Bachelor of Science Thesis
KTH School of Industrial Engineering and Management Energy Technology EGI-2018
SE-100 44 STOCKHOLM
Electricity demand analysis of a primary pump system in a district
heating system
Guizhou, China
Emil Deckner
Isabella Rambert
-2-
Electricity demand analysis of a primary pump system in a district heating system
Guizhou, China
Emil Deckner Isabella Rambert
Approved
20180606
Examiner
Viktoria Martin
Supervisor
Viktoria Martin
Commissioner Contact person
-3-
District Heating Systems (DHS) in China. Two students from the Royal Institute of Technology (KTH) performed this field study at Zhejiang University (ZJU), in the year of 2018.
During the last decades, China has made a substantial economic development with a dramatic increase in BNP which is transforming China into an industrialized country. Improving the life conditions, and managing the urbanization of such a big population, causes challenges when this transformation must be sustainable to meet the climate goals. In this transformation process, DHS play an important role in terms of providing fair living conditions. To optimize these systems, and thus make energy reductions when possible, is the key foundation that will enable the expansion of effective DHS and enhance sustainable development.
By modeling and simulating DHS, different configurations can be compared and the best configuration alternatives can be found. This report focuses on comparing Conventional Central Circulating Pump system (CCCP system) and Distributed Variable Speed Pumps system (DVSP system) to determine the most electric energy efficient configuration of a DHS. The results were used to compare the outcome of previous studies claiming that the DVSP system is more efficient than the CCCP system. In this project, a primary pump system of a DHS in the Guizhou province, China, was investigated. In order to compare the two different configurations of the DHS, two steady state models were created in the district energy optimization software Termis. One model for each configuration. Further, four scenarios were simulated for a specified set of input data. In scenario I and II, the electricity demand in a DVSP system during winter and summer respectively, were investigated. In scenario III and IV, the same simulations were made but for a CCCP system.
All simulations indicated that the electricity demand in a DVSP system were lower compared to a CCCP system. This is due to the fact that no throttling of hydraulic head is needed in a DVSP system because it provides the correct pressure change necessary for every consumer heat load. Further, the largest reductions occurred during the winter period when the heat load was higher. This is because a greater heat load is correlated with a greater saving potential.
Accordingly, the simulation results confirmed previous studies about DVSP systems being a more electric energy efficient configuration of a District Heating (DH) network.
-4- Sammanfattning
Denna rapport redovisar data och analyser inom området för modellering och simulering av fjärrvärmesystem i Kina. Två studenter från Kungliga Tekniska Högskolan (KTH) genomförde detta projekt som ett utbyte på Zhejiang University (ZJU) år 2018.
Under de senaste årtiondena har Kina gjort en avsevärd ekonomisk utveckling vilket har transformerat landet till ett industriland. Att förbättra livsvillkoren och möjliggöra urbaniseringen som sker i Kina medför stora utmaningar för att denna transformering ska ske hållbart, i syfte att möta klimatmålen. I denna process utgör fjärrvärmesystem en viktig roll för att förbättra levnadsvillkoren. Att utveckla dessa system och göra energibesparingar kommer vara en nyckelfaktor för att utveckla hållbara fjärrvärmesystem.
Genom att modellera och simulera fjärrvärmesystem kan olika konfigurationer jämföras så att den bästa konfigurationen identifieras. Den här rapporten fokuserar på att jämföra Conventional Central Circulating Pump system (CCCP-system) och Distributed Variable Speed Pumps system (DVSP-system) för att identifiera vilken som är den mest elektriska energieffektiva varianten i ett fjärrvärmesystem. Resultaten från denna studie jämfördes med tidigare studier inom området som visar att DVSP-system är mer energieffektiva. I detta projekt studerades det primära pumpsystemet i ett fjärrvärmesystem i Guizhouprovinsen i Kina. För att jämföra de två olika konfigurationerna av fjärrvärmesystemet skapades två steady state modeller i Termis, ett energioptimeringsprogram för fjärrvärmesystem. En modell för respektive konfiguration. Fyra olika scenarion simulerades för specifik given indata. I scenario I och II studerades elektricitetsbehovet för ett DVSP-system för vinter respektive sommarförhållanden. I scenario III och IV, gjordes samma simuleringar men för ett CCCP-system.
Alla simuleringar indikerade på att elektricitetsbehovet i ett DVSP-system är lägre jämfört med ett CCCP-system. Detta beroende på att ingen strypning av trycket i ledningar sker i ett DVSP- system då de utspridda pumparna förser varje sub-station med ett korrekt tryckfall. Den största reduceringen av elektricitetsbehov inträffade under vinterperioden när värmebehovet var större.
Detta då ett större värmebehov ger utrymme för större besparingar. Slutsatsen blev således att resultaten bekräftade tidigare studier som visar att DVSP-system är den mest eletricitetseffektiva konfigurationen av ett fjärrvärmesystem.
