Evaluating the Precision of Well Location Datasets Using ArcGIS

Don Downey
Chevron ETC, San Ramon, CA

The advent of global positioning satellites and geographical information systems has increased the accuracy of spatial measurements by orders of magnitude. The luxury of these accurate measurements may induce us to forget some basic cartographic principles of data precision which govern older spatial datasets. A study of 700+ well locations from Iraq show that data rounding and truncation, not cartographic datum shifts are the most likely source of well location errors.

Using the lengths of each latitude and longitude number string, an index of lengths of data strings versus data precision was built. Multiplying the latitude and longitude data precision for each well location coordinate by the lengths of a degree of latitude and longitude at each latitude, an estimate of the maximum rounding error in meters is calculated. The combination of latitude and longitude errors creates an ellipse of uncertainty around a well location and ArcGIS displays a circular “buffer” graphic feature around the locations.

The study shows that for this dataset of 764 well locations, 30% of the locations could be rounded by >100m, 16% could be rounded by >1000m and 2% could be rounded by 10,000 to 56,000m. These distances are greater than the typical datum shift errors which are typically <330m. Unfortunately, most of the wells with large rounding errors are older, rank-wildcat wells that are critical for regional stratigraphic studies. Precision analysis combined with calculation of the distance and bearing between well locations identifies the best-quality data at significant cost savings compared to using satellite imagery.