Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Simulation-based environmental-impact assessment of glycerol-to-hydrogen conversion technologies

Simulation-based environmental-impact assessment of glycerol-to-hydrogen conversion technologies Environmental Impact Glycerol Reforming Assessment Technologies Aqueous-Phase TRACI Mid-Point Impact Categories Reforming (APR) Acidification Human toxicity, carcinogens Ecotoxicity Glycerol from Auto-Thermal Eutrophication Human toxicity, Bio-Diesel non-carcinogens Reforming (ART) Production Global warming Ozone depletion Human health particulate matter Smog formation Supercritical Water Reforming (SCWR) Aspen HYSYS Simulation GaBi Platform for Life of Reforming Technologies Cycle Assessment (LCA) Keywords: glycerol; hydrogen; impact categories; reforming; supercritical water; valorization Need for greener H production to curb CO emissions Introduction 2 2 According to the US Environmental Protection Agency Background [3], the transportation sector was responsible for ≈27% of Hydrogen utilization GHG emissions in 2015 and, within this sector, commer - Currently, hydrogen (H) is mainly used in industrial cial aircraft were responsible for ≈9% of GHG emissions in processes rather than for energy production. The main 2015 [3]. The US Environmental Protection Agency (EPA) industrial uses of H include ammonia production (≈54% reported that light-duty internal combustion engine (ICE) of the overall H consumption) and oil refineries (≈35% vehicles were responsible for 60% of the GHG emissions at- of H consumption). Other uses including chemicals syn- tributed to the US transportation sector in 2015. As a result, thesis (e.g. methanol production) and the food industry hydrogen has been sought as a promising energy carrier are responsible for the rest of the H consumption [1]. for road transport. Direct H use to replace petrol-based As an energy carrier, H is used for the transport sector fuel in ICE vehicles or as a feedstock for on-board PEM fuel and as a feedstock to proton exchange membrane (PEM) cells can contribute to curbing GHG emissions from this fuel cells for electricity production. For the latter appli- sector. Also,in situ renewable H production in refuelling cation, in 2016, ≈62 000 fuel-cell systems (equivalent to stations has been pursued to eliminate GHG emissions as- 500 MW of power) were shipped worldwide [2]. Globally, sociated with H transport from production sources to re- 7.2 EJ of energy from H was consumed in 2013, which fuelling stations [2]. represents ≈1.3% of the world primary energy consump- The UN Intergovernmental Panel on Climate Change tion [1]. (IPCC) reported that the civil aviation industry currently contributes ≈2.5% of the world’s manmade CO emissions [4, 5]. Over the past three decades, the annual growth of Hydrogen-production sources: fossil-based and water world commercial air transport has averaged ≈5% and is electrolysis projected to double over the next decade or two [6–8]. As the Currently, H-production technologies are mainly fossil- civil aviation fleet grows to meet increasing air-transport based, which leads to significant CO emissions. For ex- demands, the IPCC predicts that the contribution of the ample, ≈48% of H production comes from steam-methane aviation industry to global manmade CO emissions will reforming (SMR), 30% is derived from petroleum refining increase to ≈3% in 2050 [9]. To curb GHG emissions associ- and 18% is produced by coal gasification. In the USA only, ated with the civilian aviation sector, there is an emerging ≈10 million tons of annual H production comes from SMR interest in the development of regional hybrid electric technology. Due to its fossil origin, H production causes aircraft (HEA) powered by combinations of conventional ≈60 million metric tons of CO emissions annually, which gas turbines and electric propulsion systems, comprising accounts for ≈2% of the energy-related CO emissions [3]. Water electrolysis, which enables H production without CO emissions, accounts for the remaining 4% of this Aircraft with 30–90 PAX (aircraft passengers), 4–15 tons of gross greenhouse-gas (GHG) emission [1]. take-off weight and energy of 30 000–70 000 kWh. Downloaded from https://academic.oup.com/ce/article/5/3/387/6317702 by DeepDyve user on 13 July 2021 Clean Energy | 389 motors powered by either batteries or PEM fuel cells [10], during biomass cultivation, pretreatment, oil extraction with the latter technology requiring H storage on board and refining, the transesterification process and glycerol- aircraft. In addition to the emerging HEA technology, the purification steps, CO is also absorbed during biomass full-electric aircraft technology remains a long-term vision farming, hence reducing the net emissions of CO. of aircraft manufacturers like Airbus and Boeing. It is also worth mentioning that members of the EU aviation organ- 1.2 Objective izations and industry counterparts created the ‘Flightpath 2050’ vision to reduce global CO , nitrogen oxides (NO ) and This research aims to perform comparative assessments of 2 x noise emissions such that, by 2050, civil aircraft should energy consumption, environmental burdens and adverse pollute 75% less CO, 90% less NO and 65% less noise [11]. human-health impacts of glycerol-based H production via 2 x The aforementioned background information explains the three reforming technologies, viz. supercritical water the currently observed global efforts to pursue H pro- reforming (SCWR), APR and autothermal reforming (ATR), duction from sustainable sources to curb GHG emissions, respectively. To achieve this objective, life-cycle impact as- other gaseous air pollutants and particulate matter, all of sessment (LCIA) is performed for base case (BC) scenarios which have adverse impacts on human health and the en- in which the US electricity-grid mix is assumed to be the vironment. In particular, bio-based H production via re- electricity supply source to power pumps and compressors forming technologies seems to offer a viable alternative in the reforming processes. The US grid electricity mix [20] fuel compared to the use of petroleum-based fuels (viz. is as follows: 39.5% natural gas, 23% coal, 20% nuclear and gasoline and diesel fuel) in the road transport sector and 17.5% renewables (of which 7.2% is wind power, 6.4% hydro- jet fuels in the aviation sector. power, 1.7% solar, 1.3% biomass and 0.4% geothermal). Also, natural-gas burning is assumed to supply the thermal en- ergy required for the reformer and for heating the reactant 1 Motivation, objectives and novelty of mixtures to the required reforming temperature. LCIA is research also performed for sensitivity scenarios in which wind power is assumed to be the electricity source instead of the 1.1 Motivation electricity-grid mix and biogas burning instead of natural The discussion provided in the subsection on the ‘Need gas to supply the thermal energy of reformers. Additionally, for greener H production to curb CO emission’ above to- 2 2 optimization strategies to reduce reforming energy require- gether with the following rationale for selecting glycerol as ments (both electrical and thermal) are also considered in the bio-based feedstock for greener H production are the the LCIA assessment scenarios. main motivators of this research: • Glycerol is a by-product of biodiesel production via the 1.3 Novelty transesterification of agricultural crops like soybeans, Unlike previous investigations that focused only on rapeseed, sunflower, jatropha, palm, castor, etc. [12–14]. estimating the carbon footprint associated with a single In this chemical process, ≈10 m of glycerol is produced glycerol-reforming technology, this research addresses per 90 m of rapeseed-derived biodiesel produced [15, current voids in the state of knowledge about the three 16]. In this research, we calculated ≈7.5 m of glycerol selected glycerol-reforming technologies by providing the per 90 m of soybean-derived biodiesel. According to following quantitative insights: Ciriminna et al. [17], the transesterification of vegetable oils for biodiesel production leads to ≈10 kg of glycerol by-product per 100 kg of biodiesel produced. • comparative energy consumption and midpoint impact • US biodiesel annual production is ≈5.5 million m /year. categories among these reforming technologies; In 2016, for example, the USA was the world’s biggest • efficient energy-management strategies for these biodiesel producer [18]. reforming technologies. These strategies include: (i) • EU biodiesel annual production is ≈12.5 million m /year capturing exothermic heat from the reforming reac- [19]. tion steps, (ii) burning exhaust combustible gases to • The increased production of biodiesel since 2000 to the offset the endothermal heat of the reformer and (iii) for present in both the USA and the EU led to a substantial SCWR technology, which is energy-intensive due to the reduction in glycerol prices. high-pressure requirement, the electricity-generation option in conjunction with H production has been In this author’s opinion, glycerol-reforming technolo- evaluated. gies offer a promising bio-based H source, in particular aqueous-phase reforming (APR), which could ultimately The remaining sections of this manuscript are organized replace the current use of SMR technology, which emits as follows. Section 2 discusses the key findings of the rele- ~11  kg CO -equiv. per kg of H produced [14]. In this re- vant published work, Section 3 presents the adopted re- 2 2 gard, CO emission associated with glycerol feedstock is search methods and tools, and Section 4 discusses the expected to be low because, while this GHG is emitted results of this environmental sustainability assessment. Downloaded from https://academic.oup.com/ce/article/5/3/387/6317702 by DeepDyve user on 13 July 2021 390 | Clean Energy, 2021, Vol. 5, No. 3 The principal results and major conclusions of this re- estimated that ≈3.77 kg CO -eqiv/kg H was produced after 2 2 search are summarized in Section 5. taking credit for their assumption of biogenic versus fossil- based CO emissions. Rahman [27] investigated 1 wt% glycerol (in DI water) APR over a series of nickel (Ni) and 2 Literature review copper–nickel (Cu–Ni) bimetallic catalysts supported on multiwalled carbon nanotubes (MWNT) at 240 C and 40 at- This research focuses on bio-based greener H production mospheres. His test results showed that the Cu–Ni catalyst via glycerol-reforming technologies and, hence, the scope led to 86% H selectivity and 84% glycerol conversion, and of this literature review is limited to achieving this purpose. noted that the presence of Cu in bimetallic catalysts led The three glycerol-based technologies being considered to the suppression of undesirable methanation reactions. herein are: SCWR, APR and ATR. Glycerol is a by-product Subramanian et  al. [16] studied glycerol APR over a series of biodiesel production via the transesterification of agri- of γ-Al O -supported metal nanoparticle catalysts for H 2 3 2 cultural crops. (More details and statistical data about this production in a batch reactor. They found that Pt/Al O 2 3 feedstock are provided in Section 1.1). was the most active catalyst under the test conditions. The remainder of this subsection provides brief high- The optimum conditions for H production were found to lights of key published research on H production via gly- be 240 C, 42 bar, 1000 r.p.m. stirring speed and ≥4100 sub- cerol reforming. Wen et  al. [21] studied the activities and strate/metal molar ratio for a 10-wt% glycerol feed. stabilities of Pt, Ni, Co and Cu catalysts and their sup- Schwengber et al. [19] published a review paper on gly- port substrates for H production via APR of glycerol. cerol reforming and highlighted key characteristics of the Authayanun et  al. [22] investigated the thermodynamics reported technologies. Larimi et al. [28] synthesized a series of H production via ATR of crude glycerol derived from of M-doped Pt/MgO (where M could be palladium, ruthe- the biodiesel-production processes. They assumed gly- nium, rhodium, iridium or chromium) nano-catalysts and cerol and methanol to be the primary components of the examined their performance effects on glycerol APR. Boga crude glycerol and found that H production increases as et al. [29] investigated the effect of individual components the ratio of glycerol to methanol increases and as the re- of crude glycerol on APR activity over 1 wt% Pt/Mg(Al)O, 1 forming temperature increases. Ortiz et al. [23] performed wt% Pt/AlO , 5 wt% Pt/AlO and 5 wt% Pt/C catalysts at 2 3 2 3 thermodynamic analysis of the glycerol SCWR process 29 bar and 225°C. The use of a 10-wt% alkaline crude gly- TM using AspenPlus and used the Soave–Redlich–Kwong cerol solution in water, containing 6.85 wt% glycerol, 1.62 (SRK) equation of state to identify the thermodynamically wt% soaps, 1.55 wt% methanol and 0.07 wt% ester, caused favourable operating conditions for glycerol conversion to a noticeable drop in APR activity compared to the corres- H . They used from 1 to 16  mol % (≈5–50 wt%) of glycerol ponding 6.85-wt% solution of pure glycerol in water. concentration in the feed stream, with water being the Most recently, Veluturla et  al. [30] discussed the use of balance, and conducted SCWR at 240 bar and 800 C. Zhang different types of heterogeneous catalysts for glycerol val- et al. [24] provided characterization of the Pt–Re catalysts orization through processes like carboxylation, conden- under conditions encountered during APR to obtain an sation, esterification, etherification, phase reformation, improved understanding of the role of the Rhenium (Re) selective oxidation, hydrolysis transesterification, se- catalyst. Their results showed the importance of sur - lective reduction and pyrolysis. Putra et al. [31] performed face acidity in controlling the reaction pathways during glycerol APR and phenol hydrogenation with Raney Ni at glycerol APR. Ortiz et  al. [25] simulated glycerol SCWR to 180–240°C for phenol-to-glycerol ratios from 0 to 4.88. They produce H and electricity. Avasthi et al. [26] focused their selected phenol as a model lignin monomer to demon- research on the glycerol/steam reforming (GSR) technology strate the feasibility of using renewable H from glycerol and discussed current challenges (e.g. higher prices of bio- APR for lignin upgrading. Ghani et  al. [32] reformed crude diesel compared to diesel fuel) and potential solutions. In glycerol over modified cerium–zirconium (Cs–Zr) supports this author’s opinion, none of the proposed solutions (e.g. loaded with 5  wt% nickel catalyst to generate H via ATR in situ removal of CO and H as soon as they form inside 2 2 (a combination of partial oxidation and steam reforming). the reformer) seems to be practical. They studied the effects of the reforming temperature, Rocha et al. [12] used a cradle-to-gate (C2G) life-cycle as- steam-to-carbon ratio (S/C), oxygen-to-carbon ratio (O/C), sessment (LCA) approach to quantify the environmental reduction temperature and calcination temperature, and and health impacts of the production of biodiesel from reported the following optimum operating conditions: a soybean and palm oil. Their analysis was performed using reforming temperature of 550C, S/C of 2.6, O/C of 0.5, re- SimaPro software and CML 2000 LCIA methodology. Galera duction temperature of 600 C and calcination temperature and Ortiz [14] evaluated the environmental perform- of 550 C. Bastan et al. [33] investigated the effects of the Al/ ance of H and electricity production via glycerol SCWR. Mg ratio in a series of Ni nano-catalysts supported on Al O 2 3 They used SimaPro 8.0. CML 2000 in their LCA study and and MgO on the physico-chemical characteristics of Ni/ Al O –MgO catalysts and determined the optimum cata- 2 3 Glycerol vaporization heat is 58.2 kJ/mol and its boiling point at o lyst for H production via glycerol APR. They found that the 1 a.t.m. is 290 C. 2 APR activity of synthesized catalysts strongly depended Platinum–rhenium catalyst. Downloaded from https://academic.oup.com/ce/article/5/3/387/6317702 by DeepDyve user on 13 July 2021 Clean Energy | 391 on the Al/Mg ratio and demonstrated that the Ni/Al Mg (i) Goal and scope definition step that also includes catalyst possessed the highest catalytic activity of 92% specifying the boundaries of the system being analysed glycerol conversion and selectivity towards a hydrogen and defining the functional unit (FU). The goal here is production of 76%. Charisiou et  al. [34] investigated the to conduct a quantitative environmental sustainability GSR reaction for H production and compared the per - assessment to identify which, among three glycerol- formance of Ni supported on ZrO and SiO –ZrO catalysts. reforming technologies, would be characterized as the 2 2 2 They suggested that the addition of SiO stabilizes the ZrO most sustainable and environmentally friendly tech- 2 2 monoclinic structure, restricts the sintering of Ni particles, nology for greener H production. The scope is about 2+ strengthens the interaction between Ni species/support H production from glycerol feedstock and, hence, the and influences the distribution of the gaseous products system boundary is denoted as a cradle-to-production by increasing the H yield (while not favouring the trans- gate (C2G) boundary (Fig. 1). Finally, the selected FU is formation of CO to CO). Thus, a high H /CO ratio can be assumed to be 1 kg of produced H . 