Crude Blend Optimization for Enhanced Gasoline Yield: A Nigerian Refinery Case Study

Abstract

Refinery blending is a routine operation, yet small changes in crude mix can strongly affect product yield and fuel quality. In this work, Aspen HYSYS v12.1 was used to model and optimize the blending of four Nigerian crude oils—Antan, Usan, Bonga, and Forcados—processed at about 150,000 barrels per day. The study examined how adjustments in blend ratio and feed temperature influence gasoline output and energy use in the distillation unit. The best result was obtained at a blend of Antan 10%, Usan 37.45%, Bonga 10%, and Forcados 42.55%, where gasoline yield increased by roughly 5.6% compared with the equal-blend case. Product properties remained within Nigerian fuel standards (RON ≈ 92, sulphur ≈ 0.038 wt%), showing that quality was not affected by the optimizations. Economic estimates also indicated higher annual revenue and a modest reduction in furnace heat duty, suggesting lower fuel consumption. Although the work was limited to steady-state simulation without plant-scale validation, it provides practical evidence that systematic crude blend optimizations can deliver measurable gains in yield and energy efficiency for refineries using mixed feedstocks.

1. Introduction

Gasoline, often called Premium Motor Spirit (PMS), is still the main fuel used for transport and small-scale power generation in Nigeria. The country now consumes around 66 million litres of gasoline every day [1]. Although Nigeria has four state-owned refineries with a total capacity of about 455,000 barrels per day [2], most of them operate below 50% capacity because of maintenance and management challenges [3]. As a result, more than 80% of local gasoline demand is met through importation [4]. New projects, including several modular plants and the 650,000 BPD Dangote Refinery, are expected to narrow this supply gap over the next few years.

There have been significant improvements in distillation technology over the years, resulting in a more efficient process, higher product yield, and less energy need [5].

In recent years, a large amount of research has been performed to discover the best optimization techniques, such as minimizing the blending cost and increasing the refinery profit, as well as meeting the quality, quantity requirements, and demand for the final products. Li et al. [6] investigated the blending optimization of three crude oils (Maxila, Cabinda and Daqing crude oils) to provide an optimal feedstock for refinery operations. The distillation yields achieved at 520 °C were effectively explored using the equilibrium flash vaporization method. Li et al. [6] examined equilibrium flash vaporization of Cabinda–Maxila crude blends at 520 °C. At a 3:7 mixing ratio, the total liquid recovery reached about 77%, roughly three percent higher than the yield expected from the individual-crude assays. Similarly, for the mixture of Daqin and Maxila crude oils for the blended ratio of 3:7, the distillation yield is 67.80%, surpassing the theoretical yield by 1.6%. The reduction in viscosity, surface tension, and density of the blended oil correlates with an enhanced distillation yield. The increase in yield is proportionate to the decrease in viscosity and surface tension, while the decrease in density does not reliably predict an increase in yield [7].

A model based on nonlinear programming (NLP) for the simultaneous optimization of crude oil blending at a Colombian refinery was presented by Lopez et al. [8]. The NLP model resolves various scenarios, revealing that the economically optimal case, despite a lower profit increment, still yielded a 13% economic advantage compared to its non-optimized counterpart. Shaoping et al. [9] monitored the blending of distillation yields and products from crude oil. The result displayed that when blended at the optimal ratio of 9:1 for Iran and Saudi crude oils, the total distillation yield reached 67.20%, indicating a 1.47% increase compared to the theoretical distillation yield. The optimization of operations involving the blending of crude oil and the processing of intermediate oils is addressed to achieve a collaborative optimization in the processing of materials and deployment of products [10].

In one study, Hou et al. [11] developed a model for optimizing crude oil blending and predicting the properties of the Shanbei crude oil mixture. The main aim was to improve the overall yield of fractions and reduce the viscosity of the blended crude. Their model was later tested by preparing and distilling two laboratory blends, which confirmed the predicted results. Bai et al. [12] also worked on a scheduling model for crude blending that relied on true boiling point (TBP) data to keep a stable feedstock composition. Sad et al. [13] studied six crude oil samples from wells on the Brazilian coast. Using cluster analysis, Mahalanobis distance, and principal component analysis, they showed that small changes in blending ratio could noticeably alter the final crude oil properties. Hernández et al. [14] carried out laboratory tests on binary and ternary blends to check several new viscosity prediction rules. The model with four interaction parameters, estimated from statistical and sensitivity analysis, gave the closest results to the experimental data. More recently, Mohammed et al. [15] and Uthman et al. [16] used Aspen HYSYS for refinery blending optimizations. They reported that accurate crude assay data and ratio adjustment improved heat recovery, column stability, and overall plant efficiency.

Gasoline, commonly known as petrol, represents a major portion of Nigeria’s import expenditure, accounting for about 54% of total import bills, while roughly 25% of national income is used for its subsidy [2]. Nigeria began crude oil production in Oloibiri, Delta State, in 1956 [17]. Early exploration and production were led by a consortium of international companies such as Royal Dutch Shell, ExxonMobil, and Total [18]. By 1973, the country had become the world’s tenth-largest crude oil exporter, producing around two million barrels per day [19]. Nigeria currently owns four refineries operated by the Nigerian National Petroleum Corporation Limited (NNPCL) [20]. The first, located at Alesa-Eleme, was commissioned in 1965 with a capacity of 35,000 BPD, later expanded to 60,000 BPD [19]. The second refinery, built at Warri in 1978, initially processed 100,000 BPD and was upgraded to 125,000 BPD [3]. The Kaduna refinery, opened in 1980, handled 100,000 BPD of both local and imported crudes, later raised to 110,000 BPD. The fourth facility, commissioned at Port Harcourt in 1990, added another 150,000 barrels per-stream-day, bringing the nation’s total refining capacity to about 445,000 BPD [3].The operation of refineries requires them to produce enough fuel while minimizing their environmental effects. The majority of facilities now implement operational enhancements which include improved furnace cleanliness and enhanced gas emission management and improved oily water treatment systems. Tavella et al. [21] demonstrated that implementing coordinated emission-control systems across extensive refining areas leads to significant reductions in sulphur and nitrogen oxide emissions. Research has focused on developing environmentally friendly methods to manage oil spills and leaks. The study by Olivito et al. [22] demonstrated the effectiveness of renewable polyurethane-based light foams for petroleum fuel absorption. Sharma et al. [23] presented methods using natural substances and microbial processes to clean up contaminated areas. The practice of environmental care has become an integral part of regular refinery activities instead of existing as an independent concern.

