DISCUSSION BETWEEN A RESERVOIR ENGINEER AND A GEOLOGIST:

PERMEABILITY IDENTIFICATION FROM COMPLETION TEST DATA AND

BOREHOLE IMAGE LOGS INTEGRATION
Massiot, C.1, McLean, K.2, McNamara, D.D.3, Sepulveda, F.2, Milicich, S.D.1
1GNS Science, 1 Fairway Drive, Avalon, Lower Hutt 5010
2 Contact Energy Ltd.
3 University of Ireland, Galway, Ireland.
[email protected]

Keywords: Permeability, Completion test, Borehole image Traditional completion testing forms the basis of an
log, Wairakei Geothermal Field. interpretation of the permeable depth intervals (feed zones)
and their injectivity (contribution to the well flow) in the
ABSTRACT well. Pressure, temperature, and spinner (PTS) are measured
The location of permeable zones, and preliminary at various water injection rates and during heating-up after
quantification of a geothermal well’s production or injection injection stops. Drilling losses are also sometimes
capacity are routinely interpreted from well pressure, considered. The feed zones interpreted from completion test
temperature, and flow measurements made at different data can be subject to uncertainty, which can be due to: 1)
injection rates and during heating after shut-in (also referred presence of downflows which tend to mask any permeable
to as PTS or completion test data). The spatial resolution of zones at depths below the inflow point (due to constant
feed zones interpreted from completion test data typically temperature); 2) balanced well and formation pressures
ranges from 10 – 100 m thick. By itself, the completion test preventing fluid moving from the well into the formation
dataset cannot inform on the nature of the permeability, i.e. (outflow) or from the formation into the well (inflow); 3)
fracture versus formation permeability, and is subject to large permeable zones above minor permeable zones, in
uncertainties. On the other hand, borehole image logs which case, PTS data may not have enough resolution to
provide a direct description of the fractures (location, detect minor feed zones. Pressure transient analysis is a
morphology and orientation) and of rock textures intersected valuable tool sometimes used during analysis of completion
by the borehole, to a spatial resolution of ~0.01 – 1 m. test data, a discussion of which is beyond the scope of this
paper.
Traditionally, analyses of completion test and borehole
image log data are done independently by the reservoir
engineer and geologist, respectively. We present a joint
interpretation of completion test data and borehole image
logs (acoustic and resistivity) from two boreholes in the
Wairakei Geothermal Field, New Zealand, highlighting the
advantages of a cross-discipline interpretation of a
borehole’s permeable zones. In the two studied boreholes,
we show that feed zones generally correlate with (1)
fractures of low acoustic amplitude and visible on the travel-
time image of the acoustic image log, and (2) low resistivity
fractures with a high resistivity halo on the resistivity image
log. In the latter case, fractures with haloes coincide with a
reservoir-scale fault modelled independently from 3D
modeling of stratigraphic offsets. However, not all fractures
of such appearance observed on the image logs correlate with
permeable zones. The integrated interpretation of completion
test data and image log increases the robustness in the feed
zone interpretation (location, extent, fracture or matrix
permeability components), and thus can contribute to
improving fluid flow numerical models used to sustainably
manage geothermal resources. Figure 1: Location map of wells used in this study,
nearby wells active faults mapped on surface and the
1. INTRODUCTION possible Kaiapo Fault intersection with WK271.
Permeability in volcaniclastic formations is typically
controlled by a mixture of fractures and matrix (Grant and Borehole image logs directly image formation textures and
Bixley, 2011), with the fracture component generally fractures intersecting the borehole, and allow fracture
increasing with depth. Characterising the contribution of location, orientation and apparent thickness measurements.
both fracture and matrix to permeability is important for field There are two types of borehole image logs: acoustic and
management, and for better understanding of the reservoir resistivity. Interpretation of acoustic image logs acquired
for future well siting. This knowledge can guide, for with the high-temperature acoustic televiewer ABI85 in >20
instance, the reservoir modelling decision between a single- boreholes in the Taupo Volcanic Zone (TVZ) since its
porosity (matrix-dominated) and double porosity (fracture- development in 2009, has provided invaluable observations
dominated) reservoir model. of the structural and in-situ stress settings directly in the
reservoirs (Massiot et al., 2013; McNamara et al., 2016 and
Proceedings 39th New Zealand Geothermal Workshop
22 – 24 November 2017
Rotorua, New Zealand
references therein). When the borehole can be cooled down WK271) located in the Te Mihi production zone of the
<150°C, resistivity image logs (FMI) can be acquired, which Wairakei Geothermal Field, TVZ (Figure 1) for a more
has been reported only once before in the TVZ (Halwa et al., robust interpretation of location, extent, and fracture
2013). contribution of permeable zones. We also investigate of the
characteristics of permeable fractures in borehole image
Fractures identified in borehole image logs are typically logs.
0.01-1 m thick (Massiot et al., 2017), much higher resolution
than completion test data which typically ranges from 10- 2. DATA
100m thick. Some features are known to be impermeable, Boreholes WK261 and WK271 intersected a volcaniclastic
such as high acoustic amplitude and high resistivity fractures sequence of ignimbrites, tuffs and rhyolites
which are infilled with minerals (Davatzes and Hickman, (Figures 2 and 4). The >900-m long image logs span the
2010; Milloy et al., 2015). However, it is not possible to entire open-hole section. Borehole image logs were
determine if low acoustic amplitude and low resistivity processed, quality-checked, and interpreted following the
fractures are infilled with minerals or open to fluid flow. It is workflow presented in Massiot et al. (2015). All reported
thus necessary to analyse borehole image logs in depths are with respect to the casing head flange, measured
combination with completion test data to assess fracture along the boreholes.
permeability.
WK261 was drilled vertically to a depth of 2059 m, cased to
McLean and McNamara (2011) considered three boreholes 1144 m. Completion testing (Figures 2 and 3a) with the PTS
at the Wairakei Geothermal Field for an integrated tool was carried out at three increasing injection rates (57,
interpretation of feed zones from completion test data and 114 and 170 t/h) followed by a falloff to zero flow, with
borehole image logs, illustrating the value of cross- various profiles at each injection rate. The fixed tool depth
disciplinary geological and reservoir engineering analysis. In for flow changes was 2052 m.
this paper, we present a new case study with advanced
integrated interpretation of feed zones, image logs and
subsurface modelled faults from two boreholes (WK261 and

