What is the sediment composition in the upstream end of the domain and what boundary condition for the upstream composition are you imposing? If the initial composition is unisize (it seems based on your explanation) and the boundary condition for composition is fixed, you are not allowing the necessary changes in composition for the sediment transport per size fraction to enter. 

In short, check that in the mor-file.

Also, check that you are correctly setting the boundary condition of the bed level upstream to sediment load. That is, must be 4 or 5. 

I am using ICmpCond=0 and IBedCond=4 for my model. It is running smoothly but again I see no change in the bedload transport. Am I taking wrong reference to see the impact of the sediment feed in the channel? Does the bedload transport remain constant irrespective of the additional sediment for the same discharge? I assume transport remain the same but the bedlevel might vary due to storage for additional sediment. I am a bit confused if it is with the input file or the output interpretation. What could be the best output variable to observe such changes? Thank you  

The question is then, where are you observing the cumulative load? You could add a cross-section right downstream of the upstream boundary to check the sediment that enters the model. In any case, by comparing the map results, you should find differences between runs. 

Thanks a lot. I assigned the IBedCond = 4 at upstream boundary and 1 at downstream boundary. And then I put ICmpCond=0. I can see the sediment feed now in the model. Is it a good idea to fix the bed at the downstream boundary? And how will the model be affected for ICmpCond=0 or 1? I only found a short description in the table of the manual. Would be helpful to be clear on this. Thanks once again. 

You cannot fix the bed at an outflowing boundary (in a river case, the downstream end). The bed level will be updated. Only if flow would reverse, the boundary condition would be applied. 

As for the bed-composition boundary condition, if you specify the load per size fraction it has no meaning to fix the upstream composition, you are already providing all necessary BC's. 

I made the changes as mentioned above. I assigned the IBedCond = 4 at upstream boundary and 1 at downstream boundary. And then I put ICmpCond=0.  I am using 2 fractions of sediment with 0.5m thickness each. I ran the model without sediment feed first. I extracted the maximum sediment transport at each timestep among all cross-section and multiplied this by 10 to feed into the new model. When I check the results, I found that there was no sediment feed into the channel after first timestep. The cumulative transport curve showed a big rise at the start and no sediment change in the transport. I checked the .dia file. It is full of Bed change exceeds 5% of waterdepth after ...timesteps and Erosion Shortage at NM. Does this indicate that there is no feed of sediment from the upstream boundary (which I assigned in the .mor and .bcm files) or the values for the parameters need some changes?

Maybe a sudden increase of the upstream sediment load by 10 times its original value is causing a shock wave that would need a very small time step to be resolved. Sediment may be piling up at the upstream end. Rather than decreasing the time step, I would try gradually increasing the load, if that is acceptable for your research.