A relationship between rock physics and NMR (Nuclear Magnetic Resonance) Zakir Hossain, Houston, USA

Summary

Historically nuclear magnetic resonance (NMR) is used as

a useful tool in petrophysical based reservoir evaluation.

The objective of this study is to define a relationship

between NMR T2 distribution and seismic attributes for

accurately rock properties prediction. To define rock

properties, we used laboratory NMR and ultrasonic

measured data, rock physics, and AVO analysis. From this

study, we define the following relationship:

iTR

2total

11

cutoffimicro TR ,2

cuttoffiendiPFM TTRV ,2,2

cuttoffiendi

gTTR

S,2,2

111

where, total is the total porosity, and are the Lame`

parameters related with P-impedance and S-impedance, T2i

is the NMR T2 distribution of each pore, VPFM is the

volume of pore-filling mineral (PFM), Sg is the gas

saturation, R is the or S-reflection coefficient, R

is the (reflection coefficient. This study shows

that NMR T2 distributions are directly related with seismic

attributes. Therefore, integrating NMR data with rock

physics analysis provides a solid basis for quantitative

seismic petrophysical interpretation by minimizing

interpretation risk.

Introduction

NMR is a useful tool to measure in-situ reservoir

properties. However, historically nuclear magnetic

resonance (NMR) is used for fundamental petrophysical

properties prediction including porosity, permeability,

irreducible water saturation, capillary pressure (Howard, et

al. 1993; Kenyon et al. 1995; Kenyon 1997; Hossain et al.

2011a). Recently, Hossain et al. 2011b and 2011c showed

that NMR can be used as a potential tool to understand the

fluid flow distribution and fluid related dispersion. They

described that Biot’s flow occurs only in large pores in

complex rocks while, Biot’s flow should not occur in

micro-pores. Differences of fluid flow in macro-pores and

micro-pores pores are described as the high frequency

squirt flow in complex rocks. Thus, NMR analysis helps us

to understand and quantify the different pores,

heterogeneous of pore types and their distribution, and

changing pore fluids. In contrast, rock physics analysis

helps us to understand and quantify the different

lithologies, changing pore fluids, heterogeneous of pore

types and their distribution, and elastic properties in

general. Therefore, integrating NMR data with rock physics

analysis provides a solid basis for quantitative seismic

petrophysical interpretation. The objective of this study is

to define a relationship between NMR measurement and

rock physics measurement for rock properties prediction.

Method

We used laboratory measured NMR and ultrasonic P-and

S-wave velocities measured data on brine saturated

greensand samples. All data used for this study were

published by Hossain (2011). Data representing the CO2

bearing state were calculated by using Gassmann’s

equations (Gassmann, 1951). The CO2 properties as a

function of temperature and pressure were derived based on

data from Wang et al. (2010), and brine properties were

calculated from equations of Batzle and Wang (1992)

In addition, rock physic and AVO modeling were done to

predict rock properties from ultrasonic measurement. To

predict rock properties from sonic data, we generated an

RPT (Hossain et al., 2015) which combined multiple

Figure 1: NMR measurement on fully saturated sample is

compared to the NMR measurement after centrifuging at 100 psi.

The cutoff time, which separates the T2 distribution into macro-

porosity and micro-porosity is defined as the relaxation time at the point where the cumulative porosity of the fully saturated sample

equals the irreducible water saturation. The dashed vertical line is

shown a cutoff of 5.21ms. High total porosity is a function of high

cumulative T2i low ,and low R; high micro-porosity is a

function of high cumulative T2,cutoff, high and high R,; high

volume of pore filling mineral (PFM) is function of high

high R and slow T2; high gas saturation is function of

low , low R, and fast T2. (Figure modified after Hossain

et al. 2011a).

Page 3523© 2016 SEG SEG International Exposition and 86th Annual Meeting

Dow

nloa

ded

11/1

7/16

to 2

04.2

7.21

3.16

2. R

edis

trib

utio

n su

bjec

t to

SEG

lice

nse

or c

opyr

ight

; see

Ter

ms

of U

se a

t http

://lib

rary

.seg

.org

/

A relationship between rock physics and NMR

attributes in the Ip-Vp/Vs space to describe the various rock

properties from seismic data. For AVO analysis, we

generated an RPT in the intercept-gradient space. To

generate an RPT in the intercept-gradient space, constant

shear-reflection coefficient curves (Rs) were calculated

based on the following relationship (Wiggins et at. 1983):

SP RRG 2 (1)

where, RP is the P-reflection coefficient or intercept, RS is

the S-reflection coefficient which is equivalent to R, and

G is the gradient.