Figure 2: WK261: Completion test and acoustic image log interpretation. (a) Fluid velocity (spinner) at three rates of
injection, and spinner ratio (dashed lines: interpreted areas of constant spinner ratio); (b) temperature logs at three
injection rates and during heat-up, with the pressure control point (PCP); (c) temperature gradient logs; (d) preliminary
feed zone intervals; (e) fracture thickness and morphology, and cumulative fracture thickness of all fractures, fractures of
low acoustic amplitude seen on travel-time images and the 25% thickest fractures, plotted on a log-linear scale. The final
permeable zone intervals are dashed in grey across all panels.

Proceedings 39th New Zealand Geothermal Workshop

22 – 24 November 2017
Rotorua, New Zealand
Figure 3: WK271: Completion test and resistivity image log interpretation. Same nomenclature as Figure 1. The intersection
between the Kaiapo Fault and the borehole is not precisely constrained.
WK271 was drilled in 2013 to a depth of 2268 m, cased to
1196 m and is deviated ~20° towards NNW (Figure 1). a) 0
Completion testing (Figure 3 and 4b) with the PTS tool was
carried out using three increasing injection rates (63, 114 and Inj rate Inj rate Inj rate No inj
500 57 t/h 114 t/h 170 t/h 0 t/h
159 t/h) followed by a falloff to zero flow, with various
Tool depth [mCHF]

profiles at each injection rate. The fixed tool depth for flow 1000
changes was 1120 m, within the casing shoe due to hole
stability concerns. 1500

