wie { per eS a he ransmicsion\ ‘ ASYSIEINS) 4 Tecnnsiegyandl System UN ons EN is NNContents Preface ye Power Transmission Technology Introduction 1 a Comparison of AC and DC Transmission 2 13 Application of DC Transmission 7 14 Description of DC Transmission System 8 1.5 Planning for HVDC Transmission 15 1.6 Modem Trends in DC Transmission 18 2. Thyristor Valve 2.1 _ Introduction 21 22 Thyristor Device 21 23° Thyristor Valve 30 24 Valve Tests 37 25 Recent Trends 39 3-Knalysis of HVDC Converters 3.1 Pulse Number 42 3.2 Choice of Converter Configuration 43 3.3. Simplified Analysis of Graetz Circuit 46 werter Bridge Characteristics 58 3.5 Characteristics of a Twelve Pulse Converter 61 3.6 Detailed Analysis of Converters 65 ee convection HVDC System Control iat General 76 ,Principles of DC Link Control 76 t as ‘Converter Controt Characteristics 79 3-44 System Control Hierarchy 82 45 Firing Angle Control 84 46 Current and Extinction Angle Control 89 7 Starting and Stopping of DC Link 90 48 LE ae 49 Higher Level Cont 94 a 10 Telecommunication Requirements 96 a a2 6x Contents 5. Converter Faults and Protection 5.1 52 53 54 55 56 Introduction a Be werter Faults covetion Against Overcurrents 103, Overvoltages in a Converter Station 104 Surge Arresters 106 poesia Against Overvoltages 107 6. Smoothing Reactor and DC Line 61 67 Introduction 110 Smoothing Reactors 110 DC Line 113 ‘Transient Over Voltages In DC Line 118 Protection of DC Line 121 DC Breakers 122 Monopolar Operation 126 Effects of Proximity of AC and DC Transmission Lines 127 1 ¢/ Reactive Power Control 7 72 73 TA 15 Introduction 130 Reactive Power Requirements in Steady State 130 Sources of Reactive Power 136 Static Var Systems 138 Reactive Power Control During Transients 144 JH Harmonics and Filters w4l 62 3 AAs 85 Introduction 145 Generation of Harmonics 145 Design of AC Filters 149 DC Filters 156 Carrier Frequency and RI Noise 157 9. Multiterminal DC Systems 91 92 93 94 95 Introduction 159 Potential Applications of MTDC System: Cees of MDE Systems 160 vies 1s ‘ontrol and Protection of MTDC Study of MTDC Systems 169 Systems 164 10. Qomponent Medel ar General on” the Analysis of AC/DC Systems 110 130 145 159 172ee ates peg es Contents xi “10.2 Converter Model 172 0.3 Converter Control 179 +104 Modelling of DC Network 181 a_N. 12, 14, 10.5 Modelling of AC Networks 182 Power Flow Analysis in AC/DC Systems 188 11.1 General 188 11.2. Modelling of DC Links 188 11.3 Solution of DC Load Flow 192 4 11.4 Discussion 192 11.5 Per Unit System for DC Quantities 193 11.6 Solution of AC-DC Power Flow 194 11.7 An Example : Five Terminal DC System 196 Transient Stability Analysis 199 12.1 Introduction 199 12.2 Converter Model 199 12.3, Converter Controller Models 201 12.4 DC Network Models 201 12.5 Solution Methodology 204 12.6 Direct Methods for Stability Evaluation 208 12.7 Transient Stability Improvement Using DC Link Control 209 Dynamic Stability and Power Modulation 211 13.1 Introduction 211 13.2. Power Modulation For Damping Low Frequency Oscillations : Basic Principles 211 13.3 Practical Consideration in the Application of Power Modulation Controllers 214 134 Gamma or Reactive Power Modulation 219 135 Power Modulation in MTDC Systems 220 13.6 Voltage Stability in AC/DC Systems 220 13.7 An Integrated Dynamic Model for AC/DC Systems 224 Harmonic and Torsional Interactions 231 14.1 Introduction 231 % 14.2. Harmonic Interaction 232 14.3, Torsional Interactions 237 4A Torsional Interactions with HVDC Systems 242 14.5 Counter Measures to Torsional Interactions with DC Systems 249xii Contents 15. Simulation of HVDC Systems 15.1 Introduction 252 15.2 System Simulation : Philosophy and Tools 253 153 HVDC System Simulation 254 : 15.4 Modelling of HVDC Systems for Digital Dynamic Simulation 259 16. Digital Dynamic Simulation of Converters and DC Systems 16.1 General 264 16.2 Valve Model 264 163 Gate Pulse Generation 265 16.4 Generation of Control Voltage 266 16.5 Transformer Model 267 16.6 Converter Model 270 16.7 Transient Simulation of DC and AC Systems 278 Appendix A Index 252 2871. DC Power Transmission Technology 1.1 INTRODUCTION ‘The industrial growth of a nation requires increased consumption of energy, particularly electrical energy. This has led to increase in the generation and transmission facilities to meet the increasing demand. In U.S.A., till the early seventies, the demand doubled every ten years. In developing cor like India, the demand doubles every seven years which requires conside iderable investment in electric power sector. imperative of supplying energy at reasonable costs coupled with the depleting re serves of non-renewable energy sources has led to the establishment of remote gener- ating stations - predominantly fossil-fuel fired thermal stations at pit head. Environ- mental considerations also sometimes dictate the siting of power stations at remote locations, Large hydro stations are invariably at distances of hundreds of kilometres from load centres. The need to economize on costly investments in generation reserves, sharing of benefits in utilising variability in generation mixes and load patterns havc! given rise to interconnection of neighbouring systems and development of large power grids. Remote generation and system interconnections lead to a search for efficie . eee levels. The increase in voltage level luge levels snot alway fea sil ‘problems of AC transmission particularly in long distance ‘transmission, I "led to the development of DC transmission. However, as s generation and utilisation of Power remaut ting current, the DC transmission requires conversion at two “ends, from from AC to DC at the sending end and back to AC at the receiving end. This “conversion is done at converter stations - rectifier station at the sending end and inverter Station at the receiving end. The converters are static—using high power ee ysis connected in series to give f the required vol tage ratings. The ‘Physical ‘process Conversion is such that the same station can switch from rectifier to inverter by simple control action, thus facilitating power reversal. The HVDC transmission made a modest beginning in 1954 when a 100kV, 20 MW DC link was established between Swedish mainland and the island of Gotland. Until 1970, the converter stations utilised mercut valves for rectification. The successful use of thyristors for power control in industrial devices ‘encouraged its adoption in HVDC converters by development of high power semiconductor devices. The largest device rating is now in the range of 5 kV, 3000A. The highest transmission voltage reached is +600 kV. ‘The relative merits of AC and DC transmission are reviewed in the next section.HVDC Power Transmission System: 2 12 pc ‘TRANSMISSION COMPARISON OF AC AND nd DC) which need to be ‘The relative merits of the two modes of transmission ee ee considered by a system planner are based on the following : 1) Economics of transmission 2) Technical performance 3) Reliability. ‘A major feature of power systems is ¢ creasing power demand. This implies that the ‘establishment of considered as a part of an overall long term system planning. 1s expansion necessitated by in- the continuous ¢ Pea particular fine must be 1.2.1 Economics of Power Transmission ‘The costofa transmission line includes the investmentand operational costs. Theinvest- ment includes costs of Right of Way (RoW), trarismission towers, conductors, insula- tors and tesminal equipment. ‘The operation: costs include mainly the cost of losses. "The characteristics of insulators vary with type of voltage applied. For simplicity, if it is assumed that the insulator characteristics are similar for AC and DC and depend on the peak level of voltage applied with respect to ground, then it can be shown that for S Hines designed with the same insiation level, a DC line can‘carry as t much power with two conductors (with positive and negative polarities with respect to ground) as an AC _ ine with 3 conductors of the same size. This implies that for a given power level, DC “Tine requires less RoW, simpler and cheaper towers and reduced conductor: and insulator costs. The power losses are also reduced withDC as there are’ ‘only twoconductors (about 67% of that for AC with same current carrying capacity of conductors). The absence of skin with DC is also beneficial in reducing power losses marginally. The dielectric losses in case of power cables is also very less for DC tansmission. ‘The Corona effects tend to be less significant on DC conductors than for AC and this also leads tothe choice of economic size of conductors with DC transmission. The other factors that influence the line costs are the costs of compensation and terminal equip- ment. DC lines do not require compensation but the terminal juipment costs are increased due to the presence of converters and filters. 4 OSTANCE — @"s break-even distance Fig. 1.1 Variation of costs with lineDC Power Transmission Technology 3 Figure 1.1 shows the variation of costs of transmission with distance for AC and DC transmission. AC tends to be more economical than DC for distances less than ‘break even’ distance and costlier for longer distances. The break even distances can vary from 500 to 800 km in overhead depending on the per unit line costs. 1.2.2, Technical Performance ‘The DC transmission has some positive features which are li sion, These are mainly due to the fast controllability of power in Di verter contro) The following are the advantages: "°° AT Full control over power transmitted ac2- Theability to enhance transient and dynamic stability in associated AC networks. <3, Fast control to limit fault currents in DC lines. This makes it feasible to avoid DC breakers in two terminal DC links. In addition, the DC transmission overcomes some of the problems of AC transmis- sion, These are described below: Stability limits aoa ‘The power transfer in AC lines is dependent on the angle difference between the volt- age phasorsat the two ends. For a given power level, this angle increases with distance. ‘The maximum power transfer is limited by thé considerations of steady state and. t ‘ DISTANCE —= Fig, 1.2 Power transfer capability vs. distance. sient subility. The power carrying capability ofan AC line as. function of distance is _ shown in Fig. 1.2, The same figure also shows the power carrying capability of DC lines which is unaffected by the distance of transmission, Voltage control ‘The voltage controtin AC lines is complicated by the line charging and inductive volt- age drops, The voltage profile ina AC lincis relatively flat only fora fixed level of power transfer corresponding to surge impedance loading (SIL). The voltage profile varies with the line loading. For constant voltage at the line terminals, the midpoint voltage is re-HVDC Power Transmission System 4 duced for ine loadings higher than SIL an isshowninFig.13.. 4 increased for loadings less than SIL. This x= distance from the Letength of the tine Fig, 1.3 Variation of voltage along the line, ‘The maintenance of constant voltages at the two: ends requires reactive power control. from inductive to capacitive as the line loading is increased. The reactive power require- meats increase with the increas¢ in Tine lengths. Although DC converter stations require reactive power related to the line loadings, the line itself does not require reactive power. ‘The steady-state charging currents in AC lines pose serious problems in cables. This Pats the breakeven distance for cable transmission around 40 km, Line Compensation For reasons mentioned earlier, AC lines require shunt and series compensation in ong distance transmission, mainly to overcome the problems of line charging and ‘Stability limitations. Series capacitors and shunt inductors are used for th Purpose. The ‘increase in power transfer and voltage control is also possible through the vf Var Systems (SVS). earner Sane In AC cable transmission, it is Necessary to provide shunt compensati Tegul ‘ is is a serious prdblen oe om intervals, This is a serious problem in underwater cables, ~ ae a 4 Problems of AC interconnection ‘When two power systems: . tio.) the automatic cara ate roUBh AC iss (synchronous itrconnec- generation control of * line power and frequency x é SYster@s have to be cootdinated using tie- oscillations ‘which can lead to went tripping (ii mmission of disturbances from seu ya The controllability of power flow in DC lines elimi addition, for asynchronous DC ties, Itis obvious that two systems interconnected directly with ACDC Power Transmission Technology Ground impedance In AC transmission, the existence of ground (zero sequence) current cannot be Por mitted in steady-state due to high magnitudes of ground impedance which will not only affect efficient power transfer, bt also resultin telephone interference. The ground int ~pedanceis negligible for DC currents and a DC link can operate| using one conductor with ground return (Monopolar operation). The ground return s objectionable only when bur. ‘ed metallic structures (such as pipes) are present and are subject to corrosion with DC current flow. Tris to be noted that even while operating in the monopolar mode, the AC network feeding the DC converter station operates with balanced voltages and currents. Hence, single pole operation of DC transmission systems is possible for extended periods, while in AC transmission, single phase operation (or any unbalanced operation) is not feasible for more than a second. Disadvantages of DC transmission ‘The scope of application of DC transmission is limited by the following factors. 1. The difficulty of breaking DC currents which results in high cost of DC breakers “N2, Inability to use transformers to change voltage levels "High cost of conversion equipment “4. Generation of harmonics which require AC and DC filters, adding to the cost of converter stations oe 5. Complexity of control (Over the years, there have been significant advances in DC technology, which have tried to overcome the disadvantages listed above except for (2). These are : 1, Development of DC breakers 2. Modular construction of thyristor valves Co 3. Increase in the ratings of thyristor cells that make up a valve, 4, Twelve pulse operation of converters S 5. Use of metal oxide, gapless arrestors 6. Application of digital electronics and fiber optics in control of converters ‘Some of the above advances have resulted in improving the reliability and reduction aided costs in DC systems, It can be said without exaggeration that complexity trol does not pose a problem and can actually be used to provide reliable and fast control of power transmission not only under normal conditions butalso under abnormal conditions such as line and converter faults. This has removed the need for DC current interruption in two terminal links, Even for multi-terminal operation, the requirements of current ratings of DC breakers are modest due to effective converter control. 1.2.3 Reliability The “liability of DC transmission systems is quite good and comparable to that of ACssystems. An exhaustive record of existing HVDC links in the world is available from which the reliability statistics can be computed. It must be remembered that the performarice of thyristor valves is much more reliable than mercury arc valves and farther developments in devices, control and protection improve the teliability level. For example, the devel tof direct bi ‘triggered thyristors (LTT)voc Power Transmission Syste, __qnofhigh voliage pulse transform. aes yall and anion ing em clibilit reliability Energy availability et a equivalent outage time A. ast 1 Energy availability = 100( lene valent outage time is the product of the actual outage time and the fraction of where equivalent ou system capacity lost due to outage. Transient reliability % ‘This is a factor specifying the performance of HVDC systems during recordable faults on the associated AC systems ‘ " 100x No. of times HVDC systems performed as designed ~ reget No, of recordable AC faults Recordable AC system faults are those faults which cause one or more AC bus phase voliages wo drop bélow 90% of the voltage prior to the fault. It is assumed that the short circuit level after the fault is not below the minimum specified for satisfactory converter energy availability and transient reliability of existing DC systems with thyris- valves is 95% or more. The average failure rate of thyristors in a valve is less than 0.6% per operating year. Itis common practice to provide redundant thyristors in the series string composing a HVDC valve, so that failed thyristors can be replaced during scheduled maintenance ‘(ence or twice a year). The maintenance of thyristor valves is al i = ee eta isalso much simpler than the Table 1.1 HVDC Outage Statistics ee MITF(years) MTTR(hours) Convener transformer ey 6.1 r 16.1 . bene 1a reno AC filter Oy 79 Master control 93 Pole control ‘ Pole of transmistion line — DC line switch rs MTTF = Mean time to failure os - a uels MTTR = Mean time to repair aa Some of the HVDC outa, /DC outage stisticss given of various alternatives, it must be kept in orn” AOE 1.1. In compari iabilit a double circuit AC line with the eet that isar Dc hae ait ge cane ability. is is because of the fact that > OC Power Tra failure of one Ifthe DC line pole can be p can be maint 13 APPLIC The detail nical perforOC Power Transmission Technology 7 failure of one pole does not affect the operation of the other pole (with ground. return). Ifthe DC line conductor has adequate overload rating and if the converters on the failed Pole can be paralleled with the converters on the healthy pole, the prefault power level can be maintained even with permanent outage of one pole. 13 APPLICATION OF DC TRANSMISSION The detailed comparison of AC and DC transmission in terms of economics and tech- nical performance, leads to the following areas of application for DC transmission: AC Long distance bulk power transmission” 2. Underground or underwater cables. 48 3. Asynchronous interconnetio of AC systems operating at different frequenci¢ or where independent control of systems is desired 4, Control and stabilization of power flowsin AC ties in an integrated power system. ‘The first two applications are dictated primarily by the economic advantages of DC transmission, where the concept of breakeven distance is important. To be realistic, one ‘Tust also assign a monetary value for the technical advantages of DC (or penalty costs for the drawbacks of AC). The problem of evaluation of the economic benefits is further complicated by the various alternatives that may be considered in solving problems of AC transmission - phase shifters, static var systems, series capacitors, single pole switching etc. ‘The technical superiority of DC transmission dictates its use for: ‘asynchronous inter- Connections, even when the transmission distances are negligible. Actually there are many ‘back to back’ DC links in existence where the rectification and inversion are carried out in the same converter station with no DC lines, The advantage of such DC links lies in the reduction of the overall conversion costs and improving the reliability of DC system. The alternative to DC ties may require: strengthening of existing AC network near the boundary of the two systems. This cost can be prohibitive if the capacity of the tie re- quired is moderate compared to the size of the ‘systems interconnected. In large interconnected systems, power flow in AC ties (particularly under distur- bance conditions) can be uncontrolled and lead to overloads and stability problems thus endangering system security. Strategically placed DC lines can overcome this problem ue to the controllability of power. The planning of DC transmission in such applications requires detailed study to evaluate the benefits. Presently the number of DC lines ina power grid is very small compared to the num- ber of AC lines. This indicates that DC transmission is justified only for specific appli- cations. Although advances in technology and introduction of multi-terminal DC {MTDC) systems are expected to increase the scope of application of DC transmission, itis not anticipated that AC grid will be replaced by DC power grid in future, There are two major reasons for this. Firstly, the control and protection of MIDC systems is very ‘complex and the inability of voltage transformation in DC networks i onpmi a Ue saath a networks imposes ecogomic Penalties. Secondly, the advances in DX technology have resulted in the improvement of the performance of AC transmission, through introduction of static var systems, static phase shifters, ete.e beginning. In over 16 y arc valves. The introdycy, 8, Os vDc Power transmission $555 oa, DC Power Tr The rate of growth of scare Bays vey, operation mainly dueyy | - = SOO MW OTE 8 ofthe! opt ae of. ‘owt of DC transmission, ee thyristor valves o\ are backs in mercury ar valves." Sines S065 MIE, qyence OF UNE CONSEUCin ra caansor be i jon scher ca given in Table 1.2. | ySTEM | 1.4 DESCRIPTION OF DC ‘TRANSMISSION § 1.4.1 Types of DC Links F % ? hich are defined below : “The DC Links are classified into three 1YPes ] | Pe 1). GH | (b) BIPOLAR ot stable 12 HVDC systema In serve or under contruction vroughout the werkt (©) HoMOPOLaR Fle. 1.4 DC ling 1, Monopolar link {see “Fe oy and uses ground or @] has one coe etme: mn he aly gue is also used.es (D ouwung-unis “ST ™ a 006 006 mst = = oe un oss 0s osee OL ° on (sn) sang arenbs “LI ouust Oy one 087 w Erm Ty mo (Nad) Z 210d #24n0OURA “ST SUSI — ORTON est vIbl 0 #Iyl —(VZ-OW) onedy - w##eG BOGE “ST 2, soovonpaNsy ust ol oor os ms rs = (vsn) (TNH 'V PIAS “hI USIBI os 00s see ove wa iy (Id) REIS “ET ‘UL souonputsy w6t ose oe 08 = os Z (Nao) 24rd Pd TI samapsls onyea sola “a mst O ow oor ry 3 o (aD) quowstury “11 tut 6991 ozo she 068 ° 068 (Nad) I 2tedig zany vos}aN “OT 061009 091 one ze ° ze (sn) onsacy aupoeg 6 Ones61 ue ue 9c wL fermo Ip Te (ND) 1 ed 42anCOURA -g 1961 oe oz oz ee i t6e. Pero CTT eTUMpIES “L $961 009 009 osee 609 6 ous (ZN) punesz MON 9 S961 OE 006 mst ee - - (pvmmps +5 Be $961 su ose ost ost s8 56 (SxO wASTHOY y g 59-2961 Ooze Oz o0rF ony 0 oly (ns) ssequog-peSoBjon “£, 3 1961 oor or oorF 89 s9 o 89) 1 Papp naz g OLivs61 Of 08 ost 6 6 ° ‘PURTUREY YSIPIMS-PUEPOD *] § sumpsis aus, 20e-Lan -¥ i = = : & frie thee ign Set SRO 2 ban < smeumey myn smonumucs — ruwoN aon, ae é mn Pare ‘sounnsrp worssrussuey), 8 Pl4oM a7 ynowBnoayy woponsysuo> sepun Jo aojases wy sUmISKS CAH TT Oe eee neers aaeiaseeenaicioeica| voc Power Transmission System 10 Ayordeo es mma mmr emmes ONAN ay rl ocean on/se6t 0002 txoF tL uw o G-€9) Z fouumyD #80: “Oy ‘LL snouosyoutsy $861 OOST-00S == OOP 9s 0 9st (SA) 9D 271TEM “6E ‘9LL, snouosypuksy S861 002, a i > oe (sn) MD WN “8E ‘LL snouonpuksy S861 ose ver at - 7 (NGO) PIEMEPEY “LE ‘UL snouompusy oof YN = . a wunding-2021) “Of 9861 os 002. = rz ad @) dey worn “66 Uy, snouanpusy 1961 oss svt = = a (V) 0-PIS tM YE s86t oor cost = = a (sn) onion oye “¢¢ ‘au, snowanpuksy 861 007 @ = a a (sn) ommenryg ze 3S861 o0e9 7X 009% sosiesL =O s08/EsL (aa) edreq te ‘Au, snouonppuKsy 861 001 ovr — 4 a (Nad) Anmiaromn “Of va61 SLi o0eF sosesL =O sos/esL Qi@ edreq “6z ‘AL mouanpatsy ‘e261 oz we = t - (vsn) © Appa “8z €361 S91 oe ost 86 16 L PUNE WHIMS Z PUNPOD Lz SAL, snowonypuksy esor ee oss sve = - ~ (W) mpmmq 92 7961 = 9s 00st oozt ° ont Rava) vars ey “sz ‘AA snouonpaksy ‘7861 out EX Ser -_ at - (pors-ussn) Boas, yz 1861 “4 oor ‘oov/001 90 90 = (vsn) wonns pedwop Midd “£7 2173 09/05 1861 os ot s =o oS (ua-ka) Aemy zz OB/6L61 OE 00s ose ast » a (© nqsw0H- op “Tz 6161 ool 0001 007 016 0 ow (sn(snodeoummy poomsepun) > OZ _ 8L61 0001 006 oset 066 0 086, (ND) Z afodig soary wospN 61 AW Ut 7 Pus a TOL —a1gD_—_praWHOAD rmouminy Amotauey weninns eee AAT womée- SAH oN A porey ‘sounstp votssrusuE BF .opams weg PUETIN WINOS “ED ow ose owe sl St ; = on sor oct ost 86 6 Pum urpang sn shisha ser ee ok = = - Cepay etapa “19 eat ‘casi cd Pog Kpuns-plopoue “09 ‘AL, snououyouAsy L186 os (wunustsy-yq) tunes) “65 1861 0001 Oost 0001 a 0001 (epep mec-pureTa “aS 16-2861 oozt ose 0801 = 0801 (wud) OMD UEN-BqNOGZID “LS oon cost (sm) 1 antong o1peg 95, 6361 oo oz (Aen) z-100VS “ss 24, snowaiyoucsy S36 oz 9s = = F (so) 8u8H “os 2, snounnyucsy S861 oe 9s = = < (sn) semen “Es 4], mounnpatsy. 