Optimization of Electromechanical Studies for the Connection of Hydro Generation

(1)

IN

DEGREE PROJECT ELECTRICAL ENGINEERING, SECOND CYCLE, 30 CREDITS

STOCKHOLM SWEDEN 2018,

Optimization of Electromechanical Studies for the Connection of

Hydro Generation

MATHIEU GROULT

KTH ROYAL INSTITUTE OF TECHNOLOGY SCHOOL OF ELECTRICAL ENGINEERING

(2)

Master Thesis Report

Optimization of Electromechanical Studies for the Connection of Hydro Generation

Date: July, 3^rd 2017 – December, 22^nd 2017 Student: Mathieu GROULT

EDF Supervisors: Vincent BOIZEAU – Lionel JAMY KTH Examiner: Mehrdad GHANDHARI

KTH Supervisor: Dimitrios ZOGRAFOS

(3)

EDF EFESE

R12 - CIST

Optimization of electromechanical studies for the connection of hydro generation

Degree Project in Electric Power Systems,

Second Cycle (EG230X)

Page 2 / 121

ABSTRACT

The current model for electricity generation is based on power plants connected to the transmission network. This provides electricity to the distribution network and after that to the consumers. To ensure the security of the electrical network and prevent a blackout, the performance of every electricity generation unit connected to the network is quantified in grid codes. In the case of the French transmission system, the requirements regarding the performance are written in a document produced by the French Transmission System Operator (TSO). Various events with various configurations of connection to the network have to be simulated and the corresponding performance has to be evaluated. The aim of these simulations is to determine the stability of the generators and key elements, including the response time on the active power after events such as a short circuit.

Taking into account the amount of generators connected to the transmission network, the need for optimization appears and is the purpose of this Master Thesis. To perform those simulations in an efficient way on all the generators owned by the main French electricity producer, EDF, this Master Thesis contributes with a tool called AuDySim coded with the softwares MATLAB and EUROSTAG. The implemented tool allows the user to configure an electricity generation unit before realizing all the simulations specified by the TSO and produces a report containing the results by means of curves and data. The simulations and the production of the report are achieved automatically to create a gain of time and resources.

In order to validate the performance of the tool, two case studies are performed on different types of power plants. The two case studies analyzed present a hydraulic and a nuclear power plant. In the results the performance of each type of power plant is assessed focusing on the rotor angle stability of the machine and key elements, such as the voltage and the active power. These results lead to the conclusion that AuDySim fulfills its mission, by achieving automatically an analysis of the performance of an electrical generation unit and producing it in a report.

Keywords: Active power response time, EUROSTAG, Hydraulic power plant, Grid codes, MATLAB, Nuclear power plant, Rotor angle stability

(4)

EDF EFESE

R12 - CIST

Page 3 / 121

SAMMANFATTNING

Den nuvarande elproduktionsmodellen baseras på kraftverk som är direktkopplade till stamnätet. Stamnätet i sin tur matar distributionsnätet som därefter levererar el till slutkonsumenterna. För att säkerställa stamnätets integritet samt säkerhet och undvika strömavbrott kvantifieras prestandan hos varje generator som är ansluten till det med hjälp av nätkoder. När det gäller det franska stamnätet skrivs prestandakraven i ett dokument som utfärdas av den franska transmissionssystemoperatören (TSO). Olika händelser med olika anslutningskonfigurationer måste simuleras där dess prestanda ska utvärderats. Syftet med dessa simuleringar är att identifiera stabiliteten vid varje elproduktionsenhet med bl. a. dess reaktionstid för den aktiva effekten efter kortslutningar.

Med tanke på antalet generatorer som är anslutna till stamnätet framträder ett behov för överföringsoptimering vilket är syftet med detta examensarbete. För att utföra dessa simuleringar på ett effektivt sätt på alla generatorer som ägs av den ledande franska elproducenten, EDF, bidrar denna avhandling med ett verktyg som heter AuDySim kodat i mjukvarorna MATLAB och EUROSTAG. Verktyget gör det möjligt för användaren att konfigurera en elproduktionsenhet innan man utför alla simuleringar som specificeras av TSO:n och samtidigt producerar en rapport som innehåller grafisk- och data resultat. Både simuleringar och rapporten produceras automatiskt för att optimera en bearbetningstid och resursanvändning.

För att validera verktygets prestanda utförs två fallstudier på olika typer av kraftverk. De två fallstudierna fokuserar på ett hydraulisk- respektive ett kärnkraftverk. I resultaten utvärderas prestanda för varje typ av kraftverk, med fokus på maskinens rotorvinkelstabilitet och andra viktiga faktorer, såsom spänning och aktiv effekt. Resultat leder till slutsatsen att AuDySim uppfyller sitt uppdrag genom att automatiskt analysera prestanda hos en elektrisk generationsenhet och presentera analysen i en rapport.

Nyckelord: aktiv effekt tidrespons, EUROSTAG, hydraulisk kraftverk, nätkoder, MATLAB, kärnkraftverk, rotorn vinkelstabilitet

(5)

EDF EFESE

R12 - CIST

Page 4 / 121

ACKNOWLEDGEMENTS

First, I would like to thank the CIST, the EFESE department, and Xavier LEGRAND, head of the R12 group in the company EDF, for welcoming me in the group and allowing me to carry out my Master Thesis at EDF.

Then, I would like to express my gratitude to my supervisors at EDF, Vincent BOIZEAU and Lionel JAMY for giving me the opportunity of joining the group and for their guidance and confidence for the whole duration of the project.

I am thankful as well to all the members of the team working in the R12 group for their welcome and for integrating me. Their support and their help was essential in my project.

Finally, I want to thank my supervisor at KTH Dimitrios ZOGRAFOS for his help and my examiner Mehrdad GHANDHARI who agreed to supervise my work.

