Coal is an exceptionally versatile energy resource and was a driver of the Industrial Revolution that transformed the economies of Europe and North America and changed the trajectory of civilization. The United States possessed the largest demonstrated coal reserves in the world with a value of 0.42 trillion MT in 2021 (EIA 2022). Nevertheless, the production of coal in the United States has been slowly declining because of lower natural gas prices and stagnant electricity demand. However, this decline, combined with the abundant coal reserves in the country, has created an opportunity to assess the potential of alternative uses of coal, particularly the use of coal as a feedstock for manufacturing high-value products such as graphite electrodes for industrial applications and energy storage devices, carbon fibers, activated carbon, etc. According to a 2021 National Coal Council report, all the coal-derived carbon products identified in Fig. 1 have the potential to create thousands of jobs in the United States throughout the coal-to-products supply chain, including employment in the coal-mining sector and the carbon product–manufacturing sector.

US market value and potential of carbon products derived from coal (Andrews 2021)

Carbon fibers are high-performance materials with physical and mechanical properties that can be tailored by changing processing conditions. For example, carbon fibers with high tensile strength and stiffness are used to reinforce polymers, metals, and ceramics to render composite materials that are widely used in aerospace, energy, military, automotive, and recreational sectors, among others. The primary advantage of carbon fibers is the material’s high strength-to-weight ratio, which enables the creation of lightweight structures. In the case of automobiles, every 10% decrease in weight may result in fuel savings of 6–7% (NIST 2014). Because of their light weights, carbon fiber–reinforced plastics may also enable the lengths of wind turbine blades to be increased, resulting in turbines with greater power-generating capacity.

Most of the economic studies published in the existing body of literature have been focused on carbon fiber manufactured from polyacrylonitrile (PAN) and studies focused on alternative precursors such as mesophase pitch, lignin, nylon, rayon, polyolefins, cellulosic biomass, and glycerol are rare (Mao et al. 2022). A brief literature review of cost analyses of carbon fiber manufacturing from PAN is depicted in Fig. 2. These economic analyses conducted between 1986 and 2023 reveal the manufacturing costs of PAN to vary widely between $13 and $120/kg of carbon fiber manufactured. It should be noted that all the costs presented in Fig. 2 have been adjusted to 2023 dollars.

Review of cost analyses to produce carbon fiber (CF) from polyacrylonitrile (PAN) adjusted to 2023 values

One of the earliest cost analyses was conducted at the Massachusetts Institute of Technology in 1986 by Goss in which the author developed a spreadsheet model for carbon fiber manufacturing revealing a cost breakdown not only by different process steps such as precursor, treatment, stabilization and carbonization but also by general cost categories including precursor, energy, material, labor, equipment, maintenance, insurance, and taxes, as shown in Fig. 2. In the early 2000s, similar methodology-based cost models were also developed by Das (2001) and Kline and Company (2004), followed by more recent analyses by Harper International (2009), Trützschler (2012) and Das and Nagapurkar (2021) . Some of these analyses, specifically those by Trützschler and Harper International, were reported by equipment manufacturers, while the rest were developed by researchers in close collaboration with industry.

The results of cost models shown in Fig. 2 reveal a couple of important trends. First, carbon fiber manufacturing costs have broadly decreased since 1986, as the cost reported by Nunna et al. in 2019 was nearly 1/10 of that reported by Goss in 1986. Furthermore, the cost of PAN precursor has increased broadly from 34% of total manufacturing costs (Goss 1986) to 53%-54% as reported in recent studies by Nunna et al. (2019) and Elringman et al. (2016). This indicates that the production costs of carbon fiber manufacture could be significantly reduced by replacing PAN with lower-cost precursors. For instance, using lignin in place of PAN have shown to reduce carbon fiber manufacturing cost by 50% (Mainka et al. 2015). The production cost of coal-based mesophase pitch is 40% of the average value of PAN (Giraud-Carrier and Barlow 2022; Wang 2019)  making it a promising low cost alternative to PAN. Despite the lower initial costs of precursors, few studies in the literature have examined the carbon fiber manufacturing costs derived from alternative precursors, particularly coal-based isotropic and mesophase pitch. The United States possesses the largest estimated proved recoverable reserves of coal in the world, offering a significant opportunity to replace PAN as a carbon fiber precursor.

A review of the economic models available in the open literature also revealed several inconsistencies. For example, analysis by Das and Warren et al. in 2012 showed the total share of manufacturing costs for combined steps of stabilization and carbonization to be 21%, whereas earlier studies by Goss (1986), Das (2001), and Kline and Company (2004) showed the combined share to be 39%–43%. Similar discrepancies were also found for other process steps, such as surface treatment, sizing, and winding and handling. These discrepancies could be attributed to three main factors: (1) absence of data on capital costs and technical specifications of equipment being considered, (2) differences in chosen process parameters, and (3) inadequate data on chemicals and material consumables. Also, most of the reviewed studies focused on cost assessment only of carbon fiber derived from PAN and not from alternative precursors such as lignin, coal, glycerol, etc. Furthermore, the existing literature also lacks information about the energy needed to manufacture carbon fiber from alternative precursors as well as direct comparisons between carbon fiber produced from  these alternatives and those derived from PAN. Additionally, none of the existing economic models have examined the impact of cost on the scale-up of manufacturing processes. These gaps have been addressed in the present study.

