**1. Introduction**

Disposal of large amounts of flower waste is one of the major concerns that leads to environmental issues. Cultural and heritage activities, especially in India, have produced the highest flower waste worldwide [1]. India has a population of religious devotees who commonly practice religious customs using flowers, as they are tokens of devotion in South Asian cultures [2]. Thus, large amounts of flowers are left as offerings and accumulate at religious sites like temples. According to Statista 2021, India's flower production reached about 3 million metric tonnes in the financial year 2021 [1]. Meanwhile, Varanasi is the holiest city in India, which generates around 3.5 to 4 tonnes of flower waste from temples daily. Other than religious activities, flower waste is also generated during different ceremonies, functions and festivals. Most of the flowers utilised in these activities are left unused, resulting in a common source of waste. Viewing the issues above, it is important to have an effective waste disposal policy for the recovery of such waste, with an aim towards reducing the environmental impacts. In current practice, most wilted and unused flowers in India are discarded into landfills or water bodies [3]. This causes water pollution as toxic pesticides and insecticides remaining in the flower waste would seep into the water bodies.

**Citation:** Emily Hau Yan Chong, Viknesh Andiappan, Lik Yin Ng, Parimala Shivaprasad and Denny K. S. Ng Synthesis of Integrated Flower Waste Biorefinery: Multi-Objective Optimisation with Economic and Environmental Consideration. *Processes* **2022**, *10*, 2240. https:// doi.org/10.3390/pr10112240

Academic Editor: Blaž Likozar

Received: 11 August 2022 Accepted: 17 October 2022 Published: 1 November 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

In addition, organic matter in flower waste will cause negative impacts on water quality by depleting oxygen levels in water, threatening the marine ecosystem.

To address the above-mentioned issues, flower waste can be recovered and converted into value-added products. Recently, various technologies have been developed to recover organic waste materials and convert them into value-added products. To maximise the material and energy recovery potential, an integrated flower-waste biorefinery that integrates multiple technologies will be introduced. Detailed discussions on the potential of flower-waste valorisation and synthesis of a sustainable integrated flower-waste biorefinery will be presented in the following sub-sections.

#### *1.1. Valorisation of Flower Waste*

Waste valorisation is an application of industrial ecology that allows closed-loop manufacturing facilities to gain the benefits of the cyclic pattern of the used material present in the consumption sectors [4]. The waste generated by using flowers for decorative or religious purposes can be converted into value-added products for different applications: agricultural, bioenergy, food, beverage, pharmaceutical, etc. For example, phenolic compounds present in flowers could provide a high antioxidant capacity compared to fruits and vegetables. It provides positive effects on oxidative stress-related diseases such as cardiovascular diseases, neurodegenerative diseases, cancer, diabetes, obesity, and liver diseases [5]. Hence, to conserve the medical constituents in flowers, a proper drying method is required. Drying can improve the edible nature of flowers and prolong their shelf life. The dry-flower industry has the great potential to provide employment opportunities to thousands of people, especially to housewives and rural women in India, as unlimited aesthetic value and decorative products can be created by using the dry-flower technology [6]. In addition to being utilised for food and pharmaceutical applications [7,8], flower waste can be processed through different technologies for a variety of purposes. Slavov et al. [9] and Dutta et al. [10] reviewed valorisation opportunities of flower waste to essential oil, recovery of valuable biologically active substances, biofuels production, activated carbon and their application to food industry or medicine. Research work on investigating various strategies for recovery of flower waste have been done over the past years. Table 1 summarises the possible technologies for the utilisation of flower waste.


**Table 1.** Possible technologies for utilisation of flower waste.

#### *1.2. Synthesis of Sustainable Integrated Biorefinery*

According to Kamm (1997), a biorefinery is defined as a complex system of sustainable, environmental and resources-friendly technologies for fully utilising and exploiting biological raw materials [11]. All technologies stated in Table 1 may be applied to convert biological raw materials into value-added products. Research based on different types of biological raw materials has been conducted in the past. To maximise the energy and material recovery, an integrated biorefinery was introduced [12]. An integrated biorefinery

consists of multiple technologies including feedstock handling, pretreatment processes, and different biomass conversion (e.g., thermochemical, biological, physical, etc.) and upgrading processes to produce value-added products and energy [13].