(6)

EDF EFESE

R12 - CIST

Page 5 / 121

LIST OF FIGURES

Figure 1: Electrical power network [1] ... 12

Figure 2: Frequencies during the 2006 event [4] ... 14

Figure 3: Steam turbine [14] ... 20

Figure 4: Gas turbine [14] ... 20

Figure 5: Principle of a Pelton wheel [16] ... 21

Figure 6: Detail of a Pelton wheel [15] ... 22

Figure 7: Design of a Francis and a Kaplan turbine [17] ... 22

Figure 8: Principle of a Kaplan turbine [14] ... 23

Figure 9: Structure of a bulb turbine [18] ... 23

Figure 10: No load Round-rotor machine [14] ... 24

Figure 11: Loaded Round-rotor machine [14] ... 25

Figure 12: Salient-pole generator [14] ... 26

Figure 13: Variation of the air-gap reluctance [14] ... 26

Figure 14: Saturation curve ... 29

Figure 15: SMIB system ... 32

Figure 16: SMIB equivalent system ... 33

Figure 17: Two-lines SMIB system ... 35

Figure 18: Pre-fault to post-fault state, maximum rotor angle ... 36

Figure 19: Post-fault state, maximum and minimum rotor angle... 37

Figure 20: SMIB transiently unstable ... 38

Figure 21: Equal Area Criterion ... 39

Figure 22: Model from EUROSTAG ... 42

Figure 23: Generic network ... 45

Figure 24: Short circuit network ... 50

Figure 25: Flowchart – Determination of the reactance b ... 51

Figure 26: Flowchart – Short Circuit ... 53

Figure 27: Stator voltage variation network ... 55

Figure 28: Flowchart – Stator Voltage Variation ... 56

Figure 29: Load-transfer network ... 58

Figure 30: Flowchart – Load Transfer ... 59

Figure 31: Frequency variation network ... 60

Figure 32: Flowchart – Frequency Variation ... 62

Figure 33: EUROSTAG network ... 64

Figure 34: Main window of AuDySim_HYD ... 66

Figure 35: Introduction window ... 70

Figure 36: Choice of the production site ... 70

Figure 37: Configuration of the production site – Machine, transformer and line ... 71

Figure 38: Configuration of a new machine... 71

Figure 39: Configuration of the production site – the regulators ... 72

Figure 40: Main window of AuDySim_NUC ... 73

Figure 41: Synthesis of the report ... 74

Figure 42: Presentation of the short-circuit test... 74

Figure 43: Hydraulic voltage regulator and excitation system ... 76

Figure 44: Nuclear voltage regulation and excitation system... 78

Figure 45: Nuclear speed regulator ... 78

Figure 46: Load flow – Hydro generation site... 80

Figure 47: Active power at PDC – PERF-STAB – Case 1 ... 81

Figure 48: Active power at PDC – PERF-STAB – Case 2 ... 81

Figure 49: Active power at PDC – PERF-STAB CC – Case 1 ... 83

Figure 50: Active power at PDC – PERF-STAB CC – Case 2 ... 83

Figure 51: Voltage at PDC – PERF-STAB CC – Case 1 ... 84

Figure 52: Voltage at PDC – PERF-STAB CC – Case 2 ... 84

Figure 53: Reactive power at PDC – PERF-STAB CC – Case 1 ... 85

Figure 54: Reactive power at PDC – PERF-STAB CC – Case 2 ... 85

Figure 55: Rotor speed – PERF-STAB CC – Case 1 ... 86

Figure 56: Rotor speed – PERF-STAB CC – Case 2 ... 86

(9)