9861 002. a = = (sn) Aaupis “75 9861 069 Oser Sut su Propouny-suomED soq “1S wor9861 0L02/069 OSbE SLESLI SLEISLI Purydug MINOR “OF Aig conay-uneIn “6 06/s861 081/006 ose oz 0 oz (XaW) w2se0om gy L8I7661 0002 Oost 066 0 086 € atodig 2ARy UOSTON “Ly 3 06-6861 0077/0091 OOSFIV9EF OOF 0 Or (sn) peaw Auoqr] “oy 3 2361 091 cost $6. o OL (sn) ermouromy “sy 3 68-8861 ore ' == OnE ove wt eu (CAM ZTeMBEIS “yy * & 06-6861 ose. 087, ss o¢ st (ID Weg arg “cp 3 88-5861 0009 osee ove 0 ove. Qassn) anvap-seuseang Zy i 68/8861 oz ost oot $8 56 (S¥q) z weas-nUEY “Ip é simau 2a 3 as MW MW o rol, 1D ROYER, é sueuwoy — SumojsTuwoy seonapu00 eee ee SS “ert r wasks-SGAH “ON 8 ~W amy sounsip vorssusuesy, = LL eee “henner ineyy, snovosyautsy Ly, snouaiysuksy ‘1 snowonypusy ‘uy, snouonpuksy ‘ak snouosyuksy 06/6861 1861 8861 18/9861 16-1861 ose ost OL 00st 00st 9s 9s Osbe Osbe Oost Oost OOsFi9 EF Oost 00€ 087 ser ose 0801 SL SLESLI ozs, 06 oy v6L ove ss. oot Purfdugq MaN-aqINd) AND conesy-unegny, (Xa) wosroont + € sjodig 290ry UOSTON (sp) vrmmowrawy (NAD 7 TUPErIS * C1) 28g ams + Qassn) anueg-smseqnyg Ze (SAM Z ASTOR “Ie Fesye sss 9 pour wrpomg-€ pumOD 29 9 (SA) Aoupis ZS s DC Power Transmission Technology swounue3, soa yo ONE AM aieyo,, ‘soURSIp UotsstusuEL] wasks-3GAH ON, Ssvoc Power Transmission System positive and the other nega. ‘ch terminal has two sets of The junction betw tive, Bach dential ratings, mseries OMT ends rep, bot see two sets of converters is grounded at one or both et al rent flowing under aoe crate at equal currents and hence there is zero groun' ese conditions. 3, Homopolar Link [see Fig. 1.4 (¢)] hast polarity (usually negative) and always Because of the desirability of ‘operating@ DClink without ground return, bipolar links Tink has theadvantageof reduced insulation costs, are mostcommonly used. Homopolar li CE ently, the but the disadvat ofearth return outweigh the 2 effects in aD Tinea ially less with negative polarit of the c ‘pared 10 the positive polarit zs ; ‘The monopolar operation is usedin the firststage of the development of abipolarline, as the investments on converters can be: deferréd until the. growth of load which requires bipolar operation at double the capacity of a monopolar Link. 2 conductors: one i 10 2, Bipolar link [seeFig. 14 ()]haS WO Ea fines. tive. Each may bea double co! juctor in E] bia wo or more conductors all having the same erated with ground oF metallic return, 1.4.2 Converter Station ‘The major components of a HVDC transmission system are converter stations where a Fale 1 CO—/s ® CS) Z E he é f {it | Co x a Lap—z © : aha ——_ Poe? 12 Pulse converter Transformer Smoothing reactors, OC Fitters, Tired AC Fitters HP AC Fitters Fig. 15 Schematic agram of a pica, VDC coy SHEE sation, DC Pow conver; tifier ar verter. At 15.Tt Conve al2p are tvDC Power Transmission Technology 1B conversi ‘ i i performed arom AC to DC Rectifier station) and from DC to AC (Inverter station) are ; 4. A pointto point ransmission requis tifier and inverter station ission requires twoconverter stations. The role of rec- verter control. Atypical converter station with t, i i inFi ; two 12 pulse converter units per pole, is shownin Fig. 1.5. The various components ofa converter station are dseteeed pelea: ws 'scan be reversed (resulting in power reversals) by suitable con- Converter unit This usually consists of two three pha: i in seri s Phase converter bridges connected in series to form a 1 pulse ee unit as shown in Fig, The total number of valves in such a unit are ve. The valves can‘be packaged as single valve, double valve or quadrivalve 1-Single valve Quadrivaive AC Bus Fig. 1.6 A Twelve pulse converter unit arrangements. Each valve is used to switch in asegmentof an AC voltage waveform. The converter is fed by converter transformers connected in star/star and star/delta arrange- ments. The valves are cooled by air, oil, water or freon. Liquid cooling using deionized water ismore efficient and results in the reduction of station losses. The ratingsof a valve group are limited more by the permissible short circuit currents than steady state load require- ments. The design of valves is based on the modular concept where each module con- tains a limited number of series connected thyristor levels, Valve firing signals are generated in the converter control at ground potential and are transmitted to each thyristor in the valve through a fiber optic light guide system, The ight signal received at the thyristor level is converted isan elect Sigal using ignal using gate drive amplifiers with pulse transformers, - 2 The valves are protected using snubber circuits, protective fring and gapless surge arresters, Some of the details of the control and protection of thyristor valves are givenvoc Power Transmission Syste 14 ions - (i) three phase, two Converter Transformer er can have different configurations a ae iding CO sing pase re winding, (i SDB Pmt ungrounded. ‘On the AC side, ane ooeoeoel Tr erarand delta with neutral POT The leakage reactance i | ground formers: i el with neutra! any waist onthe Se ‘init the short circuit currents pros ao ie ofthe transformes i emer are designed 10 wihsiand DC voltag sand meters Saas ici 1s, One problem that can arise is due creased eddy current losses due (0 to the DC magnetization of the core due to, ney Inback toback links, which: aredesigned for! LN Se ie cee configuration can result in identical transforms Sn a lovever, be pesfore. units, This results in the reduction of the spare capacity’ ance of extended delta transformers in practice is still to be tested. ‘DC voliage levels, an extended delta Filters ‘There are three types of filters used : 1. AC filters : These are passive circuits used to provide low impedance, shunt paths for AC harmonic currents. Both tuned and damped filter arrangements are used. 2. DC filters : These are similar to AC filters and are used for the filtering of DC har- monics. 3. High frequency (RF/PLC) filters: These are connected between the converter transformer and the station AC bus to suppress any high frequency currents. Sometimes such filters are provided on high-voltage DC bus connected between the DC filter and DC line and also on the neutral side. _ Reactive power source Converter stations require reactive power suppl is r e ly that is dependent on the active Power loading (about 50 0 60% of the active power). Fortunately, part of this reactive power requirement is povided by ac filters. In addition, shunt (switched) capacitors. synchronous condensors ani F f syncteonees os Sialic var systems are used depending on the speed of Smoothing reactor A sufficiently large series reactor is us 5 i ised on DC si for protection. The eactoris design das ilivarreseere josmodt DC current and also neutral side or at intermediate location, ndisconnected on the line side, DC switchgear This is usually a modified AC equi ¥ ployed as disconnecting switches), De eased to interrupt small DC currents (em- (MRTB) are used, if required for interruption of el Metallic return transfer breakers In addition o the equipment described above nee load currents, ‘meat for protection and measurement are also pect Switchgear and- DC Power Transmission Technology 1.5 PLANNING FOR HVDC TRANSMISSION ‘The system planner must consider DC alternative in transmission expansion. The fac- tors to be considered are (i) cost, (ii) technical performance, and (iii) reliability. Generally, the last two factors are considered as constraints to be met and the minimum cost option is selected among various alternatives that meet the specifications on technical performance and reliability. For submarine, cable transmission and interconnecting two systems of different nominal frequencies, the choice of DC is obvious. In other cases, the choice is tobe based. on detailed techno-economic comparison. The considerations in the planning for DC depends on the application. Two applica- tions can be considered as representative, ‘These are: 1, Long distance bulk power transmission 2. Interconnection between two adjacent systems In the first application, the DC and AC alternatives for the same level of system se- curity and reliability are likely to have the same power carrying capability. Thus the cost comparisons would form the basis for the selection of the DC (or AC) alternative, if the requirements regarding technical performance are not critical. ATW IITUTE OF Le] Zann 8 Tey AY % (a) TWO TERMINAL OC UNK f(* Ave. we 19576 ; Ss 4, "4 a LERARy, WARANOM-O tb) BACK TO BACK DC LINK ALONG WITH AC FEEDER (e) BACK TO BACK DC LINK AT BORDER Fig 1.7 Different configurations for asynchronous interconnection.: _ 16 In the second appli¢ation, cases. For the same level of system a fr interconnection will be smych ore ee i of DC power modulation). Thus following considerations. 1. Small fluctuations inthe vollnge can be set at any desired val 2, The system security ca Having settled on the DC link for interco! rations for interconnection. These are: 1. A two terminal transmission somewhere within the network an 2. Abackto back HVDC station (a ‘where within one of the network ‘mon station. 3, A back to back station local isa special case of the above. ‘These are illustrated in fig. 1.7. ” es ‘ tion AAC imeroonnection Fo ignot Oe ignettion will be based on the intercom and frequency 4070 a valmanced by fast contzol OF DC power. where each termins Iso called HVDC couplin: ‘andan AC line from the: HVD! several .c Power Transmission System problems in certain Pee he equired capacity of AC afte nnection, there are 1d connected by the beneficial aspects ct the power flow which three possible configu- al is located at a suitable place a DC overhead line or cable. ig station) located some- other network to the com- ted close to the border between the two systems. This Inthe choice between the firstand second configuration, itis ‘to be noted that converter ‘costs are less for the common coupling station and the AC ine costs are greater than the DC line costs. If the distances involved are less than 200 km, the second configuration is to be preferred. If the short circuit ratio (SCR) is acceptable, then the third alternative -will bé the most economic. _ ‘The specifications and design of DC system require an understanding of the various interactions between the DC and AC systems. The interruption (or reduction) of power ~ es link a ‘oceur due to (i) DC line faults (ii) AC system faults. speed of recovery from transient DC lines faults is of concer in maintaini = aintaining the daveily te oro agen me a tom and stability studies are used in dis . The recov 1 m AC system faults is. more complex. The \eatewal Me Teo ate ane bus can lead to commutation failure and Joss of inDC poveroupatcan adie edoston of AC veg: a aS (due to corresponding i iv oa asnoeee Voltage and failure of commutation Power can be determined from stability study, This system characteristics. Dynamic overvoltages ly of the s; Harmonic generation and design of filters 2. 3. 4, Damping of low frequency , y and 5. Carrier frequency interference caus ibsynchy of conduction) due othe discharge of, euY urTents The converter control plays a major role in should be such as to improve the overal “ HVDC simulators are used for planning : these in System perf \ Performance, and design studies, ). An optimum rate of increase in DC sis influenced by control strategy and The following aspects also require a detailed stud Var requirements of converter stations ystem interactions. onoUs torsional oscillations Fen ut valves (at the beginnii Stray inning Y capacitances and athe esis teractions and d the control strategy + Digital simulation and DC Pow Choice For the tot cost of i we eesDC Power Transmission Technology Choice of voltage level 17 For long distance bulk power transmission, the voltage level is chosen to minimize the total costs for a given power level (P), The total costs include investment (C,) and ost of losses (C,). The investment costs per unit length are modelled as C\=Ay+ A, nV +A, ng where Vis the voltage level with respect to ground nis the number of conductors is the total cross-section of each conductor Ay. A, and A, are constants. The cost of losses per unit length is given by C,= [n(PinV¥ pTLp\iq where = conductor resistivity T= total operation time in a year L= loss load factor ‘p= Cost per unit energy C, can be simplified as C= (A, (P/VpVing optimum system voltage timum line voltage line cost LINE COSTS + (minimutn separation equals minimum cost total } CONVERTER COSTS. Fig, 1.8 Selection of Optiroum System Voltage for A Fixed Power Transfer, (ay (1.2) (1.3)“* water cooling has now become a standard voc Power Transmission System 18 pir erm in Cy W° have, By minimizing the sum of C, and the Ih! a) Pv) B (15) nq = /RIAs?) , J= PKnqV) & pi sing (1.5) can be written as is nt density. The 101 P where J is the curret fnnol8) (1.6) ion (1.6) ignores the variation ection of optimum system 2, C=C,+C, = At Ain * zo C. The equall 1.8 shows the sel Osis. , ‘psent. Hence the voltage level is ily much lower than that in the i nimi ‘The voltage level Vis chosen to minit of terminal costs with the voltage. Figure ater voltage to minimize the sum of converter an‘ x In case of back to back DC ties, the line costs are ‘chosen to minimize converter costs. This level is genera presence of an overhead line. 1L6 MODERN TRENDS IN DC TRANSMISSION : ‘The continuing technological developments in the areas of power semiconductor devices, digital electronics, adaptive control, DC protection equipment have increased the pace of application of DC transmission. The major contribution of these develop- ‘meats is to reduce the cost of converter stations while improving the reliability and per- formance. \POrer semiconductors and valves ‘The cost of the converters can come down if the number of devices to be connected in series and parallel can be brought down. The size of the devices has gone up to 100 mm (in diameters) and there is no need for parallel connection. The increase in the _Surteat rating of the devices has made it possible to provide higher overload capabi a reasonable coss ad reduce the lower limis on transformer {eakage-ietodanee thereby improving the power factor. The voltage ratings are also on the increase. The Jight_ tis thyristors should also. imy rove th iability of Converter operation. The cost of the valves is also reduc i ficaien te mn ed by icati i ‘oxide gapless amresters and protective firing methods. ee ‘The power rating of thyristors is increased hase flow using forced vaporization is also being inves ooo (Osses in Cooling. Two -In this davies oS iMeBrated circuit created on switched off by asmall gate current, The ‘urn-off time of very large line cure nace that of GTOs. However, MCTs are arnt Mri t ‘The cost of silicon used in the manufin te ary stage ‘ea es than one hid brought down (by 15 10 20 per con ire of 1) from i wer semiconductor devi emp eee ateDC Power Transmission Technology 8 instead of the conventional FZ (float zone) method. Research is also underway in reducing this packaging cost of a device. wnverter Control The development of micro-computer based converter control equipment has now matie it possible to design systems with completelyyfedundantkonverter conyyol with automatic transfer between sysioms in the case of a malfunction. Not only is the forced ‘outage rate of control equipment reduced but it is also possible to perform scheduled pieventive maintenance on the stand-by system when the converter is in operation. The use of a mini-simulator will make it feasible to check vital control and protection functions. ‘The micro-computer based control also has the flexibility to try adaptive control algorithms or even the use of expert. systems for fault diagnosis and protection. DC breakers With the development and testing of prototype DC breakers, it will be possible to go in for tapping an existing DC link or the development of new MTDC systems. Parallel, rather than series operation of converters is likely as it allows certain flexibility in the planned growth of a system. The DC breaker ratings are not likely to exceed the full load ratings as the control intervention is expected to limit the fault current. ‘The control and protection of MTDC systems is not a straightforward extension of that used in the two-terminal DC systems. The possibility of decentralized control necessitated by communication failure, the coordination of control and protection are some of the issues currently being studied. Conversion of existing AC lines ‘Theconstraints on RoW are forcing some utilities olook into the option of converting existing AC circuits to DC in order to increase the power'transfer limit. There could be ‘some operational problems due to electromagnetic induction from AC circuits operating in the same RoW,” ‘Anexperimental project of converting a single circuit of a double circuit 220KV line is currently under commissioning stage in India. Operation with weak AC systems 4 ‘The strength of AC systems connected to the terminals of a DC link is measured in terms of short circuit ratio (SCR) which is defined as SCR = Short circuit level at the converter bus Rated DC power ~ If SCR is less than 3, the AC system is said to be weak, The conventional constant ‘extinction angle control may not be satisfactory with weak AC system. The recovery of inverters following the clearing of fault in the connected “AC system can also be problematic, ‘Constant reactive current control or AC voltage conwrol have been Suggested to Overcome some of the problems of weak AC systems. The use of fast reactive power20 HVDC Power Transmission System control at the converter bus by applying s! Limiting dynamic overvoltages through col becoming a standard practice. ee The power modulation techniques used to improve dynamic stability of power systems will have to be modified in the presence ‘of weak AC systems. Coordinated reactive and active power modulation has been suggested to overcome the problems of voltage variations that can limit the effectiveness of power modulation. tatic var systems is another alternative. nverter control during load rejection is References and Bibliography {1} C. Adamson and N.G. Hingoran, High Voltage Direct Current Power Transmission (London) Garraway, 1960. [2] EW. Kimbark, Direct Current Transmission, Vol. I (New York) John Wiley, 1971. [3] E Unimann, Power Transmission by Direct Curren, (Berlin) Springer-Verlag, 1975. [4] J. Anillaga, High Voltage Direct Current Transmission, (London) Peter Peregrinus, 1983. [5] VS. Lava, ‘An Annotated Bibliography of High Voltage Direct Current Transmission’, Portland, Oregon : Bonneville Power Administration, 1966-68, 1969-1983. {6} BJ. Cory, “High Voltage Direct Current Power Converters and Systems, London: McDonald, 1965. 