We used following relationships to calculate -reflection

coefficient and (reflection coefficient:

12

12

R (2)

)()(

)()(

1122

1122

R (3)

where, and (are related with lp and ls. and

(represent cap rock properties, whereas and

(represent reservoir rock properties.

Equations (1)-(3) were used to generate constant

reflection coefficient curves as well as constant

(-reflection coefficient curves in the intercept-

gradient space (Figure 4).

Initially, intercept and gradient were calculated for brine

saturated samples and shale interface. Then intercept

gradient were calculated for CO2 saturated samples and

shale interface. Intercept and gradient were calculated

based on Castagna and Smith (1994). The shale represents

the cap-rock for the greensand. Shale data for AVO curves

were obtained from the studied Nini 1A well (Hossain et al.

2012).

Results

The NMR T2 distributions are presented in graphical form

for each sample (Figure 1 and Figure 2). All greensand

have bimodal T2 distributions. Each T2 time corresponds to

a particular pore size. For the present greensand samples, a

peak close to 1 ms should correspond to glauconite water,

whereas all samples also present a second peak close to 100

ms that corresponds to movable fluid (Hossain et al.,

Figure 2: Geological properties defined from NMR measurements. (a) and (c) BSE images and conceptual models of two types of greensand

from the North Sea. Scale bar of these images is 200 m and the images represent macro-porosity, quartz and glauconite grains and micro-

porosity within glauconite. (a) Weakly cemented greensand (c) Micro crystalline quartz and pore-filling berthierine cemented greensand (Images

and rock model after Hossain et al., 2011). (b) NMR T2 distributions are presented in graphical form for weakly cemented and cemented samples. It is noticeable that weakly cemented samples show larger amplitude in the movable fluid than cemented samples; whereas highly diagenetically

altered samples show slightly larger amplitude in glauconite water (NMR data from Hossain, 2011).

Page 3524© 2016 SEG SEG International Exposition and 86th Annual Meeting

Dow

nloa

ded

11/1

7/16

to 2

04.2

7.21

3.16

2. R

edis

trib

utio

n su

bjec

t to

SEG

lice

nse

or c

opyr

ight

; see

Ter

ms

of U

se a

t http

://lib

rary

.seg

.org

/

A relationship between rock physics and NMR

2011a). Three factors control the rock properties defined

from NMR T2 distributions: amplitude of micro-pores,

amplitude of macro-pores and T2. Comparing

Backscattered Electron (BSE) images with NMR

measurement, it is noticeable that clean sample show larger

amplitude in the movable fluid; whereas highly

diagenetically altered sample show slightly larger

amplitude in glauconite water. Therefore, amplitude of

micro-pores can be described as a function of amount of

glauconite grains, amplitude of macro-pores can be

described as a function of effective porosity (Hossain,

2011). Furthermore, it is also noticeable that highly

diagenetically altered sample show longer T2 than clean

sample. Therefore, longer T2 can be described as a function

of pore-filling mineral. Moreover, NMR measurement on

fully saturated sample is compared to the NMR

measurement after centrifuging at 100 psi to define that

fully brine saturated sample has longer T2 than faster T2 of

partially air saturated sample.

Rock properties from acoustic measurements using an

RPT

Generated RPT was used to define rock properties as

function of seismic attributes from ultrasonic measured

data (Figure 2b). Porosity of studied samples can be

described as a function of ls(). For example, porosity

ranges from 34.2 to 37.4 can be modeled by constant ls

ranges from 2.2 to 1.5 and porosity ranges from 29.4 to

30.1 can be modeled by constant ls ranges from 4 to 3.

