Measurements of the Density and Viscosity of Heavy Oil and Water-in-Oil Emulsions Over a Wide Temperature Range

Density and dynamic viscosity of extra viscous heavy crude oil and water-in-oil (W/O) suspensions based on Ashalchinskaya (Tatarstan, Russian Federation) oil have been measured as a function of temperature and concentration of water at atmospheric pressure. The measurements were made at temperatures from (293 to 463) for density and from (293 to 367) K for viscosity with various concentrations of water (from 0 % to 30 % volume fraction). Measurements were made using modified hydrostatic weighing for density and falling body techniques for viscosity. The combined expanded uncertainty of the density, viscosity, pressure, and temperature measurements at 0.95 confidence level with a coverage factor of k = 2 is estimated to be Uρ\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$U\left( \rho \right)$$\end{document} = 0.16 % and Uη\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$U\left( \eta \right)$$\end{document} = 1.0 %, UP\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$U\left( P \right)$$\end{document} = 1.0 %, and UT\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$U\left( T \right)$$\end{document} = 0.02 K, respectively. The reliability and accuracy of the new experimental method and correct operation of the modified experimental apparatus was confirmed with measurements using different methods (pycnometric, capillary flow, commercial standard instruments, Brookfield rotational viscometer). The effect of temperature and concentration of water on the measured values of density and viscosity of W/O suspensions were studied. Using crude dry oil, the effective viscosities of several synthetic W/O emulsions are measured at atmospheric pressure using a commercial standard instrument, Brookfield rotational viscometer and falling body technique for different shear rates, temperatures and volume fractions of the dispersed phase. The various correlation equations for describing viscosity as a function of temperature and dispersed phase volume fraction is developed. A number of factors such as water content, shear rate, shear stress, and temperature and their effects on the density and dynamic viscosity of dry crude oil and W/O emulsions were assessed.


