Supply Chain Optimization of Integrated Glycerol Biorefinery: GlyThink

May 15, 2017 - Supply Chain Optimization of Integrated Glycerol Biorefinery: GlyThink Model Development and Application ... To further advance the dev...
0 downloads 10 Views 2MB Size
Subscriber access provided by CORNELL UNIVERSITY LIBRARY

Article

Supply chain optimization of integrated glycerol biorefinery: GlyThink model development and application Carina Lira Gargalo, Ana Carvalho, Krist V. Gernaey, and Gurkan Sin Ind. Eng. Chem. Res., Just Accepted Manuscript • Publication Date (Web): 15 May 2017 Downloaded from http://pubs.acs.org on May 22, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Industrial & Engineering Chemistry Research is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

Supply chain optimization of integrated glycerol biorefinery: GlyThink model development and application Carina L. Gargalo1, Ana Carvalho2, Krist V. Gernaey1 and Gürkan Sin1* 1Process

and Systems Engineering Center (PROSYS), Department of Chemical and Biochemical Engineering,

Technical University of Denmark, Building 229, 2800 Kgs. Lyngby, Denmark 2CEG-IST,

Instituto Superior Técnico, Universidade de Lisboa, Av. Rovisco Pais,1, 1049-001 Lisboa,

Portugal *[email protected]

Abstract: To further advance the development and implementation of glycerol based biorefinery concepts, it is critical to analyze the glycerol conversion into high value-added products in a holistic manner, considering both production as well as the logistics aspects related to the supply chain structure. To address the optimal design and planning of the glycerol-based biorefinery supply chain, in this work, we propose a multi-period, multistage and multi-product Mixed Integer Linear Programming optimization model, called GlyThink, based upon the maximization of the Net Present Value (NPV). The proposed model is able to identify operational decisions, including locations, capacity levels, technologies and product portfolio; as well as strategic decisions such as inventory levels, production amounts and transportation to the final markets. Several technologies are considered for the glycerol valorization to high value-added products. Existing countries with major production and consumption of biodiesel in Europe are considered as candidates for the facility sites and demand markets, and their spatial distribution is also carefully studied. The results showed that: (i) the optimal solution which provides the best NPV is obtained by establishing a multi-plant supply chain for the glycerol-based integrated biorefinery, built upon four plant site locations (Germany, France, the Netherlands and Italy); (ii) if a single-plant alternative is to be selected, Germany stands out as potentially the best location for the integrated biorefinery; (iii) government incentives might play a decisive role in the growth of a glycerol-based economy showing improved economic feasibility; and, lastly, (iv) the optimal product portfolio suggested is based on the production of succinic acid and lactic acid, followed by epichlorohydrin and poly-3hydroxybutyrate (PHB). Keywords: glycerol; biorefinery, value added bioproducts; supply chain optimization; product portfolio design;

1. Introduction Global challenges regarding climate change, social distress, strong dependence on fossil fuels and the tireless