A 2022 study by Giraud-Carrier analyzed the carbon fiber manufacturing process from coal-based mesophase pitch via a novel Ecocoke process, estimating a production cost of $11.9/kg carbon fiber. Despite the authors’ detailed analysis of production costs, the study did not examine the embodied energy of carbon fiber production. Embodied energy can be defined as the cumulative energy required to extract raw materials from nature in addition to the energy utilized in manufacturing activities. Dér et al. (2021) performed a gate-to-gate energy assessment of carbon fiber production from PAN-based materials but did not extend it to carbon fiber produced from coal-based precursors. Coal-based materials can be an attractive precursor from an embodied energy standpoint because they possess 30–70% lower embodied energy than acrylonitrile based on a per kilogram raw precursor basis (Das and Nagapurkar 2021). By specifically analyzing coal-based materials as potential precursors for carbon fiber production, the the present study overcomes the limitations of previous studies.

The main features of the present analysis are as follows:

1. (1)

   Energy estimations were made at major steps of the solvent extraction process based on experimental mass balance data of the solvent extraction process to produce mesophase pitch.

2. (2)

   An economic analysis was performed for the production of mesophase pitch from coal and the production of carbon fiber from pitch.

3. (3)

   The impact of plant scale up on the cost of mesophase pitch manufacturing  was also analyzed.

4. (4)

   A cradle-to-gate energy assessment of carbon fiber produced from coal via a solvent extraction pathway was also conducted  and compared with previous studies.

The manuscript is structured as follows. Section 1 reports a literature survey of existing economic and energy analyses for carbon fiber production. Section 2 focuses on the economic and energy assessment methodology used to analyze carbon fiber production from coal via a solvent extraction pathway. Sections 3 and 4 include a discussion and conclusions, respectively, of this work.

2.1 Technoeconomic analysis framework of mesophase pitch production from coal via solvent extraction process

The technoeconomic analysis was conducted based on experimental investigations carried out at the University of Kentucky’s Center for Applied Energy Research in 2022. Based on the mass–energy balances, economic analysis of the solvent extraction process was performed as shown in Fig. 3 based on the three major steps of the mesophase pitch production process. The cost estimation was performed using the methodologies and framework outlined by the University of Michigan in collaboration with the National Energy Technology Laboratory (NETL), National Renewable Technology Laboratory, Argonne National Laboratory, Technische Universitat Berlin, and others (Faber et al. 2022). These methodologies were used because they were applicable to low (i.e., close to 3 or 4) technology readiness levels (TRLs). The solvent extraction process to produce mesophase pitch conforms to the TRL of 3 because it was successfully conducted on a laboratory scale.

Technoeconomic and cost framework for the mesophase pitch production process via the University of Kentucky’s solvent extraction process

Based on mass–energy balances, economic analysis was initially performed for a small-scale pilot plant with a production capacity of 43 MT/year of mesophase pitch to determine the manufacturing cost of 1 kg of mesophase pitch. A pilot plant capacity was chosen for cost estimation because the proposed solvent extraction technology has been validated only at the laboratory scale. A pilot plant capacity of 43 MT/year was determined based on the maximum capacity of equipment available from vendors.

The pilot production capacity was subsequently expanded to a large industrial-size plant capacity of 10,000 MT/year to determine the large-scale manufacturing cost of mesophase pitch. Cost analysis was performed for the categories of raw materials, capital equipment, operating energy, labor, and other fixed costs. A plant period life of 15 years was assumed in this analysis, with 320 operating days and a 7% interest rate for borrowed capital.

2.2 Experimental investigations at the University of Kentucky

The mass–energy balances of the solvent extraction process developed by the Center for Applied Energy Research at the University of Kentucky are depicted in Fig. 4. According to this process, Blue gem coal was converted to mesophase pitch via three major steps: (1) mild solvent extraction, (2) filtration, and (3) mesophication. Blue gem coal was chosen owing to its low ash yield and sulfur content (Cakmak et al. 2023). In the first step, the coal was dissolved into the solvent, which was decant oil. The mixture was introduced into the mild solvent extraction reactor, and coal liquids were obtained at conditions of 690 kPa and 400 °C with a reactor residence time of 0.5 h. A 1:3 mass ratio of coal to decant oil was also maintained in the reactor.

Mass–energy balances of the coal solvent extraction process to produce mesophase pitch (MP) by the University of Kentucky

The decant oil played an important role in liquefying coal. It is the heaviest–molecular weight fraction from a fluid catalytic cracking unit in a petroleum refinery. Around 400 °C, the decant oil solvent disrupts the hydrogen bonds within the coal’s macromolecular network undergoing thermal decomposition. The role of decant oil was to relax the coal matrix and extract coal molecules from the coal to the bulk solvent phase (Hernández et al. 2012). Compared with a direct coal-liquefaction process, the decomposition of coal’s macromolecular structure from solvent extraction is typically limited. The thermal decomposition from direct coal liquefaction was higher because of harsher pressure conditions of 15,000 kPa and because of the presence of catalysts and hydrogen gas that were not present in the solvent extraction process. These harsher conditions facilitates extensive scission of covalent bonds within coal’s macromolecular structure (Karri 2011).

2.3 Estimating the energy consumption of the mesophase pitch production process

This study employed fundamental principles-based empirical equations to approximate energy consumption as experimental-level energy consumption data were not available. The following sections outline the methodology used to estimate the energy consumption of the mild solvent reactor, filtration unit, and mesophication reactor. The energy consumption estimated in this work was assumed to be steady-state  energy consumption and excluded the energy needed during start-up.

2.3.1 Energy consumption of the mild solvent reactor unit

The thermal energy demand of the mild solvent reactor unit was estimated to heat the mixture of coal and decant oil to 400 °C. These calculations were performed based on the specific heat capacities of coal and decant oil. For the purpose of energy calculations, these components were assumed to be heated separately to 400 °C from 25 °C and then introduced into the reactor in a 1:3 mass ratio of coal to decant oil.