17) RJ. Meredith et al, “Methodology for Integration of HVDC Links in Large AC Systems - Phase I : Referee Mana al Repo 1-300, RP 196-1, Hectic Power Resch Ine, March 18) ees al. ‘AC or DC? One Utilities Approach’, IEE Conf. Publication No. 205, December {91 L.Carisson, Recent Developments in HVDC Converter Station Design’, IEEE Trans. Vol. PAS-1"? No. 8, pp 2166-71, August 1984, Saae eee2. Thyristor Valve 2.1 INTRODUCTION HVDC converters are an assembly of valves which have the property of conducting in the forward direction and blocking in the reverse direction. The term ‘valve’, carried ‘over from the mercury arc valve days, is applied even now for thyristor valves which are made up of series and parallel connection of many thyristor cells or devices. ‘Themajor problem with the mercury arc valves is the occurrence of arc backs (or back fire) which results in the destruction of the rectifying property of the valves. Arc backs are random phenomena which result in failure to block in the reverse direction. Although the incidence of arc backs can be reduced by carefully controlling the factors that influence them, complete elimination is impossible and the valve cost is also increased. Furthermore, arc backs are non self-clearing and result in line to line faults which stress transformer windings and anodes in the valve. The maintenance requirements for the valves go up and lead to poor reliability. Thyristor valves which were developed in the late sixties have eliminated all these problems. They have now completely displaced mercury arc valves in HVDC transmission. Thyristors that constitute the valves are also not perfect devices. The major: ‘problem is that their ratings cannot be exceeded even for short durations. However, there is continuing development in the field of power semiconductors which has brought down the cost while improving the reliability. This chapter reviews the principles of operation, characteristics and control of thyristor devices. Some of the design aspects, protection and testing of thyristor valves for DC transmission applications are also presented. 2.2 THYRISTOR DEVICE 2.2.1 Description Thyristor is now defined as a generic term applicable to the whole range of four layer (PNPN) semiconductor switches, It is also known commercially as silicon controlled CATHODE, GATE ANODE Fig. 21. Symbol and structure of a thyristor‘HVDC Power Transmission System z it nals and its electrical i a thyristor with the three terminals and 1s @ ST seer The aioe can carry current only in one direction from symbol are shown in Fig. 21 The do duction can be controlled by Se SA to withstand, when turned off) is now in i fi anode tocathode and the instant 0 ‘The voltage rating ofa thyristor: (the ability Aa i oe pao ‘cessity of parallel connection, the hile thecurrentratinghas gone’ uy ing is insufficient to make upahigh voltage valve. Thus, series connec! voltage rang i ryaihyrisiorvalveis necessary and introduces some problems that have pepper Pe design of valves and protection. | ; Increasing the voltage rating of a thyristor is feasible, butitis atthe cost of increased ¢ losses, turn-off times: ‘and reduced peak allowable junction temperatures. The capability ofathyristorto withstand high voltages critically depends on the quality ofsiliconcrystal from which the device is made - the more uniform the crystal, the better the voltage veidhstand capability. The irradiation of silicon crystal using low energy neutron or gamma rays results in light, precisely controlled doping. 2.2.2 Principle of Operation ‘The principle of operation of thyristors can be explained by the two transistoranalogy shown in Fig, 2.2 Here a thyristor is replaced by a PNP and a NPN transistor connected +'s 6 t lea Jp Fig. 22 Two transistor analogy for a thyristor. in regenerative feedback. If the gate curent /, is injected i base of transistor 7 i Aft is injected into the i its collector current J., amplifies the collector current To, of see This in turn ms ores gate current Evenualy, 7, and T, go into complete saturation anid all ‘The equations for the two transistors are : Tey= Oy Ly + Legg, 1642 Oy 1g + Hea TenNt Ig Qa) @2) @3) Thyriste where. base leThyristor Valve 2B where a, and 0, are the common-base current gains and jg, aNd I-p9, are the common- base leakage currents of T, and T,, respectively. From (2.1) to (2.3), we have = Selo +t lox * coo (2.4) 2 info) A silicon transistor has the property that otis very low at low emitter current and rises rapidly as the emitter current builds up. When the device is off, /, = 0, and /, will be the leakage current. If it is possible to raise the emitter currents of T, and T,, such that (0, +0) al es unity, then the device triggers into saturation. There are several means of achieving this : 1. Injection of gate curent (normal turn-on) 2. By increasing the forward voltage above a limit, V,, called break-over voltage. In this case, the minority-carrier leakage current at middle junction increases due to avalanche effect. 3. By increasing the anode voltage atarate such that the depletion layer capacitance at the middle junction will create a displacement current (dv/dt turn-on). 4, Ata high enough junction temperature, the leakage current increases and causes atum-on, 5. Direct irradiation of light on silicon creates electron-hole pairs, which under the influence of electric field result in a current to trigger the thyristor. Triggering the device into saturation is called tum-on.Controlled turn-on without damaging the device is only feasible through gated turn-on. The device remains in a conducting state until the current is maintained by the circuit action, above the holding current. During this period, the gate has no control on the conduction. The turn-off process which results in the deivce regaining its blocking state is achieved either by : (i) line commutation or (ii) forced commutation. In both cases, the circuit voltage source is reversed which in turn will drive the current tozero. After a time lapse off, the turn-off time, the voltage can be reversed again, when the device regains its blocking state. 2.2.3 Device Characteristics ‘The device can be in one of the three following states : (i) Forward biased and blocking Forward biased and conducting ii) Reverse biased and blocking. ‘The transition from the first to the second state is called turn-on, while the transition from the second to the third state is called turn-off. The characteristics of the device refer to the parameters of the device both in steady-state and transient conditions (during the transition of state). Steady state characteristics Off-State : The volt-ampere characteristics of the device are shown inFig.. .3. During the off-state (both forward and reverse blocking), only a small magnitude of leakageHVDC Power Transmission System bi ‘ he order of 100 mA). The blocking capability with gate open is current flows (of # (Visguy) OF reverse (Vea) Voltages. specified in terms of limiting repetitive peak forward ( fi Fig. 23 V-I characteristics of a thyristor ‘There is also a non-repetitive peak reverse voltage rating (V,,4) which is specified. The voltage ratings are specified for power frequency (50 or 60 Hz) half-cycle sinusoidal voltages and rated junction temperature of the thyristor (typically 125°C). The variation of the voltage ratings with junction temperature is shown in Fig. 2.4. slopes01% /*C Fig. 24 Typical thyristor off-state voltage characteristic as a function of junction temperature ‘The behaviour of thyrii i i sig i ee 'yristors under transient voltages is not well understood. However, Particular study (10, 11), the following conclusions can be drawn: 1, The transi a ree ‘breakover voltage of a thyristor is independent of its voltage rating. forward breakover voltage of a thyristor under a transient voltage may be charge and inherent time delay i < y Of the thyristor, an reverse breakover voltage levels of a thyristor under sloy i (30/600 11s) can be significantly lower compared to those tained 4 ‘fast wansient voltages (1.2/50 us). This behaviour has been attri < ‘Statistical and formative time lags of avalanche formulation, fee eee: Thyristor Valve The thyrist losses which i the voltage m: of a thyristor Cumulative « degradation 0 On-State: the on-state tstem en is ages, sERe Thyristor Valve 25 ‘The thyristor capability in the reverse direction is related to the permissible energy losses which in turn is dependent on the variation of the reverse avalanche current with the voltage magnitude of the reverse voltage, The transient voltage blocking capability of a thyristor is perhaps related to the critical power needed to damage the device. Cumulative effects due to transients with less than critical Power may result in degradation of the device, On-State: There are a number of electrical and thermal parameters that characterize the on-state behaviour. Some of these are as follows: 1, On-state voltage 2, Mean (average) on-state current /,4, 3. Root mean square value of the on-state current ny, 4. Surge (non-repetitive) on-state current osu) 5. Non-repetitive survival rating (fi Br? ) 6. Holding curent (/,) 7. Operating temperature range 8. Junction to case thermal resistance ruse) 9. Contact thermal resistance The on-state voltage is the anode to cathode voltage of a thyristor in the forward conducting state, It is also referred to as. the forward voltage drop. This is an ‘important characteristic affecting the power losses during on-state and the parallel ‘operation of On-state voltage depends upon a number of factors such as the width of various reeions, life time and mobility of minority carriers, the physical mechenisere ot ‘combination, etc. The power loss at current densities around 100A/em? (conditions MAX. Tj (BLOCKING) MAX. PREFAULT Tj TRANSIENT OVERLOAD THYRISTOR (STEADY STATE) COOLING PLANT MAX. AMBIENT TEMP. Fig. 2 Temperature build-up in a thyrision MAX, STEADY STATE T] THYRISTOR CASE TEMP. COOLANT TEMP,—— ; HVDC Power Transmission System 6 comes nuous rating) is ding to maximum continuous ‘ core spejensites around 1000A/em? (Sure condi ae, afers of smal thickness and long carrier life times give rise to low on-state Silicon w id Long cs i -off time. Trying increasing the carrier life time also increases turn-off time. Sebo maf ut in igh reverse recovery cure which must be ings are determined by the junction temperature On-statecurent ratings fo ensure tha it can block the recovery voltage aera ae vale overourrent. Fig. 2.5 shows the temperature build-up ina thyristor wor ce rahi llowing an overcurrent is essentially ahigh tempers: ra Sage failure produced by intense local heating induced by excessive leakage ean ability ofa thyristor is based on its filamentation temperature auwhigh mesoplasmasare formed. However, the requirement to have surge suppression capability (voltage blocking following the surge current) will result in the operation of the thyristor at reduced junction temperature, The maximum junction temperature attained due to the cumulative effects caused by the passage of repeated current surges ‘should be below the thyristor filamentation temperature. ‘The holding current, /,, is defined as the minimum current required to maintain the thyristor in the on-state, It is the forward current below which it will cease to conduct. As the forward current is reduced, the turn-off occurs when recombination causes the due to recombinations, whereas at tions), it is mainly due to ohmic __ minority carrier level to fall below the base-region doping level. The holding current reduces with increase in junction temperature. Switching characteristics Turn-on : When a gate drive is applied with the forward voltage above latching voltage (the minimum anode to cathode voltage that will successfully turn-on a thyristor with a given gate drive), tum-on occurs. Because of the finite sheet resistance of the p- base region, only those regions of the cathode nearest to the gate are influenced by the gate current. Regenerative switching action is initially restricted to these regions. The establishment of the equilibrium current flow over the cathode area follows by outward Spreading from this conducting plasma, by diffusion. The plasma spreading is relatively slow and occurs with a typical velocit 'y of 0.1 mm/s. When the area of conduction ii small, the voltage across the device is consi i ing at Goa ieacaer in device is considerable andexcessive local heating will occur é igh. The device has an upper limit a ‘modem amplifying gate thyristors is upto 500 A/us, cane ice oe, difds, Meet os 1 DELAY ‘a A 1 RSE ae 111 SPREADING ee VU" Time Fig. 2.6 Tum-on characteristics Thyristor Ve thyristors 4 to stray cal ‘Thereai the establi depends uy time. The. spread tim using on-s seconds. Ungat Thisis to on by app usually a Turn- regions ¢ out by ar Wher thecurre junction peak val anode ju with the manner current proces: B29 _jan System, hereas at 0 ohmic on-state Trying must be e after a thyristor empera- leakage perature mression ation of erature t surges tain the onduct. ses the current, tching yristor the p- by the s. The tward tively ion is occur ich in th the Thyristor Valve 27 thyristors are used to limit di/dt, particularly arising from the discharge of current due to stray capacitances and snubber circuits. ‘There are three phases of turn-on shown in Fig. 2.6. The delay time is associated with the establishment of regenerative action in response to the gate current. Its duration depends upon the level of the gate drive. Regeneration is well established during the rise time. The current continues to increase during the spread-time. The losses during the Spread time can result in increase of 10-20% of that predicted for normal conduction using on-state voltage of the thyristor. The spreading phase may last over hundred micro- seconds. Ungated turn-on can occur due to overvoltage, dv/dtor incomplete forward recovery. This is to be avoided because of the damage it can cause to the device. Protective tum- on by applying a gate pulse whenever the possibility of ungated turn-on is detected, is usually adopted. Turn-off : All the three junctions are forward biased during on-state and the base regions contain excess minority and majority charge. This charge must either be swept out by an electric field or decay through regenerative processes within the silicon. When the circuit voltage is reversed, the current falls to zero at a certain rate. Once the current reaches zero, the flow reverses, since the minority carrier concentration at the junctions can support this current by diffusion without build-up of depletion layer. The peak value of this reverse current is reached when the excess hole concentration at the anode junction has fallen to zero. At this time, the voltage across the thyristor reverses with the development of the depletion layer and the current decays in a near exponential manner as a result of charge recombination within the n-base region. The decay of current is dependent on the mean life-time of carriers in the n-base region. The tum-off process is shown in Fig. 2.7. Fig, 2.7 Tum-off characteristics Immediately after current zero, a thyristor is unable to support forward voltage, Gradually, the thyristor acquires some forward blocking capability but its ability to withstand forward dv/dt is severely limited. Off-state dv/de capability is attained only after a millisecond or so has elapsed from current zero. This characteristic is circuit and28 Current ! i SHBY ‘Spontaneous furn=on Fig, 2.8 Current and voltage waveforms during positive recovery temperature dependent. Fig. 2.8 also shows the current and voltage waveforms for a recovery and spontaneous tum-on for a fixed dv/di transient: 22.4 Gate Drive ‘harddrive wit sharp rise time is necessary to turn-on the device quickly with large initial dé. A single short pulse is adequate to turn-on the device, provided the current in he device does not go below the holding current during the conduction period (about Pulses (or a pulse train lasting the required Contection period) are applied to avoid the blocking ofthe device due to discontinuous 120°) of the device. Sometimes, long — Fig. 29 Gae characteristics The gate v-i characteristics ae Points is bounded by twolines Anan TFB. 2.9. The locus gate voltage and current. The ee are als The thyristors in a modem the cireuitsatthe groundpowe OC potential. These sj it i see Tlic etn >Y @ Power, Thyristor V which is c ‘The diagr gate pulse ‘transmitteora Thyristor Valve 29 which is charged from the forward voltage across the device when it is not conducting. The diagram of typical module with one thyristor per module is shown in Fig. 2.10. The Bate pulses are generated in the gating and logic units in Tesponse to optical signals, transmitted via fibre-optic light guides separately to each module. The gating and OPTICAL TRIGGER L,-SATURABLE REACTOR Co+Rp- AC VOLTAGE GRADING AND DAMPING CIRCUIT Roc - DC VOLTAGE GRADING RESISTOR Cy-Ry- HIGH FREQUENCY VOLTAGE GRADING 1 = THYRISTOR VOLTAGE MONITOR POWER SUPPLY UNIT BREAK OVER DIODE (80D) GATING AND LOGIC UNIT GATING STATUS LED THYRISTOR STATUS LED ig. 2.10 Typical valve module (Source : Reference 22) thyristor status are also monitored at the ground level from the signals sent via another Set of light guides. The breakover diode (BOD) is used to sense overvoltage across the device and protect it by generating a gating pulse, BOD firing is also used as a back-up if the gating and logic unit fails, P pei Ales AC voltage grading is provided by resistor R, and capacitor, in series. This circuit also damps the oscillations produced by step changes in the voltage, which occur during Commutation of current from one valve to the next. Italso limits the rate of rise of voltage on a late firing module.30 HVDC Power Transmission System i i i this circuit is also used for ‘The resistor R,,. provides for DC voltage grading and ) 4 voltage measurement. ‘The saturable reactor is used to protect the device against: difdt. C, and R,, are used as grading circuits for high frequency voltages. 2.3 THYRISTOR VALVE 2.3.1 General A thyristor valve is made up of a number of devices connected in series to provide the Sciedvaigs rating and aa of devices connected in parallel to provide the required current rating. With modern devices of high current ratings, the need for Parallel connection does not arise. The number of series connected thyristors is determined by device ratings, transient overvollages and protection philosophy. ‘The valves are usually placed indoor in a valve hall which is protected against contamination and dust. Valves are generally base mounted’ in a single, double or quadrivalve configuration. The latter results in a compact valve hall. “The valvesare usually air insulated and cooled using air, water, oil or freon. The water cooling is normally used in modern converter stations as the power losses are reduced. Inavalve, the heat sinks and damping resistors are cooled by the water flowing in ducts. ‘The main objective of the cooling system is to reduce the total thermal resistance for the heat sink. 2.3.2 Valve Firing ‘The basic valve firing scheme is shown in Fig. 2.11. The valvecontrol, i 5 e i u 2.11. generates firing signals. Each thyristor level receives the signal directly from a separate fibre-optic cable. STOR RONICS, THY RISI ELECT! LReL up IGHT RECEIVER a LE -LIGHT EMITTER Hones EQUPMENT Fig. 2.11 Valve firing i ure : Reference 22)Thyristor Valve 31 Thus each thyristor level is independent, sharing only a duplicated light source at the ground potential. The valve control unit also includes many monitoring and protective functions. The retum pulse system coupled with short pulse firing scheme is used in present day valve control units. A separate light guide is used to send a return pulse whenever the voltage across. thyristor is sufficient and the power supply unit is charged. If at that time, firing pulses are demanded from the valve control, the light signals are sent to all the thyristor control units simultaneously. During normal operation, only one set of light pulses are generated in acycle foreach valve, However, during operation at low direct currents, many light pulses are generated due to discontinuous current. 