (December 2002). Simplified models for direct

and indirect contact cooling-towers and evaporative condensers. Liège: 6th International conference on

System Simulation in Buildings.

Lebrun, J., Bourdouxhe, J.-P., & Grodent, M. (1999). HVAC1 KIT: A toolkit for primary HVAC system energy

calculation. Atlanta, GA: American Society of Heating , Refrigerating and Air-Conditioning Engineers.


Marchio, D., & Morisot, O. (1999). Modelisation simplifiée d'un aéroréfrigerant dans l'optique d'un calcul

des consommations d'énergie d'une installation de climatisation dans un bâtiment tertiaire. ENEA/CVA-

n°98.162R, CSTB.

Morisot, O., & Marchio, D. (1998). CCSIMOL : A few inputs model of variable flows cooling coil. Liège,

Belgium: Proceedings of the fifth International Conference on System Simulation in Buildings.

Morisot, O.; Marchio, D. (2002). Simplified Model for the Operation of Chilled Water Cooling Coils Under

Nonnominal Conditions. HVAC&R Research 8(2): 135-158.

Pietsch, J. (1990). Water-loop heat pump systems assesement. ASHRAE Trans. 96, pp. 1029–1038.

Sfeir, A., Bernier, M., & Joly, A. (2005). A methodology pumping energy consumption in GCHP systems.

ASHRAE Transactions 111(2): 714-729.

Stabat, P., & Marchio, D. (2004). Simplified model for indirect-contact evaporative cooling-tower behavior.

Applied Energy 78: 433-451.

Stanley, B., & Geoffrey, A. (2000). Foundations of Financial Management. Ninth Edition, Irwin Mc Graw Hill.

Underwood, C., & Yik, F. (2004). Modelling Method for Energy in Buildings. Blackwell Publishing.

Winandy, E. (2008). Production de froid et chaleur basse température, Notes de cours. Université de Liège.


APPENDIX 1: Recovery and reversibility potentials

A.1.1 Reversibility potential

The reversibility potential (REVpot) is defined, in absolute value, as the part of the total heating demand that can be covered by the chiller in heat pump mode, in MWh. The reversibility potential depends on the heating power capacity which can be reached by converting a chiller into a heat pump and on the non simultaneous demand of cooling and heating. In order to standardize this value, it can be divided by the total heating demand, obtaining the relative reversibility potential (Irev,pot).

These two quantities are calculated hour by hour, and they represent the value of the reversibility potential such as :

- The ―chiller‖ operates in priority in cooling mode; so reversibility is possible only when no cooling is required ;

- The maximum heating power available is assessed to be 0.8 maximum cooling power of the chiller9, the supplementary demand is assumed to be covered by the boiler.

No consideration on the emitter temperature levels required is taken into account at this stage.

A.1.2 Recovery potential

The recovery potential depends on the simultaneous heating and cooling demand and on the heat power available on chiller condenser (only consideration on energy recovery for space heating is taken into account here in office buildings, but heat recovery could also possible for Hot Domestic water and/or humidification). The recovery potential is calculated hour by hour as the percentage of heating demand which could be provided by a chiller condenser under the following conditions :

the ―chiller‖ is in operation in order to provide the cooling demand;

The maximum heating power available on the condenser is calculated based on energy

conservation principle such as (EER+1)/EER cooling power provided by the chiller at the step time, the supplementary heating demand is assumed to be covered by the boiler.

9 This value is based on data from manufacturer’s catalogues. Some differences can be noticed between chillers. For the selected reversible chiller, one can notice that at nominal conditions the cooling power is about 15% lower than the heating power. However, the nominal heating conditions are at 7°C outside air temperature which do not correspond to the worst conditions whereas the nominal cooling conditions are imposed to 35°C that is close to the worst conditions. If we considered the heating power at –5°C, the heating power is about 20% lower than cooling power at nominal conditions.


No consideration on the emitter temperature levels required is done here. The % of recovery is calculated as the ratio between the total heat recovery for space heating and the total space heating demand.


APPENDIX 2: Components modeling

A.2.1 Heat pumps

A.2.1.1 DOE-2 method

Chiller and heat pump plants are complex and dynamical systems which are difficult to accurately model. In this work several methods were investigated in order to simulate the behaviour of a heat pump plant. At the beginning the first method explained by Underwood and Yik (2004) was considered. In order to calculate the consumption of the heat pump, they recommend directly using a single equation based on eight correlation coefficients used with two independent variables such as the cooling capacity and the ambient air temperature. In the case of an air-to-water chiller, for instance, the correlation could be.

where is the current capacity and the electricity consumption of the chiller. However this method has several drawbacks:

Even if the values of the coefficients are given in the reference book, there is information only

on few types of heat pump systems.

The temperature of the exhaust water is not considered in this method, so an independent

variable is missing.

