SPE 114676

Potential Microbial Enhanced Oil Recovery Processes:

A Critical Analysis
Murray R Gray, Anthony Yeung, and Julia M Foght, University of Alberta, and Harvey W Yarranton, SPE,
University of Calgary

Copyright 2008, Society of Petroleum Engineers

This paper was prepared for presentation at the 2008 SPE Annual Technical Conference and Exhibition held in Denver, Colorado, USA, 2124 September 2008.
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
This paper summarizes a critical review of possible microbially enhanced oil recovery (MEOR) methods and mechanisms to
identify the most plausible utilization of microbial technology to enhance oil recovery. This paper is intended to stimulate
discussion about broad applications of MEOR to field-wide improvement of oil displacement and recovery.
The potential benefits of different MEOR mechanisms were examined for a representative North Sea reservoir of sandstone
containing a light crude oil. In each case, the material input requirements for nutrients and inoculum were calculated in
relation to the projected incremental oil production, assuming that the key components (bacteria, nutrients, or bacterial
products) were uniformly distributed in the swept zone of the reservoir.
The capillary number for the reservoir suggested that incremental oil recovery by biosurfactant production in situ would be
modest under ideal conditions. The achievable yields would be lower because the Microbiol. in the reservoir would not be
controllable to achieve sustained surfactant production. Losses of biosurfactants by adsorption to reservoir rocks and in situ
biodegradation would further limit performance. The stimulation of surface active bacteria in the reservoir may affect the
flow of fluids by producing emulsion droplets of oil coated with bacteria. The formation of a biofilm at the oil water interface
changes the rheology of the interface, and may provide a useful mechanism to control mobility and areal sweep in reservoirs.
Any mechanisms that required changes to large volumes of reservoir material, to produce gases, solvents, acids, or to
significantly alter permeability, were not considered to be feasible. The limiting case of plugging fractures with bacteria and
their polymeric byproducts does have considerable potential, but only for reservoirs which offer significant improvements in
production with minimal injected volumes. The biological deposits must retain longer term resistance in situ to ensure that
the anticipated production increase is achieved.
This study provides a methodology for the systematic assessment of MEOR proposals using well-established reservoir
engineering principles.
Introduction
The objective of this study is to determine which mechanisms of microbial enhanced oil recovery (MEOR) have potential for
successful application to a typical sandstone North Sea petroleum reservoir with light crude oil. The scope is to examine in
situ processes by which microbial activity in a petroleum reservoir can be modified, augmented or introduced to benefit oil
recovery. Possible inputs to the reservoir for these processes include addition of specific bacterial strains and addition of
nutrients. The emphasis throughout the study is to provide quantitative estimates of the potential increase in crude oil
recovery by MEOR.
A number of mechanisms have been proposed for MEOR that fall into two broad categories: A. Alteration of oil/water/rock
interfacial properties, and B. Changes in flow behavior. Each biological mechanism was assumed to change the initial
conditions in the reservoir by altering one or more fluid, rock, or interfacial properties. In each case, the incremental oil
recovery arising from the change in properties was determined using simple and well-established methods for reservoir
analysis. Transport of bacteria and nutrients into the reservoir was assumed to be uniform and complete at the start of the
MEOR process. The analysis only considered reservoir-wide displacement mechanisms that could be used to change crude

SPE 114676

oil recovery on a large scale. Single-well huff and puff treatments have been documented frequently, and are available
commercially, but these localized near-well bore treatments were not considered in the analysis.
The microbial nutrient requirements were determined for each case and a material balance performed to determine the oil
recovery per nutrient mass. In order to be considered promising for further assessment and research, a mechanism must
achieve two goals:
1. The mechanism must have the potential for a significant increase in oil recovery.
2. The yield ratio, defined as the incremental yield of oil divided by the input material for MEOR, must be economically
attractive.
If the yield and yield ratio were favorable for a mechanism, accounting for changes in crude oil properties and reservoir
properties due to the MEOR operation, then the mechanism was considered to be a good prospect for application. These
mechanisms were then considered in more detail for feasibility and robustness in the field. If the yield ratio was near unity,
then the prospects for the mechanism were rated as poor, because the value of the oil would be comparable or less than the
input materials. Mechanisms with very low yield ratios, much less than one, were considered to be very poor prospects.
Reservoir Data
Reservoir data were compiled from information on two representative North Sea reservoirs. Net pay, porosity, and
permeability were obtained from an open-hole log. The net pay is 40 m and the porosity is 24%; both were assumed to be
invariant for the purposes of this study. The permeability distribution is given in Table 1. The composite reservoir has a
typical sandstone porosity, intermediate permeability, and moderate permeability variation.
Table 1: Permeability Distribution.
Thickness (m)
Permeability (mD)
2
110
6
30
5
50
2
120
5
0
2
100
2
500
13
200
8
80
4
300
3
100
2
10
2
90
5
20
3
300
10
100
6
50
3
500

The relative permeability data are summarized in Table 2 and indicate an intermediate wettability. Sandstones are generally
more water-wet, although intermediate wettability is not unusual. The residual oil saturation of 15% is quite low. Typical
sandstone residual oil saturations are in the order of 30%. The mobility ratio for this reservoir is 2.6.
Table 2: Relative Permeability Data:
Property
Irreducible water saturation, Swi
Residual oil saturation to water, Sorw
krow @ Swi
krw @ Sorw
Corey oil exponent, m
Corey water exponent, n

Value
0.23
0.15
0.80
0.50
3.5
2.5

SPE 114676

Capillary pressure is shown in Figure 1 and was an average from representative North Sea fields. The pore size distribution
determined from the capillary pressure data is shown in Figure 2. The average pore size is 5.0 microns and the median pore
size is 1.6 microns.

Capillary Pressure, kPa

10000

1000

100

10
0.0

0.2

0.4

0.6

0.8

1.0

Water Saturation

Figure 1: Capillary pressure curve for sandstone reservoir

Cumulative Volume Fraction

1.0

0.8

0.6

0.4

0.2

0.0
0

10

100

Capillary Radius, microns

Figure 2: Pore size distribution for sandstone reservoir

PVT data are provided in Table 3. The example reservoir fluid is a highly undersaturated light oil with low viscosity. The
water analysis is provided in Table 4. We assumed that the water chemistry would shift due to waterflooding, so that at 90%
water cut in a mature waterflood, the water chemistry would be very similar to the inlet seawater. The detailed water
chemistry did not enter into our analysis; the main impact is on sulfate concentration, which would stimulate bacteria as the
seawater permeates the reservoir.

Table 3: PVT data.

Property
Reservoir temperature - initial
Reservoir temperature after inj.
Initial pressure, Pi
Bubble Point, Pb
Boi
Rsi
Bgi
o @ Pb
w
o

SPE 114676

Value
99C (210 F)
66C (150 F)
38.6 MPa (5600 psia)
5.8 MPa (835 psia)
1.2 m/scm
32 m/m (180 scf/stb)
0.00367 m/m (0.654 stb/Mscf)
1. 1 cp
0.3 cp
758 m/m

Table 4: Water Analysis

Ion
mg/l
Sodium
8500
Potassium
220
Calcium
221
Magnesium
40
Barium
50
Strontium
8
Chloride
12600
Sulphate
9
Bicarbonate
665
The following production data were obtained for a representative pool:
Fluid rate per well, q
2000 m/d
Current water cut
90 %
Current recovery factor overall 28 %
Current recovery factor main area 34 %
The example reservoir is in the late stage of a waterflood (90% water cut). The recovery factors are within the normal range
for a waterflood. In this case, MEOR is considered as a tertiary recovery method.
Comments on Example Reservoir
The rock properties of the example reservoir are fairly typical for sandstones. The reservoir has only moderate permeability
contrast. Conformance modification is expected to more effective for reservoirs with higher permeability contrasts. The
reservoir does have a low residual oil saturation to water. Incremental oil production from surfactant generation is expected to
be higher for higher residual oil saturations.
The reservoir is highly undersaturated. The effect of biogenerated gas is expected to be higher for an undersaturated oil than
for a saturated oil. The viscosity of the crude oil is low at reservoir conditions, therefore, there is little or no benefit in further
reduction in viscosity. The benefits of viscosity reduction may be greater for a more viscous oil, but the initial viscosity
would need to be at least on order of magnitude higher. The mobility ratio of 2.6 is reasonably favorable for water flooding.
Poorer displacement and sweep efficiency is expected at higher mobility ratios. Hence, the potential benefits of viscosity
reduction and wettability alterations are higher for reservoirs with poorer mobility ratios.
The reservoir is in the late stages of a waterflood. Total potential for incremental oil recovery from MEOR depends on the
ultimate volume of the swept zone and the residual oil in place after waterflood. Hence, the benefits in the ultimate recovery
due to MEOR, either by improved sweep or changes in residual oil saturation, would be similar at any stage of the
waterflood, except that the timing of the incremental recovery will differ. The incremental recovery per mass of injected
microbes and nutrients is not expected to be sensitive to the time at which MEOR is initiated during the waterflood operation.
The benefits of MEOR as a tertiary technique would be comparable to the incremental benefits of beginning MEOR at the
start of a waterflood.

SPE 114676

A. Alteration of Oil/Water/Rock Interfacial Properties

A.1 Reduction in Interfacial Tension by Microbial Surfactants
The production of biosurfactants by microbes in situ has been proposed by many investigators as an effective mechanism for
enhanced oil recovery. These biosurfactants can reduce the interfacial tension (IFT) between the flowing aqueous phase and
the residual oil saturation, and potentially increase oil recovery. In this section, we evaluate the IFT required for effective
increases in oil recovery, and assess the potential of in situ production of biosurfactants.
A.1.1 Requirements to Mobilize Oil by Reduced IFT:
It is thought that residual oil exists as globules dispersed within the pores of the reservoir (Willhite and Green, 1998). Under
existing reservoir conditions, there is insufficient pressure gradient to move these globules through the surrounding pore
throats. The viscous force promoting flow is opposed by the capillary force acting on the globule. In this situation, a useful
correlating parameter is the capillary number, NCa, given by:

N Ca =

(1)

where u is velocity (m/s), is viscosity (Pa.s), and is interfacial tension (N/m). The capillary number is the ratio of viscous
to capillary forces. The larger the capillary number, the greater the likelihood of mobilizing the residual oil. A reduction in
the residual oil will give higher recovery from the reservoir. This outcome can be achieved by changing the capillary number,
for example, by surfactants. A waterflood can never reduce the oil saturation to below the residual oil to water, but a
surfactant flood can recover more.

Residual Oil Saturation, %

Surfactants increase the capillary number by reducing the interfacial tension. Core floods for miscible floods indicate that the
capillary number must be increased to 10-5 to 10-4 to mobilize residual oil, Figure 3 (Stalkup, 1984). The capillary number for
most reservoirs under waterflood is in the order of 10-7 (Willhite and Green, 1998). Hence, the surfactant must reduce the
interfacial tension by at least two orders of magnitude to mobilize the oil.