4 2.3.3. Valve Design Considerations The design of the valve must consider the voltage and current stresses that occur during normal and abnormal conditions. The overvoltages across a valve may be generated internally during switching action or the result of external causes such as short circuits on AC and DC systems. The fundamental frequency dynamic overvoltage is crucial in determining the voltage ratings and this arises when the load is thrown off due to blocking of converters. The magnitude of the dynamic overvoltage is dependent on the AC system strength. ‘The overcurrents in a valve arise from short circuits across a valve or a converter bridge. The surge current rating of a valve is limited both by transformer leakage and system impedances. The transformer leakage impedances have been reduced to 0.12 pu with higher rated devices. ‘The overload rating of a valve isa function of the size of the device as well asambient temperature and the cooling system. For example, it is possible to increase the direct current by 20% for a decrease in ambient temperature by 8°C, with water cooling. ‘The losses in a valve include (i) the losses during on-state and switching losses, (ii) damper and grading circuit losses and (iii) losses due to auxiliary power required for cooling. The resistances of the damper and grading circuits are also watercooled. Their " losses are functions of converter delay and overlap angles. The calculation of internally generated overvoltages and the design of valve compo- nents such as damper circuits is facilitated by digital computer programmes which can —< tps) valve last thyristor fired fired Fig. 2.12 Voliage on last firing thycistor.ser Transmission 3ysiem 32 HVDC Pows | 5a For | ‘ stor devices in a valve. j Thyrist model the non-coherent turn-on and tum-off of the efit ofindividual levels ; example, the non-coherent turn-on ‘ofa valve due to prot | SCRe j comm ; Th ci |, opera ta, j due to | high reacte by pu | Th essen 23.4 nT over strin ager Fig. 2:13 Effect of mismatch of reverse recovery charges Goring tum-off forn cap con resultsina voltage waveform across the last level to be fired, as shown in Fig. 2.12. The : {ifferenoes in the stored charges result in varying commutation margins in individual Gevices in an inverter valve. This is shown in Fig. 2.13 which also shows that the device ‘with the largest stored charge will experience a premature voltage zero crossing which ‘can be even hundreds of microseconds in advance of the valve voltage which is usually used as the reference for inverter extinction angle control, ‘The recovery voliage across the valve that has turned-off shows an overshoot that is ‘due to the charging of the stray capacitance through the source inductance (see Fig. 2.14). e le J ji \ Rp 1 Cp Fig. 2.14 Simplified equivalent circuit during tum-off The overshoot is controlled by the i The 0 s Proper choice of dam capacitances. This overvoltage is maximum for the vicinit Tpertncionnshonckciraon SCR ends ort emai oe com nal Ive, poner this does not necessarily reduce the stess ona vilee ne ne device occur due to overvoltage (following the clearing of the Gain, en per circuit resistance andThyristor Valve 33 ‘SCR can also result in nonsinusoidal voltage at the converter bus which can give rise to commutation failures, ‘The valve can be subjected to high stress commutation resulting from high di/dt when operating at low bridge currents. This is due to the premature currentextinction ofa valve due to discharging of stray capacitance. The discontinuous conduction can also result in high voltages across a valve due to the. lightly damped oscillations caused by smoothing Teactor exchanging energy with the line capacitance, Discontinuous operation isavoided by putting a minimum limit on the bridge direct current. ‘The control of electrostatic and electromagnetic fields surrounding a valve is essential to avoid corona discharges and interference with sensitive electronic circuits. 2.3.4 Grading and Damper Circuit Design ‘The intemal overvoltage arises from turn-on and tum-off Process. The turn-on overvoltage results from dispersion in the turn-on times of various thyristors in the series ‘string and can be controlled by suitably designed grading circuit. The turn-off overvolt- age results from two causes (i) the dispersion inthe reverse recovery charges of thyristors forming the series string, (ii) the voltage oscillations in the circuit Consisting of stray capacitance (across the valve) and the transformer. leakage inductance. The latter can be controlled by a suitably designed damper circuit, The voltage across a valve immediately following its extinction is given by v= v4 €°1" (cos, t~ (6,/a,) sin w,t) +14 €*2' (cosw,t -(0,/0,) sin o,t) ~ 1] 2s) @..=1/. SLC = 0775 O, V,, = jump at extinction in the sinusoidal voltage across the valve Ry, Cp are the parameters of the damper circuit, The damper circuit design is also affected by the losses in the damper as these are ttt a ae Ts oe Wn Fig. 2.15 Valve voliage waveform34 HVDC Power Transmission Syste, significant components of the overall losses The damper losses are prim: ily due to the I aril II losses. See no tinuity in the (ideal) valve voltage wave! form shown in Fig. The voltage discontinuity in Fig. 2.15. jumps increase in magnitude with the increase in the delay angle. ‘ s been presented by Ainsworth [13] A simplified formula forthe loss in a damper ha and is as follows : (2.6) where P,=Py+Py 2 } 7) 2 in? a+ sin? (a + u), Pye LI5f Vy (Cp +Cs) [sin at a2 875 [sin 20+ 3sin Xo. +u) - 2u] Pyntef'v,, Cy Ry {24640875 + 0.433 [cos 20 + cos 2(a + u)] } (28) where f= fundamental frequency in Hz C, = stray capacitance in F a= delay angle u= overlap angle ‘The formula neglects the transients in the valve: voltage waveform. Different methods of including this factor are given in [17]. The effect of overlap angle on the damping Circuit loss for two different values of 0, for different methods of calculations are shown in Fig. 2.16. entation Thyris acycle Thyristor Valve 35 The typical current waveform in the damper circuit is shown in Fig. 2.17. This shows that the current is appreciable only around the voltage jumps which are 8 in number in acycle, & DAMPING CIRCUT CURRENT(A} 100 150 200-250 300 350 wt(DEGREES) Fig. 2.17 Computed damping circuit current waveform 2.3.5 Valve Protection Overvoltage Protection AHVDC valve must be designed to withstand (a) internal and (b) external overvolt- ages. The causes for the internal voltages have been mentioned in the previous section. The external sources of overvoltage across a valve can be due to (i) lighting and switching surges, (ii) dynamic overvoltages caused by load rejection, (iii) low order harmonic resonance in the AC system, and (iv) injection of AC voltage on the DC line due to converter faults. Zinc oxide, gapless DC surge arresters across a valve can protect the valve against transient overvoltages. The overvoltage in the forward direction can be controlled by Protective firing of the thyristors in a valve. This is achieved by utilizing the voltage sensitive switching action of a high voltage break-over diode across each thyristor. The protective firing thus is independent for each thyristor level. If there is some malfunction in the main firing system, the protective firing circuit is designed to work as a back-up, working continuously until the next maintenance period. ‘The overvoltages in the forward direction can also arise from partial recovery of a valve following tum-off process, (in inverters). In bypass pair mode operation, uneven distribution of voltage in a series connected string can arise due to dispersion in the holding current (f,) of the thyristors, In such cases, forward dv/dr protection is also required by sensing the dv/dr measured across a level and turning on the device if the threshold is exceeded. The threshold dv/dr level can be made a function of the junction temperature.HVDC Power Transmission System 36 rtemperature Protection [6, 12] : ; 4 = ing of a valve ig determined by the ability of LT be Leste ages following ‘a worst case, credible fault c rf the recovery vol o determined by the junction temperature. ltis possible to have reliable thermal and el ae an accurate prediction of junction temperature jcal models of a thyristor from which ae pony if pre-fault and post-fault thermal mS ra ee usually characterized by ils transient thermal impedance Ris described bythe following equations (61 mn (2.9) S A, e, 0 TW=D Are 2 d@ fa +B, o,=B, w(t) where Tis the temperature rise W=V A,B, = parameters determined from tests V=on-sue voluuge drop =V,(1+0r+ pr) +R, (i+a+e 7) 1 +E, 273+D im I+F, (1+ yrs or") VT Qa The last equation is a semi-empirical one, where V,,R,, Ey. Fy, 0, B, ¥, £, 8 and 8 are ‘constants for a device based on its construction and physical parameters, When W (0) is a unit step function, T@=R, @) = ZA, P-e7 (-8, 1] (12) seeaeee ‘modelling of the device, itis possible to predict on a continuous conih " mea ts eee ‘of the worst-case thyristor. This enables to Goss exceeds acrtical Capability of a valve, Also, when the junction blocked), the optical fans ee (above which the recovery voltages may not be pause, head (0 all the thyristors in a valve are inhibited, If high heavy current fault occurs, iimaybe reas blocked andthe AC breaker is ripped. in the affected valve until the teehier tripe order continuous firing of thyristors thyristors are connected in provided to ensure thei Parallel, current sharis 29% thos socezsapaiy a Of atansient disturbance, one of the mee ; guish, Ths is done by regat : i parallel thyristors Thyristor Valve 2.4 VALVE TES The testing of a: a valve to perfé Dielectric Tes @ Testons @ D @) Si © Li (i) Valve @ OD @) A © s @ 1 @® § © } A valve s installed an the tests, Hi ‘The val withstand vistang um is vhich fault ™mal 2.9) -10) Thyristor Valve ‘37 2.4 VALVE TESTS [7] The testing of a valve before commissioning is carried out to check the capability of a valve to perform effectively under the conditions specified. The valve testing Procedures are undergoing changes with beter understanding of the stresses to which a valve is exposed. The tests are broadly classified as : @ dielectric tests: and (ii) ‘operational tests, ‘The dielectric tests are meant to verify the withstand and voltage related character- istics of a valve under various: ‘overvoltage conditions, The operational tests are intended 0 show the correct operatién and capabilities of a valve under various operating Conditions including the worst case fault conditions, A further classification of the tests is Gi) type tests which verify that the valve design will meet the requirements specified and agreed upon between the purchaser and supplier, (ii) production sample tests which verify that the units as, manufactured conform to the design. All dielectric tests are type tests, Dielectric Tests @ Test on valve structures (@) DC corona voltage test (©) Switching impulse voltage test © Lightning impuise voltage test Gi) Valve withstand voltage tests @) DC withstand voltage test (b) AC withstand voltage test (©) Switching impulse withstand voltage test (@) Lightning impulse withstand voltage test (©) Steep front impulse withstand voltage test (© Non-periodic firing test. ‘The tests on valve Structure are intended to verify the design of the valve structure regarding DC corona extinction voltage levels and impulse voltage withstand levels as related to its insulation coordination. The tests are performed with the portions of the ‘Structure short-circuited as necessary to simulate maximum service stress conditions, Grounded shields are to be Suitably arranged around the valve structure to simulate effects of nearby building steel, grounding rods, etc. which influence the stray capaci- tance to ground of the test structure, A valve structure refers to complete multi-valve assembly with all the components installed and the valve arresters, disconnected. All the modules need not be installed for the tests. However, their grading circuits should be simulated. ‘The valve withstand voltage tests are intended to verify the design of the valve to withstand various types of ‘overvoltages. The tests must verify that : 1. sufficient insulation is provided and test the effectiveness of internal overvoltage Protective circuits. 2. partial discharges will notoccur undernormalconditionsand any such discharges under high voltage conditions are within safe limits " 3. the grading and damping circuits have sufficient power ratingHVDC Power Transmission Systen, i thyristors string is satisfactory ' eries connected thy 4, the voltage dist Lees valve are able to withstand be gece i : ¥ ; and the various ane to interference ar es son a geet . the gating circ high overvoltage con the valve canbe fet te stg of ihyristor is adequate $0 as not to Overstress coherence in the fi fired thyristor. , ie ipa alve assembly with surge arresters disconnected yee a the the come ls such as valve reactors are kept in place so e ice tests. Alle er ‘orrectly simulated. The valve electronics are energized during conditons of valve are OWT ved before and after each test sequence, The testis ome fore a os Se el ifit meets the criteria given below : — vee Not more than 1% ofthe tal number of thyristors in avalve fail during any one 38 au test a a 2 Ro more than 2% fail forthe entire series of type tests Itis assumed that all the failures are random in nature. Operational Tests A) Type tests @ Short circuit current without subsequent blocking test Gi) Heat run test (ii) Minimum alternating voltage test (iv) Intermittent direct current test (v) Valve losses test (vi) Thyristor case temperature rise test (vii) Current sharing test. B) Production sample test : () Periodic firing and extinction test Gi) Short circuit current with subsequent blocking test ‘The operational tests can be performed on valve modules instead of complete valves. Avalve module (or section) is defined as the smallest assembly comprising a number of thyristors and their immediate auxiliaries for firing and protection, voltage dividing Components, distributed valve reactors, from which the value is built up and which exhibits the same electrical properties as the complete valve, But it can withstand only portion of the full voltage blocking capability of the valve. A valve is built up from eeimnecton of anumber of valve modules. The ‘operational tests must demonstrate 1. the valve functions properly (no commutation fai curren failure, correct voltage and it ser fos, Correct operation of all internal circuits) under various specified test 2, the wrn-on, tumn-off voltage and current thin the capabiliti stresses thyristors and other internal circuits 2 Satie Ronee 3 the cooling provided is adequate and no component is overheated under worst case fault conditions ae net ic . valve losses are within specified limits . the surge i ili i aes separa capability of the thyristors for the worst fault condition is we Thyristor Valve ‘The criteria for 1, Not more th 2. No failure 0 3. Thereareno guides. 2.5 RECENT TE ‘The recent devel HVDC valves. TI 1, Developm triggered t Light Triggere Both in US/ tested. LTT ha 1. Infinite 2. Total no 3, Faster t 4, Elimina ‘The light s diodes. ‘One of the by (i) a separ equivalent of Althoug covervoltag would bee valves. amber of dividing d which ind only up from onstrate current fied test s of the der the tion is Thyristor Valve 39 ‘The criteria for the success of a type test are the following : 1. Not more than 1% of thyristors have failed 2. No failure of electronic or other components . 5 3. Thereareno disruptive discharges across dielectric mzterial, cooling tubesor light guides, 2.5 RECENT TRENDS ‘The recent developments are expected to improve reliability and reduce the cost of HVDC valves. These are mainly : 1, Development in high power semiconductor devices - these include direct light triggered thyristors (LTT) and metal- oxide semiconductor controlled thyristors (wcT) 2. Better cooling techniques suchas forced vaporization (two phase flow) asameans of reducing thermal resistance between the heat sink and the ambient. 3. Suspension of quadrivalve assembly from ceiling to withstand seismic forces. Light Triggered Thyristor (3, 4, 8, 18] Both in USA and Japan, high power, high voltage LTT have been developed and tested. LTT has the following advantage over electrically triggered thyristors (ETT). 1. Infinite gate isolation 2. Total noise immunity for the control circuit 3. Faster turn-on time 4, Elimination of high voltage pulse transformers and auxiliary power supplies. ‘The light sources used are either gallium arsenide light emitting diodes (LED) or laser diodes. One of the problems of LTT is the reduced difdt rating. This problem can be solved by (i) a separate gate LTT to fire an ETT or (ii) controlled LTT turn-on. The electrical equivalent of the latter is shown in Fig. 2,18. k Ry Ro oe ie MAIN SCR SCR A Fig. 2.18 Light triggered thyristor Although the power supply circuits (atthe thyristor level) in LTT are eliminated, the overvoltage protection circuits are still required. To improve the system reliability, it ‘would be advantageous to eliminate these also. With this objective, a novel LTT with \HVDC Power Transmission Systen, itachi, Japan recently (g) . elf ection streture has been reported by me eee eae 8 SE tors are Of PES () Se eg tha te SE ah phenomenon. It is claim non, (i) using punch- ation in breakover vol ecteaca is about 10% when the tempera device tums on safely at 7300V. trolled tor (MCT) [9] i See ae by GE, USA and Siemens, West Germany is expected This device being dove OPO rough in the use of high power devices. In MCT, a to provide evan ee yws the thyristor to be turned-off. It is better than a GTO (gate MOSFET on eee oe much less gate current.to turn-off. Also the device has low Sn ee saee alia it suitable for use with high current densities, on-state resistance, Static Induction Thyristor [19] ic induction (SI) thyristors have been developed in Japan as an alternative to ors patialaty oe to thet advantages of high speed switching, high turn-on and turn-off current gain, etc. The concept of optically controlled, light triggered and light quenched (TQ) SI thyristors for DC transmission has been reported [19]. Electrically controlled SI thyristors are available in the ratings upto 4000V and 400A. Asymmetrical Thyristor [20] ture is varied from 23°C to 100°C. The ~ The cost and reliability of thyristor valves depend also on the number of devices that have to be connected in series, apart from the complexity of the control circuits. While the lamer aspect is considered by LTT with built-in overvoltage protection, the first aspect is tackled by increasing the blocking voltage of a device. However, this is not simple as there are constraints introduced by power loss considerations. It has been shown that there is an optimum blocking voltage which, for a symmetrical thyristor is in the range of 4 10 5 KV: To shift the optimum blocking voltage to a higher val e, a seri ion of asymmetrical thyrisior (ASCR)a tage gl lue, a series connection of Suspended Valve The use of quadri-valve configurati, quadri-valves configuration as ae tall suuctures (upio 16 cewacin Wntbeet valve hall. However, high seismic forces have Itage (Vpo) of the device is kept small and the temperature | Thyristor Val tobe withstc ceiling inste 1) SKC (2) Bday 1) V.AK Public (4 LRI Thyris {5} NL mY. Prove 19) DY. > fio) Pcs Stati (11) P. Cc Presi 12] ML Deli 13) BJ. 196: Aini 114] G.I TEE 115) Gk Val 16) Y.1 Vol 17) Re Thy 18) H. Lo 19) 3.1 30 (20) H. va Rn Gi (22) F a (23) P. Q Juive to mn and d light ically ’s that While > first is not tor is on of 1also end- -s the -113 min- isa and ver, ave Thyristor Valve tobe withstood (upto 0.5 g acceleration), itis convenient to suspend the valve from the ceiling instead of base mounting. The suspension arrangement has spring and damper mechanism for mechanical isolation of the valve from the building vibrations. The valve structure is made flexible and the connections to the wall bushings and between the valves are also made by flexible bus. References and Bibliography ru RQ) Bl 4) o @ m (8) eo (10) oy 02) 03] 4) 15) 6 o7 (18) 9} (20) 21) 22) 23) S.K. Gandhi, ‘Semiconductor Power Devices’, New York, Wiley, 1967. . ‘Modem Power Devices", New York, Wiley, 1987. a V.A.K. Temple, ‘Advanced Light Triggered Thyristors for Electric Power Systems’, IEE Conf. LR. Lowry, Guygi L., G.N, Hingorani and Nylander F.L., ‘Field Testing of Light-Triggered B. Jayant Ba Publication 205, pp 86-91, 1981. ‘Thyristors in Electric Utility Applications ibid pp 92-95, 1981. NLL. Diesko et al, ‘Application of a Digital Computer Program to Transient Analysis and Design of HVDC and AC Thyristor Valves', ibid, pp 167-170, 1981. 1H. Gibson et al, ‘Characterization, Evaluation and Modelling of Thyristors for H.V.D.C. Converter valves’, IEE Conf. Publication 255 London, pp. 320-324, 1985. RE. Harrison, Shemie RK., and Krishnayya P.CS., ‘A Proposed Test Specification for HVDC ‘Thytistor Valve’, IEEE Trans,, Vol. PAS-97, No. Y. Shimza et al Protection Structure’, ‘Nov/Dec. 1978, pp 2207-2218. *A High Voltage Light, Activated Thyristors with Novel Overvoltages Self- EEE Conference Proceeding (PESC) pp 99-103, 1987. D.Y. Chen, ‘Power Semiconductors: Fast, Tough and Compact’, IEEE Spectrum, September, 1987. P.CS. Krishnayya, ‘Important Characteristics of Thyristors of Valves for HVDC Transmission and ‘Static Var Compensstor’, CIGRE Conference Proceedings paper, 14-10, 1984. P. Chowdhuri, ‘Behaviour of Thyristors Under Transient Overvoltage’, Paper No.A80, 114-9 Presented .at IEEE PES Winter Meeting, Feb. 1980. M.L. Woodhouse, ‘Voltage and Current Stresses on HVDC Valves’, IEEE Trans. on Power Delivery, Vol. PWRD-2, No.1, pp 199-206, 1987. BJ. Cory, ‘High Voltage Direct Current Converters and Systems’, (Edited) MacDonald, Landon, 1965, Chapter 7, ‘Filters, Damping Circuits and Reactive Volt-Amps in HVDC Converters’ by 1.D. ‘Ainsworth, pp 137-174, G. Karady and Gilsig, T., ‘The Calculation of Tum-off Overvoltages in 1 HVDC Thyristor Valve’, IEEE Trans., Vol. PAS-90, pp 2802-2811, 1971. G. Karady and Gilsig T., “The Calculation of Tum-on Overvoltages in a High Voltage DC Thyristor ‘Valve’, IEEE Ts Vol. PAS-91, No. 2, pp 565-573, 1972. Y. Beausejour and Karady G., "Valve Damping Circuit Design for HVDC Systems’, IEEE Trans., Vol. PAS 92 No. 5, pp 1615-1621, 1973. Rambsbu Adape and Padiyar K.R,, ‘Aliemative Evaluation of Damping Circuit Losses in HVDC ‘Thyrinor Valves’, IEEE Trans, Vol. PWRD-3, No Hi. Meh and Temple V.1 London, pp. 310-314, 198: 43. Nishizawa and Tamamushi T. 304-309, 1985. 