EDF EFESE

R12 - CIST

Page 8 / 121

Figure 57: EAC – PERF-STAB CC – Case 1 ... 87

Figure 58: EAC – PERF-STAB CC – Case 2 ... 87

Figure 59: Active power at PDC – PERF-STAB PT MV a – Case 1 ... 89

Figure 60: Active power at PDC – PERF-STAB PT MV a – Case 2 ... 89

Figure 61: Stator voltage – PERF-STAB PT MV a – Case 1 ... 90

Figure 62: Stator voltage – PERF-STAB PT MV a – Case 2 ... 90

Figure 63: Active power at PDC – PERF-STAB PT MV b – Case 1 ... 91

Figure 64: Active power at PDC – PERF-STAB PT MV b – Case 2 ... 91

Figure 65: Stator voltage – PERF-STAB PT MV b – Case 1 ... 92

Figure 66: Stator voltage – PERF-STAB PT MV b – Case 2 ... 92

Figure 67: Load flow – Nuclear generation site ... 94

Figure 68: Active power at PDC – PERF-STAB – Nuclear Case 1 ... 96

Figure 69: Active power at PDC – PERF-STAB – Nuclear Case 2 ... 96

Figure 70: Active power at PDC – PERF-STAB CC – Nuclear Case 1... 97

Figure 71: Active power at PDC – PERF-STAB CC – Nuclear Case 2... 97

Figure 72: Voltage at PDC – PERF-STAB CC – Nuclear Case 1 ... 98

Figure 73: Voltage at PDC – PERF-STAB CC – Nuclear Case 2 ... 98

Figure 74: Reactive power at PDC – PERF-STAB CC – Nuclear Case 1 ... 99

Figure 75: Reactive power at PDC – PERF-STAB CC – Nuclear Case 2 ... 99

Figure 76: Rotor speed – PERF-STAB CC – Nuclear Case 1 ... 100

Figure 77: Rotor speed – PERF-STAB CC – Nuclear Case 2 ... 100

Figure 78: Active power at PDC – PERF-STAB PT MV a – Nuclear Case 1 ... 101

Figure 79: Active power at PDC – PERF-STAB PT MV a – Nuclear Case 2 ... 101

Figure 80: Stator voltage – PERF-STAB PT MV a – Nuclear Case 1 ... 102

Figure 81: Stator voltage – PERF-STAB PT MV a – Nuclear Case 2 ... 102

Figure 82: Active power at PDC – PERF-STAB PT MV b – Nuclear Case 1 ... 103

Figure 83: Active power at PDC – PERF-STAB PT MV b – Nuclear Case 2 ... 103

Figure 84: Stator voltage – PERF-STAB PT MV b – Nuclear Case 1 ... 104

Figure 85: Stator voltage – PERF-STAB PT MV b – Nuclear Case 2 ... 104

Figure 86: Active power at PDC – PERF-STAB REPORT CH – Nuclear Case 1 ... 105

Figure 87: Active power at PDC – PERF-STAB REPORT CH – Nuclear Case 2 ... 105

Figure 88: Reactive power at PDC – PERF-STAB REPORT CH – Nuclear Case 1 ... 106

Figure 89: Reactive power at PDC – PERF-STAB REPORT CH – Nuclear Case 2 ... 106

Figure 90: Frequency – PERF-STAB U SUR ΔF – Nuclear Case 1 ... 108

Figure 91: Frequency – PERF-STAB U SUR ΔF – Nuclear Case 2 ... 108

Figure 92: Active power at PDC – PERF-STAB U SUR ΔF – Nuclear Case 1 ... 109

Figure 93: Active power at PDC – PERF-STAB U SUR ΔF – Nuclear Case 2 ... 109

Figure 94: Stator voltage – PERF-STAB U SUR ΔF – Nuclear Case 1 ... 110

Figure 95: Stator voltage – PERF-STAB U SUR ΔF – Nuclear Case 2 ... 110

Figure 96: Frequency – Frequency drop – Nuclear Case 1 ... 111

Figure 97: Active power – Frequency drop – Nuclear Case 1 ... 112

Figure 98: Internal angle – Frequency drop – Nuclear Case 1 ... 112

Figure 99: Stator voltage – Frequency drop – Nuclear Case 1 ... 113

Figure 100: Stator voltage – Comparison of curves ... 114

Figure 101: Rotor speed – Comparison of curves ... 115

Figure 102: Active power injected by the generator – Comparison of curves ... 116

(10)

EDF EFESE

R12 - CIST

Page 9 / 121

LIST OF TABLES

Table 1: Parameters of a hydraulic salient-pole generator ... 28

Table 2: Parameters of a nuclear round rotor generator ... 28

Table 3: Grid Code tests ... 44

Table 4: Dimensioning voltage ... 46

Table 5: Initial load flow inputs ... 47

Table 6: Equivalent generator parameters ... 61

Table 7: Hydro generator parameters ... 76

Table 8: Nuclear generator parameters ... 77

Table 9: Hydraulic cases ... 79

Table 10: Nuclear cases ... 93

(11)

EDF EFESE

R12 - CIST

Page 10 / 121

ABBREVIATIONS – NOMENCLATURE

AC: Alternating Current

API: Application Programming Interface CCG: Combined Cycle Gas

CIST: “Centre d’Ingénierie Système Transport” – Engineering Center for Transportation System

DC: Direct Current

dll: Dynamic Link Library

DSO: Distribution System Operator

DTR: “Documentation Technique de Référence” – Technical Document of Reference EAC: Equal Area Criterion

EDF: “Electricité De France” – Electricity of France

EFESE: “Economie, Fonctionnement et Etudes des Systèmes Energétiques” – Economic and Technical Analysis of Energy Systems

emf: Electromotive Force

ENTSOE: European Network of Transmission System Operators for Electricity GUI: Graphical User Interface

GUIDE: Graphical User Interface Development Environment IEC: International Electrotechnical Commission

IEEE: Institute of Electrical and Electronics Engineers mmf: Magnetomotive Force

PDC: “Point De Connexion” – Point of Connection PSS: Power System Stabilizer

pu: per-unit

RTE: “Réseau de Transport d’Electricité” – Electricity Transmission Network SMIB: Single Machine Infinite Bus

TSO: Transmission System Operator

(12)

EDF EFESE

R12 - CIST

Page 11 / 121

UCTE: Union for the Coordination of the Transmission of Electricity

(13)

EDF EFESE

R12 - CIST

Page 12 / 121

1. INTRODUCTION

1.1. Background

The electrical power system can be represented by a supply chain that can be divided into three main activities. On one side, the electricity is produced thanks to power plants using hydraulic, nuclear, fossil or renewable resources. In the middle the electricity has to be transported through the transmission and then the distribution networks consisting of lines which are based on alternating or direct current with various levels of voltage. The consumption is on the other side with industries or domestic use. A representation of the operation of the network is given in Figure 1. To ensure the delivery of electricity to the consumers at any time, the entire system is monitored and studies are realized to prevent major issues. This is particularly true for the electricity generators connected to the network. A part of these studies consists of a set of simulations to analyze the behavior of the synchronous generators when subjected to a predetermined event. These simulations compose the main subject of this Master Thesis.

