- Email: [email protected]
Chapter 4. Interpretation of Chemical Analyses of Oilfield Waters
Chapter 4.
INTERPRETATION OF CHEMICAL ANALYSES OF OILFIELD WATERS
Water analyses may be used to identify the source of a water. In the oilfield one of the prime uses of these analyses is to determine the source of extraneous water in an oil well, so that casing can be set and cemented to prevent such water from flooding the oil or gas horizons. In some wells a leak may develop in the casing or cement, and water analyses are used to identify the water-bearing horizon so that the leaking area can be repaired. With the present emphasis on water pollution prevention, it is very important to locate the source of a polluting brine, so that remedial action can be taken. Comparisons of water-analysis data are tedious and time-consuming; therefore, graphical methods are commonly used for positive, rapid identification. A number of systems have been developed, all of which have some merit. Calculating probable compounds The hypothetical combinations of dissolved constituents found in waters are commonly calculated by combining the positive and negative radicals in the following order: calcium magnesium sodium potassium
bicarbonate sulfate chloride nitrate
Calcium is combined with bicarbonate, and if more calcium is available than that consumed by bicarbonate, it is combined with sulfate, chloride, and nitrate until exhausted. Conversely, any excess bicarbonate is combined with magnesium, sodium, and potassium until consumed. Other radicals can and should be added for most petroleum reservoir waters. These include lithium, strontium, barium, iron, borate, phosphate, bromide, and iodide. They can be grouped in the appropriate column and then in the calculations each positive and negative radical is totally combined, the next following radical is combined until both the cations and anions are exhausted. If the analysis is correct, the cations and anions will be present in approximately equivalent amounts. To calculate the hypothetical combinations, the reacting values of the positive and negative radicals or ions are calculated as follows: reacting
INTERPRETATION OF CHEMICAL ANALYSES
126 TABLE 4.1 Reaction coefficients Cation
Anion
Calcium Magnesium Iron Sodium
0.0499 0.0823 0.0358 0.0435
bicarbonate sulfate chloride
Cation (mg/l)
RV
Anion (mg/l)
Ca 4,000x Mg 3,000x Fe 100 x Na 9,400 x
199.6 246.8 3.6 408.9
HC03
0.0164 0.0208 0.0282
TABLE 4.11 Reacting values (RV)
0.0499 = 0.0823 = 0.0358 = 0.0435 =
so4
C1
RV
500 x 0.0164 = 8.2 200 x 0.0208 = 4.2 30,000 x 0.0282 = 846.3
858.9
858.7
TABLE 4.111 Reacting value distribution Ca Ca Ca Mg Fe Na
as as as as as as
calcium bicarbonate calcium sulfate calcium chloride magnesium chloride iron chloride sodium chloride
8.2 4.2 187.2 246.8 3.6 408.9 858.9
values (RV) or equivalents per million (epm) = mg/l of ion x valence of ion/ molecular weight of ion. The term valence of ion/molecular weight of ion is called “reaction coefficient” and the positive and negative ions have values as shown in Table 4.1. Table 4.11 indicates how the results of a water analysis are converted t o reacting values. The reacting values are a measure of the cations and anions dissolved in the water. The 4,000 mg/l of calcium with a reacting value of 199.6 can combine with all the bicarbonate, all the sulfate, and 187.2 epm of the chloride. Magnesium will combine with 246.8,iron with 3.6, and sodium with 408.9 epm of chloride. Thus the reacting values can be considered to be distributed as shown in Table 4.111.
DETERMINING A SOUGHT COMPOUND
127
TABLE 4.IV Combination factors Reaction values given
Compound sought
Combination factor
Ca o r C 0 3 Ca or SO4 Ca or C1 Mg or C 0 3 Mg or SO4 Mg or C1 Fe or C03 Fe o r S 0 4 Fe orC1 Na or C03 Na or SO4 Na or C1
CaC03 CaS04 CaClz MgCO3 MgS04 MgClz FeC03 FeS04 Fa12 Na~C03 Naz SO4 NaCl
50.1 68.1 55.5 42.2 60.1 47.6 57.8 76.0 63.4 53.1 71.0 58.4
TABLE 4.V Hypothetical combinations Ca(HC03 ) to CaC03 CaS04 CaC12 MgCh FeClz NaCl
8.2 x 50.1 4 . 2 68.1 ~ 187.2 x 55.5 246.8 x 47.6 3.6 x 63.4 858.9 x 58.4
= 411 CaC03* = 286CaS04 = 10,390 CaC12 = 11,748 MgClz = =
228 FeCIz 50,160 NaCl
*In mg/l.
