Chan et al. (2005) hypothesized that rinded concretions in the Navajo Sandstone are de novo mixing products of oxidized and reduced fluids. Dana (1896) described similar rinded concretions that also showed evidence of inward thickening, terming them centripetal structures. We argue these concretions had siderite precursors and that the reduced and oxidized mineral phases reflect groundwater chemistry that shifted from reducing to oxidizing as CO2 and CH4 were swept from the Navajo reservoir in the Escalante anticline.


We are not surprised that Chan et al. (2011) did not find Fe(II) in the bands of cement that comprise the rinds. These inward-thickening rinds formed when Fe(II) migrated to the perimeter of siderite-cemented concretion cores, precipitating as Fe(III)-oxide minerals (Loope et al., 2010).We have found abundant siderite pseudomorphs in the interiors of the largest iron-oxide concretions in the Navajo (Loope et al., 2011). Unlike the Fe-oxide minerals in the rinds, the Fe-oxides in the pseudomorphs formed by replacement of siderite crystals.

We wouldn't expect petrographic relationships to show that Fe(III) oxides and Fe(II) carbonates precipitated in equilibrium. Joint-associated, poikilotopic ferroan calcite concretions are abundant in the southeastern portion of our study area—independent evidence for a Navajo aquifer that was reducing and carrying both Fe(II) and CO2 in solution.


Goethite and lepidocrocite are known products of siderite oxidation (Cornell and Schwertmann, 2003; p. 361), and lepidocrocite precipitation is inhibited only at high HCO3- concentrations (Cornell and Schwertmann, 2003, their figure 13.7). A halo of lepidocrocite around the concretions (Parry, 2011) is fully consistent with our model.

We cannot evaluate Chan et al.’s (2011) statement regarding geochemical calculations—the relevant geochemical variables (P, T, X) are not specified—but the comment is a reasonable description of the bleaching and Fe(II) transport events we described. Bleaching will only occur if “primary” Fe(III)-oxide phases in the Navajo aquifer are under-saturated. We contend that the reduction of Fe(III) occurred prior to siderite precipitation; Fe(II) was transported down-gradient to sites of siderite precipitation.

It is possible that Fe(II) remaining in aqueous solution (after siderite precipitation) was oxidized down-dip, but much of it precipitated as the poikilotopic ferroan carbonate concretions that we mapped (Loope et al. 2010). It certainly would have been oxidized as it flowed out of the Navajo aquifer into the ancestral Colorado River. The chemistry of a solution is controlled by the values of P, T, and X (not by the values those variables will hold in the future).

Contrary to Chan et al. (2011), Parry et al. (2009) did not report siderite from the Covenant oil field. Siderite was used only as a geochemical construct.

The absence of siderite in spring precipitates is to be expected. Fe(II) transported by such systems precipitates as Fe(III)-oxide (Shipton et al., 2004). These springs provide direct evidence of Fe(II) transport in the Navajo aquifer by CO2-charged solutions.


Northwest-to-southeast orientation of comet tails (Loope et al., 2010) shows that Fe was transported down-dip. This fluid was not buoyant. Chan et al.’s (2011) statement, “the fact that aqueous fluids with even a small amount of dissolved CO2 and CH4 are less dense than fluids without the dissolved gases” is incorrect for nearly all PTX space. Both CO2 and CH4 are relatively insoluble in water; they do not form hydrogen bonds with adjacent water molecules. In general, dissolution of CH4 in liquid water will decrease the density of the solution whereas dissolution of CO2 causes solution density to increase (Neufeld et al., 2010). The tendency of aqueous solutions to become more dense upon dissolution of CO2 allowed it (and CH4) to accumulate in the bottom waters of Lake Nyos, eventually leading to the gaseous eruption that killed 1700 people in 1986 (Giggenbach, 1990). Methane is relatively insoluble in low-temperature waters compared to CO2 (e.g., Cui et al., 2004); thus, CH4-dominant waters will only be produced if the gas with which the solution is in equilibrium has a CH4:CO2 ratio greater than 20:1.

Chan et al. (2011) state that joints in the study area formed 23.3 m.y. ago, but their references indicate this is just the oldest possible age of the joints–the beginning of the Miocene. Further, the precise age of the joints is moot because the geochemical trap that led to Fe precipitation required fracture continuity to the surface and appropriate hydraulic pressure—conditions that could have developed any time after fracturing.