Phot

o: A

. A. Ó

ladó

ttir

Renewability Assessment of the Reykjanes Geothermal System, SW-Iceland

Gudni Axelsson et al. (see next slide)Iceland GeoSurvey (ÍSOR)

2GGW 2016 Gudni Axelsson et al.

Contributors

Iceland GeoSurvey (ÍSOR): Gudni Axelsson, Egill Á. Gudnason, Ragna Karlsdóttir and Ingvar Th. Magnússon

Institute of Earth Sciences, University of Iceland: Sigrún Hreinsdóttir, Karolina L. Michalczewska and Freysteinn Sigmundsson

Vatnaskil Consulting Engineers: Andri Arnaldsson and Jean-Claude C. Berthet

GNS-Science, New Zealand: Chris J. Bromley and Sigrún Hreinsdóttir

HS-Orka: Ómar Sigurdsson

Financial support by the GEORG Research Fund in Iceland is acknowledged

3

Renewability of geothermal resources

Geothermal resources are generally classified as renewable This is an oversimplification, classification is too simple In essence of a double nature, i.e. a combination of:

a) energy current (through heat convection and conduction) andb) vast stored energy

Renewability of these aspects is quite different:a) energy current is steady and fully renewableb) stored energy is renewed relatively slowly by heat conduction

Relative importance of the two components depends on both the geological nature of a system and the rate of energy extraction during utilization

GGW 2016 Gudni Axelsson et al.

4

Project purpose

Main objective of project was to add significantly to the understanding of the nature of geothermal resources

Particular emphasis on their recharge and mass balance under production, i.e. to improve understanding of their renewability

Done through unifying analysis and modelling of data from different sources Emphasis on the Reykjanes geothermal system in SW-Iceland Purpose to evaluate the relative importance of the two renewability aspects

(energy current vs. stored energy) for the Reykjanes system, in particular, under the current state of utilization

GGW 2016 Gudni Axelsson et al.

5

Project background

Based on compilation of reservoir monitoring data, as well as collection and analysis of micro-gravity and geodetic data

Consequently the data were jointly interpretedi) by simple modelling andii) by simulating data by an up-to-date

numerical reservoir model of the system Also repeated TEM-resistivity surveying to

try to follow the growth of a steam-zone at the top of the geothermal system

GGW 2016 Gudni Axelsson et al.

Photo courtesy of HS-Orka

6

Project phases

The project aimed to join together the results of several different scientific methods/ disciplines to address the issue in question, in particular the following methods:

A) High-resolution 3-D surface deformation monitoring (InSAR and GPS monitoring)

B) Micro-gravity monitoring

C) Repeated TEM (transient electromagnetic) resistivity surveying

D) Reservoir pressure- and temperature monitoring

E) Chemical content monitoring

F) Dynamic geothermal reservoir modelling, to jointly interpret data

GGW 2016 Gudni Axelsson et al.

7

Reykjanes, Iceland

GGW 2016 Gudni Axelsson et al.

Photo: O. Sigurdsson

8

Reykjanes development

GGW 2016 Gudni Axelsson et al.

Characterized by SW-NE striking tectonic and volcanic activity as well as steam-vents, mud-pools and warm ground in an area of about 2 km2

Reservoir temperature 280 – 350°C The reservoir fluid is hydrothermally modified sea-water Development started as early as 1956 with shallow drilling Seven wells drilled during 1968 – 1970; deepest well 1750 m Followed by intermittent, small-scale industrial utilization; salt and sea-mineral

production along with fish drying Exploration and development picked up again in 1998 Included drilling of 14 deep production wells A 100 MWe capacity geothermal power plant commissioned in May 2006

9

Reykjanes production history

Average yearly mass production from the

Reykjanes geothermal system from 1970 up to

2013; the operation of the 100 MWe power-plant started in 2006, while significant reinjection

started in 2009

GGW 2016 Gudni Axelsson et al.