Figure 1: Electrical power network [1]

The electric power system contains interconnected generators that provide electricity to the consumers using the high voltage transmission network and medium and low voltage distribution lines. The generators are synchronized on a unique frequency common to the whole

(14)

EDF EFESE

R12 - CIST

Page 13 / 121

grid. In the European grid, which includes continental Europe from Portugal to Poland in the North-East and Greece in the South-East except for the islands, the common frequency is aligned at 50 𝐻𝑧. This grid is composed of lines, either transporting AC (Alternating Current) for the most common or DC (Direct Current) for long transmission lines. On AC lines in France, various levels of voltages are present. Extra high voltage corresponds to 400 𝑘𝑉 lines while high voltage corresponds to 225 𝑘𝑉 and 150 𝑘𝑉. When a machine is connected to the high voltage network, the voltage at the output has to be elevated from a value within the 10 𝑘𝑉 range to the high voltage of the evacuation line. To do so, transformers are installed between the generator and the connection point to the grid. The generators are composed of an alternator and modules, whose purpose is to control the output of the generator. Typically a generator is equipped with an excitation system, a voltage regulator and eventually a speed regulator [2]. In addition to the generator and the transformer, the generation unit can contain auxiliaries alimented by the generator, whose purpose is to ensure the security of the electricity production site.

The interconnected European electricity network can be subjected to failures or faults and has to remain operational even through these events. Sometimes, a multiplication of linked or independent events can be a threat to the grid and cause severe disturbances. Two of the biggest examples of disturbances in the European grid happened in 2003 in Italy and in 2006 in the whole grid. On the 28^th of September, 2003, Italy was importing 6.4 𝐺𝑊 of power from its neighbors [3]. Due to a tree flashover, a key line connecting Switzerland to Italy was tripped.

As it could not reconnect, a second line close to the first one was disconnected due to overload and it created a cascading effect which led to the disconnection of the transmission lines connecting Italy to the rest of the European network. With a deficit of production amounting to 6.4 𝐺𝑊 and some decisions not taken at the right time, the Italian network collapsed and the country reached a blackout state. The analyses made after this event showed malfunctioning in the communication between the TSOs (“Transmission System Operator”) as well as a lack of risk-assessment. During this event, some generators did not work as planned and found themselves disconnected when they were supposed to still be active. This is a problem that needed to be solved and led to studies to ensure that this behavior would not happen again. This Master Thesis lies in the continuity of those studies. In 2006 on the 4^th of November, another major event occurred affecting most of the interconnected European countries. A lack of communication between the German TSOs on the planned opening of a transmission line coupled to a high consumption in the Western countries led to a separation of the European grid

(15)

EDF EFESE

R12 - CIST

Page 14 / 121

into three parts having different frequencies. The aggravating factor of this event was a large import of 9.26 𝐺𝑊 of active power by the Western part of Europe from North-Eastern Europe and Germany where a large part of the production corresponded to wind power [4]. As in the event of 2003, a cascading effect of disconnection of key lines led to a power imbalance in Europe which was separated into three zones corresponding to the West, North-East and South- East of Europe. A complete blackout was avoided thanks to load shedding which deprived 15 millions of Europeans from electricity for minutes. As a result, the three parts of Europe remained in operation at different frequencies as it can be seen in Figure 2 showing the frequency for each part during the event. Later a resynchronization of the separated regions occurred to restore the system.

Figure 2: Frequencies during the 2006 event [4]

To avoid these situations, the TSOs have created a set of rules gathered in documents called the Grid Codes. These documents written by each TSO and applicable on its network contain the requirements to ensure a normal operation of the grid [5] [6]. The companies which own elements connected to the network or want to connect a new element have to demonstrate that these elements satisfy various criteria. The situation of an interconnected grid with possibly different rules of connection in every part of it can trigger debates [7]. These debates, in addition to approximations at the creation of these codes [8], led to a constant evolution of the grid codes.

The current trend is in the direction of a harmonization between the different TSOs applying these rules as well as the international organizations proposing standards like the IEC

(16)

EDF EFESE

R12 - CIST

Page 15 / 121

(International Electrotechnical Commission), the IEEE (Institute of Electrical and Electronics Engineers) and the ENTSOE (European Network of Transmission System Operators for Electricity) [9] [10] [11]. The events presented in the previous paragraph led to evolutions of these grid codes. The analyses made after the failures demonstrated the need for a more complete risk-assessment, part of which consists of simulations in case of such events [3]. A concrete example of such an evolution concerns the wind farms after the event of 2006. During the failure, the wind power plants were automatically disconnected from the grid when the frequency got under 49.5 𝐻𝑧 aggravating the drop of frequency as explained by a report from the Union for the Coordination of the Transmission of Electricity (UCTE) [4]. Then, when the frequency increased the wind farms reconnected without any control from the TSOs. All this was done according to the grid codes and led to a worsening of the situation. After that the grid codes were adapted for a smoother connection and disconnection of wind farms.

The grid code applied by the French TSO, RTE (“Réseau de Transport d’Electricité”), on the French network is written in the DTR (“Documentation Technique de Référence”). It contains all the requirements from the TSO regarding the data that every connected element has to provide. When it comes to generating units it includes operating data of the generator and, the relevant section for this Master Thesis, simulations showing the behavior of the generation unit when subjected to perturbations [12]. Like every grid code, the DTR keeps evolving and the version that will be used in this Master Thesis was published and applicable on the 13^th of April, 2017.

As the main French electricity producer and provider, EDF (“Électricité de France”) is involved in these studies. Its activities are focused on providing energy to its clients in France as well as Europe and all over the world. The company covers a large spectrum of activities including the production, the distribution, sales of energy and the development of innovative solutions. In France, the production of electricity is mainly achieved using nuclear and hydro power which leads to a low rate of CO2 emissions. However, the means of production also include thermal power plants based on fossil fuels (e.g. gas) and an increasing share of renewables, led by wind farms and photovoltaic panels.

EDF and all its customers are directly impacted by the requirements from the DTR. In addition to new studies for every newly connected generating unit, the company must also provide RTE with documents showing the performance of existing generating units which have been in operation for more than ten years. These studies have to be conducted on nuclear, fossil-

(17)

EDF EFESE

R12 - CIST

Page 16 / 121

fired and hydraulic generators. The CIST (“Centre d’Ingénierie Système Transport”) is the entity at EDF responsible for delivering the documents to RTE. These studies are essential to demonstrate the ability of the connected synchronous generators to withstand failures and maintain an active electricity distribution for the European grid and all the connected consumers. They are also dedicated to the detection of abnormal behaviors from generators which could harm the system.

