A series of standard water samples with a theoretical COD_Cr value of 500 mg・L⁻¹ and Cl⁻ mass concentrations of 100, 500, 1000, 2000, 4000, 8000, 10,000, 15,000 and 20,000 mg・L⁻¹ were prepared as simulated high chloride oilfield wastewater by using standard KC₈H₅O₄ and standard NaCl. According to the classical K₂Cr₂O₇ method, 20 mL of each water sample was transferred into a COD digestion tube, then 0.4 g HgSO₄ and 30 mL AgSO₄-H₂SO₄ were added to the sample solution, and finally the mixture of each water was oxidized with 10 mL of 0.25 mol・L⁻¹ K₂Cr₂O₇ standard solution. After heating reflux for about 2 h, the residual content of K₂Cr₂O₇ was back titrated by corresponding 0.l mol・L⁻¹ [(NH₄)₂Fe(SO₄)₂・6H₂O] standard solution. The obtained COD_Cr results in the prepared water samples are summarized in Table 1.

As can be seen from Table 1, for the simulated high chloride oilfield wastewater samples with a theoretical COD_Cr of 500 mg・L⁻¹ and high level of Cl⁻, when the mass concentration of Cl⁻ was less than 2000 mg・L⁻¹, the relative errors of COD_Cr determined by the classical K₂Cr₂O₇ method were relatively small (< 5.7%), which almost accorded with the national requirements that the relative error of COD_Cr determination should be within ±5% [18] [19] . This is because the addition of 0.4 g HgSO₄ in 20 mL water samples was perfectly able to coordinate 2000 mg・L⁻¹ Cl⁻ in wastewater, and then the interferences of Cl⁻ on COD_Cr determination can be virtually eliminated. However, with the increase of Cl⁻ mass concentration in water samples, the Cl⁻ in water samples could not be completely coordinated by 0.4 g HgSO₄, resulting in the obtained concentration of COD_Cr in water samples increased gradually, and the differences between the experimental values and theoretical ones also increased gradually. All the above results indicated that pseudo COD_Cr was introduced during the determination of COD_Cr in high chloride wastewater by the classical K₂Cr₂O₇ method. As a result, if the interferences of Cl⁻ cannot be effectively eliminated, the obtained inaccurate COD_Cr results will cause wrong expert conclusion for water quality.

A series of standard water samples with a theoretical COD_Cr value of 100 mg・L⁻¹ and Cl⁻ mass concentrations of 100, 500, 1000, 2000, 4000, 8000, 10,000, 15,000 and 20,000 mg・L⁻¹ were prepared as simulated high chloride oilfield wastewater by using standard KC₈H₅O₄ and standard NaCl. According to the classical K₂Cr₂O₇ method, 20 mL of each water sample was transferred into a COD digestion tube, then 0.4 g HgSO₄ and 30 mL AgSO₄-H₂SO₄ were added to the sample solution, and finally the mixture of each water was oxidized with 10 mL of 0.25 mol・L⁻¹ K₂Cr₂O₇ standard solution. After heating reflux for about 2 h, the residual content of K₂Cr₂O₇ was back titrated by corresponding 0.1 mol・L⁻¹ [(NH₄)₂Fe(SO₄)₂∙6H₂O] standard solution. The obtained COD_Cr results in the prepared water samples are summarized in Table 2.

As can be seen from Table 2, for the simulated high chloride oilfield wastewater samples with a theoretical COD_Cr of 100 mg・L⁻¹ and high level of Cl⁻, when the mass concentration of Cl⁻ was less than 500 mg・L⁻¹, the relative errors of COD_Cr determined by the classical K₂Cr₂O₇ method almost conformed to the allowable error range required by the Chinese government. But with the increase of Cl⁻ mass concentration in water samples, not only the COD_Cr values determined by the classical K₂Cr₂O₇ method increased gradually, but also the differences between the experimental values and theoretical ones increased gradually. And the relative errors obtained in low COD water samples are more remarkable than those of high COD water samples as described in Section 3.1.1, which is mainly due to the higher relative concentration of Cl⁻ in low COD water samples than those in high COD water samples under the same mass concentration of Cl⁻. Therefore, it is very necessary for pollutant discharging companies, environmental monitoring and legislation departments to develop a method for the accurate determination of COD in high chloride wastewater.