-5- List of diagrams
Diagram 1, Electricity demand in the four simulated scenarios [kW] ... 33
Diagram 2, Reductions during summer and winter [%] ... 34
Diagram 3, Reductions during summer and winter [kW]... 34
Diagram 4, Electricity demand in the four simulated scenarios, sensitive analysis [kW] ... 35
Diagram 5, Reductions during summer and winter, sensitive analysis [%] ... 35
Diagram 6, Reductions during summer and winter, sensitive analysis [kW] ... 36
Diagram 7, Percentage change of the results ... 36
-6- List of equations
Equation 1, Law of Kirchoff ... 23
Equation 2, Electricity demand for water pump ... 23
Equation 3, Pressure change through valve ... 24
Equation 4, Heat conduction through cylindrical layer 1(2) ... 25
Equation 5, Heat conduction through cylindrical layer 2(2) ... 25
Equation 6, Heat loss from insulated pipe ... 25
Equation 7, Total Thermal Resistance... 25
Equation 8, Pressure loss using the Colebrook and White friction ... 26
Equation 9, Pressure loss using the Hazen-Williams friction ... 26
Equation 10, Colebrook and White’s formula ... 26
Equation 11, Darcy friction factor... 27
Equation 12, Mass flow at substations ... 28
-7- List of figures
Figure 1, System overview, middle line ... 18
Figure 2, Structure of the middle line (CCCP) ... 19
Figure 3, Schematic of CCCP DHS ... 20
Figure 4, Diagram of hydraulic balance (CCCP) ... 20
Figure 5, Structure of the middle line (DVSP) ... 21
Figure 6, Schematic of DVSP DHS ... 21
Figure 7, Diagram of hydraulic balance (DVSP) ... 22
Figure 7, Heat loss through an insulated pipe ... 26
Figure 8, Illustration of area E1-B (winter) ... 32
Figure 9, Illustration of area F7 (winter) ... 32
Figure 10, Comparison of hydraulic balances (CCCP) ... 38
-8- List of tables
Table 1, Parameters of the pipes 1(2) ... 30
Table 2, Parameters of the pipes 2(2) ... 30
Table 3, Parameters of water pumps ... 30
Table 4, Parameters of the consumers (substations) ... 30
Table 5, Parameters of heat pump ... 31
Table 6, Temperature data ... 31
-9- Nomenclature
𝐴 area normal to the direction of heat transfer [m2]
𝐶 Hazen Williams roughness coefficient [-]
𝑐$ specific heat capacity [J/kgK]
𝐶% flow coefficient [GPM]
𝐷 internal diameter of the pipe [m]
𝑓 friction factor [-]
𝑔 acceleration due to gravity [m/s2]
ℎ heat transfer coefficient [W/m2K]
𝐻 head [m]
𝑘 thermal conductivity [W/mK]
𝐿 length of pipe [m]
𝑚̇ mass flow rate [kg/s]
𝑃 electricity demand [W]
∆𝑃 pressure change [Pa]
𝑄 flow rate or discharge [m3/s]
𝑄̇ heat flow (net rate of heat transfer) [W]
𝑟 radius [m]
𝑅 resistance [K/W]
𝑅𝑒 Reynolds number [-]
𝑆𝐺 specific gravity [-]
𝑇 temperature on the absolute scale, (Kelvin scale) [K]
𝑇8 surrounding temperature [K]
𝑉 bulk velocity [m/s]
𝑧 elevation [m]
Symbols
𝜁 local pressure loss coefficient [-]
𝜅 pipe wall roughness [m]
𝜂 efficiency [%]
𝜌 mean density of the fluid in the pipe [kg/m3]
-10- Abbreviations
CCCP Conventional Central Circulating Pump DH District Heating
DHS District Heating System(s)
DVSP Distributed Variable Speed Pumps KTH Royal Institute of Technology NUH Northern Urban Heating
SCADA Supervisory Control and Data Acquisition ZJU Zhejiang University
-11- Table of contents
Abstract ... 3
Sammanfattning ... 4
List of diagrams... 5
List of equations ... 6
List of figures ... 7
List of tables ... 8
Nomenclature ... 9
1 Introduction ... 13
1.1 Background ... 13
1.1.1 Demand of house heating in China ... 13
1.1.2 DHS in China ... 14
1.2 Aim... 16
1.3 Research questions ... 16
1.4 Scope ... 17
1.5 Limitations ... 17
2 Theoretical model ... 17
2.1 Model of primary pump system in Guizhou... 17
2.1.1 CCCP model ... 19
2.1.2 DVSP model ... 20
2.2 Components and function ... 22
2.2.1 Pressure requirements ... 22
2.2.2 Water pumps ... 23
2.2.3 Control (throttling) valves ... 23
2.2.4 Pipes ... 24
2.2.4.1 Steady heat conduction in cylinders ... 24
2.2.4.2 Heat losses through an insulated pipe ... 25
2.2.4.3 Pressure drop over a circular pipe ... 26
2.2.5 Water source heat pumps ... 27
2.2.6 Heat exchange substations ... 27
3 Methodology ... 28
3.1 Termis ... 28
3.2 Model setup ... 28
3.2.1 CCCP model ... 29
-12-
3.2.2 DVSP model ... 29
3.3 Parameters ... 29
3.4 Assumptions ... 31
3.5 Sensitive analysis ... 32
4 Results ... 33
4.1 Sensitive analysis ... 34
5 Discussion and conclusion ... 37
5.1 CCCP versus DVSP configuration ... 37
5.2 Sensitive analysis ... 38
5.2.1 Losses due to pipes ... 39
5.3 Sources of error ... 39
5.4 Future impact ... 40
5.5 Conclusion ... 40
6 Future work ... 41
References ... 43
Appendix ... 46
-13- 1 Introduction
In this section, an introduction of District Heating Systems (DHS) in China is given, together with results from previous studies regarding the electricity demand and consumption of Conventional Central Circulation Pump (CCCP) and Distributed Variable Speed Pumps (DVSP) systems. “Electricity demand is measured in kilowatts (kW) and represents the rate at which electricity is consumed. Electricity consumption, on the other hand, is measured in kilowatt-hours (kWh) and represents the amount of electricity that has been consumed over a certain period.” (Havens, 2017). Further, the aim, research questions, scope and limitations of this report are presented.
1.1 Background
To establish basic understanding about DHS and its importance in the sustainable development of China, the demand of house heating in China will be described together with the future estimated growth of the heat demand. Further, the existing DHS in China will be described together with results from previous studies of CCCP and DVSP systems. These results will be compared to the results of this report.
1.1.1 Demand of house heating in China
China is divided into five climate zones where the different climates have considerable impact on the need for building heating. Two of these zones are the cold and severe cold zones, located in the north, where indoor heating is required by law. These zones are part of the Northern Urban Heating (NUH) area, where efficient DHS are of importance to provide welfare service in terms of good living conditions (IEA, 2017). About 41% of the Chinese population is estimated to live within this area, resulting in a big demand of efficient DHS (Gong & Werner, 2014). Between 2001-2013, DHS in the NUH area increased from 5 billion m2 to 12 billion m2
“due to the growth of income and population” (IEA, 2016). Today, the total building floor area in China is approximately 57 billion m2 whereof 13 million m2 are located in the NUH area. By 2050 the total building area, i.e. the space taken up by a building, in China is estimated to increase by 40%. This is leading to even higher demand on DHS since 95% of the new building area in the NUH area is covered with DHS (IEA, 2017).
-14- 1.1.2 DHS in China
District Heating (DH) is the most common way to provide urban heating in China, and the great demand of DHS indicates that the market of DHS is huge. For 2015, with a population of 1,397 billion (Worldometers, 2018), the power statistics showed that the district heat produced in China was 4 092 TWh (IEA, 2017). To indicate the great scale of DH in China, the statistics for 2015 can be compared with Sweden for the same year, with a population of 9,851 million (SCB, ND). 51,98 TWh district heat was produced in Sweden whereof 48,83 TWh heat was delivered to consumers (SCB, 2017). Further, China hosts the largest and fastest growing DHS in the world with the total installed heat capacity of 650 gigawatts in 2015 (IEA, 2017). 14%
of this capacity was used for steam production, while the rest was used for hot water production.
