Collision Avoidance, Cones of Uncertainty, and Appropriate CPA

[David M]

All of this is being made unnecessarily complex. It's not.

[evm1024]

Originally Posted by David M

All of this is being made unnecessarily complex. It's not.

Of course it is not all that complex if you want to navigate by the seat of your pants. (no slight intended in saying it that way)

But some want to be more involved in >>knowing<< the details of a crossing situation. That is what this thread could be about.

Take the case of a crossing where the boat is doing 5 kts and the ship is doing 20 but leave out the intent to do a 180' CPA.

Most sailors without AIS would have no idea that a crossing was in the works. The ship is just a "smuge" 4.1 nm or so away (assuming a 90 crossing aspect) and really not part of their gameplan. They do not realize that in just 12 minutes or so they will be in close quarters with the ship. (Or not if the angles or speed are different. ) And have very little idea of how close they will get to each other.

I think that most of us when we think about collisions we think about ships in a cone some 30 degrees either side of our bow.


On another topic - which RV? I've done some work with the Point Sur and a fair bit with Wecoma.

[kristjan]

What ping is saying is actually mostly correct. You take a bearing on a ship, then you begin to aim at it, and slowly come back towards your initial course as it the ship passes across your bow, before collision you slip behind it as close as you dare to. If the ship is indeed going four times your speed it will, soon be long gone.

[kristjan]

Ping also mentions a convoy, presumably in a TSS, where ships will not ever in your wildest dreams change course or speed for you, your job is to cross as close as possible to a 90 degree track and duck behind the ships as they come. Anyone who has crossed the English Channel will confirm this.

[evm1024]

Perhaps my most scary crossing (and I should say my stupidest crossing) was one night about 3 am while heading south down the coast of Washington State.

There are lots of fishing vessels and a fair bit of coastal traffic. Most of the fishing and crabbing vessels were closer in on the shelf (we were about 15 nm offshore).

For the last 3 hours we had a ship about 25 degrees off our starboard bow. It would appear to be for-reaching on us at times and at others falling aft. Its deck lights were lit so is was a bright orange glow in the moonless night.

At first because of the more or less constant bearing it was flagged as a possible collision course target. But because of the time that it hung there off the bow we (I) moved it into the parallel course category. Keep and eye on it but no worry.

Also, the deck light did not appear to be getting brighter or dimmer. And in fact it may have been parallel or at a very acute angle to our path.

As the hours progressed we became used to it hanging there. Every now and again (15 minutes) as is part of the watch keeping "standard orders" I would do a sweep of the horizon and take a look at anything with binoculars.

Normally with binoculars I could see her navigation lights (red in this case) and sometimes her crew working on the deck.

But on this sweep I could see her crew with the naked eye. We were close! I'm not sure that they had anyone on watch. If they had they would likely hailed us on the VHF or sounded their horn 5 blasts. But nothing. Out relative paths were likely just 5-10-15 degrees different and /or they altered course to port to head for Gray's Harbor after making their way down the coast.

Rude awakening from my complacency! Rather than slowing down I did a 360 turn to port (more or less a wear ship) and then crossed her stern at a 90 degree angle as best that I could make it.

At the time it appeared that we were just a few hundred feet from each other but in reality we were more like 1000' from each other. We were on a collision course but it is likely that they would have crossed out path 500' ahead of us. But because of acute angle we would have run side by side for a while.

Scared the crap out of me.

No excuses on my part. I was give way. It goes to show that you can become complacent.

Also, and somewhat noteworthy. I was bringing my new boat home on that cruise. No electronics (that worked) came with the boat. I had a handheld gps bungie corded to the help. The depth sounder was a hummingbird with the transducer epoxied to the hull. The autopilot was attached with duct tape. Well not the wheel portion. I installed that with a drill in Neah bay.

Regards.

[evm1024]

Originally Posted by kristjan

What ping is saying is actually mostly correct. You take a bearing on a ship, then you begin to aim at it, and slowly come back towards your initial course as it the ship passes across your bow, before collision you slip behind it as close as you dare to. If the ship is indeed going four times your speed it will, soon be long gone.

True - also true for the VTS.

The reality is that by aiming for the stern you end up with a 300', 400', 500' CPA. This is just due to the geometry and relative speeds of the 2 vessels.

Most of us get that.

[Lodesman]

Originally Posted by kristjan

What ping is saying is actually mostly correct. You take a bearing on a ship, then you begin to aim at it, and slowly come back towards your initial course as it the ship passes across your bow, before collision you slip behind it as close as you dare to. If the ship is indeed going four times your speed it will, soon be long gone.