Matrix supported glauconite grains of studied samples can

be described as a function of constant . For example,

sample with the highest amount glauconite grain (24.82)

can be modeled by constant=25, sample with the

intermediate amount glauconite grain (21.45) can be

modeled by constant =16, and sample with the lowest

amount glauconite grain (19.94) can be modeled by

constant =11. As micro-porosity is proportional to the

amount of glauconite grains, therefore can be also used

to describe micro-porosity of studied samples. Pore-filling

minerals of studied samples can be described as a function

of constant For example, sample with the highest

Figure 3: Rock properties analysis by combining NMR and RPT

(rock physics template). (a) NMR T2 distribution in porosity

units (p.u.), (b) RPT in the lp-Vp/Vs space (NMR and brine

saturated sonic measured data from Hossain, 2011). CO2 saturated data with red symbols were calculated using Gassmann

method. In the RPT porosity was modeled by constant ; matrix supported glauconite grains or micro-porosity were

modeled by constant , diagenesis features including pore-

filling minerals (PFM) were modeled by constant and

calculated CO2 saturated data were also modeled by constant

. Rock properties defined from NMR data are well agreed

with the rock properties defined from the RPT analysis.

Figure 4: RPT in the intercept-gradient space was used to

describe rock and fluid properties in seismic scale (sonic

measured data from Hossain, 2011). In the RPT porosity was

modeled by constant R; matrix supported glauconite grains or

micro-porosity were modeled by constant R, diagenesis features

including pore-filling minerals (PFM) were modeled by constant

R and calculated CO2 saturated data was also modeled by

constant R.

Page 3525© 2016 SEG SEG International Exposition and 86th Annual Meeting

Dow

nloa

ded

11/1

7/16

to 2

04.2

7.21

3.16

2. R

edis

trib

utio

n su

bjec

t to

SEG

lice

nse

or c

opyr

ight

; see

Ter

ms

of U

se a

t http

://lib

rary

.seg

.org

/

A relationship between rock physics and NMR

amount pore-filling mineral can be modeled by

constant =13, and sample with the lowest amount

pore-filling mineral can be modeled by constant

=7.7. Calculated CO2 saturated data can be modeled by

constant ranges from 3 to -8.

Rock properties from combined NMR and acoustic

measurements

In NMR measurement, amplitude of micro-pores

corresponds to porous grains and cumulative T2,cutoff

corresponds to micro-porosity (Figure 3a), whereas in the

RPT, describes the micro-porosity within porous grains

(Figure 2b). Therefore, cumulative T2,cutoff should

correspond to . Similarly, cumulative T2i corresponds to

total porosity (Figure 2a), whereas describes the total

porosity in the RPT (Figure 2b). Therefore, cumulative T2i

should correspond to 1/. Moreover, longer T2

corresponds to the PFM (Figure 2b), whereas

describes PFM in the RPT Therefore, longer T2

should correspond to high . Finally, fast T2

corresponds to the gas or air saturation (Figure 1), whereas

describes pore-fluids in the RPT Therefore, fast T2

should correspond to low .

Rock properties from AVO analysis

In the template in Figure 3b, porosity of studied samples

was described by , matrix supported glauconite grain or

micro-porosity was described by , pore-filling mineral

was by , CO2 saturated data was described by

. Likewise, in the template in Figure 4, R

describes porosity, R describes micro-porosity, R

describes PFM, and R describes CO2 saturated rock

properties. CO2 saturation can be clearly defined using

R. CO2 saturated greensand can be classified as the

AVO classes II and IV. Higher porosity and lower

diagenesis bearing samples show AVO class IV, whereas

lower porosity and higher diagenesis bearing samples show

AVO class II. Table 1 shows relationship we established

from rock physics and AVO analysis.

Conclusions

We showed how NMR measurements are related with

ultrasonic measurements as function of rock properties. To

define rock properties, we used laboratory NMR measured

data, ultrasonic measured data, rock physics template

(RPT) analysis, and AVO analysis. Our study shows that

that total porosity is a function seismic attribute, , AVO

attributes, Rs, and cumulative T2i e.g. high total porosity,

low low Rs, and large cumulative T2i. Similarly, we

found micro-porosity is a function seismic attribute, ,

AVO attributes, R and cumulative T2,cutoff e.g. high

micro-porosity, high high R and large cumulative

T2,cutoff. Moreover, we found pore-filling mineral (PFM) is a

function seismic attribute, , AVO attributes, R

and T2 e.g. high PFM, high high R and longer

T2. Finally, we described pore-fluid is a function seismic

attribute, , AVO attributes, R and T2 e.g. high

CO2 saturation, low low R and fast T2. This

study demonstrates that NMR measurement is a potential

tool for rock physics properties prediction in seismic

petrophysics based reservoir characterization. Therefore,

integrating NMR data with rock physics measurements can

successfully be used to predict the accurate rock properties:

total porosity, micro-porosity, pore-filling mineral and pore

fluid for reservoir evaluation.

