Real-time drill mud gas logging at the USDP-4 drilling, Unzen volcano, Japan

https://doi.org/10.1016/j.jvolgeores.2008.03.031Get rights and content

Abstract

During the Unzen conduit drilling project USDP-4, the gas phase dissolved in the drill mud was continuously analyzed. Starting from the volcano's north flank an almost complete gas profile was achieved to the final depth of 1995.75 m in July 2004. Limitations were given due to the extremely difficult drilling conditions. The highly fractured rock formation led to loss of drill mud circulation in the shallow parts of the drill hole. Significant fluid inflow horizons did not occur above 800 m (drill string length). Starting from that depth on, invading fluids were detected with the real-time gas monitoring system. Major variations in the mud gas composition occurred only below a depth of 1000 m. Of major importance are fluid inflow zones with high 3He/4He at depths of 1555 m (7.3 RA), 1755.5 m (7.48 RA) and 1977.4 m (6.21 RA). These values indicate a significant influence of fluids with mantle signature. Furthermore, enhanced methane, radon and helium concentrations were also detected at specific depth.

These detected major fluid- and gas inflow horizons may be able to explain magmatic degassing processes, related to the Unzen's eruption mechanism. This is generally true for the main fluid inflow zones and especially for the detected inflows at 1555 m, 1755.5 m and 1977.4 m.

Furthermore, a correlation between lithology and gas composition was observed. Higher H2S concentrations were detected while drilling in pyrite-rich rocks. Cracks and fissures as well as lithological changes are often correlated with increasing amounts of gas.

Trends with depth, from a minor to a more magmatic influenced regime were observed together with a change in hydrothermal alteration of the surrounding rock. This corresponds with the magmatic conduit zone which was penetrated at a depth of 1600 m, and supports the model of a high influence of hydrothermal fluid, accelerating cooling and mineral alteration.

Introduction

The Unzen Volcano, located in the Shimabara Peninsula, Kyushu, southwest Japan, started its volcanic activity about 500 000 years ago (Hoshizumi et al., 1999). The volcano was nominated as one of the Decade Volcanoes of the United Nations International Decade of Natural Disaster Reduction. It represents an important type of dangerous volcano. The last eruption (1990–95) was monitored and investigated in detail. Several models on the eruption mechanisms and magmatic processes were obtained (e.g. Nakada et al., 1999). The focus of the “Unzen scientific drilling project” (USDP) was to further clarify the regional crustal structure and magma evolution processes that control the manner of growth and failure of the volcano. In addition, processes of conduit formation, magma degassing, and magma's interaction with groundwater were investigated. The USDP-4 was conducted within the framework of the “International Continental Scientific Drilling Program” (ICDP), and was financed by the Japanese Ministry of Education, Culture, Sports, Science. and Technology (“MEXT”) and ICDP.

The fluid volume of a volcanic system is of significant importance for the formation process of magmas, the melting behaviour of rocks and the eruptive behaviour of a volcano. Circulating fluids are a primary medium for material and heat transport processes in a volcanic edifice. Furthermore, the rheological behaviour of the magma as well as melting and crystallisation processes are also strongly influenced by fluids. Petrological investigations show that ‑ after a magma mixing event prior to the ascend – the post-mixing dacitic melt is considered to contain 6 ± 1 wt.% H2O (Holtz et al., 2005, Sato et al., 2005). At the surface (the lava dome), almost no water was present. The water was liberated at less than 1–2 km deep. In view of these constraints, different physico‑chemical models show a strong increase in magma viscosity and a pressure drop directly below the growing dome. Ascension paths for the released gases are cracks and fissures inside the volcanic edifice (Nakada et al., 1997). The degassing process of the magma at this depth is, however, still not yet fully understood. Measurements of the summit fumarolic gases after the eruption of Unzen volcano have also been investigated for evaluation of the volcanic gas components (Shinohara et al., 2008-this issue). It is assumed that the volcanic gas that escaped through the conduit wall was discharged from the volcanic edifice (Kazahaya, 2000, Nakada, 2000, Shinohara, 2000, Ohsawa et al., 2002). As the center of activity of the last eruption is located between two geological faults (Chijiwa and Akamatsu-dani fault), escape paths of volcanic gases into the volcanic edifice through the conduit wall are discussed. The δ13CCO2 measurements in cold springs and shallow wells indicate a large influence of a volcanic CO2 inflow into the groundwater system. However, calculations also showed that the fumarolic CO2 flux is one or two orders of magnitude higher than the CO2 discharged by groundwater runoff (Ohsawa et al., 2002).

