Impact of butanol-sunflower oil biodiesel–diesel blends on engine performance, combustion, and after-treatment emissions (DOC + SCR)

Abstract

This study investigates the effects of n-butanol–sunflower biodiesel–diesel ternary blends on the combustion, performance, and emission characteristics of a single-cylinder compression ignition (CI) engine equipped with an integrated diesel oxidation catalyst and selective catalytic reduction (DOC + SCR) aftertreatment system. Eight fuel formulations were evaluated at 1800 rpm under four engine loads (10.98, 16.48, 21.97, and 27.46 Nm).: neat diesel (D100); binary n-butanol–diesel blends containing 5, 10, and 15 vol% butanol (D95BUT5, D90BUT10, and D85BUT15); a sunflower biodiesel–diesel blend containing 30 vol% biodiesel (B30D70); and ternary blends combining B30 with 5, 10, and 15 vol% butanol (B30D65BUT5, B30D60BUT10, and B30D55BUT15). Combustion parameters were derived from 50-cycle ensemble-averaged pressure data, while emissions were quantified upstream (P₁) and downstream (P₃) of the aftertreatment system. Results showed that D90BUT10 achieved the lowest specific fuel consumption (241.01 g/kWh) and highest brake thermal efficiency (33.97%) at 21.97 Nm, corresponding to a 4.4 percentage point improvement over D100 at the same condition. D85BUT15 yielded the lowest HC emissions, reaching 54.77 ppm at 10.98 Nm, whereas B30D55BUT15 produced the lowest CO emissions (0.032 vol%) under low-load conditions. However, B30-based blends generated the highest NO_x concentrations, with a peak value of 2109 ppm at 16.48 Nm. Following DOC + SCR treatment, CO emissions were completely eliminated for all oxygenated blends, HC emissions decreased by 10–35% depending on load and fuel composition, and NO_x conversion efficiency ranged between 40 and 68%, with the highest conversion observed at 27.46 Nm. An anomalous deterioration in combustion and performance was identified for B30D60BUT10 at intermediate loads, attributed to the combined effects of biodiesel-induced atomization limitations and butanol-related ignition delay. Overall, the study provides a systematic characterization of sunflower biodiesel–n-butanol ternary blend interactions within an integrated DOC + SCR framework and highlights the non-linear coupling between oxygenated fuel composition and aftertreatment performance.

Introduction

Escalating global energy demand and increasingly stringent emission regulations have intensified research into alternative fuels for compression ignition (CI) engines. Beyond resource constraints, the vulnerability of critical energy supply routes to geopolitical disruptions has introduced significant volatility in fuel and urea markets, emphasizing the need for localized and resilient energy solutions¹. In this regard, leveraging Turkey’s capacity for sunflower oil methyl ester (SOME) production represents a promising pathway to enhance energy security while maintaining effective emission control strategies.

Among renewable fuel options, biodiesel and higher alcohols such as n-butanol have emerged as extensively investigated candidates for CI engine applications². While each fuel exhibits distinct advantages when blended individually with diesel, their binary formulations introduce specific physicochemical limitations that necessitate the transition to ternary mixtures.

Sunflower oil methyl ester (SOME) is an attractive feedstock due to its generally compatible density, viscosity, and cetane properties. Its oxygen content promotes more complete combustion, reducing CO and HC emissions. However, higher SOME ratios often lead to moderate torque reductions (6–18%) and inferior cold-flow properties. Furthermore, fuel-bound oxygen elevates local flame temperatures, accelerating thermal NO_x formation via the Zeldovich mechanism (N₂ + O → NO + N). Simultaneously, the higher kinematic viscosity of SOME imposes viscosity-driven atomization constraints, producing a coarser droplet population with an increased Sauter mean diameter (SMD). This extends combustion duration and promotes PM formation via a nucleation–surface growth–agglomeration pathway, particularly in locally fuel-rich zones^1–3.

n-Butanol offers a higher energy density (33.1 MJ/kg), lower hygroscopicity, and complete miscibility without co-solvents compared to lower alcohols such as ethanol. It provides improved ignition quality control, higher thermodynamic stability, and reduced volatility-related handling issues. Its high latent heat of vaporization (582 kJ/kg) induces significant evaporative charge cooling, reducing peak temperatures below the Zeldovich threshold. Critically, n-butanol acts as an effective soot suppressor by improving spray atomization and reducing the carbonaceous load entering downstream aftertreatment, thereby mitigating DOC fouling and SCR catalyst deactivation³. Nevertheless, binary butanol blends are constrained by ignition quality deficits and lower calorific values.

In the ternary biodiesel–n-butanol–diesel system, these limitations are mutually compensated through synergistic interactions: (i) SOME’s high cetane number counteracts butanol’s ignition penalty; (ii) butanol’s charge-cooling offsets biodiesel’s NO_x increase; and (iii) butanol’s lower viscosity improves atomization, partially compensating for the SOME penalty^3–5. These interactions are inherently non-linear and produce complex synergistic effects on combustion phasing and emission formation that cannot be predicted from individual fuel properties alone.

Beyond engine-out emissions, compliance with Euro VI requires integrated aftertreatment. The Langmuir–Hinshelwood surface oxidation mechanism governs CO and HC conversion within the DOC, which also partially converts NO to NO₂, aiming for an optimal NO₂/NO_x ratio (~ 0.5) to trigger the kinetically dominant fast-SCR reaction pathway. Additionally, NO₂ generated within the DOC can enhance downstream soot oxidation through NO₂-assisted carbon oxidation pathways, thereby influencing particulate removal efficiency and reducing the carbonaceous load entering the SCR system. The efficiency of the SCR system is strongly governed by the catalyst temperature window (typically 200–450 °C) and by transient ammonia storage-desorption dynamics, which together determine NO_x conversion stability under varying engine load conditions. Furthermore, the exothermic oxidation of HC and CO promotes a critical temperature rise from the DOC inlet (T₁) to the SCR inlet (T₂), maintaining the aftertreatment system above the activation threshold, especially under low-load conditions. In alignment with recent advances that incorporate such evaluations within broader sustainability metrics including exergy, thermoeconomic, and environmental analyses^4–6, the current research integrates combustion and aftertreatment characterization to evaluate the energy efficiency and emission control trade-offs inherent to oxygenated multi-component fuel systems.

This objective is pursued through a comprehensive systematic investigation that simultaneously quantifies emission transformations at three discrete positions, namely upstream of the DOC (P₁), between the DOC and SCR (P₂), and downstream of the SCR (P₃), across a load-resolved matrix. This approach enables detailed evaluation of the non-linear interaction and load-dependent responses of sunflower biodiesel–n-butanol ternary blends within an integrated DOC + SCR framework. Previous studies on biodiesel-fueled DOC + SCR systems have demonstrated the effectiveness of integrated aftertreatment configurations for reducing regulated emissions under CI engine operating conditions⁷.

Accordingly, the present study experimentally investigates the combustion, performance, and exhaust emissions of eight fuel formulations: neat diesel, three n-butanol–diesel blends (5, 10, and 15 vol% butanol), one sunflower biodiesel–diesel blend (30 vol% biodiesel), and three ternary blends (B30 combined with 5, 10, and 15 vol% butanol). Experiments were conducted at 1800 rpm under four engine loads using a single-cylinder direct-injection CI engine equipped with a custom-designed DOC + SCR system. The specific objectives are to: (i) analyze in-cylinder combustion parameters including peak pressure, heat release rate, ID, combustion duration, CA10, CA50, and ringing intensity; (ii) evaluate engine performance (BSFC and BTE); (iii) quantify engine-out emissions (CO, CO₂, HC, and NO_x); and (iv) assess the conversion efficiency of the DOC + SCR system at three aforementioned measurement positions. The study additionally identifies a distinct load-dependent combustion and emission response for the B30D60BUT10 ternary blend, highlighting the non-linear interaction behavior that can emerge in oxygenated multi-component fuel systems. This multi-stage characterization provides new insight into the interaction between oxygenated ternary fuels, DOC exothermic behavior, and SCR conversion performance in agricultural CI engine platforms.

Experimental setup and procedures

Test engine and dynamometer

All experiments were conducted at the Engine Testing Laboratory of Afyon Kocatepe University. The test engine was a single-cylinder, four-stroke, direct-injection, water-cooled compression ignition engine (Antor AD510 BS). The Antor AD510 BS engine (Lombardini 3LD 510 licensed production), was specifically selected for this study because it is a widely utilized single-cylinder power unit in the Turkish agricultural sector, commonly integrated into power tillers (hoeing machines), water pumps, portable generators, and small-scale irrigation systems. A closed-loop coolant recirculation circuit, regulated by a water temperature sensor, maintained engine thermal stability throughout the tests. The principal technical specifications of the engine are summarized in Table 1. The four load conditions were set by adjusting the dynamometer brake load to 4, 6, 8, and 10 kg, corresponding to brake torque values of 10.98, 16.48, 21.97, and 27.46 Nm, respectively, as calculated from with g = 9.81 m/s² and L = 0.28 m. For reference, the maximum rated torque of the test engine (32.8 Nm) corresponds to a dynamometer load of approximately 11.94 kg under the same configuration.

Table 1.

Technical specifications of the test engine.

Parameter	Value
Manufacturer / Model	Antor AD510 BS (Lombardini 3LD 510)
Engine type	Single-cylinder, 4-stroke, DI, water-cooled CI
Displacement (cm3)	510
Compression ratio	17.5:1
Bore × Stroke (mm)	85 × 90
Rated power (kW)	8.83 @ 3000 rpm
Maximum torque (Nm)	32.8 @ 1800 rpm
Fuel injection timing (SOI)	24°CA BTDC
Injector opening pressure	190–200 bar

Engine loading was applied through a direct-current regenerative dynamometer with a rated power absorption capacity of 15 kW. Braking torque was quantified by an STCS S-type load cell (50 kg capacity, C3 accuracy class, 0.003% linearity error) mounted to the dynamometer housing. Fuel consumption was determined gravimetrically using a high-precision electronic balance (30 kg capacity, 1 g resolution) in conjunction with a digital stopwatch.

Combustion measurement system

In-cylinder pressure was acquired with an AVL QC34C piezoelectric pressure transducer (operating range: 0–250 bar; linearity: ± 0.2%) flush-mounted in the cylinder head. The charge signal from the sensor was conditioned by a Kistler 5018A charge amplifier, whose output voltage was sampled in synchrony with crank angle. Crank angle resolution was provided by an Opkon PRI 50AR8 optical encoder (360 pulses per revolution) coupled to the crankshaft. Both the amplified pressure signal and the encoder pulse train were digitized by a National Instruments USB-6351 data acquisition card and transferred to a PC. To minimize cycle-to-cycle variability effects, pressure traces over 50 consecutive engine cycles were ensemble-averaged for each test condition, and the mean cycle was used for all combustion analyses.

Heat release analysis (HRR, CHR, CA10, CA50, CA90, CD)

The apparent heat release rate (HRR), a key indicator of combustion dynamics, was derived from the ensemble-averaged in-cylinder pressure trace (50 consecutive cycles) using a single-zone, first-law-based differential model (Eq. 1)^8–10.

1

where Q_gr denotes gross heat release (J), θ the crank angle (°CA), dθ crank angle variation, k the specific heat ratio, P the instantaneous cylinder pressure (Pa), V the instantaneous cylinder volume (m³) and the heat transferred to the cylinder wall (J). In addition, , , and show the produced gross heat, cylinder volume change, in-cylinder pressure change and transferred heat depending on crank angle respectively. Equation (2) was utilized to determine the heat transfer to cylinder wall^9,10.

2

where n denotes the engine speed (rpm), the instantaneous heat transfer area (m²), the convective heat transfer coefficient (W/m²·K), the bulk in-cylinder gas temperature (K), and the cylinder wall temperature (K). Equation (3) was used to calculate the instantaneous heat transfer coefficient ( based on the model proposed by Hohenberg¹¹.

3

where denotes the mean piston speed (m/s). The empirical constants 130 and 1.4 are as originally proposed by Hohenberg¹¹.

The cumulative heat release (CHR) was calculated by integrating the HRR with respect to crank angle (Eq. 4)⁹:

4

where denotes the crank angle at the start of injection and the crank angle at the end of combustion (CA90).

The CA10, CA50, and CA90 parameters were calculated from the CHR curve. The start of combustion (SOC) is defined as the crank angle at which the HRR transitions from negative to positive values, marking the point where exothermic reactions overcome the evaporative cooling effect of the injected fuel^9,10,12. CA10, defined as the crank angle at which 10% of the total heat has been released, is widely adopted as a practical proxy for the effective start of combustion, given the difficulty of precisely identifying the HRR sign-change point under experimental conditions^13–15. To avoid conceptual ambiguity, a clear distinction is maintained between SOC and CA10: SOC represents the physical onset of exothermic reactions, whereas CA10 denotes the crank angle at which combustion becomes thermodynamically established; the interval between them is referred to as the early flame development phase^9,12. CA50 represents the crank angle at which 50% of the total heat has been released and serves as the primary indicator of combustion phasing¹⁶. CA90, denoting the crank angle at which 90% of the total heat has been released, is used to represent the end of combustion (EOC)^14,15. Combustion duration (CD) is defined as the crank angle interval from CA10 to CA90¹⁷.

