Copyright 2006, Society of Petroleum Engineers This paper was prepared for presentation at the 2006 SPE/DOE Symposium on Improved Oil Recovery held in Tulsa, Oklahoma, U.S.A., 2226 April 2006. 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, as presented, have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material, as presented, does not necessarily reflect any position of the Society of Petroleum Engineers, its officers, or members. Papers presented at SPE meetings are subject to publication review by Editorial Committees of the Society of Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes 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 where and by whom the paper was presented. Write Librarian, SPE, P.O. Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.

Abstract We propose an improved procedure for measuring acid numbers and illustrate the significance of the results by correlating with oil/brine interfacial properties.

Introduction Chemical methods of improved oil recovery are not equally effective in all reservoirs. An important factor that can influence a project's success is crude oil composition. Since crude oils are complex mixtures, evaluation of oil composition in a way that is meaningful with respect to specific chemical recovery processes can present many problems. In particular, there is a need for improvements in acid number (AN), also known as total acid number (TAN) measurements.

Acid numbers (AN) are important in evaluating crude oils for alkaline and surfactant processes, but in order to be useful, measurements must be comparable from one laboratory to another and must also capture chemically meaningful information about the crude oil. Standardization (e.g., the current ASTM recommended procedure1) should assist with the first requirement, that different labs be able to reproduce the AN value within some reasonable tolerance. Standardization does not, however, ensure that the measurement captures information about a crude oil that can be used to predict its interactions in chemical recovery processes.

Acid number measurements attempt to characterize an oil with respect to concentration of strong and weak acids by means of non-aqueous potentiometric titration. The standard procedure1 is designed to measure ANs in the range of 0.05 to 250 mg KOH/g oil. Stock tank samples of crude oil usually have ANs that are at the low end of this range; strong acids are not encountered. Thus the sensitivity of the ASTM method is barely adequate for many samples of interest. According to the ASTM procedure, 20 g of oil should be used if AN is less than 1 mg KOH/g oil. Unfortunately, high quality samples of crude oil are expensive to obtain and the quantity is very

limited. Using 20 g for AN measurement would often preclude making any other measurements. The usefulness of AN data is greatly increased if it forms part of a matrix of information that includes, at a minimum, base number (BN), SARA fraction data, and information about asphaltene stability. There are few, if any, interfacial phenomena that correlate exclusively to AN.

Basic constituents of an oil can also be assessed by non-aqueous potentiometric titration, but end-points are often more difficult to detect because the organic bases that occur in crude oils have a wide spread of dissociation constants. More than a decade ago, Dubey and Doe2 published recommendations for improved base number measurements by adding a known amount of quinoline to force a readily detectible titration end-point. Base numbers measured using spiked oil samples were significantly higher than those measured by the ASTM method and the higher base numbers were shown to correlate, together with AN for the same oils, with observations of wetting reversal on silica surfaces. A similar procedure was shown to improve the precision of AN titrations using stearic acid as the spiking agent for routine AN measurements.3 Precipitated material was observed for some crude oils in the standard solvent (50% toluene, 49.5% isopropanol or IPA, and 0.5% water). Stearic acid and o-nitrophenol were used as spiking agents by Zheng and Powers.4

No precipitation was reported in the base number solvent, methyl isobutyl ketone (MIBK),3 which has been used as a solvent for both acidic and basic titrations.4,5 Substitution of tetrabutyl ammonium hydroxide (TBAOH) for KOH in the titrating solution4,6 and MIBK for IPA as the titration solvent4 have been reported. A variety of electrodes have been tested to overcome problems with slow electrode response times in the non-aqueous environment.

In this work we have used stearic acid as a spiking agent and varied solvents, titrants, and electrodes to optimize AN measurement.

Experimental Materials and Methods Crude oil samples. Acid numbers have been measured for more than 250 crude oil samples. Table 1 summarizes selected properties of the xx oil samples used as examples in this paper. For each titration, a sample of 0.5 to 1.0 g of crude oil was used.

SPE 99884

Acid Number Measurements Revisited T. Fan and J.S. Buckley, New Mexico Tech.