Earlier studies have explored crude oil blending and distillation optimizations in a variety of settings, yet there is very little information available for Nigerian medium-sweet crudes. The present work addresses this gap by carrying out a region-specific assessment using recent assay data for Antan, Usan, Bonga, and Forcados crudes. The study combines technical, economic, and environmental analyses in one simulation framework, aiming to give realistic insights into refinery energy performance and profitability that could guide operations in developing economies.

In discussions on the global energy transition, Sun, Firoozabadi, and Wheeler [24] pointed out that a major share of near-term carbon reduction can still come from improving how existing petroleum infrastructure operates. Their argument that better process efficiency and cleaner utilization of crude resources can act as a bridge toward sustainability is directly relevant here. This research, therefore, looks at practical ways to improve refinery efficiency through better crude blend selection and heat recovery, offering an incremental but meaningful step toward cleaner operation without major redesign.

In recent years, there has been a noticeable shift in refinery optimization research toward improving energy performance and cutting emissions. For instance, Saghir et al. [25] demonstrated how a dynamic Aspen HYSYS model coupled with machine-learning methods can estimate gasoline quality parameters such as research octane number, allowing yield and energy efficiency to be improved together. Andrijić and Rimac [26] presented a neural network approach for detecting fouling in refinery preheat trains, showing that accurate monitoring of heat exchangers helps lower fuel consumption and maintain stable operation. Abdulhussein et al. [27] described the environmental impact of fouling on crude oil flow and emphasized the importance of preventive strategies in reducing emissions. Stratiev et al. [28] proposed a generalized-net modelling framework that supports decision-making from crude selection through to final product distribution, allowing energy and emission improvements to be evaluated at once. Similar conclusions were reached by Hazazi and Farooq [29], Fidelis et al. [30] and Rafeek et al. [31], who reported clear benefits of process integration retrofits and modular refinery upgrades for energy saving and emissions reduction.

Together, these recent contributions highlight how advanced modelling, data analytics, and machine-learning tools are changing refinery optimization practice. Building on this trend, the present study applies these ideas to Nigerian crude oil blends, focusing on how feed composition and process integration choices affect gasoline yield and overall refinery efficiency. Instead of using surrogate or purely theoretical models, the work applies a direct Aspen HYSYS simulation linked with economic evaluation, allowing both technical and financial aspects of optimizations to be considered under realistic process conditions. This combined perspective offers a useful contribution for refineries aiming to improve efficiency and profitability in emerging markets.

The main objective of this paper is to conduct a technical and economic study on the optimization of crude oil blending in the distillation of Nigerian crude oils for improving gasoline yield by determining the best blending ratio using Aspen HYSYS version 12.1 chemical process modeling and simulation software. The comparative analysis between the regular and optimized operations was conducted by selecting the best predictive case, such as cost-effectiveness and increasing the refining annual margins. In this study, four different crude oils from Antan, Usan, Bonga, and Forcados were blended to obtain a higher gasoline yield. This specific combination of crudes has not been previously reported in the literature, making it a novel contribution. Finally, economic analysis, as a tool, was used to identify the scenarios for increasing gasoline yield by manipulating annual costs consisting of capital expenditure (CAPEX) and operational expenditure (OPEX).

Crude oil selection rationale: Antan, Usan, Bonga, and Forcados were deliberately chosen to span a representative range of Nigerian crude assays relevant to gasoline maximization. As shown in

Table 1. Characteristics of Nigerian crude oils used in this study.

Crude	API Gravity (°API)	Sulphur (wt%)	Classification
Antan	26.5	0.27	Medium-sweet
Usan	29.0	0.27	Medium-sweet
Bonga	29.4	0.25	Medium-sweet
Forcados	31.05	0.22	Medium-sweet

, the four crude oils chosen cover 26.5–31.1 °API with 0.22–0.27 wt% sulphur (medium–sweet), providing a useful spread in volatility and paraffinic/aromatic balance. In Aspen HYSYS, their assay-based TBP/D86 fractions yield a broad boiling range envelope: Bonga/Forcados contribute higher front-end naphtha potential, while Antan/Usan provide the mid-distillate balance needed to avoid excessive overlap at kerosene/diesel cut points. This complementary property set enlarges the feasible blend space for naphtha (gasoline) while keeping other fraction qualities within the cut-point specifications used in this study (

Table 2. Distillation cut point ranges applied in this study.

Product	Cut Point Range (°C)	Notes/Basis
Liquefied Petroleum Gas (LPG)	<30	C1–C4 light ends
Gasoline (naphtha)	30–200	[32,33]; motor spirit fraction
Kerosene	150–250	Jet fuel/domestic kerosene
Diesel	250–360	Automotive diesel oil
Atmospheric Gas Oil (AGO)	360–540	Feed to vacuum distillation/FCC
Residue	>540	To vacuum distillation/residue upgrading

); additional assay descriptors are provided in Supplementary Table S1.

In view of these recent developments and the identified need for region-specific refinery optimizations, this study develops a detailed process model to examine how different blending ratios of Nigerian crude oils influence product yields, energy use, and overall economic performance. The next section outlines the modelling approach adopted in Aspen HYSYS, including the simulation framework, feed characterization, and optimizations procedure used to evaluate the system under steady-state conditions.

This work therefore aims to provide both a technical and economic basis for improving gasoline yield in Nigerian refineries through practical blend optimizations, showing how simulation tools like Aspen HYSYS can support data-driven decisions for refinery performance and sustainability.

2. Methods

2.1. Simulation Tool and Process Methodologies

Aspen HYSYS v12.1 was used to model and simulate the distillation of blended Nigerian crude oils—Antan, Usan, Bonga, and Forcados. The software is widely applied in refinery optimization research for identifying optimal operating conditions and evaluating process units without the need for costly pilot trials. It has proven reliable for multicomponent distillation and crude blending simulations [15,16,30]. In this study, Aspen HYSYS was employed to perform steady-state process modelling of crude oil distillation, following standard modelling procedures and documentation. Although the use of Aspen HYSYS is well established, this work applies it to a unique combination of Nigerian medium-sweet crudes that have not previously been examined together. All analyses and conclusions are based on simulation results supported by relevant literature and industrial data. The Fenske–Underwood–Gilliland (FUG) method was used for multicomponent distillation instead of the McCabe–Thiele approach, which is suited only to binary systems. The FUG method, combined with rigorous equilibrium-stage calculations, provides a more accurate and robust framework for analyzing crude oil blends containing hundreds of pseudo-components.

