Ships at sea

At last: a Follow-Up to the Preliminary Act - James M. Turner KC


Updates to CPR Part 61 introduce welcome and overdue reforms to Collision Statements of Case, but leave unaddressed the real bottleneck to the full articulation of the issues in ship collision disputes.

The Backdrop

The Admiralty Court has always been special.  Before the reforming zeal of the Victorian age put an end to such anomalies, it was part of the “wills, wives and wrecks” jurisdiction of the court at Doctors’ Commons, a civilian enclave distant in training, thought and geography from the rough-and-tumble of Westminster Hall.  

Among the vestiges of its history is its unique treatment of ship collision disputes.  Known until the CPR came into force as the “Preliminary Act”, the Collision Statement of Case (“CSoC”) comes in two parts.  The first part, which is not technically a statement of case at all, requires each party to make a series of binding admissions regarding key features of the collision: the ships involved; their drafts, speed, course, cargo and engine power; the place and time of the collision; when and how the other ship was first seen; and how the party’s ship was navigated in response.  The answers given to the questions posed in Part 1 cannot be departed from without the Court’s permission – which is rarely and grudgingly given, and even then (generally) only if the amendment makes the amending party’s position worse.

Part 2 of the CSoC is more recognisably a pleading (to use the old school terminology), although one written to a particular formula: Part 1 is repeated; it is alleged that the collision was caused by the other ship’s negligent navigation; and somewhat formulaic allegations are made about that negligence, rounded off with an incantation of the provisions of the Collision Regulations said to have been offended.

There was and remains an art to drafting a CSoC.  That is due partly to the significance of the Part 1 answers, but it is also born of the fact that, then and now, they are settled “blind”, i.e., without sight of the other side’s case.  The parties only get to see their opponent’s CSoC once both (or all) have been filed.  

Before the advent of voyage data recorders (“VDR”), it was very common for the combined effect of the parties’ CSoCs to produce no collision at all – whereas one had most definitely occurred.  The principal enquiry at trial was how the ships had been navigated into collision, leaving relatively limited scope for enquiry as to why.  Yet the why, of course, is at the root of any case on negligence.  Nowadays, the near-ubiquity and ready availability of tracking data has changed all that: the how is usually crystal clear, leaving the way clear to investigate the why much more closely.

That investigation is given rocket boosters by VDR audio (of which more below).  But VDR audio creates case management and pleading issues that are left untouched by the recent reforms to CPR Part 61 (which governs proceedings in the Admiralty Court).  This article outlines those issues and suggests how they might have been addressed.

The Reforms

The term “Preliminary Act” is odd enough.  Odder still, however, is the expectation that it provokes that there must be more to come, that in some way the document is a prelude to something.  But there never was any follow up, at any rate in statement of case terms.  After the CSoCs comes the CMC (case management conference), with nothing in between.  The Court has to distil the real issues in the case from the not-infrequently formulaic and inevitably non-responsive Parts 2 of the CSoCs (helped to some extent by the List of Issues).  This led to the proliferation of Requests for Further Information, Voluntary Particulars and occasionally exchanges of statements of case being ordered on particular issues, but that is all just responsible mitigation of an imperfect system, not a good system working well.  

Reforms to address this state of affairs (and others) have now been made to CPR Part 61, its Practice Direction and Form ADM3 (the CSoC form).  The reforms, which apply in all collision actions commenced after 8 April 2023, come in four parts:

  1. The existing requirement that electronic track data be exchanged pre-CSoC has been bolstered by a requirement that, if only one party has such data, it must provide it.
  2. A further 8 questions have been added to Part 1, the most interesting of which are:
    1. Who was on the bridge for the half hour leading to the collision?  This may reflect the recent judgment in Pacific Pearl v NYK Orpheus [2022] EWHC 2828 (Admlty), in which the judge held that until very shortly before the collision only a pilot was on the bridge of the Panamax Alexander, with no crew there at all.
    2. A full description of the ship’s navigational equipment, whether it was being used and further particulars regarding radar use.
    3. The real-time availability of electronic track data.
    4. Whether a bridge audio recording exists.
    5. Whether the vessel was broadcasting AIS signals and, if not, why not.
  3. Part 2 is required to be fuller and punchier, setting out the sequence of material events pre-collision, all allegations of causative negligence and other fault, and any other material facts and matters relied on.
  4. There must now be a response to the CSoC: a defence must be filed within 28 days.  Any reply (which is optional) must be served within 21 days of the defence.

