1 Introduction

In the present scenario, both developed and develo** countries are having substantial energy demand. Functioning of both industrial as well as automobile sectors is dominated by fossil fuels [1]. However, improper combustion of fossil fuels leads to considerable environmental pollution [2], affecting the health of people in the developed nations adversely [3]. Available emission curtailing techniques such as pretreatment of fuel sources, additives, and modification of engines are costly. It is the reason behind the increasing significance of research work based on alternative fuels. The key benefits of biodiesel are efficiency improvement, sustainability, and reduction in harmful gases. On top of that, the limited reservoirs of fossil fuels will not be adequate to meet world-wide energy consumption, while biodiesel can be a great substitute [4]. It is more environment friendly, oxygenated, sulphur-free, biodegradable, releasing lower emissions, more prominent among biofuels [5, 6]. Apart from this, there are certain disadvantages of biodiesel usage, such as viscos nature, denser than fossil fuels, higher modulus of elasticity, lower heating value, unequal atomization, and higher level of NOX emission [7, 8], which need to be tackled. Nanoparticles are found to be one of the most effective solutions among other additives improving the overall engine performance. They are enhancing engine performance parameters and reducing emission parameters dramatically [9].

Conventional methods of biodiesel production are having many drawbacks [10]. Some examples are pyrolysis; high temperatures should be maintained. So, the materials used for the experiments are expensive. Moreover, the purity of biodiesel is low [11]. For micro-emulsion and dilution, the product is not volatile and would’t have enough stability to form a perfect product [12]. For transesterification, biodiesel prepared by this method must be washed and neutralized for further use. There is a risk of side reactions, also it is difficult to separate the products obtained from transesterification [13]. It is also time consuming and has an impurity mixture of biodiesel [14]. Some of the drawbacks can be removed by using a costlier catalyst, but it is necessary to find an economical solution for the production of biodiesel. For feedstocks like waste cooking oil, which has a high acid value containing free fatty acids (FFA) and water, a routine transesterification procedure is ineffective. Traditional methods could not work as a sustainable solution. Moreover, the sustainability of biodiesel is largely affected by the methodology of production. In addition, traditional biodiesel production methods need reconsideration and modification for effective usage. To conquer the loopholes of traditional methods, research work is going in this direction extensively. Exploring a new methodology for biodiesel production is becoming necessary to minimize energy losses and increase efficiency. The most important step in the preparation of biodiesel is the mixing of raw materials and catalyst with solvent. In new trends of preparation of biodiesel, various reactors are developed for intensification of chemical reaction making it quicker and cheaper [15]. The study indicates that the increased production capacity can be achieved by shortening the time for separation and purification. To conquer the loopholes of traditional methods, a novel patented device based on the principle of hydrodynamic cavitation, is used for biodiesel production in the current investigations. An acid transesterification reaction almost takes 2–8 h whereas base transesterification reaction using two-step process takes 60–120 min. In addition, the current methodology is very time effective, as using the current methodology, the transesterification reaction time is reduced up to 40 min only.

After preparation, another important aspect for biodiesel preparation is the selection of feedstock [16, 17]. Determination of the cost of biodiesel depends on the selection of feedstock. Compared to consumable and non-consumable raw materials, waste cooking oil (WCO) is a potential substitute and far cheaper. In the current research studies, WCO is used as a feedstock for the preparation of biodiesel.

Nanotechnology has been playing a very crucial role by providing efficient solutions in the form of different nanoparticles such as in the case of cost economic karanja and jatropha oils, lithium impregnated solid catalyst is used for enhancement in properties [18]. By adding carbon coated aluminum nanoparticles, it shows the improving performance of the engine by reducing BSFC and NOx emission [2.6 Correlative exploration of emissions of blends consisting diesel and WCO biodiesel without CuO nanodopes

Harmful emissions of CO of WCO biodiesel blend and diesel fuel are recorded using AVL DIGAS (444N) 5 Gas Analyzer. In this experimental study, it was found that the WCO biodiesel blend B30 testing resulted in lesser CO emission compared to diesel with CI engine testing with compression ratio 18 and constant speed of 1500 rpm, varying the load from 0–6 to 12 kg, blend 0–30 to 100. CO emission of B100 blend is found to be 69.23% lower than diesel, whereas CO emission of B30 were 61.54% lower than pure diesel.