At present, the classical K₂Cr₂O₇ method (GB 11914-89) is used to determine COD in water sample by concentrated H₂SO₄-assisted heating reflux. During the reaction process, not only organic compounds but also Cl⁻ in water sample were oxidized by K₂Cr₂O₇. The interferences of Cl⁻ are mainly attributed to the following two aspects:

1) Consumption of oxidant K₂Cr₂O₇

Regarding water sample containing Cl⁻, apart from the reaction between organic compounds and K₂Cr₂O₇, the Cl⁻ can be also oxidized by K₂Cr₂O₇ and thus consumes some oxidants, which will affect the accuracy of COD determination. The corresponding reaction equation about Cl⁻ in the water sample can be expressed as follows:

2 Cl − + 1 2 O 2 → Cl 2 + [ O ] 2 − (1)

Theoretically, 0.226 mg oxygen is needed for complete oxidation of 1 mg Cl⁻ [15.999/(2 × 35.453) = 0.226]. Taking the above reason into account, Cl⁻ can cause huge positive interferences on COD_Cr determination because of its consumption of oxidant K₂Cr₂O₇ under the condition of Chinese national standard GB 11914-89.

2) Consumption of catalyst Ag₂SO₄

In the mixture system of H₂SO₄-K₂Cr₂O₇, acidified K₂Cr₂O₇ can just oxidize most of the organic compounds in the water sample. Only after adding catalyst Ag₂SO₄, can the straight chain aliphatic compounds be completely oxidized, which increases the oxidation rate of organic compounds and makes the oxidation reaction proceed more completely. However, if the water sample contains Cl⁻, it will combine with Ag⁺ to produce AgCl precipitate, resulting in poisoning of the catalyst, then decreasing the oxidation rate of the sample, and finally affecting the final determination results of COD. The reaction between Cl⁻ and Ag₂SO₄ is expressed as follows:

Ag 2 S O 4 + 2 C l − → 2 Ag C l ↓ +   S O 4 2 −         ( K s p = 1.77 × 10 − 10 , 25 ˚ C ) (2)

In addition, the resulting AgCl precipitate can be further oxidized by K₂Cr₂O₇, which also consumes some oxidants. Moreover, the white precipitate makes the solution color at the end of titration become grey and difficult to be titrated accurately.

During the COD determination, the interference reaction of Cl⁻ is as follows:

C r 2 O 7 2 − + 14 H + + 6 C l − = 2 C r 3 + + 7 H 2 O + 3 C l 2 ↑ (3)

In Equation (3), the electrode reaction of electron pair C r 2 O 7 2 − / C r 3 + is:

C r 2 O 7 2 − + 14 H + + 6 e − = 2 C r 3 + + 7 H 2 O (4)

And the electrode reaction of electron pair Cl₂/Cl⁻ is:

C l 2 + 2 e − = 2 C l − (5)

According to the Nernst equation, the electrode potential of electron pair C r 2 O 7 2 − / C r 3 + can be calculated as follows:

φ C r 2 O 7 2 − / C r 3 + = φ C r 2 O 7 2 − / C r 3 + θ + R T n F ln α C r 2 O 7 2 − ⋅ α H + 14 α C r 3 + 2 = φ C r 2 O 7 2 − / C r 3 + θ + R T n F ln α H + 14 + R T n F ln α C r 2 O 7 2 − α C r 3 + 2 (6)

Similarly, the electrode potential of electron pair Cl₂/Cl⁻ can be calculated as:

φ C l 2 / C l − = φ C l 2 / C l − θ + R T n F ln p C l 2 / p θ α C l − 2 (7)

In the determination of COD_Cr, the conditional electrode potential of oxidant K₂Cr₂O₇ can reach to φ C r 2 O 7 2 − / C r 3 + = 1.55   V due to the strong acid medium (in non strong acid medium φ C r 2 O 7 2 − / C r 3 + = 1.33     V ) [20] , thus the oxidation reaction of Cl⁻ can occur completely ( φ C l 2 / C l − = 1.36   V ). And the higher the oxidation potential is, the easier the Cl⁻ will be oxidized, and consequently the greater the error introduced is.

