convergence problem
 

I am doing simulation on a laminar flow in a capillary tube. However, i have problems on convergence. At first instance, the mass continuity residual increase and i lower the under-relaxation factor as suggested by the Fluent Guide book. It seems alright buy after about 200 iterations, the residual starts to increase again, from 2.5e-2 to 3.2e-2. May i know what's the possible problem again? Do i need to stop and again lower the under relaxation factor? or any othe rpossible reasons?