EUROCONTROL Description of the CAPAN Method

DESCRIPTION OF THE CAPAN Method

1.1.EUROCONTROL Airspace Model and RAMS

The method developed for the EUROCONTROL Airspace Model/CAPAN is used with the RAMS simulation engine.

The CAPAN Method has four main types of control and input data:

• Airspace Structure and Route Network,
• Traffic Samples,
• ATC Logic and Procedures,
• Controller Task Definitions.

A typical CAPAN Method Study will start with initial runs for data verification and debugging. The results presented will however be the average of a series of simulation runs – normally 10 to 20 simulation runs are used – where the entry times of the flights and the aircraft performances are made to vary, so as to create different situations in the ATC Sectors analysed. This reduces the possibility that the traffic sample creates a too complex or a not enough complex situation. In the case sectors are not sufficiently loaded, flights can be cloned in a proportional way so as to create a traffic load sufficient enough as to calculate theoretical sector capacities.

The Reorganised ATC Mathematical Simulator (RAMS) is used as a simulation engine to determine controller workload generated for a given traffic sample. This is a critical event model which during the simulation treats a number of defined events in the life-cycle of a simulated flight, for example, entry into the first simulated sector, exit from a sector, conflict search and conflict resolution.

1.1.1.CAPANMethod Workload Thresholds

The CAPAN Method produces values representing the loading in the simulated control positions. These values are used in determining the sector capacities are crucial for the CAPAN Method.

The determination in modelling of qualitative values (heavy load, light load, etc.) from quantitative values (numbers) is always one of empirical experimentation and is a function of the “realism” or “fidelity” of the model being used to the real world that is being simulated. The thresholds used by the ATC Capacity Analyser have been validated and calibrated by several Real Time simulation studies.

The quantitative threshold values used and their corresponding qualitative interpretations are:

Figure 61 – Workload Thresholds

It is important to note that the ATC Capacity Analyser records those workloads associated with identifiable control tasks defined to the model. It does not for example, records a specific task for general radar surveillance of traffic within a sector, nor are recuperation times recorded. The 70% threshold corresponds to 42 minutes measured working time in one hour, leaving 18 minutes time available for other tasks not defined within the model and also for general recuperation.

In development of the ATC Capacity Analyser tool, a number of real time simulations have been conducted using the results of CAPAN (including the actual traffic samples derived). The real time simulations and other operational trials have confirmed the validity of these thresholds in determining sector traffic handling capacity.

1.2.CAPAN Methodology

This section describes the methodology used by CAPAN for the determination of the sector capacities.

1.2.1.Traffic Demand and Controller Workloads

During the simulation process, RAMS calculates the flight profiles, identifies the sectors penetrated, detects and solve the conflicts and records the ATC Workloads generated by the traffic demand on the sector Controllers.

At the end of the simulation, a detailed picture of the ATC Organisation is available for further analysis of the distribution in time of the traffic and controller workloads.

As an illustration of the post simulation analysis, figure 3.1 below shows the activity of the Koksy-Low sector during the 24 hour time period.

Figure 62 - Traffic Demand and Controller Workloads

• The blue curve represents the hourly rates of entry in the sector (number of flights).
• The red curve represents the Executive Controller (EC) workload (expressed as a percentage of 1 hour).
• The green curve represents the Planning controller (PC) workload (in percentage).
• The red line represents the overload threshold of 70% used by CAPAN to identify overloaded working positions.

When considering the analysis presented in figure Figure 6.2 the following comments can be made:

• The Planning Controller (PC) workloads are almost directly proportional to traffic demand. This trend is very common in modern air traffic systems where tools such as SYSCO and OLDI assist the planning function.
• The Executive Controller workloads are not directly proportional to traffic demand. It is clear from this curve that the function EC workload vs. Traffic demand is not a linear function. A deeper analysis shows that EC workload is more dependant on complexity of traffic than on demand.
• The peak period for the Koksy Low sector was observed around 08:30 with about 100% workload for a demand of 70 aircraft. It can be observed that other peak periods were observed in this sector with lower traffic demand. For example around 14H00 and 16H00, the EC workloads exceeded 90% with traffic demands of 55 aircraft and 65 aircraft respectively. This shows that, the peak EC workload does not always correspond to the maximum number of Ac/Hour, but to a combination between number of aircraft and traffic complexity.
• It should be noted that these values are obtained with unconstrained traffic demand and CAPAN has recorded overload values on the EC position for the observed traffic demand

These observations suggest that the sector capacity based on controller workloads varies over a 24 hour period and is directly depending on traffic complexity and not on pure traffic demand.

The derived sector capacity by regression analysis will take into account the average traffic pattern of the sector based on a 24 hour period and is detailed below.

1.2.2.Sector Capacity by Regression Analysis

If we plot the values obtained for the Koksy Low sector during the 24 hour period, and if we draw the best regression line between Traffic demand and EC Workload, we obtain the sector characteristic function EC Work = F( Traffic demand) with the 24 hour traffic pattern. The intersection of this curve with the overload threshold at 70% provides a capacity value which has smoothed out the peak values observed during the peak workload hour.

Figure 63 – Capacity by Regression Analysis

This Regression Analysis sector capacity value is based on the 24 hour average traffic complexity of the sector.

It should be noted that when the workloads recorded during the 24 hour simulation are light, the capacity by regression analysis may give too high values. In fact, as the traffic demand is low, the behaviour of the sector during difficult periods has not been measured and the function traffic/workload is almost linear. For this capacity calculation a minimum of ten iterations of the traffic sample would normally be used.