Professional Documents
Culture Documents
Biomass and Bioenergy: Research Paper
Biomass and Bioenergy: Research Paper
Research paper
A R T I C LE I N FO A B S T R A C T
Keywords: The objective of this study was to analyze the effects of the total organic concertation and the proportion of
Anaerobic co-digestion microalgae to optimize biomethane production during anaerobic co-digestion of microalgae and septic tank
Biomethane production sludge. To achieve this goal, a statistical tool, central composite design, was applied to design the co-digestion
Microalgal biomass experiments. Using multiple variance analysis, the experimental data were analyzed, and the results were
Septic tank sludge
presented by the response surface methodology, which illustrated that peak methane production could be
Volatile fatty acids
Ammonia inhibition
achieved when the total volatile solids concentration was in the range of 16–20 g/L and the proportion of
microalgal biomass ranged from 26 to 47%. An inhibitory effect was observed from volatile fatty acids accu-
mulation when the total volatile solids concentration was supplied as high as 35 g/L. Organic nitrogen miner-
alization took place in all groups, ranging from 23 to 62%. There was no direct evidence showing that produced
ammonium ion nitrogen or the free ammonia nitrogen had inhibitive effects on methanogenic activity.
∗
Corresponding author.
E-mail addresses: Dingnan_lu@student.uml.edu (D. Lu), xiao_liu@student.uml.edu (X. Liu), onur_apul@uml.edu (O.G. Apul), lin.zhang@tamucc.edu (L. Zhang),
David_Ryan@uml.edu (D.K. Ryan), Jackie_zhang@uml.edu (X. Zhang).
https://doi.org/10.1016/j.biombioe.2019.105266
Received 18 September 2018; Received in revised form 23 May 2019; Accepted 5 June 2019
0961-9534/ © 2019 Elsevier Ltd. All rights reserved.
D. Lu, et al. Biomass and Bioenergy 127 (2019) 105266
activated sludge with Chlorella sp. They observed that with the addition
of activated sludge, not only the net biogas yields improved but the
biogas production rate was accelerated significantly. All the examples
listed above demonstrate advantages in methane production when
performing anaerobic co-digestion of microalgae with a carbon-rich
waste co-substrate.
The feasibility of using septic tank sludge as a co-substrate to digest
with microalgae has been proven [14]. The exceptional performance of
septic tank sludge can be attributed to its high concentration of organic
carbon (∼10,000 mg/L) and the presence of important hydrolytic en-
zymes and bacteria [15]. Particularly, the addition of septic tank sludge
to microalgae can result in a more favorable C/N ratio (20:1-30:1) to
balance the high protein content of microalgal biomass. Besides, the
introduced enzymatic activities from septic tank sludge can sig-
nificantly accelerate hydrolysis of microalgal biomass, which resulted
in a higher glucose release rate (0.14 mg/L/min) in the early anaerobic
digestion period [14]. Many anaerobic bacteria, such as Clostridium
acetobutylicum, Clostridium cellulolyticum, and Acetivibrio cellulolyticus,
have been proven to actively hydrolyze cellulose. And these re-
presentative anaerobic bacteria can be easily found in soil, sewage, and Fig. 1. Microscopic image of the harvested Chlorella sp. (1000× magnifica-
animal feces [16] which are the potential components of the septic tank tion).
sludge. It is reasonable to believe that the cellulolytic activities from the
septic tank sludge is fundamentally triggered by the presence of these 2. Methods
anaerobic bacteria. Furthermore, methanogens and various bacterial
species involved in the anaerobic degradation of organic matter, in- 2.1. Raw materials
cluding hydrolytic, fermentative, acidogenic, acetogenic, homo-
acetogenic and sulfate-reducing bacteria should also be present [17]. The microalgae strain was isolated from the primary clarifiers at the
Although the co-digestion of microalgae and septic tank sludge has Lowell Regional Wastewater Utility (LRWU) (Lowell, MA). The ob-
been proven to outperform the biogas production from the digestion of tained green spherical-shaped microalgae, identified as Chlorella sp.
