COLREGs – Restricted Visibility

 

Visibility does not reduce the rules.
It removes the ones most mariners trust.

 

Introduction

Every crossing, overtaking, and head-on scenario practised in clear weather relies on a single architectural assumption: that both vessels can see each other. The moment that assumption fails — fog, heavy rain, snow, sandstorm, smoke — the framework built by Rules 13 through 18 ceases to apply. What replaces it is Rule 19, and it is a fundamentally different system of obligation.

Most mariners know this in the abstract. Fewer have internalised the operational consequences. In restricted visibility there is no stand-on vessel. There is no give-way vessel. There is no right of way. There are only vessels detected by radar, each independently obliged to take action that avoids a close-quarters situation. The psychology is different, the timing is different, and the errors are different — and they are routinely fatal.

This article addresses Rule 19 in depth, the distinction between sound signals in sight and in fog, the practical meaning of safe speed when the world is a radar display, and the specific failures that keep appearing in casualty investigations despite decades of supposed training.

 

Contents

• 1. The Architecture of Rule 19
• 2. Why the Give-Way / Stand-On Framework Disappears
• 3. Safe Speed Under Radar — Not a Number, a Discipline
• 4. Sound Signals: Rule 35 vs Rule 34 — Two Different Languages
• 5. The Port Helm Problem
• 6. AIS as a Classification Crutch
• 7. Radar Detection Is Not Situational Awareness
• 8. Closing Reality

 

1. The Architecture of Rule 19

Rule 19 applies to vessels not in sight of one another when navigating in or near an area of restricted visibility. That phrase — in or near — matters. A vessel in clear air approaching a fog bank is already governed by Rule 19 if other vessels within that fog bank could be affected. The trigger is not the master’s windscreen. It is the condition of the operating environment.

Rule 19(a) sets the scope. Rule 19(b) demands safe speed adapted to the prevailing circumstances and conditions of restricted visibility — a restatement of Rule 6 with specific contextual force. Rule 19(c) requires that when complying with Section I (the steering and sailing rules applicable in any condition of visibility), a vessel shall have due regard to the prevailing circumstances and conditions of restricted visibility.

Rules 19(d) and (e) contain the collision avoidance logic itself, and they are where the operational divergence from clear-weather rules becomes absolute.

Rule 19(d): a vessel which detects by radar alone the presence of another vessel shall determine if a close-quarters situation is developing and/or risk of collision exists. If so, she shall take avoiding action in ample time. If that action consists of an alteration of course, the following shall so far as possible be avoided:

• (i) an alteration of course to port for a vessel forward of the beam, other than for a vessel being overtaken;
• (ii) an alteration of course towards a vessel abeam or abaft the beam.

Rule 19(e): except where it has been determined that a risk of collision does not exist, every vessel which hears apparently forward of her beam the fog signal of another vessel, or which cannot avoid a close-quarters situation with another vessel forward of her beam, shall reduce her speed to the minimum at which she can be kept on course. She shall if necessary take all her way off and navigate with extreme caution until the danger of collision is over.

That is the entire framework. No stand-on obligation. No give-way obligation. No crossing rule. No head-on rule.

Rule 19 does not assign roles. It assigns responsibilities — to every vessel independently.

 

2. Why the Give-Way / Stand-On Framework Disappears

The logic is straightforward once stated plainly. Rules 13–18 depend on aspect. A crossing situation requires one vessel to see the other’s sidelights or determine the relative heading visually. A head-on situation requires both vessels to see corresponding lights or daylight aspect. An overtaking situation requires the overtaking vessel to see the stern light or to assess that she is coming up from abaft the beam of the other vessel.

None of that assessment is possible by radar alone.

A radar contact is a bearing, a range, and — with ARPA — a calculated course, speed, CPA, and TCPA. It is not an aspect. The blip at 015 degrees relative could be heading 180, 270, or 045. ARPA provides an estimate, but ARPA estimates carry lag, error, and assumption. They are not equivalent to seeing a red sidelight.

This is why the give-way/stand-on model cannot apply. It is not a policy decision. It is an epistemic limit. The information required to invoke Rules 13–18 does not exist when the only detection method is radar.

The consequence is bilateral obligation. Both vessels must act. Neither has the right — or the duty — to maintain course and speed. A master who holds course because ARPA suggests the other vessel is the give-way vessel in a crossing situation is applying clear-weather logic to a fog-weather problem.

The stand-on obligation was never designed for a target detected only on a screen.

