2.2.4. Cetane Number

Cetane number (CN) is one of key parameters to define the ignition quality of a fuel in a diesel engine. In general, CN of fuels must be above 51, as according norm EN 590, to facilitate autoignition and provide short ignition delay. However, CN values too high can cause the ignition delay is very short and combustion may start before the fuel and air are properly mixed, leading to an incomplete combustion [34]. Experimental determination of the CN of a fuel is a procedure tedious and expensive, and therefore estimation of cetane number of mixtures is carried out using the following simple equation:

$$\text{CN} = \sum\_{i} \text{CN} \text{K}i \tag{2}$$

where *CNi* is the cetane number of each component and *Xi* is the volumetric fraction of every component [33].

#### *2.3. Performance of a Diesel Engine-Electrogenerator Set Fuelled with EA*/*SVO and D*/*EA*/*SVO Blends*

The performance and soot emissions of a diesel engine-electric generator set have been analyzed following the same experimental methodology previously reported [15–17]. Engine specifications are shown in Table 3. In addition, the operating conditions of the engine were not modified during the tests.


**Table 3.** Specifications of the Diesel Engine-Electrogenerator Set.

The electrical power ( *P*) in watts can be easily calculated using a voltmeter-ammeter by application of Equation (3):

$$P = V \cdot I \tag{3}$$

where *V* is the potential di fference or voltage (in volts) and *I* is the electric current intensity or amperage (in amps).

The contamination degree (measured in Bosch number units) was determined from the opacity of the smoke generated in the combustion process, using a smoke density tester. In this research, the smoke density was measured by an opacimeter-type TESTO 338 density gauge (or Bosch smoke meter), Figure 2, according to standard method ASTM D-2156. The measurement range for smoke density is 0 to 2.5 with ±0.03 accuracy, where the value 0 represents total clarity on the paper and 2.5 is the corresponding value to 100% cloudy.

**Figure 2.** Mechanical and environmental characterization of diesel engine-electrogenerator set [17].

The fuel consumption of diesel engine (in liters per hour) is calculated supplying to engine an identical fuel volume (0.5 L) of each proposed (bio)fuel and measuring the volume consumed after a specified time. These measurements were performed at three engine loads that are representative of low (1 kW), medium (3 kW), and high (5 kW) power demands of the engine. Tests were done in triplicate and the results are shown as average along with standard deviation, represented as error bars.

## **3. Results and Discussion**

#### *3.1. Fuel Properties of EA*/*SVO Double Blends, and D*/*EA*/*SVO Triple Blends*

The kinematic viscosity results for ethyl acetate/sunflower oil (EA/SO) and ethyl acetate/castor oil (EA/CO) double blends are shown in Table 4. As can be noticed, a remarkable decrease in the viscosity values of SVOs was achieved by the use of ethyl acetate as solvent, as expected. In fact, only a 20% of EA reduce the viscosity of CO from 226.2 cSt to 26.26 cSt, while the same proportion of EA ge<sup>t</sup> to decrease the SO viscosity value from 37.80 to 11.52 cSt. Thus, increasing the volume of ethyl acetate in the blends we are able to fulfill the viscosity values for being employed as fuels in C.I. engines, as establish the European standard EN 590 ISO 3104. It is noticeable the stronger influence of EA on castor oil, in comparison with SO, to reduce the very higher viscosity value of this oil. Following the European normative (viscosities between 2.0 and 4.5 cSt), the blends with suitable viscosity values which were selected for the reformulation of diesel are obtained with a proportion 40/60 of EA/SO and 45/55 of EA/CO, Table 4.

**Table 4.** Viscosity values at 40 ◦C (cSt, centistokes) of ethyl acetate (EA)/sunflower oil (SO) and EA/castor oil (CO) double blends. Errors expressed as standard deviation have been calculated from average three measurement.


The best EA/SO and EA/CO double blends, containing 40% an 45% of ethyl acetate, respectively, were employed to prepare the D/EA/SVO triple mixtures by addition of different volumetric proportions to fossil diesel, which vary from 20% to 80% (B20 to B80). The kinematic viscosity results of triple mixtures are shown in Figure 3. As can be seen, the incorporation of the EA/SVO biofuels, from B0

to B100, promotes a slight increment in viscosity values of all blends, which was expected since they exhibit higher viscosities than fossil diesel. In general, the viscosities of the blends fulfill with European regulations EN 590, being in the range of 3.23–4.47 cSt.

