posiva 2002 22 working report web

Upload: hazel-fransiskus-sitanggang

Post on 06-Jul-2018

230 views

Category:

Documents


0 download

TRANSCRIPT

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    1/107

    W o r k i n g R e p o r t 2 0 0 2 2 2

    Fracture

    Maooing from Olkiluoto

    Borehole

    Image

    Data

    2001

    Eve l i ina

    T a m m i s t o

    T o m a s

    L eh t im ak i

    J o r m a

    P a l m e n

    Pi r jo He l la

    Eero H e i k k i n e n

    i n t ac t

    y

    ay

    2 0 0 3

    Working Reports contain information

    on work in

    progress

    or

    pending

    completion.

    The

    conclusions and viewpoints presented in the report

    are

    those

    of

    author s)

    and

    do not necessarily

    coincide wi th those of Posiva.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    2/107

    TEKIJAORGANISAA TIO:

    TILAAJA:

    Fintact y

    Hopeatie 1 B

    00440 Helsinki

    Posiva Oy

    Toolonkatu 4

    00100 Helsinki

    TILAUSNUMEROT:

    Fin tact Oy: 9646/0 1/HH

    TILAAJAN YHDYSHENKILO:

    Heikki Hinkkanen

    Posiva Oy

    TEKIJAORGANISAATION YHDYSHENKILO:

    TEKIJAT:

    TYORAPORTTI 2002-22

    FRACTURE MAPPING FROM OLKILUOTO BOREHOLE

    IMAGE DATA 2001

    Eveliina Tammisto Tomas Lehtimaki Jorma Palmen Pirjo Hella Eero

    Heikkinen

    TARKASTAJA:

    Henry Ahokas

    Fintact Oy

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    3/107

    Tammisto, E., Lehtimaki, T., Palmen, J., Hella, P. Heikkinen, E. 2002. Fracture

    mapping from Olkiluoto borehole image data, 2001. Working Report 2002-22. Posiva

    Oy, Helsinki. 109 p.

    BSTR CT

    As a part of the complementary site investigations for the underground disposal of spent

    nuclear fuel at Olkiluoto site, fracture mapping from optical imaging data was done in

    boreholes KR1, KR2, KR4, KR9, KR12, KR13 and KR14. The data contained BIP

    imaging, carried out in 1996, from sparsely fractured sections in KR1, KR2 and KR4

    that have not been included into previous interpretation (745 m), and OPTV imaging

    results from KR9, KR12, KR13 and KR14 in 2001 (2306 m).

    The task included fracture mapping from the image material and determination of

    orientation of the fractures using the Well CAD utility, determination of properties of the

    fractures, compilation

    of

    print-outs, and data delivery. The following fracture properties

    were mapped: fracture type, aperture and filling thickness, fracture shape and colour.

    The results of this independent mapping task provide supplementary information, i.e.

    complete orientation coverage, apertures, etc. not found from core mapping data. The

    observations were checked with comparing them to the core samples.

    n total6138 fractures were mapped. These represent roughly 80 of the core mapping

    fractures. Not every fracture reported from core can be observed from the images, but

    on the other hand, specifically from frequently fractured and core loss sections also few

    possible previously non-reported fractures have been encountered.

    The open or partly open fractures, some of which might have also filling, make on

    average 5

    of

    all fractures.

    In

    less fractured

    KR1

    and KR2 these are rare, and in KR4

    and KR13 more frequent (9 ) than for the others. Fractures with filling form two-third

    of all fractures, and closed ones one-third. Aperture and thickness both range from 0.5

    mm to 120 mm. Thickness has been reported for most

    of

    the fractures.

    More than a h lf of the fractures are planar and smooth in their shape (55 ). Slightly

    less (one-fifth) are found both planar and rough, and undulating and smooth. The

    fracture colours are determined mostly as a grey scale (60-80 . Black fillings are seen

    less than others probably due to dark background, grey and white are more frequent. For

    KR2 and KR13 the colours brown, green and yellow are found relatively often.

    Seepage has been found from 0 - 49 fractures per borehole, most frequently in KR12.

    For almost

    h lf

    of these seepage locations, a hydraulic conductivity indication has been

    found with good depth agreement. Based on seepage, these are potential inflow

    locations in open borehole conditions.

    Keywords

    Borehole optical imaging, fracture interpretation, nuclear waste disposal

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    4/107

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    5/107

      ONTENTS

    BSTR CT

    TIIVISTELM

    CONTENTS

    PREF CE

    INTRODUCTION 3

    2 IMAGE MATERIAL 7

    2 1 Image logging arrangements 7

    2 2 The image properties 10

    2 3 The image quality due

    to

    borehole conditions

    14

    2 4 The preconditioning of the image material

    23

    3 MAPPING THE FRACTURES 25

    3 1

    Preparation of the material

    25

    3 2

    Picking

    of

    the features 25

    3 3 Quality criteria of the picking

    26

    4 DETERMINATION OF THE PROPERTIES

    31

    4 1

    Design of property parametres

    31

    4 2 The distribution of mapping results according to level of confidence

    71

    5 CHECKING THE OBSERVATIONS WITH THE CORE SAMPLES 73

    5 1 The checking procedures 73

    5 2

    The results

    of

    the checking

    73

    6 RESULTS 75

    6 1

    Contents

    of

    the results

    75

    6 2 Fracture type distributions 75

    6 3 Fracture aperture

    and

    thickness 80

    6 4 Fracture forms

    80

    6 5 Fracture filling colours

    81

    6 6 Inflow indications 83

    6 7 Presentation of results

    and

    data delivery

    88

    6 8 Comments on results 89

    7 SUMMARY

    91

    REFERENCES 93

    APPENDICES

    97

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    6/107

    2

    PREF CE

    This report describes fracture interpretation of borehole imaging data from Olkiluoto in

    2 1 (supplementary mapping

    of

    KRl KR2 and KR4, and mapping of KR9, KR12,

    KR13 and KR14). This work belongs to Posiva site characterization programme for

    spent nuclear fuel disposal. The work has been ordered from and supervised by Posiva.

    The authors wish to thank Posiva s contact persons Mr. Heikki Hinkkanen and Mrs.

    Liisa Wikstrom for their useful comments and guidance. We also thank the experts

    who reviewed the work and results on their essential communication: Dr. Alan

    Geoffrey Milnes

    of

    GEA Consulting, Dr. John Hudson of Rock Engineering

    Consultants, Mr. Kai Front of VTT Technical Research Centre of Finland, and Mrs.

    Ursula Sievanen of Saanio Riekkola Consulting Engineers.

    Mr. Raimo Ruotsalainen and Mr. Esa Raivonen of

    Geological Survey, and Mr. Aimo

    Hiironen

    of

    Posiva deserve special thanks for arranging the review

    of

    the core samples.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    7/107

    3

    INTRODUCTION

    Orientations and properties of fractures were mapped from borehole imaging data of

    Olkiluoto boreholes KR9, KR12, KR13 and KR14 and partly boreholes KR1, KR2 and

    KR4. The data contained 1996 BIP imaging results from those KR1, KR2 and KR4

    sparsely fractured sections, that have not been included into previous interpretation

    745 m, Stn1hle 1996) and OPTV imaging results from KR9, KR12, KR13 and KR14

    in 2001 2306 m, Wild et al. 2002).

    t was necessary to carry out the image mapping as a separate task, because the drilling

    and core mapping has been performed well before some months to 2 years) this

    work. On general view, as core mapping is traditionally performed immediately on

    site, and for the imaging can be performed due to optical conditions not earlier than

    h lf a month after drilling, the most efficient mapping procedure has been recognized a

    separate work, divided into subtasks that are checked numerically between the steps.

    This also saves time and preserves consistency on mapping of the image properties.

    Any direct comparison to the actual core mapping data is a separate task and has been

    left to the interpretation phase during later works. Background data is referred to

    adequate level.

    The fracture mapping used extensively the WellCAD software from Advanced Logic

    Technology ALT 2001). WellCAD can handle and represent different kind of

    borehole data and store the data linked to the presentations. All logs

    on

    a single file are

    displayed on the same scale. The original image data together with borehole deviation

    data were imported to WellCAD

    by

    Ms. Eveliina Tammisto and Mr. Tomas Lehtimaki

    Chapter 2). Before mapping, the core samples and images were visually examined to

    understand the appearance of

    the features on the images. The fractures were mapped

    and data transferred to determination

    of

    properties

    by

    Ms. Tammisto KR2, KR4, KR9,

    KR13), Mr. Eero Heikkinen KR14) and Mr. Lehtimaki KR1, KR12) Chapter 3).

    Determination

    of

    fracture properties from image was completely done by Ms.

    Tammisto, a geologist. This was required by the Client, with an intention to maintain

    consistency in the properties see Chapter 4). Principles on determination were agreed

    with Mr. Jorma Palmen, a geologist. The data was first checked by Mrs. Pirjo Hella

    mathematician) to select the observations to be confirmed or rejected. The selected

    observations were checked over with the core samples see Chapter 5) at Loppi core

    archive KR1, KR2, KR4, KR9, KR12) and in Olkiluoto KR13 and KR14). Ms.

