Influence of wall-rock alteration and fluid mixing mechanisms in the chemistry of thermal fluids and mud-pool sediments at Caldeiras da Ribeira Grande (S. Miguel Island, Azores)

Abstract

The Caldeiras da Ribeira Grande site is placed at the north-eastern side of the Ribeira Grande geothermal system (S. Miguel Island) and includes a fumarolic field with multiple discharge points, closely associated with a mud pool. The major discharge is artificially confined to a tank, also fed by a cold (≈19 °C) spring water, allowing the accumulation of diluted thermal fluid and mud. In the open fumarole site placed to the NE of the tank another persistent thermal fluid discharge can be observed, whose temperature (70 °C), pH (≈2–3) and electrical conductivity (185.4–3,300 μS/cm) values, as well as major element contents, are within ranges bracketed by measurements performed at the same site over time. Despite of differences imposed by inhomogeneous dilution processes within tank, the compositional fingerprints of both thermal fluids are similar, pointing to a common origin. Variations in Br, Mo, W, Nb, Sn and Sb contents are minimal, but the thermal fluid discharging at the open fumarole site reveals concentrations above 100 μg/l in Zn, Ti, Mn, Rb, Y, Ce and Nd and up to 11 μg/l in As, Hg and Pb. The chondrite-normalised REE patterns are equivalent to those usually reported for acid, SO₄ ²⁻-bearing geothermal fluids and the available data show that the fluid chemistry strongly depends on the alteration experienced by the hosting volcanic rocks. This alteration led to relative [Hg, S, Ag, Cu, ±Ge] enrichments, but a significant part of the [K, Na, Ca, Mn, Mg, Zn, Co, ±W ± REE] original rock contents was carried out by the modified fluid. Kaolinite and alunite(-jarosite), besides fine-grained relics of K-feldspar and quartz, are the fundamental components of the mud forming the boiling pool or the mounts accumulated in tank. This mineralogical similarity is consistent with bulk chemical compositions and differences in concentration rarely exceed 1.5–2 times for the large majority of the elements analysed. Some compositional features displayed by thermal fluids and coexisting muds are sensitive to pH and Eh variations, reflecting also changes in fluid/rock reaction paths due to variations in acid steam production/composition and relative proportion of steam subjected to condensation. These chemical features can be used as proxies to monitor the Ribeira Grande geothermal system, provided that adequate time series data exist for pertinent parameters in representative sampling sites.

Appendix

The fluid–rock interaction recorded at CRG site alters the parent rock (trachyte), produces different fine-grained mineral phases able to accumulate as mud sediment at the place of thermal fluid discharge, and influences the fluid composition. Having the chemical composition of fresh (RF) and altered (RA) rock, as well as of the mud sediment (SD) and the thermal fluid modified during trachyte alteration (FM), the original fluid composition (FO) involved in the hydrothermal process can be roughly assessed by means of the following mass balance:

$$M_{\text{RF}} C_{\text{RF}}^{i} + M_{\text{FO}} C_{\text{FO}}^{i} \approx M_{\text{RA}} C_{\text{RA}}^{i} + M_{\text{SD}} C_{\text{SD}}^{i} + M_{\text{FM}} C_{\text{FM}}^{i}$$

(1)

where M refers to the mass of each equation member and C ⁱ to the concentration of element i measured in it. Taking into account the long-lasting thermal fluid discharge at CRG and the extent of the hydrothermal alteration, one may assume that fluid mass variations are negligible; consequently:

$$M_{\text{FO}} \approx M_{\text{FM}}$$

(2)

In these circumstances,

$$M_{\text{RF}} \approx M_{\text{RA}} + M_{\text{SD}}$$

(3)

or

$$M_{\text{SD}} \approx M_{\text{RF}} - M_{\text{RA}}$$

(4)

This means that M _SD depends mostly on the hydrothermal alteration experienced by trachyte rocks interacting with FO, and that other possible mass contributions to M _SD (from decanting of fine solid particles in suspension or destabilisation of colloidal suspensions or particular ionic complexes) are minor. In consequence, replacing (4) in (1) and rearranging the resulting equation:

$$M_{\text{FO}} C_{\text{FO}}^{i} \approx M_{\text{RA}} \left( {C_{\text{RA}}^{i} - C_{\text{SD}}^{i} } \right) + M_{\text{RF}} \left( {C_{\text{SD}}^{i} - C_{\text{RF}}^{i} } \right) + M_{\text{FO}} C_{\text{FM}}^{i}$$

(5)

or

$$C_{\text{FO}}^{i} \approx \frac{{M_{\text{RA}} }}{{M_{\text{FO}} }}\left( {C_{\text{RA}}^{i} - C_{\text{SD}}^{i} } \right) + \frac{{M_{\text{RF}} }}{{M_{\text{FO}} }}\left( {C_{\text{SD}}^{i} - C_{\text{RF}}^{i} } \right) + C_{\text{FM}}^{i}$$

(6)

Considering now the transformation RF ⇒ RA, one may establish the relationship between M _RF and M _RA, as reported in Grant (1986):

$$M_{\text{RA}} = \frac{{M_{\text{RF}} }}{{C_{\text{RA}}^{i} }}\left( {C_{\text{RF}}^{i} + \varDelta C_{{{\text{RF}} - {\text{RA}}}}^{i} } \right)$$

(7)

Thus, by replacing (7) in (6), the final equation results:

$$C_{\text{FO}}^{i} \approx \frac{{M_{\text{RF}} }}{{M_{\text{FO}} }}\left[ {\frac{{\left( {C_{\text{RF}}^{i} + \Delta C_{{{\text{RF}}\_{\text{RA}}}}^{i} } \right)}}{{C_{\text{RA}}^{i} }} \times \left( {C_{\text{RA}}^{i} - C_{\text{SD}}^{i} } \right) + \left( {C_{\text{RA}}^{i} - C_{\text{RF}}^{i} } \right)} \right] + C_{\text{FM}}^{i}$$

(8)

allowing the estimation of C ⁱ_FO in function of M _RF/M _FO. The latter ratio corresponds to the mass proportion between the fresh rock and the original fluid involved in the hydrothermal alteration process, therefore, providing an approach to evaluate the relative amount of FO needed to accomplished the transformation fresh rock + original fluid ⇒ altered rock + mud + modified fluid.