*4.2. Result Analysis*

As mentioned previously, in this work, three scenarios are solved to illustrate the proposed model, which are:


Scenario 1: Optimisation of flower waste biorefinery with maximum economic performance.

Solving Equation (11) subject to Equations (1)–(8) in a commercial optimisation model, LINGO version 18 with global solver.

$$\text{Maximise } EP \tag{11}$$

Figure 3 illustrates the optimised pathway for scenario 1. Based on the optimisation model, 68,493 kg/h of flower waste undergo a drying process in the pretreatment section to reduce the moisture content to 11.55 wt%. As a result, 23,445 kg/h of flower waste is formed and sent into the biomass conversion process.

**Figure 3.** Optimisation based on maximum economic potential—scenario 1.

Next, flower waste is extracted with a supercritical carbon dioxide extraction method to be further converted into different intermediate products such as biomass waste, wastewater and flower essential oil. The amount of intermediate product is 468 kg/h, 703 kg/h and 22,273 kg/h, respectively.

Then, both biomass waste and wastewater are sent to the upgrading process and waste treatment process. The composting method is selected among the upgrading processes (e.g., fermentation, biogas production, biosorbent production, etc.) to form 362 kg/h of compost. Flower essential oil is converted into 21,159 kg/h of geranyl acetate via the transesterification process. The rest of the waste produced is sent to a landfill. The overall economic potential and total carbon emission in scenario 1 are summarised in Table 10.

**Table 10.** Economic potential and environmental impact in scenario 1.


Scenario 1 shows that the pathways chosen to have significantly high revenue, but the total carbon dioxide released is extremely high as the environmental impact is being neglected. The carbon generated in scenario 1 is mostly from the transesterification process, which is 76,930 kg CO2/h, as this process requires the use of machinery and electricity.

Table 11 shows the parameters that affect revenue in scenario 1. The waste generated in scenario 1 is relatively less, around 1924 kg/h. This is due to the biomass formed after extraction process being sent to the upgrading process to form other value-added products, which are compost and geranyl acetate, both with extremely high selling costs. Thus, the main revenue in scenario 1 only depends on the product selling price; treatment cost in this scenario does not have a big impact on the economic potential.

**Table 11.** Parameters that affect revenue.


Scenario 2: Optimisation of flower waste biorefinery with minimum environmental impact.

The synthesis of an integrated flower-waste biorefinery in scenario 2 is solved based on the optimisation objective listed in Equation (12). With this, the pathway of choosing the minimum environmental is determined.

$$\text{Minimise } TCE \tag{12}$$

Figure 4 shows the summarised refinery pathway based on the lowest environmental impact generated. First, 68,493 kg/h of flower waste will undergo a two-stage drying; first, sun drying to produce 23,445 kg/h of flower waste and, second, glycerine drying to form 15,904 kg/h of flower waste. Next, in the biomass conversion process, the flowers are coated to form the exact amount of flowers for decoration purposes. The overall economic potential and total carbon emission in scenario 2 is summarised in Table 12.

**Figure 4.** Optimisation based on minimum environmental impact—scenario 2.

**Table 12.** Economic potential and environmental impact in scenario 2.


According to optimised results, the flower-waste mass flow is accumulated and sent to the glycerine drying and coating process, which requires no energy consumption nor release of carbon dioxide into the atmosphere. In this case, zero emissions can be demonstrated. Table 12 shows that the economic potential and environmental impact of optimised pathway in scenario 2. Table 13 shows carbon emission from landfill.

**Table 13.** Carbon emission from landfill.


Scenario 3: Optimisation of flower waste biorefinery with maximum economic potential and minimum environmental impact.

Scenario 1 and scenario 2 can only achieve one objective at a time. Thus, fuzzy optimisation is adopted for the synthesis of an integrated flower-waste biorefinery that considers both objectives simultaneously. The value of total revenue and total carbon emission from both scenario 1 and scenario 2 shown in Table 14 are taken as the upper and lower limits for fuzzy optimization equations shown in Equations (13) and (14), where *X*U, *X*L, *Y*<sup>U</sup> and *Y*<sup>L</sup> are the upper and lower limits of total revenue and total carbon emissions, respectively.

**Table 14.** Parameters required for fuzzy optimisation.


Equations (13) and (14) describe the basis of fuzzy:

TR <sup>−</sup> <sup>X</sup><sup>L</sup> XU <sup>−</sup> <sup>X</sup><sup>L</sup> <sup>≥</sup> <sup>λ</sup> (13)

$$\frac{\mathbf{Y^U} - \text{TCE}}{\mathbf{Y^U} - \mathbf{Y^L}} \ge \lambda \tag{14}$$

with this, the pathway of choosing the minimum environmental is determined as:

$$\text{Minimise } \lambda \tag{15}$$

Figure 5 shows the optimised pathway for scenario 3. The pathways chosen by Lingo are similar to the combination of both scenarios 1 and 2, but with a different mass distribution and upgrading process. The amount of pre-dried flower waste in pretreatment is first divided into half and sent to two pathways. First, 11,521 kg/h of pre-dried flower undergo another stage of drying and coating process to form 7815 kg/h of flowers for decorative purposes. Second, another 11,923 kg/h of pre-dried flower is sent to a supercritical carbon dioxide extraction process to form 238 kg/h of biomass waste, 358 kg/h of wastewater and 11,327 kg/h of flower essential oil as intermediate products. Next, biomass waste undergoes fermentation to form 184 kg/h of bioethanol; flower essential oil goes through a transesterification process and eventually forms 10,761 kg/h of geranyl acetate.

**Figure 5.** Optimisation based on maximum economic potential and minimum environmental impact scenario 3.

The overall economic potential and total carbon emission in scenario 3 is summarised in Table 15. This tables shows that both revenue and environmental impact have achieved a balance between scenarios 1 and 2.

**Table 15.** Economic potential and environment impact in scenario 3.


#### **5. Conclusions**

This paper presented a generic mathematical optimisation approach for synthesis of a sustainable integrated flower-waste biorefinery. An Indian case study was solved to illustrate the proposed optimisation model. Multi-objective optimisation (fuzzy optimisation) is adopted as a trade-off between the economic and environmental performances of the integrated biorefinery. Note that the proposed approach can be modified to develop an integrated flower-waste biorefinery based on different geographical constraints and availability of flower waste. This proposed approach demonstrates a preliminary feasibility study of integrated biorefineries based on different parameters, such as conversion ratio, electricity demand product unit cost, etc. Future work involves integrating sensitivity analyses to the above approach on selected parameters to understand their economic impact towards the selection of technologies and processes.

**Author Contributions:** Conceptualization, E.H.Y.C., V.A., L.Y.N., P.S. and D.K.S.N.; methodology, E.H.Y.C., V.A., L.Y.N. and D.K.S.N.; software, E.H.Y.C. and D.K.S.N.; validation, V.A., L.Y.N., P.S. and D.K.S.N.; formal analysis, E.H.Y.C.; investigation, E.H.Y.C.; resources, D.K.S.N.; data curation, E.H.Y.C.; writing—original draft preparation, E.H.Y.C.; writing—review and editing, E.H.Y.C., V.A., L.Y.N., P.S. and D.K.S.N.; visualization, E.H.Y.C.; supervision, V.A., L.Y.N., P.S. and D.K.S.N.; project administration, V.A., L.Y.N. and D.K.S.N.; funding acquisition, D.K.S.N. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Acknowledgments:** This research was funded by Royal Academy of Engineering under the Frontiers of Development scheme.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