Determining a sought compound It is necessary to multiply the reacting value by a combination factor to determine a hypothetical compound. This factor is necessary to convert the reported radical into the desired compound. For example, the factor for converting Ca to CaCO, is 2.50 and the reaction coefficient for Ca is 0.0499. Therefore, the combination factor to convert the reacting value for Ca to CaCO, is 2.50 + 0.0499 = 50.1. Table 4.IV illustrates some combination factors. The combination factors given in Table 4.IV can be used to calculate the hypothetical combinations shown in Table 4.V, using the analysis shown in Table 4.111.
INTERPRETATION OF CHEMICAL ANALYSES
128
Graphic plots Graphic plots of the reacting values can be made to illustrate the relative amount of each radical present. The graphical presentation is an aid t o rapid identification of a water, and classification as t o its type, and there are several methods that have been developed.
Tickell diagram The Tickell (1921) diagram was developed using a 6-axis system or star diagram. Percentage reaction values of the ions are plotted on the axes. The percentage values are calculated by summing the epm’s of all the ions, dividing the epm of a given ion by the sum of the total epm’s, and multiplying by 100. Na
Ca+Mg
Na
ci (a)
CI RV=49.92%
(b)
Ca+Mg
2-
\
So4
h 9 2 ma / I i tar
Fig. 4.1. Tickell (a) and modified Tickell (b) diagram for Gulf Coast water, sample No.1.
Na
(a)
Ca + Mg
s
Ca+Mg
CI ~ v = 4 9 . 2 9 %
(b)
1i07 ma/ lltar
Fig. 4.2. Tickell (a) and modified Tickell (b) diagram for Anadarko Basin water, sample No. 2.
GRAPHIC PLOTS
129
H
so4 C I RV=49.92 %
(a)
so4 5,708 m e / l i t e r
(b)
Fig. 4.3. Tickell (a) and modified Tickell (b) diagram for Williston Basin water, sample N0.3.
No
Ca+Ma
Co+Mg
$
c i\ (b)
so4 1.769 me / liter
Fig. 4.4. Tickell (a) and modified Tickell (b) diagram for Gulf Coast and Anadarko Basin waters, mixed 1:l. Na
Co+Mg
CI
so4
Na
Ca+Mg
‘7 $.
(b) C I
2870 me / i i t r r
Fig. 4.6. Tickell (a) and modified Tickell (b) diagram for Gulf Coast, Williston, and Anadarko Basin waters, mixed 1 :1 :1.
130
INTERPRETATION OF CHEMICAL ANALYSES
Fig. 4.1.illustrates the Tickell diagram using reaction values in percentage in the diagram on the left, and total reaction values in the diagram on the right. The plots of total reaction values, rather than of percentage reaction values, are often more useful in water identification because the percentage values do not take into account the actual 'ion concentrations. Water differing only in concentrations of dissolved constituents cannot be distinguished. To illustrate differences in patterns for different waters, Fig. 4.1-5 were prepared using the Tickell method. Fig. 4.1 represents a water from the Gulf Coast Basin, taken from the Wilcox formation of Eocene age. Fig. 4.2 is of a sample from the Mer?.mec formation of Mississipian age in the Anadarko Basin. Fig. 4.3 is of sample from a Devonian age formation in the Williston Basin. Fig. 4.4 represents a 1:l mixture of waters of the Gulf Coast and Anadarko Basins, and Fig. 4.5 is a 1:1:1 mixture of all three waters.
REISTLE SYSTEM
Fig. 4.6. Water-analysis interpretation, Reistle system the samples of Fig. 4.1-3.
- sample numbers correspond to
GRAPHIC PLOTS
131
Reistle diagram Reistle (1927) devised a method of plotting water analyses using the ion concentrations as shown in Fig. 4.6. The data are plotted on a vertical diagram, with the cations plotted above the central zero line and the anions below. This type of diagram often is useful in making regional correlations or studying lateral variations in the water of a single formation, because several analyses can be plotted on a large sheet of paper.
