SPE 107138 The Effect of Positive Coupling and Negative Inertia On Deliverability of Gas Condensate Wells
SPE 107138 The Effect of Positive Coupling and Negative Inertia On Deliverability of Gas Condensate Wells
by:
Flow after flow pressure test has been used to demonstrate that
a great part of wells in the field are producing under the c
negative effect of inertia and these wells may reach optimum β= (2)
φ ak b
condition of flow increasing velocity to let that the positive
coupling effect overcomes the inertia. This paper explains how
For a gas condensate reservoir if dew point is below the
well test analysis can be one via to detect the influence of rate
wellbore flowing pressure, the liquid is saturated near to well
over the gas condensate wells deliverability.
generating a bank of liquid that interrupts the gas flow
efficiency and increasing the inertial flow. However, when
Background
velocity is increased and interfacial tension is practically low,
The geological setting of the reservoir is a complex structure
the capillary number increases significantly the fluid flow for
belongs to the compressive trend of El Furrial, in North of
both gas and liquid. The capillary number (Nc) is determined
Monagas. It is the result of several phases of deformation that
by:
originated from the northwest. The structure has a high level
of complexity due to the repeated activity of the Pirital μVg
Nc = (3)
megathrust located to the north and of the existence of Urica σ
fault that altered the development of the structure (Figure 1).
The main productive formations are Naricual and It has been demonstrated by different studies, which focused
Cretaceous, which compound a thickness of 2500-ft (Figure on the measurement and correlation of gas and condensate
2). Naricual Formation is sandstone of fine grains with relative permeabilities at near well bore conditions, that the
porosity from 20 to 12% and absolute permeability from 50 to rate effect as well as the negative inertial effect were both
500 mD. More than 70% of total daily production is coming significant in many cases. Nontheless, the negative inertial
from 5 main sand bodies. The Cretaceous Formation is effect over a wide range of conditions can be subordinated to
sandstone of grains very fine with intercalated shale that affect the positive rate effect and therefore the gas relative
porosity and permeability, the average permeability is 10 mD permeability can increase with increasing velocity when
and porosity 10%. significant quantity of condensate is present in the region near
According to data collected in the field show that the the wellbore. Henderson et al.[2, 4, 6] were first to report the
reservoir has a complex column of degraded fluid at under- improvement of relative permeability of condensing systems
saturated condition varying from gas condensate (75% total due to an increase in velocity as well as that caused by a
pore volume) and volatile oil in the lower part of the reservoir. reduction in interfacial tension. This flow behaviour, named as
The dew point pressure is 8,500 psia in the upper part while the positive coupling effect, was subsequently confirmed
the bubble point is 3,800 psia in oil zone. This field produces a experimentally by other investigators [3-5].
rich condensate gas with an average liquid dropout of 25%,
reaching 40% near the transition zone of the reservoir (Figure Methodology
3 presents LDO of the reservoir). Fluids obtained at surface To analyse the effect of velocity, it is necessary to design a
have unusual gravity of 37°API and GOR varies from 4 to 14 test that observes the change in rate and flowing pressure
MSCF/STB. (Pwf) denominated as flow after flow test (FAF). This kind of
test combined with continuous chromatography and
Fundamental Concepts monitoring of gas and oil (at first separation stage) can prove
Gas well deliverability is significantly affected by Non-darcy information while the velocity profiles is changed. Osorio et
flow regime (Non linear flow), when rate is increasing the al. (8) designed a series of pressure transient tests and
total skin grows linearly due to inertial effect then the gas determined that FAF test may be a good choice to evaluate the
relative permeability is reduced with rate (Ali et al, 5). rate effect.
Conceptually, the inertia forces are responsible for keeping As this reservoir has high value of liquid dropout and the
the fluid molecules on the track when moving around the grain Pwf is under dew pressure, we presume that the wells are
minerals. In a porous media, a continuous redirection of flow highly potential to observe the non-darcy flow regime
is taking place as the hydrocarbon molecules are moving considering both inertia and coupling competent each other.
between the grain minerals causing a chaotic flow when Yet, to observe this mechanism; the well should have certain
velocity is increasing (Ursine et al. 7). To analysis the inertia range of conditions for example; gas condensate fluid along to
effect, Reynolds number (Re) is used to recognize the levels of whole column of well. At low rate, it must have low
flow regime using the follow equation: drowdown pressure, no water production, neither hydraulic
kρβq nor naturally fractured, permeability and porosity less than
Re = (1)
Aμ 100 mD and 15% respectively, and no asphaltene/ sand
production record history. As a result, 3 wells have been
When Re is greater than 1.0 it is understood that flow regime selected after a screening made in more than 70 wells.