Combustion stability (MPRR, RI)

The maximum pressure rise rate (MPRR, bar/°CA), defined as the maximum rate of increase in in-cylinder pressure over a crankshaft angle during a combustion cycle, was calculated using Eq. (5)⁹. MPRR is widely adopted as a critical indicator for evaluating combustion intensity, structural loading on engine components, and knock tendency.

5

where P the instantaneous in-cylinder pressure (bar) and θ the crank angle (°CA). The ringing intensity (RI, MW/m²), which quantifies the severity of pressure oscillations and knock events during combustion, was calculated using Eq. (6)^18,19.

6

where R the specific gas constant (J/kg·K), the maximum in-cylinder gas temperature (K), the maximum in-cylinder pressure (Pa), and β the pressure rise duration (s).

Engine performance (P_b, BSFC, BTE)

The brake power output (, kW) was determined from the load cell measurement using Eq. (7):

7

where T the brake torque (Nm), calculated as , in which m denotes the load cell reading (kg), g the gravitational acceleration (9.81 m/s²), and L the dynamometer arm length (0.28 m); the correspondence between load cell readings and brake torque values is provided in Table 1. The brake specific fuel consumption (BSFC, g/kWh), representing the fuel mass flow rate required to produce one unit of brake power, was calculated using Eq. (8)⁹.

8

where the mass flow rate of fuel (g/h). The brake thermal efficiency (BTE, %), representing the fraction of fuel chemical energy converted into useful brake work, was calculated using Eq. (9)⁹.

9

where the lower heating value of the test fuel (kJ/g).

Exhaust gas aftertreatment system

DOC + SCR system configuration

Although the Antor AD510 BS engine is not equipped with emission aftertreatment technologies in its standard commercial configuration, a dedicated two-stage DOC + SCR system was custom-designed and fabricated specifically for this engine as part of an extensive experimental development program^20,21. Conventional ceramic honeycomb DOC and vanadium-based (V₂O₅-WO₃/TiO₂) SCR catalyst substrates were acquired and cut to a precise volume of 400 cm³ each to match the 510 cm³ engine displacement²⁰. The fabrication involved a six-section exhaust assembly in which the catalysts were welded into the line using conical reducers to optimize flow transition, alongside dedicated mounting ports for the DEF injector, a static mixer, and dual NO_x sensors²⁰.

The DOC employs a platinum-group metal catalyst deposited on a ceramic honeycomb monolith; it oxidizes CO and unburned hydrocarbons (HC) and partially converts NO to NO₂, thereby enhancing the fast-SCR reaction (R1) kinetics in the downstream unit. The SCR catalyst is a vanadium-based (V₂O₅-WO₃/TiO₂) honeycomb monolith with a cell density of 400 cpsi. A 32.5% aqueous urea solution (diesel exhaust fluid, DEF) served as the reductant and was delivered upstream of the SCR brick by a Bosch 2nd Generation (deNOx 2.2) dosing injector. A static mixer installed immediately downstream of the DEF injection point ensured homogeneous ammonia distribution across the catalyst face²². Exhaust gas temperature (EGT) was recorded at three discrete locations by NiCr-Ni (K-type) thermocouples: T₁ at the DOC inlet, T₂ between the DOC outlet and the SCR inlet, and T₃ at the SCR outlet.

Urea dosing and control strategy

The DEF dosing pump, SCR control unit, and injector were operated via a purpose-built MATLAB-based supervisory interface, enabling real-time monitoring and adjustment of the urea injection parameters independently of the engine management system²¹. The theoretical dosing requirement was established from the stoichiometry of the fast-SCR reaction pathway (Equation R1), which represents the most kinetically efficient NO_x reduction route under typical exhaust operating temperatures²³:

R1

Based on the respective molecular weights (NO = 30, NO₂ = 46, NH₃ = 17 g/mol), the reduction of 76 g of NO_x requires 34 g of NH₃, yielding a theoretical demand of 0.448 g NH₃ per gram of NO_x²⁰. Since one mole of urea (CO(NH₂)₂) yields two moles of NH₃ via successive thermolysis and hydrolysis reactions, 0.789 g of pure urea is required per gram of NO_x²³. Adjusting for the 32.5% standard DEF concentration gives a theoretical dosing factor of 2.43 g DEF per gram of NO_x^20,23. The instantaneous mass flow rate of NO_x (g/h) was calculated by integrating the measured NO_x concentrations (ppm) with the exhaust mass flow rate, which was derived from the measured fuel consumption and the excess air ratio (λ) data. Although the actual NO₂/ NO_x ratio at the SCR inlet was not measured in this study, the fast-SCR pathway was adopted as the design basis, as it represents the kinetically dominant reduction route under the exhaust temperatures achieved (T₂ > 285 °C throughout the test matrix), and the NH₃/NO_x molar ratio of 1.0 derived from its stoichiometry is equally valid for the standard SCR pathway²³. The DEF injection rate was pre-characterized from injector calibration curves established prior to the main experiments using a precision balance (SF-400 C, 0.01 g resolution), following the procedure described in²⁰. The pre-characterized flow rate confirmed that the selected injection setting delivered an NH₃/NO_x molar ratio of approximately 1.0 under the nominal operating conditions of the test, consistent with the theoretical dosing requirement^20,21. At each engine load condition, post-SCR exhaust emissions were recorded in three consecutive measurement phases: a pre-injection phase without DEF, an active DEF injection phase, and a post-injection recovery phase without DEF. The reported post-SCR emission values (P₃) represent the arithmetic mean of all three phases. The close agreement observed among the three phase measurements is attributable to the ammonia storage capacity of the vanadium-based SCR catalyst, which sustains NO_x reduction activity beyond the active injection interval through adsorbed NH₃ surface coverage, thereby confirming stable catalyst operation and validating the representativeness of the averaged post-SCR values^21,23.

Exhaust gas emission analyzer

Gaseous exhaust emissions were measured with a Bilsa MOD 2210 five-component analyzer at two sampling points: P₁, located at the engine exhaust outlet upstream of the DOC, and P₃, located downstream of the SCR catalyst. The operating ranges and measurement accuracies of the instrument are listed in Table 2.

Table 2.

Measurement specifications of the Bilsa MOD 2210 exhaust gas analyzer.

Component	Measurement range	Accuracy
CO	0–10 vol%	± 0.01 vol%
CO2	0–20 vol%	± 0.01 vol%
HC	0–10,000 ppm	± 1 ppm
O2	0–25 vol%	± 0.01 vol%
NOx	0–5000 ppm	± 1 ppm
Excess air coefficient (λ)	0–5.2	± 0.001

Uncertainty analysis

To assess the statistical reliability of the experimental measurements obtained in this study, a comprehensive uncertainty analysis was conducted in accordance with the error propagation framework established by Holman²⁴ and Kline and McClintock²⁵. The combined relative uncertainty of a dependent variable R that depends on n independent measured quantities x₁, x₂,…, x_n is expressed by Eq. (11):

11

where x_i denotes the i-th independent variable, R is the dependent variable of interest, U_i is the uncertainty associated with the i-th measured quantity, and n is the number of independent parameters.

The total uncertainty of each measured parameter was computed by combining two distinct uncertainty components. The first component, instrument uncertainty (u_instrument), was derived from manufacturer-provided calibration certificates and equipment specifications. The second component, repeatability uncertainty (u_{repeatability}), was evaluated from the standard deviation of three successive measurements performed under identical experimental conditions. These two components were combined using the root-sum-square (RSS) method^25,26, as given in Eq. (12):

12

where u_total is the overall combined uncertainty for a given measurement parameter, u_instrument accounts for calibration-related errors of the measuring device, and u_{repeatability} captures the variability arising from repeated experimental runs.

The relative uncertainty (expressed as a percentage) for each parameter was subsequently calculated by normalising u_total with respect to the mean value of the repeated measurements, according to Eq. (13):

13

where represents the arithmetic mean of the repeated measurements for the parameter under consideration.

The overall system uncertainty was then determined by propagating all individual percentage uncertainties through the RSS formulation²⁴, as presented in Eq. (14):

14

where U_CO2, U_CO, U_O2, U_HC, U_NOx, U_encoder, U_balance, and U_pressure refer to the individual percentage uncertainties of CO₂, CO, O₂, HC, NO_x, the crank-angle encoder, the precision balance, and the cylinder pressure transducer, respectively, as summarized in Table 3. The resulting overall system uncertainty of ± 1.93% is well within the generally accepted tolerance for internal combustion engine experiments, thereby confirming the precision and reproducibility of the reported measurements.

Table 3.

Individual percentage uncertainties of the measured parameters.

Parameter	Uncertainty (%)
CO2 (% vol.)	± 0.84
CO (% vol.)	± 0.84
O2 (% vol.)	± 0.83
HC (ppm)	± 0.83
NOx (ppm)	± 0.83
Encoder (°CA)	± 0.10
Precision balance (g)	± 0.001
Pressure sensor (bar)	± 0.50

Test fuels

Biodiesel production

Sunflower oil biodiesel was produced via alkaline transesterification. Sodium hydroxide (NaOH, 0.5 wt.% relative to oil mass) was dissolved in anhydrous methanol (20 wt.% relative to oil mass) to prepare the catalyst solution. Prior to the reaction, the sunflower oil feedstock was heated to 120 °C to reduce residual water content and then allowed to cool to approximately 60 °C. The oil was subsequently combined with the catalyst-methanol mixture and stirred continuously with a magnetic stirrer at 60 °C for 90–120 min to complete the transesterification. Following phase separation, the lower glycerol layer was withdrawn, and the crude biodiesel phase was washed with warm distilled water until a neutral pH was attained and then dried under mild vacuum to remove residual moisture.

Fuel blending and properties

Eight fuel blends were prepared volumetrically for the engine tests: neat diesel (D100); three butanol-diesel binary blends (D95BUT5, D90BUT10, and D85BUT15, containing 5, 10, and 15 vol% n-butanol, respectively); one biodiesel-diesel binary blend (B30, containing 30 vol% sunflower oil biodiesel); and three biodiesel-butanol-diesel ternary blends (B30D65BUT5, B30D60BUT10, and B30D55BUT15, each containing 30 vol% biodiesel combined with 5, 10, and 15 vol% n-butanol, with the remainder being diesel).

In this study, a 30 vol% sunflower biodiesel blend (B30) was adopted as the constant base fuel for all ternary formulations, based on converging evidence from the existing literature. Previous experimental and simulation studies have consistently shown that increasing the biodiesel fraction systematically reduces effective power and torque while raising BSFC, with B30 representing a practical upper limit before these performance penalties become prohibitive for unmodified agricultural engines^27–29. Several engine studies further indicate that blend ratios above 30 vol% may induce combustion instability or hardware incompatibility in standard injection systems, confirming that higher fractions are not suitable as a baseline without engine modifications²⁹. From a fuel-property perspective, the progressive increase in kinematic viscosity and surface tension with biodiesel content degrades atomization quality and increases injection system stress; experimental and numerical analyses identify viscosity thresholds above which feed-system issues become likely unless preheating or hardware changes are introduced^30–32. Studies on comparable single-cylinder platforms report that exceeding the 30 vol% biodiesel threshold typically results in torque losses of 7 to 18% and measurable deterioration in spray characteristics, with B100 producing reductions of up to 7.06% in torque under full-load conditions^27,29. At the same time, B30 provides a sufficiently oxygenated exhaust environment to challenge the DOC + SCR assembly under elevated engine-out NO_x loading, and previous aftertreatment studies demonstrate that moderate-to-high biodiesel fractions can enhance DOC oxidation activity and SCR NO_x reduction performance^33,34. Consequently, B30 was selected as a literature-consistent upper-limit blend that maximizes the renewable fuel share and exhaust oxygenation while preserving acceptable torque capability, injection quality, and durability margins for extended DOC + SCR characterization on an unmodified single-cylinder platform.

The key thermophysical properties of the base fuels are presented in Table 4.

Table 4.

Thermophysical properties of the base fuels.

Property	ASTM Standard	Diesel (D100)	Sunflower biodiesel (B100)
Density (kg/m3 @20 °C)	ASTM D4052	831.88	882.73
Kinematic viscosity (mm2/s @40 °C)	ASTM D445	3.085	4.264
Lower heating value (kJ/kg)	ASTM D240	44,891	40665a
Flash point (°C)	ASTM D93	60	100–170b
Cetane number	ASTM D613	≥ 51	51–54c,
Water content (wt.%)	EN ISO 12,937	0.0033	0.0418

^a Data obtained from Ref.³⁵.

^b Data obtained from Ref.³⁶.

^c Data obtained from Ref.³⁵.

Butanol, added to diesel–biodiesel blends, is an additive with 21.6% oxygen content. The basic physical and chemical properties of butanol are shown in Table 5, while the abbreviations and mixture ratios of the triple fuel blends used in combustion experiments are shown in Fig. 1.

Table 5.

Properties of n-butanol^37,38.

