Delshad 2009

Download as pdf or txt
Download as pdf or txt
You are on page 1of 14

SPE 118951

Development of a Reservoir Simulator for Souring Predictions


Farhadinia, M. A., Bryant, S. L., Sepehrnoori, K., Delshad, M., The University of Texas at Austin

Copyright 2009, Society of Petroleum Engineers

This paper was prepared for presentation at the 2009 SPE Reservoir Simulation Symposium held in The Woodlands, Texas, USA, 2–4 February 2009.

This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have not been
reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, its
officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Society of Petroleum Engineers is prohibited. Permission to
reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.

Abstract
During the last two decades, several simple models have been developed to predict the onset of reservoir souring in seawater
injected reservoirs. The key mechanism to generate hydrogen sulfide is a biological reaction between sulfate in the injection
water and volatile fatty acids in the formation water in the presence of sulfate reducing bacteria (SRB). The produced
hydrogen sulfide interacts with rock and partitions between oil and water phases. Comparison of field data with reservoir
souring model predictions often shows inconsistent results. We present the development of a comprehensive reservoir
souring model in a chemical flooding simulator that accounts for these mechanisms.
Being able to estimate the likelihood and timing of the onset of H2S production would permit more realistic assessments
of project economics. We incorporated mechanisms of generation and transportation of H2S in porous media to develop a
reservoir souring simulator to predict the onset of souring in oil reservoirs.
We implemented a general souring model in a 3D finite difference compositional non-isothermal reservoir simulator. The
results indicate that depending on the type of SRB, the temperature propagation in the formation determines the onset of
biological reactions and consequently, the generation of different hydrogen sulfide concentrations. The lag in the temperature
front with respect to the injection front can cause a delay in the observed souring if the temperature is not favorable for SRB
activation.
A predictive model would enable operators to make better decisions for remedial actions to either prevent souring or to
mitigate its impact. The developed finite difference simulator is 3D and accurately accounts for variation and impact of in-
situ concentrations, temperature, pressure, and reservoir heterogeneity on H2S transport and production.

Introduction
Reservoir souring is the process of the production of hydrogen sulfide in a seawater injected reservoir. Using the knowledge
of the mechanisms of generation and transportation of hydrogen sulfide in the reservoir, several reservoir souring models
have been developed (Ligthelm et al., 1991; Sunde et al., 1993; Eden et al., 1993). The degree of exactness and reliability of
these models depend on their capabilities to mimic the essential parameters which determine the generation and
transportation of the hydrogen sulfide in the porous media.
Figure 1 shows the process of reservoir souring. While injecting cold sea water which contains sulfate, nitrate, phosphate,
and SRB into the hot formation, which provides organic acids, in the presence of SRB, sulfate reacts with organic acids to
produce hydrogen sulfide. The produced hydrogen sulfide interacts with rock surfaces and partitions between oil and water
phases (Ligthelm et al., 1991). The expected concentrations and temperature profiles are shown in Figure 2. The temperature
distribution ranges from sea water (Tw) to the reservoir (Tres) temperatures. The activities of SRB, which are responsible for
souring, depend on the temperature distribution and available nutrients (Herbert et al., 1985). At low temperatures,
mesophiles, and at high temperatures thermophiles or hyperthermophiles, are activated and the biological reaction between
sulfate and organic acids will initiate. Table 1 shows the range of activation of the discussed SRB.

Modeling of the Reservoir Souring Process


Modeling and simulation of reservoir souring can be summarized in two processes: generation and transportation of hydrogen
sulfide in the porous media. These two processes are described below:

Biological Generation of Hydrogen Sulfide. It is well established that some kind of SRB is responsible for the H2S
generation in water flooded reservoirs. Although it is possible that some inorganic reactions can produce H2S, under the
prevailing reservoir conditions they cannot be considered as a source of souring (Khatib et al., 1997).
2 SPE 118951

SRB
Substrate + Electron acceptor Cells + H2S + CO2 (1)
(Formation water) (Injection water)
-3
(CH3COOH, PO4 ) (Sulfate, NO3-, PO4-3, SRB)

Transportation of Hydrogen Sulfide. The flow and transport of reactive components, such as hydrogen sulfide, can be
modeled using the retardation factor and the boundary conditions imposed by the injection and production wells. The
retardation factor may be the result of interaction with rock surfaces (e.g., adsorption) or due to partitioning between oil and
water phases, or a combination of both. Modeling of the partitioning and adsorption of a component is explained in the
following.