This Thesis focuses on a particular article from the DTR where a set of tests is described [12]. These tests correspond to the connection of a generator to a simplified model of the grid and the creation of events such as a short circuit on a line. Then the behavior of the generator is analyzed emphasizing on the stability and the response time for the active power and/or the voltage.

1.2. Objectives

In this context, the CIST has to realize simulations for each generating unit connected to the grid. The important number of production sites especially for hydro generation – more than 60 in France – has been the trigger for this Master Thesis. The objective is to create a tool in the shape of a software to automatically realize some designated tests and write the report on the performance for each production site. After each simulation, the stability as well as the performance of the power system has to be analyzed. This tool should create a fast method to determine the performance of a generation unit on events designated in the grid codes.

The implemented tool would allow an electricity producer to perform in-depth analyses of its synchronous generation units to make sure that they comply with the requirements from the TSO. The idea is to detect any possible anomaly in the results of a simulation to warn the producer and the TSO of the potential danger and take measures to prevent it. This would be beneficial for the companies involved as well as all the individuals using electricity from a synchronous grid. It asserts that the electrical power system is secured and avoids events such as the 2003 and 2006 blackouts in the European grid.

Finally, the tool should give the possibility to evaluate the quality of the models created to represent the regulations and the excitation system. Its modularity should also enable the user to test the validity of the data concerning the synchronous machine and the transformer.

(18)

EDF EFESE

R12 - CIST

Page 17 / 121

1.3. Limitations

Due to confidentiality issues, the parameters used in the synchronous machines are not displayed in this report. Only a range of variation for each parameter is given from values found in the literature.

1.4. Previous work

To perform the tests presented in the DTR or in similar grid codes, it is suspected that electricity producing companies other than EDF might have developed their own tools or scripts to realize the simulations. However no trace of it has been found in the available documentation.

Internally at EDF, a tool to perform similar simulations including the redaction of a report created exclusively with MATLAB and Simulink had already been developed [13]. This tool was in the finalization stage of its development. It allowed to configure a generation unit by introducing the data for the synchronous machine and the transformer. The models for the components as well as the regulations had been created using the Simscape Power Systems library in Simulink. Then the tests taken from a former version of the DTR were realized using Simulink. Once all the simulations were finished, the curves were plotted automatically and a report containing them was self-written. However, each set of simulations including the redaction of the report took between half an hour and an hour and half. The development of the tool from this Master Thesis will be based on elements from this existing tool, with improvements and complete additions of elements such as the utilization of EUROSTAG.

1.5. Contributions

This Master Thesis contributes to the electricity power system by the creation of a tool allowing the automatization of electromechanical simulations stipulated in the grid codes. The original focus was on hydraulic generators but the tool has been developed in four versions corresponding to hydraulic, nuclear and fossil-fired generators and a last version allowing the comparison of generation units. With each version, the latest simulations stipulated in the DTR are realized using the powerful capacities of EUROSTAG. The use of this software reduces the duration of the simulations for a generation site to two to five minutes compared to the previous works. Then a complete analysis of the results is automatically performed for each simulation and the results as well as the conclusions are displayed in a report. This tool corresponds to an update of the previous works with various enhancements comprising time efficient simulations and detailed results. To sum up the contributions, the outcome of this Master Thesis participates

(19)

EDF EFESE

R12 - CIST

Page 18 / 121

to the improvement and the security of any grid containing synchronous generators.

1.6. Disposition

The report is organized in the following sections. Chapter 1 presents the introduction containing the context and the objectives of this Master Thesis. The relevant theory used throughout this Master Thesis is in Chapter 2. Chapter 3 covers the methodology used in the study which consists of the development of the tool. Besides the results from the study are presented in Chapter 4 to verify the performance of the tool. Then Chapter 5 develops a discussion on the results and the performance of the implemented tool. Finally the conclusions close this report in Chapter 6.

(20)

EDF EFESE

R12 - CIST

Page 19 / 121

2. THEORETICAL BACKGROUND

In this Master Thesis, the stability of electricity production sites is analyzed. In order to understand the context of the study, a brief description of steam, gas and hydraulic turbines as well as round-rotor and salient-pole generators is made in the two following sections. The last section deals with the most important phenomenon of this Master Thesis which is the rotor angle stability under all its aspects.

2.1. Turbines

2.1.1. Steam and Gas Turbines – Nuclear, Fossil Fired power plants

Before describing the hydraulic turbine in the next part, this section presents the steam and the gas turbines. These turbines are used in the conventional electricity power plants which are the coal-fired and nuclear ones for the steam turbine and gas turbines for gas-fired power plants. An additional kind of power plant, the Combined Cycle Gas (CCG) plant, uses both a steam and a gas turbine.

The steam and the gas turbine share a common way of operating. They use a high pressure component (e.g. steam or gas) guided in an axial flow turbine. This component presents a high momentum that is transmitted to the blades present on the turbine. Thanks to this exchange of momentum, the blades rotate thus creating a torque on the axis of the turbine [14] [15]. A single-reheat steam turbine is presented in Figure 3. It consists of a close-circuit where the steam navigates in different states. The governing system (GOV) pilots the boiler and the valves to regulate the process. The boiler heats the water to create high temperature steam thanks to a heat source. This source can either come from burning coal or gas, or from a nuclear fission reaction. Once the steam leaves the boiler, it is conducted through valves in the high-pressure (HP) section of the turbine containing blades. This section provides typically 30 % of the total torque of the turbine [14]. Then the steam is reheated before going through the intermediate-pressure (IP) section. This part of the turbines contributes to 40 % of the total torque. Finally, the steam leaving the IP section enters directly the low-pressure (LP) section where the last 30 % of the torque are produced. In all the sections blades of adapted shapes are present. The turbine presented in Figure 3 presents a single reheating circuit. It is the most common type but non-reheat and double-reheat steam turbines can also be found in the industry.

