2.2.1. Viscosity Measurements

The kinematic viscosity measurements were performed in an Ostwald-Cannon-Fenske capillary viscometer (Proton Routine Viscometer 33200, size 150) at 40 ◦C, by determining the flow time (t), expressed in seconds, required for a certain volume of liquid to pass under gravity between two marked points on the instrument, placed in an upright position. The kinematic viscosity (υ) expressed in centistokes (cSt) is obtained from the equation υ = C·t, where C is the calibration constant of the measurement system, supplied by the manufacturer (0.037150 (mm<sup>2</sup>/s)/s = cSt at 40 ◦C) [28,33]. The procedure for the determination of the kinematic viscosity meets with the specifications established by the European standard (EN 590 ISO 3104). All the viscosities values reported here are the media of three determinations.

#### 2.2.2. Determination of Pour Point and Cloud Point

Cloud Point (EN 23015 and ASTM D-2500) and Pour Point (ASTM D-97) were determined according to specifications required by standard methods. Firstly, the double or triple blends, of di fferent compositions, were introduced in a flat-bottomed glass tube. The tube was tightly closed with the help of a cork carrying a thermometer with a temperature measuring in the range of −36 to 120 ◦C. The tube was introduced in a digitally controlled temperature refrigerator for twenty-four hours; the tubes were brought out from time to time and checked until the oil did not show any movement when the tube was horizontally tilted for 5 s. After this time, the loss of transparency of the solutions is evaluated. The appearance of turbidity in the samples is indicative that the CP temperature has been reached. After a progressive decrease in temperature, the samples are kept under observation until they stop flowing (PP) [31,32]. All values are the media of duplicate determinations.

2.2.3. Mechanical and Environmental Characterization of a Diesel Engine Electric Generator Fuelled with Di fferent Biofuel Double and Triple Blends

Following the experimental methodology previously described [32], the energy performance and pollutant emissions generated in a C.I. diesel engine, has been carried out, working at a rate of 3000 rpm coupled to an AYERBE AY4000MN electric generator with a power of 5 KVA 230 V, for the generation of electricity, operating under di fferent degrees of demand for electrical power. This is

achieved by connecting heating plates of 1000 watts each (Figure 1a). The diesel engine operated at a constant rate of rotation of the crankshaft and torque so that the di fferent values of electrical power obtained are an exact consequence of the mechanical power obtained after the combustion of the corresponding biofuel. Tests were carried out by providing to the engine double and triple blends of di fferent biofuels. The electrical power generated can be easily determined from the product of the potential di fference (or voltage) and the electric current intensity (or amperage), Equation 3, both obtained utilizing a voltmeter-ammeter, Figure 1b:

$$\text{Electrical power generated (watts)} = \text{voltage (volts)} \times \text{amperage (amps)}\tag{3}$$

The consumption of the diesel engine, fueled with the di fferent biofuels studied, was calculated estimating the speed of consumption of the engine when it operates under a determined demand of electric power (1, 3 or 5 kW). Thus, the operation times are achieved by operating under the same fuel volume (0.5 L).

The contamination degree is evaluated from the opacity of the smoke generated in the combustion process, which is measured by smoke opacity meters. The smoke opacity meters are instruments capable of measuring the optical properties of diesel exhaust. These instruments have been designed to quantify the visible black smoke emission making use of physical phenomena like the extinction of a light beam by scattering and absorption. There are two groups of instruments: opacity meters, which evaluate smoke in the exhaust gas, and smoke number meters, which optically evaluate soot collected on paper filters. The density gauge is a handheld instrument for determining the filter smoke number (FSN), the Bosch number, and the soot concentration of diesel engines. This instrument is composed of an optical sensor (photodiode) and a di fferential pressure sensor. The photodiode calculates the paper blackening based on the reflected light intensity by a white LED. The more soot is deposited on the filter paper, the less light is reflected. The probe volume determined by the di fferential pressure sensor is used to calculate the probe volume under reference conditions with the input height and the temperature measured by the instrument. This probe volume and the measured paper blackening are then used to calculate the FSN (filter smoke number), soot concentration (mg/m3) or Bosch smoke number. Herein, the exhaust emissions were measured by a Bosch smoke meter or opacimeter-type smoke tester TESTO 338 density gauge, following the EU Directive 2004/108/EC, at the operating conditions previously reported [32], Figure 1c. The Bosch number is a standardized unit which is calculated from the level of soot on the paper (e ffective filter loading) [48]. The instrument evaluates smoke density on a scale from 0 to 2.5, where the value 0 represents total clarity on the paper and 2.5 is the value corresponding to 100% cloudy, as established by ASTM D 2156-94, Standard Test Method for Smoke Density in Flue Gases from Burning Distillate Fuels.