2 SPE 99884

Table 1. Oil Properties API

gravity resins n-C6 asph AN BN

() (cP) (wt%) (wt%) mg KOH/g oil Mars-Pink 16.6 481 25.8 6.0 3.92 2.3o MY4-02 28.1 21.6 9.9 1.1 0.22 1.23 SQ-95 37.2 5.8 13.9 2.6 0.17 0.62

Solutions. Solvents. The solvent for most titrations was a mixture of

50% toluene (HPLC grade), 49.4% IPA (HPLC grade) and 0.06% deionized distilled water (DDW). MIBK was tested as a possible alternative solvent with improved solubilization of crude oil components.

A sample of crude oil was dissolved in 50 to 100 ml of the solvent. More dilute solutions were prepared for heavy oils whereas for light oils, more concentrated solutions were used.

Spiking solution. A solution was prepared of 0.02 M stearic acid (Aldrich, 98+%) in the solvent described above. Each AN measurement was spiked with 1 ml of this solution.

Titrant. Oil samples were titrated either with alcoholic KOH (Aldrich, 99.99%) diluted to 0.05M in IPA or with TBAOH (Aldrich, 1M solution in methanol) diluted to 0.05M with ethanol. Titrants were calibrated with a solution of potassium hydrogen phthalate (KHP) (Aldrich, ACS primary standard) at a concentration of 0.002M in DDW.

Equipment. Two electrode systems were tested. The first was an Orion 8101 glass pH electrode and an Orion 8005 reference electrode with saturated KCl in DDW as the reference electrolyte. The second was an Orion 81-02 combination pH electrode with saturated Li Cl in ethanol as the reference electrolyte. Comparable results were obtained with both electrode systems.

A Brinkmann model 350 buret/dispenser was used to deliver titrant at rates from 0.1 to 0.5 ml/min. Electrode potentials (EMF in mV) were monitored with an Orion model 520 pH meter and were recorded by a computerized data acquisition system.

Acid number calculation. Acid number can be calculated from titration endpoints and equation (1).

( )

WMWMVV

AN bi= (1)

where AN is acid number (mg KOH/g oil), Vi is the volume of titrant at the sample inflection point (ml), Vb is the volume of titrant at the blank inflection point (ml), M is molar concentration of KOH titrant (mol/L), MW is the molecular weight of KOH (56.1 g/mol), and W is the amount of the oil sample (g).

Results and Discussion Detecting the inflection point. If the ASTM method1 was followed, only one of 40 titrations of different crude oil samples had a clear inflection point. The titration of SQ-95 shown in Fig. 1 is typical.

-600

-400

-200

0

0 0.5 1 1.5

volume of KOH titrant (ml)

EMF

(mV)

SQ-95

Fig. 1. Potentiometric titration of SQ-95 crude oil shows no clear inflection point.

Addition of stearic acid forces a sharp inflection point. In Fig. 2, titrations of a spiked mixture with SQ-95 are compared to a blank titration of the spiked mixture without crude oil. The AN can be obtained from the difference between the endpoints with and without crude oil. The value of the AN thus obtained for SQ-95 is 0.2 mg KOH/g oil.

-600

-500

-400

-300

-200

-100

0

0 0.2 0.4 0.6 0.8 1

volume of KOH titrant (ml)

EMF

(mV)

stearic acid in solvent + SQ-95

stearic acid in solvent

Fig. 2. Potentiometric titration of SQ-95 crude oil spiked with steric acid compared to a blank titration of stearic acid without crude oil. Arrows mark approximate titration endpoints.

An alternative estimation technique is described in ASTM D664 for samples without a clear inflection point.1 It involves titrating standard solutions to determine fixed endpoints. The alternative method gives a much higher value of 0.7 mg KOH/g oil for this oil. Some further investigation of the effects of adding stearic acid was undertaken to ensure that the spiked solution results are not affected by specific interactions between stearic acid and crude oil components.

