Network issues, not a thing of the past

The DP equipment class 3 vessel was engaged in saturation diving activities, with two divers in the water when this DP incident occurred.

The vessel had been on location for approximately 12 hours. At the time of the event the vessel was operating in auto DP with 4 of 6 thrusters online and enabled in DP. The vessel was operating in open bus-tie configuration with 2 of 4 generators running.

The environment data just prior to the incident was shown as visibility good, initial heading 278°, significant wave height 0.5 m, wind speed 9 knots from 076° and DP current 1.2 knots to 126°.

At 18:23, when the vessel was conducting dive number 2 on the portside, the vessel experienced loss of all online thrusters, thrusters 2, 4, 5, and 6, from DP.

The DPO was notified by numerous alarms activating simultaneously.

Dive Control was notified and instructed to recover the divers. Communication was established with all stations and amber DP alert activated on the DP alert panel.

The Captain attempted to transfer controls to the IJS system until the thrusters became available and were selected into the DP system. The auto DP was re-established 4 minutes after the first thruster was lost in DP.

All divers were safely recovered to the vessel, and the vessel moved outside the platform’s 500m zone to conduct a thorough investigation of the failure.

During the incident, the vessel drifted approximately 10 metres towards the platform, no one was injured. No impact to the environment and there was no damage to equipment or subsea assets.

The vessel had a data event within the thruster control network which resulted in a build-up of data and as a result the four thrusters that were in DP auto-pos mode sequentially lost their DP ready signal and were automatically moved to back-up control.

• Thrusters in back-up control are not available to DP control system or IJS. They are available on manual levers only.
  
  
• Order in which thrusters were lost during event follows ring philosophy, clockwise direction T2 -> T4 -> T6 -> T5 (T1 and T3 tunnel thrusters not running at time of event).

Initial actions:

• Dive control notified (18:26)
  
  
• Divers recovered (18:36 both divers in the bell)
  
  
• All other downlines recovered (tugger, airline, TW) (18:47)
  
  
• ROV on deck (18:54)
  
  
• Move out of 500m zone (exit at 19:15)

At 18:23 hrs data shows the vessel approx. 92.80 metres away from the platform. Vessel regained station keeping at 18:27 hrs. 82.69 metres from platform. A drift of 10.11 metres in 4 mins.

The fibre optic network ring crosses over both DP zones and maintains communication between control cabinets, monitors status of thrusters, and sends communication to bridge panels.

The watchdog function constantly monitors communication between the main controller (DP) and back-up (manual) controllers. The main controller sends a periodic heartbeat signal to the back-up controller to confirm communication is still in place. If the signal is not received within the specified timeframe, the system activates fail-safe condition = back-up control takes over control and deactivates ‘DP Ready’ signal.

If this happens to all thruster(s) under auto DP, then the control goes to the manual levers.

Following the investigation the OEM issued a technical notice that described the effects that can occur if a network data event happens.

• Network re-configuration: Re-configuration of the network topology can occur due to signal loss or erratic performance of the fibre optic cables. During this process, the absence of a defined data path blocks all internal traffic on each switch until a new path has been verified.
  
  
• Switch reboot: When managed switches are rebooted, their internal ports remain inactive until the boot process completes. The duration of the boot sequence can exceed the time required for the safety mechanism in the back-up controller to engage.

To provide continuous uninterrupted Watchdog function in communication between normal controller and backup controller in thruster cabinets the OEM introduced unmanaged switch (dumb switch) in bottom layer of the network hierarchy. This will ensure internal communication / watchdog function between the controllers remains operational even if there is a network issue. Thrusters will remain in DP control in case of similar data event in the future.