• [y] I have checked that this issue has not already been reported.

• [y] I have confirmed this bug exists on the latest version of pandas.

• (optional) I have confirmed this bug exists on the master branch of pandas.

Note: Please read this guide detailing how to provide the necessary information for us to reproduce your bug.

Code Sample, a copy-pastable example

this is a tiny example

tmp = pd.DataFrame({'id': ['1624477460271-3908654213', '1624477460271-3908654213'], 'dt':[np.nan, pd.to_timedelta('0:0:2')]})
tmp
Out[6]: 
                         id              dt
0  1624477460271-3908654213             NaT
1  1624477460271-3908654213 0 days 00:00:02

Problem description

dataframe.groupby('some_column').timedelta.sum() results wrong when timedelta contains NaT

tmp.groupby('id').dt.sum()
Out[7]: 
id
1624477460271-3908654213   -106752 days +00:12:45.145224192
Name: dt, dtype: timedelta64[ns]

Expected Output

tmp.dt.sum()
Out[8]: Timedelta('0 days 00:00:02')

Output of pd.show_versions()