Properties	n-butanol
Molecular formula	C4H9OH
Density (kg/m3 @ 15 °C)	813
Viscosity (mm2/s @ 40 °C)	2.63
Latent heat (kJ/kg @ 25 °C)	582
Lower heating value (kJ/kg)	33,100
Boiling temperature (°C)	117.7–118
Auto-ignition temperature (°C)	343–385
Flash point at closed cup (°C)	35
Oxygen content (wt.%)	21.6
Stoichiometric ratio	11.21
Cetane number	15.9–25
Octane number	96

Fig. 1.

Test fuels abbreviations and mixture ratios.

Experimental procedure

Before each test sequence, the engine was operated on neat diesel under no-load conditions for approximately 25 min in order to raise the coolant temperature to 85 °C and to stabilize the lubricating oil temperature as well as the exhaust backpressure within the aftertreatment system. All experiments were performed at a constant engine speed of 1800 rpm, which corresponds to the rated maximum torque point of the Antor AD510 BS engine⁷. This specific speed was selected primarily because 1800 rpm represents the most prevalent and critical operating condition for this engine model in real-world agricultural applications, including power tillers, water pumps, and small-scale irrigation systems that are widely utilized across the Turkish agricultural sector⁷. By performing the evaluation at this speed, the study ensures that the synergy between the ternary fuel blends and the DOC + SCR system is characterized under thermodynamically representative conditions that align with the engine’s peak mechanical loading and typical field operations.

The experiments were conducted under four progressive load conditions of 10.98, 16.48, 21.97, and 27.46 Nm. The maximum load was intentionally limited to 27.46 Nm, which is below the rated maximum torque of 32.8 Nm, to provide a safety margin considering the lower flash points and higher volatility of butanol-containing blends and to ensure engine durability throughout the extended test campaign. Each fuel was tested, and prior to each measurement, the fuel system was thoroughly purged and the engine was allowed to reach steady-state thermal conditions.

At each load point, emissions data were recorded simultaneously at P₁ and P₃ after allowing 30 s of steady-state operation following stabilization. Figure 2 illustrates the real-time output of the time-resolved measurement profile for one experimental repetition of the B30D60BUT10 blend at 10.98 Nm. The profile shows the engine-out NO_x concentration measured at P₁ (approximately 1800 ppm), followed by the post-SCR concentration measured at P₃ (approximately 950 ppm) after the exhaust probe was relocated to the SCR outlet. Furthermore, the graphs in Fig. 2 depict the simultaneous time-dependent variations of fuel mass, T₁, T₂, and T₃ exhaust gas temperatures (EGT), and engine load (in kg), with the data updated every second. The stability of these thermal profiles throughout the measurement window confirms steady-state operation during data acquisition.

Fig. 2.

Real-time variation of NO_x emissions, exhaust temperatures, fuel mass, and engine load during sampling probe displacement.

In-cylinder pressure traces (50 consecutive cycles) were acquired in parallel with the emission measurements, while the corresponding exhaust temperatures were logged continuously. A schematic representation of the complete experimental arrangement is provided in Fig. 3.

Fig. 3.

The schematic view of the engine test setup.

Results and discussion

Combustion and performance analyses

In-cylinder pressure and heat release rate (HRR)

Figure 4 presents the variation of in-cylinder pressure and HRR curves of the test fuels under different engine load conditions. Across all loads, both in-cylinder pressure and HRR increased markedly with increasing engine load. While peak cylinder pressure values ranged between approximately 64–66 bar at 10.98 Nm, they increased to 71–75 bar at 27.46 Nm. This behavior is mechanistically governed by the greater total fuel energy input at elevated loads: a larger injected fuel mass releases more chemical energy per cycle, elevating both the adiabatic flame temperature and the mean in-cylinder pressure during the expansion stroke^9,10,39.

Fig. 4.

The influences of test fuels on cylinder pressure and HRR.

At the lowest load condition (10.98 Nm), neat diesel (D100) exhibited the highest peak cylinder pressure at 66.31 bar. Among the butanol-diesel blends, D95BUT5 (66.23 bar) and D85BUT15 (66.03 bar) produced values close to D100, whereas increasing butanol content showed a slight tendency to reduce peak pressure. This attenuation can be attributed to the lower calorific value of n-butanol (33.1 MJ/kg vs. 43.4 MJ/kg for diesel), which reduces the total heat input per cycle despite the improved premixed combustion quality. Blends containing B30 generally exhibited lower peak pressure values than D100 across all load conditions. This behavior originates from two compounding physical effects: first, the higher kinematic viscosity of SOME impairs fuel atomization, producing larger droplet sizes and slower evaporation rates that limit the fuel mass participating in the premixed phase; second, the lower net heating value of biodiesel directly reduces the peak pressure rise per unit of injected fuel mass^29,40,41.

At 16.48 Nm, D100 (71.95 bar) continued to produce the highest cylinder pressure. At 21.97 Nm, the D90BUT10 blend (71.80 bar) reached a peak pressure nearly identical to D100 (71.76 bar), demonstrating that a moderate butanol fraction of 10 vol% represents a thermodynamic optimum at medium-to-high loads. The underlying mechanism involves the high latent heat of vaporization of n-butanol (582 kJ/kg vs. 232 kJ/kg for diesel), which promotes charge cooling during the intake and compression strokes, increases the effective air-to-fuel ratio in the premixed zone, and extends the ignition delay, collectively increasing the fuel mass burned in the premixed phase and producing a sharper, more energetic heat release event^13,42,43. Under the highest load condition (27.46 Nm), D85BUT15 produced the highest peak cylinder pressure at 75.04 bar, followed by B30D70 (74.90 bar) and D90BUT10 (74.48 bar). At this load, the elevated in-cylinder temperature sufficiently overcomes the evaporative penalty of both butanol and biodiesel, thereby allowing their fuel-bound oxygen to promote more complete oxidation and sustain a high peak pressure despite the lower calorific values of the blends^42,43.

A notable anomaly is observed for the B30D60BUT10 blend, which exhibited significantly lower peak pressures at 16.48 Nm (67.20 bar) and 21.97 Nm (63.66 bar). This behavior reflects a critical synergistic interaction between the physical and chemical properties of the two alternative fuel components. The high viscosity of B30 impairs spray penetration and droplet breakup, while the low cetane number of n-butanol (CN ≈ 17) simultaneously extends the chemical ignition delay. Under medium-load conditions, the in-cylinder temperature at the end of compression is insufficient to fully compensate for these compounding limitations: the thermal environment neither provides adequate energy for rapid droplet evaporation nor sufficient activation energy for butanol autoignition within the available time window before TDC. Consequently, a disproportionately large fraction of the fuel–air mixture burns in the post-TDC diffusion-controlled phase, producing a lower and broader pressure trace. At low load, reduced fuel quantity masks this interaction, while at high load, elevated temperatures and pressures restore adequate atomization and ignition kinetics. This load-specific vulnerability of the B30D60BUT10 blend confirms that the interplay between viscosity-driven atomization limitations and cetane-driven ignition chemistry is non-linear and depends critically on the thermodynamic state of the cylinder charge^40,41,44.

An evaluation of the HRR curves reveals that butanol-blended diesel fuels enhance the premixed combustion phase, particularly under medium and high load conditions¹³. This enhancement originates from the dual physicochemical role of n-butanol: its lower boiling point (117.7 °C vs. ~ 180–360 °C for diesel) accelerates fuel vaporization and promotes a more homogeneous charge distribution during the ignition delay period, while its molecular oxygen content (21.6 wt.%) increases the local equivalence ratio toward unity in the premixed zone, thereby accelerating the elementary reaction rates during the initial combustion phase⁴⁵. At elevated butanol concentrations (15 vol%), these effects produce a discernible double-peak pattern in the HRR diagram: the first peak, occurring before TDC, reflects the rapid oxidation of the well-mixed premixed charge; the second peak corresponds to the diffusion-controlled combustion of fuel injected after the onset of heat release⁴⁵. At 10.98 Nm, HRR peaks occurred before TDC, approximately between − 6 and − 7 °CA. Among the tested fuels, D95BUT5 exhibited the highest HRR peak at 33.57 J/°CA, whereas B30-based blends remained at relatively lower levels of 29–30 J/°CA^13,39.

At 16.48 Nm, D85BUT15 (37.49 J/°CA) and D90BUT10 (37.19 J/°CA) exceeded the HRR value of D100 (36.78 J/°CA), confirming that the charge-cooling and homogenization effects of butanol become increasingly effective as the in-cylinder temperature rises with load, thereby providing a stronger thermal driving force for premixed combustion. Notably, the leading fuel at this load shifted from D95BUT5 (which led at 10.98 Nm) to D85BUT15 and D90BUT10, reflecting the load-dependent balance between two competing effects: the charge-cooling benefit of higher butanol concentrations, which intensifies the premixed combustion peak, and the calorific value penalty, which limits total heat release per cycle. At medium loads, the 10–15 vol% butanol range therefore represents the optimal blending window, maximizing the premixed fuel fraction without incurring an excessive reduction in total energy input^13,39.

Under higher load conditions (21.97 and 27.46 Nm), D90BUT10 achieved the highest HRR peak values among all fuel-load combinations, reaching 38.99 and 39.08 J/°CA, respectively. The superiority of D90BUT10 over D85BUT15 at these loads suggests that 10 vol% butanol represents the optimum blending ratio under the tested conditions: beyond this concentration, the calorific value deficit and the increasingly oxygen-diluted diffusion flame begin to suppress the total heat release rate, partially offsetting the premixed combustion gains^13,42. In contrast, B30D60BUT10 exhibited the lowest HRR value at 21.97 Nm (27.00 J/°CA), consistent with the anomaly described above for cylinder pressure.

Although B30-based blends generally produced lower HRR values compared to D100 and butanol-diesel blends across all load conditions, the B30D55BUT15 blend showed relatively improved HRR at high load, reaching 35.49 J/°CA at 27.46 Nm. This partial recovery suggests that the elevated butanol fraction (15 vol%) provides sufficient charge-cooling and oxygen enrichment to partially offset the viscosity-induced atomization penalty of the B30 component at high in-cylinder temperatures, where improved spray breakup narrows the performance gap between biodiesel-containing and pure butanol-diesel blends^13,39,42.

Ignition delay (ID) and combustion duration (CD)

An analysis of the ID results presented in Fig. 5a shows that, for all fuels, ID generally exhibits a slight increase with increasing engine load, although this trend tends to stabilize at higher loads. This behavior reflects the competing thermochemical effects governing autoignition: on one hand, the elevated end-of-compression temperature and pressure at higher loads accelerate pre-ignition reaction kinetics in accordance with Arrhenius-type rate expressions, reducing the chemical component of the ignition delay; on the other hand, the greater injected fuel mass increases the total evaporative heat demand and extends droplet lifetime within the spray, thereby prolonging the physical component of the ignition delay associated with fuel–air mixing. The near-stabilization of ID at high loads indicates that the thermodynamic acceleration of chemical kinetics progressively dominates over the physical evaporation delay as in-cylinder conditions intensify⁴⁶.

Fig. 5.

Variations of (a) ID and (b) CD with test fuels under different engine loads.

From a fuel-type perspective, butanol-blended diesel fuels, particularly D90BUT10 and D85BUT15, exhibited longer ID durations than D100, reaching approximately 14.0°CA at 16.48 Nm. This extension is primarily attributable to the low cetane number of n-butanol (CN ≈ 17 vs. CN ≈ 51 for diesel), which reflects its inherently slower pre-ignition chemistry. The cetane number governs the ease with which low-temperature oxidation reactions, specifically the cool-flame reactions associated with alkylperoxy radical isomerization pathways, proceed prior to the main ignition event. In n-butanol, the absence of low-temperature reactivity pathways characteristic of long-chain n-alkanes means that the chemical ignition delay is substantially extended, requiring higher activation energies before the transition to hot ignition occurs^9,46. In addition to this chemical mechanism, the high latent heat of vaporization of n-butanol (582 kJ/kg vs. 232 kJ/kg for diesel) absorbs thermal energy from the surrounding charge during evaporation, thereby locally reducing the gas temperature in the spray zone and imposing a supplementary thermal component on the ignition delay. Taken together, these two effects, namely the suppression of cool-flame pre-ignition chemistry and the evaporative charge cooling, provide additional time for fuel–air mixture formation prior to TDC and directly contribute to the higher premixed HRR peaks observed in Fig. 4⁴⁶.

In contrast, the B30D70 blend exhibited the shortest ID value among all test fuels, reaching 11.0°CA at 10.98 Nm. This behavior is mechanistically linked to the molecular structure of SOME: the ester functional group (–COOCH₃) incorporated within the fatty acid chains provides fuel-bound oxygen that promotes the formation of reactive OH radicals during the pre-ignition phase, lowering the effective activation energy for ignition and advancing the onset of exothermic reactions^40,46. Furthermore, the higher cetane number of SOME (CN ≈ 52–56) reflects its greater susceptibility to low-temperature oxidation pathways, facilitating earlier cool-flame activity and shortening the overall ignition delay. For ternary B30-based blends containing butanol, the cetane-enhancing effect of biodiesel and the cetane-reducing effect of butanol exert opposing influences on the pre-ignition chemistry, yielding intermediate ID values that shift progressively with butanol fraction. However, the cetane contribution of SOME is insufficient to fully counteract the chemical ignition delay penalty of butanol, resulting in a net ID extension relative to B30D70 across all load conditions^40,46.

