Chemical Engineering & Processing: Process Intensification 129 (2018) 171–180

Contents lists available at ScienceDirect

Chemical Engineering & Processing: Process Intensification

journal homepage: www.elsevier.com/locate/cep

The impact of hydraulic retention time and operating temperature on T

biofuel production and process wastewater treatment
⁎
Haider Al-Rubayea, Joseph D. Smitha, , Manohar M. Shivashankaraiaha, Jia Yua,
Shruti Karambelkara, Mahyar Ghorbanianb
a
Chemical and Biochemical Engineering Department, Missouri University of Science and Technology, Rolla, MO, 65409, USA
b
Research and Development, LG&E and KU Energy, Louisville, KY, 40202, USA

A R T I C LE I N FO A B S T R A C T

Keywords: Breweries wastewater containing high concentration of organic and inorganic compounds ranks them among the
Anaerobic digestion top pollution generating industries. Anaerobic wastewater treatment with high organic loading rates can be
Hydraulic retention time (HRT) achieved with lower COD strength at higher flowrates using a two–stage expanded granular sludge bed reactor.
Organic loading rate (OLR) Hydraulic retention time (HRT), pH, temperature, and COD strength were varied for process optimization.
Chemical oxygen demand (COD)
Brewery wastewater with 20, 30, and 40 g COD/L as a substrate for two temperature ranges were evaluated.
Under mesophilic conditions (36 °C), results show COD removal efficiency (R%) and biogas production rate
increased by 6% and 40% respectively as HRTs increased, maintaining a constant OLR. Results imply for
equivalent OLRs, better reactor performance is achieved when running high concentration COD at slower rate
compared with a lower concentration at higher rate. This implies diffusion limitation where complex proteins
and fats are passed through the reactor faster than their metabolism rate in the digester. Under thermophilic
conditions (50 °C), results show COD removal efficiency (R%) and biogas production rate increased by 4% and
40% respectively as the HRTs increased, while maintaining a constant OLR. This implies the higher and stronger
population of anaerobes are present under thermophilic condition rather than mesophilic condition.

1. Introduction environment [6].

The anaerobic digestion process (AD) as a treatment for wastewater
For developing countries, coal, natural gas, and crude oil (fossil has been used since the 19th century [7],[8]. It has been used for biogas
fuels) considered the cheapest sources for energy. But, it also con- production as well [7], which can be useful for heating and power
sidered as the main resources for the greenhouse gases (GHG) which generation [9],[10]. Liquor industries generate a large amount of
has a tremendous effect on the global warming. Biofuels can be an es- wastewater that is discharged to the sewer system [11–14]. This was-
sential alternative fuel that could help reducing the climate change and tewater contains enormous amounts of carbon which can be assessed by
pollution due to use the fossil fuel. Besides that, designing a bio-refinery measuring the chemical oxygen demand (COD). COD is the amount of
that could help changing the organics and biomass feed stocks in to oxygen required to oxidize the wastewater in a specific volume [12].
usable liquid and gas fuel and selling biofuels in a market dominated by Environmental agencies and government issued regulations can
low-cost fossil fuels are key challenges to establish vigorous biofuel reduce the discharge of these pollutants to natural water resources and
industry [1–5]. start imposing charges on companies that continue discharging without
Due to the enormous growth in the brewery and distillery in- any pollutant reducing treatment [15]. For example, in the production
dustries, massive quantities of process wastewater (WW) have been of wine, an anaerobic digestion process as a treatment for the waste-
discharged to sewer systems without any treatment. This wastewater water before discharging it to the sewer system would be beneficial. In
contains many organic, inorganic, and solid compounds that pollute the the United States, anaerobic digester (AD) can be used to treat food

Abbreviations: WW, waste water; AD, anaerobic digestion; COD, chemical oxygen demand; AFR, anaerobic fluidized reactor; ABR, anaerobic baffled reactor; P&ID, process and
instrumentation diagram; PFD, process flow diagram; TCs, thermocouples; VFAs, volatile fatty acids; VSS, volatile suspended solids; TSS, total suspended solids; TS, total solids; TDS, total
dissolved solids; PA, pre-acidification; EGSB, expanded granular sludge bed; CH3COOH, acetic acid; NaOH, sodium hydroxide; NaHCO3, sodium bicarbonate; Effluent, the wastewater
flowing out of the system; Influent, the wastewater flow in to the system; Hrs, hours; gal, gallons; min, minute
⁎
Corresponding author.
E-mail addresses: [email protected] (H. Al-Rubaye), [email protected] (J.D. Smith), [email protected] (M.M. Shivashankaraiah), [email protected] (J. Yu),
[email protected] (S. Karambelkar), [email protected] (M. Ghorbanian).

https://doi.org/10.1016/j.cep.2018.04.036
Received 18 February 2018; Received in revised form 30 April 2018; Accepted 30 April 2018
Available online 03 May 2018
0255-2701/ © 2018 Elsevier B.V. All rights reserved.
H. Al-Rubaye et al. Chemical Engineering & Processing: Process Intensification 129 (2018) 171–180

