You are on page 1of 21

Resume of :

Van Everdingen, A. F and Hurst W., “The Application of Laplace Transformation to Flow Problems in
Reservoirs”, Trans., AIME (1949), 186, 305-324.

Problem:
Reservoir engineers could need a source to obtain fundamental theory and data on the flow of fluids through
permeable media in the unsteady state. The data on the unsteady state flow are composed of solutions of the
equation :

………………………….(1)

Two sets of solutions of this equation are developed, namely, for "the constant terminal pressure case" and "the
constant terminal rate case." In the constant terminal pressure case the pressure at the terminal boundary is
lowered by unity at zero time, kept constant thereafter, and the cumulative amount of fluid flowing across the
boundary is computed, as a function of the time. In the constant terminal rate case a unit rate of production is
made to flow across the terminal boundary (from time zero onward) and the ensuing pressure drop is computed
as a function of the time. Then these two sets of equation are developed by using laplace transform.

Assumption and Limitations :


1. Incompressible radial fluid flow
2. Laminar flow
3. Constant permeability and isotropic
4. Constant rock compressibility
5. Isothermal
6. Constant viscosity and porosity
7. Constant and small fluid compressibility
8. Disregarding the product of gradient compressibility to pressure square
9. Disregarding gravity effect.
Practical Applications :
The application of this case is to approximate the case of :
1. Constant at terminal rate solution
2. Constant pressure at outer boundary
3. Constant rate production case
4. Bounden (No-Flow) outer boundary constant pressure production case.
Discussion:
1. Constant Terminal Pressure Case
The constant terminal presure case is defined as follows: At time zero the pressure at all points in the formation
is constant and equal to unity, and when the well or reservoir is opened, the pressure at the well or reservoir
boundary, r = 1, immediately drops to zero and remains zero for the duration of the production history.
Before continuing, the laplace transform of eq. 1 is :

………………………..(2)
If we treat the constant terminal pressure case symbolically, the solution of the problem at any radius and time
is given by P = P(r,t). The rate of fluid influx per unit sand thickness under these conditions is given by Darcy's
equation :

……………………………(3)

If we wish to determine the cumulative influx of fluid in absolute time T, and having expressed time in the
diffusivity equation as t = KT/fμcRb2 then :

…………………………(4)

Where,

………………………………(5)

In brief, knowing the general solution implied by Eq. 5, which expresses the integration in dimensionless time,
t, of the pressure gradient at radius unity for a pressure drop of one atmosphere, the cumulative influx into the
well bore or into the oil-bearing portion of the field can be determined by Eq. 4. Furthermore, for any pressure
drop, ΔP, Eq. IV·2 expresses the cumulative influx as per unit sand thickness:

……………………………….(6)

When an oil reservoir and the adjoining water-bearing formations are contained between two parallel and
sealing faulting planes, the flow of fluid is essentially parallel to these planes and is "linear." The constant
terminal pressure case can also be applied to this case, then the rate of fluid influx across the original water oil
contact per unit of cross-sectional area for every pressure drop is expressed by:

………………………………..(7)

2. Constant Terminal Rate Case


In the constant terminal rate case it is likewise assumed that initially the pressure everywhere in the formation is
constant but that from the time zero onward the fluid is withdrawn from the well bore or reservoir boundary at a
unit rate. By applying Darcy's equation for the rate of fluid flowing into the well or reservoir per unit sand
thickness :
………………………………….(8)
Which simplifies to :

………………………………(9)

Therefore, for any constant rate of production the cumulative pressure drop at the field radius is given by :

……………………………….(10)

Similarly, for the constant rate of production in linear flow, the cumulative pressure drop is expressed by :

……………………………………(11)

where q(T) is the rate of water encroachment per unit area of crosssection, and P(t) is the cumulative pressure
drop at the sand face per unit rate of production.