40

Berea Sandstone 1

30
Rocky Mtn
Reservoir

20

10

Berea Sandstone 2

0
1.E-08 1.E-07 1.E-06 1.E-05 1.E-04 1.E-03 1.E-02 1.E-01 1.E+00

Capillary Number, u/
Figure 3: Residual oil saturation as a function of capillary number (adapted from Stalkup, 1984).
The interfacial tension between hydrocarbons and water is typically in the order of 30 to 40 mN/m. Therefore, the surfactant
must reduce interfacial tension at least to below 0.4 mN/m to have any effect. For example, core floods with commercial
surfactants with interfacial tensions of 0.03 and 0.002 mN/m were able to mobilize 50 and 70% of the residual oil,
respectively, after one pore volume of injection. The recommended interfacial tension for chemical surfactant floods is 0.001
to 0.01 mN/m (Willhite and Green, 1998).
A.1.2 Interfacial Tension of Microbial Surfactants
Table 5summarizes the typical range of interfacial tensions (IFT) reported for surfactants from bacteria. Typically, the IFT
values are reported at the critical micelle concentration or above, where the aqueous phase is saturated with free surfactant.
The liquid phases used varied considerably between these studies, ranging from n-hexadecane to crude oil. Although a
variety of rhamnolipid biosurfactants have been studied from different bacterial sources, only one paper reported interfacial

SPE 114676

tensions below 0.06 mN/m. All other reports, including the widely studied lipopeptides called surfactins, gave interfacial
tensions above this value at neutral pH.
Table 5: Interfacial tension of biosurfactants
Biosurfactant
Interfacial tension,
mN/m
Rhamnolipid PG201
4.5

Liquid Phase

Reference and notes

Urum and Pekdemir 2004

Rhamnolipid PG201,

0.08 at pH 5.4

Rhamnolipid Dyna 201

Mannosylerythritol lipid
Trehalose glycolipid
Lipopeptide Surfactin
Lipopeptide Surfactin

0.06 at pH 5.7
0.94 at pH 5.7
0.75 at pH 7
4
1.32
4.5
0.006 at pH 6
0.023 at pH 7
0.1
1.5
0.2
2.03

Ekofisk crude
oil
dodecane
benzene
Isooctane
benzene
hexadecane
Crude oil
hexadecane
hexadecane
Decane/5%
NaCl
kerosene
hexadecane
Diesel oil
dodecane

Mixed culture
Mixed culture

18.83
0.0061

Crude oil
hexadecane

Rhamnolipid PG201
Rhamnolipid
Rhamnolipid
Bacillus biosurfactant
Bacillus lipopeptide JF-2

Hung and Shreve 2001

Hung and Shreve 2001
Olsen and Janshekar, 1985
Li et al., 2004
Benincasa, Abalos et al. 2004
Huszcza and Burczyk 2003
Lin, Minton et al. 1994
Kim, Jeon et al. 2002
Choi, Kim et al. 1999
Makkar and Cameotra 1999
Deleu, Razafindralambo et al.
1999
Kowalewski et al., 2006
Kowalewski et al. 2005

1. IFT measured for bacteria and surfactant by laser light scattering method.
Most of these studies considered biosurfactants produced under aerobic fermentation conditions, but rhamnolipids are also
formed under anaerobic denitrifying conditions (Chayabutra et al., 2001) as were lipopeptides (Davis, Lynch et al. 1999).
We assumed, therefore, that similar biosurfactants could be produced under the anaerobic conditions of a petroleum reservoir.
The one study by Kowalewski et al. (2005), which involved a mixed culture from oilfield samples, showed an IFT value ten
times lower than any other published report. With the exception of this study, no other reports of any biosurfactants were
found that indicated IFT values below 0.01 mN/m at neutral pH, which is required for effective mobilization of crude oil by
reduction of interfacial tension. Kowalewski et al. (2005) measured interfacial tension in the presence of bacteria using a
laser-light scattering technique. Bacteria can change the properties of an interface significantly, not by changing the
interfacial tension but by interacting with one another to change the elasticity of the interface (Kang et al., 2008). Such
changes in elasticity could be crucial for a laser technique that analyzes capillary waves at the interface, and this factor should
be considered in calibrating and interpreting the measurements.
The majority of the biosurfactants studied at low pH had IFT values of order 10-1 N/m. One biosurfactant from Bacillus gave
a lower value of 0.006 N/m under slightly acidic conditions of pH 6. (Lin, Minton et al. 1994; Hung and Shreve 2001). This
biosurfactant had a very narrow pH optimum; at pH 5.0 the IFT was 0.1 N/m, and at pH 7 it was 0.07 N/m. The dissolution of
carbon dioxide in the aqueous phase at reservoir pressure in the sandstone case would reduce the pH relative to surface
conditions. In light of the range of literature values, we selected an IFT of 0.06 N/m for analysis of the displacement.
A.1.3 Application of Biosurfactants to Subject Reservoir:
The first step is to calculate the capillary number prior to the introduction of a surfactant. Radial inflow of 2000 m/d was
assumed with a boundary radius of 50 meters (based on well spacing). The rock properties described previously were used.
The water viscosity was taken to be 0.001 Pa.s and the water-oil interfacial tension was assumed to be 0.035 N/m, a typical
value for water-oil systems with no surfactants.
The cross-sectional area for flow is:

A = 2Rh (1 S wi S orw ) = 1870 m

and the fluid velocity is:

(2)

SPE 114676

u=

q
= 1.24 10 5 m / s
A

(3)

The capillary number is then:

N Ca =

= 3.5 10 7

(4)

The required capillary number to begin displacing a significant fraction of the residual oil is 10-4 to 10-5. Hence, the
interfacial tension must be reduced by 2-3 orders of magnitude; that is, to 0.03 to 0.3 mN/m. There is no consistent evidence
in the literature (see Table 3) that the low end of this range of interfacial tensions can be achieved with biosurfactants at
neutral pH. However, a few studies indicated interfacial tensions in the 0.06 to 0.1 mN/m range are attainable. The best case
for biosurfactant appears to be a capillary number of 2 x 10-4, based on an IFT of 0.06 mN/m. As illustrated in Figure 3, much
larger values are required to significantly reduce the residual oil saturation.
Let us assume an optimistic case, where the surfactant is able to recover 50% of the residual oil in the swept zone. The
residual oil saturation to water for this reservoir is 0.15. Hence, the incremental recovery in the swept zone is 7.5%. The
amount of surfactant required is based on how much adsorbs on the rock surface (Willhite and Green, 1998). Typical
adsorption values for sandstones are in the range of 0.1 to 1 mg surfactant per gram of rock. Hence, the mass of surfactant
required for a given incremental oil volume is:

ms
C s ma (1 ) BV
(0.1 x 10 3 )(2860)(1 0.24)
=
=
= 19 kg / m
Qo 0.075 * (1 S wi )BV / Boi
0.075(1 0.23)(0.24) / 1.2

(5)

where ms is the mass of surfactant, Qo is the incremental oil volume, Cs is the mass of surfactant per mass of rock, ma is the
rock matrix density, is porosity, Swi is the initial water saturation, Boi is the oil volume factor, and BV is the bulk volume.
Consider an arbitrary volume of reservoir; for example, consider a reservoir pattern with a swept zone bulk volume of 106 m.
BV = 1000000 m
OOIP = BV(1-Swi)/Boi = (1000000 m)(1-0.23)(0.24)/1.2 = 154,000 m
Qo = 0.075*OOIP = 0.075(154,000 m) = 11550 m
Rock Volume = BV(1-) = 1000000 m)(1-0.24) = 760,000 m
Rock Mass (mma) = (2860 kg/m)(760,000 m) = 2036 106 kg
Surfactant Mass = Csmma = (0.1 kg/1000kg)(2036 106 kg) = 204,000 kg
Even if the residual oil saturation is 30% and the incremental oil recovery is then 15%, only 23000 m of incremental oil is
recovered per 204,000 kg of surfactant required.
A..1.4 Material Balance on Surfactant:
For the material balance, we consider two cases, first the conversion of n-alkane components of the crude oil in place to
produce surfactants, and second the conversion of an added carbon sources.
Case 1. Conversion of Alkanes from Crude Oil:
Yield of surfactant on substrate (based on data for rhamnolipids):
Aerobic growth on n-alkanes = 0.15 kg surfactant/kg of substrate (Lang and Wullbrandt 1999)
Anaerobic growth on fatty acids = 0.02 kg surfactant/kg of substrate (Chayabutra, Wu et al. 2001)
From the above reservoir case,
Incremental oil production:
11550 m
Surfactant required:
204,000 kg
Alkanes required to produce surfactant:
Aerobic = 204,000/0.15 = 1.4 x 106 kg
Anaerobic = 204,000/0.02 = 1.1 x 107 kg

SPE 114676

From this material balance, under aerobic conditions the incremental recovery of crude oil is 8.5 x 10-3 m3 crude oil/kg
alkanes converted to biosurfactant, or 5.3 x 10-2 bbl/kg. The corresponding incremental recovery under anaerobic conditions
in the reservoir would be 1.1 x 10-3 m3 crude oil/kg alkanes converted to biosurfactant, or 6.9 x 10-3 bbl/kg. Since conversion
of the reservoir to aerobic conditions is completely unrealistic, the latter yields are a reasonable estimate. Under anaerobic
conditions, an incremental yield of 6.9 x 10-3 bbl crude oil/kg of alkanes converted is quite low; the inverse is 146 kg of
alkanes converted to yield 1 barrel of crude oil.
Two factors make this scenario very unattractive from a material balance viewpoint:
1.

Significant amounts of nutrients are required to stimulate the conversion of crude oil components to surfactants,
particularly nitrates to reduce the possibilility of biological souring.

2.

The high mass of alkane required would significantly change the properties of the residual oil, making the recovery
more difficult. The conversion of the alkanes in the swept zone would dramatically increase the viscosity of the
remaining oil, by a factor of 2 or more (Peries et al., 1988) rendering the remaining oil immobile.

Consequently, this case is unfavorable from a simple material balance perspective.

Case 2. Added Carbon Nutrients
The yield of surfactant on glucose = 0.115 kg/kg (McInerney et al., 2005). Given an estimate of 11,550 m3 of crude oil from
204,000 kg of surfactant, this yield gives a requirement of 153 kg glucose/m3 of crude oil or 24 kg glucose/bbl of incremental
crude oil production. In this case, a microbe and other nutrients such as nitrate also need to be added to the reservoir at a
significant concentration, which further affects the materials requirements. For example, Youssef et al. (2007) reported the
use of 8.35 l of mixed Bacillus inoculum per kg of glucose, which increases the material input requirements from 24 kg
glucose/bbl of incremental crude oil production to 204 kg (glucose + inoculum)/bbl of incremental crude oil production.
A.1.5 Feasibility of Mechanism:
This analysis treated the behavior of biosurfactants as an extension of the well-known properties of chemical surfactants.
Consequently, a biosurfactant flood will have all of the characteristics of a chemical surfactant flood plus the added
biological interactions within the reservoir. The Achilles heel of both chemical and biological surfactants for reservoir
applications is adsorption to the mineral surfaces. The tendency of surfactants to adsorb to the reservoir rock results in large
mass requirements for mobilization of the residual crude oil, and we expect the same problem to affect biosurfactants. This
characteristic has received almost no attention in the extensive literature on biosurfactants. For example, McInerney et al.
(2005) suggested that their core flood data reflect losses by adsorption, but these losses were never quantified. Daoshan et al.
(2004) concluded that rhamnolipid biosurfactant could act as a sacrificial agent to reduce adsorptive losses of synthetic
surfactant to sandstone cores by 2530%. This observation implies that the rhamnolipid sorbed more effectively to the
sandstone than the synthetic surfactant, however the losses of biosurfactant in flooding of sandstone formations is a major
issue that requires much more careful attention before scale-up of these methods is attempted.
Treating the biosurfactant mechanism in the same way as chemical surfactants gives low estimates for incremental crude oil
production because the capillary number is not reduced enough to give significant mobilization of the crude oil. Some past
studies that predicted high recoveries of crude oil from biosurfactants assumed simultaneous co-production of surfactants,
alcohols, and a viscosifying polymer in the flowing aqueous phase in order to achieve low capillary numbers to mobilize all
of the residual oil (Sarkar at el, 1991; McInerny et al., 2005). This combination of bioproducts from a microbial culture has
never been demonstrated, therefore, its feasibility is low. Even if it were demonstrated, the input material requirements would
be much higher than for biosurfactants alone, making such a combination infeasible.
The simulation study of Sarkar et al. (1991) assumed a combination of low IFT due to biosurfactant and alcohols, and an
increase in the viscosity of the aqueous phase by a hypothetical biopolymer, to give a capillary number of ca. 1.5 x 10-4. They
estimated 11% increase in oil recovery (11% OOIP). Our optimistic estimates were somewhat lower at a similar capillary
number of 2 x 10-4, due to averaging between the swept and unswept zones. The biggest differences were in yield of
surfactants and estimates of losses due to adsorption. Our analysis used a very optimistic estimate of incremental recovery,
given the capillary number, but also indicted significant materials input requirements to achieve this recovery. As reviewed
by Bryant and Lockhart (2000), in most core studies of biosurfactants, the introduction of microbes achieved only 10-20%
recovery of the residual oil. Recoveries as high as 30-60% have been reported but are suspect because the pressure increased
during the experiment. Many core floods required tens of pore volumes of injection to achieve the reported recoveries.
Although the production of surfactants from crude oil components would seem attractive, the material balance analysis
suggests that too much of the crude oil would be required to achieve this goal, giving a much heavier residual oil that would
be more difficult to mobilize. Consequently, the approach of adding specific surfactant-producing cultures and nutrients
appears to be preferred. In addition to the physical interactions of the biosurfactant with the mineral surfaces, this approach

SPE 114676

faces significant additional biological pitfalls for application to a reservoir scale, as opposed to a single-well treatment, as
follows:

achieving the production of desired biosurfactants in situ in the presence of other organisms to compete for the
added nutrients. The production must be sustained over a period of time in order to pass the surfactant through the
reservoir, saturating the surfaces with adsorbed surfactant along the way. Furthermore, Bryant and Lockhart (2000)
pointed out the difficulties in maintain nutrient levels throughout a reservoir for long periods of time.

biosurfactants may be degraded by other bacteria in mixed cultures under aerobic, fermentative, nitrate-reducing and
sulfate-reducing conditions (Mohan et al. 2006 a, b); therefore, the survival of the produced surfactant is a
significant issue. The soluble surfactant is a much more accessible source of nutrient than the crude oil.