In the DHS in China year 2016, coal-fired boilers accounted for 33% of the total heat production while 51% were covered by co-generation, mainly from coal, and 12% were covered by gas fired boilers. The remaining 4% were covered by other energy sources such as geothermal energy, biomass, industrial excess heat etc. The large dependence of coal in the DHS leads to high pollutant emissions (ibid.).
The huge building area covered with DHS also entails a large electricity demand of pumping water. In 2008, the average electricity consumption was estimated to be around 2,0 kWh/m2 DH area, leading to a total consumption of 159 TWh that year (Yan et al., 2013). Due to the rapid expansion of DHS in China, this electricity consumption can be assumed to have increased rapidly too.
A DHS located in Beijing is representative for the overall DHS configuration in China (Sun, 2018). This system uses Combined Heat and Power (CHP) and natural gas-based peak load boilers as heating sources. CHP is the primary heating source and the boiler-houses are additional sources in order to provide heat during cold winter days. In big cities, such as Beijing and Shanghai, the CHP are using natural gas as heating source in purpose to reduce the pollutant emissions. Although, in smaller cities coal is still a common heating source for the DHS.
Further, in the DHS located in Beijing there are three centralized water pumps connected to the CHP stations on the supply water line, one to each CHP. This configuration, using centralized pumps and throttling valves, i.e. Conventional Central Circulation Pump (CCCP) is called CCCP system and is nowadays the most common application of water pumps in DHS in China
-15-
(Sun, 2018). In Sweden, a state of the art nation in the process of developing DHS systems, the CCCP system also is the most common configuration of DHS. Although, this is because Swedish DHS normally do not have any heat exchange substations. Instead, the customers are directly linked to the primary network with heat exchangers located in their buildings (Dalgren, 2018).
Beyond the growing urbanization and the increasing demand of DH in China, there are many policies and requirements set by the government regarding environmental sustainability. For example, the policies outlined in the 13th Five-Year plan starting in 2016. These policies aim to improve energy efficiency and one of the plans’ target is to reduce the energy intensity (energy consumption per unit of GDP) by 15% in 2020 (Seligsohn and Hsu, 2016). In order to meet both these demands: a growing urbanization and sustainability policies, large improvements of current DHS as well as sustainable expansions are crucial. Accordingly, it is necessary to study the electricity demand of DHS with different configurations, to find what improvements and expansions that can be made. In China, the currently most considered improvement is to replace the centralized pumps with a multi-level system in terms of Distributed Variable Speed Pumps (DVSP), i.e. local water pumps at heat substations. In 2013, 10 million m2 of building area in China was estimated to be provided with heat from DHS with DVSP configuration (Shi, 2013).
Today, this building area has increased to 237 million m2 which is a tremendous development.
Example of provinces in China using DHS with DVSP configurations are Liaoning, Hebei, Shanxi (Sheng and Duanmu, 2016; Na, ND; Hong et al., ND).
In contrary to a CCCP system, the pumps at the substations in a DVSP system “operate according to the users’ variable heat demand” (Wang et al., 2017). This means that the DVSP system provides the correct pressure change necessary for every substation, and no rich hydraulic head of water needs to be throttled. In other words, the hydraulic performance will be fundamentally different by changing DHS from a CCCP to DVSP configuration.
Consequently, this will affect the electricity demand of the water pumps. Previous studies claim that the overall electricity consumption of the system can be reduced by 50% or more by changing the arrangement of the pumps from a CCCP to DVSP configuration (Shi, 2013). In a specific case in Dalian, China, changing the CCCP system to DVSP resulted in 49,41%
electrical energy reductions (Sheng and Duanmu, 2016). Given this information, DVSP systems
-16-
in China enable large reductions of the overall electricity demand of pumping in DHS because of the DHS’ existing size and rapid expansion (Yan et al., 2013). Therefore, using DVSP configuration to improve existing, and to build new sustainable, DHS have attracted extensive attention.
To enable the implementation of these improvements and expansions of DHS, the government invest a lot of money to support this work. Understandably, there is a high level of competition among companies providing these improvements and expansions. Companies are using modeling tools to create models of potential improved or expand systems and enable simulations of their function. The simulations make it possible to determine the optimal number of distributed pumps, their locations and other important parameters and to calculate the reduction of electricity demand of any improvement. This information is used as investment research for the government who will make the final decision about which company is contracted to run the building project of the DHS. The strong competition indicates the importance of developing good modeling tools (Sun, 2018).
1.2 Aim
The aim with this bachelor thesis is to provide an electricity demand analysis of the two different configurations, CCCP and DVSP, of a primary pump system of a DHS in the Guizhou province, China. The results will be used to, for this province, confirm previous studies about DVSP systems being a more electric energy efficient configuration of a DHS, and to get more statistics of the electric energy reduction that can be made.
1.3 Research questions
The project needs to gradually solve the following questions:
• What is the electricity demand in the primary pump system in the four different scenarios, using data given for a specific time? (1) Summer (CCCP), (2) Summer (DVSP), (3) Winter (CCCP) and (4) Winter (DVSP)
• What reductions of the electricity demand can be made when using a DVSP system instead of a CCCP system?
• During which period will the largest reductions occur?
-17- 1.4 Scope
This project is focusing on sustainability in terms of electricity demand of DHS. It is an analysis of different configurations of a primary pump system of a DHS. Two steady state models are sought to be created, one for CCCP and another for DVSP configuration, in order to simulate different scenarios with given parameters and compare the results. The project is an ongoing collaboration with Zhejiang University (ZJU) during a period of two months and it will be completed by the end of May, 2018.
1.5 Limitations
In this project selected parts of the primary pump system are studied. Even though the system consists of three lines: the north, middle and south, this study is delimited to the middle line only. Further, only the components which interact and affect the electricity demand of the water pumps and their function in the DVSP and CCCP configuration of the DHS respectively, are considered. The water circulation pump in the plant, which compensates for the internal pressure loss, is excluded from the calculations since its electricity demand is the same in both of the models. Moreover, the perspective of economic profitability of the electricity demand reductions is not investigated.
2 Theoretical model
In this project, theoretical simulations are made of a DHS in Guizhou, to investigate the differences in electricity demand between a CCCP and a DVSP configuration of the DHS. A theoretical basis for the calculations are given below. The simulations are performed in the district energy optimization software Termis. The basic operations of this software are described in section 3.
