Subhash C Ayirala is a Post-Doctoral Research Associate in the Craft and Hawkins Department of Petroleum Engineering, Louisiana State University, Baton Rouge, US.

He previously worked as a Process Engineer in Amara Raja Batteries Limited, Tirupati, India. Subhash holds a B.Tech with Distinction in Chemical Engineering from Srivenkateswara University, Tirupati, India and an M Tech in Chemical Engineering with Petroleum Refinery Engineering specialization from the Indian Institute of Technology, Kharagpur, India.
He also holds an MS and PhD in Petroleum Engineering from Louisiana State University, Baton Rouge, USA. Subhash is a member of the Society of Petroleum Engineers and Pi Epsilon Tau, a petroleum engineering honorable society.
Dandina N Rao is an Associate Professor in the Craft and Hawkins Department of Petroleum Engineering, Louisiana State University, Baton Rouge, USA and holds the Emmett G Wells Jr Distinguished Professorship from the LSU College of Engineering.
He previously worked as a Principal Petroleum Engineer with TRVU/BDM Petroleum Technologies (Formerly NIPER); as a Staff Engineer with Hydro flame Engineering Consultants; as a Staff Research Engineer with the Petroleum Recovery Institute; and as a Research Engineer with Shell Canada. He holds a B Tech with Distinction from India, an MS from the University of Saskatchewan (Canada) and a PhD from the University of Calgary (Canada).
He is a registered Professional Engineer in Alberta (Canada) since 1986. In addition to being a member of the Petroleum Society (Canada), SPE, Pi Epsilon Tau, and Sigma Xi, he has served as a member of the Editional Review Board of the Journal of Canadian Petroleum Technology for over seven years and has served as the Technical Editor o f the SPE Journal for five years.

ABSTRACT
The use of surfactants has long been considered in the oil industry for oil recovery enhancement only through reduction in oil-water interfacial tension. The potential capability of these surfactants to enhance oil recovery through wettability alteration remains largely ignored.
Hence the purpose of this study is to experimentally determine the influence of surfactant type and concentration on oil recovery, oil-water relative permeabilities and wettability in reservoir rocks. Several coreflooding experiments have been conducted using Yates reservoir fluids in Berea rocks and two types of surfactants (nonionic and anionic) in varying concentrations.
A coreflood simulator has been used to calculate oil-water relative permeabilities by history matching the recovery and the pressure drop data obtained from the corefloods. The relative permeability variations have been interpreted to characterise wettability alterations induced by the surfactants. The very high oil recoveries and the relative shifts in the oil-water relative permeability ratio curves to the right obtained, indicate the capability of both of the surfactants to develop a special kind of heterogeneous wettability known as “mixed-wettability” for potential oil recovery improvements.
Thus this study opens up a new avenue for the use of low-cost surface-active chemicals in improved oil recovery applications, utilizing their ability to alter wettability instead of the conventional approach of using them for interfacial tension reduction.

INTRODUCTION
Nearly two-thirds of original oil in place found in crude oil reservoirs remains nonrecovered, even at the end of primary depletion and secondary waterfloods. This is mainly due to the capillary forces that prevent oil from flowing within the pores of reservoir rock, thereby trapping large amounts of mobile oil. The effect of these capillary forces on trapping of oil within the pores of reservoir rock can be generalised by the use of a dimensionless number, called the capillary number. The capillary number (Nca) is defined as the ratio of viscous to capillary forces, given by,

Nca = Viscous Forces  =  vµ
 Capillary Forces  scosq (1)