The second method investigated in this approach is called the DOE-2 method which has been developed by Hydeman et al. (2002). Its aim is to accurately determine the energy consumed by the heat pump given some fundamental inputs which could be easily measured by the user. The choice of this method has been motivated by the fact that the error prediction is 1.8% in terms of RMS 10 which is quite negligible compared to other methods in which an error which could reach 7% (Hydeman, et al., 2002). Also as mentioned above it utilises easily available measurements which brings a higher accuracy to the model. The sequence of this method is in 3 steps:

1) Firstly determining the CAPFT curve. Such curve represents the available capacity as a

function of evaporator and condenser temperatures thanks to six coefficients and two

independent variables. The method which will follow is described in the case of a chiller

where these variables are the entering condenser air temperature and the leaving evaporator

water temperature. However in the case of a heat pump variables are defined as the leaving

water temperature at the condenser and the supplying evaporator water temperature. The

CAPFT curve is calculated thanks to six full-load performance data points as follows:

where is defined as the cooling capacity at the evaporator in certain full load conditions. The value of will be discussed later. With six points of the CAPFT curve we can

determine the correlation equation (having the six coefficients) :

10 RMS= Root Mean Square


2) Secondly determining the EIRFT curve. Such curve represents the full-load efficiency as a

function of the evaporator and condenser temperatures. The principle is exactly the same as

in the case of the CAPFT curve but absorbed electrical power is now considered.

The value of will be explained by the end of this section. With six points of the EIRFT

curve we can determine the correlation equation (thanks to six coefficients):

3) Finally determining the EIRPLR curve. Such curve represents the efficiency as a function of

the unloading percentage. It could be found using several part load points in order to define:

A law for could be drawn as a function of the using several data points at part load.

where is defined as:

This curve is mainly determined by the behaviour of the compressor used in the heat pump as it defines whether the heat pump is operating in part load or in full load. Moreover, the shape of this curve could be differently based on the type of compressor and part load control system used in the heat pump.

Thanks to this method it is now possible to calculate the consumption of the chiller (or heat pump) for a certain capacity (which determines the PLR) and under certain operating conditions thanks to the following equation:

Having the consumption the EER (or COP) of the system can be determined. The DOE-2 Method can be schematically summarised as followed:

Figure 22: Schematic view of the DOE-2 Method

For the above mentioned and corresponding values we can either chose the nominal or the

maximum capacity and power related to the heat pump. Each of these possibilities has advantages and drawbacks. Choosing nominal values as reference values could be more commonly accepted but

Inputs

Six points under different

conditions at full load

Three points at part load

DOE – 2

METHOD

Outputs

CAPFT, EIRFT and

EIRFPLR curve

Electrical consumption of

the heat pump

Parameters


the conditions under which these values have been determined could vary from one heat pump to another. On the other hand maximal capacity is often given at some extreme conditions which do not represent the usual operation of the heat pump. However it has the main advantage to keep the PLR and CAPFT below the value of 1 so that all heat pumps curves can often be normalized. In the remainder of this study the maximum value will be used as the reference value.

A.2.1.2 Air cooled chiller

This type of heat pump is made of an air-cooled condenser and will only be used as a cold producer in the reference model (called Model 0). Many manufacturers provide information on this type of heat pump (Trane, Carrier …). In this work it has been decided to select the model with sufficient data provided by the manufacturer to develop an accurate model of the heat pump. In order not to promote any commercial model in particular this system will be designated ―Heat pump A‖. Also the power supplied by this heat pump is in the range of the demand which will be considered further on (around 500 kW).

Figure 23: Schematic and real view of an air-cooled chiller

Thanks to the amount of data that we collected it has been possible to develop the CAPFT and the EIRFT curve which can be represented as a function of the leaving water temperature and of the ambient air temperature.

Entering air temperature [°C]


Figure 24: The CAPFT and the EIRFT curve

Both curves significantly change when approaching some extreme conditions. Furthermore in these curves the extremes low and high points are symmetrically opposed versus the leaving water temperature. This result is easily understandable as the CAPFT is greater in the most favourable conditions while it is the opposite for the EIRFT. These conclusions could be applied to all types of heat pump which will be considered later. In this case the manufacturer provided sufficient information to determine the EIRFPLR curve which is often defined for the type of compressor used. It is important to keep in mind that manufacturers rarely provided data at part load. The way to overcome this issue is to use general compressor curves. Part of these curves have already been developed using the data coming from the COOLTOOLS source. It has been demonstrated that the nature of the curve is significantly the same for a defined compressor type no matter who the manufacturer is.

Figure 25: The EIRFPLR curve of ARL 144 from Mc Quay

A.2.1.3 Reversible air-to-water heat pump

This system has the possibility to work either as a heat pump or as a chiller thanks to a 4 ways valve which can change the direction flow of the refrigerant. This heat pump will be called heat pump B. It is important to keep in mind that the correlation is completely different in case the heat pump

Entering air temperature [°C]


operates in cooling mode or in heating mode. Two different fictive systems with their own correlations depending on the mode are considered further on. As expected no data about part load operation were available from the manufacturer. To overcome this problem the chiller Scroll model of the IEA Annex 43 project has been used to generate the EIRPLR curve. In conclusion the CAPFT and the EIRPLR curves are available via the manufacturer’s data and the EIRPLR curve has been taken from another project.