The specific heat of coal typically depends on its carbon, moisture, and volatile content. The following three equations were used to estimate the enthalpy needed to raise the temperature of coal from 25 to 400 °C as defined by Postrzednik (2013) based on experimental measurements (Lesniak et al. 2013). For a temperature \(T\) when \(25^\circ {\text{C}} < T < 100^\circ {\text{C}}\),

where \({C}_{{\text{p}}1}\) is the specific heat capacity in J/kg of coal at constant pressure in a dry, ash-free basis, and \({V}^{{\text{daf}}}\) is the volatile matter content (%) in coal as a dry, ash-free basis. This value was taken as 40% of Blue Gem coal’s volatile matter content as guided by the experimental investigation values provided by University of Kentucky. When \(100^\circ {\text{C}} < T < 300^\circ {\text{C}}\),

where \({C}_{{\text{p}}2}\) is the specific heat capacity in J/kg of coal at constant pressure in a dry-ash free basis, and \({c}_{0}\) and \({c}_{1}\) are coefficients defined as follows:

When \(300^\circ {\text{C}} < T < 1,100^\circ {\text{C}}\),

where \({C}_{{\text{p}}3}\) is the specific heat capacity in J/kg of coal at constant pressure in a dry, ash-free basis, and \({d}_{0}\), \({d}_{1}\), and \({d}_{2}\) are coefficients defined as follows:

Based on these formulas, the values of \({C}_{{\text{p}}1}\), \({C}_{{\text{p}}2}\), and \({C}_{{\text{p}}3}\) were calculated to be within 1.12–1.91 kJ/kg coal/K (after unit conversions) and were similar to values reported by Merrick (1983), as shown in Fig. 5. Based on these calculated values, the thermal energy needed to raise the temperature of 1 kg of coal from 25 to 400 °C was determined to be 672 kJ/kg of coal. Notably, the thermal energy demand may increase with increasing volatile matter content of coal. This is because coal’s specific heat capacity increases with volatile matter, as shown in Fig. 5 until about 600 °C after which it decreases. Coal with higher volatile matter content may contain higher concentrations of lighter hydrocarbons such as methane, ethane, and propane that possess higher specific heats relative to the solid carbon and ash content in the coal, thereby increasing thermal energy demand. It should be noted that the energy calculation of coal excludes the energy needed to induce volatilization of coal.

Specific heat capacities of coal as a variation of temperature and the coal’s volatile mineral (VM) content (Merrick 1983)

The specific heat capacity of decant oil in cal/g/°C, \({C}_{{\text{p}},\text{decant oil}}\), as a function of temperature and specific gravity was estimated using the following correlation (US Department of Commerce 1929):

where \(\rho\) is the specific gravity of decant oil at 15 °C, assumed to be 1.1 (American Petroleum Institute 2012). Using this correlation, the thermal energy needed to raise the temperature of 1 kg of decant oil from 25 °C to 400° was determined to be 860 kJ/kg decant oil after unit conversions. This energy is 22% more than the amount of energy needed to elevate identical amounts of coal (1 kg) from 25 to 400 °C. This is because decant oil’s specific heat capacity is higher than coal’s specific heat capacity. This indicates that the addition of decant oil would increase the thermal energy required more rapidly than the addition of coal into the mild solvent reaction mixture. Increasing the solvent-to-coal ratio from 3:1 to 5:1 or 10:1 as reported by Griffith et al. (2009) would increase the thermal energy demand of the mixture. The total thermal energy of coal and decant oil in a 1:3 ratio in the mild solvent extraction reactor was estimated to be 4,336 kJ/kg coal, which also incorporates the reactor’s thermal heating efficiency of 75% (Piccinno et al. 2016). The energy requirement of the mild solvent reaction mixture was estimated to be 1.35 kWh/kg of mesophase pitch produced.

The electrical demand for agitating the reaction mixture present in the mild solvent reactor was also estimated. An axial flow impeller was chosen as a hydrofoil-type agitator to ensure uniform mixing of coal and decant oil in the reactor. Agitation energy consumption typically depends on the mixing fluid’s power number, \({N}_{\text{p}}\); the impeller diameter, \(d\); rotational velocity, \(N\); density of the reaction mixture, \({\rho }_{{\text{mix}}}\); and reaction time, \(t\). The agitation energy was estimated by the following correlation (Piccinno et al. 2016):

where \({E}_{\text{stir}}\) is stirring energy in J, \({N}_{\text{p}}\) is 0.79 (assuming a turbulent flow) (Piccinno et al. 2016), \({\rho }_{{\text{mix}}}\) is 1,185 kg/m³, \(N\) is 1.42 rotations/s (85 rpm), \(d\) is 0.101 m (Jagani et al. 2010), and \(t\) is 1,800 s (30 min). The energy consumption of the agitator was determined to be 2.57 × 10⁻⁶ kWh/kg of mesophase pitch.

2.3.2 Energy consumption of the filtration unit

In the second step of the coal-to-mesophase-pitch production process, the quinoline insolubles (QIs) were removed from the obtained liquids via a filtration process. The QIs are solid materials typically consisting of char, ashes, cenospheres, and pyrolytic carbon and are highly aromatic in nature ( Suárez-Ruiz and Crelling 2008). QI is formed as a by-product during the mild solvent extraction process of coal and are undesirable products because they form a needle-shaped (uniaxially oriented) structure within the final pitch (An et al. 2021). Because high concentrations of QIs within the pitch affect the pitch’s flowability strength and other properties, QIs are removed via a filtration process to ensure high–tensile strength derived products. The mass yield of filtered coal liquids was 93.5 wt % as shown in Fig. 4 for the University of Kentucky’s solvent extraction process.