Introduction
Rapid economic development has greatly increased the demand for fossil fuel. As well-known that oil price changes continuously due to affect supply and demand. The market price for heavy crude oil is two times lower than light crude oil [1]. Heavy oil is growing rapidly. With the ever-increasing demand for energy and depletion of easy-to-produced light oil, heavy oil resources contribute to the crude oil supplement in the current energy market [2]. By growing depletion of conventional oil reservoirs potential, heavy and extra-heavy reserves have gained attention [3]. Heavy oil reservoirs contain more than 80 % of current petroleum resources [4,5]. Even after production, high viscosity crude oil to be transported through production equipment [6]. Viscosity is extremely important property for controlling production of heavy oils. High content of impurities, typically CO 2 and H 2 S, and high molecular weight species, particularly asphaltene, lead to high viscosity of heavy oils [7]. These form hydrocarbon compounds characterized by long, very complex (long) molecules, which are impart high internal friction resulting in high viscosity. The complexity of the crude oils can cause a variety of difficulties during the production, separation, transportation and refining of oil [8,9]. For heavy oil transportation we need to change physical properties. Demand for heavy and extra-heavy oil has been marginal because of their high viscosity and composition complexity that make them difficult and expensive to produce, transport and refine. Usually, the viscosity of crude oils at room temperature is > 100 mPa·s. Generally, crude oil with viscosity < 400 mPa·s is the classical maximum desired pipeline viscosity [10][11][12][13]. Oil pipeline transportation [14] has become a complex and highly technical operation due to high viscosity, even extremely viscous (extra viscous), i.e., are not pumped easily through the pipelines This is one of the major difficulties in the pipeline transportation that require efficient and economical ways to transfer the heavy crude oil [1,8,9], because of the high viscosity, especially if there is a high concentration of sulfur and some metals. To overcome this problem, reducing oil viscosity is a promising approach. Different methods were used in order to reduce the viscosity of the heavy crude oils for the pipeline transportation. There are various methods to reduce viscosity of heavy oil. Heating (thermal remediation) is a common method utilized to overcome the problem of transporting heavy and extra-heavy oil by pipeline by viscosity reduction [1,8,[15][16][17][18][19]. Crude oil pre-heating is the most attractive method, due to rapid reduction of oil viscosity [20]. This method based on the fact that as heavy oil is heated, its viscosity is reduced and thus made easier to pump. It is apparent that temperature dependence of the viscosity of crude heavy oil needs to solve the problem using heating method. Therefore, it is important to heat the oil to a point (to optimal temperature) where the oil has a substantially reduced viscosity to make easier for transportation. This is very important to make the method more efficiency. Thus, the temperature dependence of the viscosity of crude heavy oils is the key problem to pipeline transportation. Also, effect of temperature on the viscosity behavior controls the production of heavy oils (to improve the thermal method of enhanced heavy oil recovery), etc.
The role of temperature effect on reservoir crude oil density and viscosity for reservoir evaluation, thermal methods of enhanced heavy oil recovery, evaluation of hydrocarbon reserves and designing production equipment and pipelines makes its accurate determination necessary [21][22][23]. Viscosity and density are key properties of oils for characterization, evaluation, controlling and development of petroleum reservoirs. For design of oil processing equipment and petroleum reservoir simulation requires wide-ranged correlations of the viscosity and density as a function of temperature [21]. As well-known, thermal method is one of the effective methods of heavy oil recovery [22][23][24][25][26][27]. A thermal oil recovery process in which steam is injected into a heavy oil-bearing formation through a horizontally fdrilled injection well and oil is produced through a horizontal production well parallel to the injection well. Heat from the steam lowers the viscosity of the heavy crude wherein the crude is then produced to the surface via conventional lift arrangements. Under reservoir conditions, heavy oils are normally flowable in the formations. However, their flowability will be dramatically changed as temperature and pressure drop during lifting in the production wells due to viscosity changes with temperature decreasing. Unfortunately, this uncertain issue renders the heavy oils problematic on the transport-ability and even makes production wells blocked. Therefore, prior to proposing effective solutions, the accurate prediction of the heavy oil viscosity during lifting impacted by temperature and pressure is very desirable. Viscosity and density play an important role in the calculations of fluid flow through reservoir rock, pressure loss (with implications for the designs of tubing and pipelines), and the design of surface facilities, reservoir simulations, and predictions of oil recovery [24]. The reservoir heavy oil mobility (permeability/viscosity ratio) depend on viscosity and defining the efficiency of exploration and design optimization of production [28]. Vast quantities of heavy oil are trapped in shallow, accessible reservoirs, but are difficult to extract due to low flowability [29]. Viscous oil flows through a reservoir very slowly, therefore, heavy oil wells produce at lower rates than light oil wells. Therefore, the accurate viscosity and density data at various temperatures are very useful for heavy oil recovery methods including primary and thermal productions, and enhanced oil recovery (EOR). In all above-mentioned processes accurate and reliable density and viscosity data are require for heavy crude oil to improve the heavy oil production, reducing operating costs; optimization of pipeline construction, etc. Crude oil density and viscosity are a strong function of the reservoir temperature, pressure, bubble point pressure, oil gravity, gas gravity, gas solubility, and composition of the crude oil, therefore oil density and viscosity should be determined by laboratory measurements at reservoir temperatures and pressures [30,31].
Thus, crude oils are complex fluids that can cause a variety of difficulties during the production, separation, transportation and refining [8,9]. Crude oil is often mixed with water when it comes out from a well. In oil and gas industries, crude oil tends to produce together with water from the reservoir and the amount of water increased towards the end of reservoir life, particularly if the reservoir is driven by water aquifer (Lim et al. [32]). In crude oil production, oil and coproduced water are mixed due to high shear forces through the wellbore, pumps, choke valves and pipelines from the reservoir to the separation facilities. The formation of emulsion is problem that occurs in the petroleum industry where water content in the W/O emulsions can be reach to 60 % in volume. W/O emulsion forms naturally during crude oil production or can form during the refining operations and storage. Under standard oilfield conditions the most common form of emulsion is a W/O emulsion a dispersion of water droplets in oil. Emulsions of crude oil and water can be encountered at many stages during drilling, producing, transporting and processing of crude oils. This can be found in hydrocarbon reservoirs, well bores, surface facilities, transportation systems and refineries. Therefore, the formation of W/O emulsions is a usual process in oil field developments [33]. Emulsification, specifically formation of W/O emulsions, is a very common occurrence in the oil and gas industries which tend to occur naturally given the conditions of the industries. However, it is an unwanted phenomenon as it brings numerous harmful effects to the industries [34]. Formation of emulsions is undesirable in the oil and gas industries as it brings significant negative effects to the petroleum industries [35]. The rheology of W/O emulsions is of great interest in many industrial applications [35][36][37][38][39][40][41]. Property design and operation of the transport pipelines requires knowledge of the thermophysical properties of crude W/O mixture [40]. A good knowledge of petroleum emulsions, their properties and effect on operation processes is therefore necessary for controlling and improving processes at all stages [42]. As the oil-water mixture passes over chokes and valves, mechanical input leads to the formation of W/O emulsions [37,43,44]. The formation of crude oil emulsion is a prevalent and costly oil field problem that can cause significant flow assurance issues during oil production, operational problems like treatment and transportation difficulties, production loss, tripping of separation equipment, production of off-specification crude oil, petroleum refining operations [16,36,39,45,46]. The presence of emulsions in flow creating high pressure drops in pipes and/or flowlines, indicating higher energy loss as higher pumping energy is needed in order to maintain the desired mass flow rate [37]. Since the formation of W/O emulsions is inevitable and they bring many serious problems to the oil industry, therefore, study of the emulsions characteristics and their impact on oil treatment efficiency is extremely important. In the petroleum industry (in the processing, mixing, storage, and pipeline transportation of emulsions) the rheological properties of W/O emulsions are significant for modeling energy requirements, pipeline design and flow quality [37,38] assessment during transportation. In particular, it is important to be able to predict the viscosity of the emulsion as a function of the dispersed phase volume fraction. Formation of W/O emulsions is a problem for the oil industry due to their stability because of the presence of natural surfactants and stabilizers (polar compounds existing in the oil phase) such as asphaltenes, resins, and naphthenic acids, etc. [45][46][47][48][49][50][51][52]. Solid particles also play significant roles in stabilizing emulsions. For optimal design and operation of the pipelines, information on flow resistance, which is strongly depend on viscosity is required. There are many publications on the rheological properties of W/O emulsions, (see, for example [29,34,[53][54][55][56]) where were studies the effect of various factors such as droplet size distribution, viscosity of continuous phase (dry crude oil itself), viscosity of dispersed phase (water), temperature, etc., on viscoelastic properties.
It is well-known that water volume ration considerable increases the viscosity of W/O emulsion [57,58], while decreasing with temperature. Also, viscosity of W/O emulsion depends on shear rate, average droplet sizes and their distribution, viscosity and density of oil (continuing phase, dry oil). It is well-known [59] that the rheological properties of emulsions and their stability are essentially strongly depends by the volume fraction of dispersed phase (water), thermodynamic factors and chemical composition (concentration of asphaltenes and resins, volatile aromatic components, salt concentration, etc.) of oil and water.
It is well-known that on the walls of the pores of oil-saturated rocks there are adsorption-solvation layers with anomalous viscous properties. These layers have a noticeable effect on the filtration of oil, on the completeness of its displacement from the rocks. The development of large deposits of paraffinic oils using edge and in-loop flooding, the problem arose of changing the properties of oil in a porous medium and wells during cooling and release of dissolved gas. Studies have shown that the cooling of oil to a temperature below the paraffin saturation is accompanied by the formation of spatial structures, as a result of which the oil acquires structural and mechanical properties. The viscosity of such oils turns out to be unstable, depending on the shear stress. Viscosity anomalies are especially noticeable at low shear rates. Such oils are called anomalous, which means that do not obey Newton's law when they flow at low speeds.
The attention of researchers was attracted by the anomalies in the viscosity of asphaltene-containing oils during pipeline transportation. It is impossible to explain these anomalies by the properties of the adsorption-solvation layers, since their thickness is considerable small compared to the diameter of the pipes. The reason for the viscosity anomaly in such oils could be the formation of a three-dimensional structural network of asphaltene particles. If anomalies in the viscosity of asphaltene-containing oil are noticeable when it moves in pipes, then these anomalies obviously have a much stronger effect on filtration in a porous medium. Indeed, experiments on the filtration of such oil in a porous medium have shown that at low filtration rates, the mobility of the oil and its viscosity depend on the pressure gradient. It was studied how the content of asphaltenes affects the viscosity of oil and its filtration, the influence of pressure, temperature, content of dissolved gases and resins, rock permeability on the rheological characteristics of oil was studied. Oil viscosity anomalies, violations of Newton's law and Darcy's law during filtration can be the cause of low oil recovery. Statistical analysis of the results of the development of a large number of oil fields showed that for deposits of abnormal oils, oil recovery is much lower than for oil deposits, during the filtration of which viscosity anomalies are not observed. The reserves of asphaltene-containing oils are very large. In order to efficiently use them, it is necessary to take into account the anomalous properties of oil when designing the development and operation of such deposits. In many cases, this can be done on the basis of studies of the rheological properties of high-resinous oils.
The density and viscosity of anomalous oils and natural bitumen in reservoir conditions exceeds 10 thousand mPa·s. The main difference between bitumen and lowviscosity oils in the low content of light fractions (0 wt % to 2 wt %) and in the high 7 Page 6 of 46 content (25 wt % to 75 wt %) of asphaltene-resinous components, which is the reason for their high density (965 kg/m 3 to 1220 kg/m 3 ) and almost immobile in the reservoir.
The uncertainty of heavy oil property measurements affects the quality of the data, which in turn affects the accuracy of the production forecast and the recovery processes [60]. Also, to develop accurate prediction techniques of the physical properties (viscosity and density, for example) of various type of oils around the world reliable experimental data are required [61,62]. The variation in density and dynamic viscosity with temperature changes is typically predicted empirically based on reliable data [63]. The reservoir oil viscosity is typically measured isothermally at the reservoir temperature. However, at temperatures other than the reservoir condition, these data are estimated using empirical correlations [61,64]. High viscosity and low fluidity, which are governed by their fundamental physical and chemical properties of main (key) components of crude oils such as (i.e., saturates, aromatics, resins and asphaltenes, traces of sulfur, nitrogen, chlorine and metal compounds), and the complicated interactions among different oil species [65][66][67][68][69]. The viscosity of crude oil varies depending on its origin and type, as well as the nature of its chemical composition, particularly the polar components, for which intermolecular interactions can occur. For this reason, developing a comprehensive model of viscosity to include different regions of the world appears to be a difficult task. However, the published literature has lack of reliable methods for high-density liquids measurements at elevated temperatures. In the present work the effective viscosity of several synthetic W/O emulsions have been measured using crude oil from Ashalchinskaya oil filed (Tatarstan, Russian Federation), varying shear rate, temperature and volume fraction of dispersed phase. The main objective of the present study is to experimental study of the temperature and v/v fraction of water on the viscosity of W/O emulsions based on the Ashalchinskaya crude heavy oil and develop a method to describe the variation of dynamic viscosity of W/O emulsions with temperature and water v/v fraction. The detailed description of the experimental procedure is presented. Proposed modified falling-body technique (new design of the rheoviscometer) allows to accurately measure of the viscosity of Newtonian and non-Newtonian liquids. The measured values of density and viscosity of Ashalchinskaya crude heavy oil were used to develop a reliable correlation model. The models for W/O suspension viscosity as a function of temperature (VFT, Masuko-Magill, Arrhenius type models) and dispersed phase concentration (v/v fraction of water, Krieger-Dougherty and Mooney models) are described based on the measured viscosity data. The proposed models are compared with reported literature correlations. Thus, the objective of this work is to experimental study and modeling of the influence of the water cuts ranging from 0 v/v % to 30 v/v % and the temperature on the W/O emulsions properties (density and viscosity). Previous rheological studies showed that temperature (T), shear rate ( ̇ ), water volume fraction ( ), and API gravity have an important impact on the viscosity of W/O emulsions [70].

