Document [original]

Citation: Pluschke, J.; Faßlrinner, K.;

Hadrich, F.; Loukil, S.; Chamkha, M.;

Geißen, S.-U.; Sayadi, S. Anaerobic

Digestion of Olive Mill Wastewater

and Process Derivatives—Biomethane

Potential, Operation of a Continuous

Fixed Bed Digester, and Germination

Index. Appl. Sci. 2023,13, 9613.

https://doi.org/10.3390/app13179613

Academic Editors: Krzysztof Pilarski

and Agnieszka Pilarska

Received: 20 June 2023

Revised: 18 August 2023

Accepted: 19 August 2023

Published: 25 August 2023

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/).

applied

sciences

Article

Anaerobic Digestion of Olive Mill Wastewater and Process

Derivatives—Biomethane Potential, Operation of a Continuous

Fixed Bed Digester, and Germination Index

Jonas Pluschke 1,* , Katharina Faßlrinner 1, Fatma Hadrich 2, Slim Loukil 2, Mohamed Chamkha 2,

Sven-Uwe Geißen 1and Sami Sayadi 3

1Environmental Process Engineering, Technische Universität Berlin (TUB), 10623 Berlin, Germany;

[email protected] (K.F.)

Laboratory of Environmental Bioprocesses, Centre of Biotechnology of Sfax, P.O. Box 1177, Sfax 3018, Tunisia;

[email protected] (M.C.)

Biotechnology Program, Center for Sustainable Development, College of Arts and Sciences, Qatar University,

Doha 2713, Qatar

*Correspondence: [email protected]

Abstract:

Olive mill wastewater (OMW) management is an economic and environmental challenge

for olive oil-producing countries. The recovery of components with high added value, such as

antioxidants, is a highly researched approach that could help refinance performant wastewater

treatment systems. Anaerobic (co-)digestion is a suitable process to valorize the energetic and

nutritional content of OMW and OMW-derived waste streams from resource recovery processes.

Issues of process stability, operation, and yields discourage industrial application. Deepening the

understanding of biomethane potential, continuous anaerobic digester operational parameters, and

co-substrates is key to large-scale implementation. The biomethane potential of different OMW-

derived samples and organic solid market waste as co-substrate was 106–350 NL methane per kg

volatile solids (VS). The highest yields were obtained with the co-substrate and depolyphenolized

OMW mixed with retentate from an ultrafiltration pretreatment. Over 150 days, an anaerobic

fixed-bed 300 L digester was operated with different OMW-derived substrates, including OMW

with selectively reduced polyphenol concentrations. Different combinations of organic loading rate

and hydraulic retention time were set. The biogas yields ranged from 0.97 to 0.99 L of biogas per

g of volatile solids (VS) eliminated, with an average methane content in the produced biogas of

64%. Potential inhibition of the process due to high polyphenol concentrations or over-acidification

through volatile fatty acids was avoided in the continuous process through process and substrate

manipulation. High concentrations of potassium and low concentrations of nitrogen and phosphate

end up in the digestate. Sulfate reduction results in high H

S concentrations in the biogas. The

digestate was tested for phytotoxic properties via the germination index. Diluted digestate samples

improved germination by up to 50%.

Keywords:

olive mill wastewater; biomethane potential; fixed bed biogas digester polyphenols;

germination index

1. Introduction

Olive mill wastewater (OMW) is a highly relevant agricultural waste stream in the

Mediterranean area. More than 30 million m

arise each season worldwide. Tunisia has a

share of about 8% [

]. The challenges related to the management of OMW were one of the

drivers for the conversion of olive mills to more efficient 2-phase decanters that separate

the oil from a wet pomace instead of OMW and dry solids. However, millions of cubic

meters of OMW arise every season in olive oil-producing countries. They are responsible

Appl. Sci. 2023,13, 9613. https://doi.org/10.3390/app13179613 https://www.mdpi.com/journal/applsci

Appl. Sci. 2023,13, 9613 2 of 14

for vast amounts of GHG emissions, contamination of surface and groundwater, and smelly

nuances [2–4].

The anaerobic (co-)digestion of OMW has been researched since the 1980s, however

with limited industrial applications [

–

]. Much like in the case of cow and swine manure

in Central Europe, anaerobic digestion is a suitable valorization path for this important

waste stream. The technology is fully developed and optimized on an industrial scale.