    Tammisto, Mr. Lehtimaki and Mr. Heikkinen participated into this work. The sole

    purpose

    of

    checking was to remove most

    of

    the features encountered

    on

    imagery that

    are not real fractures in sense of core mapping.

    Planning

    of

    the layout for printouts Chapter 6), and data delivery were done partly by

    Mr. Heikkinen, partly by Mrs. Hella and Ms Tammisto. Summary tables of different

    fracture properties were compiled by Mrs. Hella. Practical arrangements for printing

    and data compilation were done

    by

    Ms. Tammisto.

    The preceding data is listed in the Table 1 below. The fracture locations and

    intersection angles from core mapping were used in confirming the image observation

    before review with the core sample.

    Other background data has not been used directly,

    but understanding of these has been required to perform the mapping.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    8/107

    4

    The mapping of core sample has listed all penetrative mechanical discontinuities with

    their form and roughness on the sample, the mineralogical mapping specifically also

    the coating minerals. Hydraulic conductivity is mapped with the flow logging. For

    KRl

    KR2 and KR4 there has been also dipmeter mapping available, that has brought

    out valuable although incomplete orientation data.

    Table 1 The preceding material from mapped boreholes

    Method

    Systematic

    core

    mapptng

    Systematic

    mineralogy

    mapptng

    Hydraulic

    flow logging

    data

    Geophysical

    logging

    Electrical

    dipmeter

    BIP imaging

    and mapping

    Fracture

    database

    FDB)

    OPTV

    image

    Properties

    Length location primary

    reference), type open,

    filled, tight, slickensided),

    intersection angle, partial

    orientation, form,

    roughness, tentative

    mineralogy, frequency,

    RQD. All observed ones.

    Length, type open,

    breakout, slickensided,

    filled); thorough fracture

    mineralogy and

    appearance, thickness of

    filling. Not all fractures.

    Location and hydraulic

    conductivity

    of

    the

    conductive fractures

    Reference to lithology,

    fracture zones, hydrology

    Fractures, general

    orientation of banding;

    interpretation

    of

    orientation clusters

    Image from whole

    borehole, mapped data

    from selected sections

    those not included to this

    work); fracture location,

    type, orientation,

    aperture, form, tentative

    mineralogy, alteration

    Merging

    of

    above

    mentioned fracture

    properties to same

    observations

    Non-mapped image data,

    mapped in this work

    Borehole

    KR1

    KR2

    KR4

    KR9

    KR12

    KR13

    KR14

    KR1

    KR2 KR4

    KR9

    KR12

    KR13,

    KR14

    KR12

    KR1, 2, 4,

    9

    KR13, 14

    all

    Reference

    Suomen Malmi Oy 1989a

    SMOY 1989b, Rautio 1995a

    SMOY

    1990, Rautio 1995b

    Rautio 1996

    Niinimaki 2000

    Niiinimaki 2001a

    Niinimaki 2001 b

    Blomqvist et al1992 Gehor et al. 2001a

    Gehor et al. 1996

    Gehor et al. 1997

    Gehor et al. 200 1b

    Not available during mapping

    Pollanen Rouhiainen 2000.

    Rouhiainen 1999.

    Not

    available during mapping

    Niva 1989,

    SMOY

    1989c, d, 1990b,

    Julkunen et al. 1995, 1996, 2000, Lahti

    et al. 2001

    KR1, 2, 4 Lowit et al 1996

    KR9, 12, Not available

    13, 14

    KR1, 2, 4 Strahle 1996; mapping supplemented by

    Karanko

    et

    al. 2000. Systematic mapping

    similar to this work performed for

    70o/o

    of imagery.

    KR 1 2, 4 Saksa et al. 1997, Karanko et al. 2000

    KR9, 12, Wild et al. 2002

    13, 14

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    9/107

    5

    Most important data not received from other than the image mapping are the thorough

    orientation, and the aperture o the fractures, combined with the observation type seen

    in image. The type o observation is different from each mapping method, and will

    rather be supplementing each other than being exclusive.

    It was expected that there is a proportion o fractures not seen in the imagery (Saksa et

    al. 1997, Karanko et al. 2000).

    n

    this sense both the other mapping data and the image

    mapping results are in their part inadequate to describe the fracture properties alone.

    Without a direct comparison to the core data (performed in fracture database

    approach), it cannot be said which (kind of) core fractures are not included into the

    current observations. A priori assumption is that these may be thin, dark fractures on a

    dark background, and probably often perpendicular to the borehole (horizontal lines on

    image). Despite o this potential statistical bias, the coverage o core fractures is

    expected to be reasonable.

    The work in hand concentrated on mapping o the imagery to produce material for later

    geological usage. Results were referred with the core sample for quality s sake. These

    results should not be considered as the primary data source o fracturing. A relevant

    data source these become only after the fracture database has linked core fractures to

    the image data. For orientation distributions, the image mapping data is the only

    available thorough data set until the database comes available.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    10/107

    7

    IMAGE MATERIAL

    Borehole wall images

    of

    KR1, KR2 and KR4 BIP investigations (Stn1hle 1996) and

    OPTV investigations in 2001 from KR9, KR12, KR13 and KR14 (Wild et al. 2002)

    were analysed

    in

    this work. Previous interpretation

    of

    BIP data (Strahle 1996) was

    supplemented from least fractured sections, omitted from the preceding work. The

    interpretation included for each borehole the sections shown in Table 2.

    Table

    2

    The extent

    o

    he interpretation work or this report

    Borehole Borehole length range, m

    Length,

    m Image

    type Reference

    KR1

    341 -500 (sections d e in

    267.5 BIP

    Strahle 1996

    field data),

    644-

    740 (field data section g)

    KR2 300-

    390 (sections c d),

    2 1

    BIP

    Strahle 1996

    440-

    539 (section e)

    KR4

    390-

    480 (sections d and e),

    276

    BIP Strahle 1996

    570 - 739 (sections f g and j)

    KR9 40-602

    559

    OPTV Wild et al. 2002

    KR12

    40-

    793 751 OPTV

    Wild et al. 2002

    KR13

    6-499

    494 OPTV

    Wild et al. 2002

    KR14 9-513.5 502

    OPTV

    Wild et al. 2002

    Total 3050.5

    This chapter lists the techniques used, the properties of the images, and the observed

    factors in image quality, which have influenced to the mapping and

    may

    affect to the

    results.

    2 1

    Image logging arrangements

    The images consist of bitmaps arranged by depth and orientation. Images have been

    acquired by slowly lowering or raising a wire-line camera probe in a borehole and

    digitally sampling the borehole image at a dense, constant depth rate.

    n

    the probe the

    light of a LED-string is reflected from borehole wall to a mirror and further projected

    onto CCD-element (a video camera). The acquisition geometry, and the parameters of

    the image data content are shown in Figure

    1.

    Each 360 or 720 pixels' row in the image presents a full circle around the borehole

    wall captured from a circle on the image. This produces a radial resolution

    of

    1 or 0.5

    degrees, respectively. The specifications of the tools used are presented in Table 3.

    The capturing has been performed at a rate of

    113

    or

    115

    frames

    per

    second. The

    resulting recording rate or the depth resolution of the image lines is 0.5 or 1.0 mm,

    depending on available resolution and logging speed. Recorded image lines were

    stored into a computer and transferred to a MO or compact disc. Images contain RGB

    data at a colour resolution of 0 ...255 units for each Red, Green and Blue channels

    ( millions

    of

    colours ).

    Due

    to their very large size bitmap images were stored

    in 3

    150 m slightly overlapping sections during the field work. These data sections

    ( logging runs ) are indicated

    by

    letter indexes

    a

    through

    k

    in

    each borehole.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    11/107

    8

    able 3 The tool and operation specification details for BIP and PTV images

    Tool Details BIP OPTV

    Probe manufacturer

    RaaX Ltd Japan).

    Robertson Geologging UK)/

    OyO Japan)

    Operator

    SKB/Geosigma Strahle

    Robertson Geologging

    Ltd.

    1996), Sweden

    Wild et al.

    2002), UK.

    Vertical sampling used 1 mm max 0,25 mm)

    0,5 mm

    Horizontal sampling used

    1° 360 pixels) or 0.49 0.5° 720 pixels) or 0.24 mm

    mm at

    a 56 mm at 56 mm and 0.33 mm at

    76

    diameter borehole

    mm borehole diameter

    Storage size

    1Mb/metre

    4Mb metre

    Data format BIP binary

    Initially LGX binary),

    received as

    Dill

    Nominal logging rate

    1 5 m/minute

    0,5 m/minute

    Illumination

    LED at mirror LED at camera objective

    Recording geometry

    Conical mirror Hyperbolical mirror

    Cable

    Optical Coaxial steel armored

    Logging direction

    Upwards

    Mainly upwards

    An image has been oriented to the down side

    of

    inclined borehole BIP image, Strahle

    1996), or

    by

    a magnetic 3-axial compass Flux gate) to the magnetic North, then

    adjusted to the site North OPTV image). Because

    of

    the rapid changes in local

    magnetic field orientation in presence of magnetized minerals in the host rock, the

    initial OPTV image has been straightened up by the operator Wild et al. 2002) by

    replacing the orientation with Maxibor or Fotobor orientation data Rautio 1996,

    Niinimaki 2000, Niinimaki 2001 a b).