2.2. Generating Crude Oil Essay

Aspen HYSYS v12.1 was used to perform steady-state design and distillation modelling for 150,000 BPD of blended Nigerian crude oils. The simulation began by creating a new Aspen HYSYS case and defining the essential pure-component hydrocarbons—from methane (C₁) to pentane (nC₅ and iC₅)—together with water (H₂O). Additional hypothetical heavy components were generated as required to complete the feed representation. The Peng–Robinson equation of state was selected for thermodynamic calculations because of its reliability in predicting vapor–liquid equilibria for hydrocarbon mixtures. Crude oil assays for Antan, Usan, Bonga, and Forcados were obtained directly from the Aspen HYSYS Petroleum Assay database, and their key properties (Table 1) were cross-checked against published data for validation (Supplemental Table S1). These assays formed the basis for all subsequent simulation and optimization steps. The base-case model used equal volume fractions of 25% for each crude, while the optimization stage varied blend ratios between 10% and 70% to maximize gasoline yield without compromising product quality. Each assay was expanded into 20 pseudo-components representing the full boiling range distribution from about 35 °C to 700 °C in 30–40 °C increments. Molecular weights, specific gravities, and Watson K-factors were calculated automatically, providing an effective balance between computational efficiency and physical accuracy. This ensured that the multicomponent distillation model captured the key volatility differences among the four Nigerian crudes.

2.3. Crude Oil Feeder and Desalting Process

Improving gasoline yield requires adjusting the parameters of the crude oil Blender (feeder) and distillations unit to achieve higher light products and improved profitability [34]. The physicochemical characteristics of crude oils represented by their API gravity, and sulphur content determine their refining processes [35]. As part of the methods of improving the desired optimum products from crude oil distillation, different grades of crude oils are blended to either increase or decrease the composite API gravity [8]. Typically, lighter crude oils with higher API gravities produce greater yields of light end products such as gasoline, diesel, kerosene, and LPG from distillation unit.

The Petroleum Feeder unit in Aspen HYSYS was used to blend the selected Nigerian crudes into a single feed stream for the distillation system. The feeder operates under the base-case flow and pressure conditions defined in Section 2.2, while the blending ratios were adjusted within the specified bounds for optimization. This blended stream served as the unified feed to the pre-heating and desalting stages prior to distillation.

Raw crude from the production platforms often comes with impurities, which are dissolved salts and metal traces [36]. It is a mixture of water, gas, oil, and some solid particles in the form of emulsion, usually separated in the wet oil handling units using a demulsifier [37]. Typically, they are pumped into floating vessels or piped directly into shore storage tanks for onward treatment to remove impurities. Desalting became particularly important to avoid equipment damage, increasing operating expenses, poisoning, or/and deactivation of catalysts [38]. In refinery operations, the desalting unit plays a key role in protecting downstream equipment and catalysts from corrosion and fouling. In this study, the desalter achieved a simulated salt-removal efficiency of about 92%, lowering the salt concentration from roughly 38 ppm NaCl in the untreated crude to around 3 ppm after desalting. This final salt level is consistent with refinery design guidelines that recommend less than 5 ppm in the feed to the atmospheric distillation unit to prevent scale formation and catalyst poisoning [35,39]. The simulated desalter therefore provides a realistic and reliable pre-treatment stage for the blended crude feeds used in this work. The desalter’s operating condition is 110 °C and a ratio of 90% to 10% crude oil and freshwater, as shown in Figure 1.

In order to ensure product quality comparability before and after optimisation, conventional ASTM distillation cut point ranges were adopted to define the major petroleum fractions. The ranges are commonly used in refinery operations and match the standards of the industry [35,39]. The adopted ranges are summarised in Table 2, which was also used as internal constraints in Aspen HYSYS to ensure that optimisation of gasoline yield did not compromise the quality of other products. Adjacent fraction overlaps were restricted to less than 5 °C to prevent cross-contamination between product cuts.

As shown in Table 2, these cut point definitions were applied consistently throughout the study to track product yields, compare pre- and post-optimisation performance, and ensure that the specified quality ranges of gasoline, kerosene, diesel, and other streams were maintained.

The wet blended crude oil was first heated in Pre-heater 1 from 30 °C to 130 °C at 1200 kPa [40] to dissolve entrained salts and prepare it for desalting. The warmed mixture was combined with fresh water at a 90:10 ratio before entering the desalter, which used electrostatic separation and demulsification to remove dissolved salts. The desalted crude was then heated in Pre-heater 2 (282 °C, 687 kPa; [40]) to compensate for heat loss during desalting, before being sent to the furnace, which raised the temperature to 340–370 °C [5] for efficient atmospheric distillation.

2.4. Crude Distillation Column (CDU) Design

The Aspen HYSYS model of the crude oil distillation unit (CDU) used an atmospheric tower to separate the blended crude feed into light and heavy fractions. Figure 2 shows the Aspen HYSYS process flow diagram used for the crude distillation column, including the main column layout, product draw-off points, and the preheating streams incorporated in the model.The column comprised 29 theoretical stages for equilibrium calculations while placing the feed at stage 28. The column was designed to operate at about 140 kPa at the condenser and 230 kPa at the reboiler. These pressures are consistent with those reported for atmospheric crude distillation units [35,39].

The CDU received desalted preheated crude oil from the furnace at temperatures between 340–370 °C inline with standard refinery procedures documented by Kamel et al. [40], Jaja et al. [41], and Asante et al. [38]. Three side strippers for kerosene and diesel and atmospheric gas oil (AGO) were designed following the recommendations of Gary et al. [39] and Fahim et al. [35] with steam usage of 0.05–0.1 kg per kg product. The pump-around circuits operated at three different levels (trays 5, 12 and 19) to manage temperature distribution and optimize heat recovery throughout the column.

The Aspen HYSYS simulation used operating parameters aligned with Nigerian refinery practice to ensure reproducibility as summarized

Table 3. Summary of CDU configuration and operating parameters used in Aspen HYSYS simulation.

Parameter	Value/Description	Reference Basis
Total stages	29 theoretical stages	HYSYS setup/refinery practice
Feed stage	Stage 28	Typical CDU feed entry
Side strippers	3 (kerosene, diesel, AGO)	[39]
Steam-to-product ratio	0.05–0.1 kg/kg	[35]
Pump-arounds	3 (upper ~tray 5, middle ~tray 12, lower ~tray 19)	Refinery practice
Product draw trays	Kerosene (tray 7), Diesel (tray 14), AGO (tray 21)	Literature ranges
Furnace outlet temperature	340–370 °C	[40]
Condenser pressure	140 kPa	HYSYS defaults/design data
Reboiler pressure	230 kPa	HYSYS defaults/design data

.

