GEOL. C ROAT. 49/2 3 I 7 - 3221 4 Figs. 1 ZAOREB 1996 

Professional paper 

Bright Spot, AVO Inversion and Synthetic on 
Grubisno Polje Locality 

Marija VIDOVIC. MiUenko RUKAVINA, Miros lav SYRINEK, Jasmina BUELIC 
ane! Mirjana MILINOVIC-KOLUNDZIC 

PROCEEDINGS 

Key wo rd s: Bright spot, A YO inversion and anal ysis, 
Restric ted gradient stack, Synthetic. 

Abst ract 
Bright spot analysis is generally used in exploration fo r hydroc<lr­

bons. especially in the SC:lrch fo r gas. However, many seismic am pli­
Il1de anomal ies have nOI been call sed by gas accumulat ions, but they 
arc re l:lled 10 lithological changes. A VO methods (amplitude vs. off­
se t rnClho(15) can be used to di stingu ish between Ihe presence of gas 
and litholog ical cb;mgcs. 

The Grubi SIlO polje fie ld has been defined and mapped by th ese 
methods. and a we ll. drilled afterward co nfirm ed th e res ults . After 
'lcqui ring the we ll -logg ing data a Synthetic was comp ut ed. The cor­
re lation of the Sy nthetic with lhe reproccssed scism ic sect ions and the 
A VO data enabled the more precise vertical and hori zon tal defi nitioJl 
of Ihe g'lS lIcCurtWlalion. Th is method and resu lts contribute to lhe 
more rational exploration and developmenl ofa gas field. 

L INTROD UCTION 

Bright spot analyses have been accepted in explo­
ration 1'0 1' hydrocarbons, especially in gas bearing areas 
for twenty years. A Brigh l spot is related to gas accu­
Illu lati ons in the beds , which significant ly a ffec t the 
se ismi c Hmp litudes (re fl ec tion coefficients) and ca use 
iHl1p lil ude anomalies on seismic sections. 

Howe ver, seism ic amplitude anomalies can also bc 
ge nerat cd by litholog ica l changes. To di sting ui sh 
between the presence of gas and litholog ica l changes 
A VO inve rs ion s (ampli tude vs . offset invers ion, i.e. 
amplitude re lated to the source-receiver distance), have 
been lIsed during the last ten years. 

T he A YO inversion and the Bri ght spot methods 
have been lIscd together to analyse the G rubisno Po lje 
fi e ld. T he location and struc tural con tours o f this fi eld 
can be seen on Fig. L. It is important to mention that the 
analys is of AVO inversion does not imply the gas quan­
tity in a bed. Afte r analyzing the drillin g re sults, and 
c rea ti on of a Synthetic, the com plexity of th is gas accu­
mulation and the advantage of jo int methods became 
clearer. 

tNA-NafwpJin, Subiccva 29, HR- 10000 Zagreb, Croa tia. 

Kljucn e r ij eci: bright spot, AVO pre lvorba i analiza, 
ograniceni gradient stack, sintetski seizmogram. 

Saietak 
Anatiza "Bright Spot" opec je prihv<leenll 1I iSl razivanj u uglj iko­

vodika, nmoci to pJina. Med lHi m, mnoge seizmicke ampli ludne ano­
ma lijc Ilisu uuokovane plinskim aku mu lacijama, ncgo su odraz li lo­
loskih pro mjena. Za njihovo razlikov a nje s luze AVO melode 
(ponasanje mnpti1Uda glede lldaljenosti izvor-prijemnik). 

Loka lilel Grubi sno Polje je lllv rden i okonlU ren ovim rnelodama, 
locirana jc busotina. koja je potvrdi la njihove rezllltate. Nakoll izmjc­
rcnih karo taznih podataka izraden je s int etsk i scizmogram. Korelac ije 
s int ctskog seizmog rama i seizmick ih profila (mog uce rcob radc i 
kore kc ij e se izmi cki h pod alaka i mjerenih brzi na) i "AVO" ll1 etode 
pre tvorbi , omogucile su tocnije okof1turi v<lnjc lezisla u vertikalnom i 
hor izontaln0111 smis lu. Sve 10 doprinos i racionaJnij em istraziva nju i 
razradi lokalitcta. 

