Download - ESTIMATION OF PETROPHYSICAL PROPERTIES INTEGRATING ROCK … · permeability), thin sections (for rock-fabric typing), a cycle-based stratigraphic framework (for rock-fabric distribution),

Even-texturedpore space

500 mm

Even-texturedcarbonate cement

Carbonatecement

500 mm

QAd4360(a)x

7100

7200

7300

3

Sequence L1

HFS L2.0,Sequence L2

1 2 3 4

ARFN0 100

30 0Depth(ft)

0

1000 0.1

1000 0.130

GR (API)

Core porosity(%)

Wireline logporosity

(%)

Core permeability(md)

Rock-fabricpermeability

(md)

Excellent match between measured core permeability(red) and permeability estimated using apparent rock-fabric numbers and total porosity (black).

Capillary Pressure Models for Estimation of Original Water Saturation

Discrepancies betweenpredicted permeability and corepermeability may be due totouching-vug pore systemassociated with karst breccia.

Rock-fabric-specific capillary pressuremodels were developed by Lucia (1995,1999). New data acquired in this studycompare well with the existing Class 1model. However, new data from ARFN 2and 3 samples do not fit previous models.We developed new models for ARFN 2 and3 rocks using the Thomeer method.

Saturation was calculated for the reservoirusing these capillary pressure models,wireline log porosity, ARFN framework, andreservoir column height.

Overall excellent match supports use of capillary pressure model. Areas of poor matchare the result of inaccurate resistivity tool response in intervals of low porosity.

When available wireline logs are insufficient to distinguishinterparticle porosity from vuggy porosity, total porositymust be used. Rock-fabric numbers defined by thisapproach (termed “Apparent Rock-Fabric Numbers”[ARFN]) are less accurate and may overstate permeabilityand saturation if vuggy porosity is present. Because mostwells at Fullerton lack acoustic logs and usable resistivitylogs (making it impossible to define interparticle porosityand vuggy porosity), the Apparent Rock-Fabric Number(ARFN) system was employed.

We think the most accurate calculations of permeabilityand original water saturation are derived from good- qualitywireline porosity data using transforms defined from rock-fabric relationships. Requirements of this technique aregood-quality core analyses (to define porosity andpermeability), thin sections (for rock-fabric typing), a cycle-based stratigraphic framework (for rock-fabric distribution),and complete log suites (including acoustic logs).

Rock-Fabric Technique

Water Saturation Model Comparison:Capillary Pressure vs. Log Resistivity

Lower Clear Fork L2.2 Rock Fabrics/Petrophysics

Lower Clear Fork L2.1 Rock Fabrics/Petrophysics

Saturation ModelingPermeability Estimation

ESTIMATION OF PETROPHYSICAL PROPERTIES INTEGRATING ROCK-FABRIC APPROACH WITH STRATIGRAPHY AND WIRELINE POROSITY

Porosity Log Calibration Rock Fabric Distribution

Following conventional log normalization andcalibration, porosity logs were spatiallynormalized to minimize individual well acquisitionand calibration errors. Maps below illustrate theeffects of spatial normalization of sidewallneutron log values using various search radiifor contouring average porosity data of entirereservoir interval for all wells.

Differences between normalized values andactual well values were used to adjust log dataover the entire reservoir interval.

ARFN-based Permeability Estimationvs. Core Data

Stratigraphic DistributionData Quality

Spatial Normalization of Porosity Logs

The rock-fabric method of petrophysical characterizationis based on relationships that exist between pore type,pore size, particle size, and sorting. Once rock fabrics areidentified, they can be assigned to petrophysical classest h a t h a v e r o c k - f a b r i c - s p e c i f i c p o r o s i t y /permeability transforms (rock-fabric numbers [RFN]).

The direct relationship between interparticle porosity andpermeability (Lucia, 1995, 1999) allows permeability to beaccurately calculated when interparticle porosity and rock-fabric number (RFN) can be defined (below).

Core analysis data commonly contain suspect data, asevidenced by low-porosity/high-permeability data points(below). Many of these data are the result of poor samplecleaning. For this study we obtained new plugs that werecarefully cleaned and analyzed. Note the difference in thetwo data sets from the same core.

3000-ft search radius: provides detail andreduces random error. This was used tonormalize raw wireline log porosity datain the field.