Sustained injection of nutrients would be required in order to flood a reservoir with biosurfactant, and this situation would
encourage the growth of competing organisms in the reservoir, which would consume nutrients without producing surfactant,
or consume the biosurfactant product. Even if such populations were not active initially, they would inevitably develop with
time in a non-sterile environment. The final characteristic of any surfactant treatment is the tendency to form emulsions. The
combination of bacteria, surfactants and crude oil is likely to give emulsion formation in situ, which could lead to blockage of
pores in the formation.
A.2 Alteration of Wettability of Formation by Microbes
The wettability of the minerals in a reservoir rock can be altered by biological activity, for example, by the production of
biopolymers or biosurfactants. The three-phase contact angle between the oil, the water and the mineral surface should be
important for determining the local pressure gradient and interfacial tension required to mobilize the oil in a waterflood. The
presence of bacteria in a reservoir can lead to changes in the exposed mineral surfaces by several different mechanisms:

Direct attachment of bacteria to the surface. In this case, the contact angle will be due to a heterogeneous mixture of
mineral and bacterial surface properties. The hysteresis of the contact angle will also increase because the surface will be less
uniform, unless bacteria colonize the entire mineral surface. This mechanism is limited to pores that are large enough to be
accessible to bacteria, in the range of 1 m or more.

Adsorption of bacterial metabolites to mineral surfaces. Bacterial products such as surfactants can adsorb to mineral
surfaces, and have the potential to alter the wettability by making the surface more or less hydrophobic. This mechanism can
operate at any pore diameter, and may provide an alternate mode for biosurfactants to alter the mobility of the oil, since the
known biosurfactants do not appear to reduce the interfacial tension enough to mobilize the oil directly.

Coating of the mineral with exopolysaccharides or other biopolymers. This case is similar to the direct attachment of
bacteria to the mineral surface. These polymers are associated with cells; therefore, this mechanism is closely linked to cell
attachment. These polymers can be left behind on mineral surfaces even if the cells are released or die (Wanger et al. 2006).
The mention of surface wettability raises immediately the imagery of a clean fluid-fluid interface intersecting a molecularly
smooth solid substrate at a specific angle; the angle of contact is, according to the theory of surface thermodynamics, a welldefined physical property (Adamson and Gast, 1997). A unique angle of contact will in turn lead to capillary effects such as
the rise of a liquid up a vertical tube, or the spontaneous imbibition of oil in porous structures. Unfortunately, this
idealization can quickly become unrealistic as imperfections set in. These imperfections include: (1) surface roughness and
heterogeneities on the solid substrate, which can lead to severe hysteresis effects in contact angle measurement, and (2) the
adsorption of colloidal structures at the fluid-fluid interface, which may create an interfacial skin with complicated surface
rheological properties (in addition to the simple IFT). With the incorporation of microbial species into the system, the abovementioned imperfections are almost certain to occur. Adsorption of bacteria to an oil-water interface can give an elastic skin
(Kang et al., 2008), which can have profound ramifications on the traditional notion of wettability. For example, contact
angle can be measured by placing a liquid drop on a solid surface and determining the angle of contact. If the surface of the
liquid drop possesses an elastic skin, the final equilibrium shape will be significantly altered as additional energy is required
to stretch the skin; this is exactly analogous to the resting of a water balloon on a flat surface. The resulting contact angle
will bear no relevance to wettability. With a surface skin present at the oil-water interface, the notion of imbibition
becomes invalid.
In sandstone formations, the wettability of the mineral surfaces plays a secondary role in oil displacement; therefore,
wettability reversal by microbial activity is not a promising mechanism for enhancing oil recovery. In contrast, wettability
changes and imbibition are much more significant for oil recovery from carbonate reservoirs. A change from oil-wet to
water-wet conditions in a carbonate reservoir has significant benefit for recovery. Achieving such a reversal by adsorbing
bacterial metabolic products, such as biosurfactants, to the mineral surfaces could be an attractive mechanism. Despite this
opportunity to use biological mechanisms to change the wettability of carbonate formations, and the potential compatibility
of the mineralogy with MEOR operations (Tanner et al., 1991), there have been almost no reported studies on the wettability
of carbonate rocks in the presence of bacterial cells. Such studies would need to consider carefully the role of bacteria at the
oil water interface, as noted above.

10

SPE 114676

A.3 Interfacial Rheology and Its Effects on Oil Displacement

The role of interfacial tension in oil displacement is well understood, but bacteria can alter interfacial properties in a very
different way by adhering to the oil-water interface. The impact of bacteria is not to change the interfacial tension, but to
change the viscosity and elasticity of the interface (Kang et al., 2008). Whereas the IFT is related to the force required to
change interface area at equilibrium, the elasticity will determine the response of the interface to dynamic changes. Some
previous studies suggested that interfacial viscosities had the effect of increasing the resistance to oil displacement, regardless
of the wetting conditions (Slattery, 1973). Consequently, the creation of bacterial layers at the oil-water interface will tend to
retard the mobility of oil in the formation, which in turn can reduce the rate of oil recovery. The development of a large
population of bacteria in a reservoir, for example, to produce biosurfactants, could have the unintended effect of retarding the
mobility of oil and hence nullifying the benefits of the biosurfactants. If the bacteria attach to the oil-water interface, then the
above results suggest that the oil mobility can be significantly affected in ways that cannot be predicted from the fluid
properties alone. Bacteria at the oil-water interface of oil droplets may give effective plugging of small pores in the
formation, but this mechanism would not plug vugs or fractures.
The Titan Process (www.titanoilrecovery.com) is a commercial protocol that proposes using the attachment of bacteria to the
oil interface to give a combination of biosurfactant production and bacterial-induced emulsification, to enhance oil recovery.
Bacteria are also claimed to migrate to the poorly swept zones that contain the residual oil. Leaving aside this dubious claim
of bacterial transport within the reservoir, and the role of changes in interfacial properties already discussed in Section A.1,
the essence of the claim is that the combination of interfacial rheology, wettability, and plugging of high-permeability zones
by emulsified oil. Given the recent work on bacteria at interfaces (Kang et al., 2008), some of these ideas may have merit, but
the integration with reservoir performance and Microbiol. requires further study. In particular, the benefit of the bacteria in
mobilizing oil in pores of different sizes needs to be investigated. The above discussion of interfacial rheology suggests that
oil droplets coated with bacteria will tend to plug pores in the formation. Droplets with attached bacteria would need to be at
least 10 microns in diameter to allow the formation of a surface skin, therefore, this mechanism would be active in only the
largest 10% or less of the pore volume in the reservoir pore volume in the example North Sea reservoir, and larger fractures
would be plugged.
Given the lack of available data on the role of interfacial rheology in oil displacement, material balance calculations could not
be performed on this mechanism. The information to date suggests a negative impact on oil mobility, which is a point that
deserves careful consideration in light of MEOR schemes that seek to boost the concentrations of microbes in the reservoir.
Adhesion of hydrophobic bacteria to the oil interfaces could considerably reduce mobility.
B. Changes in Flow Behavior
B.1 Alteration of Macroscopic Conformance by Plugging of High-Permeability Zones
The mechanism of MEOR that has consistently received the most attention has been plugging of high permeability zones by
biological material. The literature clearly shows that bacteria can penetrate very long distances along fractures and high
permeability zones (Beeder et al., 1996; Powell et al, 2003). If biological products can be induced along such highpermeability channels, then the sweep efficiency of the waterflood can be improved. This mechanism has a large literature,
which will not be reviewed here, and the requirements for successful application were clearly defined in the 1980s
(Jenneman et al. 1984; Lappin-Scott et al. 1988). A review of this extensive literature is beyond our scope, therefore, we will
focus on a recent well documented case study which clearly defines the requirements for success (Fujiwara et al. 2004).
B.1.1 Biopolymer Plugging of High Permeability Zones:
Fujiwara et al. (2004) set out a series of requirements for success, both biological and physical that echoed previous studies
of plugging (Jenneman et al. 1984; Lappin-Scott et al. 1988):

Effective treatment requires the formation of a long-lasting flow barrier within the formation. The biomass material
that is formed must be resistant to degradation by the indigenous microbes; otherwise the treatment would be
required at frequent intervals. For this reason, simple injection of a nutrient, such as molasses, may give a temporary
change but not a permanent blockage.

Addition of organisms to the reservoir is feasible, since the added microbes will tend to permeate the very high-flow
zones that are targeted. The added organisms must be able to survive at the reservoir conditions of temperature and
salinity, and in the presence of the crude oil.

The added organisms must be able to perform their task without removing the pre-existing organisms, and over a
range of nutrient concentration due to attenuation of added nutrients in the reservoir.

The reservoir must have high permeability zones or fractures between injectors and producers that result in poor
sweep efficiency.

SPE 114676

11

In their study of the Fuyu oil field in China, Fujiwara et al. (2004) found a strain of bacteria in a sample of reservoir rock
from the target reservoir, an Enterobacter species they designated CJF-002, that produced an insoluble cellulose polymer that
adhered to the reservoir rock. When these cells were present in the normal reservoir population, they were not effective.
When they were present at high local concentration in the presence of molasses, as a nutrient, they were able to out-compete
the other organisms present for the nutrients and produce the polymer. The other native organisms were not easily able to
degrade the polymer material. The simple addition of molasses was ineffective, but adding a sufficient concentration of
molasses with cells of CJF-002 was effective in forming polymer to plug a high permeability zone in a laboratory test. In this
case, the investigators significantly increased their chances of success by isolating an organism from the target reservoir. This
approach of bioaugmentation by indigenous bacteria ensured that the selected strain was able to survive at actual reservoir
conditions.
B.1.2 Reservoir Model for Improved Sweep Efficiency:
Figure 4 illustrates vertical sweep in a layered reservoir without cross-flow between layers. In principle, if the high
permeability layers are blocked, better sweep of the lower permeability layers is possible. In this situation, vertical sweep
efficiency and permeability blocking can be modeled using Stiles method (Craig, 1971; Stiles, 1949). The approach is
optimistic because neither displacement efficiency nor cross-flow is accounted for. None-the-less, the method provides an
order of magnitude estimate of the potential benefit of blocking high permeability streaks to improve sweep efficiency.

injector
k = 50 mD
k = 200 mD
k = 100 mD
k = 50 mD

Figure 4: Vertical sweep of water injection into a layered reservoir with no cross-flow.
The reservoir is divided into layers of different permeability and a thickness is assigned to each layer. It is assumed that water
sweeps through the highest permeability layer first. As each layer is swept, the oil recovery is determined and a new water-oil
ratio is calculated with the following equations:

k avg

RF = (1 S wi S orw ) H i +
(1 Ci )
ki

WOR =

Ci
M
1 Ci

(6)

(7)

where
i

Hi =

hj
j =1
n

h
j =1

Ci =

(kh)
j =1
n

(kh)
j =1

(8)
j

and h is pay thickness, k is permeability, and M is the mobility ratio.