wastes as a co-digestion besides the WW treatment. temperature ranges. The mesophilic temperature range is the operating
There are two main types of AD, conventional and high-rate [10], temperature for the AD and its range is between 30–38 °C, while the
[16]. The conventional AD (e.g., a continuous stirred tank reactor thermophilic temperature range is between 48–55 °C.
[CSTR]) can operate as a batch reactor where all four steps that re- The AD steps are hydrolysis, acidogenesis, acetogenesis, and me-
present the AD process occur in one stage, so there is no input and the thanogenesis, and each one involves microorganisms to carry out the
only output would be the biogas stream. This CSTR system can also run step. The hydrolysis step can be defined as a conversion of the organic
as a continuous process. This type of AD reactor requires a significant polymers (undissolved) like proteins, fats, and sugar into smaller
footprint for the pilot plant besides, the acclimatization time required compounds, such as amino acids, long chain fatty acids, and glucose.
for the microorganisms to start producing biogas would be in orders of This conversion takes place by extracellular enzymes and mainly de-
days (10–20 days) [17]. Since the CSTR reactor includes a mixer, so the pends on the pH of the substrates and the presence of the organic
hydraulic retention time (HRT) and the solid retention time will be compounds [18]. The second step is acidogenesis, which simply con-
equal. Also due to the mixing process, the possibility of flushing the verts the monomers into volatile fatty acids such as propionic acid,
substrate out of the reactor before it got degraded is high. butyric acid, and acetic acid. The third step is acetogenesis, where the
The second type of AD is the high-rate AD process which, as the volatile fatty acids get converted into acetic acid, hydrogen, and carbon
name implies, generates a higher rate of biogas within a shorter hy- dioxide [15]. The last step is the methanogenesis. In this step, there are
draulic retention time that used for the conventional reactor. A two types of microorganisms. The acetoclastic methanogens convert
common part in the high-rate AD systems is the internals, which are acetates and carbon dioxide into methane, and the hydrogenotrophic
either support internals for the microorganisms like natural zeolite in methanogens convert hydrogen and carbon dioxide into methane [26].
the anaerobic fluidized reactor (AFR) [12], or a particulate sludge The literature available neither explained the effect of utilizing
called the biomass granular particles where the microorganisms reside different high COD concentrations, HRTs, and OLRs [12], nor studied
[7],[10],[18]. the temperature ranges effect in a high rate AD reactor.
The first high rate AD reactor configuration developed was the up- In the present investigation, there was a need to study the effect of
flow anaerobic sludge blanket reactor by Lettinga in 1970 [19,20]. This the (HRT) and high COD concentration of the substrate on biogas
reactor consists of cylindrical vessel and contains the biomass granular production with a wide range OLRs, running under mesophilic and
particles that have all the microorganisms for AD process. The waste- thermophilic temperature ranges on the same reactor configuration,
water is flow through the biomass particles and the microorganisms will which will give a better understanding for the temperature factor. A
start feeding on the wastewater and generate the biogas. distillery WW with different COD strengths (20, 30, and 40 g COD/L)
Recently, the expanded granular sludge bed reactor (EGSB) was was utilized in an expanded granular sludge bed reactor. Investigating
developed, which is considered as a modified version of the UASB. The these parameters in EGSB reactor configuration would be really im-
difference between the UASB and the EGSB reactor are the upward portant since the HRT for it are shorter than that in other AD reactor
velocity onside the reactor will be much higher for the EGSB reactor. configurations.
This high rate velocity permits partial bed expansion (semi fluidization)
of the biomass granular particles. This high rate upward velocity will 2. Materials and methods
results a shorter hydraulic retention time of the substrate to spend in-
side the reactor. Also, this expansion can enhance the contact between 2.1. Experimental setup
the substrate and the biomass granular particles and the hydraulic
mixing as well. This high substrate velocity will occurred by recycling Figs. 1 and 2 show the schematic diagram and the piping and in-
part of the effluent into the system [21,22]. Since the hydraulic re- strumentation diagram (P&ID) for the anaerobic digestion process, re-
tention time for the EGSB reactor is shorter than that for other AD re- spectively. An expanded granular sludge bed reactor was designed,
actor configurations, some of the substrates utilized could possibly pass built, and utilized for wastewater treatment. The design was done by
through to the effluent stream before it get the chance to be degraded as using Solid Works software before proceeding with building the setup,
it needs longer time. Because of that, the EGSB AD used for low and which is shown in Fig. 3. It consists of two stages: the pre-acidification
medium strength substrates. stage, where the hydrolysis and acidogenesis steps occur, and the main
The AD process depends on the diffusion of the wastewater into the reactor, where acetogenesis and methanogenesis take place to produce
biomass where the reactions occur to consume the organic content and biogas. The reason for conducting the treatment in two stages is to
generate the biogas, and this should happen under a low suspended increase the stability of the process and optimizing the conditions of the
solids level. Various investigations have been conducted on the AD. process, where the growth rate for the acidogenesis is higher than the
Ghorbanian [6],[15] studied the impact of the hydraulic retention time acetogens. That occurs if the OLR increases, eventually accumulation of
on the biogas rate and composition. He found that for an equivalent the volatile fatty acids (VFAs) will occur in the process, causing in-
organic loading rate (OLR), running an AD of high strength COD at a hibition [6],[7],[24],[25]. The VFAs like butyric acid, propionic acid,
lower rate was much more beneficial than running a low strength COD and acetic acid are the intermediate products after the acidogenesis
at a higher rate, which also resulted in longer retention time, higher step.
biogas production and higher COD removal rate [6]. Montalvo [12] A distillery WW, the residue from the distillation process, was
studied the red wine wastewater treatment by an anaerobic fluidized picked. A couple of analyses were conducted to estimate the organic
bed reactor using zeolite particles. A COD removal efficiency (R%) over matter in the WW expressed by the COD. After determining the raw
80% was achieved, and the volatile fatty acids were less than 400 mg/L. COD strength, three different COD concentrations of WW were used
Budiyono [23] also studied the impact of the slaughterhouse WW on (20, 30, and 40 g COD/L). The WW at a specific COD strength was kept
biogas production and found that the WW has the potency for produ- in the pre-acidification (PA) stage. The flowrate in the pre-acidification
cing a total amount of biogas of 2.472 m3/m3. tank was controlled by a floating valve installed inside the PA tank.
For simulation, Serrano [24] developed the first model for the The PA tank is a stainless-steel tank with a volume of 33 gal, and
anaerobic digestion process using Aspen Plus. The model considered the 18 gal was the active volume used for the wastewater. A mixer (marine-
substrate to be hydrolyzed and thus is comprised of three steps that are type mixer) made of stainless steel material was installed in the PA tank
separated into two stages to account for the optimum pH value for each and connected to electrical motor to maintain uniform temperature and
step. Al-Rubaye [25] recently developed a full model for the AD process mass distribution along with an immersed heater to provide heat to the
using Aspen Plus. This model has been designed to accommodate di- reactor. This mixer is 12” in diameter and was installed vertically. The
verse types of substrates under mesophilic and thermophilic electric motor was connected to propeller by a belt, was energized by

172
H. Al-Rubaye et al. Chemical Engineering & Processing: Process Intensification 129 (2018) 171–180

Fig. 1. Schematic Diagram for the Two Stages Anaerobic Digestion Process.

the temperature controller besides the heating element. A heavy-duty maintain the activity of the biomass. It contained mineral base I, mi-
pH sensor from Omega was connected to a pH controller to maintain neral base II, nutrient base, and a buffer base [21]. Table 1 shows the
the required pH level by adding 1 N of sodium hydroxide (NaOH) using composition of the nutrients added to the EGSB to maintain the activity
a Milwaukee MC122 pH meter with an automatic peristaltic pump. In of the biomass granular particles.
the pre-acidification (PA) reactor, large polymers or molecules were The EGSB is 12 gal in volume and it consisted of three main parts,
degraded. Besides that, the pH, nutrients, and temperature were ad- the aluminum plenum, which is the lower part where the WW first is
justed to the required levels. Before running the system, the WW was introduced to the EGSB, a T-shaped pipe installed in the plenum to
kept in the PA for 24–48 h before it was pumped into the main reactor distribute the liquid evenly in the reactor, and a gas diffuser installed
using the BT600s basic variable speed peristaltic pump (Golander). inside the plenum to inject nitrogen gas into the reactor to ensure there
Processed wastewater from PA was pumped based on a range of OLRs is no trapping of air inside in order to maintain a condition conducive
varying from OLR 2.0 g COD/L per day to OLR 9 g COD/L per day. for the AD process to occur. A liquid distributor was attached at the top
Nutrient medium and mineral base were added on a regular basis to of the plenum containing about 171 holes of 2 mm ID to support the

Fig. 2. Process and Instrumentation Diagram for the Anaerobic Digestion.

