ABSTRACT:
A profitable waterflood depends on the proper operation of
individual wells, maintaining the balance between water injection
and oil production over the entire field or project, and preventing
well failures. To achieve these optimum recovery conditions,
Tadeusz Patzek, Dmitriy Silin, and Asoke De have designed
a unique software-integrated waterflood surveillance system
that automatically collects, processes, stores, and analyzes
input from new and/or conventional instrumentation to provide
intelligent control of field operations.
The system offers multiple levels of complexity depending
on available resources. Injection pressures and rates are
minimum inputs, with water production rate and salinity, and
satellite images of surface displacement on a field scale
adding increasing sophistication to the model. A network of
model-based injector controllers predicts the optimum water
injection rate that can be achieved without causing excessive
formation damage. The controller can be used either as an
off-line smart advisor or in an on-line fully automated system.
Existing controllers are not model-based and injection rates
are assigned more or less arbitrarily. They must also be reset
by hand because they become unstable with time. In some projects
controllers arent used at all and water injection is
a hit-or-miss operation. In these cases each injector is set
manually, a laborious and dangerous procedure. The new system
makes better use of field engineers time by freeing
them to focus on strategic issues rather than routine tasks.
|