Linking compatible systems in the European grid
The CENTREL group of power utilities–Poland, Hungary, Slovakia and the Czech Republic–have successfully interconnected with the UCPTE power system
Power utilities of Central Europe joined in their efforts to interconnect their power systems to the Union for the Coordination of Production and Transmission of Electricity (UCPTE) power system. This initiative was accepted by the UCPTE in 1992. CENTREL`s group of power systems was established in October 1992 by four companies:
1. Ceske Energetycke Zavody a.s. from the Czech Republic
2. Magyar Villamos Muvek from Hungary
3. Polskie Sieci Elektroenergetyczne S.A. from Poland and
4. Slovensky Energetycky Podnik from the Slovak Republic?today Slovenske Elektrarne, a.s.
The main objectives of the group were to:
– coordinate the interconnection with the UCPTE
– coordinate systems development
– improve the operational conditions of the power systems
– monitor the quality and reliability of the electricity supply and
– cooperate with the other European interconnected systems.
In October 1992, the UCPTE-CENTREL executive committee, consisting of the four CENTREL power utilities and their seven UCPTE neighbors, was set up. This committee formulated a Catalogue of Measures (Massnahmenkatalog) with technical, techno-economic and organizational requirements, which had to be fulfilled before interconnection with the UCPTE. This document, accepted by each member of the committee, was signed on October 12, 1992, in Prague.
Each CENTREL member prepared an individual feasibility study for the interconnection. In cooperation with a German-Austrian consortium (BAG, PE, VEAG and ÖVG), the study was made on implementation agreement between CENTREL and UCPTE confirming that the interconnection of CENTREL could take place no later than 1997. During 1993-1995, the German Deutsche Verbundgesellshaft from Heidelberg performed both static and dynamic stability calculations of CENTREL systems interconnected to the UCPTE.
CENTREL systems wanted to test their ability for autonomous parallel operation. The tests were performed in 1993 together with German VEAG system (former East Germany area) operating simultaneously with the CENTREL systems. After successful test results, during 1994, CENTREL systems made enormous progress in the quality of their power system operation.
CENTREL and VEAG autonomous tests
The main aims of the tests included:
– checking the effectiveness of primary and secondary controls in investigated isolated systems
– observing isolated systems operation with natural power load variation (deviations of frequency and exchanged power total)
– observing isolated systems operation with sudden active power outages of several hundred MWs in each of the systems and
– determining static and dynamic frequency characteristics for the investigated systems.
With the installed capacity at about 81,000 MW, electrical energy production from conventional thermal power stations was approximately 85 percent; nuclear power stations was approximately 12 percent; and hydro power stations was approximately 3 percent. The majority of production was from coal and lignite-fired power stations.
Nine dynamic tests (Table 1) were performed on September 29-30, 1993, with generation or imported power ranging from 300 to 500 MW. The tests included isolation of the interconnected system for islanded operation, values of power transmitted in the tie lines, coordination of operations during the tests and characteristics of central load frequency controllers. The primary control dead bands was preset to zero.
Preparation of the systems for islanded operation preceded the tests. Disconnection from NPS the Ukraina 750 kV tie line Zapadoukrainskaya–Albertirsa carrying 378 MW–was the first dynamic test. A non-scheduled test Sept. 30, 1993, switched the power unit off with 273-MW load on the VEAG (BEWAG) system.
Each power utility participating in the test was responsible for measurements in its own system. The basic measurements were performed using a cycle time ranging from 0.1 to 1.0 second (s). Direct measurements of instantaneous values of voltages and currents were taken with a 0.1-s cycle time. The basic measurements were performed on all intersystem tie lines (in the border substations), as well as in selected basic thermal and hydro pumped-storage power stations to check operation of primary and secondary controls.
Independent of the previously mentioned measuring systems, the existing measurement systems in national control centers were in operation. Synchronization of the measurements was performed using Frankfurt Radio timing signals (DZF 77).