To account for the effect of blocking a high permeability layer, it was assumed that all of the mobile oil in that layer had been
recovered and that the flow through that layer became zero. Then all of the injection water was redistributed into the open
layers. This methodology gives an unrealistic immediate boost to the current oil recovery but provides a reasonably accurate
estimate of the potential recovery at higher water cuts.
A 90 foot productive interval from the subject reservoir was divided into permeability layers based on the permeability
profile from the interpreted well log. The distribution is given in Table 6 along with the Stiles analysis results. The predicted
recovery factor at a 90% water cut was interpolated to be 32% in surprisingly good agreement with the reported recovery
factors. The predicted recovery factors at 95 and 98% water cuts are 43.8 and 46.6%, respectively.

12

SPE 114676

The Stiles calculations, assuming that 5 feet and 12 feet of the highest permeability layers are shut-off, are given in Tables 7
and 8, respectively. If 5 feet are blocked (5/85 = 5.9% of pore volume), the predicted recovery factors at 95 and 98% water
cuts are 48.5 and 48.7%. The average permeability of the flowing reservoir is reduced from 128 to 107 mD. If 12 feet are
blocked (12/85 = 14.1% of pore volume), the predicted recovery factors at 95 and 98% water cuts are 49.6 and 51.8%,
respectively. In this case, the permeability is further reduced to 90 mD. Hence, the incremental recovery from blocking high
permeability streaks is in the order of 5%. From 6 to 14% of the pore volume must be blocked and a 15 to 30% drop in
injectivity is also expected.
The predicted incremental recovery may be optimistic because displacement efficiency and cross-flow are not accounted for.
On the other hand, blocking high permeability streaks may also improve areal sweep efficiency leading to higher incremental
oil recovery. If the reservoir is fractured or contains high permeability streaks, a similar incremental recovery could be
obtained with less pore volume blocked but with a greater decrease in injectivity.
Table 6. Unadjusted Stiles analysis (Swi = 0.23, Sorw = 0.15).
h
(ft)
3
2
4
3
13
2
2
3
10
2
2
8
7
6
5
6
5
2

perm
(mD)
500
400
300
250
200
120
110
105
100
95
90
80
60
50
40
30
20
10

kh
(mD.ft)
1500
800
1200
750
2600
240
220
315
1000
190
180
640
420
300
200
180
100
20

kh/sumkh

h/sumh

RF

WOR

wcut
(%)
11.1
25.9
52.3
66.7
93.4
94.5
95.4
96.4
98.6
98.9
99.1
99.7
99.9
100.0
100.0
100.0
100.0

0.138
0.074
0.111
0.069
0.240
0.022
0.020
0.029
0.092
0.018
0.017
0.059
0.039
0.028
0.018
0.017
0.009
0.002

0.138
0.212
0.322
0.392
0.631
0.653
0.673
0.702
0.795
0.812
0.829
0.888
0.926
0.954
0.972
0.989
0.998
1.000

0.035
0.024
0.047
0.035
0.153
0.024
0.024
0.035
0.118
0.024
0.024
0.094
0.082
0.071
0.059
0.071
0.059
0.024

0.035
0.059
0.106
0.141
0.294
0.318
0.341
0.376
0.494
0.518
0.541
0.635
0.718
0.788
0.847
0.918
0.976
1.000

0.158
0.192
0.244
0.280
0.328
0.426
0.447
0.458
0.469
0.478
0.486
0.505
0.542
0.562
0.580
0.598
0.613
0.620

0.35
0.59
1.05
1.42
3.76
4.14
4.54
5.19
8.51
9.51
10.64
17.37
27.65
45.56
77.40
196.81
1191.85

WOR

wcut
(%)

0.36
0.65
2.50
2.80
3.11
3.63
6.24
7.03
7.92
13.23
21.33
35.44
60.54
154.64
938.85

11.4
29.7
86.2
88.7
90.6
92.9
97.5
98.0
98.4
99.4
99.8
99.9
100.0
100.0
100.0

Table 7. Stiles analysis with 5 feet of highest permeability blocked.

h
(ft)
3
2
4
3
13
2
2
3
10
2
2
8
7
6
5
6
5
2

perm
(mD)
500
400
300
250
200
120
110
105
100
95
90
80
60
50
40
30
20
10

kh
(mD.ft)

1200
750
2600
240
220
315
1000
190
180
640
420
300
200
180
100
20

kh/sumkh

h/sumh

RF

0.000
0.000
0.140
0.088
0.304
0.028
0.026
0.037
0.117
0.022
0.021
0.075
0.049
0.035
0.023
0.021
0.012
0.002

0.000
0.000
0.140
0.228
0.532
0.560
0.586
0.622
0.739
0.762
0.783
0.857
0.906
0.942
0.965
0.986
0.998
1.000

0.035
0.024
0.047
0.035
0.153
0.024
0.024
0.035
0.118
0.024
0.024
0.094
0.082
0.071
0.059
0.071
0.059
0.024

0.035
0.059
0.106
0.141
0.294
0.318
0.341
0.376
0.494
0.518
0.541
0.635
0.718
0.788
0.847
0.918
0.976
1.000

0.154
0.202
0.256
0.292
0.338
0.440
0.461
0.472
0.479
0.487
0.496
0.512
0.548
0.566
0.583
0.600
0.613
0.620

SPE 114676

13

Table8. Stiles analysis with 12 feet of highest permeability blocked.

h
(ft)
3
2
4
3
13
2
2
3
10
2
2
8
7
6
5
6
5
2

perm
(mD)
500
400
300
250
200
120
110
105
100
95
90
80
60
50
40
30
20
10

kh
(mD.ft)

2600
240
220
315
1000
190
180
640
420
300
200
180
100
20

kh/sumkh

h/sumh

RF

0.000
0.000
0.000
0.000
0.394
0.036
0.033
0.048
0.151
0.029
0.027
0.097
0.064
0.045
0.030
0.027
0.015
0.003

0.000
0.000
0.000
0.000
0.394
0.430
0.463
0.511
0.662
0.691
0.718
0.815
0.879
0.924
0.955
0.982
0.997
1.000

0.035
0.024
0.047
0.035
0.153
0.024
0.024
0.035
0.118
0.024
0.024
0.094
0.082
0.071
0.059
0.071
0.059
0.024

0.035
0.059
0.106
0.141
0.294
0.318
0.341
0.376
0.494
0.518
0.541
0.635
0.718
0.788
0.847
0.918
0.976
1.000

0.134
0.177
0.253
0.312
0.352
0.463
0.485
0.495
0.496
0.503
0.511
0.523
0.558
0.574
0.589
0.603
0.614
0.620

WOR

wcut
(%)

1.43
1.66
1.90
2.30
4.32
4.92
5.61
9.71
15.96
26.86
46.24
118.89
724.35

67.1
73.4
78.3
84.1
94.9
96.0
96.9
99.0
99.6
99.9
100.0
100.0
100.0

B.1.3 Material Balance:

From the Stiles analysis we assume the optimistic case that blocking 6% of the pore volume of the reservoir will give 5%
incremental recovery from the reservoir. Given a conversion factor of 1.2 m at reservoir conditions/standard cubic meter,
this result indicates that 0.7 m3 of crude oil will be recovered for every m3 of biofilm (cells plus exopolysaccharide) formed
in the pore space. Data on the effective density of bacterial biofilms are scarce. Staudt et al. (2004) found a minimum
concentration of 200 mg dry weight per mL of biofilm, for biofilms grown on a range of carbon sources. A reasonable yield
of exopolysaccharide is 0.5 g COD/g COD of substrate (Alpkvist et al. 2006), where COD is chemical oxygen demand, a
measure of concentration of convertible organic material. From these data, we would expect yields of 0.0025-0.005 L/g of
nutrient. From the volumetric yield of 0.7 m3/m3, we estimate a yield of 1.3-2.6 L crude oil/kg of nutrient added. In contrast,
the field test results of Fujiwara et al. (2004) indicated a yield ratio of 18 L/kg of molasses.
B.1.4 Feasibility of the Mechanism:
The field test of Fujiwara et al. (2004) suggested that the conformance of a waterflood can be improved by selective bacterial
plugging, with proper protocols for selection and injection of an organism and nutrients. The Stiles analysis gave a much less
impressive result, possibly because a less heterogeneous reservoir was considered. As well, the production data presented by
Fujiwara et al. indicted some incremental production from operational changes may have boosted the reported incremental
recovery. This analysis focused on vertical conformance, and did not consider areal sweep. Overall, we suggest that the Stiles
analysis prediction of 1.3-2.6 L/kg is conservative for reservoirs with high-permeability sandstone layers, while 18 L/kg is
likely at the maximum end of the possible range where both vertical and areal conformance are degraded by the presence of
fractures.
The worse the performance of a waterflood, due to channeling of water through high-permeability zones or fractures, the
more likely bacterial plugging will be to give significant improvement. As the volume of the high-permeability zones
increases relative to the total pore volume, the less attractive the yields. The analysis of Section B.1.2 suggests that reservoirs
with much less than 6% of the pore volume in high-permeability layers would be the best prospects for bacterial plugging.
Obviously fractured reservoirs are the best prospect for conformance control by bacteria. Selective plugging of fractures
gives the best possible ratio of beneficial oil recovery in proportion to the volume of reservoir that is affected by the
treatment. The smaller the volume, i.e. a fracture, the more favorable the result of material balances of incremental oil
recovered to volume of material injected. Given the benefits of plugging fractures, a key question is the maximum size of
fracture than could be effectively blocked by bacterial action. The physical strength and stability of the bacterial plug are an
important part of this analysis, as suggested by Sarkar et al. (1995). Growth of native cells in a fracture would give very
different strength and stability in comparison to the more focused production of a plug of biopolymer as practiced by
Fujiwara et al. (2004). Even with formation of a stable biopolymer, such as cellulose, some fractures would be too large for
bacteria and their products to provide effective blockage. Other treatments would be required to block fractures in this case.
The study by Fujiwara et al. (2004) suggested some key microbiological issues for further examination:
1.

Addition of rich nutrients, such as molasses, must be balanced by the addition of nitrate in order to control souring
by sulfate reducing bacteria

2.

Bioaugmentation was apparently successful; in this case the cultivation of an organism from the target reservoir on
the surface, then injecting large volumes of this organism into the formation to control the local population in the

14

SPE 114676

presence of the added nutrients. Their analysis did not seem to consider the cost of preparing and handling batches
of the organism to inject into the formation. Depending on the volume of water in an actual field, the cost and
logistics of this inoculum preparation could be significant
3.

With time biopolymers are degraded, but the study by Fujiwara et al. (2004) did not consider this issue beyond a
time scale of a few months. The injection of stains that produce biopolymer will inevitably tend to encourage strains
of bacteria that could degrade this material.

4.

The focus of this analysis was the injection of rich carbohydrates, like molasses, which can be readily converted into
biopolymers like cellulose in a matter of hours. Any approach that would use in-situ hydrocarbon would be much
more challenging, both in terms of the Microbiol. to stimulate a particular population, and the rate of the metabolic
processes.