There are several reasons why, with the exception of a few large-scale plants in Italy,

Spain, and Greece, there is no large-scale implementation of anaerobic technology for

the valorization of OMW as the main or co-substrate in agricultural biogas digesters:

(1) The seasonal and uneven occurrence of OMW over the year; (2) the weather dependent

properties of the substrate; (3) the potential inhibition of the biodegradation by high

molecular weight polyphenols or potassium and with its high organic load (COD can

exceed 100 kg

−3

); and (4) the risk of over-acidification due to the low initial pH. The

general feasibility of the process is well-researched and documented. To be a viable

substrate and energy source, the challenges of process stability, yield, co-fermentation, and

digestate utilization need further investigation [

]. Only an in-depth understanding of

inhibition effects, optimal operating parameters, maximum biogas potentials, and holistic

digestate concepts can give plant operators and planners the necessary certainty that

large-scale plants will continue to operate stably when OMW is added.

OMW carries enormous potential regarding nutrients, water, and energetic valoriza-

tion. Cow slurry is now an incremental part of the nutrient supply for agriculture in

Germany. Around 3000 biogas plants produced 4.0 TWh of electricity in 2016 [

]. A sim-

ilar scenario with OMW in the Mediterranean is conceivable if the process control and

inhibition issues are well understood and managed.

The effect of a newly developed polyphenol extraction process for selective removal

of low molecular weight polyphenols inhibiting anaerobic degradation was investigated

in biomethane potential batches and an anaerobic continuous fixed bed digester. All

OMW-derived waste streams of the process (sediment, ultrafiltration retentate, swim layer,

flushing water, and depolyphenolized OMW), as well as a slurry of organic market waste,

were investigated. The phytotoxic properties of the resulting digestate were quantified

using the germination index.

2. Materials and Methods

Olive mill wastewater:

The olive mill wastewater samples were all derived from four

different three-phase or traditional mills in Sfax, Tunisia, in the season ‘22/’23. OMW 1

was collected from the same olive mill on two different dates within one month. Numerous

olive kinds were mixed from many groves, but the most common is Chemlali [

]. The

samples for analysis were taken and stored at 4

◦

C. The substrates for feeding the biogas

reactor were stored outdoors under the sun and rain shelter for several weeks, depending

on the feeding schedule of the biogas reactor.

Depolyphenolized OMW samples were subjected to mechanical pretreatment, ul-

trafiltration, and a selective polyphenol extraction process by adsorption. The extracted

polyphenols were further purified for valorization in industrial applications, such as feed

additives or cosmetics [

]. The residue from the ultrafiltration process was reduced to

roughly 40% of the feed volume. All OMW (derived) samples used for the experiments are

characterized in Table 1.

Appl. Sci. 2023,13, 9613 3 of 14

Table 1. Composition and timetable of the substrates of the anaerobic bioreactor.

OMW Experiment aTotal Polyphenols

b/Hydroxytyrosol c

TOC/

DOC TS/VS COD SO42−K+PO4−2Total

Nitrogen Treatment and Mixture

Unit Experiment

(Days) g·L−1g·L−1g·L−1g·L−1g·L−1g·L−1g·L−1g·L−1-

OMW 1 raw BMP, GI, BOD 3.36/-

37.6/24.0

74.1/60.6 90.3 - - 0.45 0.60 OMW 1 Permeate after adsorption

(to recover polyphenols)

OMW 1retentate

BMP, BOD 2.64/-

59.8/25.0

100.5/86.3 193.3 - - 0.50 1.26 OMW 1 sieved + filtered with

ceramic ultrafiltration membranes

OMW 1 treated I AD: (40),

GI, BOD, BMP 1.4–2.4/0.18–0.54

17.9/16.8

39.1/27.6 49.1 1.3–1.7 4.8 0.31–0.41 0.18

100% 3-phase OMW after membrane

filtration and adsorption

OMW 2_1

diluted II AD: (16) 1.32/-

12.1/10.9

25.0/19.6 40.0 0.8 2.7 0.15 0.11 43% sieved OMW 2 (traditional

method), 57% water

OMW 2_2

diluted III AD: (22) 1.66

15.2/13.6

31.4/24.6 50.2 1.0 3.4 0.19 0.14 53% sieved OMW 2, 47% water

OMW 3treated V AD: (43) 1.56/0.57 18.8/11.9 39.5/29.0 56.6 2.1 3.3 0.27 0.37 75% 3-phase OMW after membrane

