|
|
|
Update:
Our Latest Forum Schedule
|
FREE
Email
Newsletter
Monthly
Magazine
Member Of
Download
Our
Issues
|
POWER SYSTEM PLANNING REPORT
Reliability Analysis For Power System Planning and Operation
By Bernard Dumont, Armin Sorg and Thomas Weber
The liberalization of electricity markets in countries all over the world has lead to tremendous changes for electric utilities. This evolution calls for enhanced power system planning tools. The cost pressure on electric utilities is increasing and so is the push for economical optimisation of business units like power system operation, planning and maintenance. This is creating a need for new supporting tools. The user is looking for tools, which can be handled as easily as possible, combined with very expressive results. This article provides an overview of system reliability analysis in a liberalised energy market, and illustrates an application on an existing transmission system in Europe.
In Canada, the successful operation of an electricity market is reliant on adequate available transmission capacity. [10] This is equally important to energy producers and energy consumers alike. "Available transmission capacity" is one necessary type assessment of transmission adequacy, and "voltage support" and "congestion" are other areas which need to be assessed.
In the short term, up to a 24-month horizon, transmission system planned outages are reviewed to identify which could be rescheduled in order to meet system reliability requirements. Over the longer term, strategic transmission system reinforcements could be added to meet system reliability requirements. Hence, transmission system reliability could be managed by transmitter-initiated or by independent operator-initiated infrastructure reinforcements or operational procedures.
Transmitters are often best positioned to identify contingencies, which will adversely impact the reliability of the system, and to take calculated risks in arranging outages or in phasing of transmission system reinforcements. Based on each contingency, the resulting impact on load levels is estimated by considering the extent to which load is interrupted, and the duration of such interruption. The contingencies that are assessed should include faults or outages to any transformer, double circuit line, bus or any two cables in the same trench.
Reliability analysis tools are an important supplement to classical power system analysis tools like power flow or short circuit calculation, especially in light of the circumstances of the liberalised market structures. Two major advantages are achieved by reliability analysis:
- The investigations lead to quantitative results given by basic reliability indices such as "mean interruption frequency" per year, "mean duration of supply interruptions" in minutes per year or "unavailability" in minutes per year for nodes or the complete system. Higher accumulated indices can be created very easily by using these basis indices.
- The investigations deliver detailed feedback (e.g. to detect weak points in the system). This is evaluated by having a system-oriented view or a node or customer oriented view. The results indicate the type of the highest impact outages, as well as information on which component in the power system is the cause.
Application of Reliability Analysis
The list of application fields for reliability analysis is wide spread. Classical tasks are well known such as a comparison of variants for a system expansion or a substation configuration.
More "enhanced" applications are given by new problems occurring due to the deregulation of electricity markets. As examples, the evaluation of penalties [1] could be done in addition to the determination of interruption costs [2][3]. A very interesting enhanced application is the evaluation of insurance models [4]. Such models need a very detailed risk analysis regarding details of the power system investigated. Here, a combination of power system planning and risk analysis mathematics generate value added results.
Reliability Analysis Tool
Our reliability analysis tool ZUBER [5] consists of a calculation part which generates the component failure combinations and models their sequence up to complete restoration of supply, the analysis part which conducts a detailed analysis on the basis of the calculated failure sequences, and the visualisation tool ZUBERView presents results in a user-friendly format. Fig. 1 gives an overview of the ZUBER program structure.
A clear separation of calculation and analysis of the failure combination sequences has been carried out with the advantage that the calculation -- by far the most time-consuming part of the power system analysis -- needs to be performed only once in a reliability analysis. Typically, the calculation results then are analysed in several different ways. As the calculation results are already stored in a database, the runtime for the analysis is quite short.
The input data includes three main parts. The power system data contain all load flow and short circuit relevant data extended by detailed substation topology information (e.g. double busbar) and overhead line topology information (e.g. multiple circuit lines for the modelling of common mode outages). Operational boundaries like limits for currents, voltages, PQ-diagrams of generators are considered. Sorted and standardized annual load duration curves or standardized daily load curves define the time dependency of the loads.
The reliability data [7] contain data for network components, for power stations and remedial measures. The reliability data for network components contain the full set of failure models according to the German Statistics of Incidents VDEW [8].