It is of general interest to determine the nature of

2000
permeability for permeable wells. The injectivity index (or
injectivity, for short) of a well provides a quantitative 2500
measure of a well’s permeability (e.g. Siega et al., 2014) . 0 1 2 3 4 5 6 7 8 9
Injectivity is routinely obtained from three stable pressures Elapsed time [hr]
measured at a fixed depth (generally, the depth of the main
feed in the well) for three injection rates. WK261 and b) 0
WK271 have injectivities of 121 and 91 t/h/bar, respectively;
Inj rate Inj rate Inj rate No inj
Figure 5). In the case of WK271 there is a significant 500 63 t/h 114 t/h 159 t/h 0 t/h
Total depth [mCHF]

downflow present at all injection rates, which has been

accounted for in the injectivity assessment via an enthalpy 1000
balance calculation:
1500
𝑚̇ 3ℎ3 = 𝑚̇ 1ℎ1 + 𝑚̇ 2ℎ2
2000
where: ṁ = mass flow (kg/s), h = enthalpy (kJ/kg) and the
numbering refers to locations: 1 = at the wellhead, 2 = the 2500
inflow, 3 = total flow below the inflow (injection + inflow 0 1 2 3 4 5 6 7 8
combined). Elapsed time [hr]

Figure 4: Depth-time charts for PTS tool during completion

testing, showing flow rate changes, and profiles run at
different speeds, for: a) WK261; b) WK271.

Proceedings 39th New Zealand Geothermal Workshop

22 – 24 November 2017
Rotorua, New Zealand
a)
180
 Fractures of high resistivity, which are likely filled
with non-conductive minerals such as quartz or
Flow rate [t/h]
160
140
120 calcite and cannot conduct fluid flow
100
80 y = 121.41x - 17383
60  Fractures of low resistivity, which may be either
40 open to fluid flow or at least partially filled with
143.4 143.6 143.8 144 144.2 144.4 144.6 144.8
minerals such as clay (Halwa et al., 2013)
Pressure @ 2052 mCHF [bara]

 Fractures of low resistivity displaying a halo of

b) high resistivity. This high resistivity rim can be
220
200
Flow rate [t/h]

180 caused by the build-up of excess current along the

160
140
120
fracture due to materials with unusually high or
100
80 y = 90.609x - 10000 low conductivity within the fracture (Lofts and
60 Bourke, 1999).
40
111.2 111.4 111.6 111.8 112 112.2 112.4 112.6 112.8
Pressure @ 1691 mCHF [bara] 3.3 Discussion between the reservoir engineer and the
geologist
Figure 5: Injectivity charts: a) WK261; b) WK271 The step where both analysts discuss their findings is critical
to a robust analysis of the location and extent of permeable
zones. The reservoir engineer has a good understanding of
the uncertainty of their interpretation, and can look for an
3. METHOD observable candidate on the image log justifying the
The integrated analysis starts with independent preliminary existence of a feed zone. On the other hand, the geologist can
analysis of completion tests and image logs. Then, both propose what constitutes a good contender (either fractures
specialists work together and discuss respective or lithological features) to be permeable, and the reservoir
interpretations, using them as input into a revised engineer can in turn investigate if there are correlating
interpretation. permeability indicators in the completion test data. Given the
uncertainty and sampling rate of completion test data, it is
3.1 Completion test data analysis sometimes adequate to extend a permeable zone by a few
The preliminary completion test data analysis involves metres to include a major feature visible on the image logs
checking the available data for anomalies which could be due that is a good candidate to be permeable.
to feed zones. Data types used in this study are: 1) changes
in temperature gradient during injection; 2) changes in 4. INTEGRATED INTERPRETATION
temperature gradient during heatup; 3) fluid velocity profiles 4.1 Preliminary completion test data analysis
from spinner data; 4) ratio of fluid velocity profiles from For WK261, there are 10 interpreted feed zones in the
different injection rates; 5) pressure control point (PCP) preliminary analysis (Table 1, Figure 2). The major feed
during heatup. zone is #7 which exhibits anomalies for all data types and
contains the pressure control point. Drilling losses were
3.2 Image log analysis excluded from the analysis as multiple (14) lost circulation
The image log interpretation is made following the workflow events in the last 300m of the borehole were all brief, do not
detailed in Massiot et al. (2015). Care was taken to correlate with other data and are more likely to be due to
differentiate natural fractures from stress-induced features insufficient air during this period of aerated drilling.
and layering.
For WK271 the interpretation of the completion test data is
Fractures identified on the acoustic image log (WK261) are not as straightforward as for WK261 due to the presence of
separated and plotted into three categories based on their a large downflow entering the well around 1300 m resulting
appearance on the image log (Figure 2): in a long isothermal section down to around 2050 m. This
isothermal section is present at all injection rates and masks
 Fractures of high acoustic amplitude, which are the temperature response which might be expected from any
mineralised and cannot conduct fluid flow feed zones within this section.