2.1 Model of primary pump system in Guizhou
The primary network of the DHS system in Guizhou consists of six water-source heat pumps, three different lines for primary net distribution and two natural gas-fired peak load boilers. The water-source heat pumps are used as a sustainable alternative to the CHP. The two boilers
-18-
should only be running during the peak demand in the winter as complementary heat sources and are not taken into consideration in the steady state models. The middle line, which is considered in the project, connects three of the heat pumps with three heating areas: F7, E1-A and E1-B. The area F7 connects the primary network to the secondary network with two heat exchange stations while the E1-areas are connected with three heat exchange substations each.
The middle line is constructed of pipes insulated with polyurethane and a high-density polyethylene protector shell. However, the pipes have varying diameter and length.
Figure 1, System overview, middle line
The nodes are named after the area to which they belong; Pl (plant), F7, E1-A and E1-B. The
“main” pipes are named with numbers while the pipes in every subarea are named after the area they belong to, as for the nodes.
The water circulation in the model is managed by water pumps and valves. The water pumps have varying characteristics as flow, pump head and power which is described in the methodology section. The location of the pumps and the usage of valves differ between the CCCP and the DVSP model, which is described later in this section. Water circulation pumps are also included in the plant. These are only used to compensate for pressure drop in the heat plant. Accordingly, the pressure at the inlet and outlet of the plant is the same.
-19- 2.1.1 CCCP model
In the CCCP model there are four water circulation pumps located on the supply line. In Termis, the four water circulation pumps are graphically represented by one single pump. However, in the pump settings the actual number of pumps, i.e. four, are set. Further, for the substations to work properly, the centralized pumps are regulated to supply all heat exchange substations with the minimum pressure difference. Consequently, the water pumps are controlled according to the substation with the lowest pressure difference.
Valves are installed at the supply side, adjacent to the heat exchange substations, to regulate the pressure at substations with a too high pressure difference. Thus, the model meets the required pressure difference. The structure of the middle line in the primary pump system, using CCCP configuration, is shown in figure 1 below.
Figure 2, Structure of the middle line (CCCP)
A schematic diagram of a DHS with CCCP configuration is shown in figure 2 below, illustrating the placing of the components in the system.
-20-
Figure 3, Schematic of CCCP DHS
The hydraulic balance of the CCCP system is shown in figure 3 below. It illustrates a continuously decreasing pressure due to transportation through the pipes on both the supply and return side. Also, the diagram shows the head that needs to be throttled in order to meet the pressure difference requirement for every substation. At heat exchanger 𝑛, i.e. the node with the lowest pressure, throttling is not necessary.
Figure 4, Diagram of hydraulic balance (CCCP)
2.1.2 DVSP model
In the DVSP model there are three variable speed pumps located on the supply line, adjacent to each heat exchange substation. This means that the distribution network system is a multilevel pump heating system. The water pumps are regulated to meet the minimum pressure change required at every heat exchange substation. No valves or centralized pump are used in this
-21-
model. The structure of the middle line in the primary pump system, using DVSP configuration, is shown in figure 4 below.
Figure 5, Structure of the middle line (DVSP)
A schematic diagram of a DHS with DVSP configuration is shown in figure 5 below, illustrating the placing of the components in the system.
Figure 6, Schematic of DVSP DHS
The hydraulic balance of the DVSP system is shown in figure 7 below. From the heat pumps, the pressure drops due transportation through the pipes. Then, the pressure increases at every substation to meet the required pressure difference and to compensate for the pressure drop in
-22-
the pipes. This is shown by the arrows pointing upwards at the pumps. Thus, the distributed pumps further away from the plant are required to operate at a higher head. This results in a lower pressure in the supply line than in the return line.
Figure 7, Diagram of hydraulic balance (DVSP)
2.2 Components and function
The components in the DHS that affect the electricity demand are; heat plants, heat exchange substations, water circulation pumps, valves and the pipes. Additionally, a sufficient pressure in the DHS is required to deliver the heat to the consumers. The fundamental equations for calculating the electricity demand of the pump and to evaluate the pressure balance in the system are given below.
2.2.1 Pressure requirements
In the DHS, the pressure needs to be controlled in order for the system to work properly. The pressure within the system must always exceed the minimum pressure allowed, and the pressure change at every substation must be sufficient to supply the required energy (Sun, 2018).
Furthermore, the pressure at the inlet of every pump must exceed the net positive suction head required, i.e. the minimum pressure allowed at the inlet. If this requirement is not fulfilled, cavitation occurs, leading to breakdowns of the pumps (Grundfors, ND).
-23-
The pressure within the system follows the law of Kirchoff according to the fluid network analogy. The simplified statement of the law says that “the algebraic sum of all pressure drops around a closed path, or mesh, in the network must be zero, having taken into account the effects of pumps.“ (Yan et al., 2013). This can be summarized in equation 1 below, where ∆𝑃 is the sum of all pressure drops in the system and 𝐻$ is the sum of the head of all circulation pumps in the system (ibid.).
Equation 1, Law of Kirchoff
@ ∆𝑃 + @ 𝐻$ = 0
2.2.2 Water pumps
The circulation water pumps are necessary components in order to achieve the pressure requirements of the system. In the middle of the water pump there is an impeller with a series of curved vanes fitted inside shroud plates, making the surrounding fluid to circulate.
Accordingly, the water particles are exposed to centrifugal force whereby the water moves radially out from the impeller (Araner, 2017).
For circulation pumps, the electricity demand can be calculated using equation 2 below, where 𝜂 denotes the pump efficiency and 𝜂D the motor capacity.
Equation 2, Electricity demand for water pump
𝑃E = 𝑄𝑑𝑃
3600 ∙ 𝜂 ∙ 𝜂D
2.2.3 Control (throttling) valves
In the model for the CCCP configuration of the primary pump system, control valves are connected to the supply water pipes. These valves are necessary components since they preserve operation conditions in the system in terms of flow, pressure, temperature and liquid level by throttling the fluid (Emerson, 2018). For an example, the valves are used to control the amount of the circulating water in the system and ensure that the flow is within the required operation range.
-24-
The pressure change depends on the mass flow through the valve, the specific gravity of water as well as the opening of the valve and can be calculated using the formula below. When a specific pressure change is set for the valve, the flow coefficient 𝐶% is calculated and used for the simulation.
