Ecological optimization and LCA of TiO 2 -SiC/ water hybrid nanofluid in a shell and tube heat exchanger by ANN

Abstract

The current study estimates the environmental impact and cost model for the early-stage design of a shell and tube heat exchanger (STHE). Additionally, the lifecycle analysis (LCA) compares the effects of use of nanofluid in the chemical processing industry to increase the effectiveness of the heat transfer to the typical STHE system. Nanofluids increase heat transfer, as evidenced by the results of the experiments. The Nusselt number and pressure drop were both increased by roughly 65% and 8%, respectively, using 0.5 vol% of TiO₂-SiC/water hybrid nanofluid compared to water as coolant. The ecological impact of STHE was estimated as s attribute of Ecological (E_F) and total cost estimation function using artificial neural networks (ANN). Finally, the IdematLightLCA (2.8.6 version) software was used to calculate energy content, Carbon footprint, and cost over the lifespan of the baseline. The results show that nanofluid-based STHE has an average payback period considerably longer than typical coolant systems but saves superior economic value at the end of its lifetime. According to the findings, the suggested novel method is both cost-effective and environmentally feasible.

1. Introduction

One of the most difficult aspects of many industrial applications is the development of energy-efficient and greater heat transfer rate technologies. ¹ To ensure optimum performance and safe processes, cooling heat transfer equipment is essential. ² Due to its small size and high load to capacity ratio, the heating rate reaches 2500 W.cm⁻². Improve the efficiency of heat transfer equipment rather than adding more heating elements ³ and supporting materials such as oil chillers ⁴ and cooling ducts, ⁵ which will result in a major rise in cost, ⁶ land congestion, ⁷ and environmental damage from exhaust gas emissions. ⁸ To achieve that goal, a much-needed alternative is to improve the coolant qualities, ⁷ which will be both economically and environmentally feasible by conserving energy.^9,10 Recent efforts have focused on improving heat exchanger design, ¹¹ notably kinetic stabilization, ⁹ boosting the cooling rate^12,13 and applying antifouling materials, ¹⁴ modifying thermal performance structures ¹⁵ and materials, ¹⁶ and suitable conservation strategy. ¹⁷

Traditional coolants have limited heat transfer capabilities. ¹⁸ As a result, the demand for fluids with superior heat transport properties is growing. ¹⁹ Furthermore, the properties of heat transfer fluids play an important character in boosting the efficiency of such high heat exchangers. ²⁰ Nanofluids are colloidal suspensions of concentrated nanoparticles that form a two-phase system with the solid phase extending into the liquid. ²¹ Nanofluids have greater thermal conductivity, thermal diffusivity, ²² viscosity, and convective heat transfer coefficients ²³ than base fluids. It has proven to be useful in several heat transfer applications. ²⁴

1.1. Analysis

The life cycle assessment (LCA) is a useful method for analyzing several aspects of a system's environmental impact, from the initial resource allocation to disposal/reuse after consumer use. Based on the environmental effect, this is a limited LCA that focuses on the primary components of production and operation of the nanofluid coolant based STHE. ²⁵

During the creation of manufactured products, cost estimate is critical. Designers can use preliminary capital costs based on a combination of general criteria to make decisions about materials, manufacturing procedures, and, most crucially, the product's morphological traits. According to studies, the best opportunities for cost savings occur during the early design stages, when up to 80% of a product's cost is determined. ²⁶ Given that the project planning accounts for such a small portion of total development expenses, emphasizing design to expense is a practical and necessary step against lowering manufacturing costs. According to a review of the cost estimation literature, an increasing number of situations using artificial neural network (ANN) methodologies have been documented. ²⁷ These strategies are the most up-to-date methods for evaluating the cost of manufactured goods. ²⁸ The major goal of utilizing ANN to estimate pricing is to emulate the behavior of a human system expert in determining the critical characteristics that influence the ultimate cost of a manufactured item. ²⁹ So far, the most commonly used ANN approach in cost analysis has been scenario-based reasoning. ANN techniques allow for the modeling, storage, and reusing of data, as well as the capture of relative product knowledge. ³⁰

Various researchers combined the power output and entropy generation ³¹ to estimate the entransy, ecological and environmental impacts of heat exchangers and termed it the Ecological function.^32–34 Unlike traditional heat exchanger design techniques, which use geometry and operational conditions as input variables to estimate the equipment's heat capacity, the current studies have used this ecological function to optimize the overall cost of well-designed STHE. ³⁵

Our purpose is to compare the environmental and economic performance of the new hybrid nanofluid-based STHE with current conventional cooling technologies utilizing the LCA and artificial neural network (ANN). The current research adds to the STHE cooling literature by examining a plant-scale nanofluid-based heat exchanger whose heat transfer capability is tested and validated in our lab using a scaled-down prototype. We are aware of no other LCA/ANN investigation of TiO₂-SiC/water hybrid nanofluid in a STHE system.

