Waste Management xxx (2018) xxx–xxx

Contents lists available at ScienceDirect

Waste Management
journal homepage: www.elsevier.com/locate/wasman

Comparative environmental and human health evaluations of

thermolysis and solvolysis recycling technologies of carbon fiber
reinforced polymer waste
Y.F. Khalil
Physical Sciences Department, United Technologies Research Center (UTRC), 411 Silver Lane, East Hartford, CT 06018, USA

a r t i c l e i n f o a b s t r a c t

Article history: This quantitative research aims to compare environmental and human health impacts associated with
Received 17 December 2017 two recycling technologies of CFRP waste. The ‘baseline’ recycling technology is the conventional ther-
Revised 8 March 2018 molysis process via pyrolysis and the ‘alternative’ recycling technology is an emerging chemical treatment
Accepted 13 March 2018
via solvolysis using supercritical water (SCW) to digest the thermoset matrix. Two Gate-to-Gate recycling
Available online xxxx
models are developed using GaBi LCA platform. The selected functional unit (FU) is 1 kg CFRP waste and
the geographical boundary of this comparative LCIA is defined to be within the U.S. The results of this
Keywords:
comparative assessment brought to light new insights about the environmental and human health
Thermolysis
Pyrolysis
impacts of CFRP waste recycling via solvolysis using SCW and, therefore, helped close a gap in the current
Solvolysis state of knowledge about sustainability of SCW-based solvolysis as compared to pyrolysis. Two research
Supercritical water questions are posed to identify whether solvolysis recycling offers more environmental and human
CFRP waste health gains relative to the conventional pyrolysis recycling. These research questions lay the basis for
EOL waste formulating two null hypotheses (H0,1 and H0,2) and their associated research hypotheses (H1,1 and
H1,2). LCIA results interpretation included ‘base case’ scenarios, ‘sensitivity studies,’ and ‘scenarios analysis.’
The results revealed that: (a) recycling via solvolysis using SCW exhibits no gains in environmental and
human health impacts relative to those impacts associated with recycling via pyrolysis and (b) use of nat-
ural gas in lieu of electricity for pyrolyzer’s heating reduces the environmental and human health impacts
by 37% (lowest) and up to 95.7% (highest). It is recommended that on-going experimental efforts that
focus only on identifying the best solvent for solvolysis-based recycling should also consider quantifica-
tion of the energy intensity as well as environmental and human health impacts of the proposed solvents.
Ó 2018 Elsevier Ltd. All rights reserved.

1. Introduction Das et al., 2016; Li et al., 2016; Pillain et al., 2017; Meng et al.,
2017). This lightweight structural material has many industrial
1.1. Background applications including commercial and military aircraft, automo-
tive, electronics, construction, sporting goods, etc. (Khalil, 2017).
Global consumption of carbon fiber reinforced polymer (CFRP) Another CFRP application is manufacturing CF-based Type-III and
continues to increase and is estimated to reach
209,800 tons by Type-IV tanks (Khalil et al., 2009, 2010) for on-board vehicular
2020 (Yuyan et al., 2009; Witik et al., 2013; La Rosa et al. 2016; hydrogen storage (whether H2 is stored as compressed gas or in
solids-state forms such as metal hydrides, chemical hydrides, or
adsorbed on activated carbon). For the commercial aircraft indus-
Abbreviations: CF, carbon fibers; CFRP, carbon fiber reinforced polymer; CTUe, try, good cases in point which demonstrate use of CFRP are Airbus
comparative toxic unit for ecotoxicity impacts such freshwater toxicity; CTUh, A350XWB and Boeing 787 Dreamliner (Pimenta and Pinho, 2011;
comparative toxic unit for human toxicity impacts; EOL, end of life; EP, epoxy resin;
Yang et al., 2012; Shuaib et al., 2015), where
50% of the aircraft
FU, functional unit; LCA, life cycle assessment; LCI, life cycle inventory; LCIA, life
cycle impact assessment; nCF, neat carbon fibers; NG, natural gas; PAN, polyacry- weight is composite. Use CFRP in unmanned aerial vehicles
lonitrile; rCF, recycled carbon fibers; RQ, research question; SCF, supercritical fluids; (UAV) is an example of how the military leverages this lightweight
SCW, supercritical water; TRACI 2.1, tool for reduction and assessment of chemicals structural material. Demand for carbon fiber (CF) in the aerospace
and other environmental impacts; TRL, technology readiness level; vCF, virgin defense sector is forecasted to reach 18,462 tons by 2020 com-
carbon fibers.
pared to 7694 tons in 2011 (Robert, 2017).
E-mail address: khalilyf@utrc.utc.com

https://doi.org/10.1016/j.wasman.2018.03.026
0956-053X/Ó 2018 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Khalil, Y.F. Comparative environmental and human health evaluations of thermolysis and solvolysis recycling technolo-
gies of carbon fiber reinforced polymer waste. Waste Management (2018), https://doi.org/10.1016/j.wasman.2018.03.026
2 Y.F. Khalil / Waste Management xxx (2018) xxx–xxx

