Table 3 Assumptions for the design of a large-scale indigo bioprocess with their values and rationale. It is assumed that the platform relies on the bioprocess sequence of Flowsheet A. (Fig. 10). For detailed calculations see Additional file 1

Values crucial for biological production

Titre

18 g/L

The indigo synthesis titre depends heavily on the upstream choices, such as the choice of bacterial strain, for example, one with available or not high-flux tryptophan pathway modifications, and the enzyme chosen for indigo bioconversion. From small-scale studies, good titres could be assessed as between 0.3 g/L to 18 g/L, when the tryptophan pathway modifications are, respectively, not harnessed (in multiple studies) and harnessed (by Berry et al. 2002) (See “Upstream processing (USP) considerations” section)

For simplicity, the titer figure is assumed as pure indigo, and this value already accounts for the incorrectly formed product, such as trace amounts of indirubin (Berry et al. 2002)

Fermenter Size

250,000 L (250 m³)

In the industrial biology sector, reactors between 100,000 and 250,000 L are common (Meyer et al. 2017; Li et al. 2020). Exemplary applications incl. the production of commodity antibiotics, such as erythromycin A (Zou et al. 2012), butanol (Lee et al. 2008); biofuels like bioethanol (Humbird et al. 2015), citric acid, lactic acid, pigments, such as astaxanthin (Panis and Carreon 2016), amino acids such as glutamate or l-lysine, for example, the latter being produced in 500 m³ bioreactors (Eggeling and Bott 2015)

For simplicity, the fermenter size is considered here as the working volume

Fermenter Number

2

Industrial setup of low-value commodities usually harnesses several bioreactors to achieve significant capacities and using several fermenters is common (Doran 1995; Meyer et al. 2017)

Purification (DSP) yield (%)

75%

DSP usually comprises several unit operations, as can be seen on flowsheets proposed for indigo purification as well (Fig. 10). The overall DSP yield is dictated by the number of unit operations and their individual yields. For 3 DSP steps, the overall yield could often oscillate around 60–90% (NBC2, 2016), and 75% has been assumed for simplification

Facility capacity (L)

500,000 L (500 m³)

Equation: (fermenter number) × (single fermenter size). The capacity size was also benchmarked against possible and technically feasible real-life biotechnological plants (Eggeling and Bott 2015; Humbird 2015)

Fermentation time

72 h

Culture time required to achieve earlier assumed 18 g/L yield (Berry et al. 2002)

Batch runs per year (for a single fermenter)

80

The achievable number of batches is influenced by the growth rate of the employed microorganism, as it influences the fermentation duration (Doran 1995)

18 g/L has been achieved with 3-day long E. coli fermentation; therefore, 2 batches per week have been assumed as possible. To account time for planned or unexpected downtime, 40 weeks of operation per year has been assumed for the bioprocess plant, giving a total of 80 batches scheduled per annum. A study that presented a different scenario with similar assumption may be seen in Pollock et al. (2013)

In real-life the schedule batch is also influenced by the longest step of the whole bioprocess, which for industrial scale may not be fermentation. It may be the DSP, as its steps are typically based on flowrates, marking how much feed volume can be run on a certain equipment beyond which the performance substantially decreases. However, for simplification it is assumed that the DSP would not constrain the batch scheduling, which is not unlikely given that for cheaper biocommodities the DSP is not extensive, as when compared to sophisticated processing of high-value commodities (Budzianowski 2017)

Calculated indigo output from a single fermenter from a single run (after DSP)

3.375 tonnes

Equation: (titre) × (single fermenter size) × (DSP yield)

Calculated annual indigo after-DSP output for values given in this table—for 2 ferementers, each run 80 times

540 tonnes

Equation: (fermenter number) × (output from single fermenter per batch) x (batch runs/year)