Use of lignin side-streams from biorefineries as fuel or co-product? Life cycle analysis of bio-ethanol and pulp production processes

Secchi, M., Castellani, V., Orlandi, M., and Collina, E. (2019)."Use of lignin side-streams from biorefineries as fuel or co-product? Life cycle analysis of bio-ethanol and pulp production processes," BioRes. 14(2), 4832-4865.

Abstract

Increasing the profitability of lignin side-streams is a challenge in the scientific community. Lignin residue originates from black liquor and lignin cake, which are residues from pulp and bio-ethanol production. This paper presents a life cycle assessment study to investigate how pulp and bio-ethanol processes vary in their environmental performance when a fraction of lignin is removed and to identify the best alternative energy source. Fossil energy, natural gas, and cogeneration were evaluated as heat and power alternative sources. The results showed that lignin removal does not considerably affect the environmental performance of the baseline systems and does not generate a relevant risk of “burdens shifting.” Natural gas was the best alternative of power source in a bio-ethanol system, whereas cogeneration showed better compatibility with pulp mills. For the analyzed systems, the necessary allocation distributed the impact contributions between the main products (bio-ethanol/pulp) and the co-products (lignin-cake/black liquor), counterbalancing the impact increase due to the introduction of the new heat, electricity supply, and additional treatment aimed at lignin extraction. Finally, sensitivity analyses confirmed the low influence on the results of the substitution ratio.

Download PDF

Full Article

Use of Lignin Side-Streams from Biorefineries as Fuel or Co-product? Life Cycle Analysis of Bio-ethanol and Pulp Production Processes

Michela Secchi, Valentina Castellani, Marco Orlandi,* and Elena Collina

Keywords: Lignin; Side-streams valorization; Biorefinery; Life cycle assessment; Allocation criteria; Green chemistry; Industrial symbiosis

Contact information: Department of Earth and Environmental Sciences, University of Milano-Bicocca, Piazza della Scienza 1, 20126 Milano, Italy; *Corresponding author: marco.orlandi@unimib.it

INTRODUCTION

Second generation bio-ethanol conversion and pulp production are both well-known technologies. There is an abundance of literature reviews dealing with these industry processes and their sustainability (Aresta et al. 2012; Aditiya et al. 2016). Many aspects, such as the choice of energy sources for pulp mills, the problem of competition in land use for food crops, or the energy profitability of bio-ethanol compared to traditional fuel sources, have been investigated from a life cycle perspective (Bai et al. 2010; Cherubini and Ulgiati 2010; Gaudreault et al.2010; Naik et al. 2010; Cherubini and Strømman 2011; Fazio and Monti 2011; Aresta et al. 2012; Tonini et al. 2012).

One of the current challenges in the scientific community is the use of lignin residue coming from these activities (Sannigrahi et al. 2010; González-García et al. 2016). Lignin, the second most abundant polymer in nature, is the generic term for a large group of aromatic polymers resulting from the oxidative combinatorial coupling of 4-hydroxyphenylpropanoids (Boerjan 2005). Lignin is one of the main compounds present in black liquor and lignin cake, which are residues from pulp and bio-ethanol production processes. As aromatic structures are still present, they have sustainable and economic potential in bio-based chemicals and materials and in the energy sector; for these reasons, new lignin fates are being studied. After a necessary reduction in lignin molecular complexity, purification, or fractionation (Schmiedl et al. 2012; Zoia et al. 2017), the possible applications of lignin as a sustainable alternative to fossil feedstocks include phenolic and epoxy resins (Lora and Glasser 2002; Salanti et al. 2018), green composites (Thakur et al. 2014), and fillers or additives in tires (Frigerio et al. 2014; Barana et al. 2018). These applications could lead to the development and improvement of green chemistry and industrial symbiosis solutions, where the residual streams coming from the biorefinery are sent to other industries and transformed in higher value products. Industrial symbiosis is the clustering of activities to prevent by-products from becoming wastes (EC COM/2014/0398 final).

While only recently lignin extraction from lignin cake residue has been investigated (Argyropoulos 2014), the extraction from black liquor in pulp industry has been explored in the kraft process. One extraction method is the Lignoboost process (Tomani et al. 2011; Tomani 2013). This technology allows a highly purified lignin to be obtained, often used for combustion as an alternative fuel, alone or together with a common liquid fossil fuel, increasing the onsite energy production.