The solid extraction yield from the initial coal typically varies according to coal rank, softening temperature, solvent properties, and reaction conditions of temperature and pressure. For instance, a solvent mixture of CS₂ and N-methylpyrrolidone in the extraction of 29 bituminous coal samples had mass yields between 30 and 66% (Karri 2011).

Many experimental studies have used unit operations such as gravitational settling, centrifugal process, and continuous flow filtration to separate QIs from coal liquids (Hima Bindu et al. 2021). For instance, Smith (1978) used a pressure leaf filter and a rotary drum pressure precoat filter to separate coal particles of 0.1 μm size and the slurry comprising up to 7 wt % of solids with a temperature of 260–315 °C. A different study proposed a vacuum filter to filter QIs using a filter size of 5–10 μm (Sunago and Migitaka). In the present work, a quote from leading Us based vendor for filtration systems for a rotary spindle–driven filter was used to obtain the power rating of the equipment, from which the electricity consumption of the equipment was calculated to be 0.01 kWh/kg of feed slurry. This value is between the values for filtration equipment’s average energy consumption reported in the literature (0.001–0.01 kWh/kg of slurry) (Piccinno et al. 2016). The energy consumption of the filtration was determined to be 4.41 × 10⁻² kWh/kg of mesophase pitch. The filtration was assumed to occur at 20 °C, well below the maximum allowable temperature of filtration equipment of 250 °C.

2.3.3 Mesophication reactor

In the third step (Fig. 4) of the coal-to-mesophase-pitch production process, the isotropic pitch was converted from a homogeneous isotropic structure to a well-ordered anisotropic mesophase pitch structure in a reactor at 410 °C and 100 kPa as shown in Fig. 4. The conversion was performed to produce a precursor that can eventually generate high-modulus carbon fibers compared with carbon fibers generated from isotropic pitch, which possess lower strength and stiffness (Park 2015). The experimentally reported mass yield of mesophase pitch from isotropic pitch was 23.9% for the University of Kentucky’s solvent extraction process. The obtained mass yield was similar to the literature-reported mesophase pitch yield of nearly 20% from coal tar–based isotropic pitch (Özel and K. D. Bartle 2002). The thermal energy demand of the reactor to raise the reactor feed from 25 to 410 °C was estimated using the following specific heat energy correlation for coal tar pitch (Hyman and Kay 1949):

where \(d\) is the specific gravity of the mesophication reactor mixture (assumed to be same as that of coal tar pitch, 1.332), and \(T\) is temperature in °F. Based on this correlation, the energy demand was determined to be 540 kJ/kg of reactor mixture or feed after unit conversions. This specific heat capacity was similar to the value reported in a different study performed on coal tar distillate mixtures, which possessed a specific heat capacity of 0.14 kWh/kg (Briggs and Popper 1957). Assuming a reactor heating efficiency of 75%, the energy requirement was estimated to be 0.842 kWh/kg of mesophase pitch produced.

The agitation energy of the mesophication reactor was estimated in a similar way as that of the mild solvent extraction reactor. The following equation was used to determine the agitation energy (Piccinno et al. 2016):

where \({E}_{{\text{stir}}}\) is the stirring energy in J, \({N}_{\text{p}}\) is 0.79 (assuming a turbulent flow) (Piccinno et al. 2016), \({\rho }_{\text{mix}}\) is the density of the coal and decant oil mixture (1332 kg/m³) (Jang et al. 2017), \(N\) is 1.42 rotations/s (85 rpm), \(d\) is 0.173 m (Piccinno et al. 2016), and \(t\) is 10,800 s (3 h). The energy consumption of the agitator was determined to be 5.59 × 10⁻⁵ kWh/kg of mesophase pitch. The energy consumption of each piece of equipment is summarized in Table 1. It should be noted that the energy source for process heating was natural gas, but low-carbon electricity could also be used for heating to further decarbonize the process. Electrification of process heat is one of four key technological pillars that can achieve industrial decarbonization as outlined in the US Department of Energy’s roadmap report released in 2022 (US Department of Energy 2022).

2.4 Economic analysis of mesophase pitch production

2.4.1 Capital cost estimation

The capital costs of the solvent extraction process were determined for the three major steps of the solvent extraction process: mild solvent extraction, filtration, and mesophication. The capital equipment cost included categories such as equipment purchase, installation, piping, and instrumentation. The equipment purchase cost can typically be determined in two ways, using either equipment cost curves or vendor quotes. In this analysis, the vendor-cost method of estimation was chosen because it typically provides much more accurate cost values for equipment than the theoretical method of cost curves (Symister 2016). It should be noted that this manuscript does not endorse any particular vendor company or their equipment. The cost and capacity information for each major step of the process are stated in Table 2.

A simple multiplicative factor or bare module factor of two was assumed to estimate total capital costs from equipment purchase costs (Faber et al. 2022). This ensured that in addition to equipment purchase costs, additional cost components such as equipment installation, piping, and instrumentation were incorporated in total capital costs. The total annual costs were subsequently amortized annually based on 7% interest and a 15-year loan payment period. Based on these assumptions, the contribution of total capital cost to the total mesophase pitch manufacturing cost was estimated to be $1.91/kg of mesophase pitch.

2.4.2 Raw material costs

The annual raw material costs for the mild solvent extraction process were estimated based on market prices of coal and decant oil. Based on the experimentally reported 1:3 mass ratio of coal to decant oil, the annual consumption of coal and decant oil were estimated for an annual plant capacity of 43 MT/year, and these values are stated in Table 3.

2.4.3 Operating costs

Operating cost comprises costs for energy consumption, labor, and other components such as plant maintenance, contingency, and engineering fees.