Figure 26: Schematic and real view of a reversible air-to-water heat pump

A.2.1.4 Water-to-water heat pump

This type of equipment is used in the exhaust-air heat pump. It could work in two different modes depending on the location of the useful effect required: either at the exhaust of the evaporator or at the exhaust of the condenser. Even if two modes could be observed the machine always works in the same way: absorbing energy at the evaporator and releasing it at the condenser. This simplifies our modelling of the heat pump as only one correlation set has to be defined. This system will be called heat pump C.

Figure 27: Schematic and real view of a water-to-water heat pump

According to the DOE-2 method the two independent variables are now the exhaust evaporator water temperature and the exhaust condenser water temperature. The EIRFPLR curve of the screw compressor has been taken from Colen (2009). One of the aspects of the modelling to be highlighted is passing from the evaporator power to the condenser power. Depending on the operating conditions we must determine the condenser power from the evaporator power or the other way around. Theoretically another set of laws should be generated with other variables to simulate the condenser power as the device work is chiller mode. However due to the lack of manufacturer’s information a simplified equation will be used:

with the EER determined thanks to the DOE-2 method. In this concrete case no data on the heat power rejected at the condenser was available from the manufacturer; so it was not possible to evaluate the error linked to using this equation. It will be done later for another type of heat pump.


A.2.1.5 Dual condenser heat pump

This type of heat pump is only used in the dual condenser system which will be described later. The heat pump has two different condensers and is always in a chiller mode but with two different modes:

When there is no heat demand the air condenser works like in any air-to-water chiller.

When there is a heat demand the main condenser becomes the water condenser and the

system works in the recovery mode. In such case there are two possibilities:

o The heat recovered at the condenser is lower than the heat demand and the water flow

takes as much energy as it can;

o Or the heat recovery is higher than the heat demand. In that case one will assume that

both condensers are functioning together with the air condenser rejecting the difference of

the heat recovered and the heat demand.

However (according to Philippe Struvay11) the control between both condensers is done switching on and off alternatively each condensers. In this work a perfect control on the condensers is assumed so that the utilisation of the condensers exactly matches the demand. This system will be called heat pump D.

Figure 28: Schematic and real view of dual condenser heat pump

Calculating the consumption of this type of heat pump has been somewhat more complex than in the other cases. With no heat demand the DOE-2 method provides the consumption and obviously the EER as the system is considered as a simple air-to-water chiller. When the water condenser is used the correlations change as the system is now considered as a water-to-water heat pump with the difference that we must add the fan consumption when the air condenser is used which is not given in the manufacturer’s data. An easy way to solve this problem was to consider that the fan consumption is proportional to the power of the condenser. This allows determining a coefficient based on the nominal capacity and power.

Thanks to this coefficient it has been possible to determine the additional amount of energy consumed by the fan when the system operates in a recovery mode and finally the COP of the system. As it was the case in the water-to-water heat pump, we consider the relation between the capacity of the evaporator and of the condenser as followed:

11 Senior Engineer at Camair Air conditioning.


In this case, it was possible to determine the error generated using the equation as the manufacturer provided some values at full load for the heat rejected at the condenser. The followed graph can be drawn for the error at full load.

Figure 29: Analysis of the error by using the simple equation at full load

The graph shows that the error strongly varies when approaching the extreme conditions and reaches 10 % at 4°C-55°C water temperatures which are the worst conditions. On the other hand when water temperatures are not extreme the error value sharply decreases to reach 0%. Even if the value can reach 10% at a specific point the mean of the error on the whole range of values is 3% which shows that the above equation can be reasonably used. It is important to keep in mind that this analysis of the error calculation is done at full load; the error at part load is not known.

A.2.1.6 Reversible air-to-water heat pump used in a water loop

As stated above this type of heat pump is only used in water loops but is very different than the air-to-water heat pump already presented. In fact this type needs its own correlations as the range of power (around 5,000 W) is completely different than the other air-to-water model developed earlier (around 500 kW). This type of heat pump is used as an individual room conditioner and takes its heat from the water loop and rejects it into the air flow. There are mainly three capacity ranges for this type of heat pump: 2,000 W, from 2,000 to 5,000 W and from 5,000 to 20,000W. It can also function in reversible mode and it is characterised by a high EER (up to 8 in certain conditions). In this type of heat pump only little Scroll compressors are used. This system will be called heat pump E.