A map of the tested systems operation area (Figure 1) indicates the intersystem measurement points on tie lines and localization of the switched-off power. The steady value of frequency deviation (f steady) varies from -52 to 46 mHz and was calculated as mean values in the interval 20 to 30 s from the instant of sudden active power balance break time period t0.
The values of frequency derivative (df/dt) for the dynamic processes are determined as the maximum 1 s mean value for an initial 2-s interval of the frequency dynamic process. The values range from -43.1 to 31.6 mHz/s. The starting point of the 1-s interval starts 0.1 s after t0.
Frequency vs. time diagrams used for calculation of the measurement results were determined from measurements performed in the 400-kV substation in Wielopole (Poland). The majority of frequency data are comparable with the ones obtained in other measurement systems. Differences in particular cases ranged from zero to several mHz.
A diagram of frequency vs. time for the 30-s intervals (Figure 2) resulted from dynamic tests carried out in Dunamenti (test # 2). A similar test was made in Boxberg (test # 8) which showed similar results. Test results include minimal and maximal fluctuation of frequency. A 36-hour histogram showed about 28 percent in the 50-cycle range. The next most dominant value was 13 percent at a frequency of between 49.990 to 49.995 Hz.
One-hour-mean values of frequency from 36-hour tests of islanded operation were 49.984 Hz to 50.017 Hz, while maximal frequency deviation with duration time exceeding 3 s (with 30-s period since t0 ignored for dynamic tests) was within +73 mHz and -77 mHz.
The results obtained both for dynamic and static tests were good. The system statism (ss) for particular tests ranged from 6.8 percent to 11 percent. The coefficient of power system regulating energy (ks) for a majority of the tests was approximately 9,000 MW/Hz over 20 percent of system power load. These results testify proper operation of primary control in the investigated system.
As the result of primary control operation, the frequency variations, both in static and dynamic states, were relatively small (from several to several tens of mHz). Primary control also caused considerable improvement in power system operation by decreasing frequency deviations when compared with the state before isolation of the investigated system.
It is necessary to point out that the investigated system of 45,000-MW capacity during the test is characterized in a natural way by larger frequency deviations as compared with a large system with several hundred thousands of MW of active power capacity.
Operation of secondary control was also positively qualified. After sudden active power outages, recovery of the frequency to rated value, in a majority of cases, occurred within 10 minutes to the level found in a majority of static states–50 Hz ?20 mHz in 4 to 7 minutes. During the 36-hour test, approximately 81 percent of the frequency deviations in the static states were within 50 Hz ?20 mHz.
Power quality improvement
In September 1993, CENTREL and VEAG power systems achieved great success, and since then, continuous upgrading of the primary and secondary controls have been made.
During the 1993/94 winter, shortages of electric energy occurred in the Ukraine and Bulgaria. Due to this fact, the integrated power system was split into parts operating in asynchronous mode, and CENTREL started permanent synchronous, autonomous operation with VEAG and with a small part of the Ukrainian power system (peak load below 1,000 MW). The quality of this mode of CENTREL systems operation was easily verified on a day-to-day basis, and the operational results formed the documentation of the long-run test.
The frequency histogram of the CENTREL and VEAG systems before activation of primary and secondary controls is shown in Figure 3a. The histogram presented in Figure 3b shows the results obtained for a K-factor of 8 percent in the secondary control formula of the central controller. The results obtained after modification of K-factor value from 8 percent to 12 percent are presented in Figure 3c. The primary control share in peak load amounted in CENTREL systems from ?2.4 percent to ?2.6 percent.
Analyses of real-time operations confirm that CENTREL operates according to the UCPTE standards with sufficient power balance and primary and secondary controls without overloading the network. In March 1995, UCPTE put an action plan into effect for synchronous operation with the CENTREL systems. A trial synchronous interconnection with UCPTE and VEAG is planned for September 1995.
Editor`s Note: The article in Power Engineering International, “HVDC converter station links European grid,” Sept./Oct. 1994, page 19, contained some misleading data about the actual condition of the CENTREL power network.