Determination of Average Reservoir Pressure From

Build-Up Surveys

D. N. DIETZ KON/NKLlJKEISHELL LABORATORIUM

MEMBER A/ME RIJSWIJK, THE NETHERLANDS

ABSTRACT
A method for determining average reservoir pressure is
q =- rrr.' h¢c ( ~ ) g (2)

presented, which is simpler to apply than that devised by

Matthews, Brons and Hazebroek. For bounded reservoirs, and therefore, combining Eqs. 1 and 2 we have
identical results are obtained if stabilized-flow conditions
prevail. The present method yields inferior results in the ~(r~)=
'Or 'Or
(3 )
transient state. The method can, with a slight modification,
also be used for water-drive reservoirs. which can be integrated to

INTRODUCTION
op q0 , r +S (4)
or 2rrkh r.· r
In the method proposed by Matthews, Brons and Haze-
broek' for determining average reservoir pressure in a mul- The boundary condition can be introduced as
ti-well reservoir, the cumulative production (the produc-
tion time of each well) enters twice: once when the build- op
-= 0 at r = rb •
up is plotted against In (t + t::,.t) / M to arrive at p*, and ,or -
a second time when the correction, p* - p is determined
with one of the several formulas for differently shaped so that Eq. 4 can be rewritten
drainage areas.
Once a steady state has been attained, however, the prev-
op _ ~(_1
,'Or - 2rrkh
__ ~)
r r,,'
(6)
ious production history should be immaterial. The same
pressure distribution could have been arrived at after dif- A second integration gives
ferent cumulative productions of the individual wells. In
principle, therefore, it should be possible to determine av- qv ( r' ) (7)
p = 2rrkh In r - 2r,,' + C, .
erage pressures without referring to cumulative produc-
tions.
AVERAGE PRESSURE OF
In the following paragraphs an expression is presented THE DRAINAGE AREA
for the difference between the pressure in a producing well
The average pressure can be expressed as
and the average pressure of its drainage area. Then the
build-up time needed to overcome this difference is indi- VIJ Vb

cated.
P = ~Ip
7rr lJ
27Trdr = -.;.I
rb
prdr (8)
o o
CASE OF A CIRCULAR DRAINAGE AREA
AND A CENTRAL WELL Substitution of Eq. 7 and integration yields
PRESSURE DISTRIBUTION
BEFORE SHUT-IN __ q,.L { In rb 3 }
The general differential equation of radial flow (see P = rrkh -2--8 + C, (9)
Muskat,' Eq. 10.2) may be written
BUILD·UP
k
-2rrh - 0 ( r -op) op.
- = 2rrh¢c r - (1) If Eq. 7 is applied to the well bore at r w, and if C2 is
o or or at eliminated by the combination with Eq. 9, we have for the
pressure at the well under steady-state conditions
In the steady state, the rate of production of the well is
equal to the rate of expansion of the fluid contained in the . q0 ( r. 3 )
the drainage area; thus Pw = P - 2rrkh In-;::- - 4 . (10)

Original manuscript received in Society of Petroleum Engineers office After shut-in, and as long as the physical boundaries of
March 22, 1965. Revised manuscript of SPE 1156 received June 24,
1965. the entire reservoir have no influence, the pressure in-
lReferences given at end of paper. creases according to the well-known expression
Discussion of this and all following technical papers is invited. Dis-
cussion in writing (three copies) may be sent to. the office of the Jour-
nal of Petroleu'n/, Technology, Any discussion offered after Dec. 31.
1965, should be in the form of a new paper. No discussion should exceed (11)
10 per cent of the manuscript being discussed.

AUGUST. 1965 955

which, for CPJ)'cr,//4kM < 0.01, can be approximated by consideration that Eq. 16 is reduced to
qp. ( cpfJ,cr",,) p* - p kt
t::.pw = - 47rkh . 0.5772 + In 4kt::.t· (12) ---:-c~--;-
qf.J,/47rkh
= In C 4 for - - =
cpf.J,CA
(20)