microalgae alone, there is scarce biochemical information available (see Fig. 1), was then cultivated using the modified BBM in the study.
regarding the co-digestion of microalgae and septic tank sludge. Spe- The detailed procedures of microalgae isolation, identification, and
cifically, it is not clear how the organic loading and the split between cultivation were described in our previous works [14,19]. The micro-
the co-substrates will affect methane production when conducting the algal stock was prepared by centrifugation at 3000 rpm for 10 min
anaerobic co-digestion of microalgae and septic tank sludge. followed by washing with distilled water. The centrifugation-washing
The main objective of this study was to investigate the effects of two was repeated three times to ensure the microalgae culture was free from
variables, namely the VS concentration (g/L) and the percentage of the chemical residues of modified BBM. The prepared microalgae stock
microalgal biomass, on the anaerobic co-digestion of microalgae and was kept suspended in tap water and stored in sealed beakers at 4 °C.
septic tank sludge. Aiming to study both parameters simultaneously, Septic tank sludge was obtained directly from a hauling truck at the
CCD followed by RSM was performed. CCD is a useful statistical tool to LRWU and transported immediately to the environmental laboratory at
arrange experiment groups when multiple factors were studied in a the University of Massachusetts Lowell. The septic tank sludge was
limited number of runs, and RSM can help visualize the statistical data collected from residential septic tanks near the city of Lowell, MA.
that was generated by the experimental arrangement from CCD [18]. The microalgae stock and septic tank sludge were characterized by
There are no commonly agreed methods to design and conduct anae- measuring pH, VS, TCOD, TKN, and TOC (see Table 1). The con-
robic co-digestion studies, which adds difficulty to compare different centration ratio of TOC and TKN were used to calculate the initial C/N
co-digestion studies. Therefore, this study can be served as an example ratio.
to help researchers design other anaerobic co-digestion studies, which
might involve two or more co-substrates. Moreover, the concentrations
2.2. Mesophilic anaerobic experiment
of VFAs and ammonia were monitored to help understand the me-
chanisms during the anaerobic co-digestion of microalgae and septic
The selected variables for the study were the substrate concentra-
tank sludge.
tion measured in terms of VS (g/L) and the proportion of microalgal
biomass added in terms of percentage of volume (%). The initial sub-
strate concentration range was arranged in accordance with Ward et al.
[7], who reported a wide range of initial VS concentrations from dif-
ferent studies performed anaerobic digestion of Chlorella sp. On the
other hand, the selected range for proportions of microalgal biomass
2
D. Lu, et al. Biomass and Bioenergy 127 (2019) 105266
was 0–100% for the co-digestion experiment. In order to prepare a RSM was performed for the result optimization of the CCD analysis.
desirable VS concentration and microalgae proportion in a digester, the RSM is a collection of statistical and mathematical techniques which
raw feedstocks of the microalgae slurry and the septic tank sludge had can visualize a response influenced by several variables [21]. This
gone through three sequential processes: 1. Prepare both raw feedstocks method uses a second-order polynomial to describe the response, see
into the same VS concentration (i.e., 15 g/L) by centrifugation; 2. Dis- Eq. (3):
card supernatant (or fill additional distilled water) in proportion to a VS
difference between a centrifuged raw feedstocks and a desirable level; Y = β0 + β1 X1 + β2 X2 + β11 X12 + β22 X22 + β12 X1 X2 + E (3)
3. Shake centrifuge tubes to resuspend settled sludge and pour prepared where Y represents the predicted response, β0 , β1, β2 , β11, β22 , and β12 are
microalgae slurry and septic tank sludge into a digester based on vo- the regression coefficients. X1 and X2 are the factors' coded values. E is
lume ratio to meet a request for specific microalgae proportion. As an the standard error.
example, take the digesters consisting 30 g/L of VS and 85% of mi- The CV was calculated as the ratio of the standard error of the es-
croalgae. The raw feedstocks were first centrifuged to obtain the VS timate to the mean value of the observed response. The R2 was calcu-
concentration of 15 g/L. Then the clean supernatant was carefully dis- lated to evaluate the proportion of data variability which is predicted
carded from centrifuge tubes by half of the total volume to reach the by the second-order polynomial model. The p-value of each regression
desirable VS concentration (i.e., 30 g/L). Lastly, 85 mL of microalgae coefficient was calculated to validate the model. Excel v.2016 and
slurry and 15 mL of septic tank sludge (both had been shaken to re- STATISTICA v.10 were used to perform the multiple regression and
suspend) were added into the serum bottles to have these digesters graphical analyses.
consisting 30 g/L of VS and 85% of microalgae.
