So, does ID identify a specific record or is it an identifier for a specific
group of records?

Do you have more than one period of time (more than  one day) in your table?

Does your time field only have the time in it or does it also include the
date?

Perhaps this will get you started.  It assumes only times in the field and
only one days worth of data

SELECT A.YourTimeField, Sum(B.TotalField) as RollingTotal
FROM A.YourTable = B.YourTable
WHERE A.YourTimeField >= B.YourTimeField
AND A.YourTimeField <= DateAdd("n",65,B.YourTimeField)
GROUP BY A.YourTimeField

SELECT Top 1 Q.YourTimeField, RollingTotal
FROM TheAboveQuery as Q
ORDER BY RollingTotal Desc

If you care to post the actual field and table names, perhaps that can all
be rolled up into one query.  Also answer the questions about your data.

When you say highest total hour, can that cross hh:00? (i.e., 12:30 am,
12:45 am, 1:00 am, 1:15 am, making 12x4 total groupings) or do you just want
to evaluate all hh:00, hh:15, hh:30, hh:45 groupings as 12 separate
entities?

If the former, you'd need a fair amount of code. I might create an array to
hold all 48 values and then a loop to total each group of 4, comparing the
Current4Total to the previous High4Total and storing High4Time in a variable
each time a new High4Total is identified.  Probably an easier way, that's
just off the top of my head.

If the latter, you can simply create a totals query, sorted Descending, that
only returns the Top record:

Something like:

SELECT TOP 1 Hour([MyDateTime]) AS FullHour, Sum([MyTotal]) AS TotalValue
FROM MyTable
GROUP BY Hour([MyDateTime])
ORDER BY Sum([MyTotal]) DESC;