American Oil and Gas Reporter - November 2015 - 44

SpecialReport: Oil & Gas Computing

Analytics Lessen Uncertainty And Risk
By Keith Holdaway
and Moray Laing

CARY, N.C.-By applying analytics,
oil companies forced to explore in geologically complex and remote areas in
their efforts to exploit unconventional hydrocarbon deposits are optimizing a wide
range of upstream data to empower effective, efficient and timely decisions that
affect the intrinsic value of the exploration,
drilling and production value chain.
In areas with intrinsically poor data
quality or challenging geologic features,
the cost associated with seismic imaging
and drilling rise, as do operational and
economic risks. To counter these issues,
companies need to integrate disciplines
and data to better access and quantify risk.
Soft computing methods offer an excellent opportunity to integrate information
from various sources with varying degrees
of uncertainty. These methods also help
to establish relationships between measurements and reservoir properties in a
multidimensional, multivariate and stochastic environment, and to assign risk
factors to predictions.
For the upstream domain, analytics
can address questions related to risk and
uncertainty, and can strengthen confidence
in reserves information while quantifying
the impact of exploitation plans on attaining predefined targets. Analytics with
descriptive and predictive modeling, forecasting, and advanced data-driven models
will help companies better understand
their facilities, wells, and reservoir systems.
Advanced visualization tools encapsulate exploration and production data
across disparate geoscientific disciplines.
Advanced analytics, underpinned by highperformance computing, positions upstream companies to act quickly so they
can capitalize on opportunities in the
face of increasing data volumes.
Seven ways in which advanced analytics helps companies gain value from
data are:
* Manage data efficiently;
* Adopt and use sample, explore,
modify, model and assess (SEMMA) sequential techniques to mine and analyze
data;
* Employ visual analytics to make
better sense of multivariate, multidimensional, and multivariant parameters;
44 THE AMERICAN OIL & GAS REPORTER

* Use predictive and prescriptive analytics to save time and money;
* Estimate reserves and detect hidden
patterns that hint at declining wells using
probabilistic forecasting techniques;
* Employ deep learning methods to
derive value from abstract data; and
* Layer in high-performance analytics
to gain insight quickly.

SEMMA Approach To E&P
Oil and gas operators spend 70-80
percent of their time managing and organizing an explosion of historical and
real-time data. Ensuring the integrity of
this big data has become a pressing issue
for the industry.
Successful and credible advanced analytical methodologies require a robust,
consistent suite of data. A stable and
flexible data management platform is a
prerequisite for any soft computing analysis. Once this foundation is established,
multiple analytical data marts can be
spawned from the master data management
(MDM) environment. Solutions that include data integration, data quality, MDM,
data governance, and enterprise data
access are necessary.
SEMMA techniques use a proven analytical workflow that generates exploratory analysis and helps reveal insights about
data. The SEMMA data mining process
provides methodologies for diverse business problems such as maximizing well
location, optimizing production, ascertaining maximum recovery factor, and
identifying an optimum hydraulic fracture
strategy in unconventional reservoirs, as
well as undertaking field segmentation,
risk analysis, pump failure prediction,
and well portfolio analysis.
SEMMA allows organizations to sample data by extracting and preparing a
portion of them for model building, using
one or more data tables. Sampling includes
operations that define or subset rows of
data. The samples should be large enough
to efficiently contain the significant information. It's best to include the complete
and comprehensive dataset for the "explore" step because of hidden patterns
and trends that are discovered only when
all the data are analyzed.
SEMMA enables companies to explore
data by searching for anticipated relationships, unanticipated trends, and anomalies. This returns deeper understanding

and ideas that insinuate hypotheses worth
modeling.
Using the technique, it is possible to
modify data by creating, selecting and
transforming variables to focus the model
selection process on the most valuable
attributes. This focuses the model selection
process on those variables that display
significant attributes vis-à-vis the objective
function or target variable(s).
SEMMA may be used to model data
employing analytical techniques to search
for a combination of the data that reliably
predicts a desired outcome.
It will assess the data by evaluating
the usefulness and reliability of the findings
from the data mining process. SEMMA
can compare different models and statistically differentiate and grade those models
to ascertain an optimum range of probabilistic results, delivered under uncertainty.

Explore Visually
With 3-D and 4-D surveys accelerating
across workflows for new field exploration
as well as in-field development and enhanced recovery operations, seismic data
are pivotal for today's oil and gas operations. In addition to customary seismic
wavelet data processing, it's increasingly
important to fully appreciate the hundreds
of seismic attributes that exist, and to
build a seismic data mart for advanced
analysis.
Analytical workflows that map seismic
attributes to reservoir properties are a
key way to create more credible and reliable reservoir characterization definitions
that underpin field development/redevelopment. These workflows should integrate-and supplement-data from well
logs. To maintain the spatial integrity of
large-areal reservoirs-and to identify both
stratigraphic and structural traps-requires
high-resolution seismic attributes and an
in-depth understanding of those attributes.
The process of characterizing the reservoir(s) of a potential or mature field requires analyzing large datasets collated
from well tests, production history and
core analysis results, enhanced by highresolution mapping of seismic attributes
to reservoir properties. To comprehend
the structure of the data, it's imperative
to capture the more subtle observations
inherent in these datasets.

American Oil and Gas Reporter - November 2015

Table of Contents for the Digital Edition of American Oil and Gas Reporter - November 2015