Page 188 - Intelligent Digital Oil And Gas Fields

144 Intelligent Digital Oil and Gas Fields

the capability to develop, monitor, and optimize drilling KPIs—such as
cycle time, NPT, and ROP—to lower overall costs and to identify key con-
tributors to rig performance (rig, personnel/process, wired drill pipe, equip-
ment, vendors, etc.). They further emphasized that the additional high value
of real-time data in a drilling knowledge base enable current drilling param-
eters to be displayed next to offset values, along with any other data in a sin-
gle collaborative workspace, which is updated automatically and in real time
and requires minimal manipulation by drilling engineers.
e
In another paper, Brul (2013) has introduced a new paradigm for
analyzing massive amounts of data, (semi)structured and unstructured, at
ultrahigh speeds and frequencies, for Big Data analytics and continuous
model updating in E&P. This new paradigm is based on a real-time adaptive
analytics and data-flow architecture, which combines stream computing
(Fig. 4.20), Hadoop/NoSQL, and Map/Reduce, and massive parallel
processing data warehouses (MPP DW). As indicated in Fig. 4.20, stream
computing applications are (or can be) represented as data-flow network
graphs (Leskovec et al., 2014) composed of operators, interconnected by
streams. The external data feeds can, for example, represent high-resolution
imagery, IoT sensor readings, stream of headline news, or market informa-
tion, such as securities and commodities. The operators implement algo-
rithms for data analysis, such as parsing, filtering, imputation, feature
extraction, and classification.
The E&P version of the IoT represents a complex network of sensors and
control and automation systems in oil and gas field operations. For example,
drilling surveillance, analysis, and optimization vibrational and acoustic data

Operator
High-resolution Stream
imagery

IoT sensor
readings

Headline news

Market
information
Fig. 4.20 Stream computing applies to high-volume and high-velocity data, whether
structured, semi-structured or unstructured. (Modified from Brul e, M.R., 2013. Big Data
in E&P: Real-Time Adaptive Analytics and Data-Flow Architecture. SPE-163721–MS,
https://doi.org/10.2118/163721-MS.)