It is commonly observed that in the case of diesel, the CO2 emissions are maximum, as compared to blends of biodiesel [38, 39]. As diesel fuels have the maximum calorific value as compared to biodiesel blends, so the combustion of diesel fuel gives the maximum generation of carbon dioxide in the engine exhaust. In this experimental study, it was found that the blend B0 (diesel) testing resulted in lesser CO2 emission compared to biodiesel with CI engine testing with compression ratio 18 and constant speed of 1500 rpm, varying the load from 0–6 to 12 kg, blend 0–30 to 100. CO2 emission of B100 is 68.96% and B30 are 51.72% lower than diesel, respectively. During testing WCO biodiesel was added in blending with diesel, the carbon to hydrogen ratio of the mixture was increased. This is the main reason behind the increase in CO2 emissions for the B30 blend of biodiesel.

In the case of nitrogen gas is a stable component at normal aerial condition [40]. But the higher temperature generated during the combustion process supports the formation of nitrogen oxide due to the breakage of nitrogen molecules. The NOx is maximum in case of diesel as compared to biodiesel blends due to the higher rate of combustion process. High rate of combustion process of injected fuel particles shows the phenomenon of maximum combustion temperature as compared to biodiesel blend B30. Moreover, the load is having a significant impact on NOx emissions. As the load increases, NOx emission was found to increase in all readings. NOx emissions of B100 are 66.51% and B30 is 48.16% lower than pure diesel, respectively.

2.7 Synthesis of CuO nanoparticles

The average diameter 32 nm ± 12.56 nm is obtained as shown in Fig. 2 a, revealing the scanning electron microscopy (SEM) image of CuO nanoparticles under biological synthesis circumstances. Figure 2b reveals the confirmation of the diameter of CuO nanoparticles under ImageJ software, where x axis is representing the diameter of CuO nanoparticles and y axis is representing the frequency distribution of CuO nanoparticles’ diameter. In biosynthesis Magnolia Kobus leaf extract is utilized as a biological source. It is treated with aqueous solution of CuSO4.5H2O. Finally, CuO nanoparticles are synthesized through SEM.

Fig. 2
figure 2

a FESEM image of CuO nanoparticles. b Image analysis

Full size image

Correlative exploration of emissions of blends consisting diesel and WCO biodiesel with CuO nanodopes.

In further testing of emission reduction, the CuO nanoparticles do** is varied as 50–100-150 ppm. From the nanoparticles do** results, it is concluded that 100 ppm is given almost 0% CO emission compared to 50 ppm and even 150 ppm. Out of three different additions of CuO nanoparticles in B30 blend such as blend B30 + CuO (50 ppm), blend B30 + CuO (100 ppm) and blend B30 + CuO (150 ppm), CO emissions of blend B30 + CuO (50 ppm) and blend B30 + CuO (100 ppm) were found to be equally lowest, which is 92.30% lower than diesel. The blend B30 doped with CuO nanoparticles in different ppm resulted in the effective CO emission reduction. Figure 3 which is a graphical representation with load parameter in kg on x axis and CO emission in terms of % vol. on y axis, demonstrate the effect of load and CuO nanoparticles do** on CO emission variation of diesel and biodiesel blends. On the condition of CO2 emission, 50 ppm CuO nanoparticles do** result [41] were giving almost 0% CO2 emission compared to 100 ppm and even 150 ppm do**. Out of three different additions of CuO nanoparticles in B30 blend such as blend B30 + CuO (50 ppm), B30 + CuO (100 ppm), and blend B30 + CuO (150 ppm), CO2 emissions of blend B30 + CuO (50 ppm) and blend B30 + CuO (100 ppm) are found to be equally lowest, which is 94.83% lower than the diesel. Figure 4, which is a graphical representation with load parameter in kg on x axis and CO2 emission in terms of % vol. on y axis, demonstrates the effect of load and CuO nanoparticles do** in different proportions on CO2 emission variations. In the emission case of NOx, blend B30 blend of WCO biodiesel added with CuO nanoparticles resulted in a decrease in NOx as the load on the engine increases compared to WCO biodiesel blend without nanoparticles [42, 43]. From CuO nanoparticles do** results, it was concluded that 50 ppm was given almost 0% NOx emission compared to 100 ppm and even 150 ppm. Out of three different additions of CuO nanoparticles in B30 blend such as blend B30 + CuO (50 ppm), blend B30 + CuO (100 ppm) and blend B30 + CuO (150 ppm), NOx emissions of blend B30 + CuO (50 ppm) was found to be equally lowest, which was 96% lower than pure diesel. Figure 5 which is a graphical representation with load parameter in kg on x axis and NOx emission in terms of % vol. on y axis, demonstrate the effect of load and CuO nanoparticles do** in different proportions on NOx emission variation of diesel and biodiesel blends