5,000-ft search radius: masks too much small-scale variability

1000-ft search radius: too many bull’s eyes (wellacquisition problems or calibration errors)

Rock-Fabric Method Based on Total Porosity

log k = [9.7982 – (12.0838 x log RFN)] + [8.6711 – (8.2965 x log RFN)] x log φIP

Rock-Fabric Porosity/Permeability Relationship

0.05 0.10 0.20 0.30 0.400.1

1.0

10

100

1000

Interparticle Porosity (fraction)

1.0

2.0

3.0

QAd2746c

Perm

eabi

lity

(md)

Class 3Field

Class 1Field

0.5

1.5

2.5

4.0

Rock Fabric Numbers

Class 2Field

Rock-Fabric Method Based on Interparticle Porosity 0.01

0.1

1

10

100

1000

0.01 0.1

Perm

eabi

lity

(md)

Total porosity (fraction)

Few suspectdata

New Core Analysis

0.01

0.1

1

10

100

1000

0.01 0.1Total porosity (fraction)

Perm

eabi

lity

(md)

Suspect data

Old Core Analysis

Estimation of Permeability and Original Water Saturation

0.08

0.08

0.06

0.06

0.04

CI = 1%SNP porosity

0 4000 ft0 1000 m

UNIVERSITY LANDS BLOCK 13

PSL BLOCK A-48

PSL BLOCK A-37

PSL BLOCK A-31

PSL BLOCK A-26

Clear Fork Unit

PSL BLOCK A-47

PSL BLOCK A-32

0.070

0.065

0 . 06 0

CI = 0.5%SNP porosity

0.065

0.060

0.070

CI = 0.5%SNP porosity

Nonreservoir

ARFN

GR (API)

0 100

Depth(ft)


(%)30 0


(md)1000 0.1

L

ower

Cle

ar F

ork

Wic

hita

1 2 3

L2.3

L2.2

L2.1

Rock-Fabric/PetrophysicalClassification

StratigraphicSequence

L1

L2.0

6800

6900

7100

7200

7300

ARFN 3

ARFN 1in most areas

ARFN 2–2.5in limestone areas

ARFN 3

ARFN 2in most areas

ARFN 1in some areas

7000

Type log showing general stratigraphic position of apparentrock-fabric numbers.

Subt

idal

Perit

idal

Subt

idal

Perit

idal

Perit

idal

HFS L 2.2 is dominated by petrophysical Class 2 medium-crystalline anhydritic dolostone. Because of the differential effect of patchy, poikilotopic anhydriteon porosity and permeability, these rocks generally plot in the Class 1 field and are characterized by an ARFN of 1. Less common moldic grainstones andgrain-dominated packstones are best represented by an ARFN of 2 to 2.5.

1 mm

Grain-dominated Packstone

ARFN1:Medium-crystalline Dolostones with

Poikilotopic Anhydrite ARFN 2- 2.5 limestones

Note tight clustering of data alongARFN 2.5 transform.

Grainstone

Total porosity (fraction)0.1 0.1

Perm

eabi

lity

(md)

0.1

1

10

100

1000ARFN 2.5

TRANSFORM

Spatial Distribution

Perm

eabi

lity

(md)


ARFN1 TRANSFORM

0.1

1

10

100

1000

0.01 0.1

Cored well

New core analysisand thin sections

ARFN 2Limestone

ARFN 1 Dolostone

ARFN 2.5Limestone

1 mm

1 mm

1 mm0 4000 ft

1000 m0

Lower Clear Fork L2.2 Rock-Fabric 1 Dolostone

6800

1

GR (API)

Core porosity(%)


(%)



(md)1 2 3 4

ARFN0 100

30 0Depth(ft)

0

1000 0.1

1000 0.130

Lower Clear Fork L2.1

Wichita (L1 and L2.0)

Lower Clear Fork L2.2 Rock-Fabric 2.5 Limestone

68502.5

GR (API)

Core porosity(%)


(%)



(md)1 2 3 4

ARFN0 100

30 0Depth(ft)

0

1000 0.1

1000 0.130

6900

7000

HFS L2.1peritidal

HFS L2.1subtidal

3

2

1 2 3 4

ARFN0 100

30 0Depth

(ft)0

1000 0.1

1000 0.130

GR (API)

Core porosity(%)


(%)



(md)

ARFN 3TRANSFORM

0.01

0.1

1

10

100

1000


Perm

eabi

lity

(md)

Tidal-flat facies

Fine-crystallinedolostone

Medium-crystallinedolostone

New Wichita Data

Typical Class 3 Fine-crystalline Dolowackestone Fabrics

Thin-section descriptions are mostly Class 3 fine-crystalline wackestoneswith little vuggy porosity. Therefore, most of the Wichita can be chararcterizedusing a rock-fabric number and an ARFN of 3. The rare class 2 grain-dominated dolopackstone and vuggy tidal-flat fabrics in the Wichita will notbe properly classified using this method (see crossplot below).