**Figure 1.** (**a**) Electrogenerator AYERBE AY4000MN, 5 KVA, 230 V connected to heating plates of 1000 watts of power each (**b**) voltmeter-ammeter devise; (**c**) TESTO 338 smoke density tester.

The data here compiled are the media of three repeated measures, attaining an experimental error lower than 9%. The results obtained with the biofuels evaluated were compared with the measurements obtained when conventional diesel was fueled.

## **3. Results and Discussion**

#### *3.1. Properties of DEE*/*Oil Double Blends, and Diesel*/*DEE*/*Oil Triple Blends*

The high viscosity values that the vegetable oils exhibit, 10–20 times greater than fossil diesel fuel, prevents its use like biofuels in conventional diesel engines. Most of them have viscosity values in the range of 30–45 cSt, concretely, castor oil has a much higher value, 226.2 cSt, very superior to values required by European standard EN 590 ISO 3104. Therefore, to achieve adequate viscosity values, the proportion in which the oils must be mixed with DEE has been investigated.

The viscosity values of the DEE/SVOs double blends are shown in Figure 2. As expected, an increase in the DEE content in the blends contributes to decreasing the viscosity values. It is interesting to confirm how DEE can promote very strong action on castor oil, in comparison with sunflower oil, in terms of reducing the viscosity of their blends, since castor oil has initially a much higher viscosity than sunflower oil. As can be observed, only 20% of DEE reduces considerably the kinematic viscosity of both oils, from 226.2 to 17.1 cSt in the case of castor oil and from 37.8 to 13.7 cSt when DEE is added to sunflower oil. For both oils, the viscosity values required by UNE EN 14214 ISO 3104, in the range from 2.0 to 4.5 mm<sup>2</sup>/s, were achieved with an analogous DEE/vegetable oil ratio (45/55).

**Figure 2.** Kinematic viscosity values at 40 ◦C of different blends of DEE/SVOs.

According to the previous study of the kinematic viscosity in double blends, the optimum and most favorable blend ratio was found to be DEE/SVO 45/55 (% by volume), independently of the oil employed. Hence, different proportions of biofuel containing 45% of DEE were mixed with fossil diesel to obtain the diesel/DEE/oil triple blends. The kinematic viscosity, cloud point, pour point, flash point and calorific values of the investigated triple blends are collected in Tables 2 and 3. A higher amount of biofuel in the blend, from B0 to B100, generates higher viscosity values, as expected since the viscosity of the added biofuels is slightly higher than that of diesel (3.20 cSt). Kinematic viscosity values are in the range of 3.20−4.25 cSt, so these biofuels present suitable values for being employed in diesel engines, complying with the European regulations EN 590, which establishes that viscosity at 40 ◦C must be in the range of 2.0–4.5 cSt.

**Table 2.** Physicochemical properties (kinematic viscosity at 40 ◦C, cloud point, pour point, flash point, and calorific values) of diesel/DEE/sunflower oil triple blends, obtained by adding different proportions of fossil diesel to the DEE/sunflower oil double blend containing 45% diethyl ether. All values are calculated as the average of three measurements.


\* The flash point values and calorific values were calculated by using Equations 1 and 2, respectively.

**Table 3.** Physicochemical properties (kinematic viscosity at 40 ◦C, cloud point, pour point, flash point, and calorific values) of diesel/DEE/castor oil triple blends, obtained by adding different proportions of fossil diesel to the DEE/castor oil double blend containing 45% diethyl ether. All values are calculated as the average of three measurements.


\* The flash point values and calorific values were calculated by using Equations 1 and 2, respectively.

Regarding the cold flow properties, the DEE has an important effect in the cloud and pour point of the triple blends, independently of the vegetable oil employed. In fact, it is found that only 9% of DEE (B20 blend) is enough to reduce the PP value 6 ◦C (from −6 to −15 ◦C) and the CP value 10 ◦C (from −6 to −16 ◦C). The best CP and PP values were obtained adding 18% of DEE, with 22% of castor oil and 60% of diesel, resulting in a CP of −16.8 ◦C and a PP of −22.8 ◦C. That is, the DEE promotes a significant improvement, with regard to fossil diesel, on the cold flow properties in all the blends studied, maintaining the appropriate viscosity to be used as a biofuel in conventional diesel engines. Additionally, the use of DEE overcomes one of the major challenges when using biodiesel as an alternative to fossil diesel in current engines, which is its poor cold flow properties.

