Rapid Quantification of Oil in Wastewater on Off-Shore Oil Platforms

During the oil production process on off-shore platforms, large amounts of waste water are accumulated. In order to ensure an environmentally safe waste dump, the quality of the waste water is regulated by law. Therefore special parameters like the total hydrocarbon content have to be determined.

The current standard measurement technique is based on the gas-chromatography (GC)-analysis of a pentane extract of the waste water. Due to the fact that pentane is highly flammable and forms explosive gas-air mixtures, its use is not desirable on oil platforms. The waste water sample thus, needs to be transferred to a mainland lab by helicopter. Obviously, this form of water analysis is highly expensive and time consuming. As an alternative to the GC-based measurements, FT-IR based methods have already been established and used.

The main disadvantage is the need for C-H free extraction solvents like carbon tetrachloride (CCl₄) or tetrachloroethylene (C₂Cl₄).Comparatively, the ALPHA-based technique is a highly effective approach for the fast screening of waste water samples, however it does not replace the OSPAR GC-based method when official results are required. It uses a simple extraction method with small amounts of a less harmful solvent, chloroform. Since the chloroform extraction is not prone to emulsification, as oppose to the case of pentane, there is no need to add salt or to centrifuge the extract. With the exception of filtration using a disposable syringe filter, there is actually no need for further treatments of the extract. Subsequent to this easy step, the extract measurement can be performed directly on the oil platform with an ALPHA spectrometer in combination with a liquid flow through the cell.

In contrast to the case of GC-analysis the sample does not need to be transported to a mainland lab. The measurement is therefore extremely cheap and fast. The complete test including the sample preparation, the measurement and the analysis takes less than 10 minutes. Plus, the intuitive and easy-to-use OPUS-LAB software allows getting reliable results even when the measurement is performed by an untrained person.

Instrumentation

The highly compact and robust ALPHA FT-IR spectrometer has proven to be highly suitable for standard applications. The samples are measured by means of a specifically designed liquid flow through cell. With the aid of multivariate “Quant 2” calibrations, it is possible to measure hydrocarbons in chloroform despite the fact that chloroform itself has C-H absorbance. This approach enables a very accurate determination of the hydrocarbon content with a typical error of prediction in the range of 2-3 ppm. As an additional benefit the OPUS-LAB based user interface provides a very intuitive way to perform the IR analysis procedure.

Figure 1. ALPHA spectrometer with flow through cell.

The ALPHA FT-IR spectrometer ensures high reliability of the data. A permanent online diagnostics of the spectrometer, by the PerformanceGuardTM, provides a “real time” display of the instrument status. The instrument validation (OQ/PQ) is performed by fully automated test routines to ensure that the instrument is constantly operating within the specifications. The OPUS software is fully compliant to cGMP and 21 CFR part 11 when operated in a validated environment.

Example Application

Determining the content of hydrocarbons in a waste water sample comprises essentially two steps: sample extraction and measurement of the resulting extract with the ALPHA FT-IR spectrometer. The evaluation and documentation of the results is performed automatically by the OPUS-LAB software. The straightforward procedure to analyze the waste water is as follows:

• At the sampling point, ensuring that the water flows for one minute, then taking a sample of one liter waste water.
• Adding 50 ml of chloroform to the water sample and shaking for 5 minutes.
• Separating the organic chloroform phase via a separation funnel.
• Flushing the organic phase through a disposable syringe filter into the flow through cell.

Figure 2. OPUS-LAB user interface.

The measurement itself is extremely simple:

• Click on “Measurements”, then on “Start”.
• When the sample is filled into the flow through cell simply press “Measure”. The measurement takes about 20 seconds. When completed, the report is automatically generated as a PDF file and can be printed:
• Once a day a background measurement against air is required.

Figure 3. Automatically generated report.

Details on the Measurement Evaluation

Figure 4 shows various samples with different amounts of petroleum hydrocarbons in chloroform. All spectra contain similar spectral features that originate from the chloroform spectrum. The differences due to the varying content of petroleum hydrocarbons are noticeable in the region below 3000 cm^-1 where the vibrations of aliphatic C-H bonds are located. The region between 3000 and 2820 cm^-1 is used for the multivariate calibration. The enlarged version of this region can be seen in Figure 5.

Figure 4. Spectra of chloroform extracts with a hydrocarbon content between 0 and 100 ppm.

Figure 5. Zoom of the calibration region of the Quant 2 method.

Figure 6 shows the ppm values of reference samples predicted by the calibration model. These values are plotted against the hydrocarbon content determined with the GC-method. The chart is the result of an internal cross validation where one spectrum is taken out of the calibration spectra set and its value is predicted by a model created out of the residual calibration spectra. This is done for each single spectrum out of the calibration spectra set. The predicted values are marked as green spots and the true value is indicated by the straight green line. The deviations of the predicted values are very small. Consequently, the Root Mean Square Error of Cross Validation (RMSECV) is very low with a value of 1.98 ppm; the coefficient of correlation is 99.59 %. As a further test the model was used to predict seven external samples.

Figure 6. Result of the cross validation. Predicted ppm values compared to the true ones.

The results of this test are listed in table 1. True and predicted values are in good agreement with each other; the deviation of the measurement values is in all cases less than 3 ppm. This error includes systematic and random errors.

Table 1. The hydrocarbon values of various samples as predicted by the Quant 2 model compared to the true values.

Summary

By using the ALPHA spectrometer system with a specially designed flow through cell a fast and accurate study of the hydrocarbon content in waste water is possible directly on oil platforms. The robust instrumental setup, the simple sample preparation and the user-friendly measurement procedure allows performing the analysis even by untrained users. The complete test takes less than 10 minutes and is therefore ideal as a monitoring method for large sample numbers.

This information has been sourced, reviewed and adapted from materials provided by Bruker Optics.

For more information on this source, please visit Bruker Optics.