2. INTERPRETATION 

T he GrubisIlO Polj e fic ld, wh ich is a pa rI o f the 
Bje lovar Depresssio n, is situated in central Croat ia, 
approx imately 120 km east of Zagreb. In the region sur­
rounding this field the well s Or· 1 (1940), Or·2 and Or-3 
(1965) we re drilled on the basis of con tempo rary 
knowl edge by gravity surveys pe rformed before 1940 
and be tween 1962-1965. T he survey dens it y was 1 
point/km2. The data from the seismic survey performed 
by single fold and analogue equ ipment were a lso used. 
All three wells were barren . 

Be tween 1975 and 198 1 a de ta il ed gravi ty survey 
was performed over a broader a rea with the dens ity of 
10 point s/km2 which resulted ill eas ie r recogn it ion of 
the gravi metric maxim um. In 1990 more in tense regio­
na l seismic exploration of the Bjelovar Depression was 
perfo rmed , while the Grub isno Po lje fi e ld has been 
in tensively explored by seismic methods since 1992. 

Based on the newly recorded regional line V. Zden­
ci-2V -92 and the re-processed older lines which implied 
the ex istence of the Bri g ht spot, the int erpre ta tion of 
Gru bisno Polje locality was perfo rmed. ]n 1994 a new 
well Gr· 1 z was drilled. At -654 m depth the well 
encount e red the gas accum ul ation , with the gas/wate r 
contac t at -740 m. Du ring 1994 and J995 add it ional 



31 K Geologi:l Cro:lliC:l 49n 

1) ~ 2) - 3) • 4) ___ · 1000 --- 5) 6) _ GNl ·140 m-- 7) .Gr.1Z 8) 

Fig. I Structural map of the EK- marker Rs7 with the contoured main reservoir. Legend: I) gas reservoir; 2) normal fault; 3) reverse fault: 4) 
isodepth lines; 5) sei smic lines: 6) estimated gas/water contact; 7) exploration well: R) analysed part of seismic section. 

seismic lines were shot, processed and interpreted. The 
result of this interpretation enabled more precise gas 
accumulation contouring (Fig. 1). 

During the interpretation the following seismic dis­
plays were used: RAP-stack (Reflection amplitudc 
preservation -stack), Phase and Frequency attribut ions, 
AVO inversion data (A YO angle trace analysis), Gradi­
ent stack, and Restricted gradient stack. As all the inter­
pretations cannot be presented in this paper, only the 
RAP-stack section, AVO angle trace analysis (Fig. 2a), 
Restricted gradient stack (Fig. 2b), and correlat ion of 
well and seismic data (Fig. 3) have been included. 

Figure 2a represents the RAP-stack section of the 
seismic profile GRP-2Y-9S. In the central part along 
the well Gr- I Z, three Bright spots are visible in the 
interval between 600-674 mscc. The shallowest Bright 
spot was caused by a gas pool, and a 100% relative gas 
concentration was registered during the well drilling 
(OP - overpressure) in the lower part of the equivalent 
of Bregi sandstones of EK-marker Rep - silty sandstone 

(Fig. 3). The deeper Bright spot was caused by entering 
the Klostar Ivanic formation (EK -marker L1), which is 
lithologically represented by clast fragments and fossil 
karst. The deepest Bright spot is related to the low 
velocity zone within the Klostar Tvanic formation (sonic 
log in the Fig. 3), probably to the marker Rv. These 
Bright spots are accompanied by the presence of gas 
and the whole interval is overpressured l

. 

