Interpretation of causation factors in IWRAP

By Markus Porthin, VTT, 18 February, 2010

The main idea on estimating grounding and collision frequencies in IWRAP (extracted from [1])

Today most risk models for estimating the grounding or collision frequency are rooted in the approach defined by Fujii et al. and by MacDuff. That is, the potential number of ship grounding or ship-ship collisions is first determined as if no aversive manoeuvres are made. This potential number of ship accidents is based on:

1. an assumed or prespecified geometric distribution of the ship traffic over the waterway and
2. on the assumption that the vessels are navigating blindly as these are operating at the considered waterway.

The thus obtained number of potential accident candidates (often called the geometric number of collision candidates) is then multiplied by a specified causation probability to find the actual number of accidents.

The causation probability, which acts as a thinning probability on the accident candidates, is estimated conditional on the defined blind navigation.

Interpretations of causation factors - Collision scenarios.

Accident type	Scenario	Interpretation of causation factor (PC)	Remarks
Headon collision	On a leg segment, two ships sailing in opposite directions would collide, if sailing exactly as defined by the lateral distributions. Ships fail to make evasive actions in order to avoid the collision.	Probability that the ships fail to make evasive actions, in a situation when they would collide if they would do nothing. The causation factor is a combination of the individual factors assigned to the two ships involved:	The theory on head-ons is solid and well documented. However, it should be noted that due to the modelling assumptions it is possible e.g. for a single ship to collide with itself.
Overtaking collision	On a leg segment, two ships sailing in same direction would collide due to different speeds, if sailing exactly as defined by the lateral distribution. Ships fail to make evasive actions in order to avoid the collision.	Same as in Headon collision.	The theory on overtakings is solid and well documented.
Crossing collision	Two ships sailing in different directions are on collision course in a crossing situation (waypoint connecting four legs, or more). Ships fail to make evasive actions in order to avoid the collision.	Same as in Headon collision.	We use the information of “from-leg – to-leg” for the two vessels to calculate the crossing angle between the two ships
Merging collision	Two ships sailing in different directions are on collision course in a merging situation (waypoint connecting three legs). Ships fail to make evasive actions in order to avoid the collision.	Same as in Headon collision.	See Crossing collision.
Bend collision	Two ships sailing in opposite directions meet in a bend (waypoint connecting two legs). One of the ships fails to change course at the waypoint, resulting in the ships ending up on collision course, see footnote(1). Ships fail to make evasive actions in order to avoid the collision.	Same as in Headon collision.	The probability of omitting to change course at the intersection is taken as 0.01, see footnote(2). The exact calculation of this scenario is documented somewhat ambiguously in [1] (equation missing). The bend collision model does not account for the lateral distance between the traffic in opposite directions. This leads to too high collision estimates e.g. in TSSs. The possibility of collisions in a bend between two ships sailing in same direction is neither accounted for in the model. PFH (20100311): It is stated in ref. 1 that bend collisions are treated as crossing collisions, but is using a different causation factor. With respect to the comment on the analysis not accounting for the distance between the two lanes, it is noted that the applied causation factor account for the navigator failing to make a turn. The distance between the two lanes may have an influence on the occurrence on a collision, provided that the navigator realises that he has failed to turn and managed to react in time to avoid the collision. To model the situation we would need a checking-time (as used in the powered grounding analysis) – we find this modelling too detailed for a bend analysis. The user is recommended to adjust the causation factor for bend such that it accounts for the increased awareness in the traffic separation scheme.

• Footnote (1) To be precise, only the omission of changing course of the ship in the inner curve is critical. If the ship in the outer curve forgets to turn, the two ships do not end up on collision course.
• Footnote (2) See [1], page 28.

Interpretations of causation factors - Powered grounding scenarios.

Ref. [1] specifies four grounding categories, two of which are illustrated here:

.

Grounding categories I and II are referred to as "groundings" in IWRAP. The same grounding causation factor in IWRAP is applied to powered groundings of both category I and II.

Accident type	Scenario	Interpretation of causation factor (PC)	Remarks
Category I grounding	On a leg segment, the ship would run aground, if sailing exactly as defined by the lateral distribution. Ship fails to make evasive actions in order to avoid the ground.	Omitting to avoid ground, in a situation when the ship would run aground if it would do nothing.	IWRAP Mk2 uses the same Pc for both Category I and II
Category II grounding	The ship would run aground, if continuing straight forward after the leg, omitting the waypoint. Ship fails to change course at the waypoint. Ship fails to notice the omitted change of course before running on ground, or notices the omission too late to be able to avoid the grounding.	Omitting to change course at waypoint.	The ships that notice the omitted turn before running aground are all assumed to be able to avoid the grounding. Another important parameter is the Mean time between checks.

References

[1] Peter Friis-Hansen: IWRAP MK II. Basic Modelling Principles for Prediction of Collision and Grounding Frequencies. Working document, Technical University of Denmark, Date: 2007.08.01, Rev. 4: 2008.03.09

[2] IWRAP Wiki site, IALA. ( this site) http://www.ialathree.org/iwrap