Anton A. Kiss, Edwin Zondervan, Richard Lakerveld, Leyla Özkan (Eds.) Proceedings of the 29th European Symposium on Computer Aided Process Engineering June 16th to 19th, 2019, Eindhoven, The Netherlands. © 2019 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/B978-0-12-818634-3.50297-6
The use of optimization tools for the Hydrogen Circular Economy M. Yáñez,a A. Ortiz,a B. Brunaud,b I.E. Grossmann,b I. Ortiz a,* a
Chemical and Biomolecular Engineering Department, University of Cantabria, Av. los Castros s/n, 39005, Santander, Spain b
Carnegie Mellon University, 15213, Pittsburgh, Pennsylvania, USA
[email protected]
Abstract Hydrogen losses in industrial waste gas streams, estimated as 10 billion Nm3 per year in Europe, constitute potential sources for hydrogen recovery. Further use of this waste hydrogen in fuelled devices promotes a world powered by hydrogen while reinforcing the paradigm of the Circular Economy. This will require the availability of effective technologies for hydrogen recovery and purification as well as the techno-economic assessment of integrating the upcycled gas into a sustainable supply chain using decisionmaking tools. This work using a mixed-integer programming model (MILP), and scenario analyses, develops the techno-economic modelling over the 2020-2050 period, at a regional scale comprising the north of Spain. Two main industrial waste streams are considered; one produced at integrated steel mills and coke making industries and, the second one, at chlor-alkali plants. The proposed optimization model integrates the following items: i) technology selection and operation, ii) hydrogen demand forecast, iii) geographical information, iv) capital investment models, and v) economic models. The optimal solutions that arise from the combination of all the infrastructure elements into the mathematical formulation, define the gradual infrastructure investments over time that are required for the transition towards a sustainable future energy mix, including surplus hydrogen. The results confirm that the critical factor in the configuration of the proposed energy system is the availability of industrial surplus hydrogen, contrary to conventional hydrogen energy systems that are mainly controlled by hydrogen demand. Additionally, this study confirms that the optimal levelized cost of upcycled hydrogen is in the range of 0.35 to 1.09 €/kg H2, which is 1.5 to 2 times lower than the price of hydrogen obtained by steam conversion of natural gas. Keywords: Hydrogen recovery, surplus hydrogen, MILP optimization model, hydrogen infrastructure
1. Introduction Hydrogen-based energy storage systems could play in the future a key role as a bridge between intermittent electricity provided by alternative sources and the common fossil fuel-based energy system. The versatility and unique properties of hydrogen open the way to accomplish this goal. Although in recent years, the prospects of a shift to a hydrogen economy have created great interest in the scientific community and social stakeholders, the success relies on the availability of the necessary infrastructures. A number of works focused on the use of decision-support tools for the design and operation of hydrogen
1778
M. Yáñez et al.
supply chains (HSC), have been reported addressing questions such as the design of the hydrogen fuel infrastructure applied at country, region and city levels [1–3]. Likewise, industrial waste streams with a hydrogen content higher than 50% are considered potential and promising sources for hydrogen recovery using cost-effective separation techniques [4]. Meanwhile, among the list of hydrogen-containing waste streams, some studies concentrated on the management, optimization, and utilization of steel-work off gases in integrated iron and steel plants [5,6]. However, little is focused on the optimization of various by-product gases to embed sustainability into HSC. To the best of our knowledge, reported optimization models for HSCs do not consider the competitiveness of upcycling hydrogen-rich waste gas sources for its reuse in both transportation and residential sectors. Hence, the originality of this study is to report the techno-economic feasibility of a HSC with contribution of upcycled hydrogen-rich waste gas sources to fuel both stationary and road transport applications [7].
2. Methodology The methodology framework of this work is proposed in Figure 1. For that purpose, a mathematical formulation with the objective of maximizing the net present value (NPV) is proposed. The corresponding problem is stated as follows. Given: • the potential sources for hydrogen recovery composition and their quality; • a set of suppliers with their corresponding time-dependent maximum supply; • locations of the key stakeholders in the target region: suppliers, merchants, and customers; • a set of allowed routes between the three stakeholders, the transportation mode between them, the delivery distance between both routes; supplier-to-merchant and merchant-to-customer; • hydrogen demand forecast by customer for both transport and residential sectors; • raw material and product prices; • a set of production, purification and conditioning technologies, and their yields to upgrade raw materials to hydrogen product, as well as their capacity at different scales; • investment and operating costs of each intermediate technology, transportation mode, depreciation, and the residual values at the end of the time horizon; • financial data (such as discount and tax rates). The goal of the proposed model is to provide the optimum answer to the following questions: how much, where and when stakeholders shall make their investments. The outputs provided by the model are: • • • •
optimal investment plan for all the merchants considered and related logistics; location (single- or multiplant), type, scale, and number of installed technologies. sourcing and supply routes for the raw materials and product considered; connections between the stakeholders, and hydrogen flows through the network.
Within the network presented, the optimization model was developed by integrating the technology selection and operation, the hydrogen demand forecast, geographical information, capital investment models, and economic models.
