科研 Skill

继电保护级差整定

protection-coordination-study

Checks relay/breaker/fuse time-current coordination and selectivity margins from short-circuit results. Use when sizing protective devices or diagnosing a miscoordination.

发布者 Codar安全审核通过v0.1.0更新于 2026-11-18

SKILL.md

Protection Coordination Study

Determines whether a power system's protective devices (relays, circuit breakers, fuses) will isolate a fault with the correct device, in the correct order, without unnecessarily tripping healthy equipment upstream. This is a distinct engineering discipline from the steady-state and dynamic studies the rest of this plugin covers (power flow, stability, hosting capacity): those tell you the system is healthy in normal and contingency operation; this skill tells you it will fail safely and selectively when something actually breaks.

When to use this

  • Sizing or specifying breakers, relays, or fuses for a new feeder, substation, or interconnection.
  • A time-current coordination study shows overlapping curves, a miscoordination (the wrong device trips first), or no margin between an upstream and downstream device.
  • Investigating a nuisance trip (a breaker opened for a fault that a downstream device should have cleared) or a failure to trip (a fault persisted because no device operated).
  • Following up a short-circuit-mitigation finding that breaker interrupting duty is close to its rating, or an interconnection-impact-mitigation finding that a new source changes fault current seen by existing relays.

This skill assumes a short-circuit study has already produced fault-current magnitudes at each bus (from surge, pandapower, psse, pslf, powerworld, or a hand calculation); it does not replace that calculation, it consumes its output.

Core concepts

  • Protective device types and their curves.
    • Fuses: a fixed, non-adjustable time-current characteristic (TCC) from the manufacturer, published as minimum-melt and total-clearing curves.
    • Thermal-magnetic/electronic trip breakers: TCC with adjustable pickup and, on electronic trips, adjustable long-time, short-time, and instantaneous settings.
    • Protective relays (overcurrent 50/51, differential 87, distance 21, etc. in IEEE device-function numbers): programmable pickup, time dial, and curve shape (IEC or IEEE/ANSI standard inverse-time curves), driving a separate breaker.
  • Standard inverse-time curve families (needed to reproduce a relay's TCC numerically):
    • IEC 60255: standard inverse, very inverse, extremely inverse — t = TMS × (k / ((I/Ipickup)^α − 1)) with (k, α) = (0.14, 0.02), (13.5, 1), (80, 2) respectively for the three curve types.
    • IEEE C37.112 (ANSI) curves: moderately/very/extremely inverse, with their own published (A, B, p) coefficient sets per curve. Always confirm the exact coefficients and the relay's actual implementation against its manual — vendors sometimes offer both IEC and IEEE curve tables in the same relay, selected by a setting.
  • Selectivity (discrimination) margin. The industry rule of thumb is a minimum 0.2–0.4 s time margin between an upstream device's curve and the downstream device's total-clearing curve, at the maximum fault current the downstream device will see, to allow for breaker operating-time tolerance and relay overshoot. Confirm the exact margin your utility or standard requires — it is a design choice, not a universal constant.
  • Three zones of a well-coordinated study: (1) downstream device clears a fault on its own section without upstream operation, (2) if it fails, the next device upstream clears with acceptable additional outage, (3) no device at any level exceeds its interrupting rating for the maximum fault current it can see (this is the short-circuit duty check).

Procedure

  1. Gather fault-current data per bus/zone. Pull maximum (bolted three-phase, for device duty) and minimum (often line-to-ground at the far end of a zone, for pickup sensitivity) fault current at every bus the study covers. Use the plugin's power-flow/short-circuit tools (surge, pandapower, psse/pslf via PowerMCP) or a dedicated ANSI/IEC short-circuit program (SKM, EasyPower, ETAP) for a production-grade number — these tools differ (ANSI vs. IEC short-circuit methods give different multiplying factors), so note which method was used.

  2. Build or import each device's TCC. For library devices, scripts/tcc_plot.py --device references/tcc-library/<vendor>/<model>.json plots the curve from a stored (pickup, time-dial, curve-type) parameter set using the IEC/IEEE formulas above. For a non-library device, digitize the manufacturer's published TCC (most vendors publish them as PDF log-log plots; some publish the coefficient table directly — prefer that when available over digitizing a plot by eye).

  3. Overlay adjacent devices on one log-log plot. scripts/coordination_plot.py --devices device_a.json device_b.json --fault-range <Imin> <Imax> draws every device in a protection chain on one current-vs-time axis and highlights the fault-current range the two devices must discriminate over (from step 1's minimum/maximum per zone).