Fig. 3
figure 3

Effect of Load on CO emission in % of blends consisting diesel and WCO biodiesel with CuO nanodopes

Full size image
Fig. 4
figure 4

Effect of Load on CO2 emission in % of blends consisting diesel and WCO biodiesel with CuO nanodopes

Full size image
Fig. 5
figure 5

Effect of Load on NOx emission in ppm of blends consisting diesel and WCO biodiesel with CuO nanodopes

Full size image

2.8 Analysis of L9 Taguchi Optimization

The response surface methodology (RSM) is utilized as an effective tool using the data obtained in the experimental investigation as shown in Table 3A and B to fit the polynomial equation generated by using Minitab 2019 [44].

Figure 6 is a graphical representation of the main effect plots for each determined parameter highlighting the variation of the blend (B0, B30, and B100, where low level is represented by 1, medium by 2, and high by 3), load (0–12 kg with variation of 6 kg), and IMEP (3.5–7.5 bar with variation of 2 bar) on x axis and means of CO emission variation on y axis. As per the results, diesel (B0) is having the highest emission of CO, it reduces exponentially in blend B30 but after that it has started increasing again when the biodiesel percentage in diesel has increased. Similarly, in 0 kg and 12 kg load CO emission is very high, whereas at half load (6 kg) CO emission is lowest. Figure 7 is a graphical representation of the main effect plots for each determined parameter highlighting the variation of CuO do** (50 -150 ppm with variation of 50 ppm, where low level is represented by 1, medium by 2, and high by 3), load (0–12 kg with variation of 6 kg, where low level is represented by 1, medium by 2 and high by 3), and IMEP (3.5–7.5 bar with variation of 2 bar, where low level is represented by 1, medium by 2 and high by 3) on x axis and means of CO emission variation on y axis. For CuO do** (Fig. 7) between 50 and 100 ppm, CO emission is very less, almost constant, whereas it is going on increasing exponentially with the increment of the nanoparticles. Variations of load are showing almost similar variation in both cases (Figs. 6 and 7). However, CO emission is much lower in CuO nanoparticles case compared to blends. In the case of IMEP, almost inverse graphs are obtained in the case of blends and nanoparticles.

Fig. 6
figure 6

Investigation by Taguchi: CO versus blend, load (kg), IMEP (bar)

Full size image
Fig. 7
figure 7

Investigation by Taguchi: CO versus CuO (nano), load (kg), IMEP (bar)

Full size image

Figure 8 is a graphical representation of the main effect plots for each determined parameter highlighting the variation of blends (B0, B30, and B100, where low level is represented by 1, medium by 2, and high by 3), load (0–12 kg with variation of 6 kg), and IMEP (3.5–7.5 bar with variation of 2 bar) on x axis and means of CO2 emission variation on y axis. In this Figure, it was shown that diesel (B0) was having the highest emission of CO2, it reduces with inducing more proportion of biodiesel. Whereas in 0 kg load CO2 emission was the lowest and it was almost proportional to the load parameter. In the case of IMEP, after 5.45 bar CO2 emission got increased. Figure 9 is a graphical representation of the main effect plots for each determined parameter highlighting the variation of CuO do** (50–150 ppm with variation of 50 ppm, where low level is represented by 1, medium by 2, and high by 3), load (0–12 kg with variation of 6 kg, where low level is represented by 1, medium by 2 and high by 3), and IMEP (3.5 -7.5 bar with variation of 2 bar, where low level is represented by 1, medium by 2 and high by 3) on x axis and means of CO2 emission variation on y axis. Figure 9 which is representing the case of CuO nanoparticles between 50 and 100 ppm, CO2 emission was very less, almost constant whereas they were going on increasing exponentially with the increment of the nanoparticles.