filtration and adsorption, 25%

retentate from the membrane

filtration treating OMW

OMW 4treated VI AD: (26) 1.56/0.98 12.2/11.5 24.5/17.7 33.9 0.4 3.0 0.28 0.16

Co-Substrate BMP - - 89.2 d/81.8 d85.0 - - - - Solid organic food-market waste

Inoculum BMP -

12.37/1.86

28.6/17.5 - 0.01 1.76 Digestate from an anaerobic reactor

that is fed with market waste

Digestateday 37 GI 0.68

2.34/2.00

22.3/11.4 8.79 0.19 4.79 0.08 0.14 Digestate of the fixed bed reactor

used for the continuous experiment

Experiment: BOD: biological oxygen demand; BMP: biomethane potential; AD: anaerobic digester; GI: germination index.

: expressed in gallic acid equivalents.

Hydroxytyrosol is

the strongest antioxidant of polyphenols (generally in nature). dg/kg.

Appl. Sci. 2023,13, 9613 4 of 14

Biomethane potential:

An adapted method based on Hollinger et al. [

] and VDI

4630 (2016) was applied to determine the biomethane potential of the treated and untreated

olive mill wastewater. Duplicates of five different samples were measured in 1 L Schott

glass bottles, filled 2/3 (666 mL) with an inoculum-substrate ratio of 1:1. The digestate

of an anaerobic biodigester treating market waste was used as an inoculum and mixed

with three substrates derived from the same OMW sample, as characterized in Table 2. All

samples are flushed with nitrogen prior to the experiment, sealed, and incubated at 37

◦

The produced biogas was measured using the liquid replacement method, utilizing NaOH

to remove CO2.

Table 2. Composition of the BMP samples (all OMW samples derive from OMW 1).

1 Inoculum 2 Raw 3 Depolyphenolized OMW 4 Co-Substrate 5 Retentate

Volume VS Volume VS Volume VS Volume VS Volume VS

Sample Name L g L g L g L g L g

Inoculum 0.35 6.11 0.35 6.11 0.35 6.11 0.35 6.11 0.35 6.11

OMW 0.10 6.11 0.07 4.28

Depolyphenolized

OMW 0.21 6.11 0.10 3.06

Co-Substrate 0.02 1.83

Retentate 0.04 3.06

Water 0.32 0.22 0.11 0.22 0.18

Total 0.67 6.11 0.67 12.23 0.67 12.23 0.67 12.23 0.67 12.23

Anaerobic digester:

The digester used for the upscaled experiments [

] is a 1.95 m

high, double-jacketed, thermostated stainless-steel column with an internal diameter of

0.45 m, resulting in an active fixed bed volume of around 270 L. It is filled with about

70 kg of carriers for biomass retention (Hiflow PVC rings—size 38-1), resulting in a specific

surface of around 200 m

2·

−3

and a gap degree of over 92%. The temperature was kept

at 37

◦

C with a thermostat. The feed was dosed with a BG 600 FJ-S peristaltic pump at

a flow rate of 0.7 L

min

−1

(November–January) and 0.8 L

min

−1

(February). The inflow

pH was adjusted with a 5 mol L

−1

-NaOH solution to pH 6. Biogas and digestate samples

were taken irregularly to determine the methane content by gas chromatography. The

feed substrates and durations are depicted in Table 1. The effects of operating parameters

such as hydraulic retention times and organic loading rates on degradation efficiency and

biomethane production were observed and determined over 150 days.

Germination Index:

The method is based on a modified Zucconi test and DIN EN

16086-2: Soil improvers and growing media—Determination of plant response—Part 2.

Triplicates of 20 seeds are placed into Petri dishes with filter paper in the dish and in the lid.

Lettuce Romaine seeds are used (Variety: NADER, Brand: Enza Zaden, Batch No. 6.121.702,

Germination and Purity: 99%, Origin: South Africa). A total of 5 mL of sample is added per

Petri dish, and 1 mL of distilled water is put on the filter paper in the lid of the Petri dish.