is dominated by pure inertia, which it can occur in dry gas
wells. In Equation 1 all the parameters are known except for Test design: the current pressure of the reservoir is 7,200 psia
and the flowing pressure at lower choke is about 7,000 psia, it
inertia coefficient (β) that expresses the resistance of rock in
let to design a test with more than 6 changes of 6 hours one,
permitting to fluid flow. The general expression is estimated
ending up with a build up test of 15 hours shutting at surface.
SPE 107138 The Effect of Positive Coupling and Negative Inertia on Deliverability of Gas Condensate Wells 3
For dry gas system, the conventional graph to analyse non- more representative in terms of stability and time. After that
Darcy flow is superposition test, using parallelism theory for we relate to the rest of curves. It is imperative to estimate the
each drawdown may be estimated the skin value for each flow. Non-Darcy flow coefficient (D) for analytic proposes. Figure
Yet, it is totally reliable if skin (S) and gas effective 8 depicts superposition method using parallelism to observe
permeability (ke) are only rely on pressure, but in case of gas the radial flow and determine S and ke.
condensed flow the S and krg are dependent also velocity (4). Figure 9 shows the method of normalized skin versus
Hence, the applicability of superposition test is not suitable to normalised flow rate of real case to estimate mechanical skin
estimate the value of S or krg. However, this procedure will be (Sm) and D, there are 3 points that following a straight line
used only to determine the non-Darcy coefficient (D). and 3 in scattered position. It is important to highlight that
In contrast, the Bourdet Derivative is applied to determine after the choke 5/8” skin value commences to decrease.
the skin and krg for each flow, the values are estimated in the According to this graphic the D is 1.727E-04 Mscfpd-1 and Sm
MTR when radial flow is reached. After that, we build an is 5.25, which corresponds to the last BU taken in Feb-2005.
adjusted Non-Darcy flow plot (skin vs Rate) to evaluate the Nevertheless, we thought that these values estimate of S and
performance of total skin with rate, and finally the ke are not reliable, hence it is important to consider the log-log
deliverability is plotted under pure inertia and combine effects method for those cases no affected seriously by wellbore
of inertia and coupling to diagnostic the well AOF. storage.
Choice of wells: Well #1 produces 2,000 STB/d of liquid and Log-Log Plot for DD flow
12 MMscfpd of gas at 16500 ft on the base of Naricual To evaluate the velocity effect, each flow was analysed using
Formation. Table 1 shows detailed information about rock and the Bourdet Derivative and Horner Graphic. As we mentioned
fluid properties. Figure 4 presents the petrophysic properties before, the non-parallelism observed in semi-log plot and the
along 100 feet of open interval, showing the best average is decreasing in skin value suggest that there is another
about 50mD and 12% for absolute permeability and porosity. mechanism affecting the gas flow. Therefore, at early time
The BU test carried out in Feb-2005 shown a skin of 23 region (ETR) we analyse the effect of velocity. From 6 DD´s
ascribed to condensing banking. Additionally, the EOS, at flows, only 3 of them exhibite the better behaviour in terms of
flowing pressure, describes condensate saturation near de stability, wellbore storage and the presence of radial flow at
wellbore is about 30%. Well #2 produces 950 STB/d of liquid ETR. Figure 10 displays two graphics with the last period
and 5 MMscfpd at 15800 ft. Well#3 was closed by high GOR. flows, figure 10.a shows Horner Plot detecting the radial flow
Even though the reservoir are subjected to pressure at each period; it can observe that when the chokes are
maintenance at 7300 psia and these wells are so far from increasing the ΔΨ/Δq is decreasing. Furthermore, in figure
injector wells, notice in Figure 5, the wells exhibit an 10.b shows Bourdet Plot we observe that the difference
increasing in GOR with time, inferring that condensing between pressure response and its derivative is shorter as flow
banking is formed. is increased. Meaning that the skin is reduced and effective
permeability is increased. As a result, well deliverability
exhibits better performance when velocity is increasing.