2.4.3.1 Energy consumption costs

Based on the energy consumption of each unit operation and source of energy used as shown in Table 1, the energy consumption costs were estimated. The natural gas consumption unit costs were assumed to be $4/MMBtu, whereas electricity unit costs were $0.1/kWh. Importantly, energy (electricity and natural gas) costs are sensitive to location. For example, in 2022, the average industrial electricity rate in southern states such as Alabama, Kentucky, Mississippi, and Tennessee was $0.07/kWh, which was nearly half the average electricity rate for the northern states of Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, and Vermont (EIA 2023). Therefore, a carbon fiber manufacturing plant in a southern state rather than a northern state would have more favorable economics with regard to energy costs because of lower utility costs.

2.4.3.2 Labor costs

Labor costs typically depend on several factors such as plant capacity, unit operations, level of automation, operating mode of process (batch vs. continuous), and geographical location of the plant. The number of personnel in a plant can be estimated using methods such as those provided by Silla (2003), but in the present work the number of labor personnel was estimated based on expert opinions of industry professionals for 43 MT/year of mesophase pitch. One full-time equivalent employee was assumed to dedicate 100% of their time and efforts to the operation of this small-scale plant. This assumption was reasonable because of the small capacities of the equipment (Table 2) and small hourly production rate of 8.16 kg/h in a batch process mode. For a larger-scale plant with a production capacity of 10,000 MT/year of mesophase pitch, 65 employees were estimated based on the suggestions of industry professionals.

For economic modeling purposes, an hourly rate of $29 was assumed for personnel based on the pay information for industrial engineering technologists and technicians in the US Bureau of Labor Statistics (2021) data for West Virginia or Kentucky in 2021. This is only slightly more than the average rate in the US Appalachia region, $24/h (Bowen et al. 2020). West Virginia was chosen as the likely location of the plant because of the state’s abundance of coal resources. Indirect labor and the fringe coefficient for managerial and direct labor were assumed to be 50% and 42% of total labor costs, respectively.

2.4.3.3 Other fixed costs

The other cost components (which included outside battery limits [OSBLs], engineering, contingency, maintenance, and insurance) were mainly based on the plant’s total capital costs and were calculated using the factors listed in Table 4. These cost components were amortized annually based on a 7% interest rate and 15-year plant project life.

2.4.4 Plant scale-up costs

In this analysis, the plant was scaled up from a small-scale production capacity of 43 MT/year of mesophase pitch to 10,000 MT/year of mesophase pitch to examine the effect of scale-up on manufacturing costs. The capital costs for a 10,000 MT/year plant were estimated from the costs for 43 MT/year using the following correlation (Faber et al. 2022):

Scaling up plant capacity can significantly decrease manufacturing costs over time. For example, in the case of semiconductor production, the manufacturing cost of PV decreased by 85% from 2010 to 2020 (NREL 2021) because of a phenomenon called learning by doing, which resulted from standardization in supply chains and fundamental improvements in manufacturing processes and design. In this analysis, the mesophase manufacturing cost was scaled up beyond the capacity of 10,000 MT/year to examine its effect on manufacturing mesophase pitch from coal. The following correlations were used to estimate the manufacturing cost (McQueen 2021):

where \(C\left(x\right)\) is the manufacturing cost after a cumulative total production of \(x\), \(b\) is the reduction rate constant, \(a\) is the technology constant, and \(LR\) is the learning rate. (Fast and slow learning rates of 15% and 10%, respectively, were assumed to examine the effect of learning rate on manufacturing cost [McQueen 2021].)

2.4.5 Carbon fiber manufacturing cost

In this analysis, the produced mesophase pitch (precursor) was converted to 50,000-tow carbon fiber via the 8 process steps shown in Fig. 6. The process steps to produce carbon fibers from coal are very similar to the steps to manufacture commodity-grade PAN-derived fibers except for the exclusion of a pretreatment step that is needed only for a PAN fiber precursor. The major assumptions for carbon fiber manufacturing are stated in the appendix.

Coal-to-carbon-fiber production process. (LT stands for low temperature, and HT stands for high temperature)

2.5 Cradle-to-gate energy assessment of carbon fiber manufacturing

The embodied energy of carbon fibers based on a cradle-to-gate life cycle analysis was estimated via four pathways existing in literature. The scope and boundary of the life cycle energy analysis has been depicted in Fig. 7 (Zhao 2017). The four analyzed pathways were (1) a PAN-based pathway, (2) traditional coke synthesis, (3) Ramaco Carbon liquefaction, and (4) the University of Kentucky’s solvent extraction process. The data for pathways (1),(2) and (3) were taken from Das and Nagapurkar (2021). Each pathway’s embodied energy was analyzed based on three major steps: (1) the initial stage of raw precursor material production, (2) conversion of the raw precursor material into precursor fiber, and (3) conversion of the precursor fiber into carbon fiber. Three types of raw precursor materials were analyzed: petroleum-based acrylonitrile (ACN) and two coal-based materials, coal tar and raw coal. OpenLCA software and Ecoinvent database v3.5 were used to perform the cradle-to-gate energy assessment.

Scope and boundary of life cycle energy analysis

3.1 Economic analysis of mesophase pitch production process

3.1.1 Capital costs

A unit operation breakdown of capital costs for a small-plant capacity of 43 MT/year and a large-plant capacity of 10,000 MT/year is depicted in Fig. 8. In addition to equipment purchase costs, the costs displayed in Fig. 8 for each piece of equipment include the additional costs for installation, piping, instrumentation, and so forth. For both plant capacities, the equipment cost of the filtration unit was the highest of all three pieces of equipment; it represented nearly 50% of total capital cost, whereas the mesophication reactor represented 35% and the mild solvent extraction reactor represented 21%. As the plant capacity expanded from 43 to 10,000 MT/year, the total capital equipment cost decreased by nearly 20% (from $1.90/kg to $0.372/kg of mesophase pitch) because of the economies-of-scale effect.

