The Baltic Ring Study Transmission System Simulations in the Baltic Sea Region

Transcription

1 The Baltic Ring Study Transmission System Simulations in the Baltic Sea Region Dr. Hans Knudsen NESA, Denmark Lars-Ove Ellus Vattenfall Transmission AB, Sweden Mikko Koskinen Fingrid, Finland Dr. Matthias Luther PreussenElektra Netz, Germany Georg Schröder PreussenElektra, Germany Abstract - In the Baltic Ring Study load flow and dynamic investigations were performed for the three synchronous systems UCPTE/CENTREL, NORDEL, and IPS/UPS surrounding the Baltic Sea. A combined network model - including DC links between the separate synchronous areas - was modeled and a number of development alternatives up to the year 2010 have been investigated. Among these alternatives are a synchronization scenario of the Baltic countries to the UCPTE/CENTREL system and a MTDC link connecting Russia and Germany. Keywords: Baltic Ring, Baltic Sea Region, Trans- European Networks (TEN), steady state and transient stability, eigenvalue analysis, MTDC system, interconnection variants between IPS Baltics/UPS and UCPTE/CENTREL. I. INTRODUCTION The Baltic Ring study was performed in the period as a co-operation project between 18 electric power companies from the 11 countries in the Baltic Sea Region (BSR). The existing and planned or discussed links in the BSR are shown in Fig. 1. The study was partly financed by the European Union as a Trans European Network (TEN) project, and the purpose of the study was to investigate the possibilities to form a common energy market in the BSR. Since the main topic of this study was exchange of energy, the main focus was of course on the production system. Information regarding production costs, loads, transmission capacities and losses have been collected, and this information was used to determine the economically optimal production schedules. In this way the trade potential in the BSR was estimated. II. POWER SYSTEM ANALYSIS Detailed power system simulations were carried out in parallel to the optimization of the energy exchange. The major tasks of this work were: to determine the actual power transmission capacities over the tie lines in the BSR considering both static and dynamic conditions, to determine approximate costs for the expected reinforcements in the transmission system, and to determine necessary additional reinforcements as well as the costs for obtaining given transmission capacities necessary to realize an optimal production schedule. By comparing these costs with the economical gain that a transmission capacity would create in an optimal production schedule it can be determined whether it is good economy to create these transmission capacities. III. NETWORK MODELS Detailed equivalent network models from all involved countries were collected and verified for a winter and a summer load situation in the stages 2000 and 2010; altogether 4 different models. These models were joined together into big models of the entire BSR. Equivalents of remote areas in the UCPTE/CENTREL and in the UPS were also included to ensure correct dynamic behavior. The network models represent today s network plus the expected network reinforcements in the time horizon up to Each model consists of approximately 750 nodes, 250 generators, and - depending on the stage - up to 12 HVDC links, spanning over 11 countries and 3 different synchronous areas; see Fig. 2. NO 35/ 12 DE PE / VEAG PE/VEAG 135/ 33 SE 33/ 16 50/ 16 76/ 26 22/ 14 LT 33/ 16 60/ 19 37/ 13 88/ 36 89/ 13 51/ 33 extended UCPTE NORDEL Fig. 1: The European interconnected systems in the Baltic Sea Region (BSR). total number of nodes/generators: 701/256 IPS/UPS Fig. 2: BSR simulation model with nodes/generators.