Equation 3, Pressure change through valve
∆𝑃 =𝑚̇J∙ 𝐺 𝐶%J
2.2.4 Pipes
A DHS consists of an insulated underground pipeline network used to transport the heated water. The insulation, as well as the dimension of the pipes, affect the heat loss in a DH network (Frederiksen S, Wener S., 1993). Furthermore, the friction that occurs in pipes causes pressure drop. When creating the two models for CCCP and DVSP configuration respectively, it is only the location of the pumps that will change, i.e. not the geometry or characteristics of the pipes.
Accordingly, the heat losses and pressure changes theoretically should remain the same in the two models which in turn would generate the same water flow through the pumps.
2.2.4.1 Steady heat conduction in cylinders
When steady heat conduction through a hot-water pipe is considered, heat is continuously lost to the surroundings through the wall of the pipe (Çengel & Ghajar, 2015). For long cylindrical pipes, commonly used in DHS, following approximations regarding the temperature are made;
(1) it only depends on the radial r-direction which is expressed as 𝑇 = 𝑇(𝑟), (2) it is independent of the azimuthal angle/axial distance (ibid.). Accordingly, the heat transfer through a pipe can be considered as constant since the rate into the pipe is equivalent to the rate out of it, i.e.
𝑄̇MNOP,MRS = 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡. Using Fourier’s law of heat conduction for heat transfer through a cylindrical layer, 𝑄̇MNOP,MRS can be rewritten as (ibid.):
-25-
Equation 4, Heat conduction through cylindrical layer 1(2)
𝑄̇MNOP,MRS = −𝑘𝐴𝑑𝑇 𝑑𝑟 = [
𝐼𝑛𝑡𝑒𝑔𝑟𝑎𝑡𝑒 𝑢𝑠𝑖𝑛𝑔 𝑇(𝑟_) = 𝑇_
𝑇(𝑟J) = 𝑇J ` ⟹ b 𝑄̇MNOP,MRS 𝐴
cd cecf
𝑑𝑟 = − b 𝑘
gd gegf
𝑑𝑇
By substituting 𝐴 = 2𝜋𝑟𝐿 and integrating the equation, 𝑄̇MNOP,MRS can be written as:
Equation 5, Heat conduction through cylindrical layer 2(2)
𝑄̇MNOP,MRS = 2𝜋𝐿𝑘𝑇_− 𝑇J
ln l𝑟𝑟_Jm = [𝑅MRS = ln l𝑟𝑟J_m
2𝜋𝐿𝑘` =𝑇_− 𝑇J 𝑅MRS
Where 𝑅MRS is the thermal resistance of the cylindrical layer against heat conduction.
2.2.4.2 Heat losses through an insulated pipe
The insulation acts as an additional layer and therefore, an additional resistance in the steady heat transfer through the pipe appears. Also, the fluid pipe is subjected to convection on both surfaces (Çengel & Ghajar, 2015). Accordingly, the equation for the steady rate of heat loss from the fluid can be expressed as (ibid.):
Equation 6, Heat loss from insulated pipe
𝑄̇nOopSqrsP $n$s =𝑇8_− 𝑇8J 𝑅rNrqS
𝑇8_ is the temperature of the fluid inside the pipe, 𝑇8J is the temperature of the surrounding and 𝑅rNrqS is the total thermal resistance sum and can be written as (ibid.):
Equation 7, Total Thermal Resistance
𝑅rNrqS = 𝑅n + 𝑅_+ 𝑅J+ 𝑅N = 𝑅MNO%,_+ 𝑅$n$s+ 𝑅nOopSqrnNO+ 𝑅J,MNO%
= 1
ℎ_𝐴_+𝑙𝑛 (𝑟J/𝑟_)
2𝜋𝑘_𝐿 +𝑙𝑛 (𝑟w/𝑟J) 2𝜋𝑘J𝐿 + 1
ℎJ𝐴w
-26-
The different radiuses and temperatures are shown in figure 7 below.
Figure 8, Heat loss through an insulated pipe
2.2.4.3 Pressure drop over a circular pipe
While fluid transfers through a pipe a pressure change, ∆𝑃$n$s, occurs. This can be considered as the sum of the pressure loss due to friction, gravity and the local pressure loss (Schneider Electric, 2012). By using the equations below, ∆𝑃$n$s can be calculated (ibid.).
Equation 8, Pressure loss using the Colebrook and White friction
∆𝑃 = −𝜌2
𝐷|𝑉|𝑉𝐿 + 𝑔(𝑧y− 𝑧z)𝜌 −1
2𝜁𝜌|𝑉|𝑉
Equation 9, Pressure loss using the Hazen-Williams friction
∆𝑃 = −10.67 ∙ 𝐷|}.~•∙ 𝐶|_.~€J∙ 𝑄_.~€J∙ 𝐿𝜌𝑔 + 𝑔(𝑧y− 𝑧z)𝜌 −1 2𝜁𝜌𝑉J
Using the Colebrook and White’s formula, the friction factor _
•‚ is calculated (ibid.):
Equation 10, Colebrook and White’s formula
1
•𝑓= −4 ∙ 𝑙𝑜𝑔_„… 𝜅
3.7 ∙ 𝐷+ 1.413 𝑅𝑒 ∙ •𝑓†
-27-
For small Reynolds numbers, 𝑅𝑒 < 2300, the friction factor can be calculated from a simplified formula (ibid.):
Equation 11, Darcy friction factor
𝑓 =16 𝑅𝑒
2.2.5 Water source heat pumps
In a water source heat pump, energy from a hot water source is used to heat the water in the primary network of a DHS. The heat is transferred from the refrigerant in the heat pump to the water in the primary system through a heat exchanger. Due to resistance of the heat exchanger, a pressure drop occurs. A water circulation pump is located adjacent to a water source heat pump to compensate for this pressure drop (Sun, 2018). Since the pressure drop is related to the mass flow, which is equivalent for the CCCP and DVSP model respectively, the pressure drop will also be consistent. Consequently, this leads to an equivalent electricity demand for water pumping in both models.
Further, the supply temperature of the heat pump is set to a fixed value and the water flow is adjusted to deliver the required heat to the consumers and to keep the return water at the same temperature. The variation of the mass flow will affect the electricity demand of the water pumps in the system as descried by equation 2.
2.2.6 Heat exchange substations
In order to transfer heat between the primary and secondary network, plate heat exchangers are used in the heat exchange substations. This allows the primary and secondary network to operate at different temperatures (Çengel & Ghajar, 2015). With a known heat load and temperature difference between the supply and return temperature at every heat exchange substation, the mass flow for these locations can be calculated using equation 12 (ibid.):
-28-
Equation 12, Mass flow at substations
𝑚̇ =𝑄̇Esqr SNqP 𝑐$∙ Δ𝑇
For the heat exchange substations to work properly, a certain pressure change over the heat exchanger is required. The pressure in the DHS needs to be regulated by water circulation pumps and valves so that every heat substation meets the pressure change requirement.