An examination of the CD results presented in Fig. 5b indicates that CD increases consistently with increasing engine load for all fuels, ranging from 55–59°CA at 10.98 Nm to 65–72°CA at 27.46 Nm¹³. This trend is governed by the increasing fuel mass injected at higher loads, which not only raises the total chemical energy that must be oxidized per cycle but also sustains a larger diffusion-controlled combustion zone in the post-TDC period, where combustion rate is limited by the local mixing rate between fuel vapor and oxidizer rather than by chemical kinetics¹³.

In terms of fuel type, D100 produced the shortest or near-shortest CD values under all load conditions, whereas B30-based blends generally exhibited longer combustion durations¹³. This behavior is mechanistically rooted in the higher kinematic viscosity and density of SOME relative to diesel: increased viscosity impairs spray atomization by reducing the aerodynamic breakup forces acting on fuel ligaments, resulting in a larger Sauter mean diameter (SMD) of the droplet population. Larger droplets require a proportionally longer evaporation time before sufficient vapor concentration builds up for sustained combustion, thereby extending the diffusion-controlled tail of the heat release event and widening the overall CD^13,47.

A particularly notable finding is that the B30D60BUT10 blend yielded the longest CD value among all test fuels at 21.97 Nm, reaching 68°CA. As discussed in “In-cylinder pressure and heat release rate (HRR)” Section, this anomaly reflects the adverse synergistic interaction between the viscosity-driven atomization limitations of B30 and the cetane-number-driven ignition delay of butanol under medium-load thermodynamic conditions, which collectively shift a disproportionately large fraction of the total heat release into the late diffusion-controlled combustion phase, broadening the CD beyond that of either component blended individually¹³.

For butanol-blended diesel fuels, a different trend was observed at high load. At 27.46 Nm, D85BUT15 and D95BUT5 produced CD values of 70 and 71°CA, respectively, compared to 65°CA for D100. This prolongation reflects the slower oxidation kinetics of the n-butanol molecule in the diffusion combustion phase: the C₄ alcohol chain, while reactive in the premixed zone due to its oxygen content, exhibits a lower adiabatic flame temperature and a reduced laminar burning velocity compared to the long-chain alkanes dominant in diesel, resulting in a slower burnout of the diffusion-controlled fuel fraction at the periphery of the spray^13,42. Importantly, however, this extended late-combustion phase does not translate into reduced efficiency at medium loads, because, as demonstrated in Fig. 4, the stronger premixed combustion peak generated by the extended ID of butanol blends releases a greater proportion of the total energy before TDC, partially compensating for the slower diffusion-phase burnout^13,39.

Combustion phasing (CA10, CA50)

Figure 6a presents the variation of CA10 across test fuels and load conditions. With increasing engine load, the CA10 position shifts progressively closer to TDC for all fuels, with values decreasing from 7–8°CA before TDC at 10.98 Nm to 4–6°CA before TDC at 27.46 Nm. Although this trend may appear counterintuitive given that greater fuel mass is associated with longer ID, it reflects a fundamental shift in the balance between physical and chemical ignition delay components at elevated loads. The substantially higher end-of-compression temperature and pressure at increased loads accelerate the Arrhenius-governed pre-ignition reaction rates to a degree that more than compensates for the additional evaporative demand of the larger fuel quantity, resulting in a net advancement of the autoignition event and consequently an earlier CA10^9,40. This finding is fully consistent with the ID convergence behavior discussed in “Ignition delay (ID) and combustion duration (CD)” Section, wherein the chemical kinetic acceleration at high loads progressively dominates over the physical mixing delay.

Fig. 6.

Variations of (a) CA10 and (b) CA50 with test fuels under different engine loads.

In terms of fuel type, B30D70 and other B30-based blends exhibited earlier or comparable CA10 values relative to D100 under low and medium load conditions. This advancement is mechanistically driven by the molecular oxygen embedded within the ester chains of SOME, which promotes local OH radical production during the pre-ignition phase, lowering the effective activation barrier for cool-flame initiation and advancing the transition to the main combustion event. Additionally, the higher cetane number of SOME facilitates alkylperoxy radical isomerization at lower temperatures, further contributing to an earlier CA10 under conditions where in-cylinder temperatures are still moderate^16,29.

In contrast, butanol-blended diesel fuels (D95BUT5, D90BUT10, D85BUT15) showed CA10 values delayed by approximately 1–2°CA compared to D100, particularly at higher loads. This retardation reflects the suppression of low-temperature pre-ignition chemistry by n-butanol: the β-scission decomposition pathways dominant in butanol oxidation generate primarily C₂-C₃ radical fragments rather than the reactive peroxy intermediates responsible for cool-flame activity in n-alkanes, thereby delaying the energy release onset. Furthermore, the charge-cooling effect of butanol’s high latent heat of vaporization reduces the local gas temperature in the spray zone during the ignition delay period, adding a thermal component to the retardation that becomes more pronounced as butanol concentration increases^13,40,48. The delayed CA10 of butanol blends is thus mechanistically linked to the extended ID reported in “Ignition delay (ID) and combustion duration (CD)” Section and directly explains the enhanced premixed HRR peaks observed in Fig. 4.

As shown in Fig. 6b, CA50 shifts progressively toward later crank angles with increasing engine load for all fuels, increasing from 6 to 8°CA after TDC at 10.98 Nm to 15–16°CA after TDC at 27.46 Nm. This retardation of CA50 with load is governed by the increasing dominance of diffusion-controlled combustion at higher fuel injection quantities: as the injected fuel mass grows, a larger fraction of the total charge must be oxidized through the slower mixing-limited diffusion pathway, which extends the crank angle interval required to reach the 50% mass fraction burned milestone and shifts CA50 progressively away from TDC^13,48.

At the lowest load condition, D100, D90BUT10, and D85BUT15 exhibited the earliest CA50 values at 6°CA after TDC, whereas all B30-based blends showed delayed combustion phases at approximately 8°CA. This 2°CA retardation of B30-containing blends reflects the spray quality penalty imposed by the higher kinematic viscosity of SOME: impaired atomization produces a coarser droplet population with a larger Sauter mean diameter, which slows the rate of fuel vapor generation in the diffusion zone and delays the consumption of the second half of the fuel mass, ultimately retarding CA50. The thermodynamic consequence of this delayed CA50 is a reduction in the effective expansion ratio during the power stroke, which contributes to the lower thermal efficiency of B30-based blends at low load conditions⁴⁰.

With increasing load, the fuel-type divergence in CA50 became more pronounced before subsequently narrowing. At 16.48 Nm, B30D65BUT5 and B30D60BUT10 exhibited the most delayed CA50 values at 13°CA after TDC, compared to 9°CA for D100. This amplified divergence arises from the compounding of two load-sensitive effects: the viscosity-driven atomization penalty of B30 intensifies as the injected fuel mass increases, producing a larger absolute quantity of coarse droplets that must evaporate through diffusion, while the simultaneously extended ID of the butanol component shifts a greater proportion of the heat release to the post-TDC period. The pronounced delay of B30D60BUT10 at this load is fully consistent with the anomalous pressure and HRR behavior identified in “In-cylinder pressure and heat release rate (HRR)” Section, confirming that medium-load conditions represent the most unfavorable operating window for this blend ratio^40,49.

At higher load conditions (21.97 and 27.46 Nm), CA50 values converged strongly across all fuels. This thermodynamic convergence occurs because the elevated in-cylinder temperature and pressure at high loads accelerate both droplet evaporation and diffusion flame oxidation rates to the point where fuel-specific differences in viscosity, cetane number, and latent heat become relatively insignificant compared to the overwhelming influence of the thermodynamic state of the cylinder charge. In essence, the high-load thermal environment enforces a common combustion phasing on all fuels regardless of their individual physicochemical properties, explaining the collapse of inter-fuel CA50 differences observed experimentally^13,48.

Combustion phasing diagram (CA10, CA50, CA90, CD)

Figure 7 presents a comprehensive crank-angle-based combustion phasing diagram consolidating the sequential progression of three distinct combustion phases, namely the ID, the early-to-mid combustion phase (CA10 → CA50), and the late burnout phase (CA50 → CA90), for all test fuels across the four engine load conditions. CD is accordingly defined as the crank angle interval from CA10 to CA90. By integrating the individual combustion parameters discussed in “In-cylinder pressure and heat release rate (HRR)” through “Combustion phasing (CA10, CA50)” Section into a unified visual framework, the diagram enables a direct comparison of overall combustion phasing and reveals which phase is most sensitive to fuel composition at each load condition.

Fig. 7.

Crank-angle-based combustion phasing diagram.

Regarding the ignition delay phase, B30D70 exhibits the longest ID at the lowest load (13.0°CA at 10.98 Nm), reflecting the compounding effects of the higher kinematic viscosity of SOME, which delays physical fuel evaporation, and its paradoxically higher cetane number, which partially counters this effect but is insufficient to overcome the viscosity penalty at low in-cylinder temperatures. Butanol–diesel blends (D95BUT5–D85BUT15) yield slightly shorter IDs of 11.0°CA at the same condition, consistent with the cetane-number and latent-heat mechanisms established in “Ignition delay (ID) and combustion duration (CD)” Section. As engine load increases, ID converges across all fuels to a narrow range of 10.0–11.0°CA at 27.46 Nm, reflecting the progressive dominance of Arrhenius-governed pre-ignition kinetics over fuel-specific physical evaporation delays as end-of-compression temperatures and pressures intensify^41,48.

The CA10 → CA50 phase widens moderately with increasing load for all fuels, ranging from 13–16°CA at 10.98 Nm to 20–22°CA at 27.46 Nm, with B30-based blends consistently occupying the upper bound of this range due to their coarser droplet population and slower fuel vapor generation rate, as discussed in “Combustion phasing (CA10, CA50)” Section. The late burnout phase (CA50 → CA90) constitutes the dominant portion of the total combustion event, spanning 42–44°CA across all fuels at low load. At higher loads, this phase extends notably for oxygenated blends, reaching 48–51°CA for butanol- and biodiesel-containing fuels compared to 45°CA for D100. This extension reflects two compounding mechanisms: the slower C₄-chain oxidation kinetics of n-butanol in the diffusion combustion phase, which reduces the laminar burning velocity of the peripheral flame zone, and the increasing dominance of mixing-limited diffusion combustion at higher fuel injection quantities, which spreads the tail of the heat release event over a progressively wider crank angle interval^41,48.

Consequently, total CD ranges from 55–59°CA at 10.98 Nm to 65–72°CA at 27.46 Nm. D100 exhibits the shortest CD across all loads (65°CA at 27.46 Nm), while B30D70 and B30D65BUT5 yield the longest (72°CA), consistent with the CA50 and CD analyses reported in “Ignition delay (ID) and combustion duration (CD)” and “Combustion phasing (CA10, CA50)” Section^40,41,48. The phasing diagram thus confirms that fuel composition exerts its most pronounced influence on the late burnout phase and total combustion duration at higher loads, where the viscosity-driven atomization penalty of biodiesel and the slower diffusion-phase oxidation kinetics of butanol collectively extend CA50 → CA90, while the ignition delay phase converges as elevated in-cylinder thermodynamic conditions become the overriding governing factor^41,48.

Combustion stability (MPRR and RI)

An examination of the maximum pressure rise rate (MPRR) results presented in Fig. 8a reveals that the overall trend exhibits a complex dependence on both engine load and fuel type. MPRR is mathematically defined as the maximum of dP/dθ over the combustion cycle and is therefore directly governed by the rate at which chemical energy is released in the premixed combustion phase: a larger premixed fuel mass burning over a shorter crank angle interval produces a steeper pressure gradient and a higher MPRR. Accordingly, MPRR tends to increase with engine load as the total fuel mass and the associated premixed fraction both grow, subject to the fuel-specific ignition characteristics¹⁶

Fig. 8.

Variations of (a) MPRR and (b) RI with test fuels under different engine loads.

Under low and medium load conditions (10.98 and 16.48 Nm), D100 and butanol-blended diesel fuels produced higher MPRR values than B30-based blends. In particular, D90BUT10 and D100 reached the highest values at 16.48 Nm, with 8.41 and 9.09 bar/°CA, respectively. This behavior is mechanistically consistent with the longer ignition delays of butanol-diesel blends established in “Ignition delay (ID) and combustion duration (CD)” Section: the extended ID allows a greater mass of fuel to accumulate in the vapor phase prior to autoignition, so that when combustion initiates, a larger premixed charge burns simultaneously and generates a steeper dP/dθ. The positive correlation between ID and MPRR observed here is also supported by the elevated HRR peaks documented in Fig. 4^16,40.

At high load (27.46 Nm), this trend became more pronounced, with D85BUT15 (9.02 bar/°CA) and D90BUT10 (8.94 bar/°CA) exhibiting the highest MPRR values. As the butanol fraction increases, both the cetane-number-driven extension of chemical ignition delay and the latent-heat-driven evaporative charge cooling act cumulatively to increase the premixed fuel accumulation period, amplifying the pressure rise gradient upon ignition. While elevated MPRR values indicate more intense and efficient premixed combustion, values exceeding approximately 10 bar/°CA are generally associated with increased structural loading on engine components and heightened knock risk; the fact that all tested blends remained below this threshold confirms the mechanical safety of the evaluated fuel formulations^13,40.