173
H. Al-Rubaye et al. Chemical Engineering & Processing: Process Intensification 129 (2018) 171–180

Fig. 3. Two-Stage Anaerobic Digestion Process.

Table 1 A temperature controller was installed with the heating tank to

Nutrient Medium Composition. provide the required heat to the reactor. Eight thermocouples were
Medium Component Quantity (mg/mL) connected to a data logger used to measure temperatures for the bio-
mass, the substrate, the effluent, the hot water inlet (to the jacket), and
Mineral Base I Cobalt (Co) 0.062 the hot water outlet (of the reactor). One thermocouple was used to
Iron (Fe) 1.126 measure the temperature in the heating tank. The thermocouple types
Manganese (Mn) 0.0139
Boron (B) 0.0044
were J and T. The pressure inside the reactor was monitored using an
Zinc (Zn) 0.0119 Omega pressure transducer PX-304 with the indicator DP-350.
Molybdenum (Mo) 0.0020
Nickel (Ni) 0.0062
Selenium (Se) 0.0104 2.2. Wastewater and granular biomass particle characteristics
Copper (Cu) 0.0026

Mineral Base II Calcium (Ca) 5.4 The biomass used in the process was obtained from Anheuser-Busch
Magnesium (Mg) 2.36
Beverage Company with a particle diameter of 2–5 mm. A couple of
Nutrient Base Nitrogen (N) 13.9
Phosphorus (P) 11.4 analyses were done on the biomass, such as volatile suspended solids
Sulphur (S) 6.76 VSS, total solids TS, total dissolved solids TDS, and the total suspended
Buffer Base Sodium Bicarbonate (NaHCO3) 40 solids TSS. The protocol followed for these analyses was made by U.S.
Geological Survey (USGS, 1989) [15]. For the biomass granular parti-
cles, the VSS was 63,914 mg VSS/L, the TSS was 358 mg/L, and the TDS
biomass particles and distribute the substrate wastewater uniformly was 2399 mg/L. The particle size and the pH were 2–5 mm and 6.9–7.2,
inside the reactor. respectively. The wastewater was collected from a distillery Company
Above the plenum is the main part of the EGSB, where the biomass in Missouri. Simply, it was the stillage waste obtained after a distillation
granular particles reside. The main part of the reactor was made of an process and was mixed with water during the cleaning of the distillation
acrylic material with a diameter of 7.5 in. and 64 in. in height, sur- vessel. Similar analyses were done for the wastewater to check its
rounded by a jacket with a diameter of 11 in., about 1.5 in. of clearance characterization before treatment. The VSS), TSS), TDS, and the pH
was the annular space where hot water flow to maintain a constant were 25 mg/L, 1118 mg/L, 70,626 mg/L, and 3–4, respectively. All
temperature in the reactor. The hot water jacket was used to maintain these analyses were conducted based on the protocols from the USGS
the temperature inside the reactor, as it further helps to avoid the direct report. A laboratory spectrophotometer (HACH model DR3900) was
contact of the heating element with the biomass particles. The active used to measure the COD concentration, the volatile fatty acids, the
volume of the reactor was about 12 gallons. alkalinity, total nitrogen, total ammonia, phosphorus, and sulfate. The
The upper part of the main reactor, the gas-liquid-solid separator, vials used for these analyses were TNT 823, TNT 872, TNT 870, TNT
was also made of acrylic materials. The main objective of this part was 828, TNT 833, TNT 845, and TNT 865, respectively. The COD con-
to separate the generated biogas from the effluent and retain the bio- centration of the wastewater was approximately 90 g COD/L, which
mass inside the reactor. Part of the effluent was recycled back to the was diluted with tap water before being fed to the reactor to get the
main reactor by using the second peristaltic pump at a rate of 30% of required COD strength. The waste water was diluted with dilution
the fresh substrate feed, while the rest was discharged to the sewer factor of 4.5, 3, and 2.25 to get 20,000, 30,000, and 40,000 mg COD/L,
system. Lastly, a water heating tank of 23 gal was used to maintain the respectively. The effluent COD concentration was analyzed to monitor
temperature of the EGSB reactor. Two centrifugal pumps (one was the process stability and calculate the COD removal efficiency. For the
working while the other one was on standby) at ½ horsepower were biogas production rate measurement, a water displacement method was
connected to a timer power switch to make each pump run for 30 min at used to measure the amount of generated biogas. Its composition was
a time. tested by two methods: an analytical protocol by using potassium

174
H. Al-Rubaye et al. Chemical Engineering & Processing: Process Intensification 129 (2018) 171–180

hydroxide protocol by Young et al. [27], and by using gas chromato- 3. Results and discussion
graphy (Varian CP-3800).
3.1. Reactor operating under mesophilic temperature range

2.3. Remediation process after inhibition situation The main advantages of using a high-rate anaerobic digestion
system includes its small footprint and the high OLR that can be
Due to the inhibition, the alkalinity level inside the reactor dropped achieved. COD strengths of 20, 30, and 40 g COD/L were used. Based on
below 100 mg CaCO3/l, while it was above 1500 mg CaCO3/l during that and the reactor’s volume, nine different OLRs (2, 2.5, 3, 4, 5, 6, 7,
steady state operation. According to literature, the ratio of the VFA 8, and 9 g COD/L.d) were tested under the mesophilic temperature
concentration to the process alkalinity should be below 0.5 to confirm range (PA reactor’s temperature was 35 °C with the EGSB running at
the stability of the process [1],[28]. The remediation process facilitates 36 °C). Based on experimental experience and literatures’ results
harvesting the liquid medium of biomass particles and replacing it with [29–31], the selection of the COD was based on the VSS concentration
fresh water. This process is accompanied with injecting a sodium bi- of the biomass granular particles, where the ratio of COD to VSS was
carbonate solution to increase the system alkalinity [6],[15]. The fresh less than 1, which ensures stable operation of the AD system. The in-
liquid medium used is harvested from existing biomass kept on standby. crease in the OLR should be less than 50% of the previous value so it
The HRT for injecting the liquid medium inside the reactor should be does not shock the biomass [32].
low enough to retain all the microorganisms and the biomass inside the In the PA reactor, some of the organic matter was converted into
reactor. The alkalinity of the process is usually monitored every day in volatile fatty acids (VFAs), like acetic acid, propionic acid, and butyric
addition to the COD and the VFA from the effluent stream. The pH is acid [15]. The percentage conversion in the PA was calculated using a
also monitored every day and a sodium hydroxide solution 1 N was formula called the degree of the pre-acidification [33]:
added gradually when needed to increase the pH level inside the re-
actor. Once the pH level reaches about 7.0, the alkalinity is above CODPA ⎞
Pre−acidification degree (PA) = ⎛1− ⎜ * 100 ⎟