Fig. 8
figure 8

Investigation by Taguchi: CO2 versus blend, load (kg), IMEP (bar)

Full size image
Fig. 9
figure 9

Investigation by Taguchi: CO2 versus CuO (nano), load (kg), IMEP (bar)

Full size image

Figure 10 is a graphical representation of the main effect plots for each determined parameter highlighting the variation of the blend (B0, B30, and B100, where low level is represented by 1, medium by 2, and high by 3), load (0–12 kg with variation of 6 kg), and IMEP (3.5–7.5 bar with variation of 2 bar) on x axis and means of NOx emission variation on y axis. In Fig. 10, it was shown that diesel (B0) was having the highest emission of NOx, and it reduced with inducing more proportion of biodiesel. Whereas in 0 kg load, NOx emission was the lowest and it was almost proportional to the load parameter. In the case of IMEP after 5.45 bar NOx emission increased. Figure 11 is a graphical representation of the main effect plots for each determined parameter highlighting the variation of CuO do** (50 -150 ppm with variation of 50 ppm, where low level is represented by 1, medium by 2, and high by 3), load (0–12 kg with variation of 6 kg, where low level is represented by 1, medium by 2 and high by 3), and IMEP (3.5 -7.5 bar with variation of 2 bar, where low level is represented by 1, medium by 2 and high by 3) on x axis and means of NOx emission variation on y axis. Figure 11, which is representing the case of CuO do** between 50 and 100 ppm, NOx emission was very less, almost constant whereas they were going on increasing exponentially with the increment of the nanoparticles.

Fig. 10
figure 10

Investigation by Taguchi: NOx versus blend, load (kg), IMEP (bar)

Full size image
Fig. 11
figure 11

Investigation by Taguchi: NOx versus CuO (nano), load (kg), IMEP (bar)

Full size image

The response model which is fitted in the equations is as described. CO, CO2 and NOx with and without nanoparticles, are the response factors. Regression Eqs. (1–6) are as follows:

$${\text{CO }} = \, 0.0484 \, - 0.0167 {\text{Blend }} + 0.00078 {\text{Load}} \left( {{\text{kg}}} \right) \, + 0.00592 {\text{IMEP }}\left( {{\text{bar}}} \right)$$
(1)
$${\text{CO }} = \, - 0.0678 + 0.0483 {\text{CuO}} \left( {{\text{nano}}} \right) + 0.0083 {\text{Load}} \left( {{\text{kg}}} \right) - 0.0000 {\text{IMEP }}\left( {{\text{bar}}} \right)$$
(2)
$${\text{CO}}_{2} = \, 0.13 - 0.667 {\text{Blend}} + 0.2331 {\text{Load}} \left( {{\text{kg}}} \right) \, + 0.296 {\text{IMEP}} \left( {{\text{bar}}} \right)$$
(3)
$$CO_{2} = \, - 2.56 \, + 1.767 {\text{CuO}} \left( {{\text{nano}}} \right) \, + 0.850 {\text{Load}} \left( {{\text{kg}}} \right) \, - 0.617 {\text{IMEP}} \left( {{\text{bar}}} \right)$$
(4)
$${\text{NOx }} = \, - 56 - 145.2 {\text{Blend}} + 60.6 {\text{Load}} \left( {{\text{kg}}} \right) + 68.2 {\text{IMEP}} \left( {{\text{bar}}} \right)$$
(5)
$${\text{NOx }} = \, - 463 + 339 {\text{CuO}} \left( {{\text{nano}}} \right) + 203 {\text{Load}} \left( {{\text{kg}}} \right) \, - 170 {\text{IMEP (bar)}}$$
(6)

The design matrices are described in Tables 2A, B and 3A, B. The optimization was required for minimal condition emission of CO, CO2, and NOx. Based on Tables 5, 6, and 7, the signal-to-noise ratio plots are shown in Figs. 6, 7, 8, 9, 10, and 11. With the help of these plots, nature of the responses is governed by the variation of crucial parameters.