    Examples

    of

    the images are in Figures 2 and 3, and their colour histograms in Figures 4

    and 5 in Section 2.2 below.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    12/107

    a) The down-hole tool assembly

    NON M  GNE11C _

    CEN f iV\ ISERS

    T R N

    SP

    A

    RE

    NT

    WINDOW

    SONDEHE O

    )

      XI

    X

    ACCELEf\O t.IETEf\

    c) Screen (camera) view

    9

    -

    Camera field

    of

    view

    Image of

    circular mirror

    Sample circle

    Hypothetical X Y

    ~ ~ ; ; . . . _ _ _ . pixel grid

    b) The

    OPTV

    functional arrangement

    CD

    '0

    s

    r

    Video Camera

    c) Image

    arrangement

    Azimuth

    N

    E

    s

    Lighting Ring

    w

    Figure 1 The nomenclature used in the image presentations and alignment. a) the

    PTV image tool and b) the camera functional arrangement with c) the

    screen view during logging. The 3-component inclinometer and

    magnetometer provide the tool orientation reference to magnetic and site

    North and

    to

    borehole deviation. This is used to align the Sample circle in

    c) to arrange the image at subsequent depth lines to show the pixels

    clockwise, each row starting at reference direction for OPTV the True

    North; for BIP the down side) modified from Wild et al 2002). The

    sinusoid line on wrapped-out image shows a planar fracture projection.

    N

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    13/107

    10

    2.2 he image properties

    The image properties are differing slightly with respect to the technique used in

    logging. Figures 2 - 5 illustrate these differences. Generally, the difference in pixel

    resolution can be seen in approximately 1

    2

    vertical and

    1:

    1 horizontal scale images in

    Figures 2 and

    3

    This scale is typical for the on-screen mapping and image quality

    assessment, but usually not for presentations or printing. The features discussed on this

    Section are most clearly visible on computer screen and suffer on quality when printed.

    The hair cracks on pegmatite granite, having dark coating, appear slightly jagged in 1

    degree

    y

    1 mm BIP image Figure 2), and the overall resolution gives a bit blurred

    impression on this scale. The individual image pixels can be clearly seen in this scale.

    The two-fold sharper resolution of OPTV image 0.5 degrees

    y

    0.5 mm) on less

    fractured tonalite gneiss gives a more sharp impression, the individual pixels can not

    be clearly distinguished on the image on this scale, and the minor cracks and grain

    boundaries do not appear as jagged

    as

    in BIP. Even the foliation can be seen. Due

    to

    better resolution the OPTV data storage size is four times larger than for the BIP.

    The above-mentioned properties are a result of selecting the data acquisition

    parameters and the logging speed rate rather than due to tool functional properties. The

    dark coating visible on Figure 2, and the opaque band on Figure 3 are due to borehole

    conditions see Section 2.3 in more detail), and not related to the tool performance.

    Further to the tool and logging resolution, the tool arrangement has influenced to the

    image appearance. This can be seen in Figures 2 and 3 in the differences of lightness

    and color saturation. The differences are shown in Figures 4 and 5, in sense of image

    analysis, in the RGB channel plots of Figures 2 and 3, respectively. For a shadow cast

    on the fracture trace due to the tool arrangement see below), the difference between

    BIP and OPTV can be best viewed on images shown in Chapter 4 Figures 15 and 22).

    In the BIP tool the light source is located adjacent to the conical mirror, to where the

    light reflects perpendicularly from the borehole wall, and further to the camera

    objective. The light source is powerful and the distance short, so in places there has

    occurred saturation of the colour channels Figure 4). Usually there is also due to large

    variation

    of

    dark and light rock types, an effect

    of

    automated exposure or shutter speed

    light intensity) adjustment. No later adjustments to the images have been applied.

    In the OPTV tool the light source is located behind the camera objective, and the

    oblique light beam reflects first from the borehole wall to a hyperbolical mirror image

    is similar to that from a fish-eye lens), and then to the objective. Despite

    of

    the strong

    light source, the image is rather dark Figure 5). Light intensity is higher with larger

    76

    mm diameter. t was required to adjustment RGB scale contrast, brightness and colour

    intensity). The adjustments were performed for optimal printing on photographic paper

    for reporting Wild et al. 2002), and have caused the image colours to convert a bit

    false, with slightly too high relative intensity in blue and red components. The light

    intensity range is a bit narrow due

    to

    the initial darkness

    of

    the image, but the contrast

    is adequately sharp for on-screen mapping. No later image enhancement has been

    applied for presentations.

    The oblique angle of illumination allowed many fractures to cast a shadow on the

    image, so these can be observed pretty clearly the difference is seen best on Figures

    6a,

    15

    and 22).

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    14/107

     

    igure

    2

    Example ofBIP image data from depth level

    340.40

    360.66 m of KR2,

    granite pegmatite with hair cracks. Scale is approximately 1:2 vertically

    and 1:1 horizontally. The pixels on the image can be seen quite well on this

    scale. The light grains are slightly overexposed see also Figure 4

    for

    reference). Thin lines appear jagged.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    15/107

      2

    igure 3

    Example of PTV image data from depth level 40.50

    40

    76 m ofKR9,

    tonalite gneiss. Scale is approximately 1:2 vertically and 1:1 horizontally.

    The pixels on image are not distinct in this scale. Image is sharp and the

    grain boundaries and hair cracks appear rather smooth. Foliation and even

    the vertical tool marks can be distinguished. On the other hand, the colour

    range is narrow, and the dark grains are a bit underexposed see also

    Figure 5 below).

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    16/107

      3

    hannel Red

    Figure

    4

    RGB colour channels o the BIP image

    in

    Figure 2 The colour range

    is

    well covered although the red channel is partly saturated resulting

    to

    slightly reddish image and partial overexposure.

    Figure 5 RGB colour channels

    o

    he OPTV image

    in

    Figure

    3

    The colour range

    is

    narrow and show a slight underexposure on all components specifically on

    red. This has lead to a slightly bluish image quality. Differences arise from

    the tool geometrical arrangements.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    17/107

    14

    2.3 he image quality due to borehole conditions

    Generally the quality

    of

    images is good and accurate. There are several technical

    factors influencing the images that also affect the mapping see Table 4).

    Table 4 The factors on image quality

    Factors on image quality

    Reason

    Figure

    Vertical banding,

    Tool decentralization Figure 6a

    opaque bands

    Flow of

    mud in water Figure 6b

    Tool marks

    Spiral banding, reamer Figure 7a

    Vertical scratch lines Figure 7b

    Optical conditions

    Cloudy water Figure 8a

    Coated borehole wall Figure 8b

    Stick-slip motion ofprobe

    Uneven movement, due to Figures 9a, b

    coarse

    surface,

    cable

    tension, flow, low weight

    of

    tool, or other

    Typically the quality deeper in some boreholes has suffered from cloudy water, dark

    opaque bands on image or stick-slip motion. Vertical illumination differences see

    Figure 6a) occur due to the decentralization of the tool especially for OPTV images

    from larger 76 mm diameter boreholes KR13 and KR14). Other vertical banding

    occurs on the bottom side of the borehole due to an opaque material flowing in the

    borehole water Figure 6b

    .

    This line helps to confirm the image orientation.

    Drilling with an oversize reamer has left spiral marks occurring as 3-5 cm wide

    patches, repeated at 13-15 cm period, of different colour than surrounding image see

    Figure 7a). On the other hand, tool marks can be seen also as light and narrow vertical

    lines Figure

    7b

    . These occur both in the BIP and the OPTV images.

    Dirty borehole water can be seen in many locations, especially at fracture zones. The

    feature is more severe when the borehole is freshly drilled, or when other measurement

    have been performed shortly before image logging Figure 8a). Boreholes drilled long

    before image logging have prominent sedimentation

    of

    rust or other dark material onto

    the wall Figure 8b).

    The stick-slip motion of the imaging probe can be seen as horizontal lines of short

    vertical banding. The probe had jammed instantaneously but depth record changed

    continuously, and the probe has then suddenly jumped over some 5 mm OPTV,

    Figure 9b) to

    10

    cm BIP, Figure 9a), where true image pixels are missing because

    they have been replaced with a single value in each column.

    The length of each slip movement is smaller in OPTV image than in BIP. This may get

    an explanation from the slower 0.5 m/min probe run rate of OPTV compared to 1.5

    m/min of BIP. The phenomenon is expected to be caused y friction on the probe due

    to coarse wall fractures etc.), or possibly a strong flow along the borehole; or cable

    tension, or the light weight

    of

    the probe itself causing uneven movement, among the

    other possible reasons. No direct explanation for this has been available so far.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    18/107

      5

    igure 6a Vertical banding. Probe decentralization KR13, depth

    36.70

    36.97 m .

    Light intensity is higher on the high side of borehole larger distance from probe

    produces better illumination due to distance and less narrow reflection angle).

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    19/107

      6

    igure 6b Vertical banding Dark material flow along the borehole KR9, depth

    231.60 231.95 m). The phenomenon is worst below fracture zones

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    20/107

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    21/107

      8

    igure

    7b. Tool marks on the image. Narrow vertical tool scratch lines KR13,

    401.10

    401.35

    m). The water is also rather cloudy dirty)

    in

    this image, and there is an

    opaque band on down side

    o

    the borehole

    to

    the east on borehole inclined to the

    west).

