Fracture Height

8/18/2019 Fracture Height

http://slidepdf.com/reader/full/fracture-height 1/11

S

Society o f

Petro eun ngineers

SPE 21833

Estimating Fracture Height From Gamma Ray Spectroscopy of

Radioactive Tracers: A Case Study

G.B. Willis, Halliburton Logging Services

SPE Member

Copyright 1991, Society of Petroleum Engineers, Inc.

This paper was prepared for presentation at the Rocky Mountain Regional Meeting and Low·Permeability Reservoirs Symposium held in Denver, Colorado, April 15-17, 1991.

This paper was selected

lor

presentation by an SPE Program Committee following review of information contained in

n

abstract submitted by the author s . Contents of the paper,

as presented, have not been reviewed by the Society of Petroleum Engineers and are sUbject to correction by the author s . The material, as presented, does not necessarily reflect

any position of the Societyof Petroleum Engineers, its officers, or members. Papers presented at SPE meetings are subjectto publication review by Editorial Committees of the Society

of Petroleum Engineers. Permission to copy is restrictedto n abstract of notmore than 300 words. Illustrations may not e copied. Theabstract should contain conspicuous acknowledgment

of where and by whom the paper is presented. Write Publications Manager, SPE, P.O. Box 833836, Richardson, TX 75083-3836 U.S.A. Telex, 730989 SPEDAL.

ABSTRACT

Five wells in the Big Horn

Basin of

Wyoming

were perforated

in

the

Phosphoria

formation

and

subsequently treated with

acid. Each

of

the acids

was t agged with a radioactive

is otope. Spectra l gamma ray logs were then

run to

determine

the

effectiveness of the

tr ea tmen ts . A fte r

processing and analyzing

the data, two facts became

evident:

1)

Several

pit l ls

need

to

be avoided

in

selecting

the

isotope and designing how

the

acid is tagged. For example ,

the

correct

isotope must be

selected with

the

correct

specific gravity

to allow

the isotope

to

stay suspended

in

the treatment fluid and

the borehole fluid must be displaced with

non-radioacti

ve fluid. 2) Looking at

total gamma ray counts alone to determine

isotope distribution

is

inconclusive and

sometimes misleading

because

this

does

not

distinguish between

tracer material in the

borehole from tracer material in

the

formation.

When gamma rays pass through any median

such

as

formation,

cement,

and casing,

Compton downscattering

occurs. l By

app

lying a

weighted

least squares WLS)

al

gorithm to

the gamma

spectroscopy data,

the

compton downscattering was taken into

References and i l lustrat ions

a t

end

of

paper.

275

account

to determine both borehole and fo

mation components.

The geology of the formation, treatment

design,

isotope design

and

isotope

se

ection are discussed

as they apply to

t

log

data.

The logging procedure, pro

cessing

of the log data, and log examp

are presented in more detail as this

is

t

main thrust of this paper.

CKGROUND

All five of the wells in this

study

are

the Cottonwood Creek Field located in t

southern end of Wyoming s Big Horn Basi

The formation treated is a late Permian a

carbonate called the Phosphoria. The Ph

sphoria

has intercrystall ine, moldic, a

vuggy porosity

reaching 30 ,

though 8-10

porosi

ty

with 1 md. of

permeability

typical. 2 Trapping mechanisms for t

hydrocarbons

are

s tr a tigraph ic, s tructur

and

combination

of the

two.

Acid

treatments

have been done in

t

Phosphoria for a number o f years

with

mix

results . The

tracer

program describ

below was run on th ese wel ls to determi

i

the

effectiveness of

these

frac jobs

due to

the acid

treatment

staying

in zon

or other

factors.

All five wells in th

project were

treated

with 120,000

150,000

gallons

of 15-20 HCL

acid.



ESTIM TING FR CTURE HEIGHT FROM

MM R Y

SPECTROSCOPY OF

R DIO CTIVE

TR CERS A

C SE

STUDY

SPE

2183

Three of

the

acids

were

tagged with

iridium.

The

other

two were

tagged

with

scandium.

Iridium

192 has major gamma

ray

energy peaks a t

311,468

and 603 Kev., while

scandium

has

ene rgy peaks

a t

809 and 1121

Kev. Scandium was used

in part in

an

effort

to

see deeper in the formation since the

higher energy gamma rays are

more

pen-

etrating than

those

from iridium.

Iridium

and scandium

are

metallic chemicals

which gives them a chemical affinity

for

deposition

on

th e

casing.

To avoid this

problem, the iso topes

were

distributed

in

a

40

mesh

particulate in the

form

of glass

beads.

The

specific gravity of

the

beads

con ta in ing the i so topes

was

1.5

G/cc. This

compared

to the acids

used

of

approxi-

mately

1.1

G/cc, depending on

the acid

concentration

and

temperature.