Brief Review of the Previously Studies of the Ashalchinskaya Crude Oil Properties
Previously reported density and dynamic viscosity data of Ashalchinskaya crude heavy oil is presented in Tables 1 and 2. The reported density and viscosity data covers the temperature range from (243 to 393) K and (243 to 363) K, respectively (see Tables 1, 2). As Table 1 shows, most density measurements for the Ashalchinskaya heavy crude oil have been performed at room temperature (at 293 K) and at atmospheric pressure using pycnometric method. Only one data source is available in the literature for the density of Ashalchinskaya heavy oil reported by Gussamov et al. [89] as a function of temperature.
Dynamic viscosity of Ashalchinskaya heavy oil has been previously measured by many authors (see Table 2). However, almost all measurements were made at single temperature of 293 K and at atmospheric pressure using Rotational viscometer. The typical uncertainty of the reported data are about (3 to 4) %. Ilyin et al. [84] studied of the rheological properties for typical heavy and light crude oils from Ashalchinskoe oil field (Transneft, JSC, Tatarstan, Russia). Measurements of the viscous properties of heavy crude oil were performed over a temperature range from (243 to 363) K. The measured viscosities were fitted to WLF model [90]. The measurements were performed with various API gravity over the temperature range from (293 to 433) K using various types of viscometers (VISCO lab 3000, Cambridge Viscosity, Inc. USA; Physica MCR 301 rotational rheometer from Anton Paar). The uncertainty of the viscosity measurements was 0.009 mPa·s. We have previously studied [88] the density and viscosity of Ashalchinskaya heavy oil over the temperature range from (293 to 459) K and from (293 to 348) K, respectively. The rheological properties and viscosity reduction of crude oil from various Oil Fields around the world (South China Sea, Iranian, Kuwaiti, Russia, etc.) have been previously studied by several authors [66,84,[90][91][92].
The present research can lead to a comprehensive understanding of the behavior of different emulsified systems. Emulsions often behave as non-Newtonian fluids and their viscosity can be several orders of magnitude higher than the continuous phase which causes considerable reduce the full capacity of oil production.

Materials
Dewatered (dry) crude oil used in the present study was obtained from Ashalchinskaya oil reservoir (Tatarstan Oil Field, Russia). As a preliminary study, the physic-chemical properties of the sample, were carried out to determine the density, viscosity and API gravity.

Main Physical-Chemical Characteristics of Ashalchinskaya Heavy Oil
The main physical-chemical properties (density, dynamic viscosity, refraction index, API gravity, chemical composition) at 293.15 K and atmospheric pressure, 0.101 MPa, which are characterizing the Ashalchinskaya heavy oil sample used in this study, are summarized in Tables 3 and 4. These characteristics of the sample were measured in the present work using pycnometric method (PZh-2-10-KSh7/16 GOST 22524-77), programming Brookfield rotational viscometer (DV-II + PRO, LVD-II + PRO) with a set of 4 spindles (SSL), and Reflectometer-densitometer RM-40. The uncertainty of the refraction index measurements using RM-40 is 0.0001. About 0.4 wt % water content was observed for Ashalchinskaya heavy oil sample. The gravity of the samples was characterized according to the relation of the American Petroleum Institute in API grades as (Speight [9]) where ρ = 0.95,651 g/cm 3 is the density of a sample at 15.6 °C. Therefore, API gravity for these samples is 16.4. Heavy crude oils are those for which API gravity lies in the range 10 < API < 22. Therefore, Ashalchinskaya oil sample in the present study is classified as heavy oil. Table 4 summarizes SARA analysis of the Ashalchinskaya crude oil sample under study (see also below Sect. 2.1.2).

Chemical Composition of the Ashalchinskaya Oil
Ashalchinskaya oil is a complex mixture containing a huge number of components. The content of asphaltenes, resins, paraffin, sulfur, nitrogen, carbon, hydrogen, vanadium, nickel, and mechanical impurities, has been determined. The results of the chemical composition (wt %) determination of the Ashalchinskaya heavy oil are summarized of in Table 5 together with earlier reported data. This table includes fractional composition; acid; coking capacity; the content of asphaltenes, resins, paraffin, sulfur, nitrogen, carbon, hydrogen, mechanical impurities, vanadium and nickel. The oil analysis has been started with the selection of a representative sample and the determination of its light hydrocarbon content C 1 to C 4 . The oil sample is dehydrated (and desalted) to a mass water content of not more than 0.5 wt %. ) with the selection of 3 % (by volume) or 10-degree fractions to a temperature of (723 to 773) K (start of decomposition). Then the residue is distilled according to the method [93] to (833 to 853) K with the selection of fractions 723 K to 773 K, 773 K to 793 K, 793 K to 813 K, 813 K to 833 K, 833 K to 853 K. For all 3 % narrow fractions, the density, refractive index, molecular weight, kinematic viscosity at 293 K, 323 K and 373 K, pour point, and total sulfur content were determined. For fractions boiling above 723 K to 773 K, the density, viscosity at 323 K, 353 K, and 373 K, pour point and flash point, coking capacity, sulfur content and softening point are determined by the "ring and ball" method [93,94]. For gasoline fractions with a boiling temperature up to 423 K to 453 K, the individual hydrocarbon composition is determined, as well as the content of n-alkanes, cyclanes, arenes and alkenes. For fractions boiling above 473 K, the content of paraffin-naphthenic, individual groups of arene hydrocarbons, the content of resins, alkanes, cyclanes, solid paraffin and structural group composition are determined. Comparison of the SARA analysis (resins, oils, and asphaltenes) of the Ashalchinskaya heavy oil sample under study using chemical and NMR methods (with a relative uncertainty of 3 % to 4 %) are presented in Table 6.
It should be noted from Table 6 that any of the three oil fragments determined by the NMR method is greater than their content in the oil composition determined by the chemical method. It can be explained by several reasons: (1) higher sensitivity of the system of nuclear spins of oil fragments when using the pulsed NMR 7 Page 12 of 46 0 -method compared to the standard (chemical) method. This is especially observed for the resin content in the sample; (2) the presence of paramagnetic impurities; (3) different molecular mobility of fragments of group composition of oils, especially for the resins containing oils. The difference between the resins content in the chemical analysis of the oil samples and when using the pulsed NMR method is 8.12 %; and (4) the presence of mechanical impurities in oils plays the role of active fillers for low molecular weight chains of oil fragments. This can lead to the aggregation of oil components, especially in the presence of 8 % to 10 % of asphaltenes. Thus,  The branch points of these chains can be porous formations that affect the molecular mobility of the group composition of oil chains. Porous formations can have different sizes and distributions within resins, oils, and even asphaltenes. They also participate in the processes of spin exchange and diffusion, because inside the pores there can be localized mechanical impurities, as well as salts of various metals available in the oils. This is indicated by the found values of the proton population, P a , P b , P c , as well as the observed decays of the nuclear magnetization signals in oil samples at different temperatures. Thus, it can be assumed that the structure of the Ashalchinskaya oil has a partially pronounced porous structure, and this should also affect other properties of the oils, for example, dynamic viscosity and density.

