Anda di halaman 1dari 3

Rock Compressibility and Failure as

~=e~ervoirM-echanisms
-.----—-. in Geopressured
Gas Reservoirs
David W. Harville,* SPE-AIME, Louisianastate U.
Murrsy F. Hawkins, Jr., SPE-AIME,Louisiana State U.
“.

Introduction
Rock compressibility has long been recognized as an that because of low net Overburden pressures, mek
important factor in material balance calculations of compressibilities in geopressured reservoirs are con-
Qfl im place for closed reservoirs producing above siderably greater than for similar rocks in normally
bubble-point pressure. 1 For example, if the pore vol- pressured reservoirs. We further suggest that as reser-
ume compressibtity of the reservoir rock is half of voir pressure is depleted, the increase in net over-
t~e g~m-pm~sih~ty Of tie undersaturated oil, neglect burden pressure initially causes inelastic rock com-
of the rock compressibtity term results in about a 50 paction or rock failure. As failure continues with de-
percent overestimation of oil in place. In general, it creasing pore pressure, rock compressibility deaeaaes
may be stated that in material balance calculations on and eventually reaches normal values in the range of
closed reservoirs, consideration of rock compressibti- 6 microsip.
ty becomes increasingly importantas the fluid com-
pressibility decreases. For this reason the effect of North Ossun F]eld, Louisiana
rock compressibility is commonly neglected in studies The mechanisms proposed in the previous paragraph
on gas reservoirs where gas compressibfity is usually are believed to be illustrated by the performance of
great. the NS2B reservoir of the North Ossun field, Lafay-
Because gas compressibilities decrease with in- ette Parish, La. This is a geopressured gas reservoir
creasing pressures, the consideration of rock compres- with an initial pore pressure of 8,921 psia at 1Z,5W
sibility becomes increasingly important for deeper, ft subsea depth, or a gradient of 0.725 psi/ft. Table
high pressure gas reservoirs. For example, the com- 1 gives pertinent data on this reservoir. Oood geologic
pressibility of the gas in the reservoir to be discussed control is indicated by the structure map, Fig. 1. Al-
is 30 microsip** at an initial reservoir pressure of though a gas-water contact exists, it is doubtful that
8,921 psia. For a nominal pore volume rock com- the associated aquifer is very large because the rese-
pressibtity of 6 microsip, neglect of rock compressi- rvoir appears to shale out on the west. III addition,
bility in material balance calculations on a closed considerable complex faulting in the area almost eer-
reservoir will result in a 20 percent overestimation of tairdy closes the reservoir with a small aaswiated
initial gas in place. If the rock compressibfity is aquifer.
larger than 6 microsip, then a still larger overestima- Good core and log data have been used to cakm-
dcn ~f ~u~~ ~!ace ~.~l~t~$In this study we propose 1... . . -u-
;.;t;~l ..J.--.—--—
lCWall hvtkmarkn
-— volume of 583 mil-
—rnnre
“Now with Phillips Patmleum Co., Mo~an City, La. lion cu ft, and, with PVT dataj to calculate an initial
. .Abb~aVia~~n f~~ 10-6 ~i-1, ~i~m being 10-c and sip (aqll are

inches per pound) being psi-l.


gas in place of 114 Bscf. The pressure-production

A study of the North Ossun field, Louisiana, reveals that as reservoir pressure is
depleted the increase in net overburden pressure initially causes rock failure and as the
failure continues with decreasing pore pressure, rock compressibility decreases until
eventually it reaches a normal value.

