Comparing the Thermometric and Potentiometric Titrations of Acidic Crude Oils

Efficient process control and reliable plant operations are essential for the successful running of a petroleum refinery. One of the main causes of decreased efficiency in the refining process is corrosion. The effects of corrosion translate to an increase in economic and production costs. The two main sources of efficiency reducing corrosion in the refining process are naphthenic acids and sulfur species.

Substantial annual savings by eliminating unplanned shutdowns and conserving costly treatment chemicals can be achieved by monitoring the acidity of crude oil and other related process oils.

One of the oldest and preferred methods for analyzing the total acid number (TAN) in petroleum products is titration. Conventional methods like ASTMD664 are not optimized for analysis of petroleum feed stocks, refinery fractions and crude oil.

The measuring surface of the potentiometric electrode that is employed in conventional analysis methods is often coated with waxy crude oil or the asphaltenes precipitates contained in it. Formation of such coatings results in a decline in the response time of the electrode. The solvent causes dehydration of the critical hydration layer that is necessary for obtaining stable potentiometric readings. Replenishing this hydration layer is carried out at the cost of an increase in the analysis time per sample by 2-3 minutes.

Potentiometric TAN titrations may require up to 120mL of the solvent, to allow titration to an alternate buffer endpoint when a true inflection is not visible. Another disadvantage of traditional potentiometric titrations is electrode fouling. In the case of thermometric titrations the analysis is improved by the use of a sensor that is unaffected by difficult matrices.

Additionally, thermometric titrations require lesser volumes of solvent and often require complete sample analysis within two minutes. Apart from these advantages it has also been observed that the thermometric TAN titration data bears a close correlation with potentiometric TAN titration data, which enables easy implementation of this method in refineries. Figure 1 shows the Metrohm Thermometric TAN Analyzer.

Figure 1. Metrohm Thermometric TAN Analyzer

Samples

The expected concentration of the crude oil samples was 0.8-1.2 mg KOH/g. The two oils used as samples were desalted crude and raw crude process oils. The expected concentration of the process oils was 1.2-1.8 mg KOH/g. The different types of process oils taken were vacuum light gas oil, atmospheric heavy gas oil, vacuum heavy gas oil and 650 endpoint gas oil.

Instruments

The four principal components of the Metrohm Thermometric TAN Analyzer are:

1. Titrotherm thermometric titrator
2. tiamo™ Titration Software
3. Thermoprobe sensor
4. Dosino™ dosing system

Fast and responsive titrations can be performed based on the powerful data processing capabilities provided by the Titrotherm thermometric titrator. The Metrohm tiamo™ Titration Software that operates the titrator is capable of processing vast amounts of data at the rate of 10 measurements per second, which is required for the reliable detection of the endpoint. The Metrohm Thermoprobe is extremely sensitive and provides instant responses.

It does not require a reference system or calibration (since the emphasis is on ΔT and not absolute temperature), or maintenance. This sensor does not have a membrane measuring surface or a diaphragm that is likely to get clogged. For simplifying compliance tracking, the sensor is fully integrated with traceability functions.

Owing to its resilient design, the Thermoprobe may simply be dipped in the stirring solvent for cleaning between titrations. Responses from the sensor will continue as long as the sample flows through the protective cage, even with coating on the electrode. The patented Dosino dosing technology from Metrohm is by far the most accurate liquid handling system in the industry. Truly unassisted analysis is made possible with top-down dispensing that eliminates the influence of air bubbles.

Two configurations of the Metrohm Thermometric TAN Analyzer are being used – stand alone and automated. The former is characterized by a small foot print, and can be used in process areas that need a walk-up analysis station. In case of the automated configuration, the indicator is mixed with the solvent and fed as slurry in a single step.

This configuration is best suited for lab environments where analysts take up many tasks, sample batches, and for safety optimization by bringing down the analyst contact with sample, solvent and indicator. In the case of automation, each sample and electrode are treated consistently between titrations, enabling high accuracy and precision levels. The tables below list down the set up for thermometric and potentiometric titrations. Figure 2 shows the automated configuration for the thermometric analyzer.

Figure 2. Metrohm Automated Thermometric TAN Analyzer

The Solvotrode easyClean mentioned above is a combined potentiometric electrode specially designed for non-aqueous titrations. The static buildup which is commonly observed in organic solvents is avoided by using an electrically shielded shaft. For removal of the blockages of the reference a flexible ground joint diaphragm is used.