Water-in-Oil Emulsion Preparation and Stability
Water is present in almost all steps of crude oil production including exploitation, upgrading, transportation and processing [96] (see also Sect. 1). Therefore, it is important to evaluate the effect of water on thermodynamic (density) and transport (viscosity) properties oil-water emulsions. In the present work we used stable W/O emulsions with various water contents. Oil emulsions were prepared as a model system on the bases of Ashalchinskaya heavy oil which well-known composition and structure (see Sect. 2.1.2). Water-in-oil emulsions were prepared using the following procedure. The defined volume of water of the required volume concentration was gradually added to the original dry Ashalchinskaya heavy crude oil at the constant mixing at 2000 rpm for 20 min using an US 2000 A. The aqueous phase volume fraction ( ) used was 10, 20, and 30 (v/v) %. It is well-known (see, for example, Ref. [25]) that the rheological properties of W/O emulsions and their stability are affected by the volume fraction, and droplet structure of the dispersed phase, and chemical composition of each phase. Emulsions were prepared at the room temperature (25 °C). The natural surfactants (i.e., asphaltenes, resins, naphthenic acids, and the total mass of volatile aromatic components, etc.), which are contained in crude Emulsions were analyzed at 25.00 °C after preparation. An optical microscope Levenhuk 640 T is used to obtain visual images (morphology and dispersion of emulsions) of the prepared emulsion samples immediately after emulsification at 64× amplification. The digital images were analyzed by Image-Pro Plus 4.5 software in order to obtain mean droplet size. Droplet size distributions of the emulsion samples were measured using a 23 MHz low-field nuclear magnetic resonance (NMR). The drop size strongly affects the stability of the emulsion against coalescence [61]. The average droplet size observed at room temperature is within (8 to 10) μm. The higher the dispersity of emulsions, the more they are stable. The emulsion prepared based on Ashalchinskii heavy oil form quite highly stable emulsions with water. It was proved that emulsions with the mean droplet diameter of 8 µm to 10 µm stay stable for one year without coalescence and phase separation. Figure 1 shows micro photos of W/O emulsions studied. The obtained W/O emulsion, which show crushed droplets of the aqueous phase (see Fig. 1), i.e., large water drops are breaking into smaller drops and forming stable and homogeneous emulsion. Afterwards, the emulsion was taken for the measurements.

Density Measurements
In our previous several publications (Sagdeev et al. [97][98][99][100][101][102][103][104][105]) various versions of a new apparatus to simultaneously measure of the density and viscosity of liquids have been designed and constructed based on combination of both well-known hydrostatic weighing (for density) and falling-body (for viscosity) principles. In the work [105] we have designed a new combined experimental setup that implements the methods of a falling body and hydrostatic weighing to simultaneously measure of the dynamic viscosity and density of high-density liquids in the temperature range from (293 to 473) K and pressures from 0.098 MPa to 250 MPa, with an expanded uncertainty of 0.15 % and 2 %, respectively. In our recent work (Sagdeev et al. [106]) we have developed a new design of the densimeter based on the hydrostatic weighing principle for accurate and fast measurement of the high-density (extra viscous) working liquids (oils for diffusion vacuum pumps, ionic liquids) over the wide temperature range from (273 to 500) K at atmospheric pressure. The physical basis and theory of the method, the apparatus, the experimental procedures, and an uncertainty assessment of the density ( ) measurements have been detailed described in our earlier publication (Sagdeev et al. [106]). In this work we have modified conventional hydrostatic weighing method applied for ordinary liquids to use for extra highdensity fluids (extra viscous heavy oil), i.e., the method was adapted for high density crude oils (see also, Ref. [107]). In the present work the same method and apparatus have been employed for density measurements of the Ashalchinskaya crude heavy 7 Page 16 of 46 oil sample and W/O emulsions (10 v/v %, 20 v/v %, and 30 v/v % of water) over the wide temperature range at atmospheric pressure. Only a brief review and essential information will be given here. Density of the heavy oil samples and W/O emulsion as a function of temperature has been measured using modified hydrostatic weighing densimeter (HWD, Sagdee et al. [106]).
The densimeter is based on hydrostatic weighing principle. Most available HWD has problem to carry out measurements for extra viscous liquids due to too long float equilibrium time due to low speed of float motion. The new designed HWD [106] is overcoming the problem. The main advantage of the present new designed HWD [106] is that it allows one-push calibration, i.e., considerable simplifying the calibration procedure. Also, the method is simpler and faster (takes short measurement time) and at the same time accuracy of the density measurements is comparable with conventional techniques. The method applicable for high density and extra viscous liquids like Ashalchinskaya crude heavy oils. Schematic diagram of the HWD for the density measurements at atmospheric pressure was presented in our previous publication (Sagdeev et al. [106]). The final working equation for density measurement in this method was derived and detailed described in our several publications (Sagdeev et al. [97,98,100,101,105]). A density measuring system consisting of a ring, a wire and a float, made of a BT-6 grade titanium alloy is suspended to the lower hook of the balance. The float with a diameter of 10 mm and length of 100 mm was used. At the top, the float has a tab for hanging on a pre-annealed constantan wire with a diameter of 0.15 mm and a length of 200 mm. A ring with a diameter of 25 mm, located above and connected by a constantan wire with a titanium float, disengages from lifting the plate. This allows the electronic balance to be adjusted to zero. The float is placed in a cylindrical vessel with an internal diameter of 25 mm, which, in turn, is inserted into the copper block placed in the heatexchanger through the channels the thermostating liquid is pumped. The thermostating was carried out with a liquid polymethylsiloxane (PMS-20), which came from the ultra-thermostat (U-10) with an accuracy of ± 0.02 K. To control the temperature in the copper block, copper-constantan thermocouples, placed in the copper block of the thermostatic system, were used.

