SQL Server 2017

By using db<>fiddle, you agree to license everything you submit by Creative Commons CC0.

create table t (groupid int, FromDate datetime, ToDate datetime);
insert into t values
(1,'2019-09-30 11:13:00','2019-09-30 11:13:50'), 
(1,'2019-09-30 11:13:20','2019-09-30 11:14:10'), 
(2,'2019-09-30 11:20:00','2019-09-30 11:20:20'), 
(1,'2019-09-30 11:20:10','2019-09-30 11:20:20'), 
(3,'2019-09-30 11:25:00','2019-09-30 11:25:30');
5 rows affected

select groupid, min(fromdate), max(todate)
from (select t.*,
             sum(case when todate > prev_fromdate then 0 else 1 end) over
                 (partition by groupid order by fromdate) as grp
      from (select t.*,
                   max(todate) over (partition by groupid
                                     order by fromdate
                                     rows between unbounded preceding and 1 preceding
                                    ) as prev_fromdate
            from t
           ) t
     ) t
group by groupid, grp;


groupid
(No column name)
(No column name)


1
2019-09-30 11:13:00.000
2019-09-30 11:20:20.000

2
2019-09-30 11:20:00.000
2019-09-30 11:20:20.000

3
2019-09-30 11:25:00.000
2019-09-30 11:25:30.000


select groupid, sum(datediff(second, fromdate, todate))
from (select groupid, min(fromdate) as fromdate, max(todate) as todate
      from (select t.*,
                   sum(case when todate > prev_fromdate then 0 else 1 end) over
                       (partition by groupid order by fromdate) as grp
            from (select t.*,
                         max(todate) over (partition by groupid
                                           order by fromdate
                                           rows between unbounded preceding and 1 preceding
                                          ) as prev_fromdate
                  from t
                 ) t
           ) t
      group by groupid, grp
     ) t
group by groupid


groupid
(No column name)


1
440

2
20

3
30

groupid	(No column name)	(No column name)
1	2019-09-30 11:13:00.000	2019-09-30 11:20:20.000
2	2019-09-30 11:20:00.000	2019-09-30 11:20:20.000
3	2019-09-30 11:25:00.000	2019-09-30 11:25:30.000