Addition of Eqs. 10 and 12 yields an expression for the Therefore, 1n C, can be read from the straight-line parts
straight-line part of the build-up curve if pressure is plotted of the curves in Figs. 2 through 8 of the reference, or from
against 1nt::.t, thus their extrapolations, at the abscissa value of kt/cpfJ,cA = 1.
The results are listed in Table 1.
p,,(t::.t) =
-
p -
qfJ,
47rkh
r In r,,'
r '
3
-2 + 0.5772 For smaller values of t, where the curves of Figs. 2
w through 8 of Ref. 1 are not straight lines, Eq. 16 does not
indicate an actual curve. In those cases, the p following
+ In cp~tcr",' ] . (13) from Eq. 19, which depends on Eq. 16 representing an
4kM
actual curve, cannot be corrected. The range of validity of
The well radius cancels out, as a skin-factor would have the present method can therefore be established by observ-
done if it had been introduced; therefore ing where the graphs in Figs. 2 through 8 of Ref. 1 start to
deviate from straight lines. The limits of validity thus
-
p",(.t::.t) -
-
P - 47rkh
qfJ, [,cpf.J,cr,,'
In 4kt::.t - 0.9228
J. (14) found are presented in the last column of Table 1.

On this straight line, or on its extrapolation, the value of APPLICATION OF METHOD

p is found at the point where the form in brackets van-
ishes. For this purpose, the line should be read at 1. Divide the reservoir on a map into drainage areas
proportional to offtake rates per unit sand thickness of
cpf.J,cr,' cpf.J,cr,,' cp,fJ,cA wells, as prescribed in Ref. 1.
4k eO.",,; (15)
10.07 k = 31.6k 2. Plot pressure build-ups against 1n ,M, and determine
k/ f.J, is known from the slope of the build-up curve, and k/fJ, from slopes.
cP, c and A can be determined by known methods. 3. Determine t, cP, c and A by generally known methods.
4. Check applicability of method by comparing kt/
DRAINAGE AREAS HAVING DIVERSE SHAPES cpfJ,cA with required values in the table.
AND WELL POSITIONS 5. Select C A from the table and read straight-line parts
Eq. 15 is applicable only to a circular drainage area and of pressure build-ups or their extrapolations at t::.tp =
a central well. Although similar expressions could be ob- cpfJ,cA/Ck for p.
tained in a like manner for other drainage areas, it is much
more convenient to use some of the figures presented in DISCUSSION
Ref. 1. The curves of Figs. 2 through 8 in Ref. 1 become
straight lines for sufficiently large values of t. These In a discussion with Matthews, Brcllls and Haze~roek
straight-line parts can be represented by equations such as and with other build-up experts, it was pointed out that the
Matthews-Brons-Hazebroek method worked accurately alsJ
p* - P In---,
Ckt in the transient state, although the assumption of steady
(16)
qfJ,/47rkh cpfJ,cA state is used in the division into drainage areas. The pres-
ent method relies more heavily on this assumption, and, in
where C A is a constant dependent on the shape of the area the transient state, becomes increasingly inaccurate for
and on the well position. smaller production times. Identical results are obtained in
p* is a point at .t::.t = 00 on a straight-line extrapolation the steady state, and in this region the present method may
of the build-up plot vs In (t+ t::.t) / M. P can be consid- be preferred for its simplicity.
ered as a point at .Mp on a straight-line extrapolation of the
build-up plot vs In t/t::.t, since the early part of the build-
ADAPTATION TO WATER-DRIVE RESERVOIRS
up can be approximated by
COMPLETE WATER DRIVE
At) *- qfJ, 1 t + t::.t
Pw ( u = P 47rkh n t::.t Under complete water drive the pressure at any point
tends to become constant. Drainage areas, defined in the
;:=:::p* - ~ In_t_ usual sense, have very irregular shapes, each one having to
(17) be in contact with the advancing water front. In this case
47rkh ,t::.t
it is preferable to divide the reservoir as regularly as pos-
Extrapolating the straight-line part of this type of plot sible into what, in accordance with D. R. Horner,' can be
is equivalent to using the latter expression beyond the val- called associated reservoir areas allocated to the wells.
idity of the above approximation. By definition, the point In a regularly drilled field the associated areas can be
(p, .Mp) is on the straight-line extrapolation, and this pair
approximated by circles. The pressure distribution can be
of values should satisfy Eq. 17, which leads to
irregular along the boundary of such an area, but it should
- . qfJ, t be constant with time. The pressure at any point in the
p= p " - - - l n - - (18) area will be the sum of a pressure P" due to the pressure
47rkh Mp
distribution at the boundary if the well were not produc-
Combination of Eqs. 16 and 18 yields ing, and a negative pressure p, due to the withdrawal by
the well, zero pressure being assumed along the boundary.
cpfJ,cA
.Mp= c:k' (19) According to Frank and v. Mises,' (Eq. 12),
(21)
which is a generalized form of Eq. 15.
The shape factor C A can be obtained from Ref. 1 by the As the total flew originates from the bcmndary, the sec-