4, 1988, pp 1823-1831. few Thyristor Application to DC Power Transmission’ ‘Advanced Light Triggered Thyristors’, IEE Conf. Publication 255, H.Mitlehner, Sack J. and Schulze H., ‘High Voltage Thyristor for HVDC Transmission and Static Var Compensators’, IEEE PESC 88 Record, 1988, Gil Addis etal, ‘Advanced HVDC Valve’, IEEE: and Power, IEE (London), pp. 375-378. My 1982. pp 934-939. P. Lips and G. Thiele, ‘Design and teting of thyristor valves for the H) Chsteaguay’, Proceedings of Int. Conf. DC Power Jane 1984. ‘Transmission, Montreal, ‘Conf. Publication 255, London, pp. 340-345, 1985. F.G.Goodrich and A.W. Tozer, “HVDC cunverter valves forthe new cross-channel link’, VDC back-to-back tie Canada, pp. 228-233,| Analysis of HVE voltages onto . | shownin Fig. | omatany inst i itch which . nverters Peeing is of HVDC Co | thyrinte awi 3. Analys' actually obi | (both for dio : ‘The outpu i - i is in HVDC converter stations frequency is Dee Conversion from AC to DC and vice versa S ct the bridge (also called | by using three phase bridge converters. The config | 3.2 CHOIC! ‘The configu and transfor The cont ration can Fig. 3.1 Graetz Circuit Graetz circuit) is as shown in Fig. 3.1. This is a six pulse converter and the 12 pulse converter is composed of two bridges in series supplied from two different (three phase) transformers with voltages differing in phase by 30° 1 PULSE NUMBER ‘The pulse number of aconverter is defined as the number of pulsations (cycles of ripple) Of direct voltage per cycle of alternating voltage. ‘The conversion from AC to DC involves switching sequentially different sinusoidal conne comm series (Note (negl Fig. 3.2 A ‘p' pulse convener0 called Pulse hase) pple) oidal Analysis of HVDC Converters a voltages onto the DC circuit. One possible configuration for a "p‘ pulse converter Is shown in Fig. 3.2. A valve can be treated as a (controllable) switch which can be wmed on at any instant, provided the voltage across it is positive. A diode is an uncontrolled switch which will turn on immediately after the voltage becomes positive whereas the thyristor switching can be delayed by an angle a (alpha). The voltage sources are actually obtained from the transformer secondary windings. The opening of the switch (both for diode and thyristor) occurs at the current zero (neglecting the tum-off time). ‘The output voltageV, of the converter consists af a DC componentand a ripple whose y is determined by the pulse number. 3.2 CHOICE OF CONVERTER CONFIGURATION ‘The configuration for a given pulse number is selected in such a way that both the valve and transformer (feeding the converter) utilization are maximized. ‘The configuration shown in Fig. 3.2 is not the best. In general, a converter configu- ration can be defined by the basic commutation group and the number of such groups Fig. 3.3 Convener made up of series und parallel connection of commutation groupe connected in series and parallel (sce Fig. 3.3). If there are ‘g’ valves in a basic commutation group and r of these are connected in parallel and s of them connected in series, then pee rs BD (Note: A commutation group is defined as the group of valves in which only one (neglecting overlap) conducts at »).44 HVDC Power Transmission System | » 3.2.1 Valve Rating a ] Analysis, Abas eae i Itage ithas to | ‘The The valve voltage rating is specified in terms of peak inverse vO" ) it has withstand. The ratio of PIV to the average dc voltage is an index’ of ee valve utilization, i The average maximum de voltage across the converter is given by s i ma { 4 i =s— ost dwt Va“5 3m J Fae | nla =e, sin + | 32) i vol TV) across a valve can be obtained as follows = eee aeen ie ea inverse voltage occurs when the valve with a phase a displacement of x radian (180°) is conducting and this is given by ofqw PIV=2E, ; 3a) the A If‘q' is odd, maximum inverse voltage occurs when the valve with a phase shift of Fa nt a is conducting. In this case, PIV=2E, cos 3.3) - ‘The valve utilization factor is given by =e Fort FIV -—22 tor qeven @ sqsin— q =—+ for qodd ; ‘Table 3.1 shows the valve utilization factor for different six pulse converter configu- ‘ations. The best valve utilization is obtained for configurations 1 and 3. Table 3.1 Valve Utilization Factor SLNo. ‘ ‘ . ely, S do 1 2 2 1 3 1.047 i i 8 1 3.142 4 1 2 1.047 ; 3 2 1 2.094 $ 1 1 2.094 Cransformer Rating ‘The current rating of indi : ing valve (as wellas the transformer winding supplying it) is given mayHVDC Power Transmission System, 44 3.2.1 Valve Rating 3 “The valve voltage rating is specified in terms of peak inverse vliage ev) ie has tg withstand. The ratio of PIV to the average dc voltage isan ne ny ization, ‘The average maximum de voltage across the converter is gh we sq mi seb, sing fine (2) inverse voltage (PIV) across a valve can be obtained as follows : noe then oe ects inverse voltage ‘occurs when the valve with a phase displacement of x radian (180°) is conducting and this is'given by PIV=2E, If“q' is odd, maximum inverse voltage occurs when the valve with a phase shift of @G.3a) xt = is conducting. In this case, PIV = 2E, cos a 3b) ‘The valve utilization factor is given by aw =n for q even © sqsine = for qodd sq sin Table 3.1 shows the valve utilization factor for different six pulse: 7 lv pulse converter configu- ‘ations. The best valve utilization is obtained for configurations 1 and 3. Table 3.1 Valve Utilization Factor ‘SLNo. PIV _ 4 ® s do 1 2 7 , 3 1.047 z 3 * 1 3.142 m4 3 i. 2 1.047 5 é 2 1 2.094 a 1 2.094 ‘Where /, is the DC current which is assumed to be constant, j | of q whic the AC pcAnalysis of HVDC Converters 45 ‘The transformer rating on the valve side (in volt amperes) is given by En S, = pel, wo Be tt G5) , V2 Va sing s ‘The transformer utilization factor ) isoily a function of g. The optimum value ol of ¢ which results in maximum utilization is equal to 3. Itis a fortunate coincidence that the AC power supply is3 phase and the commutation group of 3 valves is easily arranged. For q=3, S, cr = 1481 G6) woe ‘The transformer utilization can be improved further if two valve groups can share a single transformer winding. In this case, the current rating of the winding can be increased by a factor of V2 while decreasing the number of windings by a factor of 2. For this case, 5, Ve 1 = 1047 672) For a6 pulse converter, this can be easily arranged. The Graetz circuit shown in Fig. 3.1 is obtained when the two windings (shown in Fig. 3.4) are combined into one, Fig. 34 Six pulse converter* HVDC Power Transmission System “ aaale a = “Thus, itis shown that both from valve and transformer utilization considerations, Graetzciruit is the best circuit fora six pulse converter eS HVDC transmission, the series conduction of converter groups! prefered because: 3 i se eassot oil ‘and proveetion as wel asthe requirements of high voltage ing Tusa 12 pulse converter is obtained by the series connection of two bridges, The “Sor phase > displacement Detween the two sets of nn re ered by the i bridge for feeding the second transformer connections, Y/Y for feeding one one % i v six pulse converter because of the werterispreferable over the six pt ¢ scone ‘tis oe ae However, increase in pulse number beyond 12 is not Seal because the non-characteristic harmonics (which arise due to asymmetry in in supply etc.) are not eliminated. ‘At any instant,"two valves are conducting in the bridge, one from the upper ‘commutation group and the second from the lower commutation group. The firing of the next valve in a particular group results in the turning off of the valve that is already ‘conducting. This assumption - that there is no overlap between the two valves in a group is incorrect and will be relaxed later. However, the analysis without overlap which is simpler, gives an insight into the working of the converter. The valves are numbered in the sequence in which they are fired. Thus valve 2 is fired 60° after the firing of valve 1 and the valve 3 is fired 60° after the firing of the second valve. Each valve conducts for 120? and the interval between consecutive firing pulse is 60° in steady state. “The following assumptions are made to simplify the analysis “The de current is constant 2. The valves can be modelled as ideal switches with zero impedance when on (Conducting) and with infinite impedance when off (not conducting) 3. The AC voltages at the converter bus are sinusoidal and remain constant. One period of the AC supply voltage can be divided into 6 intervals - each coresponting 10 the conduction of a pait of valves. The dc voltage waveform repeats for each interval. Thus, for the calculation of the average de voltage, it is necessary to aie ‘only one interval (ay the interval corresponding to the conduction of valves 2 Assuming the firing of valve 3 is delayed by an angle is . 7 G, (that is a° after the zero ‘crossing of the commutation voltage for valve 3 - voltage ¢,.), the instantaneous de voltage v, during the interval is given by . “a =~ ee =e, AS Mrsa+ 6 Lae, = VIE, snot Then'e,, = V2 E,, sin (os + 60°) ‘Analysis of HVDC Converters ‘Average de voluge =v, 3 RE A Yes Equation G8) indicates is 180° (from 0° to 180°) same converter can acta8 is positive or negative is conduction of eurrent by DC Voltage Waveform “The de voliage wavefor ‘the supply frequency. Thi the order henp ‘where pis the pulse num ‘The rms value of the h*On System, erations Analysis of HVDC Converters a referreg a+60* Voltage Average de voltage =, =3 J V7 E,, sin (Ot + 60°) dor Bes. The a I by the =EV2E, [cos (a+ 60°) - cos (a + 120°)] Second z v. < ¥ cos 1.35E), cosa & SC Of the dun > alr 88) 2 fs noe aN, ra netry in Equation (3.8 indicates thi for different valdes ofc, V,is variable, The range of a is 180° (from 0° to 180°) and correspondingly V, can vary from V,, to~V.,. Thus the same converter can act asa rectifier or inverter depending upon whether the dc voltage is positive or negative. tis tobe noted that tit is based on the assumption of continuous conduction of current by any valve over the 120° interval. DC Voltage Waveform ae ‘ ~_‘Thedc voltage waveform contains aripple whose fundamental frequency is six times aa the supply frequency. This can be analysed in Fourier series and contains harmonics of lready hoon ich is ae red i \ where p is the pulse number and n is an integer. alge ‘The rms value of the h* order harmonic in de voltage is riven by i 2 ducts V2 [rnlatoad anol 69) ib 4 ~The waveform of the direct voliage for different values of a. are shown in Fig. 3.5. ‘The waveforms of the valve voltage are also shown in the same figure. The figure shows non 2 1 yeas WY | \ ry to es 2 | 1 on "ero ae 5 dic | tens} AD ! sist a ead Fig. 3.5 DC and valve vohage waveformsHVDC Power 1 rurrrsosers oysten, “ ion from one valve to the next j ich ari to the commutation i ext in ince Son es San jumps areall of the same magnitude given by the same’ commu . d Vv a VIE, sin a= (HIV 49 SMO _ B.10) : contribute to the voltage sureses(2¥-) across the valve. Valve Analysis of H ‘The rms ‘These voltage jumps 0° damping circuits are provi Although a can vary from 0 et ensure the firing of all the series connect -@ greater than 2er0, (Sa) ided to relieve these stresses. e 0 to 180°, the full range cannot be utilized. In order to istors, it is necessary to provide a ‘Also, in order to allow for the tum off a nninimum li 80°. The dela P, Fi to provide an imit less than 180°. ry angle timc a pond 180" =) where 7 is geld the extinction angle where cos ¢ rm is n Cael margin angle). The minimum va be aa eee ee TheDC istypically 10° although in normal operation as an inverter, itis not llowed to go below we get 15° or 18°. Inthe absence of forced commutation, a. cannot be made negative. oh ‘Substit AC Current Waveform aiaa Itis assumed that the direct current has no ripple (or harmonics). This is new © -valid because of the smoothing reactor provided in series with the bridge circuit. The AC The carreants flowing through the valve (secondary) and primary windings of the converter i consumes 33.2 Wit ams 1 4Ta Duete “a ao LT supply 1 a from one the curre Fig. 3.6 AC Current Waveform are cont (commu Each transformer contain harmonics. The wavefi is form of thecurrentina valve winding i 3.7). In in Fig. 3.6. The. Ive winding is shown es eee fundamental component of the current is given by 12 he Vat 4 1, e040 VE), Gull) Whereas the rms value of the current is r= fB1 ae = 3.12) = warmonics contained in the current waveform two: are ‘ h=npt) Of the order given by a where nis an integer, pis the pulse an inte . . number. . ACharmonieare571113andhigherones phases abe converte, the oder of” ee for each one of the first four harmonics and ; are filtered out by using tuned filters eae i fand a high pass filter forthe rest,;normaty it. TheAC converter gisshown ven by G.I) @.12) order of ed filters Analysis of HVDC Converters 4 ‘The rms value of h* harmonic is given by (3.13) ‘The Power Factor ‘The AC power supplied to the converter is given by Pyce VEE, 1, 0080 where cos ¢ is the power factor. ‘The DC power must match the AC power ignoring the losses in the converter. Thus, we get hem Pag Vip = V3 E,, 1, cos @.14) ‘Substituting for V,, and J, from equations (3.8) and (3.11) in the above equation we obtain cos} = cos G.15) The reactive power requirements are increased as t is increased from zero (or reduced from 180°). When c= 90°, the power factor is zero and only reactive power is consumed. 332 With Overlap Due to the leakage inductance of the converter transformers and the impedance in the supply network, the current in a valve cannot change suddenly and thus commutation from one valve to the next cannot be instantaneous. For example, when valve 3 is fired, the current transfer from valve 1 to valve 3 takes a finite period during which both valves are conducting. This is called overlap and its duration is measured by the overlap (Commutation) angle ‘us Each interval of the period of supply can be divided into two subintervals (see Fig. 3.7). In the first subinterval, three valves are cohducting and in the second subinterval, 3 6 8 oe 8 Fig. 3.7 Timing diagram two valves are conducting. This is based on the assumption that the overlap angle is less than 60°. As the overlap angle increases to 60°, there is no instant when only two valves ‘fe conducting, As the overlap angle increases beyond 60°, there is finite period during an interval, when four valves conduct and the rest of the interval during which three valves conduct. Thus there are three modes of the converter as follows: 1. Mode 1 - Two and three valve conduction (u < 60°)50 HVDC Power Transmission System 2, Mode 2 - Three valve conduction (u= 60°) - 3. Mode 3 - Three and four valve conduction (u > 60°) 1e 2 may be just a point on the boundary Depending upon the delay angle o, the mod of modes 1 and 3. This is explained later. With the assumptionsstated carlicr for s only one interval of 60 degree duration - sa of (conduction of) valve 3. _Srhoaiyss of Two and Three Valve Conduction Mode For the interval: teady state analysis, it issufficient toconsider y the interval corresponding to the initiation considered, the bridge circuitcan be reduced to that shown in Fig.3.8, Fig. 3.8 Equivalent circuit for 3 valve conduction (3 “,) Fon ae .16) ‘The LHS. in the above equation is called the commutating emf whose value is given by =v2E,, snot G19 which is also the voltage across valve 3 just. before it starts conducting. Since neue G18) We get, VIE, sinwt = 2L as 4 ;@ G9; Solving the above equation, we get i, =1, (C08 0-cosww), as oSa+u where, . VIE 1” eT 3.21) 3.20) Analysis of HVDC C Note that the s i,or=¢ Atwrs atu, Fig. 39 Average Dire ‘The averaAnalysis of HVDC Converters 5 51 boundan, Note that the solution given in (3.20) is obtained from the initial condition sek i, (@t= 0) =0 3.22) initigh Aton = 0+ u, i, =/,, This gives tion i 142, [08 0: cos (a+ u)) 3.23) The waveforms of direct véltage and the valve currents during commutation for a rectifier and an inverter are shown in Fig, 3.9(a) and3,9(b) E pectively. Itis tobe noted Fig. 3.3. that during commutation, the instantaneous de voltage i instead of (e, -¢,). 1 (b) Fig. 3.9 Voltage and current waveform during firing of a valve (a) rectifier (b) invener Average Direct Voltage The average direct voltage can be obtained as (3.16) is i a a given vy, -3| J e,a(ot)+ f (47%) 4 «| x on . a+60 Boe ae: aoe = 3 if e,4@)- | —y sw] a 5 a+60° em Ce, -€,) 2 -f 4“den =F e, d@t) J z 3.18) | J te : ase =v, cosa - J v7E, sin odor : (3.24) ; : 5 ae 2V,, cosa - > V2 E,, [cosa- cos (a+ u)] 20) B.25)HVDC Power Transmission Syston, 52 ‘The equation (323) can be substituted in @.24) £0 yy Analysis of HVDC Conve i, eal ‘two major jumps occu sl a- ve¥y[ee- ah %6 3.26) v,,2v2 where . V,.=V2 R= 3 ol, G27, “The remaining 6 ju R_ iscalled equivalent commutdho ZGistance. It is something analogous to armature B, is ale ea achines inthe sense that it only Tepresents & voWABS Grop and nog janes Om 6 oa ———— power loss. i in. magnitudes —— qT Re g hg | werter Equations 2 uato i DepreeesA Ngo cose ‘a : cs B=x-a and use opposite po Fig. 3.10 Equivalent circuit for a bridge converter ‘current. Thus | -Vv “The equivalent circuit of the bridge rectifier is shown in Fig. 3.10. v= DC Voltage and Valve Voltage Wave Forms WNIT wt v2" 180" 240" 31 36 Fig. 3.11 DC Voltage for'a = 15°, u= 15° Fig. 3.11 shows the waveform of the = “ of the voltage across the converter bridge (v,). The pe volage waveform s shown Fig 312 Thevuvevohags na ae nx Te eee to be noted that ¥ AC Current and ‘The waveforr distorted. Thus,t for the case with derived from Fo Fig. 3.12 Voluge across Valve 1 for @ = 15°,u = 152 : :Analysis of HVDC Converters 53 ‘two major ji G25) major jumps occur at the firing and the turning off of the valve and are given by V2 V2E,, sina V2 V28, sin (+ u) G27 nats ‘The remaining 6 jumps can be divided into two groups, the first group composed of ‘ : ; jumps that are equal in ft j - magne to 11 and the second group of jumps with magnitudes 12 . These voltage jumps result in extra losses in the damper cient there ‘Equations For an inverter, itis usual to define an advance angle Bas Bt B=n-o * * 28) ants opps poaiy for the de voltage with voltage rise opposite to the direction of, ee [cosa + cos (a +u)) - Vwi ’ = [cos (nt - B) + cos (we - 9) Fs ‘ — = HE [eosB + cosy] f - 329) where the extinction angle 7/is defined as yeBousnoon G30) Similarly, it can be shown that Vg “Vig OSB R, Ty oF =v, 8 - Rly Gap vy). The ‘The subscript “i” refers to thertfiverter. . aps. The Canaanteg equations (29) and (3.31) with 3:25) and (3:26) shows et eT cqnaa weobained by substting for ain the rece equations, However i's to be noted that while 0: is directly controllable, 7 is not. AC Current and DC Voltage Harmonics ‘The waveforms of the valve currents and the current in the valve winding are Se Ths the expression for the fundametaleomponcn.o the AC currentderived aia, Tus esp nt vel, The ata ees fre cen derived from Fourier analysis and is given by 12 af (3.32) a [i a 14) :54 HVDC Power Transmission Sy, where v6 lornsae ru] G33 2 ‘ H te] G4) } -v6, usin 2a sn28| m “4/4 (Cosa — cosd) j where @ is the power factor angle and § = a+ u. j From the above expressions, the power factor angle can be obtained as 2u + sin 20 ~ sin 28 635) cos 20.-cos 25 ‘The harmonic components in the AC currentare also altered. These are reduced from the values calculated with no overlap. The reduction factor is given by 1/2 tan $= 1 phe hla? x? - on cos (20+u)] G36) where sin (h +1) sin (h - 1) peOa Dens eames F loos « - os (a +4) 1 ve! =e + = harmonic component with no overlap Fig. 3.13 Variation of AC current harmonics with overlap Figure 3.13 shows this reduction factor for the case o: = 30°. It can be seen that all the harmonics especially of higher order, decrease sharply with increasing values of u and the reduction factor lies in the range 0.1 to 0.2. Analysis of HV The harmo that Vv v where Figure 3. (ii) Three a Althougt during DC| be larger an ‘When th three and th commutatic For examp! equivalent |3.33) 334) 35) 36) ae ‘Analysis of HVDC Converters 55 ied harmonics in the direct voltage are also altered due to overlap. It can be shown v EP De od nn wr -{ze +67 -2FG cos Qa+uy)} 639) where _ 008 (h + 1} cos (h ~ 1) Tim aaeadta Ghee hela G40) 0800 Tada pelt 0.078) + ! £80980) + 0.02 ‘ | 1 2 ae A Fig. 3.14 Variation of DC voltage harmonies with overlap Vv, Figure 3.14 showe(7=] for different values of u and 0.. ae (ii) Three and Four valve conduction mode Although two and three valve conduction mode is the normal mode of operation, during DC line faults oradip in the AC voltage, itis likely thatthe overlap angle u would be larger and may exceed 60°. ‘When the overlap angle exceeds 60°, the minimum number of valves conducting are three and there are intervals when four valves are conducting. This is because when a commutation process is started, the previous’ commutation process isnot yet completed. For example, when valve 3 is fired, the valves 1,6, and 2 are still conducting. The equivalent circuit for this condition is shown in Fig. 3.15. * Big. 3.15 Equivalent cireuit for four valve conduction56 Foriasa@rsa+u-60° i, =1, sin(@t + 609 +A G4 =I aie sin or +C B42) Ey 2 wheel ate” Jr" et “The constant A can be determined from the instial condition HVDC Power Transmission Sym, i (or =a)=1, =/, sin (+ 60°) +A 3.43) 1 ‘The constant C can be determined from the final condition j i i, (or =a+u - 60%) = O= 1-1, sin (a+ u - 60°) +C =0 Bas) | | For, a+u -60°S wt S$ a+ 60° i= 1, os or+B cus] ‘The constant B can be determined from the continuity equation i(mt=a+u=609=/, sin(atu) +A =I, os (a + u ~60°)+B ” From symmetry, we have iy 4 (fF = 0 +60°)=i, (wt = a) 1, cosa + 609 =B=1,-1; sino +C G47) After some manipulations, we can obtain bs 1, = [eos (a - 309 ~ cos(o + u +309) (3.48) Average Direct Voltage The e: ic i i or oe average direct voltage is obtained as ee 3 Ynz f -ded er) 41 -60° €, =-E, wos wr 33 , Vv, =#-T En [sin (© +60°)—sin (@+u-60°y ae 3. Vet ye ty fom © ~ 30°) +008 (a + y +309) (3.49) From (3.48) and (3.49) we get v, “t4[ VF onto-s09- 4] * Analysis 9 Comp resistance ‘When the I, inéreas 1 Atu= u would of a< 3 overlap a beyond 6 Fig ting, the ata poten raising th mode wh in direct DC and) The inSyste, B4 1 G42) (3.43) (3.44) (3.45) (3.46) (3.47) (3.48) 3.49) Aiehjet of HVDC Converters 57 = V3V,, 005 (a ~ 309)- 3R, J, G50) * Comparing equation (3.50) with (3.26) shows that the equivalent commutating resistance for this case is three times that for the case with overlap angle less than 60°. ‘When the delay angle is zero, the overlap angle u reaches a value of 60° when the current T, increases to ES 1 1, = 1, 0 ~ cos 60°} = > 1, ‘Atu=60®, three valves are conducting all the time. If the currents increased further, x would tend to increase further, however, this is not possible unless 230°. For values of a < 30°, an increase in direct current is possible only when o: is increased and the overlap angle is maintained at 60°, The reason why the overlap cannot be increased beyond 60° for a. < 30° is evident from Fig. 3.16. When valves 2 and 6 are still condu- Fig. 3.16 Figure explaining minimum delay angle = 30? if overlap angle = 60°or greater +e, ¢ c —e Ps cting, the anode of valve 3 is ata potential of (-2——) = —5# and the cathode is ata potential of e,. Hence unless e, is negative, the valve 3 cannot conduct (even after raising the gate pulse) and this is possible only for 0.2 30°. Thus, there is an intermediate mode when the value of overlap angle remains constant at 60°. In this mode, the increase in direct current is accompanied by automatic increase in the firing angle. DC and Valve Voltage Wave Forms The instantaneous voltage across the converter bridge (v,) and the valve voltage for 0 60" Fig. 3.17 DC voltage for a = 30° u = 90° 720° 180° 240" 300° 360"HVDC Power Transmission Sy, 58 the 3 and 4 valve conduction modes ares! instantaneous voltage across the bridge ¢# in Fi 3.18 respectively, sis of HVDC C. hown in Figures 3.17 and 1 The | Analys ‘an have both positive and negative eXCUTSiony For mode (2) 1 equation is given t so" if o® 60° 120° 180° 240° 300° 360° wt = z j V . vhere Vi, = g ion (3.5 Fig. 3.18 Voltage Across Valve I for a = 30°, u = 90° Equation (: 4 r 4 k followed by periods of zero magnitude, The valve voltage has 6 jumps with three of them 7 having the magnitude ede a a V,,= SE, sin (a- 309) ‘The other three jumps have the magnitudes given by vi, . ‘ Vi a= tg yy sin (+ u + 30°) 3.4CONVERTER BRIDGE CHARACTERISTICS. A) Rectifier : The rectifier in general has three modes. 