1 The appellant’s vessel, the Navigator Aries (the “NA”), is a liquefied petroleum gas tanker. The respondent’s vessel, the Leo Perdana (the “LP”), is a container vessel. Both vessels collided in the Surabaya Strait (the “Strait”) on 28 June 2015, at 23:17:51 hours local time. This appeal arises from a consolidated action where both vessels’ owners sought to apportion liability for the collision.

2 The vessels had been travelling on reciprocal courses within the Strait. The NA was on a northerly voyage towards the Strait’s entrance, and the LP was on a southerly voyage from the same entrance. Both vessels were under compulsory pilotage.

3 The Strait was marked by lateral buoys. Parties agreed that the buoyed channel represented a narrow channel to which Rule 9 of the International Regulations for Preventing Collisions at Sea 1972 (the “COLREGS”) applied.

4 Within the buoyed channel was a dredged channel about 150m in width and up to 13m in depth. Dredging works were completed only shortly before the collision, and the British Admiralty Chart used on board both vessels did not reflect the channel’s post-dredging bathymetry. Nevertheless, the parties’ navigation experts – Captain John Nicholas Duncan Simpson (“Capt Simpson”) for the appellant and Captain Keith Hart (“Capt Hart”) for the respondent – opined that the vessels’ pilots would likely have been familiar with the channel’s updated dredged conditions.

5 At around 23:12:23 hours, both pilots agreed over VHF radio that their vessels would pass port-to-port. Over the next few minutes, they ordered a series of minor alterations of course to starboard. Up to the last minute before collision, the bridge on both vessels was relatively calm.

6 At 23:16:40 hours, the LP’s pilot gave a “midships” rudder order, which entailed taking off her starboard helm. The LP’s rudder eased from 17.4 to starboard at 23:16:40 hours, to 0.1 to starboard (effectively amidships) at 23:16:44 hours. At 23:16:46 hours, the LP’s pilot gave a “steady” order, which left her helmsman to decide how to steer the vessel with a view to maintaining her existing heading. Her helmsman responded by applying a port helm. The LP also started to experience a port sheer at 23:16:46 hours. This was a relatively slow, but ultimately uncontrollable, swing to port. Her engine was at full ahead, and her speed over ground was 13.9 knots. Over the next minute, the LP’s pilot gave various orders for the vessel to be brought to starboard, and subsequently, for her engine to be stopped. Those on board the NA also noticed the sheer and took avoidance action by altering hard-to-starboard.

7 These manoeuvres ultimately failed to prevent the collision, and at 23:17:51 hours, the LP’s bow struck the NA’s port side at an acute angle.

8 The High Court judge (the “Judge”) found the collision’s proximate cause to be the LP’s port sheer. In turn, the sheer resulted from hydrodynamic interaction between the LP and a bank lying on her starboard side – a phenomenon known as “bow cushion effect”. However, it was the NA that created the situation of difficulty or danger by failing to comply with Rule 9 of the COLREGS and forcing the LP to a position where she would experience bow cushion effect. The NA was moreover in breach of Rules 5 and 7 by taking insufficient action despite her radar showing that she would not achieve the agreed passing, and by taking too long to react to the LP’s port sheer. Against this, the LP had been travelling at an excessive speed, which contributed to the bow cushion effect. Ultimately, the NA’s faults bore greater causative potency and culpability, and the Judge apportioned liability at 70:30 in the LP’s favour.

9 On appeal, the appellant did not dispute the NA’s breaches of Rules 5, 7 and 9(a) of the COLREGS. Its case was that the LP’s pilot’s “midships” order amounted to a novus actus interveniens that broke the chain of causation between the bow cushion effect and the port sheer. Furthermore, it was not the NA’s breach of Rule 9(a) that led the LP to the edge of the dredged channel where she experienced the bow cushion effect; the LP would have been at that location anyway had the LP complied with Rule 9(a). Furthermore, the Judge erred in failing to factor in the LP’s other causative faults (namely, breaches of Rules 2, 8, 9 and 14 of the COLREGS) and in failing to attribute greater causative potency to the LP’s excessive speed.

