Deep Roots of Geothermal Systems Understanding and Utilizing Gudni Axelsson, Iceland GeoSurvey Warwick Kissling, GNS Science Chris Bromley, GNS Science Central and South American Workshop on Geothermal Energy, Cuernavaca, Mexico, April 2016 ANNEX 12 Technology Collaboration Project on Deep Roots Axelsson Kissling & Bromley Cuernavaca Workshop April 2016
CONTENTS The deeper & hotter roots of volcanic geothermal systems Heat transfer from deep heat sources to shallower levels Motivation: sustainability of conventional reservoirs? Much greater energy output from super-critical T&P wells. Examples of on-going research aimed at deep roots : Part A: Iceland - DRG project Part B: New Zealand TVZ hot deep roots Others: Italy, etc.. Conclusions Axelsson Kissling & Bromley Cuernavaca Workshop April 2016 Photo A. Sigurdarson
EXAMPLES of VOLCANIC-TYPE GEOTHERMAL SYSTEMS, with ENERGY PRODUCTION, WORLDWIDE Miravalles, Costa Rica Ahuachapan, El Salvador Berlin, El Salvador San Jacinto, Nicaragua Momotombo, Nicaragua Bouillante, Guadeloupe Los Azufres, Mexico Los Humeros, Mexico The Geysers, California, USA Puna, Hawaii, USA Matsukawa, Japan Hatchobaru, Japan MakBan, The Philippines Palinpinion, The Philippines Kamojang, Indonesia Wayang Windu, Indonesia Wairakei, New Zealand Rotokawa, New Zealand Olkaria, Kenya Aluto-Langano, Ethiopia Larderello, Italy Pico Alto, The Azores Hengill, Iceland Krafla, Iceland >90% of global geothermal power production Axelsson Kissling & Bromley Cuernavaca Workshop April 2016
(1) The basic concept of heat transfer by volcanic geothermal systems hasn t changed much in 50 years White (1967) Sæmundsson (2014) (2) The main heat sources are deep-seated magmachambers, cooling plutons or smaller intrusions (dykes, sills) Super-critical target Axelsson Kissling & Bromley Cuernavaca Workshop April 2016
(4) Either : a) deeply circulating water must come into direct contact with deep magma chamber/pluton, or b) smaller intrusions at shallower levels must come into direct contact with groundwater, or both White (1967) Sæmundsson (2014) (3) Heat conduction alone is much too slow to explain rapid heattransfer from roots to exploitable levels.therefore. Axelsson Kissling & Bromley Cuernavaca Workshop April 2016
HEAT TRANSFER FROM ROOTS Transfer of heat up to shallower (exploitable) levels is a complicated process involving flow of magma, flow of fluids (two-phase or supercritical fluids and/or superheated steam), heat conduction and convection, as well as thermo-elastic rock mechanics (brittle-ductile) and chemical processes So difficult to explore, hasn t been drilled yet and can t be modelled with conventional modelling tools ie challenging..! Axelsson Kissling & Bromley Cuernavaca Workshop April 2016
Convective Downward Migration of Fractures Why is heat transfer from hot intrusions, or magma, into geothermal systems so rapid? Cooling water migrates into hot rock through fractures that open up by contraction; energy that is derived from the cooling is transported upwards by convection (Bödvarsson,1951-81, Lister 1972-82) Drilling through lava in Iceland has demonstrated this and revealed an extremely thin conductive boundary layer Lister Axelsson Kissling & Bromley
THE DEEP ROOTS CHALLENGE What is the nature of the heat sources? Deep magma chambers or shallower partial-melt in dykes/sills? Heat transfer up to reservoirs: by deep fluid circulation (down to magma chambers) or by interaction between shallower intrusions and water circulation? Is there sufficient deep permeability for extraction through deep wells? Can stimulation by cooling help? Are supercritical conditions prevailing at great depth? Is there superheated steam at shallower levels? How can fluid with high temperature, high pressure and corrosive chemistry (HCl, HF) be utilized? Axelsson Kissling & Bromley Cuernavaca Workshop April 2016
RELEVANCE FOR FUTURE UTILIZATION OF GEOTHERMAL RESOURCES Greater output from wells drilled into the roots, if sufficient permeability can be found, because of higher temperature and pressure, especially if supercritical or producing super-heated steam perhaps more than 5x normal output Extends resource vertically rather than laterally Less environmental effects expected because of greater depth and smaller horizontal extent Potential for applying reinjection-production doublets, or EGStechnology, if permeability is limited Also for reinjecting deep (4-5 km) and extracting shallow (above 3 km depth) Still numerous technical problems to be overcome Axelsson Kissling & Bromley Cuernavaca Workshop April 2016