Table 5 Signal -to-noise ratio for CO (the smaller is better)
Full size table
Table 6 Signal -to-noise ratio for CO2 (smaller is better)
Full size table
Table 7 Signal to noise ratios for NOx, (smaller is better)
Full size table

Investigation with Taguchi for CO versus blend, load (kg), IMEP (bar) and CO versus CuO, load (kg), IMEP (bar): with and without nanoparticles is shown in Table 5. Investigation with Taguchi for CO2 versus blend, load (kg), IMEP (bar) and CO2 versus CuO, load (kg), IMEP (bar): with and without nanoparticles is shown in Table 6. Taguchi Analysis for NOx versus blend, load (kg), IMEP (bar) and NOx versus CuO, load (kg), IMEP (bar) with and without nanoparticles is shown in Table 7.

In the Pareto chart (Fig. 12), for CO emission without nanoparticles, which is a graphical representation of the standardization effect on the x axis and the predefined predictor effect on y axis, the predefined predictors were A (blend), B (load), and C (IMEP). From the influencing parameters in Pareto chart, the sequence A > C > B was considered as the outcome of predefined predictors for regression analysis of CO versus Blend, Load (kg), and IMEP (bar). The observation showed that A (blend) is the highest influence factor on CO emission. From the influencing parameters in the previous Pareto chart (Fig. 12), in the next Pareto chart (Fig. 13) for CO emission from nanoparticles, which is a graphical representation of the standardization effect on the x axis and the predefined predictor effect on y axis, the predefined predictors were A (CuO), B (load), and C (IMEP). From the influencing parameters in this Pareto chart, the sequence A > B > C was considered as the outcome of predefined predictors for regression analysis of CO versus CuO nanoparticles, Load (kg), and IMEP (bar), where A represents CuO mixed blend B30 kee** the other influencing parameters the same. The observation showed that A (CuO) is the highest influence factor on CO emission. The coefficient of determination is almost equal to 1, which reveals the chosen optimization technique is very accurate in prediction.

Fig. 12
figure 12

Regression analysis: CO versus blend, load (kg), IMEP (bar) without nanoparticle

Full size image
Fig. 13
figure 13

Regression analysis: CO versus CuO (nano), load (kg), IMEP (bar) with nanoparticles

Full size image

Residual plot is a category of scatter plots representing the input of data or independent variables on the x axis and residuals on the y axis. In this four-in-one residual plot (Figs. 14, 18, and 22) for CO, CO2, and NOx emissions without nanoparticles, also in the residual plots for CO, CO2, and NOx emissions with nanoparticles (Figs. 15, 19 and 23) the residuals were reaching almost to a linear nature in normal probability plot. In addition, an approximate symmetric nature of the histogram indicated that the residuals were normally distributed. Residuals were possessing constant variance as they were scattered randomly around zero in residual versus fitted values. Residuals exhibited no clear pattern in the residual versus order plot (there is no undesirable effect).

Fig. 14
figure 14

Residual plot for CO without nanoparticles

Full size image
Fig. 15
figure 15

Residual plot for CO with nanoparticles

Full size image
Fig. 16
figure 16

Regression analysis: CO2 versus blend, load (kg), IMEP (bar) without nanoparticles

Full size image
Fig. 17
figure 17

Regression analysis: CO2 versus CuO (nano), load (kg), IMEP (bar) with nanoparticles

Full size image
Fig. 18
figure 18

Residual plot for CO2 without nanoparticles

Full size image
Fig. 19
figure 19

Residual plot for CO2 with nanoparticles

Full size image

In the Pareto chart (Fig. 16), for CO2 emission without nanoparticles, which is a graphical representation of the standardization effect on the x axis and the predefined predictor effect on y axis, the predefined predictors were A (blend), B (load), and C (IMEP). From the influencing parameters in Pareto chart, the sequence B > A > C was considered as the outcome of the predefined predictors for regression analysis of CO2 versus Blend, Load (kg), and IMEP (bar). The observation showed that B (load) is the highest influence factor on CO2 emission. In the Pareto chart (Fig. 17) for CO2 emission from nanoparticles, which is a graphical representation of the standardization effect on the x axis and the predefined predictor effect on y axis, the predefined predictors were A (CuO), B (load), and C (IMEP). From the influencing parameters in this Pareto chart, the sequence A > B > C was considered as the outcome of a predefined predictor for regression analysis, where A represents CuO mixed blend B30 kee** the other influencing parameters the same. The observation showed that A (CuO) is the highest influence factor on CO2 emission.

