The volume of air traffic is increasing year by year. Currently, Air Traffic Control (ATC) ensures safe separation when merging air traffic flows by stretching flight paths using RADAR vectoring. In the future under 4-dimensional trajectory-based operations, ATC will instead instruct an arrival time at a specific waypoint, and aircraft will achieve the arrival time using the Required Time of Arrival (RTA) function of the Flight Management System (FMS). The RTA function controls arrival time by changing airspeed, and with some FMS models it can operate during all flight phases, while others can change only cruise speed. When the arrival time is not achievable by RTA, ATC may provide RADAR vectors or even instruct holding. However, RADAR vectoring is often accompanied by speed instructions that cause an aircraft to operate at above its optimal speed. We therefore suppose that holding is more efficient than vectoring, and that holding procedures with 4-dimensional trajectory will be similar to current holding because holding is a kind of time-based operation. In this study, we compared the fuel consumption of actual flights under RADAR vectoring with simulated fuel optimal flights which instead executed holding. The results show that holding was able can reduce the fuel consumption of all aircraft, but we observed many conflicts at the merging point. We propose different holding altitudes in order to resolve these conflicts, and the proposed modification showed that average fuel consumption can be still reduced with a limited deterioration.
