Both the introduction of service observing, and the use of the NOT for
educational support have of course an effect of the amount of support
required. This can be easily been seen for the service observing.
Given an average of 
3 nights per observing run (see Table 3) we
normally would have had 7 regular support nights, compared to 22 full
working nights.
To make a rough estimate of the additional effort spend, I have made the following calculation:
Support night: In this case, the support astronomer will spend the night on the mountain, where s/he typically is at the telescope until midnight and is on-call the rest of the night. In principle, a day including a night on the mountain is counted as 2 working days to account for the extended work hours and the need to stay over-night on the mountain and be on-call. On the other hand, some time during these days are left for other tasks and I estimate the equivalent effort for 1 support night to 1.5 working days.
Technical or Service night: In this case, the support astronomer will spend the whole night at the telescope and the few hours in the day time are needed to prepare for the night. Also here staying a day including a night on the mountain is counted as 2 working days, which in this case are fully spend on astronomical support.
In the last semester there were a total of 51 support nights and 28 technical nights, which correspond to 132.5 working days. In this semester there were a total 48 support nights, 23.5 technical nights and 22 service nights, corresponding to 163 working days. Apart from the support nights for the instruments and telescope for the two course which are included under the normal support nights, there were an additional 4 nights of full support by the staff member, plus the effort by the students which I estimate at 1 full night support. Taking these numbers, I get a total amount of support for this semester corresponding to 173 working days, i.e. an increase of just over 30% (the biggest part of 23% is for the service observing alone).
Note that for the service observing I did not include the additional effort needed to obtain and process the information from the PIs of the proposals, plus collecting and distributing the data and details of the observations to the PIs, which would not be needed for a regular observing or technical night. In this semester this also included a significant amount of development work to set-up the system to define and schedule the observations (see above). Furthermore, the effort spend in advance preparing for the summer school was also not included.
As this rather conservative estimate show a very significant increase in effort (also on the technical staff!), one should realize that providing these services does have its implications for the available effort for other (development) tasks.