 |
The Performance Assessment and Technology Validation group (within the EEC's ATC research area) is currently using model based simulations to evaluate the performance of a new technique for merging traffic streams known as Point Merge. This work is performed in conjunction with operational feasibility studies using real-time simulations with TMA controllers from potential implementation centres.
Today, in busy TMAs, air traffic controllers tend to merge several arrival streams by adjusting the individual paths of aircraft so that they fit together with a safe longitudinal spacing. Often paths are stretched tactically using several open loop vectoring instructions to divert aircraft from planned routes and then return them. During these tactical manoeuvres it is difficult to predict an accurate path distance to the runway. Aircraft are descended early in a stepwise fashion just in case their paths have to be decreased at short notice. The earlier aircraft descend the longer they spend at low altitudes where fuel efficiency is poor and the greater the noise levels transmitted to the ground.
The Point Merge technique works on the principal that the default P-RNAV route has already been ‘stretched’ so that at anytime the controller can tactically shorten the path using one simple ‘direct to waypoint’ instruction i.e. short-circuiting a number of intermediate route points. In addition the shape of the route (sequencing leg) has the form of the sector of a circle and the ‘direct to waypoint’ instruction is always to the centre of that circle (merge point). The constant radius property of a circle means that the distance from where the ‘direct to waypoint’ instruction is issued to the runway has always the same fixed value. Optimum descent profiles (Continuous Descent Approaches) are therefore feasible from the moment the direct to merge point instruction is given. |
 |
| Figure 1: Point Merge design with two arrival flows |
|
|
|
|
|
|
Recorded trajectories from real time controller in the loop simulations were used to model the merging of four arrival streams on to a single runway with the fast-time simulation platform RAMS Plus. Two scenarios were modelled: vectoring (figure 2) and Point Merge (see figure 3) using a discrete set of alternative predefined arrival routes corresponding closely to the tracks in the real-time simulation, and two controllers (approach followed by final). The models were comparable in granularity: the vectoring model was composed of 27 routes and the Point Merge model was composed of 28 routes. Not all alternative routes were used: for example both models used 16 of the routes in zero disturbance conditions. For the Point Merge model the radial leg distances were 20 NM and 22 NM and the flows through North and South merge points (see figure 3) were vertically separated by 1,000 feet.
TMA performance measurements were made between the initial approach fixes (10,000 and 11,000 feet) and the final approach fix (3,000 feet). The minimum separation in both models was set to 3.5 NM horizontally and 1,000 feet vertically. The model was checked for realism by operational staff involved in the corresponding real-time simulations. It was judged suitable for comparing number of instructions, controller workload, fuel efficiency and distances travelled, but there were reservations about the spacing accuracy.
A series of trials based on perturbing traffic from real-time simulations was run for each of the two TMA models (no wind). Each TMA model was run with over 40 hours of traffic based on an aircraft type mix of 80% medium and 20% heavy.
|
 |
| Figure 2: RAMS Plus model of Vectoring (baseline) |
|
|
|
|
|
 |
| Figure 3: RAMS Plus model of Point Merge with four arrival flows |
|
|
|
|
Figure 4 shows the number of instructions of each type per aircraft for vectoring and Point Merge. The number of lateral instructions was reduced from an average of about 4 heading instructions for the vectoring model to about 1 direct to instruction in the Point Merge. The number of speed instructions increased by about 10% for point merge and the number of level instructions was reduced from an average of about 2.5 per aircraft for vectoring to about 1 for Point Merge, which is consistent with results from real-time simulations.
|
 |
| Figure 4: Number of instructions by type for vectoring and Point Merge |
|
|
|
|
| RAMS Plus has a controller workload model which takes into account coordination, radio communication, flight data management and radar activity. Figure 5 shows approach and final controllers’ workloads were on average about 30% less with Point Merge than with vectoring. |
 |
| Figure 5: Approach (APP) and final (FIN) controller workload for vectoring and Point Merge |
|
|
|
|
Altitude profiles from initial approach fix to runway threshold were used with fuel flow rates from corresponding BADA aircraft models (EUROCONTROL base of aircraft data) to calculate fuel consumed per aircraft. The Point Merge model used on average about 100 kg or 10% less than vectoring (figure 6) over similar distances (average 75NM with standard deviations of 10NM for vectoring and 9NM for Point Merge) (figure 7).
|
 |
| Figure 6: Fuel consumption per aircraft for vectoring and Point Merge |
|
|
|
|
 |
| Figure 7: Distance flown per aircraft in TMA for vectoring and Point Merge |
|
|
|
|
The results of this initial modelling work indicate that the Point Merge technique could have significant performance advantages over conventional vectoring. Airlines could save up to about 100 kg of fuel per aircraft which in terms of pollution in a busy TMA (single runway with over 500 arrivals per day) corresponds to the order of 50 tonnes less fuel burned per day. As well as being able to keep lateral navigation engaged for longer (i.e. no vectoring), the pilots’ task may also be simplified with about 40% fewer tactical manoeuvring instructions. Results indicate that controller workload using Point Merge is no more than with vectoring and could be as much as 30% less.
A small number of TMAs have a pre-defined discrete set of arrival routes for path-stretching during the merging of traffic streams (e.g. trombone shape at Frankfurt). Model based experiments are under development to compare the performance of Point Merge with trombone route based merging. The effects of TMA size, navigation accuracy and traffic uncertainty on the Point Merge technique are also being investigated. This P-RNAV routes based evaluation is part of a TMA performance assessment programme which will look at other concept elements such as adherence to network operations plan target time of arrival and 4D trajectory management in the context of SESAR.
|
|
|
|
| |
|
 |
|
Last validation: 11/03/2008
|
|
|
|
|
|
