Can you please clarify how interval data should be reported?

In the SR test scripts, there are up to three different values reported for
interval usage:

1.  The value reported in the QTY02 of the PTD*BO loop
2.  The value calculated from data provided in the PTD*BO loop [Ending Read
(MEA06) - Beginning Read (MEA05) * Meter Multiplier (MEA03)]
3.  Sum of intervals in PTD*PM loop

Most of these differences are due to rounding issues.  Meaning that the
number of decimal places is inconsistently applied.  Some of these rounding
issues can be significant on large usage values.  What is the standard for
these variances or the percent of tolerance?

My expectation is a clean tie (validation) between the three values listed
above.  And, for the validated usage value from the 867 to tie out to the
usage reported on the 810.  Is this an unrealistic expectation?

Thanks a bunch,

Allyson Hafner
EDI Manager
CSC Energy Services
614-737-1160