Transcript Philippe Theys` presentation

Data Quality Conference (New Orleans 11-2/3-99)
• Metadata is the key to Data Quality.
• Data models should be based on business
data requirements.
• Data shelf life is longer than any product
currently manufactured.
dq-s4- 1
What are the business requirements for geodata?
• The main purpose of geodata is to help
management take decisions.
Decisions require:
– The information used to take the decision,
– The decision threshold and rules,
– The interval of confidence of the information.
• Other purpose is to serve as a baseline for
future analysis (???).
dq-s4- 2
What are the business requirements for geodata?
• The main purpose of geodata is to help
management take decisions.
• Decisions require:
– The information used to take the decision,
– Example: Measurement of porosity,
– The decision threshold and rules,
– Example: Minimum porosity required to complete
reservoir.
– The interval of confidence of the information,
– Example: Precision and accuracy of measured porosity
dq-s4- 3
What are the business requirements for geodata?
• The main purpose of geodata is to help
management take decisions.
• Decisions require:
–
–
–
–
–
–
–
–
–
The information used to take the decision,
Example: Measurement of porosity,
Available from logging company
The decision threshold and rules,
Example: Minimum porosity required to complete
reservoir.
Available from O & G company (is it stored?)
The interval of confidence of the information,
Example: Precision and accuracy of measured porosity
Not currently delivered by logging company.
dq-s4- 4
Long term usage of geodata
Geodata may be used decades after acquisition
(field study, time-lapse techniques)
Business requirements decades from now are
difficult or impossible to define at the time of
acquisition
Need to keep all acquisition data and
description of the acquisition conditions
dq-s4- 5
Examples
• Observations of volcanic eruptions by the
Chinese (confirmed by ice cores in Antartica
and Greenland). Observations of comets,
eclipses and star explosions.
• Observations of finches by Darwin (long
before he thought about the theory of
evolution).
• Description of an experiment in physics.
• Keeping a “log” to describe life on a ship.
• Keeping all AIT data before the modeling
software (using all raw TR curves) was made
available. This software does need all these
raw curves.
dq-s4- 6
Proposed data “scheme”
• Short term use
• Business requirements can
be defined. Should be done
at job planning phase.
• Requirements vary with
users.
• There could be multiple
users/partners/departments
(with conflicting
requirements).
• Format needs to be
optimized for efficiency and
size.
• Service company (and
PWLS) can propose default
format
• Long term use
• Business requirements are
not defined.
• The service company in
best position to assess
what could be useful.
• Everything, even
apparently irrelevant items
need to be kept.
• Optimization and size are
not an urgent concern.
dq-s4- 7
Data scheme
Acq
Sa
S
S’
L
S”
dq-s4- 8