Breakdown of capital costs for the mild solvent extraction process for plant capacities of 43 MT/year and 10,000 MT/year

3.1.2 Raw material costs

The total raw material cost per kilogram of produced mesophase pitch (broken down by raw material in Fig. 9) was determined to be $1.91/kg of mesophase pitch for both the plant capacities. These total raw material costs were computed based on market prices of decant oil ($0.55/kg) and coal ($0.06/kg). Importantly, the market prices of these raw materials may change because of market volatility or bulk material discounts offered by different vendors.

Breakdown of raw material costs for 43 and 10,000 MT/year mesophase pitch production plants without solvent recycle

The breakdown in Fig. 9 is attributed to the high unit price of decant oil and high consumption of decant oil relative to coal. The unit price of decant oil ($0.55/kg) was nearly 9 times that of coal ($0.06/kg), and the mass consumption of decant oil was 3 times as that of coal.

The analysis assumed decant oil was not recycled; however, initial experimental investigations suggest that 50%–70% of the solvent can be recycled using recovered distillate (from mesophase processing). This could reduce the raw material cost contribution to 0.98–0.61/kg of mesophase. 3.1.3 Operating costs.

The total operating costs for the entire process were estimated to be $0.04/kg of mesophase pitch as shown in Fig. 10. Most of the operating energy costs (52%) arose because of the thermal energy demand of the mild solvent reactor; 33% were from the mesophication reactor's thermal demand, and 10% were from the filtration unit. The agitation electrical energy was found to be negligible in both the reactors.

Breakdown of operating energy costs for a 43 MT/year plant and a 10,000 MT/year plant

The energy costs associated with the mild solvent extraction reactor were the highest among all major process steps because of the high thermal energy demand of mild solvent extraction. High thermal energy demand arose because of high reactor mass loading (4.5 kg reactor feed/kg of mesophase pitch produced, the highest among all process unit operations).

The reaction mixture exiting the mild solvent reactor was cooled from 400 to 25 °C before it was introduced into the filtration unit. The cooling energy consumption was assumed to be the same as the energy needed to elevate the mixture from 25 to 400 °C in the mild solvent reaction mixture (i.e., 1.35 kWh/kg of mesophase pitch produced). Based on the cooling water unit costs of $0.35/GJ, the cooling water costs were determined to be $0.002/kg of mesophase pitch (Turton et al. 1998).

The operating costs (dollar per kilogram of mesophase pitch) were identical for the small and large plants because identical energy consumption was assumed, as shown in Fig. 10. However, pilot-scale investigations need to be conducted to validate this assumption because the specific energy consumption may vary with the expansion of plant capacity. Opportunities for heat integration and optimizations were beyond the scope of this study but would be helpful in further driving down the thermal energy requirement and the associated operating energy–related costs.

3.1.3 Other fixed costs

The total labor costs for the 43 and 10,000 MT/year plants were estimated to be $2.68/kg and $0.75/kg of mesophase pitch, respectively. The labor costs decreased by nearly 72% because of the expansion of plant capacity and the economies-of-scale effect. The total of the other fixed costs also decreased by 72% as the plant expanded in capacity from 43 to 10000 MT/year as shown in Fig. 11.

Breakdown of other fixed costs for 43 and 10,000 MT/year plant capacities

The other fixed costs were expected to decrease because all the contributors to these costs (such as engineering, maintenance, and contingency) were all calculated based on capital equipment costs. Because plant expansion caused capital costs to decline, the other fixed costs also decreased. For both the plant capacities, OSBLs were the largest share of other fixed costs (61%) followed by engineering (22%), contingency (11%), maintenance (4%), and insurance (1%). The OSBL costs were higher because of the high cost factor assumed during their calculation as shown in Table 4. The OSBL cost factor was high because it included costs for utilities such as electric power plants, natural gas, steam, and water supply as well as wastewater treatment and sewage plants.

3.1.4 Total cost breakdown to produce mesophase pitch from coal

The total cost to produce 1 kg mesophase pitch for a 43 MT/year plant capacity was estimated to be $8.20, whereas that cost for a 10,000 MT/year plant was $3.30, a 60% decrease. These costs are depicted in Fig. 12. The cost of mesophase pitch for the 10,000 MT/year plant was lower because of lower capital, labor, and other fixed costs, which were 83%, 72%, and 83% lower, respectively, than the corresponding costs for the small-capacity plant, demonstrating the economies-of-scale effect. Other remaining cost components such as raw materials and operating costs were identical for both the plant capacities. As the plant increased in manufacturing capacity, all the cost components of mesophase pitch manufacturing decreased in value except for raw materials, which remained the same (i.e., $1.91/kg).

Breakdown of mesophase pitch manufacturing costs via the solvent extraction process

For the small-plant capacity, the cost component that contributed the most to the mesophase pitch manufacturing cost was labor, but its contribution decreased as the plant capacity expanded. The raw material cost dominated the mesophase pitch manufacturing cost for the large-plant capacity. This suggests that raw material cost is a major cost contributor for the process. Various strategies such as recycling of decant oil or ordering of bulk raw materials to receive discounted prices could be adopted to drive down mesophase pitch manufacturing costs.