 Fractures of low acoustic amplitude, which may be For WK271 there are 7 interpreted feed zones in the
either open to fluid flow or at least partially filled preliminary analysis (Table 2, Figure 3). It is not possible to
with minerals such as clay, iron oxides, and pyrite identify the major feed zone: while the PCP coincides with
(Milloy et al., 2015) FZ #4 (spinner anomalies but not temperature due to
masking), this may not be the major zone as the PCP is also
 Fractures of low acoustic amplitude which are seen approximately at the midpoint between the source of the
on the travel-time image. While these fractures downflow (FZ #1-3) and the end of the isothermal section
may also be filled with clay minerals, they have the (FZ #6).
highest probability of being open to fluid flow as a
topography is observed on the borehole wall

Fractures identified on the resistivity image log (WK271) are

separated and plotted into three categories based on their
appearance on the image log (Figure 3):

Proceedings 39th New Zealand Geothermal Workshop

22 – 24 November 2017
Rotorua, New Zealand
 Table 1: WK261 preliminary feed zone interpretation * Y1 (yellow) = yes, certain; Y2 (brown) = yes, uncertain; N
(white) = no.
Data types – presence of anomalies
Feed Depth Temp. Temp. 4.2 Borehole image log analysis
Spin
zone range Grad. Grad. Spin PCP In WK261, 699 fractures were identified, and are dominantly
ratio
ID m Inj. Heat. of low acoustic amplitude (Table 3), of which 97 are also
Y1 Y1 observed on the travel-time image. In WK271, significantly
1150-
1 1150- 1150- N N N more (2299) fractures were identified on the resistivity image
1180
1180 1180 log, and are dominantly of low resistivity. Eighty-one of
Y1 Y1 these are of low resistivity fractures and display a high
1280-
2 1280- 1280- N N N resistivity halo. Excluding fractures of high acoustic
1310
1310 1310 amplitude in WK261, and those of high resistivity in
Y1 Y1 Y2 WK271, which are most likely infilled with minerals, yields
1390-
3 1390- 1390- N 1400- N 698 potentially permeable fractures in borehole WK261, and
1420
1410 1410 1420 2143 in borehole WK271.
Y1 Y1 Y1
1450- In WK261, fracture thickness measured perpendicular to the
4 1450- 1450- N 1470- N
1490 fracture walls varies is <895 mm, with 75% of fractures
1470 1470 1490
Y1 Y1 <12.4 mm. In WK271, fracture thickness is <169 mm, with
1560- 75% of fractures <6.2 mm. In both cases, the thickest
5 1560- 1560- N N N
1630 fractures likely form a fractured zone rather than individual
1630 1630
Y1 Y1 structures. The ability of resistivity image logs to detect
1670- thinner fractures than acoustic image logs relates to the
6 1670- 1670- N N N
1690 different types of signals acquired, resistivity image logs
1690 1690
Y2 Y2 Y2 Y1 Y1 being more susceptible to fine variations of mineral
7 1810- properties, while acoustic image logs are more susceptible to
1810- 1810- 1800- 1820- 1850-
major 1880 the borehole roughness as caused by large open fractures
1850 1850 1850 1840 1880
Y1 (Davatzes and Hickman, 2011).
1890-
8 N 1890- N N N
1910 Table 3: Length of imaged interval and number of
1910
fracture interpreted on each image log. Fracture
Y2 morphology: HA: high acoustic amplitude; LA:
1980-
9 N N 1980- N N low acoustic amplitude; LAT: low acoustic
2010
2010 amplitude and visible on the travel-time image;
Y1 Y1 HR: high-resistivity; LR: low resistivity; H: low
2050-
10 2040- 2040- n/a n/a N resistivity with high resistivity halo.
2059
2059 2059
* Y1 (yellow) = yes, certain; Y2 (brown) = yes, uncertain; N Length of
(white) = no. imaged Number of fractures on image
Borehole
interval logs
Table 2: WK271 preliminary feed zone interpretation (m)