3. Superposition Theorem
With these fundamental relationships available. it remains to be shown how the constant pressure case can be
interpreted for variable terminal pressures, or in the constant rate case, for variable rates. The linearity of the
diffusivity equation allows the application of the superposition theorem as a sequence of constant terminal
pressures or constant rates in such a fashion that it reproduces the pressure or production history at the
boundary, r = 1.
The physical significance can best be realized by an application. Fig. 1 shows the pressure decline in the well
bore or a field that has been flowing and for which we wish to obtain the amount of fluid produced. As shown,
the pressure history is reproduced as a series of pressure plateaus which represent a sequence of constant
terminal pressures. Therefore, By the application of Eq. 6, the cumulative fluid produced in time t by the
pressure drop ΔP0, operative since zero time, is expressed by . If we next consider
the pressure drop ΔP1 which occurs in time t1 and treat this as a separate entity, but
take cognizance of its time of inception t1 then the cumulative fluid produced by this increment of pressure drop
by superimposing all these effects of pressure changes, the total influx in time t is expressed as :

……………………….(12)

when t > t3,. To reproduce the smooth curve relationship of Fig. I-A, these pressure plateaus can be taken as
infinitesimally small, which give the summation of Eq. IV-13 by the integral:

…………………………….(13)
By considering variable rates of fluid production, such as shown in Fig. I-B, and reproducing these rates as a
series of constant rate plateaus, for determination of cummulative pressure drop, then Eq. 10 can be changed
into:

………………………………..(14)

If the increments are infinitesimal, or the smooth curve relationship applies, Eq. 14 becomes :

……………………………………..(15)

If q(o) = 0, Eq. 15 can also be expressed as:

…………………………………………(16)

Since Eqs. 12 and 13 are of such simple algebraic forms, they are most practical to use with production history
in making reservoir studies. In applying the pressure or rate plateaus as shown in Fig. 1, it must be realized that
the time interval for each plateau should be taken as small as possible, so as to reproduce within engineering
accuracy the trend of the curves. Naturally, if an exact interpretation is desired, Eqs. 13 and 15 apply.
Figure 1. a) Sequence constant terminal pressures, b) Sequence constant terminal rates
4. The constant rate case, (Infinite Medium)
Boundary conditions for the constant rate case in an infinite medium are that (1) the pressure drop P (r, t) is
zero initially at every point in the formation, and (2) at the radius of the field (r = l) we have:

.........................................(17)

Then, A reference to a text on Bessel functions, such as Karman and Biot pp. 61-63, shows that the general
solution for Eq. 2 is given by:

.....................................(18)

where I o (r √ p) and K o ( r √ p ) are modified Bessel functions of the first and second kind, respectively, and of
zero order. A and B are two constants which satisfy a second order differential equation. Since P(r.p) is the
transform of the pressure drop at a point in the formation, and because at a point not yet affected by
production the absolute pressure equals the initial pressure, it is required that P(r,p) should approach zero as
r becomes large. Therefore, to obey the initial condition, the constant A must equal zero and Eq. 18 become :

..........................................................(19)

By using the laplace transformation for second boundary condition for rate production unit obtained :

................................................................(20)

Then for general solution, Eq. 19 becomes :

.........................................................(21)

By inversion from Eq. 21 with line source solution where t >100, obtained :

...........................................................(22)

For Pressure drop at boundary conditions where t value is large, then :

.........................................................(23)

Then, for linear flow, diffusivity equation from laplace transformation is obtained:

...............................................................(24)

Then, the general solution of Eq. 24, written as :

.................................................(25)
By using the simillar technique, do transformation for pressure drop at x = 0 resulted :

..............................................................(26)

Equation of General solution from Eq. 21 can be changed by using residue methods then resulted:

.....................................................(27)

For, mathematical convienience, eq. 27 then changed into :

..........................................................(28)

5. Constant Terminal Pressure Case (Infinite Medium)


In the constant terminal pressure case it is assumed that at all times the pressure drop at r = 1 will be unity, with
conditon of Eq. 18 which is expressed as a :

