Escort Planning

Towing Nautical Institute “Seaways” Article
by Captain Gregory Brooks MNI

To date, large engineering companies, research institutes, research departments of some shipyards and classification societies, working with the major international shippers have done excellent work using live trials and simulation work to quantify the science behind shiphandling and documenting for the mariner its many variables.  Currently, all ships have maneuvering information readily available to the ship officers and pilots that was documented as part of their builder’s trials.  Further, professional mariners are also provided with many formulas for predicting ship behavior such as calculating squat or the amount of bollard pull that will be required to maneuver a ship safely against a certain wind or current force, to give two examples.  Using this plethora of data and information, a mariner can calculate and plan, with reasonable accuracy, his port passage and the maneuvers that will be required as part of that plan. 

However there is one important area of shiphandling that has been left in the dark ages, with many pilots and masters saying to themselves, “I think this will work”, or worse “I hope this will work.”  This important area is the ability to accurately plan an escort transit.

The better escort systems around the world set maximum allowable speed limits based on computer simulations and or live trial data.  However, it is important to note that these transit speeds are broken down into various classes of vessels based on gross, deadweight or displacement tons.  Other systems simply require a certain amount of tug bollard pull based on the ships size and / or expected current conditions during the passage.  Most systems fail to specifically match the escort tug’s ability to apply steering and/or braking forces through its towline to the ship at the allowed transit speed based on the energy demands should the ship suffer a serious mechanical failure at that speed.  [While escort systems are generally designed to address mechanical failures there may also be the need for the escort tug to respond to an incident caused by a human failure on the ship’s bridge.]

Further, most escort systems do not specifically describe to the shipmaster or vessel owner what the intended results of their escort system is.  For example, is the system, as designed, simply to mitigate the incident by reducing the speed of an incident, or is it capable of preventing a grounding if only the ship’s engine fails as opposed to total system failure including engine, bowthruster and rudder.  The ultimate escort systems, such as the one in Prince William Sound, are designed to be able to “save” a ship traveling at the approved escort transit speed if it suffers a hard over rudder failure before the ship touches the grounding line to either side of the ship.  The master of the ship should know what kind of escort system his ship is participating in, for him to ensure that escort systems worst case scenario is acceptable to his owner.  If the escort’s mission does not meet the owner’s safety desires some adjustments to the escort should be made to ensure a reasonable chance of success as defined by the ship owner.  However, if this shipmaster asked the pilot to alter the escort plan based on the escorting information that is generally known today they would be stymied by the lack of information.  This lack of information also inhibits planners ashore from evaluating the effectiveness of various escort systems or in the planning of new terminals such as the many LNG facilities being constructed around the world at the present time.

First, most escort tugs around the world proudly advertise their bollard pull both ahead and astern.  While this information is very useful in planning how to handle the ship in the prevailing wind and current when maneuvering into or out of its berth, using this measure to predict escorting results can be inaccurate and misleading, especially at speeds above 5 knots as these forces are measured at zero speed.  The mechanical power of the tractor is only one factor in its ability to apply high towline forces to a ship at higher speeds (above five knots).  As the speed of the escort goes up the ability of the tractor’s hull to also safely create towline forces becomes a very important factor.  Some escort tugs, like a well designed Voith Schneider water tractor, or a modern well designed Z-drive tractor[1]) will be able to apply greater forces than its rated bollard pull at escort speeds of up to ten knots, but there are tugs being offered for escort service that cannot do this due to limitation in their design.  This one size fits all based on bollard pull approach to escorting simply does not meet the safety standards required by today’s professional mariner.

What the mariner actually needs, but generally cannot get, are towline force predictions for the tractor at various speeds[2].  Shown below are two graphs produced by Glosten and Associates of Seattle, WA that can be developed for most tractors for this purpose.  In the first graph the tractor’s maximum steering and breaking forces that the boat can produce are predicted across the speed range from five to eleven knots.  On the second graph, the computer has been used to predict the towline angle to the ship’s centerline that the tug would use to produce these steering and braking forces [this angle would indicate to the tug captain the maneuver that he must perform with the tug to attain this performance].  So we see that the tug performance information can be made available if the customer requests it from the towing companies.  However, this data is only half of the information that the Pilot or Ship Master will need to calculate a truly safe escort speed.

Further complicating this issue is the recognition that these performance estimates are based on calm water conditions that may not exist during the escort passage.  In reality the tug’s performance will degrade as the sea state increases.  Currently, this degradation factor is poorly understood and is another area where mariners are applying their mariner’s eye but help is on the way.  The Maritime Research Institute (Marin) is currently hosting a research project called SAFE TUG where they will use their simulator to begin to identify these degradation factors in tug performance.

