Chapter 1
Rheological Characterization of Crude Oil and Related Products

Flávio H. Marchesini

Pontifical Catholic University of Rio de Janeiro

1.1 Introduction

Crude oil and related products undergo different transport processes from extraction to end use. For example, crude oils may be transported through pipelines before the refining process (Petrellis and Flumerfelt, 1973; Smith and Ramsden, 1978; Rønningsen et al., 1991; Wardhaugh and Boger, 1991a), fuel oils are injected into combustion engines to produce mechanical work (Graboski and McCormick, 1998, Ramadhas et al., 2004; Agarwal, 2007; Joshi and Pegg, 2007), and lubricant oils are used to reduce friction between mechanical parts in contact (Dyson, 1965; Webber, 1999, 2001).

The design of each of these processes requires the rheological properties of the oils, as the pumping power and the dimensions of the lines, connections, and mechanical parts are defined assuming that the oil has a viscosity within a specific range. If this range is not properly set during the design stage and the process starts running with an oil having a viscosity out of the appropriate range, different issues can arise. For example, severe flow assurance issues can be faced during the restart flow of crude oils in pipelines (Petrellis and Flumerfelt, 1973; Smith and Ramsden, 1978; Wardhaugh and Boger, 1991a; Rønningsen et al., 1991), and filters and lines can be plugged, preventing an engine from starting (Graboski and McCormick, 1998; Ramadhas et al., 2004; Agarwal, 2007; Joshi and Pegg, 2007). Therefore, to guarantee that the process works properly, the rheological properties of these oils must be known as accurately as possible, in representative process conditions.

In general, at high enough temperatures, crude oil and related products behave as simple Newtonian liquids, whose viscosities depend solely on temperature. However, at low enough temperatures, the rheological behavior of these oils usually becomes quite complex due to precipitation of higher-molecular-weight compounds, which gives rise to a gelation phenomenon when a certain amount of crystals is present. At this low temperature range, the oil viscosity increases significantly and depends not only on temperature but also on time, shear, and thermal and shear histories (Petrellis and Flumerfelt, 1973; Smith and Ramsden, 1978; Wardhaugh and Boger, 1987, 1991b; Rønningsen et al., 1991; Rønningsen, 1992; Chang et al., 1998, 2000; Webber, 1999, 2001; Venkatesan et al., 2005).

This complex rheological behavior at low temperatures may introduce difficulties in performing the rheological characterization of these oils. A number of precautions must be taken to get accurate properties during rheological measurements with these oils (Wardhaugh and Boger, 1987, 1991a; Marchesini et al., 2012; Alicke et al., 2015). Thus, we discuss in this chapter how to prepare samples for rheological measurements (in Section 1.2), the most common rheological tests performed with these oils and how to interpret the data (in Section 1.2), and the potential sources of errors in rheological measurements and how to avoid them (in Section 1.3).

1.2 Sample Preparation for Rheological Characterization

As described in this section, the sample preparation procedure for rheological characterization can be divided into four main steps: (i) ensuring the chemical stability (Section 1.2.1), (ii) choosing the rheometer geometry (Section 1.2.2), (iii) erasing the thermal memory (Section 1.2.3), and (iv) performing the cooling process (Section 1.2.4).

1.2.1 Ensuring the Chemical Stability

The first step of the sample preparation procedure is to make sure that the crude oil or related product is not going to evaporate or lose significant amounts of lightweight compounds under the temperature and pressure conditions in which the rheological test is going to be performed. This step is intended to guarantee the chemical stability of the sample during the test, thus avoiding evaporation effects on the time-dependent rheological properties being measured (Wardhaugh and Boger, 1987).

If the oil is not stable enough at the test conditions, a pretreatment can be applied to the oil to evaporate light ends before loading a sample into the rheometer or viscometer used. The pretreatment usually consists of heating the oil at a temperature within the temperature range of the process of interest (Smith and Ramsden, 1978; Wardhaugh and Boger, 1987; Marchesini et al., 2012).