Current estimates show that
3000 tons of CFRP waste are gen- Thermolysis involves thermal decomposition of the thermoset
erated annually in Europe and the U.S. (Vicki, 2010; Ye et al., 2013; resin matrix to recover CF (Cunliffe et al., 2003; Pickering, 2006;
Shuaib et al., 2015). Sources of this waste include tows cuttings Song et al., 2009; Pompidou et al., 2012; Yang et al., 2012; Greco
during manufacturing, expired prepregs during storage, and EOL et al., 2012; Morin et al., 2012; López et al., 2013; Witik et al.,
CFRP components. Literature data (Lester et al., 2004) shows that 2013). Song et al. (2009), for example, reported that
2.8 MJ of
waste from tows cutting could be as high as 40%.1 Moreover world- energy would be needed to pyrolize 1 kg of CFRP waste and that
wide,
8500 commercial aircraft are expected to retire and disman-
19 MJ/kg CFRP waste could be recovered from pyrolysis’s recy-
tled by 2025 (Carberry, 2008). In this regard, Yang et al. (2012) noted clate byproducts. Witik et al. (2013) reported that use of pyrolysis
that the aerospace industry alone is estimated to reclaim between to recover CF from CFRP waste would consume
5–10% of the
4.5 and 6.8 million kg of rCF (both from manufacturing and EOL) energy required to produce neat CF (nCF). Applying this insight
by 2029. to Khalil’s (2017) estimated energy consumption of 301.3 MJ to
The forecasted rise in CFRP waste generation signals an urgent produce 1 kg CFRP (60 wt% CF), the energy required to produce 1
need for identification of sustainable technologies to process the kg rCF via pyrolysis would range from 15 to 30 MJ/kg rCF. More-
anticipated thousands of tons of CFRP waste. Section 1.2 summa- over, studies on microwave heating of CFRP (Lester et al., 2004;
rizes current and emergent methods for managing CFRP waste. Shuaib et al., 2015). Obunai et al. (2015) reported this emerging
technology is technically feasible, more energy efficient, and faster
compared to the conventional thermolysis. Lester et al. (2004)
1.2. Literature review reported that microwave heating would require about 10 MJ/kg
of CFRP which less than the energy required for conventional
The published literature covers numerous studies on waste pyrolysis (without heat recovery) by about 33% less that the energy
management of CFRP manufacturing waste and EOL components. required for conventional pyrolysis (Shuaib et al., 2015). However,
The reported waste management approaches include landfilling, neither Lester et al. (2004) nor Shuaib et al. (2015) have com-
incineration (with and without heat recovery) and recycling mented on the cost of the microwave recycling technology. With
(Shuaib et al., 2015; Khalil, 2017). Historically, CFRP waste has been respect to the tensile strength of rCF, microwave heating can retain
disposed of in landfills but since 2004 both the U.S. Environmental about 79% of the fiber’s original tensile strength compared to about
Protection Agency (EPA), Land Disposal Restrictions (LDR), and 96% for conventional pyrolysis technology (Shuaib et al., 2015).
European regulation (EU Directive 99/31/EC) continued to impose Thermolysis via Pyrolysis is done by heating the CFRP waste in
constraints on disposal of organic materials (like CFRP) in landfills an inert environment such as nitrogen (N2) at temperatures
(Shuaib et al., 2015; La Rosa et al., 2016; Meng et al., 2017). Aircraft between 400 °C and up to 800 °C (Pickering, 2006). Pyrolysis also
manufacturers are anticipating future regulations on aircraft EOL produces byproduct recyclates, namely, gases (such hydrogen,
composite waste disposal in a manner similar to current regulatory methane, and other non-condensable hydrocarbons), oils and
restrictions on CFRP waste from scrapped vehicles. Incineration of wax (the condensable byproduct), and carbonaceous solid residue
composite waste has its own environmental problems and, thus, (char). The liquid byproduct has a relatively high caloric value
is viewed as an unsustainable solution. Additionally, high cost (embodied energy) similar to fuel oil (
30–40 MJ/kg) and the gas-
and high energy intensity of CF production from the Pan-based pre- eous byproduct has a relatively lower calorific value
15–20 MJ/kg
cursor (Shuaib et al., 2015; Khalil, 2017) has driven the need for (Pickering, 2006). Nunes et al. (2017) examined CFRP waste recy-
recycling of CFRP waste. In this regard, numerous studies have cling via thermolysis using steam and nitrogen gas by a process
reported on CF recycling technologies including microwave heating developed in France. They used 1.4 kg N2 gas and 1.6 kg water
(Lester et al., 2004; Shuaib et al., 2015; Khalil, 2017), pyrolysis (steam) to treat 1 kg of CFRP waste. The total electric energy con-
(Cunliffe et al., 2003), hydro-thermolysis (Pinero-Hernanz et al., sumption in this process was 71.64 MJ (
20 kWh) per kg CFRP
2008), chemical solvolysis (Pinero-Hernanz et al., 2008; Goto, waste. Their LCIA results showed environmental advantages com-
2009). Those studies among others reached the same conclusion pared to waste landfilling.
that recycling CF from manufacturing waste and EOL CFRP compo- Recycling CFRP using the gasification technology involves using
nents has environmental and economic benefits. superheated steam at about 600 °C at atmospheric pressure. The
The CFRP waste recycling approach includes mechanical recy- authors reported that this technology is efficient in removing the
cling, thermolysis, and solvolysis using supercritical fluids (SCFs)2 epoxy resin used in CFRP. Their observation was based on a
like water, alcohols (e.g., methanol, ethanol, n-propanol, etc.), and bench-scale experimental investigation. As a mild ocident, super-
other organic solvents (such as acetone and acetic acid) under differ- heated steam decomposes the polymer matrix into low molecular
ent operating temperatures and pressures (Pimenta and Pinho, 2011; weight hydrocarbons with emission of gases including CO, CO2, H2
Khalil, 2017). and CH4. The authors judged, based on their experimental observa-
Mechanical recycling involves CFRP waste grinding and sieving tions, that this technology could be technically and economically
to separate CF from the resin matrix (Oliveux et al., 2015; Shuaib feasible. However, they arrived at this conclusion without provid-
et al., 2015). The associated source of environmental and human ing a rigorous scaling-up study from their bench-scale results to a
health burdens comes from the electrical energy (MJ/kg CFRP typical industrial setting.
waste) expended in the grinding and sieving processes. Because Researchers like Pinero-Hernanz et al. (2008), Yuyan et al.
mechanical recycling is out of scope of the current research, it will (2009), Morin et al. (2012), Knight (2013), Pinçaud et al. (2014),
not be further discussed in the remainder of this section in order to Yildirir et al. (2014), Dauguet et al. (2015), Henry et al. (2016), La
focus only on recycling via thermolysis and solvolysis using super- Rosa et al. (2016), Keith et al. (2016), and Okajima and Sako
critical water (SCW). (2017) had experimentally examined CFRP waste solvolysis using
SCFs as solvents to digest the thermoset resin into liquid-phase
1
Vartega Carbon Fiber Recycling, December 2016. depolymerized monomers. The removal efficiency of the resin
2
A supercritical fluid (SCF) is a substance at temperatures and pressures above its matrix could be 98% (Pinero-Hernanz et al., 2008) and Yuyan
critical point, where distinct liquid and gas phases do not exist and where the liquid et al. (2009) reported 100% decomposition efficiency of the therm-
and gas phases disappear to become a single supercritical phase. SCFs effuse through
solids just like a gas, dissolve materials just like a liquid, and have characteristics that
set resin using SCW. The operating temperature and pressure of
are useful in many industrial applications including digesting the polymeric resin this chemical treatment are above the critical temperature (TC)
matrix in CFRP waste. and critical pressure (PC) of the solvent. For example, water has

Please cite this article in press as: Khalil, Y.F. Comparative environmental and human health evaluations of thermolysis and solvolysis recycling technolo-
gies of carbon fiber reinforced polymer waste. Waste Management (2018), https://doi.org/10.1016/j.wasman.2018.03.026
 Y.F. Khalil / Waste Management xxx (2018) xxx–xxx 3