B.2 Increase in Permeability by Organic Acid Production

One benefit ascribed to MEOR is an improvement in reservoir permeability due to the formation of acids, which dissolve
carbonate minerals. This mechanism could be potentially beneficial in a sandstone formation if it is cemented by carbonates;
therefore, we consider the requirements for opening permeability by bacterial action. As discussed in Section B.1, growth of
microbes in a reservoir can give plugging due to the accumulation of cells and exopolysaccharides (bacterial polymers) in the
reservoir pore space. An additional effect due to cells attached to oil droplets was considered in Section A.3, which would
tend to further reduce permeability. The key question, therefore, is whether the organic acids produced by bacteria can open
more pore volume than will be filled by the cells themselves.
B.2.1 Balance between Biomass and Acid Production
Production of organic acids from sugars is a normal phase of anaerobic fermentation. Bacteria such as Clostridum species can
produce 0.0034 mol of acid/kg of molasses (Lenz and Moreira, 1980). Similar values have been observed for mixed cultures
under anaerobic conditions, so the specific strain of organism is probably not important (Rodriguez et al., 2006). Each mol of
calcium carbonate requires 2 mol of acid to dissolve, so with a density of 2700 kg/m3, the amount of volume cleared is 6.39
x10-5 m3 CaCO3 removed/kg molasses. The same amount of sugars will produce growth of bacteria. Even at high pressure of
17 MPa and under anaerobic conditions, the yield of bacteria was 0.09 kg of dry cells/kg of sugar (Bothun et al., 2004).
Assuming that the cells are 80% water, the yield of cells would be approximately 4.5 x 10-4 m3 of cell biomass/kg sugar. On
this basis, the volume of biomass produced would be at least 7 times the volume of calcium carbonate that could be
dissolved. Any production of exopolysaccharides would make this ratio even larger.
For the mechanism of acid production, therefore, the density and composition of cells in comparison to the density of
carbonates dictates that more volume will be occupied than will be opened by dissolution. These two volumes need not be in
the same location in the reservoir, for example, an active population in the vicinity of an injection well could produce acids
that would then pass into the swept zone of the reservoir. Given the volume ratio of the biomass produced to the carbonate
dissolved, it is difficult to see the benefit of completely plugging one portion of the reservoir in order to gain incremental
benefits elsewhere. One potential benefit from this mechanism would be if the acid widens pore throats (possible) while the
biomass deposits in pore cavities (unlikely). Another possible benefit would be if the cells block the large pores that are
largely swept free of oil by the waterflood, and open the smaller pores that are too small to admit bacteria. Due to the small
amount of carbonate removed relative to the volume of cells, this latter mechanism would still give a drastic decrease in
overall permeability.
Both of the biological products would be transient in the reservoir due to further biological conversion. The initial mass of
cells from nutrient injection would die off, and their cell material would be reprocessed by other organisms. If the cells
produce polymers, this material could last longer in the formation, as discussed in section D.1. The organic acids would also
be consumed by a variety of cells in the reservoir, in competition with the reaction with carbonates. For example, acetic acid
would be consumed by sulfate-reducing bacteria (Section B).
We conclude that the basic physical properties of carbonates versus biomass requires that the mechanism of permeability
improvement by organic acids be rejected completely. The net impact is to fill the reservoir volume with biomass and thereby
reduce permeability.
B.3 Microbial Gas Production In Situ
Several studies have suggested that conversion of light crude oil to heavy crude oil and methane and carbon dioxide gas is a
natural biological process in oil reservoirs (Head et al., 2003; Aitken et al., 2004; Jones, et al., 2008). It has been proposed
that this gas produced by microbes can increase oil recovery. Three mechanisms have been suggested: 1) reduction of the
residual oil saturation, 2) reduction of oil viscosity, 3) repressurization.
Reduction of Residual Oil Saturation: Assume that, after waterflooding, the remaining oil saturation is equal to the
residual oil saturation to water. The residual oil is dispersed in the pore network as blobs that cannot connect with each

SPE 114676

15

other and flow. Miscible gas injection can potentially recover all of this residual oil (Green and Wilhite, 1998).
Immiscible gas injection has been shown to cause the oil to spread into a film (Kantzas et al., 1988a,b; Oren et al., 1994)
which can flow and be recovered. Immiscible gas injection, however, suffers from gravity override and poor sweep
efficiency and the potential for recovering the residual oil is limited. Some improvement in sweep efficiency after
waterflooding is possible, particularly if gravity under-ride has occurred. A significant reduction in residual oil saturation
is possible with a vertical gas flood (Kantzas et al., 1988b). For effective sweep, at least 0.1 to 0.3 pore volumes of gas are
required for immiscible water-alternating gas (WAG)-type processes and at least one pore volume of gas for a miscible
displacement (Green and Wilhite, 1998). Also note that effective permeability is reduced in three-phase flow.
Reduction of Oil Viscosity: In undersaturated reservoirs, some microbial gas is likely to dissolve in the oil and reduce its
viscosity. Methanogens produce about 60% methane, 40% carbon dioxide. Methane will partition between the oil and gas
phases but carbon dioxide will partition to the water phase as well. Sulfate reducers produce mainly carbon dioxide,
which can significantly reduce the viscosity, and hydrogen sulfide which would be less effective in changing the oil
properties. Lower oil viscosity improves the mobility ratio. While the residual oil saturation to water will not decrease, the
oil recovery at an economically limiting water-oil ratio could increase. Better sweep efficiency is also possible. Both
factors are only likely to be significant in reservoirs with an unfavorable mobility ratio. Oil swelling could also liberate
some residual oil.
Repressurization: Since gas has a lower density than oil at reservoir conditions, the microbial conversion of oil to gas has
the potential to increase the reservoir pressure. An increase in reservoir pressure would normally increase production
rates. However, if a free gas phase forms, the rate improvement would be offset by the loss of relative oil permeability in
three phase flow. Also the mobility of gas is higher than that of oil or water and therefore it is likely that the generated gas
would be produced before any significant pressure increase occurred.
Note that if the microbes consume some oil to produce microbial gas, the viscosity of the remaining oil will be higher than
the original oil. As the light oil fractions are consumed, the gas will be less soluble in the remaining higher boiling material.
Eventually an immobile heavy oil residue will remain and only gas will flow. The use of microbes to convert immobile oil to
flowing gas is outside the scope of this study. The following analysis is concerned only with potential increases in ultimate
oil recovery. The time required to achieve this recovery and the challenges in delivering and maintaining the microbes in the
reservoir are outside the scope of the study and are not considered.
B.3.1 Application to the Subject Reservoir
The reservoir is highly undersaturated (Pb = 835 psia, Pi = 5600 psia). Therefore, if the microbial gas does not flow as soon as
it is generated, the initial gas will likely dissolve in the oil rather than form a distinct gas phase. The potential benefits are
then an increase in sweep efficiency and oil rate due to a reduction in the oil viscosity and a reduction in the residual oil
saturation due to swelling of the oil. From this perspective, the example reservoir presents a very favorable potential
application of MEOR by gas dissolution, because of its undersaturation. If a pore volume of gas is generated, a miscible
displacement could occur. If the gas does not dissolve in the oil, then a WAG-like displacement may be possible. For a
saturated or slightly undersaturated reservoir, the generated gas will likely flow immediately and only the miscible
displacement or WAG-like mechanisms are relevant. We consider three cases that correspond to the potential behavior of the
biologically-generated gas; reduction in oil viscosity, swelling of the oil by the gas, and gas flooding.
B 3.1.1 Viscosity Reduction:
The viscosity of the oil at saturated reservoir conditions is 1.1 mPa.s. The potential reduction of viscosity was assumed to be
50%. This reduction corresponds to plausible solubilities of methane or carbon dioxide in the oil as will be discussed later.
Viscosity impacts oil displacement. Oil displacement in a waterflood can be expressed as a product of the displacement
efficiency, Ed, and the macroscopic sweep efficiency, Ev:

ROIP = E d E v OOIP

(9)

where ROIP is the recoverable oil in place from waterflooding and OOIP is the original oil in place. Both the displacement
and sweep efficiencies depend partly on the oil viscosity.
Displacement Efficiency at Economic Limit: Displacement efficiency is a measure of how much oil is swept by water in the
flooded zone. It is a function of the oil and water mobilities and therefore depends on the relative permeability relationships
and the fluid viscosities. Displacement efficiency can be determined from fractional flow defined as follows:

16

SPE 114676

fw =

1
k
1+ o w
kw o

(10)

The fractional flow values, fw, are equivalent to water cut based on reservoir volumes.
In this case, the relative permeabilities were determined from the Corey equations:

k row = k rowe (1 S wD )
k rw = k rwe (S wD )

(11)

(12)

where

S w S wi
1 S wi S orw

S wD =

(13)

The Corey parameters and viscosity data are given in Table 9. A fractional flow curve was generated from the data provided,
as shown in Figure 5. The Welge method was used to estimate the average water saturation at various fractional flow values.
The displacement efficiency is calculated from the average water saturation as follows:

Ed =

S w,avg S wi

(14)

1 S wi

Displacement efficiencies at each water cut are given in Table 10. The calculations were repeated assuming that the oil
viscosity was reduced to 0.55 mPa.s, with the results also shown in Table 10. The reduction in oil viscosity could potentially
increase the displacement efficiency by approximately 5 percentage units.
1.0
0.8

fw

0.6
0.4
0.2
0.0
0.0

0.2

0.4

0.6

0.8

1.0

Sw

Figure 5. Determination of the average saturation (Sw) in a linear waterflood displacement using the Welge Method.
An oil viscosity of 1.1 mPa.s was used in this example. fw is defined in equation 10.

SPE 114676

17

Table 9. Input parameters for fractional flow curve.

Parameter
Value
1.1
o (mPa.s)
0.3
w (mPa.s)
Swi
0.23
Sorw
0.15
krow @ Swi
0.80
0.50
krw @ 1-Sorw
m
3.5
n
2.5
Table 10. Effect of oil viscosity on displacement efficiencies at fractional flow values of 0.90, 0.95, and 0.98.
fw
0.90
0.95
0.98

o = 1.1 mPa.s
Sw,avg
Ed
0.621
0.63
0.652
0.68
0.689
0.74

o = 0.55 mPa.s
Sw,avg
Ed
0.655
0.69
0.680
0.73
0.711
0.78

Sweep Efficiency: The reservoir is currently operating at a 90% water cut. The viscosity of the oil at saturated reservoir
conditions is 1.1 mPa.s and the water viscosity is approximately 0.3 mPa.s. The given relative permeability end points are 0.8
and 0.5 for oil and water, respectively. Hence, the mobility ratio is approximately 2.6. Areal sweep correlations (Craig, 1971,
Willhite 1986) indicate that the recovery factor at breakthrough can be substantially improved if the mobility ratio is reduced.
However, the potential to increase sweep efficiency at a 90% water cut is limited. For example, if the mobility ratio were
reduced to unity (oil viscosity reduced by 60%), the sweep efficiency could be increased by approximately 2 to 5 percentage
units.
Overall: The potential improvement in recovery factor is given by:

RF = (E d E v )new (E d E v )current

(15)

The current recovery factor is 28% overall and 37% in the main area. The expected displacement efficiency at the current
90% water cut is 63%. Therefore, the sweep efficiency is currently 44% overall and 59% in the main area. If both the
displacement and sweep efficiencies were improved by 5 percentage units, the incremental recovery factors are:
Overall: RF = (0.68)(0.49) (0.63)(0.44) = 0.056
Main: RF = (0.68)(0.64) (0.63)(0.59) = 0.064
Hence, the maximum potential overall improvement in recovery factor if oil viscosity was reduced by 50% is in the order of
6%.
The microbial gas requirements are as follows. The viscosity of methane at saturated reservoir conditions is approximately
0.0005 mPa.s. If we assume the following mixing rule:

= exp{ o ln o + CH 4 ln CH 4 }

(16)

where is volume fraction, then the methane must make up 9% of the oil volume to reduce the oil viscosity by 50%. If only
methane were generated, the gas requirement per OOIP is:

G gen
OOIP

0.09Vr Boi 0.09 1.2

=
= 201 scf / stb
(1 0.09)Vr B gi 0.91 0.00059

(17)

where Ggen is the surface volume of gas required and Vr is the reservoir volume occupied by the oil and dissolved gas.
Saturation and viscosity data for CO2 in stock tank oils are reported by Whitson and Brule (2000). At 5500 psia and 210F,
the viscosity of a 1 mPa.s oil can be reduced by 50% at a CO2 content of approximately 22 mol%. For the example reservoir
fluid, 22 mol% is equivalent to 8 vol%. If only CO2 were generated, the required amount of gas per OOIP is:

18

SPE 114676

Ginj
OOIP

0.08Vr Boi 0.08 1.2

=
= 208 scf / stb
(1 0.08)Vr B gi 0.92 0.00050

(18)

Note that Equations 10 and 11 assume that all of the gas partitions to the oil.
B 3.1.2 Oil Swelling:
If gas dissolves in the residual oil, the oil will swell. It is possible that some of the excess oil volume will then be swept out of
the reservoir. An order of magnitude estimate of the potential increase in oil recovery can be made with the following
assumptions:

Generated gas that dissolves in the oil will increase the bubble point of the mixture. The process will continue until
the bubble point equals the current reservoir pressure.