The second piece of data that the mariner requires to plan his escort is the amount of towline pull that will be required to overpower a failed rudder and return the ship to its original heading before it touches the grounding line[3].  This data will be ship specific as the handling characteristics for every ship, even of the same tonnage, can be remarkably different.  Here the ship owner could apply his own corporate standard of safety.  If, for example, he simply wants to miss the grounding line he can elect to use a minimal safety margin.  Or an owner who wants to guard against the human factor may want to build a 25% safety margin into this equation.

The escort requirement curves should be based on a comprehensive calculation that would take into consideration factors such issues as:

  1. the directional stability of the ship, based on the ship’s unique underwater shape,

  2. its rudder size and effectiveness,

  3. the ship’s speed and propulsion characteristics,

  4. the kinetic energy of the ship at that speed, based on the ship’s displacement,

  5. the distance to the grounding line,

  6. the type and severity of the disablement, e.g. hard-over steering failure, loss of propulsion, etc.

  7. the typical environmental conditions; such as tidal currents, average wind speed and direction, and wind waves or swells if they are present,

  8. an assumption on how quickly the ship’s crew or pilot will recognize the failure and react (this will determine the amount of rotational momentum that will develop), and

  9. an assumption on how quickly the escort tug will be able to get into position and apply the required force. 

  10.  the assumption that the escort tug(s) will be tethered.

Based on this input, an engineering firm could produce a series of simple graphs that would contain several curves for different displacements that the mariner could use to identify the tractor performance required to escort at a certain speed based on the available channel width.  Conversely, if a tractor of that performance was not available it could be used to identify at what speed the escort must use to stay within the capabilities of the available tractor.

The writer proposes that the IMO consider adding the above mentioned escort requirement curves to the ship’s maneuvering data. 

Of course our industry would have to establish an international standard set of definitions and/or other criteria for these calculations to be consistently documented.  Examples of some of these definitions are:

  1. Ship’s speed for the escort calculations is speed through the water, not over the ground.

  1. A standard recognition / reaction delay to simulate the human factor before the application of the escort tug’s steering forces would be 15s / 15s, e.g. 15 sec from the onset of the failure to the crew’s recognition that a disablement is imminent and 15 seconds from the time of recognition to implementing a response on board and informing the escort tug(s).  A standard delay for the tug to be in position and working must also be considered and adopted.

  1. The off track error of the ship would include the swept path of the ship before bringing it under control.

  1. The steering forces of the tug are to be applied to the transom of the ship.

  1. Speed curves for each ship should be calculated to illustrate the steering forces that would be required to keep the ship safely in the channel should a hard over rudder failure take place for transit speeds between five and ten knots.  Further, these curves should terminate when the off track error exceeds 1000 meters.

  1. The speed curves should be calculated for deep water as this will produce the worst off track errors due to the ship’s more sluggish maneuvering in shallow water.

With this new tool a ship’s master could evaluate the capabilities of the available escort tugs as part of his Port Passage Plan and decide, based on science, what a reasonable safe transit speed should be according to his owner’s safety instructions. 

These suggestions are not submitted as the ultimate solution to this issue, but rather to get the subject onto the Nautical Institute’s table for discussion.  Further, the article is not intended to disparage any current escort system or the pilots that service them, but simply to identify an area in our maritime knowledge base that needs more scientific input.  I look forward to a lively discussion on the merits of this suggestion.



[1] For the purposes of this article the term “tractor” will be used for all tugs capable of providing a steering force to a ship, using a towline, at speeds of five knots and above.  The author recognizes that there are many different terms used for escort tugs today such as “tractor”, “reverse tractor”, “ASD”, “rotor tugs”, etc., but it is felt that these terms do not serve our industry well, as the cross pollination of the better design features into other designs have made these terms useless for a mariner.  This subject deserves a separate article in itself.

[2] Currently Voith Schneider tractors are supplied with engineering predictions on the tractors ability to produce towline forces at 0, 2, 4, 6, 8, and 10 knots.  ABS will also issue an Escort Certificate that certifies engineering performance data submitted to it.  Finally, DNV will issue an Escort Certificate certifying a tractors escorting performance, but only after observing live instrumented trials.  It is the author’s opinion that this latter method should be adopted by all formal escort systems.

[3] This assumes that the ship is operating in a narrow channel.  In a more open area, the pilot might elect to use the tractor to make a round turn with the ship.

The ability to accurately plan an escort transit has been left in the dark ages, with many pilots and masters saying to themselves, “I think this will work”, or worse “I hope this will work.”

Also of Interest