2 2 2 achieved with a negligible CO/CO ratio. Shejale and Yadav (ii) LCI analysis that includes input/output materials and [44] investigated the GSR and compared two catalysts, viz. energy flows and direct emissions. During this step, Ni–Cu/La O –MgO and Ni–Co/La O –MgO, prepared by the the process input/output mass and energy flows as 2 3 2 3 co-precipitation and impregnation techniques for the GSR well as compositions are calculated and data gaps reaction. are identified. In this regard, LCI data fall into two With respect to the valorization of glycerol to produce categories: primary (viz. foreground) and secondary value-added products, Kaur et  al. [39] provided a compre- (viz. background) data. In this research, the pri- hensive assessment for the production of value-added mary data for each of the three glycerol-reforming products from crude glycerol. Their assessment included technologies is produced using Aspen HYSYS process the environmental and economic aspects of different simulation. The secondary data (such as the environ- glycerol-conversion routes such as the chemical and bio- mental and human-health burdens associated with chemical methods. the glycerol by-product of biodiesel production via the transesterification of biomass) is directly obtained from the GaBi LCI database. 3 Methods (iii) LCIA, where the environmental burdens of the mate- Sections 3.1 and 3.2 present the methods and tools em- rial and energy flows as well as human-health impacts ployed in this research. Section 3.3 presents the roadmap are quantified.  for integrating Aspen HYSYS simulation results into GaBi The impact assessment of interest to this research product system models to calculate environmental bur - focuses on human health and the natural environ- dens and human-health impacts, and Section 3.4 pro - ment. As discussed in Khalil [8 , 35] as well as in the vides life-cycle inventory (LCI) primary data generated ILCD Handbook [36], there are many methodologies for using Aspen HYSYS simulations of SCWR, APR and ATR conducting LCIA such as CML 2002, Eco-Indicator 99, EPS technologies. 2000, Impact 2002+, ReCiPe 2008, USEtox, TRACI, etc. However, we selected to adopt the TRACI methodology that is currently used in the GaBi LCA software package. 3.1 LCA Also, the midpoint characterization methodology has This research follows the LCA methodology described in lower uncertainty in the results compared to the end- ISO 14040:2006 standards [40]. The LCA methodology com- point characterization used in other methodologies prises four interrelated steps as follows [8, 35]: such as Eco-Indicator 99 [37]. The author of this work Gaseous Glycerol emissions Glycerol Reforming Plant Other reactants Bio-based (SCWR, APR, or ATR) (water, air) H2 gas Water from Thermal and gas/liquid electrical energy System Boundary separator Fig. 1: System boundary of the cradle-to-gate (hydrogen-production plant) system 4 5 GaBi ThinkStep LCA platform. Source: http://www.gabi-software. TRACI: A Tool for Reduction and Assessment of Chemicals and com. Other Environmental Impacts. Downloaded from https://academic.oup.com/ce/article/5/3/387/6317702 by DeepDyve user on 13 July 2021 392 | Clean Energy, 2021, Vol. 5, No. 3 also used TRACI LCIA methodology in his previous re- H , compositions of gaseous and liquid process streams, search [8, 35].  pressure and temperature conditions of the input and TRACI 2.1 LCIA methodology—the US EPA developed output streams, as well as operating conditions inside the impact-assessment methodology (viz. TRACI) for the simulated unit operations (viz. reformer, water–gas the purpose of LCA studies [38]. The impact categories shift (WGS) reactors, gas/liquid separator and pressure- are characterized at the midpoint level draw simple swing adsorber). The Aspen HYSYS CPA fluid property cause–effect chains to show the point at which each package has been selected for performing the thermo- impact category is characterized and reflect the cur - dynamic calculations in the simulated glycerol-reforming rent state of developments, consistently with EPA technologies. regulations and policies, as well as the best available Aspen HYSYS simulations also allowed modelling LCIA practices. The impact categories used in the energy-management strategies such as capturing exo- TRACI methodology are shown in Table 1. The GaBi thermic heat from some of the reforming reaction steps platform includes TRACI 2.1 (which is the latest v -er and from burning exhaust combustible gases to offset the sion of this database). overall endothermal heat of the reformer. For SCWR tech- (iv ) Interpretation of the analysis results including sensi- nology, which is energy-intensive due to the high-pressure tivity studies and scenarios analysis. In this research, requirement, the in situ electricity-generation option in three BC scenarios and four sensitivity scenarios are conjunction with H production has been evaluated using developed and analysed, as discussed in Section 4. Aspen HYSYS simulation. As Table 1 shows, there are two impact categories that dir - 3.1.2 Integration of process simulation with LCA models ectly refer to human health, namely the carcinogens and Because GaBi LCI databases do not contain LCI data for the non-carcinogens impact categories. However, other im- bio-based production of H from glycerol via SCWR, APR pact categories also affect human health, such as particu- or ATR, it was necessary to use the Aspen HYSYS platform late matter, smog formation, ozone depletion and global to simulate each of the three reforming processes. Fig. 1 warming [37]. Also, NO emissions are represented in the shows the HYSYS simulation of the APR process. HYSYS eutrophication and acidification impact categories. Note simulations of the SCWR and ATR have unit operations that SO and NO are contributors to acid gases, which are x x similar to those shown in Fig. 1. However, the main dif- represented in the acidification potential (measured in kg ferences among the three HYSYS simulations include the SO -equiv.). operating conditions and feed streams of the reformer. In this research, the GaBi LCA tool (2018 edition) has For example, SCWR uses supercritical water and glycerol, been used to develop C2G product system models for the and ATR uses air (in addition to water and glycerol) in the glycerol-reforming technologies being evaluated. GaBi reformer. product system models account for the environmental Description of the unit operations depicted in Fig. 2. A mixer and human-health burdens associated with glycerol (a is used to blend the glycerol and water (in the required by-product of the biodiesel process), which is an input feed weight percentages), a pump to compress the feed mix- stream to GaBi product system models. ture to the desired operating pressure of the reformer and a heat exchanger to heat the feed to the desired 3.1.1 Process design and simulation operating temperature of the reformer. Downstream of The use of Aspen HYSYS is deemed necessary to simu- the reformer is an expander valve to lower the pres- late each of the three glycerol-reforming technologies sure of the product gases followed by a heat exchanger and provide useful quantitative information including (Heater-1) upstream of the high-temperature gas–water input/output mass and energy flows per 1 kg of produced shift reactor (GWS-1). The latter unit operation is fol- lowed by a cooling heat exchanger (Cooler-1) to lower Table 1: TRACI 2.1 midpoint impact categories the temperature of the gases before entering the low-temperature WSG-2. TRACI 2.1 midpoint impact category Units The gaseous stream leaving the WGS-2 is then cooled Acidification kg SO -equiv. in Cooler-2 before entering the gas separator, which Ecotoxicity CTUe separates the gases from condensed water. A pressure- Eutrophication kg N-equiv. swing absorber unit is used to separate the product H Global warming kg CO -equiv. from other combustible gases (a mixture of CO and some Human-health particulate matter kg PM2.5-equiv. H ) and CO . The heat generated from the combustion of Human toxicity, carcinogens CTUh 2 2 CO and remaining H is used to offset the heat required Human toxicity, non-carcinogens CTUh for the glycerol reformer. The stream of exhaust gases Ozone depletion kg CFC 11-equiv. Smog formation kg O -equiv. is then cooled (in Cooler-3) before it goes to the stack. CTUe, comparative toxic unit for ecotoxicity impacts (e.g. freshwater toxicity); CTUh, comparative toxic unit for human-toxicity impacts. CPA = Cubic-Plus-Association equation of state. Downloaded from https://academic.oup.com/ce/article/5/3/387/6317702 by DeepDyve user on 13 July 2021 Clean Energy | 393 ADJ-2 SPRDSHT-1 From To Reformer Heater-1 Valve Heater Duty APR Glycerol Pump Pump To Feed Glycerol-Water Out Reformer Mixing Heater Unit Water Pump Reform Electrical Energy Reformer Energy Bottoms ADJ-1 Heat Heat ADJ-2 Removed Added to To from WGS-2 Heater-1 Cooler-1 Cooler-1 Out Heat To To To WGS-2 E WGS-2 Removed Cooler-1 Heater-1 WGS-1 from WGS-2 WGS-1 Cooler-2 Heater-1 Energy Cooler-2 WGS-2 WGS-1 WGS-1 To Bottoms Energy Bottoms Separator H2 Gas (Product) ADJ-2 Heat To PSA Removed CO2 PSA Unit from Cooler-3 PSA Combustible Combustion off-Gases Hot Gases to Cooler-3 Combustion Gas-Liquid Stack Gases Separator Air Furnace Separated Water Furnace Furnace Duty Bottoms Fig. 2: Reforming process simulation of the aqueous-phase reforming technology for hydrogen production The three glycerol-reforming processes (SCWR, APR and CO + CH ⇌ 2H + 2CO (ΔH) at 25 C = 247.02 KJ/mole 2 4 2 R ATR) for H production involve the following primary and R4: Glycerol-oxidation reaction (unique to the ATR secondary stoichiometric reactions: process) R1: Glycerol reforming (primary reaction) C H O + 1.5O ⇌ 3CO + 4H (ΔH) at 25 C  =  –597.73  KJ/ 3 8 3 2 2 2 R C H O + 3H O ⇌ 7H + 3CO (ΔH) at 25 C = 127.75 KJ/mole mole 3 8 3 2 2 2 R o o In addition to the primary reforming reaction (R1), sec- High-temperature (350C) and low-temperature (200 C) ondary reactions may also occur (as described by R2 WGS reactors are included in the three reforming pro- and R3). cesses to convert CO (in the presence of HO) into CO 2 2 R2: Methanation reaction and H . CO + 3H ⇌ CH + H O (ΔH) at 25 C = –205.88 KJ/mole R5: WGS reaction 2 4 2 R R3: Methane dry reforming reaction CO + H O ⇌ H + CO (ΔH) at 25 C = –41.14 KJ/mole 2 2 2 R Note: Calculations of the reaction enthalpies at 25 C and 1 a.t.m. are performed using HSC Chemistry® soft- Note that CO that shows as a reactant in R2 could come from ware (version 10) [42]. the direct decomposition of CH O to CO and H . This decom- 3 8 3 2 Fig. 3 shows how Aspen HYSYS results are inte- position reaction is endothermic with (∆H) of ≈250 kJ/mole at grated with GaBi product system models to quantify 25 C and 1 a.t.m. [41]. Downloaded from https://academic.oup.com/ce/article/5/3/387/6317702 by DeepDyve user on 13 July 2021 394 | Clean Energy, 2021, Vol. 5, No. 3 Aspen HYSYS Glycerol reforming process simulation simulation for production platform of 1 FU of H2 Environmental burdens and adverse human health impacts (per 1 FU of H2): Acidification Ecotoxicity Eutrophication Calculated mass and energy flows & streams Global warming (emissions)chemical Human health, particulates compostions per Human toxicity, carcinogenics 1 FU of H2 Human toxicity, non- carcinogenics Ozone depletion Smog formation Construct life cycle inventory (LCI) for production of 1 FU of H2 Product GaBi LCA system platform model Fig. 3: Integration of the reforming process simulation with the LCA product system model to calculate the midpoint impact categories associated with glycerol reforming for hydrogen production Table 2: Primary life-cycle inventory data associated with the supercritical water reforming (SCWR) of glycerol for 1  kg of H production Mass and energy inputs Mass and energy outputs Glycerol feedstock: 13.3 kg H gas: 1 kg Water to reformer: 44.78 kg CO gas: 13.28 kg Air to combustible-gases furnace: 0.123 kg Water: 41.436 Water/glycerol mass ratio = 3.37 Stack emissions: 2.487 kg Reformer operation conditions: • CH gas: 1.286 kg • Temperature: 800C • CO gas: 0.695 kg • Pressure: 240 bar • Water vapour: 0.392 kg • N gas: 0.0943 kg • O gas: 0.0198 kg Total mass in: 58.203 kg Total mass out: 58.203 kg Thermal energy required = 218 MJ Thermal energy available for harvesting at 90% thermal Electrical energy required = 1.8 MJ efficiency = 139 MJ (to operate process pumps/compressors) Net thermal energy required is ≈81 MJ assuming credit is given to the thermal energy available for harvesting the environmental burdens and adverse human-health to generate the LCI primary data required as the input to impacts of each of the three glycerol-based hydrogen- GaBi product system models. The primary data for SCWR, production technologies. APR and ATR are provided in Tables 2, 3 and 4, respectively. For both Aspen HYSYS simulations and GaBi product system models, 1 kg of H produced has been selected as the FU. Hence, all mass and energy flows and mid-term 4 Results and discussion impact categories are calculated per 1 FU of produced H . The following subsections present results of Aspen HYSYS simulations and GaBi LCIA for the three selected 3.1.3 LCI primary data glycerol-reforming technologies for H production. Also The three glycerol-reforming technologies (SCWR, APR discussed is the interpretation of the generated results. and ATR) are simulated using the Aspen HYSYS platform Per ISO 14040:2006 [40], the interpretation step of the LCA Downloaded from https://academic.oup.com/ce/article/5/3/387/6317702 by DeepDyve user on 13 July 2021 Clean Energy | 395 framework requires that the assessment results should be selected bio-based H-production technologies; (ii) sensi- reported by the most informative means possible. In light tivity scenarios that quantify the effect on the calculated of this ISO reporting requirement, this research adopts midpoint impact categories as a result of variations in pro- the following approaches for interpretation of its LCIA nu- cess electricity sources (viz. US electricity-grid mix versus merical results, namely: (i) BC comparative assessments wind power) and thermal energy sources (viz. natural gas of TRACI 2.1 LCIA results per impact category across the versus biogas). Moreover, the distinctively high operating Table 3: Primary life-cycle inventory data associated with the aqueous-phase reforming (APR) of glycerol for 1 kg of H production Mass and energy inputs Mass and energy outputs Glycerol feedstock: 7.7 kg H gas: 1 kg Water to reformer: 69.3 kg CO gas: 955 kg Air to combustible-gases furnace: 0.123 kg Water: 65.4 Water/glycerol mass ratio = 9.0 Stack emissions: 1.173 kg Reformer operation conditions: • CO gas: 0.051 kg • Temperature: 290C • Water vapour: 0.972 kg • Pressure: 52.69 bar • N gas: 0.0943 kg • CH gas: 0.056 Total mass in: 77.123 kg Total mass out: 77.123 kg Thermal energy required = 216.67 MJ Thermal energy available for harvesting at 90% thermal Electrical energy required = 0.522 MJ efficiency = 201.08 MJ (to operate process pumps/compressors) Net thermal energy required is ≈16.11 MJ assuming credit is given to the thermal energy available for harvesting Table 4: Primary life-cycle inventory data associated with the autothermal reforming (ATR) of glycerol for 1 kg of H production Mass and energy inputs Mass and energy outputs Glycerol feedstock: 9.21 kg