Within the Pro-Lignin “high-value products from lignin side-streams of modern biorefineries” project framework, lignin by-products from current biomass fractionating processes with distinct structural features, e.g., softwood and hardwood kraft and steam-exploded lignin, were assessed with the aim of evaluating the potential valorization routes by also considering the environmental sustainability aspect. The final scope of the project was to improve the profitability of lignocellulosic-based bio-ethanol and pulp production processes by the upgrading of lignin side-streams to high-value products (dispersants, resins, adhesives, foams, and thermoplastics, e.g., for coatings, wood products, composites, and construction materials).

The recovery of lignin as a by-product represents a case of industrial symbiosis because in the current production processes for pulp and bio-ethanol the streams containing lignin (black liquor and lignin cake, respectively) are not discarded as waste, but are used instead as an energy source. Therefore, the potential environmental gain coming from their reuse in other production processes has to be compared with the potential additional burdens coming from the substitution of lignin as a fuel in the original production process.

The present study is one of the outputs of the above-mentioned project, and it investigates how pulp and bio-ethanol processes vary in their environmental performance when lignin is removed from these closed loops where it is one of the main energy sources. Life cycle assessment (LCA) methodology, the most reliable tool to assess the environmental feasibility of eco-innovation strategies, was applied to the current production processes and to the alternative processes where lignin is removed and substituted by other energy sources. The study focused on the “burdens shifting” risk, namely reducing the environmental impact at one point in the life cycle while increasing it at another point. A critical issue addressed is the choice of allocation criteria, i.e., how the environmental burdens arising from a production process are partitioned between products and co-products. In addition, a comparison among different energy sources helps to define the most environmentally friendly option for the substitution of lignin.

The goals of the study were as follows:

to assess the possible risk of “burdens shifting” derived from lignin removal from pulp and bio-ethanol production process;
to evaluate the influence of allocation criteria when lignin cake and black liquor become co-products respectively of bio-ethanol and pulp;
to identify the best alternative to lignin energy source, from an environmental point of view.

The focus of the paper is entirely on the current bio-ethanol and pulp production processes and their modifications for lignin extraction and substitution as an energy source, independently from the use of lignin after its removal. The results may support the pulp and bio-ethanol industry in finding the best solution for the substitution of lignin as an energy source within the innovative production process.

EXPERIMENTAL

The production processes considered in the study are: bio-ethanol production (system 1); softwood (SW) pulp production (system 2); hardwood (HW) pulp production (system 3). For each of the analyzed systems, the study compares the LCA results of the pulp and bio-ethanol production systems in the baseline scenario (B1, B2, and B3), corresponding to the current closed-loop process, with the innovative scenarios where the lignin is partially removed from the production process as a co-product of bio-ethanol and paper (scenarios B1*, B2*, and B3*). The definition of removal scenarios considers several parameters that may influence the result, such as the energy source (or the mix of energy sources) used to replace lignin and the type of allocation applied to the system (e.g., by mass, economic value, or energy content). Therefore, several removal scenarios are developed, assessed, and compared. Table 1 summarizes the options considered and all the scenarios assessed in the study.

The main assumptions tested in each scenario are illustrated in the section, Main Assumptions. Additionally, the Monte Carlo analysis was run to see the combined influence on the impact quantification of the uncertainty related to the various data inputs. This type of analysis makes it possible to calculate the range extent in the LCA actual results. In the Monte Carlo approach, the software builds the uncertainty distribution through a series of repeated calculations (in the present work 1000 times), by taking into account every time a random variable for each value within the uncertainty range related to the specific input data. Finally, two sensitivity analyses were performed to test the effect of the assumptions made on two parameters used in the systems modeling: the amount of lignin removed from the main system and the market price assumed for lignin as a co-product.

Table 1. List of the Evaluated Scenarios ^

^ Abbreviations used to define the names of scenarios: Baseline = “B”, co-generation = “cogen”, natural gas = “gas”, grid electricity = “grid”, heavy fuel oil = “fossil”, allocation by mass = “m”, allocation by economic value = “ec”, allocation by energy content = “en”

Scope Definition, System Boundaries, and Functional Unit of the Studied Systems

As defined in the ISO 14040 (2006) series, the first step in LCA is to define the general aspects on which the analysis is conducted: main scope, system boundaries, and functional unit.