For the control, 5 mL of distilled water was added to the seeds instead of the sample. The

closed Petri dish is kept in a dark place at room temperature until day 5. Every day, the

number of germinated seeds is counted. On the fifth day, the root length of all germinated

seeds was measured. The Germination Index is calculated with the Formula (1):

GI (%) = (NGS·ARLS·NGC

−1×ARLC

−1)·100 (1)

GI: germination index; NG: number of germinated seeds; ARL: average root length; S

refers to the sample; and C refers to the control.

Analytical methods:

The Analytik Jena TOC analyzer multi N/C 3100 (Jena, Ger-

many) was used to determine total organic carbon (TOC), dissolved organic carbon (DOC),

and total nitrogen (TN). For DOC measurements, the sample was filtered with a Whatman

0.45 µm membrane filter (Kent, UK). The chemical oxygen demand (COD) was measured

Appl. Sci. 2023,13, 9613 5 of 14

based on the thermal disintegration of the samples at 1200

◦

C using the QuickCODlab-

03D0318 analyzer and autosampler from LAR Process Analysers AG. The pH and con-

ductivity were measured using an OHAUS Starter 2100 and a Consort C651 sensor. The

Folin–Ciocalteau method was used to determine the polyphenol concentration in gallic

acid equivalents. The following protocol, which was optimized at the Technical University

of Berlin (TUB), was used to prepare 96-well plates: (1) add 90

L ultrapure water; (2) add

L of gallic acid standard or sample diluted to the calibrated concentration; (3) add

L Folin–Ciocalteau reagent; and (4) add 100

L Na

(10.75%). The reaction time

is 30 min in the dark (first 15 min shaking at 100 rpm). The light absorption at 620 nm is

measured. The blank sample was subtracted, and gallic acid equivalents were calculated

based on a linear calibration curve (gallic acid: 800, 400, 200, 100, 50, and 25 mg

−1

). The

dry matter (DM) was determined in accordance with DIN 12,880 by drying the sample

for 24 h and weighing the mass difference. The dried samples were oxidized at 600

◦

for 6 h to remove all organic components and re-weighted to determine the organic dry

matter (VS). To quantify the methane content in the biogas, the sample vials are flushed

for 5 min with nitrogen gas. The vial is then flushed with the produced biogas for at least

5 min and afterward analyzed in a gas chromatography system (Agilent Technologies

7890A GC System with a Bora bond q column and Agilent Technologies G188, Network

Headspace Sampler). Based on the Nordmann method and a titration test adapted by the

German Federal Agricultural Research Center (FAL), the ratio between the volatile fatty

acids and the total alkalic carbonates (FOS/TAC) was measured. A total of 20 mL of sample

is placed on a magnetic stirrer and homogenized consistently. Using the SCHOTT TitroLine

Easy, 0.1 N H

(=0.05 mol

−1

) was titrated until the sample reached pH 5. The time

and amount of acid needed are recorded. The titration is then continued until pH 4.4 is

reached, and the needed time and amount of acid are recorded again. FOS and TAC are

then calculated with Formulas (2) and (3):

TAC = H2SO4-Volume added from start to pH 5 in mL ×250 (2)

FOS = (H2SO4-Volume added from pH 5 to pH 4.4 in mL ×1.66 −0.15) ×50 (3)

Ion concentrations were determined via ion chromatography by Metrohm using a

Metrosep A Supp 17—150/4.0 column for anions and a Metrosep C 4—150/4.0 column

for cations. All samples were filtered and diluted to the calibrated concentrations of

20–100 ppm

. The determination of the biological oxygen demand 30 (BOD

) was carried

out in accordance with DIN EN 1899-2 using Oxi Top Control of WTW for four samples, all

deriving from different treating steps of the OMW. The test was carried out in triplicate

with two different dilutions for each sample. The incubation period was 30 days at a

temperature of 20 ◦C.