TC = 374 °C and PC = 22 MPa and, hence, SCW should be at temper- low-temperature solvolysis using acetic acid-water solution is
atures and pressures above 374.1 °C and 22 MPa (Knight, 2013). In relatively low.
the case of using SCW as the solvent, the resulting liquid waste The gap analysis performed in this research revealed that: (a)
byproduct should contain a mixture of water and organics from Current published literature lacks detailed comparative LCIA-
the depolymerization of the thermoset resin. When chemical treat- based environmental and human health impacts assessment of
ment of CFRP waste involves use of supercritical alcohols such as the emerging solvolysis-based technologies for recycling CFRP
methanol, ethanol, and n-propanol, the operating temperatures waste and (b) There is a crucial need for determining which among
may range from 300 °C to 450 °C and the operating pressure may several recycling technologies offers the least environmental and
range from 5 MPa to 17 MPa (Yang et al., 2012). Some researchers human health burdens.
proposed use of acids to dissolve the thermoset matrix, for exam-
ple, La Rosa et al. (2016) chemically treated CFRP waste with acetic
2. Research problem statement, objective, and originality
acid (25 vol% solution at 85 °C for 1.5 h) to reclaim the reinforcing
CF from the polymer matrix. Depolymerization of the thermoset
2.1. Problem statement
resin in CFRP waste was also carried out in ethylene glycol and
ethylene glycol/water mixtures at near-critical conditions of the
As a result of growing annual global consumption of CFRP, thou-
two solvents (Yildirir et al., 2014). Currently, however, solvolysis
sands of tons of CFRP waste are anticipated to be generated annu-
using SCFs remains in early stages of development and (Shuaib
ally from several industries, including the aerospace, automotive,
et al., 2015; Rybicka et al., 2016) assigned this chemical treatment
and wind turbine industries. Managing CFRP waste by landfilling
technology as a low technology readiness level (TRL 2–4) com-
is facing continued regulatory restrictions, and incineration is
pared to thermolysis which is a more matured technology (i.e.,
viewed as an unsustainable practice. Also, CF is conventionally
high TRL level in the range 8–9). Recycling via solvolysis also faces
manufactured from a costly petroleum-based precursor called
other technical challenges including high cost and material com-
polyacrylonitrile (PAN) through an energy intensive process
patibility issues due to their highly corrosive nature (Yuyan et al.,
(Khalil, 2017). Accordingly, recycling CFRP waste is considered to
2009).
be the optimal waste management strategy. Presently, there is a
Life cycle impact assessment (LCIA) studies reported in the pub-
crucial need for determining which among several recycling tech-
lished literature were limited to comparing the environmental bur-
nologies offers the least environmental and human health burdens.
dens and human toxicity of various thermolysis processes (namely,
LCIA appears to be the appropriate methodology to address this
pyrolysis and fluidized beds) versus landfilling and incineration
central sustainability question.
with and without heat recovery (Bjorklund and Finnveden, 2005).
For example, Witik et al. (2013) compared the environmental via-
bility of recycling via pyrolysis against incineration and landfilling. 2.2. Objective
Pinçaud et al. (2014) is the only research team who compared the
environmental burdens of CFRP waste recycling via solvolysis The objective of this research is to employ LCIA quantitative
using SCW versus landfilling. Their study concluded that the methods and tools to compare the environmental and human
solvolysis process is greener despite the associated high electricity health impacts associated with the two CFRP waste recycling tech-
consumption (assumed to be dominated by nuclear power). More- nologies, namely, conventional thermolysis via pyrolysis (i.e., ther-
over, Pinçaud et al. (2014) did not report either the amount of elec- mal decomposition in an inert environment) and the emerging
trical energy (MJ) needed to produce 1 kg of SCW, or SCW mass solvolysis technology which uses supercritical water (SCW) to
required per 1 kg CFRP waste. Finally, the study did not discuss chemically digest the thermoset polymer matrix contained in CFRP
the technical challenge associated with material compatibility waste.
issues as SCW is known to be a highly corrosive reaction medium.
Moreover, in our opinion, use of landfilling as a benchmark in those 2.3. Originality and impact
comparative LCIA studies is not scientifically sound since ther-
moset polymers are generally non-biodegradable and may require Our gap analysis of relevant published literature revealed, to the
hundreds of years to gradually biodegrade and release greenhouse best of our knowledge, the current lack of the following: (a) A com-
gases (GHG) like methane. In this regard, Witik et al. (2013) had to prehensive and quantitative environmental and human health
assume a 1% degradation rate over the first 100 years of CFRP impacts assessment which compares the conventional pyrolysis
waste disposal in landfills and other studies assumed
26% degra- technology vs. emerging solvolysis recycling technology of CFRP
dation of 60,000 years. Hence, deriving conclusions based on waste, (b) Identification of which among several recycling tech-
benchmarking against CFRP waste disposal in landfills is subjective nologies offers the least environmental and human health burdens,
and dependent on assumptions made (Witik et al., 2013). and (c) Reporting the energy intensity and technical challenges
La Rosa et al. (2016) used acetic acid-water binary solution (25 associated with use of SCFs to depolymerize the thermoset resin
wt% acetic acid) at 80 °C to recover CF from waste CFRP. Their matric and recover CF. In this regard, it is worth mentioning that
chemical treatment seems to be less energy intensive compared the relevant published work on recycling via solvolysis have invari-
to recycling via solvolysis using CCFs. In our research, we applied ably highlighted only merits of solvolysis in terms of producing
GaBi LCA platform to model the recycling process employed by cleaner rCF (compared to the pyrolysis process which may leave
La Rosa et al. (2016). Our LCIA calculations showed that the some char residue on rCF) and improved mechanical properties
cradle-to-gate primary energy consumption to produce 1 kg of of the recovered CF. Unfortunately, nothing was reported to the
acetic acid-water binary solution (25 wt% acetic acid) is about energy intensity associated with producing those SCFs (such as
42.16 MJ. Also, about 2.37 kg CO2-Equiv would be emitted to pro- SCW).
duce 1 kg acetic acid–water binary solution at 80 °C. Since La The originality of this research are threefold as follows:
Rosa et al. (2016) used 1 kg acetic acid solution (25 wt% acetic acid)
per 0.577 kg CFRP, we calculate 1.73 kg acetic acid solution to treat  Performance of comparative environmental and human health
1 kg CFRP, which will lead to emission of 4.1 kg CO2-Equiv/kg- impacts assessment of CFRP waste recycling via pyrolysis
CFRP. Accordingly without performing cradle-to-gate LCIA calcula- (which is the conventional and matured technology) vs. emerg-
tions, it might seem that the energy required to perform ing recycling via solvolysis using supercritical water (SCW).

Please cite this article in press as: Khalil, Y.F. Comparative environmental and human health evaluations of thermolysis and solvolysis recycling technolo-
gies of carbon fiber reinforced polymer waste. Waste Management (2018), https://doi.org/10.1016/j.wasman.2018.03.026
4 Y.F. Khalil / Waste Management xxx (2018) xxx–xxx

 Adoption of a comprehensive approach to LCIA results interpre- Null hypothesis (H0,2):