1) First mode : Two and three valve conduction mode (u< 60°) 2) Second mode : Three valve conduction mode only for & < 30°, (u = 60°) 3) Third mode: Three and four valve conduction mode o> 30°, (60 30° 120° Vi 0 1 va ah Wt 2260 HVDC Power Transmission Syston B) Inverter Analysis of HV. The inverter characteristics are similar to the rectifier characteristic, However, 1b) 60°: operation as an inverter requires a minimum commutation ee ae Which pee the voltage across the valve S negative. Hence the operating regi inverter jg ‘salons dil tifier. a: iseameane aie ‘angle (Gis equal to the extinction angle (y) only for value, 1(c) 90° of BS60°. Thisisclear from Fig. 3.20(a). The voltage across the valve hasa positive den, 60° D because of the succeeding commutation. a & Ly Ft are FIRST RANGE SECOND RANGE THIRD RANGE (a) (b) (c) Fig. 3.20 Relationship between inverter angles ‘This dent normally occurs (for B < 60°) after the sinusiodal voltage on which itis superposed has become positive. . ‘Hence under this condition & = y. With increased overlap and consequently earlier ignition (firing) of the valve, the dent encroaches on the period in which the valve vol would otherwise be negative [see Fig. 3.20(b)] and this makes <'y. After the front the deat becomes entirely negative, further advance of the dent (Fig. 3.20(c)] does nat decrease commutation margin further. Thus, the margin angle (€) has different relation- ship to depending on the range of operation, which is summarized below. First range: B< 60° bey G56) Mode 2: Second Range: 60°< 8 < 90° . $= 60°-u=y-B-60") on Mode 3: Third Range: B>90° : ~ 30° Mode Ree itis — . tion, it is necessary to maintain a certain minimum margin: &,, This results in 3 submodes of the 1st mode. i istics pe The overall inverter characteristics Mode 5: Mode 1 1(a) B < 60° for values of u < (60° -,) ‘ oe y=%=b Poaus. on The characteristics are linear defined by a V,= cos %-7, G3),* SYsteom Ver, the 5 Which, CMter ig Values ive dent ich it is | earlier volt front joes not elation- (3.56) G57) (3.58) in angle tics are as) ‘Analysis of HVDC Converters 6 1(b) 60° 60° corresponding to > 90° +, ‘The characteristics are again linear but with a different slope and defined by V,= 13 cosy, —3i, c 61) ‘The boundary of the inverter operation for a value of &, = 15° is shown in Fig. 3.19. In the normal operation of the converter J, is in the range of 0.08 to 0.1. Hence, only the first linear portion of the characteristics (both for rectifier and inverter) is of importance in the converter control. 3.5 CHARACTERISTICS OF A TWELVE PULSE CONVERTER ‘As long as the AC voltages at the converter bus remain sinusoidal (with effective filtering), the operation of one bridge is unaffected by the operation of the other bridge. connected in series. In this case, the converter characteristics are as shown in Fig. 3.19 with the assumption that the AC voltages at the converter bus remain constant. The region of rectifier operation can be divided into five modes as follows : Mode 1: 4 and 5 valve conduction 0 mode are elliptical as explained in the previous section, area circuit of the twelve pulse converter is the series combination of The equi’ cuits for the two bridges. This is because the two bridges are i, equivalent ent. ‘side and in parallel on the AC side. The current waveform ( ing see he primary ‘winding of the star/star connected transformer isas; sean inne or asl jis similar to that shown in Fig. 3.6. The current waveform in the prim g 3 SL viv | (a) other t cand the | conduc valve ¢ ‘The inverte 3mode effect ¢ in the) Fig. 3.21 AC current waveform in a 12 Pulse convener side of the star/delta connected transformer is shown in Fig. 3.21 (b). The waveform of the line current injected into the converter bus is shown in Fig. 3.21(c). Incan be show? typical that the Sth, 7th, 17th, 19th harmonics (in general, when h + 12n + 1) produced by Seats two bridges cancel each other. Thus, the lowest AC harmonic tin a 12 pul? converter is of order 11. PresentAnalysis of HVDC Converters 63 Effect of Source Reactance on converter without AC filters ‘When the source reactance is not zero and no AC filters are provided (see Fig. 3.22). the operation of either bridge is affected by the commutation process taking place in the we — et Ott via Cc Fig. 3.22 A 12 Pulse convener unit with source reactance included other bridge. In this case, the operation of the twelve pulse converter is quite complex and there could be two additional modes (i) 5 valve conduction and (ji) 6-7-8-7 valve conduction. Also there could be anew mode of 5-6-7-6 valve conduction (instead of 6 valve conduction), depending on the value of a coupling factor K defined by x Ss rece, x, = ‘The different modes of operation as 4 rectifier are summarized in Table 3.3. The inverter operation has only 4 major modes with 3 sub modes corresponding to the first 3 modes. The various modes for the inverter operation are summarized in Table 3.4. The effect of the common source inductance for the two bridges results in additional dents in the valve voltage waveforms, caused by the commutation in the adjacent bridge. A G62) ofn150?, uat8°, 6 #18" Fig. 323 Valve Voltage 4-5 Valve Conduction Mode of Inversion typical valve voltage waveform is shown in Fig. 3.23 D" and Di are the two additional dents caused by the common source reactance.wo HVDC Power Transmission Sy, thon for rectification ‘Table 33 Summary of different modes of operat : “Analysis of HVDC Conver ion of 1. @ ‘coupling factor _ —_—_—— ® S.No Name of mode oa nr osuce Oskat L ASyate 2 Suave, Osasa, uk Osks1 ay conection F “ b) Sub mode-2 30) S6vahe asa, Reuck Osks.s16 e bbues 30) Sénbe a2, EeucBna+a, si6 Vol transition iss Place whenever the link current starts falling. To avoid Variabj geaemort eee eee eailen, the currentreference at the inverter is kept below that ~ at the rect mount called the ‘cx Sy ee tee of z Leonia, See aie current margin’, This is typically about 10% ‘arying [The power reversal inthe link can take i saa i : Place by the reversal of the DC voltage, This FO] Over is done easily by increasing the delay angle at the station initially operating as the thyristoy, “rectifier, while reducing the delay angle at the station initially operating, as the inverter althou, us, it is necessary to provide both CEA and CC controllers termi gh rollers at both terminals. The on-load tap changer control at the inverter is used ‘mainly to maintain a constant DC voltage (when the inverter is in CEA control). The tap changer control atthe rectifier ulate the isdesigned to maintain the delay angle within certain limits (say 10° to 20°) in order to termi maintain certain voltage margin for the purposes of current control. A voltage margin of rite 3% is generally considered to be adequate to meet any sudden demand for the increase pe mC jn the link current. al the ‘The feedback control of power in a DC link is not desirable for the following reasons: 1, Atlow DC voltages, the currentrequired is excessive to maintain the required level OVves the of power. This can be counter productive because of the excessive requirements € power on the reactive power, which depresses the voltage further. 2. The constant power characteristic contributes to negative damping and degrades reactive dynamic stability, mi 43 CONVERTER CONTROL CHARACTERISTICS °s in the 43.1 Basic Characteristics The basic principles of the control of DC link have been stated in the previous section. ‘The control characteristics of both stations are illustrated in Fig. 4.2 which shows the tion in pavoid nangle tation losses | is the lage 2 5 could Fig. 4.2 Converter controller characteristic DC voltage at the station II versus DC current. Each station characteristic has three parts a as given below: Ci : i Station ‘Type’ > link Sesion yee n constant current pooth be af ed fete SeaHVDC Power Transmission 5 Yen, 80 i i i determines the intersection of the two characteristics (point A) Mode combnat anes i Satin T operating as rectifier with constant current control and Station” a Ee = operating at constant (minimum) extinction angle ; qu 43.2 severe canbe three modes of operation ofthe link (For the same direction of ‘The previo flow) depending on the ceiling voltage ‘of the rectifier which determines the poin, quadrant of th intersection of the ‘two characteristics. ‘These are defined below : of fwoother requ L cc arrecierand CEA a iver Coperaing PO A) which is the normal mag Mode Stabil of operation. 2. With slightdipinthe AC voliage, thepointofintersection drifts toC which imps; peers rminimum ot at rectifier and minimum vy at the inverter. ie intersection o! 3._With Jower AC voltage at the Toor the mode of operation shifts to poi hich implies CC ale verter with minimum, ot at the rectifier. Point 8 istic ab has, generally, more negative slope than characteristi simile vahesot Rand R, This is because of the fact that the slope of ab ah combined resisance (R, +R.) white the slope of fe is due to R,,. However, for = tothe atthe invener, the slope of fe could be more negative than that of ab. it tow icy will be thre control whi this probler the current adequate teleco mmunicati e mai Set ica tO a ag telecommunication vent inadvertent pid transmission of margin requires wectiferopmader nets Power reversal in the tine caren oF Pw of the inverier (100° w 110, oe toprevent then ink due 10 failure of é ‘wit1e9, by Puting minimum ine transition tothe CEA tes its on the de! (see lay ang be greate:lig tB for ae Converter and HVDC System Control 4.3.2 Modification of the Control Characteristics The previous discussion has outlined then \ced to restrict the control region to the first quadrant of the V,—7, plane to avoid unwanted reve iti twootherrequireineatswnic eno: ersalof power. In addition, there are Mode Stabilization : The slope of ab and fe are nearly equal which can lead to iti i l ab and . Poor definition of the intersection of point ‘C’. Further, ithe slope of fe exceeds that of ab (see Fig. 4.4), there 'g2 Ian la Fig. 4.4 Mlustration of 3-point instability will be three possible operating points A, A' and A". This implies instability of the anol which wil result in hunting between different modes of operation. To eliminate this problem, the inverter characteristics are modified and given a Positive slope when the current is between J, .and I, (See Fig. 4.5(a)). Alternate solution is to modify the | ir = - min.¥ lye Igy haa la Fig. 4.5 Modification of inverter characteristics inventer control to. intain a constant DC voltage with back-up control of minimum, CEA (see Fig. 45(0), This requires te normal operating value of extinction angle a be greater than the minimum value.HVDC Power Transmission Sy, ) 82 C x nt limit: ance va enn re nity ein he AC 7 ‘ ™ ate ee ae The low AC oe incre of the overlap angle, In seen commun Faure DECAUSEST ny unl the conditions that lg cases, itis necessary 10 reduce the DE the reduction of current relieves those vaya teefaced DC volage ate ee crcnuous current flow in them, ich are overst : 4 is faults on the rectifier side AC system, ng inverter has . Ifthe low eee frotor causing excessive consumo at reactive Dove, operate at very low i tu! 1o modify charact soos Ts i nits This is shown in Fig. 4.6 which also shows current mits, toinclude voltage dependent ‘ul, 7 faa a2 bn Fig. 4.6 Control characteristics including VDCOL qa Current exror characteristics to stabilize the mode when ‘operating with DC current between I, and I, : ‘The characteristic cc’ and c'c show the limitation of current due to the reduction in voltage. The DC current is reduced from J,, to’, linearly and maintained at/",, below the voltage V.,. The inverter characteristic alsc follows the rectifier characteristic 10 maintain the current margin except for hd” which is due to the lower limit imposed 0 the: delay ‘angle of the inverter, SYSTEM CONTROL HIERARCHY seat Contor functions required for the HVDC link located at one ofthe wes Shown in Fig, 4.7. The master controller fot a bipole is controller (from energy sands with the power order (P..) from the syste rs Converts meee Centre). Italso has other information such as AC voltag’ (40th pole contami see, TM MABE Contller tansmian oe vale oops (conver) The veh Orde King angle rder nthe wav monitoring and fring cg the op Catverer contol also oversees val selectdn logic, commutation falursprvee ee It also ineludes bypass pai sequences, margin switching and vaive peoe cirewes ONL converter starvsto? are performed using the Converter ani converter, the compl voltage ai ‘Y- meas Circuit Va ret~THtre~ aystem Controt Master = anao-e Communication Controt to remote terminal Pole 11 Control VGC - Valve Group Control Fig: 4.7 Hierarchical Control Structure for a DC link The nale control alén inearnaratac nale nentantinn TNO Kina mentantinn and anstnmatConverter and HVDC System Control Communication. to remote terminal Pole II Controt [ise] [ove] [ec] t VGC - Valve Group Control Fig. 4.7 Hierarchical Control Structure for a DC link The pole control also incorporates pole protection, DC line protection and optional converter paralleling and deparalleling sequences. The master controller which oversees the complete bipole includes the functions of frequency control, power modulation, AC voltage and reactive power control and torsional frequency damping control. It also84 HVDC Power Transmission System ‘oversees pole paralleling sequences, transient pole current increases: for outage of a pole, balancing of pole currents and communication of the power/current order to the remote ‘terminal. ‘The block diagram of the pole and converter controls is shown in Fig. 4.8. This shows. the basic control functions. The currentor extinction angle controller generates acontrol signal V, which is related to the firing angle required. The firing angle controller generates gate pulses in response to the control signal V,. The selector picks the smaller Of the 0: determined by the current and CEA. controllers. _4SFIRING ANGLE CONTROL ‘The operation of CC and CEA controllers is closely linked with the method of generation of gate pulses forthe valves in a converter. The following are the two basic requirements for the firing pulse generation of HVDC valves. 1. The firing instant forall the valvesare determined at ground potential and the firing signals sent individual thyristors by lightsignals through fibre-optic cables. The required gate power is made available at the potential of individual thyristor. 2. While a single pulse is adequate to tum-on a thyristor, the gate pulse generator must send a pulse whenever required, if the particular valve is to be kept in a ‘conducting state. This is of importance when operating at low DC currents anda transient might reduce the current below the holding current. ‘There are two basic firing schemes, namely: Individual phase control (IPC) ‘Equidistant pulse control (EPC). IPC was used in the past and has now been replaced by EPC for reasons that will be 4.5,\ Individaal Phase Control _This was used in the early HVDC projects. The main feature of this scheme is that the oa Baise Beneration foreach phase (or valve) is independent ofeach other and the Lepid = rigid! ly synchronized with the commutation voltages. ‘There Ze are wo ways in which this ean be achieved-constant «control and inverse Constant a Control In this scheme (see Fig. 4.9) six timing (commutation) voltages are derived from the Inverse cosit There are 4,10(@)), Sixt From AC line Voltage by 90° a Sum of t Fig. 4.1¢ Control y linearly Shift atoa Converter and HVDC System Control : ba converter AC bus via voltage transformers and the si verter AC bi six gate pulses are generated at aonicallyidea delay times subsequent to the respective voltage zero crossings. The insantofzero crossing ‘of a particular commutation voltage corresponds to a.= 0 for that valve. ays are produced by independent delay circuits and controlled by & common control voltage V derived from the current controllers. Inverse cosine control * There are ‘several variations of this, but in one common arrangement (see Fig. 4.10(@)), six timing voltages, (obtained as in constant ct control) are each phase shifted Voltage Zero Crossing Detector To Valve gate “Ve, Firing instant Fig, 4.10 Inverse cosine control by 90° and added separately to a common control voltage V . The zero crossing of the sum of the two vollages initiates the ring pulse forthe particular valve considered (see Fig. 4.10(b)). The delay angle 0: is nominally proportional to the inverse cosine of the Control voltdge. Italso depends on the AC system voltage amplitudeand shape. The main Advantage of this control scheme is that the average BC voliage seapos Se brncen ates linearly with the control voltage V,. It is essential in this scheme to maintain the phase shift at 90° for variations in the supply frequency.8 HVDC Power Transmission System, 86 Drawbacks of IPC Scheme E ‘The major drawback of IPC scheme is the aggravation of the harmonic stability sit es was encountered particularly in systems with low short circuit ratios (less pee 4). The harmonic instability, unlike instability in control systems, Is a problem that +haract ificati isti ics in steady-state. is i nification of noncharacteristic harmonics in s| . Sealy i t any distortion in the system voltage leads to is is mainly duc to the fuct that Mm ; ee nase crossings which affect the instants of firing pulses in IPC vesseme: Thisimpliesthateven when the fundamental frequency voltage components are balanced, the firing pulses are not equidistant in steady-state. This in tum leads to the jon of noncharacteristic harmonics (harmonics of order h # np * 1) in the AC current which can amplify the harmonic: content of the AC voltage at the converter bus, ‘The problem is aggravated atthe frequencies for which the filler impedance and the it ‘are in parallel resonance. Te cna ane instability can be overcome by the following measures: 1. Influencing the harmonic behaviour of AC network impedance seen by the 7 converter (through the provision of synchronous condensers or additional filters for filtering out noncharacteristic harmonics). % Use of filters in control circuit to filter out noncharacteristic harmonics in the commutation voltages. This can, however, be problematic due to variations in the supply frequency and wiladd to the control delays. 3/ The use of firing angle control independent of the zero crossings of the AC voltages. This is the most attractive solution and leads to the equidistant pulse firing scheme described next. Equidistant Pulse Control (EPC) In this scheme, the firing pulses are generated in steady state at equal intervals of Heals ass Counter. This control scheme was first suggested by Ainsworth [3] as eee oscillator to generate the firing pulses. There are three variations Pulse frequency control (PFC) x Pulse period contro! Pulse phase control (PPC) Pulse Frequency Control (PFC) Inthis scheme, a voltage controlled oscillator (VCO) is used, the fi i is determined by the control voltage V, which is related to the eine eae (current, extinction angle or DC voltage) being regulated. The frequency in. eae ae pearaun is equal of, where, isthe nominal frequency ofthe AC system, PFC system ‘an integral characteristi i rc — res leristic and has tobe used along witha feedback control system for Figure 4.11 shows a simplified block diagram of the PFC s: system (11), controlled oscillator (VCO) consists of an integrator, comparator and: oe Converter ¢ Theoutp instant ( where V In stead Sine in st system is obta:$ 87 Converter and HVDC System Control Fig. 4.11 Block diagram of PFC system ‘The output pulses of the generator drive the ring counter and also reset the: ‘integrator. The instant (1) of the firing pulse is determined from the following equations: ! J ; KV. +V,).a =v, . (4.10) where V, is a bias (constant) voltage and V, is proportional to the system period. In steady-state, V, = 0, and from Eq (4.10) we get EV Gon fev, (4.11) Since 1,-1..