Fig. 20
figure 20

Regression analysis: NOx versus blend, load (kg), IMEP (bar) without nanoparticles

Full size image
Fig. 21
figure 21

Regression analysis: NOx versus CuO (nano), load (kg), IMEP (bar) with nanoparticle

Full size image

In the Pareto chart (Fig. 20), for NOx emission without nanoparticles, which is a graphical representation of the standardization effect on the x axis and the predefined predictor effect on y axis, the predefined predictors were A (blend), B (load), and C (IMEP). From the influencing parameters in Pareto chart, the sequence B > A > C was considered as the outcome of the predefined predictors for regression analysis of NOx versus Blend, Load (kg), and IMEP (bar). The observation showed that B (load) is the highest influence factor on NOx emission. In the Pareto chart (Fig. 21), NOx emission from nanoparticles, which is graphical representation of the standardization effect on the x axis and the predefined predictor effect on y axis, the predefined predictors were A (CuO), B (load), and C (IMEP). From the influencing parameters in this Pareto chart, the sequence A > B > C was considered as the outcome of a predefined predictor for regression analysis of NOx, where A represents CuO mixed blend B30, kee** the other influencing parameters the same. The observation showed that A (CuO) is the highest influence factor on NOx emission.

Fig. 22
figure 22

Residual plot for NOx without nanoparticles

Full size image
Fig. 23
figure 23

Residual plot for NOx with nanoparticles

Full size image

As per the observation from Pareto chart for CO emission without nanoparticles, sequence, A > C > B was considered for regression analysis of CO versus Blend, Load (kg), and IMEP (bar). A (blend) was the highest influential factor before C (IMEP). In Figs. 24 and 25, the contour plot and surface plot of CO are shown without CuO nanoparticles, respectively, which are the graphical representations of predefined parameters, where x axis is representing the blend variation, and y axis is representing IMEP in the bar, and z axis is representing the variation of CO as the combined outcome of variation of IMEP and blend, highlighting the 2D and 3D response contour plot and surface plot of CO, plotted using response surface methodology approach. From both experimental as well as RSM approach, Minimum CO emission is obtained at 6.28 kg of blend (B30) and IMEP (5.75 bar).

Fig. 24
figure 24

Contour plot of CO without nanoparticles

Full size image
Fig. 25
figure 25

Surface plot of CO without nanoparticles

Full size image

As per the observation from Pareto chart for CO emission from nanoparticles, the sequence A > B > C was obtained for regression analysis. A (CuO nanoparticles) was the highest influence factor. In Figs. 26 and 27, the contour plot and surface plot of CO are shown for CuO nanoparticles, respectively, which are the graphical representations of predefined parameters, where x axis is representing the variation of CuO nanoparticles, and y axis is representing the load variation, and z axis is representing the variation of CO as combined outcome of variation of CuO nanoparticles and load, highlighting the 2D and 3D response contour plot and surface plot of CO, plotted using response surface methodology approach. From both experimental as well as RSM approach, the minimum CO emission from nanoparticles is obtained at the lowest do** of nanoparticles (50 ppm) in the blend B30 even at the highest load 12.17 kg.

Fig. 26
figure 26

Contour plot of CO with nanoparticles

Full size image
Fig. 27
figure 27

Surface plot of CO with nanoparticles

Full size image

As per the observation from Pareto chart for CO2 emission without nanoparticles, B (load) was the highest influence factor, after that blend was the second highest. In Figs. 28 and 29, the contour plot and surface plot of CO2 are shown without CuO nanoparticles, respectively, which are the graphical representations of predefined parameters, where x axis is representing the blend variation, and y axis is representing the load in kg, and z axis is representing the variation of CO2 as combined outcome of variation of load and blend, highlighting the 2D and 3D response contour plot and surface plot of CO2, plotted using response surface methodology approach. From both experimental as well as RSM approach, minimum CO2 emission without nanoparticles is obtained at 0 kg load and 3.48 bar IMEP for B0 (diesel), B30 and B100 blends.