Upcycling surplus hydrogen to design a sustainable supply chain
1779
Figure 1. Methodology to optimize a sustainable hydrogen supply chain
3. Optimization model An optimization modelling approach based on multi-scenario MILP has been developed. The mathematical model was implemented in JuMP (Julia for Mathematical Optimization) and the optimization solver used was Gurobi 7.0.2. The objective of the model is to maximize the NPV over 30 years (2020-2050). Furthermore, the operational planning model regarding plant capacity, production transportation, and mass balance relationships, is considered together with the constraints of these activities. Because of the complexity of the proposed model, a two-stage hierarchical approach has been used in order to solve the MILP model in reasonable computational time, achieving near-optimal solutions (5% optimality gap) in less than 2 h. The first step consists of the solution of a relaxed single-period problem to determine the location of production plants at the end of the horizon. From this initial assessment, merchant companies that are not selected in the first step are eliminated. Next, in the second step, the 30-year horizon problem is solved with a reduced set of merchants. The optimality gaps have been set to 2% and 5% for the first and second step, respectively.
4. Case study: Upcycled waste gas-based HSC for the North of Spain This work is focused on the techno-economic feasibility of the upcycling of hydrogencontaining multicomponent gas mixtures to feed stationary and portable fuel cells using optimization tools, geographically located in the north of Spain. The proposed model is focused on two main industrial waste streams. The first hydrogen source corresponds to high purity hydrogen off gases of the chloralkali industry denoted as raw material R99. The second most valuable by-product considered in the optimization model, is coke oven gas (COG) which is produced at integrated steel mills and coke making industries. Hereafter, this raw material has been denoted as R50 because the average hydrogen composition is between 36-62% vol.
1780
M. Yáñez et al.
Moreover, the geographic distribution of the future hydrogen market comprises a number of stakeholders that correspond to the three nodes of the hydrogen network, as illustrated in Figure 2. Suppliers, which are industrial companies that produce the surplus hydrogen. Merchants, which are the major industrial gases producers and responsible of the raw material transformation into the final product. However, surplus hydrogen transformation onsite was also considered at supplier’s plant sites. Finally, the potential customers that are urban areas with more than a hundred thousand inhabitants.
Figure 2. Waste gas streams-based HSC studied for the north of Spain 4.1. Data collection In this study, two scenarios concerning two levels of demand for road vehicle transportation and residential/commercial sectors have been considered (see Table 1) [8]. Regarding the intermediate technologies, steam methane, CH4, reforming (SMR) with carbon capture and storage (CCS) has been considered as benchmark technology in order to satisfy the expected demand for hydrogen [9]. With regard to the upcycling of surplus hydrogen, we have selected a combination of two of the most mature technologies for hydrogen purification: membrane technology (MEM) followed by pressure swing adsorption (PSA) [10]. The final product P99, pure liquefied hydrogen, requires a liquefaction stage (LIQ.). Each plant type incurs in fixed capital and unit production costs, as function of its capacity. Furthermore, the transportation costs depend on the selected mode and distance [11]. We considered that raw materials are transported as compressed gas hydrogen (CH2) by tube trailer or pipeline (already installed), and the final hydrogen products are shipped as liquid hydrogen (LH2) by truck. We have considered the corresponding unit transportation cost for each type of hydrogen delivery mode. Table 1. Demand scenarios of hydrogen market penetration by end users and timeframe Scenario (S) Hydrogen market penetration (%) Hydrogen demand
End-use (e) e1: Transport sector e2: Residential/Service sector e1: Transport sector S2 e2: Residential/Service sector Total S1 (tons H2 per year) Total S2 (tons H2 per year) S1
2020 2030 2040 2050 0.7 8.4 16.6 25.3 0.0 0.7 3.0 6.7 1.4 16.8 33.2 50.6 1.0 6.0 10.0 13.5 8.9E+03 1.2E+05 2.7E+05 4.6E+05 4.0E+04 3.4E+05 6.3E+05 9.2E+05
4.2. Results and Discussion This section shows the main results obtained by application of the proposed multiperiod mixed-integer programming model to the optimization of the network infrastructure for the fulfilment of low hydrogen demand (S1). The solution of the model determines: i) the amount of hydrogen-rich waste streams (R50 and R99) converted into liquefied hydrogen at the supplier’s plants, and ii) the optimum SMR-CCS plant site locations.