The component types of substations and lines have to be distinguished. For power stations, data for the six-state model can be considered. The data for remedial measures like busbar transfers or start-up of injection units contain the feasibility and duration of the measures. The protection data needed are the types, location and operative direction of protection relays defining the protection tripping areas for each component. These data enable the determination of non-selectivity in the protection system limitations.
Overview
The complete Serbian bulk power system was investigated under operational and planning scenarios, with focus on the Western region. Here the effect of the replacement of some 400kV lines is checked. The reliability indices are calculated for the actual state of the power system as well as for an expansion planning scenario 2010. The main objective is to compare the results of different nodes within this power system but also to benchmark to systems in Europe. Considering the existing problems in the power system, the effects of the proposed improvements are highlighted.
The Western part of the Serbian bulk power system includes the main load centre, Belgrad and the hydro power plants in Bajina Basta. The connection to the load centre is mainly by four 220kV lines. The pump storage plant and the thermal power plant in Obrenovac are strongly connected, so the pump energy is delivered mainly by this coal-fired plant in low load times. During peak load conditions the pump storage plant helps to cover the energy demand in the Belgrad region.
Comparing the expansion planning scenario 2010 to the power system status 2001 notes two major changes, a new power plant two x 350 MW in Kolubara and two 220kV lines to be replaced by 400kV lines. A new 400kV substation in Mitrovica is also considered in order to connect the new 400kV line to the power system. After the proposed power system improvements, the pump storage power plant is regarded as connected to the 400kV grid. Two overhead lines are replaced by 400kV lines, one connected to the planned 400kV substation, Mitrovica the other connected to the new Kolubara power plant.
Modelling of the Power System
The Serbian bulk power transmission system is interconnected to several neighbouring power systems, which are modelled as 100 per cent reliable to avoid any influence on the Serbian power system. The power plants are regarded using a two-state representation instead of a more detailed six-state model due to data acquisition and computing time constraints.
Most busbars are not protected by busbar differential protection, so the distance protection of the corresponding substations has to trip in case of busbar outages. The lines are protected by distance protection, the transformers by differential protection. Data for the Serbian bulk power system are used and supplements provided by available data based on the German VDEW Statistics of Incidents [7,8] or experience values.
Information on remedial measures are included, such as the time for a busbar transfer and the time for switching operations. For all remedial measures a duration of 0.5 hours is considered to account for the fact that switching actions have to be done by staff located in each substation in accordance with the dispatching centre.
The Analysis
The AC power flow is used as the power system state analysis tool within the reliability calculation. So, for example, reactive power problems can be calculated. The AC power flow as used in load flow analysis ensures a good modelling of the power system behaviour.
The minimal state probability for the investigations is set to Qmin = 10-8 with a maximum outage order of three components. For each variant, nearly 145.000 outage combinations are investigated. The computation duration of the calculation part was 21 hours and for the analysis part, three hours on a 600 MHz Personal Computer with a Windows NT 4.0 operating system. The results of the calculations are expected values as well as standard deviation for the reliability indices.
Results of the Reliability Analysis
The effect of the power system improvements is given in detail for two nodes in Figures 2 and 3 for the reliability index unavailability. The total value of the unavailability is divided according to the affected components. For example, the results are extracted to find out the outage of which component has an influence on the investigated node.
Figure 2a illustrates the actual situation for the load node Srbobran. The values for the unavailability are unacceptably high. Because of high lengths of the connecting 220kV lines, the outage of one of the lines leads to a voltage collapse situation in case of maximum load. The most influencing outages are overhead line outages (L xxx). The influence of the generators (Gxxx) is given by regarding overlapping independent single outages in the system, such as the outage of a generator and the outage of a line. Regarding the power system changes as described above, the results lead to an improved level of unavailability as can be seen in Figure 2b.
The new 440kV station Mitrovica leads to a reduced connection length for the 220kV lines. As can be seen in Figure 2b, the outage of the transformer 440kV / 220kV TMITRO42 delivers the most severe impact on the unavailability of Srbobran. Comparing the values of Figures 2a and 2b, the drastic reduction and the positive effect of the system improvements can be seen.
