SPATIAL AND TEMPORAL TRENDS IN THE FOOTPRINT OF OIL AND GAS DEVELOPMENT IN THE PICEANCE BASIN, WESTERN COLORADO

Development of oil and gas resources results in landscape level land use and land cover (LULC) change through the construction of energy related infrastructure. Expected future development in several economically important energy plays has placed renewed interest in understanding the LULC changes associated with energy development. A major data gap limiting our ability to examine LULC changes is the lack of datasets describing the spatial distribution of energy infrastructure. One such area lacking spatially explicit data is the Piceance Basin in northwest Colorado and northeast Utah. Energy development in the Piceance Basin began in the early 1900’s and is expected to continue into the foreseeable future, with recent estimates of 66.3 trillion cubic feet of recoverable natural gas within the Mancos Shale alone.

To facilitate analyses on energy development in the Piceance Basin, we digitized all well pads in the Colorado portion of the basin, determined the previous land use of converted areas through time, and explored potential relationship between the number of wells per pad and well pad area to model future development. A total of 4,499 unique well pads were digitized within the Piceance Basin from imagery obtained through the 2015 National Agricultural Imagery Program. The total amount of land converted to well pads was over 21 million m², with the largest amount of conversion occurring in shrub/scrub (7.9 million m²), evergreen (2.1 million m²), and deciduous (1.3 million m²) forest environments. Energy development has shifted to pad drilling, where multiple wells are drilled from a single pad, increasing the average number of wells per pad from 2.5 prior to 2001, to 9.1 between 2012 through 2016. However, during the same time period, the pad area per well has also increased from 2,030 m² to 3,504 m². Estimated kernel density curves of the relationship between the number of wells per pad and the pad site area were shown to be approximately log-normally distributed. These methods could be used to model the land use requirements of predicted future development in the Piceance Basin. The data and results from this study will facilitate future studies assessing the spatial and temporal effects of energy development on ecosystem function in the Piceance Basin.