The analysis was performed on 1 kg of final product (bio-ethanol or kraft pulp). This functional unit allows addressing the study on the production process and possible innovations thereof. The analysis is cradle-to-factory gate, which means that it focuses on the production stages of pulp and bio-ethanol supply chains, from the raw materials to the final product. According to its scope, the assessment does not consider product use and disposal stages but only the handling of the biomass input (cultivation and transport) and its treatments (water, energy, and chemicals input), energy recovery, waste, and emissions coming from the abovementioned life cycle stages. The discussion about possible uses of lignin and the development of future scenarios is out of the scope of this paper.

Main Assumptions

Allocation

The scope of the analysis included the lignin extraction stage and the whole production process. Therefore, the allocation rules were adopted to assign the respective share of impact to the products and co-products (bio-ethanol and lignin cake or pulp and black liquor).

Although ISO 14044 (2006) and the International Reference Life Cycle Data System (ILCD) Handbook (Institute for Environment and Sustainability 2010) recommend a system expansion, it is not possible to apply it to the present case studies. Lignin has a wide range of potential uses, but currently none of them is at the market stage. Hence, it is not possible to identify and model a single option of substitution to be included in the models. Moreover, the focus of the analysis is on bio-ethanol and pulp industries and on the effect of lignin removal (and substitution) on the environmental profile of the main products (bio-ethanol and pulp), independently of the final use of the lignin removed. This is reflected also by the functional unit chosen (i.e. 1 kg of main product).

Allocation can be a critical issue when dealing with the LCA of innovations (Hospido et al. 2010; Hetherington et al. 2014). In particular, the choice of the allocation approach could have a remarkable influence on the results, especially when dealing with biorefineries and their multifunctional products (Singh et al. 2010; Cherubini et al. 2011; Wang et al. 2011; Ahlgren et al. 2015; Karka et al. 2015; Sandin et al. 2015; Djomo et al. 2017). Therefore, different allocation approaches are evaluated and compared for the systems considered in this study. The assumptions used to define the allocation values for the products and co-products are described below.

For the bio-ethanol system, the mass allocation is the first choice, as suggested by ILCD Handbook (Institute for Environment and Sustainability 2010) and ISO 14044:2006, but also the energy content criterion has been evaluated. Mass allocation is performed according to De Bari et al. (2008): 3.6 kg of lignin cake are produced for 1 kg of bio-ethanol, therefore 28% of impact is allocated to bio-ethanol.

Energy allocation is performed according to the lignin cake energy content (9850 MJ/t) and 91.9% of the impacts is allocated to the bio-ethanol.

For a SW and HW pulp system, following the EPD Product Category Rules (EPD 2011) for the product basic organic chemicals, an economic allocation is applied based on market price. Because the potential fates of purified lignin could assume different economic values, depending on the material it could replace, economic allocation is performed according to following assumptions:

The prices of pulp and lignin are based on the literature data according to the best knowledge and judgment of the researchers.
The mean monthly price for wood pulp was calculated over the period of October 2013 to May 2015: 688 €/t (IndexMundi 2016).
As there is no list price for lignin and the price is dependent on source and quality of lignin as well as final product, it is not easy to select only one proper price for lignin. Therefore, the price range for lignin will be used to give the best and worst scenarios.
According to the public data available, a price range of 1000 €/t to 2000 €/t was selected for pure high quality kraft lignin to be used to replace high value compounds.
Reasoning for the selected price range is as follows:

According to Lake(2010) and Hodásová et al. (2015), lignin with 1500 USD/t price (1100 €/t) could still be a cost competitive alternative to replace phenol.
According to Nikishkina (2014), the market price of phenol 1500 USD/t is expected to increase in the future, justifying also the use of lignin prices higher than 1100 €/t.
Lake(2010) and Hodásová et al. (2015) report that lignin prices around 2000 USD/t (1500 €/t) could be tolerated in PU resins, whereas Nikishkina (2014) indicated market prices of 6000 USD/t to 8000 USD/t for epoxy resins. Therefore, we assumed that the epoxy resins could tolerate higher prices of 2000 €/t for the lignin raw material.
Even though lignin-based dispersants (such as lignosulfonates) are considered lower/medium value products, Higson (2011) gives price ranges of 250 £/t to 2000 £/t (310 €/ton to 2500 €/ton) for lignin and lignosulfonates in general. Furthermore, Hodásová et al. (2015) report that kraft lignin and lignosulphonates price starts from around 200 USD/t (about 184 €/t).
In Finland, given a margin, the assumed lignin price range could be of 1000 €/t to 2000 €/t, as calculated starting from the data above and in accordance to a personal communication from the companies involved in the Pro-Lignin project.