The anaerobic co-digestion was carried out in a series of 200 mL
2.5. Ammonia (NH3) and ammonium (NH4+) speciation analysis
serum bottles with a liquid mixture volume of 110 mL and a headspace
of 90 mL for the gas. The serum bottles were sealed by rubber stoppers
In order to quantify the proportions of molecular ammonia and
and secured with aluminum foil to keep them airtight. After the set-up
ammonium ion, pKa for ammonium ion at specific was calculated using
of each reactor, the headspace was flushed with nitrogen for 15 min to
the following equation [22]:
remove residual oxygen in the headspace. Anaerobic digestion was
performed by placing the serum bottles in a reciprocal shaker bath 2729.92
pK a = 0.09018 +
(VWR™ model 1227) and run at 120 oscillations per minute at 35 °C for T + 273.15 (4)
30 days. The volume of produced biogas was calculated by measuring o
where T is the temperature of culture solution ( C). The ammonia re-
the pressure in the bottle's headspace. The experimental design is moval rate was determined using the pKa calculated by Eq. (4), and the
shown in Table 2. All the reactors were carried out in triplicate. proportional relation between FAN (mg/L) and TAN (mg/L) can be
Anaerobic inoculum consisted of digested sludge from a mesophilic expressed as the following:
anaerobic reactor (New Brunswick Scientific Co., Inc.) fed with artifi-
cial wastewater [20] under semi-continuous operation. Prior to the co- FAN 1
=
digestion experiment, the inoculum (VS at 15 g/L) had been degassed at TAN 1 + 10 (pK a− pH ) (5)
35 °C for one week. Observation indicated no significant methane (0.06 where pKa is the negative logarithm of the dissociation constant of
mL/gVS) being produced after the one-week incubation. 8.95 at 35 °C (calculated by Eq. (4)), and pH is the negative logarithm of
3
D. Lu, et al. Biomass and Bioenergy 127 (2019) 105266
the hydrogen ion activity on the selected day. Therefore, Eq. (5) can be specifically, both the first-order coefficients (β1 and β2) of x1 (i.e., VS
rewritten by inputting the calculated pKa value at 35 °C, then the final concentration) and x2 (i.e., microalgal proportion) and the second-
equations for FAN and AIN can be expressed as the following: order coefficients (β11 and β22) of x12 and x 22 . The β1 and β2 had close
negative values, which means that both x1 and x2 had negative linear
FAN 1
= correlations with Y. In practical terms, with a fixed microalgae pro-
TAN 1 + 10 (8.95 − pH ) (6)
portion then an increase in VS loading will lead to a decrease in biogas
AIN 10 (8.95 − pH ) production. Similarly, with a static supply of VS loading, biogas pro-
=
TAN 1 + 10 (8.95 − pH ) (7) duction will decrease when increasing in microalgae proportion. The
second-order coefficients β11 and β12 also had negative values, indicated
From above Eq. (6) and Eq. (7), it can be found when a digester is
the parabolas open downwards in their respective two-dimensional
running at a constant temperature, the ammonia molecular and am-
surface. Since this quadratic function inputs, x1 and x2, were an ordered
monia ion speciation totally depends on the pH.
pair of real numbers and a function's curve is the collection of all or-
dered triples (x1, x2 , f (x1, x2) ) , the resulted function was a three-di-
2.6. Analytical techniques
mensional surface with its mouth opening downwards. In other words,
the vertex had the highest methane production on the three-dimen-
VS was measured in accordance with the Standard Methods [23].