Fig. 28
figure 28

Contour plot of CO2 without nanoparticles

Full size image
Fig. 29
figure 29

Surface plot of CO2 without nanoparticles

Full size image

As per the observation from Pareto chart for CO2 emission from nanoparticles, A (CuO) was the highest influence factor, after that load was the second highest. In Figs. 30 and 31, the contour plot and surface plot of CO2 are shown on CuO nanoparticles, respectively, which are the graphical representations of predefined parameters, where x axis is representing the variation of CuO nanoparticles, and y axis is representing the load variation, and z axis is representing the variation of CO2 as combined outcome of variation of CuO nanoparticles and load, highlighting the 2D and 3D response contour plot and surface plot of CO2, plotted using response surface methodology approach. From both experimental as well as RSM approach, the minimum CO2 emission from nanoparticles is obtained at the lowest do** of nanoparticles (50 ppm) in the blend B30 even at the highest load 12.17 kg.

Fig. 30
figure 30

Contour plot of CO2 with nanoparticles

Full size image
Fig. 31
figure 31

Surface plot of CO2 with nanoparticles

Full size image

As per the observation from Pareto chart for NOx emission without nanoparticles, B (load) was the highest influence factor, after that blend was the second highest. In Figs. 32 and 33, the contour plot and surface plot of NOx are shown without CuO nanoparticles, respectively, which are the graphical representations of predefined parameters, where x axis is representing the blend variation, and y axis is representing the load in the kg, and z axis is representing the variation of NOx as combined outcome of variation of load and blend, highlighting the 2D and 3D response contour plot and surface plot of NOx, plotted using the response surface methodology perspective. From both experimental as well as RSM approach, minimum NOx emission without nanoparticles is obtained at 0 kg load and 3.48 bar IMEP for B0 (diesel), B30 and B100 blends.

Fig. 32
figure 32

Contour plot of NOx without nanoparticles

Full size image
Fig. 33
figure 33

Surface plot of NOx without nanoparticles

Full size image

As per the observation from Pareto chart for NOx emission from nanoparticles, A (CuO) was the highest influence factor after that load was the second highest. In Figs. 34 and 35, the contour plot and surface plot of NOx are shown with CuO nanoparticles, respectively, which are the graphical representations of predefined parameters, where x axis is representing the variation of CuO nanoparticles, and y axis is representing the load variation, and z axis is representing the variation of NOx as combined outcome of variation of CuO nanoparticles and load, highlighting the 2D and 3D response contour plot and surface plot of NOx, plotted using response surface methodology approach. From both experimental as well as RSM approach, minimum NOx emission from nanoparticles is obtained at the lowest do** of nanoparticles (50 ppm) in the blend B30 even at the highest load 12.17 kg.

Fig. 34
figure 34

Contour plot of NOx with nanoparticles

Full size image
Fig. 35
figure 35

Surface plot of NOx with nanoparticles

Full size image

2.9 Key findings

In the present investigation, economical techniques are discussed for the reduction in harmful GHG emissions due to automobiles. Newly designed device based on hydrodynamic cavitation principal is used for the preparation of biodiesel conquering the flaws of traditional methods of biodiesel preparation. WCO biodiesel used in the current work can be a cost-effective substitute for diesel. Giancarlo Chiatti et al. [45] also agreed that WCO biodiesel can be a sustainable source of energy in their study of impact of WCO biodiesel on particle size distribution. In the experimental analysis during testing WCO biodiesel blend as shown in Table 2A, it is found that CO emissions of diesel fuel are higher than WCO biodiesel blends, and CO2 emissions are found to increase as the blend proportion of WCO biodiesel blend is increasing. Lai Fatt Chuah et al. [46] also stated the same pattern in the emissions of CO and CO2 during the comparative analysis of biodiesel and diesel. P. Karthikeyan et al. [47] also stated on the addition of nanoparticles, the reduction in NOx emission in biodiesel blends compared to diesel. In the current investigation, for better emission reduction results, the enhancement is done in WCO biodiesel blend with CuO nanoparticles which is hel** in order to achieve the complete combustion and improving the quality of combustion process reducing CO emission in comparison to diesel fuel. Moreover, speed of the combustion process, in the case of CuO nanoparticles doped WCO blend B30 is improved due to the high surface to volume ratio of CuO nanoparticles resulting in good atomization and rapid evaporation of fuel. Rapid evaporation might be the main reason behind the decreasing emission of CO2 in the engine exhaust compared to B30 blend of WCO biodiesel without nanoparticles. In addition to that, a good catalytic reaction due to CuO nanoparticles in the B30 blend of WCO biodiesel increases the heat transfer rate in the combustion chamber, which means the combustion temperature in the cylinder is less. Lower combustion temperature does not support the formation of NOx in the cylinder. In the case of B30 blend of WCO biodiesel added with CuO nanoparticles, there is a decrease in the NOx as the load on the engine increases compared to WCO biodiesel blend without nanoparticles. Silambarasan Rajendran [48] also agreed that effective mitigation of the emission of NOx is possible with effective antioxidant additives in biodiesel blends. According to the research study of Sukumar S. et al. [49], CuO is exhibiting good antioxidant properties. Moreover, CuO nanoparticles have a tendency to improve oxidation stability, this might be one of the reasons behind preventing oxidation and reducing NOx emission in the combustion process [50]. Moreover, the load is having significant impact on NOx emission. As the load increases, NOx emission is found to increase in all readings.