 

3. Safe Speed Under Radar — Not a Number, a Discipline

Rule 6 lists the factors relevant to determining safe speed. For vessels with operational radar, Rule 6(b) adds: the characteristics, efficiency and limitations of the radar equipment; any constraints imposed by the radar range scale in use; the effect on radar detection of the sea state, weather, and other sources of interference; the possibility that small vessels, ice, and other floating objects may not be detected by radar at an adequate range; the number, location, and movement of vessels detected by radar; and the more exact assessment of visibility that may be possible when radar is used to determine the range of vessels or other objects in the vicinity.

Rule 19(b) reactivates this entire calculus with particular urgency.

Safe speed in restricted visibility is not a fixed number. It is not half-ahead. It is the speed at which the vessel can detect a developing close-quarters situation and take effective avoiding action before that situation becomes a collision.

This means safe speed is a function of radar performance, traffic density, sea room, the vessel’s stopping distance at the current speed, and the rate at which new contacts appear. In a TSS approach in fog with heavy traffic on a 6-mile range scale, safe speed may be six knots. In open ocean with clear radar and no contacts, it may be full sea speed. The point is not the number. The point is the reasoning.

What the investigations show repeatedly: speed was not reduced early enough. The vessel entered fog at passage speed and maintained it because the radar picture looked clean. Then a contact appeared at four miles with a high closing rate, and there was not enough time or space to manoeuvre. The speed that was safe five minutes ago was not safe once the situation changed — and nobody reassessed.

Safe speed is not a setting. It is a continuous judgement.

The obligation under Rule 19(e) to reduce to bare steerage way — or to stop entirely — when a fog signal is heard forward of the beam is the hard backstop. It exists because, by the time a vessel hears a whistle signal forward without radar confirmation of what made it, the situation is already dangerously close. The Rule does not suggest reducing speed. It demands it.

 

4. Sound Signals: Rule 35 vs Rule 34 — Two Different Languages

This is one of the most commonly blurred distinctions in the COLREGs, and the blurring has real consequences.

Rule 34 governs manoeuvring and warning signals made by vessels in sight of one another. One short blast: I am altering my course to starboard. Two short blasts: I am altering my course to port. Three short blasts: I am operating astern propulsion. Five or more short and rapid blasts: doubt signal. These are action signals. They communicate intentions and manoeuvres between vessels that can see each other.

Rule 35 governs sound signals in restricted visibility. They are identity and status signals, not action signals. One prolonged blast at intervals of not more than two minutes: power-driven vessel making way through the water. Two prolonged blasts at intervals of not more than two minutes: power-driven vessel underway but stopped and making no way. One prolonged followed by two short blasts: vessel not under command, restricted in ability to manoeuvre, constrained by draught, sailing vessel, vessel engaged in fishing, vessel engaged in towing or pushing. And so on for specific categories.

The critical distinction: Rule 35 signals do not communicate manoeuvring intentions. A prolonged blast in fog means I exist, I am here, I am a power-driven vessel making way. It does not mean I am maintaining course. It does not mean I intend to pass on your starboard side. It carries no information about what the vessel is about to do.

A vessel in fog that hears a prolonged blast and interprets it as a Rule 34 signal — or expects the other vessel to respond to a manoeuvring signal — is operating under a fundamental misunderstanding. The two rule sets are entirely separate communication protocols for entirely separate visibility regimes.

A fog signal tells you what something is. Not what it intends.

The practical implication: in restricted visibility, VHF is sometimes used to try to negotiate passing arrangements. This is dangerous for reasons discussed elsewhere on this site. But the temptation to use VHF in fog often arises precisely because Rule 35 signals carry no action content. Masters want to know what the other vessel is going to do. The answer is: so does the other master. Rule 19 requires both to act independently. The absence of a negotiation mechanism is not a gap in the rules. It is the design.

 

5. The Port Helm Problem

Rule 19(d)(i) is one of the most important prohibitions in the COLREGs, and it is one of the most frequently violated.

For a vessel detected by radar forward of the beam — other than one being overtaken — an alteration of course to port shall, so far as possible, be avoided.

The logic is geometric and probabilistic. If a contact is forward of the beam and on the starboard side, a port alteration increases the rate of convergence. If the contact is forward of the beam and on the port side, and the other vessel is also altering to starboard as Rule 19(d) would encourage, a port alteration by the first vessel leads both vessels towards each other. The only scenario in which a port alteration for a contact forward of the beam reliably opens CPA is when the geometry is fully known — and in restricted visibility, it is not.

Starboard alterations are strongly favoured because they increase the probability of a safe outcome across the widest range of unknown geometries.