Reagents

For thermometric titrations, the reagents used were:

• Titrant : c(KOH) = 0.1 mol/L in IPA
• Sample Solvent – 75:25 Xylene:IPA
• Indicator – Paraformadehyde, >95% pure, SigmaAldrich, Cat. 158127

In case of potentiometric titrations, the reagents used were:

• Titrant: c(KOH) = 0.1 mol/L in IPA
• Sample Solvent – 5 mL H2O, 495 mL IPA, 500 mL toluene

Sample Preparation

There was no additional sample preparation steps involved in this process.

Analysis

Thermometric

Sample aliquots with masses varying between 3 and 9 g were weighed into a disposable plastic beaker. The next step was dosing of 30mL of the solvent, followed by mixing of the sample. Next, roughly 0.5g of dry paraformaldehyde was added to the sample manually. In the case of automated configurations, these steps may be combined by mixing 17g of the indicator with 1L of the sample solvent.

With the help of a peristaltic pump, 30ml of this solution was dispensed. Titration of the sample to a thermometric end point with c(KOH) = 0.1 mol/L in IPA was performed after mixing. After analysis of sample was complete, a stream of xylene was used to clean the electrode.

Potentiometric

Similar to thermometric analysis, sample aliquots with masses varying from 2 and 9g were weighed into a disposable plastic beaker. Following this step, 120 mL of ASTM D664 TAN solvent was dosed. Thorough mixing of the sample was allowed, followed by titration to potentiometric inflection point with c(KOH) = 0.1 mol/L in IPA.

A stream of TAN solvent was used to clean the electrode after completion of analysis. Rehydration of the electrode membrane was done using DI water for 2 minutes followed by rinsing with IPA. The tables below list the parameters in these titrations.

Thermometric

Potentiometric

Calculation

TAN is calculated using the following expression:

Where, TAN = Total acid number in mg KOH/g; V_EP1 = Titrant consumption in mL; Blk = Blank volume in mL; c(titrant) = Concentration of titrant in mol/L; M[KOH] = Formula mass of KOH 56.11g/mol; and WS = Sample weight in g.

Results and Discussion

Across all the 12 blank determinations it was observed that the thermometric blank volumes were similar and the mean blank value was 0.073mL. The absolute deviation value for these blanks was 0.017mL. This deviation amounts to ±0.018 mg KOH/g change in concentration in a 5g sample.

The recorded data proved that a blank value was not required over a variety of process samples and crude oil. In case of a blank value exceeding 0.1mL the analyst needed to either re-analyze or prepare fresh titrant and solvent. No manual integration was required for the selection of endpoint as it was done automatically.

A derivative formula was applied instantly by the tiamo™ Titration Software for confirming the automatic endpoint selection. The average time taken for reaching thermometric endpoints was 59.4s in contrast to an average time of more than 3.5 minutes per sample by potentiometric titrations. The thermometric titrations required minimal volumes of sample with the average requirement being 2.15 mL. This time did not include the cleaning and rehydration time (in case of potentiometric titration). Table 1 shows the sample of comparison data. Table 2 shows the comparison of titration time.

Table 1. Sample of comparison data, TAN mg KOH/g

Table 2. Titration time comparison, seconds (s) to EP for 5g sample

Note: Analysis times do not include time to clean and prepare electrode between titrations.

Figure 3 shows the thermometric titration curve for raw crude, and Figure 4 shows the corresponding potentiometric titration curve.

Figure 3. Thermometric titration curve for raw crude

Figure 4. Potentiometric titration curve for raw crude

Figure 5 depicts the data correlation plot on the expanded sample set.

Figure 5. Data correlation plot on expanded sample set

Conclusion

Based on the mentioned observations, it can be concluded that thermometric titration method produces highly reliable and accurate results for the evaluation of acidity in crude and refinery process oils. Samples presented in the highly repeatable TAN results at anticipated concentrations over a range of sample weights can be effectively analyzed using thermometric titration.

Titrations through the thermometric technique can be carried out quickly with minimal solvent, sample and titrant requirements as opposed to conventional methods. Evaluation of acidity of crude oils and refinery process oils is far more economical if the thermometric titration method is followed.

This information has been sourced, reviewed and adapted from materials provided by Metrohm AG.

For more information on this source, please visit Metrohm AG.