**Figure 3.** Kinematic viscosity values at 40 ◦C of diesel/ethyl acetate/sunflower oil (D/EA/SO) and diesel/ethyl acetate/castor oil (D/EA/CO) triple blends. The results are the average of three measurements and errors are represented as standard deviation through error bars.

Cloud point, pour point, calorific values, and cetane number of triple blends with SO are collected in Table 5, whereas those with CO are collected in Table 6. Flow properties of diesel at low temperatures are greatly improved by the presence of EA in the D/EA/SVO blends, independently on the vegetable oil employed. In fact, a small percentage of ethyl acetate (8% in B20 blends with SO) generates a significant decrease in the pour point (PP), from −16.0 to −19.9 ◦C, and in the cloud point (CP), from −6.0 to −12.0 ◦C. Likewise, the triple blend with 11% of castor oil and 9% of EA reaches a reduction in PP and CP of up to 5 ◦C and 7 ◦C, respectively. Based on these results, the blends which exhibit the best behavior at low temperature are composed by 40% of biofuel, i.e., 60/16/24 (D/EA/SO) and 60/18/22 (D/EA/CO), which has conducted to a reduction of up to 4.2–6.0 ◦C in PP and 7.6–9.8 ◦C in CP. Particularly, blends with castor oil as SVO allow to obtain a slight improvement of CP and PP than their counterparts with sunflower oil. This behavior is very similar to that found using others compounds such as diethyl ether [16] or acetone [17] as additives in triple mixtures with the SVOs.

**Table 5.** Cloud point, pour point, calorific value and cetane number of diesel/ethyl acetate/sunflower oil triple blends (EA = 40%). Errors have been calculated from average three measurement and expressed as standard deviation.



**Table 6.** Cloud point, pour point, calorific value and cetane number of diesel/ethyl acetate/castor oil triple blends (EA = 45%). Errors have been calculated from average three measurements and expressed as standard deviation.

Regarding the calorific values for triple blends (Tables 5 and 6), as the percentage of ethyl acetate in the blend increase, the calorific value decrease. This is logical since EA exhibits a lower calorific value than both, diesel, and SVOs. Therefore, the most favourable results are found in the B20 triple blends, which have the highest calorific values, 34.51 MJ/L for the blends with sunflower oil (2.84% lower than diesel), and 34.54 MJ/L for the blends with castor oil (2.76% lower than diesel), while the B80 triple blends with highest biofuel content (80%) show the lowest calorific values of 31.49 and 31.58 MJ/L, i.e., 11.35 and 11.09% lower than diesel, for SO and CO blends respectively. As can be observed, there is no appreciable di fferences between calorific values of blends containing either sunflower or castor oil, since both oils display a comparable calorific power (see Table 1). Overall, the calorific value of biofuels EA/SVO studied was around 14% lower than that of diesel (B0).

The results related to cetane number (Tables 5 and 6) of blends with SO and CO were very similar among them, showing a decrease in cetane number values as EA/SVO ratio increases. It can be seen that the cetane number of all blends is below 51, which is the minimum cetane number of diesel in European standard EN 14214. Therefore, the ignition delay of these fuels is expected to be longer respect to conventional diesel fuel.

#### *3.2. Performance of a Diesel Engine Operating as Electric Generator*

According to the characterization results, the mixtures proposed as (bio)fuels that comply with kinematic viscosity requirements stablished by European normative EN 590, were tested in a diesel engine. The engine loads employed for the tests were 0, 1000, 2000, 3000, 4000, and 5000 W. Figure 4 illustrates the impact of engine load on engine performance fueled with the di fferent D/EA/SVO triple blends, containing SO (Figure 4a) and CO (Figure 4b). Furthermore, the performance of EA/SVO (B100) and fossil diesel were also included for comparison.