Data types – presence of anomalies WK261

699
Feed Depth Temp. Temp. 902
Spin (acoustic)
zone range Grad. Grad. Spin PCP (HA: 1; LA:601; LAT: 97)
ratio
ID mCHF Inj. Heat.
Y1 Y1 WK271 2299
1230- 955
1 1230- 1230- N N N (resistivity)
1255 (HR: 156; LR: 2062; H: 81)
1255 1255
Y1 Y1
1265-
2 1265- 1265- N N
1290 Y1
1290 1290 4.3 Integrated analysis of permeable zones in WK261
1260-
Y1 Y2 with acoustic image log interpretation
1305- 1350
3 1305- 1305- N N The cumulative fracture thickness displays large steps
1330
1330 1330 coinciding with nearly all permeable zones initially
Y2 Y1 Y2 interpreted from completion tests, despite the datasets being
1610-
4 N N 1680- 1650- 1610- independent (Figure 2d, in log-linear scale). There is a large
1730
1730 1720 1670 contribution of these steps from the thickest fractures (as
Y1 displayed by the cumulative fracture thickness for the 25%
1910-
5 1910- N thickest fractures), rather than by numerous closely-spaced
1980 Y2
1980 thin fractures. These thickest fractures are dominantly of low
1910-
Y1 Y1 Y1 acoustic amplitude and seen on the travel time image. The
2030- 2050
6 1910- 1910- 2030- N location and extent of each feed zone was thus confirmed by
2090
2180 2180 2090 the combination of both datasets. Although it is not possible
Y2 Y2 to rule out a partial contribution to permeability by matrix,
2130-
7 2130- 2130- N all feed zones have some fracture permeability.
2150
2150 2150

Proceedings 39th New Zealand Geothermal Workshop

22 – 24 November 2017
Rotorua, New Zealand
The major feed zone (FZ#7), initially 70-m thick can be Based on the apparent correlation between fracture
directly linked to two specific, thick fractures seen on the type/thickness and permeability indicators in acoustic image
travel-time image, and other thinner fractures of the same logs, an additional step not undertaken in this study would
morphology. Similarly, the sharp temperature increase consist of extracting very minor zones of permeability, such
observed on temperature logs during injection and heat-up in as in FZ#5 in WK261. In this FZ, individual thick fractures
FZ#4 can be related to a single thick fracture seen on the can be associated with specific temperature gradient
travel time image. changes, which qualifies them as permeable. While such
detailed analysis of individual permeable fractures is not
Feed zones which do not display any significant thick critical for reservoir management purposes, it is the base for
fractures visible on travel-time image are minor (feed zones fundamental research studies aiming at understanding the
1 and 8). Permeability may be dominantly controlled by the tectonic settings and interactions with fluid circulations in a
matrix (tuff lithology) rather than by fractures. given stress state (Barton et al., 1995). In turn, these
structural and geomechanical studies help to site wells which
Interpretation of image logs provide an added, independent will have an increased success of finding high permeability.
line of evidence to confirm the presence and extent of
permeable zones controlled by fractures. Due to the high 5.2 What does a permeable fracture look like?
quality of the completion test data and therefore the Figure 6a displays fractures located in the major FZ# 7 in
preliminary interpretation, the feed zones have in this case
borehole WK261. Two fractures, respectively 0.5 and 0.9m-
not changed much by integrating with the image log
thick, appear clearly on the travel-time image which
interpretation data. However, the feed zone interpretation is indicates that the borehole wall is rough, likely caused by
more robust. spalling from the borehole wall at the intersection between
these open fractures and the borehole wall.
4.4 Integrated analysis of permeable zones in WK271
with resistivity image log interpretation Similar features observed in WK261 (Figures 6b-c) are not
In WK271, it is more difficult to interpret the completion test associated with either a temperature gradient change, or a
data due to the presence of a large downflow, masking any clear spinner drop. At 1910-1920 m ((Figures 6b), the
potential temperature anomalies. This shortcoming in the spalling from the borehole wall is so intense that it was
completion test data creates an ideal circumstance for the difficult to identify individual fractures, and this zone was
fracture dataset to make a significant contribution to the suggested by the geologist as a good contender for fracture
refinement of the location of feed zones. permeability. This confirms that acoustic image logs alone
cannot identify permeable fractures, but rather identifies
As discussed in Section 4.1 the major feed zone is not clear potential permeable fractures which can be validated (or not)
from completion test data alone, however with fracture data by completion tests.
it can be confidently interpreted as FZ#4. There is a good
correlation between the occurrence of fractures of low
resistivity displaying a high resistivity halo and feed zones
delineated independently from completion tests. FZ#4
contains 43% of all fractures with halo identified in the
borehols, though this interval covers only 13% of the total
imaged borehole length. The 3D geological model, based on
stratigraphic offsets and detailed LiDAR surface fault
mapping (Villamor et al., 2015), indicates that the Kaiapo
Fault intersects WK271 near this feed zone (precise depth
unconstrained due to the similar borehole trend and fault dip
direction). Fractures with haloes may thus be related to the
damage zone of the Kaiapo fault. The other fractures with
haloes are dominantly located in feed zones 5 and 6. The
correlation between thick fractures and feed zones identified
in WK271 resistivity image log is not as clear as in WK261
acoustic image log.