tation by including not only the base case scenarios, but also
include sensitivity studies (SS) and scenarios analysis (SA). Solvolysis-based recycling of CFRP waste does not reduce
Inclusion of SS and SA in interpreting LCIA quantification results adverse human safety impacts relative to those associated with
is critically important as it helps addressing uncertainties asso- thermal recycling.
ciated with both primary and secondary data used in the devel-
oped recycling models. Research hypothesis (H1,2):
 Identification of where additional research efforts should be
directed with respect to solvolysis using SCFs for CFRP waste Solvolysis-based recycling of CFRP waste could reduce adverse
recycling. human safety impacts relative to those associated with thermal
recycling.
If the provided evidence supports the claim made by the two
2.4. Research impact
research hypotheses, then the posed claim should be accepted.
Otherwise, the research hypotheses should be rejected in favor of
The insights gained from this research help close a gap in the
the claim made by the null hypotheses. It should be noted that
current state of knowledge about sustainability of SCW-based
whether claims made by the research hypotheses are accepted or
solvolysis as a recycling technology for CFRP waste and how it
rejected, new insights are generated that can help advancing cur-
compares to recycling via pyrolysis.
rent state of knowledge about the topic being investigated. The
generated insights should be used by the researcher to shed new
3. Research methodology light on where additional research efforts and funding should be
directed.
The adopted research methodology is composed of two
steps that guided this comparative assessment. These two steps 3.3. LCA methods and tools
are:
The LCA methodology is an internationally recognized approach
 Formulation of the research questions. described in ISO Standards 14040-2006 and ISO 14044-2006
 Hypothesis testing. (Bovea and Vidal, 2004; Khalil, 2017). The LCA quantitative tool
being employed in this investigation is the GaBi platform,3 GaBi
3.1. Research questions (RQs) thinkstep life cycle inventory (LCI) databases,4 and TRACI 2.1 LCIA
methodology.5
The following two research questions are formulated to LCA methods and tools have been employed to generate quan-
address sustainability of CFRP waste recycling via solvolysis using tifiable evidence that can either support or reject the claims of the
SCW: proposed H1,1 and H1,2. If the quantifiable evidence fails to support
the claims of H1,1 and H1,2, then claims made by H0,1, and H0,2
(a) To what extent is solvolysis more environmentally sustain- should be accepted and vice versa. Rejection of claims of H1,1 and
able compared to thermal recycling? H1,2 is equally valuable to the case of accepting these claims since
(b) To what extent is solvolysis safer from the human health in both cases new scientific insights are revealed to help closing
perspective compared to thermal recycling? the identified knowledge gap about the environmental and human
health impacts associated with emerging vs. conventional recy-
cling technologies of CFRP waste.
3.2. Hypothesis testing (HT)
3.4. LCA recycling models (i.e., product systems) and assumptions
Having formulated the aforementioned RQs, hypothesis test-
ing methodology is then applied to translate each of the pro-
Two CFRP waste recycling models are developed using GaBi LCA
posed research questions into an equivalent null hypothesis
platform. These models and their system boundaries are shown in
(denoted as H0) and an associated alternative hypothesis
Figs. 1 and 2. As can be seen, the scope of this comparative assess-
(denoted as H1). The null hypothesis states a specific claim that
ment excludes other life cycle phases, namely, CFRP production
is assumed to be true unless there is a strong evidence against it.
phase and use phase. Fig. 1 shows two input mass flows (CFRP
The research (or alternative) hypothesis articulates a new claim
waste and N2 gas to provide the inerting environment inside the
that can be accepted if the researcher can provide evidence to
pyrolyzer). The two input energy flows represents the electricity
support this claim.
required for CFRP waste grinding and either electricity or natural
Based on the RQs established in Section 3.1, two null hypothe-
gas to heat the pyrolyzer and its content to the required tempera-
ses H0,1 and H0,2) and two research hypotheses (H1,1 and H1,2) are
ture. The output mass flows (recyclates) represent rCF, pyrolysis
formulated as follows:
gases (containing hydrogen, methane, and other hydrocarbon
gases), oils & wax, and carbonaceous solid residue (char). In theory,
Null hypothesis (H0,1):
the pyrolysis gases could be used as fuel to heat the pyrolyzer and,
hence, could offset some of the electricity (or natural gas) energy
Solvolysis-based recycling of CFRP waste does not offer reduc-
required for pyrolyzer’s heating. Also, the oils & wax could be used
tion in environmental impacts relative to those associated with
as fuels or as feedstock for manufacturing other chemicals. How-
thermal recycling.
ever, these options are not included in this study.

Research hypothesis (H1,1):

3
GaBi ts LCA platform. Source: http://www.gabi-software.com.
4
GaBi databases 2017 edition. Source: http://www.gabi-software.com/uploads/me-
Solvolysis-based recycling of CFRP waste could offer reduction dia/GaBi_Databases_2017_Upgrades_and_Improvements.pdf.
in environmental impacts relative to those associated with thermal 5
Tool for the Reduction and Assessment of Chemical and Other Environmental
recycling. Impacts (TRACI), Version 2.1. User’s Manual; EPA/600/R-12/554 2012.

Please cite this article in press as: Khalil, Y.F. Comparative environmental and human health evaluations of thermolysis and solvolysis recycling technolo-
gies of carbon fiber reinforced polymer waste. Waste Management (2018), https://doi.org/10.1016/j.wasman.2018.03.026
 Y.F. Khalil / Waste Management xxx (2018) xxx–xxx 5

Fig. 1. Model for CFRP waste recycling via pyrolysis.

Fig. 2. Model for CFRP waste recycling via solvolysis using SCW.

Fig. 2 shows two input mass flows (CFRP waste and deionized could be further treated to produce useful chemical products.
water to provide SCW). The two input energy flows represent the However, this possibility is not included in the current study.
electricity required for CFRP waste grinding and to operate the Finally, the following assumptions have been made and applied
pumps and compressors to produce steam at the required super- to the recycling models (Figs. 1 and 2):
critical pressure. Also, natural gas burning is required to produce
steam from the deionized feedwater. The output mass flows (recy-  Recycling facility is located at the same site where CFRP waste is
clates) represent rCF, liquid-phase mixture of organic monomers collected. The deliberate Exclusion of CFRP waste transport to
(from depolymerizing the thermoset resin) mixed with water, the recycling facility is intended to avoid introducing sources
and carbonaceous solid residue (char). In theory, the monomers of subjectivity into this study. To be included into the recycling

Please cite this article in press as: Khalil, Y.F. Comparative environmental and human health evaluations of thermolysis and solvolysis recycling technolo-
gies of carbon fiber reinforced polymer waste. Waste Management (2018), https://doi.org/10.1016/j.wasman.2018.03.026
6 Y.F. Khalil / Waste Management xxx (2018) xxx–xxx

Table 1
Primary and secondary data for thermolysis (via pyrolysis) model (Fig. 1).

Input and Output Flows Recycling via Thermolysis (Pyrolysis) at 700 °C Primary Data Sources Secondary Data Sources
Input mass & energy flows Input mass flows: 1 kg CFRP and 0.94 kg N2 gas  Pickering (2006) GaBi LCI databases for U.S.:
Input energy flows: 0.53 MJ (grinding energy) and 12 MJ for  Morin et al. (2012)  Grid electricity mix
pyrolyzer  Yang et al. (2012)  Thermal energy from natural
Output mass flows rCF: 0.6 kg  Witik et al. (2013) gas
(recyclates) Pyrolysis gases1: 0.02 kg  Howarth et al.
N2 gas: 0.94 kg (2014)
Oils & wax2: 0.12 kg  Khalil (2017)
Carbonaceous solid residue3: 0.26 kg  Meng et al. (2017)
1
Pyrolysis gases represent the non-condensable part and contain hydrogen, methane, and other hydrocarbons.
2
Oils and wax represent the condensable part of pyrolysis recyclates and contain a mixture of complex organic compounds.
3
This solid recyclate is the char residue resulting from thermally decomposing the thermoset resin matrix.

