(Reviewed by Alex Apostolov, USA)
The brochure surveys the fault location techniques in practical use on AC transmission and distribution lines, examines how each behaves across different network topologies, and reports the results of a questionnaire completed by forty utilities. It acknowledges the IEEE guide for determining fault location on AC transmission and distribution lines as a foundation.
The introduction defines fault location error in both absolute and relative terms, and lists fifteen factors that affect the quality of a result — among them fault type and resistance, pre-fault power flow, source strength behind the line ends, line imbalance from lack of transposition, inaccuracy in line parameters, fault inception angle, CT saturation, and the quality of A/D conversion. It notes that fault locators, unlike protection, need not operate in real time, and can therefore use more complex algorithms and achieve higher accuracy, at the cost of results arriving in seconds or minutes. Sampling frequencies across technologies covered span 1 kHz to 10 MHz.
Chapter 2 describes four families. Impedance-based methods are treated in single-ended and double-ended forms, with the unsynchronized double-ended case requiring iterative solution of a nonlinear problem. Travelling wave methods are covered in single-ended and double-ended forms, including the difficulty of distinguishing reflections from the fault and from the remote terminal. Current measurement and voltage drop techniques are also described, as are faulted circuit indicators and line patrol, which remain dominant in distribution practice.
Chapter 3 opens with instrument transformers and their interfaces, summarizing in a table which measurement principles suit which technique — drawing on the earlier B5.24 brochure, and including conventional transformers with merging units. The chapter then works through topologies: two-terminal circuits (homogeneous, non-homogeneous, series-compensated, and un-transposed), three-terminal and multi-terminal circuits, and distribution networks (solidly grounded radial with laterals, compensated, interconnected radial, and ring-main). Series-compensated lines are identified as particularly difficult because of metal-oxide varistor nonlinearity; compensated distribution networks are difficult because earth faults are often intermittent.
Record retrieval and its maintenance implications are discussed, from direct port access through substation computers to WAN delivery, with recordings stored predominantly in COMTRADE format. Time synchronization errors between terminals are identified as a persistent obstacle to double-ended calculation.
Chapter 4 covers improving confidence in a result by correlating it with geographic maps, lightning tracking systems, and satellite-based vegetation fire tracking. Chapter 5 addresses what happens after the calculation: auto-reclosure practice and its effect on interpretation, and the distribution of results. It includes a matrix mapping conditions — unsuccessful auto-reclosure, lost load, initiated scrub fire, imposed operating constraints — onto operational, engineering, regulatory, customer and executive stakeholders, with short- and medium-term notification timeframes.
Chapter 6 covers synchrophasor based location in four categories: multi-terminal methods assuming dense PMU coverage, wide-area methods assuming sparse coverage and using a two-stage narrow-then-pinpoint procedure, distribution methods addressing DER penetration and meshed configurations, and special cases for series-compensated and double-circuit lines. It then covers centralized model-based and self-learning methods, which simulate fault scenarios and select the closest match, noting their computational burden and difficulty with meshed networks.
Forty utilities from around twenty countries responded, mostly TSOs. On average six people per company are involved in fault location analysis. Single-ended impedance is the most widely applied transmission method, reported by roughly 85% of respondents for two-terminal circuits, and is overwhelmingly implemented inside protection relays rather than in stand-alone devices; line fault indicators are predominantly stand-alone. Reported impedance-based errors cluster below 10% and often below 5%, while travelling wave errors are typically within 300 m.
Respondents identified differing vendor sampling rates and inconsistent filtering as a barrier to double-ended impedance calculation, travelling wave systems triggering on only 60% of feeder faults in one utility’s experience, and inaccurate impedance data and mutual coupling as sources of single-ended unreliability.
The working group concludes that initial estimates from relays and recorders are frequently inaccurate, that improved estimates should draw on all techniques relevant to the topology together with geographical, environmental and customer data, and that the search for the fault should begin in the vicinity of the estimate.
Working Group B5.52, December 2021 — 137 pages
Convenor: S. Sefidpour (SE)
Secretary: A. Montenegro (US)
ISBN 978-2-85873-559-4