Figure 3 shows a comparison of the unavailability for the load node Belgrad 3 for both scenarios. The influence of the proposed measures leads to a strong reduction of the unavailability of the load node Belgrad 3 from scenario 2001 (left bars) to 2010 (right bars). The main reason for the reduction is avoiding overload situations of transformers T T1/T2 BGD8 by a new connection to the load centre reducing the load of these transformers. So, in case of an outage, reserve capacity is available.
Additionally, Figure 4 summarizes the results for the complete investigated power system. For all load nodes in the power system, in most cases the 110kV nodes of the transformers, the reliability indices frequency of interruption as well as the unavailability are shown. The upper part of Figure 4 shows the nodes with high values for the reliability indices, the lower part includes the low values. Because low values of reliability indices describe a high power supply quality, the upper part of Figure 4 includes the weak points of the Serbian bulk power system.
Comparing the results for the scenario of the years 2001 and 2010, shows for several nodes a high degree of reduction of the reliability indices. This describes the positive effect of the proposed planning activities on the power system reliability.
The proposed power system improvements solve many existing problems in the Serbian bulk power system. The replacement of existing 220kV lines with new 400kV ones increases the reliability of the Serbian power system as well as for different load nodes.
The overall results, as given in Figure 4, show that besides the problems solved, additional planning effort is necessary to further improve the system. The aim of future power system planning has to be that the reliability level be comparable to middle European conditions to ensure a high-level power supply in a common liberalized electricity market [6].
Conclusions
The added value of reliability analysis for power system planning is illustrated. The increased cost pressure given by the open market activities lead to the necessity to extend the classical planning tools by reliability analysis.
Modern reliability analysis tools allow a detailed analysis of the results after calculating the reliability indices. Besides overall values for the power system, the results for different load nodes can by analyzed individually. This includes the important information of which outage and of which component in the system is mainly responsible for supply interruptions of the investigated node. Based on this information, weak points in the system can be detected and improvement measures taken.
In conclusion, it can be stated that reliability analysis is a powerful tool in addressing the requirements of the new market.
References
1) Ofgem. "Utilities Act 2000-Financial Penalties." Office of Gas and Electricity Markets. London (2000)
2) R. Billinton, G. Tollefson and G.Wacker. "Comprehensive Bibliography on Reliability Worth and Electrical Service Interruption Costs 1980-1990." IEEE Trans. Power Systems, Vol.:PWRS-6, No.: 4, (1999), pp. 1508-1514
3) R. N. Allan and K. K. Kariuki. "Factors affecting customer outage costs due to electric service interruptions." IEEE Proceedings Generation Transmission Distribution, Vol.: 143, No.: 6, (1996), pp. 521-528
4) A.Sorg, I. Stover, Th. Weber, W.H. Wellssow and M. Zdrallek. "Investigation of a Supply Interruption Insurance System." 6th International Conference on Probabilistic Methods Applied to Power Systems, (PMAPS), Funchal. Madeira/Portugal (2000)
5) FGH, ZUBER. Program for Calculation of the Supply Reliability in Electric Power Systems, Manual of Version 3.05. V., Mannheim (2000)
6) A. Sorg, Th. Weber. "Application Examples of Reliability Analyses According to the Requirements of a Liberalized Electricity Market." Proceedings of the 1st Balkan Power Conference (BPC). Bled/Slovenia (2001)
7) C. Bose, R. Hugel, H. Lebeau, Th. Weber and. W.H. Wellssow. "The new VDEW statistic of incidents- A source for component reliability indices of electric power systems." Proceedings of the International Conference on Safety and Reliability (ESREL). Lisbon, Portugal (1997), pp. 1127-1134
8) VDEW, VDEW statistic of incidents 1994-1998. VWEW, Frankfurt a.M., Germany
9) Sorg, A.; Weber, Th.; Kaltenborn, U.; Marinkovic, M.; Mijuskovic, N. "Improvements of the Reliability of the Serbian Bulk Power System -- Results of a Case Study". 2nd Balkan Power Conference (BPC) -- Power System Control and Deregulation of Electricity Market. Belgrad (Serbia): 2002
Armin Sorg and Thomas Weber are with ALSTOM Energietechnik GmbH/Network Planning Business in Frankfurt, Germany.Bernard Dumont is with Alstom Canada.
For more information on ALSTOM Network Consulting and Service offering or inquiries contact your local representative or send us an email at TDS.canada@tde.alstom.com. ET
|
|
|
|
|
LATEST ISSUE