Kind of missing the point of the whole thread. If the ship is going 4X your speed, and risk of collision exists, you are about 15º on that ship's bow; you can't see its stern; pointing at it should keep it from hitting you, but stepping around its stern?!? You're going to pass it on a near-reciprocal course (presumably red to red); by the time his stern is at your beam, the risk is clear and past and you can resume your original course.

[kristjan]

Originally Posted by Lodesman

Kind of missing the point of the whole thread. If the ship is going 4X your speed, and risk of collision exists, you are about 15º on that ship's bow; you can't see its stern; pointing at it should keep it from hitting you, but stepping around its stern?!? You're going to pass it on a near-reciprocal course (presumably red to red); by the time his stern is at your beam, the risk is clear and past and you can resume your original course.

That would be what I described

[Dockhead]

Originally Posted by evm1024

True - also true for the VTS.

The reality is that by aiming for the stern you end up with a 300', 400', 500' CPA. This is just due to the geometry and relative speeds of the 2 vessels.

Most of us get that.

Sure. But if there is a big difference in speed, you'll never get nearly that close to him, by aiming at his stern, from anywhere you can SEE his stern. As I think you explain in a different post.

That's why Ping's method works well. Keep the stern of a faster vessel in view, and aim at it, and you can't go wrong. Depending on the difference in speed, you won't get anywhere near him which, of course, is very safe.

[Exile]

Originally Posted by Dockhead

Sure. But if there is a big difference in speed, you'll never get nearly that close to him, by aiming at his stern, from anywhere you can SEE his stern. As I think you explain in a different post.

That's why Ping's method works well. Keep the stern of a faster vessel in view, and aim at it, and you can't go wrong. Depending on the difference in speed, you won't get anywhere near him which, of course, is very safe.

This approach is a bit ironic for me to be reading about here, since it's always what I've generally done w/o any instruction, except I've usually aimed a bit off the big ship's stern rather than right on it. But that's only because I wasn't all that experienced and so wanted more of a margin. My only point being that it must be more of an intuitive approach rather than one coming from the books. Then again I've generally only experienced one-on-one encounters in open water. I'm sure it can get more complicated.

[Dockhead]

Originally Posted by Exile

This approach is a bit ironic for me to be reading about here, since it's always what I've generally done w/o any instruction, except I've usually aimed a bit off the big ship's stern rather than right on it. But that's only because I wasn't all that experienced and so wanted more of a margin. My only point being that it must be more of an intuitive approach rather than one coming from the books. Then again I've generally only experienced one-on-one encounters in open water. I'm sure it can get more complicated.

Sure. Easy cases are -- easy. Most crossings are not collision courses.

Where you need the knowledge and technique is for the hard cases.

It's easy to assume that you just see a ship and dodge around him. See a steady bearing against a stanchion, and just turn to go behind him. After all, that's how you do collision avoidance in cars -- just stop at red lights and don't run into anything.

And then one day you find yourself in a situation where you can't see which way to turn, or whether you can even get out of the way in time. And the light bulb goes off, that's there's more to it. EVM1024 described his own experience above. I had mine on my very first Channel crossing. That's for those of us, like me, like most of us I guess, who did not get adequate training in collision avoidance.

[Dockhead]

You can see here:

US Navy destroyer collision - Page 62 - Cruisers & Sailing Forums

What can happen when you wait too late to take action, and when you don't have a good grasp of good procedure.

This should be a good warning to those who think you really don't need to know anything -- just dodge out of the way. Sometimes that doesn't work.

There is a point in a crossing on a collision course, where avoiding a collision becomes very difficult, depending on knowing exactly what to do -- which way to turn -- which you might not be able to see. And there's another point, where a collision becomes unavoidable. Note well that those points occur EARLIER for us, because in dealing with faster vessels, it takes more time for us, than it does for them, to open up distance from the point of collision.

Another great lesson from the McCain-Crystal collision is the risk of ships maneuvering into each other. This is one of the main reasons why we are supposed to take action EARLY -- if you leave it until there is an in extremis situation, and both vessels are maneuvering at the same time, you may not be able to avoid the other vessel's maneuvering in an unexpected way.

[Juho]

Originally Posted by Exile

This approach is a bit ironic for me to be reading about here, since it's always what I've generally done w/o any instruction, except I've usually aimed a bit off the big ship's stern rather than right on it. But that's only because I wasn't all that experienced and so wanted more of a margin. My only point being that it must be more of an intuitive approach rather than one coming from the books. Then again I've generally only experienced one-on-one encounters in open water. I'm sure it can get more complicated.