Uncertainty Assessment of the Density Measurements Using HWD Technique
The details of the uncertainty evaluation were discussed in our previous publication (Sagdeev et al. [50,105]). The uncertainty assessment of the measured densities for Ashalchinskaya oil were was performed according the recommendations [108]. The experimental density data for crude oil samples were evaluated using working equation for the method [50,105]. Assuming that all of the input parameters (measured quantities) X i ( m a ,m L are the mass of ring, wire and float in air and liquid, respectively; m c 1 is the total mass of ring, wire and float at calibration of float; * L is the density of liquid under study obtained with pycnometer; a is the density of air; L is the density of liquid under study; T is the thermal expansion coefficient of float material at experimental temperature; V w is the volume of the wire made from constantan; V r is the volume of the aluminum ring; m is the mass of the suspended system; V is the volume of suspended system; V c f is the float volume at calibration; t is the experimental temperature; t 20 is the calibrating temperature; m 1 is the total mass of ring, wire and float; L,t is the density of the liquid under study at experimental temperature, see Table 7) are independent, the variance of density (ρ) is where N is the number of input parameters (measured quantities) in the working equation of the method [106] and the combined standard uncertainty is the square root of the variance (see Ref. [108]). The uncertainty of single density measurement is a function of uncertainties of the input parameters (measured quantities) entering the density evaluation procedure. The uncertainty evaluation of the single density measurements at selected temperature of 444 K as an example for Ashalchinskaya oil are presented in Table 7. All input parameters used to uncertainty analysis are given in Table 7. This table provides the value of each input parameter X i and their estimated standard uncertainties. Based on the data from Table 7 the total expanded uncertainty of the density measurement at 0.95 confidence level with a coverage factor of k = 2 is estimated to be 0.16 % including calibration effect.
In order to check the accuracy of the method, correct operation of the new designed HWD, and confirm the reliability of density data measured by the present method for the Ashalchinskaya crude heavy oil sample and W/O emulsions, the measurements were made on well-studied liquids such as pure n-heptane in the temperature range from (293 to 357) K at atmospheric pressure, ethane-1,2-diol and propane-1,2-diol from (273 to 473) K, and the hexadecane over a temperature range from (283 to 443) K. The measurements were made using two different techniques, namely, pycnometric (Kivilis [109]) and the present HWD method. We used standard pycnometer (PZH2-10-KSH 7/16) with volume of 10 ml. The pycnometer was thermostated using ultra-thermostat type of U-10 which provides temperature stability within 0.02 K. The results of both techniques were compared with the most reliable reported data (see, for example, Sagdeev et al. [97,103,105]) where comprehensive review all of the data sources from NIST/TRC/TDE Data Base (Frenkel et al. [108]) were provided for n-heptane and with the values calculated from  [110]). This means that the result obtained using the both methods are in good agreement and consistent with reference data (REF-PROP, Lemmon et al. [110]; Frenkel et al. [111]). Thus, this result is verifying the reliability of the method to accurate measure of the density of liquids and correct operating of the new designed HWD instrument used in the present work for Ashalchinskaya crude heavy oil and W/O emulsions. As was mentioned above, new developed HWD for measurement of the density high -viscosity and high-density liquids (IL, oils, etc.) has certain advantages over other existing densimeters, for example, constant-volume piezometer, VTD, pycnometer, etc. In particular, much cheaper in compare with VTD, easier to use, especially for high viscosity liquids (at low temperature measurements), faster (short measurement time for single data point), the results are the same accuracy, and wide temperature range of applicability (from 273 K to 500 K). Also, all VTD (Anton Paar) instruments are requiring viscosity correction to measured densities, if the sample under study of higher viscosities, i.e., for VTD the measured density of the liquid is a function of viscosity. Most cases there are no accurate viscosity data for the same sample. The most widely used densimeters in compare with the present HWD have very limited temperature range of applicability, for example, pycnometers-from room temperature to 368 K, and most VTD -usually from room temperature to 353 K. The present method and new HWD was successfully used to accurate measurements of the density various molecular liquids including high-density oils (Sagdeev et al. [106]) in the temperature range from (273 to 500) K.

Measurements of the Viscosity of Ashalchinskaya Heavy Oil and W/O Emulsions
There are very limited and reliable experimental data on viscosity of viscous heavy oils in a wide temperature range (see, for example our previous publications [88,107], Table 1, Sect. 1.1). To determine the dynamic viscosity as a function of temperature using the falling body technique, the density of the same oils is required as a function of temperature within the same range of temperature. A new designed falling-body viscometer (rheo-viscometer) for measurements of the dynamic viscosity of high and extra viscous liquids (Newtonian and non-Newtonian) over a wide temperature range (from 273 K to 500 K) has been developed in our previous publication [97,105,107,112]. The rheo-viscometer combines the advantages of both falling-body and capillary techniques of viscosity measurements (see details in Ref. [107]). The method based on the principle of measuring two coupling flow characteristics: (1) the stress on the piston and (2) the time required for liquid of volume V to flow through the capillary. Both characteristics are using to calculate shear stress on the capillary wall and gradient of velocity (including non-Newtonian liquids). Brief review of the method is given below. The main parts of the rheo-viscometer are: (1) the system of viscosity measurements; (2) the thermostating and electrical measuring system; (3) the system for body falling time registration; (4) rheo-viscometer loading system; and (5) balancing system. The system of viscosity measurements consists of (1) a measuring cup in which the measuring piston (see Fig. 2) falls coaxially. The measuring cup is made of titanium alloy BT-6 and has the following dimensions: length-160 mm, length of the measuring section-140 mm, ID-25 mm, OD-30 mm. The rheo-viscometer's pistons characteristics can be finding in our previous publication [107]. The piston has a cylindrical shape with straight channels along the piston, 75 mm long, with an OD of 24.8 mm and an ID of 10 mm to be placed inside the gate. The channels are made with high precision on a CNC gear milling machine of the Japanese company "HAMAI-120". This design allows the center of gravity of the measuring piston to be shifted for automatic centering in the measuring cup. The length of the shutter is 86 mm, the diameter is 6 mm. The shutter is needed to ensure the self-centering of the piston in the autoclave when it falls down or rises up when the piston moves. The modification of the rheo-viscometer was aimed at designing and manufacturing 11 measuring pistons with the number of teeth from 0 to 100 in increments of 10. The photographs in Fig. 2. The piston was based on a workpiece in the form of a bar with diameter of 25 mm of high accuracy and quality and made of titanium grade BT-6. The details of the design of the rheo-viscometer are given in our previous publication [105]. The viscosity was measured using coaxial cylinder (pistons) with 7 Page 20 of 46 various geometry (see Fig. 2). Auxiliary units of the installation are: (1) a system for setting the instrument for verticality and horizontality; (2) centering system; (3) system of vertical movement of measuring pistons; (4) piston position tracking system; (5) launch system; and (6) waxed cord. The modification of the experimental viscosity apparatus was aimed at improving the system for measuring viscosity in order to expand the range of measurements of the dynamic viscosity of liquids [113]. The temperature control system (thermostating) allows to maintain a temperature in the range from (293 to 473) K with an accuracy of ± 0.02 K, by supplying a thermostatic liquid (polymethylsiloxane, grade ПMC-20) to a heat exchanger. Two optical sensors CG-FD, located at a distance of 30 mm were designed to measure of the piston immersion time into the liquid under study. The shutter located on the thread has a slot that allows the photosensor to record and measure the piston falling time. The working equation for the method has been developed in our previous publication [107] where the values of measuring cell parameters a i (i = 0, 2) were determined using the calibration procedure using various liquids with the well-known viscosities (see Table 8); d e = 2.91 778 × 10 -4 m is the equivalent diameter has been determined using calibrating procedure [107]; w b is the falling body velocity, is the falling time, L is the density of oil, D is the inner diameter of the measuring tube). The purpose of the calibration procedure is to determine the viscometer constants, which are allow taking into account the change in the geometric dimensions of the measuring unit and the hydrodynamic condition during measuring the dynamic viscosity of the oil in a wide range of temperature and pressure changes. Calibration is reduced to measuring the time of falling body in the measuring cell filled with liquid under study (oil) at temperature T = 293 K and atmospheric pressure P = 0.101 MPa. Four different calibration procedures to determine the calibration constant have been developed in our previous paper [107]. To calibrate the viscometer, liquids were used, the viscosity of which varied from (20 to 1166) mPa·s (see Table 8).
The dependence of oil viscosity on the velocity gradient becomes noticeable at temperatures (293 to 296) K above the pour point. At a temperature close to the pour  Fig. 2 Measuring pistons point, oil acquires a static shear stress, which increases with temperature decreases. As the temperature rises, the rheological properties normalize, i.e., oil can turn into a Newtonian fluid due to the fact that the solid paraffins of oil begin to melt, the crystal lattice is destroyed and the viscosity decreases. In this case, the stability of the emulsions deteriorates, droplets coalesce and the emulsion breaks down, which is explained by a decrease in the mechanical strength of the adsorption shells. This method, in comparison with the capillary, falling and rolling ball methods as well as rotational [114], and oscillatory [115][116][117] methods, has a several of advantages. For example, (1) allows to create a compact setup with a small measuring cell, which requires a small amount of the sample; (2) negligible small thermal inertia of the device which allows to the rapid reach the stationary mode; (3) the laminar flow of liquid through the cylindrical annular gap is provided by the selection of a falling body required size; and (4) the present falling-body rheo-viscometer has wider temperature (from room to 473 K) and pressure ranges of applicability (up to 250 MPa). In this method, measurements of the dynamic viscosity are required density of the samples (oil and W/O suspensions) at the experimental conditions. For this purpose, to determine the density of oil and W/O suspensions, the method of hydrostatic weighing described above (see Sect. 2.2) has been used (see also details in Ref. [50,118]) which allows measurements over the wide temperature range.