3.1.5 Effect of solvent recycling on mesophase pitch production—A sensitivity analysis

Recycling decant oil can considerably reduce the mesophase pitch manufacturing cost because 97% of the raw material cost was attributed to decant oil. A simple sensitivity analysis for a 10,000 MT/year plant revealed that recycling 50% and 70% of decant oil can reduce raw material cost to $1.9/kg and $0.62/kg of mesophase pitch, respectively, compared with $3.3/kg of mesophase pitch without decant oil recycling, as shown in Fig. 13. As a result, the total mesophase pitch manufacturing cost could be reduced to $2.38/kg of mesophase pitch. At a nearly 70% solvent recycle rate, the manufacturing cost of mesophase pitch via solvent extraction decreased to $2/kg of mesophase pitch, which is 33% lower than the cost of commercial spinnable mesophase pitch produced from Advanced Carbon Products (ACP), LLC, which also likely uses decant oil to produce pitch. Furthermore, the manufacturing cost of mesophase pitch of $3.30/kg for 10,000 MT/year obtained in this study is considerably lower than the PAN price available in the literature, which is $5.00–$6.60/kg depending on commodity oil prices (Choi et al. 2019; Giraud-Carrier and Barlow 2022). This suggests that mesophase pitch produced from solvent extraction could be cost competitive with commercial mesophase pitch currently available in the US market. Despite the economic benefits of solvent recycling process, experimental investigations may be necessary to ascertain the carbon fiber quality would not be negatively impacted due to high recycle rates of solvent.

Effect of solvent (decant oil) recycling on mesophase pitch total manufacturing cost ($/kg) for a 10,000 MT/year plant capacity

3.1.6 Effect of plant scale-up on manufacturing cost of mesophase pitch

The effect of scaling up production capacities beyond 10,000 MT/year of mesophase pitch via the solvent extraction process is shown in Fig. 14. Increasing the production capacities was found to decrease the initial mesophase pitch manufacturing cost ($3.30/kg) of mesophase pitch by nearly 21–50% depending on slow or fast learning rates of scaling up the manufacturing process. The actual rate of decrease would be determined by the level of standardization in supply chains and fundamental improvements in manufacturing and the design of processes. For a fast learning rate, the mesophase pitch manufacturing cost dropped to $2.10/kg at a cumulative production rate of 60,000 MT/year, which is the anticipated total carbon fiber demand within a particular region or country. For a slow learning rate, the pitch manufacturing cost fell to $2.52/kg. Both these costs are well below that of the mesophase pitch produced by ACP, LLC, at $3.30/kg and the PAN market prices of $5.00–$6.60/kg (Choi et al. 2019; Giraud-Carrier and Barlow 2022). However, the mesophase pitch manufacturing cost would still be significantly higher than the production cost ($1.32/kg) of mesophase pitch produced via the novel Ecocoke process (Giraud-Carrier and Barlow 2022). It should be noted that the scaling-up costs were not analyzed in detail here as this is beyond the scope of the study.

Effect of production capacities beyond 10,000 MT/year on mesophase pitch manufacturing cost via the solvent extraction process

3.1.7 Carbon fiber manufacturing cost

The cost contributions of other steps in the carbon fiber manufacturing process from mesophase pitch produced via the solvent extraction process are illustrated in Fig. 15. These costs were estimated based on a 2021 report from Das and Nagapurkar on carbon fiber manufacturing. The costs shown include the materials, capital, and energy to manufacture 50,000-tow pitch carbon fibers with a production capacity of 3750 MT/year from coal-based mesophase pitch. The components of the carbon fiber manufacturing process that contributed most significantly to the total cost were the purchase of raw materials, solvent extraction, and melt spinning, whereas the remaining eight components such as oxidation, carbonization, graphitization, abatement, and winding/inspection/shipping had relatively low individual cost shares of the total carbon fiber manufacturing cost. Each of these remaining components was less than 10% (< $1.10/kg of carbon fibers) of the overall carbon fiber manufacturing cost, as illustrated in Fig. 15.

Breakdown of different cost contributions to the cost of carbon fiber (CF) manufacturing via mesophase pitch from the solvent extraction process (no solvent recycle); costs based on Das and Nagapurkar 2021

Realizing the high sensitivity of decant oil price to total carbon fiber manufacturing cost, a simple sensitivity analysis revealed that a variation in the price of decant oil by ± 30% from a base price of $0.55/kg decant oil would  vary the total carbon fiber production cost ($/kg carbon fiber) by only a minor share of ± 7%. This is because the share of decant oil to total carbon fiber manufacturing costs was only 22% as can be seen from Fig. 16. The precursor manufacture cost (mesophase pitch) contributed the most to the total carbon fiber manufacture cost at nearly 39%; the remaining contributors were other unit operations such as melt spinning (24%), oxidation (10%), low temperature–high temperature carbonization (7%), winding/inspection/shipping (7%), abatement (4%), surface treatment (3%), and sizing (3%).

The rank of coal, its chemical composition and its fluidity may strongly influence the quality of carbon fiber manufactured from it. For instance for coals with higher aliphatic content i.e. with methyl groups may exhibit more linear structure that could enhance its spinnability into fibers. The aliphatic content in coal depends on its rank as low rank coals (sub-bituminous, lignite) that typically have higher aliphatic content than high rank coals, such as bituminous and anthracite. The distribution of molecular weight also affects the quality of carbon fiber produced. A high concentration of low molecular weight compounds within coal may lead to easy breakage of carbon fibers. A uniform molecular weight distribution within coal-based pitch would aid in the manufacture of unform thickness fibers thereby leading to better quality fibers. The fluidity of coal also plays a crucial role in formation of precursor pitch. Furthermore, an increase in the viscosity of mesophase pitch may result in the increase of fiber diameter thereby negatively impacting its spinnability into fibers (Keboletse et al. 2021) (Fig. 16).

Cost breakdown of different processes to produce carbon fiber (CF) from coal

3.1.8 Comparison of carbon fiber manufacturing cost with other literature studies

The total cost to produce carbon fibers from coal and decant oil via the solvent extraction process was estimated to be $11.50/kg for 50,000-tow pitch carbon fiber with a production volume of 3,750 MT/year. The estimated carbon fiber cost is significantly lower than the current PAN-based carbon fiber price ($20/kg–$30/kg) available in the market (Alibaba 2022).