sional plot. The magnitude of a second-order coefficient directly de-
The TCOD and TAN were carried out by using a Hach DR/2000 spec-
termines the shape of a parabola. In other words, it is able to indicate
trophotometer (Hach Inc., Loveland, CO, USA). The TKN assay was
the sensitivity of methane production affected by both variables, which
done by using Hach DR/3900 spectrophotometer (Hach Inc., Loveland,
will be discussed in the next two paragraphs. Lastly, the coefficient of
CO, USA) coupled with a DRB 200 block digester (Hach Inc., Loveland,
x1x2 (β12) had the p-value greater than 0.05, which implied the inter-
CO, USA). The TOC measurement was performed by using an Ionics
action effect between the two selected variables were relatively insig-
model 1555B carbon analyzer (GE Analytical Instruments, Boulder, CO,
nificant to the result of methane production.
USA). The biogas pressure in the bottle's headspace was measured every
For the graphical interpretation of the interactions, the use of a
other day with a digital pressure meter Cole Parmer 500PSI (Chicago,
three-dimensional plot of the regression model is highly useful [25]. In
IL), and the biogas composition (CH4 and CO2) was evaluated si-
this study, the two variables x1 and x2 and the result of y (i.e., methane
multaneously by using a gas chromatograph: SRI Model 8610 equipped
production) were used as the axes of the surface response plot (see
with a flame ionization detector (FID) for CH4 and CO2. The con-
Fig. 3A). The predicted values for the response of methane production
centration of VFAs, including formate, acetate, propionate, and buty-
are listed in Table 3. The three-dimensional shape of the response
rate, was determined by using a Dionex ion chromatography (IC)
surface was a hump, indicated a peak methane production had occurred
system equipped with an ED50 electrochemical detector.
inside the designed CCD boundary. A contour plot, made by projecting
the three-dimensional plot on the plane of x1−x2, was used to depict
3. Results and discussion
the range of the correlated x1 and x2 values, in which the resulted
methane production was higher than a specific value, such as, 180, 210,
3.1. Statistical and graphical analyses
240, 270, 300, and 330 mL/gVSfed (see Fig. 3B). Therefore, the most
4
D. Lu, et al. Biomass and Bioenergy 127 (2019) 105266
Table 4
Statistical analysis of the evaluated factors (Initial VS and microalgae %) re-
garding the response of methane production.
Analysis of variance (ANOVA)
Variable Value
R2 0.96
Adj. R2 0.89
F 1.04
Fcritical 3.44
Coef. of variance (%) 5.14
Coefficient probability
Coefficient p-value*
∗
p-value < 0.05 is statistically significant at 95% level of confidence.
5
D. Lu, et al. Biomass and Bioenergy 127 (2019) 105266
3.2. Methane production and VFAs concentration group R7 due to lower initial organic loading.
6
D. Lu, et al. Biomass and Bioenergy 127 (2019) 105266
AIN and FAN are the two principal forms of inorganic ammonia
nitrogen in aqueous solution, and FAN has been suggested to be the
major cause of inhibition since it is freely membrane-permeable
[30,31]. The hydrophobic FAN molecules may diffuse passively into
cells, causing a proton imbalance and/or potassium deficiency [32–34].
To determine the proportions of AIN and FAN in the digester, NH3 and
NH4+ speciation analysis was done by using Eq. (4), Eq. (6) and Eq. (7).
From the above equations, it is worth noting that temperature and pH
are the two fundamental environmental factors that govern the spe-
ciation process of NH3 and NH4+ [35]. To suppress the formation of
FAN and reduce the impact of FAN inhibition, interactions between
temperature, pH, and FAN speciation were estimated and compared
mathematically. Fig. 7. Concentration of organic nitrogen (mg/L) at the beginning (whit
The temperature of this anaerobic digestion study was pre- column) and end (gray column) of the experiment, and reduction (%) for
determined to be 35 °C. Comparing to thermophilic conditions, this R0(VS20 ALG50%), R5(VS20 ALG0%), calculated average (R5 and R6), and
mesophilic temperature might result in a relatively lower kinetic R6(VS20 ALG100%).