1500 mg CaCO3/L, and both COD and VFA for the effluent are about ⎝ CODin ⎠ (1)
2000 g COD/L, and 500 mg CH3COOH/L, respectively, and the reactor
is ready for wastewater treatment again. where CODin is the COD concentration of the feed introduced into the
This investigation was ran at equivalent OLR. Maintaining a con- pre-acidification tank and CODPA is the remaining COD concentration
stant OLR was done either by letting high substrate concentration flow after the pre-acidification stage. Due to the ongoing flow of the sub-
at lower rate or lower substrate concentration flow at higher rate. The strate into and out of the PA reactor based on the OLR of the system, the
OLR values that tested were 2, 3, 4, 5, 6, 7, and 8 g COD/L.day, and the wastewater had a variable retention time between 5–24 days. As for the
substrate concentration used were 20, 30, and 40 g COD/L. monitoring OLR of 2 g COD/L.d, the retention time in the PA reactor was 24 days,
the performance of the system was done by monitoring and analyzing while for OLR of 8 g COD/L.d, the retention time was about 7 days. The
the effluent, the biogas production and analysis, and checking the HRT applied on the PA reactor will be based on the flowrate of the
acidity level of the system. The means, standard deviation, and coeffi- substrate into the EGSB reactor. Table 3 shows the pre-acidification
cient of variations (the ration of the standard deviation to the mean) degrees and the hydraulic retention time for all the COD strengths
were calculated for each parameter recorded. The operating parameters tested. Fig. 4 shows the relationship between the HRT for the EGSB and
used (OLRs, COD strengths) with the associated HRT and the substrates PA% in the PA reactor. The PA% for the high COD strength (40 g COD/
flow are shown in Table 2. l) was the highest based on the HRT, as the substrate spent longer time
in the PA reactor than the (30 g COD/l). The PA% for 30 g COD/l
strength was higher than that obtained for the 20 g COD/l. That means
more volatile fatty acids were produced for the 40 g COD/l than were
Table 2 produced for the 30 g COD/l, and the volatile fatty acids formation for
Operating Parameters Used in the Investigation. the 30 g COD/l was higher than that produced for 20 g COD/l.
Organic COD Strength, Hydraulic Substrate Recycle The HRT will be connected to the volumetric flowrate of the sub-
Loading Rate, mg COD/L Retention Time, Feed, L/day Rate, L/day strate that flows from the PA reactor to the EGSB. The higher the HRT,
mg COD/L.day day the lower the flowrate of the substrate and the more VFAs produced in
the PA reactor.
2,000 20,000 10 4.5 1.35
30,000 15 3 0.9 Using two-stage AD is more beneficial in stabilizing the system,
40,000 20 2.25 0.675 especially if there is an OLR shock, which could cause an inhibition in
3,000 20,000 6.666 6.75 2.025
the system as the acetogens grow at a lower rate than the acidogens
30,000 10 4.5 1.35 [33]. Tables 4a, 4b and 4c , show the substrate concentration based on
40,000 13.333 3.375 1.0125 the COD concentration for different OLRs for the influent (i.e., influent
4,000 20,000 5 9 2.7 is the substrate feed into the PA reactor), the PA (i.e., the substrate feed
30,000 7.5 6 1.8
40,000 10 4.5 1.35 Table 3
5,000 20,000 4 11.25 3.375 Pre-acidification Degree Values and the Hydraulic Retention Time for COD
30,000 6 7.5 2.25 Strengths 20, 30, and 40 mg COD/l.
40,000 8 5.62 1.686
COD = 20 g COD/l COD = 30 g COD/l COD = 40 g COD/l
6,000 20,000 3.333 13.5 4.05
30,000 5 9 2.7 HRT PA% HRT PA% HRT PA%
40,000 6.666 6.75 2.025
10 32.482 15 50.323 20 63.741
7,000 20,000 2.85 15.78 4.734
6.66 27.587 10 51.283 13.33 63.836
30,000 4.28 10.51 3.153
5 27.413 7.5 50.347 10 63.359
40,000 5.71 7.88 2.364
4 26.456 6 51.157 8 63.789
8,000 20,000 2.5 18 5.4 3.33 23.2 5 51.522 6.66 63.965
30,000 3.75 12 3.6 2.85 27.512 4.285 49.862 5.7 63.872
40,000 5 8 2.4 2.5 25.582 3.75 49.841 5 63.852