2 It is to be noted that it is the first time ever that a joint network model of the entire BSR has been made. IV. ANALYSIS OF PRESENT DEVELOPMENT ANS Based on the known transfer capacities in today s transmission system and evaluating the effects of expected new transmission lines the maximum transfer capacities in the BSR in 2010 were estimated to be as shown in Fig x600 N 1040 T (W) 2000 D PE T T, (E) D VEAG SW T 2000 T 1400 KA VS 600 VS 600 o.r. Nordel IPS CIS T = Thermal Limit winter limit summer limit in MW UCPTE CENTREL IPS Baltija = Dynamic Stability VS = Voltage Stability, Voltage Collapse o.r. = other reasons LT 2400 T 70 o.r. 160 o.r T 800 VS 1200 VS 900 T VS 2200 T 50 o.r VS 2000 VS 1000 VS Fig. 3: Transmission Capacities 2010 Kolenergo Lenenergo North-West Region of Russia Pskovenergo Using these transfer capacity limits the Baltic Ring Study project could determine the optimal production schedules over the entire year. It is obviously impossible to check whether each and every load flow over the year is actually stable, considering both static and dynamic stability. Therefore two snapshots were taken out for closer examination for both stage 2000 and stage These were a winter and a summer load situation. These two situations were considered to cover the extremes, so that if there are no stability problems in these two situations then it is likely that there in general will be no stability problems for the proposed production schedules. The considered power exchange for the winter load 2010 are shown in Fig. 4. The power surplus or deficit in MW in each region is indicated for each region. The main objective of the load flow analysis was to evaluate the power systems according to: the fulfilling of the (n-1) criterion, and the sensitivity of the power systems against voltage drops N SE E (W) KA LT PE VEAG Nordel 2150 North-West Region of Russia Fig. 4: Considered Power Exchange for Winter Peak Load 2010 IPS CIS IPS Baltija All parties in the study agreed on common criteria to define static stability. These were that in load flow simulations under single fault conditions no lines should be loaded higher than 120%, nor should any substation voltages deviate more than ±10% compared to the prefault voltage. Nodes and lines in the BSR exceeding these limits have to be checked by more detailed investigations. Under no-fault conditions, the voltage profile should be trimmed in such a way that all substations are within 5% of its rated voltage and all generator voltages are within 1 p.u.±0.05. The assumed generation schedule, the transmission network topology and the initial voltage profile for the exchanges given in Fig. 4 have been verified by all partners concerned. According to the above given methodology, no single fault condition has been detected which would violate the static stability of the system, assuming additional reactive power compensation in Belarus to compensate for a reduced use of the installed generation capacity there, where this reduced use was a consequence of the economical optimization. The intention of the dynamic investigations of the present development plans was to illustrate the static stability of the overall interconnected power systems around the Baltic Sea and especially the damping of inter-area oscillations. It has been checked whether or not the economic power exchanges in the present development plans can be realized and, if not, what kind of reinforcements will become necessary. In addition, detailed investigations including different load flow scenarios have been carried out for the IPS/UPS system.

3 NO 10 DE 13 SW LT Fig. 5: Investigated Fault Locations for the present development plans To cover the whole Baltic Sea Region, temporary 3-phase line faults (fault clearing time 100 or 150 ms depending on the system) with successful fast auto-reclosing have been simulated at different locations for all study cases, see Fig. 5. No extreme test faults have been applied because the main interest is static stability and because some parts of the system are represented with equivalents. In the IPS/UPS system additional transient stability studies were carried out for a 2-phase short circuits on a 750 kv line without auto-reclosing, for 3-phase short circuits on 330 kv interconnecting lines and for a trip of a 1000 MW generation unit after a single-phase short circuit on a 750 kv line. In order to investigate the steady state stability a simplified method of eigenvalue analysis was performed by post-processing of selected output from dynamic simulations. The result was the most important of the real eigenvalues σ and of the complex eigenvalues σ + j ω inasmuch as they could be observed in the recorded signal. By recording a number of different signals in the simulation it was assured that all important eigenvalues related to inter-area oscillations were determined. The minimum observed damping index σ has been collected in Table 1 for each AC interface between countries. Extra stabilizers in HVDC links or in series capacitors are not applied. 