Tables 2 and 3 show the calorific values of triple blends containing sunflower and castor oil, respectively. The calorific values decreased as the percentage of diethyl ether in the blend increased. As can be observed, there is no notable difference in the respective values for blends with either sunflower or castor oil, since these oils have similar calorific power. The results show that the B20 triple blends exhibited the highest calorific value, 41.6 MJ/kg in the case of sunflower oil and 41.4 MJ/kg for castor oil. In addition, an inverse correlation between calorific value and kinematic viscosity can be observed, since the calorific value of the biofuels increases as the kinematic viscosity decreases.

The results of the flash point of the analyzed mixtures (Tables 2 and 3) show an increment of this value as the DEE/oil ratio is greater, with both sunflower and castor oil. As it can be seen, the incorporation of SVOs in mixtures allows the FPs values to improve, since these oils exhibit a higher FP than both diesel and DEE. The highest FP was presented by B100 blends containing castor oil (132.6 ◦C) and sunflower oil (128.2 ◦C). The FP values are situated in the range 78.4−128.2 ◦C for diesel/DEE/sunflower oil blends, and 79.3−132.6 ◦C for diesel/DEE/castor oil blends. These values can be compared with FP requirements specified by EN 590 standard, which establishes that the fuel must have a minimum FP of 55 ◦C. In this case, each of the blends has FPs above the required value, so they are compliant with the requirements. Additionally, these biofuels do not have too high a

flash point that they can self−ignite thereby causing no safety problems during handling, storage or transportation, and they are recommended for use in CI engines.

#### *3.2. Mechanical Performance of the Diesel Engine*

To determine the optimal proportion of DEE that guarantees an adequate engine performance, the biofuel blends characterized in Tables 2 and 3 have been tested in a compression ignition engine. In this study, it was also included for comparative purposes, conventional diesel fuel as a reference. Hence, Figure 3 shows the power generated at different power demanded for by the triple blends employing either sunflower oil (3a) or castor oil (3b). In all cases, the power generated increased as the power supplied to the engine also increased from 1 to 3 kW and then, a stabilization of the power generated occurred, slightly decreasing when the maximum value of power supplied (5 kW) was reached. This behavior was improved using castor oil, where the power generated increased as the power supplied was also increased up to 4 kW.

**Figure 3.** Power generated (in Watts) based on the power demanded (in Watts) by the triple blends diesel/DEE/sunflower oil (**a**) or diesel/DEE/castor oil (**b**).

It is noteworthy the excellent results obtained with the B20 and B40 blends, achieving similar (diesel/DEE/sunflower oil) or even higher values (diesel/DEE/castor oil) of power generated than that obtained with fossil diesel at the highest values of demanded power (4 and 5 kW). Independently on the oil employed, a general trend was observed for B20 and B40 blends, where the rise of the DEE content caused the higher power generated. It is important to note that, despite obtaining a very slight increase in the power generated from B20 to B40 blends, when the proportion of DEE in triple mixtures increases above 25%, i.e., B60, B80, and B100 blends, the diesel C.I. engine did not work correctly. This behavior could be associated with the low calorific power of the DEE, confirming the previous results obtained with diesel/ethanol/oil [32].

## *3.3. Smoke Opacity Emissions*

Regarding the pollutants emission (Figure 4), the results obtained showed a significant reduction in smoke emissions in diesel engines, compared to those of conventional diesel, especially for B40 blends. This fact was even more evident with castor oil, achieving a reduction of 77% in the pollutant emission at the highest demand employed (5 kW). This important change is mainly due to the contribution of DEE to higher oxygen content in the blend, which improves the combustion and reduces the contamination. So, improved and complete combustion could be the reason for obtaining lower smoke

opacity values when the oxygenated additive is added to the blends. Additionally, as can be seen in Figure 4b, a slight but noticeable decrease in smoke emissions is observed when mixtures containing castor oil are employed, compared with the mixtures using sunflower oil in the same proportion (Figure 4a). In previous studies, it has been shown that the presence of unsaturations influences soot formation [49]. So, this may be explained by the differences in the fatty acid composition for the two vegetable oils used, since the linoleic acid present in sunflower oil has a greater number of double bonds than the ricinoleic acid present in castor oil.

**Figure 4.** Smoke opacity (Bosch number) generated as a function of the power demanded for different triple blends diesel/DEE/sunflower oil (**a**) or diesel/DEE/castor oil (**b**).

It should be noted that independently of the blend tested, from demanded powers of 1000 W onwards, all the blends performed better than fossil diesel in terms of achieving better combustion, which allows a reduction of emissions. The lowest smoke emissions are mainly obtained at medium and high demand (from 2000 W onward). Concretely, a reduction in smoke opacity values of about 66% is achieved with diesel/DEE/sunflower oil blends, whereas the same blends containing castor oil reduce emissions up to 77%.