Figure 30: Schematic and real view of a little reversible air-to-water heat pump

Condenser leaving water temperature [°C]


For the modelling of these heat pumps, correlations have to be modified as in this case the only variable is the temperature of the water loop. The entering air temperature cannot vary or at least only in a small range as it is set as the comfort temperature of the room (varying between 23°C is heating mode to 27 °C in cooling mode). Correlations are now defined as:

The regulation of this type of heat pump is very complex as it relies on many parameters. Also in a single building, as behaviour of each heat pump can vary depending on how it is utilised, it is not possible to determine the required capacity for each heat pump. In order to simplify this behaviour of the entire set of heat pumps the PLR will always be taken at a value of 1. In this context the EIRFPLR curve is not used anymore.

A.2.1.7 Heat pumps description

Description Air-to-water Chiller

Name Heat pump A

Reference Capacity 418 500 W

Reference Power 132 100 W

Compressor type Screw

Refrigerant R 407C

Table 2: Description of the Heat pump A

Description Reversible Air-to-water Heat pump

Name Heat pump B

Reference Capacity-cooling 451 400 W

Reference Capacity-heating 467 400 W

Reference Power-cooling 165 100 W

Reference Power-heating 172 000 W

Compressor type Scroll

Refrigerant R 407C

Table 3: Description of Heat pump B

Description Water-to-water Heat pump

Name Heat pump C

Reference Capacity 469 000 W

Reference Power 103 000 W


Refrigerant R 134a

Table 4: Description of the Heat pump C

Description Dual condenser Heat pump

Name Heat pump D


Reference Capacity-recovery 438 000 W

Reference Power-cooling 110 200 W

Reference Power-recovery 106 000 W


Refrigerant R 134a

Table 5: Description of the Heat pump D


Description Reversible air-to-water Heat pump

Name Heat pump E


Reference Capacity-recovery 3 100 W

Reference Power-cooling 260 W

Reference Power-recovery 600 W

Compressor type Reciprocating

Refrigerant R 22

Table 6: Description of Heat pump E

A.2.2 Water Pump

This model is easier to apply as it uses correlations based on experiments reported in the literature. Depending on the type of systems studied, the pump used in the hydraulic network has to be distinguished from the ones used in the water loop, as this latter type is used in much larger hydraulic networks.

A.2.2.1 Hydraulic network pump

This model was easy to establish, as ASHRAE directly provides correlations between the hydraulic network pump and the water flow passing through it (ASHRAE, 2004). This reference gives the following data for non-residential buildings:

Type of hydraulic network Correlation Cold water network 0.3487 kW/(kg/s) Hot water network 0.3012 kW/(kg/s)

Table 7: Hydraulic network pump consumption correlations

A.2.2.2 Water loop pump

Concerning this type of pump the aim of the study was to develop the same kind of correlation for the power consumed in relation to the water flow. To reach this goal, we used data developed by Cane et al. (1998) for various water loop systems. In its paper Cane studied many types of water loop systems and their specifications. Thanks to this reference, it has been possible to develop a correlation curve using such data.

0 10 20 30 40 50 60 70 80 900

10

20

30

40

50

60

70

80

Flowrate [kg/s]

Pu

mp

co

nsu

mp

tio

n [k

W]

Pumpconsumption=-1.22376 + 0.726844·FlowratePumpconsumption=-1.22376 + 0.726844·Flowrate

Figure 31: Water loop pump consumption correlations (Cane,et al., 1998)

R²=80%


The following graph represents the pump consumption versus the water flow rate. The water loop pump curve increases much more rapidly than the other ones which makes sense as the hydraulic network is larger in the case of a water loop system.

Figure 32: Pump consumptions in function of the water flow rate

A.2.3 Boiler

A.2.3.1 General hypothesis

Concerning this boiler model some points have to be highlighted. Firstly a constant efficiency boiler model will not been considered; the efficiency of the boiler will be determined through the model. Secondly, according to Cooper (1994), the supply water temperature should not be below a certain point defined close to 60°C for a usual boiler. This limit highly depends on the type of boiler used and can decrease in the case of a condensing boiler. However in the case of a hot hydraulic network considered in the work the temperature needed is in the range of 40-50 °C. This problem will been addressed with the assumption that a mixing hydraulic network is placed before entering the boiler and can perfectly control the valves in order to keep the right temperature difference and the right water flow. This hydraulic network can be described as followed:

Figure 33: Boiler regulation network

In this network the left valve (A) controls the incoming water flows in order to have a supply temperature high enough not to damage the boiler. The right valve (B) perfectly controls both incoming water flows in order to have the right exhaust temperature This mixing network can been modelled through equations using the indication ―1‖ for temperature of the water directly entering or going out of the boiler.

A

B


Firstly the nominal conditions have to be defined. Generally the temperature range of 60 to 80 °C is used referring to the supply and exhaust conditions. These nominal conditions will characterise the water flow going through the boiler which will always remain constant.

Secondly the network temperatures have to be defined.

The following mixing equations will be used in order to identify the water flow rate and the proportion of the water flow which is sent to the boiler under nominal conditions The following equations are the mixing equations of respectively the A and the B valve.