10 Under s 1(1) of the Maritime Conventions Act 1911 (Cap IA3, 2004 Rev Ed), liability is apportioned based on a broad, commonsensical and qualitative assessment of the culpability and causative potency of both vessels. Culpability is concerned with the nature and quality of each vessel’s faults, and not the number of faults as such. Only causative fault is relevant: at [54]–[58].

11 The Judge correctly concluded that the collision’s immediate cause was the LP’s port sheer, the physical cause of which was the bow cushion effect experienced by the LP. Based on the expert evidence, due to the sloping edge of the dredged channel on the LP’s starboard side, the LP had a smaller underkeel clearance on her starboard side than her port side. The resulting uneven pressure field caused the LP to sheer to port: at [62]–[67].

12 While the port sheer commenced at 23:16:46 hours, the evidence suggested that the bow cushion effect operated on the LP as early as 23:14 hours. From 23:14:40 hours, the LP consistently failed to achieve the headings ordered by her pilot in a timely manner. Given the LP’s speed, the minor alterations ordered should have only taken a few seconds to achieve. Moreover, her pilot saw a need to order further alterations even before earlier ordered headings were achieved. These were not due to inadequate helm action or defective steering gear and machinery. The objective inference was that the LP was experiencing some form of rejection on her starboard side: at [71]–[76].

13 The presence of bow cushion effect from about 23:14 hours should have been readily apparent to those on board the LP, particularly her pilot, considering the regular alterations ordered, the large rudder angles applied, and the fact that the vessel was proceeding at full speed ahead: at [77]–[78].

14 The LP’s pilot should not have given the “midships” order at 23:16:40 hours. This was a deliberate navigational order, which the pilot would have known would remove the starboard helm and affected the rudder’s ability to counteract the hydrodynamic forces. There was no good reason for the order. The Court preferred Capt Simpson’s evidence that the vessel’s wheel should have been at 20° to starboard to maintain her heading. Capt Hart’s evidence that “midships” and “steady” are normal navigational orders ignored the fact that the LP was not navigating under normal conditions at the material time, but already experiencing bow cushion effect: at [81]–[94].

15 The LP’s pilot’s “midships” order substantially contributed to the port sheer and was in breach of Rules 8(a), 8(c) and 8(d) of the COLREGS. Arresting the LP’s alteration to starboard brought both vessels closer together in what was already shaping to be a close quarters situation: at [96].

16 The Court made no finding on whether the “midships” order also breached Rule 14 of the COLREGS. In any event, a finding of breach would not have added much to the appellant’s case, given the Court’s findings on Rule 8: at [97]–[103].

17 It was unclear whether the appellant was maintaining its argument that the “midships” order amounted to a novus actus interveniens. In any case, the “midships” order did not rise to the level of a novus actus interveniens: at [104]–[105].

18 Rule 9(a) of the COLREGS requires a vessel to keep as near to the outer limit of a narrow channel that lies on her starboard side as is safe and practicable (the “Limit Requirement”). This departs from the position under Rule 25(a) of the International Regulations for Preventing Collisions at Sea 1960, which notionally divided narrow channels into dual carriageways and required a vessel to navigate on the “lane” to her starboard side (the “Lane Requirement”). This departure is clearly and unambiguously reflected in the plain wording of Rule 9(a) of the COLREGS. The Limit Requirement also gives effect to the object and purpose of the COLREGS, ie, the promotion of safe navigation and prevention of collisions and close quarter situations. It promotes safe navigation by positively mandating that vessels achieve the widest clearance that is safe and practicable, and avoiding the difficult exercise of having to determine the mid-channel line of the narrow channel. The flexible and situational character of the Limit Requirement also better accommodates the diversity of localised concerns and changing conditions that narrow channels present. While Rule 9(a) does not embody the Lane Requirement, as a practical matter, a vessel found in the “lane” on her port side of the channel would likely face difficulties in demonstrating that she was as far to the starboard limit as was safe and practicable: at [113]–[136].