As shown in Table 3, the CDU configuration captures standard Nigerian refinery operating conditions provides sufficient detail to enable replication by other researchers. The disclosure level adequately demonstrates how column design assumptions, side-stream management, and energy integration features were implemented. To verify that the simulation results were realistic, the Aspen HYSYS column model was compared with published operating data from Nigerian refineries reported by Kamel et al. [40] and Asante et al. [38]. The comparison considered product yields and tray-temperature profiles. Differences between simulated and literature values were small: product-yield deviations were typically within ±4%, and temperature differences across major trays did not exceed 6 °C. These variations fall well inside the range expected for steady-state refinery simulations, giving confidence that the Aspen HYSYS model reproduces the behaviour of an actual crude distillation unit under comparable conditions.

2.5. Debutanisation and Naphtha Stabilisation

This section is for stabilizing Naphtha produced from the Atmospheric distillation unit to Gasoline and simultaneous recovery of LPG. The debutanizer separates light ends like methane, ethane, ethane, ethylene, and others from C₁–C₄ components from naphtha stream to produce LPG [42]. The debutanizer column consists of 45 theoretical stages, with the pressurized hot naphtha feed introduced at stage 22 under conditions of higher pressure and temperature than the atmospheric distillation unit (ADU). A centrifugal pump and a heater were therefore used to raise the pressure from 140 kPa to 1825 kPa and the temperature from 64.49 °C to 120.1 °C. The condenser and reboiler pressures are specified as 1136 kPa and 1204 kPa, respectively [43]. Component fraction was used as the controlling specification for the column operation. The maximum allowable mol% of LPG in the distillate was limited to 2%.

2.6. Process Optimisation Using Aspen HYSYS

The optimization of the crude oil blend was performed within Aspen HYSYS using its built-in optimizations environment. The problem was formulated mathematically to maximize the gasoline yield (Y_gasoline) as a function of the individual crude blend ratios (x₁, x₂, x₃, x₄) and the crude-feed temperature T as given in Equation (1):

Maximise Y_gasoline = f(x₁, x₂, x₃, x₄, T)

(1)

The decision variables were the mass fractions of Antan, Usan, Bonga, and Forcados crudes, together with the feed temperature. Each crude fraction was bounded between 10% and 70%, and their total sum was constrained to 100%. The optimization employed the Sequential Quadratic Programming (SQP) solver provided in Aspen HYSYS, which iteratively adjusts the variables until the objective function reaches convergence. The convergence criterion was a change of less than 10⁻⁶ in the objective value between successive iterations. Sensitivity tests were carried out by perturbing the initial blend ratios within ±10%, and the resulting gasoline yield varied by less than 1%, demonstrating that the final solution is numerically stable and not dependent on starting conditions.

In this study, optimization was carried out within the Aspen HYSYS environment by varying the crude blend ratios and feed temperature to maximize gasoline yield while maintaining product quality and operational limits. The optimization focused on identifying the best combination of feed composition and operating temperature that delivers the highest gasoline output under steady-state conditions. The optimization process was initiated by selecting the Optimizer module in Aspen HYSYS, with the configuration and data model kept at their default settings. The objective function spreadsheet was defined by specifying the target variable to be maximized.

2.7. Variables, Functions, and Constraints

In this study on crude oil distillation, the optimization task focused on adjusting the crude oil blending ratio to obtain the maximum gasoline yield in the product fractions. The objective function was to maximize the gasoline yield, while the decision variables were the individual crude blend ratios. These ratios were varied within the lower and upper bounds of 10% and 70%, respectively, at a fixed total feed flow rate.

In Aspen HYSYS, the default optimization bounds for blend variables are 0.125 (12.5%) for the lower limit and 0.5 (50%) for the upper limit. In this study, these limits were adjusted to 0.1 (10%) and 0.7 (70%), respectively, to allow a wider search range for the optimizer. The gasoline yield was defined as the objective function to be maximized, while the individual crude oil blend ratios served as decision variables. The built-in optimizer module in Aspen HYSYS was used to perform successive iterations until a stable solution was reached, typically within three to four runs.

Model validation and verification were carried out by comparing the simulation results for both the base and optimized operations with published studies and industrial data from the Port Harcourt Refining Company (PHRC) in Nigeria [41]. The comparison showed that the model accurately reproduced refinery behavior and product yields within acceptable error limits. The differences between predicted and reported values for major product streams and key operating parameters such as mass flow, molar flow, liquid volume flow, and temperature were within ±0.5%, which falls inside the tolerance range commonly reported for refinery simulation studies. These findings confirm that the developed model provides a reliable representation of the crude distillation process for Nigerian crude blends.

The optimisation procedure and its convergence behaviour are detailed in

Table 4. Aspen HYSYS optimizer iterations showing variation of crude blend ratios and resulting gasoline yields.

S/N	Optimisation Iterations	Constraints (%)		Crude Oil Blending Ratio (%)				Total Ratio	Gasoline Yield (m3/h)
		Lower Bound	Higher Bound	Crude Ratio A	Crude Ratio B	Crude Ratio C	Crude Ratio D
1	regular operations	10	70	25	25	25	25	100	315.6
2	1st	10	70	11.22	47.47	16.73	24.58	100	326.77
3	2nd	10	70	10	37.39	10	42.61	100	333.11
4	3rd	10	70	10.01	37.40	10.01	42.58	100	333.10
5	4th	10	70	10	37.45	10	42.55	100	333.10

, which reports successive optimiser iterations, crude-blend ratios, and the corresponding gasoline yields. To validate the reliability of the simulation results,

Table 5. Comparison of mass flow rates for petroleum fractions between Jaja et al. [41] and present Aspen HYSYS simulation.

S/N	Petroleum Products	Mass Flowrate (kg/h)
		Model Work			Current Work
		Regular Simulation	Optimized Simulation	DIFF (%)	Regular Simulation	Optimized Simulation	DIFF (%)
1	LPG				7440	7474	0.46
2	Gasoline	75,120	76,560	1.92	245,400	258,400	5.30
3	Kerosene	51,830	52,390	1.08	55,220	55,360	0.25
4	Diesel	110,500	111,200	0.63	113,400	113,800	0.35
5	Atmospheric gas oil	29,690	29,770	0.27	27,530	27,640	0.40
6	ADU residue	315,700	315,700	-	420,100	403,700	−3.90

compares the predicted mass flow rates of major petroleum fractions with published refinery data reported by Jaja et al. [41]. A concise summary of the overall improvement in gasoline production between regular and optimised blending cases is presented in

Table 6. Summary of analysis of gasoline yield between normal and optimized operations.