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    22/107

    19

    igure 8a Water quality. Dirty water

    in

    the borehole KR14, 512.50 512.80

    m .

    Flowing clay particles can be seen

    on

    the cent er line

    o

    the image down side o the

    borehole inclined to the north).

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    23/107

    20

    igure 8b Water quality. Coated surface o borehole wall KRJ, 342.80 343.05 m .

    The coating will accumulate during long period, and is best preserved on the spiral

    tool mark.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    24/107

    2

    igure

    9a Stick-slip motion. BIP image depth

    643.90- 644.15,

    KRJ). The length

    of

    each slippage is

    20 30

    mm, sometimes even larger, and damages the image severely

    when observed. Occurrence is rather rare on sparsely fractured bedrock.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    25/107

    22

    Figure 9b. Stick-slip motion. PTV image KR14, 505.55 - 505.80 m). The length of

    each slippage is only 3-5 mm, so the mapping is not totally blocked. The phenomenon

    can be thorough over long sections.

    In general the BIP image had suffered less from the dirty water and does not indicate

    severe stick-slip motion on sections treated in this work. In 1996 the water in the

    boreholes had settled down reasonably long. Also the colour intensity is good. On the

    other hand the image pixel resolution is clearly poorer than in OPTV.

    or

    OPTV image

    the pixel resolution is good but the image is dark and a minor stick-slip motion makes

    the quality at the deeper parts slightly poor. The short time after drilling and other

    borehole operations and probably the upward sequence of OPTV logging made other

    boreholes than long time settled KR9 to suffer from dirty water conditions. KR12 was

    reasonably good but for freshly drilled KR13 and KR14 the condition was less good.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    26/107

      3

    Also the decentralization causes more uneven light distribution in 76 mm diameter

    boreholes compared to 56 mm, although the image quality is otherwise better with

    larger diameter.

    2 4 The preconditioning of the image material

    Before the mapping the images were necessary to transport into the software system

    used in the mapping.

    Original OPTV and BIP data was imported in 10 m sections into WellCAD (version

    3.0) working logs and stored into files containing the depth and original measurement

    section information (index a - k ) in the file names. During the logging there was at

    each run section a 1 - 2 m overlap reproduced in the images. Data import preserved the

    original data resolution and image orientation, and allowed scalability on the depth and

    the radial axis.

    The BIP data was contained in binary files

    of

    100 m in each, including recorded depth

    and orientation. The depth reading was adjusted according to values obtained from the

    original report (Stn\hle 1996), representative at the top of each 10 m section. The

    adjustments are linear between the top and the bottom of the each 100 m run section.

    Residual differences of 10 - 20 cm to core sample can occur due to local deviation

    from the trend. The stretch

    of

    the image over 1 m section is typically 5 - 15 cm. In

    places the residual difference may reach 30 cm due to deviation in core mapping depth

    (reconstruction of sample has failed),

    or

    a slight jamming of the probe.

    The OPTV image was read in as 10 m sections of Device Independent Bitmap (DIB)

    files. Original OPTV records were adjusted

    by

    setting top

    of

    subsequent log to the

    bottom of previous log. The DIB files do not contain depth or orientation data, but left

    edge of image is at the true North, and the top and the bottom depth values for each 1

    m section were taken from the field work report (Wild et al. 2002) during the import.

    The difference to the core sample is typically less than 10 cm, and at largest less than

    30cm

    The depth differences were found during the fracture mapping, but were not

    systematically corrected for in the images. A proper adjustment would require trend

    analysis for each section and non-linear

    or

    piecewise linear adjustments on that basis.

    This was not task of the work.

    The imported images were checked for possible errors and proper orientation.

    Supplementary data including rock types, crushed borehole intervals, drilling fractures

    with orientation and intersection angle, drilling fracture frequency, and the borehole

    orientation (azimuth from the North and inclination from the vertical) were presented

    in the Well CAD working logs to aid the mapping.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    27/107

    25

    MAPPING THE FRACTURES

    In the planning phase o the work, typical appearance

    o

    the fractures in the images

    were compared with those from core sample mapping at Loppi core archive (KR2 and

    KR4, KR12)

    o

    Geological Survey and in Olkiluoto (KR14). The task was to

    distinguish the real fractures (penetrative mechanical discontinuities) from narrow

    mica bands, veins, foliation etc. possibly appearing similar way on the images.

    3 1 Preparation of the material

    In Well CAD the images were set to a depth scale o

    1:1

    - 1:2 and stretched on the log

    so that the horizontal scale was approximately the same as the vertical (image 17 - 22

    cm wide on screen; aspect ratio close to

    1:

    1

    .

    With this setting the pixels on screen

    (0.25 mm) were close to the true image resolution.

    A new Structure Log to be used in mapping was created on the WellCAD file. The

    structure log and the BIP or OPTV image log were set on top

    o

    each other in exactly

    same horizontal and vertical scales. The borehole diameter was set to a correct value

    (56 or 76 mm) in the structure log. The tadpole presentations (arrow position on depth

    shows the location, on horizontal scale the dip, and the tail points to dip direction,

    north to up, east to left etc.)

    o

    drilling observed fractures (both the 3-D orientation and

    the intersection angle) were set on the side

    o

    these two logs.

    3.2

    Picking of the features

    On the structure logs were drawn numerically the sinusoidal projection traces o the

    mapped fracture planes. The software stored the orientation data

    o

    the planes into data

    table. Any (near to) planar feature appears on wrapped-out image as a sinusoid trace,

    where the lowest point indicates the dip direction and amplitude (difference from

    highest to lowest point) indicates the dip when correlated to the borehole diameter.

    When the interpreted traces were combined to the borehole deviation data, the true

    orientations o the fractures were obtained and converted to the tables. The data was

    presented in a tadpole log beside the image log and compared to the core logging data.

    A discontinuity feature in the image was considered to be a potential fracture, when it

    differed clearly from the background image; was continuous over the image width, and

    was not easily confused to a vein, band o mica plates, foliation or schistosity, or other

    thin, similarly orientated features. These criteria are in accordance with those applied

    in core mapping. In core mapping the features that penetrate thoroughly over the

    sample are reported as fractures. This cannot always be seen reliably for weak class

    o

    features in the image. Features

    o

    incomplete continuity (partial fractures - terminating

    to another, faulting one; discordant cracks between two adjacent fractures) were agreed

    to be left out from the image mapping in this work, although these may be o

    importance. Mapping can be continued for these features later according

    to

    the

    requirements.

    Typical indications o a fracture are often light in colour over a dark rock type, or dark

    in colour over a light rock. These can be few pixels to several millimetres wide, or

    even o a sub-pixel width (less than

    0.3

    -

    0.5

    mm) when showing clear continuity. A

    (drilling induced) breakout or washout o filling material or the borehole wall will

    make the observation easier. A black line in an image would indicate open space

    (water filled) between the fracture surfaces, a coloured line on the other hand some

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    28/107

    26

    kind

    of

    filling. Examples

    of

    clearly visible (probable) and poorly visible (potential)

    fractures are shown in Figure

    10.

    Fracture observations were picked from activated structure logs y clicking with

    mouse pointer in minimum three points (usually more) on the sinusoidal trace. The

    software then calculated the fracture plane matching to the points and displayed the

    corresponding projection trace on the log. The orientation data was stored into a file for

    editing and exporting. The assigned depth is the midpoint of the trace depth (average

    of

    high and low points). The trace can be activated and graphically edited to represent the

    observations in an optimal way.

    3.3

    uality criteria of the picking

    Depending on the curvature and roughness of the real fracture surface pretty large

    deviations from the orientation can occur. The software has directly calculated the

    orientation

    of

    the plane relative

    to

    the borehole coordinates. The values only need to be

    corrected for borehole deviation.

    Typical accuracy of the plane fit is some

    1 5

    degrees. The relative dip is most accurate

    when the plane intersects the borehole image nearly perpendicularly. The relative dip

    direction is most accurate when the trace is aligned with borehole axis. For a stepped

    or undulating trace, a visually well-representative trace was fitted, and there can be

    quite a large deviation locally from the plane. This also means that the orientation is

    more inaccurate, error can be even several tens of degrees. The final accuracy will

    depend on the orientation of the fracture. The true dip direction is very insensitive for

    near-horizontal fractures, even differences of some 30 degrees need to be allowed; on

    the other hand for near vertical ones the dip will greatly depend on the accuracy of the

    plane fit - small curvature

    of

    the trace or change in the intersection angle with borehole

    axis will cause comparably large difference on the true dip.

    It

    has to be borne in mind

    that also the core mapping orientation may in cases be inadequately defined.

    The core mapping data from the drilling reports were used to guide the image mapping.

    The mapping was done from top to bottom in order to detect depth differences. At this

    point it was not possible to combine the mapped fractures together with the fracture

    data from the drilling. The features that were clear and continuous on the image were

    mapped most easily. The core data were mostly used for checking the correct depth

    and orientation of the image, and when poorly visible features were mapped. Extremity

    points of traces were avoided to be used in plane fitting.