models, the transport distance and transport type between [kg SO2-Equiv], ecotoxicity [CTUe], eutrophication [kg N-Equiv],
CFRP waste collection site and recycling facility will have to global warming [kg CO2-Equiv], human health-particulates [kg
be subjectively assumed. PM2.5-Equiv], human toxicity-carcinogenics [CTUh], human
 Recycling facility is geographically located within the U.S. health-non-carcinogenics [CTUh], ozone depletion [kg CFC 11-
Hence, the electricity required for recycling (input energy flows Equiv], and smog [kg O3-Equiv].
in Figs. 1 and 2) is assumed to be provided by the U.S. grid elec-
tricity mix. Similarly, the thermal energy production via natural 4.1. Recycling via pyrolysis
gas burning (input energy flows in Figs. 1 and 2) is assumed to
be in the U.S. It should be noted that the U.S. electricity grid mix 4.1.1. Base case scenario
is represented by a portfolio of 20% nuclear, 65% fossil (of which The pyrolysis base case scenario assumes 0.53 MJ/kg CFRP
30% from coal, 34% from natural gas, and 1% from oil) and 15% waste for the grinding energy and 12 MJ/kg CFRP waste for pyroly-
renewables.6 zer’s heating using NG. The sources of these primary data are
 CFRP waste is assumed to contain 60 wt% CF and 40 wt% ther- depicted in Table 1.
moset (epoxy resin) matrix. Fig. 3 shows the calculated mid-point impact categories using
 CFRP waste recycling via thermolysis without energy recovery TRACI 2.1 LCIA methodology. Two observations are to be noted in
has been considered in this study. this figure. Firstly, the Y-axis is plotted using a logarithmic scale
 The emerging recycling via solvolysis using SCW as a solvent is in order to include values of all the impact categories and secondly,
considered to be the alternative scenario for rCF production. the plotted impact categories have different units as shown on the
 Sources of CFRP waste include expired and/or off-cutting man- x-axis. For example, the carbon footprint (signified by global
ufacturing prepregs and EOL CFRP components. warming) associated with this scenario is
0.96 kg CO2 equiv.
 Materials of construction of the two recycling facilities (pyroly- per kg CFRP waste and ecotoxicity is
0.03 CTUe units per kg CFRP
sis and solvolysis) are not taken into account in the recycling waste.
models. This assumption has been made to avoid another Another key takeaway is that the energy required to recover 1
source of uncertainty in LCIA calculations. kg rCF using pyrolysis (with NG for pyrolyzer’s heating) is
5.5% (if
NG is used for pyrolyzer’s heating) and
11.7% (if electricity is used
3.5. LCA functional unit (FU) for pyrolyzer’s heating) of the energy required to produce 1 kg nCF
from PAN precursor (Khalil, 2017).
Per ISO Standards 14040 and 14044, studies that use LCA meth-
ods should define a functional unit (FU) upon which LCIA calcula- 4.1.2. Sensitivity studies (SS)
tions are based. In this research, the selected FU for the baseline Grinding energy (mechanical size reduction) has been selected
scenario and alternative scenario is 1 kg of CFRP waste. The choice as the sensitivity parameter due to the associated wide range of
of this mass-based FU is intended to make the assessment results values (0.27–2.03 MJ/kg with an intermediate value of 0.53 MJ/kg
easily scalable. CFRP waste cited in the published literature for this parameter
(Witik et al., 2013; Howarth et al., 2014; Meng et al., 2017).
3.6. Life cycle inventory (LCI): sources of primary and secondary data Fig. 4 displays the pyrolysis sensitivity study using grinding
energy as the sensitivity parameter. The origin point (0%, 0%) in
Input mass and energy flows as well as output flows (recy- Fig. 4 represents the case where the grinding energy is 0.53 MJ/
clates) for the two recycling models (Figs. 1 and 2) are shown in kg CFRP waste (i.e., base case scenario). It is assumed that the pyr-
Tables 1 and 2. The primary data are extracted from the relevant olyzer is heated by NG as recommended in this research. As Fig. 4
literature sources as noted in Tables 1 and 2 and the secondary shows, ozone depletion is the most sensitive impact category (line
data which represent the input energy flows (Figs. 6 and 7) are with the steepest slope) to changes in the grinding energy followed
taken from GaBi LCI databases. by human health (particulates). The ecotoxicity category is the
least sensitive category (line with the lowest slope) to changes in
4. Results and discussion the grinding energy. The key insight that can be obtained from this
sensitivity study is that use of energy-efficient grinding equipment
The comparative LCIA is carried out using GaBi software plat- (which consumes less than 0.53 MJ/kg CFRP waste) can have a
form and TRACI 2.1 database. The mid-point impact categories major impacts on reducing the ozone depletion category.
being calculated in this comparative assessment are: acidification
4.1.3. Scenarios analysis (SA)
6
GaBi LCI database for U.S. electricity grid mix and the U.S. Energy Information TRACE 2.1 LCIA methodology for environmental and human
Administration at www.eia.gov. health impacts are compared for the following two scenarios:

Please cite this article in press as: Khalil, Y.F. Comparative environmental and human health evaluations of thermolysis and solvolysis recycling technolo-
gies of carbon fiber reinforced polymer waste. Waste Management (2018), https://doi.org/10.1016/j.wasman.2018.03.026
 Y.F. Khalil / Waste Management xxx (2018) xxx–xxx 7

Table 2
Primary and secondary data for solvolysis model (Fig. 2).

Input and Output Flows Recycling via Solvolysis using SCW at 380 Primary Data Sources Secondary Data Sources
°C and 24 MPa
Input mass & energy flows Input mass flows: 1 kg CFRP and 4.67 kg  Pinero-Hernanz et al. (2008) GaBi LCI databases for U.S.:
SCW (plus added 0.131 kg KOH catalyst)  Morin et al. (2012)  Grid electricity mix
Input energy flows: 0.53 MJ (grinding  Knight (2013)  Thermal energy from natural gas
energy). Energy to produce 4.67 kg SCW  Howarth et al. (2014)
consists of 145.34 MJ thermal energy from  Dauguet et al. (2015)
natural gas and 25 MJ of electrical energy  Henry et al. (2016)
Output mass flows (recyclates) rCF: 0.6 kgLiquid-phase (mixture of  La Rosa et al. (2016)
monomers and water)  Khalil (2017)
: 5.062 kg  Meng et al. (2017)
Solid residue: 0.139 kg

Fig. 3. Base case pyrolysis scenario (0.53 MJ grinding energy and 12 MJ thermal energy from NG). Calculations are per 1 FU, i.e., 1 kg CFRP waste.

 Pyrolyzer is electrically heated (Morin et al., 2012; Yang et al., ity are among the desired properties of SCW (Loppinet-Serani et al.,
2012; Witik et al., 2013) using the U.S grid mix. 2008). Moreover compared to other SCFs (like methanol, ethanol,
 Pyrolyzer is heated via NG burning as proposed in this research. n-propanol, acetone, acetic acid, nitric acid, and ethylene glycol),
water is thought to be a greener reaction medium that is abundant
Based on the comparative impact assessment results shown in and less costly, and have no toxic effects. The aforementioned
Fig. 5, use of NG heating in lieu of U.S. electricity mix for pyroly- desired properties of SCW support our rationale for selecting
zer’s heating would lead to
95.7% reduction on ozone depletion SCW as the solvent that can efficiently digest the thermoset resin
(highest), followed by
91.6% reduction in human health (particu- matrix in CFPF waste. Nonetheless, as this research reveals, pro-
lates), and
37% reduction in ecotoxicity (lowest). The reduction in duction of SCW is an energy intensive process with adverse envi-
carbon footprint (global warming) is
55.4%. ronmental and human health impacts that exceed those impacts
associated with recycling CFRP waste via pyrolysis.
4.2. Recycling via solvolysis using SCW
4.2.1. Base case scenario
When water is compressed and heated to pressures and tem- The calculated values of environmental and human health
peratures above its critical point (PC = 22.1 PMa and TC = 374 °C), impact categories of solvolysis using SCW base case scenario are
SCW is produced. Under such conditions, SCW becomes an excel- plotted in Fig. 6. Note that in order to plot all the values of these
lent solvent with viscosities, densities and other physical proper- impact categories, a logarithmic scale is used for the Y-axis. All cal-
ties that are intermediate between those of its gaseous and culations are based on 1 FU (i.e., 1 kg CFRP waste). According to
liquid states. Good heat-transporting properties and high diffusiv- Fig. 6,
16.2 kg CO2-Equiv is emitted per kg of CFRP waste being

Please cite this article in press as: Khalil, Y.F. Comparative environmental and human health evaluations of thermolysis and solvolysis recycling technolo-
gies of carbon fiber reinforced polymer waste. Waste Management (2018), https://doi.org/10.1016/j.wasman.2018.03.026
8 Y.F. Khalil / Waste Management xxx (2018) xxx–xxx

Fig. 4. Pyrolysis sensitivity study: grinding energy as the sensitivity parameter. Legend: (1) Ozone depletion [kg CFC 11-Equiv.], (2) Human health, particulates [kg PM2.5-Equiv.],
(3) Acidification [kg SO2-Equiv.], (4) Eutrophication [kg N-Equiv.], (5) Human toxicity, non-carcinogenics [CTUh], (6) Human toxicity, carcinogenics [CTUh], (7) Smog [kg O3-Equiv.],
(8) Global warming [kg CO2-Equiv.], and (9) Ecotoxicity [CTUe].