All generated gas immediately dissolves in the oil until Pb=P (i.e. miscible gas)

The volume of swelling equals the volume of dissolved gas at reservoir conditions.

The residual oil saturation does not change and therefore the volume added by swelling will be produced.

The composition of this produced volume is the average composition of the swollen oil. This composition will
change as more gas dissolves in the oil but will be approximated as the composition of a mixture with a bubble point
halfway between the current bubble point and the current reservoir pressure.

The first step is to determine how much gas dissolves in the oil until the bubble point equals the reservoir pressure. A
rigorous calculation would require a compositional fluid model. A simple order of magnitude estimate can be obtained from
the black oil PVT data. The amount of gas that dissolves to reach a new bubble point is given by:

SGIP
= Rs ( Pr ) Rs ( Pb )
OOIP

(19)

where SGIP is solution gas oil in place, Rs is the solution gas-oil ratio, Pr is the reservoir pressure, and Pb is the original
bubble point pressure. The solution gas-oil ratio at the original bubble point pressure of 835 psia is 180 scf/stb. This point
was fit with the Vasquez-Beggs correlation, as shown in Figure 6. The same correlation was then used to calculate the
solution gas-oil ratio for a bubble point pressure equal to the current reservoir pressure of 5500 psia. The calculated solution
gas-oil ratio at the new bubble point is 1710 scf/stb. Hence, the amount of microbial gas that can dissolve in the oil is 1710
180 = 1530 scf per stb of oil.
The volume of dissolved microbial gas at reservoir conditions is given by:

Vdg
RmOIP

SGIP
Bg
RmOIP

(20)

And if the residual oil saturation does not change, the volume of displaced fluid equals Vdg. If all of the displaced fluid is oil,
the maximum change in oil recovery is given by:

RF =

SGIP B g
RmOIP Bo

(21)

At 5500 psia, the calculated values of Bo and Bg are 1.86 rb/stb and 0.654 rb/Mscf. The incremental oil recovery is then:

RF =

SGIP B g
0.000654
= 1530
= 0.54
1.86
RmOIP Bo

(22)

SPE 114676

19

2000
PVT data
Vasquez-Beggs, Pb=Psat
Vasquez-Beggs, Pb=Pr

Rs (SCF/STB)

1500

1000

500

0
0

2000

4000

6000

Pressure (psia)
Figure 6. Fitting and extrapolation of solution gas-oil ratio with the Vasquez-Beggs correlation.

The predicted maximum incremental oil recovery is 54%. This estimate is too high because the dissolved gas will be
produced with the oil. When the bubble point reaches 5500 psia, the additional dissolved gas makes up approximately 50% of
the fluid volume at reservoir conditions. Therefore, while displaced fluid is flowing, the average dissolved gas makes up
approximately 25% of the fluid. Hence, the incremental oil recovery should only be 75% of the maximum value; that is,
0.75x0.54 = 0.40.
This analysis suggests that if 1530 scf of microbial gas can be generated per stb of oil, then 40% of the remaining oil could be
recovered. Assume that only the currently swept zones are affected by this mechanism (oil is still capable of flowing in
unswept areas). The remaining oil saturation in the swept zones is in the order of 38% (see Table 10). The potential
improvement in recovery factor is then 0.40x0.38=0.15 or 15% of the OOIP.
B 3.1.3 Gas Flood:
Here, the most optimistic assumption is that no gas dissolves in the oil. Assume that 0.20 pore volumes of gas are required.
Immiscible gas floods have increased the recovery factor of existing waterfloods by up to 20% but more typically by 5%
(Christensen et al., 2001; Awan et al., 2006). Assume a 10% increase in oil recovery. Note, the estimated recovery is
optimistic because vertical displacement is unlikely given that the bacteria are to be distributed throughout the reservoir.
The gas requirement per OOIP is:

(0.2 PV )gas
=
OOIP ((1 S wi ) PV )oil
G gen

Boi
0.2
1.2

=
= 477 scf / stb
B
0
.
77
0
.
000654
gi

(23)

At a 28% current recovery factor and assuming that the generated gas requirement is still 0.2 pore volumes, the required
amount of gas per remaining OIP is

(0.2 PV )gas

=
RmOIP ((1 RF )(1 S wi ) PV )oil
G gen

Boi
0.2
1.2

B = (0.72)(0.77) 0.000654 = 662 scf / stb

gi

(24)

B.3.2 Material Balance Calculations

Few yield factors for gas formation from anaerobic decomposition of hydrocarbons have been reported, therefore, we adapted
data from two sources:
Case 1. Formation of methane and carbon dioxide from the anaerobic fermentation of oleic acid, which has a
predominant hydrocarbon composition with an acid functional group (octadecenoic acid). The yield factors calculated
from Pereira et al. (2002) were:
Biomass yield

1.50 g biomass/g hydrocarbon

20

SPE 114676

Maximum total gas yield

Methane yield

487 Std. L/kg (7.9 SCF/lb) (methane + CO2)

292 Std. L/kg (4.7 SCF/lb)

Case 2. Formation of carbon dioxide from degradation of aromatics (benzene) under nitrate reducing conditions from
Burland and Edwards (1999). They observed a minimum requirement of 10 mol of nitrate consumed for every mol of
benzene. Assuming that petroleum degrades with the same yield factors as pure benzene, we obtain:
Biomass yield
Maximum CO2 yield

0.11 kg dry biomass/kg hydrocarbon

1571 std L/kg hydrocarbon
170 std L/kg nitrate (as ammonium nitrate)

These yield factors were used to calculate the conversion of oil required from the above cases, taking the optimistic
case that all of the gas product was carbon dioxide.
Case 3. Addition of carbohydrate
As an alternative to conversion of the oil in place, we also considered the gas formation from adding sugars such as
molasses under anaerobic conditions. Values of circa 742 std L/kg (12 SCF/lb) of dissolved organic are common in
wastewater treatment (Bailey and Ollis, 1986).
Based on the assumptions of the reservoir engineering models, the incremental recovery from gas drive MEOR mechanisms
would range from 6% due to viscosity reduction to 15% due to swelling of the oil. Since the gas required for each mechanism
differed, the results were normalized in Table 11, on a basis of inputs required per incremental barrel of production due to
MEOR. In each case, optimistic assumptions were used by assuming that all of the gas yield was carbon dioxide, which then
dissolved only in the oil in mechanisms A and B, or which did not dissolve in oil or water for the gas drive mechanism
(Mechanism C).
Table 11. Incremental oil recovery ratios from material balance calculations

Mechanism

A. Viscosity
reduction
B. Swelling
C. Gas drive

Gas required

Kg crude oil converted/bbl

recovered

Incremental
recovery,
%OOIP
6

SCF/b
bl

Std
L/bbl

Case 1
Alkanes

Case 2
Aromatics

208

5844

200

62

Case 2
Nitrate
required
Kg
nitrate/bbl
recovered
571

15
10

1530
662

42985
18599

588
382

182
118

1680
1091

Case 3
Carbohydrate/bbl
recovered
Kg carbohydrate/bbl
recovered
131
386
250

If the microbes convert residual oil in place, the cases considered required conversion of 62-588 kg crude oil/barrel of crude
oil recovered. At first glance, these lower bound of this range may appear to be a reasonable target. The requirement for
nitrate, however, puts this approach into perspective. Nitrate is required for the respiration of the microbes in the reservoir,
and to ensure that souring is avoided. Burland and Edwards (1999) observed that at least 10 mol of nitrate were consumed for
every mol of benzene under anaerobic conditions. Their data indicated values as high as 40 mol nitrate/mol benzene
converted. Consequently, although the gas yields from this Case 2 are high per unit of crude oil converted, the cost in terms
of nitrate addition is very high. Given the data of Burland and Edwards (1999), the yield of carbon dioxide is 170 std L/kg
nitrate (as ammonium nitrate). For the best case listed above, the addition of at least 571 kg of nitrate would be required for
each incremental barrel of crude oil recovered. One barrel of the example North Sea crude oil contains 137 kg of oil,
therefore, the amounts of nitrate required are clearly much too high for these mechanisms to be feasible.
Use of carbohydrates as nutrients instead of crude oil would give intermediate yields of gas; lower than conversion of
aromatics but higher than n-alkanes. The amounts of carbohydrates required for incremental production are all in the range of
1 to 3 kg carbohydrate/kg incremental crude oil recovered, which makes this Case 3 completely infeasible. This conclusion is
completely consistent with the previous study by Sarkar et al. (1991).
These unfavorable ratios are the best case, and do not consider a number of biological barriers, such as the problem of placing
the bacteria and the added nutrient material throughout the reservoir volume, to avoid plugging the injectors. For cases 1 and
2, the conversion of the lightest fractions of the crude oil in the swept zone would significantly increase the viscosity of the
remaining material and give shrinkage that would invalidate the above analysis. Conversion of only 5% of the crude, all of
which would be in the 350 C- fraction of the crude (C20 or less) would increase viscosity by 1.5 times (Peries et al. 1988).
The most optimistic case in Table 11 requires conversion of 2.7% of the original oil in place, or 7% of the oil in the swept
zone. Consequently, this mechanism would increase the viscosity of the oil by at least 1.5 times in the swept zone, on a dead

SPE 114676

21

oil basis. The benefit of the dissolved gas would offset this result, but the net effect would be less improvement in viscosity
and a lower incremental recovery than indicated in Table 11.
These calculations show that biological generation of gas from the oil in place is not a feasible method to recover additional
crude oil, whether by swelling of the oil, reducing its viscosity, or by gas flood. In particular, the requirement for nitrate to
control souring and to sustain large scale modifications to the reservoir composition makes the approach infeasible. This
conclusion does not rule out the possibility of gas production from residual oil as an objective in it own right; but this
objective was outside the scope of the study.
B.4 Microbial Solvent Production In Situ
Another suggested mechanism for MEOR is the production of solvents in situ, which would dissolve in the crude oil and
reduce its viscosity. A microbially generated liquid solvent such as butanol would likely partition between the oil and water
phases in the reservoir. The portion of the solvent that dissolves in the oil may reduce the oil viscosity, increasing its
mobility. As discussed in Section B.3, increased oil mobility can lead to increased oil recovery at a given economically
limiting water-oil ratio.
However, as discussed in Section B.3, a 50% reduction in oil viscosity is required to gain 6% incremental oil recovery for the
example reservoir. The oil viscosity of this reservoir is 1.1 mPa.s. The viscosity of n-butanol at 99C is 0.55 mPa.s (Reid et
al., 1977). There is so little difference in viscosity between the oil and the solvent, that a significant reduction in viscosity is
not possible. Even a small potential gain in oil recovery would require very large volumes of solvent. For the case of the
example North Sea reservoir, this mechanism offers no benefit in terms of oil recovery.
Discussion
Prospects for Individual Mechanisms:
The prospects for MEOR mechanisms are summarized in Table 12, based on the analysis of material balance calculations and
an assessment of reservoir performance from this study. The incremental oil yield in each case is given on a consistent basis
of L of incremental crude oil per kg of added material. The mechanisms were rated as good, poor, or very poor on the basis of
the yield ratio, which was the ratio of incremental oil production to the input of nutrients or other materials to the reservoir. A
mechanism with an estimated yield ratio significantly greater than unity, under best case conditions, was rated as a good
prospect for consideration or development. A mechanism with a yield ratio near unity was rated as a poor prospect, because
our analysis constituted an optimistic best-case scenario for the performance of each mechanism in terms of reservoir
transport and microbial effectiveness. Mechanisms with yield ratios of much less than one were rated as very poor.
Table 12 Summary of MEOR Mechanisms for Light Crude Oil in Sandstone Reservoirs
Mechanism