.......................................................(29)

Then, Laplace transform in determining pressure at any point with condition of Eq .19 is expressed by :

......................................................(30)

Then, Flow rates in Laplace domain, can be written as :

...........................................................(31)

With application of Mellin’s Inversion in Eq. 31, then obtained :

..................................................(32)

6. No Fluid Flow Across Exterior Boundary (Limited Reservoirs)


The first example considered is the constant terminal pressure case for radial flow of limited extent. The
boundary conditions are such that at the well bore or field's edge, r = 1, the cumulative pressure drop is unity,
and at some distance removed from this boundary at a point in the reservoir r = R, there exists a restriction such
that no fluid can flow past this barrier so that at that point :
............................................................................(33)

The general solution of Eq. 18 still applies, but to fulfill the boundary conditions it is necessary to re-determine
values for constants A and B. The transformation of the boundary condition at r = 1 is expressed as :
....................................................................(34)
and at r = R the condition is
.................................(35)

And then, the solution for A and B variable is obtained :

................................(36)

................................(37)

To find Q(t) the cumulative fluid produced for unit pressure drop, then the transform for the pressure gradient at
r = 1 is obtained as follows:

........................................(38)

Where the negative sign is introduced in order to make Q(t) positive. Theorem B shows that the integration
with respect to time introduces an additional operator p in the denominator to give:

........................................................(39)

Then, by using residue method, obtained:


2

R2−1

e−a t J 12 (an R)
n

Q(t )= −2 ∑ 2 2 ...............................................(40)
2 2
a ,a a n [ J o ( a n )−J 1 ¿ an R ¿ ]
1 2

Likewise from the foregoing work it can be easily shown that the transform of the pressure drop at any point in
the formation in a limited reservoir for the constant terminal rate case, is expressed by :

...................................................(41)

Then, by applying the Cauchy’s theorem to the Meillin’s formula and also applying the Residue Method,
cummulative pressure drop at r = 1 can be calculated using Eq. 42

........................(42)
7. Pressure Fixed at Exterior Boundary (Limited Reservoir)

As a variation on the condition, we may assume that the pressure at r = R is constant. In effect, this assumption
helps to explain approximately the pressure history of flowing a well at a constant rate when, upon opening, the
bottom hole pressure drops very rapidly and then levels out to become constant with time. The case has been
developed by Hurst using a cylinder source and by Muskat using a point source solution. When developing the
solution by means of the Laplace transformation, it is assumed that the exterior boundary r = R, P(R,p) = 0,
which fixes the pressure at the exterior boundary as constant.

Cylindrical source :

................................................ (43)

λ is obtained from :

..................................(44)

Point source solution :

..................................................(45)
Resume of :
Romboutsos, A., & Stewart, G. (1988). A Direct Deconvolution or Convolution Algorithm for Well Test
Analysis", paper SPE 18157 presented at the 63rd SPE Annual Technical Conference and Exhibition.
Houston, TX, Oct, 2-5

Problem :
The convolutional values of pressure and flow rate can be changed so as to provide a response value for the
variable rate. A deconvolution process for transformation allows a constant rate response of the same value to
be determined from the measured pressure and available rate data. Based on this concept, a general equation is
developed to perform and analyze a well test process for variable rates in heterogeneous systems. The Stefest
method is an option from the inversion process of the product of the Laplace transformation solved by
numerical inversion.

Assumption and Limitations :


For Understanding convienience, the assumptions and limits are only carried out in the build-up condition
where the sandface rate is constant and the wellbore storage coefficient are considered to be the same.