175
H. Al-Rubaye et al. Chemical Engineering & Processing: Process Intensification 129 (2018) 171–180

Fig. 5. Effluent pH Variation with the Organic Loading Rate at different HRTs
for COD Strength of 30 g COD/L.
Fig. 4. Pre-Acidification Degree in PA Reactor Relation with Hydraulic
Retention Time (HRT).
out of the PA tank), and the effluent (i.e., the remaining substrate out of
the system) streams’ substrate.
Table 4A
From Tables 4a, 4b and 4c, it can be shown that the effluent COD
COD Concentration for the AD Streams and the Removal Efficiency for
20,000 mg/l COD Strength.
concentration does not follow a specific trend with increasing OLR for
all COD strengths that tested. That means the R% also does not change
OLR, mg 20 g COD/l with the increase in OLR and it will remain within a close range, but it
COD/
L.day HRT, day Influent, mg PA, mg Effluent, mg R%
got improved with the COD strength increase, which is similar to what
COD/L COD/L COD/L was found in the literature.
For example, Ghorbanian [6] in his investigation found that the
2,000 10 20,000 13103.67 1601 91.995 total chemical R% will increase significantly with increase the hy-
3,000 6.66 20,000 14482.67 1583.41 92.08
draulic retention time but not with the OLR changing. The pH level in
4,000 5 20,000 15557.33 1512.7 92.43
5,000 4 20,000 14517.33 1558.17 92.209 the PA tank was adjusted to 4.9–5.2 by adding sodium hydroxide 1 N.
6,000 3.33 20,000 15360 1472.54 92.6373 As it seen in Fig. 5, for COD strength of (30 g COD/L), the pH value
7,000 2.85 20,000 14497.67 1213.333 93.93333 for the effluent didn’t affect that much by the increase in the OLR. The
8,000 2.5 20,000 14883.67 1279 93.605
pH of the effluent was about 7.0–7.4 for the OLRs from 2 to 8 g COD/
L.d, but under the highest OLR (9 g COD/L.d), the pH value decreased
∼5.5 as the rate of the acidogenic microorganisms crossed higher than
Table 4B
that required for the acetogens. The VFA concentration will be higher
COD Concentration for the AD Streams and the Removal Efficiency for
30,000 mg/l COD Strength. and cause a drop in the pH level inside the EGSB, resulting in an in-
hibition inside the main reactor [12]. This trend is similar to what
OLR, mg 30 g COD/l
found in the literature [12].
COD/
L.day HRT, day Influent, mg PA, mg Effluent, mg R%
As shown in Fig. 6, for COD strength 30 g COD/l, the R% was in-
COD/L COD/L COD/L creased with the increase in HRT; however, with higher OLR (i.e.,
OLR = 9 g COD/l.d), the R% decreased. For the 30 g COD/l, the R%
2,000 15 30,000 14,903 2084.667 93.05111 was increased from 93% to 97% for OLR from 2 g COD/L.d to 7 g COD/
3,000 10 30,000 14367 2159 92.80333
4,000 7.5 30,000 14055.33 2230 92.56667
L.d respectively; then, it decreased to 95% at OLR 8 g COD/L.d. At OLR
5,000 6 30,000 14615 1629.25 94.56917 9 g COD/L.d, the R% decreased dramatically to 64%. This decrease was
6,000 5 30,000 14655 1394.75 95.35083
7,000 4.28 30,000 14386.25 1231.25 95.89583
8,000 3.75 30,000 14,167.25 1646 94.51333

Table 4C
COD Concentration for the AD Streams and the Removal Efficiency for
40,000 mg/l COD Strength.
OLR, mg 40 g COD/l
COD/
L.day HRT, day Influent, mg PA, mg Effluent, mg R%
COD/L COD/L COD/L

2,000 20 40,000 14503.67 1406 96.485

3,000 13.33 40,000 14465.67 1588 96.03
4,000 10 40,000 14656.33 1619.667 95.95083
5,000 8 40,000 14484.33 1346.333 96.63417
6,000 6.66 40,000 14414 1517 96.2075
7,000 5.71 40,000 14451.33 1592 96.02
8,000 5 40,000 14459.33 1380.667 96.54833
Fig. 6. COD Removal Efficiency Variation with the Organic Loading rate within
The Experimentation Time for COD Strength of 30 g COD/L.

176
H. Al-Rubaye et al. Chemical Engineering & Processing: Process Intensification 129 (2018) 171–180

Fig. 7. Variation of the Pre-Acidification COD with the Organic Loading Rate Fig. 9. Variation of the Effluent Volatile Fatty Acids with the Organic Loading
within the Experimentation Time for COD Strength of 30 g COD/l. Rate within the Experimentation Time.

mainly due to the insufficient residence time required by the micro- This behavior was attributed to the VFA overfeeding problem; and
organisms to consume the organic content inside the substrate as it the bacterial population had reached their process feeding limit. The
decreased from 12 days at the lowest OLR 2 g COD/L.d to 3 days at OLR higher the VFA formation, the higher the consumption of the COD
9 g COD/L.d. which is the highest OLR tested in this investigation. input. This is demonstrated from Fig. 4, Tables 3 and 4c . For the 40 g
In Fig. 7, for a COD strength of 30 g COD/L, the COD concentration COD/L, the pre-acidification degree which represents the highest
in the PA reactor shows a decrease when the OLR increased. As the OLR amount of the COD converted in to the VFA (about 63%), and the COD
increased from 2.5 g COD/L.d to 8 g COD/L.d, the COD from the PA R% for the same COD strength (40 g COD/L) was also the highest
decreased from 14,903 g COD/L to 14,167 g COD/L in about 96 days. (about 96%).
At OLR 9 gCOD/L.d, the COD in the PA increased up to 20,755 g COD/ The treatment of the industrial wastewater will be accompanied by
L. biogas production [35],[36]. The biogas will be produced by the
This sudden increase caused an inhibition to the EGSB system as the acetoclastic methanogens, which feed on the acetic acid generated from
residence time was too short for the PA reactor to convert the COD into the acetogenesis step. Besides this, the hydrogen gas produced from the
VFA. The amount of the VFA injected into the EGSB was higher than the acidogenesis and acetogenesis steps will be converted into methane by
ability of the acetogens to convert it into the acetate [32],[34]. the hydrogenotrophic methanogens. Fig. 10 shows the biogas produc-
Fig. 8 shows the COD for the effluent stream decreased from 2085 g tion rate along with the OLR through the experimentation time. It can
COD/l at OLR 2.5 g COD/l.d to 1650 g COD/l at OLR 8 g COD/l.d. This be noted that the biogas production will keep increasing from 10.8
is due to the acclimatization time required by the microorganisms to average gal/day at OLR 2.5 g COD/L.d to 23.26 gal/day at OLR 8 g
stabilize. At OLR 9 g COD/l.d, the COD for the effluent stream increased COD/L.d, so the biogas production almost doubled with the increase in
from 1650 g COD/l to 10,350 g COD/l, due to the low residence time in OLR within about 92 days. For the next OLR (9 g COD/L.d), the biogas
the EGSB reactor required to assimilate the organic matter and the in- production rate decreased to about 20.2 average gal/day due to the
crease in the acidity of the system which causes inhibition to the overfeeding situation. When the amount of substrate flow through the
granular biomass particles. EGSB was higher than the ability of the microorganisms to convert all
Fig. 9 shows the VFA for the effluent, which is an essential para- the substrate, the acidity of the reactor will increase and cause an in-
meter used to check the stability of the process. When the OLR increases hibition to the system.
from 2.5 g COD/L.d to 8 g COD/L.d, the volatile fatty acids represented At a steady state process, the production rate of the methane was
by the acetic acid (CH3COOH) concentration were between found to increase with increasing OLR until reaching the maximum,
250–300 mg CH3COOH/L. At OLR 9 g COD/L.d, the VFA concentration where the HRT inside the reactor was found to be the same as the time
raised from 261 mg CH3COOH/L to about 2600 mg CH3COOH/L.

Fig. 8. Variation of the Effluent COD with the Organic Loading Rate within Fig. 10. Variation in the Biogas gas with the Organic Loading Rate within the
Experimentation Time. Experimentation Time.

