George Habek (SAS Institute) presented a paper
at the SAS Global Forum last month in Washington DC on the use of SAS
in predictive asset maintenance subtitled, ‘find out why before it’s
too late!’ Habek showed how SAS predictive asset maintenance (PAM) was
used by Kuwait Oil to optimize steam injection in an offshore heavy oil
project.
Steam
injection proved challenging and a significant fall-off in production
occurred a couple of months after operations commenced. Production and
injection data was transferred from KOC’s systems of reference to an
SAS data mart for analysis with PAM. PAM’s data exploration
functionality was used to visualize tags and events and to pinpoint
anomalies. SAS Enterprise Miner identified the main culprits for
the failing production, injection rates that were too high, or too much
chemical additives. The solution is now delivered as a PAM ‘portlet’ to
manage the whole injection program.
© Oil IT Journal - all rights reserved.