Figure 3 shows tests in which a constant amount of stearic acid was added to titrations of varying amounts of two quite

SPE 99884 3

different crude oils (see Table 1). The value of Vi varies linearly with oil sample size for both SQ-95 (AN = 0.2 mg KOH/g oil) and Mars-Pink (AN = 3.9 mg KOH/g oil). The slopes are proportional to the AN of each oil and both linear trends extrapolate to the value of Vb in the absence of oil.

0

1

2

3

4

5

0 0.5 1 1.5 2 2.5

weight of oil sample (g)

volu

me

of ti

trant

Vi (

ml) Mars-Pink

SQ-95

Fig. 3. The volume of titrant varies linearly with the amount of oil titrated. There is no evidence of specific interactions between stearic acid and components of either Mars-Pink or SQ-95. (Note: titrant is TBAOH in ethanol.)

Blank titrations of the mixed solvent are difficult to reproduce. When the blank is spiked with stearic acid, however, the results are more consistent. Figure 4 shows, the effect of solvent when a constant amount of stearic acid (0.022 mmol) is diluted to different concentrations. The titrant reacted with stearic acid alone, obtained by extrapolation to zero solvent, corresponds to 0.21 mmol. Thus spiking improves the accuracy of both the crude oil and blank endpoint determinations. The AN measured with spiked samples is much more accurate than that estimated from unspiked samples or from predetermined endpoints.

Titrants and solvents.

Comparison of KOH and TBAOH titrants. The ASTM standard titrant, alcoholic KOH, must be freshly made and standardized frequently. Figure 4a compares titrations of a known amount of stearic acid by freshly made KOH titrant and the same titrant three weeks later. The results with the aged titrant are clearly unacceptable. If, however, the alcoholic KOH is replaced by an alcoholic solution of TBAOH, the titrant is much more stable. Results with the freshly made and a three-month-old solution are comparable (Fig. 4b).

-600

-500

-400

-300

-200

-100

0

100

200

0 0.2 0.4 0.6 0.8

volume of titrant (ml)

EMF

(mv)

KOH (freshly prepared)

KOH (aged 3 weeks)

(a) titrant is KOH in IPA

-600

-500

-400

-300

-200

-100

0

100

200

0 0.2 0.4 0.6 0.8

volume of titrant (ml)

EMF

(mv)

TBAOH (freshly prepared)

TBAOH (aged 3 months)

(b) titrant is TBAOH in ethanol

Fig. 4. Effect of titrant age on titrations of stearic acid. TBAOH in ethanol is much more stable than KOH in IPA.

Acid numbers, measured in triplicate, are compared in Table 2 for these two titrants. Values are the same, within the accuracy of the measurements and standard deviations are comparable. Where there is a difference, TBAOH is more reproducible. Considering the improved stability shown by solutions of TBAOH, we have adopted it as our standard titrant.

4 SPE 99884

Table 2. Acid Number Measured with Different Titrants

Crude oil Titrant =KOH Ttitrant =TBAOH AN (mg KOH/g oil) AN (mg KOH/g oil)

Mars-Pink 3.92 0.05 3.95 0.02 SQ-95 0.17 0.02 0.22 0.02

Alternative solvents. MIBK, the solvent used in base

number titrations, was tested as an alternative to the standard mixture of toluene/IPA/DDW. As noted previously, no precipitation was reported in MIBK solutions of crude oils. Figure 5 shows the results of titrations for MY4-02 crude oil solutions spiked with stearic acid in standard solvent and MIBK. There is a clear endpoint in the standard solvent, but not in MIBK. Other oil samples showed similar trends. Even blank samples of stearic acid alone failed to produce clear endpoints. We conclude that the standard solvent should not be replaced by MIBK for AN measurements.

-700

-600

-500

-400

-300

-200

-100

0

0 0.5 1 1.5 2 2.5

volume of TBAOH titrant (ml)

EMF

(mV)

MY4-02 in standard solvent mixture

MY4-02 in MIBK

Fig. 5. Titration of MY4-02 crude oil, spiked with stearic acid, and dissolved either in the standard mixture of toluene/IPA/DDW or in MIBK. The standard solvent mixture produces much clearer endpoints than does MIBK.