Practical Applications :
Case study used on this paper is the Phillips-1 Oil Well, and below is the example of the application:
1. Previously, hydraulic fracturing was performed at initial completion. Furthermore, the well test process
is carried out by means of the buildup test. In deconvolution the total fluid rate data is used to be used in
this case.
2. Figure 2 shows the results of the log-log plot of real data and deconvolution data. The point that can be
seen is a drastic change in the form of early time to the unit slope line equal to 1/2. Deconvolution data
are analyzed using conventional techniques depending on the type of pressure response obtained, which
in this case is a solution for the uniform flux fracture. This is because there is no variable rate effect.
3. In Figure 3 on the semilog plot a straight line can be drawn at the end of the curve. So then we can get
the slope value (m0 = 611,765 where (kh / μ) becomes 87.2 md-ft / cp and it is indicated that there is a
skin effect (S) = -0.6.
4. In Figure 4 is a plot of the deconvolution of the value ΔP vs Δt1 / 2 where a straight line can be drawn at
the early time data giving a slope, m = 19.6, from which the fracture half length is obtained, x f = 81.98
ft. The quality of the data obtained from the tools used and the results of the deconvolution only show a
small noise effect.

Figure 2. Log-Log plot graph between deconvolution and real data


Figure 2. Semilog graph deconvolution data

Figure 3. Cartesian graph plot ΔP vs Δt1/2 using Deconvolution data

Discussion :
1. The concept of numerical Laplace transform with a linear approach can be a solution in obtaining the
transformation of the time function. This is because in the process of convolution and deconvolution the
transformation of pressure and flow rate is easier to do with numerical or analytical transformations.
2. The concept of the standard CRD method can be used in the analysis of reducing the variable rate
response to a constant that is equivalent to the sandface response. Based on this study, the
transformation product can be inverted with the Stehfest algorithm and the Durbin algorithm.
3. The fundamental constant rate response for a certain set of parameters (b and h D) is generated as a PD-tD
table and convoluted with a synthetic flow that is suitable for ideal wellbore storage. This model has an
analytical solution, which can be converted into a laplace domain, so that responses to storage and skins
can be calculated analytically. However, this solution is very complex and time consuming in
computation because it involves the addition of infinite series. The addition of storage and skins by
numerical convolution is very accurate, while computation time is reduced by a factor of 75.
4. In the deconvolution method, the equivalent constant sand face rate response is calculated from one
variable rate. To verify the deconvolution method, conventional analytical models are used to produce
both the ideal constant sand face rate response and the pressure response to the welIbore storage
wellbore storage, which is typically estimated at about five segments per log cycle. In the case of double
porosity, further observations are made. The pressure response with storage is synthetically
deconvoluted using known wellbore storage coefficients. The result can be seen that the deconvolution
data other than the first two points are almost exactly on the theoretical CRD response. Deconvolution
in this case represents the initial slope unit corresponding to the flow through the fracture system.

Point (2A) Solution


Inner Boundary Condition 1

For Inner Boundary Condition 1, Equation showed below:

Wellbore storage effect………..(3)

In the equation 3 explained relationship between dimensionless wellbore


unloading rate with dimensionless sanface flow will be become resulted 1, equation for wellbore storage
constant, C, by which purposed by Van Everdingen hurst is showed below :

………………(6)

The equation above shows the volume of fluid in the borehole that has accumulated. This value is influenced by
compressibility or changes in liquid level. The distortion problem due to wellbore storage can be shown as a
form of flow inequality in early time production between the surface flow rate (q) and the flow rate from
reservoir to wellbore (qsf) which can affect the analysis and interpretation of pressure tests in the wells,
especially when the effect of wellbore storage dominates the pressure response in the wellbore. This can happen
because when the well is shut-in the pressure difference between the reservoir and the wellbore will support the
liquid column to flow into the wellbore to a certain height. At the beginning of production, therefore, the flow
read is the fluid flow that was previously in the well. Then over time the flow rate on the surface (q) with the
flow rate from the reservoir to the wellbore (qsf) will be the same and then the flow rate at the surface (q) will be
constant. The mass equibrillium on the wellbore can be define as:

…………………………………..(7)