Fig. 5. Scenarios analysis: Pyrolyzer’s heating using U.S. electricity grid mix vs. natural gas (calculations per 1 FU, i.e., 1 kg CFRP waste).

chemical treated by the emerging solvolysis using SCW Dauguet et al., 2015). Appendix A provides more details on SCW
technology. production.
The x-axis in Fig. 7 shows percent changes in the electrical
energy in the positive and negative directions around the origin
4.2.2. Sensitivity studies (SS) point (0%, 0%). The y-axis shows the corresponding percent change
Two sensitivity studies, SS-1 and SS-2, are performed for the in the calculated environmental and human health impacts in
SCW-based solvolysis technology. The selected sensitivity parame- response to changes in the x-axis. As can be seen from Fig. 7, the
ter for SS-1 is the electrical energy (25 MJ) used for operating the most sensitive impact category to changes in the electrical energy
pumps and compressors associated with production of 4.67 kg (line with steepest slope) is human health, particulates [kg PM2.5-
SCW required for depolymerizing 1 kg CFRP waste (Knight, 2013; Equiv.] followed by ecotoxicity [CTUe] and acidification [kg SO2-

Please cite this article in press as: Khalil, Y.F. Comparative environmental and human health evaluations of thermolysis and solvolysis recycling technolo-
gies of carbon fiber reinforced polymer waste. Waste Management (2018), https://doi.org/10.1016/j.wasman.2018.03.026
 Y.F. Khalil / Waste Management xxx (2018) xxx–xxx 9

Fig. 6. Impact categories associated with solvolysis using SCW: base case scenario. (Calculations based on 1 FU, i.e., 1 kg CFRP waste).

Equiv.]. The least sensitive parameter to changes in the electrical mal energy from NG to produce the required amount of SCW (4.67
energy (line with lowest slop) is global warming [kg CO2-Equiv.]. kg) for this solvolysis process (Knight, 2013; Dauguet et al., 2015).
The main take away from SS-1 is that reduction of used electrical The selected sensitivity parameter for SS-2 (Fig. 8) is the ther-
energy below 25 MJ/kg CFRP waste (which is the base case value) mal energy from NG (145 MJ/kg CFRP waste) used for production
would lead to environmental and human health gains (the nega- of 4.67 kg SCW required for dissolving of 1 kg CFRP waste
tive side of the y-axis). For example, more energy-efficient pumps (Knight, 2013; Dauguet et al., 2015).
and compressors could be used in the SCW-based solvolysis pro- The x-axis in Fig. 8 represents the percent change in NG energy
cess. It should be noted that the origin point (0%, 0%) in Fig. 7 rep- in the positive and negative directions around the origin point (0%,
resents the base case energy consumption for solvolysis of 1 kg 0%). The most sensitive impact category to changes in NG energy
CFRP waste, namely: 25 MJ of electrical energy and 145 MJ of ther-

Fig. 7. Solvolysis sensitivity study SS-1: Percent change in calculated impact Fig. 8. Solvolysis sensitivity study SS-2: Percent change in calculated impact
categories vs. percent change in electrical energy (Calculations based on 1 FU, i.e., 1 categories vs. percent change in electrical energy. Calculations based on 1 FU, i.e., 1
kg CFRP waste). Legend: (1) Human Health, Particulates [kg PM2.5-Equiv.], (2) kg CFRP waste. Legend: (1) Global Warming [kg CO2-Equiv.], (2) Eutrophication [kg N-
Ecotoxicity [CTUe], (3) Acidification [kg SO2-Equiv.], (4) Human toxicity, non-carcino- Equiv.], (3) Human toxicity, carcinogenics [CTUh], (4) Smog [kg O3-Equiv.], (5) Human
genics [CTUh], (5) Human toxicity, carcinogenics [CTUh], (6) Ozone Depletion [kg CFC toxicity, non-carcinogenics [CTUh], (6) Ecotoxicity [CTUe], (7) Acidification [kg SO2-
11-Equiv.], (7) Eutrophication [kg N-Equiv.], (8) Smog [kg O3-Equiv.], and (9) Global Equiv.], (8) Human Health, Particulates [kg PM2.5-Equiv.], and (9) Ozone Depletion [kg
Warming [kg CO2-Equiv.] CFC 11-Equiv.].

Please cite this article in press as: Khalil, Y.F. Comparative environmental and human health evaluations of thermolysis and solvolysis recycling technolo-
gies of carbon fiber reinforced polymer waste. Waste Management (2018), https://doi.org/10.1016/j.wasman.2018.03.026
10 Y.F. Khalil / Waste Management xxx (2018) xxx–xxx

Table 3 Fig. 9 shows the impacts assessment results for the three sce-
Scenarios analysis for SCW-based solvolysis. narios provided in Table 3. The main takeaways from this scenarios
Base Case Scenario 1 Scenario 2 analysis are twofold:
Scenario
Thermal energy from NG, 145 72.5(50% of 36.25(25% of (a) Environmental and human health gains can be achieved by
MJ/kg CFRP waste1 base case) base case) reducing the energy expenditure in the SCW-based solvoly-
Electrical energy, MJ/kg 25 12.5(50% of 6.25(25% of sis below the energy consumption associated with the base
CFRP waste2 base case) base case
case scenario. For example, a comparison between scenario
1
Thermal energy required to generate steam from feedwater to the boiler (see 1 and the base case scenario reveals that the highest reduc-
Fig. x in Appendix xx for SCW production Aspen HYSYS simulation).
2
tion would be
49.8% and correspond to the ecotoxicity cat-
Electrical energy required to operate FW pump and two compressors (see Fig. x
egory followed by 49.4% reduction which corresponds to
in Appendix xx for SCW production Aspen HYSYS simulation).
global warming. The lowest reduction is 31.1% and corre-
(line with steepest slope) is global warming [kg CO2-Equiv.] fol- sponds to ozone depletion.
lowed by eutrophication [kg N-Equiv.] and human toxicity, car- (b) Comparison of impact categories of SCW-based solvolysis
cinogenics [CTUh]. The least sensitive parameter to changes in scenarios (Fig. 9) vs. pyrolysis scenarios analysis (Fig. 5)
NG energy (line with lowest slope) is ozone depletion [kg CFC would reveal that even if energy requirement for SCW-
11-Equiv. The main take away from SS-2 is that reduction of used based solvolysis can be reduced to 25% of the energy
NG energy (below 145 MJ which is the base case value) would lead requirement for the base case scenario (Table 3), the impact
to environmental and human health gains (the negative side of the categories associated with pyrolysis (whether pyrolyzer’s
y-axis in Fig. 8). For example, use of a more efficient-energy steam heater comes from electricity or from NG burning) remain
boiler would be recommended. It should be noted that the origin below those associated with solvolysis scenarios (Fig. 9).
point (0%, 0%) in Fig. 8 represents the base case energy consump- For example, the carbon footprint associated with solvolysis
tion for solvolysis of 1 kg CFRP waste, namely, 25 MJ of electrical ‘Scenario 2’ is about 6 times higher than that associated with
energy and 145 MJ of thermal energy from NG to produce the scenario where pyrolyzer is electrically heated and about 58
required amount of SCW for this solvolysis process. Appendix A times higher than that associated with scenario where pyro-
provides more detailed HYSYS simulation model for SCW lyzer’s heat is from NG burning (Fig. 5).
production.
4.3. Comparative assessment: pyrolysis vs. solvolysis using SCW
4.2.3. Scenarios analysis (SA)
Comparative environmental and human health impacts assess- Qualitative arguments have been made about merits and draw-
ments have been conducted for three scenarios shown in Table 3. backs of CFRP waste recycling via pyrolysis versus solvolysis (Keith
Scenarios 1 and 2 hypothetically assume that improvements have et al., 2016). For example, a major drawback of pyrolysis is inability
been made to reduce the energy intensity of SCW production by to recover the thermoset resin since it is thermally destroyed and,
50% and 25%, respectively, compared to the base case scenario. hence, the life cycle loop is not closed. However, potential use of