St iff diagra m Stiff (1951) plotted the reaction values of the ions on a system of rectangular coordinates as illustrated in Fig. 4.7. The cations are plotted to the left and the anions to the right of a vertical zero line. The end points then are connected by straight lines to form a closed diagram, sometimes called a “butterfly” diagram. To emphasize a constituent that may be a key t o interpretation, the scales may be varied by changing the denominator of the
Fig. 4.7. Water-analysis interpretation, Stiff method - sample numbers correspond to the samples of Fig. 4.1-3.
132
INTERPRETATION OF CHEMICAL ANALYSES
ion fraction usually in multiples of 10. However, when looking at a group of waters all must be plotted on the same scale. Many investigators believe that this is the best method of comparing oilfield water analyses. The method is simple, and nontechnical personnel can be easily trained t o construct the diagrams.
Other methods Several other water identification diagrams have been developed, primarily for use with fresh waters, and they will not be discussed here. The Piper (1953)diagram and the Stiff (1951)diagram were adapted to automatic data processing by Morgan et al. (1966),and Morgan and McNellis (1969).The Piper (1953)diagram uses a multiple trilinear plot t o depict the water analysis, and this quaternary diagram shows the chemical composition of the water in terms of cations and anions. Angino and Morgan (1966)applied the automated Stiff and Piper diagrams to some oilfield brines and obtained good results.
References Angino, E.E. and Morgan, C.O., 1966. Application of pattern analysis t o the classification of oilfield brines. Kans. State Geol. Sum.,Comput. Contrib., No.7, pp.53-56. Morgan, C.O. and McNellis, J.M., 1969. Stiff diagrams of water-quality data programmed for the digital computer. Kuns. State Geol. Sum., Spec. Distrib. Publ., No.43, 27 pp. Morgan, C.O., Dingman, R.J. and McNellis, J.M., 1966. Digital computer methods for water-quality data. Ground Water, 4:35-42. Piper, A.M., 1953. A graphic procedure in the geochemical interpretation of water analyses. US.Geol. Surv. Ground Water Note, No.12, 1 4 pp. Reistle, C.E., 1927. Identification of oilfield waters by chemical analysis. U.S.Bur. Min. Tech. Paper, No.404, 25 pp. Stiff, H.A., 1951. The interpretation of chemical water analysis by means of patterns. J. Pet. Technol., 3:15-17. Tickell, F.G., 1921. A method for graphical interpretation of water analysis. Calif. State Oil Gas Superv., 6:5-11.
INTERPRETATION OF CHEMICAL ANALYSES OF OILFIELD WATERS
Water analyses may be used to identify the source of a water. In the oilfield one of the prime uses of these analyses is to determine the source of extraneous water in an oil well, so that casing can be set and cemented to prevent such water from flooding the oil or gas horizons. In some wells a leak may develop in the casing or cement, and water analyses are used to identify the water-bearing horizon so that the leaking area can be repaired. With the present emphasis on water pollution prevention, it is very important to locate the source of a polluting brine, so that remedial action can be taken. Comparisons of water-analysis data are tedious and time-consuming; therefore, graphical methods are commonly used for positive, rapid identification. A number of systems have been developed, all of which have some merit. Calculating probable compounds The hypothetical combinations of dissolved constituents found in waters are commonly calculated by combining the positive and negative radicals in the following order: calcium magnesium sodium potassium
bicarbonate sulfate chloride nitrate
Calcium is combined with bicarbonate, and if more calcium is available than that consumed by bicarbonate, it is combined with sulfate, chloride, and nitrate until exhausted. Conversely, any excess bicarbonate is combined with magnesium, sodium, and potassium until consumed. Other radicals can and should be added for most petroleum reservoir waters. These include lithium, strontium, barium, iron, borate, phosphate, bromide, and iodide. They can be grouped in the appropriate column and then in the calculations each positive and negative radical is totally combined, the next following radical is combined until both the cations and anions are exhausted. If the analysis is correct, the cations and anions will be present in approximately equivalent amounts. To calculate the hypothetical combinations, the reacting values of the positive and negative radicals or ions are calculated as follows: reacting
INTERPRETATION OF CHEMICAL ANALYSES
126 TABLE 4.1 Reaction coefficients Cation
Anion
Calcium Magnesium Iron Sodium
0.0499 0.0823 0.0358 0.0435
bicarbonate sulfate chloride
Cation (mg/l)
RV
Anion (mg/l)
Ca 4,000x Mg 3,000x Fe 100 x Na 9,400 x
199.6 246.8 3.6 408.9
HC03
0.0164 0.0208 0.0282
TABLE 4.11 Reacting values (RV)
0.0499 = 0.0823 = 0.0358 = 0.0435 =
so4
C1
RV
500 x 0.0164 = 8.2 200 x 0.0208 = 4.2 30,000 x 0.0282 = 846.3
858.9
858.7
TABLE 4.111 Reacting value distribution Ca Ca Ca Mg Fe Na
as as as as as as
calcium bicarbonate calcium sulfate calcium chloride magnesium chloride iron chloride sodium chloride
8.2 4.2 187.2 246.8 3.6 408.9 858.9
values (RV) or equivalents per million (epm) = mg/l of ion x valence of ion/ molecular weight of ion. The term valence of ion/molecular weight of ion is called “reaction coefficient” and the positive and negative ions have values as shown in Table 4.1. Table 4.11 indicates how the results of a water analysis are converted t o reacting values. The reacting values are a measure of the cations and anions dissolved in the water. The 4,000 mg/l of calcium with a reacting value of 199.6 can combine with all the bicarbonate, all the sulfate, and 187.2 epm of the chloride. Magnesium will combine with 246.8,iron with 3.6, and sodium with 408.9 epm of chloride. Thus the reacting values can be considered to be distributed as shown in Table 4.111.