Viewing the importance of sustainable production and consumption in the industry, systematic synthesis and optimisation of integration biorefinery have been developed since the last decade. A number of systematic tools have been developed, ranging from hierarchical approaches [14,15], algorithmic approaches [16,17], mathematical optimisation approaches and hybrid approaches. For example, Ng et al. [14] proposed a hierarchical approach known as the forward-reverse synthesis tree to synthesise and analyse integrated biorefineries. Most recently, Tey et al. [15] presented an extended hierarchical decomposition approach, which was developed for chemical process synthesis for the synthesis of biorefinery processes. Based on the benefit of algorithmic approaches to execute a sequential set of actions based on automated reasoning, calculation and data processing, Pgraph was adapted to synthesise an integrated biorefinery [16] and integrated palm-based biorefinery based on the circular economy concept [17]. To address a complex synthesis problem with multi-criteria requirements, mathematical optimisation approaches have been developed. For example, a superstructure-based shortcut approach was developed by Bao et al. for the conceptual design of integrated biorefineries [18]. A superstructure optimisation model was then adapted in the synthesis of integrated biorefinery for different feedstock, such as a palm-based biorefinery [19], microalgal biorefinery [20], seaweed-based biorefinery [21], etc. A detailed review of the synthesis tools can be found in the recent review papers by Chemmangattuvalappil et al. [22].

#### Environmental Assessment on Integrated Flower-Waste Biorefinery

To synthesise a sustainable integrated biorefinery, an environmental assessment is one of the important factors. As shown by Romero-García et al. [23], various methods can be adopted to evaluate the environmental impact of an integrated biorefinery. These include life cycle assessment (LCA), potential environmental impact (PEI), greenhouse gases (GHG) emission, footprints (carbon, water and sustainability), etc. LCA is a systematic, scientific method for calculating and evaluating long-term environmental impact [24]. LCA has been applied in various integrated biorefineries such as cyclamen plants in greenhouse cultivation [25] and Ethiopian rose cultivation [26]. Russo et al. [25] applied LCA towards roses and cyclamens in greenhouse cultivation. According to Saraiva [27], there are a number of parameters that need to be considered when doing life-cycle assessment, as shown in Table 2.


**Table 2.** Parameters considered in measuring environmental impacts.

Alig et al. [28] applied LCA to determine the environmental impacts of production of different types of cut roses from Holland, Kenya and Ecuador. In the previous work [28], the emission of gases from the packaging and transportation of cut roses are considered when performing LCA.

Limited research has been done on the synthesis of a flower-waste integrated biorefinery. Thus, this work aims to synthesise an integrated flower-waste biorefinery. To design

a sustainable integrated flower-waste biorefinery, other than an environmental assessment, the economic performance of the synthesised biorefinery should also be considered. Therefore, such a design problem is required to be solved via multi-objective optimisation approaches. Note that a number of works have been developed to synthesise an integrated biorefinery with multi-objective functions, such as trade-off between environmental and economic performances [29–32], under uncertainty [33], chemical product design [34], inherent safety and health [35], etc.

According to Tay and Ng [29], fuzzy optimisation identifies the ideal alternatives in decision-making problems by solving an objective function on a set of alternatives given by fixed constraints; the preferable alternatives have the more desired maximum or minimum objective function values. Thus, a fuzzy optimisation is adapted in this work.

#### **2. Problem Statement**

Given a set of flower waste based on various compositions (i.e., petal, leaf, stem), *i* ∈ *I* - *Fi* Biomass that may be transformed into intermediates *<sup>k</sup>* <sup>∈</sup> *<sup>K</sup>* and then further converted into products *k* ∈ *K* via technologies *j* and *j* based on fixed conversion factors, these conversions factors for intermediates *k* and final product *k* are denoted as X*ijk* and X*kj k* . Based on the conversion factors, the flowrates *F* of each intermediate *k* and product *k'* can then be determined. A generic superstructure of the model is illustrated in Figure 1.

**Figure 1.** Generic superstructure for flower-waste biorefinery.

The objective of this work is to produce an integrated flower-waste biorefinery with economic and environmental considerations. Three scenarios are analysed in the case study: (i) maximum economic performance, (ii) minimum environmental impact and (iii) multi-objective optimisation with maximum economic performance and minimum environmental impact.

#### **3. Mathematical Optimisation Model**

Before formulating the mathematical optimisation model, data collection (e.g., conversion performance, emission factor, material and energy requirement, price of product and cost of materials, etc.) for all possible technologies and processes that valorise flower waste into value-added products were collected. Then, a superstructure model that includes all possible process alternatives was generated and a mathematical optimisation model was developed based on different optimisation objectives.