According to the assumptions, 65% of the impacts are allocated to pulp. A sensitivity analysis on this value was run for softwood pulp production.

In addition, a mass allocation is performed and compared to the economic allocation. According to average production data (44000 t/y of dried black liquor is produced for 181818 t/y pulp), 80.5% of impacts is allocated to pulp.

Amount of lignin removed

The amount of lignin that is removed from the main system (bio-ethanol or pulp production) should be a trade-off between the market value of lignin as a co-product and the cost of its substitution with other energy sources. Because there is currently no market for lignin, the amount of lignin assumed in the scenarios is based on estimations made by companies supporting the work within the Pro-Lignin project. According to these data (personal communication), the amount of lignin removed from the bio-ethanol production process is assumed to be 40%, whereas the amount of lignin removed from the pulp production process is 50%. A sensitivity analysis is run on the softwood pulp system, varying the amount of lignin removed from 50% to 30% (worst scenario, in case the cost of alternative energy sources would rise in the future).

Life Cycle Inventories of the Production Processes

The inventories (LCI) are built with both primary data (i.e. directly collected at manufacturers) as well as secondary data, such as literature and ecoinvent libraries v. 2.2 (Frischknecht et al. 2007). The LCI data includes energy and water consumption, biomass cultivation, chemicals, air and water emissions, disposal, and wastewater treatments.

Bio-ethanol

The bio-ethanol plant is outlined in Fig. 1a. Figure 1b illustrates the same system where a partial lignin removal is taking place.

Fig. 1. System boundaries in bio-ethanol production. (a) Baseline scenario, (b) partial lignin removal scenario

Bio-ethanol production in this study is modelled according to an average plant located in Europe, using simultaneous saccharification and fermentation (SSF) technology. Bio-ethanol derives 50% from crop residues (wheat straw) and 50% from a perennial crop (Arundo donax). Further details about the production process are reported in the supplementary material (See Appendix).

To build the inventory, the ecoinvent (Frischknecht et al. 2007) process “ethanol, 95% in H₂O, from corn, at distillery” has been modified according to the study performed by De Bari et al. (2008) to model a second-generation production process, starting from 50% agricultural residues (wheat straw) and 50% perennial energy corps (Arundo donax).

The wheat straw dataset belongs to the Agrifootprint database (Durlinger et al. 2014) and derives from a multi-output process that describes the average yearly production of wheat grain and straw on a hectare of agricultural soil. Mass allocation based on dry matter content is adopted. Because no data about Arundo donax was available, Miscanthus rhizome was used as proxy for Arundo donax seeds, as they are both perennial energy crops and require similar raw materials and energy input (Fazio and Monti 2011; Salanti et al. 2012). The inventory for Arundo donax cultivation is built according to Monti et al. (2009) and is reported in Table 2. The functional unit of the process is 1 ha of cultivation and allows evaluating the perennial crop end use.

Table 2. LCI for Giant Reed (Arundo donax) Cultivation

Ecoinvent process data for Miscanthus considers the possible benefits derived from using it for phytoremediation, to adsorb heavy metals emissions to soil and water (e.g. copper and zinc) derived from the agricultural phase (Nemecek and Schnetzer 2011). Because it was not possible to verify whether the same can be assumed also for Arundo donax cultivation, emissions of copper, chromium, zinc, and nickel to soil were set to zero, and no avoided impact was inventoried. No additional drying process was included.

In the baseline bio-ethanol process, the lignin cake underwent combustion and supplied 10 kWh electricity to the system. According to expert judgment, the maximum amount of lignin-cake that can be removed without seriously affecting the system is 40%. This quantity corresponds to 4 kWh of electricity that need to be substituted by grid electricity, cogeneration, or natural gas in the alternative scenarios analysed.

(a)

(b)

Fig. 2. System boundaries in softwood and hardwood pulp production (modified from: Suhr et al. 2015). (a) Baseline scenario, (b) partial lignin removal scenario

Pulp (SW and HW)

The pulp plant outline is shown in Fig. 2a. Figure 2b illustrates the same system in a situation of partial lignin removal.