Partitioning. Partitioning of a component between two phases is modeled using a constant partition coefficient (K-factor
approach), which asserts that the ratio of component concentrations in the two phases is a constant
CTotal ,Oil
( K ( mass ) = ). The retardation factor due to partitioning is
CTotal ,Water
CT2
A1 = 1 + (2)
CT1
For H2S, the partitioning coefficient (K) is a function of temperature, but only depends weakly on pressure (Ligthelm et
al., 1991; Eden et al., 1993; UTCHEM Technical documentation, 2000).

Adsorption. The adsorption capacity of a component on the formation rock surface is defined as grams of the adsorbed
component to the gram of rock (aT). DS , the retardation factor parameter, is the ratio of the average concentration of the
adsorbed component to its concentration in the flowing phase, as expressed in Equation 2:
CT (1 − φ )ρr a T
Ds = = (3)
C Tl φρl CTl
The retardation due to adsorption is formulated as follows (UTCHEM Technical documentation, 2000):
RET=1+ DS (4)

A comprehensive predictive simulator should have the capabilities to describe the mechanisms of generation and
transportation of hydrogen sulfide under different reservoir conditions and characteristics.

UTCHEM Simulator
UTCHEM is a three-dimensional, multicomponent, and multiphase reservoir simulator. UTCHEM was developed at The
University of Texas at Austin for enhanced oil recovery processes which use surfactants and polymers (UTCHEM Technical
documentation, 2000). The simulator handles geochemical and biological reactions and includes an energy balance. These
features are vital for modeling reservoir souring.
In the simulation process, UTCHEM first solves for pressure distribution implicitly, then the concentration profiles are
solved explicitly. After solving the energy balance equations, the biological reactions which generate H2S are handled. This is
the system of ordinary differential equations which describes the reaction rate for each species solved in each time-step for
every gridblock (see Appendix).

Biodegradation Reactions. The biodegradation option of UTCHEM is based on Molz et al., 1986 (de Blanc et al., 1996). In
order to simulate a typical biological reaction we used the simplified form of the biological option in UTCHEM (Delshad et
al., 1996):
dS μ X ⎛ S ⎞⎛ A ⎞
= − m ax ⎜⎜ ⎟⎟ ⎜⎜ ⎟⎟ − K a b io S (5)
dt Y ⎝ KS +S ⎠ ⎝ K A+ A ⎠
dA μ XE ⎛ S ⎞⎛ A ⎞
= − m ax ⎜⎜ ⎟⎟ ⎜⎜ ⎟⎟ (6)
dt Y ⎝ KS +S ⎠ ⎝ K A+ A ⎠
SPE 118951 3

When nutrients such as nitrogen and phosphorous limit the reaction, the substrate utilization must be modified similar to
the following equation:
dS μ X ⎛ S ⎞⎛ A ⎞⎛ N ⎞
r = = − max ⎜⎜ ⎟⎟ ⎜⎜ ⎟⎟ ⎜⎜ ⎟⎟ (7)
⎝ KS +S ⎠ ⎝ K A + A ⎠ ⎝ K N + N
s dt Y ⎠
In order to simulate the reservoir souring process, the partitioning and adsorption capabilities are also needed, in addition
to the current biological option. Both the partitioning and retardation due to adsorption is modeled using the tracer option in
the UTCHEM simulator.

Field Application of the Developed Model and Simulator


In this section, the capabilities of our simulator in the prediction of the reservoir souring in 3D reservoirs are discussed. This
simulation study shows that the developed simulator is capable of handling a complicated case which may be encountered in
real fields.
In this case study, a reservoir which consists of three layers of varying properties (stochastic estimation of properties as
indicated in Tables 2 and 3) and thirteen wells has been defined. The reservoir has a dimension of 623.2ft by 623.2ft by 40 ft.
The reservoir properties and conditions are given in Tables 2-7. Whereas Tables 2-4 show reservoir characteristcs and
conditions, Table 5 provides the thermal properties of rock and fluids. Table 6 gives the simulation conditions and Table 7
represents the initial concentrations of the chemical and biological species engaged in the reservoir souring process. Figure 3
also illustrates the location of the wells in the model where the circles represent the production wells, triangles represent the
injection wells, and wi represents the well number.