Generally, as the power supplied increases up to 4000 W, the power output also increases, whereas the highest engine load (5000 W) generated a drop in power output of engine. This behaviour is observed with D/EA/SO blends containing up to 60% of biofuel, and with D/EA/SO blends containing up to 80% of biofuel. On the contrary, blends B80 with SO and EA/SVO double blends display a di fferent trend: firstly, the power generated increases from 0 to 3000 W; then, it falls down from 3000 to 4000 W; and finally, it remains stable when the highest load is applied to engine. It is noteworthy that, at the highest load conditions, i.e., 4000 and 5000 W, the blends D/EA/CO exhibit very similar or even higher power output values than diesel. Thus, fuels composed by up to 60% of biofuel, EA/SO (B20–B60), and up to 80% of biofuel EA/CO (B20–B80) revealed a very notable e fficiency on engine. However, higher concentrations of EA/SVO led to a worst engine performance, in comparison with conventional diesel or analogous mixtures with lower biofuel content. Be that as it may, it is very interesting the fact that the EA/SVO double blends allow the running on engine without employing any fossil diesel content, which means a 100% of diesel substitution for renewable compounds. This achievement is even more remarkable for biofuel containing castor oil, since the decrease in the power output is between 7% and 30% in respect to diesel, which is lower than that obtained for biofuel with sunflower oil (50–74% lower than diesel). Usually, triple mixtures with castor oil exhibit better behavior as

(bio)fuels than those containing sunflower oil in any proportion investigated. In this sense, the usage of castor oil as part of these triple mixtures not only improves the power results as compared to sunflower oil, but also achieves better results with regard to fossil diesel, even with higher percentages of ethyl acetate up to 36% (B20–B80 blends). Taking into account that both SVOs exhibit very similar calorific values (Table 1), the better behavior of CO could be due to the higher cetane number of this oil, which improves the combustion quality. The progressive reduction in the engine performance observed with the blends B20 to B100 could be attributed to the low calorific value that ethyl acetate exhibit. Hence, a higher proportion of EA in the blend would promote the reduction in the energy content of blends (Table 1). This fact agreed with the results reported in previous investigations where triple blends containing ethanol [6], diethyl ether [16], or acetone [17] were employed as LVLCs. Nonetheless, the influence of other operating parameters on engine performance cannot be ruled out.

**Figure 4.** Power output (W) at different engine loads (1000–5000 W) for (**a**) diesel/ethyl acetate/sunflower oil; (**b**) diesel/ethyl acetate/castor oil blends. The error in measurements was always less than 3%.

## *3.3. Smoke Emissions: Opacity*

Figure 5 shows the smoke emissions (expressed in Bosch number) of the different blends tested. As can be seen, the higher the amount of EA/SVO in the blend, the higher the reduction of soot emissions. As we previously reported, this behavior is explained by the increase of the oxygen content in the fuel. Oxygen acts reducing the formation of rich zones and promoting the oxidation of soot nuclei during fuel combustion. Although the oxygen content is the dominant factor on soot emissions, in turn a lower cetane number has also demonstrated to decrease soot particles emissions due to that a longer ignition delay provides more time for the premixing of fuel and air prior to the start of combustion, which increases the oxidation of soot particles [35].

Accordingly, the lowest opacity values are obtained with pure biofuels EA/SVO (B100), independently on the vegetable oil used, since they have the highest concentration of EA and the lowest CN. These B100 blends decrease soot emissions up to 85% when sunflower oil is employed and up to 96% for the blends containing castor oil as SVO. For B20–B80 D/EA/SO triple blends, the opacity is reduced from 16 to 80% of the total opacity value attained with fossil diesel, Figure 5a. For its part, the reduction is even higher when B20–B80 D/EA/CO triple blends are employed, up to 94% lower than opacity obtained with diesel for B80 blend, Figure 5b. The slightly better behavior obtained with CO blends can be attributed to the lower presence of unsaturation that ricinoleic acid of CO exhibit, in comparison to that exhibit by linoleic acid of SO. As it is well-known, unsaturation in fuels contributes to the formation of soot precursor species [36]. The results are in agreemen<sup>t</sup> with previous studies where triple mixtures with other oxygenated compounds reported a similar behavior in term of emissions reduction [6,16,17].

**Figure 5.** Smoke emissions (Bosch number) at different engine loads (1000–5000 kW) for (**a**) diesel/ethyl acetate/sunflower oil; (**b**) diesel/ethyl acetate/castor oil blends.