where v and µ are the velocity and viscosity, respectively of the displacing fluid, s is the oil-water interfacial tension and q is the contact angle (wettability).
The effect of capillary number on residual oil saturation) shows that an increase of four to five orders of magnitude in capillary number is required in any improved oil recovery process to reduce the residual oil saturation significantly.
The capillary number can be increased by either reducing the interfacial tension (s) or by letting the value of contact angle (q) approach 90° which means an intermediate wettability of the rock-fluids system. Most of the previous work done in this area has ignored the effect of contact angle on capillary number by setting the contact angle equal to zero. This is equivalent to assuming strongly water-wet conditions in all the reservoirs ignoring the fact that there are more non water-wet reservoirs than water-wet ones2.
Surfactants have long been considered in the oil industry for oil recovery enhancement. There are mainly two mechanisms associated with the surfactants that aid improved oil recovery: reduction in oil-water interfacial tension and alteration of reservoir rock wettability. Conventional surfactant floods have been widely practiced in industry utilising mainly their ability to reduce oil-water interfacial tension.
For significant enhancements in oil recovery by the interfacial tension reduction mechanism, four to five orders of magnitude reduction in interfacial tension is required. Surfactants capable of providing such large reductions in interfacial tension are expensive and are required in large quantities, resulting in uneconomical field applications.
This is the major reason behind the continuous decline and subsequent extinction of chemically improved oil recovery projects in the world over the past few decades3. Therefore, the objective of this study is to investigate the other largely ignored beneficial aspect of surfactants, namely alteration of wettability, on improved oil recovery in reservoir rock-fluid systems by measuring oil-water relative permeabilities.
For this purpose, coreflooding experiments have been conducted using Berea sandstone cores, stocktank crude oil from Yates reservoir (West Texas, US) and two types of low cost surfactants, one nonionic ethoxy alcohol and the other anionic ethoxy sulphate, at ambient conditions of temperature and pressure. Synthetic brine matching the Yates reservoir brine composition is used in all corefloods.
A coreflood simulator has been used to calculate oil-water relative permeabilities by history matching the oil recovery and pressure drop experimental data obtained from corefloods. The variations in oil-water relative permeabilities have been then utilized to interpret the surfactant-induced wettability alterations and their effects on oil recovery.

EXPERIMENTAL DETAILS
Design and Assembly of Coreflood Apparatus
Fig 1 shows the coreflood apparatus assembled for use in this study. Part A in the picture is the pump used for injecting different fluids into the core. Part B is the pressure gauge that is used to measure the pressure drop across the core during the flood. Part C is the core holder inside which the core is placed. Part D is the measuring cylinder used to measure the flow rates at the outlet. The dead volumes of all the flow lines were measured and accounted for in all the material balance calculations.

Experimental Procedure
Coreflooding experiments were conducted on Yates reservoir fluids in Berea rocks using two types of surfactants (nonionic and anionic) in varying concentrations. The first set of experiments with nonionic surfactant was conducted on a shorter Berea core (1.5 inch in diameter and 3.0 inch in length). A larger Berea core (2.0 inch in diameter and 6.0 inch in length) was used in the latter set of experiments with anionic surfactant to increase the pore volume. The absolute permeability of these Berea cores was about 400 and and porosity was 21 per cent. Rapaport and Leas4 criterion (LVµ = 1.0, where L is the system length in cm, V is the flow rate per unit cross sectional area in cm/min and u is the displacing phase viscosity in cP) was used to calculate the stable volumetric flow rates for use in all the experiments, since it was essential to conduct all the floods in a flow regime where the recovery is independent of injection flow rate. Otherwise, the results will be clouded with uncertainty of the extent of flow rate effects in rockfluids system with changing interfacial tension and wettability.
At first the core was saturated with brine to determine its porosity and the absolute permeability. Then oil was injected at a flow rate of 2.0 cc/min for two pore volumes and 6.0 cc/min for five pore volumes to bring the core to initial water saturation (Swi). Then a waterflood was conducted. The core was then brought back to Swi by flooding with Yates crude oil.
Then the effect of surfactant concentration on oil recovery was studied by carrying out several floods with synthetic brine containing different concentrations (500, 1,500, 3,500 and 5,000 ppm) of nonionic (ethoxy alcohol) and anionic (ethoxy sulphate) surfactants. During each of these floods, pressure drop and oil and brine productions were continuously monitored.
Calculation of Oil-Water Relative Permeabilities from Coreflood Data.
A coreflood simulator was used to simulate the experimentally measured pressure-drop and recovery histories using the oil-water relative permeabilities described by,

kro  =  krom (1 - S) eo  (2)
krw = krwm Sew   (3)

S =  Sw – Swi   
  ––––––––– 
 1 – Sor – Swi   (4)
where Sw is the brine saturation, Sw is the irreducible brine saturation, Swm, is the residual oil saturation, Sor, is the maximum brine saturation or (1-Sor), kro is the relative permeability to oil, krw is the relative permeability to brine, krom is the relative permeability to oil at Swi, krwm is the relative permeability to brine at Sor and eo and ew are the Corey exponents.
This semi-analytical model uses the fractional flow theory of Buckley and Leverette5 to calculate the recovery and pressure drop data at any given time after the start of the displacement and then calculates the relative permeabilities by minimising the sum-of-squares of weighted deviations of the experimental pressure and  production histories from the calculated values.
As an example case, the history match of recovery and pressure drop data obtained from the simulator during oilwater relative permeability calculations for waterflood of Yates crude oil in Berea rock with no surfactant are shown in Fig 2. An excellent match between the experimental and the predicted coreflood data can be seen in Fig 2. Similar results were observed from the simulator for all the corefloods conducted. The relative permeability data obtained from the ,simulator was then interpreted using Craig’s rules-of-thumb6, 7, 8 to discern the wettability alterations induced by the surfactant solutions of varying concentrations. Craig’s rules-of-thumb, used in this study to characterise wettability, are summarized in Table 1.
The fractional water flow was computed using the computed oil-water relative permeabilities from the simulator with the simple conventional approach, given by:

