Narrative:

First officer attempted to start APU while I was taxiing in to gate. Starter hung at 25% RPM. Waited over 1 minute for second start. Second start also hung at 25%. Waited 2-3 minutes for third start attempt; at which time APU started successfully. While deplaning passengers; noticed fuel smell in cockpit but not cabin. I went outside aircraft to check and noticed no smell or leaks and nothing else out of the ordinary. Called maintenance out and he also noticed fuel smell in cockpit; shut down the APU and went back to check. Maintenance man noticed fire in APU exhaust and grabbed a fire extinguisher and put out the fire. I called d-tower to inform them and they called fire crews out. We exceeded the starting limitation on the third start attempt. I am not sure that was the cause of the fire. I knew that it would take a long time to get a gpu to the gate and I was reluctant to deplane the passengers without the APU running; because my options were to deplane with #2 engine running; which I absolutely did not want to do on the crowded ramp; or deplane on emergency lights with battery power only on the aircraft. These factors plus the task of taxiing the aircraft while my first officer was trying to start the APU created a high workload environment. I should have waited out on the ramp while we got the APU started before I taxied in to the gate.

Google
 

Original NASA ASRS Text

Title: A CRJ200 flight crew reported experiencing an APU fire on taxi in when they made a third start attempt contrary to limitations.

Narrative: First Officer attempted to start APU while I was taxiing in to gate. Starter hung at 25% RPM. Waited over 1 minute for second start. Second start also hung at 25%. Waited 2-3 minutes for third start attempt; at which time APU started successfully. While deplaning passengers; noticed fuel smell in cockpit but not cabin. I went outside aircraft to check and noticed no smell or leaks and nothing else out of the ordinary. Called maintenance out and he also noticed fuel smell in cockpit; shut down the APU and went back to check. Maintenance man noticed fire in APU exhaust and grabbed a fire extinguisher and put out the fire. I called D-Tower to inform them and they called fire crews out. We exceeded the starting limitation on the third start attempt. I am not sure that was the cause of the fire. I knew that it would take a long time to get a GPU to the gate and I was reluctant to deplane the passengers without the APU running; because my options were to deplane with #2 engine running; which I absolutely did not want to do on the crowded ramp; or deplane on emergency lights with battery power only on the aircraft. These factors plus the task of taxiing the aircraft while my First Officer was trying to start the APU created a high workload environment. I should have waited out on the ramp while we got the APU started before I taxied in to the gate.

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.