The

specific

gravi ty o f the

beads

containing the isotope

was

minimized in order

to optimize

the

mixing

with

the acid

and reduce

set t l ing.

Ideally i t

is

des irable to

have

an isotope

wi

th the

same density

as the

frac fluid.

This is d iff icu lt to achieve because

the

carrier

of

the isotope

must be chemically

and

physically

stable with

respect

to

the

conditions.

Using a

relat ively

low

density

carrier for the

isotope

proved

to be an

important

part

of the

t racer

design

which

is discussed

further

in the examples.

LOGGING PROCEDURES ND

PRESENTATIONS

A

before su rvey base spectral log

was

not

run

on

these wells for both

economic and

technical reasons. Since each well

was

being tagged with only

one

radioactive

isotope

and

i t has been

well documented

that

a single

t racer isotope can effec-

tively be deconvolved

from

the

natural

gamma ray

energy spectrum by

the use of the

weighted

least .

squares

Wt.S

l,3,4)algo-

rithm, the base log

was

not necessa ry .

A 1 11/16 TracerScan tool was

chosen over

a

large diameter tool, since there

was a

requirement

to go

through

2

7/8

tubing.

Five

passes

were made over

the zones

of

interest

on

l l five wells to

minimize

s t t i s t ic l

errors and improve

log quality.

Figure

1 shows

overlays

of ll five passes

on two

of the

wells.

One well with

276

scandium and one

with iridium

were

cho

for the

example. These examples show g

repeatability, but also show that

a sin

pass

has too

many s t t i s t ic l

fluctuati

to

be

used

alone.

This clearly demonstra

the need f or mul tip le passes

and

data

eraging. From th ese fiv e passes compo

data

were

compiled

and

used for fi

analysis.

Composi

te

log data presented

on

the

r

time

log

and

in figure

1

included

the

f

lowing

curves:

1. Total T race r

Curve:

This

curve

gives

a

calculated tota

tracer

concentration

as

result

of

weighted

least squares algorithm

W

processing.

1 This gives

a very cl

indication

of the

vertical

distribut

of the iso tope.

f multiple

t race

had

been used,

a

total tracer

curv

would

have

been presented for each

isotope.

2. Relative

Distance Curve: 4

By comparing

the

count rate

in

a h

energy range 350 Kev. to 3000 Kev

relat ive

to the

low

energy coun t

r

150 Kev. to 350 Kev.),

the

effects

Compton downscattering

can

be obser

and a Compton

rat io

Rc

determined.

gamma

rays

emanating from

the

trac

material pass through

any medium s

as

formation, cement and pipe),

th

more

intervening material

present,

more downscattering

t akes p lace , he

affecting

a computed Compton

ratio

From this data

the

relative distan

between

the

source of the radiation

the

detector

can

be determined and

presented as

a

relat ive

distance

cur

For ease

of interpretation,

the

cas

value

is represented

in the

log prese

ation by

the l ines defining the de

track.

This

makes

i t easy

to

visual

whether

the

isotope is inside or outside

the pipe.

3. Estimated

Gamma

Ray:



833

G

WILLIS

3

By

deconvolving the

isotope

counts

from

the

natural

gamma counts using the W S

alogorithum

a formation gamma ray

curve is e s ~ i m t e and presented in

track one in API units. Figure 2 com-

pares

a

previously

run gamma ray log

with

th e estimated gamma ray from th e

CSNG. Also presented is the

tota l

tracer curve. The estimated gamma ray

compares very well with the previously

run log except where the total tracer

reading

is ext remely high, making

the

natural activity hard

to

isolate. This

is

i ll us tr ate d j us t above

X900,

where

the

total

tracer

curve exceeds

10 000

API units.

X MPL S

1

well was treated

with

120 000 gallons

f acid and tagged with

.36

mill icurie s of

d ium per

1,000

gallons.

The

acid

was

out through

perforations between

to X806.

looking

at the tota l tracer curve

i t

evident

a large quantity of tracer

is

present

at

the

perforations.

hidden by the 0

to

10 000

API

unit

is

the

lower

tracer

concentration

the

perforations. This is much more

on

the

relative distance curve.

curve indicates tracer material

up the outer casing annulus as

as X678.

e relative

distance

would

appear to

show

iridium went further out

in

the

from X678 to

X754 than

i t

did

th e perforations. This however

may

t be the case. The h igh counts at the

are

caused

by a high

concen-

of

tracer

material

in the p erfo r-

This makes i t

st t is t ic l ly

to see

the weaker

signal

coming

the

formation. The fact that

the

distance curve

is

showing form-

and not

borehole signal

across the

forat ions indicates

that a large

part

of

e

treatment

did

go out in

the

formation

some

acid

channeled up and went out

f

zone.