The main gas accumulation is situated in the Mosti 
member of the Moslovacka gom f0n11ation (EK marker 
Rs7), and in the basal Tertiary (BT). The Mosti member 
is represented by biocalcirudites - bioealearenites, while 
the basal Tertiary is represented by metamorphics. The 
top of the main reservoir is not represented on a RAP­
stack section with a Bright spot, because of the consid­
erably higher velocity of the carbonate matrix. Due to 
the gas accumulation the reflection coefficients are 
weakened (Dim spot). The posi tive amplitude is barely 
visible on the section. Entering the main gas accumula­
tion, the overpressure is lost. 

I Vuk:ldinovic, M. (1995): Zavrsno geoiosko izvjesce za istraznu busotinu Grubisno Polje- I zapad.- Unpublished report, Ina-Naftaplin Library. 



Vidovic. Rukavina. Syrinck Rijclic & Milinovic· Kolundiic: Bright Spot. AVO invcn;ion :mJ SYll1helic 011 .. 319 

, 
... 
III 

a 

COP 
85' 

b 

., 
...... _..!'.' . 

AVO 

AVO 
767 

. ' . . ' . 

G .... 1Z 

Gr-1Z 

' .. ~ .. 

AVO 

AVO 
710 

• 

COP 
656 

500 

u 

'" '" e 

"' 
BT 

1000 

R, 

500 

"" .., 
'" E 

III 

1000 

Fig. 2 a) Interprctation of RAP stack section wit h two AVO analyses (eMP 767 and 710) included: b) inlcrprclation of Restricled Gmdient 
Slack. Red allom;. J i e.~ indicate the potential presence of gas. 

The basal Tertiary is charac terized by a slightly 
higher amplilude. The gas/watcr cont ac t is re latcd to a 
mild Flat SpOL The bottom of the Gr-IZ well is at 790 
msec within the Tertiary basement. 

Below the botiom of the well, at approx imately 820 
msec latera ll y to the well there is a Bright spot, whi ch 

represents a probable lithological change, maybe cvcn 
an erosional unconformit y, which has not yet bcen 
renehed by any of the wells. Thi s Bright spot is weak­
cned in it s centra l part because of the absorption of 
seismic energy (Shadow zonc below Bright spot ). 

On the RAP-stack section the in terpreta tion of EK-



320 Geologia Croatica 49(2 

I I - - ~ ... ""--_:- SONIC (usekim) DENSITY (!;CI!l) RC AVO COP RP :tP'" 

soo o 1.25 

Fig.::; Comparison of well, seismic data alld schematic lithologic.1I column. Legend : I ) mctalllorphilcs (g"<ll1il Illica selll),I): 2) bioc;ticirudilC to 
bioca lcarenit e: 3) sandy argillaceoLls marl 10 calcareous clay, c la sti cs and fossil karst fragments: 4) marl and sill y sandstoll e; 5) sil lY to 
argillaceous marl: 6) argillaceolls marl with sandstone interbeds; 7) marly clay with sand interbeds: R) main reservoir. 

marke rs: 6., R57, BT and a probable lithological change 
"X" arc shown. There are a lso two A VO angle trace 
analyses, on CDI' (Common Depth Point) 767 and 7 10 
wilh angles of 8 to 24 degrees, to illustrate Ihe possible 
gas saturations in the sediment complex. 

Figure 2b (Restric ted Gradient s tack) is d isc ussed 
below and ind icates that the results of AVO analyses of 
pre-s tack seismic data show good correlat ion with the 
RAP-stack section (Fig. 2a) confirming the presence of 
gas. 

3. AVO INVERSION 

It has been proven that the P-wave reflection coeffi­
cie nt s on the bo undary of two clastic media changc 
with the angles of incidence, or with the offsets because 
of the var iolls re lative values of Poisson's rat io across 
the rellect ing interface. 

These fac ts were al ready noticed on the seismic da ta 
pri or to s tacking (KOEFOED, 1955), and were late r 
confirmed theoretically and by laboratory meas ure­
ment s. The res ults of the laboratory measurements 
showed that the highly porous, gas saturated sa nd ­
stones, give a vc ry low Poisson' s ratio. In the natural 
underground condition s in sandy sediment s, if gas is 
present, il is proved that the P-wavc reflect ion coe lTi ­
cient increases wilh the angle of incidence (OSTRAN­
DER, 1984; DOMENICO, 1976; R UTHERFORD & 
WILLIAMS , 1989). 