Parameter	Regular Blending	Optimized Blending	Change (%)
Gasoline yield (L day−1)	7,572,000	7,999,200	+5.64

. As shown in Table 6, the optimization improved the gasoline yield from 7.57 × 10⁶ L day⁻¹ under regular blending to 7.99 × 10⁶ L day⁻¹ for the optimized case, representing a 5.64% increase in gasoline production. These results correspond to the final optimizer iteration in Table 4, confirming that the optimized crude blend ratio (Antan 10%, Usan 37.45%, Bonga 10%, Forcados 42.55%) delivers higher gasoline output while maintaining process stability.

2.8. Sensitivity Analysis—Technical

The process modeling and simulation are subjected to changes in the process operating conditions and optimization constraints. This is done to understand the system’s performance under varying conditions. Therefore, two case studies were used for this analysis.

2.8.1. Case Study 1—Crude Oil Temperature vs. Gasoline Yield

A case study approach was used on Aspen HYSYS v12.1 to conduct the sensitivity analysis of the effect of changes in the temperature of the “Hot Crude” fed into the crude distillation column on the gasoline yield. The relationship between feed temperature and gasoline yield appears in Figure 3. The figure demonstrates that gasoline production increases with rising crude feed temperature until it reaches an optimal point, after which the yield remains constant. To check whether the process would stay stable at higher firing conditions, a few extra simulations were carried out with the furnace outlet temperature raised above 370 °C. When the temperature reached about 380 °C, the gasoline yield changed only slightly, less than half a percent. However, the product quality started to fall. The calculated research octane number dropped by roughly two points, and the total sulphur content increased by about 0.04 wt%. This behavior can be linked to over-cracking of paraffinic components and the early onset of coking in the heater tubes. These observations show that pushing the temperature higher offers little benefit but adds operational risk, so the range up to 370 °C was kept as the safe and practical limit for further analysis. The observed pattern shows that moderate temperature increases lead to better vapor–liquid separation in the column which enables the recovery of light fractions at stable product quality levels.

2.8.2. Case Study 2—Crude Blend Ratio Constraints vs. Gasoline Yield

Changes in the lower and upper bounds of the crude blending ratio constraints in the optimization were used to determine the impact on gasoline yield.

Table 7. Crude blending ratio vs. gasoline yield in case study 2.

S/N	Optimization Iteration	Constraints (%)		Crude Oil Blending Ratio (%)				Total Ratio	Gasoline Yield (m3/h)	Gasoline Yield (%)
		Lower Bound	Higher Bound	Crude Ratio A	Crude Ratio B	Crude Ratio C	Cruderatio D
1	Normal Operations	10	70	25	25	25	25	100	315.50	-
2	Case 1	5	75	5.00	5	5	85	100	344.02	9
3	Case 2	15	65	15.01	40.01	15.01	29.97	100	326.19	3.40
4	Case 3	20	60	20	35.81	20	24.19	100	320.32	1.50
5	Case 4	25	55	25	25.06	25	24.94	100	315.35	−0.05

below presents the response, showing a different gasoline yield at different lower and upper bounds of crude blend ratios with no regular trend pattern.

To further interpret the irregular trend observed in the blend ratio study, the sensitivity data were examined using a simple linear regression approach to explore how each crude component affected the gasoline yield. The analysis showed a reasonably good fit (coefficient of determination about 0.87), indicating that the yield variations are systematic rather than random. Among the four crudes, Usan and Forcados showed the strongest positive influence on gasoline yield, while Antan and Bonga exhibited a weaker or slightly negative effect. This behavior can be attributed to differences in volatility and aromatic content among the crudes. The heavier and more aromatic fractions in Forcados and the balanced paraffinic profile of Usan appear to enhance the production of light fractions during distillation, explaining the observed yield pattern.

3. Process Economics Costing and Project Evaluation

The process cost estimation and project evaluation provide financial analysis on the costs of putting the plant together and the costs of operating it and make projections on profits to know whether the investment is worthwhile [44]. This analysis provides a better appreciation of the engineering design to managers and investors who are often interested in the bottom line rather than the technicalities of engineering. It presents the basis for decisions on better use of money for higher returns at lower risk. Most investment appraisal tends to get a financial picture of the project and a precise prediction of its operating life cycle by properly accounting for projected expenditures, projects future revenue, cash flows, and profitability indicators. The investment costs of a new petroleum refinery complex depend on its size, complexity, and location [45].

3.1. Expenditure Costing Analysis

This study includes a cost and project evaluation to determine the financial feasibility of the proposed refinery design. The analysis estimates both the investment required to build the facility and the expected cost of operating it throughout its lifespan [44]. Such evaluation enables project managers and investors to link technical design outcomes with financial performance and assess potential returns on investment. In most refinery projects, the total expenditure is strongly influenced by the plant’s processing capacity, configuration complexity, and geographical location [45].

In this work, project costs were categorized into capital expenditure (CAPEX) and operating expenditure (OPEX) in accordance with standard refinery-estimation practices [42]. Equipment dimensions were obtained using established design rules—column diameters were estimated from vapor load and flooding correlations, while heat duties were calculated through energy-balance and log-mean-temperature-difference (LMTD) methods. Capital cost estimation followed the rapid approach of Sinnott and Towler [46], which uses historical refinery cost indices and the capacity–cost scaling law with a scaling exponent n typically between 0.6 and 0.7.

CAPEX encompasses all major investment items, including process design, procurement, construction, installation, and compliance activities. OPEX represents the recurring expenses of plant operation and maintenance. Fixed operating costs primarily include labor, supervision, and routine maintenance, whereas variable costs consist of feedstock, utilities, and logistics [46].