Acknowledgments

All laboratory experiment results used for this study were

published in the PhD Thesis, Technical University of

Denmark.

Table 2: Rock properties prediction by combining NMR and RPT

Page 3526© 2016 SEG SEG International Exposition and 86th Annual Meeting

Dow

nloa

ded

11/1

7/16

to 2

04.2

7.21

3.16

2. R

edis

trib

utio

n su

bjec

t to

SEG

lice

nse

or c

opyr

ight

; see

Ter

ms

of U

se a

t http

://lib

rary

.seg

.org

/

EDITED REFERENCES Note: This reference list is a copyedited version of the reference list submitted by the author. Reference lists for the 2016

SEG Technical Program Expanded Abstracts have been copyedited so that references provided with the online metadata for each paper will achieve a high degree of linking to cited sources that appear on the Web.

REFERENCES Batzle, M., and Z. Wang, 1992, Seismic properties of pore fluids: Geophysics, 57, 1396–1408,

http://dx.doi.org/10.1190/1.1443207. Castagna, J. P., and S. W. Smith, 1994, Comparison of AVO indicators: A modeling study. Geophysics,

59, 1849–1855, http://dx.doi.org/10.1190/1.1443572. Gassmann, F., 1951, Elastic waves through a packing of spheres. Geophysics, 16, 673–685,

http://dx.doi.org/10.1190/1.1437859. Hossain, Z., 2011, Rock Physics modelling of the North Sea greensand: Ph.D thesis, Technical University

of Denmark. Hossain, Z., I. L. Fabricius, A. C. Grattoni, and M. Solymar, 2011a, Petrophysical properties of greensand

as predicted from NMR measurements. Petroleum Geoscience, 7, 211–225, http://dx.doi.org/10.3997/2214-4609.201401334.

Hossain, Z., T. Mukerji, and I. L. Fabricius, 2011b, Influence of pore fluid and frequency on elastic properties of greensand as interpreted using NMR data. 81st Annual International Meeting, SEG, Expanded Abstracts.

Hossain, Z., T. Mukerji, and I. L. Fabricius, 2011c, Biot’s and squirt flow mechanism of greensand as interpreted using NMR: Extended Abstract, 1IWRP, August 7–12, Colorado School of Mines.

Hossain, Z., T. Mukerji, and I. L. Fabricius, 2012, Vp–Vs relationship of glauconitic greensand. Geophysical Prospecting, 60, 117–137, http://dx.doi.org/10.1111/j.1365-2478.2011.00968.x.

Hossain, Z., S. Volterrani, F. Diaz, and P. Constance, 2015, Integration of rock physics template to improve Bayes’ facies classification: 85th Annual International Meeting, SEG, Expanded Abstracts, 2760–2764.

Howard, J. J., W. E. Kenyon, and C. Straley, 1993, Proton magnetic resonance and pore size variations in reservoir sandstones. SPE Formation Evaluation, 8, 194–200, http://dx.doi.org/10.2118/20600-PA.

Kenyon, W. E., 1997, Petrophysical principles of applications of NMR logging. The Log Analyst, 38, 21–43.

Kenyon, B., R. Kleinberg, C. Straley and C. Morriss, 1995, Nuclear magnetic resonance imaging — Technology for the 21st Century: Oilfield Review, 7, 19–30.

Wang, Z., M. Sun, and M. Batzle, 2010, CO2 velocity measurement and models for temperatures up to 200 °C and pressure up to 100 MPa. Geophysics, 75, no. 3, 123–129, http://dx.doi.org/10.1190/1.3383324.

Wiggins, R., G. S. Kenny, and C. D. McClure, 1983, A method for determining and displaying the shear-velocity reflectivities of a geologic formation: European Patent Application 0113944.

Page 3527© 2016 SEG SEG International Exposition and 86th Annual Meeting

Dow

nloa

ded

11/1

7/16

to 2

04.2

7.21

3.16

2. R

edis

trib

utio

n su

bjec

t to

SEG

lice

nse

or c

opyr

ight

; see

Ter

ms

of U

se a

t http

://lib

rary

.seg

.org

/