In contrast, B30-based blends consistently showed lower MPRR values across all load conditions. The higher kinematic viscosity of SOME impairs spray atomization and reduces the rate of fuel evaporation, thereby limiting the premixed fuel fraction and producing a more gradual pressure rise characteristic of diffusion-dominated combustion. The lower net calorific value of biodiesel further constrains the total energy release per unit crank angle, collectively resulting in a more controlled and mechanically benign pressure rise profile^40,50.

A particularly notable observation is that the B30D60BUT10 blend produced the lowest MPRR value among all test fuels at 21.97 Nm, with 5.92 bar/°CA. This result is the direct mechanistic consequence of the anomaly described in “In-cylinder pressure and heat release rate (HRR)” and “Ignition delay (ID) and combustion duration (CD)” Sections: under medium-load conditions, the compounding of biodiesel viscosity and butanol cetane-number effects shifts combustion disproportionately toward the post-TDC diffusion-controlled phase, where the fuel burns at a slower rate governed by turbulent mixing rather than by chemical kinetics. The resulting heat release is spread over a wider crank angle interval, flattening the dP/dθ profile and yielding an anomalously low MPRR that reflects degraded rather than controlled combustion⁴⁰.

The ringing intensity (RI) results presented in Fig. 8b are broadly consistent with the MPRR trends, as expected from the mathematical structure of the RI expression (Eq. 6), in which RI scales with the square of MPRR and is additionally modulated by the maximum in-cylinder gas temperature () and pressure (). D100 and butanol-blended diesel fuels exhibited higher RI values than B30-based blends under low and medium load conditions, with D100 producing the highest RI of 7.64 MW/m² at 10.98 Nm. The elevated RI of butanol blends at medium-to-high loads reflects both their higher MPRR and their higher T_max, the latter arising from the more intense premixed combustion phase generated by the extended ID¹⁶.

At higher load conditions (21.97 and 27.46 Nm), D85BUT15 emerged as the fuel with the highest RI, reaching 8.36 and 8.69 MW/m², respectively. This outcome reflects the compounding of three load-sensitive effects: the increasingly extended ID of D85BUT15 produces greater premixed fuel accumulation and a higher MPRR; the stronger premixed combustion raises T_max, which amplifies RI through the temperature-dependent term in the RI formula; and the elevated in-cylinder pressure at high loads further modulates the RI through its P_max term, albeit in the denominator^13,42.

B30-based blends generally exhibited lower RI values across all conditions, consistent with their lower MPRR and the moderate peak temperatures associated with diffusion-dominated combustion. The reduced knocking tendency of biodiesel blends is further supported by the higher cetane number of SOME, which promotes earlier and more gradual heat release, distributing the energy input over a wider crank angle window and limiting the instantaneous pressure oscillation amplitude²⁷.

However, a notable exception was observed at 16.48 Nm, where the B30D65BUT5 blend produced an unexpectedly high RI value of 10.06 MW/m², exceeding all other fuels at this load including D100. This anomaly is mechanistically attributed to a specific thermodynamic interaction between the 5 vol% butanol fraction and the B30 biodiesel base at this intermediate load condition: at 16.48 Nm, the in-cylinder temperature is sufficient to activate the charge-cooling and evaporation-enhancement effects of butanol without yet providing the high-temperature thermal environment that would moderate the pressure oscillation amplitude at full load. The result is a localized intensification of the premixed combustion event, potentially associated with micro-explosion phenomena at the droplet scale where butanol and biodiesel components evaporate at different rates, generating abrupt local pressure gradients that manifest as elevated RI^40,49.

Engine performance (BSFC, BTE)

An examination of Fig. 9a, which presents the brake specific fuel consumption (BSFC) results, shows that BSFC values for all fuels decrease consistently as engine load increases from 10.98 Nm to 21.97 Nm, after which a slight increase is observed at 27.46 Nm^10,46. The decreasing BSFC trend at low-to-medium loads reflects the improving combustion efficiency associated with higher in-cylinder temperatures and pressures: as load increases, the thermodynamic conditions accelerate fuel oxidation kinetics, reduce the fraction of fuel energy lost to incomplete combustion, and improve the indicated thermal work per cycle. At maximum load, however, the slight BSFC increase is attributable to the combined effects of extended combustion duration beyond TDC, increased friction and pumping losses at higher cylinder pressures, and the enhanced convective heat transfer to the cylinder walls driven by elevated in-cylinder gas temperatures, which collectively reduce the fraction of released chemical energy converted to useful brake work⁹.

Fig. 9.

Effects of test fuels on (a) BSFC and (b) BTE under different engine loads.

From a fuel-type perspective, the D90BUT10 blend achieved the lowest BSFC value among all fuels at 21.97 Nm, with 241.01 g/kWh, followed by D85BUT15 (254.59 g/kWh) and D95BUT5 (259.83 g/kWh). This result is mechanistically explained by the dual role of n-butanol at moderate blending ratios: its lower boiling point and surface tension improve spray atomization and droplet breakup relative to neat diesel, producing a finer droplet population that evaporates more rapidly and generates a more homogeneous air–fuel mixture; simultaneously, its molecular oxygen content enhances local oxidation rates in the premixed zone, reducing the quantity of partially oxidized species that would otherwise represent thermodynamic losses. Together, these effects increase the fraction of fuel chemical energy converted to useful brake work per unit mass of fuel consumed, yielding a net reduction in BSFC despite the lower calorific value of butanol^13,40.

On the other hand, B30-based blends exhibited higher BSFC values than D100 under all load conditions. This behavior is governed by two compounding mechanisms: first, the lower net calorific value of SOME (approximately 37.5 MJ/kg vs. 43.4 MJ/kg for diesel) requires a proportionally greater fuel mass to deliver an equivalent energy input per cycle; second, the higher kinematic viscosity of B30 blends impairs spray atomization, increases droplet size, and slows fuel evaporation, collectively reducing combustion efficiency and further increasing the mass of fuel required to sustain the target brake output. In particular, the B30D65BUT5 blend reached the highest BSFC value of 307.81 g/kWh at 10.98 Nm, where the adverse viscosity effects are most pronounced due to the lower in-cylinder temperatures prevailing at minimal load^13,27.

The BTE results presented in Fig. 9b exhibit a trend fully consistent with the BSFC findings, as BTE is the thermodynamic inverse of BSFC when normalized by fuel heating value. For all fuels, BTE increases with engine load up to 21.97 Nm and then decreases slightly at 27.46 Nm, reflecting the same balance between improving combustion efficiency at moderate loads and increasing thermal and mechanical losses at maximum load^16,46.

The most notable result is that D90BUT10 achieved the highest BTE of the entire test matrix at 21.97 Nm, reaching 33.97%, which is 4.4 percentage points higher than D100 (29.53%). This performance advantage originates from the thermodynamic synergy between extended ignition delay and improved premixed combustion phasing: the longer ID of D90BUT10, driven by the cetane-number and latent-heat effects of butanol described in “Ignition delay (ID) and combustion duration (CD)” Section, promotes a more homogeneous charge distribution before ignition, concentrating a greater proportion of the heat release near TDC where the in-cylinder volume is minimum and the thermodynamic work conversion efficiency is maximum. The D85BUT15 blend also achieved a high BTE of 32.49% at the same load, confirming that the 10–15 vol% butanol range consistently maximizes thermal efficiency under medium-to-high load conditions in this engine^13,42.

In contrast, B30-based blends generally produced BTE values comparable to or slightly below D100. Notably, B30D55BUT15 achieved the highest BTE among B30 blends at 10.98 Nm with 29.66%, confirming that butanol addition partially mitigates the adverse effects of biodiesel by improving spray quality and advancing the premixed combustion fraction. However, B30D65BUT5 and B30D70 exhibited lower BTE than D100 at high load conditions, indicating that the viscosity-driven combustion quality penalty of high biodiesel fractions progressively outweighs the oxygen-content benefit as the thermodynamic demand on the fuel injection system intensifies^13,40.

Emissions analyses

Carbon monoxide (CO) emissions

The effects of the test fuels and the DOC + SCR system on CO emissions are presented in Fig. 10. An examination of the pre-DOC + SCR (P₁) results shows that CO emissions increase consistently with rising engine load for all test fuels, reaching a pronounced peak at 27.46 Nm, where D100 recorded 1.092 vol%^29,42. This behavior is governed by the progressive deterioration of local air–fuel stoichiometry at elevated loads: the greater injected fuel mass creates locally fuel-rich regions within the spray where the available oxygen is consumed by the initial oxidation steps, leaving insufficient O₂ for the complete oxidation of carbon intermediates through the CO → CO₂ pathway. CO formation is thermodynamically favored over CO₂ production in oxygen-deficient zones because the reaction C + ½O₂ → CO is kinetically faster than the subsequent oxidation step CO + ½O₂ → CO₂, which requires a higher activation energy and is strongly inhibited by the low local oxygen partial pressure⁵¹.

Fig. 10.

Combined effects of test fuels and engine load on CO emissions (a) before DOC + SCR (P₁) and (b) after DOC + SCR (P₃).

From a fuel-type perspective, blends containing biodiesel and butanol generally produced lower CO emissions than D100, owing to their higher inherent oxygen content and lower carbon-to-hydrogen (C/H) ratio^29,42. The fuel-bound oxygen in both SOME and n-butanol serves as an in-situ oxidant within the diffusion flame: oxygen atoms covalently bonded to the fuel molecule are released during the initial pyrolysis and decomposition reactions, creating a localized oxygen surplus in the immediate vicinity of the burning droplet that promotes the conversion of CO intermediates to CO₂ before they can escape the primary reaction zone. This mechanism is particularly effective for the B30D55BUT15 blend, which yielded the lowest CO values at both 10.98 Nm (0.032 vol%) and 21.97 Nm (0.107 vol%), reflecting the additive contributions of biodiesel oxygen (approximately 11 wt.%) and butanol oxygen (21.6 wt.%) to the in-cylinder oxidation environment^49,52.

As shown in Fig. 10b, the post-DOC + SCR (P₃) results demonstrate a dramatic reduction in CO emissions following aftertreatment. For all oxygenated blends, CO emissions decreased to zero, reflecting the highly effective catalytic oxidation of CO within the DOC substrate. The DOC facilitates the heterogeneous surface reaction CO + ½O₂ → CO₂ over the platinum-group metal catalyst, a process that proceeds at high efficiency once the catalyst surface temperature exceeds the CO light-off threshold of approximately 150–200 °C⁵³. Since the exhaust gas temperature at the DOC inlet (T₁) consistently exceeded 267 °C across all test conditions, the catalyst operated well above its light-off temperature, ensuring near-complete CO conversion for all oxygenated fuel blends whose engine-out CO concentrations were within the catalyst’s oxidative capacity⁵³.

For D100, a small residual CO value of 0.015 vol% was measured at 27.46 Nm, corresponding to a DOC conversion efficiency of approximately 98.6%. This residual CO is attributable to the mass transfer limitation of the catalytic reaction at high exhaust flow rates: under maximum load conditions, the increased volumetric flow of exhaust gas reduces the residence time of CO molecules within the catalyst channels, limiting the contact time available for the heterogeneous oxidation reaction and allowing a small fraction of CO to pass through unconverted despite the favorable operating temperature⁵³.

Carbon dioxide (CO₂) emissions

The CO₂ emission results are presented in Fig. 11. An examination of the pre-DOC + SCR (P₁) values in Fig. 11a indicates a consistent increase in CO₂ emissions with rising engine load for all test fuels^28,29,46. Specifically, CO₂ levels ranged between 4.31–4.60 vol% at 10.98 Nm and increased to 10.78–11.56 vol% at 27.46 Nm. This trend reflects the direct relationship between fuel mass flow rate and the total carbon oxidized per unit time: at higher loads, a proportionally greater quantity of fuel carbon undergoes complete combustion to CO₂, while the simultaneous reduction in excess air ratio (as documented in “Exhaust oxygen content and excess air ratio (O₂ and λ)” Section) concentrates CO₂ within a smaller total exhaust gas volume, amplifying the measured vol% values.

Fig. 11.

Combined effects of test fuels and engine load on CO₂ emissions (a) before DOC + SCR (P₁) and (b) after DOC + SCR (P₃).

D100 produced comparatively lower CO₂ emissions at low and medium load conditions relative to B30-based blends. This behavior is mechanistically linked to the lower oxygen content of neat diesel: in oxygen-deficient combustion zones, a portion of the carbon that would otherwise be fully oxidized to CO₂ is instead converted to CO or remains as unburned hydrocarbons, as discussed in “Carbon monoxide (CO) emissions” Section. Consequently, the incomplete carbon oxidation pathway diverts carbon away from CO₂ production, yielding lower measured CO₂ concentrations^41,51. In contrast, B30-based blends promoted more complete carbon oxidation due to the fuel-bound oxygen of SOME, which sustains CO → CO₂ conversion even in locally fuel-rich zones, resulting in higher CO₂ emissions at low and medium loads. Accordingly, B30D60BUT10 and B30D65BUT5 generated higher CO₂ levels than D100 at 10.98 and 16.48 Nm, consistent with their superior CO oxidation performance documented in Fig. 10.