With decant oil recycle rates of 50% and 70% in the solvent extraction process, the manufacturing cost of carbon fiber was estimated to be even lower at $9.90/kg and $9.50/kg of carbon fiber, respectively, as shown in Fig. 17. These total manufacturing costs were lower than the costs of carbon fibers produced from other mesophase pitch–based materials or PAN-based materials. For instance, carbon fibers produced via the Ecocoke process had a manufacturing cost of $11.90/kg of carbon fibers, whereas carbon fibers produced via mesophase pitch produced from traditional coke manufacturing was $10.30/kg of carbon fibers (Das and Nagapurkar 2021). Even the conventional manufacturing cost of carbon fiber from PAN from a 2019 study by Nunna et al. was $10.81/kg of carbon fibers, which was higher than the manufacturing cost of carbon fibers produced via mesophase pitch from solvent extraction with recycling ($9.50/kg–$9.90/kg of carbon fibers) (Nunna et al. 2019). This suggests that carbon fibers produced via mesophase pitch from solvent extraction could be economically competitive.

Precursor cost share in total carbon fiber (CF) manufacturing cost

3.2 Cradle-to-gate energy assessment of carbon fiber manufacturing

The results of this work reveal that the carbon fiber produced from coal via solvent extraction process possessed an embodied energy of 510 MJ/kg, nearly half that of carbon fiber produced conventionally from ACN. Of the four carbon fiber production pathways presented in the study, the carbon fibers produced from ACN possessed the highest embodied energy of 1,190 MJ/kg of carbon fibers, whereas carbon fibers produced from coal tar had the lowest embodied energy at 460 MJ/kg of carbon fibers, as shown in Fig. 18. The embodied energy factors for the processes depicted in Fig. 16 have been provided in the appendix section of the manuscript.

Embodied energy analysis of carbon fibers produced via four pathways. The four analyzed pathways were PAN-based pathway, traditional coke synthesis, Ramaco Carbon (RC) liquefaction, and the University of Kentucky’s solvent extraction process; based on Das and Nagapurkar (2021) and analysis performed in this study

The carbon fibers produced from coal-based materials, on average, possessed 60% lower embodied energy than conventional carbon fibers produced via ACN. This suggests that coal-based carbon fiber manufacturing pathways are more energy efficient than the conventional ACN-based pathway. The embodied energy of carbon fibers via mesophase pitch produced from solvent extraction was 58% lower than that of carbon fibers produced via PAN. This was primarily because the conversion percentage of coal-based mesophase pitch fibers was 1.6 times higher than that of PAN fibers. The weight conversion of mesophase pitch fiber precursor to carbon fibers was 74%, whereas the weight conversion of PAN fibers to carbon fibers was only 45% (Das and Nagapurkar 2021). A different study by Harper International (Harper 2020) also noted the higher yield (> 70%) of pitch-based carbon fiber relative to PAN-based fiber (~ 50%). Because of this higher yield, the carbon fibers manufactured from coal-based material possessed less specific embodied energy than carbon fibers produced from the conventional ACN pathway.

Additionally, other factors contributed to the reduction of embodied energy values in coal-based materials compared with the ACN-based pathway, especially in cases involving mesophase pitch produced through the traditional coke-making synthesis process and Ramaco Carbon’s liquefaction process. This was because of the lower initial embodied energy values of the raw precursors and the reduced energy required during the conversion of these raw precursors into precursor fibers. Consequently, the resulting carbon fibers exhibited lower embodied energy values. For example, coal tar’s specific embodied energy (39 MJ/kg) was nearly half that of ACN (90 MJ/kg), whereas the processing energy for the conversion of coal tar to mesophase pitch fibers was considerably lower than the processing energy of converting ACN to PAN fibers (245–394 MJ/kg of PAN) (Dér et al. 2021). Because of these factors, the carbon fibers produced from coal tar possessed nearly 61% lower total embodied energy than carbon fibers generated via the conventional ACN pathway, as shown in Fig. 18.

In this work, a technoeconomic analysis was performed for the manufacture of carbon fibers by melt spinning a mesophase pitch derived from coal via solvent extraction process using decant oil. For a plant capacity of 10,000 MT/year of mesophase pitch, economic analysis revealed the cost to be $3.30/kg mesophase pitch, with 57% of the cost arising due to decant oil. Sensitivity analysis revealed that recycling 50% and 70% of decant oil reduced the total mesophase pitch manufacturing cost by 28% and 39%, respectively, with initial experimental investigations showing promising results to achieve solvent recyclability.

The total cost to produce carbon fibers was estimated to be $11.50/kg for 50,000-tow pitch carbon fiber with a production volume of 3750 MT/year. The estimated carbon fiber cost is significantly lower than the current PAN-based carbon fiber price ($20–$30/kg) available in the market. With decant oil recycle rates of 50% and 70% in the solvent extraction process, the manufacturing cost of carbon fiber was estimated to be $9.90/kg and $9.50/kg of carbon fiber, respectively.

The cradle-to-gate energy assessment results revealed that the carbon fiber produced from coal via the solvent extraction process possessed an embodied energy of 510 MJ/kg of carbon fiber, which is nearly half that of carbon fiber produced conventionally from PAN. This was primarily because the conversion percentage of coal-based mesophase pitch fibers to carbon fiber was 1.6 times higher than that of PAN fibers. This suggests that carbon fiber produced from coal via the solvent extraction process may be more energy efficient than carbon fiber produced via the conventional can-based pathway.