H gas: 1 kg Water to reformer: 4.33 kg CO gas: 8.81 kg Air to combustible-gases furnace: 5.53 kg Water: 2.07 Water/glycerol mass ratio = 0.47 Stack emissions: 7.19 kg Reformer operation conditions [43]: • CO gas: 0.0997 kg • Temperature: 650C • Water vapour: 0.0 kg • Pressure: 1.013 bar • N gas: 4.24 kg • CH gas: 2.85 Total mass in: 19.07 kg Total mass out: 19.07 kg Electrical energy required = 41.1 MJ (to operate process pumps/compressors) If credit is given to thermal-energy harvesting (at 90% efficiency) in the ATR process, it would be sufficient to offset the thermal energy required for the reformer Air to Burn Combustible Gases before Stack Gaseous Emissions (N2, CH4, ect.) to Stack Thermal Energy Source Glycerol Feedstock GaBi-Generared Product System Model for H2 Gas (FU = 1 kg) H2 Production via Glycerol Reforming Technology (SCWR) CO2 Gas Byproduct Water to Reformer U.S. Electricity Grid Mix U.S. Diesel-Powered Trucks Water from Process Gas/Liquid Separator Fig. 4: Product system model for H production using glycerol SCWR Input and output mass and energy data for the SCWR technology are presented in Table 2. Downloaded from https://academic.oup.com/ce/article/5/3/387/6317702 by DeepDyve user on 13 July 2021 396 | Clean Energy, 2021, Vol. 5, No. 3 pressure (240 bar) of the SCWR technology allowed for in the HYSYS simulations. More details about the energy- analysing the effect of in situ electricity generation to optimization strategies are discussed in Section 4.2. offset some of the required electricity on the LCIA impact The mass and energy flows calculated using Aspen categories. HYSYS to produce 1 kg H are provided in Tables 2, 3 and 4, respectively. Also, with respect to SCWR of glycerol, Khalil [35] estimated that ~79 MJ of energy would be required to 4.1 Aspen HYSYS simulations: BC and SS produce 1  kg of supercritical water (SCW). This energy is To meet the research objectives discussed in Section 2.2, consumed in heating the water to temperatures above its Aspen HYSYS models are developed to simulate the fol- critical point (374 C) and compressing the resulting steam lowing glycerol-reforming scenarios: (i) BC scenarios in to pressures above its critical pressure (220 bar). which the US electricity-grid mix is assumed to supply the electrical power to operate the pumps and compres- 4.2 Quantifications of GaBi LCA product system sors in the glycerol-reforming processes; also, natural- models for H production gas burning is assumed to supply the required reformer’s 2 thermal energy and for heating the reactants to the re- 4.2.1 SCWR for H production quired reforming temperature; (ii) SS in which wind power, Fig. 4 shows the input and output streams of the GaBi instead of the electricity-grid mix, is assumed to supply the product system model for the production of 1  kg of H electrical power for the process pumps and compressors, (i.e. 1 FU) using SCWR technology. Input and output mass and biogas burning, instead of natural gas, to supply the and energy data are presented in Table 2. The input mass thermal energy required for the reformer; (iii) optimiza- streams represent the glycerol and water feedstocks. The tion strategies to reduce the reforming energy require- input energy streams include the thermal energy required ments (both electrical and thermal) are also considered by the reformer and electricity (from the US grid mix) to Table 5: Midpoint impact categories associated with the supercritical water reforming (SCWR) of glycerol for 1 kg of H production a b TRACI 2.1 midpoint impact categories Scenario 1 Scenario 2 Impact reduction –2 –2 Acidification (kg SO -equiv.) 1.14 × 10 1.07 × 10 6.1% –2 –2 Ecotoxicity (CTUe) 6.37 × 10 5.55 × 10 12.9% –2 –2 Eutrophication (kg N-equiv.) 1.53 × 10 1.50 × 10 0.2% Global warming (kg CO-equiv.) 51.9 51.5 0.7% –4 –4 Human-health particulates (kg PM2.5-equiv.) 5.26 × 10 4.66 × 10 11.2% –9 –9 Human toxicity, carcinogenic (CTUh) 1.69 × 10 1.64 × 10 2.9% –8 –8 Human toxicity, non-carcinogenic (CTUh) 2.10 × 10 1.61 × 10 23.3% –11 –11 Ozone depletion (kg CFC 11-equiv.) 9.91 × 10 1.89 × 10 80.9% –1 –1 Smog (kg O -equiv.) 4.47 × 10 4.41 × 10 1.3% Scenario 1: This is the base case (BC) scenario with electricity generation from the US electricity-grid mix to power the process pumps and compressors. Scenario 2: This is a sensitivity scenario (SS) with in situ electricity generation from SCWR (using a turbine downstream of the reformer where the hot gases at 240 bar are allowed to expand in the turbine to generate electrical power) to supply the electric loads of the process pumps and compressors. Air to Burn Combustible Gases before Stack Gaseous Emissions (N2, CH4, ect.) to Stack Thermal Energy Source Glycerol Feedstock GaBi-Generared Product System Model for H2 Gas (FU = 1 kg) H2 Production via Glycerol Reforming Technology (APR) CO2 Gas Byproduct Water to Reformer U.S. Electricity Grid Mix U.S. Diesel-Powered Trucks Water from Process Gas/Liquid Separator Fig. 5: Product system model for H production using glycerol APR 2 Downloaded from https://academic.oup.com/ce/article/5/3/387/6317702 by DeepDyve user on 13 July 2021 Clean Energy | 397 provide the electrical power to the process pumps and required to operate the process pumps/compressors. The compressors. It is assumed that diesel-powered trucks are associated global-warming potential is 51.9  kg CO -equiv. used for the transportation of glycerol from the biodiesel for Scenario 1 and 51.5 kg CO-equiv. for Scenario 2. Clearly, plant to the reforming plant (distance is assumed to be 50 this energy-management strategy led to an ≈0.8% carbon- km). The output streams include H, gaseous emissions footprint reduction (namely 0.4  kg CO -equiv. per kg of 2 2 (N , CH , CO , etc.) and water removed by the gas/liquid produced H ). In both scenarios, the biogenic carbon rep- 2 4 2 2 separator. These unit operations were modelled in Aspen resents ≈27% of the associated global warming potential HYSYS simulations (as illustrated in Fig. 2). (GWP) (measured in kg CO -equiv. per kg of H produced). 2 2 The reformer is modelled by a conversion reactor (at As Table 3 shows, the highest impact category reduction 800 C and 240 bar) and the two WGS reactors (WGS-1 and is related to ozone depletion (≈81%) and the lowest im- WGS-2) are modelled by equilibrium reactors. WGS-1 oper - pact reduction is related to the eutrophication (≈0.2%) o o ates at 350C and 15 bar and WGS-2 operates at 200 C and impact category. The reduction in the human-toxicity 15 bar. (non-carcinogenic) impact category is ≈23%. Table 5 provides the nine TRACI 2.1 midpoint impact categories associated with the production of 1 FU (viz. 1 kg) 4.2.2 APR for H production of H using SCWR for Scenarios 1 and 2, respectively. Fig. 5 depicts the input and output streams of the GaBi The results summarized in Table 5 indicate some reduc- product system model for the production of 1 kg of H (i.e. tions in the burdens associated with human health and 1 FU) using APR technology. Input and output mass and en- the environment as a result of generating electrical power ergy data are presented in Table 5. The input mass streams (by expanding the hot exhaust gases from the reformer represent glycerol and water feedstocks. The input energy through an installed turbine) to offset the electric loads streams include thermal energy required by the reformer, Table 6: Midpoint impact categories associated with aqueous-phase reforming (APR) of glycerol for 1 kg of H production a b c TRACI 2.1 midpoint impact categories Scenario 3 Scenario 4 Scenario 5 –3 –3 –3 Acidification (kg SO -equiv.) 2.08 × 10 9.66 × 10 1.83 × 10 –2 –1 –2 Ecotoxicity (CTUe) 6.46 × 10 1.25 × 10 6.28 × 10 –4 –3 –4 Eutrophication (kg N-equiv.) 3.19 × 10 2.76 × 10 3.06 × 10 Global warming (kg CO-equiv.) 4.11 4.18 3.88 –4 –4 –4 Human-health particulates (kg PM2.5-equiv.) 2.74 × 10 5.00 × 10 2.55 × 10 –10 –9 –10 Human toxicity, carcinogenic (CTUh) 7.59 × 10 2.00 × 10 7.45 × 10 –8 –7 –8 Human toxicity, non-carcinogenic (CTUh) 1.76 × 10 3.93 × 10 1.65 × 10 –11 –11 –11 Ozone depletion (kg CFC 11-equiv.) 4.96 × 10 5.53 × 10 1.46 × 10 –1 –1 –1 Smog (kg O -equiv.) 1.03 × 10 1.56 × 10 1.01 × 10 Scenario 3: This is the base case scenario in which natural gas is used to provide the required thermal energy in the process. The US electricity-grid mix provides the required process electricity. Scenario 4: In this scenario, biogas is used instead of natural gas to provide the required thermal energy for the process. The US electricity-grid mix provides the required process electricity. Scenario 5: In this scenario, wind-turbine power is used instead of the electricity-grid mix as the source of electricity for operating the process compressors and pumps. Air to Burn Combustible Gases before Stack Gaseous Emissions (N2, CH4, ect.) to Stack Thermal Energy Source Glycerol Feedstock H2 Gas (FU = 1 kg) GaBi-Generared Product System Model for Water to Reformer H2 Production via Glycerol Reforming Technology (ATR) CO2 Gas Byproduct Air to ATR Reformer U.S. Electricity Grid Mix Water from Process Gas/Liquid Separator U.S. Diesel-Powered Trucks Fig. 6: Product system model for H production using glycerol ATR 2 Downloaded from https://academic.oup.com/ce/article/5/3/387/6317702 by DeepDyve user on 13 July 2021 398 | Clean Energy, 2021, Vol. 5, No. 3 Table 7: Midpoint impact categories associated with autothermal reforming (ATR) of glycerol for 1 kg of H production a b TRACI 2.1 midpoint impact categories Scenario 6 Scenario 7 Impact reduction –2 –4 Acidification (kg SO -equiv.) 2.14 × 10 7.17 × 10 96.6% –1 –2 Ecotoxicity (CTUe) 1.57 × 10 1.38 × 10 91.2% Eutrophication (kg N-equiv.) 4.18 4.18 0.0% Global warming (kg CO-equiv.) 87.2 80.3 7.9% –3 –5 Human-health particulates (kg PM2.5-equiv.) 1.59 × 10 8.73 × 10 94.5% –9 –10 Human toxicity, carcinogenic (CTUh) 1.19 × 10 1.08 × 10 90.9% –8 –8 Human toxicity, non-carcinogenic (CTUh) 9.77 × 10 1.09 × 10 88.8% –9 –12 Ozone depletion (kg CFC 11-equiv.) 2.75 × 10 2.39 × 10 99.9% –1 –1 Smog (kg O -equiv.) 3.01 × 10 1.22 × 10 59.5% Scenario 6: This is the base case (BC) scenario in which electricity from the US grid mix is used to power the process pumps and compressors. Scenario 7: Electricity from wind turbines (instead of electricity from the grid mix) is used to power the process pumps and compressors. 4.2.3 ATR for H production Table 8: Comparative primary energy consumption to pro- Fig. 6 shows the input and output streams of the GaBi duce 1 kg of H gas via glycerol reforming product system model for the production of 1 kg of H (i.e. Primary energy 1 FU) using ATR technology. Input and output mass and en- Glycerol-reforming consumption ergy data are presented in Table 6. The input mass streams technology (MJ/kg H ) represent glycerol and water feedstocks as well as air for Supercritical water reforming (SCWR): the ATR reformer. The input energy streams include the • Scenario 1, viz. (BC) 33.17 thermal energy required by the reformer, the US electri- SCWR • Sensitivity Scenario 2 (SS-2) 27.68 city grid to provide electrical power to process pumps and Aqueous-phase reforming (APR): compressors, and US diesel-powered trucks for the trans- • Scenario 3, viz. (BC) 23.76 APR portation of glycerol from the biodiesel plant to the re- • Sensitivity Scenario 4 (SS-4) 23.71 forming plant (assumed to be 50 km). The output streams • Sensitivity Scenario 5 (SS-5) 23.62 include H , gaseous emissions (N , CH , CO , etc.) and water 2 2 4 2 Autothermal reforming (ATR): removed to the gas/liquid separator. The ATR reformer is • Scenario 6, viz. (BC) 121.15 ART modelled by a conversion reactor (at 650 C and 1.013 bar) • Sensitivity Scenario 7 (SS-7) 104.43 and the two WGS reactors (WGS-1 and WGS-2) are mod- elled by equilibrium reactors. WGS-1 operates at 350 C and the US electricity grid to provide electrical power to the 15 bar and WGS-2 operates at 200 C and 15 bar. process pumps and compressors, and US diesel-powered Input and output mass and energy data for the ATR trucks for the transportation of glycerol from the biodiesel technology are presented in Table 6. plant to the reforming plant (assumed to be 50 km). The In the BC scenario, electrical power is provided from the output streams include H, gaseous emissions (N , CH , 2 2 4 US electricity-grid mix. In the sensitivity scenario, elec- CO , etc.) and water removed to the gas/liquid separator. trical power is assumed to be provided from wind power. The reformer is modelled by a conversion reactor (at Table 7 summarizes the GaBi LCIA results (midpoint im- 290 C and 52.69 bar) and the two WGS reactors (WGS-1 and pact categories associated with the production of 1 kg of H WGS-2) are modelled by equilibrium reactors. WGS-1 oper - using ATR technology). o o ates at 350C and 15 bar and WGS-2 operates at 200 C and As Table 7 shows, switching the electricity supply from 15 bar. Input and output mass and energy data for the APR the US grid mix to wind power resulted in a substantial technology are presented in Table 6. reduction in the adverse environmental impacts (with In BC Scenario 3, electrical power for the modelled pumps the exception of water eutrophication). Also, the carbon- and compressors is provided from the US electricity-grid footprint reduction is ≈8%. The notably higher value of the mix and the thermal energy for the reformer is provided by eutrophication potential compared to APR (see Table 5) and burning natural gas. In one sensitivity scenario, electrical SCWR (see Table 3) can be attributed to use of air in the power is provided from US wind power while natural gas ATR reformer and the resulting N emissions (as shown in is used to the provide the thermal energy for the reformer. Tables 2, 4 and 6, respectively). In the second sensitivity scenario, electrical power is pro- vided from the US electricity-grid mix and biogas burning (instead of natural gas) is used to provide the thermal en- 4.3 Environmental sustainability assessment ergy for the reformer. Inspection of the results depicted in Table 6 shows that, In alignment with ISO 14040:2006 [40] reporting require- with the exception of GWP, the environmental impacts de- ments for the LCA results interpretation step, Sections crease in the following order from highest to lowest: 4.2.1, 4.2.2 and 4.2.3 presented the LCIA results for SCWR, Scenario 4 > Scenario 3 > Scenario 5 APR and ATR technologies and compared their associated Downloaded from https://academic.oup.com/ce/article/5/3/387/6317702 by DeepDyve user on 13 July 2021 Clean Energy | 399 1.0E+01 (BC)_ART (BC)_SCWR (BC)_APR 1.0E+00 1.0E–01 1.0E–02 1.0E–03 1.0E–04 Acidification Ecotoxicity Eutrophication Human Health Smog [kg SO2-Equiv.] [CTUe] [kg N-Equiv.] Particulates [kg O3-Equiv.] [kg PM2.5-Equiv.] Fig. 7: Comparative midpoint impact categories of three base case (BC) scenarios of glycerol reforming via SCWR, APR and ATR processes BC scenarios versus sensitivity scenarios with respect to Other quantitative insights from Table 8 (three BC environmental and human-health impacts. Section 4.3 scenarios and four SS): shows how the aforementioned LCIA results (viz. sustain- • For SCWR: in situ electricity generation (via expanding ability indicators) of SCWR, APR and ATR compare with the high-pressure product gases from the reformer) to each other in order to identify which glycerol-reforming offset the use of electricity from the grid mix resulted in technology can be considered the most sustainable. The an ≈17% reduction in primary energy consumption per indicators of this environmental sustainability assessment kg of produced H . are: primary energy consumption and environmental and • For ATR: use of electricity from wind power instead of human-health impacts (all evaluated per 1 kg of produced electricity from the US grid mix resulted in an ≈14% re- H ). Table 8 summarizes the primary energy consumed to duction in the primary energy consumption per kg of produce 1 kg of H for the BC scenarios as well as the SS. produced H . As Table 8 shows, the primary energy consump- • For APR: replacement of natural gas with biogas as the tions (measured in MJ/kg of produced H) of the three source of thermal energy for the reforming process BC scenarios can be ordered (from highest to lowest) as led to an ≈0.21% reduction in primary energy con- follows: sumption compared to the BC scenario in which nat- ATR > SCWR > APR ural gas (NG) is the source of thermal energy for the Accordingly, APR represents