For the softwood pulp production system (details in the supplementary material), the ecoinvent process “sulphite pulp, bleached, at plant” had been modified according to data provided by the one of the biggest pulp and paper companies in Europe, to better represent the situation. For this reason, water and limestone inputs and particulate emissions were changed, and an additional step of lignin extraction (Hannus et al. 2012) was modelled only in the alternative scenario.

LignoBoost technology was adopted only in softwood pulp mills to extract the wood component lignin from the black liquor. The process description is reported in the supplementary material, whereas in Table 3 the chemical inputs are reported, as calculated according to Hannus et al. (2012). Considering that 1 t of black liquor 30% DS (dry solids) contains 0.07 t of lignin 70% DS, a plant producing 4790 t/day of black liquor was taken as reference for calculations, and the final process has been modelled as reported in Table 3. Because of the lack of specific data about energy and water consumption, only chemical inputs are included in the inventory.

Table 3. LCI for 1 kg of Lignin Precipitated with LignoBoost Technology

All data regarding pulp production from hardwood came entirely from a Brazilian company producing pulp and paper from eucalyptus fibers: their inputs were used to modify the ecoinvent dataset “sulphite pulp, bleached, at plant” (as shown in the supplementary material, Table S1) to better model the production process. Instead, for the softwood process the inventory stuck to the dataset retrieved in ecoinvent.

In both pulp production processes, the black liquor undergoes combustion to generate electricity and heat.

In hardwood pulp production systems, black liquor combustion supplies the entire energy requirement; in this case, 0.315 MJ of heat, i.e. 50% of black liquor, was substituted. Alternative energy sources analyzed were heavy fuel oil, cogeneration technology, and natural gas. Because of the lack of detailed information and the similarity of production technologies, the authors chose to model the black liquor removal by applying the same maximum fraction as in a softwood pulp mill (50%, see following paragraph).

For a softwood pulp mill, the authors supposed to remove 50% of the black liquor (maximum removable amount, information from well-acquired experience in a pulp and paper company), i.e. 5.4 MJ of heat. This amount can be alternatively supplied by heavy fuel oil, natural gas, or cogeneration technology. Only in the latter case, also 100% of electricity, i.e. 0.14 kWh, was replaced. The authors tested the effects of all these substitution options to identify the most environmentally friendly one.

A sensitivity analysis was performed to evaluate the influence of the substitution ratio on the results, evaluating a 30% hypothetical substitution rate.

RESULTS AND DISCUSSION

The life cycle inventory analysis (LCIA) was run by means of Simapro 8.0.3 software, using the ILCD impact assessment method v. 1.03 (Institute for Environment and Sustainability 2012).

Results are presented as single score calculations carried out with ILCD method, by applying ILCD normalization factors (Benini et al. 2014) and equal weighting among impact categories, to allow for an easier comparison among the scenarios.

Influence of Lignin Removal

In this section, a comparison between the baseline scenario and the alternative scenario where lignin was removed and replaced is presented for all production systems (bio-ethanol, SW pulp, and HW pulp). In the alternative scenarios, impacts were allocated to both the main product (bio-ethanol and pulp) and to the co-product (lignin). Results were presented for scenarios where the cogeneration technology was selected as the alternative energy source because this was the most probable option according to companies who operate in the two sectors involved.

In Figs. 3, 4, and 5 the baseline scenario for each supply chain was compared to the new one in which lignin was considered as a co-product and impacts were allocated according to the considered criteria, i.e. mass or energy content for bio-ethanol and mass or economic for pulp. What really changes throughout the two supply chains was the amount of lignin removed. In the bio-ethanol system, the removed fraction represented the 40% of lignin cake (which corresponds to 4 kWh of electricity), and in the pulp systems, the removed fraction represented 50% of black liquor (namely, 5.4 MJ of heat for SW pulp and 0.315 MJ for HW pulp).

Fig. 3. Bio-ethanol system: baseline vs. lignin removal (energy [B1 * cogen + en] and mass [B1 * cogen + m] allocation); Lignin substitution with cogeneration, single score

Fig. 4. SW pulp system: baseline vs. lignin removal (economic [B2 * cogen + ec] and mass [B2 * cogen + m] allocation); Lignin substitution with cogeneration, single score