Narrative:

Dfw was operating in a north flow configuration departing runway 35L and runway 36R. Weather was clear and wind was variable from the nnw. I cleared air carrier X for takeoff on an RNAV departure runway 35L to mechl. As I observed track compliance on air carrier X I noticed air carrier Y on a ferra departure off runway 36R to haskl begin to drift east. I visually observed both aircraft out the window. I noted that air carrier X was established on the RNAV track to mechl. I also observed air carrier Y turn or correct to the west. The RNAV departure has created additional work for controllers because the inherent problem with RNAV is with track compliance. When an aircraft does not fly his track for whatever reason then it is filed as a pilot deviation. Because separation is based off the RNAV track and not a heading to be flown; controllers are required to retain control of the departure to ensure track compliance. This has created situations where the controller has to determine at what point is the pilot established on the track. Also; an unintended consequence of retaining the aircraft on frequency is that pilot's are questioning controllers if they should contact departure control or simply just changing frequency without ATC authorization. This has happened numerous times. Another unintended consequence is that by having the tower controller responsible for track compliance it has drawn the controller closer to a radar monitor than a runway monitor. Taking time and eye contact from one task and redirecting it to another task requires that time share between tasks has changed. Longer periods of tending to track compliance means less time scanning runways and movement areas. The number of pilot deviations RNAV has generated can speak to the problem that it is not user friendly in its current format. I doubt the FAA is content with all the pilot deviations nor are the pilots happy about receiving them. Instead of four tracks off the airport (two off each departure runway); the simple solution is a back to basic whereby initial separation can be guaranteed using radar vectors. This places the control of the aircraft back in the hands of the pilot at a critical stage of flight and allows for course correction without affecting subsequent departures i.e.; a radar vector following an RNAV must have three or five miles spacing to avoid a converging of targets in departure airspace. Since all jets flying will be using headings then that simple standard can be used to apply in-trail separation standards. Another benefit is the aspect of go-around and missed approaches into an RNAV environment. Simply stated; the two don't mix very well. By having departure on radar vectors it is only natural that a go-around or missed-approach be assigned a radar vector. The very same reason D10 has imposed the additional spacing between RNAV/vector departures applies to the tower. The behavior of an RNAV/vector mix is the same except for the tower it occurs as soon as a go-around or missed approach takes place. It is my proposal to suspend RNAV departures and rewrite the SID to an initial vector to a common RNAV waypoint. This will help tower establish initial departure separation without the worry of a conflicting route or go-arounds in low weather conditions. RNAV should be configured so that the benefits don't outweigh the costs.

Google
 

Original NASA ASRS Text

Title: A DFW controller voiced concern regarding RNAV departure procedures. Several negative operational aspects of RNAV SIDS in use now are highlighted.

Narrative: DFW was operating in a North flow configuration departing Runway 35L and Runway 36R. Weather was clear and wind was variable from the NNW. I cleared ACR X for takeoff on an RNAV departure Runway 35L to MECHL. As I observed track compliance on ACR X I noticed ACR Y on a FERRA departure off Runway 36R to HASKL begin to drift east. I visually observed both aircraft out the window. I noted that ACR X was established on the RNAV track to MECHL. I also observed ACR Y turn or correct to the West. The RNAV departure has created additional work for controllers because the inherent problem with RNAV is with track compliance. When an aircraft does not fly his track for whatever reason then it is filed as a pilot deviation. Because separation is based off the RNAV track and not a heading to be flown; controllers are required to retain control of the departure to ensure track compliance. This has created situations where the controller has to determine at what point is the pilot established on the track. Also; an unintended consequence of retaining the aircraft on frequency is that pilot's are questioning controllers if they should contact departure control or simply just changing frequency without ATC authorization. This has happened numerous times. Another unintended consequence is that by having the tower controller responsible for track compliance it has drawn the controller closer to a RADAR monitor than a runway monitor. Taking time and eye contact from one task and redirecting it to another task requires that time share between tasks has changed. Longer periods of tending to track compliance means less time scanning runways and movement areas. The number of pilot deviations RNAV has generated can speak to the problem that it is not user friendly in its current format. I doubt the FAA is content with all the pilot deviations nor are the pilots happy about receiving them. Instead of four tracks off the airport (two off each departure runway); the simple solution is a back to basic whereby initial separation can be guaranteed using RADAR vectors. This places the control of the aircraft back in the hands of the pilot at a critical stage of flight and allows for course correction without affecting subsequent departures i.e.; a RADAR vector following an RNAV must have three or five miles spacing to avoid a converging of targets in departure airspace. Since all jets flying will be using headings then that simple standard can be used to apply in-trail separation standards. Another benefit is the aspect of go-around and missed approaches into an RNAV environment. Simply stated; the two don't mix very well. By having departure on RADAR vectors it is only natural that a go-around or missed-approach be assigned a RADAR vector. The very same reason D10 has imposed the additional spacing between RNAV/Vector departures applies to the tower. The behavior of an RNAV/Vector mix is the same except for the tower it occurs as soon as a go-around or missed approach takes place. It is my proposal to suspend RNAV departures and rewrite the SID to an initial vector to a common RNAV waypoint. This will help tower establish initial departure separation without the worry of a conflicting route or go-arounds in low weather conditions. RNAV should be configured so that the benefits don't outweigh the costs.

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.