1528
TABLE 1-RESERVOIRDATANORTHOSSUNFIELD* ~_GBS
LAFAYEITE PARISH, LA., NS2B RESERVOIR P— ~ y[l + c.(P, – p)] + (G – G,)%

Depth, ft 12,500 (l?ater) (Gas)


Pressure, psia 8,921
. . . . . . . . .. (1)
Gradient, psi/ft 0.725
Temperature, “F 248 As used in this study, tbe connate water Sw included
Gas-water contact, ft 12,580 only that water in the gas reservoir. In other calcula-
Average gross sand, tt 100 tions presented by Harville2 the connate water value
Porosity (33 s% quite uniform) 0.235
(electric ioga) 0.24 was increased to include water in associated limited
Connate water (eiectric logs) 0.34 aquifers. In one case the aquifer size was assumed to
Permeability, md (33 SS) 200 be equal to the size of the gas reservoir, and in an-
Producing walls 4 other case is was assumed to be twice the size of the
Geologic controi weils 18 gas reservoir. Pressure equalization between the res-
Dew-point pressure, psia 6,920 ervoir and the aquifer was shown to be a valid as-
initial GOR, bbi/MMacf 160 sumption for these small aquifers and for the indi-
Condensate gravity, *APi 47
cated rock and fluid properties.
Net bulk gas volume, MMcf 2,480
inithi Z fSCtO; 1.472
Fig. 4 shows the pore volume calculated by Eq. 1
~~ ~ $GEc~oEof pm ?~ssu~, ~&g no aaeoci-
Initiai gas in piace, Bscf (volumetric) 114
Initial gas compressibility 30 X iO-O p~i-’ at ated aquifer and using for. the
. . i&ial
. gas in place the
8,S$2i @a
. . - ...nti..r.a nf
volumetric estimate of i 14 BSCI. The d=rivau.. “.
history of the reservoir is shown in Fig. 2, and p/z this curve may be used to calculate the rock com-
vs cumulative production is shown in Fig, 3. For pres- pressibility at each point. As indicated in Fig. 4, the
sures below the dew point, two-phase gas deviation rock compressibfity reaches a maximum of 28 micro-
factors were calculated from the PVT data as de- sip during the early depletion where rock failure is
scribed in Chapter 2 of Ref. 1, presumed to occur. Below about 6,000 psi, according
Fig. 3 indicates, from extrapolation in the early to thk interpretation, rock failure is essentially com-
life, an initial gas in place of 220 Bscf, almost twice plete and the calculated rock compressibility drops
the volumetric estimate of 114 Bscf. It is proposed to a more usual I alue of 6 microsip. Because of lower
that during the early life the pressure is partially rock and h@er gas compressibilities at lower pore
sustained by high rock compressibfity resulting from pressures, reservoir performance is very close to that
rock failure as described previously. After the pro- of a constant volume system, and the p/z extrapola-
duction of some 20 Bscf, rock failure is essentially tion is reasonably accurate. It may be only coinci-
completed and rock compressibfity drops to a nor- dental that rock compressibility drops to more usual
mal value of about 6 microsip, at a pore pressure of values when the pore pressure reaches a normal value
about 6,500 psi. At thk pore pressure the gas com- for the reservoir depth; i.e., a gradient of about 0.50
pressibility is up to about 75 microsip, which makes psi/ ft.
rock c.ompressibfity a comparatively small factor. The foregoing presentation does not prove, of
Thus extrapolation of the curve after about 20 Bscf course, that rock failure is the major source of pres-
has been produced should be valid, and in fact it ex- sure support of this reservoir during depletion. (How-
trapolates to 118 Bscf to agree well with the volu- ever, Fig. 5, taken from work by Fatt,a provides a
metric estimate of 114 Bscf. striking similarity in the behavior of a Sespe sand-
A further analysis of thk reservoir includes the stone sample that had not undergone high overburden
calculation, with the following equation, of total ap- pressures during its geologic history.) Corroboration
parent reservoir pore volume as a function of pres- for the suggestion — that rock failure is a major
sure. source of pressure support during depletion — is

14 lg

4 ,;%-’-...::’””’ “w c“”’””
Ig
.*

b~
*>
.....
,6* “--------

0
k’
18
“’2‘i 4
0
. . ..
‘.
3

20
a s ,) ~
1 g9 25 ~
.*’ 12 0----
--- 0
-. #-
<8 20:
-------------
-.. . .-9
6 Z
07 15 z
:. 7
:
. . . . . ..-Q”-””--”””
;6 102
1: ~
5 5
+’:~~’
~o
-.7
0
59 60 61 62 63 64 65 66
Year
II& l—Structure mep, top of sand, NS2B reservoir,
North Oesun field, Le. Fig. 2—Pressure-produdion histov, NS2B r$servoir.