19 Rule 9(a) of the COLREGS did not require the LP to have gone further starboard past the western edge of the dredged channel. While the LP managed to pass further to the west of the dredged channel without incident after the collision, she was facing markedly different and difficult circumstances that would not ordinarily guide a vessel’s navigation track. There was also insufficient evidence of it being practicable for the LP to navigate in those waters at commercial speeds. The appellant’s other arguments on Rule 9(a) were inconclusive: at [137]–[155].

20 It was unnecessary to determine whether the LP breached Rule 2 of the COLREGS, as any such breach would not have been causative of the collision or relevant to apportioning liability. In any case, the appellant’s arguments on Rule 2 were speculative: at [156]–[160].

21 The LP breached Rules 6 and 8(e) of the COLREGS by travelling at an excessive speed and failing to slow down sooner. Her speed through water was consistently above 16 knots until 23:17:03 hours, and the navigation experts agreed that this was excessive. The LP’s pilot only ordered that her engine be stopped at 23:17:20 hours, some 25 seconds after the port sheer commenced and almost three minutes after he ought to have noticed the bow cushion effect: at [161]–[166].

22 The Judge correctly held that the LP’s speed increased her blameworthiness by contributing to the bow cushion effect. Contrary to the appellant’s submissions, the LP’s speed did not increase her relative blameworthiness in other ways. The NA’s speed had also been excessive. No cogent reason was provided for why the LP’s causative fault would have been greater than the NA’s: at [162] and [167]–[174].

23 The appellant did not challenge the Judge’s finding that the NA breached Rule 9(a). This breach carried causative fault. It limited the navigable sea room available to the LP in the narrow channel and constrained the ability of both vessels to execute the agreed port-to-port passing safely: at [176]–[183].

24 The appellant did not challenge the Judge’s findings that the NA was in breach of Rules 5 and 7 of the COLREGS. The Judge’s findings were left undisturbed: at [184]–[186].

25 The appropriate apportionment of liability was for both vessels to bear equal blame. The Judge’s apportionment at 70:30 in the LP’s favour placed too much weight on the NA’s breach of Rule 9(a) and none on the LP’s pilot’s “midships” order: at [175] and [192].

26 Undue weight should not have been placed on the NA’s breach of Rule 9(a). It was significant that parties had reached a passing agreement: a relatively close passing would have been expected, and the LP would have known of the NA’s position to the west of the Strait’s recommended track at the time the passing agreement was reached. The expectation must have been that both vessels, guided by experienced pilots, would have behaved in a seamanlike manner, and that each would conscientiously and progressively make the necessary alterations to starboard. Indeed, both pilots appeared confident that their vessels would pass safely, and from their courses, the vessels appeared to be shaping to complete the port-to-port passing: at [193]–[199] and [212]–[213].

27 What ultimately brought the vessels rapidly towards a collision was the LP’s port sheer. There was high causative potency in the LP’s failure to detect and react to the bow cushion effect earlier, which was exacerbated by her excessive speed. Instead, her starboard helm was deliberately and erroneously removed by her “midships” order: at [200] and [214].

28 The Judge also placed undue emphasis on comparing the number of alterations taken by the LP with that of the NA. These alterations had occurred before the port sheer, which was the immediate cause of the collision, and were thus of relatively lower causative potency: at [201].

29 For the foregoing reasons, the LP was to bear at least 50% of the blame. Against this, the NA’s breach of Rule 9 was also of real significance, in limiting the sea room and time available to the LP to overcome her port sheer. Additionally, both vessels committed serious breaches of Rule 6 by proceeding at excessive speeds. The Court concluded that a 50:50 apportionment was most appropriate in the circumstances: at [215]–[217].