Fig. 9. Scenarios analysis: SCW-based solvolysis.

Please cite this article in press as: Khalil, Y.F. Comparative environmental and human health evaluations of thermolysis and solvolysis recycling technolo-
gies of carbon fiber reinforced polymer waste. Waste Management (2018), https://doi.org/10.1016/j.wasman.2018.03.026
 Y.F. Khalil / Waste Management xxx (2018) xxx–xxx 11

Fig. 10. Percent reduction in impact categories (Eq. (1)) as a result of using pyrolysis in lieu of solvolysis for CFRP recycling.

pyrolysis gases as a fuel and use of pyrolysis oils and wax as feed- technologies of CFRP waste, namely: (a) thermolysis via pyrolysis
stock for other chemicals could, to some extent, offset the afore- and (b) solvolysis using SCW. The former technology was consid-
mentioned drawback. One of the drawbacks CRFP waste ered a ‘base line’ for benchmarking purposes with the latter ‘‘alter-
recycling via solvolysis, in addition to high cost and energy usage, native’ technology. Interpretation of LCIA results for the selected
is that resulting liquid phase (containing organic monomers) rep- recycling technologies included base case scenarios, sensitivity
resents major health and safety hazards if not chemical treated studies, and scenarios analysis. LCIA calculations were based on 1
to make other valuable products. Moreover, the presence of FU, i.e., 1 kg of CFRP waste. The principal conclusions of this
organic residue on rCF require additional treatment with chemical research are four fold: (i) The calculated nine impact categories
solvents (such as acetone) to remover this residue from the recov- associated with solvolysis are higher than those from pyrolysis
ered fiber (Keith et al., 2016). One of key merits of solvolysis, how- (Section 4.3). For example, human health impact (particulates) cat-
ever, is that the resulting rCF can have mechanical properties close egory from solvolysis is
78 times greater than that from pyrolysis
to that of nCF. and ecotoxicity from solvolysis is
76 times greater than that from
This research brings to light new insights by conducting com- pyrolysis. Carbon footprint (i.e., global warming) from solvolysis is
parative environmental and human safety impacts assessment of
17 times greater than that from pyrolysis and ozone depletion
pyrolysis (with NG for pyrolyzer’s heating) and SCW-based solvol- category from solvolysis is
3 times greater than that from pyrol-
ysis for CFRP waste recycling. Results of the carried out compara- ysis, (ii) Accordingly, the comparative assessment results provided
tive impacts assessment are summarized in Fig. 10. quantitative evidence to support rejection of the research hypothe-
Fig. 10 delineates the percent reduction in impact categories as ses (H1,1 and H1,2) in favor of the associated null hypotheses (H0,1
a result of using pyrolysis in lieu of solvolysis for recycling CFRP and H0,2). Details are provided in ‘Section 3.2,’ (iii) Use of NG in lieu
waste. The Y-axis is calculated using Eq. (1) as follows: of electricity for pyrolyzer’s heating would lead to
95.7% reduc-
tion in ozone depletion (highest), followed by
91.6% reduction in
Percent reduction in impact category
human health (particulates), and
37% reduction in ecotoxicity
ðImpacts CategoryÞjSolvolysis  ðImpact CategoryÞjPyrolysis (lowest). The reduction in carbon footprint is
55.4% (Sec-
¼ ð1Þ
ðImpact CategoryÞjPyrolysis tion 4.1.3), and (iv) The comparative assessment results provided
quantitative evidence that CFRP waste recycling via pyrolysis is
Eq. (1) has been applied to each of the nine impact categories more advantageous from environmental and human health per-
(shown on x-axis) to calculate the corresponding percent reduction spectives compared to solvolysis using SCW. The aforementioned
(shown on y-axis). The percent reduction in impact categories advantages could be further amplified by: (a) Use of pyrolyzer’s
ranges from 62.2% (lowest) for ozone depletion and up to 98.7% gaseous recyclate as a fuel to displace some of the thermal energy
(highest) for ecotoxicity and human health (particulates). required for pyrolyzer’s heating and (b) Use of pyrolyzer’s oils and
wax recyclate as fuels or intermediates for production of other
5. Conclusions chemicals. Finally, in addition to the on-going experimental inves-
tigations that focus only on identifying the best solvent for the
In the present research, comparative environmental and human solvolysis technology, it is equally important that future research
health impacts assessments have been performed for two recycling should consider quantification of the energy intensity as well as

Please cite this article in press as: Khalil, Y.F. Comparative environmental and human health evaluations of thermolysis and solvolysis recycling technolo-
gies of carbon fiber reinforced polymer waste. Waste Management (2018), https://doi.org/10.1016/j.wasman.2018.03.026
12 Y.F. Khalil / Waste Management xxx (2018) xxx–xxx

