PTQ Q2 2025 Issue

Inputs

Model parameterisation with appropriate mixing rules

Predictions for blend

Titration experiment (assess onsets for crude compatibility)

Oil 1 Onset

Oil 2 Onset

Other oil components

Predicted onset

Resins

m L of titrant/g of oil

Asphaltenes

Asphaltene phase (wt%) (assess precipitation amounts in the renery streams)

Oil 1

Oil 2

Blend

Figure 1 Flowchart of methodology to model asphaltene precipitation in oil blends

blending and process optimisation. This enables refiners to make informed decisions quickly and effectively. Asphaltene model At the core of the MFCCT is a robust thermodynamic framework that considers vapour, liquid, and asphaltene as equilibrium phases.7 The model employs the cubic plus association equation of state (CPA-EOS) for phase equilib- rium calculations. Asphaltene precipitation is modelled by considering two key mechanisms: asphaltene-asphaltene self-association and asphaltene-resin cross-association. At least one onset data is required to parameterise the model to account for the association behaviour of asphaltenes. In terms of fluid characterisation, the refinery assay and the properties are represented as a defined number of pseudo-components, including the assignment of com- ponent properties for phase equilibrium calculations. The key steps involved in the characterisation are generating a boiling curve from the crude assay, translating the boil- ing curve into single carbon number (SCN) fractions, and lumping of SCN fractions into a specific number of pseu - do-components using appropriate estimation methods and correlations. The inclusion of saturates, aromatics, resins, and asphaltenes (SARA) or paraffins, naphthenes, and aromatics (PNA) data will enhance the characterisa- tion accuracy. Note that this asphaltene phase behaviour model applies only to a single-source oil when calculating precipitation amounts upon dilution with a pure precipitating solvent (for example, n-heptane). When a source oil is blended with other source oils, the association behaviour and the solubil- ity of asphaltenes are affected. Adaptation of the model for oil blends Figure 1 summarises the methodology used for phase behaviour modelling of oil blends. Fluid characterisation and parameterisation are first performed individually for each source oil. Each source oil requires a crude assay with asphaltene content and onset data. Simulated distillation data and bulk density are acceptable when full crude assay is absent. Second, assays for the source oils are blended based on their proportions, followed by fluid characterisation. Third,

model parameters for each source oil are blended using appropriate mixing rules to determine the blend’s param- eters. Finally, the phase equilibrium calculations use the characterised fluid and the blend parameters. The calcula - tions simulate the titration experiment, thereby predicting the onset for the blend. If the blend is unstable, the pre- cipitation amounts are calculated. Note that the Petro-SIM simulator automates these calculations once the source oil streams are defined. MFCCT validation Validation of the MFCCT was carried out using data from several refiners. A set of source oils was selected to meas - ure their composition, properties, and onset data. These source oils were then blended in different proportions to create a select number of blends, for which the onsets were also measured. All the source oil streams were set up and synthesised in the Petro-SIM software, including their onset data, to estab- lish the parameters for each source oil. A product blend stream was then created using appropriate source oil pro- portions, together with normal refinery property blending rules and MFCCT mixing rules, to calculate its composition and properties, including the predicted onset if the blend was stable. If the blend was unstable, the amount of pre- cipitation (asphaltene phase) was calculated. The predicted onset was validated by comparing it to the measured onsets. Figure 2 compares predicted and measured onsets for blends of Oil 1 and Oil 2. Due to proprietary constraints, their composition and properties are not disclosed. Onsets for Oil 1 and Oil 2 were fitted, and blend onsets (Oil 1/Oil2 in wt%) at 75/25, 50/50, and 25/75 were predicted. The weight-averaged onsets for blends, based on source oil

Compatibility assessments based on the BCI

Range

Interpretation

BCI = 0.0

Incompatible source oil proportions

0.0 < BCI < 0.7

Potentially incompatible source oil proportions (prone to unstable stream with processing) Potentially compatible source oil proportions

0.7 ≤ BCI < 1.0

BCI ≥ 1.0

Compatible source oil proportions

Table 1

PTQ Q2 2025