Upcycling surplus hydrogen to design a sustainable supply chain
1781
Investment Network: Case S1 leads to a solution with a NPV of 941 MM , where the revenue derived from hydrogen sales (3370 MM€) is able to absorb the costs (2030 MM ). No single hydrogen production method will be profitable to produce enough hydrogen to fulfil the expected demand on its own. The optimal solution of the integrated surplus HSC leads to the installation of ten units of different technologies until 2050 in northern Spain: W1 (7 units), W2 (2 units) and W3 (1 unit), as shown Table 2. Table 2. Investment plan obtained for scenario S1 j߳J 12 14 16 6 4 7 6 5 6 3
Location Asturias Huesca Cantabria Salamanca BCN (Rubí) Santander Salamanca Huesca Salamanca BCN (Parets del Vallés)
Period 2020 2020 2020 2022 2028 2033 2037 2041 2045 2048
Technology W3 W2 W2 W1 W1 W1 W1 W1 W1 W1
Size (ton H2/year) Investment (MM ) 50000 116.3 1000 8 1000 8 200000 605.2 200000 605.2 200000 605.2 200000 605.2 200000 605.2 200000 605.2 200000 605.2
Thus, decentralized on-site hydrogen production by the upcycling of industrial surplus hydrogen is the best choice for market uptake and for avoiding costly distribution infrastructure until the demand increases. Surplus hydrogen flowrates: As summarized in Figure 3, in Case S1, the full amount of R99 is utilized with an inflow of 293.400 tons over the next 30 years, and R99 can meet 0.5% of the total hydrogen demand in the north of Spain for the whole time period. Whereas the amount of liquefied hydrogen produced from R50, which is 1.497.000 tons of R50, is able to cover a much larger hydrogen demand accounting for 10.1% of the total hydrogen demand. Consequently, the rest of the hydrogen produced to fulfil the total demand is obtained from CH4 using SMR with CCS as benchmark technology while producing the least CO2 emissions compared to the rest of the commercially available technologies.
Figure 3. Share of pure H2 produced per raw material r ࣅ R; CH4; R99; R50 However, the use of inexpensive surplus hydrogen sources may have a central role in the early phase of hydrogen infrastructure build up in the north of Spain. Therefore, industrialized hydrogen will also play an important role in initiating the transition to a hydrogen economy with localized plants of SMR with CCS; this will support the demand before expanding to less populous areas forming a more decentralized green hydrogen production. Analysing the surplus hydrogen flowrates by customer, it can be observed that though R50 is partially marketed to all final end-users, it has a pivotal contribution when the production of the final product is closer to the customers.
1782
M. Yáñez et al.
5. Conclusions This research provides the methodology to assess the techno-economic feasibility of upgrading and reusing surplus hydrogen gases promoting the shift to the Circular Economy. The analysis has been performed over two scenarios of hydrogen demand (S1 and S2) and the results show that as long as both scenarios of hydrogen demand (S1 and S2) apply, all generated case studies lead to a solution with positive NPVs. The results confirm that the use of inexpensive surplus hydrogen sources such as R50 and R99 offers an economic solution to cover hydrogen demand in the very early stage of transition to the future global hydrogen-incorporated economy, especially when the industrialized hydrogen generation is closer to the demand markets.
Acknowledgments This research was supported by the projects CTQ2015-66078-R (MINECO/FEDER, Spain) and SOE1/P1/E0293 (INTERREG SUDOE /FEDER, UE), “Energy Sustainability at the Sudoe Region: Red PEMFC-Sudoe”.
References [1] Almansoori, A. and Shah, N. (2009) Design and operation of a future hydrogen supply chain: Multi-period model. Int. J. Hydrogen Energy. 34 (19),. [2] Ochoa Bique, A. and Zondervan, E. (2018) An outlook towards hydrogen supply chain networks in 2050 — Design of novel fuel infrastructures in Germany. Chemical Engineering Research and Design. 134 90–103. [3] Sabio, N., Gadalla, M., Jiménez, L., and Guillén-Gosálbez, G. (2010) Multiobjective optimization of a HSC for vehicle use including economic and financial risk metrics. A case study of Spain. Computer Aided Chemical Engineering. [4] Shalygin, M.G., Abramov, S.M., Netrusov, A.I., and Teplyakov, V. V. (2015) Membrane recovery of hydrogen from gaseous mixtures of biogenic and technogenic origin. Int. J. Hydrogen Energy. 40 (8), 3438–3451. [5] Lundgren, J., Ekbom, T., Hulteberg, C., Larsson, M., Grip, C.-E., Nilsson, L., et al. (2013) Methanol production from steel-work off-gases and biomass based synthesis gas. Appl. Energy. 112 431–439. [6] Chen, Q., Gu, Y., Tang, Z., Wei, W., and Sun, Y. (2018) Assessment of low-carbon iron and steel production with CO2 recycling and utilization technologies: A case study in China. Appl. Energy. 220 192–207. [7] Yáñez, M., Ortiz, A., Brunaud, B., Grossmann, I.E., and Ortiz, I. (2018) Contribution of upcycling surplus hydrogen to design a sustainable supply chain: The case study of Northern Spain. Appl. Energy. 231. [8] HyWays (2007) The European Hydrogen Energy Roadmap. HyWays. [9] Sahdai, M. (2017) Review of modelling approaches used in the HSC context for the UK. Int. J. Hydrogen Energy. 42 (39), 24927–24938. [10] Alqaheem, Y., Alomair, A., Vinoba, M., and Pérez, A. (2017) Polymeric GasSeparation Membranes for Petroleum Refining. 2017. [11] Reuß, M., Grube, T., Robinius, M., Preuster, P., Wasserscheid, P., and Stolten, D. (2017) Seasonal storage and alternative carriers: A flexible hydrogen supply chain model. Appl. Energy. 200 290–302.