As can be known from Equation (6), the concentration of C r 2 O 7 2 − and Cr³⁺ is one of the main factors that affect the electrode potential. Under certain conditions of COD_Cr determination, the higher the oxidant concentration is, the more residual amount of oxidant will be after the reaction, which makes the ratio of C r 2 O 7 2 − / C r 3 + larger, and results in a higher oxidation potential. Theoretically speaking, the proper reduction of the oxidant concentration is helpful to reduce the oxidation potential of K₂Cr₂O₇ and thus reduce the interferences of Cl⁻ on COD_Cr determination.

In order to investigate the effect of oxidant concentration on Cl⁻ oxidation, a series of NaCl solutions with Cl⁻ mass concentrations of 1000, 2000, 4000, 8000, 15,000 and 20,000 mg・L⁻¹ were prepared by standard NaCl. 20 mL of each water sample was transferred into a COD digestion tube, then appropriate amounts of HgSO₄ [m(HgSO₄): m(Cl⁻)=10:1] were added to the sample solution, and finally the mixture of each water was oxidized with 10 mL of 0.25, 0.1 and 0.025 mol・L⁻¹ K₂Cr₂O₇ standard solution, respectively. After heating reflux for about 2 h, the residual content of K₂Cr₂O₇ was back titrated by corresponding 0.1, 0.04 and 0.01 mol・L⁻¹ [(NH₄)₂Fe(SO₄)₂・6H₂O] standard solution, respectively. The obtained COD_Cr results in the prepared water samples are plotted in Figure 1.

As can be seen from Figure 1, on the one hand, under the same Cl⁻ mass concentration, both the oxidation rate of Cl⁻ and obtained COD_Cr values increased gradually with the increase of oxidant concentration; on the other hand, under the same oxidant concentration, the obtained COD_Cr values and their relative errors also increased gradually with the increase of Cl⁻ concentration. When the oxidant concentration was about 0.025 mol・L⁻¹, the oxidation rate of Cl⁻ was the smallest. These results indicated that the oxidant concentration plays an very important role during the determination of COD in high chloride wastewater. Therefore, the proper reduction of oxidant concentration is helpful to reduce the oxidation rate of Cl⁻ and improve the accuracy of COD_Cr determination, which is highly consistent with our theoretical analysis in Section 3.2.1.

In order to investigate the effect of oxidant concentration on the oxidation of organic compounds, a series of standard KC₈H₅O₄ solutions with theoretical COD_Cr values of 50, 100, 300 and 600 mg・L⁻¹ were prepared by standard KC₈H₅O₄. After proper dilution, 20 mL of each water sample was transferred into a COD digestion tube, then 30 mL AgSO₄-H₂SO₄ was added to the sample solution, and finally the mixture of each water was oxidized with 10 mL of 0.25, 0.1 and 0.025 mol・L⁻¹ K₂Cr₂O₇ standard solution, respectively. After heating reflux for about 2 h, the residual content of K₂Cr₂O₇ was back titrated by corresponding 0.1, 0.04 and 0.01 mol・L⁻¹ [(NH₄)₂Fe(SO₄)₂・6H₂O] standard solution, respectively. The obtained COD_Cr results in the prepared water samples are summarized in Table 3.

Table 3 revealed that under the same condition of COD_Cr determination, different concentrations of oxidants exhibit almost the same oxidation ability on the standard KC₈H₅O₄, and the measured COD_Cr values are nearly consistent with the theoretical ones. In addition, with the increase of theoretical COD_Cr value in the samples, the measured results by different concentrations of oxidants are almost the same, which implied that the concentration of oxidant has little effect on the oxidation of organic compounds as long as the amount of K₂Cr₂O₇ is enough.

