Do not use min() to truncate time_to_death

[yiwen-h]

At the moment when calculating time to death throughout the model we truncate to the end of the sim_duration using the min function if the time_to_death exceeds the sim duration. This biases values by chopping of the end of the interval.

There were concerns that when time_to_death is allowed to be longer than the sim_duration, this can result in very large output Excel files with way more columns than needed (in the output Excel files we only care about results for the sim duration).

We could address this by:

1. remove use of min() function to truncate time to event
2. Ensure that outputs CSVs are limited to the number of columns for sim_duration only

See discussion here