Results of Simulation Studies


The overall results of the prediction of hydrogen sulfide in the production wells are given in Figures 4 and 5. The results
show that maximum concentration of the observed hydrogen sulfide in well number 7 occurs around 400 days (the earlier
breakthrough) and well number 10 is about 900 days (the late breakthrough). Figure 3 confirms that the earlier observed
souring in well number 7 compared to other wells is due to the shortest distance between injectors and producers.
Furthermore, well number 7 is surrounded by four injectors which cause higher production rates and consequently earlier
breakthrough.
Figures 6-13 show the propagation of the essential physical and chemical variables which control the reservoir souring in
this reservoir model. Figures 6 shows the distribution of a water tracer after 2 years of seawater injection. The tracer will
distribute between injectors and producers while the corners of the reservoir remain untouched. Similar distribution can be
observed for the temperature fronts as shown in Figure 7. Figure 8 shows the hydrogen sulfide distribution in the reservoir
after 2 years of sea water injection. Hydrogen sulfide distribution shows that after 2 years (Figure 8) the mixing zone has
passed the production wells (Figures 4 and 5). Thus, the maximum hydrogen sulfide concentration is not within the reservoir
and as time passes the remaining concentration of hydrogen sulfide in the reservoir will decrease.
Distribution of the nitrate (from injected water) and phosphate (in formation water) after two years are shown in Figures 9
and 10, respectively. The concentrations in the corners for nitrate and phosphate are in reverse order, while distribution of
nitrate from injection water (Figure 9) is less in corner points which is less affected by injection water. On the other hand, in
the corners the distribution of phosphate in formation water remains unchanged at its initial concentration and decreases
where fluid flow occurs. The SRB, which is produced within the reservoir, has a distribution (Figure 11) similar to phosphate
(Figure 10). This observation also can be explained in terms of migration of the mixing zone and the unaffected corners. The
distribution of sulfate (from injected water) and acetate (in formation) are illustrated in Figure 12 and 13, respectively. As
expected, the acetate distribution (Figure 12) shows higher concentration in unaffected regions, like corner areas, while the
concentration decreases where fluid flow occurs. This conclusion is reversed for sulfate which is injected by seawater (Figure
13).

Conclusions
• A three-dimensional simulator for the prediction of the onset of reservoir souring in seawater injected reservoir was
developed.
• The effects of variables such as temperature and available nutrients on the generation of hydrogen sulfide in porous
media can be included in the model.
• Interaction of hydrogen sulfide with rock surfaces and partitioning between oil and water phases also are considered
in simulation.
• The simulator can consider the heterogeneity of the reservoir which is necessary for field applications.
4 SPE 118951

Nomenclature
A electron acceptor concentration (mg/l)
A electron acceptor concentration in attached biomass (ML3)
A1 retardation coefficient
aT microgram/gram rock
b endogenous decay coefficient (T-1)
CT2 concentration in phase 2, oil
CT1 concentration in phase 1, water
C T average adsorbed concentration
CTl flowing concentration in phase l
Ds ratio of adsorbed concentration to flowing concentration
K partitioning coefficient
E mass of electron acceptor consumed per mass of substrate biodegraded
K electron acceptor half-saturation coefficient (ML3)
A

K abio first-order reaction rate coefficient (for abiotic decay reactions, T1)

KN limiting nutrient half-saturation coefficient concentration (ML3)

KP nutrient half saturation (mg/l)

KS substrate half saturation coefficient (mg/l)