#### *Model Formulation*

Referring to Figure 1, the total amount of flower waste *i* is first sent to various biomass conversion technologies *j* to produce intermediates *k*. The intermediate then undergoes upgrading process *j* to be further converted into products *p* or *K* . The mass flow distribution within the integrated biorefinery is represented by Equation (1).

$$F\_{\text{l}}^{\text{Biomass}} = \sum\_{\text{j}} f\_{\text{ij}} \,\forall \,\,\text{i} \tag{1}$$

where *fij* is the mass flowrate of the flower waste to technologies *j*.

$$F\_k = \sum\_{\mathbf{j}} \sum\_{\mathbf{i}} f\_{\mathbf{i}\mathbf{j}} \mathsf{X}\_{\mathbf{i}\mathbf{j}\mathbf{k}} \,\forall \, k \tag{2}$$

where X*ijk* is the conversion factor of flower waste *i* to intermediate *k*.

For intermediate *k*, which requires further processing in technologies *j'*, the mass flowrate of final product *k'* is given as below:

$$F\_k = \sum\_{j'} f\_{kj'} \,\forall \, k \tag{3}$$

where *fkj'* is the mass flowrate of flower waste intermediate to technologies *j'*.

$$F\_{k'} = \sum\_{j'} \sum\_{k} f\_{kj'} \chi\_{kj'k'} \,\forall \, k' \tag{4}$$

where X*kj*'*k*' X*kj <sup>k</sup>* is conversion factor of intermediate *k* to final product *k'*.

Note that the above mathematical model can be modified to suit different technologies *j* and *j*'. For example, in the drying process, water within the biomass conversion is performed to determine the water lost during the process and total mass changed. For the technologies that convert the biological material into bioenergy, the conversion factor X can be replaced by an energy conversion factor.

Other than tracking the final products, waste generated throughout the biorefinery is also determined. In this work, it is assumed that all the waste that cannot be converted into value-added products will be sent to a landfill. The equation below describes the mass flow of the collected waste to a landfill, from intermediates *k* and final product *k'* in technologies *j* and *j'*.

$$F^{\text{Landfill}} = \sum\_{j} \sum\_{i} f\_{i\bar{j}} \left(1 - \chi\_{i\bar{j}k}\right) + \sum\_{j'} \sum\_{k} f\_{k\bar{j}'} \left(1 - \chi\_{k\bar{j}'k'}\right) \tag{5}$$

The economic performance of the integrated biorefinery (*EP*), the revenue (*REV*) and landfill cost (*COST*Landfill) are determined as below:

$$REV = \sum\_{k} F\_{k} \times \mathbf{P}\_{k} + \sum\_{k'} F\_{k'} \times \mathbf{P}\_{k'} \tag{6}$$

$$\text{CGST}^{\text{Landfill}} = \text{F}^{\text{Landfill}} \times \text{TR} \tag{7}$$

$$EP = REV - COST^{\text{Landfill}} \tag{8}$$

where P*<sup>k</sup>* and P*k*', TR are the selling price of intermediate *k*, final product *k'* and disposure cost, respectively.

Environmental impact is one of the crucial factors when developing a process. In this work, the parameters that affect the total environmental impact *IMP*Total are the energy required for different technologies and carbon emissions based on one unit of electricity usage. The overall environmental impact in this work is determined via Equation (9):

$$IMP^{\text{Total}} = \sum\_{i} \sum\_{j} T\_{j} \times \text{CE}\_{i\bar{j}k} + \sum\_{k} \sum\_{j'} T\_{j'} \times \text{CE}\_{k\bar{j}'k'} \tag{9}$$

where *Tj* and *Tj'* is the energy required for technologies in the biomass conversion process and upgrading process, while CE*ijk* and CE*kj*'k' are the carbon emissions when one unit of electricity is utilised in technologies *j* and *j'*.

#### **4. Case Study**

A case study from India has been used to illustrate the proposed systematic optimisation approach. As mentioned previously, valorisation of flower waste in an integrated biorefinery involves various types of technologies to transform flower waste into valuable product. Based on a detailed literature review, all available technology pathways such as drying, extraction and production to generate bioenergy are summarised in Table 3. Note that the technologies can be widely divided into three categories, which are drying, extraction and generation of bioenergy.


