1. Frankfurt (Germany), 6-9 June 2011 Gerhard Theil Technical University Vienna Institute of Energy Systems and Electric Drives gerhard.theil@tuwien.ac.at Andreas Theil Wien Energie Stromnetz andreas.theil@wienenergie-stromnetz.at Marta Theil marta@theilnet.net THEIL – Austria – Session S5 – Paper 0314 MEDIUM VOLTAGE NETWORK RELIABILITY EVALUATION: SIMULATION OF PRACTICALLY APPLIED SUPPLY RESTORATION STRATEGIES FOR DOUBLE-FAILURE EVENTS

2. Frankfurt (Germany), 6-9 June 2011 Objective Reliability evaluation of medium voltage overhead-line and cable networks Main topics Assessment of multiple-fault frequency in networks exposed to progressed component aging Simulation of efficient supply restoration procedures for cases when fault-clearing personnel has to expect occurrence of more than one fault location per failure- event THEIL – Austria – Session S5 – Paper 0314

3. Frankfurt (Germany), 6-9 June 2011 Multiple-fault frequency model Approach: Modeling short-circuit exposition levels caused by past-time failure events for individual network components – “stress factors”. Multiple outage frequency of network components is determined by “stress factors”. THEIL – Austria – Session S5 – Paper 0314

4. Frankfurt (Germany), 6-9 June 2011 Classical multiple-fault frequency model Classical conditional probability approach for frequency of second-order fault: (primary event at component i secondary event at component k) fe Frequency of the primary event, e.g. single-phase-fault pe Conditional probability of the second event given a first one has occurred l i, l k Line lengths Substitution of conditional probability by the stress-factor results in the: Stress-based multiple-fault frequency model THEIL – Austria – Session S5 – Paper 0314

5. Frankfurt (Germany), 6-9 June 2011 Stress-based multiple-fault frequency model Sm Stress factor representing the second-order-event sensitivity of component k with respect to a first-order event occurring at component i, tp Age of component k ch Calibration factor Stress factor is dependent on:  Expectation of fault number accumulated during age of component k  Position of components i and k within network structure. (Components positioned near an injection source usually carry larger short-circuit- currents than components in remote positions). THEIL – Austria – Session S5 – Paper 0314

6. Frankfurt (Germany), 6-9 June 2011 Medium-voltage network reliability evaluation System reliability index: Not supplied energy nc Number of network components nn Number of consumer substations affected by the failure event f Component outage frequencies t U Interruption time of the i-th consumer affected by the failure event L U Not supplied load of the i-th consumer THEIL – Austria – Session S5 – Paper 0314

7. Frankfurt (Germany), 6-9 June 2011 Fault management strategies Failure-search procedure A) Fault-clearing personnel activities are conducted by minimization of not supplied energy. B) Binary search. Supply restoration 1.) Switching-in of lines for supply restoration is performed during the entire fault-clearing procedure irrespective of the number of switching actions. 2.) Supply restoration actions are performed for identified intact parts of the line only. THEIL – Austria – Session S5 – Paper 0314

8. Frankfurt (Germany), 6-9 June 2011 Fault management strategies Single-phase-fault handling - Line remains in operation, short-time consumer interruptions occur because of switching performed during failure search. + Line is switched off immediately after fault indication. Short-time switching-in actions are performed for fault locating. Thus, exposure time to failure effects is lower than in case '-'. * Single-phase-faults are transformed to short-circuit-faults by activating low-impedance-grounding. Subsequently, information provided by fault-locators is used during failure-location procedure. THEIL – Austria – Session S5 – Paper 0314

9. Frankfurt (Germany), 6-9 June 2011 Study System Total line length: 41 km Number of supply stations: 100 Single-outage frequency: overhead-line: 0,069/km.a cable: 0,021/km.a Double-outage frequency: 10-9 to 10-5 /a THEIL – Austria – Session S5 – Paper 0314

10. Frankfurt (Germany), 6-9 June 2011 Comparison of LOLE for different supply restoration strategies – 2 lines Line 1 Line 2 Simulation of single and double component outages Upper dark column parts: Contribution of double-faults THEIL – Austria – Session S5 – Paper 0314

11. Frankfurt (Germany), 6-9 June 2011 Comparison of LOLE for different supply restoration strategies Remarks A1-, B1- Proportions of reliability indices are dependent on line structure and load distribution along line. For lines without lateral branches (line L1), strategy A is the preferable one. For lines with branches (line L2) binary search procedure is converted to strategy A. A1-, A2- Reduction of stress exposition (A2) leads to longer supply interruption times. A1-, A1+ Disconnecting the line immediately after fault occurrence (A1+) possibly leads to reduction of the probability of double and multiple faults. Provided that this reduction is sufficiently large, improvement of system reliability can be reached. Otherwise an increase of LOLE has to be expected. A1-, A1* Activating low-impedance-grounding can result in similar effects. A1+, A1* Information delivered by fault-locators (A1*) is available for acceleration of fault-location and supply-restoration procedures. Additional reduction of LOLE can be observed (line 1). THEIL – Austria – Session S5 – Paper 0314

12. Frankfurt (Germany), 6-9 June 2011 Conclusions  Multiple-fault frequency in networks exposed to progressed component aging is modeled by taking stress accumulated during component lifetime into account.  Algorithms for simulation of supply restoration procedures handling multiple failure events were developed and implemented into medium-voltage network reliability evaluation software. THEIL – Austria – Session S5 – Paper 0314

13. Frankfurt (Germany), 6-9 June 2011 Conclusions  Taking solely first-order-failure events into consideration, supply restoration procedures without limitation of not successful switching actions result in the lowest total not-supplied energy. However significant lifetime reductions of components may be the consequence.  Taking second-order-failure events into consideration, strategies resulting in lower times of exposure to single-phase-faults seem to be more profitable. Possibility of significant multiple-outage frequency reductions and substantial system reliability improvements exists.  Low-impedance-grounding activated immediately after occurrence of single-phase-faults seems to be a promising measure. THEIL – Austria – Session S5 – Paper 0314