As illustrated in Fig. 11b, the post-DOC + SCR (P₃) CO₂ concentrations were substantially lower than the corresponding P₁ values across all fuels, with reductions ranging from 30 to 60% depending on fuel type and load condition. Importantly, this reduction does not reflect a removal or chemical conversion of CO₂ within the aftertreatment system, as CO₂ is a thermodynamically stable end-product of carbon oxidation that is not catalytically reduced under typical exhaust conditions. Rather, the observed decrease in CO₂ vol% is primarily attributable to the dilution of the exhaust gas stream by the injected DEF: the aqueous urea solution introduces a substantial water and ammonia vapor load into the exhaust, increasing the total molar flow of gas through the SCR unit and thereby reducing the molar fraction of CO₂ in the overall mixture. Additionally, the SCR reactions produce N₂ and H₂O as products, further expanding the total gas volume and diluting all species concentrations including CO₂^48,53. At the highest load (27.46 Nm), the lowest CO₂ concentration after DOC + SCR was recorded for B30D60BUT10 at 3.844 vol%, followed by B30D65BUT5 (4.234 vol%) and D95BUT5 (4.749 vol%), a ranking consistent with the relative DEF injection rates at this load condition. Unlike CO, which is catalytically oxidized to near-zero levels within the DOC, CO₂ cannot be reduced to zero by any mechanism operating within the aftertreatment system, as it represents the fully oxidized terminal state of all carbon-containing species in the exhaust.

Hydrocarbon (HC) emissions

The effects of the test fuels and the DOC + SCR system on hydrocarbon (HC) emissions are presented in Fig. 12. An evaluation of the pre-DOC + SCR (P₁) results in Fig. 12a shows that D100 produced the highest HC emissions under all load conditions, with values increasing from 151.44 ppm to 215.78 ppm as engine load increased. HC emissions in diesel engines originate from multiple physicochemical sources: the quench layer adjacent to the cold cylinder wall, where the flame is extinguished before complete oxidation; the crevice volumes between the piston rings and cylinder liner, where the fuel–air mixture is trapped and cannot sustain combustion; and locally over-lean or over-rich zones within the spray that either fall outside the flammability limits or suffer from insufficient oxygen for complete oxidation. The increase in HC with load for D100 reflects the growing contribution of fuel-rich spray regions under higher injection quantities, where the low inherent oxygen content of diesel is insufficient to sustain complete carbon chain oxidation^9,29,35.

Fig. 12.

Combined effects of test fuels and engine load on HC emissions (a) before DOC + SCR (P₁) and (b) after DOC + SCR (P₃).

Butanol-blended diesel fuels exhibited a pronounced reduction in HC emissions, with D85BUT15 yielding the lowest HC levels across all load conditions and reaching a minimum of 54.77 ppm at 10.98 Nm. This reduction is mechanistically driven by two synergistic properties of n-butanol: first, its lower kinematic viscosity and surface tension relative to diesel improve spray atomization by reducing the energy barrier to droplet breakup, producing a finer droplet population with a smaller Sauter mean diameter that evaporates more rapidly and reduces the likelihood of wall wetting and crevice trapping; second, the molecular oxygen content of n-butanol (21.6 wt.%) provides an in-situ oxidant that sustains the oxidation of hydrocarbon radicals within the diffusion flame, preventing their escape from the primary reaction zone as unburned species. Together, these mechanisms reduce HC emissions proportionally with butanol concentration, explaining the monotonic decrease observed from D95BUT5 to D85BUT15^16,39,54.

B30-based blends, although producing lower HC emissions than D100, showed higher values compared to butanol-diesel blends across all load conditions. This behavior is rooted in the molecular structure of SOME: the long-chain fatty acid methyl ester (FAME) molecules (C16-C18) have inherently slower oxidation kinetics compared to the shorter-chain alkanes dominant in diesel, requiring higher activation energies and longer residence times for complete combustion. Furthermore, the higher kinematic viscosity of SOME impairs spray atomization and increases the probability of droplet impingement on the cylinder wall, where fuel deposits form a liquid film that burns incompletely during the expansion stroke, contributing disproportionately to HC emissions^43,51.

As illustrated in Fig. 12b, the post-DOC + SCR (P₃) results indicate that the system reduced HC emissions by approximately 10–35% across all fuel and load combinations^23,55. This conversion efficiency is substantially lower than that achieved for CO (approximately 98.6%), reflecting the fundamental difference in catalytic oxidation kinetics between CO and unburned hydrocarbons. While CO oxidation over platinum-group metal catalysts proceeds through a simple Langmuir–Hinshelwood surface mechanism with a relatively low activation energy and a light-off temperature of approximately 150–200 °C, the catalytic oxidation of long-chain hydrocarbon molecules requires a significantly higher surface temperature to overcome their greater activation energy barrier, with HC light-off typically occurring above 250–300 °C depending on the molecular weight and structure of the hydrocarbon species. Since the T₁ values at low loads (267–294 °C) approach but do not substantially exceed the HC light-off threshold, the DOC operates in a thermally marginal regime for HC oxidation at these conditions, explaining the lower and more variable conversion efficiencies relative to CO⁵³. At the highest load condition (27.46 Nm), the lowest HC after DOC + SCR was recorded for D85BUT15 at 62.58 ppm, followed by D90BUT10 at 69.00 ppm, consistent with their lower engine-out HC concentrations entering the DOC.

While the measured HC conversion efficiency of 10–35% highlights the relative resistance of long-chain hydrocarbon molecules to catalytic oxidation within the tested operating range, this does not indicate that ternary blends are unsuitable for integrated aftertreatment systems^9,53. Although partially converted high-molecular-weight hydrocarbons from sunflower biodiesel could theoretically deposit on the DOC and SCR catalyst surfaces through coking and pore-plugging mechanisms, potentially reducing the active catalyst surface area and inducing long-term deactivation, the presence of n-butanol in the ternary blend serves as a critical mitigating factor^23,53. n-Butanol reduces the total carbonaceous load entering the DOC by acting as a soot suppressor through two complementary pathways: its lower viscosity and higher vapor pressure promote more complete in-cylinder combustion, reducing the mass of unburned hydrocarbon species reaching the catalyst; and its molecular oxygen content lowers the C/H ratio of the combustion intermediates, shifting the product distribution away from high-molecular-weight polycyclic species toward lighter, more readily oxidized fragments^13,39. Furthermore, the combined fuel-bound oxygen of the biodiesel-butanol-diesel triad ensures that engine-out HC concentrations are already substantially lower than those of neat diesel, reducing the thermal and chemical stress imposed on the downstream catalysts^40,52. To comprehensively address long-term durability concerns, future research involving dedicated endurance tests and post-test catalyst characterization is recommended to quantify the deactivation rate of vanadium-based SCR systems under sustained operation with oxygenated ternary fuel blends^23,55.

Nitrogen oxides (NO_x) emissions

The effects of the test fuels and the DOC + SCR system on NO_x emissions are presented in Fig. 13, where the upper section illustrates the pre-DOC + SCR (P₁) values and the lower section displays the post-DOC + SCR (P₃) results. An analysis of the P₁ results reveals that B30-based blends produced significantly higher NO_x emissions than D100 under all load conditions. In particular, B30D65BUT5 and B30D60BUT10 reached the highest values in the test series at 16.48 Nm, with 2109.6 ppm and 2109.2 ppm, respectively. This behavior is mechanistically governed by the extended Zeldovich thermal NO_x formation pathway: the fuel-bound oxygen in SOME increases the local oxygen availability within the diffusion flame, elevating the adiabatic flame temperature and accelerating the rate-limiting reaction N₂ + O → NO + N, which exhibits a strong exponential dependence on temperature. The resulting NO_x increase of 60–109% relative to D100 reflects the compounding of higher flame temperature and extended residence time of combustion gases above the critical NO_x formation threshold of approximately 1800 K^29,51,53.

Fig. 13.

Combined effects of test fuels and engine load on NO_x emissions (a) before DOC + SCR (P₁) and (b) after DOC + SCR (P₃).

In contrast, butanol-blended diesel fuels (D90BUT10 and D85BUT15) exhibited lower NO_x emissions compared to D100 across most load conditions. This suppression originates from two thermochemical mechanisms operating simultaneously: first, the high latent heat of vaporization of n-butanol (582 kJ/kg) absorbs thermal energy from the surrounding charge during evaporation, reducing the peak adiabatic flame temperature and limiting the kinetic driving force for Zeldovich NO formation; second, the extended ID associated with the low cetane number of butanol promotes a more homogeneous charge distribution before ignition, reducing the peak local temperatures in the diffusion flame zone where thermal NO_x formation is most intense^13,16,40. For all fuels, NO_x emissions peaked within the 16.48–21.97 Nm load range, followed by a decline at 27.46 Nm. This decline is attributable to the richer air–fuel mixture prevailing at maximum injection rates, which creates locally oxygen-deficient zones that kinetically inhibit NO formation despite the elevated bulk gas temperature, as the Zeldovich pathway requires both high temperature and sufficient oxygen partial pressure to proceed at significant rates⁵⁵.

As shown in the lower section of Fig. 13, the post-DOC + SCR (P₃) data demonstrate that the system reduced NO_x emissions by approximately 40–68% across all fuel and load combinations. The highest conversion efficiencies were achieved at 27.46 Nm, reaching 68% for B30D60BUT10 and 66% for B30D65BUT5. The lowest absolute NO_x emission after DOC + SCR was recorded for D85BUT15 at 637.6 ppm at 27.46 Nm, reflecting the synergistic effect of butanol in simultaneously suppressing engine-out NO_x formation and promoting SCR catalyst efficiency through enhanced NO oxidation to NO₂ in the upstream DOC stage^7,23.

A crucial mechanism governing SCR performance is the NO₂/NO_x ratio at the SCR inlet, which determines the relative contributions of the standard SCR (4NO + 4NH₃ + O₂ → 4N₂ + 6H₂O) and the fast-SCR (NO + NO₂ + 2NH₃ → 2N₂ + 3H₂O) reaction pathways. The fast-SCR reaction proceeds at a substantially higher rate than the standard pathway because NO₂ is a stronger oxidant than O₂ for the reduction of NH₃, and the reaction is kinetically favorable at the operating temperatures of vanadium-based catalysts (250–450 °C). The upstream DOC partially oxidizes NO to NO₂ over its platinum-group metal surface, shifting the NO₂/NO_x ratio toward the stoichiometric optimum of 0.5 required for maximal fast-SCR activity^9,23. Although the emission analyzer employed in this study measures total NO_x without speciation, the superior conversion efficiencies recorded for oxygenated ternary blends suggest that their higher engine-out HC and CO concentrations provide additional reducing species within the DOC that promote the selective oxidation of NO to NO₂, thereby enhancing the NO₂/NO_x ratio entering the SCR^7,53.

Unlike the near-complete CO conversion observed in “Carbon monoxide (CO) emissions” Section, zero NO_x emission levels could not be achieved. This limitation reflects the inherent sensitivity of SCR performance to the precise stoichiometric balance between injected urea and instantaneous NO_x load, the minimum catalyst temperature required for efficient urea hydrolysis, and the residence time available for the heterogeneous reduction reactions within the catalyst bed. At low load conditions, where exhaust temperatures are closer to the SCR light-off threshold, these constraints become more binding and limit the achievable conversion efficiency^23,55.

Exhaust oxygen content and excess air ratio (O₂ and λ)

The effects of the test fuels and the DOC + SCR system on exhaust O₂ concentration are presented in Fig. 14a. An examination of the pre-DOC + SCR (P₁) results indicates a pronounced decrease in exhaust O₂ levels with increasing engine load, with concentrations declining from 13.68–14.12 vol% at 10.98 Nm to 3.46–4.16 vol% at 27.46 Nm^29,41. This reduction reflects the stoichiometric oxygen demand of the combustion process: as the injected fuel mass increases with load, a proportionally greater quantity of oxygen is consumed within the cylinder to sustain hydrocarbon oxidation, reducing the residual O₂ concentration in the exhaust. The progressive depletion of exhaust O₂ with load is thermodynamically consistent with the simultaneous increase in CO emissions (“Carbon monoxide (CO) emissions” Section) and the rise in NO_x at medium loads (“Nitrogen oxides (NO_x) emissions” Section), both of which reflect the tightening oxygen availability as the air–fuel mixture approaches stoichiometric conditions^29,46.

Fig. 14.

Combined effects of test fuels and engine load on (a) O₂ content (%-vol) and (b) excess air coefficient (λ) before DOC + SCR (P₁) and after DOC + SCR (P₃).

Fuel-type differences in exhaust O₂ became more pronounced at medium load conditions. At 21.97 Nm, D90BUT10 exhibited the highest P₁ O₂ concentration at 8.78 vol%, whereas B30D70 produced the lowest at 7.56 vol%. This divergence is mechanistically linked to the combustion efficiency differences established in “Engine performance (BSFC, BTE)” Section: D90BUT10, benefiting from its superior atomization and premixed combustion quality, converts a greater fraction of the injected fuel carbon to CO₂ and H₂O within the primary flame zone, consuming oxygen more completely per unit of carbon oxidized. In contrast, the higher viscosity of B30D70 impairs atomization and promotes diffusion-dominated combustion with locally fuel-rich zones, where a portion of the oxygen is consumed in partial oxidation reactions rather than complete carbon burnout, yet the overall oxygen utilization per unit fuel mass is lower due to the fuel-bound oxygen of biodiesel contributing to local oxidation without drawing from the air supply^13,53.

