If you want a 95% probability of starting rounds on time, assume a 60-move game. If you want 99%, assume 80 moves.
Also, allow an extra half hour between rounds to make pairings, get the players settled, compensate for late-starting games in previous rounds, etc.
For example, with game/90, inc/30, a 60-move game could take 2 hours per player, so allow 4 and a half hours between rounds, such as 10:00 and 2:30.
Forget about multiple controls, such as 40/60, inc/30, then SD/30, inc/30. The whole idea of increment is to get rid of “artificial time pressure” at move 40 (or any other intermediate point). Also, some clocks may not support increment with multiple controls.
Forget about the w.c. break. The player on the move would have an unfair advantage, as he could use the extra 10 minutes to ponder his move. To get around this, you’d have to go the whole nine yards with adjournments, sealed move envelopes, etc, which I really don’t think you’d want to deal with these days.
One thing you MUST consider is what to do in games played with clocks which are not increment-capable. One alternative would be to require each player to furnish an increment-capable clock, but this could reduce your turnout. Another alternative would be an alternative time control. For example, game/90, inc/30 could be played at game/120 in games played with older equipment. (Obviously, increment-capable equipment would be preferable, i.e. either player who furnishes such equipment could insist on using it.)
Increment is feasible for one or two rounds per day. Three is stretching it. With four or more, forget increment – revert to the 5-second delay.
Bill Smythe