Field Example Application
The methodology has been applied to 3 wells, which 2 show
results easier to analyze and one has problem with the Analysis of Results
collected data. For practice proposes, we present the analysis Table 3 shows the results of S and ke estimated from Log-Log
and results of Well #1 in order to avoid crowed. Table 2 plot for each flow, we infer that the reduction of S and the
shows the results of each flow and chromatographic and increasing of ke are due to the competition between inertia and
Figure 6 displays the behavior of pressure when rate is coupling effects.
increasing. The FF test consists of 6 changes of flow and BU To corroborate the analysis made here, Figure 11 (ΔP2
of 15 hours. versus gas rate) plots 6 points keeping a line of 45 degree and
continuity among points. Moreover, notice that the rate is
Semi-log graphic increasing and the differential between ΔP2 is reduced in each
Figure 7 compares the behavior expected and actual of each flow. Suggesting the flowing pressure is enhanced when
DD. In the expected case, we observe that DD’s is increasing velocity is increased.
as rate increases, in the (MTR) when radial flow is stabilized Another plot was built to finalize our analysis, the
straight line are marked at the first flow keeping a parallelism deliverability index (Pressure over flow rate), as it is shown in
in the rest of flows. That means that S is varying linearly. The Figure 12. There are two curves showing pure inertia case and
application of parallelism method permits to calculate the skin inertia/coupling case. The first two points have a good match
at each step. However, Figure 7.b the actual case shows with pure inertia case given an AOF of 95 MMscfpd.
discontinuous behavior as rate is increased. There is not However, notice that the beginning to the flow # 3 the straight
parallelism among curves, what indicating variation of skin line is deviated due to the impact of positive coupling,
and permeability with rate. signifying that this mechanism becomes more important.
Therefore, the gas deliverability is improved at AOF of 114
Non-Darcy flow analysis MMscfpd. Based on the figures 11 and 12, we realize gas and
Eventhough the semilog graphic did not present very well condensate flows are dominated by inertia at lower rate and by
parallelism among curves, we consider the first flow period as
4 G.A. Carvajal, C. Gonzalez, E. Arreaza, C. Cesin, M. Fernandez and J. Bello SPE 107138
coupling at higher rate. Before FAF test, the well has kept Ruíz, Yelitza Ruíz, Angel Caballos and Adbel Mata for their
with a choke of 1/2”, indicating that it has been under the compromise with this study.
inertial effect.
Derived from Log-log analysis, Figure 13 (Skin* versus Nomenclature
normalized gas rate) depicts the new calculation of skin. The A = Cross sectional area
first rates are over the straight line of non-darcy flow meaning k = Absolute permeability
it is in pure inertia zone, but the last 3 point are deviated from ke = effective permeability
non-Darcy line indicating that the gas flow reaches a transition ρ = Fluid density
zone between inertia and coupling being the later more β = Inertial coefficiente
dominant. σ = Interfacial tension
Additionally, the BU final of the test does not reveal the D = Νon-Darcy flow coefficient
existence of two or more zones. Figure 13 shows a BU taken q = Flow rate
in Dec-2005 and compares with the test made in Feb-2005, we z = two phase gas compressibility factor
notice that there are two radial flows indicating the presence μ = Fluid viscosity
of 2 zones. The BU of FAF test does not exhibit a radial flow, Vg = gas velocity
which is likelihood due to either wellbore storage effect or the φ = Porosity
condensing banking is partially removed by those Nc = capillary number
mechanisms. To support the last comment, it can observe in Re = Reynolds number
table 2 that C7+ increases from 20 to 26% as well as the gas CGR = Condensate gas ratio
flow. Nonetheless, further works will be done to support this GOR = Gas Oil Ratio
paragraph using analytic calculation and numerical simulation. LDO = Liquid dropout
Effective permeability has been estimated for each flow. MLDO = maximum liquid dropout
However this information is not suitable for primary proposes, CVV = Condensate vapour volume ratio
and we will do a series of numerical analysis to compar the S = non-Darcy flow Skin
field and simulation data which will be reported in further Sm = mechanical Skin
works.
Finally, the Well#1 is producing under the inertial effects Subscripts
and may be changed to better condition which coupling effect a = porosity exponent for inertial coefficient
is important in order to improve the gas well deliverability. b = permeability exponent for inertial coefficient
The same recomendation can be spread out the rest of wells, g,o = gas or oil
which more than 70% produce at lower rata.