2. Methodology

2.1. Preparation of hybrid nanofluid

Nanofluid preparation is essential for obtaining superior thermophysical characteristics. Even when using the very same nanoparticles and base fluid, differing preparation conditions employed in nanofluid synthesis can produce varied outcomes. ³⁶ Nanoparticle aggregation in the base fluid affects the nanofluid's stability, thermal conductivity, and viscosity, all of which reduce thermal performance. ³⁷ Ultrasonication is a more typical method for dispersing nanoparticles and breaking up agglomerates. The current study used the two-step procedure for the preparation of hybrid nanofluids (Figure S1). ³⁸ The ultrasonic cavitation technique is used to disseminate a known quantity of TiO₂ (Titania, Nanoshel, 95% purity) and SiC (Sigma Aldrich, 99% purity) nanoparticles in water. ³⁹ Table 1 shows the thermo-physical properties of the nanofluid.

Table 1.

Physical properties of the nanoparticles and nanofluid at 25 °C.

Properties	SiC	TiO2	Water	0.5 vol% SiC-TiO2hybrid nanofluid
Thermal conductivity (Wm−1k−1)	120	15	0.6	0.65
Density (kg/m3)	3.2	5.6	998	1168.1
Particle size (nm)	<  ±  30	<  ±  20	-	<  ±  30
Specific heat (J/kg.K)	737	692	4180	4738.48

2.2. Characterization

The size and form of hybrid nanoparticles were determined using a FESEM-EDS (JEOL-JSM-7610F). The zeta potential, size, and stability of the generated nanofluid were investigated using dynamic light scattering (DLS, ZEN 3600, Malvern, UK). A thermal property analyzer was used to ascertain the thermal conductivity of the hybrid nanofluid (KD2-Pro, DECAGON devices, USA). An ultrasonic processor (CROMA UP-1200) was used to disseminate the nanoparticles into the base fluid at various intensities.

2.3. Experimental setup

1 pass shell and tube heat exchanger (STHE; Figure 1) was designed to test the performance of convective heat transfer and was built to the specifications listed in Table 2. The hot and cold fluids enter the heat exchanger at one end and exit at the other. In a closed-loop system, the STHE is provided with data logging function for a pressure display.

Figure 1.

The designed STHE with data logging function.

Table 2.

Geometrical specifications of the shell-and-tube heat exchanger (STHE).

Parameter	Specification
Shell outer and inner diameter	0.13 m and 0.125 m
Length of the shell	0.75 m
Shell material	Stainless steel
Tube outer and inner diameter	0.0137 m and 0.0107 m
Tube length (L)	0.6 m
Tube material	Stainless steel
Tube pitch	Triangular, 30°
Type of tube layout	Circular layout
Type of tube arrangement	Inline arrangement
No of tubes inside the shell (N)	8
No. of passes	1
Baffle specification	4 baffles, 25% of horizontal cut segmental baffle with a 0.15 m baffle spacing.
Header size	50 mm
Material of construction	Stainless steel

The trials were carried out in STHE according to Table S1 , with the RPMs of the pumping system was varied concurrently to sustain the appropriate N_Re. The N_Re for the nanofluid was 4.61% lower than the water coolant system due to the enhanced viscosity of the TiO₂-SiC/water hybrid nanofluid.

2.4. Data processing

The heat transfer parameters were defined as follows,

$$Q_{hf}={\dot{m}}_{hf}{C}_{P_{hf}}(T_{in}-T_{out}{)}_{hf}$$

(1)

$$Q_{nf}={\dot{m}}_{nf}{C}_{P_{nf}}(T_{in}-T_{out}{)}_{nf}$$

(2)

$$Q_{avg}={\displaystyle \frac{Q_{hf}+Q_{nf}}{2}}$$

(3)

$$Overallheattransfercoefficiebt(U_0)={\displaystyle \frac{Q_{avg}}{A_0\mathrm{\Delta }{T}_{LMTD}}}$$