As mentioned earlier we will consider that the boiler water flow rate will be constant in the future; however the supply water temperature of the boiler and the proportion of the water flow rate which is bypassed can vary. However the exhaust temperature of the boiler and the exhaust boiler network temperature which is determined by the application remain constant.

Therefore, depending on the network supply water temperature the proportion of bypass

and the boiler supply water temperature will vary according to the following

equations.

Considering an exhaust boiler temperature of 80 °C and an exhaust boiler network temperature of 45°C, the supply boiler temperature and the percentage of bypass can been drawn as followed versus the supply boiler network temperature.

10 15 20 25 30 35 40 450

10

20

30

40

50

60

70

80

0.4

0.6

0.8

1

tw,su,boiler

t w,e

x,b

oiler,

1,n

, t

w,e

x,b

oiler,

t w,s

u,b

oiler,

1

tw,ex,boiler,1,ntw,ex,boiler,1,n

tw,ex,boilertw,ex,boiler

tw,su,boiler,1tw,su,boiler,1

Xb

yp

ass

XbypassXbypass


Figure 34: and versus

It is clear that the more is close to the more the percentage of bypass increased as less water needs to be heated up. Moreover in the same way the supply boiler temperature approaches the exhaust one.

A.2.3.2 Reference boiler model

The first boiler model is essentially based on the model described in ASHRAE Toolkit (Lebrun et al., 1999). According to Lebrun et al., the model of the boiler mainly uses an adiabatic burner, a water-gases heat exchanger and a water-ambient air heat exchanger.

A.2.3.3 Adiabatic burner

When considering the burner as adiabatic no heat transfer with the ambient air is considered. This part of the model can be developed as having the air and the fuel as inputs and the exhaust gases as output. More precisely, we calculate first the difference of power required to bring the initial temperature of the fuel and the air to the reference temperature. This step is required as the LHV is evaluated at this temperature. In this case the reference temperature is set at 25°C.

The combustion process is determined by adding the energy delivered by the fuel and considering that the combustion is complete (which is represented by the fourth heat transfer).

Considering a complete combustion

The determination of the exhaust gases temperature is done through the heat equilibrium of the burner (hypothesis of adiabatic burner).

and considering that:

It is important to keep in mind that the mass conservation principle acts through:

where , and are respectively the flow rate of the air, of the fuel and of the exhaust gases.

A.2.3.4 Water-gases exchanger

This part of the model aims at determining the exhaust water temperature passing through the

burner. The principle of the modelling is based on the method completed by Lebrun et al. in the HVAC Toolkit (1999). This method will evaluate the efficiency of the exchanger using some heat transfer coefficients. Firstly, the maximum and minimum capacity flows have to be determined. A capacity flow is the multiplication of the mass flow rate by the specific heat of the fluid. Its dimensions are [W/°C].


Secondly the NTU (Number of transfer unit) is determined using the heat transfer coefficient. According to Lebrun et al.(1999), a value of

can be determined. However this value varies depending on the gases flow rate according to the equation developed by Lebrun et al. (1999):

The following equations allow determining the effectiveness of the exchanger.

assuming a counter flow heat exchanger. Using the program it is now possible to determine the heat transfer and the theoretical water exhaust temperature of the boiler ( ).

where is the water temperature at the exhaust of the first heat exchanger.

A.2.3.5 Water-air exchanger

In reality the temperature found in the previous section is not the actual one as a certain heat loss has to be taken into account. To model this heat loss a fictitious heat exchanger between the water and the air is considered. Using this model is a good trade off because in boilers water pipes are placed around the core and they are the first element which losses energy through the environment. According to Lebrun et al. (1999), a value of

can be supplied, which hopefully is much lower than the previous .

As the air temperature is constant the model of the exchanger becomes a semi-isotherm heat exchanger which changes the equations of the effectiveness.


The last equation provides the exhaust water temperature.

A.2.3.6 Operation of the boiler

The question which must be asked at this point is: ―Is the exhaust water temperature high enough?‖. The answer will depend from the set point temperature and will determine the mode of operation of the boiler. Indeed, if the exhaust water temperature is above the set temperature, the boiler will function part of the time to have an average exhaust water temperature equals to the set temperature. The ratio of the operating time to the non-operating time (θ) is determined by:

where is the exhaust water temperature when the boiler is not operating. If is smaller

than , the boiler will always operate at full load. As determines the percentage of operation, the same equation can be used to know the mean of the fuel consumption.

where is the consumption of the boiler when it functions at full load and is

obviously equal to zero. Many boilers do not only have a ON/OFF mode but more than these two simple modes. Considering the case of a boiler which has a ON/MIN/OFF mode the following algorithm can be drawn. IF THEN

The boiler always functions at full load. ELSE

IF THEN

is defined as the ratio between the time at full load and the time at minimum load. The boiler operates cycling between both loads. ELSE