3 Methodology
Firstly, information about the investigated primary pump system is collected. This information is accessed through existing data for a specified time of the DHS, provided by the project team at ZJU. Secondly, two steady state models of the primary pump system are created in Termis.
One model for the CCCP configuration and another, revised, model for the DVSP configuration. Lastly, simulations of the two models, based on given operational data corresponding to the specified operation conditions of summer and winter status, are made.
3.1 Termis
Termis is a district heating optimization software developed by Schneider Electric. The system can be used for static modeling, simulation and optimization of DHS as well as dynamic real time simulations based on real time data from the SCADA (Supervisory Control and Data Acquisition) system and forecast information (Steiner, 2018). This means that the software can be used to create the most sustainable and cost-effective systems by comparing different configuration alternatives. Based on the model and the filled in parameters, the software uses thermodynamic relations to make calculations and provide results for different scenarios. The Termis User Guide Version 5.0 as well as 6.0 have been used in order to learn about the system and how to create models and run simulations.
3.2 Model setup
Two models are created in Termis based on given parameters. The heat load and ambient temperature differs between the summer and winter simulations, while the other values remain
-29-
constant. By filling in given parameters in Termis, flows and pressure in all nodes are calculated by the software. Then, the pressure drop at the substations is manually adjusted to meet the required pressure value. This is done differently for the CCCP respectively DVSP model.
3.2.1 CCCP model
In the CCCP model, the pressure drop at the substations is controlled by adjusting the fixed pressure change in the central circulating water pump. The pressure change is set to meet the pressure change requirement at the substation with the lowest pressure drop value. Then, the pressure at the substations exceeding the pressure drop requirement, is throttled by adjusting the fixed pressure drops at the valves, adjacent to the substations. With correct pressure drops at all substations, the total electricity demand for the primary system can be achieved by studying the electricity demand in the centralized pump.
3.2.2 DVSP model
In the DVSP model, the pressure drop at the substations is controlled by adjusting the fixed pressure change in the distributed water circulation pumps. The pressure change is set to provide the required pressure drop at the substation, located adjacent to the pump. With correct pressure drops at all substations, the total electricity demand for the primary system can be achieved by summarizing the electricity demand of the distributed water pumps.
3.3 Parameters
In order to create two functional models there is a need of parameters of the components in the primary pump system. In this section, parameters are provided for the following components;
pipes, water source heat pumps, water pumps and consumers. Also, data for the ambient temperature is presented. The information is provided by the project team at ZJU.
Pipe Outer
diameter
Wall thickness
Inner diameter
Heat transfer
coefficient Roughness
mm mm mm mm W/m/K mm
300 325 7,00 318 0,10 0,01
Nominal diameter
Denomination
-30-
350 377 7,00 370 0,10 0,01
450 478 8,00 470 0,10 0,01
500 529 9,00 520 0,10 0,01
Table 1, Parameters of the pipes 1(2)
Pipe Nominal
Diameter
Length m
Pl_1 DN500 380
Pl_2 DN300 200
Pl_3 DN450 240
Pl_4 DN350 150
PI_5 DN350 20
Pipes at substation areas DN350 5
Table 2, Parameters of the pipes 2(2)
Area Flow Pump head Power Nominal
Rotation Number of pumps
m3/h m kW rpm
F7 115 32,0 18,5 1450 3
F7 52 32,0 11,0 2900 3
E1-A 72 33,0 11,0 2900 3
E1-A 72 33,0 11,0 2900 3
E1-A 72 33,0 11,0 2900 3
E1-B 69 30,0 11,0 2900 3
E1-B 69 30,0 11,0 2900 3
E1-B 69 30,0 11,0 2900 3
Plant 400 62,0 110,0 1480 4
Table 3, Parameters of water pumps
Area F7_1 F7_2 E1-A E1-B
Heat load
(winter) MW 2,380 1,330 4,650 4,650
Heat load
(summer) MW 1,904 1,064 3,720 3,720
Pressure drop bar 3 3 3
Elevation m 1 082,0 1 113,8 1 129,0
Table 4, Parameters of the consumers (substations)
-31-
Parameter Unit Value
Nominal temperature of return water °C 38,2
Nominal temperature of supply water °C 45,0
Number of heat pumps - 3
Pressure drop bar 1,63
Elevation m 1031,60
Table 5, Parameters of heat pump
Ambient temperature Degrees [°C]
Winter 5,0
Summer 20,0
Table 6, Temperature data
3.4 Assumptions
When creating the models, there have been assumptions made regarding the nodes in the system. The given data for the elevation of the nodes in table 5 only includes four of all the nodes in the middle line, the main nodes. Therefore, the models are created to assume that all nodes connected to a specific area have the same elevation. For example, all the nodes connected to E1-B have the same elevation (1129 m).
Further, pipe lengths are only given for the main pipes in the system. The lengths of the remaining pipes, which ones connect the main pipes to the substations, are assumed to be 5 meters long each. The lengths of the supply and the return pipe are assumed to be equal, as well as the roughness and heat transfer coefficient.
Furthermore, data for the initial power for each node is required while creating the models in Termis. The initial power is provided from table 5 in terms of the heat load for every area.
Although, it does only tell the heat load for the whole area and not in every specific “end” node connected to respectively area. Since all of the pumps belonging to E1-A and E1-B have the same characteristics, the total heat load for area E1-A and E1-B respectively, has been divided with the number of connected end nodes to respectively area. Accordingly, the model assumes
-32-
that each of these connected end nodes has the same amount of heat load. This is illustrated in figure 8 below for area E1-B with a total heat load of 4,65 MW, found in table 5.
Figure 9, Illustration of area E1-B (winter)
The two pumps located in area F7 have different characteristics. In order to get a proportional heat load in each end node connected to this area, the heat load is assumed to be proportional to the nominal flow of the pump adjacent to respectively end node. This is illustrated in figure 9 below.
Figure 10, Illustration of area F7 (winter)
3.5 Sensitive analysis
When doing the sensitive analysis, the consumers heat load are increasing with a percentage change of 3%. Accordingly, the models will provide a range of outcomes for the electricity demand and the impact of the new heat load values will be determined. Also, the sensitive
1,55 MW
1,55 MW
1,55 MW 0 MW
0 MW
2,5548 MW 115 m3/h 1,1552 MW
0 MW 52 m3/h
-33-
analysis aims to consider the outcome of heat loss and pressure change in the system when running the CCCP and DVSP model respectively. The simulation results will be used to validate that these losses do not deviate between the two models.