Paper
Section
Modification of interfacial properties
Biosurfactants
A.1

Alteration of wettability

A.2

Interfacial bacterial films

A.3

Yield Factor, L of crude

oil/kg added material

Prospects
for MEOR

Reasons

Yield from bacterial crude oil

conversion: 1.1 L/kg of oil
converted
Yield from added nutrients: 0.8
L/kg of added material
<<1 L/kg (little evidence for
dependence in sandstones)

Poor

a) Insufficient change in IFT

(capillary number too large)
b) Issues for generation and loss
of biosurfactant

Very Poor

Unknown

Unknown

Lack of data showing

dependence of recovery on
wettability, and lack of methods to
control wettability
Mechanisms of reservoir plugging
by bacteria + emulsified oil need
more study

Bulk modification of fluid or reservoir properties

Plugging of highB.1
18 L/kg nutrient (fractures) to
permeability zones
1.3 L/kg (permeable layers)
Organic acid to open flow
B.2
<0.3 L/kg
paths
Gas production to mobilize
B.3
Nitrate addition for oil
oil
conversion <0.3 L/kg
Carbohydrate addition < 1.2
L/kg (Table D3.3)
Solvent production in situ
B.4
<< 1 L/kg (little dependence)

Good
Very poor
Very poor

Very poor

Good potential only for fractured

reservoirs
Biomass volume exceeds volume
opened by acid
Large amounts of materials
(nitrates and other nutrients) are
required
Large amounts of materials
(nitrates and other nutrients) are
required

22

SPE 114676

The first group of mechanisms has the attraction of modifying only interfacial properties, offering the potential for maximum
impact with minimum input required from the surface. Although the study of biosurfactants over many years has been
justified partly for MEOR applications, the production of biosurfactant alone did not improve the capillary number enough
for the example reservoir to give significant mobilization of the oil. Biosurfactants and chemical surfactants share this
common characteristic. Given only a marginal increase in production, the amount of input material makes this approach
unattractive. Also several key issues have not been seriously considered in the biosurfactant literature including the losses of
surfactant to the surface of the reservoir minerals, and biodegradation of surfactants by other species already present in the
reservoir. Stimulation of biosurfactant production over an extended period to sweep a reservoir would inevitably stimulate a
population of biosurfactant degraders.
Changes in the wettability of a sandstone formation are unlikely to give a significant benefit. The residual oil saturation of
sandstones does not vary significantly with wettability. Also, wettability effects are likely to be dwarfed by the simultaneous
effect of a reduction in interfacial tension. The most interesting prospect is the use of bacteria to attach to the crude oil
interface, giving emulsified droplets that could plug permeable zones in the reservoir. This mechanism would not plug
fractures, but it could offer control of flow into high permeability sandstone zones. Enhancing the conformance by promoting
flow of water into low-permeability zones would increase production. Further experimental investigation of this potential
mechanism could be fruitful.
The second group of mechanisms required the modification of the bulk properties of the reservoir or the reservoir fluids. In
every case, material balance calculations showed that attempts to change the properties of a significant volume of reservoir
rock or reservoir fluids were impractical. The amount of input material required would exceed the incremental production of
crude oil. The one exception is the use of bacteria and nutrients to plug high permeability fractures with biopolymer and cells.
In this case, the choice of a suitable reservoir has the potential to enable a significant improvement in the areal sweep
efficiency with a minimal addition of material (bacteria and nutrients) from the surface. Suitable reservoirs would need to be
identified where fracturing is controlling, that offer blockable fractures which control conformance. The selection of suitable
organisms to produce a stable biopolymer appears essential to the success of this approach. The performance of these
candidate bacteria must be assessed under the conditions of the reservoir, including the presence of the crude oil and the
actual local pH. The latter parameter is often reduced in situ by the dissolution of carbon dioxide at the high pressure that
exists in the reservoirs. The operation of the waterflood must also take into account the reduction in injectivity caused by
plugging of highly permeable fractures. If the flow rate is maintained constant by increasing the injection pressure, then new
fractures will tend to form, negating the benefit of the MEOR treatment.
The main disagreement between this study and previous work is our conclusion that the potential for the use of biosurfactants
is poor. For example, although Sarkar et al. (1991) agreed that bulk modification of the reservoir or its fluids by injecting a
carbon source such as glucose was impractical, they proposed that the production of biosurfactants could achieve significant
improvement in recovery. Unfortunately, their calculations and simulations to support this conclusion assumed that the
bacteria produced not only surfactant, but also alcohols and polymer to enable a combination of low IFT and viscous
displacing fluid. Their calculations assumed a capillary number of ca. 1.5 x 10-4 due to the combination of these factors,
which is comparable to our best estimate for biosurfactant, based on the lowest IFTs reported for biosurfactant alone (the
most optimistic estimate for biosurfactant was a capillary number of 2 x 10-4, based on an IFT of 0.06 mN/m). While our
estimates were similar in the magnitude of incremental oil recovery (Sarkar et al. (1991) estimated that 11% incremental oil
recovery could be achieved) our calculations showed that the material input requirements to achieve this goal were very
large. One major unknown in this analysis, which requires better data, is the loss of biosurfactant due to adsorption. This
effect is well known in the EOR application of chemical surfactants, but has received little attention in the literature on
biosurfactants. The second major unknown is the rate of biodegradation of the surfactant by other organisms. Such
degradation will be enhanced by sustained addition of nutrients to achieve a flood of biosurfactant formed in situ.
Regardless of estimates of the benefit of interfacial tension, the growth of organisms in situ on the oil water interface will
have a significant role in producing complex emulsions that are stabilized by cells and surfactant. The flow of such emulsions
through the reservoir will be complex, and may lead to plugging as discussed in Section A.3.
The conclusions of this study are almost completely independent of the specific details of the reservoir; we would expect
similar results for a range of permeabilities and pore size distributions for other sandstone reservoirs with light crude oils.
Mechanisms that affect the mobility ratio, such as viscosity reduction, depend on the relative permeability curves and the
permeability variation of the reservoir. Somewhat higher incremental oil recovery may be possible for reservoirs with
reservoir heterogeneity. Mechanisms that mobilize residual oil, such as surfactant addition, will be more effective for higher
residual oil saturation. As well, the adsorption of surfactant on the reservoir rock varies considerably between reservoirs
depending on the clay content. However, since optimistic evaluations were performed in all cases, these differences are
unlikely to alter the conclusions of this study.

SPE 114676

23

In every case, the analysis assumed effective transport of the bacteria in the reservoir to consider the most optimistic possible
case for each mechanism. Consequently, the exact size and adhesion properties of the bacteria did not affect the screening of
mechanisms. Similarly, the details of salinity and temperature did not enter our analysis; we assumed that active organisms
with the desired potential were available in each case. The pressures of oilfield operation have an indirect impact on bacterial
growth and/or by-product formation through the dissolution of carbon dioxide at elevated pressure. This variable must be
considered in screening bacteria for in situ operations.
Combinations of Mechanisms:
The prospects for combining mechanisms for bulk modification of the reservoir or its fluids are as poor as each mechanism
independently. Additive combinations of mechanisms that each require the addition of large quantities of material, for
example, gas production plus solvent production, will still fail due to the overall material balance. Combinations of
mechanisms that give synergistic or non-linear interactions are more interesting, but also more difficult to access. First, we
will consider the mechanism of biosurfactant production, which was rated as a poor prospect in our analysis but which has
been promoted by many previous investigators. Two possible combinations could enhance the effectiveness of biosurfactant:
1. Combine biosurfactant with an alcohol as co-surfactant and a polymer to increase the viscosity. This combination provides
lower IFT than the surfactant alone, and the polymer will increase the viscosity of the aqueous phase to achieve more
efficient displacement of the oil. This combination was used by Sarkar et al. (1991) in their analysis of MEOR, and the data
were presented in Figure 3 show that the large capillary numbers achieved by this combination could indeed give significant
mobilization of residual oil. This combination of mechanisms requires simultaneous release of surfactant, alcohol, and a
soluble polymer in the aqueous phase. Unfortunately, none of the studies of biosurfactant production have demonstrated such
a combination. In addition, this combination requires a significant change in the bulk composition of the mobile aqueous
phase, to give a polymer bank with a significant fraction of the pore volume of the reservoir. This combination no longer
requires a modification of a thin interface between the oil and the water, rather, it requires a much more significant shift in
composition, which in turn requires the addition of large quantities of nutrient material from the surface. Even if organisms
could be isolated to simultaneously provide all three components to create a surfactant/polymer flood, the attendant material
requirements and biomass production would be a serious issue.
2. Control the pH in order to enhance the effectiveness of the surfactant or to suppress adsorption to mineral surfaces. Some
of the lipopeptide surfactants give lower IFT at pH 5 to 6 (Table 5). In the sandstone case, such a low pH may occur naturally
depending on the amount of carbon dioxide in the gas phase that is available to dissolve. Adding acid to reduce the pH of an
entire reservoir is not an attractive option. Conversely, the use of pH >7 has been proposed to control the adsorption of
anionic surfactants to mineral surfaces, particularly in carbonate reservoirs (Zhang et al. 2006). Addition of a base, such as
sodium carbonate, is as unattractive as adding an acid. The biosurfactants are all non-ionic compounds, therefore, their
adsorption would not be as sensitive to pH as non-ionic surfactants. These comments suggest that future evaluation of
biosurfactants must carefully consider the in situ pH of the target reservoir, as well as the salts and mineral chemistry that
influence surface adsorption.
Evaluation of New MEOR Mechanisms:
We suggest that proposed new MEOR mechanisms can be evaluated on a preliminary basis by following the approach taken
in this study. We recommend the following sequence of analysis of new proposals. The first step is to define the potential
mechanism in enough detail to allow material balance calculations using a simple model of selected reservoir type. If the
mechanism is analogous to established EOR methods, then documentation on these designs can allow an assessment of
performance. Where these data are lacking, then an experimental plan will be required to measure the intended mechanism
and check the proof of concept. An example is the role of bacterial surface skins on the plugging of pores by oil droplets,
where only the most preliminary results are available in the open literature.
The heart of the assessment process is to check the mass balance, to determine the potential oil recovery versus the amount of
input materials required. The literature on MEOR contains many examples where a simple mass balance has not been
considered in evaluating a proposed scheme. If the yield factor is favorable at this point, then the feasibility of the proposed
biology must be assessed. The robustness of any added organisms must be addressed under in situ conditions, as well as the
ability of the mechanism to function in the presence of other organisms indigenous to the reservoir. Given the almost certain
presence of an active mixed population in a reservoir, then specific organisms or reaction paths must out-compete the
indigenous organisms. The recommended practice is to consider indigenous organisms isolated from the reservoir for MEOR
applications due to their tolerance for the specific conditions in the reservoir. These organisms can then be cultured at the
surface and injected into the formation.

