The applicability of the Complexity Score in Oceanic Control Areas

It would be of great value to define a similar complexity score for oceanic ANSPs. The complexity score was however designed for continental Europe. Thus there are several issues that must be considered regarding the applicability of this metric in other areas such as oceanic areas.

Considering the dimensions of the cells used to calculate the complexity score, it is necessary to compare these dimensions with the separation minima in each control area.

The Reduced Vertical Separation Minima (RVSM)96 is 1.000 feet. The RVSM airspace is between flight levels 290 and 410 and is applied in Europe including the Reykjavik Control Area.

For en-route traffic, the minimum horizontal separation within radar coverage is from 5 NM (9,3 km) to 10 NM (18,5 km).

The minimum lateral separation outside radar coverage is from 50 NM up to 120 NM under certain conditions.

The longitudinal separation outside radar coverage is between 10 to 30 minutes, depending on the type of aircraft and separation technique used. The longitudinal separation between two aircraft can be reduced where the speed of the trailing aircraft is lower. Depending on the difference in speed the longitudinal separation minima can be reduced down to 5 minutes. Oceanic airspace is usually outside radar range although the Reykjavik CTA has radar coverage in the busiest part of the airspace.

95

H.of Vertical Int. is an abbreviation of Hours of Vertical Interactions.

96 _{Airspace is referred to as RVSM airspace if it is allowed to use 1000 feet vertical separation instead of 2000}

A cell of horizontal dimensions 20 x 20 NM is probably too small to capture interactions between aircraft which are required to fulfil lateral separation minima of 50 NM and longitudinal separation of 10 minutes97.

Considering only one flight level a maximum of 6 aircraft could pass through a 20 x 20 NM cell in one hour (with 10 minute longitudinal separation) using procedural separation. If the same situation in radar separation is considered, presuming the average speed to be 830 km/hour, a maximum of 356 aircraft could pass through the cell in one hour98.

Using too small cell size in procedural separation could therefore lead to very few interactions captured, because of the separation minimum which applies in procedural separation.

The size of the cell was selected by the ACE working group on Complexity because it mapped the boundaries better than larger cell. It would be possible to select larger cells for a study of complexity in oceanic areas. To capture similar portion of interactions, i.e. using the ratio of the difference in separation minima, the cell would have to be 200 x 200 NM. This could however be too large with respect to other factors such as the boundaries between control areas. When the complexity score was established, three sizes of cells were considered 20x20, 40x40 and 60x60 NM. The smallest was chosen because it was considered to map the boundaries of smaller control areas best. The oceanic control areas are large in comparison with the smaller areas in continental Europe and larger cell sizes might therefore be considered.

It should be noted that the selection of cell size has great effect on the calculations of the interactions and the adjusted density. Figure 5.5 showed how the adjusted density is calculated for a 20 x 20 NM cell, presuming that all aircraft are at the same speed and pass through the cell in 3 minutes. In figure 5.8 similar calculations are performed but here a comparison is made between two centres with identical air traffic (and identical size) but in one case there is one 40 x 40 NM cell while in the other there are four 20 x 20 NM cells.

97

If the speed of two aircraft is 830 km/hour, 10 minutes separation is about 74 NM.

98 _{Presuming that the aircraft are 5 NM apart from each other and a line of 4 aircraft enters the cell every 0,669}

Figure 5.8: Adjusted density calculated for different cell size (40x40 in Centre 1 and 20x20 in Centre 2).

When the cell size is doubled, the time in the cell is also doubled, i.e. 6 minutes instead of 3 minutes. The number of interactions increases considerably by expanding the cell size. The results of these calculations show that the cell size affects the number of interactions and the adjusted density; hence, the complexity score is affected. If the cell size is too small, potential horizontal, vertical and speed interactions might not be captured,

It is therefore not possible to compare the complexity score between two ANSPs if the cell size used in the calculation is different.

A further concern regarding the cells is the use of Albers equal area projection; it should be considered whether this projection is applicable in the areas where comparison is to be made99 or whether other kind of projection might be more suitable.

As mentioned previously there are several factors which make comparison between the performance of continental and oceanic ANSPs difficult. It would therefore be of great advantage if a similar complexity metric would be defined for oceanic airspace on the basis of the methodology used in the complexity score. The horizontal, vertical and speed interactions are all relevant in oceanic air traffic. There may be different strategies used in the control centres with respect to route structure. Where there are parallel tracks with few crossing points, the horizontal interactions may become close to zero. This may be done in order to increase the capacity. As a result adjusted density may be relatively high. It would therefore be very interesting to perform these calculations for the oceanic centres which are compared in the CANSO benchmarking report. Presently the ANSPs are classified only by the total IFR flight hours that are controlled over a given period.