In contrast, the post-DOC + SCR (P₃) O₂ concentrations were consistently and substantially higher than the corresponding P₁ values across all fuel and load combinations, ranging from 16.23 to 17.08 vol% at 10.98 Nm and 12.65–15.16 vol% at 27.46 Nm. This counterintuitive increase does not reflect a generation of oxygen within the aftertreatment system, as the DOC reactions (CO + ½O₂ → CO₂ and HC + O₂ → CO₂ + H₂O) actually consume oxygen. Rather, the elevated P₃ O₂ concentrations arise from two compounding measurement effects: first, the emission analyzer operates on a dry-gas basis, removing water vapor before quantification; since DEF injection introduces a substantial water vapor load into the exhaust stream, its removal before measurement concentrates all remaining dry-gas species, including O₂, proportionally increasing their apparent vol% values; second, the SCR reactions convert NO_x molecules (which constituted a measurable fraction of the total exhaust gas at P₁) to N₂ and H₂O, reducing the total mole count of non-O₂ species in the dry gas matrix and thereby increasing the relative O₂ fraction^23,56.

An evaluation of Fig. 14b, which presents the excess air ratio (λ) results, shows that the pre-DOC + SCR (P₁) values decrease consistently with increasing engine load. At 10.98 Nm, λ values ranged between 3.04 and 3.25 for all fuels, indicating a highly lean air–fuel mixture, whereas at 27.46 Nm, these values decreased to 1.15–1.22, approaching stoichiometric conditions. This behavior reflects the fundamental load control mechanism of unthrottled diesel engines: since intake air quantity is not throttled, increasing the injected fuel mass at higher loads progressively reduces the global excess air ratio while the total air mass per cycle remains approximately constant^9,10. The measured λ values at maximum load (1.15–1.22) are consistent with the smoke-limited operating boundary of direct-injection diesel engines, below which incomplete combustion of the diffusion-controlled fuel fraction generates visible particulate emissions⁹.

Among the tested fuels, D90BUT10 produced the leanest mixture at 10.98 Nm with λ = 3.25, reflecting its superior combustion efficiency and more complete oxygen utilization per unit fuel mass, which leaves a greater residual O₂ fraction in the exhaust. In contrast, B30D70 yielded the lowest λ at 21.97 Nm with λ = 1.57, indicating a relatively richer effective mixture attributable to the higher fuel-to-air mass ratio required to compensate for the lower calorific value of the B30 blend and maintain equivalent brake torque output^13,46.

The post-DOC + SCR (P₃) λ values were consistently higher than the corresponding P₁ values under all conditions, with the highest value observed for B30D65BUT5 at 10.98 Nm, reaching λ = 6.03. It is important to note that the P₃ λ values do not represent the true combustion air–fuel ratio, as λ at the SCR outlet is influenced by the same DEF-induced dilution and dry-basis measurement effects described for O₂ above, rather than by the in-cylinder combustion stoichiometry. Nonetheless, the elevated P₃ λ values qualitatively confirm that the aftertreatment system maintains a strongly oxidizing atmosphere in the post-treatment exhaust stream, which is a prerequisite for efficient DOC operation and is consistent with the near-complete CO conversion reported in “Carbon monoxide (CO) emissions” Section^56,57.

BSFC-NO_x and temperature-NO_x relationships

Figure 15a presents the BSFC-NO_x trade-off diagram, with the upper region corresponding to pre-DOC + SCR (P₁) conditions and the lower region to post-DOC + SCR (P₃) conditions.

Fig. 15.

Combined effects of test fuels and DOC + SCR system on NOx emissions as a function of (a) brake specific fuel consumption (BSFC) and (b) exhaust gas temperature at T₂ (between DOC and SCR), showing P₁ before DOC + SCR (upper region) and P₃ after DOC + SCR (lower region).

An examination of the P₁ region reveals a general inverse relationship between BSFC and NO_x emissions: fuels with lower specific fuel consumption tend to exhibit higher NO_x concentrations. This trade-off reflects the fundamental thermodynamic coupling between combustion efficiency and thermal NO_x formation: improved combustion quality increases the mean in-cylinder gas temperature and the residence time of combustion gases above the Zeldovich threshold (~ 1800 K), thereby simultaneously reducing fuel consumption and accelerating thermal NO_x production. This behavior is well-established in the literature for oxygenated fuel blends and represents an inherent constraint of conventional diffusion-flame diesel combustion^13,29.

B30-based blends exhibited a characteristic clustering in the upper-right region of the P₁ diagram, combining high NO_x emissions (1724–2109 ppm) with relatively elevated BSFC values (282–315 g/kWh). This apparently paradoxical combination, in which both high NO_x and high fuel consumption occur simultaneously, is mechanistically explained by the dual penalty of biodiesel: the fuel-bound oxygen of SOME elevates local flame temperatures and promotes Zeldovich NO_x formation, while the lower net calorific value and higher viscosity of B30 blends simultaneously increase the fuel mass required to sustain the target brake torque, raising BSFC. Consequently, B30-based blends occupy a thermodynamically unfavorable position that combines the NO_x penalty of high-oxygen combustion with the efficiency penalty of impaired atomization^13,40.

In contrast, butanol-blended diesel fuels clustered in the lower-left region, yielding both reduced BSFC (241–296 g/kWh) and lower NO_x emissions (1322–1743 ppm), consistent with the findings reported in “Engine performance (BSFC, BTE) and “Nitrogen oxides (NO_x) emissions” Sections. This favorable positioning reflects the charge-cooling effect of butanol’s high latent heat of vaporization, which simultaneously suppresses peak flame temperatures (reducing NO_x) and enhances premixed combustion efficiency (reducing BSFC)^13,39.

Following DOC + SCR application (P₃ region), NO_x emissions for all fuels decreased substantially to the range of 569–1022 ppm, while BSFC remained unchanged, as expected given that the aftertreatment system operates exclusively on the exhaust stream without influencing the in-cylinder combustion process. The overall structure of the BSFC-NO_x trade-off was therefore preserved in the P₃ region, with B30-based blends retaining their relatively higher post-treatment NO_x levels compared to butanol-diesel blends. This confirms that the fuel-composition-dependent NO_x formation advantage of butanol blends is maintained through the full aftertreatment chain^13,53.

Figure 15b presents the T₂-NO_x relationship across both the P₁ (upper) and P₃ (lower) regions, using T₂ as the reference temperature since it reflects the thermal state of the exhaust stream entering the SCR catalyst after DOC oxidation. An analysis of the P₁ region shows that NO_x emissions generally increase with rising T₂, consistent with the Arrhenius-governed exponential temperature dependence of the Zeldovich mechanism. Within the T₂ range of 285–365 °C, NO_x values varied between 1322 and 2109 ppm, with B30-based blends clustering at higher T₂ values due to the more intense exothermic oxidation reactions within the DOC that elevate T₂ beyond T₁, as discussed in “Exhaust gas temperature (EGT)” Section^29,53.

However, at the highest load condition (27.46 Nm), where T₂ temperatures exceeded 400 °C, a relative decrease in engine-out NO_x was observed despite the elevated temperature. This apparent reversal of the T₂-NO_x trend is not attributable to aftertreatment effects, as these are P₁ measurements taken upstream of the DOC. Rather, it reflects the richer air–fuel mixture prevailing at maximum fuel injection rates, where the local oxygen partial pressure in the spray zone is insufficient to sustain the bimolecular Zeldovich reactions (N₂ + O → NO + N; O₂ + N → NO + O) at their maximum rate despite the elevated temperature, as both high temperature and adequate oxygen availability are required simultaneously for significant thermal NO_x formation^9,57.

In the P₃ region, following DOC + SCR application, NO_x emissions for all fuels decreased to the range of 569–1022 ppm. The persistence of the T₂-NO_x positive correlation in the P₃ region confirms the strong dependence of SCR conversion efficiency on exhaust gas temperature: higher T₂ values accelerate urea thermolysis and hydrolysis kinetics within the SCR bed, increase the surface reaction rates of the fast-SCR and standard SCR pathways on the vanadium-based catalyst, and shift the equilibrium of the NH₃ adsorption–desorption balance toward higher surface coverage, collectively enhancing NO_x reduction efficiency at elevated temperatures^23,56.

Exhaust gas temperature (EGT)

Exhaust gas temperature (EGT) measurements were conducted at three locations along the aftertreatment circuit: upstream of the DOC (T₁), between the DOC outlet and the SCR inlet (T₂), and downstream of the SCR (T₃). The results are presented in Fig. 16.

Fig. 16.

EGT values versus engine load for (a) T₁ (before DOC) and (b) T₂ (between DOC and SCR) and T₃ (after SCR) with test fuels.

An analysis of T₁ values reveals a consistent increase in exhaust temperature with increasing engine load for all fuels, with values ranging from 267–294 °C at 10.98 Nm to 413–435 °C at 27.46 Nm^9,29. This rise reflects the direct thermodynamic relationship between fuel mass flow rate and total heat release per cycle: at higher loads, the greater quantity of fuel burned per cycle increases both the peak combustion temperature and the sensible enthalpy of the exhaust gas leaving the cylinder, elevating T₁ in proportion to the total energy input⁹.

In terms of fuel type, butanol-blended diesel fuels produced T₁ values broadly comparable to D100 under low and medium load conditions, consistent with their similar total heat release profiles. At higher loads, D85BUT15 produced notably elevated T₁ values, reaching 358 °C at 21.97 Nm and 426 °C at 27.46 Nm. This behavior is mechanistically linked to its extended combustion duration and retarded CA50, as established in “Ignition delay (ID) and combustion duration (CD)” and “Combustion phasing (CA10, CA50)” Section: the prolonged late-phase heat release of D85BUT15 sustains elevated gas temperatures through a greater portion of the expansion stroke, increasing the thermal energy content of the exhaust gas at the cylinder exit⁴⁸. B30-based blends generally exhibited slightly lower T₁ values than D100 at low loads, a consequence of their diffusion-dominated combustion profile which distributes heat release over a wider crank angle interval and reduces the peak exhaust enthalpy flux at minimal load conditions⁴⁰.

T₂ values measured downstream of the DOC were consistently higher than the corresponding T₁ values across all fuel and load combinations, with the temperature rise being more pronounced for biodiesel-containing blends. This phenomenon is governed by the exothermic heterogeneous oxidation reactions occurring on the platinum-group metal surface of the DOC: CO and unburned HC present in the exhaust stream are catalytically oxidized to CO₂ and H₂O, releasing thermal energy that elevates the bulk gas temperature between T₁ and T₂^7,53. The intensified T₁ → T₂ temperature rise observed for B30-based blends is mechanistically attributable to the higher engine-out HC and CO concentrations documented in “Carbon monoxide (CO) emissions” and “Hydrocarbon (HC) emissions”Sections: the greater quantity of incompletely oxidized species entering the DOC provides a larger exothermic heat source within the catalyst substrate, proportionally elevating T₂ relative to T₁. The higher kinematic viscosity of SOME impairs in-cylinder fuel atomization, producing larger droplet sizes and a greater mass of unburned HC escaping the primary combustion zone, thereby sustaining the elevated DOC exotherm characteristic of biodiesel-containing blends^53,54. Importantly, this active heat generation within the DOC ensures that T₂ remains well above the minimum effective operating temperature of 285 °C required for efficient urea thermolysis and NO_x reduction in the downstream SCR system under all tested conditions^23,53.

T₃ values measured at the SCR outlet ranged between 141–195 °C, representing a substantial temperature drop of 230–260 °C relative to T₂ under high-load conditions. This marked decrease confirms the active thermochemical operation of the SCR system and is attributable to two compounding endothermic processes: first, the thermolysis of urea (CO(NH₂)₂ → NH₃ + HNCO, ΔH > 0) and the subsequent hydrolysis of isocyanic acid (HNCO + H₂O → NH₃ + CO₂, ΔH > 0) absorb thermal energy from the exhaust gas stream; second, the heterogeneous SCR reduction reactions on the vanadium-based catalyst surface consume additional enthalpy as the reactive N–H bonds in ammonia are utilized to reduce NO_x to N₂²³. The consistently low T₃ values relative to T₂ thus serve as indirect confirmation that the SCR system was thermally active and chemically reducing NO_x throughout all test conditions, further supported by the NO_x conversion efficiencies of 40–68% reported in “Nitrogen oxides (NO_x) emissions” Section.

Limitations and future work

Although this study provides a systematic and comprehensive evaluation of combustion characteristics, engine performance, gaseous emissions, and DOC + SCR aftertreatment conversion efficiencies across eight fuel formulations and four engine load conditions, several limitations are acknowledged that define the scope of the present work and motivate future research directions.