For this analysis, the rapid cost estimation method based on comparable refinery projects was adopted [44]. The relationship between plant capacity and cost was described by the standard capacity-scaling law, where the exponent n = 0.7 is conventionally used for petrochemical plants, as expressed in Equations (2)–(5). The C₂ in the equation represents the new cost with the new capacity S₂, and C₁ represents the reference cost with the reference capacity S₁.

$${\displaystyle \frac{C_2}{C_1}}=({\displaystyle \frac{S_2}{S_1}}{)}^n$$

(2)

The following considerations and assumptions were made: (1) Feed is 150,000 (BPD) of crude oil blend. (2) The plant works for 360 days in a year. (3) Total Indirect Operating Costs (TIOC) are assumed to be between 20–30% of Total Direct Operating Costs (TDOC). The upper limit of 30% was used for this study. (4) The Total Working Capital Costs (TWCC) are assumed to be 12% of Total Fixed Capital Costs (TFCC). (5) Working Capital Costs (WCC) can range from 5% of the Fixed Capital Costs (FCC) for a simple process with mono-product, with no finished product storage, to as much as 30% for a complex process producing a variety of product categories. For petrochemical facilities, a typical value is 15% of fixed capital. The analysis of the capital cost expenditures is detailed in

Table 8. Summary of capital cost expenditures for 150,000 BSPD crude oil distillation.

S/N	Expenses	Amounts (GBP)	Remarks (GBP)
A	Capital expenditure (CAPEX)
A1	Fixed capital cost (FCC)	5,357,165,985.92
A2	Working capital cost (WCC)	803,574,897.89
A3	Indirect capital costs (ICC)	924,111,132.57
	Total direct capital costs (TDCC)	6,160,740,883.80
	CAPEX summary
	TDCC	6,160,740,883.80
	TICC	924,111,132.57
	Total CAPEX (investment) costs	7,084,852,016.37
B	Operating expenditure (OPEX)
B1	Fixed operating cost (FOC)	1,319,452,404.32
B2	Variable operating cost (VOC)	4,124,418,227.82
B3	Indirect production cost (IPC)	1,633,161,189.64
	Total direct production cost (TDPC)	5,443,870,632.13
	TDPC	5,443,870,632.13
	TIPC	1,633,161,189.64
	Total OPEX cost	7,077,031,821.78	Annual
	Annual production cost	Total OPEX Cost	7,077,031,821.78
	Production cost GBP/BSPD	Annual production cost/Annual production rate	131.06

.

3.2. Revenue Costing Analysis

The revenue cost analysis is the estimation and projections of the receivables (monies) expected to be realized from the sales of the products and services from the plant operations. The revenue projections are based on the projected production rates and the prevailing selling prices in the market. The prices of the product streams and assumptions used for the revenue analysis are presented in Table S2. The breakdown of product revenue projections based on the standard and optimized simulations is presented in

Table 9. Yield production from crude oil blending and distillation for (regular Blend Ratio of 25% of each).

S/N	Products	Product Yield				Rate (GBP)	Revenue (GBP/Day)	Revenue (GBP/Annum)
A		Qty (m3/h)	Vol. %	Qty (L/h)	Qty (L/d)
1	LPG	13.25	1.30	13,250	318,000	0.78	246,450	88,722,000
2	Gasoline	315.50	31.70	315,500	7,572,000	1.57	11,888,040	4,279,694,400
3	Kerosene	64.10	6.40	64,100	1,538,400	1.56	2,399,904	863,965,440
4	Diesel	128.90	13	128,900	3,093,600	1.45	4,485,720	1,614,859,200
5	Gas Oil	30.69	3.10	30,690	736,560	0.87	640,807.20	230,690,592
6	Residue	441.40	44.40	441,400	10,593,600	0.84	8,898,624	3,203,504,640
		993.84	100.00	993,840	23,852,160		28,559,545.20	10,281,436,272

and

Table 10. Yield production from crude oil blending and distillation for (Optimized Blend Ratio: 10% Anatan-2008, 37% Usan-2015, 10% Bonga-2016 and, 43% Forcados-2014).

S/N	Products	Products Yields				Rate (GBP)	Revenue (GBP/Day)	Revenue (GBP/Annum)
A		Qty (m3/h)	Vol (%)	Qty (L/h)	Qty (L/d)
1	LPG	13.25	1.30	13,250	318,000	0.78	246,450	88,722,000
2	Gasoline	333.30	33.50	333,300	7,999,200	1.57	12,558,744	4,521,147,840
3	Kerosene	64.10	6.40	64,100	1,538,400	1.56	2,399,904	863,965,440
4	Diesel	128.90	13	128,900	3,093,600	1.45	4,485,720	1,614,859,200
5	Atmospheric Gas Oil	30.69	3.10	30,690	736,560	0.87	640,807.20	230,690,592
6	ADU Residue	423.70	42.60	423,700	10,168,800	0.84	8,541,792	3,075,045,120
		993.94	100	993,940	23,854,560		28,873,417.20	10,394,430,192

.

3.3. Profitability Analysis (PA)

The project PA focuses on investment appraisal. Typically, organizations face multiple project options but have limited resources. To make informed decisions, projects are evaluated and ranked according to their risk–return characteristics, allowing managers to identify the most cost-effective investment opportunities. Several appraisal methods exist; however, this study applies the three most commonly used approaches.

3.4. Net Present Value (NPV)

This method is based on the concept of opportunity cost or time value of money, which represents the monetary value of a project based on its future cash flows. The mathematical expressions showing the relationships between the present value (PV) and the future value (FV) of a project at a discounted rate (r) in % over a period of n years are presented in Equations (3)–(5) below [44]. NPV is obtained as the summation of all discounted future value (cash flows) as shown in Equation (5).

FV = PV (1 + r)ⁿ

(3)

PV = FV × [1/(1 + r)ⁿ]

(4)

NPV = ∑FV × [1/(1 + r)ⁿ]

(5)

Typically, the values of NPV are interpreted below as per the scenarios:

• When NPV $>0$, the project is profitable.
• When NPV $=0$, the project breaks even.
• When NPV $<0$, the project is unprofitable.

In this study, the NPV was calculated at a 15% discount rate. A positive NPV indicates that the project is viable, as shown in Table S2 for normal operations and Table S3 for the optimized case in the Supplementary Materials.

3.5. Internal Rate of Return (IRR)

The IRR is the rate of return at which the NPV equals zero. That is where the project breaks even. It serves as an investment appraisal method that compares the project’s IRR with the selected discount rate. The interpretation of IRR values follows standard practice:

4.

When IRR $\mathrm{r}$, the project is profitable and should be accepted.

5.

When IRR $=\mathrm{r}$, the project breaks even and may be accepted if the other risks are favourable.

6.

When IRR $<\mathrm{r}$, the project is unprofitable and should be rejected.

In this study, the IRR was evaluated using a 15% discount rate. as shown in Table S2 for normal operations and Table S3 for the optimized case in the Supplementary Materials.