Example

2

This well was treated with 130

000 allons

of

acid and tagged

with

0.3 mil

icur le s

of

iridium per

1000

gallons. The acid was

pumped

out of

per(orations

between

X852

to

X872, X889

to

X899 and X913

to XX915.

The

tota l

tracer

curve shows two

hot spots in

this

well.

One area

is

near

T

between

Y185 and Y228; the o ther area

is

across

the

perforations.

The relative distance curve

shows tracer material insid e the pipe from

th e

perforations

ll the

way to

TD.

a ls o indic ate s

the only

place in

the

well

tracer material

is

outside

of the pipe

is

across the

perforations.

Unlike example

1,

this well had a long rat-

hole.

Since

th e

specific

gravity of the

particulates containing

th e

tracer

is

1.5

l i t appears some

of the

tracer

did

not

make

the

turn and set t led

at the bottom of

the well.

This is confirmed by

the rela-

tive distance curve indicating iridium in

the

borehole

from the

perforat ions to

TD.

Example 3

This well was treated with 120 000 gallons

of acid

and

tagged

with

0.3 mill icurie s of

i ridium per

1000

gallons

of

acid. The acid

was pumped

out

through

perforations

between

Y238 to

Y248;

Y254 to

Y259;

Y278 to

Y279;

and Y290

to

Y295.

The total tracer

curve

shows tracer mate ri al i n high concentration

from T to the top

of

the

perforations.

The

smooth

decrease in

counts from

T

to

the

perfora tions indica tes some

of

th e radio-

active

beads once

again set t led

at

the

bottom of the well

inside

the pipe. This

is confirmed by

the

relat ive distance

curve.

The relative distance curve very clearly

shows

th e only

tracer

material outside of

the pipe is across the

perforations.

t

also shows that

the

bottom perforations

X290

to

X295) did not take any acid. This

well is an

excellent

example

of

processing

the

Compton

downscattering information

to

discern th e radial distribution of the is -

otope.



ESTIM TING FR CTURE HEIGHT FROM

MM R Y

SPECTROSCOPY OF R DIO CTIVE

TR CERS

A

C SE

STUDY

SPE

21833

CKNOWLEDGEMENTS

REFERENCES:

The use a

WLS algorithm for

determination of the radial

distribut

of an

isotope

is a

major

improvement

tracer

logging.

I would

l ike to thank

Bass

Enterpr ises

allowing me to present

the

log example

this paper. I would also like to

g

particular

acknowledgement

to

Dave

Sha

of

HLS, Billy Aud

of

Bass

Enterprises

Dave Ell io t t

of Protechnic

s, Internatio

Aud W W

Sullivan, R.B.,

Coal

E.B.,

Poulson, T.D., Warrenbo

P A Acid

Refracturing

Prog

Increases Reserves, Cottonwood C

Uni

t

Washakie County Wyomi

paper 21821 presented

at

the

SPE

Permability Reservoirs Sympo

April 15-17, 1991

Gadeken L.L., Smith, H D J

Seifer t ,

D.

J .

Calibration

and

Analysis of Borehole and Format

Sensitivities for Gamma

Ray Spe

roscopy Measurements with

Multip

Radioactive Tracers , paper

V p

ented a t

the

28th annual

SPWLA

Symposium

in

London

England,

J

1987

Gadeken L L and Smith, H D J

Tracerscan:

A

Spectroscopy Techn

for Determining the

Distribution

Multiple Radioactive Tracers

Downhole Operations , paper ZZ, SP

Twenty-Seventh

Annual Logging S

posium

Houston,

Texas,

June 9

1986

3.

2.

Example

5

Example 4

This well was treated with 150,000 gallons

of

acid and for the

f i rs t time scandium was

used as

the

t racer material tagging a t

a

rate

of 0.3 mill icurie s per

1000

gallon of

acid.

The well went on vacuum

at the

end

of

the acid job.

This caused most

of the

isotope

to

be flushed

into

the formation

beyond the depth of

investigation

of the

tool,

thus reducing

the

observed count

rate. The relat ive distance curve gives a

clear

indication

that t he t reatment

stayed

in

zone.

This well

was

treated

with

145,000

gallons

of acid and was aga in tagged with scandium

This well did not

have the

rathole

to allow

the t racer to

set t le

out. The relat ive

distance

and

total tracer data show

the

t reatment s tayed in zone very

well.

The

higher

concent ra tion across the

lower perf

orations probably indicates

that the

higher

specific gravity of

the particulate

con

taining

the t racer material

caused downward

migration of the isotope.

I t was learned from th es e logs , using both

the radial information

and the to ta l t racer

information, that their treatments stayed

in zone.

This

information

was

cr i t ica l

in

proving that t he t reatment f i t the

model

The large

fluctuation

of total t racer

counts

also

shows

how

high

concentration of

the tracer material

w ithin the

perforations

prevents

the relative

distance curve from

distinguishing

the deep

formation comp-

onent.