DEEP ROOTS RESEARCH Advancement of exploration methods to improve resolution at depth; seismic methods could provide better resolution than resistivity methods Combined analysis of different types of exploration data Study of extinct and eroded volcanic geothermal systems Mathematical modelling of the heat extraction and heat transfer mechanisms involved and relevant advancement of modelling methods and software Modelling of chemical processes (at depth, in wells and on surface) Advancing well design and drilling technology Developing utilization technology Axelsson Kissling & Bromley Cuernavaca Workshop April 2016
PART A: GEORG DEEP ROOTS PROJECT 3-year project, jointly funded by GEORG, 3 Icelandic power companies and Orkustofnun 1. Study of extinct/eroded roots of volcanic systems 2. High-resolution seismic monitoring, MT, deformation monitoring (GPS, INSAR) Axelsson Kissling & Bromley Cuernavaca Workshop April 2016
GEORG DEEP ROOTS PROJECT cont. 3. Advancement of modeling methods to understand physical processes and manage reservoirs 4. Material selection and component design for wells 5. Design of power processes for energy utilization and material recovery Has considerable international input and cooperation A 2-day workshop covering achievements and future plans held in Reykjavík in March 2016, see: http://www.georg.hi.is/node/255 Axelsson Kissling & Bromley Cuernavaca Workshop April 2016
MODELLING OF GEOTHERMAL SYSTEMS Conventional industry-standard reservoir models (e.g. TOUGH2) are used to assess capacity and for management purposes during utilization; usually shallow (2-3 km) and with p+t below critical point; heat-sources idealized as steady inflow of mass and energy at the bottom of the models; must be history-calibrated with well-data. There is a need to incorporate the heat sources (intrusions) more accurately in the models; include transient events where specific intrusions come into contact with the fluids in the deeper parts of geothermal systems Part of the Deep Roots Project objective was to attempt this using academic software (e.g. Hydrotherm, CSMP++) Axelsson Kissling & Bromley Cuernavaca Workshop April 2016
MODELLING ACHIEVEMNTS SO FAR THROUGH DRG Academic software has been applied to hypothetical models, to improve understanding of geothermal activity around magma intrusions. Neither Hydrotherm nor CSMP++ appear suitable for industrystyle modeling. New EOS (Equations of State) for itough2 extends applicability of the software to much higher p+t and greater depth (supercritical, >800 C & 100 MPa); this is extremely valuable Applications to actual volcanic systems in Iceland is under way Industry will greatly benefit from both increased under-standing and improved modeling tools at high p+t conditions Axelsson Kissling & Bromley Cuernavaca Workshop April 2016
Example of the outcome of the modelling part of DRG: Temperature distribution ( C) at 5,000 years after emplacement of an intrusion; calculated with a new supercritical EOS for TOUGH2/iTOUGH2 (Magnúsdóttir) Axelsson Kissling & Bromley Cuernavaca Workshop April 2016
PART B: Understanding the deep roots of geothermal systems in the Taupo Volcanic Zone (TVZ) of New Zealand Taupo Volcanic Zone young (~2Ma) rift above subducting plate High natural heat flux ~4200 MW (mean 700 mw/m 2 ) Extension 16 mm/yr at coast, decreasing to the south Partial melt beneath TVZ imaged with Magneto-tellurics (MT) This sub-section is based on material given at GEORG Deep Roots meeting Reykjavik, February 2016 by Warwick Kissling & colleagues, of GNS Science GNS Science
Taupo Volcanic Zone many high T, volcanictype, geothermal systems Subducting plate Crustal velocity structure GNS Science
Large-scale TVZ fluid-flow model (attempted in 2005) Impermeable base at 8 km Distributed heat sources on ~10 km scale no intrusions Simple geology Locates geothermal systems ~correctly Mass/heat flows approx. within a factor of ~2 Temperatures ~50 o C too low Kissling & Weir, JVGR 145, (2005) GNS Science