FZ#1 and FZ#6 occur at the interface between lithologies,

which suggests a lithological control on permeability. The
image log does not extend to the bottom of the borehole, so
the source of permeabiity in FZ#7 is not clear.

5. DISCUSSION
5.1 Efficiency of the process
Using this integrated process, the discussion of the extent and Figure 6: WK261 acoustic image log: permeable and non-
fracture/matrix components of permeable zones takes a permeable fractures of low acoustic fractures and
relatively short time compared to the prior interpretation of visible on the travel-time image (arrows). a) major
completion tests and image logs. Within this short time, FZ#7; b) spalling from the borehole wall in a non-
discussions add a greater degree of certainty to subsequent permeable zone; c) non-permeable zone.
interpretations, including the extent/location of feed zones, Statically normalized images with from left to
identification of the major feed zone(s) in a well, with a right: amplitude image, travel-time image and
compromised completion test dataset and indications of the travel-time with fracture interpretation.
fracture/matrix contribution.

Proceedings 39th New Zealand Geothermal Workshop

22 – 24 November 2017
Rotorua, New Zealand
The contribution to permeability of the numerous fractures  Increased fracturing at the base of the partially
<30 mm-thick located in feed zones (such as in Figure 6a) is welded ignimbrite potentially caused by a local
unclear. In addition, 66% of fractures detected on the increase in welding indiscernible from cuttings
WK261 image log are not located in any feed zone. These (Smyth and Sharp, 2006)
fractures may be infilled with low acoustic amplitude
material such as clay, or lack sufficient connectivity to the Fractures with haloes are also observed near this interface
permeable fracture system (Kissling and Massiot, this (Figure 3) which could be caused by increased water-rock
volume). While not critical to the reservoir management, interactions. Further textural analysis of the resistivity image
non-permeable fractures are very useful to evaluate the log is ongoing to evaluate lithological controls on fracturing
tectonic settings in relationships to the in-situ stress and permeability.
(McNamara et al., 2015). These fractures which were not
permeable at the time of the completion test may also be
reactivated by subsequent natural seismicity or borehole
stimulation, and thus may provide additional permeability at
a later stage.