DETERMINING A SOUGHT COMPOUND
127
TABLE 4.IV Combination factors Reaction values given
Compound sought
Combination factor
Ca o r C 0 3 Ca or SO4 Ca or C1 Mg or C 0 3 Mg or SO4 Mg or C1 Fe or C03 Fe o r S 0 4 Fe orC1 Na or C03 Na or SO4 Na or C1
CaC03 CaS04 CaClz MgCO3 MgS04 MgClz FeC03 FeS04 Fa12 Na~C03 Naz SO4 NaCl
50.1 68.1 55.5 42.2 60.1 47.6 57.8 76.0 63.4 53.1 71.0 58.4
TABLE 4.V Hypothetical combinations Ca(HC03 ) to CaC03 CaS04 CaC12 MgCh FeClz NaCl
8.2 x 50.1 4 . 2 68.1 ~ 187.2 x 55.5 246.8 x 47.6 3.6 x 63.4 858.9 x 58.4
= 411 CaC03* = 286CaS04 = 10,390 CaC12 = 11,748 MgClz = =
228 FeCIz 50,160 NaCl
*In mg/l.
Determining a sought compound It is necessary to multiply the reacting value by a combination factor to determine a hypothetical compound. This factor is necessary to convert the reported radical into the desired compound. For example, the factor for converting Ca to CaCO, is 2.50 and the reaction coefficient for Ca is 0.0499. Therefore, the combination factor to convert the reacting value for Ca to CaCO, is 2.50 + 0.0499 = 50.1. Table 4.IV illustrates some combination factors. The combination factors given in Table 4.IV can be used to calculate the hypothetical combinations shown in Table 4.V, using the analysis shown in Table 4.111.
INTERPRETATION OF CHEMICAL ANALYSES
128
Graphic plots Graphic plots of the reacting values can be made to illustrate the relative amount of each radical present. The graphical presentation is an aid t o rapid identification of a water, and classification as t o its type, and there are several methods that have been developed.
Tickell diagram The Tickell (1921) diagram was developed using a 6-axis system or star diagram. Percentage reaction values of the ions are plotted on the axes. The percentage values are calculated by summing the epm’s of all the ions, dividing the epm of a given ion by the sum of the total epm’s, and multiplying by 100. Na
Ca+Mg
Na
ci (a)
CI RV=49.92%
(b)
Ca+Mg
2-
\
So4
h 9 2 ma / I i tar
Fig. 4.1. Tickell (a) and modified Tickell (b) diagram for Gulf Coast water, sample No.1.
Na
(a)
Ca + Mg
s
Ca+Mg
CI ~ v = 4 9 . 2 9 %
(b)
1i07 ma/ lltar
Fig. 4.2. Tickell (a) and modified Tickell (b) diagram for Anadarko Basin water, sample No. 2.
GRAPHIC PLOTS
129
H
so4 C I RV=49.92 %
(a)
so4 5,708 m e / l i t e r
(b)
Fig. 4.3. Tickell (a) and modified Tickell (b) diagram for Williston Basin water, sample N0.3.