  4. Check each coordination pair for:

    • Selectivity: downstream curve's total-clearing time is below the upstream curve's minimum operating time by the margin from the Core Concepts section, across the whole fault-current range the pair can see.
    • Sensitivity: every device picks up (its curve is defined and operates) at the minimum fault current its zone can produce — a pickup set too high will not sense a far-end or high-impedance fault.
    • Speed: the fastest-clearing device consistent with selectivity is chosen for safety and equipment-damage reasons (e.g. arc-flash energy, which scales with clearing time); do not add margin beyond what selectivity requires.
    • Device duty: the maximum fault current at that device's location does not exceed its interrupting/withstand rating — pair with short-circuit-mitigation if it does. scripts/tcc_plot.py and scripts/coordination_plot.py both accept --report to emit a numeric table of margins at a list of test currents instead of (or alongside) the plot, for a documented pass/fail record.
  5. Check special cases explicitly, not just the normal radial chain:

    • Inrush and cold-load pickup: a transformer or motor's energizing inrush current can exceed normal load by 6–12× for a few cycles; the protective device's pickup and short-time curve must ride through this without tripping (check against the device's "no trip" region, not just its minimum pickup).
    • CT saturation: a current transformer feeding a relay can saturate at high fault current, distorting the relay's view of the fault — relevant for differential and high-set instantaneous elements; check the CT's accuracy class and knee-point voltage against the maximum fault current, not just its nameplate ratio.
    • Bidirectional/multi-source fault paths (parallel feeders, DER, ring buses): a device may need directional elements (relay function 67) because fault current can flow either direction; a plain non-directional overcurrent relay can misoperate or fail to operate depending on which source is feeding the fault. Pair with interconnection-impact-mitigation when new DER changes fault-current direction or magnitude.
  6. Document settings and margins. Write the final pickup, time-dial/ curve-type, and instantaneous settings for every device, with the coordination margin achieved at the minimum and maximum fault current in its zone, into a coordination study report. This becomes the commissioning record the relay technician sets the device to.

Data sources

  • Relay/breaker/fuse manufacturer TCC curves and setting manuals (search " time current curve" or " instruction manual").
  • IEC 60255-151 / IEEE C37.112 for standard inverse-time curve coefficients.
  • CT accuracy class and saturation data from the CT manufacturer's datasheet (ratio, class, knee-point voltage).
  • Utility interconnection requirements (often specify minimum/maximum short-circuit contribution, required protection functions, and coordination margin) from the local utility's interconnection standard.

Prerequisites

  • Fault-current magnitudes (maximum and minimum) at every bus/zone in scope, from a completed short-circuit study — see surge, pandapower, or the PowerMCP-backed psse/pslf/powerworld skills, or a dedicated ANSI/IEC short-circuit program.
  • python3 with numpy and matplotlib for scripts/tcc_plot.py and scripts/coordination_plot.py.
  • The actual device TCC data (coefficient table or digitized curve) for every device in the chain; a generic "typical" curve is not a substitute for the installed device's actual setting.

Limitations

  • This skill does not run a short-circuit study itself; it consumes one. An error in the upstream fault-current calculation (wrong source impedance, missing contribution from DER or motors) propagates directly into a wrong coordination verdict.
  • IEC and ANSI/IEEE short-circuit calculation methods can give materially different fault-current magnitudes for the same network (different treatment of pre-fault voltage, X/R ratio, and multiplying factors); mixing results from different methods across one coordination study produces an inconsistent result. Use one method consistently and say which.
  • scripts/tcc_plot.py's built-in IEC/IEEE curve formulas assume the relay implements the standard curve exactly; some electromechanical or older solid-state relays deviate from the textbook formula and need their actual published curve digitized instead.
  • Coordination margins and curve shapes do not capture everything a full arc-flash or protection-and-control (P&C) engineering review would: CT/PT wiring polarity, trip-circuit supervision, breaker failure protection, and communications-based schemes (e.g. teleprotection) are outside this skill's scope and need a licensed protection engineer's review before commissioning.
  • This is a planning/engineering aid, not a substitute for the utility's or jurisdiction's mandatory protection-coordination review and sign-off where one is required.

Quality checks before sign-off

  • Every adjacent device pair in the protection chain has a documented selectivity margin (with the specific margin value used and its source) across the full minimum-to-maximum fault-current range of its zone.
  • Every device's pickup setting is below the minimum fault current its zone must sense, and above the maximum normal load/inrush current it must ride through — both checked explicitly, not just one.
  • Every device's rating exceeds the maximum fault current at its location (cross-checked against short-circuit-mitigation if any margin is tight).
  • CT saturation and directionality checked for any device on a multi-source or ring topology, not assumed away.
  • The short-circuit method (ANSI or IEC) used for the underlying fault currents is stated in the report, and consistent throughout.

示例提示词

帮我检查这几级保护装置的时间电流曲线有没有选择性配合问题。

所属插件

电子电气与信息工程面向电气与电子工程研究:电力系统仿真(PSS/E、PSLF、PowerWorld、pandapower、PyPSA、OpenDSS、ANDES)与越限/稳定/短路处置,FPGA 时序收敛与跨时钟域检查,电路与 PCB 设计评审,硬件产品开发流程。

安装该插件会一并获得同领域的 63 个 Skill / MCP。

同插件的其他科研 Skill

常见问题

继电保护级差整定能做什么?

Checks relay/breaker/fuse time-current coordination and selectivity margins from short-circuit results. Use when sizing protective devices or diagnosing a miscoordination。

如何安装继电保护级差整定?

在科搭 Codar 左侧点击「插件」,在「组件」页签搜索「继电保护级差整定」并安装;在 Claude Code 或 Codex 中可以发送本页的安装提示词,或下载 Skill 包后解压到 ~/.claude/skills 或 ~/.codex/skills。

继电保护级差整定需要哪些权限?

该科研 Skill没有声明额外权限;执行具体操作时仍由 Agent 按自身规则请求确认。

继电保护级差整定经过官方审核吗?

是。SpeedSkills 收录的每个版本都会记录内容摘要与权限声明,当前版本已通过安全审核。