Wichita (L1, L2.0) Rock-Fabrics/Petrophysics

Fenestral pores

1 mm1 mm

1 mm


0.1

Subtidal Cycles

Intrasequence Variability: Lower Clear Fork L2.1 Spatial Trends within Subtidal Cycles of L2.1

ARFN 2 Limestone and Dolostone

ARFN 1 Dolostone

ARFN 2 Dolostone

ARFN 1 Dolostone

ARFN 2 Limestone

ARFN 2 Limestoneand Dolostone

Grain-dominated Packstone

Medium-crystalline Dolostone

1 mm

Medium-crystalline Dolostonewith Poikilotopic Anhydrite

Moldic Grainstone

Subtidal facies consist of complexassemblages of Class 2 dolostonesand limestones and Class 1dolostones

ARFN 3:Fine-crystalline Dolostone

Sequence-top peritidal facies consistof Class 3 dolostones. These rockshave limited vuggy porosity and arecharacterized by an ARFN of 3.

Transgressive (subtidal) and highstand (peritidal) legs ofhigh-frequency sequences contain different rock fabricsand thus require different porosity/permeability transforms.

Class 2 dolostones withabundant anhydrite behave asARFN 1 rocks because of theirpatchy porosity (see L 2.2above).

ARFN 1 Dolostone

This group includes Class2 dolostones andlimestones having littlevuggy porosity and Class1 moldic limestones thathave significant vuggyporosity.

Peritidal Cycles

LimestonesDolostones

HFS

L2.

1

GR (API)0 100

Depth(ft)

Wireline logporosity (%)

030

6900

7000

Perit

idal

Subt

idal

ARFN 1 TRANSFORM

Perm

eabi

lity

(md)

0.01

0.1

1

10

100

1000


ARFN 2 TRANSFORM

0.1

1

10

100

1000

0.01

Perm

eabi

lity

(md)

Class 2

0.1

1

10

100

1000


Perm

eabi

lity

(md)

0.01 0.1

ARFN 3 TRANSFORM

1 mm

1 mm1 mm

1 mm

1 mm

0

500

1000

1500

2000

0 20 40 60 80 100

25%20%15%

10%05%

Inje

ctio

n Pr

essu

re (p

sia)

Air Saturation (%)

ARFN 2 (New Thomeer Model)

Sw(Class 2) = 1-{2.71828^(-0.2/(Log(15.5064*H/Phi^(-2.8394))))}

Gamma Ray

GAPI0 150

6900

7000

7100

7200

Depth(ft)

Porosity

V/V0.4 V/V1 0

Water saturation frominverted electric logs

V/V1 0

0

Water saturation fromcapillary pressure model

Wic

hita

Perit

idal

Low

er C

lear

For

k

Perit

idal

Subt

idal

L 2.

0

Subt

idal

L 2.

1L

2.2

Pore space Grains Dolomite crystals

Rock-Fabric Petrophysical Classes

Wackestone

Note: bar is 100 mm

Crystal size>100 mm

Dolomite

Crystal size<100 mm

Crystal size>100 mm

Limestone

Limestone

Crystal size20–100 mm

Dolomite

Crystal size<20 mm

PackstoneGrainstone Packstone MudstoneMUD-DOMINATED FABRICGRAIN-DOMINATED FABRIC

QA15772(a)cx

Class 1Class 2

Class 3

Effect of Patchy Cement onPetrophysical Class

Interparticle porosity(percent)

AB

10,000

1000

100

10

110 20 30 40

Perm

eabi

lity

(md)

Class 1

Class 2

Class 3

Equalpore-throat

size

Incr

easi

ng p

ore-

thro

at s

ize

C

Lime mud

ESTIMATION OF PETROPHYSICAL PROPERTIES INTEGRATING ROCK-FABRIC APPROACH WITH STRATIGRAPHY AND WIRELINE POROSITY

Poikilotopicsulfate

Patchyanhydrite cement

500 mm

Sw(Class 1) = 0.02219 x H^-0.316 x f^-1.745

0

500

1000

1500

2000

0 20 40 60 80 100

Phi = 20%

Phi = 10%

Phi = 5%

Inje

ctio

n Pr

essu

re (p

sia)

Air Saturation (%)

ARFN 1 (Lucia, 1995, 1999)

0

500

1000

1500

2000

0 20 40 60 80 100

Inje

ctio

n Pr

essu

re (p

sia)

Air Saturation (%)Sw(Class 3) = 1-{2.71828^(-0.1/(Log(0.3827*H/Phi^(-1.9717))))}

ARFN 3 (New Thomeer Model)

25%20%

15%10%

5%

0.05 0.10 0.20 0.30 0.400.1

1.0

10

100

1000

Total Porosity (fraction)

1.0

2.0

3.0

Perm

eabi

lity

(md)

Class 3Field

Class 1Field

0.5

1.5

2.5

4.0Class 2

Field

Apparent Rock Fabric Numbers

BureauofEconomic

Geology

Cored well

New core analysisand thin sections

0 4000 ft1000 m0