KSO4 sulfate half saturation coefficient (mg/l)
mC mass of cells in a single micro colony
N concentration of a limiting nutrient (ML3)
RET retardation due to adsorption
rS rate of substrate utilization, ML-3t-1
S aqueous phase substrate concentration, ML-3
substrate concentration in attached biomass, ML-3
S
SRB sulfur reducing bacteria
Tw injection water temperature
Tres reservoir initial temperatur
t time, T
VC volume of a single micro colony, L3
X aqueous phase (unattached) biomass concentration, ML3
X attached biomass concentration; mass of attached cells per volume of aqueous phase, ML3
Y yield coefficient, mass of cells produced per mass of substrate biodegraded
Greek Symbols
β surface area of a single micro colony, L2
κ mass transfer coefficient (in biological reactions), LT-1
φ porosity
ρκ
density of pure component κ at reference phase pressure, ML −3
ρl density of flowing phase, ML −3
ρr rock density
μ specific growth rate (1/day)
μ max −1
maximum specific growth rate, t

References
de Blanc, C. P., McKinney, D. C., Speitel Jr., G. E., Sepehrnoori, K., and Delshad, M., “A 3D NAP Flow and Biodegradation Model,”
Proceedings of the Specialty Conference held in conjunction with the ASCE National Convention, Washington, D.C., November 12-
14, 1996.
Delshad, M., Pope, G. A., and Sepehrnoori, K., “A compositional simulator for modeling surfactant enhanced aquifer remediation, 1
Formulation,” Journal of Contaminant Hydrology, 23, 303-327, 1996.
Eden, B., Laycock, P. J., and Fielder, M., “Oilfield Reservoir Souring,” OTH 92 385, HSE Books, 1993.
Farhadinia, M. A., Bryant, S. L., Sepehrnoori, K., “Development and Implementation of a Multi-dimensional Reservoir Souring Module in
a Chemical Flooding Simulator,” Ph. D. Dissertation, The University of Texas at Austin, Fall 2005.
SPE 118951 5

Herbert, B. N., Gilbert, P. D., Stochdale, H., and Watkinson, R. J., “Factors Controlling the Activity of Sulphate-Reducing Bacteria in
Reservoirs During Water Injection,” Paper SPE 13978 available from SPE, Richardson, Texas, 1985.
Khatib, Z. I., and Salanitro, P. J., “Reservoir Souring: Analysis of Surveys and Experience in Sour Waterfloods,” Paper SPE 38795,
presented at the 1997 SPE Annual Technical Conference and Exhibition held in San Antonio, Texas, October 5-8, 1997.
Leu, J. Y., McGovern, C. P., Poter A. J. R., and Hmilton, W. A., “The Same Species of Sulphate-Reducing Desulfomicrobium Occur in
Different Oil Field Environments in the North Sea,” Letters in Applied Microbiology, 29, 246-252, 1999.
Ligthelm, D. J., de Boer, R. B., and Brint, J. F., “Reservoir Souring: An Analytical Model for H2S Generation and Transportation in an Oil
Reservoir Owing to Bacteria Activity,” Paper SPE 25197 presented at the 1991 Offshore Europe Conference held in Aberdeen,
September 3-6, 1991.
Molz, F. J., Widdowson, M.A., and Benefield, L. D., “Simulation of Microbial Growth Dynamics Coupled to Nutrient and Oxygen
Transport in Porous Media,” Water Resources Research, 22, No.8, 1207-1216, August 1986.
Okabe, S., and Characklists, W. G., “Effects of Temperature and Phosphorous Concentration on Microbial Sulfate Reduction by
Desulfovibrio Desulfuricans,” Biotechnology and Bioengineering, 39, 1031-1042, 1992.
Sunde, E. Beeder, J., Nilson, R.K., and Torsvik, T., “Aerobic Microbial Enhanced Oil Recovery for Ofshore Use,” Paper SPE 25197
presented at the 1992 SPE/DOE Eight Symposium on Enhanced Oil Recovery held in Tulsa, Oklahoma, April 22-24, 1992.
Sunde, E. and Thorstenson, T., “Field-Related Mathematical Model to Predict and Reduce Reservoir Souring,” Paper SPE 25197 presented
at the SPE International Symposium on Oilfield Chemistry held in New Orleans, LA, March 2-5, 1993.
UTCHEM, “UTCHEM Technical Documentation for UTCHEM – 9.0,” Reservoir Engineering Research Program, Center for Petroleum
and Geosystems Engineering, The University of Texas at Austin, 2000.