10

Reykjanes pressure decline

Pressure monitoring data from wells at Reykjanes, measured at a depth of

1500 m b.s.l.; most of the data-points are measured in

production wells during breaks in production while

some are measured in observation wells, e.g. RN-

16 at the margin of the main production field

GGW 2016 Gudni Axelsson et al.

11GGW 2016 Gudni Axelsson et al.

Reykjanes subsidence

Subsidence in Reykjanes (RNES) and Svartsengi (SVAR) estimated from GPS measurements spanning 1992 to 2014; GARD/GASK is shown here for reference

12GGW 2016 Gudni Axelsson et al.

Reykjanes subsidence (cont.)

Average subsidence rate from January 2009 to July

2013 in Reykjanes estimated from the

combination of sets of ascending and descending

TerraSAR-X InSAR interferograms

13GGW 2016 Gudni Axelsson et al.

Purpose of estimating the mass changes in the geothermal system during the period 2006–2010

Hence the renewal (recharge) of the fluid reserves in the geothermal system

Gravity surveys conducted during the summers of 2004 (prior to the start-up of the power plant), 2008 and 2010

Micro-gravity monitoring

14GGW 2016 Gudni Axelsson et al.

Gravity change modelling

The analysis involved three main steps:

1) An estimation of the mass changes in the geothermal system through a Gauss-integral of the observed gravity changes during two periods, 2004–2008 and 2008–2010; 30 – 50% during the latter period

2) A simulation of the gravity-change anomaly for 2008–2010 by two simple mass change models; center of mass change at 1300 – 1700 m depth

3) A calculation of gravity changes at the observation points of the gravity grid on basis of mass changes in the numerical model of the geothermal system – see next slides

See Gudnason et al. (WGC 2015)

15GGW 2016 Gudni Axelsson et al.

Numerical reservoir model

A TOUGH2 model Calibrated by various

reservoir data Gravity changes due to

mass changes in the model were calculated at the observation points of the gravity grid

Modelled anomaly comparable to observed one, not exactly however

16

Chemical content

GGW 2016 Gudni Axelsson et al.

No clear indications of major changes in chemical content of produced fluid (i.e. due to colder recharge) have been observed to date in Reykjanes

This result can be used to estimate the minimum volume of the Reykjanes reservoir On basis of the fluid volume extracted during the first 8 years of operation of the power

plant a volume of about 1.2 km3 is estimated (assuming a porosity of 10%) Considerably less than the minimum estimated volume of the system, which is of the

order of 3 km3

This result, along with the limited recharge, likely explains why no chemical changes have been observed so far

17

Main results

GGW 2016 Gudni Axelsson et al.

During 2008 – 2010 the renewal of reservoir fluid through recharge is estimated to have been of the order of 30 – 50%, or about 250 ± 60 kg/s on average; the renewal for 2006 – 2008 is expected to have been correspondingly less

Rough mass-balance estimates based on the limited fluid renewal in the geothermal system, during the current large-scale utilization, and the small size of the geothermal system, show that reservoir fluid content may be depleted in some decades; this identifies the need for substantial reinjection; associated research is ongoing

In spite of the limited size and recharge the energy in-place in the system is enormous; it is estimated that only a small fraction (2%) will have been extracted after 100 years under current extraction and recharge conditions

18

Conclusions/recommendations

GGW 2016 Gudni Axelsson et al.

Simple as well as finite-element modelling of observed deformation can further constrain the mass change in, and the renewability of, the Reykjanes system

Gravity change data should be used as a direct calibration parameter in numerical reservoir modelling, when possible

The ultimate goal is to set up one all-embracing model to simulate gravity change, deformation and chemical data, along with all reservoir data, in a fully coupled manner

Interpretation of the repeated TEM resistivity soundings indicates some shallow changes due to the growth of steam cap of the Reykjanes system, supporting the contention that resistivity methods may be a useful monitoring tool; in this case it didn’t yield quantitative results

19GGW 2016 Gudni Axelsson et al. Photo courtesy of HS-Orka

Thank you for listening