To make further understanding convinience, then equation below is made from flowrate form with volume of
wellbore (Vwb)

………………….(8)

From this equation, can be developed into 2 cases :


1. Fluid filled wellbore
2. Rising or falling liquid level

1. Fluid Filled Wellbore

In this case changes of wellbore volume (ΔVwb) can be define as follows :

...................................................(9)

Eq. 9, is the material balance equation for slightly compressible fluids. Then we
change the Eq.9 into limit form related with diffrential time :

Or
....... (10)

For Pseudo steady state condition, we obtained:

and resulted

then by add influx term, then resulted:

......................(11)

And then, by combining Eqs .10 and 11 gives :

.................(12)

By doing the substitution of equation 12 into 8, with qin = qsf dan qout = q, then resulted :

...................(13)

Next, to define wellbore storage coefficient on the form:

Cs = Cwb Vwb ..........................................(14)

And then, by doing substitution Equation 14 to 13 resulted :

....................(15)

2. Rising or falling liquid level

Changes of wellbore volume can be written as :

...................................(16)

In the Eqs. 16 changes of volumetric liquid is identified, because of changes of liquid level, with variable Δt
and in the limit form Δt  0, then

..........(17)

Or

..................(18)
Then, pressure changes caused by rising fluid level can be solved :

..............................................(19)

Where Δz is :

................................................(20)

Then, Eq. 20 is divide by Δt, then obtained:

Or

.........................................(21)

Subtitute Eq. 21 and 18, becomes:

......................(22)

Then with substitute Eq. 22 and 8 resulted :

.....................(23)

Wellbore storage coefficient, Cs in form:

.............................(24)

Then, Eq. 23 becomes:

.....................................(25)

By using general solution, then manipulate with inner boundary condition resulted:

...............................................(26)

Then, Eq. 26 is divided by q, resulted:


..........................................(27)

Where:

.........................................................(28)

Then, recalling dimensionless pressure and dimensionless time definition :

..................................................(29)

.............................................(30)

Table 1 tDC and PDC Value in Every Units


Darcy Units Field Units SI Units
tDC
1 2.637 x 10-4 3.557 x 10-6
PDC
2π 7.081 x 10-3 5.356 x 10-4
Then, Eq. 29 and 30 in every t and P form :
………………………….(31)

…………………………(32)

By considering dP/dt and by substitution Eq. 31 and 32 obtained:

Then, simplifies with elimination:


……………………………(33)

Then, substitute Eq. 33 to 27, obtained:

or

Then, with defining the wellbore storage coefficient:


…………………..(34)

Then obtained:
…………………………(35)

After obtaining the Eq. 35, equation above is the general solution of dimensionless
flowrate related with wellbore phase redistribution, then assuming the constant tubing pressure, then obtained:
…………………….(36)

Or
…………………………(37)

By considering qsf = 0 dan qD = 0. Combining Eq. 36 obtained:

...........................................(38)

By separating dPwf term:


........................................................(39)

With integrate derivation available, then:


.................................................(40)

For understanding can be written as:

Or
.........................................................(40)

Inner Boundary Condition 2


For inner boundary condition 2, by considering skin effect near wellbore, equation showed below:
………………………………….. (4)

By considering skin effect on steady state condition and sand face pressure
drop is proportional with the sand face flowrate, then obtained from Eq. 4 :
…………………………….(41)

Based on Eq. 41 gives value where dimensionless flowing pressure is PD(tD)


showed by Van everdingen and ramey. Then the skin effect showed by Hawkins by :
……………………………………(42)

For very large value of k1 then Eq. 42 changed:


………………………….(43)

And known as :
……………………..(44)

From Eq .4 we can conclude that:


1. Positive skin value, then Dimensionless wellbore pressure will be decrease from skin = 0 and showed by
decreasing in productivity.
2. Negative skin value, then Dimensionless wellbore pressure apparently will be increase at skin =(+) and 0,
and shows increase in pressure around wellbore caused by skin.