fw =  1   
  –––––––––––––– 
 1 + (kroµw /krwµo  ) (5)
Where Fw is the fractional water flow, µw and µo are the viscosities of displacing and displaced phases, respectively.

RESULTS AND DISCUSSION
The summary of experimental and simulator results for waterflood of Yates crude oil in Berea rock at various nonionic and anionic surfactant concentrations is shown in Tables 2 and 3, respectively. These results indicate only minor adjustments in end-point water relative permeabilities in the simulator to obtain acceptable history match during coreflood simulation.
The effect of nonionic surfactant concentration on oil recovery, pressure drop, relative permeability ratios (krw/kro) and fractional water flow is shown in Figs 3(a), 3(b), 3(c) and 3(d), respectively. The effect of anionic surfactant concentration on oil recovery, pressure drop, relative permeability ratios (krw/kro) and fractional water flow is shown in Figs 4(a), 4(b), 4(c) and 4(d), respectively.
From these experimental results, the following important observations are made.
• The oil recovery is gradually increased from 56 per cent to 94 per cent as the nonionic surfactant concentration is increased from 0 ppm to 5,000 ppm. The optimum surfactant concentration for this system is 3,500 ppm as the oil recovery remains unchanged above that value (Fig 3(a)). In the case of anionic surfactant, the oil recovery is gradually increased from 52 per cent at 0 ppm to 78 per cent at 3,500 ppm surfactant concentration (Fig 4(a)).
• The pressure drop is gradually decreased with nonionic surfactant concentration, except at 1,500 ppm, due to the tendency of this surfactant not to form oil-water emulsions (Fig 3(b)). Unlike with nonanionic surfactant, a gradual increase of pressure drop with anionic surfactant concentration can be seen, except at 500 ppm (Fig 4(b)). This is due to the emulsion forming tendency of this surfactant at concentrations above 500 ppm. The formation of oil-water emulsion was so strong at 3,500 ppm that waterflood at 5,000 ppm anionic surfactant concentration was not conducted. Fig 5 shows the picture of strong oil-water emulsion observed in the collected production stream at 3,500 ppm anionic surfactant concentration.
•  There is a significant change in the connate water saturation and it increases from 40 per cent to 65 per cent in these floods as the nonionic surfactant concentration is increased from 0 ppm to 5,000 ppm (Table 2). In the case of anionic surfactant, the connate water saturation is almost unchanged and is around 30 per cent for all the floods (Table 3).
•  The end-point relative permeability to water at residual oil saturation remains almost the same and is always less than 30 per cent in these floods at different nonionic surfactant concentrations (Table 2). Even in the case of anionic surfactant, the end-point water permeability is less than 30 per cent for all the floods (Table 3).
•  There is a significant shift to the right in the water saturation at crossover-point from 60 per cent to 90 per cent in these floods as the nonionic surfactant concentration is increased from 0 ppm to 5,000 ppm (at krw/kro= 1.0 from Fig 3(c)). In the case of anionic surfactant, the water saturation at crossover-point varies from 55 per cent to 75 per cent for all the floods (at krw/kro = 1.0 from Fig 4(c)).
•  The end-point relative permeability to oil at connate water saturation decreases from 97 per cent to 88 per cent in these floods as the nonionic surfactant concentration is increased from 0 ppm to 5,000 ppm (Table 2). In the case of anionic surfactant, at first, the end-point relative permeability to oil at connate water saturation decreases from 97 per cent to 86 per cent at 500 ppm surfactant concentration, then further decreases to 45 per cent at 1,500 ppm surfactant concentration (Table 3). A further increase in surfactant concentration to 3,500 ppm does not influence the end-point oil relative permeability.
  This significant drop in the end-point oil permeabilities is attributed mainly to the formation of emulsions at these concentrations above 500 ppm, which were clearly visible in the collected production streams.
•  The fractional water flow curves for both nonionic and surfactants are gradually shifting to right as the surfactant concentration is increased from 0 ppm (Figs 3(d) and 4(d)). This kind of rightward shift in fractional water flow curves indicates a significant decrease in fractional water cut in the production due to these surfactants.
