The paradox of asphaltene precipitation with normal paraffins

The paradox of asphaltene precipitation with normal paraffins
复制标题

DOI:
10.1021/ef0496956
复制
发表时间:
2005-07-01
期刊:
影响因子:
5.3
通讯作者:
Teclemariam, A
Teclemariam, A
中科院分区:
工程技术3区
文献类型:
--
作者:
Wiehe, IA;Yarranton, HW;Teclemariam, A

文献摘要

被引文献

相似文献

对于沥青和原油,在絮凝点(即初始沥青质沉淀点)处正构烷烃的体积随着正构烷烃碳数的增加而增加,在碳数为9或10时达到最大值,然后减小。因此,沥青质通常可以用比用正戊烷更小体积的正十六烷开始沉淀,即使大体积的正十六烷比大体积的正戊烷沉淀少得多(和更多芳族)的沥青质。对于沥青质来说,正十六烷是一种比正戊烷更好和更差的溶剂?溶剂质量的这种悖论可以通过将不同尺寸的分子的混合熵与溶解度参数的混合热相结合来解决,如由常规的Flory-Huggins模型所表示的。有了足够的表征数据,Yarranton等人的近似值和方法可以定量描述从絮凝点到正构烷烃从戊烷到十六烷大量过量的沥青质沉淀。为了仅描述沥青和原油的絮凝点数据,可以使用Wiehe的油相容性模型。尽管油相容性模型是基于对于给定油在絮凝点的溶解度参数是恒定的,但是使用“有效”溶解度参数,可以用很少的表征数据预测絮凝点。
For bitumens and crude oils, the volume of n-paraffin at the flocculation point, which is the point of incipient asphaltene precipitation, increases as the n-paraffin carbon number increases, reaching a maximum at a carbon number of 9 or 10, and then decreases. Thus, asphaltenes often can begin precipitating with a smaller volume of n-hexadecane than with n-pentane, even though large volumes of n-hexadecane precipitate much less (and more aromatic) asphaltenes than large volumes of n-pentane. How can n-hexadecane be both a better and a poorer solvent than n-pentane for asphaltenes? This paradox of solvent quality can be resolved by combining the entropy of mixing of molecules of different sizes with the heat of mixing from solubility parameters, as expressed by the regular Flory-Huggins model. With sufficient characterization data, the approximations and methods of Yarranton et al. can quantitatively describe asphaltene precipitation from the flocculation point to large excesses of n-paraffins from pentane to hexadecane. To describe only the flocculation point data of bitumens and crude oils, the oil compatibility model of Wiehe can be used. Although the oil compatibility model was derived on the basis that the solubility parameter is constant for a given oil at the flocculation point, using "effective" solubility parameters, flocculation points can be predicted with little characterization data.