Narrative:

Initial departure push; noise abatement procedures in effect; sfo was departing runways 1R and 1L; landing runway 28R; sfo ceiling was 007 ovc and oak ceiling was 003 ovc. The ci-1 initiated headings to sfo tower to eliminate/reduce departure delays for both airports. Sfo had called off a SH36 which was assigned departure instructions of heading 030 vector to fix. Ci-1 approved two fifty degree headings to sfo; both air carriers. Oak tower was released because the fifty headings off of sfo go behind oak airport. Fifty degree headings are turned to 90 or 100 degree headings after radar contact depending on sfo final traffic for noise abatement. Once the aircraft is out of 7;000 ft the pilot is cleared direct fix or assigned another vector as appropriate for sequencing to fix. Sfo local controller cleared the SH36 first. Due to the slow climb performance of this aircraft; it is common to have to assign a heading of 040 to 050 to go behind the oakland departures. Per nct/sfo tower LOA; norcal has control for IFR departure aircraft at 1;600 ft. The SH36 was approximately 2 miles upwind and I was about to assign the pilot a heading to 040 when I observed a B737 (out of sfo) get airborne. Based on my radar presentation; the first B737 appeared directly behind the SH36. The B737 was initially overtaking by 60 KTS and closing quickly. Tower had not shipped the B737 to me; so I turned the SH36 to a heading of 010 to create more separation. I could not go further left because oakland tower had traffic released. During noise abatement; oakland departure traffic is assigned heading 270 and a climb to 10;000 or 15;000 depending on outbound route. I finally got communications with the B737 and immediately turned the aircraft to a heading of 090. At this same time; oakland tower had another B737 airborne climbing on a 270 heading. The ci-1 was on the line with oakland tower; but it was not clear at that time if the second B737 would receive an altitude amendment. To avoid a head-on situation with the oak B737; I turned the SH36 to a heading of 090 to achieve lateral spacing with the oak B737 and avoid a head-on loss of separation. I suspected I might lose lateral separation with all three aircraft; but I knew that they would not converge because of the headings assigned. I decided a lateral loss of separation was better than head-on convergence with the oakland departure. The proximity event occurred between the sfo B737 and the SF36; lateral separation had been lost as a result of the turn I had to issue to protect the oakland departure. I do not have the specific altitudes; but I have been informed the separation loss was 2.74 miles lateral and 800 ft vertical and was determined to be a proximity event. I was informed by the operations manager; there was only 5 degrees of divergence between the sfo B737 and the SH36 on initial departure. Based on this; it is no wonder the sfo B737 appeared to be directly behind the SH36. The initial loss of separation between the sfo B737 and the SH36 has been previously reported. I did not expect the tower controller to apply radar departure divergence rules since it was not appropriate. Recommendation; sfo tower is responsible for initial departure separation. As it was explained to me; the tower controller claimed visual separation was applied in addition to radar departure divergence rules. I'm not sure how visual separation was applied due to the low ceilings. It appears radar departure divergence rules are not understood and were executed incorrectly. What the tower controller did was assign successive departure headings that crossed over each other. Specifically the SH36 was cleared off runway heading of 10 degrees and turned to a heading of 30 degrees. The sfo B737 was cleared off runway heading of 10 degrees and turned to a heading of 50 degrees inside the SH36. It is impossible to have had the required 15 degrees or more of course divergence within 1 mile of the runway. The play back showed 5 degrees. Aircraft performance characteristics were not understood and/or considered. It also appears that the tower controller did not consider nct has control at 1;600 ft. Briefings should be given on correct execution of this procedure. In addition; another possibility is to have sfo depart runway 28 during noise abatement when arrival demand is nil. This would allow both oakland and sfo to depart without having to hold for release. The other option is to cease headings and do basic noise abatement to avoid these types of unsafe situations. This would increase departure delays; but better to arrive late and alive than not at all. The situation above sets up a potential mid-air collision.

Google
 

Original NASA ASRS Text

Title: NCT Controller described a conflict situation caused by SFO Tower's reported inappropriate utilization of successive departure separation criteria.