956 JOURNAL OF PETROLEUM TECHNOLOGY

 TABLE 1

Stabilized Stabilized
conditions conditions
kf kf ,
farTA>
I-tc In C, C,' for CPflcA .>
~ C,
- -

In bounded reservoirs
EEj' 2.38 10.8 0.3

() 3.45 31.6 0.1

2

EB' 2
1.58 4.86 1.0

0 3.43 30.9 0.1

E£j'
2
0.73 2.07 0.8

(] 3.45 31.6 0.1 !

4
I, 1.00 2.72 0.8

!4 • II -1.46 0.232 2.5

8 3.32 27.6 0.2

I
• I -2.16 0.115 3.0

Q 3.30 27.1 0.2

m 1.22 3.39 0.6

~(tG I
3.09 21.9 0.4
~I
2
1.14 3.13 0.3

EEtfjI -0.50 0.607 LO

0' 2
3.12 22.6 0.2
2

•
4 I' 1.68 5.38 0.7
EEEE} 2
-2.20 0.111 1.2

I'
•
3[~}
2.36 0.7
0.86 -2.32 0.098 0.9
5

EE
In water·drive reservoirs

2.56 12.9 0.6

8 2.95 19.1 0.1

8j 1.52 4.57 0.5

In reservoirs of unknown production charader

AlH;UST, 1965
8 3.22 25 0.1

957
ond pressure field is described by P = reservoir pressure dependent on place and
_ qp. r time, atm
P2 - 2rrkh In r. . (22) Pw = well pressure, atm
p = average pressure in drainage area or associat-'
Integration and division by rrr,,' lead to ed reservoir area, atm
qp. p* = closed-in pressure linearly extrap<:llated on
P2 = - 4rrkh . (23) plot against In (t + M) I M for infinite c1osed-
in time, atm
Application of Eq. 22 to the well radius gives P, = pressure field due only to pressure distribu-
tion on boundary, atm
_ qp. r", p, = pressure field due only to withdrawal by well,
P'w - 2 7T kh I nrlJ- (24)
atm
The well pressure before closing in can be expressed as POI = pressure field due only to oil expansion, atm
!::..p", = pressure increase at the well since closing in,
Pm = p+ 4!~h( I +2 In :,:», (25)
(aplat)" =
atm
general rate of pressure drop in stabilized res-
ervoir, atm sec-I