IF THEN

ELSE =0 ENDIF ENDIF ENDIF Firstly, the boiler operates cycling between switching off and the minimum load. Secondly the boiler does not function if the temperature needed is below the exhaust water temperature when the boiler is switched off. Finally it is possible to evaluate the general efficiency of the boiler:

According to Lebrun et al. (1999), the electric consumption of a gaseous boiler is:


This model seems very accurate and rigorous. Unfortunately it cannot be used in practical situations as it creates numerical instabilities. More precisely, it has not been possible to implement it in the water loop system as this system is more complex and needs more iteration. The graph here below shows the evolution of the efficiency and the operating ratio of a ON/OFF boiler assuming a nominal exhaust temperature of 80°C. The efficiency slowly decreases as the supply temperature approaches the exhaust one. On the other hand, the operating ratio is constant in a certain range below 53°C and decreases to zero as the supply temperature tends to match the exhaust one. Below the limit temperature of 53°C the boiler is no more capable of furnishing an 80°C exhaust temperature which can be seen by this constant operating ratio area.

40 45 50 55 60 65 70 75 800.8

0.85

0.9

0.95

1

0

0.2

0.4

0.6

0.8

1

tw,su [C]

bo

iler

boilerboiler

Figure 35: Efficiency and functioning ratio versus supply temperature

A.2.3.7 Simplified boiler model

The problem created by the previous model in the water loop systems has been tackle in a very simple way. The principle was to create an efficiency table generated by the previous model which allows determining the fuel consumption. The second model was developed as follows. Firstly the capacity of the boiler is determined assuming that the boiler is designed to reach the exhaust conditions with given parameters. This first set differs from the first model as the previous one assumed that the exhaust temperature could not be reached if the boiler was under dimensioned.

To determine the previous mixing equations are still used. Secondly the current capacity is compared to the nominal one in order to establish the PLR (part load ratio).

Thirdly the efficiency is interpolated in the table using two variables: the supply temperature ( ) and the PLR. This table has been created using the previous model and with a fictitious boiler having a reference efficiency of 90% with the temperature difference of 60 to 80 °C operating at half capacity. Searching in the table provides the which is the total efficiency of the boiler. However if a different type of boiler is used the general efficiency of the boiler can be modified as followed:


where is determined by the table, is the efficiency of the boiler considered working at 60-80°C and is the reference efficiency (90%) used to generate the table.

Finally the fuel consumption is determined by the general equation:

Considering the table, as said before, it has been developed using the previous model under 60-80°c conditions. In fact two tables have been generated: one for the ON/OFF boiler and a second one for a two stages boiler. Also, the at PLR=0 is taken the same as in PLR=5% in order to avoid numerical instability.

Table 8: for ON/OFF boiler

A.2.4 Cooling tower

The model has been developed in three steps. 1. a sizing of the cooling tower has been done in order to determine the heat transfer coefficient;

2. the simulation of the cooling tower is done using the previously calculated heat transfer

coefficient;

3. the operating mode is determined comparing the exhaust water temperature to the set

temperature defined by the user.

A.2.4.1 Cooling tower sizing

The first part of the model is mainly based on the structure developed by Stabat et al. (1999) in the ―Consoclim‖ reference created by the ―Ecole des Mines‖ (Paris).The first part of the sizing step requires to determine the heat transfer coefficient related to only one nominal point12. At the beginning the air flow rate is determined through:

12 Such as the supply and exhaust water temperature, air temperature, water flow rate, etc.


where was determined by the ambient conditions and by a hypothesis on the exhaust air temperature in nominal conditions, considering that it is saturated . Then a sizing procedure determines the heat transfer coefficient of the water and air side considering the cooling tower as a counter flow heat exchanger. The effectiveness of the cooling tower considered as a heat exchanger is calculated according to the -NTU theory.

The second equation provides the efficiency of the heat exchanger. According to the - NTU theory and considering a counter flow heat exchanger it is possible to determine the coefficient of heat transfer using :

and considering that

while the coefficient is determined by

Marchio et al. (1999) assume that a relation could be established between the heat transfer coefficients as:

with as a repartition key (Marchio, et al., 1999) and having calculated through the the

– NTU method.

A.2.4.2 Cooling tower simulation

The simulation part is characterised by the research of the total current heat transfer coefficient according to the value given at a nominal point. According to Marchio et al. (1999) heat transfer coefficient laws are:

where is the air flow rate and the water flow rate.

is the general cooling tower heat transfer coefficient which allows computing the exhaust water temperature by solving the equation of a simple counter flow heat exchanger using the – NTU method.


A.2.4.3 Operating mode

Finally, the exhaust water temperature is compared to the exhaust temperature set point which determines the percentage of utilisation of the cooling tower.

if considering a ON/OFF control. With this percentage it is possible to compute the consumption of the cooling tower using the nominal electricity consumption (which has to be provided at the start).