3 Conclusion

Emission of GHG gases such as CO, CO2 and NOx by using diesel as a fuel in four-stroke single-cylinder CI engines is found to be very high. Available pollution control devices are costly. WCO biodiesel can be one of the cost effective and sustainable solutions. In addition to that, the loopholes of traditional methods are conquered by a novel patented device based on the principle of hydrodynamic cavitation, used for biodiesel production in the current investigation. Two different experimental investigations are carried out using WCO biodiesel as fuel in CI engine. In the first one, blend, load, IMEP are considered as independent parameters, in the second one the blend B30 from the previous case has kept constant and is mixed with CuO nanoparticles which became a new independent parameter in the second case, kee** the other two parameters (load, IMEP) similar to the first one, at three levels. By varying these parameters, after application of the L9 Taguchi approach, a minimum emission of CO, CO2, and NOx is achieved. Optimized results as per Taguchi approach are as follows:

  • For the first investigation are 0.08 ppm CO, 0.6 ppm CO2 and 30 ppm NOx emission by appertaining the condemnatory merger of inputs such as blend (B0-Diesel), load (12.24 kg) and IMEP (3.48 bar) as shown in Fig. 36.

  • Effective CuO nanoparticles do**: after do** CuO nanoparticles, percentage reduction is found to be 92.3%, 94.82%, and 96% compared to the emission of diesel in CO, CO2 and NOx gases respectively.

  • Taguchi analysis based second experimental investigation on nanoparticles CuO in blend B30, the prognosticated results of optimization were 0.03 ppm CO, 0.3 ppm CO2 and 21 ppm NOx emission, by appertaining the condemnatory merger of inputs such as CuO nanoparticles (50 ppm), load (0.06 kg) and IMEP (3.48 bar) as shown in Fig. 37.

  • It can be studied further by adding a combination of effective nanoparticle does for effective combustion reduction as well as performance improvement.

  • Current experimental investigation has proved that the blending process of pure diesel and hydrodynamic cavitation based WCO biodiesel specifically with B30 blend has reduced CO and CO2 emissions.

Fig. 36
figure 36

Optimization of process parameters without nanoparticles

Full size image
Fig. 37
figure 37

Optimization of process parameters with nanoparticles

Full size image

Further conduction of the experiments utilizing the prognosticated results of optimization and outcome results were substantiated under the 3% error to prove the efficacy of the present investigation. Few limitations of the current research investigation are for the collection of waste cooking oil, a proper system has to be developed where direct collection provision will be there for food supply chains, various packaged food industries, big as well as small hotels, mess and even households. In this new methodology of biodiesel production pretreatment of waste cooking oil such as preheating and filtering are compulsory. Moreover, the experimental study is not able to reduce the emission of NOx gas by adding nanoparticles at 150 ppm. In the current investigation, WCO biodiesel is prepared by one of the cost effective and sustainable technologies. This biodiesel production technology can be further tested and used for various edible and non-edible feedstocks. Testing of the prepared various biodiesel blends could be done with multiple nanoparticles do** for more productive results. The results of the current investigation are concluded that enhanced WCO biodiesel blends doped with CuO nanoparticles are an environmentally friendly workable solutions to automobile industry. Present optimization work will be inscribing the problems of air pollution and providing an effective solution of reduction in emissions.