7.3 Drive-Off vs Drift-Off

Two ways to leave the setpoint without intending to

Almost every DP incident discussion eventually asks: was it a drive-off or a drift-off? The distinction is not academic. The physics differ, the displays differ, and the first correct action differs. Induction MCQs and simulator debriefs love this pair because candidates who only memorise “vessel left position” cannot choose the right emergency response.

Drive-off defined

A drive-off occurs when the thruster system actively produces force that moves the vessel away from the intended position or heading. The plant has energy and thrusters are working — but the command is wrong (or the feedback/sensor picture that generates the command is wrong).

Typical drive-off causes include:

• Faulty position reference accepted with high weight (PRS jump, multipath, wrong target on laser/radar, acoustic jump).
• Reference common-mode failure not caught by voting (all selected PRS share the same error).
• Sensor faults (gyro step, wind sensor spike causing large feed-forward).
• Thruster command/feedback faults (wrong azimuth feedback, frozen feedback with live command, thruster runaway).
• Operator error (wrong setpoint step, unintended mode change, joystick in wrong frame).
• Model/estimator corruption leading to large erroneous force demand.

Mental model: the system is fighting to a wrong place.

Drift-off defined

A drift-off occurs when thrusters cannot produce enough force (or produce none) to counteract environment, so wind, current, and waves push the vessel off position. The thrusters are not driving you away; they are failing to hold you.

Typical drift-off causes include:

• Blackout or major power loss,
• multiple thruster failures / offline thrusters,
• power limitation so severe that residual thrust is inadequate,
• weather exceeding capability (including post-failure residual capability),
• fuel, cooling, or VFD failures that unload thrusters,
• intentional thruster emergency stop that leaves no force (then becomes drift).

Mental model: nature is winning because you cannot push back hard enough.

Side-by-side discrimination (memorise this table)

[!IMPORTANT] Exam trap: “Drive-off and drift-off are the same event.” They are distinct failure modes. Another trap: “Drive-off is about references and drift-off is about heading.” Both are position-loss categories defined by active wrong thrust versus insufficient thrust, not a PRS-versus-heading split.

Emergency responses: why diagnosis matters

If you treat a drive-off like a drift-off — for example, waiting only for more generators while thrusters continue to push you into a platform — you leave the erroneous thrust running. The correct family of actions for drive-off includes:

• Emergency stop thrusters or thruster emergency stops as per vessel procedure,
• Deselect Auto DP / take manual joystick control if that stops the wrong force,
• deselect the bad reference if clearly identified and procedure allows,
• gain control and move to a safe position, then rebuild the system carefully.

If you treat a drift-off like a drive-off — slamming thruster emergency stop when thrusters were already dead or barely alive — you do not fix the problem and may delay power restoration or prevent the few remaining thrusters from helping. Drift-off responses include:

• Restore power (blackout recovery priorities),
• restart thrusters when bus is healthy,
• use remaining thrusters/joystick to clear the hazard if any force exists,
• reduce concurrent risk (stop dive, inform asset),
• accept that weather may move you until capability returns.

Real life can blend modes (drive-off until estop, then drift-off). Still, the first seconds demand the right theme: stop wrong force versus restore missing force.

Classic exam and simulator scenarios

1. PRS jump drive-off: DGNSS steps 20 m; Auto DP thrusts hard toward the false position. Thrusters high; power OK. Correct theme: reject/deselect bad PRS, reduce/stop thrust, take manual control if needed.
2. Blackout drift-off: Main board dies; thrusters stop; vessel drifts with current toward a buoyed hose. Correct theme: blackout recovery, then controlled thrust to clear; not “calibrate laser first.”
3. Thruster feedback drive-off: One azimuth reports wrong angle; TAL commands a force that yaws the vessel off. Correct theme: thruster emergency stop / deselect faulty thruster; investigate feedback.
4. Weather drift-off after failure: Open bus loses one section; residual thrusters insufficient in rising wind. Correct theme: abort critical work, open distance if possible, restore plant; not assume references failed first.

Operator decision process (practical)

When the footprint explodes:

1. Look at thrusters and power first — hard working vs dead?
2. Look at PRS residuals and sensors — jump vs healthy?
3. Look at direction of motion — along thruster force or along weather?
4. Apply the matching emergency procedure (company DP emergency checklist).
5. Protect people and asset — communications, stop external work, clear structure.
6. Only then rebuild Auto DP, references, and task status.

Worked discrimination scenario

A vessel in Auto DP near a platform suddenly accelerates into the weather while thrusters show high load and all gensets healthy. PRS display shows one laser range collapsing. This pattern is drive-off (active thrust chasing a bad relative reference), not blackout drift-off. The DPO deselects the laser, takes joystick, stops the approach into the jacket, then rebuilds references. Had the same footprint growth occurred with black screens and zero thruster RPM, the diagnosis would flip to drift-off and blackout recovery.

Exam traps for drive-off / drift-off

Bottom line: drive-off = thrusters push you off by mistake; drift-off = thrusters cannot hold you so environment pushes you off. Discriminate with thruster activity, power state, and sensors — then either stop bad thrust or restore thrust and manage the drift.