No
Ca+Ma
Co+Mg
$
c i\ (b)
so4 1.769 me / liter
Fig. 4.4. Tickell (a) and modified Tickell (b) diagram for Gulf Coast and Anadarko Basin waters, mixed 1:l. Na
Co+Mg
CI
so4
Na
Ca+Mg
‘7 $.
(b) C I
2870 me / i i t r r
Fig. 4.6. Tickell (a) and modified Tickell (b) diagram for Gulf Coast, Williston, and Anadarko Basin waters, mixed 1 :1 :1.
130
INTERPRETATION OF CHEMICAL ANALYSES
Fig. 4.1.illustrates the Tickell diagram using reaction values in percentage in the diagram on the left, and total reaction values in the diagram on the right. The plots of total reaction values, rather than of percentage reaction values, are often more useful in water identification because the percentage values do not take into account the actual 'ion concentrations. Water differing only in concentrations of dissolved constituents cannot be distinguished. To illustrate differences in patterns for different waters, Fig. 4.1-5 were prepared using the Tickell method. Fig. 4.1 represents a water from the Gulf Coast Basin, taken from the Wilcox formation of Eocene age. Fig. 4.2 is of a sample from the Mer?.mec formation of Mississipian age in the Anadarko Basin. Fig. 4.3 is of sample from a Devonian age formation in the Williston Basin. Fig. 4.4 represents a 1:l mixture of waters of the Gulf Coast and Anadarko Basins, and Fig. 4.5 is a 1:1:1 mixture of all three waters.
REISTLE SYSTEM
Fig. 4.6. Water-analysis interpretation, Reistle system the samples of Fig. 4.1-3.
- sample numbers correspond to
GRAPHIC PLOTS
131
Reistle diagram Reistle (1927) devised a method of plotting water analyses using the ion concentrations as shown in Fig. 4.6. The data are plotted on a vertical diagram, with the cations plotted above the central zero line and the anions below. This type of diagram often is useful in making regional correlations or studying lateral variations in the water of a single formation, because several analyses can be plotted on a large sheet of paper.
St iff diagra m Stiff (1951) plotted the reaction values of the ions on a system of rectangular coordinates as illustrated in Fig. 4.7. The cations are plotted to the left and the anions to the right of a vertical zero line. The end points then are connected by straight lines to form a closed diagram, sometimes called a “butterfly” diagram. To emphasize a constituent that may be a key t o interpretation, the scales may be varied by changing the denominator of the
Fig. 4.7. Water-analysis interpretation, Stiff method - sample numbers correspond to the samples of Fig. 4.1-3.
132
INTERPRETATION OF CHEMICAL ANALYSES
ion fraction usually in multiples of 10. However, when looking at a group of waters all must be plotted on the same scale. Many investigators believe that this is the best method of comparing oilfield water analyses. The method is simple, and nontechnical personnel can be easily trained t o construct the diagrams.
Other methods Several other water identification diagrams have been developed, primarily for use with fresh waters, and they will not be discussed here. The Piper (1953)diagram and the Stiff (1951)diagram were adapted to automatic data processing by Morgan et al. (1966),and Morgan and McNellis (1969).The Piper (1953)diagram uses a multiple trilinear plot t o depict the water analysis, and this quaternary diagram shows the chemical composition of the water in terms of cations and anions. Angino and Morgan (1966)applied the automated Stiff and Piper diagrams to some oilfield brines and obtained good results.
References Angino, E.E. and Morgan, C.O., 1966. Application of pattern analysis t o the classification of oilfield brines. Kans. State Geol. Sum.,Comput. Contrib., No.7, pp.53-56. Morgan, C.O. and McNellis, J.M., 1969. Stiff diagrams of water-quality data programmed for the digital computer. Kuns. State Geol. Sum., Spec. Distrib. Publ., No.43, 27 pp. Morgan, C.O., Dingman, R.J. and McNellis, J.M., 1966. Digital computer methods for water-quality data. Ground Water, 4:35-42. Piper, A.M., 1953. A graphic procedure in the geochemical interpretation of water analyses. US.Geol. Surv. Ground Water Note, No.12, 1 4 pp. Reistle, C.E., 1927. Identification of oilfield waters by chemical analysis. U.S.Bur. Min. Tech. Paper, No.404, 25 pp. Stiff, H.A., 1951. The interpretation of chemical water analysis by means of patterns. J. Pet. Technol., 3:15-17. Tickell, F.G., 1921. A method for graphical interpretation of water analysis. Calif. State Oil Gas Superv., 6:5-11.