This counts Severely Errored Seconds -- one-second intervals on this OTU signal where the error density crossed a much higher threshold than a plain Errored Second. Because this reading comes from a closed, completed 24-hour history bin for this OTU signal, SES is weighted far more heavily than ES in most availability calculations, so even a handful of SES on this OTU signal can matter more to an uptime report than a large number of ordinary ES. For example, a bin showing 45 ES but only 1 SES on this OTU indicates mostly minor, scattered errors, while the same 45 ES paired with 20 SES would indicate the errors clustered into a few severe bursts, across that completed 24-hour day.