(4)

$$Convectiveheattransfercoefficient(h_h)={\displaystyle \frac{k_h{j}_H{N}_{Re}P{r}^{1/3}}{D_e}}$$

(5)

$$Pressuredrop(\mathrm{\Delta }P)={\displaystyle \frac{f_{Exp}L\rho v_{avg}^2}{2D_e}}$$

(6)

$$Frictionfactor(f)={\displaystyle \frac{0.3164}{N_{Re}^{0.25}};}$$

(7)

2.5. Entracy and ecological function

According to Angulo-Brown ³² the following are the definitions of entransy (ΔG) in kW.K and ecological function (E_F) in kW.

$$E_F=Q_{Avg}-{\displaystyle \frac{\mathrm{\Delta }{G}_T}{AMTD}}$$

(8)

Here, Q_Avg  =  Overall rate of heat transfer (kW) calculated using Eq. (3)

$$TotalEntracy\mathrm{\Delta }{G}_T=0.5U_0{A}_{Total}FLMTD[({(T_{in}-T_{out})}_{hf})-({(T_{in}-T_{out})}_{nf})]+{\left[\dot{m}{\displaystyle \frac{\mathrm{\Delta }P}{\rho }{\displaystyle \frac{{(T_{in}-T_{out})}_{hf}}{{\left({\displaystyle \frac{T_{in}}{T_{out}}}\right)}_{hf}}}}\right]}_{shell}+{\left[\dot{m}{\displaystyle \frac{\mathrm{\Delta }P}{\rho }{\displaystyle \frac{{(T_{in}-T_{out})}_{nf}}{{\left({\displaystyle \frac{T_{in}}{T_{out}}}\right)}_{nf}}}}\right]}_{tube}$$

(9)

AMTD  =  Arithmatic mean temperature difference (K)

A_T  =  Total heat transfer area (m²)  =  πD₀LN  =  0.206 m²

F  =  Correction factor (considered as 0.9)

LMTD  =  Logmean temperature difference (K)

Subscript: in, out, hf, and nf, are representing inlet, outlet, hot fluid and nanofluid.

$\dot{m}$   =  Mass flow rate (kg/sec)

ΔP  =  Pressure drop (Pa)

ρ  =  Density (kg/m³)

2.6. Overall cost estimation of shell tube heat exchanger

In general the overall cost (C_Total, $/year) of equipment is calculated using the following equation, ⁴⁰

$$C_{Total}=a{C}_{inv}+C_{op}$$

(10)

Where,

$$a=Annualdiscountrate/year={\displaystyle \frac{1}{Equipmentlife(n)}}$$

Let, n  =  10 years

$$Costoftotalinvestment(C_{inv},\mathrm{\$})=a_1+a_2{A}_T^{a_3}(11)$$

(11)

where, a₁, a₂ and a₃ are 8000, 259.2 and 0.91, respectively.

Here, a₁, a₂, and a₃ are the equipment factors based on the material of construction (stainless steel)). ²⁶

$$Totaloperatingcost(C_{OP},\mathrm{\$})=P.C_E.H$$

(12)

$$Here,Pumpingpower(P,kW)={\displaystyle \frac{1}{\eta }\left\{{\left({\displaystyle \frac{\dot{m\mathrm{\Delta }P}}{\rho }}\right)}_{shell}+{\left({\displaystyle \frac{\dot{m\mathrm{\Delta }P}}{\rho }}\right)}_{tube}\right\}}$$

(13)

Where, η is the overall pump efficiency (generally value is taken between 0.6 and 0.7)

C_E is representing the total electricity cost The average electricity price for 2021 is considered 0.111 $/kW h.

H is representing the operating time per year. In the current case the heat exchanger is considered to be operated for 330 days i.e., 7920 h.

3. Results and discussion

3.1. Characterization

DLS (dynamic light scattering) is an excellent method for determining the size of nanoparticles. The average size of nanoparticles in TiO₂-SiC hybrid nanoparticle was 26 nm, according to the size histograms (Figure 2(a)&(b)). The well-spread TiO₂ nanoparticles on the SiC nanofiber's surface are visible in a FESEM micrograph of hybrid nanofluid, which verifies the DLS analysis results. ⁴¹ The lower polydispersity index (PDI) value (lower than 0.2) of the DLS analysis assuring the mono-dispersity of the sample. ⁴²

Figure 2.

(a) The FESEM (b) DLS and (c) Zetapotential analysis of TiO₂ coated SiC/water hybrid nanofluid.