7 TABLE 1 - DAMPING COEFCIENT σ OF INTER-AREA OSCILLATIONS FOR THE PRESENT DEVELOPMENT AN Interface w2000 s2000 w2010 s ,268-0,365-0,321-0,354 DE- -0,181-0,143-0,114-0,143 DE- -0,051-0,141-0,101-0, ,315-0,404-0,311-0, ,300-0,390-0,319-0,408 -SW -0,155-0,117-0,128-0,108 LT- -0,319-0,395-0,338-0,423 LT- -0,325-0,403-0,311-0,400 -LT -0,324-0,403-0,311-0, ,322-0,404-0,320-0,413 -LT 0,050-0,192-0,071 - SW- -0,196-0,315-0,209-0,276 SW-NO -0,161-0,191-0,153-0,198 The results show very good damping of oscillations in the IPS/UPS system. Also the damping of large oscillations in the NORDEL and UCPTE systems is clearly better than the figures in the table show. The minimum damping occur typically with small oscillations having amplitude of only some megawatts. The interconnection Poland - Lithuania is unstable with increasing oscillations (negative damping) in the winter peak load 2000 case. Obviously oscillations of the Lithuania - Poland interconnection are reflected also in other interconnections in the UCPTE system. Therefore it is necessary to equip generators in Kruonis (LT) - feeding electricity to Poland - with power system stabilizers. The positive effect on the 400 kv interconnection Kruonis (LT) - Elk () of the stabilizers can be seen in Fig. 6. The simulated fault is a 3-phase, 100 ms fault with successful auto-reclosing on line Hagenwerder (DE) - Mikulowa (), fault number 9 in Fig. 5. P [MW] without PSS with PSS 0.87 Hz 280 t [s] Fig. 6: Swing curves from the plan 1 dynamic investigations

4 With this addition no undamped oscillations were detected in any of the simulated faults and operation conditions of the present development plans. However in certain cases the damping could be better - especially in summer conditions - although in general the behavior is very satisfactory. This low damping is probably caused by the equivalents which are composed for different power exchange scenarios, because in the data verification for the individual synchronous systems a good correspondence to the real system has been reported without any specific damping problems. For all considered "significant disturbances" simulated in IPS/UPS system, the stability of the power systems is not violated. Therefore no additional means have to be considered - except already planned increase of transmission capacity from Russia to Belarus between year 2000 and The results are not surprising, because the system is close to the existing system and the maximum admissible transmission capacities are well known. The optimized power exchanges are within these limits. Based on the static and dynamic stability studies of the summer and winter load networks for the stages 2000 and 2010 it was concluded that the state of the transmission system in the BSR is fairly good, when planned new investments are included. Only a few local reinforcements are necessary in order to optimize the system performance and thereby to fulfil the optimal production schedules. V. ANALYSIS OF ALTERNATIVE DEVELOPMENT ANS In order to evaluate the sensitivity of the conclusions that were based on the present development plans it was decided also to investigate a number of alternative development plans. A total of 9 plans were investigated; a base case plan, which is the plan already described in the previous section, and 8 alternative plans - numbered 2 through 9. These alternative plans show the consequences of e.g. transmission line expansion, common environmental rules, increased CHP, increased energy efficiency, etc. In the plans 7, 8, and 9 special attention was paid to the future development of the south-eastern BSR, see Fig. 7. In these plans no significant changes, as compared to the base case plan, was assumed outside the south-eastern part of the BSR. Therefore the system investigations focused only on this region. From a transmission system design point-of-view these plans are the most interesting, dealing with alternative synchronous borders in the BSR, and with new technology in form of an MTDC link going from Russia to Germany. Including the base case plan the main similarities and differences of these 4 plans can be shown as in Table 2. Fig. 7: Transmission system in the south-eastern BSR The MTDC link is under investigation in a separate TEN project [5], and is expected to connect Russia, Belarus, Lithuania, Poland, and Germany. A proposed track is shown in Fig. 8. It is expected to go via the territory of Kaliningrad () in order to provide an additional supply line to the main Russian grid. The main points of interest to take into account when studying the development of the south-eastern BSR are: that this is the missing link in the Baltic Ring, since there is neither any AC nor any DC transfer capacity between IPS/UPS and (UCPTE+CENTREL) in this region, to determine the power transfer capacity between the Baltics and, respectively between IPS/UPS and as a constraint the Kaliningrad region must have a tie line to the synchronous system of Russia and Belarus TABLE 2 - COMPARISON OF AN MAIN CHARACTERISTICS Without MTDC With MTDC Baltic s synch. w. UPS Base case plan Plan 9 Baltic s synch. w. Plan 7 Plan 8 Fig. 8: Full Extension of the MTDC System

5 In the dynamic investigations a number of critical events in the region were investigated. These events were chosen based on the knowledge and experience of the companies in the region. Two different types of events were investigated: a) 3-phase faults at one end of vital transmission lines, b) 1-phase faults at important power plant units or at HVDC or MTDC converter terminal. The fault clearing time was assumed to be 150 ms followed by a permanent trip of the unit in both cases. These stated events inflict realistic disturbances on the network and the network is consequently expected to be able to withstand these events. If not then possible reinforcements are investigated and a remedy proposed in order to maintain a stable operation. Plan 9: IPS Baltics in UPS, with MTDC link. The main difference in this plan, as compared to the base case plan, is that an MTDC link from Russia to Germany has been included. A load