the lowest energy- reformer. Also, the use of wind power instead of grid intensive glycerol-reforming technology compared to ATR mix to provide the electricity required to power the and SCWR. As shown by the three BC scenarios in Table compressors and pumps in the APR process led to an 8, the energy intensity (MJ/kg H) of APR is ≈20% of that ≈0.60% reduction in the primary energy consumption associated with ATR and ≈72% of that associated with per kg of produced H . SCWR. Also, it should be noted that the ATR process has three feed streams entering the reformer (glycerol, water Figs 7–9 compare the nine TRACI 2.1 midpoint impact and air), which have to be heated to the reformer operating categories of the three BC scenarios associated with the temperature of 650C. Hence, as Table 8 shows, there is SCWR, APR and ATR processes for glycerol-based H pro- more primary energy consumption (in MJ/kg H ) in the ATR duction. The impact categories are divided among Figs 7–9 process compared to the SCWR and APR processes. for visual clarity purposes (notice the y-axis scale differ - As a benchmark, Spath and Mann [45] reported 183.2 ences). The general trends in the three figures show that MJ/kg H for the primary energy consumption of SMR the impact categories are highest for ATR, lowest for APR, technology. with SCWR somewhere in between. Value of Mid-Point Impact Category Downloaded from https://academic.oup.com/ce/article/5/3/387/6317702 by DeepDyve user on 13 July 2021 400 | Clean Energy, 2021, Vol. 5, No. 3 1.0E–07 (BC)_ART (BC)_SCWR (BC)_APR 1.0E–08 1.0E–09 1.0E–10 1.0E–11 Human toxicity, carcinogenics Human toxicity, non-carcinogenics Ozone Depletion [CTUh] [CTUh] [kg CFC 11-Equiv.] Fig. 8: Comparative midpoint impact categories of base case (BC) scenarios of glycerol reforming via SCWR, APR and ATR processes (continued from Fig. 7) 1.0E+02 (BC)_ART (BC)_SCWR (BC)_APR 9.0E+01 8.0E+01 7.0E+01 6.0E+01 5.0E+01 4.0E+01 3.0E+01 2.0E+01 1.0E+01 0.0E+00 (BC)_ART (BC)_SCWR (BC)_APR Fig. 9: Comparative midpoint impact categories of base case (BC) scenarios of glycerol reforming via SCWR, APR and ATR processes (continued from Fig. 7) A closer inspection of Fig. 7 shows that the APR process A closer inspection of Fig. 8 shows that the APR process leads to the highest reduction in the eutrophication cat- leads to the highest reduction in ozone depletion (followed egory (followed by SCWR) compared to ATR. Quantitatively, by SCWR) compared to ATR. the eutrophication impact category associated with APR Inspection of Fig. 9 for BC scenarios shows that the APR technology is <0.01% of that associated with ATR tech- technology reduces CO emissions by ≈95% per kg of pro- nology and ≈2% of that associated with SCWR technology. duced H compared to ATR and by ≈92% per kg H com- 2 2 The notably high eutrophication category associated with pared to SCWR. As a benchmark, CO emission from SMR is ATR can be attributed primarily to the air used in the re- 11.9 kg CO -equiv./kg H [45] compared to 4.1 kg CO equiv./ 2 2 2 former and the resulting N emission. kg of H for the APR technology. However, the CO emission 2 2 2 Value of Mid-Point Impact Category Global Warming [kg CO2-Equiv.] Downloaded from https://academic.oup.com/ce/article/5/3/387/6317702 by DeepDyve user on 13 July 2021 Clean Energy | 401 from SMR remains below that of the ATR and SCWR tech- HYSYS process simulations with the GaBi LCA platform. nologies (which are 87.2 and 51.9  kg CO equiv./kg H , The sustainability indicators used in this assessment 2 2 respectively). conform to the TRACI 2.1 LCIA methodology and include Additional quantitative LCIA-based insights: the primary energy consumption and midpoint impact categories (viz. human-health and environmental im- • SCWR sensitivity Scenario 2 shows a reduction of pacts) per kg of H. ≈0.4  kg CO -equiv./kg H compared to SCWR Scenario 2 2 The results of our study revealed that H production from 1 (BC) as a result of using in situ generated electricity glycerol via APR is more environmentally sustainable than compared to using electricity from the US grid mix glycerol ATR, SCWR and conventional SMR technology. For to power the process pumps and compressors in the example, the primary energy consumption (in MJ/kg H ) can SCWR process. be ranked from highest to lowest in the following order of • APR sensitivity Scenario 4 shows a carbon-footprint re- H production technologies: SMR > ATR > SCWR > APR. The duction of ≈0.7 kg CO -equiv./kg H as a result of using 2 2 associated CO emissions (in kg CO -equiv./kg H ) can also 2 2 2 wind power (a renewable energy source) instead of the be ranked from highest to lowest in the following order of US electricity-grid mix (BC Scenario 3) to power the pro- H -production technologies: ATR > SCWR > SMR > APR. cess pumps and compressors in the APR system. Future research should focus on expanding the scope • ATR sensitivity Scenario 7 shows a reduction in the of this comparative environmental assessment to include carbon footprint of ≈6.9  kg CO-equiv./kg H as a re- 2 2 other H -production routes such as water electrolysis using sult of using wind power, instead of grid electricity, to electricity from wind turbines and solar photovoltaic, as power the process pumps and compressors in the ATR well as the nuclear-powered and solar-powered thermo- process. chemical water-splitting pathways. It is this author’s opinion that bio-based H production via glycerol APR technology is a promising alternative to fossil- based H production using SMR for the following reasons: Acknowledgements The author of this research greatly appreciates the reviews and • SMR technology consumes natural gas (a depletable re- constructive comments provided by several experts in sustain- source) and emits ≈11.9 kg CO for every kg of produced ability and LCA from the University of Oxford in the UK and from hydrogen. It should be noted that this CO emission from Harvard University in the USA. The literature materials used in SMR is fossil-based as opposed to being biogenic CO. 2 this research were made available through the Bodleian Libraries • SMR is an energy-intensive technology and would be of University of Oxford and Tate library at Harris Manchester even more energy-intensive if the SMR plant is in- College (HMC) in Oxford during this author’s tenure as a Technical Fellow at HMC in the UK. tegrated with carbon-capture and sequestration processes. • Glycerol feedstock in APR is a by-product of biomass Conflict of Interest transesterification to produce biodiesel. Also, glycerol APR technology emits ≈4.11 kg CO per kg of produced H None declared. 2 2 compared to ≈11.9 kg CO based on the SMR technology. • The carbon footprint of the APR technology can be fur - References ther reduced by integrating the bio-based hydrogen [1] DOE/EERE. Fuel Cell Technologies Market Report 2016. https:// product plant with the biodiesel plant to eliminate GHG www.energy.gov/sites/default/files/2017/10/f37/fcto_2016_ emissions associated with glycerol transport. market_report.pdf (7 May 2021, date last accessed). [2] DOE/EERE. Department of Energy (DOE) Office of Energy Efficiency and Renewable Energy (EERE), FY18 Hydrogen and Fuel 5 Conclusions Cell R&D Funding Opportunity Announcement (FOA) Number: DE-FOA-0001874. https://www.energy.gov/sites/default/ The future of hydrogen as an energy carrier is important, files/2018/08/f54/fcto-fy18-foa-selections_0.pdf (7 May 2021, but so too is the ability to quantify the energy require- date last accessed). ments and environmental burdens associated with H - [3] EPA. Fast facts on transportation greenhouse gas emissions. Source: production technologies. In this regard, LCA is a central Fast facts on transportation greenhouse gas emissions | green framework for performing comparative environmental vehicle guide | US EPA https://www.epa.gov/greenvehicles/ sustainability analyses of H -production technologies. fast-facts-transportation-greenhouse-gas-emissions (7 May Production of H from bio-based feedstocks such as gly- 2021, date last accessed). [4] OECD/IEA. World Energy Outlook 2016. Paris: Organization for cerol (a by-product of biodiesel production via biomass Economic Co-operation and Development & International transesterification) is considered a carbon-neutral tech- Energy Agency (IEA), 2016. nology because its associated CO emission is assumed to [5] Neuling  U, Kaltschmitt  M. Techno-economic and environ- be absorbed during the growth phase of the plant. mental analysis of aviation biofuels. Fuel Processing Technology, In this study, we carried out an environmental sustain- 2018, 171:54-69. ability assessment of glycerol reforming for H production [6] Airbus. Global Market Forecast: Mapping Demand 2016/2035. using a quantitative methodology that integrated Aspen Toulouse: Airbus, 2016. Downloaded from https://academic.oup.com/ce/article/5/3/387/6317702 by DeepDyve user on 13 July 2021 402 | Clean Energy, 2021, Vol. 5, No. 3 [7] Boeing. Current Market Outlook 2016–2035. Chicago, IL: Boeing, [28] Larimi AS, Kazemeini M, Khorasheh F. Highly selective doped Pt/ 2016. MgO nano-sheets for renewable hydrogen production from APR [8] Khalil  YF. Eco-efficient lightweight carbon-fiber reinforced of glycerol. Int. Journal of Hydrogen Energy, 2016, 41:17390-17398. polymer for environmentally greener commercial aviation in- [29] Boga DA, Liu F, Bruijnincx PC, et al. Aqueous-phase reforming dustry. Sustainable Production and Consumption, 2017, 12:16–26. of crude glycerol: effect of impurities on hydrogen produc- [9] EPA. Technical Support Document for Hydrogen Production: Proposed tion. Catalysis Science & Technology, 2016, 6(1):134-143. Rule for Mandatory Reporting of Greenhouse Gases. Washington, [30] Veluturla S, Archna N, Subba Rao D et , al. Catalytic valoriza- DC: US Environmental Protection Agency (EPA), Office of Air & tion of raw glycerol derived from biodiesel: a review. Biofuels, Radiation, 2008. 2018, 9(3):305-314. [10] Hoelzen J, Liu Y, Bensmann B, et al. Conceptual design of op- [31] Putra RD, Trajano HL, Liu S, et  al. In-situ glycerol aqueous eration strategies for hybrid electric aircraft. Energies, 2018, phase reforming and phenol hydrogenation over Raney Ni®. 11(1):217. Chemical Engineering Journal, 2018, 350:181-191. [11] European Commission, EC. Flightpath 2050—Europe’s Vision for [32] Ghani AA, Torabi F, Ibrahim H. Autothermal reforming process Aviation. Luxembourg City: European Commission, 2011. for efficient hydrogen production from crude glycerol using [12] Rocha  MH, Capaz  RS, Lora  EE, et  al. Life cycle assessment nickel supported catalyst: parametric and statistical analyses. (LCA) for biofuels in Brazilian conditions: a meta-analysis. Energy, 2018, 144:129-145. Renewable and Sustainable Energy Reviews, 2014, 37:435-459. [33] Bastan F, Kazemeini M, Larimi A, et al. Production of renewable [13] Zaimes GG, Vora N, Chopra SS,et al. Design of sustainable bio- hydrogen through aqueous-phase reforming of glycerol over fuel processes and supply chains: challenges and opportun- Ni/Al O MgO nano-catalyst. International Journal of Hydrogen 2 3 ities. Processes, 2015, 3(3):634-663. Energy, 2018, 43(2):614-621. [14] Galera S Gutiérrez Ortiz FJ. Life cycle assessment of hydrogen [34] Charisiou  ND, Papageridis  KN, Siakavelas  G, et  al. The influ- and power production by supercritical water reforming of gly- ence of SiO doping on the Ni/ZrO supported catalyst for 2 2 cerol. Energy Conversion and Management, 2015, 96:637-645. hydrogen production through the glycerol steam reforming [15] Thompson  JC, He  B. Characterization of crude glycerol from reaction. Catalysis Today, 2019, 319:206-219. biodiesel production from multiple feedstocks. Applied [35] Khalil YF. Sustainability assessment of solvolysis using super - Engineering in Agriculture, 2006, 22:261-265. critical fluids for carbon fiber reinforced polymers waste man- [16] Subramanian  ND, Callison  J, Catlow  CR, et  al. Optimised agement. Sustainable Production and Consumption, 2019, 17:74-84. hydrogen production by aqueous phase reforming of gly- [36] European Commission, Joint Research Centre, Institute for cerol on Pt/Al O . International Journal of Hydrogen Energy, 2016, Environment and Sustainability. ILCD Handbook: Analysing of 2 3 41(41):18441-18450. Existing Environmental Impact Assessment Methodologies for Use [17] Ciriminna  R, Pina  CD, Rossi  M, et  al. Understanding the gly- in Life Cycle Assessment. 1st edn. Ispra: European Commission, cerol market.Eur opean Journal of Lipid Science and Technology, Joint Research Centre, Institute for Environment and 2014, 116(10):1432–1439. Sustainability, 2010. [18] Statista. Leading Biodiesel Producers Worldwide in 2019, [37] Dhaliwal H. Life Cycle Impact Assessment, 2015. https://www. by Country. https://www.statista.com/statistics/271472/ earthshiftglobal.com/docs/impact-assessment-methods- biodiesel-production-in-selected-countries/ (7 May 2021, date free-lca-class_esg-sept2015.pdf (7 May 2021, date last last accessed). accessed). [19] Schwengber CA, Alves HJ, Schaffner RA, et  al. Overview of [38] Bare J. Tool for the Reduction and Assessment of Chemical glycerol reforming for hydrogen production. Renewable and and Other Environmental Impacts (TRACI). USER’S MANUAL. Sustainable Energy Reviews, 2016, 58:259-266. EPA/600/R-12/554. Cincinnati, OH: National Risk Management [20] DOE/EIA. Electricity in the United States, 2019. https://www.eia. Research Laboratory, Office of Research and Development, gov/energyexplained/electricity/electricity-in-the-us.php (7 U.S. Environmental Protection Agency, 2012, 1–24. May 2021, date last accessed). [39] Kaur J, Sarma AK, Jha MK, et al. Valorisation of crude glycerol [21] Wen G, Xu Y, Ma H, et al. Production of hydrogen by aqueous- to value-added products: perspectives of process technology, phase reforming of glycerol. International Journal of Hydrogen economics and environmental issues. Biotechnol Rep (Amst), Energy, 2008, 33(22):6657-6666. 2020, 27:e00487. [22] Authayanun S, Arpornwichanop A, Paengjuntuek W,et  al. [40] ISO. ISO 14040:2006. Environmental Management—Life Cycle Thermodynamic study of hydrogen production from crude Assessment—Principles and Framework. https://www.iso.org/ glycerol autothermal reforming for fuel cell applications. standard/37456.html (7 May 2021, date last accessed). International Journal of Hydrogen Energy, 2010, 35(13):6617-6623. [41] Almeida PM. Hydrogen production from liquid glycerol reform [23] Ortiz FG, Ollero P, Serrera A, et al. Thermodynamic study of the based on perovskite catalysts. Master’s thesis. Salvador: Federal supercritical water reforming of glycerol. International Journal University of Bahia, 2011. of Hydrogen Energy, 2011, 36(15):8994-9013. [42] Metso:Outotec. HSC Chemistry® Software. https://www. [24] Zhang L, Karim AM, Engelhard MH, et  al. Correlation of Pt–Re mogroup.com/portfolio/hsc-chemistry/?r=2 (7 May 2021, date surface properties with reaction pathways for the aqueous- last accessed). phase reforming of glycerol. Journal of Catalysis, 2012, 287:37-43. [43] Liu Y, Farrauto R, Lawal A. Autothermal reforming of glycerol [25] Ortiz FJ, Ollero P, Serrera A, et  al. Optimization of power and in a dual layer monolith catalyst. Chemical Engineering Science, hydrogen production from glycerol by supercritical water re- 2013, 89:31-39. forming. Chemical Engineering Journal, 2013, 15(218):309-318. [44] Shejale  AD, Yadav  GD. Ni–Cu and Ni–Co supported on La–Mg [26] Avasthi KS, Reddy RN, Patel S. Challenges in the production of based metal oxides prepared by coprecipitation and impregnation hydrogen from glycerol–a biodiesel byproduct via steam re- for superior hydrogen production via steam reforming of glycerol. forming process. Procedia Engineering, 2013, 51:423-9. Industrial & Engineering Chemistry Research, 2018, 57:4785-4797. [27] Rahman MM. H production from aqueous-phase reforming of [45] Spath PL, Mann MK.Life Cycle Assessment of Hydrogen Production glycerol over Cu-Ni bimetallic catalysts supported on carbon via Natural Gas Steam Reforming. NREL/TP-570–27637. Golden, nanotubes. Int. Journal of Hydrogen Energy, 2015, 40:14833-14844. CO: National Renewable Energy Laboratory, 2001, 1–33. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Clean Energy Oxford University Press

Simulation-based environmental-impact assessment of glycerol-to-hydrogen conversion technologies

Clean Energy , Volume 5 (3) – Sep 1, 2021

Loading next page...