Solution 2B
Dimensionless term is given as:

Solution of Eqs. 1 through 5 follows readily using the Laplace transformation as shown by Blackwell. The
transform of the dimensionless flowing pressure may be written as:
......................................................(6)

where Ko and K1 are the modified Bessel functions of the second kind 01 zero and unit orders. An identical
transform was published recently by Russell.

Jaeger also considered determination of the pressure distribution within the radial system, the sand face flux,
and the pressure drop across the skin (within the heat conduction analog, of course). However, these quantities
are only of incidental interest to the present purpose of this study. Blackwell and jaeger presented the heat
conduction analog to the following real inversion integral solution to Eq. 6

.....................(7)

where Jo(u) and J1(u) are the Bessel functions of the first kind of zero and unit orders, and Y 0(u) and Y1(u) are
the Bessel functions of the second kind of the respective orders. Both numerical evaluations and short and long-
time approximations for the integral in Eq. 7 have been presented by a number of authors in connection with
other problems. However, the relationships between them seem to have been largely overlooked.
Another related solution has appeared in the petroleum literature. By using the superposition principle and
representing the well as a continuous line source, the Laplace transform of the dimensionless well pressure may
be shown to be:

...........................................(8)

And Eq. 8 produces the real inversion integral:

....................................(9)
Solution 2C

Using the inverted solution from 2B, which is described as (integrated against tD)

Solution:

Figure 1. PwD vs tD For Infinite Radial System With Storage And Skin Effect.
Fig. 1 is a log-log plot such as would be required if "type-curve" matching of field performance were to be
used. Beck et al. illustrated type-curve matching of borehole unsteady-temperature data to determine thermal
constants of well systems. The similar appearance of the computed curves for various storage constants and
skin effects indicates that this sort of performance data interpretation might be rather poor for fluid flow
systems. There is one useful technique illustrated by Fig. 1, however. All curves approach a unit slope at small
times for finite, positive skin effects, and a skin effect of zero. This condition results because only the first term
of the series in very small times. Thus at very early times.

Fig. 1 indicates that the approximation holds to dimensionless times on the order of 0.5 C to C. This indicates
that early performance is controlled entirely by wellbore unloading. It further indicates that the storage constant
(but only the storage constant) might be determined from early field performance data. That is, a plot of the
logarithm of (Pi - Pwf vs the logarithm of flowing time should have a unIt slope and a value of (q/ C) at t = 1.
Alternatively, the flowing pressure could be plotted vs flowing time on Cartesian coordinates and the slope
would be -( q/ C). This approach may be useful to determine the storage constant where specific information is
not available. Perhaps more important, this analysis indicates that well pressure data cannot be used to
determine either skin effect or flow capacity before times when the dimensionless pressure on Fig. 1 begins to
bend below the early straight line.

A pertinent question is what is the duration of this initial period? If we assume that a time sufficient for the
sand-face flow rate to become 20 percent of the surface flow rate is adequate, then Eq. 10 can be solved
approximately to provide:

tD = 0.2 Cs, s ≠ 0..................................(10)

By using the same criterion, then:

tD = 0.4 C, s = 0....................................(11)

It is recommended that Eq. 11 be used for skin effects ranging from zero to +2 and Eq. 10 be used to estimate
the time for skin effects greater than +2. The reason for this recommendation may be seen by inspection of Fig.
1. The dimensionless times that pressures increase as the skin effect increases above zero. Because Eq. 10 is
only an approximate simplification of a very complex form, it would forecast smaller times than Eq. 11 if the
skin effect is between zero and +2.

In summary, data obtained from the start of a test until dimensionless times specified by the appropriate Eq, 10
or 11 may be used to estimate the wellbore storage coefficient, C. Well-test data obtained for times between the
appropriate Eqs. 10 or 11 may hopefully be interpreted for skin effect and formation flow capacity.

You might also like