Nation-crippling failures of Extra-High-Voltage (EHV) transformers from geomagnetic induced currents (GICs) caused by earth-directed solar coronal mass ejections (CMEs) have been quantified and impacts highlighted by our mass media for more than a decade. Unaddressed, durations of the power distribution outages will be determined by new/repaired transformer replacements available from very few suppliers.
Transformer protection schemes are intended to enable timely grid disconnections of transformers experiencing excessive GICs which shift excitation operating points that cause rapid thermal insulation damage (described below). Costs of the hardware associated with proposed approaches often appear attractive, but require costly high-voltage installations, inhibiting wide-scale implementation and creating vulnerability to inappropriate activation or compromise. Monitoring only GIC-caused Wye-grounding currents to justify switching (e.g., more than 30 Amperes longer than five minutes) does not account for AC load magnitudes or imbalances which determine the vulnerability of a given transformer.
Dependency upon space-weather warnings of GIC damage potential is fraught with opportunities for human error and potentially significant recurring costs. CME-induced currents depend upon local and seasonally-varying soil conductivity, and specific damage thresholds depend upon each transformer’s AC load condition. Unnecessary grid disconnections/reconnections increase opportunities for grid system instabilities.
An alternate real-time damage-averting approach is proposed that requires neither costly exposed installations of critical high-voltage components, nor installations beyond the immediate proximities of transformers at risk, affording confirmed automatic grid protection, restoration and reasonable security.
Operating-point shifts due to CME-induced currents (see Figure) can result in alternate half-cycle over-excitement of transformers in service at full power loads, exhibiting very strong second-order (and higher) harmonic winding current components and concomitant audio-frequency sounds (e.g., via magnetostriction) prior to their ultimate insulation burnout due to excess-flux induction heating of tie plates, core leg bolts, and hardware. Hence it is proposed (1) that multiple detectors of such signature transformer sounds be placed in close proximity (or attached to) such transformers, (2) that an unique Boolean Logic-based Truth-Table algorithm be implemented to quickly assess their signals and those of magnetic ~DC flux detectors and Wye-grounding currents (see “4” below) to establish a multiply-confirmed indicator of specific transformer burnout danger and accordingly (3) provide switch-control signal(s) to existing means for activation of existing switches to accomplish near-immediate disconnect(s) of the local power system from the power grid, where (4) signals from magnetic-field detectors of time-varying GIC-caused (~DC) magnetization of the transformer core – and signals from detectors of the neutral wire GIC (~DC) Wye-connection grounding currents are assessed collectively by the above algorithm to provide (5) confirmed-danger alerting/switching signal(s) --- and (6) an “all clear” signal(s) (indicating “normal,” acceptably low levels of neutral grounding wire GICs), to justify and/or activate established procedures for grid reconnection(s).
Alternate methods are envisioned for deriving real-time local danger signals for analyses and confirmed responses to GICs, affording low vulnerabilities to compromise.
- 2014 Top 100 Entries
ABOUT THE ENTRANT
Name: Carl Lundgren
Type of entry: individual
Number of times previously entering contest:2
Carl is inspired by:
My experience with LF propagation may provide a solution to the following problem:
During a Hemp event resulting in the absence of all Telecommunications Infrastructure including Internet and Cellular, current procedures using HF backup communications to report damage assessments to Emergency Response Agencies from the field will be hampered, resulting in disruption of time-critical resource management.
Software used for this entry:
PROPRIETARY BOOLEAN LOGIC TRUTH-TABLE SWITCHING ALGORITHM
Patent status: none