 
/lp/oxford-university-press/simulation-based-environmental-impact-assessment-of-glycerol-to-0XZ2IZvUhD
Publisher
Oxford University Press
Copyright
Copyright © 2021 National Institute of Clean-and-Low-Carbon Energy
ISSN
2515-4230
eISSN
2515-396X
DOI
10.1093/ce/zkab018
Publisher site
See Article on Publisher Site

Abstract

Environmental Impact Glycerol Reforming Assessment Technologies Aqueous-Phase TRACI Mid-Point Impact Categories Reforming (APR) Acidification Human toxicity, carcinogens Ecotoxicity Glycerol from Auto-Thermal Eutrophication Human toxicity, Bio-Diesel non-carcinogens Reforming (ART) Production Global warming Ozone depletion Human health particulate matter Smog formation Supercritical Water Reforming (SCWR) Aspen HYSYS Simulation GaBi Platform for Life of Reforming Technologies Cycle Assessment (LCA) Keywords: glycerol; hydrogen; impact categories; reforming; supercritical water; valorization Need for greener H production to curb CO emissions Introduction 2 2 According to the US Environmental Protection Agency Background [3], the transportation sector was responsible for ≈27% of Hydrogen utilization GHG emissions in 2015 and, within this sector, commer - Currently, hydrogen (H) is mainly used in industrial cial aircraft were responsible for ≈9% of GHG emissions in processes rather than for energy production. The main 2015 [3]. The US Environmental Protection Agency (EPA) industrial uses of H include ammonia production (≈54% reported that light-duty internal combustion engine (ICE) of the overall H consumption) and oil refineries (≈35% vehicles were responsible for 60% of the GHG emissions at- of H consumption). Other uses including chemicals syn- tributed to the US transportation sector in 2015. As a result, thesis (e.g. methanol production) and the food industry hydrogen has been sought as a promising energy carrier are responsible for the rest of the H consumption [1]. for road transport. Direct H use to replace petrol-based As an energy carrier, H is used for the transport sector fuel in ICE vehicles or as a feedstock for on-board PEM fuel and as a feedstock to proton exchange membrane (PEM) cells can contribute to curbing GHG emissions from this fuel cells for electricity production. For the latter appli- sector. Also,in situ renewable H production in refuelling cation, in 2016, ≈62 000 fuel-cell systems (equivalent to stations has been pursued to eliminate GHG emissions as- 500 MW of power) were shipped worldwide [2]. Globally, sociated with H transport from production sources to re- 7.2 EJ of energy from H was consumed in 2013, which fuelling stations [2]. represents ≈1.3% of the world primary energy consump- The UN Intergovernmental Panel on Climate Change tion [1]. (IPCC) reported that the civil aviation industry currently contributes ≈2.5% of the world’s manmade CO emissions [4, 5]. Over the past three decades, the annual growth of Hydrogen-production sources: fossil-based and water world commercial air transport has averaged ≈5% and is electrolysis projected to double over the next decade or two [6–8]. As the Currently, H-production technologies are mainly fossil- civil aviation fleet grows to meet increasing air-transport based, which leads to significant CO emissions. For ex- demands, the IPCC predicts that the contribution of the ample, ≈48% of H production comes from steam-methane aviation industry to global manmade CO emissions will reforming (SMR), 30% is derived from petroleum refining increase to ≈3% in 2050 [9]. To curb GHG emissions associ- and 18% is produced by coal gasification. In the USA only, ated with the civilian aviation sector, there is an emerging ≈10 million tons of annual H production comes from SMR interest in the development of regional hybrid electric technology. Due to its fossil origin, H production causes aircraft (HEA) powered by combinations of conventional ≈60 million metric tons of CO emissions annually, which gas turbines and electric propulsion systems, comprising accounts for ≈2% of the energy-related CO emissions [3]. Water electrolysis, which enables H production without CO emissions, accounts for the remaining 4% of this Aircraft with 30–90 PAX (aircraft passengers), 4–15 tons of gross greenhouse-gas (GHG) emission [1]. take-off weight and energy of 30 000–70 000 kWh. Downloaded from https://academic.oup.com/ce/article/5/3/387/6317702 by DeepDyve user on 13 July 2021 Clean Energy | 389 motors powered by either batteries or PEM fuel cells [10], during biomass cultivation, pretreatment, oil extraction with the latter technology requiring H storage on board and refining, the transesterification process and glycerol- aircraft. In addition to the emerging HEA technology, the purification steps, CO is also absorbed during biomass full-electric aircraft technology remains a long-term vision farming, hence reducing the net emissions of CO. of aircraft manufacturers like Airbus and Boeing. It is also worth mentioning that members of the EU aviation organ- 1.2 Objective izations and industry counterparts created the ‘Flightpath 2050’ vision to reduce global CO , nitrogen oxides (NO ) and This research aims to perform comparative assessments of 2 x noise emissions such that, by 2050, civil aircraft should energy consumption, environmental burdens and adverse pollute 75% less CO, 90% less NO and 65% less noise [11]. human-health impacts of glycerol-based H production via 2 x The aforementioned background information explains the three reforming technologies, viz. supercritical water the currently observed global efforts to pursue H pro- reforming (SCWR), APR and autothermal reforming (ATR), duction from sustainable sources to curb GHG emissions, respectively. To achieve this objective, life-cycle impact as- other gaseous air pollutants and particulate matter, all of sessment (LCIA) is performed for base case (BC) scenarios which have adverse impacts on human health and the en- in which the US electricity-grid mix is assumed to be the vironment. In particular, bio-based H production via re- electricity supply source to power pumps and compressors forming technologies seems to offer a viable alternative in the reforming processes. The US grid electricity mix [20] fuel compared to the use of petroleum-based fuels (viz. is as follows: 39.5% natural gas, 23% coal, 20% nuclear and gasoline and diesel fuel) in the road transport sector and 17.5% renewables (of which 7.2% is wind power, 6.4% hydro- jet fuels in the aviation sector. power, 1.7% solar, 1.3% biomass and 0.4% geothermal). Also, natural-gas burning is assumed to supply the thermal en- ergy required for the reformer and for heating the reactant 1 Motivation, objectives and novelty of mixtures to the required reforming temperature. LCIA is research also performed for sensitivity scenarios in which wind power is assumed to be the electricity source instead of the 1.1 Motivation electricity-grid mix and biogas burning instead of natural The discussion provided in the subsection on the ‘Need gas to supply the thermal energy of reformers. Additionally, for greener H production to curb CO emission’ above to- 2 2 optimization strategies to reduce reforming energy require- gether with the following rationale for selecting glycerol as ments (both electrical and thermal) are also considered in the bio-based feedstock for greener H production are the the LCIA assessment scenarios. main motivators of this research: • Glycerol is a by-product of biodiesel production via the 1.3 Novelty transesterification of agricultural crops like soybeans, Unlike previous investigations that focused only on rapeseed, sunflower, jatropha, palm, castor, etc. [12–14]. estimating the carbon footprint associated with a single In this chemical process, ≈10 m of glycerol is produced glycerol-reforming technology, this research addresses per 90 m of rapeseed-derived biodiesel produced [15, current voids in the state of knowledge about the three 16]. In this research, we calculated ≈7.5 m of glycerol selected glycerol-reforming technologies by providing the per 90 m of soybean-derived biodiesel. According to following quantitative insights: Ciriminna et al. [17], the transesterification of vegetable oils for biodiesel production leads to ≈10 kg of glycerol by-product per 100 kg of biodiesel produced. • comparative energy consumption and midpoint impact • US biodiesel annual production is ≈5.5 million m /year. categories among these reforming technologies; In 2016, for example, the USA was the world’s biggest • efficient energy-management strategies for these biodiesel producer [18]. reforming technologies. These strategies include: (i) • EU biodiesel annual production is ≈12.5 million m /year capturing exothermic heat from the reforming reac- [19]. tion steps, (ii) burning exhaust combustible gases to • The increased production of biodiesel since 2000 to the offset the endothermal heat of the reformer and (iii) for present in both the USA and the EU led to a substantial SCWR technology, which is energy-intensive due to the reduction in glycerol prices. high-pressure requirement, the electricity-generation option in conjunction with H production has been In this author’s opinion, glycerol-reforming technolo- evaluated. gies offer a promising bio-based H source, in particular aqueous-phase reforming (APR), which could ultimately The remaining sections of this manuscript are organized replace the current use of SMR technology, which emits as follows. Section 2 discusses the key findings of the rele- ~11  kg CO -equiv. per kg of H produced [14]. In this re- vant published work, Section 3 presents the adopted re- 2 2 gard, CO emission associated with glycerol feedstock is search methods and tools, and Section 4 discusses the expected to be low because, while this GHG is emitted results of this environmental sustainability assessment. Downloaded from https://academic.oup.com/ce/article/5/3/387/6317702 by DeepDyve user on 13 July 2021 390 | Clean Energy, 2021, Vol. 5, No. 3 The principal results and major conclusions of this re- estimated that ≈3.77 kg CO -eqiv/kg H was produced after 2 2 search are summarized in Section 5. taking credit for their assumption of biogenic versus fossil- based CO emissions. Rahman [27] investigated 1 wt% glycerol (in DI water) APR over a series of nickel (Ni) and 2 Literature review copper–nickel (Cu–Ni) bimetallic catalysts supported on multiwalled carbon nanotubes (MWNT) at 240 C and 40 at- This research focuses on bio-based greener H production mospheres. His test results showed that the Cu–Ni catalyst via glycerol-reforming technologies and, hence, the scope led to 86% H selectivity and 84% glycerol conversion, and of this literature review is limited to achieving this purpose. noted that the presence of Cu in bimetallic catalysts led The three glycerol-based technologies being considered to the suppression of undesirable methanation reactions. herein are: SCWR, APR and ATR. Glycerol is a by-product Subramanian et  al. [16] studied glycerol APR over a series of biodiesel production via the transesterification of agri- of γ-Al O -supported metal nanoparticle catalysts for H 2 3 2 cultural crops. (More details and statistical data about this production in a batch reactor. They found that Pt/Al O 2 3 feedstock are provided in Section 1.1). was the most active catalyst under the test conditions. The remainder of this subsection provides brief high- The optimum conditions for H production were found to lights of key published research on H production via gly- be 240 C, 42 bar, 1000 r.p.m. stirring speed and ≥4100 sub- cerol reforming. Wen et  al. [21] studied the activities and strate/metal molar ratio for a 10-wt% glycerol feed. stabilities of Pt, Ni, Co and Cu catalysts and their sup- Schwengber et al. [19] published a review paper on gly- port substrates for H production via APR of glycerol. cerol reforming and highlighted key characteristics of the Authayanun et  al. [22] investigated the thermodynamics reported technologies. Larimi et al. [28] synthesized a series of H production via ATR of crude glycerol derived from of M-doped Pt/MgO (where M could be palladium, ruthe- the biodiesel-production processes. They assumed gly- nium, rhodium, iridium or chromium) nano-catalysts and cerol and methanol to be the primary components of the examined their performance effects on glycerol APR. Boga crude glycerol and found that H production increases as et al. [29] investigated the effect of individual components the ratio of glycerol to methanol increases and as the re- of crude glycerol on APR activity over 1 wt% Pt/Mg(Al)O, 1 forming temperature increases. Ortiz et al. [23] performed wt% Pt/AlO , 5 wt% Pt/AlO and 5 wt% Pt/C catalysts at 2 3 2 3 thermodynamic analysis of the glycerol SCWR process 29 bar and 225°C. The use of a 10-wt% alkaline crude gly- TM using AspenPlus and used the Soave–Redlich–Kwong cerol solution in water, containing 6.85 wt% glycerol, 1.62 (SRK) equation of state to identify the thermodynamically wt% soaps, 1.55 wt% methanol and 0.07 wt% ester, caused favourable operating conditions for glycerol conversion to a noticeable drop in APR activity compared to the corres- H . They used from 1 to 16  mol % (≈5–50 wt%) of glycerol ponding 6.85-wt% solution of pure glycerol in water. concentration in the feed stream, with water being the Most recently, Veluturla et  al. [30] discussed the use of balance, and conducted SCWR at 240 bar and 800 C. Zhang different types of heterogeneous catalysts for glycerol val- et al. [24] provided characterization of the Pt–Re catalysts orization through processes like carboxylation, conden- under conditions encountered during APR to obtain an sation, esterification, etherification, phase reformation, improved understanding of the role of the Rhenium (Re) selective oxidation, hydrolysis transesterification, se- catalyst. Their results showed the importance of sur - lective reduction and pyrolysis. Putra et al. [31] performed face acidity in controlling the reaction pathways during glycerol APR and phenol hydrogenation with Raney Ni at glycerol APR. Ortiz et  al. [25] simulated glycerol SCWR to 180–240°C for phenol-to-glycerol ratios from 0 to 4.88. They produce H and electricity. Avasthi et al. [26] focused their selected phenol as a model lignin monomer to demon- research on the glycerol/steam reforming (GSR) technology strate the feasibility of using renewable H from glycerol and discussed current challenges (e.g. higher prices of bio- APR for lignin upgrading. Ghani et  al. [32] reformed crude diesel compared to diesel fuel) and potential solutions. In glycerol over modified cerium–zirconium (Cs–Zr) supports this author’s opinion, none of the proposed solutions (e.g. loaded with 5  wt% nickel catalyst to generate H via ATR in situ removal of CO and H as soon as they form inside 2 2 (a combination of partial oxidation and steam reforming). the reformer) seems to be practical. They studied the effects of the reforming temperature, Rocha et al. [12] used a cradle-to-gate (C2G) life-cycle as- steam-to-carbon ratio (S/C), oxygen-to-carbon ratio (O/C), sessment (LCA) approach to quantify the environmental reduction temperature and calcination temperature, and and health impacts of the production of biodiesel from reported the following optimum operating conditions: a soybean and palm oil. Their analysis was performed using reforming temperature of 550C, S/C of 2.6, O/C of 0.5, re- SimaPro software and CML 2000 LCIA methodology. Galera duction temperature of 600 C and calcination temperature and Ortiz [14] evaluated the environmental perform- of 550 C. Bastan et al. [33] investigated the effects of the Al/ ance of H and electricity production via glycerol SCWR. Mg ratio in a series of Ni nano-catalysts supported on Al O 2 3 They used SimaPro 8.0. CML 2000 in their LCA study and and MgO on the physico-chemical characteristics of Ni/ Al O –MgO catalysts and determined the optimum cata- 2 3 Glycerol vaporization heat is 58.2 kJ/mol and its boiling point at o lyst for H production via glycerol APR. They found that the 1 a.t.m. is 290 C. 2 APR activity of synthesized catalysts strongly depended Platinum–rhenium catalyst. Downloaded from https://academic.oup.com/ce/article/5/3/387/6317702 by DeepDyve user on 13 July 2021 Clean Energy | 391 on the Al/Mg ratio and demonstrated that the Ni/Al Mg (i) Goal and scope definition step that also includes catalyst possessed the highest catalytic activity of 92% specifying the boundaries of the system being analysed glycerol conversion and selectivity towards a hydrogen and defining the functional unit (FU). The goal here is production of 76%. Charisiou et  al. [34] investigated the to conduct a quantitative environmental sustainability GSR reaction for H production and compared the per - assessment to identify which, among three glycerol- formance of Ni supported on ZrO and SiO –ZrO catalysts. reforming technologies, would be characterized as the 2 2 2 They suggested that the addition of SiO stabilizes the ZrO most sustainable and environmentally friendly tech- 2 2 monoclinic structure, restricts the sintering of Ni particles, nology for greener H production. The scope is about 2+ strengthens the interaction between Ni species/support H production from glycerol feedstock and, hence, the and influences the distribution of the gaseous products system boundary is denoted as a cradle-to-production by increasing the H yield (while not favouring the trans- gate (C2G) boundary (Fig. 1). Finally, the selected FU is formation of CO to CO). Thus, a high H /CO ratio can be assumed to be 1 kg of produced H . 