(21)

EDF EFESE

R12 - CIST

Page 20 / 121

Figure 3: Steam turbine [14]

The difference between a steam and a gas turbine is the absence of an intermediate process in the gas turbine. There is no need for a boiler to heat up steam as the element that goes through the turbine is directly the exhaust gas from the burning of the fuel. A diagram showing the process of a gas turbine is displayed in Figure 4. A compressor directs air into the combustion chamber where it is mixed with the fuel and burnt. The exhaust gas from the combustion is directed into the turbine where it acts on rotating blades to create a torque. Then the exhaust gas leaving the turbine is directed into a heat exchanger to heat up the air before it enters the combustion chamber. Finally the exhaust is released.

A combination of the two processes has been implemented in CCG plants using the exhaust from the combustion of the gas to both drive a gas turbine and create steam by boiling water. This steam is then directed into a steam turbine on a different or the same shaft to produce an additional torque. This method increases the efficiency compared to a single gas turbine [14].

Figure 4: Gas turbine [14]

(22)

EDF EFESE

R12 - CIST

Page 21 / 121 2.1.2. Hydraulic Turbines

This section describes briefly the different types of turbine used in the hydraulic electricity production sites. The concept as well as the advantages and drawbacks of each type of turbine are depicted below. The different types of turbine can be separated by their head. The head is the height of the waterfall in the conduit from the water reserve to the turbine. The high- head turbines, for a head greater than 300 𝑚 [15], are Pelton wheels while low- and medium- head turbines, for a head lower than 360 𝑚 [16], are either Francis or Kaplan turbines.

The Pelton wheel is commonly used for high waterfalls. The principle of a Pelton wheel is described in Figure 5. The water is stored in a reserve and passes through a conduit and a penstock to make it gain kinetic energy. Then the water is directed at high velocity onto the bowl-shaped buckets thanks to the nozzle. The buckets deflect the incoming water and turn it back so that a maximum of energy is transferred to the turbine, providing mechanical torque.

The needle present on the nozzle allows a control of the size of the water jet thus controlling the power output [14]. The jet deflector is a safety component which removes the water jet from the buckets in case of a rapid load change. In practice, there are various nozzles on one Pelton wheel making various jets impacting on the same wheel as it can be seen in Figure 6. An advantage of the Pelton wheel is an easy maintenance of the wheel.

Figure 5: Principle of a Pelton wheel [16]

(23)

EDF EFESE

R12 - CIST

Page 22 / 121

Figure 6: Detail of a Pelton wheel [15]

Figure 7: Design of a Francis and a Kaplan turbine [17]

When it comes to low- and medium-head turbines, the two types of turbine involved are the Francis and the Kaplan turbines. Their design can be seen in Figure 7. The Francis turbine can be used as both a low- and a medium-head while the Kaplan turbine is a low-head turbine.

Both of these types of turbine operate at low speed implying a large diameter for the turbine.

For the Francis type, the incoming flow of water driving the turbine is radial while in the Kaplan turbine, the flow is axial. The Francis can be used either with a penstock like for the Pelton but with spiral volute to guide the water onto the turbine, or with a low-head and a large flow of water instead of a waterfall. The Kaplan turbine is only used with low-head, corresponding to a run-of-the-river hydraulic production site. Figure 8 describes the structure of a Kaplan turbine.

The water comes from the river flow, through a spiral volute, through the wicket gates or guide vanes and onto the blades. In case of a Kaplan turbine, the blades are adjustable allowing to modify the power output. In both cases the wicket gates are used to control the power output of

(24)

EDF EFESE

R12 - CIST

Page 23 / 121

the turbine. The advantage from using a turbine whose blades have a variable pitch is to get a high efficiency at any load.

Figure 8: Principle of a Kaplan turbine [14]

A last type of turbine used in the Rance tidal power station is the bulb turbine which is a derivative from the Kaplan turbine. Its structure, represented in Figure 9, is close to a horizontal axis Kaplan turbine. In this particular power station, the turbine uses the tidal flow on the blades to create the torque and then the power output.

Figure 9: Structure of a bulb turbine [18]

2.2. Generators

In the previous section, the structure of the turbine is explained. The turbine with its rotating runner is connected to a generator through the driveshaft. It is the generator which

(25)

EDF EFESE

R12 - CIST

Page 24 / 121

actually produces electrical power from rotating mechanical power. In this part the round-rotor generator is described first to get an understanding of the type of generator commonly used with steam and gas turbines. Then the theory about the salient-pole generator used with hydraulic turbines is explained based on the description of the previous type of generator.

Finally, all the parameters that are used to represent the generators in this Thesis are stated and quantified. Only the most important and interesting equations are written, the objective of this section being to describe the physical interactions and present the parameters used in the Thesis.

2.2.1. Round-Rotor Machines – Nuclear, Fossil-Fired Power Plants

The round-rotor machine is a type of synchronous generator mainly used for high speed applications for instance connected to a steam or gas turbine. It is composed of a rotating part connected to the turbine called the rotor. The DC (Direct Current) field winding is wound around this rotor. The second main part of the synchronous generator is the stator. This fixed component is around the rotor and bears the three-phase armature winding which carries the power in the direction of the electrical grid. The structure of a two-pole round-rotor machine is presented in Figure 10 without load.