ACS Paragon Plus Environment

1

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 37

price instability have motivated the growth and development of a bio-based economy. The biorefinery concept, aiming at replacing fossil sources, has been widely defined as the conversion of all types of biomass (organic residues, crops, etc.) into a broad assortment of bio-based products, such as fuels, chemicals, power and heat, among others (1,2). Considering that the market mostly relies on the petro-based chemicals, materials and fuels, biorefining can aid civilization to preserve resources, maintain regional employment and decrease the environmental burden while ensuring energy security. This leads to an increasing interest into biorefineries and it consequently highlights the need for a robust bio-industry. To this end, it is essential to assess and develop the economic and structural understanding about biorefineries, which implies careful selection of supply chain design, minimizing the threats, all the way from feedstock suppliers to downstream building block technologies and sustainable materials that are converted into higher value chemicals (3). Furthermore, to innovate in the field of bio-products, factors that affect supply chain integration and development, such as government incentives, might be promising and even crucial for the growth of the bio-based industry. Among other countries, a few states in the U.S. have some policies in place to encourage the development and establishment of bio-industry in the country. All in all, novel research on the subject of integrated biorefineries should not only explore how to evolve towards a bio-based economy, but also on how to deal with sustainability concerns so that this industry can compete efficiently with the petro-based industry. A challenge is that emerging technologies suffer not only from uncertain reliability but also are accompanied from uncertain performance characteristics, which lead to a great number of possible alternatives regarding the design, operation and product portfolio offered by biorefineries, from which the most suitable process configurations with regards to economics, environmental constraints and overall sustainability must be selected (4,5). Thus, there is an urgent need for updated models in the frame of the optimal design and planning of integrated biorefineries. Such models should support to identify and validate not only the economic viability of the potential future commercialization of a given product, but also its overall sustainability. Therefore, supply chain management through optimization has become a widespread method to address and analyze the multiplicity of potential alternatives. A supply chain is a network of facilities that perform a set of operations ranging from the acquisition of raw materials, the transformation of these raw materials, and the transportation of finished goods to the clients (6). Thus, supply chain optimization seeks to identify the optimal strategic, tactical and operational decisions in order to maximize the performance of the supply chain regarding a certain performance indicator. Several authors have conducted research in this field, and there are a few literature review articles available in the field of biorefinery optimization, with special emphasis on biomass to biofuel (mostly bioethanol) supply chain optimization problems. An extensive literature review on the topic has for example been presented by Ghaderi et al. (2016) (7). In this work, since the goal is not to provide an extensive review of the literature, and therefore we have narrowed the literature survey to only include examples of supply chain optimization problems that considered common elements of strategic long-term

ACS Paragon Plus Environment

2

Page 3 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

decisions, being: multi-period, multi-stage and/or multi-product supply chain optimization problems. Sammons et al. (2008) (8) proposed a mathematical optimization based framework that uses profitability indicators and further techno-economic metrics to determine the optimal production pathways and set of products for the biorefinery. Chen et al. (2010)

(9) suggested a two-stage programming model to

simultaneously identify the design and operational decision variables for polygeneration systems using coal and biomass to coproduce power, liquid fuels, and chemicals. Bernardi et al. (2012) (10) proposed a MILP for the strategic design and planning decisions of a multi-period and multi-stage upstream ethanol supply chain. Murillo-Alvarado et al. (2013) (11) developed a generalized disjunctive programming model that accounts for the simultaneous selection of products, feedstocks, and processing steps. Zang et al. (2014) (12) uses a mixedinteger nonlinear programming (MINLP) model to determine the optimal plant sizes, locations, and product distributions for an integrated fast pyrolysis biorefinery supply chain. Cambero et al (2015) (13) proposes a multi-period MILP model to optimize a supply chain of forest residues for the production of bioenergy and biofuels. Although there are several studies on the topic of supply chain for biorefineries, most of them focus on biomass to biofuel supply chains, focusing mainly on the upstream of the biofuels supply chain (14). Few works consider multiproduct programming problems. However, even less studies consider the production of value added bioproducts besides biofuels. Furthermore, to the best of our knowledge, there are no studies on the supply chain for the valorization of glycerol to high value added products and building blocks. Therefore, this work focuses on the optimization of the supply chain for the optimal operation of an integrated biorefinery based upon the conversion of glycerol to high value-added products. Glycerol has been vastly used in the health care, but considering that this industry is saturated, the significant growth of biodiesel production is raising a potential issue regarding the significant amount of glycerol surplus in the market. This surplus consequently resulted into a ten-fold decrease in its prices (15). Therefore, with the glycerol valorization being a realistic case study, research into the valorization of this resource has been conducted and important literature has been referring to glycerol as having a tremendous potential as a valuable building block for the production of value added products (15–17). Hence, this work differs from previous studies in the field, by proposing the GlyThink model, a MILP multi-product, multi-period (planning horizon over a discrete set of time periods) and multi-stage (decisions on multiple parts/stages of the supply chain) tool. It introduces the following novel contributions: (i)

a mathematical formulation based upon the proposed superstructure for the maximization of the economic performance across the entire value chain (maximizing Net Present Value), by using detailed cash flow analysis with taxation and capital depreciation, transportation and operating costs;