New Geophysical Results have changed our view of TVZ Deep Roots : a) Seismic tomography Active source seismic explosion data Subduction zone earthquakes at ~40-90 km depth propagate near-vertically; dense sampling of mid-crust. 40 sites deployed for 2 years ~ 5 km apart Derive : Vp, Vs, Vp/Vs, Qp, Qs Compare with resistivity/mt Conclusion: crustal V & Q inhomogeneous Bannister, Bourguignon, Sherburn & Bertrand, WGC, (2015) GNS Science
New Geophysical Results of TVZ Deep Roots: b) 3D MT inversion Are these MT inversions consistent with an under-plating or intrusion model of the TVZ heat source? Bertrand et al., JVGR (2015) GNS Science
Rotorua - Waimangu Relocated seismicity (mag >2 over 7 years) highlighted blind active fault, permeable fluid conduit to reservoir? Heise et al., JVGR Lake Rotomahana special volume Bertrand et al., GNS CR 2014/116 Caldwell et al., GNS CR 2014/93 Heise et al., GNS CR 2013/203 Hill et al., GNS CR 2013/160 Bannister et al., JVGR special volume GNS Science
Ohaaki geothermal field Bertrand et al., JVGR, (2015). Inclined deep upflow? Mroczek et al., Geothermics, (2016). Kissling and Bertrand (in prep) GNS Science
MT & Seismicity at Rotokawa Geothermal System Vertical deep upflow? Bertrand et al., JVGR, 2015. Compartment of seismicity: induced by fluid flow from injection to production Ductile, >400 o C at ~4 km depth? Sherburn et al., Geothermics, 2014 GNS Science
Ground Surface Deformation InSAR and GPS ~25 mm/yr of subsidence concentrated in the central TVZ Geothermal subsidence (up to 50 mm/yr at Wairakei ) Consistent with cooling and contraction of ~0.06 km 3 /yr rhyolitic magma sill at 6 km depth Implied heat loss ~ 5000 MW -15 0 15 (mm/yr) Hamling, Hreinsdottir and Fournier, JGR, 2015 GNS Science
Brittle-ductile transition with depth (from seismicity) Bibby et al., JVGR 68 (1995) 1995: transition at 8 km but now: Bannister et al., WGC (2015) ~3.5km beneath Rotokawa > ~10-30 km outside TVZ shallower in areas of high heat flux, deeper elsewhere (Kissling et al., 2009) GNS Science
Regional scale THM models TM code SULEC (Ellis et al., 2011) to explore under-plating, rifting, fluid flux from depth and magma intrusions TH code TGNS (Kissling, NZGW, 2014) to model convective fluid and heat transfer to the surface Detailed geology (to sub-fault scale) to characterise hydrological behaviour of different fault architectures Large scale structure beneath the TVZ rift. (after Stern et al., 2010) GNS Science
OTHER RELATED PROJECTS WORLDWIDE COTHERM, Switzerland (exploration and modelling) IMAGE, EU funded (exploration) DESCRAMBLE (Italy), EU funded (deep drilling and research) DEEPEGS, EU funded (deep stimulation and research) GEOWELL, EU funded (drilling technology) FUTUREVOLC (volcanology and hazards) Iceland deep drilling project (IDDP) Krafla Magma Drilling Project (KMDP) IPGT collaboration through modeling group (US, Aust., NZ, Swiss, Iceland) Japan : Super-critical (JBBP Beyond Brittle ) research New Zealand : HADES (hotter and deeper) research Axelsson Kissling & Bromley Cuernavaca Workshop April 2016
DESCRAMBLE (ITALY) Location & characteristics of deep drilling project (Venelle-2) into supercritical (~450 C & 250 b) 28 Present well depth is 2,2 km Location of Venelle_2 well, Larderello, Tuscany, Italy Target well depth is 3-3.5 km Bertani, 2015
CONCLUSIONS Immense energy resource in deep roots because of high temperature and pressure (>supercritical) Is there enough permeability near the roots for production? Or will utilization through deep reinjection (EGS methods) be more feasible (ductile to brittle transition thru cooling)? Exploration methods need to be upgraded to yield sufficient resolution/accuracy at depths involved Improved modelling of processes and large-scale regional models of neighbouring systems is needed Various on-going research projects worldwide are aimed at the deep roots (directly or indirectly) Collaboration between projects and countries has been beneficial and should be further encouraged Annex 12 of IEA Geothermal can certainly contribute JOIN US Axelsson Kissling & Bromley Cuernavaca Workshop April 2016
THANK YOU! Axelsson Kissling & Bromley Cuernavaca Workshop April 2016