3. Results

3.1. Biological Oxygen Demand

For the substrates used in the biomethane potential experiments, the biological oxygen

demand was determined to determine the easily and slowly degradable components

(Figure 1). For all substrates except digestate, an increase in oxygen demand is observed

after day 2, indicating that most readily degradable organics are degraded. At the end

of the measurements, after 30 days, the raw OMW and the retentate curves were still

slightly increasing, indicating that further degradation of poorly degradable organics

was continuing. Membrane filtration and adsorption improve the BOD

to COD ratio

from 1:3.2 (raw) to 1:2.2 (depolyphenolized OMW). In terms of slow-degrading substrates

(BOD

:COD ratio), the ratio for raw OMW was 1:1.6, and for depolyphenolized OMW, it

was 1:1.3. This indicated slow degradation kinetics for many organics in the wastewater.

The retentate has a BOD

to COD ratio of 1:6, with a BOD

to COD ratio of 1:2, which leads

to the conclusion that long retention times or dilution are necessary for the degradation.

Appl. Sci. 2023,13, 9613 6 of 14

As for the digestate, the BOD

to COD ratio is 1:2, indicating that the remaining organics

require more adapted biomass and longer retention times for further degradation [14].

Appl. Sci. 2023, 13, x FOR PEER REVIEW 5 of 14

Network Headspace Sampler). Based on the Nordmann method and a titration test

adapted by the German Federal Agricultural Research Center (FAL), the ratio between the

volatile fatty acids and the total alkalic carbonates (FOS/TAC) was measured. A total of 20

mL of sample is placed on a magnetic stirrer and homogenized consistently. Using the

SCHOTT TitroLine Easy, 0.1 N H2SO4 (=0.05 mol·L−1) was titrated until the sample reached

pH 5. The time and amount of acid needed are recorded. The titration is then continued

until pH 4.4 is reached, and the needed time and amount of acid are recorded again. FOS

and TAC are then calculated with Formulas (2) and (3):

TAC = H2SO4-Volume added from start to pH 5 in mL × 250 (2)

FOS = (H2SO4-Volume added from pH 5 to pH 4.4 in mL × 1.66 − 0.15) × 50 (3)

Ion concentrations were determined via ion chromatography by Metrohm using a

Metrosep A Supp 17—150/4.0 column for anions and a Metrosep C 4—150/4.0 column for

cations. All samples were filtered and diluted to the calibrated concentrations of 20–100

ppm. The determination of the biological oxygen demand 30 (BOD30) was carried out in

accordance with DIN EN 1899-2 using Oxi Top Control of WTW for four samples, all de-

riving from diﬀerent treating steps of the OMW. The test was carried out in triplicate with

two diﬀerent dilutions for each sample. The incubation period was 30 days at a tempera-

ture of 20 °C.

3. Results

3.1. Biological Oxygen Demand

For the substrates used in the biomethane potential experiments, the biological oxy-

gen demand was determined to determine the easily and slowly degradable components

(Figure 1). For all substrates except digestate, an increase in oxygen demand is observed

after day 2, indicating that most readily degradable organics are degraded. At the end of

the measurements, after 30 days, the raw OMW and the retentate curves were still slightly

increasing, indicating that further degradation of poorly degradable organics was contin-

uing. Membrane filtration and adsorption improve the BOD5 to COD ratio from 1:3.2 (raw)

to 1:2.2 (depolyphenolized OMW). In terms of slow-degrading substrates (BOD30:COD

ratio), the ratio for raw OMW was 1:1.6, and for depolyphenolized OMW, it was 1:1.3. This

indicated slow degradation kinetics for many organics in the wastewater. The retentate

has a BOD5 to COD ratio of 1:6, with a BOD30 to COD ratio of 1:2, which leads to the

conclusion that long retention times or dilution are necessary for the degradation. As for

the digestate, the BOD30 to COD ratio is 1:2, indicating that the remaining organics require

more adapted biomass and longer retention times for further degradation [14].

Figure 1. Biological oxygen demand over 30 days of OMW-derived substrates.