Figure 7a presents thick low-resistivity fractures in borehole

WK271, some with high resistivity haloes, located within
major FZ#4. Similar features are observed in zones without
any permeability indicators (Figure 7b). In this case,
hydrothermal mineralisation may have filled-in the fracture,
preventing fluid flow. Current investigation of the slip
tendency of these fractures within the in-situ stress field are
ongoing to investigate the possible nature of these fractures.

Figure 8: WK271, lithological control on permeability:

FZ#6 (grey square), at the interface (dashed line)
between a partially welded ignimbrite and a
rhyolite (as defined from cuttings), coincides with
a pervasive low-resistivity zone (dark grey).
Statically normalized resistivity image.
5.4 Strategy to identify potentially permeable fractures in
acoustic and resistivity image logs
Permeable fractures in borehole image logs in the two
boreholes used in this study have different characteristics. In
acoustic image logs, thick (>15-20 mm) fractures of low
acoustic amplitude and visible on the travel-time image are
good contenders to be permeable. Similar observations have
been made in nearby boreholes WK264 and WK266
(unpublished), as well as previously reported by McLean and
McNamara (2011) in three other boreholes in the Wairakei
Geothermal Field. So, for the Wairakei Geothermal Field,
rapid interpretation of potential fracture-dominated
permeable zones can be attempted using acoustic image logs,
and then integrated with a preliminary assessment of
Figure 7: Dynamically normalized WK271 resistivity permeable zones from completion test data. Further in-depth
image log of fractures with haloes in a) a analysis of both datasets can then be made at a later stage.
permeable zone and b) a non-permeable zone.
In a similar study of the andesitic formations hosting the
5.3 Evidence of matrix-controlled permeability
Rotokawa reservoir, a correlation between such thick low
FZ#6 in WK271 is located at the interface between a partially acoustic fractures seen on the travel-time image and
welded ignimbrite and a rhyolite lava as defined from permeable zones was demonstrated for some of the
cuttings observations. The resistivity image log displays a permeable zones (McNamara et al., 2015), but not as
zone of low resistivity (Figure 8). In this case, it is likely that systematically as at Wairakei. So, morphological
permeability occurs at the interface itself, associated with characteristics of good contenders for being permeable needs
one or several of these contributions: to be tested for each geothermal field, or at least each
 The interface itself serving as a fluid pathway reservoir rock type.
 The brecciated carapace of the rhyolite
Based on the first resistivity image log acquired in the
Wairakei Geothermal Field (WK271), low resistivity