environmental and human health impacts of the proposed Lester, E. et al., 2004. Microwave heating as a means for carbon fiber recovery from
polymer composites: a technical feasibility study. Mater. Res. Bull. 39, 1549–
solvents.
1556.
Li, X. et al., 2016. Environmental and financial performance of mechanical recycling
Acknowledgements of carbon fiber reinforced polymers and comparison with conventional disposal
routes. J. Clean. Prod. 127, 451–460.
López, F. et al., 2013. Recovery of carbon fibres by the thermolysis and gasification
The constructive comments and reviews provided during this of waste prepreg. J. Anal. Appl. Pyrol. 104, 675–683.
research by Professors Richard E. Wetzler and Marshall T. Spriggs Loppinet-Serani, A. et al., 2008. Current and foreseeable applications of supercritical
of Harvard University, Cambridge, Massachusetts are greatly water for energy and the environment. ChemSusChem 1, 486–503.
Meng, F. et al., 2017. Energy and environmental assessment and reuse of fluidised
appreciated. Also, the author would like to thank Susan Killoran, bed recycled carbon fibres. Composites Part A 100, 206–214.
a Fellow Librarian at the Harris Manchester College, University of Morin, C. et al., 2012. Near- and supercritical solvolysis of carbon fibre reinforced
Oxford, UK, for her valuable assistance during the literature review polymers (CFRPs) for recycling carbon fibers as a valuable resource: state of the
art. J. Supercrit. Fluids 66, 232–240.
phase of this search. Nunes, A. et al., 2017. Life cycle assessment of a steam thermolysis process to
recover carbon fibers from carbon fiber-reinforced polymer waste. Int. J. Life
References Cycle Assess. https://doi.org/10.1007/s11367-017-1416-6, 1-14.
Obunai, K. et al., 2015. Carbon fiber extraction from waste CFRP by microwave
irradiation. Composites Part A 78 (2015), 160–165.
Bjorklund, A., Finnveden, G., 2005. Recycling revisited life cycle comparisons of
Okajima, I., Sako, T., 2017. Recycling of carbon fiber-reinforced plastic using
global warming impact and total energy use of waste management strategies. J.
supercritical and subcritical fluids. J. Mater. Cycles Waste Manage. 19, 15–20.
Resour. Conserv. Recycl. 44 (4), 309–317.
Oliveux, G. et al., 2015. Current status of recycling of fibre reinforced polymers:
Bovea, M.D., Vidal, R., 2004. Increasing product value by integrating environmental
review of technologies, reuse and resulting properties. Prog. Mater Sci. 72, 61–
impacts, costs, and customer valuation. J. Resour. Conserv. Recycl. 41 (2), 133–
99.
145.
Pickering, S., 2006. Recycling technologies for thermoset composite materials—
Carberry, W., 2008. Airplane recycling efforts benefit Boeing operators. Boeing AERO
current status. Composites Part A 37, 1206–1215.
Mag. QRT 4 (08), 6–13.
Pillain, B. et al., 2017. Identification of key sustainability performance indicators and
Cunliffe, A. et al., 2003. Pyrolysis of composite plastic waste. Environ. Technol. 24,
related assessment methods for the carbon fiber recycling sector. Ecol. Indic. 72,
653–663.
833–847.
Das, J. et al., 2016. Global Carbon Fiber Composites Supply Chain Competitiveness
Pimenta, S., Pinho, S.T., 2011. Recycling carbon fibre reinforced polymers for
Analysis. Oak Ridge National Laboratory (ORNL), Tennessee, Knoxville.
structural applications: Technology review and market outlook. Waste Manage.
Dauguet, M. et al., 2015. Recycling of CFRP for high value applications. Proc. CIRP
(Oxford) 31, 378–392.
Life Cycle Eng., 1–7
Pinçaud, M. et al., 2014. Environmental feasibility of the recycling of carbon fibers
Goto, M., 2009. Chemical recycling of plastics using sub- and supercritical fluids. J.
from CFRPs by solvolysis using supercritical water. ACS Sustain. Chem. Eng. 2,
Supercrit. Fluid. 47, 500–507.
1498–1502.
Greco, A. et al., 2012. Thermal and chemical treatment of recycled carbon fibres for
Pinero-Hernanz, R. et al., 2008. Chemical recycling of carbon fiber reinforced
improved adhesion to polymeric matrix. J. Compos. Mater., 369–377
composites in nearcritical and supercritical water. Composites Part A 39, 454–
Henry, L. et al., 2016. Semi-continuous flow recycling method for carbon fiber
461.
reinforced thermoset polymers by near- and supercritical solvolysis. Polym.
Pompidou, S. et al., 2012. Recycling of carbon fiber. Identification of bases for a
Degrad. Stab. 133, 264–274.
synergy between recyclers and designers. In: ASME 2012 11th Biennial
Howarth, J. et al., 2014. Energy intensity and environmental analysis of mechanical
Conference on Engineering Systems Design and Analysis. ASME, Nantes, France.
recycling of carbon fiber composite. J. Clean. Prod. 81, 46–50.
Robert, T. (2017). The carbon fiber industry worldwide 2011–2020: An evaluation of
ISO 14040, 2006a. International Organization for Standardization. Environmental
current markets and future supply and demand. Retrieved from
Management – Life Cycle Assessment – Principles and Framework. ISO14000
CarbonFiber.com: <http://www.carbonfiber-report.com/Special.pdf>.
International Standards Compendium, Switzerland.
Rybicka, J. et al., 2016. Technology readiness level assessment. J. Clean. Prod. 112,
ISO 14044, 2006b. International Organization for Standardization. Environmental
1001–1012.
Management – Life Cycle Assessment – Requirements and Guidelines. ISO14000
Shuaib, N.A. et al., 2015. Resource efficiency and composite waste in UK supply
International Standards Compendium, Switzerland.
chain. Procedia CIRP 29, 662–667.
Keith, M. et al., 2016. Optimization of solvolysis for for recycling carbon fiber
Song, Y. et al., 2009. Life cycle energy analysis of fiber-reinforced composites.
reinforced composites. In: ECCM17 – 17th European Conference on Composite
Composites Part A 40, 1257–1265.
Materials, 26–30th, June, Munich, Germany, pp. 1–7.
Vicki, P., 2010. Launching the carbon fiber recycling industry. Reinf. Plast. 54, 33–37.
Khalil, Y.F. et al., 2009. Quantifying and addressing the DOE material reactivity
Witik, R. et al., 2013. Carbon fibre reinforced composite waste: an environmental
requirements with analysis & testing of hydrogen storage materials & systems.
assessment of recycling, energy recovery and landfilling. Compos. A Appl. Sci.
U.S. Department of Energy, Annual Merit Review (AMR), Paper # STP-50-Khalil,
Manuf. 49, 89–99.
May 20, 2009, Arlington, VA, pp. 1–30.
Yang, Y. et al., 2012. Recycling of composite materials. Chem. Eng. Process. 51, 53–
Khalil, Y.F. et al., 2010. Quantifying and addressing the DOE material reactivity
68.
requirements with analysis & testing of hydrogen storage materials & systems.
Ye, S.Y. et al., 2013. Parameter optimization of the steam thermolysis: a process to
U.S. Department of Energy, Annual Merit Review (AMR), Paper # ST012, June 8,
recover carbon fibers from polymer-matrix composites. Waste Biomass. Valor 4,
2010, Washington, DC.
73–86.
Khalil, Y.F., 2017. Eco-efficient lightweight carbon-fiber reinforced polymer for
Yildirir, E. et al., 2014. Recovery of carbon fibers and production of high quality fuel
environmentally greener commercial aviation industry. J. Sustain. Prod.
gas from the chemical recycling of carbon fiber reinforced plastic wastes. J.
Consump. 12, 16–26.
Supercrit. Fluids 92, 107–114.
Knight, C., 2013. Recycling High-performance Carbon Fiber Reinforced Polymer
Yuyan, L. et al., 2009. Recycling of carbon fibre reinforced composites using water in
Composites using Subcritical and Supercritical Water. UMIT Dissertation
subcritical conditions. Mater. Sci. Eng., A 520, 179–183.
Publishing, Ann Arbor, MI 48106-1346, UMI no. 3596525, ProQuest, LLC.
La Rosa, A.D. et al., 2016. Recycling treatment of carbon fiber/epoxy composites:
materials recovery and characterization and environmental impacts through
life cycle assessment. Composites Part B 104, 17–25.

Please cite this article in press as: Khalil, Y.F. Comparative environmental and human health evaluations of thermolysis and solvolysis recycling technolo-
gies of carbon fiber reinforced polymer waste. Waste Management (2018), https://doi.org/10.1016/j.wasman.2018.03.026