If the temperature supplied by the cooling tower is higher than the set temperature, will be equal to 1 and the cooling tower will always operate. It should be highlighted that the nominal consumption of the cooling tower was taken from the ASHRAE Toolkit (Lebrun, et al., 1999) with the correlation:

for an axial fan.

A.2.5 Dry cooler model

The dry cooler model is basically the same as the cooling tower model. However as we work without any water spraying, only the dry air temperature is considered and no iteration caused by the web bulb temperature is present. Like in the cooling tower description the model is divided into three parts.

A.2.5.1 Dry cooler sizing

The first part of the sizing step is to determine the heat transfer coefficient related to some nominal data. Using the logarithmical temperature difference it is possible to establish the total coefficient of transfer .

while

Moreover Marchio et al. (1999) assume that in the case of a dry cooler a relation could be established between the heat transfer coefficients:

with as a repartition key (Marchio, et al., 1999).

A.2.5.2 Dry cooler simulation

The simulation part is characterised by the research of the total heat transfer coefficient according to the value given at a nominal point. According to Marchio et al. (1999) the heat transfer coefficient laws are:


where is the volumetric flow rate of the air. The next step of the simulation is to find the exhaust water temperature by solving the equation of a counter flow heat exchanger like in the case of a cooling tower.

A.2.5.3 Operating mode

Determining the operating mode of the dry cooler is exactly the same as in the case of the cooling tower. The aim of this part is therefore to determine which is the percentage of the utilisation of the dry cooler. The nominal electrical consumption of the dry cooler has been determined through correlation using manufacturer data:

A.2.6 Cooling coil model

This model is only used in the exhaust heat pump system in order to absorb as much energy as possible from the exhaust ventilation air which energy will be used in a heat pump. Like in some other models the sizing and the simulation part are present.

A.2.6.1 Cooling coil sizing

A.2.6.2 First attempt

The first part of the sizing model is done exactly like in the dry cooler model with the aim is to determine the global heat transfer coefficient under nominal conditions thanks to the logarithmical temperature difference method explained by Lebrun (2004). According to Bolher et al. (1999) this coefficient permits to establish the heat transfer coefficient of each part of the cooling coil:

where is the specific heat of the air and is the specific heat of the fictive fluid which allows

considering the air flow loaded of water. In the sizing part, it is assumed that the exhaust air relative humidity is 90% in wet regime. This procedure seems very rigorous. However, it has a main drawback which the need to have the exhaust air temperature as an input. In most of the cases such temperature is not easily determined which explains why another sizing part model has been developed.

A.2.6.3 Second attempt

In order to simplify the model, the heat transfer coefficients has been directly determined according to the manufacturer’s data. This modification allows reducing the number of inputs needed by the


system to only describe the air mass flow rate. According to the manufacturer data the following laws have been established under nominal conditions (mainly water temperature difference = 5°C with

=23°C):

Moreover it allows determining the water flow rate needed through the cooling coil.

The following graphs provide a comparison between both models of the heat transfer coefficients and the water flow rate. Indication 1 and 2 are respectively determined for the first and second attempt.

1 2 3 4 5 6 7 8 9 100

5000

10000

15000

20000

25000

30000

35000

Ma [kg/s]

AU

[W

/K]

AUa,1AUa,1

AUa,2AUa,2

AUw,1AUw,1

AUw,2AUw,2

1 2 3 4 5 6 7 8 9 10

0

1

2

3

4

5

6

7

Ma [kg/s]

Mw

[

kg

/s]

Mw,1Mw,1

Mw,2Mw,2

Below an air flow rate of 5Kg/s all coefficients are pretty much in the same range. Above this value,

and spread apart. Hopefully in the rest of the study the air flow rate will rarely be over such limit which allows using the simplest second model instead of the first one in order to avoid determining the exhaust air temperature.

A.2.6.4 Cooling coil simulation

Having these heat transfer coefficients it is possible to establish the global heat transfer coefficient using the theory developed by Bohler et al. (1999) as well as the nominal and the real heat transfer coefficients. The rest of the simulation of the model is done using the model of a simple heat exchanger using a fictitious fluid.


In order to calculate the exhaust water temperature it is important to determine if the air condenses. According to Bohler et al. (1999) the drew point temperature can be determined using an explicit equation:

where is the humidity ratio of the air. A test must be done to see if the exhaust air is above or below this temperature. Even if in both cases typical counter flow heat exchanger equation will be applied, the value of will be changed according to Morisot et al. (2002) if the air condenses, as:

with and

Figure 36: Real view of a cooling coil

A.2.7 Ground heat exchanger

A.2.7.1 Ground heat exchanger model

The modelling of a ground exchanger could be the most difficult task to achieve because of certain rationales:

The geometry of the ground exchanger could be hard to model;

The phenomena of thermal inertia of the ground has to be taken into account;

The evaluation of the heat flow requires using Bessel functions.