**Table 3.** Summary of flower-waste biorefineries.

#### *4.1. Superstructure*

Based on data collected as shown in Table 3, all possible technology pathways are collected and presented in a superstructure shown in Figure 2. For easier comprehension, the technologies are divided into different categories based on the function of the technologies, namely drying, extraction and bioenergy generation. Note that the superstructure can be divided into several parts: first pretreatment of flower waste, a biomass conversion process that converts the pretreated flower waste into intermediates, and the intermediates are then further converted into final products via upgrading processes.

In this work, the moisture content of flower waste is identified as the major limitation in the main process; therefore, drying processes are taken as pretreatment systems. Based on a literature review, the drying process allows for the reduction in moisture content in flower waste from 40 wt% to 10 wt% [36]. As a result, a two-stage drying process is necessary as one-stage drying is insufficient to meet the requirement. A thermal treatment technique is carried out in the pretreatment section when flower waste is exposed to sunlight before it is sent into the biomass conversion process.

41

As shown in Figure 2, flower waste that undergoes pretreatment can then be sent into coating or extraction processes (biomass conversion process). The coating process is to preserve the edible flower with a longer shelf time and to provide mechanical support, preserving the original form of the flower [42]. In this process, retention of quality parameters (i.e., moisture content, colour, flavour, gaseous exchange and oxidation) are considered.

According to Report Linker [44], the U.S. essential oils market by revenue is predicted to grow at a CAGR of 9% by 2026 due to their value for medicinal and recreational purposes. As a result, this superstructure provides another pathway for extraction of flower essential oil. Intermediates like wastewater, biomass waste and flower essential oil are generated after the extraction process. Wastewater will be sent to waste treatment while biomass waste can go through intermediates process to form bioethanol, biogas, geranyl acetate, etc. Based on research work done by Shivaprasad [43], the flower essential oil produced can undergo a transesterification reaction to form geranyl acetate and citronellyl acetate.

According to Statista [1], the flower production reaches 3 million metric tonnes per year, which amounts to 8220 metric tonnes of flowers produced daily. According to Waghmode [45], 40% of India's entire flower production is sold, amounting to 3300 metric tonnes of flowers sold each day. In this case study, it is assumed that half of the flower waste sold is utilised for decorative purposes and for heritage activities; this will result in a total of 1,640,000 kg of waste produced per day, and roughly 68,500 kg every hour. The properties of flower waste, such as density, composition and weight distribution based on flower parts, are summarised in Table 4. Note that three main parts in flower waste are petals, leaves and stems. However, according to case study in India, flower waste left behind from religious offering is frequently incomplete, only consisting of flower petals and leaves. Thus, the stem parts are neglected in the flower-waste weight distribution.

**Table 4.** Properties of flower waste.


Performance of the drying processes and the conversion factor required for extraction technologies are summarised in Tables 5 and 6, respectively. Note that for the drying process, the final moisture content is used to compute the mass balance of the integrated flower-waste biorefinery.

**Table 5.** Performance of drying processes.





**Table 6.** *Cont.*

To evaluate the economic performance of the integrated flower-waste biorefinery, the selling price of all final products from different technologies are determined. According to Geitner and Block [54], an average operational hour of a production plant is taken as 8000 h per annum, which is used as the basis of operating hours in this work. To determine the number of batches per year (ndry,annual), the total operating hours in a year (tannual) can be divided with the time required for one batch of drying process (tdry,batch) shown in Equation (10):

$$\mathbf{n}^{\text{dry,annual}} = \frac{\mathbf{t}^{\text{annual}}}{\mathbf{t}^{\text{dry,batch}}} \tag{10}$$

The product cost can then be determined by the ratio of cost per batch with weight per batch for each technology. Table 7 shows the product cost for different drying processes. For processes that are not time-constrained, product cost can be determined by a literature review as listed in Table 8.

**Table 7.** Product cost for drying processes.


**Table 8.** Product cost for upgrading processes.


To evaluate the environmental impact of the integrated biorefinery, the energy consumption of each technology is required and has been summarised in Table 9.


**Table 9.** Energy required, and carbon emission for different technologies.

Note that most energy required for drying processes are labelled as N/A, as the drying processes do not require electricity nor power consumption. Sun and shade drying processes occur naturally with the aid of heat energy from the sun to reduce the moisture content in flower waste. Meanwhile, for embedded drying and glycerine drying, external material (i.e., silica gel and glycerol solution) is provided to cover the surface of flower waste, and water will be transferred into the silica body and glycerol solution. In this work, it is assumed 0.51 kg CO2 is released when one unit of electricity is utilised [80].