During the A YO invc rsion procedure special attcn ­
tion should be paid to thc quality of the seismic data , 
both field rccord ing and during se ismic process ing, 

becausc they significantly a ffect the recorded re flection 
amplitudes as a distance funct ion from the source. Dur­
ing the survey attention should be pa id to whether the 
analysed horizons arc at suitable depths, considering 
Ihe di s tance between the source and the receive r; if 
the re is enough coverage (at least three-fold on target 
horizons); how the field dala a re affeeted by noisc; and 
whether the seismic encrgy can be losl during shoOling, 
because of some natural conditions, e.g. weathering 
zone. 

Generally the following fac tors a ffect A YO inver­
sion s: angles of reflecti on incidence, re flec tor curv ing 
with sign ificant impact at small depths (il is not recom­
mended to perform the inve rsion up to ,approximately 
400 Ill), transmission coeffi cien ts, multiples and ine las­
tic smoothings. 

During the seismic process ing it is impor tant to 
restore the ene rgy lost wit h time and dis tance, because 
of the spherical spreading o f wavefronls. The ot her fac­
tors thai affect the relat ive amplitude re lations are: the 
diminishing of amplitudes because of geophone a rrays 
and sources, especiall y at shallow depths, whe re the 
angles o f incidence can be increased, whi le part of 
energy can be losl because of re fraction. This is why it 
is not recommended to ana lyse the longer offsets. The 
same happens for greate r depths at sho J1 offsets. Gcner­
ally speaking, vcry shallow and deep reflections are not 
rel iable for A YO inversions. 

The next factors affecting ampl itud e re lati ons arc 
the tuning effects. Closely si tuated re fl ec ti ons cannot be 
expressed on seismics, but thcy merge int o one, which 
could be anal yzed through the NMO eq uation and by 
higher seismic resolution (OSTRANDER, 1984; West-



Vidovic, Ruk~vina, Syrinck, Bijelic & Milinovic Kolundzic' Bright Spot i\ VO I nvcr~ion and Synthetic on 321 

Q) 
-0 

:~ 
Ci 
E 
'" c 
.g P' 
u 
Q) 

~ 

sin2~ 

Fig. 4 Lincar iii of AVO curve of rencclion ampli lude vs. s in ' ~. 

ern Geophys ical:!), These effects in th inner shall ow 
sands (approx imately 12-15 111 ) can resul t in d imin ish­
ing ampl itudes at greater distances from the source. 

Various no ises can also s ignificantl y afrec t the 
amplitudes, and in the case of long offsets they can pro­
voke dimi nishing of amplilude, Thi s is why analys is or 
the SIN ra ti o (S ignal -to-noise ralio) is recommended. 
To improve the SIN rat io in accordance with the Fres­
nel-zone width, the partial averaging of severa l orrsets 
or sever" l angles was undertaken. 

It is recom mended to use da ta recorded under the 
constant cond itions for AVO inversion s, wh ile d uring 
the seismic process ing one should keep to the ehosen 
processi ng schedule. The most important phases or the 
processing arc: treat ment of the spherical spread ing or 
wave fronts, phase correction duc to var iou s recordi ng 
cond itions, deta iled vcloc ity analyses, deconvolut ion; it 
is not recommended to pcrform statistica l deconvolu­
tion which may introd uce amp litude and phase varia­
tions . Here, surface consistent deconvo lutio n of the 
spiki ng Iype was performed, because the operat ions are 
separately created: a) operalOI' in the source domain 10 

crase the ditTerences from source to source, b) opera tor 
in tllc receiver domain that corrects the contacts during 
the recordin g rece iver/earth (coupling), c) operator in 
the COmJl1 on offsets domain, to reject reverberations. 