Narrative: Initial departure push; noise abatement procedures in effect; SFO was departing Runways 1R and 1L; landing Runway 28R; SFO ceiling was 007 OVC and OAK ceiling was 003 OVC. The CI-1 initiated headings to SFO Tower to eliminate/reduce departure delays for both airports. SFO had called off a SH36 which was assigned departure instructions of heading 030 vector to fix. CI-1 approved two fifty degree headings to SFO; both air carriers. OAK Tower was released because the fifty headings off of SFO go behind OAK Airport. Fifty degree headings are turned to 90 or 100 degree headings after RADAR contact depending on SFO final traffic for noise abatement. Once the aircraft is out of 7;000 FT the pilot is cleared direct fix or assigned another vector as appropriate for sequencing to fix. SFO Local Controller cleared the SH36 first. Due to the slow climb performance of this aircraft; it is common to have to assign a heading of 040 to 050 to go behind the Oakland departures. Per NCT/SFO Tower LOA; NORCAL has control for IFR departure aircraft at 1;600 FT. The SH36 was approximately 2 miles upwind and I was about to assign the pilot a heading to 040 when I observed a B737 (out of SFO) get airborne. Based on my RADAR presentation; the first B737 appeared directly behind the SH36. The B737 was initially overtaking by 60 KTS and closing quickly. Tower had not shipped the B737 to me; so I turned the SH36 to a heading of 010 to create more separation. I could not go further left because Oakland Tower had traffic released. During noise abatement; Oakland departure traffic is assigned heading 270 and a climb to 10;000 or 15;000 depending on outbound route. I finally got communications with the B737 and immediately turned the aircraft to a heading of 090. At this same time; Oakland Tower had another B737 airborne climbing on a 270 heading. The CI-1 was on the line with Oakland Tower; but it was not clear at that time if the second B737 would receive an altitude amendment. To avoid a head-on situation with the OAK B737; I turned the SH36 to a heading of 090 to achieve lateral spacing with the OAK B737 and avoid a head-on loss of separation. I suspected I might lose lateral separation with all three aircraft; but I knew that they would not converge because of the headings assigned. I decided a lateral loss of separation was better than head-on convergence with the Oakland departure. The proximity event occurred between the SFO B737 and the SF36; lateral separation had been lost as a result of the turn I had to issue to protect the Oakland departure. I do not have the specific altitudes; but I have been informed the separation loss was 2.74 miles lateral and 800 FT vertical and was determined to be a proximity event. I was informed by the Operations Manager; there was only 5 degrees of divergence between the SFO B737 and the SH36 on initial departure. Based on this; it is no wonder the SFO B737 appeared to be directly behind the SH36. The initial loss of separation between the SFO B737 and the SH36 has been previously reported. I did not expect the Tower Controller to apply RADAR departure divergence rules since it was not appropriate. Recommendation; SFO Tower is responsible for initial departure separation. As it was explained to me; the Tower Controller claimed visual separation was applied in addition to RADAR departure divergence rules. I'm not sure how visual separation was applied due to the low ceilings. It appears RADAR departure divergence rules are not understood and were executed incorrectly. What the Tower Controller did was assign successive departure headings that crossed over each other. Specifically the SH36 was cleared off runway heading of 10 degrees and turned to a heading of 30 degrees. The SFO B737 was cleared off runway heading of 10 degrees and turned to a heading of 50 degrees inside the SH36. It is impossible to have had the required 15 degrees or more of course divergence within 1 mile of the runway. The play back showed 5 degrees. Aircraft performance characteristics were not understood and/or considered. It also appears that the Tower Controller did not consider NCT has control at 1;600 FT. Briefings should be given on correct execution of this procedure. In addition; another possibility is to have SFO depart Runway 28 during noise abatement when arrival demand is nil. This would allow both Oakland and SFO to depart without having to hold for release. The other option is to cease headings and do basic noise abatement to avoid these types of unsafe situations. This would increase departure delays; but better to arrive late and alive than not at all. The situation above sets up a potential mid-air collision.

Data retrieved from NASA's ASRS site as of April 2012 and automatically converted to unabbreviated mixed upper/lowercase text. This report is for informational purposes with no guarantee of accuracy. See NASA's ASRS site for official report.