    For each observation a level

    of

    certainty was determined on basis

    of

    visibility on image

    and on correspondence to the core data depth and orientation. Clearly visible trace was

    nominated as probable and poorly visible as possible . Good correspondence with

    location (10 cm) and reasonable correspondence with orientation (intersection angle, or

    when available the dip direction and dip within 30 degrees) was indicated with label

    existing , poor correspondence to either location or orientation with new .

    Finally, after each depth section of a borehole was mapped, the structure logs of

    observations were copied, and the copy was transformed to the true site coordinate

    system. The data was converted to tadpole presentation and compared with oriented

    drilling fractures. The data were exported to an Excel sheet. All data sets of a borehole

    were merged together. Duplicate observations at overlapping logging run sections (1-2

    m) were discarded.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    29/107

    27

    Example o clearly and weakly visible fractures, and the mapped traces, are shown in

    pair o images in Figures 1

    Oa b.

    The probable fracture (green line) is seen as 5 mm

    wide continuous trace, with a slight irregularity and curvature on the shape. The

    possible fracture (thin black line) is seen as jagged, less planar thin bluish-white line

    on both sides

    o

    the projection line (see red arrows on Figure lOb). Similar but more

    planar feature is on the bottom o Figure

    10.

    This kind o features needed to be

    confirmed from the core.

    Figure lOa. Examples o racture picking KR9, 141.50-141.75 m). A clear feature

    upper, a probable fracture), and a less well visible feature lower, a possible fracture)

    with the fitted sinusoid curves

    o a plane projection. The poorly visible trace is

    indicated on Figure

    JOb

    arrows).

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    30/107

    28

    epth

    PTV

    ~ ~

    1:2

    E s

    w

    Figure JOb Examples o racture picking KR9, 141.50-141.75 m). A clear fracture

    upper), and a less well visible fracture lower) without sinusoid curves. The red

    arrows indicate the points where the undulating and thin, bluish white fracture trace

    can be seen on the image. Penetrative continuity can be seen best on screen.

    The mapping was done as a separate phase of work and entirely before proceeding to

    classification and property determination. To speed up the procedure, mapping was

    performed by three different persons a whole borehole for each):

    Ms Tammisto, a geologist: KR2, KR4, KR9, KR13, KR14

    Mr. Lehtimaki, a geophysicist:

    KRI

    KR12

    Mr. Heikkinen, a geophysicist: KR14

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    31/107

     

    - -

     

    ~

    - - - - - - - -

     

    -

     

    - - - - -

     

    - - ---

    9

    Due to different handwriting , experience and background of persons involved, the

    mapping was assumed to lead to slightly a different result with respect to relative

    amount and certainty

    of

    observations (specifically the less visible ones). Some persons

    tend to pick more clear features (losing some true data), and some try to find as many

    features as possible (with a risk

    of

    mapping also non-true fractures). The differences

    could have been best confirmed with a blind comparison test of a same section to avoid

    influence of image quality and rock type.

    The procedure of separate mapping of properties, and checking of observations against

    the core sample were considered essential. Differences were considered to be equalized

    by limiting the determination of properties to a single person, a geologist (Ms.

    Tammisto).

    Also, most important, using the separate phases made each phase more efficient and

    time intensive compared to a situation where all steps were carried out simultaneously.

    And, the statistical checking

    of

    picked properties was more consistent when handling a

    single phase simultaneously over a borehole before proceeding to the next phase.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    32/107

    3

    DETERMINATION OF THE PROPERTIES

    The determination of fracture properties was performed

    by

    Ms. Tammisto, a geologist,

    alone to maintain consistency of the results. The work was performed with keeping

    on

    display simultaneously an Excel spreadsheet file with depth and orientation data, and

    the WellCAD image log from corresponding depth.

    The observation was viewed from the image log and the type, aperture and filling,

    level of certainty, form and colour were determined. The possible values for properties

    were agreed beforehand. The data was further classified and rearranged during the

    checking and analysis procedure (Chapters 5 and 6).

    4 1 Design of property parametres

    The parameters were designed before the mapping to preserve similarity with the

    previous image mapping results (Stn\hle 1996) and core mapping results (Suomen

    Malmi Oy 1989a, b, 1990, Rautio 1995a, b, 1996, Niinimaki 2000, 2001 a, b) from the

    area, and

    on

    the other hand the work done on merging the fracture data from different

    mappings (Karanko et al. 2000). The technique was adjusted further to this work to

    allow thorough mapping of the information visible in the image. The mapping

    concentrated to the properties that cannot be reliably or completely obtained from the

    core mapping (aperture and orientation were the most important).

    On

    the other hand,

    the other, supporting parameters that are mapped also from the core, were expected to

    assist later comparison and merging of the different observations and techniques. The

    observation and their illustrations

    on

    pairs

    of

    Figures 11-24 are listed

    in

    Table 5 below.

    Table

    5

    The parameters used in the determination

    o

    properties

    Nos. Properties Comments

    Figures

    where

    illustrated

    1 Level

    of

    Clear visibility on

    image=

    probable , corresponding the

    certainty

    core data = existing

    Clear visibility on

    image=

    probable , not corresponding

    2

    the core data = new

    Poor visibility on image = possible , corresponding the 3

    core data = existing

    Poor visibility on

    image=

    possible , corresponding the 14

    core data = new

    2-5 Type Visual properties

    of

    feature on image; open fractures

    15a-d

    Partly open fractures

    16a-d

    Fractures with filling (no visible aperture) 17a-b

    Fractures without visual filling or aperture

    18a-b

    6-7 Aperture, Measurement principle

    of

    the properties (perpendicular to

    19a-b

    thickness

    the fracture plane)

    8-9

    Form

    Fracture form, seen in image width scale (planarity, steps, 20a f

    undulation) and in local mm scale (local smoothness)

    21a-d

    11- Col

    or

    Color or grayscale data visible in different conditions,

    22a-b

    3

    ranging from darkness or gray value to col or dye 23a-b

    14 Seepage Indication

    of

    material inflow from the surface 24

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    33/107

    32

    The determined properties and their possible values are summarized in Table 6

    Distributions o the fractures according to their class before checking are listed in

    Table 7 The details on mapping parameters are presented below.

    Level

    o

    certainty 1)

    was used as a guiding basis for internal quality analysis o the

    mapping. Also the observations to be checked were selected on this basis (Chapter 5).

    Checking maintains the probable/possible classifications on Table 7 based on the

    image visibility, but will alter the existing/new to checked/rejected when

    applicable, to Table

    9

    The rejected class is not maintained in the final results (Table

    9). The possible ones were checked systematically, the probable ones only with

    few checks. A probable feature is so clear that there is no doubt on its determination

    as a fracture. Observation was possible when it was not sure from image whether

    there was a fracture or a vein or whether the observation was penetrative or not.

    Feature was considered existing (fracture) when there were adequately similarly

    oriented observations at the same depth level in the drilling report, and new when

    there were no correspondence on either location or orientation. Also depth differences

    between the image and core sample result in possible class observations. The typical

    observations on different categories are illustrated in Figures 11 - 14.

    Fracture type

    2

    - 5)

    consists o four independent observations. These were

    openness , filling , alteration and tightness . The previously used mutually

    excluding types open-with filling-closed were considered to be inadequate so these

    parameters were mapped separately. This results to fractures that can be with filling

    and/or open. Anyhow, a closed fracture was not classified as an open fracture or a

    fracture with filling, because it cannot be judged from visual appearance on image.

    The type is a property that is observed from image at the intersection line

    o

    the plane

    (along the plane), as there is no view onto the plane surface in image like there is in the

    core. Observations are limited to their scale, i.e. features clearly smaller in thickness as

    the pixel size are assigned with property closed , although there can be filling

    material

    or

    aperture thinner than 0,25 mm. Thus a closed fracture in the image can

    quite well be a fracture with filling in core mapping, etc. The open and partly open

    fractures may be seen

    as

    black pixels in BIP OPTV, and shadow in OPTV. Filling

    can be present also when a fracture is open . Partial filling has not been separated fom

    complete in mapping. And, i there is no observation with filling , it does not mean

    there would not be filling on the surface. The Figures 15 - 18 below show examples o

    different fracture types.

    The thickness 6)

    o the visible solid filling between the fracture surfaces and the

    aperture 7)

    (a distance perpendicularly between fracture surfaces being apart each

    another) were measured in millimeters from central part o the fracture trace,

    perpendicular to the fracture plane, or, when the fracture was partly open, from the

    open part. This was thought as the most representative value. At the peak points o the

    projection (upper or

    lower end) the cut section is wider and usually more interfered by

    drilling than at the central part. Thickness does not include aperture, nor vice versa.

    Both aperture and thickness can vary considerably along the trace, so these estimates

    should not be counted together to a total width estimate.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    34/107

    33

    The aperture may partly or totally originate from drilling induced physical failure

    (break-out) or washout of material between the surfaces, or even between closely

    spaced several fractures, or from clay-filled or brecciated larger zone. f the

    observation is linked to clearly collapsed or washed wide section (a narrow zone), the

    edges have been mapped as separate observations. The fractures inside the zones have

    been picked and classified only

    if

    they appear clearly. The image is the only

    opportunity to get a direct openness indication

    of

    a fracture surface, as usually the

    reconstruction of the core sample has no means to observe it; and there are rare n

    s tu

    indications on open fractures in other than hydraulic conductivity data. Figure

    19

    shows an example ofmeasuring the aperture and thickness.