4 Results
This section provides the results of the project and answers the research questions described in section 1.3. The results have been calculated in Termis and compiled in Microsoft Excel.
The total electricity demand for the four different scenarios is presented in diagram 1. In diagram 2 and 3, the electricity demand reductions when using a DVSP instead of CCCP model is shown. Diagram 2 shows the reductions in percentage while diagram 3 shows them in kW.
Furthermore, both diagram 2 and 3 show that the largest electricity demand reductions are made during the winter.
Diagram 1, Electricity demand in the four simulated scenarios [kW]
140,86
88,27 121,93
83,80
0 20 40 60 80 100 120 140 160
Winter Summer
Electricity demand [kW]
Electricity demand [kW]
CCCP DVSP
-34-
Diagram 2, Reductions during summer and winter [%]
Diagram 3, Reductions during summer and winter [kW]
4.1 Sensitive analysis
When increasing the consumer heat load values with 3%, the models generated new simulation results regarding the electricity demand and reductions. Diagram 4-6 below, present the results in the same way as above.
13,44%
5,06%
0,00%
2,00%
4,00%
6,00%
8,00%
10,00%
12,00%
14,00%
16,00%
Reduction winter Reduction summer
Reduction [%]
Reduction of electricity demand [%]
18,93
4,47
0 2 4 6 8 10 12 14 16 18 20
Reduction winter Reduction summer
Reduction [kW]
Reduction of electricity demand [kW]
-35-
Diagram 4, Electricity demand in the four simulated scenarios, sensitive analysis [kW]
Diagram 5, Reductions during summer and winter, sensitive analysis [%]
150,43
93,9 128,81
87,85
0 20 40 60 80 100 120 140 160
Winter Summer
Electricity consumption [kW]
Elecrricity demand [kW]
CCCP DVSP
14,37%
6,44%
0,00%
2,00%
4,00%
6,00%
8,00%
10,00%
12,00%
14,00%
16,00%
Reduction winter Reduction summer
Reduction [%]
Reduction of electricity demand [%]
-36-
Diagram 6, Reductions during summer and winter, sensitive analysis [kW]
When comparing the results from the sensitivity analysis with the initial results, it is found that both the electricity demand and reductions increase. The percentage increases are shown in diagram 7 below.
Diagram 7, Percentage change of the results 21,62
6,05
0 5 10 15 20 25
Reduction winter Reduction summer
Reduction [kW]
Reduction of electricity demand [kW]
6,79%
5,64% 6,38%
4,83%
14,21%
35,35%
6,94%
27,23%
0,00%
5,00%
10,00%
15,00%
20,00%
25,00%
30,00%
35,00%
40,00%
Electricity demand,
CCCP Winter
Electricity demand,
CCCP Summer
Electricity demand,
DVSP Winter
Electricity demand,
DVSP Summer
Electricity reductions, winter
Electricity reductions, summer
Percentage reductions, winter
Percentace reductions, summer
Sensitive analysis
-37-
When studying the total heat losses through pipes, it is found that the heat loss in the summer simulations is 4,67 kW for both the CCCP and DVSP model. Also, in the winter simulations it is found that the heat loss through pipes is the same for the CCCP and DVSP model, namely 7,91 kW.
In the simulation of the CCCP model with summer parameters, it is found that the pressure drop in pipe Pl_1 is 5 bar. The equal value is found when studying the pressure drop in the DVSP model, using the summer parameters for the same pipe.
5 Discussion and conclusion
In this section, the provided results are discussed with reference to the formulated research questions. Also, a conclusion is drawn.
5.1 CCCP versus DVSP configuration
The simulation results indicate that the electricity demand is reduced when using the DVSP instead of CCCP configuration of the studied primary pump system. That is the case for both summer and winter scenarios. Thus, the results confirm previous studies claiming DVSP systems as more efficient from an electricity demand point of view. The electricity demand reduction of using the DVSP instead of the CCCP configuration for winter and summer, is 13,44% and 5,06% respectively. Accordingly, the largest electricity demand reductions are made during winter.
However, previous studies claim that the reductions of using DVSP instead of CCCP system can exceed 50% (Shi, 2013). That percentage is way higher than the ones generated from this project’s simulation results. One explanation to these dissimilarities can be the relatively small geographical coverage of the studied system in Guizhou. Since all of the heat exchange substations are located on approximately equal distance from the centralized pump in the CCCP model, the head that needs to be throttled is relatively small. If the substation with the lowest pressure drop had been located further away from the substation with the highest pressure drop,
-38-
the throttling would have been higher, and larger energy losses would have occurred. In such case, it would have been even more efficient using a DVSP configuration from an electricity demand reduction point of view. The reasoning is shown in figure 10.
Figure 11, Comparison of hydraulic balances (CCCP)
As shown in figure 10, the head that needs to be throttled is much lower in the right diagram than in the left diagram. This is because the location of the substations in the right diagram is at more equal distance from the heat plant.
5.2 Sensitive analysis
As presented in the sensitive analysis in section 3.5, varying values of the consumer heat load have been set in the models to determine and compare the simulation outcomes. When increasing the heat load with 3% for all substations, the results show that the electricity demand of the water pumps increases with 4-6%, depending on scenario. Moreover, the electricity demand reductions increase even more when increasing the heat load with 3%, especially for the summer scenarios with a lower initial heat load. This indicates that the results are sensitive to changes in heat load, and that even small changes have large effects on the electricity demand reductions that can be made in the primary pump system.
In addition, changes in the heat load in real-time scenarios are to be expected due to varying heat demand. Therefore, the results of this report are highly dependent on the specific data given for the heat load during summer and winter. If this information is non-representative for the real system, the results may not provide the actual electricity demand reductions of the system.
However, even when changing the heat load for the sensitive analysis, the DVSP model is still
-39-
the one which generates the lowest electricity demand of the water pumps. Therefore, the results of the best configuration of pump location can be said to be certain.
5.2.1 Losses due to pipes
Focusing on losses due to pipes, the results indicate there is no deviation between the CCCP and DVSP model. Regarding heat losses through the pipes, these are 7,91 kW in both cases and the pressure drop for the studied pipe, P1_1, also proves to be the same (5 bar) for the two models. Consequently, these results confirm the assumption that heat losses, as well as pressure drop, due to pipes are independent of the pump configuration and can be calculated according to the basics of thermodynamics. Thus, these types of losses will not affect the electricity demand of the pumps and do not need to be taken into consideration in the comparative analysis between the two configurations of the DHS in Guizhou.