The helium isotopic ratio 3He/4He were measured in three different hot springs located at the Unzen volcano during the 1990 eruption and thereafter. The helium isotopic ratio is often used to identify the sources of volcanic products (Notsu et al., 2001). The 3He/4He of the atmosphere is 1.39 · 10- 6 and defined as “RA”. The 3He/4He values of 3–10 RA are interpreted as values close to mantle ratios. RA values of MORB are typically of ∼ 8.2 and plume RA values of ∼ 35-38 (Moreira et al., 2001, Dunai and Porcelli, 2002). 3He/4He values of ∼ 7 RA have been found for magmatic He of the Japanese island arc (Nagao et al., 1981, Kita et al., 1993). The 3He/4He ratios of the Unzen hot springs lie within a range between ∼ 4.06 RA (Obama hot spring) and 7.07 RA (Shimabara hot spring) and vary with respect to the volcanic activity of Unzen (Notsu et al., 2001). These findings demonstrate a close connection between the groundwater and the magmatic fluid systems.

Investigations of the flux of magmatic volatiles dissolved in groundwater suggests two types of magma degassing: (1) degassing from deep reservoirs through fractures and (2) from magma conduits through aquifers (Kazahaya et al., 2004). However, new calculations based on budgets of volcanic gas components, SO2 emission rates, magma effusion rates and melt inclusions are leading to the conclusions that most ascending gases were discharged through summit degassing. Due to these results, the volatile flux could be quantitatively explained without loss to the volcanic edifice (Satoh et al., 2003, Shinohara et al., 2004). Therefore the degassing process of ascending magma can be explained with a closed-system degassing (Satoh et al., 2003).

The USDP-4 drilling was conducted as shown in Fig. 1. The online gas monitoring was carried out to obtain a gas profile over the whole borehole length in order to gain information on the volcano fluid circulation system, to detect possible ascension paths for gases produced during magma degassing and to determine hydrothermal and/or volcanic fluid inflow horizons.

As shown in Fig. 1, the drilling encountered the conduit zone at ∼ 1600 m drill string length. The conduit zone comprises of non-stratified, poorly sorted polymictic breccia and coherent dacite. The material varies in color and alteration (Goto et al., 2004). The conduit region is ∼ 500 m wide with several parallel, vertical dykes, intruding into the compact volcanic breccia. The conduit of the 1990–1995 eruption is ∼ 40 m large and in the same dimension as the lava spine (Nakada et al., 2004).

The measured temperature inside the conduit region was low (∼ 200 °C) compared to previous estimations of 600 °C (Nakada et al., 2005). This is explained by intensive hydrothermal fluid circulation within the conduit zone, accelerating cooling and alteration (Eichelberger, 2004, Nakada et al., 2004).

Section snippets

Experimental setup and data processing

The mud gas comprises mainly of the air dissolved in the circulating mud, but also contains gases entering the borehole through cracks and fissures and gases released from crushing the rocks during drilling operations (Erzinger et al., 2004). The gas was extracted using a “degasser”, consisting of a stainless steel tube with a stirring motor on top (custom-built unit), which was installed in the mud pit above the shaker screens. A membrane gas pump in the laboratory trailer produced a constant

Results

Fig. 3 displays the geological and gas concentration profile from 800–1995.75 m (drill string length). Numerous fractures and fissures in the shallow rock formations (0–800 m) led to the loss of drill mud circulation. Above 500 m almost no drill mud returned. These difficulties caused not only severe delays within the drilling schedule but made the detection of dissolved drill mud gases nearly impossible. Important fluid inflow horizons have not been observed above 800 m. From that depth on,

Discussion

Increasing CH4, He, H2S and Rn amounts in the drill mud are evidence for fluid inflows into the borehole (Zimmer and Erzinger, 1995, Erzinger et al., 2004). Cracks, fissures and veins within the rock formation as well as rocks with a higher permeability are most likely pathways for fluids. At the Unzen drilling, elevated gas concentrations were observed within the drilled rock formations. The areas of lava dikes seem to contain higher gas amounts (e.g. at 1082 m, 1124 m, 1761–1789 m, 1975 m).

Acknowledgments

We thank Samuel Niedermann and Enzio Schnabel at the GFZ for conducting the He isotope measurements, Hiroshi Shinohara, Hiroshi Shimizu, Sumio Sakuma and Kozo Uto for the many help and the fruitful discussions, the DFG for funding the project (DFG project ZI 715/1-2) and especially the GFZ for the unbureaucratic support during the 2nd drilling period in 2004.

References (33)

  • Eichelberger, J.C., 2004. The Present and Future of Volcano Research Drilling, AGU 2004 Fall Meeting. Eos,...
  • ErzingerJ. et al.