    The form and roughness

    8

    -

    9)

    of the surface have been described as a deviation

    from the plane (from the sinusoid trace) in two separate scales. In the scale

    of

    the

    wrapped-out image (200 - 250 mm), form 8) was analyzed. When the fracture trace

    follows accurately enough the sinusoid

    of

    a plane, observation is planar .

    f

    the trace

    deviates several millimeters in any point, it is undulating (i.e. curved). In case the

    surface changes suddenly several millimeters, it is denoted as stepped . In the smaller

    few millimeter scale, roughness 9) was analyzed. The fracture is smooth if there is

    no observable deviation from the sinusoid, and otherwise rough . Examples

    of

    different forms of the fractures in different scales are shown in Figures 20 - 21.

    The image observation is obtained from the side of the fracture plane, not from the

    surface as in the core mapping, so the observations may not be quite comparable to the

    surface roughness. Because the property is clearly observable, and also indicates the

    accuracy of the plane fitting, it was recorded.

    The colour 11 - 13) of fracture filling was assessed on three-level scale. Depending

    on the colour and darkness of the background, and quality of the image, the colour

    could be observed as

    brightness 11 dark or light; especially when other colours cannot be seen)

    grey intensity (12, white-grey-black)

    colour dye (13, brown-red-yellow-green-blue).

    In certain cases there are two or more separate colours, seen partially or layered on the

    trace. From OPTV data the colour was observed from initial field data records, as in

    the final data the contrast adjustment has altered the colours. In Figures 22 and 23

    examples

    of

    the colour or grey intensity observation are shown.

    Seepage 14) can be observed only when some material flows into the borehole from a

    fracture and the colour of flowing material differs from the background. Outflow

    cannot be observed. The flow from the fracture or along the borehole can be seen as a

    curved fan on the image. The fan initiates from channels on fracture surface. t can be

    guided towards the high side (light, warmer water) of the borehole, or to the low side

    (heavier, cold water or material) and tur narrower before it vanishes within few tens

    of

    centimeters. A visual observation at close probe view indicated that such trace is a

    turbulent flow of material in the water. t can also be a stain of material sedimented on

    the wall over a longer time. Direction (upwards or downwards) has been marked for

    specific fractures. The Figure 24 shows an example

    of

    seepage location.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    35/107

    34

    able 6

    The parameters treated during mapping with allowed values.

    No.

    Property

    Values omments

    1

    Level of

    Probable, Drilling fractures at same ±

    10

    cm depth range, orientation

    certainty

    existing similar to the image observation within 30 degs, clear on

    image

    Possible,

    Drilling fractures a t same ± 10 cm depth range, orientation

    existing different from the image observation, and/or not clear on

    image

    Probable,

    Drilling fractures not observed at same ± 10 cm depth range,

    new or orientation clearly different to the image observation;

    however clearly visible and fracture-like

    Possible,

    Drilling fractures are not observed at same depth range ea.

    new ±

    10

    cm, or orientation is different to the

    TV

    observation,

    and are not clearly observable from image

    2

    Openness Open Projections apart from each other

    Partly open Part of surfaces apart from each other, part are closed when

    seen in the observation scale

    Not open Surfaces are adjacent in the observation scale

    3

    Filling With filling

    Filling seen in image (coloured pixels, different from

    background)

    No filling

    No filling seen in image ( in observation scale)

    4

    Alteration Altered

    Fracture lines are discoloured or disintegrated (rust, etc).

    Fresh Fracture lines intact, no colour changes

    5

    Tightness

    Not closed

    Surfaces are either apart, separated with some material, or

    weathering is observed

    Closed

    In the observation scale 0,25- 0 5 mm) the surfaces are

    closed, there is no filling and fracture is not weathered.

    Excludes the others.

    Filling thickness

    Measured from central part of trace (in mm)

    7

    Aperture

    Central part of trace, for partly open from the open part (in

    mm)

    8

    Shape Planar

    On image width scale

    undulating

    stepped

    9

    Roughness Smooth

    On few mm scale

    Rough

    10

    Mineral Possible

    An estimate, not stored n final data

    mineral

    11

    Darkness Dark

    Only when image dark and no color or grey scale can be

    Light

    estimated

    12

    Grey intensity

    Black Other than colour dye

    Grey

    White

    13

    Colour dye Yellow

    Other than grey intensity

    Red

    Green

    Brown

    14

    Seepage

    Upwards Material flow into borehole and its direction

    Downwards

    No

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    36/107

    35

    PTV

    e

    1:2

    E

    N

    9

    Drill obs tract

    7JD

    Figure 11 Existing and probable observation KR14, 72 80 73.10 m green sinusoid

    and tadpole). Similar core fracture black tadpole) is observed close to the same depth.

    The trace is clear and continuous on image.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    37/107

    36

    epth

    PlV

    e

    1:2

    r

    N

    E

    N

    0

    90

    Drill s   fract

    90

    Figure 12a. Existing and possible observation KR14, 73.60- 73.80

    m,

    green sinusoid

    and tadpole). A core fracture is observed close to the same depth black tadpole). The

    trace on image is a bit vague, and its continuity uncertain a white undulating line).

    The dip is steep; the dip direction o a steep fracture can easily turn 180 degrees

    in

    a

    vertical borehole

    i

    he feature is undulating; real dip difference is approx. 3 0 degrees

    over the vertical.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    38/107

    37

    OPl\1

    :2

    N

    N

    N s

    Jtl

    N

    Figure 12b Existing and possible observation KR14, 73.60 73.80 m without the

    sinusoid curve. The red arrows clearly indicate the undulating white line, a possible

    fracture with filling. The real reported fracture can be at slightly different position

    i

    re-opened.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    39/107

    38

    Pmjection 11 9

    igure 13a A new and probable observation KR9, 70.00 70.20 m . Similar core

    fractures are not observed near to the same depth, but the observation is clear on

    image. The fracture clearly

    is

    there, and typically when checking from the core sample

    the observation was confirmed. It is possible that a welded fracture has not been

    reported in drilling report although it could have been observed i broken apart. The

    form deviates significantly from planar see arrows

    on

    Figure 13b).

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    40/107

    39

    Depth

    OPTV

    :2

    E

    s

    w

    Projection

    E s w

    Figure 13b. New and probable observation KR9, 70.00 70.20 m without the

    sinusoid curve. The arrows indicate the position where the continuous, undulating line

    is best observed on image.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    41/107

    4

    Pmj

    ectjon 1 9

    Figure 14a New and possible observation KR9, 71.55-71.75 m, black line and

    tadpole, on sparsely fractured rock . Similar core fractures are not observed close to

    the same depth, and the observation is not very clear on image. It is possible that there

    is a mica plate originated continuity on core, or a hair crack i seen on image is not

    penetrative; or the fracture has not been observed

    on

    the core mapping. The checking

    o these features on the core either led to rejection

    o

    observation, but

    in

    few cases a

    new real fracture has been added to the database.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    42/107

    4

    epth

    PTV

    :2

    E

    s

    w

    Projection

    E

    s

    w

    71 6

    Figure 14b. New nd possible observation KR9, 71.55-71.75 m) without the sinusoid

    curve. The arrows indicate where the faint trace

    of

    possible hair crack has been seen

    on image black, continuous undulating line between feldspar grains).

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    43/107

    42

    Figure 15a An open fracture with filling

    KR9, 13 7

    55

    13

    7

    7

    5

    m.).

    The arrow

    shows the location where some filling material has been observed. The filling is not

    penetrative in this case and may have partly washed out during drilling.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    44/107

    Depth

    :2

    137 6

    43

    E

    PTV

    s

    w

    Figure 15b. An open fracture, with fi lling KR9, 137.55 137.75 m). The sinusoid and

    arrow removed so the trace is more clearly seen.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    45/107

     

    Figure 15c An open fracture with sinusoid curve, no visible filling KR12, 713.35

    713.55). There is some indication

    o

    the drilling debris lain on bottom

    o

    the surface,

    but no compact filling between the surfaces. Either filling did not exist on this location,

    or has been completely washed away.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    46/107

    Depth

    :2

    713 4

    45

    OPTV

    s

    w

    N

    Figure 15d

    n

    open fracture without sinusoid curve, no visible fill ing KR12, 713.35

    7

    .55). The same image as in Figure 15a.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    47/107

      6

    Figure 16a A partly open fracture, with filling KR12, 738.35- 738.50 m . There

    is

    clearly compact filling calcite?) between the fracture surfaces, and

    p rt of

    he

    fracture trace is clearly open on North left) side.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    48/107

      7

    Depth

    PTV

    ~ ~

    :2

    s

    w

    738 4

    Figure 16b A partly open fracture, with filling KR12, 738.35- 738.50

    m .