5.3 Sources of error
Apart from the sensitivity of the models when changing the values of heat load, it is important to reflect upon other sources of error in the models. A potential source of error could be the data provided by the project team at ZJU. If any of the given information would show to deviate from the actual system, this would affect the results and make them less reliable.
Regarding the heat loss due to pipes, it will vary depending on the heat transfer coefficient. For the created models, the same heat transfer coefficient is used for all of the pipes, even though they have different characteristics. This could have affected the results, since losses due to pipes are correlated to the necessary mass flow in the system, and therefore also the electricity demand of the pumps. However, when comparing the total heat loss of the system for the specific winter scenario (7,91 kW) with respect to the total heat load for the same specific moment (13 011 kW), it is found that the heat losses are very small. Therefore, incorrect data regarding the heat transfer coefficient of the pipes, should not have had a significant impact on the provided results.
Further, when creating pump curves in Termis, selected values for the mass flows were taken from the given pump diagrams and used in the models. Based on the filled in values, pump curves for respectively pump were created with a built-in linear interpolation function in the
-40-
Termis software. Therefore, a higher precision could be reached by filling in more, specific, values from the given pump curves since these curves have a direct impact on the electricity demand of the pumps. Furthermore, it is found that the pumps in both the CCCP and DVSP system are over-sized with respect to the required pressure change in the systems. Thus, by using more suitable sized pumps for the required pressure change, the results will change too.
At last, the electricity demand of the water circulation pump used to compensate for the pressure change in the heat plant, is not taken into consideration. This is because the electricity demand of this pump is the same for both configurations, due to an equivalent mass flow in both models.
However, this affects the overall percentage of electricity demand reductions of using the DVSP instead of CCCP system. If this pump would not have been delimited, the reduction-ratios would have been lower.
5.4 Future impact
Even though there exist potential improvements of the created models, the overall results imply that a DVSP configuration of a DHS is correlated with electricity demand reductions. Worth noting is that the studied primary pump system is a very small system relative other DHS in China. Thus, if electricity demand reductions can be made in such a small system, it is necessary to emphasize the potential of reductions that can be made in the country as whole if DHS improvements, in terms of DVSP configuration, are made. That is due to the rapid expansion of DHS in the nation and the fact that China already hosts the largest area covered by DHS in the world, mostly made up of CCCP systems that could be changed to DVSP systems. Also, by increasing the life conditions and meeting the urbanization of a big population in a sustainable way like this, China tends to realize targets aligning with the 13th Five-Year plan in a greater extent and meet the climate goals. Hence, it is essential to acknowledge and embrace these kind of developments and improvements in order to enable a sustainable future with fair living conditions.
5.5 Conclusion
This project was aimed to contribute with an electricity demand analysis of a primary pump system in Guizhou, China and to confirm existing knowledge about DVSP system being more electric energy efficient than a CCCP system. After creating two steady state models, one for
-41-
each configuration, and simulating four different scenarios, the results showed an electricity demand of 140,86 kW and 88,27 kW for winter respectively summer for a CCCP configuration.
For the DVSP configuration, the electricity demand was 121,93 kW and 83,80 kW for winter respectively summer. Translating this into electricity demand reduction-ratios, the reduction was 13,44% and 5,06% for winter respectively summer. Accordingly, the largest reductions are made during the winter. This is by using a DVSP configuration of the examined DHS instead of a CCCP configuration. Given the results, the conclusion is that this study confirms existing knowledge about DVSP being a more electric energy efficient configuration of a DHS.
6 Future work
There is a great potential of future work related to this project. First of all, the models do only consider one of the three lines in the DHS. This could be improved by observing all of the lines in the system and include these in the models and thus make them more complex. Furthermore, only four simulations have been made. By increasing the number of simulations with different heat load and ambient temperature, the results tend to align with the reality in a greater extent.
In addition, the models are steady state. In order to use real time data such as weather, varying temperature and the use of peak load boilers, the future work should embrace the creation of a dynamic model. Thus, the whole DHS in terms of primary and secondary network can be considered more easily in Termis and the results would be representative over time and not only for a specific moment.
Further, the sensitive analysis has potential to be extended and include the effect of changing parameters such as pipe lengths. The actual system is very small due to short pipes which accordingly generates small losses. Thus, it would be relevant to study how a larger system, i.e.
longer pipes, affects the results in terms of electricity demand of the water pumps and how the amount of losses change.
Additionally, an extended analysis including smart optimization of DHS is suitable as future work. The key for solving the constant changing demand of DH issue is called “Smart Heating Systems”. This system uses historical statistics and measured values to optimize the system depending on the demand, i.e. it has a dynamic approach. The computer system will come up
-42-
with an operation scheme for the DHS and how the different parameters in the system should be adjusted to create the most efficient use of the resources. This concept of using big data and other information for making decisions is called a cyber-physical system. This means that the cyber space is used for controlling a physical system, in this case, a modelled system controlling the actual DHS. By taking the cyber-physical system into consideration, the best possible execution for the physical system can be achieved. Accordingly, the system can be optimized, and the precision of the system can increase over time when more information of the system is gathered from real time analyses from different operation schemes (Zhong, 2018).
Acknowledgements
First of all, we would like to thank for the financial support from Ångpanneföreningens Forskningsstiftelse (ÅForsk). Furthermore, there are a several persons whose work and help have contributed to this thesis and made it possible. Great thanks go to Professor Zitao Yu, Professor Wei Zhong, Ph.D. Xiaojie Lin and Expert Kaihong Sun at the Institute of Thermal Science and Power Systems, College of Energy Engineering, Zhejiang University for providing us with necessary information and continuous support. Thanks also go to Professor Viktoria Martin at the Department of Energy Technology at KTH who has supported our work from Sweden and to Johan Dalgren at Stockholm Exergi for guiding us in Termis. Gratefulness also goes to Yingfang He, Regional Advisor for China at KTH and Zian Zhang, Program Coordinator at ZJU who have enabled a great coordination of this exchange period.
Last but not least, an acknowledgement goes to the students Mingchao Li, Shuowei Lu, Sibin Lu, Jiaying Chen, Yipeng Yu and Xielei Chen at ZJU who have contributed to this thesis by always answering questions and offering a helping hand.