The zetapotential graph value of 0.5 vol% TiO₂-SiC/water hybrid nanofluid after 5 days of preparation was around 43 mV, as illustrated in Figure 2(c). The zetapotential is the electric potential that exists at the sliding plane between the particle contact and the liquid. ⁴³ The elapsed time adjustment was within the error of the zeta potential readings. ⁴⁴ The observed zeta potential value and its time-dependent variation clearly demonstrate that the sample is electrostatically stabilized and rather stable. The charge stabilization produces a strong enough potential barrier to prevent aggregation. ⁴⁵

3.2. Thermal conductivity

The KD2 Pro results revealed that adding TiO₂-SiC hybrid nanoparticles to the base fluids increased thermal conductivity. Figure 3 (a)&(b) depicts this impact. The effects of nanoparticle concentration on thermal conductivity are illustrated in this graph. It is crucial to pay attention to the particle concentration in the base fluids, which is indicated as a volume percentage of nanoparticles. ⁴⁶ The effective thermal conductivity of nanofluids improves as the temperature of the nanofluid rises. Brownian motion reduces the distance between nanoparticles as the concentration of nanoparticles grows, which is the best explanation for the effect. ⁴⁷ Furthermore, when the concentration and nanofluid temperature rise, particle to particle contact rises, resulting in increased thermal conductivity. ⁴⁸ Furthermore, numerous types of heat exchange parameters are also responsible. ⁴⁴

Figure 3.

The effect of (a) & (b) temperature on the nanofluid thermal conductivity and (c) Reynolds number on the shell side and tube side outlet temperatures.

3.3. Temperature profile of the STHE

The temperature profiles of the hot and cold fluids at varied flow rates are shown in Figure 3(c). The trend revealed that as the flow rate of the cold fluid increased, the temperature of the hot fluid at the shell outflow dropped as well, but the temperature of the nanofluid increased. Even at low Reynolds numbers, hybrid nanofluid showed a faster rate of heat transfer than water (coolant) due to its improved thermophysical properties. ⁴⁹ The enhanced heat transfer rate in the turbulent flow pattern was ensured by the increased Brownian motion of the nanoparticles. ⁵⁰

3.4. Convective heat transfer coefficient

The nanoparticles are more thermal conducting than fluid and so play an important part in improving the convective heat transfer of any fluid. In this study, the resulting convective heat transfer coefficient of hybrid nanofluid was higher than its individual nanofluid with decreased particle size and greater zetapotential and efficient thermal conductivity. ⁵¹ The convective heat transfer coefficient increased with increasing fluid velocity (N_Re), as demonstrated in Figure 4 (a). The rise in the convective coefficient of heat transfer is attributable to the thermal conductivity and thermal boundary thickness of the fluid. ⁵² When the thermal conductivity of the fluid increases due to the movement of the nanoparticles across the wall, viscosity decreases and the thermal boundary layer thickness decreases. ⁵³ The dispersion mechanism ensured appropriate dispersion (mono dispersion with decreased PDI) of TiO₂-SiC hybrid nanoparticles in the base fluid, resulting in a very stable hybrid nanofluid. Because of its extended stability period (no agglomeration), the current nanofluid has an advantage in terms of achieving a better flow region while avoiding fouling the heat exchanger. The utilization of TiO₂-SiC/water hybrid nanofluids in heat exchange devices is extremely attractive due to these properties, as well as their improved convective heat transfer.^7,19,54

Figure 4.

The effect of Reynolds number on the (a) convective heat transfer coefficient of nanofluids and (b) nusselt number.

3.5. Nusselt number

In the proposed shell and tube heat exchanger, Figure 4 (b) compares the Nusselt number of various mono and hybrid nanofluids with the base fluid. As can be shown, heat transfer for nanofluids is significantly better than for the base fluid. For 0.5 vol% TiO₂-SiC/water hybrid nanofluid, the enhancement was around 65%, compared to 25% and 43% for 0.5 vol% TiO₂/water and SiC/water, respectively. In comparison to base fluid, hybrid nanofluid has a greater Nusselt number at the same Reynolds number. ⁵⁵ Because the particles in turbulent flow are in the tumultuous motion, the flow pattern attempted to limit film development on the effective surface-area, lowering the impedance to heat transfer. With a long residence period in laminar flow, the possibilities of forming a broad insulating layer on the plane to improve H.T. resistance were greatest The experimentation correctly justified these concepts. ⁵⁶