177
H. Al-Rubaye et al. Chemical Engineering & Processing: Process Intensification 129 (2018) 171–180

Fig. 11. Variation of the pH and the Alkalinity Level within the Remediation
Time.

Fig. 13. Biogas Production Rate Variation with the Organic Loading Rate for
required by the microorganisms to produce biogas at the maximum Different COD Strengths.
rate.
Once the OLR increased above that, the HRT was shortened and
microorganisms to assimilate it than the higher COD strength.
affected the microorganisms, causing an inhibition of the process. As
the inhibition started in the reactor, a remediation process was required
to reactivate the microorganisms available inside the biomass particles, 3.2. Reactor operation under thermophilic temperature range
as they were over-feeding. Fig. 11 shows the progress in pH variation
along with the alkalinity level in the remediation time after the in- Running an anaerobic digestion system under mesophilic tempera-
hibition. ture range is rarely found for expanded granular sludge bed reactor and
Three different COD strengths, 20, 30, and 40 g COD/L, have been couple of researchers investigated and implemented that [6],[37],[38].
tested under equivalent OLRs 2, 2.5, 3, 4, 5, 6, 7, and 8 g COD/L.day. Most of the investigations implemented on EGSB reactors were done at
The liquid effluent and the gas composition have been analyzed. mesophilic temperature range (35–37 °C) due to the difficulties in
As shown in Fig. 12, the R% increased with increasing the HRT for maintaining a constant temperature higher than that. Furthermore,
each OLR applied, so for a COD of 20 g COD/L, the R% would be about running anaerobic digesters under thermophilic temperature range
90% at OLR of 3 g COD/L.d while for a COD strength of 40 g COD/L and 50–60 °C were only implemented on batch reactors [29]. The main
the same OLR, the R% was 96%. This applies for all the OLRs, and it is challenge in operating the high rate AD reactor under thermophilic
caused by the higher hydraulic retention time that lets the substrate temperature range is how to maintain the temperature of the system
spend a longer time in the reactor, and thus more COD is consumed. [39]. At thermophilic temperature range, the microbial growth rate of
From Fig. 13, it can be shown that the biogas production rate in- the methanogens will be faster than that at mesophilic temperature
creased with an increase in the OLR and at the same time, the lower range and the acclimatization time required for the microorganisms
COD strength had a higher biogas production rate than the higher COD will be between 10–15 days [40] while for the mesophilic anaerobes, it
strength [32]. For example, for an equivalent COD strength 20 g COD/ would be between 20–30 days. The thermophilic temperature range
L, at OLR of 3 g COD/L.d, the biogas production rate was 9.6 gal/day will be narrow between 50–60 °C. The transition from mesophilic to
but at OLR of 8 g COD/L.d, the production rate increased up to thermophilic temperature range can’t be at once, the increase in tem-
23.3 gal/day since the wastewater flow inside enhanced and had more perature should be at a very low rate (0.5–2) ° C/day. The methanogens
organics to digest in the reactor. Besides that, the low COD strength would be very sensitive that could be easily shocked and inhibited if it
WW requires less time for digestion, and it will be easier for the got heated too fast and hence, it affect the biogas production rate. The
heating rate for the system was about 0.5 °C/d to heat the system up
from 36 °C (mesophilic temperature range) to 55 °C (thermophilic
temperature range) to overcome all the chances of inhibition or in-
stability due to the temperature increase. During the heating interval,
the substrate was flowing at an OLR of 2 g COD/L.day (HRT of ten days)
and the biogas production rate was monitored. As it shown in Fig. 14,
the biogas was slightly increased with temperature increase at a con-
stant OLR at 20 g COD/L strength. When the reactor was running at
36 °C, the biogas production rate was about 5.6 gal/day and after it
reached about 50 °C, the production rate increased to about 7.9 gal/
day. This increase was not completely stable, but it had some fluctua-
tions as the biomass was very sensitive in the temperature transition
from mesophilic to thermophilic range. To remediate this fluctuation,
temporarily fresh substrate coming from the PA reactor should be de-
creased and increase the recycle stream to maintain the equivalent OLR.
30% of the flowrate in to the EGSB reactor was recycled normally.
The temperature mapping can be shown in Fig. 15, for the whole
EGSB reactor under thermophilic temperature range for a time duration
of four days. All the thermocouples were measuring a steady tem-
Fig. 12. COD Removal Efficiency Variation with the Hydraulic Retention Time perature for the inlet and outlet hot water streams, also the temperature
under Different OLRs. inside the reactor itself was approximately steady for the biomass

178
H. Al-Rubaye et al. Chemical Engineering & Processing: Process Intensification 129 (2018) 171–180

Fig. 16. Biogas Production Rate under Mesophilic and Thermophilic

Temperature Range and the ratio of Biogas Increase in Thermophilic Range to
the Mesophilic Range.
Fig. 14. Biogas Production Rate with Temperature Increase.

L.day the biogas production rate was about 19.457 gal/day at meso-
philic range while it was about 40.673 gal/day at thermophilic range.
At OLR 7 g COD/L.day the biogas production rate was 20.504 gal/day
under mesophilic range while it was about 46.825 gal/day under
thermophilic range. At the last OLR (8 g COD/L.day), the biogas pro-
duction rate was about 20.726 gal/day and 50.174 gal/day under me-
sophilic and thermophilic temperature ranges, respectively. so the
biogas increased about 90–100 % along the OLRs in the thermophilic
range more than that at mesophilic range, this is due to the highly
growth rate of the methanogens under thermophilic range than meso-
philic range which is compatible to what found in the literature [29],
[40].
Also, as it shown in Fig. 17, the COD R% for the system running
under thermophilic range was found to be little higher than that at
mesophilic due to the higher growth rate at thermophilic that make the
Fig. 15. Temperature Mapping in Two-Stage Anaerobic Digestion (EGSB). methanogens consume all the organic matters especially that some of
the VFA would form in the second stage of the AD process as well.
granular particles and for the substrate liquid. And the temperature The decrease in the R% when the system ran at mesophilic tem-
inside the PA reactor was also steady at 41 °C. The temperate of the perature range was due to the inhibition occurred in the system. The
effluent slightly increased even when the temperature mapping for all accumulation rate of the VFAs in the main reactor was higher than the
the other thermocouples were steady, this increase is due to the in- microorganisms’ ability to convert the VFAs into biogas and this in-
crease in the biogas production increase with the time. The biogas creased acidity level in the system and hence, the R% will decrease as
bubbles in the reactor enhanced the mixing of the substrate in the EGSB well as the biogas production.
reactor and therefore, it will increase the temperature at the top of the Table 5 shows the data collected from running the AD system under
reactor especially that no heating system was exist at the top of the thermophilic temperature range. The biogas composition produced
reactor where the effluent flow out of the system.
The pH values for the effluent when the system ran at thermophilic
range were (7.6–7.8), which were little higher than that of the meso-
philic temperature range (7.0–7.4), as the solubility of carbon dioxide
decreases with temperature increase, so less amount of carbonic acid
would be in the system. Furthermore, the production rate under ther-
mophilic range was higher than that at mesophilic range. For substrate
concentration of 20 g COD/L, running at a range of OLRs 2, 3, 4, 5, 6, 7,
and 8 g COD/L.day. Fig. 16 shows the biogas production rate under
mesophilic and thermophilic temperature range at an equivalent OLR
values. So at OLR 2 g COD/L.day, the biogas production rate was about
3.325 gal/day under mesophilic temperature range, while it was about
8.57 gal/day under thermophilic temperature range. Also, at OLR 3 g
COD/L.day, the biogas production rate was 9.582 gal/day under me-
sophilic temperature range while it was about 19.37 gal/day under
thermophilic range. At OLR 4 g COD/L.day, the production rate for the
biogas was about 12.804 gal/day under mesophilic range while it was
about 24.14 gal/day under thermophilic temperature range. At OLR 5 g
COD/L.day, the biogas production rate was about 17.278 gal/day while
at thermophilic range was about 33.91 gal/day. Next, at OLR 6 g COD/ Fig. 17. COD Removal Efficiency with the Organic Loading Rate under
Mesophilic and Thermophilic Temperature Range.