3.2. Biomethane Potential

The BMP batches were incubated for 105 days, showing the generally slow kinetics of

the anaerobic degradation. The biogas yield is shown in Figure 2, compared to the inoculum

baseline. Due to different technical issues, three samples were not quantified. The baseline

of the graph is the methane yield of the inoculum at 731

60 mL after 105 days. The

variation in the inoculum BMP results in a margin of error of approximately 16%, which

affects all results in addition to the variations in the BMP samples. Due to differences

in the obtained yields, both batches containing co-substrate are depicted separately. The

highest yields were achieved by mixing the OMW with a slurry of organic market waste

and by depolyphenolized OMW mixed with retentate, both reaching 350 NL methane per

kg of organic dry matter. The volatile solids content was reduced by 6.4 g and 7.8 g VS,

respectively. Initial inhibition was observed for depolyphenolized OMW and raw OMW

with co-substrate, with lower biomethane production rates than the inoculum for 20 and

65 days, respectively. This is unexpected and likely due to the higher-than-expected and

fluctuating biogas production of the inoculum, combined with the error margin of the

measurement method. Raw OMW, raw OMW with co-substrate, and depolyphenolyzed

OMW with retentate showed retarded biogas production, with a second steep increase in

methane production between days 30 and 45. This is likely linked to an adapted microbiome

in the samples after extensive incubation [

]. Using digestate from an anaerobic digester

that is fed with olive mill wastewater is an option to reduce this lag phase. Micoli et al.

indicated that adding biochar is a viable option to stabilize the incubation phase [

]. By

reducing the effect of the degradation of the inoculum on the overall result, increasing the

VS ratio between the inoculum and substrate can increase the accuracy of the experiment.

On day 6, the incubation temperature for all samples was 8 degrees higher than during

the rest of the period due to a technical problem. It is noticeable that on this day, there is a

decrease in methane production for most substrates, which is likely connected to the higher

pressure in the vessel due to the higher temperature. The gas yield was adapted to normal

conditions (273 K and 1.013 bar).

Appl. Sci. 2023,13, 9613 7 of 14

Appl. Sci. 2023, 13, x FOR PEER REVIEW 6 of 14

3.2. Biomethane Potential

The BMP batches were incubated for 105 days, showing the generally slow kinetics

of the anaerobic degradation. The biogas yield is shown in Figure 2, compared to the in-

oculum baseline. Due to diﬀerent technical issues, three samples were not quantified. The

baseline of the graph is the methane yield of the inoculum at 731 ± 60 mL after 105 days.

The variation in the inoculum BMP results in a margin of error of approximately 16%,

which aﬀects all results in addition to the variations in the BMP samples. Due to diﬀer-

ences in the obtained yields, both batches containing co-substrate are depicted separately.

The highest yields were achieved by mixing the OMW with a slurry of organic market

waste and by depolyphenolized OMW mixed with retentate, both reaching 350 NL me-

thane per kg of organic dry matter. The volatile solids content was reduced by 6.4 g and

7.8 g VS, respectively. Initial inhibition was observed for depolyphenolized OMW and

raw OMW with co-substrate, with lower biomethane production rates than the inoculum

for 20 and 65 days, respectively. This is unexpected and likely due to the higher-than-

expected and fluctuating biogas production of the inoculum, combined with the error

margin of the measurement method. Raw OMW, raw OMW with co-substrate, and de-

polyphenolyzed OMW with retentate showed retarded biogas production, with a second

steep increase in methane production between days 30 and 45. This is likely linked to an

adapted microbiome in the samples after extensive incubation [15]. Using digestate from

an anaerobic digester that is fed with olive mill wastewater is an option to reduce this lag

phase. Micoli et al. indicated that adding biochar is a viable option to stabilize the incuba-

tion phase [16]. By reducing the eﬀect of the degradation of the inoculum on the overall

result, increasing the VS ratio between the inoculum and substrate can increase the accu-

racy of the experiment.

On day 6, the incubation temperature for all samples was 8 degrees higher than dur-

ing the rest of the period due to a technical problem. It is noticeable that on this day, there

is a decrease in methane production for most substrates, which is likely connected to the

higher pressure in the vessel due to the higher temperature. The gas yield was adapted to

normal conditions (273 K and 1.013 bar).

Figure 2. Biomethane potential with the inoculum as a baseline.

3.3. Continuous Anaerobic Digestion of Raw and Derived OMW

The fixed-bed biogas reactor was operated over 150 days, utilizing different OMW-

derived substrates (Table 1). To investigate the effects on the control of process stability

and degradation efficiency, the organic loading rates (OLR) and hydraulic retention times

(HRT) were varied (Figure 3). The OLR varied between 1.8 and 5.8 kg COD

−3·

−1

and

the HRT between 10 and 27 days. (Figure 3).