flow investigation gave no indication of bottlenecks in the system, except for a need to strengthen the connections between Lithuania and Belarus. The only problem related to the MTDC link was that the location of the Belarussian MTDC terminal in Sever seemed unfavorable because Sever is not located near a strong load and production center in Belarus. From a dynamic point-of-view there seemed to be no problems with this plan. Neither eigenvalue analysis, nor time domain simulations with known critical events indicated any problems. This is hardly surprising, since the base case plan in itself did not display any significant problems, and since the MTDC link is connected at relatively strong points in the various networks. When investigating a fault at the MTDC converter station in Wahle, Germany it was noticed that this fault not only disturbed the transmission on the MTDC link but also the 2-terminal HVDC links connecting NORDEL with. This is shown in Fig. 9. This clearly illustrates that the entire BSR even though it consists of 3 separate synchronous networks - must be considered as one electrical region, and that simultaneous commutation failures in the future may replace loss of the largest production unit as being the most critical dimensioning event for system stability in the BSR. Plan 7: IPS Baltics in, without MTDC link. The main difference in this plan, as compared to the base case plan, is that the synchronous border between and IPS/UPS has been changed, so that the Baltics are in synchronous operation with. As a starting point two lines between Lithuania and Poland was assumed. Due to the constraint regarding the Kaliningrad region an AC tie line between Belarus and Kalingrad via the territory of Lithuania has been included as well. This plan results in a relatively weak connection with large impedances between the production centers in the Baltics and the rest of the grid. Inside the Baltics the same situation will also appear between Estonia and Latvia. In addition, the power is expected to flow towards, which is worse for the stability than the opposite direction. Therefore it was expected that strong reinforcements will be required in order to keep the machines in synchronous operation even when critical faults occur. From a load flow point of view no improvements are required besides measures to rearrange lines due to the new synchronous border. Simulations show that it is very important for the overall performance within the Baltic power system to reduce the impedance between Lithuania and Poland, as well as between Estonia and Latvia. The maximum power transfer between Lithuania and Poland was determined with respect to the transfer between Estonia and Latvia as well as with different reinforcements. The study comprised investigations of varying combination of reinforcements in the Baltics and some in Poland such as new transmission lines, series compensation, SVCs, protective power plant disconnection, fast valving, braking resistors, etc. Fig. 10 illustrates the impact of an installation of series compensation on both of the two interconnection lines and an optional third line. Fig. 9: Fault in Germany - Power on DC links between and NORDEL Fig. 10: Illustration of plan 7 dynamic investigations.

6 The figure shows clearly that an installation of series compensation increases the capacity by MW and a third line by 500 MW. Note that the studied event in Fig. 10 is not the most critical event. The most severe faults are at the production centers in the north of Estonia. An eigenvalue analysis for plan 7 with a reinforced network show that the system is stable and has an acceptable damping in winter load. The damping for summer load is a little bit worse and may be considered to be in the limit zone whether it is acceptable or not. A detailed study with a more accurate network model is however required in order to decide whether or not the damping indeed is good enough in the summer situation. In conclusion the study proposes three circuits between Lithuania and Poland, a new line between Estonia and Latvia (Sindi-Salaspils) plus an arrangement to re-use a line between Estonia and Latvia (either via Pskov in Russia, or parallel to the Russian border), and installation of series compensation for totally seven lines, see Fig. 7. In the IPS/UPS system dynamic investigations have identified additional necessary system reinforcements. Plan 8: Baltics in, with MTDC link. This plan is, so to say, a combination of the two previous plans; new synchronous border and an MTDC link from Russia to Germany. However instead of an AC tie line to Kaliningrad there is an additional converter station in Kaliningrad at the MTDC line. Note that this plan - like the conclusion of plan 7 - also includes 3 tie lines between the Baltics an the rest of, since the MTDC link constitutes a connection parallel to the 2 AC tie lines between Poland and Lithuania. As also observed in the dynamic investigations of plan 7 there is a transient stability problem in the north part of the Baltics. In case of faults in the north of Estonia the estonian generators will loose synchronism with the rest of the Baltics. A new line between Sindi,, and Salaspils, combined with series compensation on this new line and on all the existing lines connecting Baltija and Eesti, with Valmiera, (see Fig. 7) will link Estonia strongly enough to Latvia and to the rest of the Baltics. This solution leads however to a new problem, because the tie lines between Poland and Lithuania suddenly becomes a bottleneck. With the strong interconnections