Proceedings 39th New Zealand Geothermal Workshop

22 – 24 November 2017
Rotorua, New Zealand
fractures with high resistivity haloes seem to correlate with Halwa, L., Wallis, I. C., & Lozada, G. T. (2013). Geological
the major feed zone (FZ#4) also coinciding with a modelled analysis of the volcanic subsurface using borehole
fault zone, and with minor feed zones #5 and #6, making resistivity images in the Ngatamariki Geothermal
them good contenders for rapid identification of potential Field, New Zealand. New Zealand Geothermal
permeable fractures. Halwa et al. (2013) interpreted similar Workshop 2013 Proceedings.
features observed in the nearby Ngatamariki Geothermal
Field as an indicator of fractures filled with highly Kissiling, W. and Massiot, C. (submitted): Geometrical
conductive minerals, such as pyrite or magnetite. In borehole modelling of fracture networks in an andesite-hosted
WK271, the high conductivity of these permeable fractures geothermal reservoir. This volume.
with haloes may be related to the presence of fluids in the
fracture, or to actual changes of mineralogy near the fracture Lofts, J.C., and Bourke L.T. (1999) The recognition of
walls caused by intense water/rock interactions. Further artefacts from acoustic and resistivity borehole
resistivity log acquisition will undoubtedly help with imaging devices. Geological Society of London,
defining which fracture morphology is likely to be Special Publication Vol 159:59-76.
permeable. Areas of pervasive low resistivity as presented on
Figure 7 may also provide guidance for matrix-controlled Massiot, C., McNamara, D. D., & Lewis, B. (2013).
permeability. Interpretive review of the acoustic borehole image logs
acquired to date in the Wairakei-Tauhara Geothermal
6. CONCLUSION Field. GNS Science report 2013/04. 26 p.
For the Wairakei Geothermal Field, thick fractures with low Massiot, C., McNamara, D. D., & Lewis, B. (2015).
acoustic amplitude also visible on the travel time on acoustic Processing and analysis of high temperature
image logs are good contenders for being permeable. Based geothermal acoustic borehole image logs in the Taupo
on the first resistivity image log acquired at Wairakei, low Volcanic Zone, New Zealand. Geothermics, 53, 190–
resistivity fractures with high resistivity haloes are good 201. Doi: 10.1016/j.geothermics.2014.05.010
contenders for being permeable, and in this case occur
predominantly in the vicinity of a reservoir-scale fault. To Massiot, C.; Nicol, A.; McNamara, D.D.; Townend, J.
conclusively identify whether these fractures are actually (2017). Evidence for tectonic, lithologic and thermal
permeable requires independent analysis of completion test controls on fracture system geometries in an andesitic
data, where both datasets inform each other. Discussions high-temperature geothermal field. Journal of
between the reservoir engineer and geologist enhances the Geophysical Research (122).
value of these datasets, providing a significant dimension of
robustness to interpretations and work within the two McLean, K., & McNamara, D. (2011).; Fractures interpreted
disciplines, and ultimately improved reservoir models and from acoustic formation imaging technology:
field management. The individualisation of permeable correlation to permeability, Proceedings 36th
fractures in image logs is also crucial to further understand Workshop on Geothermal Reservoir Engineering,
tectonic and water-rock interaction processes in high- Stanford.
temperature geothermal systems.
McNamara, D. D., & Massiot, C. (2016). Geothermal
ACKNOWLEDGEMENTS Structural Geology in New Zealand: innovative
This study is part of a contribution from GNS Science’s characterisation and micro-analytical techniques.
Geothermal Resources of New Zealand and New Zealand Proceedings 38th New Zealand Geothermal
Geothermal Future research programmes, funding of which Workshop.
was provided by the Government of New Zealand. We thank
Contact Energy Ltd for the provision and permission to Milloy, S., Mclean, K., & Mcnamara, D. D. (2015).
publish well data. The authors acknowledge support of this Comparing Borehole Televiewer Logs with
work by Haliburton Software and Services, a Haliburton Continuous Core: An Example from New Zealand.
Company, through the use of RecallTM Borehole software. World Geothermal Congress Proceedings.

REFERENCES Siega, C., Grant, M., Bixley, P., & Mannington, W. (2014).
Quantifying the Effect of Temperature on Well
Barton, C. A., Zoback, M. D., & Moos, D. (1995). Fluid flow Injectivity. Proceedings of the 36th New Zealand
along potentially active faults in crystalline rock. Geothermal Workshop, 24-26 November, Auckland,
Geology, 23(8), 683–686. New Zealand.
Davatzes, N. C., and S. H. Hickman (2010), Stress, fracture, Smyth, R. C., & Sharp, J. M. (2006). The hydrology of tuffs.
and fluid-flow analysis using acoustic and electrical In Tuffs - Their Properties, Uses, Hydrology, and
image logs in hot fractured granites of the Coso Resources (pp. 91–111). Geological Society of
Geothermal Field, California, U.S.A., in Dipmeter and America. doi: 10.1130/SPE408
borehole image log technology: AAPG Memoir 92,
edited by M. Poppelreiter, C. Garcia-Carballido, and Villamor, P., Clark, K., Watson, M., Rosenberg, M. D.,
M. Kraaijveld, pp.259–293, AAPG Special Volumes, Lukovic, B., Ries, W., & González, Á. (2015). New
doi:10.1306/13181288M923134. Zealand Geothermal Power Plants as Critical
Facilities: an Active Fault Avoidance Study in the
Grant, M. A., & Bixley, P. F. (2011). Geothermal Reservoir Wairakei Geothermal Field, New Zealand. World
Engineering (2nd ed.). Burlington, USA: Academic Geothermal Congress 2015, 19–25.
Press.

Proceedings 39th New Zealand Geothermal Workshop

22 – 24 November 2017
Rotorua, New Zealand