Design, Construction, and Concept Validation of a Laboratory-Scale Two-phase Reactor to Valorize Whiskey Distillery By-products

The by-products generated from the whiskey distillation process consist of organic liquids with a high chemical oxygen demand (COD) and residues with a high solid content. Low-carbon strategies that repurpose and valorize such by-products are now imperative to reduce the carbon footprint of the food and beverage industries. The operation of a two-phase anaerobic digester to produce volatile fatty acids (VFAs) and biogas may enable distilleries to transition toward a low-carbon bioeconomy. An example of such a system is a leach bed reactor connected to an expanded granular sludge bed (LBR-EGSB) which was designed, commissioned, and conceptually validated in this paper. Several design improvements progress the LBR-EGSB beyond previous reactor designs. An external gas–liquid–solid separator in the EGSB was used to capture any residual gases produced by the effluent and may reduce the amount of methane slippage and biomass washout. The implementation of a siphon-actuated leachate cup is a low-cost alternative that is less prone to actuation malfunction as compared to electrically actuated solenoid valves in previous reactor designs. Furthermore, replacing fresh water with distillery’s liquid by-products as leachate promotes a circular repurpose and reuse philosophy. The system proved to be effective in generating VFAs (10.3 g VFAs L–1Leachate), in EGSB COD removal (96%), and in producing methane-rich biogas (75%vol), which is higher than the values achieved by traditional anaerobic digestion systems. The LBR-EGSB could ultimately provide more by-product valorization and decarbonization opportunities than traditional anaerobic digestion systems for a whiskey distillery.

: Whiskey by-products and characteristics used to calculate cake maize and centrate annual production The total wet weight of thick stillage ( must equal the combined wet weight of ℎ ) cake maize ( and centrate ( . and can be rewritten with ) ) the known volatile content of the wet weight (Table S1):

Appendix S2: Specific Methane Activity (SMA) of Granular Sludge
The specific methane activity (SMA) test is used to establish the methane-producing capacity of the inoculum. The SMA provides insight into the reactor's organic loading rate, particularly during an anaerobic reactor's start-up (Nizami et al., 2011).  Figure S1 shows the cumulative methane production. CUmulative methane production (mL) Time (hr) Figure S1: Cumulative methane production from the SMA test.
The SMA was estimated by using the gradient at the steepest region (Chernicharo, 2007) in ). The actual gas production from the SMA test was 241.4 L CH 4 . g COD. L -1

CH 4
Therefore, the conversion efficiency of the inoculum is ±75 %.

Appendix S3: Optimum HRT to Produce VFA-Rich Leachate
The HRT is the average duration the liquid fraction remains within a reactor. The dilution rate ( ) is the volumetric flow rate (Q) that enters the working volume (V e ) and is the inverse Φ of the HRT for LBRs. A high dilution rate, which equates to a low HRT, is desired. However, a dilution rate that is too high may result in flooding of the substrate and/or washing out of essential microbes. Flooding and microbial washout may negatively impact the leaching process; therefore, a balance must be established. The HRT for LBRs can be calculated as per respectively, for a 16L (working volume) LBR system. However, a study found that HRTs lower than 9hrs lead to pooling on the substrate surface, but this is dependent on the reactor configuration and the substrate. The leaching trial was operated in pooling/flooding conditions. Figure S2 shows the VFA composition during the preliminary leaching trials. The lowest HRT (31 min) yielded the highest VFA yield but resulted in pooling within the LBR. As VFA yield was the primary objective, the 31 min HRT, 140 L.d -1 recirculation rate, was used for the commissioning of the LBRs.

Appendix S4: EGSB Acclimatisation Experimental Trials
The EGSBs were acclimatized through experimental trials and operated over two distinct phases over 82 days as follows ( Figure S3):  Phase 1 (Day 0-56): the EGSB was fed directly with raw mWBM 'wet' independently of the LBRs. This was done to acclimatize the inoculum to mWBM 'wet' and allow inactive sludge to be washed out from the EGSB. Initially, the OLR was 0.7 g COD. L -1 . d -1 (HRT of 23 days) and gradually increased to a maximum OLR of ±2.6 g COD. L -1 . d -1 (HRT of 6 days).  Phase 2 (Day 57-82): the EGSB was fed VFA-rich leachate generated from the LBR preliminary trials (see Appendix S3) but was still disconnected from the LBRs. This was done to acclimatize the inoculum to the VFA-rich leachate. The OLR started at 2.2 g COD. L -1 . d -1 (HRT of 6 days) for 22 days and increased to ±2.6 g COD. L -1 . d -1 (HRT of 6 days). While the OLR changed, the HRT remained the same due to the increased COD concentration of the VFA-rich leachate.  Phase 3 (Day 83-112): the EGSB was connected directly to the LBR leachate after the LBR leaching trials. The EGSB influent flow rate was the same as the settings on day 82. As the COD of the leachate varied, the average COD over the previous 7 days was used to set the OLR to ±2.6 g COD. L -1 . d -1 (HRT of 6 days) for the following week.

Appendix S5: Theoretical LBR-EGSB Biogas Yield
The theoretical biogas, methane, and COD yield for the treatment of mWBM in the LBR-

II. COD Production from mWBM
It can be shown that 1 g of VS produces 1.4 g of COD. The COD production from mWBM 'dry' introduced to the LBRs can be calculated as follows: VS-to-COD conversion = 1.4 g COD g -1 VS VS conversion at 70% = 185 g VS x 1.4 g COD g -1 VS (see I.) Total COD from mWBM dry = 259 g COD from mWBM 'dry' The leachate COD consists of the COD originating in mWBM 'wet' and the COD generated from mWBM 'dry'. mWBM 'wet' consists of centrate (±40%) and thin stillage (±60%). Therefore, the total COD produced after 1 SRT (15 days) can be calculated as follows: Thin stillage added every 7 days = 60 %ww of mWBM 'wet'= 2.4 kg (4kg x 0.6) ≈ 2.4 L Thin stillage added after 15 days = 3.5 L x 3 = 7.2L Total COD from thin stillage added = 7.2 L x 33.4 g COD. L -1 of thin stillage (see Table 1) = 240 g COD A similar calculation for the COD contributed by centrate = 168 g COD Total COD from mWBM 'wet' = 240 g COD (from thin stillage) + 168 g COD (from centrate) = 408 g COD The total COD of mWBM is the sum of the COD of mWBM 'wet' and mWBM 'dry'. Total COD in mWBM = 240 g COD mWBM 'dry' + 408 g COD mWBM 'wet' = 648 g COD mWBM after 21 days

III. Biogas Production from mWBM
1 kg of COD destroyed generates 350 L of CH 4. The methane production can be estimated from the COD production and efficiency of the EGSB (95%) (Chernicharo. 2007). The methane production from the estimated COD produced from the LBR is as follows: 350 L CH 4 kg -1 COD x 648 g COD x 95% efficiency ∴ Methane produced after 21 days = 215 L CH 4 The theoretical specific methane yield of mWBM in the LBR-EGSB can be estimated by using the amount of VS introduced into the system after 3 feeds of mWBM and the methane produced after 21 days: Theoretical VS introduced = 4.67 kg of mWBM x 3 Feeds x VS of mWBM (see Box S1) Theoretical VS introduced = 4.67 kg x 3 x 6.37% = 892 g VS mWBM introduced after 21 days.