CONCLUSIONS:

This tagging

program was

successful because

the

opera to r d id a good job engineering

the

t racer program The isotope,

the

material

con ta in ing the

isotope, the specific gra

vi

ty of the carrier were

a l l

taken

into

account to optimize

the results.

Gamma

ray

spectroscopy

of radioactive

tracers provides

unique

information useful

in designing a well stimulation program

4

Gadeken

L.L.,

Smith, H.D.,

Relative Distance Indicator f

Gamma

Ray

Spectroscopy

Measurem

with Radioacti

ve Tracers

p

17962 presented a t

the

SPE Mi

East Oil

Technical

Conference

Exh ib it ion held in Manana

Bah

March 11-14, 1989

278



SPI;. 21

___

Ell

OIlMla lelllngl 8o

Tracer OR

I

Relallve Dlslance

o API 200 0 API liOOO

F...

FItError

0···.•

•·· • •••••1

CCl

Eat 08IMIII ralngl

48

80

TracerOR

I

Relative Dlltanee

o API 200 0 API

liOOO

F...

Fit Error

ii r

CCl

Ij

lft

_

t•

Ii

-. j-- -t.: _1,-

..

·tTolal Scandium

•. .

-._.-

~ A e l a l i v e

lo r r I I

T T l

Ff-- lI : Olslance·--

•-, t =:;.

- - - ~ ~ l l

} ~ ~ ~

~ ~ ; : ~ ~ ~ ~

] y 2 O O 1 I - : ~ t : _ : - ~ t : _ ~ - H ~ I : : i ~ - r ± _ t _ _ t t _ : : 1 _ _ f

_:: :: :

:f;ijII1

iii i ; ~ : = -

~ I ~ ~ G ~ ~ a

:

;; . . . .

..

_

.

.

.

...

.

..

~ y t ~

__ Aelallve

Distance

:: :» ).l:

I r I I - I

: : X / H » i ~ L : 1 :

Figure one: Overlays of multiple logging runs with CSNG tools on two wells.



192

r

TracerGR

o API 12000

~ t l r c e r

- Iridium

Gamma

o

API 200

Est. Gamma

Ci APj 2

.

\

Perfs

.

-

X900

Est. Gamma

Gamma

==

-

o

SPE

2 83

3

Gamma Est. Gamma

API

2

API

2

Gamma

Est. Gamma

Figure two: Comparison estimated gamma ray from a TracerScan to a previous

run gamma ray.

28



Est. GR 192

o

API 200 Formation

II

o

s

o

l

Formation

CCl

SPE 83

3

192

Tracer GR

o

API 12000

Total Gamma Ray

o API 12000

p

~ ~ ~

=

~

::-

Total GR

Example two: Compton ratios indicate ir idium at perforat ions is in the formation

while high counts at the bottom of the hole came from inside the pipe.

8



Est

o

API 200

CCL

Formation

s

o

o

m

Formation

192

r

Tracer

o 1000

Total Gamma Ray

o

1000

a

I i ~ ~

:

~ ~ i ~

Total GR

_

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ : : : = :

: : : :

:

: : :

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ; ~

_

SP 8

Example one: TracerScan processing indicates iridium outside of the casing went

out of zone

281



Est. GR 46Sc

o API

Formation

~ ~ :

..

:

~

a m m a ~

• •

Formation

o

Cf Cf

co co

Formation

~

.·.Relative

~ · i s t n c e

~ ~ S c a n d i u m

~ .. :

CCl

SPE

83

3

46

SC

Tracer GR

o API 5000

Total Gamma Ray

o

API 5000

Total

GR

:.

~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ r · ·

Example four: When this well went on vacuum

it

flushed most of the scandium

beyond

the depth

of investigation of the

logging tool the scandium

that is detectable

indicates the

treatment stayed in zone.

284



amma

o

API 2

CCl

o

1

Formation

l

o

m

Formation

192

r

Tracer

o

API

8

Total Gamma Ray

o

API 8

S

83

Total GR

Example three: Relative distance and

count

rates indicate some of the high density

iridium beads settled at the bottom of the well

83



Est.

GR

o API 200

46SC

Formation

.

o

•

o

D

Formation

CCL

46SC

Tracer GR

o API 5

Total Gamma Ray

o

API QQ

SP 83

, . . - ~ : r . t t ; . : .

. : ~ ~ ~ J

~ ~ ~ ~

~ : L ~ ~ \

~

•

~ - . : . :

.

.

-

.: ::

Total Scandium

==

~ ~ : :

}

~ : t ~ ·

; i =

~

,

~ T o t a l G R

:

..' ..

.

• • > r

r

~ } r . · ~ · . L ; · ;

~ ~ - A l P

Example five: This well is a good example

of

a treatment staying in zone.

Documents

Fracture Height