The failure pattern is consistent: the OOW sees a contact at, say, 020 degrees relative, ARPA showing a CPA of 0.3 miles. The OOW decides to go to port to pass astern. The reasoning mirrors clear-weather crossing logic — give way by going under the stern. But this is not a crossing situation as defined by Rule 15. Rule 15 does not apply. Rule 19(d) does. And 19(d)(i) says do not alter to port.

The phrase so far as possible is not a loophole. It acknowledges that navigational constraints — a coastline to starboard, a TSS boundary, shallow water — may physically preclude a starboard alteration. In that case, speed reduction under 19(e) or a combination of speed and minimal course change may be the only option. But the default is unambiguous.

Port helm for a forward contact in fog is not a manoeuvre. It is a gamble on unknown geometry.

 

6. AIS as a Classification Crutch

AIS provides name, MMSI, position, course over ground, speed over ground, heading, rate of turn, navigational status, destination, draught, and vessel type. It is enormously useful for traffic monitoring and identification. It is not a substitute for radar in the context of Rule 19, and treating it as one is a growing source of error.

The problem manifests in a specific way: an OOW detects a radar target, overlays AIS data, sees that the target is (for example) a fishing vessel on a roughly reciprocal course, and then begins to apply crossing or head-on logic from Rules 13–18 on the basis that AIS has revealed the other vessel’s aspect and type.

This is wrong on multiple levels.

First, AIS navigational status is manually set and frequently inaccurate. Vessels reported as underway using engine may be engaged in fishing. Vessels reported as at anchor may be making way. The data field is only as reliable as the last time someone on the other bridge updated it.

Second, AIS COG and heading are not equivalent to visual aspect. Knowing a vessel’s COG does not tell you which sidelight you would see if you could see her. It tells you the direction her ground track points, which may differ from her heading due to current, wind, or a rate of turn.

Third, and most fundamentally: Rule 19 applies when vessels are not in sight of one another. AIS does not make a vessel in sight. The phrase in sight in the COLREGs means visible by eye or, as extended by case law and practice, by visual means. A data link does not constitute visual contact. The vessel on the screen is still a radar-detected vessel subject to Rule 19, regardless of how much AIS metadata is attached to the plot.

The operational failure is subtle because the information feels like knowledge. An OOW with AIS overlay feels that they understand the situation in a way that a radar-only picture does not provide. That feeling is partly justified — the information is real and useful. But the feeling also encourages the reintroduction of give-way/stand-on thinking, which Rule 19 explicitly displaces.

AIS tells you what a target says it is. It does not tell you what it will do.

 

7. Radar Detection Is Not Situational Awareness

A clean radar picture with well-tracked ARPA targets, AIS overlays, and guard zones set does not constitute awareness of the situation. It constitutes awareness of what the system has detected and how the system has interpreted it.

Small craft — yachts, wooden fishing boats, partially submerged containers — may not appear on radar at all. Sea clutter suppression, if over-applied, removes real targets along with noise. Rain clutter masks contacts within precipitation cells. A vessel in the radar shadow of a larger contact is invisible until it emerges. And targets in the first few cables — the range that matters most in the seconds before collision — are often lost in the clutter of the minimum range ring.

The master or OOW who trusts the radar picture as a complete representation of reality is making the same category error as the one who trusts AIS as visual contact. The radar shows what it can see. It does not show what is there.

Rule 19(b) requires safe speed adapted to the circumstances. That adaptation must account for what radar cannot detect, not only for what it displays. This is the deeper meaning of the Rule 6(b) factors: they exist to force an acknowledgement that radar has limits, and those limits vary with conditions in real time.

A vessel proceeding at fifteen knots in fog because the radar shows no contacts within twelve miles has not necessarily determined a safe speed. She has determined that nothing detectable by radar is within twelve miles. That is a different statement. The gap between those two statements is where collisions live.

The screen shows the world the radar can build. Not the world that exists.

 

8. Closing Reality

Rule 19 is not a degraded version of the clear-weather rules applied in poor conditions. It is a different system, built for a different epistemic environment, in which the fundamental information required for give-way/stand-on logic does not exist.

The obligations are bilateral. The manoeuvring logic favours starboard turns and speed reductions. The sound signal regime provides identity, not intention. AIS enhances the radar picture but does not change the legal framework. And safe speed is not a number set at the start of the watch; it is a continuous assessment tied to what the radar can and cannot reveal.

Every restricted-visibility collision investigation reaches the same handful of conclusions: speed too high for the conditions, course alteration to port for a forward contact, reliance on technology to provide certainty that the environment does not permit, and the assumption — always the assumption — that the rules of clear weather still apply when the fog closes in.

They do not.

Fog does not change the rules. It reveals which ones the bridge team never properly understood.