U03-P01 ODP-IODP CORK Observatories: Designs … · The idea of sealing ODP ... download connection...
Transcript of U03-P01 ODP-IODP CORK Observatories: Designs … · The idea of sealing ODP ... download connection...
Modified Abstract: Since 1991, the Ocean Drilling Program (ODP) and Integrated Ocean Drilling Program (IODP) have instrumented 28 subseafloor boreholes with long-term sealed-hole observatories called Circulation Obviation Retrofit Kits (CORKs). As was described by other speakers in the U03 oral session, additional installations have been implemented or are planned during the first few years of the 2013-2023 International Ocean Discovery Program (also IODP), using D/V’s Chikyu and JOIDES Resolution. Understanding subseafloor hydrology and its relationship to hydrothermal and tectonic processes have been prime objectives of scientific ocean drilling since the late 1970’s. However, early experience indicated that holes that penetrated through marine sediments into underlying oceanic basement often allowed open exchange between formation fluids and ocean water, perturbing if not totally disturbing the in-situ hydrogeological state (Panel 1). This motivated the CORK approach to seal select holes and instrument them with long-term sensor strings and data loggers, to record the recovery from drilling disturbances to the in-situ state and monitor natural hydrologic, tidal, and geodynamic signals. Panel 1 also presents a brief summary of the designs of the CORK observatories through 2011 and the locations of the installations through 2018. The original design included a single seal at the seafloor, and later designs have allowed for separately monitoring multiple zones sealed by packers in a single hole. Also, legacy reentry holes can be retrofitted with less expensive “CORK-Lite” models deployed by remotely operated vehicles (ROVs). The sensor strings have always included pressure and temperature monitoring, and many have included self-contained fluid samplers driven by osmotic pumps (“OsmoSamplers”) that are customized for a range of geochemical and microbiological sampling objectives. Typically, data and samplers have been recovered and/or exchanged at average intervals of ~1-3 years using manned or unmanned research submersibles. Installations to date have been in sedimented young ocean crust or in subduction settings. Important geophysical findings to date include documenting the following: (1) small pressure and temperature differentials associated with vigorous off-axis hydrothermal circulation in highly permeable young oceanic crust (Panel 2); (2) formation response to seafloor tidal loading (Panel 3); (3) formation pressure as a proxy for plate-scale strain in response to tectonic stresses and earthquakes (Panels 3 and 4); (4) vertical seafloor deformation associated with slow and rapid fault slip (Panel 4); and (5) temperature variations associated with fluid flow events, volumetric strain within the crust, and turbidity flows and other oceanographic events at the seafloor (Panel 5).
U03-P01 ODP-IODP CORK Observatories: Designs and Geophysical Results Since 1991 Keir Becker1, Earl E. Davis2, Andrew T. Fisher3, Masataka Kinoshita4, and Heinrich Villinger5
1U. of Miami - RSMAS, 2Geological Survey of Canada, 3U. of California at Santa Cruz, 4Earthquake Research Institute, U. of Tokyo, 5U. of Bremen
1. Original Concept, Design Evolution, Locations, and Acknowledgements During the 1970’s, several DSDP reentry holes into young oceanic basement beneath sediment cover displayed strong downhole flow of ocean bottom water into oceanic basement, revealing the presence of highly permeable zones in basement but disturbing the in-situ hydrologic state (Fig. IA). The idea of sealing ODP reentry holes with long-term in-situ hydrogeological instrumentation was first sketched out by ED, KB, and Bobb Carson on a dinner napkin at a 1989 ODP panel meeting (Fig. IB). That concept included a capability to sample borehole fluids as they reequilibrated with formation fluids, as well as monitoring pressures and temperatures in the sealed holes with multiple geophysical objectives: (a) determining in-situ values, (b) sensing hydrological transients, (c) resolving subseafloor tidal loading effects, and (d) after filtering out the last, resolving transients of tectonic origin and plate strain signals. Following eleven deployments of original CORK’s in the 1990’s and a late 1997 USSAC-funded workshop, three engineering approaches were developed for sealing multiple intervals in a single hole: Advanced CORK or “ACORK”, CORK-II, and wireline CORK (Fig. IC). During IODP, the downhole Smart and Genius Plugs were developed for NanTroSEIZE, and sophisticated Long-Term Borehole Monitoring Systems (LTBMS) were developed for deployment from D/V Chikyu. Fig ID shows locations of installations to date and those planned through 2018. We gratefully acknowledge NSF, GSC, JAMSTEC, DFG, and C-DEBI for generous financial support since 1990 for CORK installations from the ODP/IODP drillships and subsequent submersible servicing operations.