A ll the figures we re produced wit hout amplitude 
equa li sati on, so Ihat relative energy on all the recorded 
receivers could remain in rea l lerlll S. Fi gure 2a shows 
the RAP-Slack display and amplit ude vs, offsets ror two 
cOlll mon dept h poinls (COP 767 and 7 10) on the sorted 
se ismic traces. It should be men tioncd that the A va 
e ffects acco rd ing to the angle of incidence are more 
obv iously expressed , than those according to the o frse t. 

Bes ide A Va invers ions of seismic data before 
stacki ng, altribtJl ive stacks were made: Grad ient stack 
and Res tric ted Gradient stack. Gradi ent Sl ack G is 
based on th e principle that the re fl ec tion ampl it ude 
R(P) of P-waves reflect ed [rom the int erface bct ween 
two clasti c media, varies linearly wi lh Ihe sin 2 p (Fi g. 4) 
accorcii ng to the equat ion: 

R(P) - P + G . sin'p 

where P is interccpt, P retlection ang le, and G gradient 

(s lope). 

The linear fi t o f AVO curve of reflection amplitude 
vs. si n:! p run correctly for refl ect ion ang le P less than 
25°, and data must be corrected for sta tics and NMO by 
carefu lly obtained veloci ties. 

The values are calculat ed on each seismi c trace 
before Slacking data and for each sampling period. The 
section of Grad ient stack G contains the information on 
the increase in amplitude changes depending on s i n 2 ~ 
(oll sels). 

Difrcrelll companies have vari olls programs for 
analysi ng the am plitude vs, offse t. T he e lement ary 
a lgorithms arc the same, the dilTerences occur in the 
var ious kinds of representati on and inversion in to 
attri butes. The Restricted grad ient stack GR (Fig. 2b) is 
recommended. This is the Grad ient or abso lute ampli ­
tudes depending on sin2p: 

GR = sgn(P) . G 

When P and G have the sam e s ign GR is posi tive, 
when they are opposite signs GR is negati ve. 

Comparing Restric ted G radi ent stack GR (Fig, 2b) 
with RAP-stack (Fig. 2a) showing Bright spot indica­
tors on the GR section , there are positive anoma lies 
(show n in red) when the absolute amplit udes increase 
with offsets. 

In a typical sand-marl sequence, the large pos itive 
anomalies ind icate the potent ial presence of gas. They 
are relatcd to Ihc increase of absolute ampli tudes in 
re lation to the increased angles, whi le the negat ive 
amplitudes deny the presence o f gas. 

On Fig, 2 b, red positive amplitudes are s ituated 
above the main gas accu nlul alion, approximately 
between 600 and 675 Illsec, support ing the above men­
ti olled in terprctal ion of th is compl ex nrea . 

4. SYNTHETIC 

Logg in g data and veloci ty measurements provided 
dma for a Synthet ic to be created a ft e r the we ll was 
completcd and logg ing and velocity data became ava il ­
ab le. T he proccdure for the crea tio n of' the Synlhet ic 
enables control of Ihe bas ic inte rpre tation assumptions 
and corrclation of the drilling results, well -log data and 
seismic profiles. 

Thc curvc correct ion of the sonic log was performed 
first, the so-called Cali bration o f the recorded sonic log 
wi th the reco rded veloc it y (ve locity surveyor VSP -
Vertica l seismic profiling) . The control point s that are 
chosen have geological importance and they arc located 
at lithological changes or stratigraphic boundaries. 

Differences in velocities can appear 1'01' a va riety of 
reaso ns, such arc the case o f two different so urces of 

~ Wes lcrn Gcophysical (1994): Manual s for Users, IIclp.- Unplibi. Di sc Documentation . 



:\22 

velocity survey. The sonic log integrated travel time, 
measured on the basis of refraction may not truly reflect 
the seismic travel time. Sonic log integrated veloci ties 
call diller from the seismic waves velocities, because 
the frequency of the seismic line is lower and its wave­
length is longer. Velocities depend on the places where 
the gcophones wcre sited in a well. Sometimes fractures 
in formations during the drilling process have a greater 
effect on seismic velocities than on the sonic log. 