(ii)

the model appropriately estimates the fixed capital investment for the facilities by employing disjunctive programming to linearize the power-law exponential cost function and reformulating it to a mixed integer linear programming problem;

ACS Paragon Plus Environment

3

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(iii)

Page 4 of 37

the model identifies, at an early stage of design, the optimal operational decisions, including crude glycerol suppliers, plant site location(s), capacity levels, technologies and product portfolio; furthermore, strategic decisions such as inventory levels, production amounts and transportation to the final markets are also supported;

(iv)

a superstructure is proposed by considering cross-country combinations for the optimal location of the plant site facilities, considering several suppliers, multiple upgrading technologies, crude glycerol acquisition and distribution logistics from the plant site location(s) to the market sinks; furthermore, several locations in Europe are pondered as potential plant site locations for the glycerol conversion into high value-added products.

In summary, GlyThink aims at being a decision-making tool through supply chain optimization, leading to the identification of optimal glycerol integrated biorefinery concepts at the early-stage of design, by maximizing the economic performance and identifying the optimal product portfolio, alongside with the most efficient strategic and design decisions. The remainder of the manuscript is organized as follows. In section 2 an overview of the GlyThink model is provided, highlighting the problem statement; in section 3 the mathematical formulation of the optimization problem is formally introduced. Section 4 introduces the case study, and in section 5 the case study results and the discussion are presented. Lastly, key conclusions are drawn and ‘take home’ messages are formulated in section 6.

2. Overview of the GlyThink model The overall network of the integrated glycerol biorefinery supply chain is illustrated in Figure 1, where both upstream and downstream parts of the supply chain are highlighted. Within the network presented, there are six significant sections as follows: 1) transport of biomass to the biodiesel production sites; 2) glycerol production in the biodiesel plants; 3) transportation of crude glycerol to the glycerol conversion plant site(s); 4) glycerol purification; 5) conversion, product separation and purification; 6) distribution of the products to the final markets.

ACS Paragon Plus Environment

4

Page 5 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

Figure 1: Structure of the multi-stage nature of the glycerol to value-added products supply chain network.

The GlyThink model focuses on the downstream part of the network presented in Figure 1. This is due to the fact that: (i) glycerol is an immediate by-product of the biodiesel industry, being produced independently of the raw material used for the biodiesel manufacture; and, (ii) it is nowadays a surplus product, with low market prices (18). Therefore, one of the main goals of this work is to - at early stage of design - identify and critically analyze the optimal integrated glycerol-based biorefinery supply chain for the valorization of glycerol into high value-added products. As presented in Figure 2, the GlyThink model was developed by integrating technology selection and operation, geographical information, capital investment models, economic models and optimization techniques. The corresponding problem can be formally stated as follows. Overall, given: •

a possible superstructure of the integrated glycerol-based biorefinery supply chain combining crude glycerol acquisition, plant site locations, upgrading technologies and distribution logistics;



the planning time period, corresponding to the typical biorefinery lifetime in terms of years;



a set of crude glycerol suppliers and a maximum supply available per supplier;



crude glycerol composition;



a set of potential products to be produced from glycerol;



a set of available production, separation and purification technologies and corresponding yields;



a set of potential locations for the construction of the biorefinery(ies);



a set of potential markets and corresponding demands;



distances between nodes of the supply chain structure;



crude glycerol and product prices;



labor cost dependency on production capacity;



transportation capacity and related costs;



upper and lower bounds for each technology’s capacity level;

ACS Paragon Plus Environment

5

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60



process and economic models (such as fixed capital investment calculation); and,



financial data (such as interest and tax rates).