In order to investigate the effect of the amount of masking agent on COD_Cr determination, in this part, a series of standard water samples with a theoretical COD_Cr value of 500 mg・L⁻¹ and Cl⁻ mass concentrations of 1000, 2000, 4000, 8000, 15,000 and 20,000 mg・L⁻¹ were prepared by standard KC₈H₅O₄ and standard NaCl. After 10-folds dilution, 20 mL of each water sample was transferred into a COD digestion tube, then different amounts of HgSO₄ based on the mass ratio of HgSO₄:Cl⁻ together with 30 mL AgSO₄-H₂SO₄ were added to the sample solution, and finally the mixture of each water was oxidized with 10 mL of 0.25 mol・L⁻¹ K₂Cr₂O₇ standard solution. After heating reflux for about 2 h, the residual content of K₂Cr₂O₇ was back titrated by corresponding 0.1 mol・L⁻¹ [(NH₄)₂Fe(SO₄)₂・6H₂O] standard solution. The obtained COD_Cr results in the prepared water samples are summarized in Table 4.

As can be seen from Table 4, the relative errors of COD_Cr results decreased gradually with the increase of the mass ratio of HgSO₄: Cl⁻. When the mass ratio of HgSO₄: Cl⁻ is 10:1, the relative errors are small (<±3%), and they are not significantly improved with the increase of the mass ratio of HgSO₄: Cl⁻. However, the environmental pollution risk would increase if large amounts of HgSO₄ were used. Therefore, 10:1 was suggested to be the suitable mass ratio of HgSO₄: Cl⁻ for COD_Cr determination in high chlorine wastewater.

*The values in bracket are the relative errors and the same meanings in Table 5 and Table 6.

A series of simulated high chlorine oilfield wastewater samples with theoretical COD_Cr values of 50 and 80 mg・L⁻¹, and Cl⁻ mass concentrations of 1000, 2000, 4000, 8000, 15,000 and 20,000 mg・L⁻¹ were prepared by standard KC₈H₅O₄ and standard NaCl. According to the classical K₂Cr₂O₇ method, 20 mL of each water sample was transferred into a COD digestion tube, then appropriate amounts of HgSO₄ [m(HgSO₄) : m(Cl⁻)=10:1] and 30 mL AgSO₄-H₂SO₄ were added to the sample solution, and finally the mixture of each wastewater was oxidized with 10 mL of 0.25 mol・L⁻¹, 0.1 mol・L⁻¹ and 0.025 mol・L⁻¹ K₂Cr₂O₇ standard solution, respectively. After heating reflux for about 2 h, the residual content of K₂Cr₂O₇ was back titrated by corresponding 0.1, 0.04 and 0.01 mol・L⁻¹ [(NH₄)₂Fe(SO₄)₂・6H₂O] standard solution, respectively. The obtained COD_Cr results in the simulated high chlorine oilfield wastewater samples are summarized in Table 5.

Table 5 shows that with the increase of Cl⁻ mass concentration in wastewater samples, the relative errors of COD_Cr determination increased gradually. However, the obtained results provided by low concentration oxidant method (0.025 mol・L⁻¹) are much better than those determined by 0.25 mol・L⁻¹ and 0.1 mol・L⁻¹ oxidants, and the relative errors conform to the national requirements for COD_Cr determination. These experimental results demonstrated that low concentration oxidant method is feasible for the determination of low COD_Cr in high chlorine wastewater when Cl⁻ concentration is not more than 20000 mg・L⁻¹ and COD<100 mg・L⁻¹.