179
H. Al-Rubaye et al. Chemical Engineering & Processing: Process Intensification 129 (2018) 171–180

Table 5 [7] A.S. Dieter Deublin, Biogas from Waste and Renewable Resources, Weinheim, 2008.
Substrate Concentrations in AD System Running under Thermophilic [8] P. Lusk, Methane recovery from animal manures, The Current Opportunities
Casebook, (1998) Washington, DC.
Temperature Range. [9] K. Moriarty, Feasibility study of anaerobic digestion of food waste in St. Bernard,
OLR, mg 20,000 mg COD/L Louisiana, Natl. Renew. Energy Lab. (January) (2013).
[10] J.H. Reith, R.H. Wijffels, H. Barten (Eds.), Bio-Methane & Bio-Hydrogen, Dutch
COD/L.day
Biological Hydrogen Foundation, c/o Energy research Centre of The Netherlands
HRT, day Influent, g PA, g COD/ Effluent, g R% ECN, 2003.
COD/L.day L.day COD/L.day [11] K.V. Rajeshwari, M. Balakrishnan, A. Kansal, K. Lata, V.V.N. Kishore, State-of-the-
art of anaerobic digestion technology for industrial wastewater treatment, Renew.
2,000 10 20,000 13205 937.6 95.312 Sustain. Energy Rev. 4 (2) (2000) 135–156.
3,000 6.66 20,000 13508 657.6 96.712 [12] S. Montalvo, L. Guerrero, R. Borja, I. Cortes, E. Sanchez, M.F. Colmenarejo,
4,000 5 20,000 14327 765 96.175 Treatment of wastewater from red and tropical fruit wine production by zeolite
5,000 4 20,000 14.517 530.6 97.347 anaerobic fluidized bed reactor, J. Environ. Sci. Health Part B— Pestic. Food
Contam. Agric. Wastes 43 (5) (2008) 437–442.
6,000 3.33 20,000 13925 614.6 96.927
[13] W. Qiao, Z. Yin, W. Wang, J. Wang, Z. Zhang, Pilot-scale experiment on thermally
7,000 2.85 20,000 14271 463 97.685 hydrolyzed sludge liquor anaerobic digestion using a mesophilic expanded granular
8,000 2.5 20,000 14832 511 97.445 sludge bed reactor, Water Sci. Technol. 68 (4) (2013) 948–955.
[14] R.T.H.K.E. Mcneil, P.J. Anderson, H. Bartholomew, Biological Treatment of
Distillery Waste, Australian Society of Sugar Cane Technology, 1982.
[15] M. Ghorbanian, Enhancement of Anaerobic Digestion of Actual Industrial
under thermophilic temperature range was found to be almost similar Wastewaters : Reactor Stability and Kinetic, University of Louisville, 2014.
to the biogas produced when the system was running under mesophilic [16] V. Zupančič, D. Gregor, Grilc, Anaerobic treatment and biogas production from
temperature range. organic waste, in: S. Kumar (Ed.), Management of Organic Waste, InTech, Slovenia,
2012p. 198.
[17] R.K. Jadhao, S.D. Dawande, Effect of hydraulic retention time and sludge retention
4. Conclusion time on membrane bioreactor: performance in summer season, Department of
Chemical Engineering, Laxminarayan Institute of Technology, Abstract, Int. J.
Chem. Phys. Sci. 2 (2) (2013) 1–18.
A pilot scale of expanded granular sludge bed reactor was designed [18] T.Z.D. De Mes, A.J.M. Stams, J.H. Reith, G. Zeeman, Methane production by
and built for industrial wastewater treatment, where 20, 30, and 40 g anaerobic digestion of wastewater and solid wastes, Reith J.H., Wijffels R.H. H.,
Barten H. Biomethane Bio-hydrogen – Status Perspect, Biol. Methane Hydrogen
COD/L substrate strengths were used for treatment. The tests were run Prod. (2003) 58–102.
under two temperature ranges. For mesophilic range, all three COD [19] M.T. Kato, S. Rebac, G. Lettinga, Anaerobic treatment of low-strength brewery
strengths were used and for thermophilic range, only one COD strength wastewater in expanded granular sludge bed reactor, Appl. Biochem. Biotechnol. 76
(1) (1999) 15–32.
was used. Under mesophilic condition (36 °C), results showed that COD [20] G. Lettinga, S. Rebac, G. Zeeman, Challenge of psychrophilic anaerobic wastewater
R% and biogas production rate increased by ∼1–6% and ∼30–40%, treatment, Trends Biotechnol. 19 (October (9)) (2017) 363–370.
[21] S.J. Lim, Comparisons between the UASB and the EGSB reactor, Iowa State Univ.
respectively, as HRTs increased by approximately four to five times, Publ. 17 (1969) (2011) 1–17.
while maintaining a constant OLR (∼2, 2.5, 3, 4, 5, 6, 7, and 8 g COD/ [22] G.D. Najafpour, A.A.L. Zinatizadeh, A.R. Mohamed, M. Hasnain Isa,
L.day). Results imply that for equivalent OLRs, better reactor perfor- H. Nasrollahzadeh, High-rate anaerobic digestion of palm oil mill effluent in an
upflow anaerobic sludge-fixed film bioreactor, Process Biochem. 41 (2) (2006)
mance is achieved when running high-concentration COD at a slower 370–379.
rate compared with a lower concentration COD at a faster rate. This [23] I.N. Budiyono, I.N. Widiasa, S. Johari, Sunarso, Study on slaughterhouse wastes
potency and characteristic for biogas production, Int. J. Waste Resour. 1 (5)
also implies a diffusion limiting process where higher molecular weight
(2011) 4–7.
organics, such as complex proteins and fats, likely are passed through [24] R.S. Peris, Biogas Process Simulation Using Aspen Plus, (2011).