= Veh, (4.12) in steady-state, the gain X, of the integrator is chosen as . K,=Ph,V,/V, 413), ‘The circuit shown in Fig. 4.11 does not incorporate frequency correction (when the system frequency deviates from, f,). The frequency corréction according to Ainsworth is obtained by deriving V, as shown in Fig, 4.12. From Fig. 4.12, we have Y=V,/045T,), Ve= KV, 0,-1.3) From pulse generator Fig. 4.12 Frequency-carrection for PFCHVDC Power Transmission Syste, 88 jod Control [4] 4 i Pulse Peri forthe way in which the control volage Vs handle This is similar to PFC extePl. same as that in Fig. 4.11, however, V, is now ‘The structure ¢ Tha the instant , of the pulse generation is given by summed with V, ins pee e Vv, +V, (4.14) 2V, +V, [ anar%s From Eq. (4.14), we get - KV 0,-4.2 Mat Me va 15) With V,=0, the interval between consecutive pulses, in steady state, is exactly equal ‘Figure 4.13 shows how an exponentially decaying V, which appears at =, shit eas by decreasing the pulse frequency between , and /,. Fig. 4.13 Firing pulse train The frequency correction in this scheme is obtained by either updating V, in response to the system frequency variation or including anottier integrator in the CC or CEA controller, Itcan be seen that conceptually, the pulse period contrd1 is similar to PFC and both have integral chanteristics, Ainsworth [11] claims that PFC gives better stability because the control voltage V. is averaged in this scheme, unlike in i 1 voltage V_ ‘i pulse period control. However, ifthe ripple in control voltage V, has harmonic components of frequency less than pf,, they are bound to affect the operation of PFC also, Pulse Phase Control (PPC) This has been suggested by Rumpf and Ranade [6], Reeve and Sevenco [7]. As ‘suggested in (6] an analog circuit is configured to generate firi ead following equation ‘generate firing pulses according to the bn | ; BY, 2h Yea MS (4.16) where V, and V,, , are the control voltages at the instants ¢, and ¢, 1 Fespectively. Converter a For pro} js constam o. control aatition © frequency The mé simiting V inverse © incorpore jimits 21 ontrols 2 intervals iscalled Drawba Althe certain li resultsit phase to breaker: which c to be er EPC oscillat this prc with IF 45.3 ¢ ‘The In the initial Howe archit in ger chant 4.6 C Tl diag T cont and equiConverter and HVDC System Control 89 For proportional current control, the steady state can be reached when the error or V, is constant. To reduce the error to zero, the authors of reference [6] recommend a slow a control with feedback signal taken from the measured delay angle. This signal in addition to a signal derived from frequency error (to compensate for variation in frequency) is used to control the voltage V,. ‘ The major advantages claimed for PPC over PFC are. (Gi) easy inclusion of climits by limiting V_ as in TPC and (ji) linearization of control characteristic by including an inverse cosine function block after the current controller. Actually, limits can also be incorporated into PFC or pulse period control system. In any case, the constraints of limits give rise to generation of nonequidistant pulses, unless special symmetrizing controls are included. This basically involves: firing the subsequent (p-1) valves at equal inervalsonceavaive is fired at, say, minimum limit, untessan increase in the firing angle is called for. Drawbacks of EPC Scheme Although EPC scheme has replaced IPC scheme in modem HVDC projects, it has certain limitations. The first drawback is that under unbalanced voltage conditions /EPC resultsin less DC voltage compared to IPC. Unbalance in the voltages results from single __ phase to ground fault in the AC system which may persist for over 10 cycles due to stuck breakers. Under such conditions, itis desirable to maximize DC power transferin the link which calls for IPC. However, transition from EPC to IPC under the fault conditions has ~ to be evaluated carefully. EPC scheme also results in higher negative damping contribution to torsional oscillations when HVDCis the major transmission link from a thermal station. However, this problem is not so serious as the problem of non-characteristic harmonics associated with IPC. 4.5.3 Control Hardware The implementation of the controls can be achieved using analog or digital circuits. In the later case, micro-processor based controls can be used. .P based controls were initially used where speed of control was not critical, but flexibility was an advantage. However, with the availability of high speed programmable controllers using bit-slice architecture, converter control using these devices is now feasible. 1P based controllers, in general, are more reliable than analog controllers and permit redundancy. The control channel not in use operates in a hot standby mode. 4.6 CURRENT AND EXTINCTION ANGLE CONTROL The current controller (see Fig. 4.8) is invariably of feedback type. A typical block diagram of the controller which is of PI type is shown in Fig. 4.14. . ‘The extinction angle controller can be of predictive type or feedback type ain FG Control, The predictive controller is considered to be less prone to commutation failure and was nao used in the early schemes. The feedback control with PFC type of quidistant pulse control overcomes the problems associated with IPC., al 90 "HVDC Power Transmission Sys—p Tmargin Fig. 4.14 Block diagram of current controller : sea dl : it was felt thar tinction angle, as opposed to current, isa discrete variable and it was felt teedback contol ofganma sles tan the predictive type In one of the prediine schemes proposed by Hingorani [18], the firing pulse generation is based on the following equation bse where e,, is the commutation voltage across valve j and ¢, is the instant of its firing. In general, the prediction of firing angle is based on the equation i B=, (4.18) where u, is the overlap angle of valve j, which is to be predicted based on the current knowledge of the commutation voltage and DC current. In feedback control of extinction angle shown in Fig. 4.8 the measured value of Passes through a non-linear block. The control is made faster when-y < ¥, and slower when > 7, In the former case, the fast control reduces the incidence of commutation failure while inthe later case, the instability due to the negative resistance characteristic Of fast y control is avoided. Under large disturbances such as a sudden dip in the AC voltage, signals derived from the derivative of voltage or DC current aid the advancing of delay angle for fast recovery from commutation failures, Fkstrom and Liss [4] describe a predictive control scheme in conjunction with EPC. , (wt) +2X, 1, 4.17) 4.7 STARTING AND STOPPING OF DC LINK 4.7.1 Energization and Deenergization of a Bridge Consider N series connected bridges at a converter station. Tf one of the bridges is to be taken out of service, there is need to not only block, but bypass the bridge: This is because ofthe fact that just blocking the pulses does not extinguish the currentin the pair of valves that are left conducting at the time of blocking, The continued conduction of this pair injects AC voltage into the link which can give rise to current and voltage comerter and ‘The bypas jvating py active it as it ‘The proce valves Zand given. Wi inthe usua command bridge. TI the valve of the by Fig. 4.16 fakes“he, felt tha, dictive Rae ec ee eee a, Converter and HVDC System Control és Fa Oe oe bridge can be done withthe help ofa separate bypass valve or ty bscaerendsts sealed the bridge (two valves inthe same arm of the bridge). The bess rangi spare Win aves where the possibilty of arc backs makes pce a ses Whe ost fan ex valves AV i tse of bypass pris he The process of desnergization of bridges explained wih reference o Fig. 4.15.The valves ‘assumed to be conducting initially when the blocking command is Fig. 4.15 A converter bridge with isolators given, With the selection of bypass pair 1 and 4, the commutation from valve 2 to 4 is inthe usual manner, but the commutation from valve 3to valve 5 is prevented. Inthe case. of a predetermined choice of the bypass path, the time lapse between the blocking command and the current transfer to bypass path can vary from 60° to 180° for arectifier : bridge. This time can be reduced (from 60° to 120°) ifthe bypass pair is chosen such that the valve with the lower valve number carrying current at the instant of blocking is included in the bypass pair, In the inverter, there is no time lag involved in the activation of the bypass pair. The voltage waveforms for the rectifier and the inverter are shown in Fig. 4.16(a) and (b) respectively. The overlap is neglected here. @) ioinianaenre ns Wide os Fig 4.16 Voltage waveforms during de-energisation for rectifier and inventer bypass pair is shunted to a mechanical switch S,, With the aid Moncon sucnen ‘can now be isolated. The isolator pair S and switch S, are interlocked such that one or both are always closed. ae Theenstguaignofablocked bridge isdone in two stages. The currentis firstdiverted from s, to the bypass pair. For this to happen. S, must generate the required-arc voltage and torhinir ase pis voltage, the circuit inductance mustbe small Incase the bypass pair— Vo" HVDC Power Transmision Sug 92 tn ce ah haieretachpnbrbaie capacity are ued dass switch i a sari ae at vat wks pins insannenusty neglecting ov, However fe wi Serer mete eee . Shown in Fig. 4.17. jee en ne Fig. 4.17 Voluge waveforms during energisation for rectifier and inverter It is to be noted that to avoid operation at high delay or extinction angles, the eenergization of a bridge at the rectifier (or inverter) station is accompanied by the ‘ ‘deenergisation of a bridge at the inverter (or rectifier) station. 4.72 Start-Up of DC Link ‘There are two different start-up procedures depending upon whether the converter Sing controle providesa short gate pulse ora long gate pulse, The Tong gate pulse lasts ‘nearly 120°, the average conduction period of a valve, Start-up with long-pulse firing 1m this case, the curent extinction during the star-up is nota Problem. The starting ‘Sequence in this case is as follows: 1, Deblock inverter at about +y = 90° 2. Deblock rectifier at a = 85° 3. Ramp up voltage by inverte Start-up with short pulse fring Jn this case the problem of current extinction du ing start-up is present je with forward bias is not put into condueti thectmeain har a a luction i : below holding cure’ mn when the current in that transiently falls ‘The starting sequence for this cas is as follows: 1. Open bypass switch at one terminal 2. Deblock that terminal and load to minimum 3. Open bypass switch atthe second terminal a to establish low direct curent * control and the current by rectifier control, Current in the rectifier mode ore nd commutate current to the bypass 4. Start the second terminal also in 5. The inverter terminal is Put into 6. Ramp up voltage and current the rectifier mode the inversion mode voc Power Transni ‘The voltage is no of the line to be chee 48 POWER CON Figure 4.18 sho eee ‘TC-Telecommuni OS-Order Setting ‘on the current o1 -{VDCOL). The 0 full current for k delay is given to | would otherwise. By providing ‘Power order from the fastest respon tion requirement. DC line voltage 1 line voltage thus referring th vin Vhemthe DC ine is very long Cannot be Using acurrent ther substationhata are yon? oe HVDC Power Transmission System “ The voltage is normally raised x SS Pair of the line to be checked before vale ting the current. This permits the insulation Ver, fash on the generator shaft. The switching Fee The ramping of power avoids stresses hig 3f thg ‘The required power ramping rate nee i the line are also reduced. Case are 8 require fast restoration of en the atrengih ofthe AC system. Weaker Unfortunately, high ramping rates in such for maintaining transient stability. to the requirement of reactive power at the conven 8 © larse voltage drops due . The permi n ipa emt ia aap rane gms gles, the ‘d by the : os HH = be bs ar A esi Ise lasts ac of other Control Signal ‘TC-Telecommunication equipment VDCL-Voltage imi m — sge dependent current limiter tarting _ | Fig. 4.18 Power and auxiliary controller on the current order are modified by the voltage dependent current order limiter AVDCOL). The objective of VDCOL is to prevent individual thyristors from carrying full current for long periods during commutation failures. However, sufficient time delay is given to prevent the action of VDCOL during normal ac system faults, which Would otherwise drastically reduce DC power. By providing both converter stations with dividing circuits and transmitting the Power order from the leading station in which the power order is set tothe trailing station, a the fastest response to DC line voltage changes is obtained without undue communica- eae tion requirement. To get equal calculated currentorders in the twostations, the measured DC line voltage must be referred to the same point on the DC line by compensation for i line voltage drop. This is done by adding the term +R J, to the measured voltage, lus referring the voltage to a common reference point. Whea he DC! re cae is largeand varies considerably,¢.g., when the overhead pass line is very long and exposed to large temperature variations, the DC line voltage drop Cannot be, individually in the two stations. This problem can be solved by Using a current order calculated in one substation only and transmitting its output to the ther substation,is HVDC Power Transmission Syuq, perwer responding rate of change of order are set with thum noe by the: ee Gasol tacanin the lead station, which may be the in . or rectifier station. The order setting may also be transmitted there by a remote cont) link from the energy control centre. The order setting unit which is a digital unit baseq on integrated logic circuitry is intended mainly for performing power stepping nq synchronization of order stepping in the two stations with the help of the communication link. The order setting unit in the trail station accepts a power order from the telecom, munication system and has a supervisory function for the power order transmission. The received power s both parity checked and checked with regard to allowed change from cone order message to the next. Ifthe received order message is accepted, a receipt signal issentback tothe lead station to allow transmission of new order messages. If thereceny signal is not generated and accordingly not received in the lead station, order stepping is interrupted and the latest power order is repeatedly transmitted until accepted inthe trail station at which, order stepping can start again. Thus, power transmission interrup. tions as a consequence of telecommunication failures are avoided. ‘The use of programmable logic for realization of power flow control function seems to be most appropriate solution for future applications. A standardized computer hardware equipment can be used and the functions may be adapted to- individual ‘Fequirement. 49 HIGHER LEVEL CONTROLLERS ‘The power in a DC link can also be controlled in response to the quantities derived from the AC system (such as frequency) in order to improve the security of the over-all system. This is achieved by an additional power (or current) order derived from an emergency power controller or auxiliary controller. The functions of such higher level controllers are discussed below: 49.1 Frequency and power/frequency control ‘The frequency control can be used in the case of (i) isolated load and (ji) isolated ‘Seneration. Inthe later case, the objective is to improve the damping and reduce the wear of the generator and governor system, ‘The nuclear power stations are very sensitive to output power and frequency fluctuations, By proper control of the power carried on the DC link in a hybrid Eansmission system consisting of both HVDC and UHV lines, the frequency variations both during steady-state and transient conditions can be minimized, The Prospects of a ccorteaie Control between nuclear station and HVDC system has been investigated in Japan. ‘When the DC link is used as a tie between two power systems, the frequency bias can bbe used to adjust the power flow over the tie to assist the system in difficulty. It is to be noted that an HVDC link has no inherent sensitivity to system frequency unless it is deliberately introduced in the control ‘system. Without it, a constant power flow can overspeed a receiving system which has become separated from part or all of its load, even if governors cut-off all Primary energy inflow to local generation. Similarly, a sending system can be broughteventually toahalt, if more power is required from it than the connected generation can produce. Thus, although it would first appear that there is little benefit in the frequency control for a DC link between two systems sO power stabil ‘ADC tie us oscil: ‘AC tie line pov tie can also Prc and can thus p 49.3. Emerge In an AC tie it interconnect toallow the dis for both or to t infeeds on whi AHVDC ii disturbances 0 is available to healthy systen the link. In ger be shared in a can be dampec amount of DC with the DC li Tesults can be 49.4 Reactiv Thereactiv the dynamic o allowing the i Teceiving end 49.5 Subsyn ‘damp low f ‘Sea Totor veloSion om enn ie ith “hum, Converter and HVDC System Controt e he iny, i atthe link cannot be e; i i ot con Jarge th rpc fee | influence the frequency of ether, itis prudent unit 1 to incorporate a 'Y Control which in the event of a partial break-up prt based ofa system, will prevent overspeed of underspeed, Pping nunicaget 492. Stabilization of AC Ties e son an tick a th Weak AC esto neighbouring systems, power varied if oie fom maintain stability if one of the AC ti trips, All 19 balance the Load flows and Signa ‘ADC tie used in parallel with an AC te can be em, : ae ployed to damp the low fr he receing ine owes Coen te, THe contol signal usedcan be the mus ct hence Stepping ‘AC te lne power (or current) or the phase angle difference acroes the nes tie. The DC ted in ne Fa rere tueneY Control for one endif the AC tie becomes discon ected interrupy. and can thus permit resynchronization of the AC tie. i ons 49.3. Emergency Control computer Inan AC tie line, the power flow is determined by conditions in the systems which dividual it interconnects. If one of the systems suffers a disturbance there are only two options; infos on which itis dependent even before the disturbance. asi AHYDC linkon the other hand, even with simplestcontols, buffers one system from eee disturbances on the other. Power flow can continue at worst unchanged but the option fences isavailable to vary power flow to assist the system in trouble to the extent to which the ea healthy system can allow without puting itsefin difficulty, subject only tothe rating of tte link. In general, with suitable control, a disturbance originating in ether system ean be shared in a predetermined manner and the oscillations occurring in the two systems ¢abe damped simultaneously. Substantial damping can be achieved with a very small ‘apount of DC power modulation. However, if a large degree of modulation is required isolated With the DC line already operating close to its full capacity, itis found that significant he wear results can be achieved simply by reduction of DC power at the appropriate instants. quency 494 Reactive Power Control hybrid The reactive power control isimportant, particularly in weak AC systems in reducing riations the dynamic overvoltages. Also in inverters, the fast reactive power control can help in Allowing the injection of increased power at times of need to improve the stability of the ‘ceiving end AC system. 49.5 Subsynchronous Damping Control A radial HVDC fink connected to a thermal generating station can contribute to the "tative damping of the torsional oscillations at subsynchronous frequency due to the snletactions with the current controller. This problem usually surfaces when there, ‘areno Parallel AC links, The power modulation controller mentioned earlier which is. designed todamp low frequency rotor oscillations can aggravate the problem. However, asuitably designed subsynchronous damping controller (SSDC) with contol signal derived from the roto velocity can help to damp torsional frequency oscillations. Ste3 & Ge SURREAL OG EEA HVDC Power Transmission Syste, : s 4.10 TELECOMMUNICATION REQUIREMENTS i ication is possible, th ing with voice communication is Possible, the sia mma et i commen he utilization of ‘inherent: : . : enna 1, a 1200 baud channel is required for power flow oe ping inerva ‘of 20 msec with a resolution of 0.05 percent of . This gives onde, which should be adequate for mostcases, oo aan cance of proper current margin is essential andi the primary objective of ‘The maintena telecommunicati ided, Certain precautions can help in overcoming ibe ni oe fre of eommurication, In raising the current order, i seo eabl wo cary it out first atthe rectifier and then atthe inverter. In reducing the deter tne inverter curent reference is reduced before that atthe rectifier, a References and Biblngraphy ; cof a coordinated control system for BWR nuclear power plant and 1 bt intaiatn uc IEEE Trans on Power Delivery, Va, PWRD-1,No3.July 1986 pp 319. fe ee er ee ee see cea ‘Vol 114, pp 949-957, 1967. : ‘3. LD. Ainsworth, “The phase locked oscillator - a new control system for controlled static converters’ IEEE Trans, Val. PAS. 87, No.3, pp 859-865, March 1968, 4. A Ekstrom and Liss. A refined HVDC control 3ystem’, IEEE Trans. Vol. PAS. 89, pp 728-732, May/ June 1970. ‘5S. L Amillaga and Galanos G. ‘Theoretical basis of a digital method of grid control for HVDC converters’, IEEE Trans, Vol. PAS. 89, No.8. pp 2049-2055, Nov/Dec. 1970. 