Page 6 of 37

The goal is to maximize the Net Present Value associated with the integrated glycerol biorefinery supply chain and determining the following operational and strategic decision variables: •

Location of glycerol suppliers and related logistics;



The number, capacities, locations (single- or multi-plant), and technologies for the biorefinery plants;



Glycerol inflow consumed for each selected biorefinery;



Product portfolio, production scale and storage levels at the plant locations for each time period; and,



Product quantities to be delivered from plants to the demand sinks.

Superstructure of glycerol valorization Literature review

Facility capital investment model

Process models

Product and raw material prices

Market demands Distances Suppliers + crude glycerol availability

Annual fixed costs Sales of products and coproducts

Crude glycerol suppliers

Plant locations

GlyThink MILP model for Glycerol valorization

Conversion technology(s) & capacity(s)

Operating costs

Optimal product portfolio

Transportation cost

Optimal flow of products to the markets

Renewable energy horizon 2020 Europe

Figure 2: Component parts and expected outcomes of the GlyThink model.

3. Mathematical Formulation The GlyThink model is formulated as a multi-period, multi-product and multi-stage mixed integer linear program (MILP), aiming at maximizing the Net Present Value associated with the optimal integrated glycerol biorefinery supply chain. From this moment on, SC will be used as a simplified terminology for the integrated glycerol biorefinery supply chain. Constraints are set up concerning: the supply of glycerol to the plant site location(s), production capacity, mass balances, technology(s) for the conversion of glycerol into the final products, operating costs, transportation capacities, and demand satisfaction. The mathematical model is described in detail below. The full nomenclature, including definitions of all sets, variables, and parameters of the model, is given at the end of this article.

ACS Paragon Plus Environment

6

Page 7 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

3.1.

Constraints

The operational planning model regarding plant capacity, production, transportation and mass balance relationships is considered together with the constraints of these activities due to the supply chain structure. Thus, the corresponding constraints and relationships are grouped into six classes, and these are: mass balances, selection of conversion technology and plant location, supply, demand satisfaction, technology capacity and cost, and non-negativity constraints. Each class is presented in more detail below. Mass balances The mathematical formulation regarding the mass balances has been adapted from a previous work developed in our group (19), where generic process models enable the description of every technology within the superstructure through a sequence of tasks that follow the same modeling structure. The overall mass balance for each component i in technology k at plant site location x at each time period t is

 set by Eq. (1). For each technology k at site location x and time period t, the inflow of i (,,, ) plus the

amount of i produced ( ,,, ), must be equal to the amount of i separated as waste ( ,,, ) plus the output

flow (,,, ) to be delivered to the customers or to be stored in location x. 











 ∪

,,, +  ,,, + , ∙ ,,, = 







,,, + ,,, , ∀" ∈ $ ∧ & ∈ ' ∧ ( ∈ )

(1)



Furthermore, in this equation, *+,- represents the set of technologies used for the conversion of raw

materials into value-added products, and finally, *./ represents the set of technologies to be used for the

separation and purification of the above-mentioned products. Also, , is the fraction of a chemical or utility mixed with the process stream (" ∈ $0 ), being 1 if the utility/chemical/solvent i is directly added to the flow stream (e.g. direct steam), and 0 otherwise (e.g. cooling water).

The amount of component i produced or consumed in the conversion technologies,  ,,, , is given in Eq. (2).



 ,,, 

*6789 6(

 = 1,,2 ∙ 3245+,,2 ∙ 245+,,, , ∀" ∧ & ∈ ' ∧ ( ∈ )

:

(2)

6(

where, 1,,2 and 1,,2 ∙ 32+,,2 represent reaction stoichiometry for each component i in technology k and reaction r, and conversion of key reactant i in technology k where r occurs , respectively.

The total amount of chemicals or utilities consumed/added i is obtained as a fraction of the total flow in the technologies k and it is given in Eq. (3) as follows.

ACS Paragon Plus Environment

7

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60