Uncertainty Assessment of the Dynamic Viscosity Measurements Using Falling Body Method (Rheo-Viscometer)
The uncertainty of dynamic viscosity measurements for Ashalchinskaya heavy oil has been evaluated using the same procedure as for density described above (see Sect. 2.2.1) according to the recommendations [111]. Example of the uncertainty estimation for single viscosity measurement of the Ashalchinskaya oil at the selected temperature is presented in Table 9.
,T ), where d e is the equivalent diameter has been determined using calibrating procedure 105 , D is the tube diameter, T is the temperature, is the sensitivity of the measured values of viscosity to the measuring parameters X i , can be calculated from  Table 9). The combined standard uncertainty is the square root of the variance (see Ref. [108]). As one can see from Table 9, the expanded uncertainty of the dynamic viscosity measurements of the Ashalchinskaya crude oil using the falling body method (Rheo-viscometer [107]) at 0.95 confidence level with a coverage factor of k = 2 is estimated to be 1.9 %. This is considerable improvement of the conventional falling body technique which typical uncertainty is within (2 to 4) %.

Density of Ashalchinskaya Heavy Oil and W/O Suspensions-Effect of Temperature and Water Concentration
Measurements In order to confirm the reliability and accuracy of the measured density data for Ashalchinskaya heavy oil and W/O suspensions the measurements were performed also using pycnometeric method (PZh-2-10-KSh7/16 GOST 22524-77) over the temperature range from (293 to 365) K at atmospheric pressure (see Table 10). The high-speed weighing (with 2-s stabilization) and analytical electronic balance HR-250AZG (A&D Co. LTD, Japan) has been used to weight of the pycnometers. The repeatability of  the electronic balance is within 0.1 mg in the mass range from (0 to 200) g. The agreement between the HWD and pycnometeric methods is within (0.03 to 0.14) % (Table 10). However, the temperature range of the pycnometric method applicability is very restricted and has some disadvantages in compare with HWD technique [50,88,106] (see, Sect. 2.2). As can be seen from Table 10 (see also Fig. 3), the data obtained by the HWD for the Ashalchinskii heavy oil are systematically lower than the data obtained by the pycnometer method by about 0.9 %, which is probably due to the removal of moisture during heating of the oil sample. The pycnometer method reduces moisture removal due to the presence of a stopper on the pycnometer. The moisture content in the Ashalchinskii heavy oil sample is about 0.5 wt %. As one can see from  Due to the presence of water in W/O suspensions and large gas evolution, the density measurements were made in the temperature range below 373 K at atmospheric pressure. At high temperatures above 367 K, we found that heating of the W/O emulsions during the measurements causes the release of gases. We experimentally observed that above 367 K the intensity of the gas release is increasing. At moderate temperatures (below 367 K) the effect of gas release on the measured properties is small and can be neglected.

Density Correlation
Because crude oil is a complex mixture of several hundred components, modeling its properties using mixture model with equations for all of the constituents in the mixture is not a practical solution. Modeling of the thermodynamic properties of   productions. Thus, the measured densities for Ashalchinskaya heavy oil have been fitted to quadratic temperature function.
The optimal values of the fitting parameters of Eq. 4 for dry Ashalchinskaya heavy oil and W/O emulsions are presented in Tables 12 and 13.
Calculated values of density for Ashalchinskaya heavy oil and W/O emulsions as a function of temperature are given in Figs. 3, 4 and 5 together with the measured values. AAD between the measured and calculated values of density is 0.03 % for dry oil and 0.005 %, 0.007 %, and 0.019 % for W/O emulsions, respectively.

Correlation of the Dynamic Viscosity of Ashalchinskaya Heavy Oil and W/O Suspensions-Effect of Temperature and Water Addition
Dynamic viscosity of Ashalchinskaya dry heavy oil (100 %), 90 % oil + 10 % water and 80 % oil + 20 % water W/O emulsions have been studied as a function of temperature and volume concentration of water over the temperature range from (293 to 366) K at atmospheric pressure. Measured values of dynamic viscosity of Ashalchinskaya heavy oil (100 %, dry oil) and W/O suspensions with the concentrations of 10 and 20 v/v % are given in Table 14 and depicted in Figs. 6, 7, 8, 9, 10, 11, 12 and 13 as a function of temperature and water v/v % concentration (in variuous projections, η-T, η-ρ, η/η 0 -φ, η/η 0 -T, and ln η-T −1 ). As one can see from Figs. 6 and 13, viscosity of W/O emulsions is decreasing with the temperature increases and rapid increases by increasing the water concentration in oil (Fig. 10). As can be noted from Figs. 6, 7, 11 viscosity of dry crude heavy oil and its aqueous mixtures (Figs. 6, 13) decreases with temperature due to the weakening of the intermolecular interactions. As Figs. 6 and 13 shows, the rate of the temperature decreases strongly depends on the temperature range. As one can see, the measured values of viscosity decrease rapidly as the temperature initially increases above ambient temperature (298 K). The viscosity of the Ashalchinskaya heavy oil sample changes (decreases) from 1339 mPa·s at a temperature of 302 K, to 36 mPa·s at high temperatures 366 K. At high temperatures, for almost all liquids, the viscosity asymptotically approaches a high temperature limit (becomes constant). For Ashalchinskaya heavy oil, the viscosity decreases markedly between room temperature and 320 K. Above about 320 K, the values of viscosity change very slightly. As can be note from Figs. 6 and 10, temperature and water volume fraction have important impacts on the viscosity of W/O emulsions. For example, between temperatures (293 and (4) (T) = a 0 + a 1 ⋅ T + a 2 ⋅ T 2 . 333) K the viscosity of Ashalchinskaya crude heavy oil changes by factor of 20, while at high temperatures (above 333 K) the viscosity changes by factor of 2. The temperature effect on viscosity strongly depends also on the API gravity of oils. At low API (below 20) the rate of viscosity changes is very high (changes by factor 40), while at high API gravity the small changes (by factor 4) of viscosity with temperature is observed [88]. The low-(T → 0 ) and high-(T → ∞) temperature limits of viscosity behavior of liquids are well-known. The low temperature asymptote (low temperature limit, T → 0 ) is η −1 → 0, while the high temperature asymptote (high temperature limit, T → ∞) is η −1 → η ∞ −1 constant). As Figs. 6 and 11 show, the temperature rate, (∂η/∂T), of the viscosity of the Ashalchinskaya crude heavy oil and W/O emulsions decreases with temperature until η asymptotically approaches a constant value (4 to 7 mPa/s/K) at high temperatures (approximately above 365 K). The rate of viscosity decreases for the dry Ashalchinskaya crude heavy oil changes from (− 377.8 to − 4.3) mPa/s/K and W/O emulsions from (− 616.8 to − 6.1) mPa/s/K for 10 v/v % and (− 745.2 to − 7.0) mPa/s/K for 10 v/v % with increasing temperature from room temperature to 365 K. Addition of water into oil increase the rate of the viscosity changes almost two times at concentration of 20 v/v %. Figure 7 shows temperature dependence of the Ashalchinskaya heavy oil as a function of temperature calculated from VTF model (Eq. 5) together with the values for light and medium oils calculated from the same equation in our previous publication [88]. The qualitative behavior of the temperature dependence of the viscosity is the same. The present measured densities for the Ashalchinskaya heavy oil (see Sect. 3.1) have been used to study of the density dependence of the dynamic  viscosity. The density dependence of the viscosity data of the present Ashalchinskaya heavy oil and for the light and medium oils from the previous studies [88] is shown in Fig. 8. As Fig. 6 shows, when increasing the temperature, it is observed a decrease in viscosity of W/O emulsions caused by a decrease in continuous phase (oil) viscosity, i.e., temperature dependence of the W/O emulsions illustrates the same behavior as pure dry oil. Results of the present experiments (see Fig. 10) showed  that viscosity of W/O emulsions noticeably increases with the water content (φ, v/v %) as result to increase the number and volume of dispersed water droplets in the continuous phase (oil) (see also Ref. [49]). When the temperature increases, it is observed a decrease in viscosity of emulsion caused by a decrease in oil viscosity (bulk viscosity, continuous phase viscosity increases). Figure 10 Fig. 10, at high temperatures (above 293 K) in the initial concentration range the rapid increases of the viscosity with concentration of water is observed. However, at low temperatures (below 293 K, low shear rate region, where W/O emulsion exhibits a non-Newtonian behavior) almost linear concentration dependence of viscosity has been observed (see Fig. 10). The present density (Table 11) and dynamic viscosity (Table 14) data for Ashalchinskaya dry oil and W/O emulsions can be used to calculate the kinematic viscosity = ∕ data for the same oil samples.