Abbreviation
Conclusion DD : draw down test
1.- The FAF tests experimented in the Santa Bárbara Field FAF : Flow after flow test
permited to evaluate the effects of velocity on deliverability BU : Build up test
gas condensate well TCF : trillion cubit foot
2.- The special well test analysis confirms that the wells OIP : Oil in place
exhibit effect of inertia at lower rate and positive coupling at GIP : Gas in place
higher rate. It has also observed that positive coupling C7+ : heptanus plus
becomes more important when rate is increasing MTR : Middle time region
3.- The deliverability index is improved when rate increases as ETR : Early time region
well as the C7+ increasing assuming condensate banking is rc : reservoir condition
being produced
4.- The conventional Non-Darcy flow analysis through the References
parallelism theory does not permit to analyses the combined 1. Briones, M. et al: “ study of Gas condensate Well
effects of inertia and coupling. peoductivity in Santa Barbara Field, venzuela, by Well test
5.- Log-Log technique gave reasonable results to understand Analysis.”, paper SPE 77538. San Antonio. Texas 2002.
the impact of coupling when rate increases 2. Henderson, G.D. et al: “The Effect of Velocity and
6.- At certain velocity and well condition, the skin shows a Interfacial Tension on the Relative Permeability of Gas
tendency to be reduced by rate. Condensate Fluids in the Wellbore Region,” 8th European
7.- The actual gas condensate deliverability of wells in Santa Symposium on Improved Oil Recovery, Vienna, Austria, May
Bárbara field are producing under the effect of inertia 1995.
8.- These wells can have better flow condition if rate is 3. Boom, W. et al: “Experimental Evidence for Improved
increased at those prevailing conditions which positive Condensate Mobility at Near-Wellbore Flow Conditions,”
coupling may overcome the inertial effects. paper SPE 30766, presented at the SPE Annual Technical
Conference and Exhibition, Dallas, 22-25 October 1995.
Acknowledgement 4. Henderson, G.D. et al: “Measurement and Correlation of
The authors thank the Reservoir Management of PDVSA E&P Gas Condensate Relative Permeability by the Steady-State
for permission to publish this paper. Also the authors want to Method,” SPEJ, (June 1996) p 191-201.
thank those engineers as Jose Ortiz, Carlos Jimenez, Vanessa
SPE 107138 The Effect of Positive Coupling and Negative Inertia on Deliverability of Gas Condensate Wells 5
Tables
Table 1. Rock and fluid properties of Santa Bárbara Field, Naricual
Formation.
Rock and reservoir properties Fluid Properties
Thickness, ft 100 CGR, STB/MMscf 160
Porosity, % 12 API° 36
Permeability, mD 50 Zb, dim 1.35
Rock compres., psi-1 4.86E-6 μg, cps 4.20E-2 @ rc
Well radius, ft 0.35 γg, dim 0.78
Temperature, °F 290 MLDO% 45 Figure 2. Stratigraphic correlation indicating the main reservoirs
Initial Pressure, psia 12,500 Bg, rcf/scf 3.50E-3
Currente Pres., psia 7200 CVV, psi/stb 920
Figure 4. Petrophysical evaluation of Well#1, the average permeability can be 50mD and Porosity 12%. ·
Oil rate
GOR
Gas rate
Figure 5. Production History of Well #·1. Notice the increasing in GOR during 1996-2001.
Figure 6. Flow after Flow test showing changes in pressure when rate is increasing
SPE 107138 The Effect of Positive Coupling and Negative Inertia on Deliverability of Gas Condensate Wells 7
Figure 7a and 7.b. Semi-log plots for expected and actual case, respectively. Notice that DD
pressure increases as flow increases continuously in expected case but and for actual case is
discontinuously.
Figure 8. Superposition method, using parallelism technique to calculate skin at radial flow.
Figure 9. Non-Darcy flow plot to determine Non-Darcy coefficient (D) and mechanical skin (Sm).
8 G.A. Carvajal, C. Gonzalez, E. Arreaza, C. Cesin, M. Fernandez and J. Bello SPE 107138
Figure 10a and 10b. Horner and Log-Log Plot for 3 flows of FF test.
Figure 11. Fetkovich Plot. Notice the straight line with 45º showing good stability of test.
Figure 13. Non-Darcy skin versus Normalized gas rate. Skin value from log-log plot showing zones
where inertia is dominant and coupling and inertia competing together.
Figure 14. Comparison of Build up tests taken December and February 2005. Showing the
radial flow and flow conditions.