This article provides a comprehensive overview of the Bentley 3500/53 (Over-Speed Card) “3-from-2" (2003) interlock logic configuration—covering everything from hardware architecture, software configuration, voting logic, and relay wiring to triggering and resetting.
I. Hardware Architecture (Based on TMR Triple Redundancy)
1. Hardware consists of 3 3500/53 overspeed cards (same model, e.g., 3500/53-02-00), installed in adjacent slots (e.g., slots 7, 8, and 9).
- 3 independent speed sensors: 3 eddy current or magnetoelectric sensors mounted on the same gear disc, each physically, electrically, and power-supply independent.
- 3500 chassis (3500/05) + dual redundant power supplies (3500/15): Redundant power supplies are mandatory and must comply with API 670.
- Backplane TMR bus: The three cards implement a “two-out-of-three" hardware-level voting mechanism via backplane hardwiring, independent of software or the network.
The three cards are fully peer-to-peer with no master-slave relationship: Each card performs independent data acquisition, independent processing, and independent output to the trip relay. A failure in a single card or probe does not affect the system as a whole: The system automatically degrades to a “two-out-of-one" configuration without tripping the unit. If any two cards simultaneously detect an overspeed condition → hard trip: Response time ≤ 30 ms.
II. Software Configuration (3500 Configuration Software, Key Steps)
1. Preliminary Software Preparation:
- 3500 Rack Configuration Software (v3.35+).
- Connection: Connect the computer’s serial port or a USB-to-serial adapter to the RIM module on the rack; set the key switch to Program mode.
- Principle: The configuration of all three card sets must be identical; configure the first card first, then synchronously copy the settings to the other two.
2. Basic Module Configuration (same for each card):
- Slot Selection: Select 3 adjacent slots; Module Type: Select 3500/53 Overspeed.
- Channel Activation: Channel → Active.
- Sensor Type: Eddy Current: Proximity (200 mV/mil); Magnetic: Magnetic.
- Range: 0–40,000 RPM (based on the unit’s rated speed, e.g., 3,000 RPM).
- Gear Parameters: Teeth per Revolution (e.g., 60), Polarity (sensor polarity).
3. Alarm / Trip Setpoints (Three Levels, Core)
Using a rated speed of 3,000 rpm as an example:
- Alert: 103% = 3,090 rpm; triggers an alarm only, does not trip the unit.
- Danger (Overspeed Trip): 110% = 3,300 rpm; triggers 2-out-of-3 voting, resulting in a forced trip.
- Trip (Ultimate Forced Trip): 114% = 3,420 rpm; the on-board hard relay trips immediately without voting.
- Delay: 0 ms for alert, 0 ms for trip (no delay for overspeed).
4. Voting Mode (2 out of 3, most critical)
- Enter the Voting Logic interface and select 2 out of 3.
- Logic definition: ≥2 cards simultaneously reach the Danger setpoint → trigger a general trip.
- 1 card failure / alarm → automatically masked, does not participate in voting.
- Only 2 normal cards remain → automatically downgraded to 1 out of 2.
- Synchronization Settings: The voting mode for all 3 cards must be exactly the same; mixing 2-out-of-3 and 1-out-of-2 is prohibited.
5. Relay Output Configuration (Hardwiring Core)
- Each 3500/53 unit provides 4 relay channels. Key configurations:
- Alarm Relay: Activates during an Alert, sending an alarm to DEH/DCS.
- Danger Relay (Trip): Activates during a “Danger" event; dry contacts (NO/NC), hardwired to ETS / rapid-closing valve.
- Relay Status: Normally closed (NC), open during a trip (NO), safety-designed.
6. Self-Diagnosis and Fault Bypass
- Sensor Fault: Open circuit / short circuit / signal anomaly → The corresponding card automatically bypasses the sensor and is excluded from voting.
- Card Fault: CPU / power supply / communication anomaly → This card is bypassed, and the remaining two cards operate in 1oo2 mode.
- Fault Alarm: A bypass signal is sent to the DEH to alert maintenance personnel; the system does not trip.
