Narrative:

During initial climb; aircraft overshot assigned altitude by about 350 ft. Departure control assigned us direct to the VOR. First officer (pilot flying) was hand flying; so I input the fix in the CDU; confirmed with my first officer and executed; then selected LNAV on the AFDS. (At this point; we had just completed flap cleanup.) we were climbing at about 4;000 FPM at climb N1 power setting. Cold weather; light load (10 passengers) contributed to high climb rate. As I input the fix and executed; I had an eye on the climb rate and my first officer's actions; anticipating a transition to level flight at 4;000 ft. At about 3;500 ft MSL; my first officer selected autopilot B command mode; the light illuminated; and I saw him look away (climb power remained set). I noted that the aircraft had not begun to level and realized quickly that the autopilot would not capture the assigned altitude. I made a brief comment about the impending overshoot and took control of the aircraft to begin the leveloff just prior to 4;000 ft. We topped out at about 4;350 ft. I called out the overshoot deviation to the departure controller and reported returning to 4;000 ft. We were then assigned a vector of 360 degrees. Shortly after; we were reassigned direct to the VOR. Once on course; I noted another aircraft at 5;000 ft on the TCAS display about 8 miles ahead. No TCAS alert or warning was issued during the overshoot event. I asked the controller if there had been a conflict and he said 'no.' on being handed off to center; I again asked if there were 'any issues' and was again told; 'no.' the rest of the flight was uneventful. I've seen this type of problem before. The common thread is the assumption that the autopilot will level the aircraft at the selected altitude; regardless of when it is engaged; or the climb rate or power setting. The biggest 'fix' is to reduce power and climb rate early under such conditions (cold; light load; high power setting). In other words; anticipate. My first officer and I reviewed our actions once in level flight. He was surprised that the airplane didn't level as he thought it would. I explained why it didn't; and the power and climb rate reduction fix. We were busy getting the airplane cleaned up; turning to the tower-assigned heading after takeoff; and then transitioning to the LNAV fix. We simply misplaced our priorities; namely; fly the airplane first; then navigate. I chalk this one up to inexperience with this type of problem on my first officer's part and an assumption that the autopilot will handle any leveloff task; regardless of climb rate or when the autopilot is engaged. Our discussion centered on what caused the problem and the remedy. For my part as pilot monitoring; I saw the high climb rate and anticipated the problem; but was late intervening with either meaningful comment or action. Things occurred pretty quickly and the dark cockpit made it a bit tough to see the first officer to gauge his actions. A possible action we could take is to run this in the simulator during initial training as a first officer loft scenario; with the instructor pointing out the problem with high power settings; high climb rates; and late autopilot activation; with emphasis on the 'pendulum effect' created by the engines being located under the wings affecting leveloffs while at climb power.

Related reports:

Google
 

Original NASA ASRS Text

Title: B737 Captain describes First Officers attempt to engage the autopilot with 500 FT to go in climb at 4000 FT per minute vertical speed; an Overshoot occured.

Narrative: During initial climb; aircraft overshot assigned altitude by about 350 FT. Departure Control assigned us direct to the VOR. First Officer (Pilot Flying) was hand flying; so I input the fix in the CDU; confirmed with my First Officer and executed; then selected LNAV on the AFDS. (At this point; we had just completed flap cleanup.) We were climbing at about 4;000 FPM at climb N1 power setting. Cold weather; light load (10 passengers) contributed to high climb rate. As I input the fix and executed; I had an eye on the climb rate and my First Officer's actions; anticipating a transition to level flight at 4;000 FT. At about 3;500 FT MSL; my First Officer selected Autopilot B command mode; the light illuminated; and I saw him look away (climb power remained set). I noted that the aircraft had not begun to level and realized quickly that the autopilot would not capture the assigned altitude. I made a brief comment about the impending overshoot and took control of the aircraft to begin the leveloff just prior to 4;000 FT. We topped out at about 4;350 FT. I called out the overshoot deviation to the Departure Controller and reported returning to 4;000 FT. We were then assigned a vector of 360 degrees. Shortly after; we were reassigned direct to the VOR. Once on course; I noted another aircraft at 5;000 FT on the TCAS display about 8 miles ahead. No TCAS alert or warning was issued during the overshoot event. I asked the Controller if there had been a conflict and he said 'no.' On being handed off to Center; I again asked if there were 'any issues' and was again told; 'no.' The rest of the flight was uneventful. I've seen this type of problem before. The common thread is the assumption that the autopilot will level the aircraft at the selected altitude; regardless of when it is engaged; or the climb rate or power setting. The biggest 'fix' is to reduce power and climb rate early under such conditions (cold; light load; high power setting). In other words; anticipate. My First Officer and I reviewed our actions once in level flight. He was surprised that the airplane didn't level as he thought it would. I explained why it didn't; and the power and climb rate reduction fix. We were busy getting the airplane cleaned up; turning to the Tower-assigned heading after takeoff; and then transitioning to the LNAV fix. We simply misplaced our priorities; namely; fly the airplane first; then navigate. I chalk this one up to inexperience with this type of problem on my First Officer's part and an assumption that the autopilot will handle any leveloff task; regardless of climb rate or when the autopilot is engaged. Our discussion centered on what caused the problem and the remedy. For my part as Pilot Monitoring; I saw the high climb rate and anticipated the problem; but was late intervening with either meaningful comment or action. Things occurred pretty quickly and the dark cockpit made it a bit tough to see the First Officer to gauge his actions. A possible action we could take is to run this in the simulator during initial training as a First Officer LOFT scenario; with the instructor pointing out the problem with high power settings; high climb rates; and late autopilot activation; with emphasis on the 'pendulum effect' created by the engines being located under the wings affecting leveloffs while at climb power.

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.