2 2 2 achieved with a negligible CO/CO ratio. Shejale and Yadav (ii) LCI analysis that includes input/output materials and [44] investigated the GSR and compared two catalysts, viz. energy flows and direct emissions. During this step, Ni–Cu/La O –MgO and Ni–Co/La O –MgO, prepared by the the process input/output mass and energy flows as 2 3 2 3 co-precipitation and impregnation techniques for the GSR well as compositions are calculated and data gaps reaction. are identified. In this regard, LCI data fall into two With respect to the valorization of glycerol to produce categories: primary (viz. foreground) and secondary value-added products, Kaur et  al. [39] provided a compre- (viz. background) data. In this research, the pri- hensive assessment for the production of value-added mary data for each of the three glycerol-reforming products from crude glycerol. Their assessment included technologies is produced using Aspen HYSYS process the environmental and economic aspects of different simulation. The secondary data (such as the environ- glycerol-conversion routes such as the chemical and bio- mental and human-health burdens associated with chemical methods. the glycerol by-product of biodiesel production via the transesterification of biomass) is directly obtained from the GaBi LCI database. 3 Methods (iii) LCIA, where the environmental burdens of the mate- Sections 3.1 and 3.2 present the methods and tools em- rial and energy flows as well as human-health impacts ployed in this research. Section 3.3 presents the roadmap are quantified.  for integrating Aspen HYSYS simulation results into GaBi The impact assessment of interest to this research product system models to calculate environmental bur - focuses on human health and the natural environ- dens and human-health impacts, and Section 3.4 pro - ment. As discussed in Khalil [8 , 35] as well as in the vides life-cycle inventory (LCI) primary data generated ILCD Handbook [36], there are many methodologies for using Aspen HYSYS simulations of SCWR, APR and ATR conducting LCIA such as CML 2002, Eco-Indicator 99, EPS technologies. 2000, Impact 2002+, ReCiPe 2008, USEtox, TRACI, etc. However, we selected to adopt the TRACI methodology that is currently used in the GaBi LCA software package. 3.1 LCA Also, the midpoint characterization methodology has This research follows the LCA methodology described in lower uncertainty in the results compared to the end- ISO 14040:2006 standards [40]. The LCA methodology com- point characterization used in other methodologies prises four interrelated steps as follows [8, 35]: such as Eco-Indicator 99 [37]. The author of this work Gaseous Glycerol emissions Glycerol Reforming Plant Other reactants Bio-based (SCWR, APR, or ATR) (water, air) H2 gas Water from Thermal and gas/liquid electrical energy System Boundary separator Fig. 1: System boundary of the cradle-to-gate (hydrogen-production plant) system 4 5 GaBi ThinkStep LCA platform. Source: http://www.gabi-software. TRACI: A Tool for Reduction and Assessment of Chemicals and com. Other Environmental Impacts. Downloaded from https://academic.oup.com/ce/article/5/3/387/6317702 by DeepDyve user on 13 July 2021 392 | Clean Energy, 2021, Vol. 5, No. 3 also used TRACI LCIA methodology in his previous re- H , compositions of gaseous and liquid process streams, search [8, 35].  pressure and temperature conditions of the input and TRACI 2.1 LCIA methodology—the US EPA developed output streams, as well as operating conditions inside the impact-assessment methodology (viz. TRACI) for the simulated unit operations (viz. reformer, water–gas the purpose of LCA studies [38]. The impact categories shift (WGS) reactors, gas/liquid separator and pressure- are characterized at the midpoint level draw simple swing adsorber). The Aspen HYSYS CPA fluid property cause–effect chains to show the point at which each package has been selected for performing the thermo- impact category is characterized and reflect the cur - dynamic calculations in the simulated glycerol-reforming rent state of developments, consistently with EPA technologies. regulations and policies, as well as the best available Aspen HYSYS simulations also allowed modelling LCIA practices. The impact categories used in the energy-management strategies such as capturing exo- TRACI methodology are shown in Table 1. The GaBi thermic heat from some of the reforming reaction steps platform includes TRACI 2.1 (which is the latest v -er and from burning exhaust combustible gases to offset the sion of this database). overall endothermal heat of the reformer. For SCWR tech- (iv ) Interpretation of the analysis results including sensi- nology, which is energy-intensive due to the high-pressure tivity studies and scenarios analysis. In this research, requirement, the in situ electricity-generation option in three BC scenarios and four sensitivity scenarios are conjunction with H production has been evaluated using developed and analysed, as discussed in Section 4. Aspen HYSYS simulation. As Table 1 shows, there are two impact categories that dir - 3.1.2 Integration of process simulation with LCA models ectly refer to human health, namely the carcinogens and Because GaBi LCI databases do not contain LCI data for the non-carcinogens impact categories. However, other im- bio-based production of H from glycerol via SCWR, APR pact categories also affect human health, such as particu- or ATR, it was necessary to use the Aspen HYSYS platform late matter, smog formation, ozone depletion and global to simulate each of the three reforming processes. Fig. 1 warming [37]. Also, NO emissions are represented in the shows the HYSYS simulation of the APR process. HYSYS eutrophication and acidification impact categories. Note simulations of the SCWR and ATR have unit operations that SO and NO are contributors to acid gases, which are x x similar to those shown in Fig. 1. However, the main dif- represented in the acidification potential (measured in kg ferences among the three HYSYS simulations include the SO -equiv.). operating conditions and feed streams of the reformer. In this research, the GaBi LCA tool (2018 edition) has For example, SCWR uses supercritical water and glycerol, been used to develop C2G product system models for the and ATR uses air (in addition to water and glycerol) in the glycerol-reforming technologies being evaluated. GaBi reformer. product system models account for the environmental Description of the unit operations depicted in Fig. 2. A mixer and human-health burdens associated with glycerol (a is used to blend the glycerol and water (in the required by-product of the biodiesel process), which is an input feed weight percentages), a pump to compress the feed mix- stream to GaBi product system models. ture to the desired operating pressure of the reformer and a heat exchanger to heat the feed to the desired 3.1.1 Process design and simulation operating temperature of the reformer. Downstream of The use of Aspen HYSYS is deemed necessary to simu- the reformer is an expander valve to lower the pres- late each of the three glycerol-reforming technologies sure of the product gases followed by a heat exchanger and provide useful quantitative information including (Heater-1) upstream of the high-temperature gas–water input/output mass and energy flows per 1 kg of produced shift reactor (GWS-1). The latter unit operation is fol- lowed by a cooling heat exchanger (Cooler-1) to lower Table 1: TRACI 2.1 midpoint impact categories the temperature of the gases before entering the low-temperature WSG-2. TRACI 2.1 midpoint impact category Units The gaseous stream leaving the WGS-2 is then cooled Acidification kg SO -equiv. in Cooler-2 before entering the gas separator, which Ecotoxicity CTUe separates the gases from condensed water. A pressure- Eutrophication kg N-equiv. swing absorber unit is used to separate the product H Global warming kg CO -equiv. from other combustible gases (a mixture of CO and some Human-health particulate matter kg PM2.5-equiv. H ) and CO . The heat generated from the combustion of Human toxicity, carcinogens CTUh 2 2 CO and remaining H is used to offset the heat required Human toxicity, non-carcinogens CTUh for the glycerol reformer. The stream of exhaust gases Ozone depletion kg CFC 11-equiv. Smog formation kg O -equiv. is then cooled (in Cooler-3) before it goes to the stack. CTUe, comparative toxic unit for ecotoxicity impacts (e.g. freshwater toxicity); CTUh, comparative toxic unit for human-toxicity impacts. CPA = Cubic-Plus-Association equation of state. Downloaded from https://academic.oup.com/ce/article/5/3/387/6317702 by DeepDyve user on 13 July 2021 Clean Energy | 393 ADJ-2 SPRDSHT-1 From To Reformer Heater-1 Valve Heater Duty APR Glycerol Pump Pump To Feed Glycerol-Water Out Reformer Mixing Heater Unit Water Pump Reform Electrical Energy Reformer Energy Bottoms ADJ-1 Heat Heat ADJ-2 Removed Added to To from WGS-2 Heater-1 Cooler-1 Cooler-1 Out Heat To To To WGS-2 E WGS-2 Removed Cooler-1 Heater-1 WGS-1 from WGS-2 WGS-1 Cooler-2 Heater-1 Energy Cooler-2 WGS-2 WGS-1 WGS-1 To Bottoms Energy Bottoms Separator H2 Gas (Product) ADJ-2 Heat To PSA Removed CO2 PSA Unit from Cooler-3 PSA Combustible Combustion off-Gases Hot Gases to Cooler-3 Combustion Gas-Liquid Stack Gases Separator Air Furnace Separated Water Furnace Furnace Duty Bottoms Fig. 2: Reforming process simulation of the aqueous-phase reforming technology for hydrogen production The three glycerol-reforming processes (SCWR, APR and CO + CH ⇌ 2H + 2CO (ΔH) at 25 C = 247.02 KJ/mole 2 4 2 R ATR) for H production involve the following primary and R4: Glycerol-oxidation reaction (unique to the ATR secondary stoichiometric reactions: process) R1: Glycerol reforming (primary reaction) C H O + 1.5O ⇌ 3CO + 4H (ΔH) at 25 C  =  –597.73  KJ/ 3 8 3 2 2 2 R C H O + 3H O ⇌ 7H + 3CO (ΔH) at 25 C = 127.75 KJ/mole mole 3 8 3 2 2 2 R o o In addition to the primary reforming reaction (R1), sec- High-temperature (350C) and low-temperature (200 C) ondary reactions may also occur (as described by R2 WGS reactors are included in the three reforming pro- and R3). cesses to convert CO (in the presence of HO) into CO 2 2 R2: Methanation reaction and H . CO + 3H ⇌ CH + H O (ΔH) at 25 C = –205.88 KJ/mole R5: WGS reaction 2 4 2 R R3: Methane dry reforming reaction CO + H O ⇌ H + CO (ΔH) at 25 C = –41.14 KJ/mole 2 2 2 R Note: Calculations of the reaction enthalpies at 25 C and 1 a.t.m. are performed using HSC Chemistry® soft- Note that CO that shows as a reactant in R2 could come from ware (version 10) [42]. the direct decomposition of CH O to CO and H . This decom- 3 8 3 2 Fig. 3 shows how Aspen HYSYS results are inte- position reaction is endothermic with (∆H) of ≈250 kJ/mole at grated with GaBi product system models to quantify 25 C and 1 a.t.m. [41]. Downloaded from https://academic.oup.com/ce/article/5/3/387/6317702 by DeepDyve user on 13 July 2021 394 | Clean Energy, 2021, Vol. 5, No. 3 Aspen HYSYS Glycerol reforming process simulation simulation for production platform of 1 FU of H2 Environmental burdens and adverse human health impacts (per 1 FU of H2): Acidification Ecotoxicity Eutrophication Calculated mass and energy flows & streams Global warming (emissions)chemical Human health, particulates compostions per Human toxicity, carcinogenics 1 FU of H2 Human toxicity, non- carcinogenics Ozone depletion Smog formation Construct life cycle inventory (LCI) for production of 1 FU of H2 Product GaBi LCA system platform model Fig. 3: Integration of the reforming process simulation with the LCA product system model to calculate the midpoint impact categories associated with glycerol reforming for hydrogen production Table 2: Primary life-cycle inventory data associated with the supercritical water reforming (SCWR) of glycerol for 1  kg of H production Mass and energy inputs Mass and energy outputs Glycerol feedstock: 13.3 kg H gas: 1 kg Water to reformer: 44.78 kg CO gas: 13.28 kg Air to combustible-gases furnace: 0.123 kg Water: 41.436 Water/glycerol mass ratio = 3.37 Stack emissions: 2.487 kg Reformer operation conditions: • CH gas: 1.286 kg • Temperature: 800C • CO gas: 0.695 kg • Pressure: 240 bar • Water vapour: 0.392 kg • N gas: 0.0943 kg • O gas: 0.0198 kg Total mass in: 58.203 kg Total mass out: 58.203 kg Thermal energy required = 218 MJ Thermal energy available for harvesting at 90% thermal Electrical energy required = 1.8 MJ efficiency = 139 MJ (to operate process pumps/compressors) Net thermal energy required is ≈81 MJ assuming credit is given to the thermal energy available for harvesting the environmental burdens and adverse human-health to generate the LCI primary data required as the input to impacts of each of the three glycerol-based hydrogen- GaBi product system models. The primary data for SCWR, production technologies. APR and ATR are provided in Tables 2, 3 and 4, respectively. For both Aspen HYSYS simulations and GaBi product system models, 1 kg of H produced has been selected as the FU. Hence, all mass and energy flows and mid-term 4 Results and discussion impact categories are calculated per 1 FU of produced H . The following subsections present results of Aspen HYSYS simulations and GaBi LCIA for the three selected 3.1.3 LCI primary data glycerol-reforming technologies for H production. Also The three glycerol-reforming technologies (SCWR, APR discussed is the interpretation of the generated results. and ATR) are simulated using the Aspen HYSYS platform Per ISO 14040:2006 [40], the interpretation step of the LCA Downloaded from https://academic.oup.com/ce/article/5/3/387/6317702 by DeepDyve user on 13 July 2021 Clean Energy | 395 framework requires that the assessment results should be selected bio-based H-production technologies; (ii) sensi- reported by the most informative means possible. In light tivity scenarios that quantify the effect on the calculated of this ISO reporting requirement, this research adopts midpoint impact categories as a result of variations in pro- the following approaches for interpretation of its LCIA nu- cess electricity sources (viz. US electricity-grid mix versus merical results, namely: (i) BC comparative assessments wind power) and thermal energy sources (viz. natural gas of TRACI 2.1 LCIA results per impact category across the versus biogas). Moreover, the distinctively high operating Table 3: Primary life-cycle inventory data associated with the aqueous-phase reforming (APR) of glycerol for 1 kg of H production Mass and energy inputs Mass and energy outputs Glycerol feedstock: 7.7 kg H gas: 1 kg Water to reformer: 69.3 kg CO gas: 955 kg Air to combustible-gases furnace: 0.123 kg Water: 65.4 Water/glycerol mass ratio = 9.0 Stack emissions: 1.173 kg Reformer operation conditions: • CO gas: 0.051 kg • Temperature: 290C • Water vapour: 0.972 kg • Pressure: 52.69 bar • N gas: 0.0943 kg • CH gas: 0.056 Total mass in: 77.123 kg Total mass out: 77.123 kg Thermal energy required = 216.67 MJ Thermal energy available for harvesting at 90% thermal Electrical energy required = 0.522 MJ efficiency = 201.08 MJ (to operate process pumps/compressors) Net thermal energy required is ≈16.11 MJ assuming credit is given to the thermal energy available for harvesting Table 4: Primary life-cycle inventory data associated with the autothermal