Data loggerData logger
Data logger
Sea floor Sea floorSea floor Sea floorRe-entrycone Re-entrycone
Zone AhydraulicsamplingportZone BhydraulicsamplingportZone ChydraulicsamplingportUncased9 7/8"RCB hole Uncased9 7/8"RCB holeUncased9 7/8"RCB hole
Packer
PackerPacker
Packer Packer
10 3/4" casing Standard10 3/4" casing16" casing 16" casing 16" casing
Recoverablebridge plugMonitoringinstrument
ReamedLWD hole
ReamedLWD hole
4 Ω” liner
Zone B
Zone BZoneC
Zone AZone A
Thermistorcable Multipletool string Multipletool string
HydraulicconduitsSeal SealSeal Seal
Original CORKSmart Plug
Advanced CORK CORK II
GroutGrout
Grout
Standard 10 3/4"casingStandard 10 3/4"casing
Seismometer andstrain gauge
Samplingports Samplingports
Fluidsampler
CONTROLVEHICLE
supportpackage
soft tether
instrument string
lower packer withlead-in package hydraulic
power unit
downloadconnection
data logger
power andtelemetry
upper packer
Becker and Davis FIgure 7
Wireline CORK
4. Co- and Inter-Seismic Strain in Oceanic Crust and Subduction Zones Once tidal loading and oceanographic signals are filtered out, CORK pressure records have revealed transients from tectonic strain events in both ocean crust and subduction zones, as well as signals of secular strain in subduction settings. A prime example in oceanic crust is the record of initial strain and hydrologic drainage or viscoelastic relaxation in the Juan de Fuca plate compressed by an extensional spreading event (Fig. 4A). Good examples in subduction settings include the records of both secular strain in the over-riding plate and responses to regional earthquakes displayed at both the Nankai Trough and Costa Rica Margin (Figs. 4B and 4C). The Costa Rica Margin record is also diagnostic of slow seafloor subsidence and long-term after-slip following a nearby 2012 Mw 7.6 earthquake (Fig. 4D).
504B,896A
1253A,1255A395A
948D,949C
808I,1173B
1200C
ORIGINAL SINGLE-SEAL CORKS, 1991-2001MULTI-LEVEL CORK-II/ACORKS, 2001-2013
TOPOGRAPHY: W. SMITH & D. SANDWELLSMART/GENIUS PLUG + LTBMS, 2010-2017
889C,U1364A
857D,858G
892B
1024C,1025C
1026B, 1027C,U1301A/B,U1362A/B
0°
30°
30°
60°
60°
0° 0°60° W60° E 120° W120° E 180°
U1382AU1383B/C
C0002G, C0010GC0006 (11/17)
U1315A
HIKURANGI (2018)
Summary Geophysical Results• Uppermost young oceanic basement is very transmissive
over regional scales and supports extensive lateral fluid flow associated with relatively small pressure differentials.• Flow directions in young oceanic crust include a
significant, structurally controlled ridge-parallel component.• In subduction zones, CORKs record (a) formation
pressures associated with compressional stress regimes and (b) formation responses to tectonic events.• In both settings, the formation tidal loading response
yields hydrologic and elastic properties that may vary under stress. • In both settings, subseafloor pressures in well-sealed
CORKs provide quantitative proxies for plate-scale strain on time scales ranging from secular to seismic. The best strain monitoring is done in isolated low-porosity formations.• Slip in shallow portions of subduction zones can occur
spontaneously, be triggered by small stress changes, and can occur with little or no seismic expression.
3. Subseafloor Tidal Loading Signals and Formation Pressure as a Proxy for Strain CORK pressures also record the attenuated and phase-lagged formation response to seafloor tidal loading. This is a combination of (a) instantaneous elastic response that depends on formation loading efficiency and (b) diffusive response that depends on
formation hydraulic diffusivity (Fig. 3A). These responses vary significantly among sites (Fig. 3B) and can change with regional stress (Fig. 3C). The formation pressures also respond to plate-scale tectonic strain, and those signals can be resolved after filtering out the tidal loading signals (examples in Panel 4). The sensitivity of formation pressures to volumetric strain depends on the elastic properties and porosity of the formation and is on the order of a few kPa per microstrain (Fig. 3D), with lower-porosity formations like basement and deeper sediments being better candidates for effective strain monitoring.