From the above observations, it is evident that the Yates fluids – Berea rock system is initially water-wet, using the Craig’s rules-of-thumb on oil-water relative permeability characteristics shown in Table 1. The very high oil recoveries observed in this rock-fluids system in the presence of both the surfactants indicate that the system is neither oil-wet nor water-wet at higher surfactant concentrations. Hence Craig’s rules-of-thumb are not applicable to infer the surfactant-induced wettability shifts in the systems used, as these rules are defined only to distinguish between strongly water-wet and oil-wet systems. Hence, the ratios of oil-water relative permeabilities (krw/kro) are used to interpret the wettability alterations caused by the surfactants. From the plots of relative permeability ratios (krw/kro) against water saturation for both nonionic and anionic surfactants (Figs 3(c) and 4(c)), it can be clearly seen that the relative permeability ratio curves are gradually shifting to the right as the surfactant concentrations are increased. This type of relative shift in the relative permeability ratio curves indicates the development of a mixed wettability condition by both the surfactants from an initial water-wet state9, l0, The very high oil recoveries at higher surfactant concentrations (Figs 3(a) and 4(a)) also indicate the development of a Salathiel type mixedwettability condition ll due to these surfactants.
The relative shifts in the relative permeability curves with anionic surfactant (Fig 4(c)) are not as large as those observed in the case of nonionic surfactant (Fig 3(c)), which did not form an oil-water emulsion. This smaller shift in relative permeability ratio curves for the anionic surfactant is caused by the competing effects of (1) formation of oil-water emulsion (which shifts the relative permeability ratio curves to the left), and (2) the development of mixed-wettability (which shifts the relative permeability ratio curves to the right). The final overall rightward shift (Fig 4(c)) clearly demonstrates that the wettability alteration to mixed-wet far outweighs the effect of oil-water emulsions.
The oil-water interfacial tensions measured between Yates crude oil and Yates synthetic brine at different surfactant concentrations of both nonionic and anionic surfactants from Ayirala et a1.12 are plotted in Fig 6. From Fig 6, it can be seen that the interfacial tension reduction observed with both the surfactants is only of two orders of magnitude (from 23.36 mN/m at 0 ppm to 0.13 mN/m at 3,500 ppm for nonionic surfactant and from 23.36 mN/m at 0 ppm to 0.48 mN/m at 3,500 ppm for anionic surfactant). For significant improvements in oil recovery as observed in this study, four to five orders of magnitude reduction in the interfacial tension is required. Hence, the wettability alteration to mixed-wet by the surfactants is the principal mechanism responsible for significant oil recovery improvements obtained. Thus this study has experimentally demonstrated beneficial effects of low cost wettabilityaltering capable surfactants for significant oil recovery enhancements, and hence opens up a new approach for the use of surface-active chemicals to alter wettability for economic field applications.
CONCLUSIONS
•  Both the nonionic and anionic surfactants used in this study have altered wettability of Berea rock-Yates reservoir fluids system in addition to reducing the oilwater interfacial tension.
•  The optimum surfactant concentration was found to be 3,500 ppm for both types of surfactants used for oil recovery enhancement.
•  Higher oil recoveries and the absence of oil-water emulsion formation were observed with the nonionic surfactant, while relatively lower oil recoveries and oil/water emulsions were observed with the anionic surfactant.
•  Nonionic surfactant at 3,500 ppm concentration is recommended as the most favorable surfactant type and concentration for the rock-fluids system studied.
•  The gradual shifts to the right in relative permeability ratio curves indicate the capability of both these surfactants to develop a special kind of heterogeneous wettability known as “mixed-wettability” for potential oil recovery enhancements.
•  Only two orders of magnitude reduction in interfacial tension obtained with both the surfactants suggests a relatively lower effect of interfacial tension reduction mechanism on improved oil recoveries.
•  Wettability alteration to mixed-wet is the principal mechanism responsible for significant oil recovery improvements observed.
•  This study opens up a new approach for the use of low cost surface-active chemicals at low concentrations to alter wettability, instead of the conventional approach of using them for interfacial tension reduction for economic field applications.