Appl. Sci. 2023, 13, x FOR PEER REVIEW 7 of 14

3.3. Continuous Anaerobic Digestion of Raw and Derived OMW

The fixed-bed biogas reactor was operated over 150 days, utilizing diﬀerent OMW-

derived substrates (Table 1). To investigate the eﬀects on the control of process stability

and degradation eﬃciency, the organic loading rates (OLR) and hydraulic retention times

(HRT) were varied (Figure 3). The OLR varied between 1.8 and 5.8 kg COD·m−3·d−1 and

the HRT between 10 and 27 days. (Figure 3).

Figure 3. Chronological overview of OLR, HRT, and used substrates time periods of the substrates:

I: OMW 1, II: OMW 2_1, III: OMW 2_2, IV: OMW 2_3, V + VI + VII: OMW 3.

Three phases can be distinguished in the substrate: The first phase lasted 40 days,

and exclusively depolyphenolized OMW was fed. In those 40 days, the organic loading

rate doubled from 2 to 4 kg COD·m−3·d−1 in two steps after 2 weeks of stable operation.

From days 41 to 78, diluted OMW was used, with an average OLR of 3.1 kg COD m−3 d−1.

Tap water was added to decrease the viscosity and concentration of potentially inhibiting

monomeric polyphenols. The third and longest phase lasted from days 79 to 148 (phases

V–VII, with substrates derived from OMW3 in the depolyphenolization process). The mix-

ing ratio of the diﬀerent OMW substrates is the same as that produced during polyphenol

extraction (25–45% retentate and 55–75% depolyphenolized OMW).

Figure 4 shows the daily biogas production and the measured methane content. The

horizontal lines indicate the average biogas production for diﬀerent OLRs, displaying a

clear correlation. An increase in the produced gas volume is visible with increasing OLR.

The performance of the reactor was stable between days 77 and 86, but the biogas produc-

tion was not recorded due to a technical issue with the biogas meter. The measured me-

thane content varies between 51 and 78%, with an average of 64%. The obtained yields

vary between 0.97 and 0.99 L biogas·(g VSelim.)−1 or 0.44–0.5 L biogas·(g CODelim.)−1 (0.4–

0.46 L biogas ·(g CODelim.)−1 at standard conditions). The methane yield is 0.29–0.33 L

CH4·(g CODelim.)−1 (0.26–0.30 L CH4·(g CODelim.)−1 at standard conditions). This corre-

sponds to 76–86% of the theoretical methane maximum that can be obtained (0.38 L·g COD

at 25 °C [17]). The yields for the period with the inflow derived from OMW 3 (0.33–0.34 L

CH4·(g CODelim.)−1) cannot be considered due to the inhomogeneous substrate. In the first

days, the reactor eﬄuent still contained digestate from the previous influent, which had

lower COD and VS values. This leads to an overestimation of the yield obtained in the

first days, which is not considered in the evaluation.

VII

Figure 3.

Chronological overview of OLR, HRT, and used substrates time periods of the substrates: I:

OMW 1, II: OMW 2_1, III: OMW 2_2, IV: OMW 2_3, V + VI + VII: OMW 3.

Three phases can be distinguished in the substrate: The first phase lasted 40 days,

and exclusively depolyphenolized OMW was fed. In those 40 days, the organic loading

rate doubled from 2 to 4 kg COD

−3·

−1

in two steps after 2 weeks of stable operation.

From days 41 to 78, diluted OMW was used, with an average OLR of

3.1 kg COD m−3d−1

Tap water was added to decrease the viscosity and concentration of potentially inhibit-

ing monomeric polyphenols. The third and longest phase lasted from days 79 to 148

(

phases V–VII

, with substrates derived from OMW3 in the depolyphenolization process).

The mixing ratio of the different OMW substrates is the same as that produced during

polyphenol extraction (25–45% retentate and 55–75% depolyphenolized OMW).

