RELATIONSHIP BETWEEN EMULSION THICKNESS AND OF DEMULSIFIERS

Water-in-Oil (W/O) Emulsion is deemed as one of the most difficult problems in colloid science. Several modelling studies have been conducted to gain a more thorough understanding of the separation process of emulsion from different configurations. Some of these phenomenological models are based on experimental published data or theoretical estimates of the coalescing process. However, authors have agreed that a general coalescence model that can describe coalescence of liquid-liquid systems that can encapsule all scaling parameters is needed. For that reason, a “simple” one dimensional mechanistic coalescence model is vital, which outlines the key parameters correctly. Unavoidably, these fitting parameters must be established by information derived from experiments. Hence, this improved empirical model, accounts for all the major phenomena of coalescence probability. Using previous works of (Noik, C.; Palmero, T.; Dalmazzone, C., 2013), the model includes droplet polydispersity and the viscosity of a liquid-liquid system, through the adoption of Sauter Mean Diameter (d32) and Einstein’s viscosity equation (𝜂𝑐−𝑐). The methodology first looked at settling of droplets through the existing model, and then introduced this to determine how coalescence behaves in the presence of demulsifiers from various authors. Application of this modified model to experimental research papers was performed and the results were recorded. As expected by enhancing the viscosity of the oil phase, this showed how fast the droplets settle for hindered settling section. The model results are somewhat acceptable for the phase distributions, albeit there were some issues with the units published by the authors which did not derive an exact replication of the stimulated results. This study outlined how considering the impact of rheological properties with droplet coalescence phenomena can be integral. The simulated results of emulsion separation profiles were differentiated with the predefined model, between experiments and tests simulation. The model sensitivity to operational parameter such as oil characteristics were tentatively analyzed. Additionally, recommendations are presented for further work which entails consideration of Population Balance Equations (PBEs) with several velocity groups, as compared to one and its effects of different demulsifiers.

油包水（Water-in-Oil, W/O）乳液被认为是胶体科学领域中最棘手的问题之一。目前已有多项建模研究相继开展，旨在更深入地理解不同构型下乳液的分离过程。其中部分现象学模型基于已发表的实验数据或聚结过程的理论估算，但学界一致认为，亟需一种能够涵盖所有尺度参数、可描述液-液体系聚结行为的通用聚结模型。

有鉴于此，一款“简洁”的一维机理聚结模型至关重要，该模型可准确刻画核心关键参数。这类拟合参数不可避免地需要通过实验衍生信息来确定，因此本改进型经验模型兼顾了聚结概率的所有核心现象。本研究依托Noik等人（2013年）的前期工作，通过引入索特平均直径（Sauter Mean Diameter, d32）与爱因斯坦黏度公式（𝜂𝑐−𝑐），将液滴多分散性与液-液体系黏度纳入模型考量。

本研究方法首先基于现有模型分析液滴沉降行为，随后引入该模型以探究不同研究者提出的破乳剂存在时的聚结特性。将该修正模型应用于已发表的实验研究并记录结果。正如预期，提高油相黏度后，该模型可直观呈现受阻沉降区间内液滴的沉降速率。尽管部分作者所使用的单位未能精准复现模拟结果，该模型在相分布预测上仍具备一定的合理性。

本研究阐明了将流变特性与液滴聚结现象相结合的重要性：将乳液分离剖面的模拟结果与预设模型进行对比，以区分实验与模拟测试的差异；同时初步分析了模型对油相特性等操作参数的敏感性。此外，本研究提出后续研究方向建议：相较于单速度组模型，应考虑采用多速度组的群体平衡方程（Population Balance Equations, PBEs），并探究不同破乳剂的影响。