the reactor faster than their metabolism rate in the granular biomass for [25] H. Al-Rubaye, S. Karambelkar, M.M. Shivashankaraiah, J.D. Smith, Process simu-
lation of two-stage anaerobic digestion for methane production, Biofuels 7269
digestion. At higher temperature (55 °C), under thermophilic condition (April) (2017) 1–11.
and 20 g COD/L strength, results showed that COD R% and biogas [26] D.J. Batstone, J. Keller, I. Angelidaki, S.V. Kalyuzhny, S.G. Pavlostathis, A. Rozzi,
production rate increased by ∼4% and ∼90–100%, respectively, W.T.M. Sanders, H. Siegrist, V.A. Vavilin, Anaerobic Digestion Model No. 1
(ADM1), IWA Publishing, 2002.
compared to mesophilic range, while maintaining a constant OLR (∼2, [27] Y. Cho, J.C. Young, Modeling the Performance of Fixed-Film Anaerobic Reactors
3, 4, 5, 6, 7, and 8 g COD/L. day). This implies the higher and stronger 792 (2018) no. 1949.
[28] Q. Zhang, J. Hu, D.-J. Lee, Biogas from anaerobic digestion processes: research
population of anaerobes present under thermophilic condition rather updates, Renew. Energy 98 (2016) 108–119.
than mesophilic condition. [29] I. Angelidaki, X. Chen, J. Cui, P. Kaparaju, L. Ellegaard, Thermophilic anaerobic
digestion of source-sorted organic fraction of household municipal solid waste:
start-up procedure for continuously stirred tank reactor, Water Res. 40 (14) (2006)
Acknowledgments 2621–2628.
[30] D. Liu, D. Liu, R.J. Zeng, I. Angelidaki, Hydrogen and methane production from
household solid waste in the two-stage fermentation process, Water Res. 40 (11)
The authors would like to gratefully mention HCED for their fi-
(2006) 2230–2236.
nancial help. We would like to thank the Chemical and Biochemical [31] G. Luo, F. Talebnia, D. Karakashev, L. Xie, Q. Zhou, I. Angelidaki, Enhanced bioe-
Engineering Dept./Missouri University of Science and Technology, nergy recovery from rapeseed plant in a biorefinery concept, Bioresour. Technol.
102 (2) (2011) 1433–1439.
Missouri State, USA for their support. Also, the authors really appre- [32] L.K. Wang, V. Ivanov, J.-H. Tay, Environmental Biotechnology vol. 10, Humana
ciated the help and efforts of Eileen Smith for technical editing support. Press, Totowa, NJ, 2010 1.
[33] Y.-H. Ahn, K.-S. Min, R.E. Speece, Pre-acidification in anaerobic sludge bed process
treating brewery wastewater, Water Res. 35 (December (18)) (2001) 4267–4276.
References [34] R. Girault, P. Rousseau, J.P. Steyer, N. Bernet, F. Béline, Combination of batch
experiments with continuous reactor data for ADM1 calibration: application to
anaerobic digestion of pig slurry, Water Sci. Technol. 63 (11) (2011) 2575–2582.
[1] X. Wang, J.M. Padgett, F.B. De la Cruz, M.A. Barlaz, Wood biodegradation in la-
[35] A. Dhir, C. Ram, Design of an anaerobic digester for wastewater treatment, Int. J.
boratory-scale landfills, Environ. Sci. Technol. 45 (August (16)) (2011) 6864–6871.
Adv. Res. Eng. Appl. Sci. 1 (5) (2012) 56–66.
[2] Q. Xiong, Y. Yang, F. Xu, Y. Pan, J. Zhang, K. Hong, G. Lorenzini, S. Wang,
[36] A.K. Kralj, D. Kralj, Methanol production from biogas, Int. J. Math. Comput. Simul.
Overview of computational fluid dynamics simulation of reactor-scale biomass
4 (2) (2010) 34–41.
pyrolysis, ACS Sustain. Chem. Eng. 5 (April (4)) (2017) 2783–2798.
[37] C. Fang, S. O-Thong, K. Boe, I. Angelidaki, Comparison of UASB and EGSB reactors
[3] C. Guenther, VanEssendeflt, K. Chaudhari, P. Nicoletti, M. Shahnam, T. Li,
performance, for treatment of raw and deoiled palm oil mill effluent (POME), J.
R. Turton, N. Weiland, P. Wang, Gasification modeling in the 21st century, Power
Hazard. Mater. 189 (1–2) (2011) 229–234.
Eng. 116 (June (6)) (2012) 80–82.
[38] C. Fang, K. Boe, I. Angelidaki, Biogas production from potato-juice, a by-product
[4] H. Golpour, T. Boravelli, J.D. Smith, H.R. Safarpour, Production of syngas from
from potato-starch processing, in upflow anaerobic sludge blanket (UASB) and
biomass using a downdraft gasifier, Int. J. Eng. Res. Appl. 7 (6) (2017) 61–71.
expanded granular sludge bed (EGSB) reactors, Bioresour. Technol. 102 (10) (2011)
[5] J. Yu, J.D. Smith, Validation and application of a kinetic model for biomass gasi-
5734–5741.
fication simulation and optimization in updraft gasifiers, Chem. Eng.
[39] G.D. Zupančič, M. Roš, Heat and energy requirements in thermophilic anaerobic
Process.—Process Intensif. 125 (2018) 214–226.
sludge digestion, Renew. Energy 28 (November (14)) (2003) 2255–2267.
[6] M. Ghorbanian, R.M. Lupitskyy, J.V. Satyavolu, R.E. Berson, Impact of hydraulic
[40] D.W. Hamilton, Anaerobic Digestion of Animal Manures: Understanding the Basic
retention time at constant organic loading rate in a two-stage expanded granular
Processes, Oklahoma (2009).
sludge bed reactor, Environ. Eng. Sci. 31 (6) (2014) 317–323.

180