3.6. Life cycle analysis and optimization

Electricity from the grid was used to test the initial prototype. The prototype includes the tube, shell, pipe, storage vessel, and a pumping system. The goal of the LCA was to calculate the total amount of energy, Environmental impacts, and expenditure of the reference point during a 25-year period. ⁵⁷ The LCA for key components: material, production, transportation, utilization, and end-of-life potential were all calculated using the IdematLightLCA (2.8.6 verison) application. Overall cost of materials was used to establish the type of equipment and mass of each part. ⁵⁸ Each component was chosen to ensure proper manufacture. Each part was chosen for the correct production procedure. The data collection for the LCA research was effectively performed according to the requirements of the method and system limits. ²⁵ With regard to 1 kg of fundamental fluid, the results indicate that nanofluid generation (nanoparticles processing, mixing, transporting and the disposal of hazardous waste) and its pumping during operation are the most important contributors to global warming. ⁵⁹

When the system is in use, the basic prototype requires 3 kW of power to run. For the length of its existence, the system will be fueled by the grid and will run for 330 days a year, 8 h a day. A subset of the recyclable materials was selected to investigate the possibility of energy harvesting at the conclusion of the product's life cycle. Depending on the outcomes of LCA, Figure 5 demonstrates the relative influence of life cycle stages. The findings of the basic prototype with electrical consumption show that the usage stage consumes the most energy (89.5%), creates the most CO₂ (75.7%), and is the most expensive (66.3%). ²⁹

Figure 5.

The life cycle analysis of the nanofluid based STHE.

The thermal analysis may be used to calculate capital expenses, maintenance costs, and offset fuel savings costs based on current electricity and natural gas prices, as shown in Table S1 (supplementary file). Due to the current market pricing of nanoparticles at $5/g, the nanofluid-based STHE incurs an additional $150 in capital expenses and an additional $50 in maintenance costs. Because of the enhanced efficiency, the overall cost reduction per year for nanofluid-based STHE is more than that of conventional STHE. With the enhanced heat transfer coefficient, the heat exchanger area can be reduced by 7%, by which the overall cost can be reduced further. ⁵⁸

To justify present estimations, a collection of neural network settings for project cost was constructed and analyzed during the early design stage. This study investigated and assessed the comparison of individual configurations and training methodologies. In a series of research, neural networks were trained and applied to a set of databases. Other control parameters, as well as the costing accuracy, were explored.

3.7. Case study

As previously stated, the purpose of this simulation is for a potential industry to estimate the cost of a proposed shell and tube heat exchanger during the early design and development stage. ⁶⁰ The investigation presumes that capital expenditures are solely determined by product features that are fully or partially known at the time of cost assessment. ⁶¹ Following the creation of data by the simulation experiment, various neural network designs were applied to the data and evaluated to see how well they fit the underlying cost function. As a result, numerous sensibility tests were performed to confirm the parameter correlation. In this example, the cost of cooling hot water from 65°C to 46°C using 0.5 vol% TiO₂-SiC/water hybrid nanofluid at N_Re  =  18000 and 25 to 42°C was estimated. Table 3 displays the ecological function values and the total estimated cost for the given case study without optimization.

Table 3.

The theoretical results before optimization.

Shell outer diameter (DSo), m	Tube outer diameter (DTo), m	Length of the tube (LT), m	Number of tubes (NT)	Ecological function (EF), kW	Total cost (CTotal), $
0.13	0.0137	0.6	8	39.83	824.17

In this study, an artificial neural network with feed forward back propagation was used to investigate the ecological function and total cost estimation of the STHE as output variables. The key design parameters and their respective value ranges for all simulations are shown in Table S2 . The network training and testing data was generated at random. The chosen input values were then subjected to neural network analysis. ⁶²

Because this is a preliminary study, only a few key parameters, such as shell and tube outer diameters, as well as the number of tubes and their length, were considered to pass the cost function. To optimize the cost function, random input values were tested on the basis of the available data. ⁴¹ After employing cost-effective predicting models such as trainlm optimization function with two hidden layers and ten neurons each, the testing set correlation coefficient reached 0.9998, confirming the efficiency of nanofluids over conventional cooling systems (Figure 6). Table 4, Table S3, and Table S4 demonstrate a direct relationship between ecological function and total cost of STHE. ⁶³ Whereas, as the number of tubes increased, both output parameter values increased. Furthermore, no variation in the ecological function was observed at the same tube length with more number of tubes. ⁶⁴ Among the other input variables, the length of the tube has the greatest influence on the total cost estimation function. ⁶⁵

Figure 6.