within the Baltics the consequence of a fault in the north of Estonia would be that the entire Baltics eventually would loose synchronism with the rest of due to a badly damped interarea oscillation. Series compensation of the tie lines between Poland and Lithuania will however remove this problem, see Fig. 11. Note that the angle difference between adjacent generators always is relatively small. In case the -LT tie lines are not series compensated the Estonian and Lithuanian generators forms a more coherent group with a large angle difference to the Polish generators. The consequence would be that the relays at the -LT tie lines would detect large currents and subsequently disconnect these lines. Fig. 11: Fault in north of Estonia generator swing curves with series compensation. Except for this no problems were observed in the or the NORDEL parts of the network. In the IPS/UPS system the following observations were made: because of the disconnection of the Baltics from the IPS/UPS system it is necessary to do some reinforcements in the St.Petersburg-Novgorod-Pskov region to replace the - for the IPS/UPS system - disconnected lines via the Baltics, in Belarus it is necessary either to make reinforcements in the region around the MTDC converter station at Sever or to decide on another location for the converter station closer to a load center in Belarus, in the autonomous power system of Kaliningrad it will be necessary to take some special measures to live up to the (n-1) criterion, because - compared to the total load - there will be only a few, relatively large production units. VI. CONCLUSION The power system simulation in the Baltic Ring Study has evaluated the dynamic system behavior for different planning scenarios in the BSR. The main results from the power system analysis can be summarized as follows: Including the already planned new investments the state of the transmission system in the BSR is fairly good. Only a few additional reinforcements in order to optimize the system performance are necessary for the present development plans. Considering alternative expansion and interconnection scenarios of the transmission systems additional reinforcements will be necessary. These reinforcements will, however, reduce both the need for investments in the production system, as well as reduce the overall running costs. For the transmission system the advantage of more interconnections is, that the total need for spinning reserve as well as reserve capacity will be reduced, and that the frequency will become more stable. The Baltic Ring study can be seen as a first step in the cooperation between the electricity companies around

7 the Baltic Sea. Further activities will be initiated, discussed, and harmonized in the new-established Baltic Ring Electricity Cooperation (BALTREL). In order to physically close the Baltic Ring the study proposes the construction of a DC link between Lithuania and Poland via Kaliningrad. The link should be constructed in such a way that it can be the first step of an MTDC link between from Russia to Germany. It was observed during simulations with a representation of the dynamic behaviour of the HVDC links that simultaneous trips of HVDC links may occur. Due to the expected large amount of DC transmission capacity in the BSR in the future, such simultaneous trips may in the future become the most critical system dimensioning event. VII. ACKNOWLEDGEMENTS The transmission system studies described in this paper was performed as a co-operation project, where all of the 18 participating companies in the Baltic Ring Study have supported the project by active participation. Further more - apart from the companies of the authors - representatives from RAO S Rossii, Russia have also taken a big part in the described investigations. VIII. REFERENCES [1] Baltic Ring Study - Main Report Volume 1: Analysis and Conclusion; TEN Study, January [2] Baltic Ring Study - Power System Analysis Report, January [3]The Tacis and Phare technical study of the interface between the extended West European power system and its Eastern neighbours. PreussenElektra AG, Bayernwerk AG, EDF, RWE Energie AG, 1996 [4] Multi-terminal HVDC Systems, Sub-report within the Baltic Ring Study, November [5] East-West High Power Electricity Transmission System - Baltic Route, TEN Study in progress. IX. BIOGRAPHIES Hans Knudsen - was born in Denmark in He received his M.Sc.E.E and his Ph.D. in 1991 and 1994 respectively from the Technical University of Denmark. He is currently employed in the Transmission Planning section of the transmission and distribution company NESA. Matthias Luther - was born in Germany in He received his M.Sc.E.E. and his Ph.D. in 1986 and 1992 respectively from the Technical University of Braunschweig. Presently, he is responsible for customer services at PreussenElektra Netz. Georg Schröder - was born in Germany in He received his M.Sc.E.E. from the Aachen University of Technology in In 1998 he joined the infrastructure team of PreussenElektra and is now responsible for process organization. Lars-Ove Ellus - was born in Sweden in He received his B.Sc.E.E in From 1985 to 1988 he studied Electrical Engineering at Technical Institute of Västeras and Royal Institute of Technology, Stockholm. He is presently employed at Vattenfall Transmission AB and responsible for network analysis and planning. Mikko Koskinen - was born in He received his M.Sc.E.E. from the Helsinki University of Technology in He is now working for Finnish Power Grid Plc in the field of transmission system reliability evaluation.