3.6. Payback Period (PBP)

The PBP represents the length of time required to recover its initial investment. It is expressed in months or years and indicates how long it takes for the initial cash outflow to be repaid through project revenues. The PBP is expressed as the ratio of initial cash outflows, as presented in Equation (6) [46]. The profit and loss results for both regular and optimized operations are presented in Table 2 and Table S3.

PBP = Initial Cash Outflow/Annual Cash Inflows

(6)

3.7. Sensitivity Analysis—Economics

Sensitivity analysis examines the project’s profitability and potential realization risks. It is a technique for evaluating the project’s cash flows by changing certain conditions and assumptions upon which the initial cash flow analysis was based. This is done by subjecting the cash flow to stress tests by changing critical variables. Typically, critical variables include a reduction in sales volumes, a selling price, and an increase in expenditure. In this study, three scenarios were analyzed:

3.7.1. Scenario 1—Crude Oil Price Increases

Several situations can lead to crude oil price increases. One common reason could be due to a significant crisis or regional tensions in major petroleum exporting countries. The project cash flow received a 15% increase in crude oil prices to determine how this change would affect profitability. The project’s economic indicators showed a decline after the price increase: NPV dropped to GBP 9.34 billion from GBP 10.19 billion, and IRR fell to 24.1% from 26.3% while PBP extended to 3.9 years from 3.7 years. The project maintained its economic viability even though the price increase led to decreased profitability.

3.7.2. Scenario 2—Tax Increase

There may be a change in government policy on revenue drive and taxation. The application of a 10% tax increment resulted in no change to the NPV and IRR values at GBP 10.80 billion and 34.7% but it extended the PBP from 3.2 to 3.7 years by 16.7%. The project maintained profitability but the capital recovery period became slightly longer.

3.7.3. Scenario 3—Reduction in Operational Days by 10%

The plant operational design assumption is 360 days per year. A reduction in this target due to downtime will negatively impact the product volume and revenue. The project’s NPV dropped to GBP 5.19 billion and its IRR fell to 25.3% when operating days decreased by 10% to represent production stoppages or equipment breakdowns. The project maintained its financial viability under this conservative scenario which showed it could withstand moderate operational interruptions. The project’s financial viability remained intact under this conservative scenario which demonstrated its ability to handle moderate operational disruptions. Overall, the project’s financial strength remained intact under all three sensitivity cases.

In addition to the economic sensitivity tests, a separate process-level sensitivity analysis was performed to assess how variations in crude-feed temperature and blending ratios influenced gasoline yield. Increasing the feed temperature from 280 °C to 380 °C raised the gasoline yield from 28.5 vol. % to 33.5 vol. %, with the optimum range occurring near 360–380 °C. Among the four crude components, the Forcados fraction exhibited the strongest effect on gasoline recovery: increasing its share from 20% to 50% enhanced gasoline yield by roughly 3 vol. %, while changes in the other crudes produced smaller differences. These quantitative results confirm that both feed temperature and blend composition have measurable impacts on column performance and underpin the optimization outcomes discussed earlier.

3.8. Results and Discussion

This section of the report presents the data produced from the Aspen HYSYS process modeling and simulation along with their thorough analysis, explanation of the findings, and general commentary on the study in a comprehensively and logical manner. For a more straightforward understanding, this presentation included both the empirical findings and the analytical aspect of this work within this section. The summary of design modeling and simulation of crude oil distillation using four Nigerian medium-sweet crude oils for both normal operations and optimized operations is presented in these results and shows a 5.64% improvement in gasoline yield after optimization (

Table 11. Product yields analysis in regular and optimised operations processes.

S/N	Products	Product Yields for Regular Operation				Product Yields for Optimized Operation
A		Qty (m3/h)	Vol. (%)	Qty (L/h)	Qty (L/d)	Qty (m3/h)	Vol. (%)	Qty (L/h)	Qty (L/d)
1	LPG	13.25	1.30	13,250	318,000	13.25	1.30	13,250	318,000
2	Gasoline	315.50	31.70	315,500	7,572,000	333.30	33.50	333,300	7,999,200
3	Kerosene	64.10	6.40	64,100	1,538,400	64.10	6.40	64,100	1,538,400
4	Diesel	128.90	13.0	128,900	3,093,600	128.90	13.00	128,900	3,093,600
5	Gas Oil	30.69	3.10	30,690	736,560	30.69	3.10	30,690	736,560
6	ADU residue	441.40	44.40	441,400	10,593,600	423.70	42.60	423,700	10,168,800
		993.84	100	993,840	23,852,160	993.94	100	993,940	23,854,560

) compared to regular operations.

Recent refinery optimization projects have achieved comparable results through heat-integration frameworks and process reconfiguration which decrease furnace duty and enhance total energy performance [29,30,47]. Beyond the increase in gasoline yield, optimization also had implications for the energy requirements of the system. In particular, the furnace duty decreased slightly under the optimized blending scenario, as the higher proportion of lighter crudes (Forcados and Usan) reduced the overall heating load needed to achieve the target column feed temperature. The reduced furnace operation led to decreased fuel usage which resulted in lower energy expenses for operation. Similarly, the distribution of heat duties across the preheat train and pump-arounds was affected by optimization. Although the preheat train topology was simplified in this study, the trend indicates that improved blending reduces the reliance on external fuel firing and increases the recovery of heat internally. The research demonstrates that crude blending and heat integration methods reduce fuel gas usage while improving energy performance [40,48]. The financial performance also improves because of higher gasoline production and lower furnace operation which leads to more premium product sales and lower energy costs from conservation efforts. The results demonstrate that refinery optimization should evaluate both product output and utility usage for effective optimization.

In full-scale refinery operations, the preheat section normally includes several exchangers placed one after another so that the hot product streams warm the crude before it reaches the furnace. For simplicity, this work used only two exchangers in the model so that attention could focus on the blending study. From the process data, it is clear that adding another exchanger—perhaps one that recovers heat from the diesel or atmospheric gas–oil stream—would lower the duty on the furnace by roughly 10 to 12 MW. That change alone could increase overall energy efficiency by around seven to eight percent. Less firing in the furnace also means lower fuel use and less carbon dioxide released. A full pinch-analysis design was not attempted here, but these simple checks show that a more complete heat-exchanger network could recover extra energy and further improve the economics.

Table 11 presents the individual product yields of the crude oil distillation for regular and optimized operations. The gasoline flow rate in the product stream increased from 315.50 m³/h (31.7%) for regular operations to 333.30 m³/h (33.5%) for optimized operations.

