Resiliency considerations in designing commercial scale systems for lignocellulosic ethanol production

https://doi.org/10.1016/j.compchemeng.2021.107239Get rights and content

Highlights

  • An optimization model is solved for a realistic biomass supply chain.

  • Drought scenarios are incorporated as probability-weighted cost in the model.

  • The results from the application reduce the actual cost up to 4.1%.

  • Drought affects the procurement strategy as well as the facility location.

  • Benefits of the model increases with increasing production target.

Abstract

An optimization model for designing a resilient biomass to energy system is applied to study commercial scale ethanol production. Uncertainty in biomass availability due to drought is calculated using historical rainfall data. The probability of drought is region-specific, whereas the magnitude of loss is specific to the feedstock. The supply system includes collection sites, Regional Biomass Pre-processing Depots (RBPD), and biorefineries. The design decisions include locations and capacities of biorefineries and RPBDs, while the operational decisions include feedstock selection and material flow. The probability-weighted expected cost of production is minimized considering ideal and disruption scenarios. The problem is solved for the state of Maharashtra in India with 33 districts and five types of biomass. The results show up to 4.1% benefit due to the proposed approach over a conventional approach that ignores drought. Crop specific scenarios showed that drought affected the facility locations, capacities as well as procurement strategies.

Introduction

Lignocellulosic ethanol is being considered as a sustainable replacement to fossil fuels globally (Giuliano et al., 2016) and several countries, including India, are considering commercialization of lignocellulosic ethanol production. In India, the proposal of achieving 20% blending of gasoline (petrol) with lignocellulosic ethanol using otherwise surplus agricultural residues by the year 2030 highlights the ongoing aggressive steps taken by the Government of India (GoI) in this direction. However, the pertaining uncertainty in the biomass availability hinders the profitability of these systems (Maheshwari, Singla, Shastri, 2017, Soren, Shastri, 2019). The uncertainty in the biomass availability can arise due to inherent factors such as the seasonality (Welfle et al., 2014) and, more critically, dependent on natural factors such as weather, rainfall patterns, and climatic conditions. It is predicted that climate change will affect the rainfall patterns, leading to more frequent extreme events such as droughts and floods. These events will affect different crops to different extents based on the crop water requirement for each crop and the maturity in terms of the growth stage. The consideration of multiple types of biomass for ethanol production will reduce the uncertainty in the supply and result in an inherently resilient system. Moreover, as different biomass have different ethanol yield (Vikash and Shastri, 2019) and different per unit price, managing these disruptions can become challenging and may lead to financial losses due to inefficient biomass use. This situation is true for a country like India, where the biomass availability is spatially distributed, and efficient biomass to energy system will require the utilization of multiple agricultural residues from different crops. These creates additional challenges, because biomass procurement decisions will be affected due to disruptions (Awudu and Zhang, 2012). Therefore, a resilient supply chain network is required, which considers the effect of these disruptions at the design stage (Sabouhi, Pishvaee, Jabalameli, 2018, Haeri, Hosseini-Motlagh, Ghatreh Samani, Rezaei, 2019, Cui, Ouyang, Shen, 2010, Vlajic, Vorst, Haijema, 2012, Mari, Lee, Memon, 2014).

Supply chain resiliency is defined as the capability of a supply chain to return to its original state or a more desirable state after being disrupted by unexpected risks (Hohenstein, Feisel, Hartmann, Giunipero, 2015, Kim, Chen, Linderman, 2015, Elleuch, Dafaoui, Elmhamedi, Chabchoub, 2016, Ribeiro, Barbosa-Povoa, 2018). The ability of the supply chain to maintain production during disruption also relates to supply chain resiliency. This aspect in the bioenergy sector is the focus of the present study. In the bioenergy sector, optimization has been used as an important tool to solve problems related to supply chain resiliency (Lo et al., 2021). The majority of disruption are assumed at the infrastructure (Marufuzzaman, Eksioglu, Li, Wang, 2014, Huang, Pang, 2014, Bai, Li, Peng, Wang, Ouyang, 2015, Liu, Wang, Ouyang, 2017) with earthquake, hurricane and flooding as the potential triggers for unexpected risk to supply chain. Huang and Pang (2014) considered the effect of earthquake at the biorefinery where five damage stages from no effect to complete failure relate to disruption scenarios whereas, Bai et al. (2015) considered flooding as a potential trigger for disruption. Marufuzzaman et al. (2014) modelled disruption at the inter-modal hubs where the transportation of biomass was disrupted due to hurricane and flooding. Liu et al. (2017) considered disruption at the collection sites as a result of flooding where the resiliency of the supply chain was improved by changing the assignments of the connected farm to an alternate location. The farm and the biorefinery locations were fixed and the uncertainty in biomass availability was due to seasonality which had no effect on the supply chain resiliency.

Prior work in designing resilient biomass supply chains has two limitations. First, in most cases, the drought is not considered to affect biomass availability. As mentioned previously, natural events are expected to disrupt agriculture more frequently in the future, which makes this assumption limiting. This is particularly true for India and many other countries where agriculture is mainly rainfed. Second, most studies have considered a single type of biomass. This again is limiting assumption considering the realities of scaling up of lignocellulosic biofuel systems. This aspect is particularly important from the resiliency perspective. Although managing multiple biomass types creates logistical challenges, it can also provide redundancy in supply, which is one of the attributes of resiliency. Therefore, diversity in feedstock may improve system resiliency. However, this effect has not been studied in literature. Moreover, a quantification method to assess the impact of disruption on biomass loss based on rainfall is also missing in the prior studies.