Viscosity Correlation-Temperature Dependence
There are many empirical and semiempirical correlation and prediction models to accurately represent the viscosity of crude oils (see, for example, Messaâdi et al. [120]; Ben Haj-Kacem et al. [121]; Petrosky and Farshad [122], see also [123][124][125]). Modified linear Arrhenius-Andrade model, proposed by Vogel-Tamman-Fulcher [126][127][128] for accurate representation low temperature behavior of viscosity Ashalchinskaya dry oil and W/O emulsions, has been used in the present work. The measured values of dynamic viscosity of Ashalchinskaya dry oil and W/O emulsions (10 and 20 v/v % of water) were fitted to Vogel-Tammann-Fulcher (VTF) correlation model [126][127][128] where ln ∞ , B , and T 0 are the VTF parameters. Derived optimal values of the VTF parameters ( ln ∞ , B , and T 0 ) are given in Table 15 together with AAD(%) = 100 , where exp is the present experimental values of viscosity and cal is the values calculated from the VTF model (Eq. 5). As one can see VTF model represents the temperature dependence of the measured viscosities for dry oil and W/O emulsions within the experimental uncertainty (AAD = 1.0 % to 1.6 %). The physically meaningingful parameters, B and T 0 , of the VTF model Eq. 5 were used to estimate the glass transition temperature,T g , of the oil and W/O emulsions based on Angell relation where log 10 g ∕ 0 = 17 is constant, = B∕T 0 (see Table 15). The derived from the present experimental viscosity data values of glass transition temperature,T g , for the Ashalchinskaya heavy oil and W/O emulsions are given in Table 15.
Deviation statistics between the measured and calculated from VTF Eq  [129], and aqueous salt solutions [130,131]. This model was also used by many other authors (see, for example [89,[132][133][134][135][136] to accurate represent measured viscosities of various molecular liquids and IL. Kolotova et al. [37] used the same VTF model to represent experimental viscosity data of W/O emulsions. Figure 11 shows measured values of the viscosity of Ashalchinskaya heavy oil together with the values calculated from VTF model (Eq. 5) and original linear Arrhenius-Andrade relation [123,124,[137][138][139][140][141][142][143][144] where b 0 is the viscosity (mPa·s), = ∞ , at the high temperature limit ( T → ∞ , i.e., viscosity of the system in vapor state);b 1 = a ∕R (viscosity Arrhenius energy or viscosity flow activation energy) is the slope of the Arrhenius plot ln vs T −1 , and a = ΔH are the flow activation energy (enthalpy of activation, related with the enthalpy of vaporization), where T is in K. The flow activation energy a and parameter ln b 0 can be directly calculated from the slope (see Fig. 12) and intersect of the straight line by the Arrhenius relationship function (Eq. 7). The intercepts and slopes of the linear plot ( ln vs T −1 ) are flow activation energy E a and parameter ln b 0 , T g T 0 = 1 + 2.303 log 10 g ∕ 0 , respectively (see Fig. 12). Derived from the present viscosity data optimal values of the Arrhenius parameters for Ashalchinskaya heavy oil are b 1 = ε a /R = 6857.00 and ln b 0 = −15.52 731. Linear Arrhenius-Andrade model Eq. 7 shows large deviations from measured viscosity data for Ashalchinskaya heavy oil (AAD = 6 %), especially at low temperatures, although most pure liquids and liquid mixtures are obeyed to the linear Arrhenius behavior (Eq. 7) of viscosity. Therefore, the original linear Arrhenius Eq. 7 is failed to accurate represent measured viscosities of Ashalchinskaya heavy oil in the low temperature range (near the glass temperature) where rapid increases of the viscosity are observing. Some liquids and liquid mixtures the experimental curve ln vs T −1 considerable deviates from the original linear Arrhenius behavior Eq. 7, especially at low temperatures (near the glass temperature, T g ) where rapid increases of the viscosity are observing. Ilyin et al. [84] also found that for some light and heavy crude oils ln vs T −1 relation deviate from linear Arrhenius law in the low temperature range, where non-Newtonian behavior of the viscosity is observing. As one can see from Fig. 11, the extrapolation property of the VTF model Eq. 5 is much better than linear Arrhenius-Andrade Eq. 7, especially at low temperatures where rapid changes of viscosity is observing. Therefore, as Fig. 11 shows, the original linear Arrhenius-Andrade model Eq. 7 cannot accurately represent rapid increases of viscosity at low temperatures. VTF model can be reasonable extrapolated to low and high temperatures (out of the present experimental temperature range). As was illustrate in our previous publication [88], extrapolation of the VTF model (Eq. 5) to low temperature range (below 293 K), where there are no experimental data, predicts the measured viscosity data within (2 to 7) %. Thus, Eq. 5 can be recommended to accurately represent experimental viscosity data for Ashalchinskaya crude heavy oil and W/O emulsions in the wide temperature range (from 302 K to 366 K), especially at low temperature range where rapid changes of viscosity is observing.
The present viscosity data for Ashalchinskaya crude heavy oil has been also fitted to the Masuko-Magill model (Masuko and Magill [125]) The derived optimal values of the fitting Masuko-Magill model parameters are ln g = 28.6313, A = 32.11 846,B = 3.48 498 and T g = 206.5 for Ashalchinskaya crude heavy oil. The accuracy of the present viscosity data representation by Masuko-Magill model Eq. 8 is slightly better than VTF model Eq. 5, AAD = 1.26 % and AAD = 1.56 %, respectively. However, Masuko-Magill model Eq. 8 contain 4 adjustable parameters, while VTF model Eq. 5 just 3 fitting parameters, therefore VTF model has more reliable extrapolation properties.
The measured values of the viscosity were used to calculate the temperature coefficient of viscosity, ( ln ∕ T) API , for Ashalchinskaya crude heavy oil, which directly related with the temperature rate of viscosity changes. The temperature coefficient of viscosity for Ashalchinskaya crude heavy oil and W/O emulsions changes is within (− 0.112 to − 0.05) K −1 , depending on temperature range. The values of temperature coefficient of viscosity, ( ln ∕ T) API , slightly increases (by 5 % at low temperatures and 4.2 % at high temperatures) with a concentration of water v/v % increases. The strong temperature dependence of dynamic viscosity at low temperatures can be explained due to the formation of stable association structures originated from intermolecular interactions of resin and asphaltene. As temperature increases, the association interactions become weaker, resulting in abrupt decrease in viscosity and less significant temperature dependence.