The most significant experimental constraint is the absence of direct particulate matter (PM) and smoke opacity measurements, as the current test bench was not equipped with a dedicated PM measurement system such as a smoke meter, filter smoke number analyzer, or particle size spectrometer. Since PM and NO_x emissions pose synergistic environmental and public health risks, being directly linked to cardiopulmonary diseases, respiratory tract infections, and adverse neurological effects, their simultaneous quantification alongside gaseous emissions is essential for a holistic assessment of ternary oxygenated blends in DOC + SCR-equipped engines^9,10. Although n-butanol is theoretically established as an effective soot suppressor through its charge-cooling, viscosity-reduction, and oxygen-enrichment effects, all of which were discussed in the context of HC emissions in “Hydrocarbon (HC) emissions” Section, direct experimental verification of PM reduction under the tested fuel-load matrix remains a primary objective for future research^23,40.

A second limitation concerns the speciation of NO_x. The emission analyzer employed in this study measured total NO_x (NO + NO₂) without providing individual species concentrations. The NO₂/NO_x ratio at the DOC outlet (T₂ position, SCR inlet) is a critical parameter governing fast-SCR reaction kinetics, as the optimal stoichiometric condition for maximum NO_x conversion requires NO₂/NO_x ≈ 0.5, as discussed in “Nitrogen oxides (NO_x) emissions” Section. Consequently, while the superior NO_x conversion efficiencies observed for oxygenated ternary blends were attributed to a favorable shift in this ratio, direct experimental verification was not possible within the current instrumentation setup. Future studies should incorporate NO/ NO₂ speciation measurement capability to quantify the fuel-composition-dependent influence on DOC oxidation selectivity and fast-SCR pathway utilization^23,57.

Third, while the exhaust mass flow rate was reliably derived from fuel consumption and lambda (λ) data for urea dosing control, the absence of direct, high-frequency intake air and exhaust flow measurements represents a technical refinement opportunity. Incorporating real-time mass flow sensors in future iterations could further enhance the transient response of the closed-loop dosing system and enable a more granular analysis of space velocity effects within the catalyst substrate. Such high-fidelity measurements would complement the current steady-state findings and support the development of even more precise aftertreatment control algorithms for variable-speed applications.

Fourth, the potential for long-term catalyst deactivation through surface coking and pore-plugging by high-molecular-weight hydrocarbon species from sunflower oil biodiesel was identified as a mechanistic concern in “Hydrocarbon (HC) emissions” Section. While the short-duration experiments conducted in this study did not reveal any measurable degradation in DOC or SCR conversion efficiency, the cumulative effect of sustained high-HC loading on vanadium-based catalyst activity under real-world operating conditions remains uncharacterized. Dedicated endurance testing and post-test catalyst characterization, including BET surface area analysis and X-ray diffraction, are therefore recommended to quantify the deactivation rate and establish operational service intervals for DOC + SCR systems operating on biodiesel-butanol-diesel ternary blends^23,55.

Finally, all experiments were conducted at a constant engine speed of 1800 rpm, which was selected as the rated speed of the Antor AD510 BS test engine and represents a typical operating condition for stationary agricultural and industrial diesel applications. While this approach enables controlled and reproducible load-sweep characterization, it does not capture the speed-dependent variations in exhaust temperature, flow rate, and emissions speciation that occur across the full engine operating map. Future research will therefore focus on extending the evaluation across a complete speed-load matrix under both steady-state and transient operating conditions to comprehensively characterize the synergistic impact of high-alcohol ternary blends and integrated DOC + SCR aftertreatment systems across the full operational envelope^7,56.

Summary of combustion, performance, and emission responses

Figure 17 provides a consolidated directional summary of all combustion, performance, and emission responses investigated in this study. The results confirm that n-butanol addition to diesel fuel consistently improved brake thermal efficiency and reduced specific fuel consumption, with D90BUT10 emerging as the most balanced formulation across the full load range. Biodiesel addition (B30D70) produced moderate deteriorations in peak cylinder pressure, HRR, and SFC due to its lower heating value and higher viscosity, while partially compensating through reduced CO, HC, and NO_x engine-out emissions. The ternary blends introduced competing effects: the B30D60BUT10 blend exhibited a notable mid-load anomaly characterized by simultaneously lower peak pressure, HRR, and MPRR, representing an unfavorable synergistic interaction between biodiesel viscosity and butanol’s low cetane number under medium-load conditions. The DOC + SCR aftertreatment system demonstrated the most decisive emission benefit, virtually eliminating CO across all fuels and load conditions, achieving HC reductions of 10–35% (limited by the catalytic light-off temperatures of long-chain hydrocarbon species) and achieving NO_x reductions of 40–68% via SCR. CO₂ concentrations decreased after the DOC + SCR primarily due to dilution from urea injection and the consumption of NH₃ in NO_x reduction reactions rather than a reduction in total carbon oxidation. Finally, the summary delineates critical research gaps and future outlooks, particularly highlighting the potential for nanoparticle-doped fuel blends and the implementation of high-fidelity mass flow sensing to further refine aftertreatment control strategies.

Fig. 17.

Directional summary of combustion, performance, and emission responses to n-butanol blending, biodiesel addition, ternary blending, and DOC + SCR aftertreatment, including identified research gaps and future outlooks (Antor AD510 BS, 1800 rpm; ↑↑/↓↓ strong, ↑/↓ moderate, ≈/ → negligible, ↕ load-dependent, —not applicable).

Conclusions

This study systematically evaluated the effects of n-butanol–sunflower oil biodiesel–diesel ternary blends on combustion characteristics, engine performance, and exhaust emissions in a single-cylinder direct-injection CI engine equipped with a custom DOC + SCR aftertreatment system across four progressive load conditions at 1800 rpm. The principal findings are summarized as follows:

1. Butanol addition at moderate fractions (10 vol%) enhanced fuel–air mixture homogeneity and premixed combustion intensity, with D90BUT10 achieving the highest HRR peaks and the lowest BSFC (241.01 g/kWh) and highest brake thermal efficiency (33.97%) at 21.97 Nm, representing a 4.4 percentage point improvement over D100. These improvements are attributed to the charge-cooling effect and molecular oxygen content of n-butanol, which collectively advance combustion phasing toward TDC and reduce thermodynamic losses.

2. B30-based blends consistently exhibited lower peak pressures, higher BSFC, and longer combustion durations than D100, owing to the lower heating value and viscosity-induced atomization penalty of sunflower oil biodiesel. A critical non-linear interaction was identified for the B30D60BUT10 blend at medium loads (16.48–21.97 Nm), where the compounding effects of biodiesel viscosity and butanol’s low cetane number produced anomalously low peak pressure, HRR, and MPRR, confirming that blend ratio optimization is essential for mid-range load conditions.

3. Engine-out CO and HC emissions were lowest for oxygenated blends, with B30D55BUT15 and D85BUT15 achieving the minimum values at low-to-medium loads. B30-based blends produced the highest NO_x concentrations due to elevated local flame temperatures promoted by biodiesel’s fuel-bound oxygen, while butanol-diesel blends (D90BUT10, D85BUT15) reduced NO_x below D100 levels through evaporative charge cooling.

4. The DOC + SCR system achieved near-complete CO elimination across all oxygenated blends and reduced HC emissions by 10–35%, with the residual HC attributable to the resistance of long-chain FAME molecules to catalytic oxidation at the prevailing exhaust temperatures. NO_x conversion efficiency ranged from 40 to 68%, with peak performance at the highest load (27.46 Nm) where elevated T₂ temperatures enhanced urea hydrolysis kinetics. D85BUT15 yielded the lowest absolute post-SCR NO_x concentration (637.6 ppm at 27.46 Nm), reflecting the synergistic benefit of butanol in suppressing engine-out NO_x formation and facilitating NO-to-NO₂ oxidation within the DOC, thereby enhancing fast-SCR efficiency.

5. The T₂ temperature consistently exceeded T₁ due to exothermic DOC oxidation, remaining above the minimum SCR activation threshold of 285 °C under all test conditions. The T₂–T₃ differential of 230–260 °C at high loads confirmed sustained SCR thermochemical activity throughout the test matrix.

6. Among all configurations, D90BUT10 represents the most balanced formulation for combustion efficiency and fuel economy, while D85BUT15 constitutes the optimal choice when minimum post-treatment NO_x is the primary criterion. These findings demonstrate that n-butanol incorporation into diesel and biodiesel-diesel blends is a practically viable strategy for simultaneous improvement of combustion efficiency and reduction of engine-out CO and HC emissions, with the DOC + SCR system effectively compensating for the NO_x penalty of biodiesel-rich formulations.

Abbreviations

ASTM

American Society for Testing and Materials

B30

30 Vol% sunflower oil biodiesel + 70 vol% diesel

B30D55BUT15

30 Vol% biodiesel + 55 vol% diesel + 15 vol% n-butanol

B30D60BUT10

30 Vol% biodiesel + 60 vol% diesel + 10 vol% n-butanol

B30D65BUT5

30 Vol% biodiesel + 65 vol% diesel + 5 vol% n-butanol

B30D70

30 Vol% biodiesel + 70 vol% diesel

BSFC

Brake specific fuel consumption (g/kWh)

BTE

Brake thermal efficiency(%)

CA

Crank angle (°CA)

CA10

Crank angle at 10% mass fraction burned (°CA)

CA50

Crank angle at 50% mass fraction burned (°CA)

CA90

Crank angle at 90% mass fraction burned (°CA)

CD

Combustion duration (°CA)

CHR

Cumulative heat release (J)

CI

Compression ignition

CO

Carbon monoxide (vol%)

CO₂

Carbon dioxide (vol%)

D100

Neat diesel fuel

D85BUT15

85 Vol% diesel + 15 vol% n-butanol

D90BUT10

90 Vol% diesel + 10 vol% n-butanol

D95BUT5

95 Vol% diesel + 5 vol% n-butanol

DAQ

Data acquisition

DEF

Diesel exhaust fluid (aqueous urea solution, 32.5%)

DI

Direct injection

DOC

Diesel oxidation catalyst

EGT

Exhaust gas temperature (°C)

EOC

End of combustion (°CA)

HC

Unburned hydrocarbons (ppm)

HCCI

Homogeneous charge compression ignition

HRR

Heat release rate (J/°CA)

ID

Ignition delay (°CA)

MPRR

Maximum pressure rise rate (bar/°CA)

NaOH

Sodium hydroxide

NO_x

Nitrogen oxides (NO + NO₂) (ppm)

O₂

Oxygen (vol%)

P₁

Exhaust sampling point upstream of DOC

P₂

Exhaust sampling point between DOC and SCR

P₃

Exhaust sampling point downstream of SCR

RI

Ringing intensity (MW/m²)

rpm

Revolutions per minute (min⁻¹)

RSS

Root-sum-square

SCR

Selective catalytic reduction

SMD

Sauter mean diameter (μm)

SOME

Sunflower oil methyl ester

SOC

Start of combustion (°CA)

T₁

Exhaust temperature at DOC inlet (°C)

T₂

Exhaust temperature between DOC outlet and SCR inlet (°C)

T₃

Exhaust temperature at SCR outlet (°C)

TDC

Top dead center

Symbols

A_w

Instantaneous heat transfer area (m²)

β

Pressure rise duration (s)

C₀

Hohenberg heat transfer constant

C₁

Hohenberg piston speed constant (m/s)

g

Gravitational acceleration (m/s²)

h_c

Convective heat transfer coefficient (W/m²·K)

k

Specific heat ratio (c_p/c_v)

L

Dynamometer arm length (m)

LHV

Lower heating value (kJ/g)

m

Load cell reading (kg)

_DEF

DEF mass flow rate (g/s)

_fuel

Fuel mass flow rate (g/s)

_NOx

NO_x mass flow rate (g/s)

n

Engine speed (rpm)

P

Instantaneous in-cylinder pressure (bar)

P_b

Brake power output (kW)

P_max

Maximum in-cylinder pressure (Pa)

Q_gr

Gross heat released (J)

Q_heat

Heat transferred to cylinder walls (J)

R

Dependent measured variable (uncertainty analysis)

T

Brake torque (Nm)

T_max

Maximum in-cylinder gas temperature (K)

T_g

Bulk in-cylinder gas temperature (K)

T_w

Cylinder wall temperature (K)

U_i

Uncertainty of the i-th measured quantity

u_instrument

Instrument calibration uncertainty

u_{repeatability}

Repeatability uncertainty

u_total

Combined measurement uncertainty

Mean piston speed (m/s)

V

Instantaneous cylinder volume (m³)

Arithmetic mean of repeated measurements

x_i

I-th independent measured variable

θ

Crank angle (°CA)

λ

Excess air ratio (air–fuel equivalence ratio)

Author contributions

The author, Şükrü Ayhan Baydır, was responsible for all aspects of this research, including conceptualization, methodology, software development, investigation, data curation, visualization, and the writing, reviewing, and editing of the manuscript.

Funding

The author received no financial support for the research, authorship, and/or publication of this article.

Data availability

The datasets used and analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Ethical approval and consent to participate

This study does not contain any studies with human or animal subjects performed by the author. Consequently, this research does not involve human subjects, and no informed consent was required.

Consent for publication

This research does not contain any individual person’s data in any form, and no human participants were involved in the study.

Consent to participate

This research does not involve human subjects.