ANN based testing and validation of the total cost estimation of STHE.

Table 4.

ANN's optimization results for the given case study.

	Factor 1	Factor 2	Factor 3	Factor 4	Response 1	Response 2
Std	A:Shell outer dia (m)	B:Tube Outer dia (m)	C:Tube length (m)	D:Number of tubes	Ecological function (kW)	Total cost ($/yr)
1	0.1	0.001	0.1	2	39.9890	818.0516
2	1.5	0.001	0.1	2	40.0043	818.0516
3	0.1	0.01	0.1	2	39.8416	818.2770
4	1.5	0.01	0.1	2	39.9945	818.2770
5	0.1	0.001	1	2	39.9890	818.2770
6	1.5	0.001	1	2	40.0043	818.2770
7	0.1	0.01	1	2	39.8416	820.1092
8	1.5	0.01	1	2	39.9945	820.1092
9	0.1	0.001	0.1	15	39.9890	818.2178
10	1.5	0.001	0.1	15	40.0043	818.2178
11	0.1	0.01	0.1	15	39.8416	819.6280
12	1.5	0.01	0.1	15	39.9945	819.6280
13	0.1	0.001	1	15	39.9890	819.6280
14	1.5	0.001	1	15	40.0043	819.6280
15	0.1	0.01	1	15	39.8416	831.0903
16	1.5	0.01	1	15	39.9945	831.0903
17	0.6	0.0055	0.55	9	39.9904	820.6459
18	2.2	0.0055	0.55	9	40.0013	820.6459
19	0.8	0.0035	0.55	9	39.9982	819.7604
20	0.8	0.0145	0.55	9	39.9757	824.3644
21	0.8	0.0055	0.35	9	39.9941	819.7604
22	0.8	0.0055	1.45	9	39.9941	824.3644
23	0.8	0.0055	0.55	5	39.9941	819.6402
24	0.8	0.0055	0.55	22	39.9941	824.1297
25	0.8	0.0055	0.55	9	39.9941	820.6459
26	0.8	0.0055	0.55	9	39.9941	820.6459
27	0.8	0.0055	0.55	9	39.9941	820.6459
28	0.8	0.0055	0.55	9	39.9941	820.6459
29	0.8	0.0055	0.55	9	39.9941	820.6459
30	0.8	0.0055	0.55	9	39.9941	820.6459

4. Conclusion

The current study uses powerful early design and development stages to conduct a detailed experimental and numerical investigation of a single pass shell and tube heat exchanger. Here, a STHE prototype was created, its heat transfer performance was evaluated, and the efficiency of traditional coolants was compared to that of TiO₂-SiC/water hybrid nanofluid. The nanofluid was created in two steps, and an ultrasonic cavitation technique was used to ensure stability while operating in the STHE. The current nanofluid has an advantage in terms of achieving a superior flow region while avoiding heat exchanger plugging due to its extended stability period (without agglomeration).

The following are the key findings of the research:

The uniform coating of TiO₂ nanoparticles on the SiC surface and mono dispersion in the base fluid were confirmed by FESEM and DLS.

The thermal conductivity of the TiO₂-SiC/water hybrid nanofluid was higher TiO₂/water, SiC/water, and water-cooling system at any nanofluid concentration and temperature.

Even at low N_Re, the heat transfer rate was increased due to improved thermophysical properties.

Because of the low density and viscosity variations, there was a small increase in pressure drop while working in STHE.

In terms of industrial applications, the TiO₂-SiC/water hybrid nanofluid in STHE has shown to be superior to traditional fluid in aspects of heat transfer rate.

The total cost of the nanofluid in costing is made up of 20% for materials, 4.5% for production, 3% for transportation, and 75.5% for use, according to LCA study.

ANN's costing function confirmed the same with an R² value of 0.9998 using the provided data set.

Among the other input variables, the length of the tube has the greatest impact on the total cost estimation function.

Supplemental Material

sj-docx-1-pie-10.1177_09544089221093304 - Supplemental material for Ecological optimization and LCA of TiO₂-SiC/ water hybrid nanofluid in a shell and tube heat exchanger by ANN

Supplemental material, sj-docx-1-pie-10.1177_09544089221093304 for Ecological optimization and LCA of TiO₂-SiC/ water hybrid nanofluid in a shell and tube heat exchanger by ANN by Manjakuppam Malika and Shriram S. Sonawane in Proceedings of the Institution of Mechanical Engineers, Part E: Journal of Process Mechanical Engineering