Thermocouple Burnout Detection Done Right: Upscale and Downscale Strategy on Emerson Ovation and Woodward Controls

Why Does an Open Thermocouple Read Noise Instead of Zero?

Thermocouples fail open more often than they fail short. Plant people call it burnout. The danger is subtle. A high-impedance input with no complete circuit picks up surrounding electrical noise. Power lines, motors, and variable frequency drives all radiate it. Therefore, a burnt-out thermocouple does not read a fixed value. It reads a wildly swinging temperature. I saw an exhaust gas thermocouple on a gas turbine swing from 620 to 950 degrees Celsius within seconds. The swing tripped an over-temperature limit in a turbine speed controller and unloaded the machine. However, the thermocouple itself was the only fault. Burnout detection exists precisely to prevent this theater.

How Does the Bias Resistor Create a Defined Failure?

Burnout detection adds a high-value resistor, typically in the mega-ohm range, across the thermocouple input. During normal operation, the thermocouple circuit shunts the resistor, so the bias current has no measurable effect. When the junction opens, the resistor path dominates. The input then drives to a defined extreme, either upscale or downscale, depending on configuration. Upscale means the reading races toward the top of the range. Downscale means it collapses toward the bottom. The choice is a safety decision, not a preference. First, ask what the control system should do on sensor failure. For furnace temperature protection, upscale forces a safe shutdown on high temperature. For feedwater temperature indication, downscale avoids spurious trips. Therefore, review each loop's failure philosophy before choosing the direction. Moreover, never choose a direction by copy-paste from another loop.

• Step 1: List every thermocouple loop and its role: control, protection, or indication only.
• Step 2: Define the safe direction for each loop with operations and process engineering.
• Step 3: Configure the burnout direction per channel, not per system default.
• Step 4: Set a bad-quality alarm on the channel so the failure announces itself immediately.

How Do You Configure Burnout on Emerson Ovation?

Emerson Ovation input cards handle thermocouple diagnostics natively. Dedicated TC and analog input modules report distinct quality states for open-circuit, short-circuit, and out-of-range conditions. First, configure the drop-out state per point in the point builder. Options typically include driving the analog value upscale, downscale, or holding last value with bad quality. For protection points, drive the value in the safe direction and alarm on the quality bit. Second, check the cold junction compensation (CJC) reference. A failed CJC sensor shifts every reading on the card by tens of degrees. That looks like a slow process drift, not a hard failure. Therefore, trend the CJC value alongside process temperatures during commissioning. Finally, use the Ovation diagnostic displays to read per-channel open-circuit counts. A channel that logs intermittent opens during vibration is a thermowell or connector problem, not a burned junction.

• Step 1: Open the point record and verify the burnout or drop-out configuration per TC point.
• Step 2: Confirm the quality alarm routes to the operator alarm list, not only to maintenance logs.
• Step 3: Trend the CJC reference temperature and compare it against ambient expectations.
• Step 4: Review channel diagnostic counters for intermittent open events before replacing sensors.

What Is the Reality of Turbine EGT Protection on Woodward Controls?

Exhaust gas temperature (EGT) protection on turbines deserves special care. Systems like Woodward MicroNet controllers poll EGT thermocouples at high rates and vote across multiples. A single open EGT probe with upscale burnout can drag the average toward a spurious over-temperature event. Therefore, configure the EGT averaging logic to exclude bad-quality channels rather than include their burnout extremes. Moreover, many plants forward EGT data to historians over Modbus TCP (port 502) or Modbus RTU over RS-485 at 9600 to 19200 bps. Verify that the bad-quality flag actually transmits, because a frozen or extreme value in the historian corrupts trend-based maintenance decisions. However, do not rely on the communication path for protection. Protection must live in the controller, on the same scan as the trip logic. Finally, prefer HART or Foundation Fieldbus temperature transmitters where process speed allows. Smart transmitters report open-element diagnostics directly, before the value ever reaches the controller.

• Step 1: Verify EGT averaging excludes channels flagged open or bad-quality.
• Step 2: Bench-test one channel by lifting a lead at the terminal strip during commissioning.
• Step 3: Confirm the operator display shows the sensor failure, not just the derived average.
• Step 4: Check that Modbus-mapped EGT values carry quality flags or companion status registers.

Conclusion & Action Advice

Burnout detection converts a random noise antenna into a predictable, alarming failure. First, audit every thermocouple point and confirm its burnout direction matches the loop's safety philosophy. Second, verify quality alarms reach operators, because a silent upscale value is a trap. Moreover, trend CJC references and channel diagnostics to catch connector and thermowell issues early. However, remember that detection is not protection. Averaging and voting logic must exclude failed channels deliberately. Therefore, test burnout behavior physically, lead by lead, before every startup. Finally, migrate critical points to smart transmitters where you can. The best burnout response is the one that also tells you which sensor failed, and when.

Author: Xu Jianguo is an industrial automation engineer with over 10 years of experience in PLC, DCS, and control systems.