Case Study – Human Factor – Position Reference Confusion

1 Overview

A DP 2 vessel was engaged in diving operations and had deployed an ROV. The vessel was on Auto DP with two DGNSS and two HPR position references selected in DP Control, all four references were weighted equally.

Both HPRs became unstable after the vessel executed a position change. After a short time, the two DGNSS references began to drift in the opposite direction of the HPRs for no apparent reason.  The bridge crew deselected one DGNSS to recalibrate, followed by the remaining references multiple times over the course of forty minutes.

Total vessel travel from desired position was measured at 37m.

2 What happened?

Soon after the position reference drift off was noticed, #1 DGNSS was deselected by the SDPO.  The reference ‘point of origin’, which was based on the position when #1 DGNSS was initially calibrated, was lost and a new ‘point of origin’ was calculated when #1 DGNSS was re-enabled a few seconds later, based on the newly measured median position.  When the first DGNSS was deselected, HPR1 and 2 took preference over the second DGNSS, and the DP model relied on the incorrect HPR position.

3 What can be concluded?

It was HPR-1 and HPR-2 that were giving erroneous position measurements. They drifted in the opposite direction to DGNSS 1 & 2.  At this stage, the weighting was equal on all four reference systems, which meant the DP model was not able to decide which reference system was sending the erroneous measurements and the DP model was calculating a median position based on the four reference inputs.

After reviewing the log files, the OEM concluded that there were no errors or faults recorded on the DGNSS’s during the event.  A comparison with the survey DGNSS’ verified this conclusion.

During the event #1 DGNSS was recalibrated several times (then subsequently the other position references) which further contributed to the DP model being inaccurate to real time events.

The HPRs started to be unstable in the period prior to the event due to minor vessel movements/thruster wash and the fact that the horizontal distance from the transducer to the transponders was quite high (approx. 3 x water depth).

According to the OEM report, water column conditions and reflections contributed to bad position data in this case.

Some considerations:

• Do not re-calibrate position references when the position reference systems (PRS) are drifting and there is no indication of which system is providing inaccurate measurements, as this will result in new coordinates for the reference ‘point of origin’ and the DP model will not be aware of any position deviation.
• If the position reference is recalibrated, the historical trace will disappear from the DP position plot making it difficult to see if the vessel has moved.
• Operating with 2 x DGPS and 2 x HPR requires extreme caution because the DP model could use equal weighing on all four PRSs, even though one system may provide incorrect inputs.
• During operations, the PRS weighting usually changes depending on the condition of the received signal. If the system had been allowed to adjust the weighting dynamically, it could have identified the poor HPR signals and adjusted accordingly.  However, if both HPR signals were moving away from the DGNSS signals, the system may struggle to determine which pair were incorrect.
• DPO training and experience should have been sufficient to understand that deselecting one PRS would weigh the ‘paired’ system against the other PRS of the same type, so having fixed weighting and deselecting one PRS is effectively guessing which one is incorrect without a thorough review of the individual PRS panels and stations to see if signals were of poor quality.

In general, regardless of whether the PRS is local or global, the first one chosen determines the Reference Origin which is used in the DP controller for internal calculations.

Please refer to the Reference Origin section of an OEMs operator manual, copied below:

To change the Reference Origin, one must deselect all PRS and then select the one PRS you want as Reference Origin.

Each time the Reference Origin is changed, the position setpoint for Auto Pos is changed, as there will be a difference between the PRS calculated position.  This will impact the geographical position of the vessel, which is not a good idea while for example in pipelaying mode or drilling with BOP down. So this change of Reference Origin is tricky.

In addition, the weighting of PRS in the DP controller can be influenced by the DPO by selecting “Reduced GPS”.

Based on experience, the GNSS systems have a higher weight than the local PRS. In some cases, the total weight of 2 GNSS was 90% and the total weight of 2 HPR was 10%.  Therefore the “reduced GPS” function was developed.

4 Guidance that would be relevant

The following IMCA Guidance would be relevant to this case study:

IMCA M117 – The Training and experience of key DP personnel

IMCA M220 – Guidance on operational planning

IMCA M252 – Guidance on position reference systems and sensors for DP Operations