These are all understandable, sensible and welcome steps.  It was faintly absurd, particularly with the shift in the focus of the enquiry from how to why, that the rules were silent about statements of case once CSoCs had been filed.  This should improve substantially the distillation of the issues by the time the case reaches a CMC.  In practice, it is not hard to imagine that extensions of time (occasionally considerable ones) will be needed for the defence, as seafarers are not always easy to get hold of.

The requirement of an admission as to the existence of bridge audio is both interesting and welcome.  It may be another consequence of the Pacific Pearl case, as the Panamax Alexander oddly had no bridge audio recording, apparently because it had been in another collision the day before, and presumably had no spare SD card to replace the one used up on that occasion.

Bridge audio, however, is the Achilles’ heel of the process – and the unaddressed elephant in the room of these reforms.

VDR Audio

The bridge of a modern commercial ship is wired for sound.  Anything up to a dozen microphones (perhaps more) are embedded to pick up what is being said throughout the bridge, on the bridge wings and over the VHF.  The audio is recorded on, typically, three or four audio channels, and stored on the ship’s Voyage Data Recorder (“VDR”) until it is overwritten when the recording device becomes full.  When there is an incident, a button is pushed and the recording of the two hours or so up to the button push is saved permanently.

It goes without saying that that recording is a potential goldmine for the cross-examining advocate – for here is the real-time recording of what was actually said and discussed on the bridge of a ship in the run up to a significant marine casualty.  It is a primary source for the why focus of collision actions in the third decade of the twenty first century.

VDR systems are typically designed around recording data, as distinct from playing it back.  Playback software is typically clunky, counterintuitive, and bespoke.  It does, though, allow the individual audio channels to be isolated and listened to one at a time.  That is often very necessary, if a conversation on a bridge wing (say) is to be heard over the din of a VHF set.  
This has consequences when it comes to preparing the evidence for trial.  A court is (almost) never interested in listening to a recording; it wants a transcript of what was said.  If seafarers only ever spoke English, things would be simple: a junior member of the legal team could spend a lot of time listening through the different channels of each ship’s recording and typing up what was said.  In practice, however, it is common to be faced with three or four languages on VDR audio: English, Mandarin, Arabic and Tagalog all featured in the Panamax Alexander’s first collision, for example.

To prepare the audio transcript, therefore, it is usually necessary to employ external linguists to listen to, transcribe and then translate the VDR audio.  Not only is that time consuming, but it is generally impractical to set up the VDR playback software on their computers.  Instead, audio files are exported from the VDR data and used instead.  That works reasonably well, provided the recording consists of only one channel at a time and not all four.

Needless to say, however, preparation of the VDR audio transcripts (one for each ship) is extremely time-consuming, difficult and contentious (and correspondingly expensive).  All too often the exercise is started too late: in The Alexandra 1 v The Ever Smart [2017] 1 Lloyd's Rep. 666, the transcripts were not finally agreed until after the trial had started.  Disputes are frequent as to what is being said and how (rendered into English) it should be understood.  It is often necessary to include a “tie breaker” provision in a procedural order for the resolution of such disputes (as a judge will generally be unwilling to have to decide such a point).

Transcripts and CSoCs

Occasionally, early drafts of the transcript are prepared by and exchanged informally between the parties before proceedings are commenced. More commonly, though, work on them does not really start until weeks or months after the CMC, and invariably takes far longer than anyone expected it to.  

And yet the audio transcripts are at the very heart of the case as regards the cross-examination – and, ultimately, the negligence – of those involved in the collision.  Indeed, the arguments that are run at trial – with the enormous added benefit of these transcripts – are often unrecognisable from the ones set out in the CSoC.  That was particularly so in Sakizaya Kalon, Osios David v Panamax Alexander [2021] 2 Lloyd's Rep. 70, where the transcripts were finally agreed about a week before the trial started, and new points were taken which could not have been dreamed of before then.

As enjoyable as this can be for the advocates, it is not conducive to the efficient resolution of disputes. It also plays into the hands of a party wishing to drag its heels.  And yet the rules always required (and do so now in, as it were, bold underlining) parties to set out their cases fully at the outset.  Logically, therefore, there is a powerful argument for requiring the production and agreement of audio transcripts either before CSoCs are settled at all, or – perhaps more suitably – between the filing of what is now Part 1 and Part 2 of the CSoC. That may perhaps not be necessary in every case, but in those (many) cases where it is, the only alternative is a spate of amendments to the pleaded case at what will all too often be a fairly late stage of the proceedings.

As it is, the new rules will be given time to bed in, and practitioners will be given time to get used to them – and this practitioner at least will be left feeling that an opportunity has been missed.


James M. Turner KC is co-author (with Justice Sarah Derrington AM FAAL) of The Law & Practice of Admiralty Matters (OUP).