DECEMBER,1%9 1529
found in the difficulty of explaining the character of
the support by generally accepted water influx the-
ories, because the strong initial support rather sud-
denly drops to a very small value. In the additional
studies by Harville, not presented here, which con-
sidered small associated aquifers, the character of the
pressure support was unchanged, and only the values
of the rock compressibil@ were reduced, during both
the initial assumed rock failure interval and the later
stage.
It has been suggested that pressure support in this
~~ of reservoir may be water influx from the over-
lying and underlying shales. Preliminary studies of
this possibility indicate that the adjacent shales can
provide neither the volume nor the character of the
pressure support indicated by Fig. 4. On the other
1 hand, rock failure rather nicely explains both.
\ l~~efiyei ~~ fi~f ~~ f~~.~ oc~~-rsh tajs tyrpe of
reservoir, it should not be forgotten that normal rock
compressibtity and connate water expansion are im-
Gas Production ( BCF) portant reservoir mechanisms at higher pore pressures
because of decreasing gas compressibility. It should
Fig. 3-p/z vs cumulative production for the also be realized that what we have just said is even
NS2B reservoir.
more important with oil reservoirs because of their
very low fluid compressibilities.
With regard to geopressured gas reservoirs, it is .
understood that many that we bdieved to be essen-
tially closed systems exhibit a large discrepancy be-
tween the volumetric estimate of gas in place and that
estimated by extrapolation of the p/z plot. Commonly
mentioned is a ratio of about 2 between the figmes,
which is the ratio found in thk study. We hope that
this study will encourage other studies that may shed
more light on this discrepancy and the reservoir
mechanism or mechanisms controlling production
from geopressured reservoirs.
Nomenclature
B, = gas volume factor at pressure p, cu ft/8cf
B,l = gas volume factor at initial pressure,
W2
Cu ft/scf
Pressure, PSIA (000 ‘s) c* = connate water compressibtity, psi-i
(used 3 X 10+ psi-’)
Fig. 4-Calculated pore volume of the NS2B resenroir,
assuming no water influx and G = 114 Bcf. G = initial gas in place, scf
G, = gas produced when pressure is p, scf
p = average pore pressure at later time, psia
Pi = initial reservoir pore pressure, psia
& = connate water saturation, fraction
v, = total pore volume, cu ft
References
1. Craft, B. C. and Hawkins, M. F., Jr.: Applied Petroleum
Reservoir Engineering, Prentice-Hall, Inc., Engiewood
cliffs, N. J. (1959) 135-138.
2. Harville, D. W.: “Rock Collapse as a Producing Mechsn-
ism in SuperpressureReservoira”,MS thesis, Louisiana
StateU., BatonRouge(1967).
3,Fatt, I.: “Compreasibdityof Sandstonesat Low to Mod-
:~5~ Preasuses”,Bull., AAPG (1954) 4% No. 8, 1924-
JWT

0 2 4 6 e 10 12 14
External Pressure, PSI (000 ‘s)
Original manuscript raceivad March 24, 1969. Raviaad manu-
script received July 9, 1969. Paper (SRZ 2600) wos pmntad at
Fig.5-Bulk volume change of a Seapa sandstone sample, Symposium on Abnormal Subsurface Praasum held st Louiaisma
zero pom pressure (after Fatt), Pore volume compreasi. Stste U., Baton Rouga, April 28, 1967.0 copyright 1969 Amaricsn
bilities are based on porosity of 25 percent. hsatituta of Minin& Matelluraicsl, and Petroleum Ssseineass, k.

1530 JOURNAL OF PETROLEUM TECHNO~OGY

Anda mungkin juga menyukai