Electrode response. Electrode responses change upon contact with most of the crude oil samples that we have tested. This is true for both two electrode and combination electrode systems. Figure 6a illustrates this effect for the combination electrode by comparing titrations of a blank solution spiked with stearic acid, before and after contact with a spiked solution of SQ-95 crude oil. Titration of the blank before contact with crude oil produces a negative, physically meaningless value for AN. Only if the blank is measured after contact with the crude oil, so that the electrode response is similarly affected for both measurements, can a realistic measure of AN be obtained. Additional comparisons (Fig. 6b) show the altered electrode response to be stable over many measurements. These included. Thorough cleaning restores the original electrode response.

-600

-500

-400

-300

-200

-100

0

0 0.2 0.4 0.6 0.8 1

volume of TBAOH titrant (ml)

EMF

(mV)

1 - blank (before)2 - SQ-953 - blank (after)

(a) A series of three measurements beginning with clean electrodes: 1) a spiked blank solution, followed by 2) SQ-95 crude oil, after which 3) a spiked blank solution was remeasured.

-600

-500

-400

-300

-200

-100

0

0 0.2 0.4 0.6 0.8 1 1.2

volume of titrant (ml)

EMV

(mv)

Run #1 - beforeRun #2 - before

Run #7 - after

Run #12 - afterRun #17 - after

(b) Measurements of stearic acid spiked blank solutions during at intervals during a series of 17 titrations. Two spiked blank solutions were measured using clean electrodes before contact of the electrode with crude oil. A spiked blank solution was remeasured at intervals between measurements of SQ-95 and Mars-Pink ANs.

Fig. 6. Exposure to crude oil immediately changes the electrode response. Additional contact with crude oil does not appear to shift electrode response further. Blank measurements should be made after crude oil titrations to cancel out changes in electrode response.

SPE 99884 5

Acid numbers measured by improved technique. Figure 7 shows AN measured by the improved techniques described above as a function of oil gravity. Nine percent had AN that were below the limit of detection (< 0.01 mg KOH/g oil). The largest number of samples (61%) was between 0.1 and 1; 17% had AN values greater than 1 mg KOH/g oil. In general, lower gravity oils had higher values of AN, although the trend is very scattered. A similarly weak relationships can be found between log(AN) and the amounts of resins and asphaltenes. The AN increases with increasing amounts of polar crude oil components. Crude oil properties are from CO-Wet, a database of crude oil properties that relate to their wetting tendencies, described previously.7 The method used to measure SARA fractions, including resins and nC6 asphaltenes, has also been published previously.8

0.01

0.1

1

10

0 20 40 60

API Gravity ()

AN

(mg

KO

H/g

oil)

Fig. 7. Overall, AN values of 255 crude oils show a decrease with increasing API gravity.

Interfacial properties correlated with AN. It is widely believed that ANs are related to low interfacial tensions (IFT) between some crude oils and displacing brine at sufficiently alkaline conditions9 although little direct evidence exists either in laboratory reports10 or field test results11 demonstrating a direct connection between AN and IFT. The reason often suggested is that perhaps a small fraction of the acids in a crude oil may play a large role in IFT reduction, whereas others may be inactive. Examination of the ASTM method for measuring AN suggests another alternative: that routine measurements of AN can be inaccurate, masking any relationships that might exist between AN and interfacial properties.

Buckley and Fan12 recently reported that IFT can be correlated to oil properties, including AN. The correlating parameters include amount of asphaltene, AN, BN, and viscosity, with different suites of variables and dependences for different brine strength and pH conditions.

In alkaline/surfactant processes where a synthetic surfactant is used to increase oil recovery, phase behavior and oil/brine IFT depend on many factors, including the relative amounts of water (containing synthetic surfactant) and oil (containing natural surfactants). Zhang and Hirasaki13 observed that optimal salinities (at which IFT is minimized)

could be correlated for a range of water/oil ratios from 1 to 10 by the ratio of natural to synthetic surfactant. Acid number measured in our laboratory was used in this correlation as the basis for calculation of the amount of natural surfactant.