    The same

    image as on Figure 16b.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    49/107

      8

    Figure 16c A partly open fracture with sinusoid curve, no visible filling KR12,

    713.15- 713.35 . The aperture is clearly seen but no indication

    o

    any filling material

    is present.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    50/107

    49

    epth

    PTV

    :2

    ~ ~

    N

    s

    w

    N

    713 .2

    Figure 16d A partly open fracture without sinusoid curve, no filling KR12, 713.15

    713.35). The same image as

    in

    Figure

    16c

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    51/107

    50

    Figure 17a

    A fracture with filling, neither open nor partly open KR9, 142.65

    142.80

    m).

    The filling, light in colour, can be seen on west right)

    p rt

    ofthe trace.

    In

    the observation scale, there cannot be seen any aperture or breakout fracture

    is

    healed).

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    52/107

      epth

    :2

    N

    5

    E

    OPTV

    w

    N

    Figure 17b. A fracture with filling, neither open nor partly open KR9,

    142 65

    142.80

    m .

    The same image as in Figure17

    a.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    53/107

    52

    epth

    OPTV

    1:2

    r

    N E S W N

    PROJE TION

    N

    E

    s w N

    Figure 18a

    A closed fracture KR9, 41.70-41.90

    m)

    with sinusoid curve. There is a

    narrow curved crack trace cutting the mineral grains that does not allow any

    aperture or filling material to be observed.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    54/107

    5

    epth

    OPTV

    2

    s

    w

    4 B

    Figure 18b A closed fracture KR9,

    41

    .70-41.90 m) without sinusoid curve. The red

    arrows show the trace. There is no aperture or filling visible on the trace length. This

    does not exclude the possibility

    o

    illing observed on the surface, or a minor less than

    0.25 mm) aperture being present.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    55/107

    54

    Figure 19a The measurement principle o aperture KRJ2, 94.95 95.20 m .

    The

    reading

    is

    taken from the mid part o the trace, between the surfaces apart in the

    image.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    56/107

     

    Figure 19b The measurement principle of thickness KR12,

    74.80

    75.00 m . The

    reading is taken from the mid par t

    of

    he trace, over the filling material.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    57/107

    56

    Figure 20a Fracture form in image width scale 200 mm), planar KR9, 147.85

    148.05 m with sinusoid curve. The form follows nicely the sinusoid curve over the

    image width.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    58/107

    Depth

    :2

    148.0

    E

    57

    OPTV

    s

    w

    igure

    20b.

    Fracture form in image width scale 200 mm), planar

    KR9,

    147.85

    148. 05 m) without sinusoid curve. The same image as in Figure 20a.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    59/107

    ~

    8

    Figure 20c

    Fracture form

    in

    image width scale 200 mm), undulating KR12, 65.55-

    65.80 m with sinusoid curve. The trace form is clearly deviating from the sinusoid

    curve o plane projection.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    60/107

    59

    Depth

    OPTV

    1:2

    1 N E S W N

    6: 6

    Figure 20d Fracture form in image width scale 200 mm), undulating KR12, 65 55

    65.80

    m

    without sinusoid curve. The image is same as

    in

    Figure 20c.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    61/107

    6

    Figure 20e

    Fracture form in image width scale 200 mm), stepped fracture KR12

    61.35 61.66 m with sinusoid curve. There is clearly a step probable brittle minor

    fault) at the middle

    o

    the trace. A differently oriented minor crack seems

    to

    cut the

    fracture mapped.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    62/107

    61

    Dep h

    OPl\1

    1 2

    61 6

    Figure 20f. Fracture form in image width scale 200 mm), stepped fracture KR12

    61.35- 61.66 m without sinusoid curve.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    63/107

    6

    Figure 2

    a

    Fracture form

    in

    local, mm scale smooth KR12, 66.00 66.10 m with

    sinusoid curve. The surface

    is

    even, no deviation from planar form is seen.

    OPTV

    1:2

    E s

    w

    Figure

    2lb

    Fracture form

    in

    local, mm scale smooth KR12,

    66 00

    66.10

    m

    without

    sinusoid curve. The same image as

    in

    Figure 21

    a.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    64/107

    6

    igure 21

    c.

    Fracture form in local, mm scale rough fracture KR12, 62.15 62.35 m

    with sinusoid curve. The local f w mm variation on the surface from the planar

    projection is clearly seen.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    65/107

    64

    Depth

    PTV

    N

    E

    s

    w

    N

    igure

    2 d

    Fracture form

    in

    local, mm scale rough fracture KR12, 62.15 62.35 m)

    without sinusoid curve. The same image as in Figure

    21 c

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    66/107

    6

    Figure 22a A white fracture KRJ , 370.00- 370.20

    m

    with sinusoid curve. The filling

    material seen between the surfaces is white calcite, kaolinite?)

    in

    eo/or.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    67/107

     

    Depth

    BIPS

    N s

    w

    Figure 22b A white fracture KRJ, 370.00- 370.20

    m

    without sinusoid curve. The

    same image as in Figure 22a.

    N

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    68/107

      7

    Figure 23a

    A grey fracture

    KR2, 313.9 314.15 m)

    with sinusoid curve. The dark

    grey eo/or

    of

    illing between the surfaces is clearly seen quartz, chlorite?).

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    69/107

      epth

    1:2

    314.0

    N

    E

    68

    IP

    s

    w

    Figure 3

    b A grey fracture

    KR2,

    313 9 314 15

    m)

    without sinusoid curve The

    same image as in Figure 23a

    N

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    70/107

      9

    Figure 24a

    A location where material flows into the borehole (KR9,

    280.54

    280.82

    m) with sinusoid curve. There is a fan ofmaterial flowing upwards, tending to reach

    the high side

    of

    he borehole (north), and a thicker line reaching to the down side

    (south)

    of

    he borehole, flowing downwards. The first is probably soluble material and

    water flowing, the latter probably solid material rending to depth, or heavier water

    fraction.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    71/107

    Depth

    m:2m R)

    N

    200 8

    7

    E

    PTV

    s

    w

    Figure 24b A location where materia/flows into the borehole K.R9, 280.54 280.82

    m without sinusoid curve. The same image as

    in

    Figure 24a.

    N

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    72/107

    71

    There was no attempt to observe existence

    of

    an alteration rim around the fracture trace

    or to measure its thickness. Neither were assessed the filling properties (grain size,

    compactness, clay, carbonate, quartz) as these were considered to be best mapped from

    the core. Both can be estimated from the image to some extent, but the latter not with a

    certainty.

    4 2 The distribution of mapping results according to level of confidence

    The original mapping resulted in 7933 fractures for 3050 m, see Table 7 for

    observations in each borehole before checking.

    Table

    7

    The distribution and amount o ractures before checking

    K.Rl, KR2,

    KR4 KR9

    KR12,

    KR13,N

    KR14,

    N( ) N( ) N( ) N( )

    N( )

    (%) N( )

    Existing/ 226 (60)

    98 (27) 99 (41) 887 (71) 1797 (71) 1073 (67) 615 (40)

    probable

    Existing/ 86 (23)

    90 (25) 81 (33) 134 (11) 321 (13)

    205 (13)

    110 (7)

    possible

    New/

    14

    (4)

    101 (28) 38 (16) 104 (8) 201 (8) 155 (10) 171(11)

    probable

    New/ 52 (14) 77 (21) 26 (11) 124 (10)

    227 (9) 161 (10)

    660 (42)

    possible

    Probable 240 (63) 199 (54) 137 (56) 991 (79) 1998 (78) 1228 (77) 786(51)

    Possible 138 (37) 167 (46) 107 (44) 258 (21) 548 (22) 366 (23)

    770 (49)

    Total 378 366 244 1249 2546

    1594 1556

    Obs. length 267.5 201 276 559 751

    494 502

    m

    Observations 1.41 1.82 0.88 2.23 3.39

    3.23 3.10

    /m

    Fractures/m 1.42

    1.68 0.85

    2.24 3.62 3.77

    1.54

    (drilling)

    Initial fracture frequencies are similar to those observed in the core mapping. The host

    rock sections (outside defined fracture zones) in KRl, KR2 and KR4 indicate less

    fractures than the completely mapped other boreholes. Less fractured KR14 is an

    exception in terms

    of

    comparison with the core mapping data.

    The proportions

    of

    probable and possible features distribute interestingly. The BIP

    mapped boreholes KRl, KR2 and KR4 show fewer probable observations

    (54-

    63%)

    than the OPTV mapped KR9, KR12, KR13 (77-79%).

    An

    exception

    of

    this trend is

    KR14 (51%), which is least fractured and was mapped by a different person. Also the

    image is poor due to stick-slip interference and dirty water. Class existing/ possible

    is more frequent

    (23-

    33%) in BIP image mapping than in OPTV mapping (7-13%).

    For new fractures there is more variation in BIP image results (4-28%) than in

    OPTV (8-11 %, except new possible 42% in KR14). These may arise from the

    correctness

    of

    the depth adjustment in BIP images or core sample.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    73/107

    72

    The certainty class was used to decide which borehole sections would be checked over

    with the core sample.

    t

    appeared that the better resolution

    o

    OPTV image allows to

    observe the features with greater certainty, although the relative amount

    o

    the

    observations does not depend so much on the image source.

    n

    analysis

    o

    the different

    certainty classes it was observed that the possible observations are more often

    closed or with filling than the others. So the checking against the core sample was

    focused into existing/possible , new/possible and new/probable , leaving the

    existing/probable that was already linked to the core sample mapping for lesser

    interest as it had external confirmation from the drilling data.