Appendix
Biodegradation Reactions
The biodegradation model is based on the Molz model (de Blanc et al., 1996). Any number of substrates, electron acceptors,
and biological species can be modeled simultaneously. The equations which relate the concentration of species both in
attached and free cell reactions are given as below (de Blanc, et al., 1996):
dS βκ X μ m ax X ⎛ S ⎞⎛ A ⎞
= − (S − S ) − ⎜⎜ ⎟⎟ ⎜⎜ ⎟⎟ − K a b io S ; (A1)
dt mc Y ⎝ K S + S ⎠⎝ K A + A ⎠
d S βκ μ ρ
m ax X ⎜ ⎛ S ⎞ ⎛ A ⎞
= (S − S ) − ⎟⎜
⎜ K + S ⎟ ⎜ K + A ⎟ − K a b io S
⎟ ; (A2)
dt V c Y ⎝ S ⎠ ⎝ A ⎠
dA βκ X μ m ax X E ⎛ S ⎞ ⎛ A ⎞
= − ( A − A) − ⎜⎜ K + S ⎟⎟ ⎜⎜ K ⎟⎟ ; (A3)
dt m c Y ⎝ S ⎠ ⎝ A + A ⎠
d A βκ μ m ax ρ X E ⎛ S ⎞⎛ A ⎞
= − ( A − A) − ⎜ ⎟⎜ ⎟ ; (A4)
⎜ ⎟ ⎜ ⎟
dt V c Y ⎝ K S + S ⎠ ⎝ K A+ A ⎠
dX ⎛ S ⎞ ⎛ A ⎞
= μ m a x X ⎜⎜ ⎟⎟ ⎜⎜ ⎟⎟ − b X ; (A5)
dt ⎝ K S + S ⎠ ⎝ K A + A ⎠
d X ⎛ ⎞ ⎛ ⎞
= μ m ax X ⎜
⎜ K
S
⎟ ⎜
⎜ K
A
⎟ − b X

. (A6)
dt ⎝ S + S ⎟⎠ ⎝ A + A ⎠

If the external mass transport is ignored, the system of six equations (Equations A1-A6) will be reduced to Equations A7
and A8, which calculate the loss of substrate and electron acceptor.
dS μ m ax X ⎛ S ⎞ ⎛ A ⎞
= − ⎜⎜ ⎟⎟ ⎜⎜ K + A ⎟⎟ − K a b i o S ; (A7)
dt Y ⎝ K s + S ⎠ ⎝ A ⎠
dA μ X E ⎛ ⎞ ⎛ ⎞
= − m ax
S A
⎜⎜ ⎟⎟ ⎜⎜ ⎟⎟ . (A8)
dt Y ⎝ Ks +S ⎠ ⎝ K A + A ⎠
When nutrients such as nitrogen and phosphorous limit the reaction, the substrate utilization needs to be modified
according to the following equation:
dS μ X ⎛ S ⎞⎛ A ⎞⎛ ⎞
= − m ax ⎜⎜
N
rs = ⎟⎟ ⎜⎜ ⎟⎟ ⎜⎜ ⎟⎟ . (A9)
dt Y ⎝ K S + S ⎠ ⎝ K A+ A ⎠ ⎝ K N+N ⎠

Conceptual Model of Reservoir Souring


The multi-dimensional reservoir souring model developed in this work has the features to include the variations in the
concentrations and temperatures within the reservoir. The model includes the biological reactions in both attached and free
cells which are responsible for H2S generation. Furthermore, the H2S retardation due to partitioning and adsorption is also
included.
6 SPE 118951

Figure A1 represents the conceptual model of the biodegradation process modeled in UTCHEM. Substrate concentrations
can change in each gridblock, and SRB can both attach to the rock surfaces (biofilm) and remain in the aqueous phase. Thus,
we accounted for both attached (sessile) and free-floating (planktonic) bacteria reactions. The temperature ranges from that of
injected seawater to reservoir temperature. The pressure also changes between injectors and producers. In Figure 1, S
represents substrate molecules in the bulk liquid that must diffuse across a stagnant liquid layer to become available to
attached biomass (biofilm and sessile bacteria). The subscript f refers to intra-biomass concentration (Farhadinia et al., 2008).