24

SPE 114676

Finally, a more complete assessment of a proposed mechanism must go beyond the present study and consider the issues of
transport in the subsurface, to ensure that the organisms, nutrients and products will be transported and distributed in the
reservoir as intended. The analysis in this study was highly simplified by assuming the uniform distribution of nutrients and
microbes into a heterogeneous flow environment in a reservoir.
Acknowledgements
The authors are grateful to Duncan Anderson of The Industry Technology Facilitator (ITF), Aberdeen, for his help and
support in securing funding for this work from the following companies: BP, Chevron, ConocoPhillips, Eni, Maersk Oil,
Shell, StatoilHydro, and Total.
References
Adamson, A.W. and Gast, A.P. (1997). Physical Chemistry of Surfaces, 6th Ed., John Wiley & Sons, New York.
Aitken, C.M., Jones, D.M. and Larter, S.R. (2004) Anaerobic hydrocarbon biodegradation in deep subsurface oil reservoirs. Nature.
431:291-294
Alpkvist, E., Picioreanu, C., van Loosdrecht, M. C. M., and Heyden, A. (2006). "Three-dimensional biofilm model with individual cells
and continuum EPS matrix." Biotechnol. Bioeng., 94:961-979.
Awan, A.R., Teigland, R., Kleppe, J., EOR Survey in the North Sea, SPE 99546, Tulsa, April 22-26, 2006.
Bailey J.E. , Ollis, D.F. Biochemical engineering fundamentals 2nd ed., New York: McGraw-Hill, 1986.
Beeder, J., Nilsen, R. K., Thorstenson, T., and Torsvik, T. (1996). "Penetration of sulfate reducers through a porous North Sea oil
reservoir." Appl. Environ.Microbiol. , 62: 3551-3553.
Benincasa, M., A. Abalos, et al. (2004). "Chemical structure, surface properties and biological activities of the biosurfactant produced by
Pseudomonas aeruginosa LBI from soapstock." Antonie Van Leeuwenhoek International J. General and Molecular Microbiol. 85: 18.
Bothun, G.D., B.L. Knutson, J.A. Berberich, H.J. Strobel, and S.E. Nokes. (2004) Metabolic selectivity and growth of Clostridium
thermocellum in continuous culture under elevated hydrostatic pressure. Appl. Microbiol. Biotech. 65:149-157.
Bryant, S. L. and T. P. Lockhart (2000). Reservoir engineering analysis of microbial enhanced oil recovery, SPE 63229. SPE Annual
Technical Conference and Exhibition, Dallas, TX, 1-4 October, 2000, SPE.
Burland, S.M., Edwards, E.A. Anaerobic benzene biodegradation linked to nitrate reduction. Appl. Environ. Microbiol. 1999, 65: 529-533.
Buzza, D.M.A. (2002). General theory for capillary waves and surface light scattering. Langmuir 18:8418-8435.
Cameotra, S.S. and R.S. Makkar. (1998) Synthesis of biosurfactants in extreme conditions. Appl. Microbiol. Biotechnol. 50:520-529.
Chayabutra, C., J. Wu, et al. (2001). "Rhamnolipid production by Pseudomonas aeruginosa under denitrification: Effects of limiting
nutrients and carbon substrates." Biotechnol. Bioeng. 72: 25-33.
Choi, K. S., S. H. Kim, et al. (1999). "Purification and characterization of biosurfactant from Tsukamurella sp. 26A." J. Microbiol.
Biotechnol. 9: 32-38.
Christensen, J.R., Stenby, E.H., Skauge, A., (2001) Review of WAG Field Experience, SPE Res. Eval. & Eng., April 2001, 97-106.
Daoshan, L., L. Shouliang, Yi, L, and W. Demin. (2004) The effect of biosurfactant on the interfacial tension and adsorption loss of
surfactant in ASP flooding. Colloids and Surfaces A: Physicochemical Engineering Aspects. 244: 53-60.
Davis, D. A., H. C. Lynch, et al. (1999). "The production of Surfactin in batch culture by Bacillus subtilis ATCC 21332 is strongly
influenced by the conditions of nitrogen metabolism." Enzyme and Microbial Technology 25: 322-329.
Deleu, M., H. Razafindralambo, et al. (1999). "Interfacial and emulsifying properties of lipopeptides from Bacillus subtilis." Colloids and
Surfaces a-Physicochemical and Engineering Aspects 152: 3-10.
Dorobantu, L.S., A. Yeung, J.M. Foght and M.R. Gray (2004). Stabilization of oil-water emulsions by hydrophobic bacteria. Appl.
Environ.Microbiol. , 70:6333-6336.
Fujiwara, K., Sugai, Y., Yazawa, N., Ohno, K., Hong, C. X., and Enomoto, H. (2004). "Biotechnological approach for development of
microbial enhanced oil recovery technique." In: Petroleum Biotechnology: Developments and Perspectives. R. Vazquez-Duhalt and R.
Quintero-Ramirez (eds.) Studies in Surface Science and Catalysis 151:405-445, Elsevier, Amsterdam.
Head, I.M., D.M. Jones, and S.R. Larter. (2003) Biological activity in the deep subsurface and the origin of heavy oil. Nature. 426:344-352.
Hung, H. C. and G. S. Shreve (2001). "Effect of the hydrocarbon phase on interfacial and thermodynamic properties of two anionic
glycolipid biosurfactants in hydrocarbon/water systems." J. Physical Chemistry B 105: 12596-12600.
Huszcza, E. and B. Burczyk (2003). "Biosurfactant production by Bacillus coagulans." J. Surf. Deter. 6(1): 61-64.
Jenneman, G. E., Knapp, R. M., McInerney, M. J., Menzie, D. E., and Revus, D. E. (1984). "Experimental studies of in situ microbial
enhanced oil recovery." Soc. Pet. Eng. J., 24:33-37.
Jones, D.M., Head, I.M., Gray, N.D., Adams, J.J., Rowan, A.K., Aitken, C.M., Bennett, B., Huang, H., Brown, A., Bowler, B.F.J.,
Oldenburg, T., Erdmann, M., Larter, S.R. (2008). Crude-oil biodegradation via methanogenesis in subsurface petroleum reservoirs.
Nature 451:176-180.
Kang, Z.; Yeung, A.; Foght, J.M.; Gray, M.R. (2008) Mechanical Properties of Hexadecane-Water Interfaces with Adsorbed Hydrophobic
Bacteria Colloid and Surfaces B. Biointerfaces, 62:273-279.
Kantzas, A., Chatzis, I., Dullien, F.A.L., Enhanced Oil Recovery by Inert Gas Injection, SPE 17379, Tulsa, April 17-20, 1988.
Kim, H. S., J. W. Jeon, et al. (2002). "Surface and physico-chemical properties of a glycolipid biosurfactant, mannosylerythritol lipid, from
Candida antarctica." Biotechnol. Lett. 24:1637-1641.
Kowalewski, E., I. Ruesltten, et al. (2005). Interpretation of Microbial Oil Recovery from Laboratory Experiments. 13th European
Symposium on Improved Oil Recovery, Budapest, Hungary, 25 - 27 April, 2005.
Kowalewski, E., I. Rueslatten, et al. (2006). "Microbial improved oil recoverybacterial induced wettability and interfacial tension effects
on oil production." J. Petroleum Science and Engineering 52:275-286.

SPE 114676

25

Lang, S. and D. Wullbrandt (1999). "Rhamnose lipids - biosynthesis, microbial production and application potential." Appl. Microbiol.
Biotechnol. 51: 22-32.
Lappin-Scott, H. M., Cusack, F., and Costerton, J. W. (1988). "Nutrient resuscitation and growth of starved cells in sandstone cores: A
novel approach to enhanced oil recovery." Appl. Environ. Microbiol., 54:1373-1382.
Lenz, T.G. and Moreira, A.R. (1980) Economic evaluation of the acetone-butanol fermentation. Ind. Eng. Chem. Res. Dev. 19: 478-483.
Lin, S.-C., M. A. Minton, et al. (1994). "Structural and immunological characterization of a biosurfactant produced by Bacillus
licheniformis JF-2" Appl. Environ.Microbiol. 60: 31-38.
Makkar, R. S. and S. S. Cameotra (1999). "Structural characterization of a biosurfactant produced by Bacillus subtilis at 45 degrees C." J.
Surf. Deter. 2:367-372.
McInerney, M.J., Duncan, K.E., Youssef, N., Fincher, T., Maudgalya, S.K., Folmsbee, M., Knapp, R.M., Simpson, D.R., Nagle, D. (2005).
Development of microorganisms with improved transport and biosurfactant activity for enhanced oil recovery. US Department of
Energy Report DE-FC-02NT15321 R02. http://www.osti.gov/energycitations/purl.cover.jsp?purl=/860919-GXQnLc/
Mohan, P.K., G. Nakhla and E.K. Hanful. (2006a) Biokinetics of biodegradation of surfactants under aerobic, anoxic and anaerobic
conditions. Water Research 40:533-540.
Mohan, P.K., G. Nakhla and E.K. Hanful. (2006b) Biodegradability of surfactants under aerobic, anoxic and anaerobic conditions. J.
Environmental Engineering. 132:279-283.
Oren, P.E., (1994)Effect of Wettability and Spreading on recovery of Waterflood Residual Oil by Immiscible Gasflooding, SPE
Formation Eval., June 1994, 149-156.
Pereira, M.A., Mota, M. Alves, M.M. (2002) Operation of an anaerobic filter and an EGSB reactor for the treatment of an oleic acid-based
effluent: influence of inoculum quality Process Biochem. 2002, 37:1025-1031.
Peries, J. P., Quignard, A., Farjon, C., and Laborde, M. (1988). "Thermal and Catalytic Asvahl Processes under Hydrogen Pressure for
Converting Heavy Crudes and Conventional Residues." Revue De L Institut Francais Du Petrole, 43:847-853.
Powell, K. L., Taylor, R. G., Cronin, A. A., Barrett, M. H., Pedley, S., Sellwood, J., Trowsdale, S. A., and Lerner, D. N. (2003). "Microbial
contamination of two urban sandstone aquifers in the UK." Water Research, 37:339-352.
Reid, R.C., Prausnitz, J.M., Sherwood, T.K., (1997) The Properties of Gases and Liquids, 3rd Ed., McGraw Hill, New York.
Sarkar, A.K.; Sharma, M.M.; Georgiou, G. (1991) Compositional numerical simulation of MEOR processes. In Microbial Enhancement
of Oil Recovery Recent Advances, Donaldson, E.C. ed., Elsevier, Amsterdam, pp. 331-343.
Sarkar, A.K.; Sharma, M.M.; Georgiou, G. (1995). "Strength and Stability of Microbial Plugs in Porous Media," Proceedings of the Fifth
International Conference on Microbial Enhanced Oil Recovery and Related Biotechnology for Solving Environmental Problems,
Plano, TX, September 1995, pp 213-234,
Scriven, L.E. (1960). Dynamics of a fluid interface. Chemical Engineering Science, 12:98-108.
Slattery, J.C. (1974). Interfacial effects in the entrapment and displacement of residual oil. AIChE J., 20: 1145-1154.
Stalkup, F.I., Jr., (1984) Miscible Displacement, SPE Monograph, SPE, Richardson, TX, p.3.
Staudt, C., Horn, H., Hempel, D. C., and Neu, T. R. (2004). "Volumetric measurements of bacterial cells and extracellular polymeric
substance glycoconjugates in biofilms." Biotechnol. Bioeng., 88: 585-592.
Stiles, W.E., (1949)Use of Permeability Distributions in Waterflood Calculations, Trans. AIME, 15: 9-12.
Tanner, R.S., Udegbunam, E.O., McInerney, M.J. and Knapp, R.M. (1991). "Microbially enhanced oil recovery from carbonate
reservoirs." GeoMicrobiol. J. 9: 169-195.
Urum, K. and T. Pekdemir (2004). "Evaluation of biosurfactants for crude oil contaminated soil washing." Chemosphere 57: 1139-1150.
Wanger, G., G. Southam, et al. (2006). "Structural and chemical characterization of a natural fracture surface from 2.8 kilometers below
land surface: Biofilms in the deep subsurface." GeoMicrobiol. J. 23: 443-452.
Whitson, C.H., Brule, R.B., (2000) Phase Behavior, SPE Monograph 20, SPE, Richardson, TX.
Willhite, G.P., Green, D.W. (1998) Enhanced Oil Recovery, SPE Textbook Series Vol.6, SPE, Richardson, TX
Willhite, G.P. (1986) Waterflooding, Society of Petroleum Engineers, Richardson, TX.
Yakimov, M. M., Amro, M. M., Bock, M., Boseker, K., Fredrickson, H. L., Kessel, D. G., and Timmis, K. N. (1997). "The potential of
Bacillus licheniformis strains for in situ enhanced oil recovery." J. Petroleum Science and Engineering, 18: 147-160.
Youssef, N., Simpson, D.R., Duncan, K.E., McInerney, M.J., Folmsbee, M., Fincher, T., Knapp, R.M. (2007). In situ biosurfactant
production by Bacillus strains injected into a limestone petroleium reservoir. Appl. Environ. Microbiol. 73: 1239-1247.
Zhang, D. L., Liu, S. H., Puerto, M., Miller, C. A., and Hirasaki, G. J. (2006). "Wettability alteration and spontaneous imbibition in oil-wet
carbonate formations." J. Petroleum Science and Engineering, 52: 213-226.