Example Cleanup Project

This example is based upon a real case - a cleanup project we had with a multinational.


The Challenge

The client had a regional dataset that they wished to clean up prior to starting a new round. The client had conscientiously tried to manage their data and had corporate standards, however the OpenWorks database had been in use for 7 years and had many generations of ‘standards’ in it. We were given 10 man days in which to clean the data.

The Approach

First we ran the RoQC Tool ‘audit_report’ to document the initial status of the project. The audit report is also a very valuable tool for discussions with the client when deciding what should be done and the priorities. In this case the priorities were:
  • Well locations (inc. deviation data)
  • Picks
  • Logs (inc. standardising the curve dictionary)
The project audit showed that even though the project had >20 stratigraphic columns loaded close to 50,000 picks were not in accordance to these columns! A very high percentage of the logs had non standard names and/or were missing units.

A review with the client resulted in a request to define the necessary stratigraphic columns and to extend the corporate log standards where necessary.

Discussions with the users/specialists resulted in refining 5 of the stratigraphic columns, defining/implementing 3 new ones and deleting the others.

The new strat columns were built, QC’d and implemented then the process of cleaning the pick/grid/fault interpretation/etc data began. Using the RoQC Tools ‘list_illegal_picks’ and ‘strat_cleanup’ all of the interpretations that did not comply with the standards(strat columns) were moved to the correct names. This process requires interaction with the users.

After cleaning the picks/interpretations to comply with the standards, ‘pick_data_manager’ was run and an official, corporate, set of picks was built.

The audit report also identified the wells that had obviously incorrect locations and where the position logs were incorrectly calculated. These wells were then corrected.

The RoQC Tool ‘log_dm’ was used to clean the log data to comply with the new extended corporate standard. Since the client had consistently bulk loaded log data from known data sources it was decided to bulk fix the curves that were missing units. These curves had their units set to match the definitions in the curve dictionary.

The Results

After 10 man days of work the client had a project with:-
  1. Clean well locations (1820)
  2. All picks/grids/fault polygons/unit attributes/etc according to corporate standards (83,000 picks)
  3. All logs according to corporate standards (10,500)
  4. Stratigraphic columns that support their business needs and are according to corporate standards
  5. All logs now have units
  6. Curve aliases implemented
  7. Example templates using curve aliases
  8. Miscellaneous other things fixed - e.g. standard company names, country set on all wells
The following image is from Correlation after the cleanup. Names have been changed to be unrecognisable but the data is real and gives a true indication of the effect of our cleanup work.



Move the mouse over the image to see what the original data looked like....
Can take 3-4 seconds.
Click on image to get a larger image.

If you do not get the 'before_cleanup' image when the mouse is over the image, please turn on 'allow Javascript' in your browser.


<< Back to Consulting

    
 
©1998-2006 RoQC DataManagement - All Rights Reserved. Updated 23 September, 2007