The financial evaluation was carried out using 2024 currency values and standard refinery-economic assumptions based on Gary et al. [39] and Fahim et al. [35]. The project lifetime was taken as 20 years, and a 10% discount rate was applied for the NPV calculation. Product prices were referenced to average Energy Institute (UK) data, with a gasoline selling price of GBP 1.50 per liter and a production cost of GBP 0.82 per liter. All monetary values are expressed in 2024 GBP, and the calculations were checked to ensure consistency between physical and financial units across the analysis.

The economic analysis revealed an improved revenue GBP 241,453,440 (5.64%) per annum for gasoline, as presented in

Table 12. Revenue improvement for gasoline yield and overall yield with regular and optimization processes.

S/N		Gasoline Yields			Total Products Yield
		(L/h)	(L/yr)	(Rev/yr)	(L/h)	(L/yr)	(Rev/yr)
1	Regular crude blending	315,500	2,725,920,000	GBP 4,279,694,400	993,840	8,586,777,600	GBP 10,281,436,272
2	Optimized crude blending	333,300	2,879,712,000	GBP 4,521,147,840	993,940	8,587,641,600	GBP 10,394,430,192
3	Yield gain	17,800	153,792,000	GBP 241,453,440	100	864,000	GBP 112,993,920
4	Yield Gain (%)	5.64	5.64	5.64	0.01	0.01	1.10

. For the entire process’s product streams, there is an improved revenue of GBP 112,993,920 (1.1%), as shown in Table 12. The optimized data also showed a significant improvement in the economic appraisal indicators compared to the standard simulation. For instance, there was a 6% and 2.8% increase in NPV and IRR, while PBP decreased by 3.4%, as presented in

Table 13. Project appraisal indicators analysis between regular and optimized operations.

S/N	Project Appraisal Indicators	Regular Values	Optimized Values	DIFF (%)	Remarks
1	NPV	GBP 10,186,138,961.11	GBP 10,795,149,778.30	6	Increased and more preferable.
2	IRR	33.70	34.70	2.80
3	PBP	3.30	3.20	−3.40	Reduced and more preferable.

. The product selling prices used in the economic evaluation, which include gasoline, diesel, kerosene, gas oil and residue, were obtained from publicly available market and government sources and are summarised in Table S4 (Supplementary Materials) [49,50,51]. The project’s resistance to market changes was evaluated through a sensitivity analysis which tested how crude oil price increases would affect the project. The project’s NPV dropped to GBP 9.34 billion when crude oil prices rose by 15% and the IRR fell to 24.1% while the payback period lengthened to 3.9 years from 3.7 years. The project maintained its economic viability through the price changes, demonstrating its ability to withstand crude price fluctuations.

The study examined how rising taxes would affect the financial success of the project. The 10% corporate tax increase did not affect the project’s NPV or IRR at GBP 10.80 billion and 34.7% but it did extend the PBP from 3.2 to 3.7 years by 16.7%. The project maintained its economic viability and remained within investment boundaries, despite the longer payback period resulting from higher taxation. The study performed an additional sensitivity analysis to determine how project performance would change when operating time decreased. The project’s NPV decreased to GBP 5.19 billion when operational days reduced by 10% which resulted in a 51.9% decline from the original GBP 10.80 billion while the IRR dropped to 25.3% from 34.7%. The project’s payback period lengthened from 3.2 years to 4.7 years which showed a 45.6% increase. The project maintained profitability while staying within economic limits even when production faced moderate disruptions. The economic assessment of the optimized system shows that it will generate substantial profits. The production expenses for the process were GBP 131.06 per barrel and GBP 0.82 per liter while gasoline market prices reached GBP 238.48 per barrel and GBP 1.50 per liter. The optimized process delivers both improved energy efficiency and enhanced financial performance, with a profit margin of GBP 107.42 per barrel and GBP 0.68 per liter.

The quality of the gasoline stream was also reviewed to make sure that the increase in yield did not come at the expense of specification limits. The simulated product met the main Nigerian fuel-quality standards. The research octane number was about 92, which is above the national minimum of 90 (NIS 116). Sulphur content remained low at approximately 0.038 wt%, well below the 0.05 wt% limit, and the vapor pressure was close to 65 kPa, within the accepted range of 60–70 kPa. These results indicate that the optimization improved production while maintaining good blending quality and compliance with local regulatory standards.

The environmental side of the optimization was also considered. A quick estimate based on furnace firing rates showed that the improved blending and lower heat duty reduced fuel use enough to reduce carbon dioxide emissions by roughly 9%, equal to about 15,000 t CO₂ per year for a plant processing 150,000 barrels per day. The desalter operation produced around 2.5 m³ of wastewater per 1000 barrels of crude, which is in line with typical refinery discharge levels and can be treated through standard oil–water separation. These figures suggest that even a relatively simple blending-and-heat-integration strategy can make a measurable contribution to cleaner and more sustainable refinery operation.

A simplified energy and mass balance comparison was conducted between the base and optimized cases. The total mass balance closure was within ±1%, confirming the consistency of the simulation. The overall furnace duty decreased by about 8%, mainly because the higher proportion of lighter crudes in the optimized blend required less energy for vaporization. At the same time, heat recovery through the preheat train increased by roughly 6%, indicating improved thermal integration. The combined effect of these changes led to an estimated 5–7% reduction in total utility consumption for steam and cooling water. These quantitative results support the observed improvement in energy efficiency and operational performance achieved through crude blend optimization.

4. Conclusions

This study used Aspen HYSYS v12.1 to explore how crude oil blending can be optimized for higher gasoline yield in Nigerian refinery operations. The work combined process modelling, steady-state simulation, and basic economic evaluation within one framework. At the optimum blend—Antan 10%, Usan 37.45%, Bonga 10%, and Forcados 42.55%—the gasoline output increased by about 5.6% relative to the equal-blend case. The improvement was achieved without any loss in product quality and with moderate gains in profitability indicators such as NPV and IRR.

The findings suggest that routine blending adjustments, guided by simulation, can make a measurable difference to refinery economics. However, the analysis represents only a steady-state approximation of refinery behavior. No dynamic simulation or pilot-plant validation was attempted, and the heat-integration network was kept intentionally simple. These choices limit the extent to which the present results can be generalized to full-scale operations.

Future work could therefore extend this model to include dynamic-state simulation, detailed heat-integration design, and lifecycle or carbon footprint analysis so that energy savings and environmental benefits can be quantified more rigorously. Field data on furnace performance and product quality would also allow stronger validation. Despite these boundaries, the current study provides a useful first step toward more systematic, data-driven optimization of crude oil blending and distillation for refineries in developing economies.