7
D. Lu, et al. Biomass and Bioenergy 127 (2019) 105266
8
D. Lu, et al. Biomass and Bioenergy 127 (2019) 105266
[22] B. Calli, B. Mertoglu, B. Inanc, O. Yenigun, Effects of high free ammonia con- review, Bioresour. Technol. 99 (10) (2008) 4044–4064.
centrations on the performances of anaerobic bioreactors, Process Biochem. 40 [32] M. Kayhanian, Ammonia inhibition in high-solids biogasification: an overview and
(3–4) (2005) 1285–1292. practical solutions, Environ. Technol. 20 (4) (1999) 355–365.
[23] APHA, AWWA, W.E.F. WPCF, Standard Methods for the Examination of Water and [33] G.D. Sprott, G.B. Patel, Ammonia toxicity in pure cultures of methanogenic bac-
Wastewater, 17 st ed., American Public Health Association (APHA), Washington, teria, Syst. Appl. Microbiol. 7 (1986) 358–363.
DC, 1989. [34] C. Gallert, S. Bauer, J. Winter, Effect of ammonia on the anaerobic degradation of
[24] Q.K. Beg, V. Sahai, R. Gupta, Statistical media optimization and alkaline protease protein by a mesophilic and thermophilic biowaste population, Appl. Microbiol.
production from Bacillus mojavensis in a bioreactor, Process Biochem. 39 (2) Biotechnol. 50 (4) (1998) 495–501.
(2003) 203–209. [35] R. Rajagopal, D.I. Masse, G. Singh, A critical review on inhibition of anaerobic
[25] R.L. Mason, R.F. Gunst, J.L. Hess, Statistical Design and Analysis of Experiments: digestion process by excess ammonia, Bioresour. Technol. 143 (2013) 632–641.
with Applications to Engineering and Science, John Wiley & Sons Inc., Hoboken, [36] J.K. Kim, B.R. Oh, Y.N. Chun, S.W. Kim, Effects of temperature and hydraulic re-
NJ, 2003. tention time on anaerobic digestion of food waste, J. Biosci. Bioeng. 102 (4) (2006)
[26] X. Guan, R. Melchers, Effect of response surface parameter variation on structural 328–332.
reliability estimates, Struct. Saf. 23 (4) (2001) 429–444. [37] I. Angelidaki, B.K. Ahring, Anaerobic thermophilic digestion of manure at different
[27] H. Takeda, Sugar composition of the cell wall and the taxonomy of Chlorella ammonia loads: effect of temperature, Water Res. 28 (3) (1994) 727–731.
(Chlorophyceae), J. Phycol. 27 (2) (1991) 224–232. [38] I. Angelidaki, B. Ahring, Thermophilic anaerobic digestion of livestock waste: the
[28] H. Takeda, Cell wall sugars of some Scenedesmus species, Phytochemistry 42 (3) effect of ammonia, Appl. Microbiol. Biotechnol. 38 (4) (1993) 560–564.
(1996) 673–675. [39] A.M. Abo-Shady, Y.A. Mohamed, T. Lasheen, Chemical composition of the cell wall
[29] I. Siegert, C. Banks, The effect of volatile fatty acid additions on the anaerobic in some green algae species, Biol. Plant. 35 (4) (1993) 629–632.
digestion of cellulose and glucose in batch reactors, Process Biochem. 40 (11) [40] P.Y. Robidoux, J. Lopez-Gastey, A. Choucri, G.I. Sunahara, Procedure to screen il-
(2005) 3412–3418. licit discharge of toxic substances in septic sludge received at a wastewater treat-
[30] E. Kroeker, D. Schulte, A. Sparling, H. Lapp, Anaerobic treatment process stability, J ment plant, Ecotoxicol. Environ. Saf. 39 (1) (1998) 31–40.
Water Pollut Con F 51 (1979) 718–727. [41] C.-Y. Lin, Effect of heavy metals on volatile fatty acid degradation in anaerobic
[31] Y. Chen, J.J. Cheng, K.S. Creamer, Inhibition of anaerobic digestion process: a digestion, Water Res. 26 (2) (1992) 177–183.