6 E Rump and Renade S. ‘Comparison of suitable control systems for HVDC stations connected to weak ‘AC symems’, Pan -New control systems, ar II - Operational Behaviour of HVDC Transmission TERE Trams., Val. PAS.91 pp 549-564, March/April 1972, 7. JReeve and JA. Sevenco, ‘An automatic control ‘scheme for HVDC transmission using digital é ‘techniques’, Part I - Principles of Operation, IEEE Trans. Vol. PAS.92, pp. 2319-2324, 1973. Praberts, Rana, S. Rumpf E. and Wess, T. ‘Control systems for Cabora bassa HVDC . Design criteria simolator sudies’, CIGRE Proc., Paper 14-06, 1976. : _— » { Mehichal, Ranade, 7. Ring H. and Meyl,D.“Panllel operation of Nelson Ri HVDC bipoles 1 and 2 conteol system simulator studi ', Paper81 SM 371. Su i Portland, Oregon, July 1981, Pape 1-4, Presented at IEBE PES Summer Meeting, 10. 5, Reeve snd Geisbrecht W.J. ‘Microprocessor control for HVDC converters’, Proc. of Int. Conf. on a See ‘i 1981 tnd variable static equipment for AC and DC transmission” (IEE 205), pp 186-189, Dec. +s80r based control circuit for a 3, Aug. 1986, pp 310317. h simple program language for mission, Monte, June 1984, pp "HVDC transnias . = 1,No, 2, Api 1986, pp 254-268, n°! AMPMEN with high reid, IEEE Trane., Vol. PWRD- 8 wick, A is ‘ements IEEE Tr, Vo PAS.AT. Meh iS pe aren BIE Onkol for ACIDC sate 5. ( 5. INTROD\ ‘Asin ACS) ipment eas lightni protected by transmission i Apart from cause the stres station, the va damage cause excessive los: In this chay described. Th seen that the 62CONVE 5.2.1 Genera There are 1 Faults: @ Ar Gi) Ar Gii) Mi (iv) Qu 2 Comm 3 Short« The arc b: temporar Teverse direc 1S a nonself incidence of2, May/ RE 5. Converter Faults and Protection 5.1 INTRODUCTION Asin AC systems, the faults in a DC system are caused by (i) the malfunctioning of the equipment and controllers and (i) the failure of insulation ‘caused by extemal sources suchas lightning, pollution, etc. The faults have to be detected and the system has to be protected by switching and control action such that the disruption in the power transmission is minimized, Apart from disrupting the normal Operation, the various faults that can occur also cause the stressing of the equipment due to: Overcurrents and overvoltages. Inaconverter station, the valves are the most critical equipment which need to be Protected against damage caused by the rise in the junction! temperature of thyristors, which is caused by excessive losses in the device and sensitivity to overvoliages. Inthis chapter, the faults Occurring in a converter station, their causes and. effects are described. The protection against overcurrents and overvoltages is also discussed. It is seen that the converter control plays a major role in protecting the equipment. G2 CONVERTER FAULTS 5.2.1 General There are three basic types of faulis that can occur in converters as given below: 1 Faults due to malfunctions of valves and controllers (i) Are backs (or back fire) in mercury arc valves i) Arc through (fire through) iii) Misfire : (iv) Quenching or current extinction a converter station ‘The are back is the failure of the valve to block in the reverse direction and results in the temporary destruction of the rectifying property of the valve due to conduction in the evetsedirection, This isamajor faultin mercury arc valves andis ofrandom nature, This isa Tonself clearing fault and results in severe stresses on transformer windings as the incidence backs i: . Rien thyristors don’t suer from are backs which has led to the exclusion of Mercury arc valves from modern converter stations. Hence this fault will not be Somectthe converter faults such as commutation failure, arc through and misfire are Self clearing if the causes that led to these faults are of transient nature. However, they Can still causea: major disturbance unless the system including the controllers is; Properly designed.98 HVDC Power Transmision sya, 5,22Commutation Failure Because of the turn-off time requirements of thyristors, there is need to mainiin , minimum value of the extinction angle defined by y= 180-a-0 Jon voliage and the Dc Gy overlap is a function of the commutation voltage and the eae ee ee the current or both can result in an comet in the overlap angle which can result in Y stage crm se (0 \etoge sera vate 8 Figure 5.1 Voluge waveforms fora single commutation faiue «= 212 BS P té The failure of two successive commutations in the same cycle, is called “double 2ommutation failure’. If the commutation failure occurs when valve 4 is fired also, the valves 1 and 2 are leftin the conducting state until the instant in the nextcycle when valve REC WARANGAL CENTRAL LIBRARY t Acc.No.119576 oo Vee eee Fig. 5.2 Bridge voltage waveforms for a double commutation failurea 100 ~ VDC Power Transmission Sys1¢y, 3 will be fired. The bridge voltage waveform can be seen that the double commutation commutation failure. ‘The following are the effects of a single commutati 1. The bridge voltage remains zero for a period exceeding ‘fof acycle, during which conducting. ‘The recovery. from 1. The response of the gamma controller at the inverter 2 The current control in the link and 3. The magnitude of the AC voltage If, on detection of a commutation failure, the angle of advance (B) is increased, there is.a good chance that subsequent commutation failures may be averted. However, this also depends upon the control of DC currentand the magnitude of AC voltage. The initial rate of rise of current in the inverter is limited by the smoothing reactor and the current controller at the rectifier helps to limit the current in the case of persistent commutation failure. Itmay beeven necessary toreduce the currentreference to limitthe overlap angle in the case of low voltages caused by faults in the AC system. While in most cases, commutation failures are self clearing, in the case of persistent ‘commutation failures, the converter differential protection helps to take the converter out of service. This protection is based on the comparison of DC and the valve side AC currents. During commutation failures when the two valves in an arm of a bridge are left conducting, the AC current goes to zero while the DC current continues to flow. ‘The commutation failure in a bridge can lead to consequential commutation failures inthe Series connected bridges unless the rate of rise of current is sufficiently limited by the series connected smoothing reactors. SE Are Through ‘This isa fault likely to occur mainly at the inverter station, where the valve voltages are positive most of the time (when they are not conducting). A malfunction in the gale pulse generator or the arrival of a spurious pulse can fire a valve which is not supposed toconduct, butis forward biased. Forexample, in a bridge, when valve 1 has successfully pane! itscurrenttovalve 3, the initial voltage across itis negative (for the duration of the extinction angle) and then becomes positive, If valve 1 is fired at this time, the current will transfer back o valve 1 from valve3, The effects of an arc through are similat to that of a commutation failure - the voltage across the bridge falls as valve 4 is fired (with valve 1 conducting) and the AC current goes to zero when valve 2 current. goes 10 Zero. The firing of valve 5 is unsuccessful and the bridge recovers to normal operation after valve 6 is fired and the subsequent firings are according to the normal sequence- Thus a single arc through is also self-clearing if ri * ‘The arc throughs in thyristor valves can . Sn hs occur di ions i 0} ‘system, the probability of which is Fe a ee ; very small. Anyway, th i i sistent arc throughs is also through the converter differentia!’ aes chee for this case is shown in Figure 5.2 ang i failure is more severe than the single ion failure a commutation failure depends on the following factors: converter Faults an ‘. fisfire while an arc t yrs when the re “while misfire in the latter case. average bridge vo results in large ct damped oscillato extinguish and re voltage and curre ‘The overvoltages (20 Wenge M71 900) 2 eeConverter Faults and Protection, . 101 speaiire While an arc through is caused by the presence of an unwanted gate pulse, misfire ‘occurs when the required gate pulse is missing and the incoming valve is unable to fire. ‘The probability ‘of the occurrence of misfire is very small in modern converter stations because of duplicated converter controls, monitoring and protective firing of valves. While misfire can occur in rectifier or inverter Stations, the effects are more severe in the latter case. This is due to the fact that in inverters, persistent misfire leads to the average bridge voltage going to zero, while an AC voltage is injected into the link. This results in large current and voltage oscillations in the DC fink as it presents a lightly damped oscillatory circuit viewed from’ tne converter. The DC current may even extinguish and result in large overvoltages across the valves. The waveforms of the DC voltage and current in the link for persistent misfire in the inverter are shcwn in Fig. 5.3. ‘The overvoltages can be controlled by controller modifications. . SAP aANT Shir’ ccc aar eA aa rT ee ew $ * tana ena. 778-08) Fig. 13 DC volige and curent for persistent misfireSS 102 HVDC Power Transmission Sy, The effects of a single misfire are similar to those of commutation failure through. When valve 3 in a bridge misfires, the valves 1 and 2 are left conduct; valve 4 is fired, However, at the end of the cycle, the normal sequence of valve Rol is restored. Thus a single misfire is also self-clearing. Ting, 528 Current Extinction The extinction of curent can occur in a valve if the current through it falls below holding current. This can arise at low values of the bridge currents when any tray can lead to current extinction. The current extinciion can result in Overvoltages scr, the valve due to current chopping in an oscillatory circuit formed by the SMoothi reactor and the DC line capacitance. ™ ‘The problem of current extinction is more severe in the case of short pulse fj method discussed in the last chapter. However, in modern converter stations, the Fetun, Pulses coming from thyristor levels to the valve group control, indicate the build upg voliage across the thyristors and initiate fresh firing pulses when the valve is suppoy tobe conducting. It may happen thata number of firing pulses may be generated during acycle when the link current is low. §34Sion Circuit in a Bridge ‘This fault also has very low probability as the valves are kept iri a valve hall withair Conditioning. However, bushing flashover can lead to a short circuit across, the bridge and produce large current peaks in the valves that are conducting. ~ The short circuit currents are significant only in rectifier bridges. The worst caseis ‘when the short circuit occurs at the instant of firing a valve at o.= 0. Assuming that the isno inductance in series with the bridge, the peak short circuit current Giyeu) is given by i i i par, (i+sinn/p) (p/3)-1 [Ff (52) as 3 Sap? B (4) ’ where P = pulse number of the converter (6 or 12) i,,= the de current at the instant of firing the valve 3) 1,= V2 E,/0X,) 6 For a six pulse converter, the peak current is i 1 - pe = 2234, + iy] ‘The bridge voltage and current waveform are shown in fig.5.4. In Eq. (5.2), the effect of network impedance in limiting the current is neglc ab ‘The maximum peak current in a valve results when itis conducting into a valve it Forexample, the maximum currentin valve 3, when itstarts conducting with short across valve 1, is given by J pat 69 = (+008) + (1, /21,) , (54) Converter Faul ‘The peak c valves must hi by blocking tt state provided voltage, addit the AC break ‘The detecti and DC curre: increase. \53 PROTEC ‘The overcu in AC system: i) selectivity i ‘The main 1 controller acti Control. The s and the conver ‘groups such t Broup (or brid Consider a Pole). The prc Protection agi ‘The basic | Provided by v the valve side the smoothing and fast detec Overcurrent re Protection to esta action Pole d Otherwise detConverter Faults and Protection 103 fa) ' t T ' ' ' ' ' M0 ty! instant of fauit initiation Fig, 5.4 Bridge voltage and current waveform during short circuit ‘The peak currents are of the order of 10 to 12 times is the rated current and the th valves must have surge current ratings above this value. The fault. clearing is performed by blocking the pulses when the fault current B0es to zero, the valve assumes blocking same provided the evoltage across itis not high. Ifthe valve is unable to block the forward , inal loops of overcurrents result and this i ‘ppit value, : is can be avoided only by tripping ‘The detection of bridge or valve short circuit is also performed by compari d paring the AC and DC currents. In this case, the DC current goes to zero while AC current tends to increase. overvoliages, for example, a fault between the valve bridge and the converter transformer. The firing of bypass pairs or closing of the bypass Switch across one converter generates overvoltages across the remaining converters, ‘The energization of the DC line from the rectifier side with the remote terminal blocked can cause high overvoltages at the inverter which is open ended. Such events must be avoided by deblocking the inverter first and limiting the rate of decrease of the delay angle. 5.5 SURGE ARRESTERS In the initial stages of application of DC technology, DC surge arresters were not available, the valves were protected by the spark gaps connected across them. Later, with the development of active spark gaps, it was possible to extinguish the arrester current withoutexceeding the protective level and DC arresters were made of nonlinear resistors in series with the active spark gaps. With the development of metal oxide resistors with high nonlinearity, the need fora Series gap has disappeared and the present DC arresters are gapless arresters. The metal oxide elements were frst applied in AC arresters in 1976, Comprising primarily of zinc oxide, but contain Se Converter Faults ae seis seen that the tem sats, but beCOmes PX curr Tye the zinc oxide 10 OP ility of the mate ility of term stabil and proces composition sof the Povervoltages i dyna dination. Howe vey losses is essentia ener sed by the crack e pable of absorbing ar OV). F MeO rany DC applic inadequate and multiph selecting discs such th current. The maximum than 0.5%. 56 PROTECTION A 5.6.1 General ‘The basic principle: systems. These are giv 1 The overvoltage limited at all tir arresters must b 2 Self restoring in no danger to the 3. The operation of Frequent discha Thisimpliestha ‘Operating voltas There must be p different pants insulation, the 1 The overvoltages § “tS On the AC sii finest? DC bus and Aitectly protected by 562 Overvoltage P - i ‘ypical arran; alt: 5.7. For a 7 ti * Per pole. Th SS hich vary. witvith rent lors or a inc a is- ng nis he ra Cowerter Faults and Protection 107 ris seen that the temperature coefficient of the material is slightly negative at low curents, but becomes Positive at currents above a few amperes, This ake it possible a ¢ the zinc oxide | elements in parallel to discharge high energy surges. The long a salty ofthe mate! a ‘satisfactory although itis, influenced considerably by disc re properties ofthe ‘Material are such that itis possible to design arresters to control dynamic overvoltages in addition to Switching surges, This Fesults in economic insula- tion coordination. However, Proper design Of the arrester based on the evaluation of the energy losses is essential. The ultimate limiton theenergy dissipation capability of adisc js imposed by the cracking of the disc under thermal shock, A single column arrester is apable of ‘absorbing around 7 kJ per kV of the maximum Continuous operating voltage NV). In many DC applications, the energy capability of a single column of discs is inadequate and multiple columns are used. A parallel column arrester is made up by selecting dises such that the voltage for each column is the same at a predetermined current. The maximum difference in the currents of parallel columns can be made less than 0.5%. 56 PROTECTION AGAINST OVERVOLTAGES 56.1 General ‘The basic principles of overvoliages protection is the same in DC systems as in AC systems. These are given below: 1 The overvoltage stresses in equipment with non self-restoring insulation must be limited at all times by providing surge arresters. The protection level of the amresters must be lower than the breakdown voltage of the insulation. 2 Self restoring insulation such as air may be allowed to breakdown where there is 0 danger to the safety of the personnel. 3 The operation of surge arresters or flashover of air insulation mustnot be frequent. Frequent discharges of arresters may damage them. This implies that the protective level of arresters must be higher than the maximum Operating voltage in the system. 2 @ ‘There must be proper coordination of the insulation and overvoltage protection in different parts of the system, taking into account the characteristics of the insulation, the nature of overvoltages, etc. The overvoltages generated on the AC side should, as far as possible, be limited by "ers on the AC side. The overvoltages generated on the DC side must be limited by PC tne, DC bus and neutral bus arrester. The critical components such as valves are directly protected by arresters connected close to the components. 42 Overvottage Protection in a Converter Station ‘Ypical arrangement of surge arresters in a converter station (for a pole) is shown ait. 57. Por a niece whens} 12-pulse conventers per pole, there are about 40 ers Per pole, The arresters are selected with adequate energy dissipation capabili- "=! Which vary with the location of the aresters. For example, the valve arresicrVDC Power Transmission Syyg, 108 be subjected to hi, tial can ‘gl , highest pore” the val sng the commutation group a NE is between the valve and protecting the when a ground fault OCS Te ortho lineanau® energies than other arresters converter transformer in the upper DC filter. wre ott f 5 cant 2th om a Oy) = oeiter sooner bY, & vis; eae 5 a wear ps tom, Sn = Fig. 5.7 Typical arrangement of surge arrestors for a converter pole (Source : Reference 9) The closing of a bypass switch across a converter results in increasing the DC volag® ae the. cee ‘converter, The converter unit arrester is stressed in such ee Protective firing of a valve is the back i is available, mages inte forward crcoien, ‘up protection that is avail References and Bibliography (1) J. Kauferte and Povh D,“ : dates mes ‘Conca of overvoltage snd overcurrent protection of HVDC conver 21 Kutodand Pov Resovey of HVDC namin afer inhe AC yon EEC rae meine friable Static Equipment for AC and DC Transmission | (3] J.P. Bowles, ‘Overvoliages in HVDC transmissic at de ci TEE Te Va PASS enue OY ome A (4) D.B. Giesner and Arrillaga J. ‘Behaviour of is : consis" Proc. IBE, Vol 119, No), 1972, pp 209.218" Ow Unk under unbalanced AC fault (5) EW. Kimbark, “Transient overvohages caused by bipolar DC" a 5 ‘monopolar ground ‘Toeory and simulation’ IEEE Trans, Vol. PAS-89, 1970, pp sen Converter Faults and P NG. Hingorai, (8 iid, 1970, pp 57 (7) A Cleric and Tas ‘Conf. Publication {8} DJ. Melvold, Od ‘monopolar line fa {9} CIGRE Study Co HVDC convener fo) HLA. Peterson, Ph (10) res IEEE (11) J-Reeve and Kay ‘Trans. Vol. PAS 12) CV. Thio field tests’ ‘Winnipeg, 1980 (13) G.D. Breuer, K coordination’, E“1 on hi, co en the yard fo Be of , ‘lve ctor ie Converter Faults and Protection 2 109 ; : faults. (6) NG. ingore “Tansient overvohages on bipolar HVDC overhead lines caused by DC line f (6 ia, 1970 pp 397-610 vs = Fee a ratini A. "Transient overvoluages caused by earth faut on bipolar DC lines’, EE + on HVDCIAC Power Transmission, Nov. 1973, pp 196-200. : a) Dd. Melvoli Odam PC. and Vithayatil, 1, Transient overvltages on HVDC bipolar ine daring 1a onopolr ie feck THER Yon P4896, No.2, MarVAprl 197, pp 381-01. [9] CIGRE Study Committee No. 33, ‘Application guide for insulation coordination and protection of HVDC convener stations 33.83 (SC 03-21, WD). 10] HA. PeterionPhadke A.G. and ReitanD.K.,Transientsin EHV DC power systems, PartI-Rectifier ‘currents’, IEEE Trans. Vol, PAS. 18, No. 7, July. 1969, pp 981-989. {11 J. Reeve and Kapoor S., Analysis of transient short circuit currents in HVDC power systems", IEEE ‘Trans. Vol. PAS-90, No. 3, 1971, pp 1174-1182, 112] CV. Thiootal. Switching overvoltages onthe Nelon River HVDC system - Studies, experience and field tests’, IEEE Conf, on ‘Overvol luges and Compensation on Integrated AC-DC systems", Winnipeg, 1980, pp 15-26. [13] GD. Brever, Krege 1.8. and Shakshang E.C., “Zine-oxide overvollage control and insulation ‘coordination’, Electric Forum, Vol. 9, No. 3, 1984, pp $0.55.