It can be inferred from the above experimental results that the determination of high COD_Cr in high chlorine wastewater is also possible by low concentration oxidant method. In order to verify this inference, a series of simulated high chlorine oilfield wastewater samples with theoretical COD_Cr values of 500, 1000 and 3000 mg・L⁻¹, and Cl⁻ mass concentrations of 5000, 10,000, 50,000, 10,000 and 200,000 mg・L⁻¹ were prepared by standard KC₈H₅O₄ and standard NaCl. After proper dilution, 20 mL of each water sample was transferred into a COD digestion tube, then appropriate amounts of HgSO₄ [m(HgSO₄): m(Cl⁻) = 10:1] and 30 mL AgSO₄-H₂SO₄ were added to the sample solution, and finally the mixture of each wastewater was oxidized with 10 mL of 0.025 mol・L⁻¹ K₂Cr₂O₇ standard solution. After heating reflux for about 2 h, the residual content of K₂Cr₂O₇ was back titrated by 0.01 mol・L⁻¹ [(NH₄)₂Fe(SO₄)₂・6H₂O] standard solution. The obtained COD_Cr results in the simulated high chlorine oilfield wastewater samples are summarized in Table 6.

As can be seen from Table 6, the low concentration oxidant method provides relatively small errors for the determination of high COD_Cr in high chlorine wastewater, and the obtained results are accurate and reliable. The above experiments proved that it is feasible to determine COD_Cr (including both low COD and high COD) in high chlorine wastewater by low concentration oxidant method. Therefore, in the next part, we intend to extend the low concentration oxidant method to determine the COD_Cr values in actual high chloride oilfield wastewater.

Six kinds of oilfield wastewater samples named JHSX, JHCL, XJ81J, XJ81C, XJ91J and XJ91C were collected from Jianghan oilfield and Xinjiang oilfield, China. The Cl⁻ contents in the six wastewater samples are 92,744.4, 74,155.4, 7578.5, 7699.7, 2086.8 and 2273 mg・L⁻¹, respectively, which were determined by a Thermo Fisher ICS-2100 ion chromatograph. It clearly shows that these wastewater samples belong to typical high chloride oilfield wastewater. Therefore, in this section, the COD_Cr values in these six high chloride oilfield wastewater samples were determined by the proposed low concentration oxidant method. According to the low concentration of oxidant method, each wastewater sample was diluted 10-folds and 20 mL of diluted wastewater sample was transferred into a COD digestion tube, then appropriate amounts of HgSO₄ [m(HgSO₄): m(Cl⁻)=10:1] and 30 mL AgSO₄-H₂SO₄ were added to the sample solution, and finally the mixture of each wastewater was oxidized with 0.025 mol・L⁻¹ K₂Cr₂O₇ standard solution. After heating reflux, the residual content of K₂Cr₂O₇ was back titrated by corresponding 0.01 mol・L⁻¹ [(NH₄)₂Fe(SO₄)₂・6H₂O] standard solution. The obtained COD_Cr results in the six kinds of actual high chloride oilfield wastewater samples are summarized in Table 7.

As shown in Table 7, the COD_Cr values of six kinds of high chloride oilfield wastewater samples are not high, which ranges from 100 to 150 mg・L⁻¹. Furthermore, spiked recovery experiments were carried out to verify the accuracy of low concentration oxidant method for COD_Cr determination in actual high chloride oilfield wastewater. Firstly, 500 mg・L⁻¹ theoretical COD_Cr was added to the above six high chloride oilfield wastewater samples, and then the COD_Cr values of these spiked oilfield wastewater samples were determined by the proposed low concentration oxidant method. The spiked recoveries were calculated as c₁/c₂ × 100%, where c₁ is the difference of measured total COD_Cr value of the spiked oilfield wastewater samples and unspiked oilfield wastewater samples; and c₂ is the theoretical COD_Cr value of added standard KC₈H₅O₄. The calculated recovery results are shown in Table 7. It can be seen from Table 7 that the spiked COD_Cr recoveries of six high chloride oilfield wastewater samples measured by the low concentration oxidant method were in the range of 95.4-106.5%, which revealed that the low concentration oxidant method is feasible and reliable for the determination of COD_Cr in high chloride oilfield wastewater.