Figure 10: No load Round-rotor machine [14]

The DC rotor windings are indicated by f1 and f2 while the three-phase stator winding is represented by the A-phase (a1 and a2 the beginning and the end of the winding), the B-phase (b1 and b2) and the C-phase (c1 and c2), each shifted by 120°. In order to get equations from this system, the d- and q-axis are defined on the rotor. The d-axis is the main magnetic axis of the field winding [14], while the q-axis is in quadrature with the d-axis. As the rotor is moving, the DC field windings it bears are also rotating, creating a rotating field flux called Φf as well as a

(26)

EDF EFESE

R12 - CIST

Page 25 / 121

field leakage flux Φfl. A last variable is defined on the rotor which is 𝐹⃗⃗⃗ the magnetomotive _𝑓 force (mmf) wave resulting from the action of the field current.

When a load is connected to the generator, the resultant rotating air-gap flux Φr appears as depicted in Figure 11. The load also implies a reaction from the armature represented by the armature reaction 𝐹⃗⃗⃗ which is an mmf wave. Thus the resultant mmf wave is 𝐹_𝑎 ⃗⃗⃗ and is aligned _𝑟 with the air-gap flux. Another parameter not represented on the schematic of the round-rotor is δfr called the torque angle between 𝐹⃗⃗⃗ and 𝐹_𝑟 ⃗⃗⃗ . _𝑓

Figure 11: Loaded Round-rotor machine [14]

Based on this description and considering the electromagnetic interactions between the stator and the rotor, the following formula can be given for the torque of the two-pole round- rotor machine [14].

It is also possible to consider an expression for the electrical power output of the round- rotor generator for transient studies based on the line-to-neutral voltage U, the electromotive force (emf) on the q axis 𝐸_𝑞^′ and the transient reactance on the d-axis 𝑥_𝑑^′ [19].

2.2.2. Salient-Pole Machines – Hydraulic Power Plants

A salient-pole machine is also a synchronous generator. The main difference between the round-rotor and salient-pole machines is the air-gap between the stator and the rotor. While the width of the air-gap is constant for a round-rotor generator, in a salient-pole machine the

𝜏 = 𝜋

2𝐹_𝑟Φ_𝑓sin 𝛿_𝑓𝑟 (2.1)

𝑃_𝑒 = 3𝐸_𝑞^′𝑈

𝑥_𝑑^′ sin 𝛿_𝑓𝑟 (2.2)

(27)

EDF EFESE

R12 - CIST

Page 26 / 121

width varies between the d- and q-axis. In practice, the air-gap is narrow on the d-axis and wide on the q-axis as it can be seen in Figure 12. Another difference resides in the utilization made of this generator. It is used for low speed applications which makes it prominent for hydroelectric generation and is usually composed of a larger diameter than for the round-rotor generator.

Figure 12: Salient-pole generator [14]

As it can be seen in Figure 12, the air-gap width is varying when the rotor is moving in the stator. The corresponding variation of the reluctance in the air-gap is described in Figure 13. It is maximal on the q-axis when the air-gap is wider and minimal on the narrow d-axis.

The problem that this phenomenon triggers is that for the round-rotor generator, the reluctance of the air-gap had been assumed to be constant. Thus the formulas valid for the round-rotor machine are not valid anymore for the salient-pole generator. The solution to circumvent this issue is to consider a constant value for the mmf (or the emf) on the d- and q-axis, different on the two axes.

Figure 13: Variation of the air-gap reluctance [14]

Knowing that the rotor tries to find the position with a minimum reluctance and using this assumption in the case of a salient-pole generator allows to find the expression for the torque as it is done in (2.1). In the following equation ℛ_𝑑 and ℛ_𝑞 are the d- and q-axis reluctances [14].

(28)

EDF EFESE

R12 - CIST

Page 27 / 121

Equation (2.3) is composed of two terms, the first term is the synchronous torque, identical to (2.1) while the second term is the reluctance torque. This formula also applies for the round-rotor generator where ℛ_𝑞 = ℛ_𝑑 with the second term disappearing in (2.3).

As it is done for the round-rotor, an expression can be found for the electrical power of the salient-pole generator with the introduction of the q-axis reactance 𝑥_𝑞 [19].

This formula is also composed of the synchronous term valid for the round-rotor machine and an additional term due to the varying air gap and thus the varying reactance on the d- and q-axis that corresponds to it.

2.2.3. Representation of the Generators

In this Master Thesis, the generators are implemented using a software called EUROSTAG. Thus the chosen representation of the machine is adapted to the parameters used in the software. The model used to describe the generator is the one of a salient-pole generator with 3-phase stator winding for hydraulic generation units. The parameters used as inputs in this model are listed in Table 1. The typical values for hydraulic generators are listed in the last column of Table 1. The following study uses values contained within the mentioned ranges.

Parameter Corresponding representation Value range [16] [20]

Stator resistance 𝑅_𝑎 0.002 – 0.02 (pu)

Stator leakage 𝑋_𝑎 0.1 – 0.2 (pu)

Direct reactance 𝑋_𝑑 0.9 – 1.5 (pu)

Direct transient reactance 𝑋_𝑑^′ 0.3 – 0.5 (pu)

Direct sub-transient reactance 𝑋_𝑑^′′ 0.25 – 0.35 (pu)

Quadrature reactance 𝑋_𝑞 0.5 – 1.1 (pu)

Quadrature sub-transient reactance 𝑋_𝑞^′′ 0.25 – 0.35 (pu)

Direct transient time constant 𝑇_𝑑0^′ 3 – 8 (s)

Direct sub-transient time constant 𝑇_𝑑0^′′ 0.006 – 0.2 (s)