Viscosity Correlation-Water Concentration Dependence for W/O Emulsions
Viscosity equations for emulsions have been reviewed by Pal [38] and Farah et al. [39]. The W/O emulsion viscosity ( ) is directly related with the viscosity of its continuous phase ( 0 , dry oil, 100 %). As Fig. 6  The present relative viscosity (see Fig. 9), r = 0 , of W/O emulsion were fitted to the model proposed by Krieger and Dougherty (KD equation) [57] where is a volume fraction of water, m is a theoretically meaningful parameter (is the maximum packing concentration of dispersed phase, depends on size of the drops), namely, the limit of the closest packing of drops in the space and has a clear physical meaning in considering a possible structure of the suspension, B * = B m , (9) r = (1 + 2.5 ),  [58]. Krieger and Dougherty model [57] is valid for high concentrations of the dispersed phase, when the viscosity of suspension becomes infinite. In the present work the values of B * and m in Eq. 11 were considered as fitting parameters. At low volume fractions ( → 0) Krieger and Dougherty Eq. 11 reduces to Einstein Eq. 9. The derived values of m and B * as a function of temperature are given in Table 16. As one can see from Table 16, the value of the parameter m = 0.6 is in good agreement with a value for random close-packing of monodisperse particles obtained by numerical simulations [146] as well as the experimental data for various real emulsions [37,38,147].
The present viscosity data for W/O emulsion have been also fitted to the Mooney equation [148] The optimal values of the fitting parameters B m and k m derived from the present viscosity data of W/O emulsions for various isotherms are given in Table 17. Benayoune et al. [40] reported that the best fit was obtained at a geometric crowing factor is k m = 0.705 and B m = 2.5. The most values of k m reported in the literature are within 1.35 to 1.91. However, as one can see from Table 17, both Mooney parameters B m and k m are function of temperature. The present result for parameter k m for various temperatures lies within the same range reported in the literature. Slightly modified form of the Eq. 12 has been used by Saito et al. [149] and Kashefi et al. [150].

Measurements of the Dynamic Viscosity of Ashalchinskaya Oil and W/O Emulsions Based on Brookfield Rotational Viscometer (DV-II + PRO)
In order to verify the reliability and accuracy of the measured values of dynamic viscosity of Ashalchinskaya crude heavy oil and W/O emulsions, the test measurements were performed using commercial cone and plate digital Brookfield rotational viscometer (model DV-II + PRO, LVD-II + PRO) coupled to a microcomputer. Programmable viscometer DV-II + PRO (Brookfield) is designed to measure the viscosity of liquids at given shear rates. Brookfield viscometers employ the principle of rotational viscometry (see our previous publication [88]). Rotational viscometers use the idea that the torque required to turn an object, such as a spindle, in a fluid, can indicate the viscosity of that fluid. The common Brookfield-type viscometer determines the required torque for rotating a disk or bob in a fluid at known speed. Cup and bob viscometers work by defining the exact volume of sample which is to be sheared within a test cell; the torque required to achieve a certain rotational speed is measured and plotted. The motor drives the measuring bob and the sample cup stands still. The viscosity is proportional to the motor torque that is required for turning the measuring bob against the fluid's viscous forces. This is called the Searle principle. Cone and plate viscometers use a cone of very shallow angle in bare contact with a flat plate. With this system the shear rate beneath the plate is constant to a modest degree of precision and deconvolution of a flow curve; a graph of shear stress (torque) against shear rate (angular velocity) yields the viscosity in a straightforward manner. The details regarding the operational system can be found elsewhere (BROOKFIELD DV-II + Pro Viscometer [151], visit also ROOKFIELD ENGINEERING LABORA-TORIES, INC). The uncertainty of the viscosity measurements with Brookfield rotational viscometer is within 1 %. This type of viscometer has been successfully used for years in various applications including oil industry for crude oil viscosity measurements [25,73,88]. Viscosity ranges for the viscometer are (for rotational speeds of 0.1 rpm through 200 rpm) within from (15 × 10 6 cP to 6 × 10 6 cP). For this method the viscosity is defined as (Pa ⋅ s) = ṡ , where ̇ s −1 is the share rate and s N∕m 2 is shear stress. The shear rate ̇ s −1 of a given measurement is determined by the rotational speed of the spindle, the size and shape of the spindle, and the container used, therefore the distance between the container wall and spindle surface. For Newtonian fluids have the same viscosity at different shear rates. The viscometer allows determining the viscosity , , shear rate ̇ and shear stress s values. The accuracy of the DV-II + Pro is verified using viscosity of standard fluids which are available from Brookfield Engineering Laboratories. Viscosity standards, calibrated at 298.15 K, for silicone and mineral oils (from 5 cP to 5 000 cP, 10 cP to 12 500 cP, 50 cP to 30 000 cP, and 100 cP to 60 000 cP). Temperature control is within 0.1 K. This type of viscometer is ideal for the measurements of the viscosity for liquids such as heavy crude oils.
The measured viscosity data for Ashalchinskaya heavy oil (100 %, dry oil) and W/O suspensions using Brookfield viscometer as a function of temperature between (293 and 343) K are given in Table 18. As one can see from Table 18, shear rate values during the viscosity measurements were varied from (0.792 to 39.60)/s for pure oil (100 %, dry oil) and from (0.792 to 15.80)/s for W/O emulsions. Figure 13 shows the comparison of the measured dynamic viscosities of Ashalchinskaya oil (100 %, dry oil) and W/O suspensions as a function of temperature derived using Brookfield viscometer (Table 18) and falling-body technique ( Table 14). As one can see from Fig. 13, the agreement between the both Brookfield and falling-body methods is good enough, deviations are within 6 % for pure dry Ashalchinskaya oil, 15 % for W/O emulsion (10 v/v %), and 5 % for W/O emulsion (v/v 20 %).

Conclusions
A new designed HWD and falling-body viscometer have been employed to study of the effect of temperature and volume fraction of water on density and dynamic viscosity of Ashalchinskaya crude heavy oil from Tatarstan Oil Filed (Russia) and (2) easier to use and much faster (short measuring time) than conventional techniques, especially for high viscosity liquids; (3) wide temperature range of applicability (from 273 K to 500 K); (4) no filling and sampling problem for high-viscosity liquids; and (5) high accuracy (0.16 %). Viscosity of the same Ashalchinskaya heavy crude oil and W/O emulsions were measured using commercial Brookfield rotational viscometer (DV-II + PRO, LVD-II + PRO) with an uncertainty of 1 %. The various correlation equations for describing viscosity as a function of temperature (VFT, Masuko-Magill, Arrhenius type models) and dispersed phase volume fraction (Krieger-Dougherty and Mooney models) are developed based on the present measured viscosity data. The proposed equations give good correlation (deviations within 1.0 % to 1.6 %) between the measured viscosities of W/O emulsions as a function of temperature and the volume fraction of water. Krieger-Dougherty parameter m can be considered as a constant of 0.6 (independent on temperature) without losing the accuracy of the representation of the measured viscosity data. It was illustrated that the measured temperature behavior of the viscosity of Ashalchinskaya heavy crude oil and W/O emulsions can be best represented by VFT and Masuko-Magill models with three and four fitting parameters, respectively. The original linear Arrhenius model is failed to accurate represent measured viscosities for Ashalchinskaya heavy oil in the low temperature range (near the glass temperature) where rapid increase of the viscosity is observing. The rate of viscosity decreases for the dry Ashalchinskaya crude heavy oil changes from (− 377.8 to − 4.3) mPa/s/K and W/O emulsions from (− 616.8 to − 6.1) mPa/ s/K for 10 v/v % and (− 745.2 to − 7.0) mPa/s/K for 10 v/v % with increasing temperature from room temperature to 365 K. Addition of water into oil increase the rate of the viscosity changes almost two times at concentration of 20 v/v %.