III. “Two out of Three" Voting Logic (Hardware-Level, No Software Delay)
1. Normal Operation (RPM < 3,300):
- 3 sensors → Independent data acquisition by 3 cards → Normal RPM → All “Danger" relays energize (NC closed).
- Trip Circuit: NC contacts of the “Danger" relays on 3 cards connected in series → Circuit closed → Safety oil established → Unit operates normally.
2. Single-card overspeed (1 card ≥ 3300 rpm):
- A single card detects overspeed → The Danger relay on that card opens.
- Only 1 card activates → Condition 2oo3 is not met → General trip is not triggered → The unit continues to operate.
- A single-card alarm is sent to the CCS as a reminder to prevent false trips.
3. Overspeed on Two or More Cards (≥2 cards at ≥3,300 rpm):
- The second card also detects overspeed → Condition 2oo3 is met.
- The “Danger" relays on both cards open simultaneously → The series circuit is broken → Safety oil pressure is relieved → Speed-closing valves / main steam valves close.
- Full-chain response time ≤30 ms, far exceeding that of PLC/DCS (200–500 ms).
4. Ultimate Hard Trip (≥3,420 rpm):
If any single card detects ≥3,420 rpm → the card’s internal hard relay actuates directly, bypassing the voting process to force a trip. This serves as the final line of defense against extreme overspeed (e.g., uncontrolled load shedding).
IV. Hard-Wired Configuration (2oo3 Circuit—Must Be Correct)
1. Relay Contact Wiring (Safety-Critical)
- Danger relay per card: NC (Normally Closed), NO (Normally Open), COM (Common).
- Two-out-of-three circuit: NC contacts from 3 cards connected in series → one end connected to 24 VDC, the other end connected to the ETS trip coil / speed-closing valve solenoid.
- Principle: Normally closed, opens upon trip, conforming to the fail-safe principle.
2. Signal Exchange (Hardwired + Communication) to DEH/DCS
- Hardwired: Trip dry contacts, alarm dry contacts, card fault / bypass dry contacts.
- Communication: Real-time speed, peak speed, cause of first trip, card status.
- From DEH/DCS Hardwired: Reset signal, test enable signal.
- Communication: Unit operating status, speed setpoint.
V. Triggering and Reset Procedures
1. Overspeed Triggering Procedure:
- Speed ≥ 3300 rpm → Detected by 2 or more cards → Passed by 2003 vote.
- 2-card “Danger" relay opens → Hard-wired circuit opens → ETS solenoid valve actuates.
- Safety oil pressure relief → Rapid closure of main steam valve / regulating valve / quick-closing valve → Unit shutdown.
- 3500 System Log: First triggering card, peak speed, action time, fault cause.
- DEH Interlock: Interlocks, pressure relief, anti-surge valve fully open, alarm pop-up.
2. Reset Procedure:
- Confirm unit shutdown, speed at 0, and fault cleared.
- Local Hard Reset: Short-circuit terminals RST/COM on the 3500/53 (independent reset for each card).
- After reset: “Danger" relay energizes → hard circuit is energized → safety oil is re-established → startup is permitted.
- Remote software reset is prohibited: must be performed locally to prevent accidental resets.
VI. Handling Typical Abnormal Conditions (Preventing False Trips / Failure to Trip)
- Single-sensor open circuit: Corresponding card bypasses → remaining two cards trigger 1oo2 → no trip.
- Single-card hardware failure: Same as above → No trip.
- Simultaneous failure of two probes: 2oo3 condition met → Trip.
- Electromagnetic interference: False alarm on a single card → Bypass → No trip.
VIII. Summary:
The 3500/53 “Three-Input, Two-Output" interlock is a hardware-level, SIL 3, TMR-redundant overspeed protection solution. Through independent data acquisition from three cards, hardware 2003 voting, and hard-wired direct tripping, it ensures no tripping on a single fault, guaranteed tripping on a double fault, and millisecond-level response. It fully complies with API 670/612 standards and serves as the final, robust line of defense against runaway in large compressors / turbines.