Figure A1 Conceptual model of the souring process

Table 1 Activation range of the different SRB types (after Okabe et al., 1992;
Leu et al., 1999; Sunde et al., 1992)
SRB Types Lower limit of Maximum growth rate Upper limit of activation
activation (°F) Temperature (°F) (°F)
Mesophilic 50 95 109
Thermophilic 100 145 170
Hyper thermophilic 163 203 219

Table 2 Reservoir characteristics


Layer # L1, L2, L3
Porosity 0.3
Perm-x 800-2300
Perm-y Perm-y = Perm-x
Perm-z Perm-z = 0.1 Perm-x

Table 3 Reservoir conditions


Water relative Oil relative
SWI Swr Sor Krw0 Kro0 permeability permeability
exponent exponent
0.32-0.72 0.25 0.15 0.2 0.95 3 2

Table 4 Reservoir conditions (continued)


Initial Initial Reservoir Rock Water Oil Water Oil
Temperature Pressure Depth Compres. Compres. Compres. Vis. Vis.
(ºF) (psia) (ft) (1/psi) (1/psi) (1/psi) (cp) (cp)
160 1770 4150 0 0 0 0.46 40
SPE 118951 7

Table 5 Reservoir data for energy balance equation


Rock Rock Rock heat Water phase Oil phase Water Oil
density thermal capacity heat capacity heat capacity density density
(lb/ft3) conduct. (Btu(lb-F)-1 (Btu(lb-ºF)-1 (Btu(lb-ºF)-1 (psi/ft) (psi/ft)
(Btu(lb-
ºF)-1
165.43 40 0.2117 1 0.5 0.4368 0.3462

Table 6 Reservoir simulation data


Injection well, Production well,
constant rate constant rate ∆x ∆y ∆z
(ft3/day) (ft3/day) (ft) (ft) (ft)
W3, 1967 W1, 679
W6, 2123 W2, 803
W8, 2244 W4, 928
W9, 1740 W5, 850 19×32.8 19×32.8 10, 20, 10
W11, 1942 W7, 2088
--- W10, 843
--- W12, 611
--- W13, 693

Table 7 Initial concentration data


SRB in seawater 0.0001 (mg/l)
SO4 2700.0 (mg/l)
POC (part. Org. C) 0.01 (mg/l)
NO3 0.6 (mg/l)
PO4 0.06 (mg/l)
CH3COOH 1000.0 (mg/l)
in formation water
PO4 0.3 (mg/l)
8 SPE 118951

Figure 1 Schematic illustration of oil field reservoir souring (after Furui and Bryant, 2005)

Figure 2 Simplified view of concentrations and temperature distributions during water flooding (after Furui and Bryant, 2005)
SPE 118951 9

NY / N X 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
1
2
3 w5○ w 2○ w 1○
4
5
6
7 w6∆ w3∆
8
9
10 w7○
11 ○w10 w4○
12
13
14 w11∆ w8∆
15
16
17 w13○ w12○
18 w8∆
19
Figure 3 Location of the injection and production wells

25

Well-1
20
Well-2
Well-4
15 Well-5
H 2 S ( m g /l)

10

0
0 200 400 600 800 1000 1200 1400
DAYS
Figure 4 History of the produced hydrogen sulfide in wells 1, 2, 4, and 5
10 SPE 118951

25

Well-7
20 Well-10
Well-12
Well-13
15
H 2 S ( m g /l)

10

0
0 200 400 600 800 1000 1200 1400
DAYS

Figure 5 History of the produced hydrogen sulfide in wells 7, 10, 12, and 13

Figure 6 Tracer distribution after 2 years of water injection


SPE 118951 11

Figure 7 Temperature distribution after 2 years of water injection

Figure 8 Hydrogen sulfide distribution after 2 years of water injection


12 SPE 118951

Figure 9 Nitrate distribution after 2 years of water injection

Figure 10 Phosphate distribution after 2 years of water injection


SPE 118951 13

Figure 11 SRB distribution after 2 years of water injection

Figure 12 Sulfate distribution after 2 years of water injection


14 SPE 118951

Figure 13 Acetate distribution after 2 years of water injection

You might also like