2007 Ground Water Summit


Monday, April 30, 2007
4:30 p.m. - 7:30 p.m.

Tuesday, May 1, 2007
9:30 a.m. - 1:30 p.m.

Tuesday, May 1, 2007
4:30 p.m. - 7:30 p.m.

Multivariate Statistical Analysis of Hydrochemical Data for Water Resources Characterization, Piceance Basin, Colorado

Tamee R. Albrecht, MS, student and Geoffrey D. Thyne, Colorado School of Mines

Portions of the arid western US are experiencing rapid development of petroleum resources in areas where water resources are scarce and may be vulnerable to impact.    Hydrochemical data from the area undergoing the fastest petroleum development in western Colorado, the Piceance Basin, was analyzed to better characterize the surface and groundwater resources. The Laramide-age basin consists of sedimentary rocks with substantial structurally-controlled faulting and is an important source of gas production from deeper sandstones. The two shallow aquifer systems include alluvium adjacent to rivers, and discontinuous sandstone lenses of the Eocene Wasatch Formation. Normalized and standardized inorganic data from 775 water samples were separated into seven statistically distinct groups using hierarchical clustering. Principal component analysis showed that most of the variability in the water chemistry could be defined by three processes or sources.

Several clusters are lower TDS, Ca-Mg-HCO3 type water consistent with natural background.  However, others were characterized as impacted with either distinctive Fe-Mn or higher TDS, Na-Cl-SO4 signatures. The Fe-Mn samples are similar to background, however most are located at a methane seep caused by gas well completion problems.  Additional samples of this water-type can be found at scattered locations across the study area suggesting impact from either natural or petroleum production related leakage of methane. The samples with the Na-Cl signature are impacted by septic tank effluent and/or leaking formation water from the gas production zone.  The remaining impacted samples show elevated sodium, chloride and sulfate and may represent mixing between natural waters and the Na-Cl source.  Spatially, no significant correlation exists between samples with elevated Na-Cl and structural trends, which supports the hypothesis that Na-Cl waters are not derived from natural leakage upward along deep-seated faults. Continuing research will explore the source of the NaCl with samples that have organic parameters derived from gas production.


The 2007 Ground Water Summit