𝜏 = 𝜋

2𝐹_𝑟Φ_𝑓sin 𝛿_𝑓𝑟+ 𝜋

4𝐹_𝑟²ℛ_𝑞− ℛ_𝑑

ℛ_𝑞ℛ_𝑑 sin 2𝛿_𝑓𝑟 (2.3)

𝑃_𝑒 = 3𝐸_𝑞^′𝑈

𝑥_𝑑^′ sin 𝛿_𝑓𝑟+ 3𝑈² 2 (1

𝑥_𝑞− 1

𝑥_𝑑^′) sin 2𝛿_𝑓𝑟 (2.4)

(29)

EDF EFESE

R12 - CIST

Page 28 / 121

Quadrature sub-transient time constant 𝑇𝑞0′′ 0.028 – 0.22 (s)

Mechanical damping coefficient 𝑑 0 – 0.2 (pu)

Constant of inertia 𝐻 2 – 4 (MW.s/MVA)

Coefficients of the saturation curve 𝑚_𝑑, 𝑚_𝑞, 𝑛_𝑑, 𝑛_𝑞 𝑚𝑑/𝑞≈ 0.1 ; 𝑛𝑑/𝑞≈ 6 Table 1: Parameters of a hydraulic salient-pole generator

Contrary to hydraulic generators, the nuclear and fossil-fired generators are represented by a round-rotor generator with 3-phase stator and damper windings. This modifies the typical values for each component and adds two parameters, the quadrature transient reactance and its time constant. The parameters and the corresponding values for a typical nuclear generator are given in Table 2.

Parameter Corresponding representation Value range [16] [20]

Stator resistance 𝑅_𝑎 0.0015 – 0.005 (pu)

Stator leakage 𝑋_𝑎 0.1 – 0.2 (pu)

Direct reactance 𝑋_𝑑 1.5 – 2.5 (pu)

Direct transient reactance 𝑋_𝑑^′ 0.2 – 0.35 (pu)

Direct sub-transient reactance 𝑋_𝑑^′′ 0.15 – 0.25 (pu)

Quadrature reactance 𝑋𝑞 1.5 – 2.5 (pu)

Quadrature transient reactance 𝑋𝑞′ 0.3 – 1 (pu)

Quadrature sub-transient reactance 𝑋𝑞′′ 0.15 – 0.25 (pu)

Direct transient time constant 𝑇𝑑0′ 8 – 12 (s)

Direct sub-transient time constant 𝑇𝑑0′′ 0.01 – 0.16 (s) Quadrature transient time constant 𝑇_𝑞0^′ 0.5 – 2 (s) Quadrature sub-transient time constant 𝑇_𝑞0^′′ 0.12 – 0.83 (s)

Mechanical damping coefficient 𝑑 0 – 0.2 (pu)

Constant of inertia 𝐻 4 – 10 (MW.s/MVA)

Coefficients of the saturation curve 𝑚_𝑑, 𝑚_𝑞, 𝑛_𝑑, 𝑛_𝑞 𝑚_𝑑/𝑞≈ 0.1 ; 𝑛_𝑑/𝑞≈ 6 Table 2: Parameters of a nuclear round rotor generator

For the mechanical damping coefficient, a value of 𝑑 = 0 𝑝𝑢 is used in all the models as it does not affect the simulations and is not provided in the available data given for each

(30)

EDF EFESE

R12 - CIST

Page 29 / 121 production site [20].

The saturation model of the synchronous generators in EUROSTAG is based on the open-circuit saturation curve. Figure 14 shows the open circuit characteristic of the machine with the stator voltage 𝑣 in function of the excitation current 𝑖 and the air gap line, tangential to the curve near the origin. This characteristic is described by equation (2.5) [20].

In this equation the saturation coefficients are 𝑚_𝑑 = 𝑚_𝑞 = 𝑚 and 𝑛_𝑑 = 𝑛_𝑞 = 𝑛. From (2.5) and the saturation curve in Figure 14 it is possible to find a formula for 𝑚 and 𝑛.

Figure 14: Saturation curve

Based on the three points referred by (𝐼_𝑎, 1), (𝐼_𝑏, 1) and (𝐼_𝑐, 𝑥), the expressions of the saturation coefficients are given by [20]:

The model for the synchronous generator used by EUROSTAG is based upon the PARK equations which are not described in this report but are available in [21].

2.3. Power System Stability

After having described the turbine and the generator of an electricity production system the next step, which is the content of this section, is to focus on the stability of these power plants when connected to the transmission grid. To do so, the basis about stability is presented before moving on to the rotor-angle stability applied to the generator.

𝑖 = 𝑣. (1 + 𝑚. 𝑣^𝑛) (2.5)

𝑚 = 𝐼_𝑏− 𝐼_𝑎

𝐼_𝑎 (2.6)

𝑛 = 1

ln 𝑥 × ln ( 𝐼_𝑐 − 𝑥. 𝐼_𝑎

𝑥. (𝐼_𝑏− 𝐼_𝑎)) (2.7)

Optimization of Electromechanical Studies for the Connection of Hydro Generation

Optimization of Electromechanical Studies for the Connection of

Hydro Generation

MATHIEU GROULT

Master Thesis Report

Optimization of Electromechanical Studies for the Connection of Hydro Generation

ABSTRACT

SAMMANFATTNING

ACKNOWLEDGEMENTS

TABLE OF CONTENTS

LIST OF FIGURES

LIST OF TABLES

ABBREVIATIONS – NOMENCLATURE

1. INTRODUCTION

1.1. Background

1.2. Objectives

1.3. Limitations

1.4. Previous work

1.5. Contributions

1.6. Disposition

2. THEORETICAL BACKGROUND

2.1. Turbines

2.2. Generators

2.3. Power System Stability