and after closing in as

q = production rate of well before closing in, cc

p,,(~t) 4!~h{
sec-1
= P+ 1+2 In ::: q" = rate of oil expansion in associated reservoir
area, cc sec-1
- 0.5772 - In ¢p.cr"'l (26) r = distance to well center, cm
4k~t ,
r" = outer radius of circular drainage area or asso-
from which it follows that ciated reservoir area, cm
r", = well radius, cm
_ ¢p.cr,,' <pp.cr,,' 9',.ucA
!::..t- - - - - , = --- = ---- . (27) t = corrected production time, defined as cumu-
p 4ke'''-' 6.1 k 19.1 k
lative well production divided by rate before
APPLICA TIo.N o.F PRo.Po.SED closing-in, sec
METHo.D TO. RESERVOIRS OF !::..t = closed-in time, sec
UNKNo.WN PRo.DUCTIo.N CHARACTER
!::..tji = defined by p", (Mp) = P on linear extrapola-
The usual purpose of average pressure determinations is tion of plot against InM, sec
to calculate by material balance the strength of the water
drive. It may therefore appear that the problem has en-
p. = viscosity of reservoir fluid, cp
tered a vicious circle. As shown in the Appendix, bounded ¢ = porosity, fraction
reservoirs also can be analyzed after division into associat-
ed reservoir areas, rather than into drainage areas. The ACKNOWLEDGMENTS
same constant C A = 31.6 for a circle is applicable in the
P. Hazebroek kindly made himself available as a spar-
latter case. When there is doubt about the amount of wa-
ring partner in discussions on the validity of the above
ter drive, it is recommended that the reservoir be divided
theories.
regularly into associated reservoir areas and that the inter-
mediate value C A = 25 be used for the circularized area. The permission of Shell Research N.V. Amsterdam, The
The inevitable range of uncertainty from 19.1 to 31.6 Netherlands, to publish this paper is acknowledged.
should lead to errors no worse than those from the uncer-
tainty in the compressibility. These errors in .Mp have but REFERENCES
little influence on p, because p is read from a semi-loga- 1. Matthews, C. S., Brons, F. and Hazebroek, P.: "A Method for
rithmic plot. Determination of Average Pressure in a Bounded Reservoir",
Trans., AIME (1954) 201,182.
NOMENCLATURE 2. Muskat, Morris: Th'e Flow 0/ Homogeneous Fluid Through
Porous Media, McGraw-Hili Book Co., Inc., New York (1937).
The formulas are suitable for any consistent system of 3. Horner, D. R.: "Average Reservoir Pressure", Proc., Fourth
units. The units indicated below will serve as an example. World Petroleum Congress, 131.
4. Frank and v. Mises: "Die Differential and Integralgleichun-
A = drainage area or associated reservoir area, gen", Teil Auflage, 691.
sq cm
C A = constant dependent on shape of area, position APPENDIX
of well and on production character
A CIRCULAR ASSOCIATED RESERVOIR AREA
C" C, = Integration constants, atm IN A BOUNDED RESERVOIR
Ei = exponential integral defined by
We introduce q,. as the rate at which oil is expanding
Cfl e" within the associated area because of the general pressure
Ei(-x) = -
J
x
-du
u
decline (aplat)". This rate will differ from the production
rate q of the well to the extent that the associated area
differs from the drainage area. In the case of a non-pro-
c = effective compressibility of reservoir fluid, ducing (observation) well, the outward flow through the
atm- ' boundary of the associated area would be q".
h = sand thickness, cm The pressure field is now split into three simpler ones:
k = permeability, darcies P, due to the pressure distribution along the boundary at

9511 JOURNAL OF PETROLEUM TECHNOLOGY

the time of closing in, while this distribution is considereJ The average value is then found to be
constant with time; p, (negative) caused by the withdrawal TO

= --,J P. 27rrdr = 8q efL

by the well, constant zero pressure being assumed along
- 1 (36)
the boundary; and PJ as it would occur for a uniform rate p .• kh '
7[' Y II 7r
of pressure drop, if the well were not producing and if o
the pressure everywhere along the boundary just reached
zero. and the value at the well
The first two fields are identical to p, and p, discussed q·fL
for water-drive reservoirs. Therefore, P"w = 47rkh . (37)

(21) Addition of the three fields yields the average pressure

(23) (38 )

and the well pressure before shut-in

(24)
(39)
The third field is governed by

k 0 (
--2r.h- r --OP,,) = 2Trh¢cr - (oP . ) (28)
or
fL or or
which can be rewritten as
at "
p," =
-
P +
qfL
47rkh
(1 + 2 In
r", )
r, q,fL
+ 87rkh' (40)

_0
or
(r _OP,,)
or
=
-
qr/-1r
Trkh r,,' .
(29)
After closing in, the well pressure can now be expressed by

- qfL { 1 r",
Integration gives Pw (~t) = P + 47rkh + 2 1n -,:;:- 0.5772

op I r.'Jl.tcrw'l q,fL
(30)
or n 4kM-( + 87rkh

The condition at the inner boundary

= P +~~_L -:- 2 In ~ - 0.5772
47rkh \ 2 rb
-
op" = 0 for r = 0 (31 )
or I q)fL Cr ",' l (q,. - q)'l

reduces Eq. 30 to
- n -4kM f+ 8·.TAh (41)

The bracket form again vanishes for

op" q,fLr
(32)
~ = 2Trkhr,,'
~t-J!uncorrected cA
-
= ¢fL (42)
31.6k'
Further integration leads to
and the true average may be expressed as
q,fL , +C (33 )
p" = - 47rkhr,,' r " -. - (q, - q)u
p('ol'rect == pUIl('orl'c('tC'd - (43)
which, after introduction of the outer boundary condition 8r.kh

p" = 0 for r = ro (34) The indicated correction is but a minor one. Moreover,
in the determination of the average pressure of the entire
becomes reservoir, it cancels out because
"i,q, = 'i,q (44)
(35)
***

AUGUST, 1965 959