FACTUAL INFORMATION The Boeing 737 was being radar vectored to intercept the final approach path for a landing on runway 16R from a right base leg, to follow a Boeing 747 which was already established on the instrument landing system (ILS) final approach. The B737 turned onto the ILS localiser track below the glideslope whilst descending to 2,500 ft some 10.5 NM from the runway threshold. Shortly after the localiser track was intercepted, the aircraft experienced several abrupt changes in bank angle, both left and right, the most severe being a roll to the right through 51 degrees to a maximum right bank of 34.8 degrees. A missed approach was carried out, followed by a normal approach and landing. A post-flight inspection found no defects which may have contributed to the occurrence. The aircraft was subsequently cleared to continue scheduled operations. A review of recorded radar data and of information derived from the flight data recorders of the B737 and the preceding B747. It showed that the B737 was about 450 ft lower than the B747 had been at the same point in space, reaching that point some 127 seconds after the B747 had passed. The longitudinal separation between the B737 and the B747 at that time was 5.5 NM. Recorded wind data, as derived from the inertial reference system of the B737, indicated the wind direction varied between 165 and 185 degrees, at a speed of 8-14 kts. ANALYSIS The circumstances described in this occurrence are very similar to those of an earlier occurrence (9500460). The following features were common to both: Both lead aircraft were B747s which were established on the localiser as well as the glideslope. Both following aircraft were B737s which were given a radar vector to intercept the localiser, below the glideslope, at 2,500 ft. This resulted in both B737s passing the same point in space some two minutes later, but 500 ft lower than the preceding B747s. Atmospheric conditions in the vicinity of the approach path at the time of both occurrences were conducive to the slow decay of wake vortices. As the localiser track is 155 degrees, there would have been little or no lateral displacement of any wake vortices produced by the B747s. Both following aircraft encountered uncommanded rolls consistent with encountering wake turbulence generated by the preceding B747. United Kingdom Civil Aviation Authority wake turbulence studies (August 1994) have shown that B747 aircraft produce high rates of wake turbulence affecting following aircraft. For sequencing purposes during VMC operations in the Sydney terminal area, most domestic aircraft arriving from the south are radar vectored to join a downwind leg when runway 16 is the duty runway. These aircraft are routinely cleared to descend to 2,500 ft whilst being radar vectored to intercept final approach about 6 NM from touchdown. International flights however, must be established on final approach at least 10 NM from the threshold. Many of these aircraft, such as the B747, are in the "heavy" category. This sequencing often results in the following domestic aircraft passing through the same lateral airspace as the preceding aircraft but some 500 ft lower. The relative positions of respective aircraft, the provision of minimum wake turbulence radar separation, and meteorological conditions conducive to the formation and slow decay of wake vortices can make it possible for aircraft to experience wake turbulence encounters whilst such procedures are being implemented. Consideration, therefore, of the vertical positioning of the following aircraft relative to the leader may provide the greatest potential for preventing accidents and incidents as a result of wake turbulence encounters. SIGNIFICANT FACTORS 1. Atmospheric conditions were conducive to the slow decay of wingtip vortices generated by the preceding B747. 2. The B737 was sequenced by ATC to intercept the localiser for runway 16 approximately 500 ft below the preceding B747. SAFETY ACTION As a result of the investigation into this occurrence and a number of other occurrences, the Bureau of Air Safety Investigation issued interim IR 960101 recommendations to the Civil Aviation Safety Authority and Airservices Australia on 7 November 1996. "1. The Bureau of Air Safety Investigation recommends that the Civil Aviation Safety Authority and Airservices Australia: "(i) Evaluate the current wake turbulence separation standards. Consideration should be given to the evaluation of technology being developed to aid in the detection, tracking and forecasting of wake vortices as a further means of reducing the risk of wake turbulence encounters. "(ii) Critically evaluate all current airport arrival and departure paths and procedures to identify and eliminate potential wake turbulence problems. "2. The Bureau of Air Safety Investigation recommends that the Civil Aviation Safety Authority re-institute a wake turbulence education program. This education program should highlight areas of possible wake turbulence encounters and advise ways to minimise the effects of the encounters".