While there is more work to be done, these studies suggest that accurate measurements of AN, together with other crude oil properties of similar accuracy, can be used to evaluate oils with respect to their interfacial properties.

Conclusions Both crude oil and blank solutions must be spiked with an

oil-soluble acid, such as stearic acid, to obtain accurate and repeatable titration endpoints in non-aqueous potentiometric titrations to determine AN of crude oils. Spiking these solutions permits accurate determinations of AN using smaller samples that those required for the standard ASTM procedure.

Substitution of tetrabutyl ammonium hydroxide for potassium hydroxide as titrant confers additional stability on the titration results.

Methyl isobutyl ketone is not an appropriate substitute for the standard solvent mixture of toluene/isopropyl alcohol/water.

Blank measurements must be made after initial crude oil titrations to compensate for changes in electrode properties that occur when the electrodes come in contact with crude oil.

Carefully measured values of AN are reproducible and correlate with interfacial properties of crude oils.

Acknowledgments This work was supported by the NPTO office of the US DOE under contract DE-FC26-01BC15164, by the State of New Mexico, by industrial sponsors including BP, ChevronTexaco, and Total. The authors acknowledge the contributions of Stphanie Monsterleet who first explored adaptations of the ASTM procedure by spiking crude oils with stearic acid and helpful discussions with George Hirasaki.

References 1. ASTM Standard Test Method D664-01: Standard Test Method

for Acid Number of Petroleum Products by Potentiometric Titration. Annual Book of ASTM Standards, Sect. 5, Am. Soc. Testing Materials, Philadelphia, (2001), P. 273-279.

2. Dubey, S.T. and Doe, P.H.: Base Number and Wetting Properties of Crude Oils, SPERE (Aug. 1993) 8, 195-200.

3. Monsterleet, S. and Buckley, J.S.: Standard Measurements of Acid and Base Numbers, PRRC 96-08 (1996).

4. Zheng, J.Z. and Powers, S.E.: Identifying the Effect of Polar Constituents in Coal-Derived NAPLs on Interfacial Tension, Env. Sci. & Tech. (2003) 37, 3090-3094.

5. Bruss, D.B. and Wyld, G.E.A.: Methyl Isobutyl Ketone as Wide-Range Solvent for Titration of Acid Mixtures and Nitrogen Bases, Anal. Chem. (1957) 29, 232-235.

6. Mediaas, H., Grande K.V., Hustad, B.M., Rasch, A., Ruesltten, H.G., and Vindstad, J.E.: The Acid-IER Method a Method for Selective Isolation of Carboxylic Acids from Crude Oils and Other Organic Solvents, paper SPE 80404 presented at the 2003 Internat, Symp, on Oilfield Scale, Aberdeen, Jan. 29-30.

7. Buckley, J.S. and Wang, J.X.: Crude Oil and Asphaltene Characterization for Prediction of Wetting Alteration, J. Pet. Sci. Eng. (2002) 33, 195-202.

6 SPE 99884

8. Fan, T. and Buckley, J.S.: Rapid and Accurate SARA Analysis of Medium Gravity Crude Oils, Energy & Fuels (2002) 16, 1571-1575.

9. Lake, L.W.: Enhanced Oil Recovery, Prentice Hall, Englewood Cliffs, NJ, (1989) pg 436.

10. Ramakrishnan, T.S. and Wasan, D.T.: A Model for Interfacial Activity of Acidic Crude Oil/Caustic Systems for Alkaline Flooding, SPEJ (Aug. 1983) 602-612.

11. Lorenz, P.B. and Peru, D.A.: Guidelines Help Select Reservoirs for NaHCO3 EOR, OGJ (Sept. 11, 1989) 53-57.

12. Buckley, J.S. and Fan, T.: Crude Oil/Brine Interfacial Tensions, paper SCA 2005-01, presented at the 2005 SCA Symposium, Toronto, 21-25 Aug.

13. Zhang, L.D. and Hirasaki, G.J.: Wettability Alteration and Spontaneous Imbibition in Oil-Wet Carbonate Formations, J. Pet. Sci. Eng. (2006) in press.