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    74/107

    73

    CHECKING THE OBSERVATIONS WITH THE CORE SAMPLES

    5 1

    The checking procedures

    The observations were checked towards the core samples at the core sample archive

    of

    Geological Survey of Finland in Loppi (boreholes KRl, KR2, KR4, KR9 and KR12),

    and at Olkiluoto (KR13 and KR14). All the observations of classes 'New, Possible

    and New, Probable were checked systematically as well as most

    of

    the observations

    Existing, Possible .

    The images and results were compared with the samples. The results

    of

    the first

    checked borehole (KR2) were analyzed to allow guidance of the remaining work.

    Checking required typically a day for 300- 400 observations. The rate depended on

    fracture frequency, proportion

    of

    observations to be checked, rock type, quality

    of

    image and the accuracy

    of

    the depth adjustment. The checked fractures received status

    probable , and are indicated also with a label confirmed in the final database. The

    others have a label unchecked . Rejected not a fracture -observations have been

    removed from the final fracture lists. Influence

    of

    checking is presented in Table 8.

    5.2

    The results of the checking

    After the checking 6138 fractures remained in the data (78% of the initial amount).

    Generally some 78-88%

    of

    observations remained, except for KR4 (62%) and KR14

    (5lo/o).

    Table

    8

    Influence

    o

    checking

    Inspection

    KRl,

    KR2, KR4, KR9, KR12, KR13,

    KR14,

    N( )

    N o/o) N( )

    N( )

    N( ) N( ) N( )

    Initially

    378 366

    244 1249 2546 1594 1556

    Final

    295 (78)

    321 (88) 148 (61) 1026 (82) 2228 1342 (84)

    778 (51)

    (88)

    Checked 153 (40) 213 (58) 151 (62) 386 (31) 804 (32) 535 (34) 969 (63)

    Rejected 83 (22) 58 (16)

    96 (39) 223 (18) 317(12)

    252 (16) 770 (50)

    Rejected

    of

    54% 27% 64% 58% 39% 47% 79%

    checked

    Exist/Cert 2 (1%, 1 (0%,

    3 (3%, 13 1 %,

    15 (1%, 10

    1

    %, 27 (4%,

    rejected N=226) N=98) N=99) N= 887) N=1797) N=1073) N=604)

    Exist/ uncert

    36 (42%,

    14 (16%, 48 (59%, 74 (55%, 99 (31%, 78 (38%, 63 (57%,

    rejected N=86) N=90) N=81) N=134) N=321)

    N= 205) N=llO)

    New/cert 5 (36%,

    11 11 %,

    21

    (55%, 44 (42%, 51 (25%, 59 (38 %, 95 (34%,

    rejected N=14) N=lOl) N=38)

    N=104) N=201) N=155) N=278)

    New/uncert 40 (77%, 32 (42%,

    24 (92%, 92 (74%,

    152 105 (65%,

    585 91 %,

    rejected

    N=52) N=77)

    N=26) N=124) (67%, N=161) N=644)

    N=227)

    New,

    0

    13

    (4%)

    0 0

    1 (0%)

    0

    12 (1.5%)

    confirmed

    Frequency

    1.10

    1.60

    0.54 1.84 pes/m 2.97 2.72 1.55

    pes/m pes/m

    pes/m pes/m pes/m pes/m

    Drilling

    1.42

    1.68

    0.85 2.24 pes/m 3.62 3.77 pes/m 1.54

    frequency pes/m pes/m pes/m pes/m pes/m

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    75/107

    74

    Varying by borehole some 40-62% of the observations were checked from BIP, and

    from the checked ones 27-64% were rejected, the others confirmed. For KR9-KR13

    OPTV 31 - 34% were checked, and 39-58% of the checked ones were rejected, others

    were confirmed. In KR14 63%

    of

    observations was checked and 79% of the checked

    ones were rejected, the remaining 21% confirmed. Typically the rejected fractures

    represented rather types closed or fractures with filling . Any of the rejected

    fractures were

    of

    types open or partly open .

    The fractures classified as Existing, Probable were checked only occasionally (2-3%)

    especially when a depth error or disorientation of the core sample were suspected. A

    proportion of 0-4%

    of

    all were rejected (some one-third of the checked ones in this

    class). Table 9 shows the proportion of fractures in the certainty classes after checking.

    The properties probable and possible practically lost their meaning in checking.

    The few possible observations are related to locations where the sample has been

    missing. The greatest remaining uncertainty is found for observations initially

    belonging to the class possible , as these have not been checked as extensively as the

    others. Generally these are linked to similarly oriented core mapping fractures at the

    same depth level and usually are very clear in the image. The information has not been

    merged in this work. According to tentative observation some 1 o of all the features in

    this class would change when checked, mostly at locations with depth or orientation

    errors.

    n

    any case the core mapping and image mapping would produce slightly

    different results due to e.g. accuracy reasons.

    The proportion 6 - 14% of new fractures suggests, that there are some fractures

    unobserved in the core sample mapping. Some

    of

    these may barely indicate

    an

    inadequate depth value in core reconstruction (deviation from the image depth

    adjustment), or an uncorrected drift on image depth value.

    Table 9 Certainty classes after checking Those rejected are not counted or in the

    proportions The possible ones confirmed have been transferred to probable class

    Borehole/ KRl

    KR

    KR4

    KR9 KR12 KR13

    KR14

    Certainty (N=295) (N=321) (N=148) (N=1026) (N=2228)

    (N=1342)

    (N=778)

    Existing/ 272 148 134

    91

    %) 943 (92%) 2029 1197 662 (85%)

    Probable

    (92 ) (46%) (91 ) (89 )

    Existing/ 5 (2%)

    30 (9%) 2 (1%) 3 (0%) 3 (0.1%)

    7 (1%) 7 (1%)

    Possible

    New/ 14 (5%)

    130

    11

    (7%) 74 (7%) 190 (8.5%) 133 (10%)

    105 (13%)

    Probable (40%)

    New/ 4 (1%)

    13

    (4%) 1 (1%)

    5 (0%)

    6 (0.3%)

    3 (0%) 4 (1%)

    Possible

  • 8/16/2019 POSIVA 2002 22 Working Report Web

    76/107

    75

    RESULTS

    6 1

    Contents of the results

    n

    total 613 8 fractures were included into the final results after checking. The

    frequencies and amount represent 81%

    of

    the core sample fractures (63% - 100% for

    different boreholes). Lowest proportion is found in KR4 (63%). Reason might emerge

    from image quality or fracture type distribution, or accuracy in the image adjustment.

    Highest proportions (95% and 100%) are in KR2 (BIP) and KR14 (OPTV). The former

    was checked first and in more detail than the others. From the latter smaller scale and

    more detailed possible fractures were mapped resulting to a greater amount of rejected

    observations at the checking phase. The proportion for other boreholes is 77 - 82%.

    There are probably some 0-5% new observations in different boreholes. The new

    fractures may be drilling induced, from core loss areas (not from triple-tube drilled

    KR13 and KR14), or features that have not been reported from sample because

    of

    being closed, but can be mapped from image.

    The coverage of previous BIP interpretation (Strahle 1996) was checked for

    comparison. Mapping has covered in average 43% of the core mapping reported

    fractures, varying between 22 - 60% depending on borehole and depth interval.

    Proportions of observed fractures were less near to surface where the fracture

    frequency is highest. Mapping in 1996 did not utilize the core mapping data but was

    based on image data alone, and recorded only the most clear indications.

    6.2

    Fracture type distributions

    The fracture type distribution was analyzed. The open , partly open and with

    filling types can exist simultaneously forming five major classes whereas the sixth

    major class closed is excluding the others by definition. The class altered is met

    only in boreholes KR2 and KR13, probably reflecting local variation in bedrock

    conditions. Distribution of types in results are in Table 10 and in Figures 25 - 32.

    Table 10 Fracture type distributions

    Borehole/

    KRl KR2

    KR4 KR9

    KR 2

    KR 3

    KR 4 All

    Types

    N=295 N=321

    N=148 N=1026 N=2228 N=1338 N=778 N=6134

    Open

    0 (0%) 0 (0%) 1 (0.6%) 6 (0.6%)

    15

    (0.7%) 29 (2.2%) 5 (0.6%) 56

    (0.9%)

    Partly 0 (0%) 0 (0%) 0 (0%) 16 56 (2.5%) 45 (3.4%)

    18

    135

    open

    (1.6%)

    (2.3%) (2.2%)

    Open,

    2 (1%)

    1 (0.3%)

    2 (1.2%)

    1 (0.1 %) 2 (0.1 %) 3 (0.2%) 3 (0.4%)

    14

    with

    (0.2%)

    fillin2

    Partly 0 (0%) 3 (1%) 10

    13

    25 (1.1%) 45 3 .3%)

    19

    115

    open, with

    (6.8%)

    (1.3%) (2.4%) (1.9%)

    fillin2

    With

    181

    256

    114 646 1400 915

    536

    4048

    fillin2

    (61%)

    (80%)

    (77%) (63%) (62