Maheshwari et al. (2017) addressed the first limitation in their work. They formulated an optimization problem that considered disruption is biomass availability due to events such as drought and flood. A limited number of farms were assumed to be effected by drought with equal probability and the disrupted farms were assumed to be have complete loss of biomass. The formulation was subsequently, extended by Soren and Shastri (2019) to consider scenarios of drought at all the farms where the probability of drought were regional and the disruption level were specific to the location. Additional factors such as short term procurement penalty and the penalty for production shortfall were also incorporated in the design. The design recommended by the proposed approach was compared to a base design that ignored disruptions using supply chain cost as a metric for the economic benefit. Mean Procurement Distance (MPD), Expected Disruption Probability (EDP), and Expected Unmet Demand (EUD) were also compared to highlight the associated design strategy.

This work extends the work by Soren and Shastri (2019) by addressing the second major limitation of considering a single biomass type. This assumption of single biomass is not practical in the Indian context due to small farm sizes and diversity in the cultivated crops. The constraints are modified to include multiple types of biomass to produce ethanol. The additional constraints relate to the difference in the availability of each biomass due to the different harvesting period. In addition to this main novelty, this work also addresses the following limitations of the model proposed by Soren and Shastri (2019):

  • 1.

    The biorefinery capacity was based on the ideal biomass availability and ignored realistic demands.

  • 2.

    The model was applied to a hypothetical scenario and did not consider drought probabilities and the disruption levels using actual data.

  • 3.

    No systematic methodology was included for the calculation of drought probabilities and disruption level based on historical rainfall data.

Addressing these limitations was important for the commercial application of the previously established model by Soren and Shastri (2019). Moreover, the methodological contribution for the calculation of disruption level will be useful in other applications where, the quantification of biomass loss will be required using the historical data of rainfall.

The model is applied to a case study in the Indian state of Maharashtra considering realistic demands, actual residue availability and rainfall data, and five different biomass types. This is another novel contribution of this work considering the limited assessment of commercial lignocellulosic ethanol production in India. The manuscript is arranged as follows. The second section describes the optimization model formulation considered in this work. The modifications over the previous work are also explained in this section. The methodologies used for calculating drought probability and disruption level are explained in third section. The fourth section describes the case study for the application of this model. Results and discussion are described in the fifth section, which is followed by the conclusions in the end.

Section snippets

Optimization model formulation and modifications

The optimization framework was formulated by Maheshwari et al. (2017) and revised by Soren and Shastri (2019). The various inputs in the framework include historical data of rainfall, biomass availability during ideal and drought instances, and the cost coefficients for calculating the production cost. The outputs from the framework is an optimized supply chain design as shown in Fig. 1.

The optimization model is briefly summarized here. The biomass to energy system includes three components:

Methodology for calculating drought probability and disruption level

This work also proposes to use realistic and accurate drought probabilities and disruption levels for each crop. In earlier work, these values were assumed. Here, we have instead proposed a methodology that uses historical rainfall data to determine both these parameters. The methodology is explained here. Although the methodology is explained in the context of crop growth seasons in India, it can be extended to other regions as well.

Case study application

The optimization model is applied to a realistic case study considering 33 districts in the state of Maharashtra in India. Maharashtra is one of the leading agricultural state in India with availability of vast quantities of residues. The collection sites are assumed to be located at the district headquarters of each district and provide multiple types of biomass for procurement. The biomass considered includes cotton stalk, sugarcane bagasse, rice straw, wheat stalk, and jowar stalk. The

Results and discussions

The scenarios with the multi-feedstock design are first discussed, which are followed by the discussion on single biomass design. The time horizon for the simulation is considered to be one year with a step size t of one month. The biomass available in each time step of the horizon is based on the period of harvest for the specific crop. It has been assumed that, the estimates of drought probability and disruption level are reliable considering the long duration of the historical data used for

Conclusions

This work presented a commercial-scale application of a previously proposed approach for designing a supply chain for lignocellulosic ethanol production. The uncertainty in the biomass availability as a result of drought was considered in the design to improve the resiliency of the supply chain. The economic benefits of a resilient supply chain using the proposed design was enumerated by comparing with a conventional design which ignores the effect of drought. The effect of variation in the

CRediT authorship contribution statement

Ashish Soren: Methodology, Software, Formal analysis, Writing - review & editing. Yogendra Shastri: Conceptualization, Methodology, Supervision.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

The rainfall data used in the calculation of drought probability and disruption level was provided by Prof. Subimal Ghosh, Department of Civil Engineering, IIT Bombay.

References (28)

  • A. Soren et al.

    Resilient design of biomass to energy system considering uncertainty in biomass supply

    Comput. Chem. Eng.

    (2019)
  • J.V. Vlajic et al.

    A framework for designing robust food supply chains

    Int. J. Prod. Econ.

    (2012)
  • A. Welfle et al.

    Increasing biomass resource availability through supply chain analysis

    Biomass Bioenergy

    (2014)
  • Y. Bai et al.

    Effects of disruption risks on biorefinery location design

    Energies

    (2015)
  • Cited by (6)

    • Biofuel supply chain management in the circular economy transition: An inclusive knowledge map of the field

      2022, Chemosphere
      Citation Excerpt :

      In addition, “ethanol”, which is the most produced biofuel at the industrial scale level (Amândio et al., 2022), has been addressed in the optimization models of several research. The optimization-based model by Punnathanam and Shastri (2021) for ethanol production from the agricultural residue to minimize the total annual cost and the optimization model by Soren and Shastri (2021) for commercial-scale ethanol production considering the cost minimization are a few examples in this regard. “Techno-economic analysis” of biofuel production and the processes involved are the main focus of some other studies.

    View full text