    Real-time mud gas logging during drilling of the SAFOD Pilot Hole in Parkfield, CA

    Geophysical Research Letters

    (2004)
  • FyfeW.S. et al.

    Fluids in the Earth's Crust: Their Significance in Metamorphic, Tectonic, and Chemical Transport Processes

    (1978)
  • Goto, Y., Yoshimoto, M., Shimano, T., Kurokawa, M., Sakuma, S., Sugimoto, T., Noguchi, S., Toramaru, A., Hoshizumi, H.,...
  • HoltzF. et al.

    Experimental petrology of the 1991–1995 Unzen Dacite, Japan. Part I: Phase relations, phase compositions and pre-eruptive conditions

    Journal of Petrology

    (2005)
  • Kajiwara, T., Ikeda, R., Nakada, S., Uto, K., Nishi, M., 2004. Unzen Volcano Scientific Drilling: Well Logging Data of...
  • Cited by (14)

    • Identification of gas inflow zones in the COSC-1 borehole (Jämtland, central Sweden) by drilling mud gas monitoring, downhole geophysical logging and drill core analysis

      2020, Applied Geochemistry
      Citation Excerpt :

      After the KTB experience, on-line monitoring of fluids and gases from circulating drilling mud has been successfully conducted in several continental scientific drilling projects in crystalline and sedimentary strata, most of which are funded and supported by the International Continental Scientific Drilling Program (ICDP). These drilling projects were focussing on fault zones (Corinth, Chelungpu, SAFOD, New Zealand (Erzinger et al., 2004; Wiersberg and Erzinger, 2007, 2008, 2011)), volcanic systems (Unzen, Long Valley (Tretner et al., 2008)), collision zones (Donghai, COSC (Luo et al., 2004; Lorenz et al., 2015)), and gas hydrates (Mallik (Wiersberg et al., 2005)). On-line drilling mud gas monitoring has also been applied in non-ICDP scientific drilling projects, e.g. the Wenchuan fault zone drilling (Gong et al., 2014).

    • Real time drilling mud gas response to small-moderate earthquakes in Wenchuan earthquake Scientific Drilling Hole-1 in SW China

      2017, Journal of Asian Earth Sciences
      Citation Excerpt :

      Furthermore, few real time drilling mud gas data have ever been applied in earthquake related hydrological analysis. The real time drilling mud gas analysis are widely used in scientific research, e.g., underground shear zone detection (Erzinger et al., 2004), fault zone permeability architectures analysis (Wiersberg and Erzinger, 2008, 2011), fault zone structure evolution analysis (Gong et al., 2015) and magma degassing process study (Tretner et al., 2008). Mechanically grinding rocks, intersecting fissures and faults during the drill will release gases trapped in pore space or cracks, and therefore leave significant gas composition changes in the returning drill mud.

    • Origin and in situ concentrations of hydrocarbons in the Kumano forearc basin from drilling mud gas monitoring during IODP NanTroSEIZE Exp. 319

      2015, Applied Geochemistry
      Citation Excerpt :

      An assembly only for scientific purpose has been developed during drilling of the KTB in Germany (Zimmer and Erzinger, 1995) and has been deployed on several scientific drilling projects of the International Continental Scientific Drilling Program ICDP. Gas was continuously extracted from returning drilling mud, analyzed in real time and sampled for off-line studies during e.g. fault zone drilling, drilling for gas hydrates, and volcanoes (Erzinger et al., 2006; Tang et al., 2013; Tretner et al., 2008; Wiersberg and Erzinger, 2007, 2008, 2011). For deep continental drilling, drilling mud is pumped through the drill string and flows back through the annulus between the drill pipe and the bore hole in order to stabilize the bore hole, to cool the drill bit, to extract drill cuttings and to insert e.g. core barrel or geophysical logging probes (Fig. 2).

    • Real time fluid analysis during drilling of the Wenchuan Earthquake Fault Scientific Drilling Project and its responding features

      2014, Tectonophysics
      Citation Excerpt :

      The electrical motor, which drives the stirring impellers, needs long term stability and good air tightness in order to stir the mud efficiently. The papers have described the gas separator and gas flow line in detail (Erzinger et al., 2006; Luo et al., 2004b,c; Tang et al., 2006, 2010; Tretner et al., 2008). The gas extracted from the drilling mud would be pumped through the PTFE tubes into the laboratory after the adjustment of air balance, which could prevent drilling mud from condensing within the gas flow line and maintain constant gas flow.

    View all citing articles on Scopus
    View full text