reforming (ATR) of glycerol for 1 kg of H production Mass and energy inputs Mass and energy outputs Glycerol feedstock: 9.21 kg H gas: 1 kg Water to reformer: 4.33 kg CO gas: 8.81 kg Air to combustible-gases furnace: 5.53 kg Water: 2.07 Water/glycerol mass ratio = 0.47 Stack emissions: 7.19 kg Reformer operation conditions [43]: • CO gas: 0.0997 kg • Temperature: 650C • Water vapour: 0.0 kg • Pressure: 1.013 bar • N gas: 4.24 kg • CH gas: 2.85 Total mass in: 19.07 kg Total mass out: 19.07 kg Electrical energy required = 41.1 MJ (to operate process pumps/compressors) If credit is given to thermal-energy harvesting (at 90% efficiency) in the ATR process, it would be sufficient to offset the thermal energy required for the reformer Air to Burn Combustible Gases before Stack Gaseous Emissions (N2, CH4, ect.) to Stack Thermal Energy Source Glycerol Feedstock GaBi-Generared Product System Model for H2 Gas (FU = 1 kg) H2 Production via Glycerol Reforming Technology (SCWR) CO2 Gas Byproduct Water to Reformer U.S. Electricity Grid Mix U.S. Diesel-Powered Trucks Water from Process Gas/Liquid Separator Fig. 4: Product system model for H production using glycerol SCWR Input and output mass and energy data for the SCWR technology are presented in Table 2. Downloaded from https://academic.oup.com/ce/article/5/3/387/6317702 by DeepDyve user on 13 July 2021 396 | Clean Energy, 2021, Vol. 5, No. 3 pressure (240 bar) of the SCWR technology allowed for in the HYSYS simulations. More details about the energy- analysing the effect of in situ electricity generation to optimization strategies are discussed in Section 4.2. offset some of the required electricity on the LCIA impact The mass and energy flows calculated using Aspen categories. HYSYS to produce 1 kg H are provided in Tables 2, 3 and 4, respectively. Also, with respect to SCWR of glycerol, Khalil [35] estimated that ~79 MJ of energy would be required to 4.1 Aspen HYSYS simulations: BC and SS produce 1  kg of supercritical water (SCW). This energy is To meet the research objectives discussed in Section 2.2, consumed in heating the water to temperatures above its Aspen HYSYS models are developed to simulate the fol- critical point (374 C) and compressing the resulting steam lowing glycerol-reforming scenarios: (i) BC scenarios in to pressures above its critical pressure (220 bar). which the US electricity-grid mix is assumed to supply the electrical power to operate the pumps and compres- 4.2 Quantifications of GaBi LCA product system sors in the glycerol-reforming processes; also, natural- models for H production gas burning is assumed to supply the required reformer’s 2 thermal energy and for heating the reactants to the re- 4.2.1 SCWR for H production quired reforming temperature; (ii) SS in which wind power, Fig. 4 shows the input and output streams of the GaBi instead of the electricity-grid mix, is assumed to supply the product system model for the production of 1  kg of H electrical power for the process pumps and compressors, (i.e. 1 FU) using SCWR technology. Input and output mass and biogas burning, instead of natural gas, to supply the and energy data are presented in Table 2. The input mass thermal energy required for the reformer; (iii) optimiza- streams represent the glycerol and water feedstocks. The tion strategies to reduce the reforming energy require- input energy streams include the thermal energy required ments (both electrical and thermal) are also considered by the reformer and electricity (from the US grid mix) to Table 5: Midpoint impact categories associated with the supercritical water reforming (SCWR) of glycerol for 1 kg of H production a b TRACI 2.1 midpoint impact categories Scenario 1 Scenario 2 Impact reduction –2 –2 Acidification (kg SO -equiv.) 1.14 × 10 1.07 × 10 6.1% –2 –2 Ecotoxicity (CTUe) 6.37 × 10 5.55 × 10 12.9% –2 –2 Eutrophication (kg N-equiv.) 1.53 × 10 1.50 × 10 0.2% Global warming (kg CO-equiv.) 51.9 51.5 0.7% –4 –4 Human-health particulates (kg PM2.5-equiv.) 5.26 × 10 4.66 × 10 11.2% –9 –9 Human toxicity, carcinogenic (CTUh) 1.69 × 10 1.64 × 10 2.9% –8 –8 Human toxicity, non-carcinogenic (CTUh) 2.10 × 10 1.61 × 10 23.3% –11 –11 Ozone depletion (kg CFC 11-equiv.) 9.91 × 10 1.89 × 10 80.9% –1 –1 Smog (kg O -equiv.) 4.47 × 10 4.41 × 10 1.3% Scenario 1: This is the base case (BC) scenario with electricity generation from the US electricity-grid mix to power the process pumps and compressors. Scenario 2: This is a sensitivity scenario (SS) with in situ electricity generation from SCWR (using a turbine downstream of the reformer where the hot gases at 240 bar are allowed to expand in the turbine to generate electrical power) to supply the electric loads of the process pumps and compressors. Air to Burn Combustible Gases before Stack Gaseous Emissions (N2, CH4, ect.) to Stack Thermal Energy Source Glycerol Feedstock GaBi-Generared Product System Model for H2 Gas (FU = 1 kg) H2 Production via Glycerol Reforming Technology (APR) CO2 Gas Byproduct Water to Reformer U.S. Electricity Grid Mix U.S. Diesel-Powered Trucks Water from Process Gas/Liquid Separator Fig. 5: Product system model for H production using glycerol APR 2 Downloaded from https://academic.oup.com/ce/article/5/3/387/6317702 by DeepDyve user on 13 July 2021 Clean Energy | 397 provide the electrical power to the process pumps and required to operate the process pumps/compressors. The compressors. It is assumed that diesel-powered trucks are associated global-warming potential is 51.9  kg CO -equiv. used for the transportation of glycerol from the biodiesel for Scenario 1 and 51.5 kg CO-equiv. for Scenario 2. Clearly, plant to the reforming plant (distance is assumed to be 50 this energy-management strategy led to an ≈0.8% carbon- km). The output streams include H, gaseous emissions footprint reduction (namely 0.4  kg CO -equiv. per kg of 2 2 (N , CH , CO , etc.) and water removed by the gas/liquid produced H ). In both scenarios, the biogenic carbon rep- 2 4 2 2 separator. These unit operations were modelled in Aspen resents ≈27% of the associated global warming potential HYSYS simulations (as illustrated in Fig. 2). (GWP) (measured in kg CO -equiv. per kg of H produced). 2 2 The reformer is modelled by a conversion reactor (at As Table 3 shows, the highest impact category reduction 800 C and 240 bar) and the two WGS reactors (WGS-1 and is related to ozone depletion (≈81%) and the lowest im- WGS-2) are modelled by equilibrium reactors. WGS-1 oper - pact reduction is related to the eutrophication (≈0.2%) o o ates at 350C and 15 bar and WGS-2 operates at 200 C and impact category. The reduction in the human-toxicity 15 bar. (non-carcinogenic) impact category is ≈23%. Table 5 provides the nine TRACI 2.1 midpoint impact categories associated with the production of 1 FU (viz. 1 kg) 4.2.2 APR for H production of H using SCWR for Scenarios 1 and 2, respectively. Fig. 5 depicts the input and output streams of the GaBi The results summarized in Table 5 indicate some reduc- product system model for the production of 1 kg of H (i.e. tions in the burdens associated with human health and 1 FU) using APR technology. Input and output mass and en- the environment as a result of generating electrical power ergy data are presented in Table 5. The input mass streams (by expanding the hot exhaust gases from the reformer represent glycerol and water feedstocks. The input energy through an installed turbine) to offset the electric loads streams include thermal energy required by the reformer, Table 6: Midpoint impact categories associated with aqueous-phase reforming (APR) of glycerol for 1 kg of H production a b c TRACI 2.1 midpoint impact categories Scenario 3 Scenario 4 Scenario 5 –3 –3 –3 Acidification (kg SO -equiv.) 2.08 × 10 9.66 × 10 1.83 × 10 –2 –1 –2 Ecotoxicity (CTUe) 6.46 × 10 1.25 × 10 6.28 × 10 –4 –3 –4 Eutrophication (kg N-equiv.) 3.19 × 10 2.76 × 10 3.06 × 10 Global warming (kg CO-equiv.) 4.11 4.18 3.88 –4 –4 –4 Human-health particulates (kg PM2.5-equiv.) 2.74 × 10 5.00 × 10 2.55 × 10 –10 –9 –10 Human toxicity, carcinogenic (CTUh) 7.59 × 10 2.00 × 10 7.45 × 10 –8 –7 –8 Human toxicity, non-carcinogenic (CTUh) 1.76 × 10 3.93 × 10 1.65 × 10 –11 –11 –11 Ozone depletion (kg CFC 11-equiv.) 4.96 × 10 5.53 × 10 1.46 × 10 –1 –1 –1 Smog (kg O -equiv.) 1.03 × 10 1.56 × 10 1.01 × 10 Scenario 3: This is the base case scenario in which natural gas is used to provide the required thermal energy in the process. The US electricity-grid mix provides the required process electricity. Scenario 4: In this scenario, biogas is used instead of natural gas to provide the required thermal energy for the process. The US electricity-grid mix provides the required process electricity. Scenario 5: In this scenario, wind-turbine power is used instead of the electricity-grid mix as the source of electricity for operating the process compressors and pumps. Air to Burn Combustible Gases before Stack Gaseous Emissions (N2, CH4, ect.) to Stack Thermal Energy Source Glycerol Feedstock H2 Gas (FU = 1 kg) GaBi-Generared Product System Model for Water to Reformer H2 Production via Glycerol Reforming Technology (ATR) CO2 Gas Byproduct Air to ATR Reformer U.S. Electricity Grid Mix Water from Process Gas/Liquid Separator U.S. Diesel-Powered Trucks Fig. 6: Product system model for H production using glycerol ATR 2 Downloaded from https://academic.oup.com/ce/article/5/3/387/6317702 by DeepDyve user on 13 July 2021 398 | Clean Energy, 2021, Vol. 5, No. 3 Table 7: Midpoint impact categories associated with autothermal reforming (ATR) of glycerol for 1 kg of H production a b TRACI 2.1 midpoint impact categories Scenario 6 Scenario 7 Impact reduction –2 –4 Acidification (kg SO -equiv.) 2.14 × 10 7.17 × 10 96.6% –1 –2 Ecotoxicity (CTUe) 1.57 × 10 1.38 × 10 91.2% Eutrophication (kg N-equiv.) 4.18 4.18 0.0% Global warming (kg CO-equiv.) 87.2 80.3 7.9% –3 –5 Human-health particulates (kg PM2.5-equiv.) 1.59 × 10 8.73 × 10 94.5% –9 –10 Human toxicity, carcinogenic (CTUh) 1.19 × 10 1.08 × 10 90.9% –8 –8 Human toxicity, non-carcinogenic (CTUh) 9.77 × 10 1.09 × 10 88.8% –9 –12 Ozone depletion (kg CFC 11-equiv.) 2.75 × 10 2.39 × 10 99.9% –1 –1 Smog (kg O -equiv.) 3.01 × 10 1.22 × 10 59.5% Scenario 6: This is the base case (BC) scenario in which electricity from the US grid mix is used to power the process pumps and compressors. Scenario 7: Electricity from wind turbines (instead of electricity from the grid mix) is used to power the process pumps and compressors. 4.2.3 ATR for H production Table 8: Comparative primary energy consumption to pro- Fig. 6 shows the input and output streams of the GaBi duce 1 kg of H gas via glycerol reforming product system model for the production of 1 kg of H (i.e. Primary energy 1 FU) using ATR technology. Input and output mass and en- Glycerol-reforming consumption ergy data are presented in Table 6. The input mass streams technology (MJ/kg H ) represent glycerol and water feedstocks as well as air for Supercritical water reforming (SCWR): the ATR reformer. The input energy streams include the • Scenario 1, viz. (BC) 33.17 thermal energy required by the reformer, the US electri- SCWR • Sensitivity Scenario 2 (SS-2) 27.68 city grid to provide electrical power to process pumps and Aqueous-phase reforming (APR): compressors, and US diesel-powered trucks for the trans- • Scenario 3, viz. (BC) 23.76 APR portation of glycerol from the biodiesel plant to the re- • Sensitivity Scenario 4 (SS-4) 23.71 forming plant (assumed to be 50 km). The output streams • Sensitivity Scenario 5 (SS-5) 23.62 include H , gaseous emissions (N , CH , CO , etc.) and water 2 2 4 2 Autothermal reforming (ATR): removed to the gas/liquid separator. The ATR reformer is • Scenario 6, viz. (BC) 121.15 ART modelled by a conversion reactor (at 650 C and 1.013 bar) • Sensitivity Scenario 7 (SS-7) 104.43 and the two WGS reactors (WGS-1 and WGS-2) are mod- elled by equilibrium reactors. WGS-1 operates at 350 C and the US electricity grid to provide electrical power to the 15 bar and WGS-2 operates at 200 C and 15 bar. process pumps and compressors, and US diesel-powered Input and output mass and energy data for the ATR trucks for the transportation of glycerol from the biodiesel technology are presented in Table 6. plant to the reforming plant (assumed to be 50 km). The In the BC scenario, electrical power is provided from the output streams include H, gaseous emissions (N , CH , 2 2 4 US electricity-grid mix. In the sensitivity scenario, elec- CO , etc.) and water removed to the gas/liquid separator. trical power is assumed to be provided from wind power. The reformer is modelled by a conversion reactor (at Table 7 summarizes the GaBi LCIA results (midpoint im- 290 C and 52.69 bar) and the two WGS reactors (WGS-1 and pact categories associated with the production of 1 kg of H WGS-2) are modelled by equilibrium reactors. WGS-1 oper - using ATR technology). o o ates at 350C and 15 bar and WGS-2 operates at 200 C and As Table 7 shows, switching the electricity supply from 15 bar. Input and output mass and energy data for the APR the US grid mix to wind power resulted in a substantial technology are presented in Table 6. reduction in the adverse environmental impacts (with In BC Scenario 3, electrical power for the modelled pumps the exception of water eutrophication). Also, the carbon- and compressors is provided from the US electricity-grid footprint reduction is ≈8%. The notably higher value of the mix and the thermal energy for the reformer is provided by eutrophication potential compared to APR (see Table 5) and burning natural gas. In one sensitivity scenario, electrical SCWR (see Table 3) can be attributed to use of air in the power is provided from US wind power while natural gas ATR reformer and the resulting N emissions (as shown in is used to the provide the thermal energy for the reformer. Tables 2, 4 and 6, respectively). In the second sensitivity scenario, electrical power is pro- vided from the US electricity-grid mix and biogas burning (instead of natural gas) is used to provide the thermal en- 4.3 Environmental sustainability assessment ergy for the reformer. Inspection of the results depicted in Table 6 shows that, In alignment with ISO 14040:2006 [40] reporting require- with the exception of GWP, the environmental impacts de- ments for the LCA results interpretation step, Sections crease in the following order from highest to lowest: 4.2.1, 4.2.2 and 4.2.3 presented the LCIA results for SCWR, Scenario 4 > Scenario 3 > Scenario 5 APR and ATR technologies and compared their associated Downloaded from https://academic.oup.com/ce/article/5/3/387/6317702 by DeepDyve user on 13 July 2021 Clean Energy | 399 1.0E+01 (BC)_ART (BC)_SCWR (BC)_APR 1.0E+00 1.0E–01 1.0E–02 1.0E–03 1.0E–04 Acidification Ecotoxicity Eutrophication Human Health Smog [kg SO2-Equiv.] [CTUe] [kg N-Equiv.] Particulates [kg O3-Equiv.] [kg PM2.5-Equiv.] Fig. 7: Comparative midpoint impact categories of three base case (BC) scenarios of glycerol reforming via SCWR, APR and ATR processes BC scenarios versus sensitivity scenarios with respect to Other quantitative insights from Table 8 (three BC environmental and human-health impacts. Section 4.3 scenarios and four SS): shows how the aforementioned LCIA results (viz. sustain- • For SCWR: in situ electricity generation (via expanding ability indicators) of SCWR, APR and ATR compare with the high-pressure product gases from the reformer) to each other in order to identify which glycerol-reforming offset the use of electricity from the grid mix resulted in technology can be considered the most sustainable. The an ≈17% reduction in primary energy consumption per indicators of this environmental sustainability assessment kg of produced H . are: primary energy consumption and environmental and • For ATR: use of electricity from wind power instead of human-health impacts (all evaluated per 1 kg of produced electricity from the US grid mix resulted in an ≈14% re- H ). Table 8 summarizes the primary energy consumed to duction in the primary energy consumption per kg of produce 1 kg of H for the BC scenarios as well as the SS. produced H . As Table 8 shows, the primary energy consump- • For APR: replacement of natural gas with biogas as the tions (measured in MJ/kg of produced H) of the three source of thermal energy for the reforming process BC scenarios can be ordered (from highest to lowest) as led to an ≈0.21% reduction in primary energy con- follows: sumption compared to the BC scenario in which nat- ATR > SCWR > APR ural gas (NG) is the source of thermal energy for the Accordingly, APR represents the lowest energy- reformer. Also, the use of wind power instead of grid intensive glycerol-reforming technology compared to ATR mix to provide the electricity required to power the and SCWR. As shown by the three BC scenarios in Table compressors and pumps in the APR process led to an 8, the energy intensity (MJ/kg H) of APR is ≈20% of that ≈0.60% reduction in the primary energy consumption associated with ATR and ≈72% of that associated with per kg of produced H . SCWR. Also, it should be noted that the ATR process has three feed streams entering the reformer (glycerol, water Figs 7–9 compare the nine TRACI 2.1 midpoint impact and air), which have to be heated to the reformer operating categories of the three BC scenarios associated with the temperature of 650C. Hence, as Table 8 shows, there is SCWR, APR and ATR processes for glycerol-based H pro- more primary energy consumption (in MJ/kg H ) in the ATR duction. The impact categories are divided among Figs 7–9 process compared to the SCWR and APR processes. for visual clarity purposes (notice the y-axis scale differ - As a benchmark, Spath and Mann [45] reported 183.2 ences). The general trends in the three figures show that MJ/kg H for the primary energy consumption of SMR the impact categories are highest for ATR, lowest for APR, technology. with SCWR somewhere in between. Value of Mid-Point Impact Category Downloaded from https://academic.oup.com/ce/article/5/3/387/6317702 by DeepDyve user on 13 July 2021 400 | Clean Energy, 2021, Vol. 5, No. 3 1.0E–07 (BC)_ART (BC)_SCWR (BC)_APR 1.0E–08 1.0E–09 1.0E–10 1.0E–11 Human toxicity, carcinogenics Human toxicity, non-carcinogenics Ozone Depletion [CTUh] [CTUh] [kg CFC 11-Equiv.] Fig. 8: Comparative midpoint impact categories of base case (BC) scenarios of glycerol reforming via SCWR, APR and ATR processes (continued from Fig. 7) 1.0E+02 (BC)_ART (BC)_SCWR (BC)_APR 9.0E+01 8.0E+01 7.0E+01 6.0E+01 5.0E+01 4.0E+01 3.0E+01 2.0E+01 1.0E+01 0.0E+00 (BC)_ART (BC)_SCWR (BC)_APR Fig. 9: Comparative midpoint impact categories of base case (BC) scenarios of glycerol reforming via SCWR, APR and ATR processes (continued from Fig. 7) A closer inspection of Fig. 7 shows that the APR process A closer inspection of Fig. 8 shows that the APR process leads to the highest reduction in the eutrophication cat- leads to the highest reduction in ozone depletion (followed egory (followed by SCWR) compared to ATR. Quantitatively, by SCWR) compared to ATR. the eutrophication impact category associated with APR Inspection of Fig. 9 for BC scenarios shows that the APR technology is <0.01% of that associated with ATR tech- technology reduces CO emissions by ≈95% per kg of pro- nology and ≈2% of that associated with SCWR technology. duced H compared to ATR and by ≈92% per kg H com- 2 2 The notably high eutrophication category associated with pared to SCWR. As a benchmark, CO emission from SMR is ATR can be attributed primarily to the air used in the re- 11.9 kg CO -equiv./kg H [45] compared to 4.1 kg CO equiv./ 2 2 2 former and the resulting N emission. kg of H for the APR technology. However, the CO emission 2 2 2 Value of Mid-Point Impact Category Global Warming [kg CO2-Equiv.] Downloaded from https://academic.oup.com/ce/article/5/3/387/6317702 by DeepDyve user on 13 July 2021 Clean Energy | 401 from SMR remains below that of the ATR and SCWR tech- HYSYS process simulations with the GaBi LCA platform. nologies (which are 87.2 and 51.9  kg CO equiv./kg H , The sustainability indicators used in this assessment 2 2 respectively). conform to the TRACI 2.1 LCIA methodology and include Additional quantitative LCIA-based insights: the primary energy consumption and midpoint impact categories (viz. human-health and environmental im- • SCWR sensitivity Scenario 2 shows a reduction of pacts) per kg of H. ≈0.4  kg CO -equiv./kg H compared to SCWR Scenario 2 2 The results of our study revealed that H production from 1 (BC) as a result of using in situ generated electricity glycerol via APR is more environmentally sustainable than compared to using electricity from the US grid mix glycerol ATR, SCWR and conventional SMR technology. For to power the process pumps and compressors in the example, the primary energy consumption (in MJ/kg H ) can SCWR process. be ranked from highest to lowest in the following order of • APR sensitivity Scenario 4 shows a carbon-footprint re- H production technologies: SMR > ATR > SCWR > APR. The duction of ≈0.7 kg CO -equiv./kg H as a result of using 2 2 associated CO emissions (in kg CO -equiv./kg H ) can also 2 2 2 wind power (a renewable energy source) instead of the be ranked from highest to lowest in the following order of US electricity-grid mix (BC Scenario 3) to power the pro- H -production technologies: ATR > SCWR > SMR > APR. cess pumps and compressors in the APR system. Future research should focus on expanding the scope • ATR sensitivity Scenario 7 shows a reduction in the of this comparative environmental assessment to include carbon footprint of ≈6.9  kg CO-equiv./kg H as a re- 2 2 other H -production routes such as water electrolysis using sult of using wind power, instead of grid electricity, to electricity from wind turbines and solar photovoltaic, as power the process pumps and compressors in the ATR well as the nuclear-powered and solar-powered thermo- process. chemical water-splitting pathways. It is this author’s opinion that bio-based H production via glycerol APR technology is a promising alternative to fossil- based H production using SMR for the following reasons: Acknowledgements The author of this research greatly appreciates the reviews and • SMR technology consumes natural gas (a depletable re- constructive comments provided by several experts in sustain- source) and emits ≈11.9 kg CO for every kg of produced ability and LCA from the University of Oxford in the UK and from hydrogen. It should be noted that this CO emission from Harvard University in the USA. The literature materials used in SMR is fossil-based as opposed to being biogenic CO. 2 this research were made available through the Bodleian Libraries • SMR is an energy-intensive technology and would be of University of Oxford and Tate library at Harris Manchester even more energy-intensive if the SMR plant is in- College (HMC) in Oxford during this author’s tenure as a Technical Fellow at HMC in the UK. tegrated with carbon-capture and sequestration processes. • Glycerol feedstock in APR is a by-product of biomass Conflict of Interest transesterification to produce biodiesel. Also, glycerol APR technology emits ≈4.11 kg CO per kg of produced H None declared. 2 2 compared to ≈11.9 kg CO based on the SMR technology. • The carbon footprint of the APR technology can be fur - References ther reduced by integrating the bio-based hydrogen [1] DOE/EERE. Fuel Cell Technologies Market Report 2016. https:// product plant with the biodiesel plant to eliminate GHG www.energy.gov/sites/default/files/2017/10/f37/fcto_2016_ emissions associated with glycerol transport. market_report.pdf (7 May 2021, date last accessed). [2] DOE/EERE. Department of Energy (DOE) Office of Energy Efficiency and Renewable Energy (EERE), FY18 Hydrogen and Fuel 5 Conclusions Cell R&D Funding Opportunity Announcement (FOA) Number: DE-FOA-0001874. https://www.energy.gov/sites/default/ The future of hydrogen as an energy carrier is important, files/2018/08/f54/fcto-fy18-foa-selections_0.pdf (7 May 2021, but so too is the ability to quantify the energy require- date last accessed). ments and environmental burdens associated with H - [3] EPA. Fast facts on transportation greenhouse gas emissions. Source: production technologies. In this regard, LCA is a central Fast facts on transportation greenhouse gas emissions | green framework for performing comparative environmental vehicle guide | US EPA https://www.epa.gov/greenvehicles/ sustainability analyses of H -production technologies. fast-facts-transportation-greenhouse-gas-emissions (7 May Production of H from bio-based feedstocks such as gly- 2021, date last accessed). [4] OECD/IEA. World Energy Outlook 2016. Paris: Organization for cerol (a by-product of biodiesel production via biomass Economic Co-operation and Development & International transesterification) is considered a carbon-neutral tech- Energy Agency (IEA), 2016. nology because its associated CO emission is assumed to [5] Neuling  U, Kaltschmitt  M. Techno-economic and environ- be absorbed during the growth phase of the plant. mental analysis of aviation biofuels. Fuel Processing Technology, In this study, we carried out an environmental sustain- 2018, 171:54-69. ability assessment of glycerol reforming for H production [6] Airbus. Global Market Forecast: Mapping Demand 2016/2035. using a quantitative methodology that integrated Aspen Toulouse: Airbus, 2016. Downloaded from https://academic.oup.com/ce/article/5/3/387/6317702 by DeepDyve user on 13 July 2021 402 | Clean Energy, 2021, Vol. 5, No. 3 [7] Boeing. Current Market Outlook 2016–2035. Chicago, IL: Boeing, [28] Larimi AS, Kazemeini M, Khorasheh F. Highly selective doped Pt/ 2016. MgO nano-sheets for renewable hydrogen production from APR [8] Khalil  YF. Eco-efficient lightweight carbon-fiber reinforced of glycerol. Int. Journal of Hydrogen Energy, 2016, 41:17390-17398. polymer for environmentally greener commercial aviation in- [29] Boga DA, Liu F, Bruijnincx PC, et al. Aqueous-phase reforming dustry. Sustainable Production and Consumption, 2017, 12:16–26. of crude glycerol: effect of impurities on hydrogen produc- [9] EPA. Technical Support Document for Hydrogen Production: Proposed tion. Catalysis Science & Technology, 2016, 6(1):134-143. Rule for Mandatory Reporting of Greenhouse Gases. Washington, [30] Veluturla S, Archna N, Subba Rao D et , al. Catalytic valoriza- DC: US Environmental Protection Agency (EPA), Office of Air & tion of raw glycerol derived from biodiesel: a review. Biofuels, Radiation, 2008. 2018, 9(3):305-314. [10] Hoelzen J, Liu Y, Bensmann B, et al. Conceptual design of op- [31] Putra RD, Trajano HL, Liu S, et  al. In-situ glycerol aqueous eration strategies for hybrid electric aircraft. Energies, 2018, phase reforming and phenol hydrogenation over Raney Ni®. 11(1):217. Chemical Engineering Journal, 2018, 350:181-191. [11] European Commission, EC. Flightpath 2050—Europe’s Vision for [32] Ghani AA, Torabi F, Ibrahim H. Autothermal reforming process Aviation. Luxembourg City: European Commission, 2011. for efficient hydrogen production from crude glycerol using [12] Rocha  MH, Capaz  RS, Lora  EE, et  al. Life cycle assessment nickel supported catalyst: parametric and statistical analyses. (LCA) for biofuels in Brazilian conditions: a meta-analysis. Energy, 2018, 144:129-145. Renewable and Sustainable Energy Reviews, 2014, 37:435-459. [33] Bastan F, Kazemeini M, Larimi A, et al. Production of renewable [13] Zaimes GG, Vora N, Chopra SS,et al. Design of sustainable bio- hydrogen through aqueous-phase reforming of glycerol over fuel processes and supply chains: challenges and opportun- Ni/Al O MgO nano-catalyst. International Journal of Hydrogen 2 3 ities. Processes, 2015, 3(3):634-663. Energy, 2018, 43(2):614-621. [14] Galera S Gutiérrez Ortiz FJ. Life cycle assessment of hydrogen [34] Charisiou  ND, Papageridis  KN, Siakavelas  G, et  al. The influ- and power production by supercritical water reforming of gly- ence of SiO doping on the Ni/ZrO supported catalyst for 2 2 cerol. Energy Conversion and Management, 2015, 96:637-645. hydrogen production through the glycerol steam reforming [15] Thompson  JC, He  B. Characterization of crude glycerol from reaction. Catalysis Today, 2019, 319:206-219. biodiesel production from multiple feedstocks. Applied [35] Khalil YF. Sustainability assessment of solvolysis using super - Engineering in Agriculture, 2006, 22:261-265. critical fluids for carbon fiber reinforced polymers waste man- [16] Subramanian  ND, Callison  J, Catlow  CR, et  al. Optimised agement. Sustainable Production and Consumption, 2019, 17:74-84. hydrogen production by aqueous phase reforming of gly- [36] European Commission, Joint Research Centre, Institute for cerol on Pt/Al O . International Journal of Hydrogen Energy, 2016, Environment and Sustainability. ILCD Handbook: Analysing of 2 3 41(41):18441-18450. Existing Environmental Impact Assessment Methodologies for Use [17] Ciriminna  R, Pina  CD, Rossi  M, et  al. Understanding the gly- in Life Cycle Assessment. 1st edn. Ispra: European Commission, cerol market.Eur opean Journal of Lipid Science and Technology, Joint Research Centre, Institute for Environment and 2014, 116(10):1432–1439. Sustainability, 2010. [18] Statista. Leading Biodiesel Producers Worldwide in 2019, [37] Dhaliwal H. Life Cycle Impact Assessment, 2015. https://www. by Country. https://www.statista.com/statistics/271472/ earthshiftglobal.com/docs/impact-assessment-methods- biodiesel-production-in-selected-countries/ (7 May 2021, date free-lca-class_esg-sept2015.pdf (7 May 2021, date last last accessed). accessed). [19] Schwengber CA, Alves HJ, Schaffner RA, et  al. Overview of [38] Bare J. Tool for the Reduction and Assessment of Chemical glycerol reforming for hydrogen production. Renewable and and Other Environmental Impacts (TRACI). USER’S MANUAL. Sustainable Energy Reviews, 2016, 58:259-266. EPA/600/R-12/554. Cincinnati, OH: National Risk Management [20] DOE/EIA. Electricity in the United States, 2019. https://www.eia. Research Laboratory, Office of Research and Development, gov/energyexplained/electricity/electricity-in-the-us.php (7 U.S. Environmental Protection Agency, 2012, 1–24. May 2021, date last accessed). [39] Kaur J, Sarma AK, Jha MK, et al. Valorisation of crude glycerol [21] Wen G, Xu Y, Ma H, et al. Production of hydrogen by aqueous- to value-added products: perspectives of process technology, phase reforming of glycerol. International Journal of Hydrogen economics and environmental issues. Biotechnol Rep (Amst), Energy, 2008, 33(22):6657-6666. 2020, 27:e00487. [22] Authayanun S, Arpornwichanop A, Paengjuntuek W,et  al. [40] ISO. ISO 14040:2006. Environmental Management—Life Cycle Thermodynamic study of hydrogen production from crude Assessment—Principles and Framework. https://www.iso.org/ glycerol autothermal reforming for fuel cell applications. standard/37456.html (7 May 2021, date last accessed). International Journal of Hydrogen Energy, 2010, 35(13):6617-6623. [41] Almeida PM. Hydrogen production from liquid glycerol reform [23] Ortiz FG, Ollero P, Serrera A, et al. Thermodynamic study of the based on perovskite catalysts. Master’s thesis. Salvador: Federal supercritical water reforming of glycerol. International Journal University of Bahia, 2011. of Hydrogen Energy, 2011, 36(15):8994-9013. [42] Metso:Outotec. HSC Chemistry® Software. https://www. [24] Zhang L, Karim AM, Engelhard MH, et  al. Correlation of Pt–Re mogroup.com/portfolio/hsc-chemistry/?r=2 (7 May 2021, date surface properties with reaction pathways for the aqueous- last accessed). phase reforming of glycerol. Journal of Catalysis, 2012, 287:37-43. [43] Liu Y, Farrauto R, Lawal A. Autothermal reforming of glycerol [25] Ortiz FJ, Ollero P, Serrera A, et  al. Optimization of power and in a dual layer monolith catalyst. Chemical Engineering Science, hydrogen production from glycerol by supercritical water re- 2013, 89:31-39. forming. Chemical Engineering Journal, 2013, 15(218):309-318. [44] Shejale  AD, Yadav  GD. Ni–Cu and Ni–Co supported on La–Mg [26] Avasthi KS, Reddy RN, Patel S. Challenges in the production of based metal oxides prepared by coprecipitation and impregnation hydrogen from glycerol–a biodiesel byproduct via steam re- for superior hydrogen production via steam reforming of glycerol. forming process. Procedia Engineering, 2013, 51:423-9. Industrial & Engineering Chemistry Research, 2018, 57:4785-4797. [27] Rahman MM. H production from aqueous-phase reforming of [45] Spath PL, Mann MK.Life Cycle Assessment of Hydrogen Production glycerol over Cu-Ni bimetallic catalysts supported on carbon via Natural Gas Steam Reforming. NREL/TP-570–27637. Golden, nanotubes. Int. Journal of Hydrogen Energy, 2015, 40:14833-14844. CO: National Renewable Energy Laboratory, 2001, 1–33.

Journal

Clean EnergyOxford University Press

Published: Sep 1, 2021

References