Also, several physical properties of the ground are quite difficult to evaluate. The model used in this work is directly based on a model developed by Bernier (2006). According to Bernier (2001), many assumptions have been made in order to develop the model:

An analytical solution of the infinite cylindrical heat source was developed to take into

account the influence of the heat rejected or extracted on the surrounding ground

temperature. (Bernier, A Review of the Cylindrical Heat Source Method for the Design and

Analysis of Vertical Ground-Coupled Heat Pump Systems, 2000)

The thermal gradient on the depth of the borehole is not taken into account. The mean

temperature of the inner surface is calculated.

The thermal convective resistance of the fluid and the thermal conductive resistance of the

pipe and the grout are taken into account. However dynamics effects are neglected.


The effect of the ambient air temperature affecting the ground temperature is neglected. This

effect is negligible after 5-10 meters deep.

The effect of three dimensional heat transfers and the interaction between two boreholes of

the same borefield are neglected. These effects are negligible for a simulation on a period

shorter than a few years. (Bernier, Chahla, & Pinel, Long-term ground temperature changes in

geo-exchange systems, 2005)

The undisturbed ground temperature is supposed to remain constant on the depth of the

borehole.

Without entering into details, this model is divided into different parts: 1. Evaluating through an algorithm the MLAA (Multiple Load Aggregation Algorithm) (Bernier,

et al., 2004). This method is based on the principle that past thermal charges also influence

the present solving of the model. This is mainly due to the thermal inertia of the ground. This

technique allows reducing the thermal ―history‖ of the ground by aggregating all the past

charges and so the number of terms to evaluate during the solving. The older the charge the

less it influences the present equations.

2. Evaluating the penalty temperature which allows estimating the impact of thermal pollution

of the ground (a modification of the ground temperature). As this impact could only occur

after 5 years of utilisation it will be neglected in a first evaluation.

3. Calculating the temperature of the inner surface of the tube via the calculation of the thermal

resistance of the ground and considering that it is the same for the whole tube.

4. Calculating the different thermal resistances thanks to the provided data and through the

evaluation of the Nusselt and Fourier numbers.

5. Calculating the exhaust water temperature using a simple model of semi-isotherm exchanger

between the fluid and the inner surface temperature.

considering a semi-isotherm heat exchanger

where is the specific heat of the glycol water and is the temperature of the borehole

inner surface.

Many assumptions have to be done for this ground exchanger; for instance on ground properties, borehole configurations, etc... Values for all these parameters have been taken as common values from different papers coming from Bernier’s references. In practice the simulation will be done for only one year, which neglects the effect of thermal pollution of the ground. Considering a sinusoidal supply temperature, from the following graph it appears that the inner surface temperature (called ) precisely follows the variation of the supply temperature. The fluid is cooled down during its journey and the exhaust temperature is always between the supply and the inner surface temperatures.

A.2.7.2 Ground heat exchanger parameters

This section describes the parameters’ value used by Bernier his ground heat exchanger model.


Parameters Value Units

Borehole

borehole diameter 0.152 [m]

borehole depth 100 [m]

borefield aspect ratio 1.25 [-]

borehole spacing 5.5 [m]

Soil

thermal diffusivity 0.082 [m^2/day]

thermal conductivity 2.1 [W/m-K]

Pipe


Interior diameter 0.0274 [m]

exterior diameter 0.0335 [m]

Grout


Glycol water

fluid specific heat 3960 [J/kg-K]

fluid density 1022 [kg/m^3]

viscosity 0.003081 [Pa-s]

Conductivity 0.49 [W/m-K]

Table 9: Ground heat exchanger parameters

A.2.8 Thermal storage

The model is based on E. Winandy (2008). According to this model two parameters which have to be provided can describe the entire thermal tank: the volume of the tank and the coefficient of heat losses. According to Pietsch (1990), the volume of the thermal storage could be split into two parts: the volume in the loop and the volume of the eventual supplementary tank. The volume in the loop is determined following the law ―14 litres per kW installed‖ and the volume of the tank should be determined with the law ―50 to 100 litres per kW installed‖. Therefore the volume of the tanks should be at least 3 times bigger than the volume of the loop itself.

Theoretically the balance of energy can be written as the following equation.

where and

It means that the increase of internal energy of the water is balanced by the difference of entering and exhausting enthalpy and heat losses. However in order to not complicate the model heat losses will be neglected in a normal operation as the thermal storage temperature is quite similar to the ambient

one. Moreover is calculated through:


where is the initial temperature of the storage. The second equation allows calculating the internal temperature of the thermal storage. The graph below shows the variation of the internal thermal storage temperature according to a sinusoidal supply water temperature.

IEA48 Simulation Tools: Reference Book - · PDF fileAuthors: Stéphane ... o Sizing of...

Documents

Transcript of IEA48 Simulation Tools: Reference Book - · PDF fileAuthors: Stéphane ... o Sizing of...