This counts Severely Errored Seconds accumulated by the path since the start of the current day -- seconds where the error rate was bad enough (or a defect like LOP occurred) to cross the severe threshold, not just any single bit error. SES is a stronger signal than plain ES, so an admin watching this field over the course of a day is really tracking how many seconds were bad enough to likely have caused a real service hit, such as a dropped voice call or retransmitted data. With the default CurrentDay SES threshold on this path set to 7, a reading of 9 by mid-afternoon would already indicate an alarm-worthy day even before the 24-hour bucket closes.