People are quite good at this, also without any instructions, since they need these skills in everyday life. When we walk on busy streets we need to take into account the movements of all the other people, and we don't want to end up in the awkward situation where we don't know if we should turn to the left or to the right, or if we should wait for the others or go first.

[evm1024]

Just for fun and for the visual aspect I've graphed out a crossing. The givens are that the ship is going 4 times your speed, you are on a collision course. The speeds do not matter too much but for sake of the argument we can keep with the 5 kts and 20 kts.

In each plot you are heading North and the ship is heading West. The top graph shows a right angle crossing and the 2 bottom show 2 non right angle crossings.

The curved plots show the path the boat would take if it aimed at the ship. Of course the I'm only plotting 4 course corrections just to make the plot easy. But they are useful anyway.

I'm showing numbers for the boat being 1 nm out from the collision point (which makes the ship 4 nm out). But you can scale it to any distance. It could be cables or whatever.

By aiming at the ship we end up with a XTE of about 2/3 nm that we have to make up. This is not a problem in open water with a long way to go. It does change the aspect for any other ships in the area. This may or may not be a problem.

An alternate is to slow the boat down to 2.5 kts which let's the ship pass ahead of the boat by 1/2 nm but depending on factors this may bring the boat closer than is comfortable to the ship.

However, the real key is - At what point to you realize that your boat and the ship are a collision (or close crossing) course? (Even more of a reason to have AIS)

At the start the ship is over 4 nm away. How many of us recreational skippers are even tracking a ship 4 nm away? That ship does a nm every 3 minutes.

If we only start to take action when the ship is 2 nm out then even pointing at the ship places us with about a 1/4 nm CPA.


Regards

[Dockhead]

Originally Posted by evm1024

Just for fun and for the visual aspect I've graphed out a crossing. The givens are that the ship is going 4 times your speed, you are on a collision course. The speeds do not matter too much but for sake of the argument we can keep with the 5 kts and 20 kts.

In each plot you are heading North and the ship is heading West. The top graph shows a right angle crossing and the 2 bottom show 2 non right angle crossings.

The curved plots show the path the boat would take if it aimed at the ship. Of course the I'm only plotting 4 course corrections just to make the plot easy. But they are useful anyway.

I'm showing numbers for the boat being 1 nm out from the collision point (which makes the ship 4 nm out). But you can scale it to any distance. It could be cables or whatever.

By aiming at the ship we end up with a XTE of about 2/3 nm that we have to make up. This is not a problem in open water with a long way to go. It does change the aspect for any other ships in the area. This may or may not be a problem.

An alternate is to slow the boat down to 2.5 kts which let's the ship pass ahead of the boat by 1/2 nm but depending on factors this may bring the boat closer than is comfortable to the ship.

However, the real key is - At what point to you realize that your boat and the ship are a collision (or close crossing) course? (Even more of a reason to have AIS)

At the start the ship is over 4 nm away. How many of us recreational skippers are even tracking a ship 4 nm away? That ship does a nm every 3 minutes.

If we only start to take action when the ship is 2 nm out then even pointing at the ship places us with about a 1/4 nm CPA.


Regards

This is really nice work!!

One thing worth mentioning -- is that to start pointing at the ship in this scenario, from a collision course, requires a large change of course -- 76 degrees to be exact, even more if you're pointing towards amidships of him.

This is what is extremely hard to visualize for many people -- a collision course with a much faster vessel cannot be eyeballed -- he's coming at you from near your beam, or even behind your beam if your courses aren't exactly perpendicular. How very right you are to say "However, the real key is - At what point to you realize that your boat and the ship are a collision (or close crossing) course? (Even more of a reason to have AIS)" very true and insightful words.

We can all perceive a steady-ish bearing, and have learned how to do that. But from four miles out, I guess you can't see the difference between passing a mile ahead and a mile behind, based on steadiness of the bearing against a stanchion from a small boat in a seaway. Of course this is why God made hand bearing compasses. But even that is fairly imprecise, and takes time and a goodly series of bearings to start to form a picture.

As to our not even perceiving ships at 4 miles out -- we surely need to perceive him and understand how we're crossing by this time. If he doesn't see us, and we're on a collision course, then 4 miles out is a fully developed risk of collision situation. If we are in fact on an exact collision course and we point towards him from 2 miles out, then you correctly say that we still miss him, although 2 1/2 cables is a pretty near run thing. BUT WHAT IF WE'RE ACTUALLY PASSING SOMEWHAT AHEAD? We might be, and we might not be able to perceive it. That can be really dangerous, because you can find yourself in a place where you can't quite get past but you can't turn back already.