Page 227 - Intelligent Digital Oil And Gas Fields

Workflow Automation and Intelligent Control 179

• Filtering and conditioning algorithms: It is referred to as doing data prepara-
tion using algorithms. Analytical models, such as IPR and Vogel equa-
tions, need a single, representative, and average model of the day/week/
month or the time period of evaluation. For this, the cleansing algo-
rithms are used to clean data up from data spikes, frozen or missing data,
system errors, typos, etc.
• Well performance model (steady-state condition): Examples include the diag-
nostic model discussed in this section. Any spreadsheet, Web-based, and
stand-alone application can be used to compute the flowing condition of
a well and estimate the IPR and VLP. This calculation can be executed
into the workflow, daily or weekly. The IPR real-time calculation does
not have a logical basis because fluctuations in both pressure and flow rate
(due to changes in flow regimes, liquid loading, and solid obstructions)
can generate misleading results in the IPR calculation.
• Model-based analytics (data driven): The prediction section can be built
using an ANN and fuzzy logic techniques to predict 1 to about 30days
of production performance. Alternatively, type curve, DCA, and other
curve-fitting methods can help predict production performance with
acceptable error and, more importantly, generate immediate actions to
prevent high water cut and diagnose underperforming wells.
• Tracking actions and well events using expert systems, pattern recogni-
tion, and predictive advisory tools.

5.4.3 UI Dashboard and Layout
The dashboard and layout should include important information affecting
the current operations, and it should be concise and well organized so that
managers, supervisor, engineers, and technician can easily read and decipher
the displayed information. First, display a geographical information system
(GIS) with latitude and longitude of well surface locations. This data can be
combined with daily production of both water and oil, and displayed as a pie
chart. In the background, display the total cumulative oil or gas to mimic the
drainage area extrapolated at reservoir conditions with active links to the
well and field data, along with metrics on well events including downtime.
The dashboard also has queries to alarm and alert management by exception.
Fig. 5.17 shows a typical dashboard showing a GIS map with production
data and infographic plots. In the middle of the screen, gas, oil, and water
production plots versus time are shown with their respective forecasting.