Figure 4shows the daily biogas production and the measured methane content. The

horizontal lines indicate the average biogas production for different OLRs, displaying

a clear correlation. An increase in the produced gas volume is visible with increasing

Appl. Sci. 2023,13, 9613 8 of 14

OLR. The performance of the reactor was stable between days 77 and 86, but the bio-

gas production was not recorded due to a technical issue with the biogas meter. The

measured methane content varies between 51 and 78%, with an average of 64%. The

obtained yields vary between 0.97 and 0.99 L biogas

(g VS

elim

−1

or 0.44–0.5 L biogas

COD

elim

−1

(0.4–0.46 L biogas

(g COD

elim

−1

at standard conditions). The methane yield

0.29–0.33 L CH4·(g CODelim.)−1

(0.26–0.30 L CH

4·

(g COD

elim

−1

at standard conditions).

This corresponds to 76–86% of the theoretical methane maximum that can be obtained

(0.38 L

g COD at 25

◦

C [

]). The yields for the period with the inflow derived from

OMW 3 (0.33–0.34 L CH

4·

(g COD

elim

−1

) cannot be considered due to the inhomogeneous

substrate. In the first days, the reactor effluent still contained digestate from the previous

influent, which had lower COD and VS values. This leads to an overestimation of the yield

obtained in the first days, which is not considered in the evaluation.

Appl. Sci. 2023, 13, x FOR PEER REVIEW 8 of 14

Figure 4. Daily biogas production, methane content, and average daily biogas production for each

OLR phase are depicted in Figure 3.

The FOS/TAC ratio is a practical indicator that can be determined with simple labor-

atory equipment and is used by many operators of biogas digesters to determine the bio-

chemical state of the digesters. The optimum is between 0.3 and 0.4. The pH of the sub-

strate was adjusted to 6 for the first 70 days of the experiment using a 20% NaOH solution

(5 mol L−1) to prevent over-acidification. After the pH adjustment of the feedstock was

terminated, the FOS/TAC increased. The data were generated only until day 96 of the ex-

periment, with a value of 0.38, slightly over the optimum. The microbiome was stable and

able to treat the acidic OMW without pH adjustment, producing a digestate with a pH

between 7.0 and 7.8 (Figure 5). Lower HRT and higher OLR resulted in higher hydrolysis

and acidification. The stable pH at the outflow indicates that higher OLRs are likely feasi-

ble without overacidification [18,19].

Figure 5. FOS, TAC, FOS/TAC, and pH feed and outflow. The grey bar indicates an ideal FOS/TAC

value between 0.2 and 0.3.

Figure 6 shows the VS feed and outflow concentrations together with the resulting

degradation eﬃciencies. The average removal of VS was 71%. A set of outliers with deg-

radation rates below 60% are partly due to hardware issues and are also due to imperfect

sampling and analytical errors. Generally, a decrease in reaction velocity is possible due

to the inhibitory character of raw OMW that was fed in the first 41 days. The DM/VS ratio

Figure 4.

Daily biogas production, methane content, and average daily biogas production for each

OLR phase are depicted in Figure 3.

The FOS/TAC ratio is a practical indicator that can be determined with simple lab-

oratory equipment and is used by many operators of biogas digesters to determine the

biochemical state of the digesters. The optimum is between 0.3 and 0.4. The pH of the

substrate was adjusted to 6 for the first 70 days of the experiment using a 20% NaOH

solution (5 mol L

−1

) to prevent over-acidification. After the pH adjustment of the feedstock

was terminated, the FOS/TAC increased. The data were generated only until day 96 of the

experiment, with a value of 0.38, slightly over the optimum. The microbiome was stable

and able to treat the acidic OMW without pH adjustment, producing a digestate with a pH

between 7.0 and 7.8 (Figure 5). Lower HRT and higher OLR resulted in higher hydrolysis

and acidification. The stable pH at the outflow indicates that higher OLRs are likely feasible

without overacidification [18,19].

Figure 6shows the VS feed and outflow concentrations together with the resulting

degradation efficiencies. The average removal of VS was 71%. A set of outliers with

degradation rates below 60% are partly due to hardware issues and are also due to imperfect

sampling and analytical errors. Generally, a decrease in reaction velocity is possible due

to the inhibitory character of raw OMW that was fed in the first 41 days. The DM/VS

ratio of 57% in the digestate indicates advanced mineralization of organic carbon. The

ratio TOC:VS:COD of the feed averaged 2:3:6. After the anaerobic digester, this changed to

1:3:4, indicating that the COD is only partially related to organic carbon and likely to the

non-carbon portion of the VS.

Appl. Sci. 2023,13, 9613 9 of 14