5. Temperature Transients In addition to documenting background values as in Panel 2, CORK formation and seafloor temperature records have revealed a number of transient effects. Figure 5A illustrates one of the very few subseafloor fluid flow events detected by CORKs to date. Figure 5B reveals basement temperature changes in Hole 1025C in response to volumetric strain events also shown by Figs. 3C and 4A. Figure 5C documents changes in CORK seafloor temperatures that follow episodic variations in deep-water currents at the Costa Rica Margin.
2. Hydrothermal Circulation in Young Oceanic Crust Multi-CORK arrays in three young, well sedimented ridge flanks (Juan de Fuca Ridge, Costa Rica Rift, Mid-Atlantic Ridge) have produced consistent observations of temperature and pressure represented by the 3.3-3.4 Ma Juan de Fuca example shown here. The 1996 pair of original CORKs (Fig. IIA) recorded nearly isothermal in-situ temperatures at the sediment-basement interface despite a big difference in sediment thickness, requiring very active fluid circulation in basement to homogenize temperatures. The CORKs also recorded relatively small pressure differentials (Fig. IIB), implying that the uppermost basement must be highly permeable. Such observations lead to a general 2-d model for ocean crustal hydrogeology (Fig. IIC). More recently, an active tracer injection experiment at the same Juan de Fuca CORK array documented a strong component of south-north flow along structure sub-parallel to the spreading axis (Fig. IID), and
also showed that this flow was largely confined to a relatively small (<1%) interconnected “effective porosity” that provides the primary connectivity for rapid solute transport.
Poroelastic theory for pressure response to seafloor loading
Load partitioning depends on constituent compressibility
contrasts Pressure propagation and flow depends on regional compressibility contrasts
and permeability
General 2-d model for ocean crustal hydrogeology: low sediment permeability ( ̴10-16 m2)
high basement permeability ( 1̴0-11 m2) low flow through sediment (mm yr-1)
high flow through basement (10s of m yr-1) recharge/discharge limited to areas of thin or absent sediment
2A
2B
2C2D
3A
1A
1B 1C 1D
3C 1025C, M2
Strain / pressure conversion efficiency (≈ 5 - 15 kPa => 10-6 strain)
Requirements for optimal sensitivity: - low porosity - high matrix compressibility - low fluid compressibility - low grain compressibility - hydrologic isolation
Using pressure as proxy for volumetric strain
(Pa-1)
3D
Instantaneous (elastic) response to seafloor loading = “tidal loading efficiency”
Amplitude response => elastic properties
Increasing alteration, decreasing loading efficiency
3B Pressure Anomalies Recorded at Nankai ACORKs
from March 2011 Tohoku Earthquake
!
!
30°
40°N
Seamounts
Shikoku
Basin
130°E 140°
Shikoku
Eurasian plate
PhilippineSeaplate
~4cm/yr
Shikoku Basin
Fossil spreading ridge
Kyushu-Palau Ridge
132°E 134° 136° 138° 140°
36° N
34°
30°
Kyus
hu
Honshu
32°
Izu-Bonin
Pacificplate
Nankai Trough
Izu-Bonin
arc
N a n k a i T r o u g h
Kinan
Fig. 6
Trench
Delayed ACORK response - time scale for stress adjustment in adjoining plates?
Importance of complementary data … examples from Costa Rica
… down-dip/up-dip connections
… long-term afterslip
… afterslip details
Secular strain accumulation observable in hydrologically isolated formations
(Sensor drift ~ 0.15 kPa/yr = 1.5 cm/yr)
0.7 kPa/yr ≈ -0.14 μstrain/yr
3.3 kPa/yr ≈ -0.7 μstrain/yr
Subseafloor Pressure Transients from Tectonic Strain Events
After tidal loading signals are filtered from the CORK formation pressure records, other signals are evident. This example: an extensional seafloor spreading event with associated plate contraction, as recorded in three off-axis CORKs in compressional quadrant. Responses show
decreasing amplitude and strain with distance from axis, followed by hydrologic drainage (and/or viscoelastic relaxation) with increasing time constants with distance from axis.
~ -200 nanostrain
~ -130 nanostrain
~ -13 nanostrain
In several hundred site-years of CORK monitoring, only a few natural fluid-flow events have been detected. A prime example was a 6-month fluid flow pulse up-fault at Hole 892B, North Hydrate Ridge, Oregon Margin, sensed only by the thermistor positioned within the fault zone.
Fluid Flow Events
One year
Neira et al., 2016, EPSL, 450, 355-365
4D
4C
4B
4A
5A
5B
5C
T2
T7