After the calibration has been performed, the Syn­
thetic can be created from the sonic log and the density 
log transforming them into two-way-time. Simultane­
ously the seismic profiles with wells marked on them 
arc loaded. Due to the better coreiation all other avail­
able logs were loaded, while the lithological column 
according to the lithological profile was compiled (Fig. 
3) . AVO analysis of seismic data CDr 744, from th e 
nearby well Gr- I Z was added to the end of Fig. 3. 

This allows comparison and identificat ion of the 
horizons on se ismic profiles, logging cu rves and on 
Synthetic. 

Wavelets for convolv ing the reflection coefficients 
call be for example Ricker wavelets, but the correlation 
wavelet can also be ex trac ted by comparing the seism ic 
pro riles and reflection coerricients. During this proce­
dure one can indicate the need ror corrections of the 
seismic profiles, sonic log or velocity survey . 

During the study of al l these comp iled data, the 
problems or phase corrections can be exp lain ed, and 
also the time shifts in se ismic data, and lithological 
impacts, which can be dillerent on th e Synthetic and 
the surface seismic. T he problems during the seismi c 
processing can be defined and any requirement for the 
reprocessing of seismic profiles can be indicated. 

On the examp le of the well Gr-I Z there was no need 
ror the correction of seismic profiles, because they were 
especially reprocessed for the needs of AVO. The Syn­
thetic was formed with the Ri cker wavelet, which rep­
resents the real picture of th e Synthetic seismogram. 
Small discrepancies are explai ned in the interpretat ion, 
according to the well data and the complexity of th is 
gas accuillulation. 

lr the Synthetic is shifted along th e profile to the 
well, the stratigraphic continuity can be followed dur­
ing the process or extracting the correlation wavelet, as 
a better or weaker correlation. 

5. CONCLUSION 

Bright spots enabled the definition of gas accumula­
tions along with the usual interpretation methods. The 
difference between gas accumulat ions and lithological 
changes can be identified by the AVO analysis inver­
sion or se ismic data, and this also contributed to the 
more rational cOlllouring of the gas accumulation and to 
the well location. 

In the complex subsurface geo logy A YO methods 
can be unreliable, whilc the results of these analyses 

Geologia Cromiea 49/2 

must be in terpreted very careful ly, so the impact or 
local changes in geology can be separated frolll the 
anomalies of the amplitudes related to the source­
-receiver distance. 

It is important to mention that these methods must 
be confirmed by other methods, and the results Illust be 
reviewed in terms of the problems of the exploration 
area. Due to explanati on of this complex gas accuillula­
tion, we include all the well data, thnt in the seism ic 
section along with Bright spots contains also Dim spots. 
This can also point 10 any need for reprocessing and 
correction of seismic data, and more accurate contour­
ing of the reservoir in the vertical and horizontal sense. 
If the various data are weI! correlated, they represent 
and confirm the in terpreta tion and indicate other 
favourable drilling locations. In all, thi s is a contribu­
tion toward more rational exploration and has a positive 
impact on business decisions. 

The processings was performed with the programs 
or Western Geophys ical and Western At las Companies 
on IBM RISC-6000. 

6. REFERENCES 

DOMENICO. S.N. (1976): Enloct of brine-gas mixture 
on velocity in an uncosol idated sand reservoir.­
Geophysics. 41, 882-894. 

KOEFOED, O. (1955): On the elleet of Poisson's ratio 
of rock strate on the relleclion coefficient of plane 
waves.- Geophysical Prospecting, 3, 38 I -387. 

OSTRANDER,W.J. (1984): P lane-wave reflection 
coefficients for gas sands and nonnormal angles of 
incidence.- Gcophysics, 49/10, 1637-1648. 

RUTHERFORD, R.S. & W1LLlAMS , R.H. (1989): 
Amplitude-ve rsus-offset variat ions in gas sands.­
Geophysics, 54, 680-688.