It is important noting that a difference between the rheological properties of the pretreated oil and the untreated oil can be observed, and higher viscosity values are usually obtained for the samples after applying a pretreatment to evaporate light ends (Wardhaugh and Boger, 1987). However, with regard to many applications, the rheological tests with the pretreated oil provide conservative data for the transport process design (Wardhaugh and Boger, 1991a). If this is not the case or if more accurate data is needed, the rheological properties of the pretreated oil can be corrected by estimating the increase in viscosity due to evaporation of light ends (Wardhaugh and Boger, 1987; Rønningsen et al., 1991).

1.2.2 Choosing the Rheometer Geometry

The second step is to choose the appropriate rheometer geometry in which the sample is going to be loaded for rheological characterization. The classical geometries used to perform the rheological characterization of materials in rotational rheometers are: (i) cone and plate, (ii) parallel plates, and (iii) concentric cylinders (also known as the Couette geometry). To decide which is the best geometry for the rheological characterization of a given oil used for a particular application, some points must be addressed.

If the rheological tests are going to be performed in a temperature range in which no crystals appear in the sample, the oil may present a Newtonian behavior. In this case, any classical geometry is expected to give the same results, so any of the three geometries can be chosen. However, if crystals are expected to appear during the test and if the oil presents the complex rheological behavior expected at low temperatures, the rheometer geometry must be carefully chosen to obtain reliable data of the bulk rheological behavior (Marchesini et al., 2012).

Even though the cone and plate geometry is widely used for the rheological characterization of crude oil and related products, this geometry may not be the best choice depending on the oil at hand and test conditions (Marchesini et al., 2012). In favor of the cone and plate there is the argument that it is the only geometry in which all parts of the sample are submitted to exactly the same shear rate (Wardhaugh and Boger, 1987). In addition, as the cone and plate geometry requires a small amount of sample, it may be easy to control the temperature inside the sample. However, the cone and plate geometry is not suitable for the rheological characterization of samples having large enough crystals suspended, as it may violate the continuum hypothesis used in the rheometer theory. In addition, there is evidence in the literature that very small gaps-as the ones of commercial cone and plate geometries-cause the precipitation of crystals at higher temperatures (Davenport and Somper, 1971; Rønningsen et al., 1991). Thus, to obtain the bulk rheological properties of these oils at low temperatures, large enough gaps are required (Marchesini et al., 2012).

In this case, the parallel plates or the concentric cylinders can be chosen. The parallel-plate geometry has the advantage of being the best geometry to vary the gap, thus making easy the task of finding the large enough gap above which the rheological data stop changing with the gap. Moreover, the parallel-plate geometry is also a convenient choice for preventing apparent wall slip during rheological measurements, as it is easy to vary the gap and roughen its surfaces (e.g. by using sandpaper). However, the parallel-plate geometry has the disadvantage of having a shear rate dependence on the radius inside the sample, which might complicate the control of the shear history in some cases. It is important to note that as the highest shear rates occur at the highest radii-the regions that contribute most to the torque being measured-the non-homogeneous flow field in the parallel-plate geometry should not be a serious issue, at least in some cases. Corrections are available in the literature to end up with more accurate data when using the parallel-plate geometry (de Souza Mendes et al., 2014).

The concentric cylinders geometry presents the advantage of having a much less significant shear rate gradient inside the sample when compared to the parallel-plate geometry, allowing for a better control of the shear history in some cases. However, the concentric cylinders require larger sample volumes, which can lead to errors in the measurements due to contraction of the sample during the test (Wardhaugh and Boger, 1987, 1991a). Besides that, to obtain gap-independent results with the concentric cylinders geometry, cylinders with different diameters ratio are needed to vary the geometry gap, which may not be available. So, the best choice of rheometer geometry to get accurate data may depend on each case (Marchesini et al., 2012).

1.2.3 Erasing the Thermal Memory

The third step is to load the oil sample into the rheometer geometry and...