Bioresource Technology 166 (2014) 373–380

Contents lists available at ScienceDirect

Bioresource Technology
journal homepage: www.elsevier.com/locate/biortech

Anaerobic digestion of pig and dairy manure under photo-dark

fermentation condition
Dongxue Yin a,b, Wei Liu a,b, Ningning Zhai b,c, Gaihe Yang b,c,⇑, Xiaojiao Wang b,c, Yongzhong Feng b,c,
Guangxin Ren b,c
a
College of Forestry, Northwest A&F University, Yangling, Shaanxi Province 712100, China
b
Research Center for Recycling Agricultural Engineering Technology of Shaanxi Province, Yangling, Shaanxi Province 712100, China
c
College of Agronomy, Northwest A&F University, Yangling, Shaanxi Province 712100, China

h i g h l i g h t s

 Digestion of pig manure (PM) and dairy manure (DM) were conducted.
 Anaerobic digestion was carried out under photo-dark and total dark condition.
 Photo-dark condition promoted anaerobic fermentation for PM.
 Total dark fermentation was more suitable for DM.
 pH value had the maximum determination coefficient with cumulative biogas production.

a r t i c l e i n f o a b s t r a c t

Article history: Anaerobic digestion (AD) with livestock manure is a promising way for biogas production. This work
Received 17 January 2014 presents the influence of photo-dark fermentation on biogas production of pig manure (PM) and dairy
Received in revised form 5 May 2014 manure (DM). All sets were conducted with temperature 35 ± 2 °C and total solid concentrations 8%:
Accepted 7 May 2014
PM1 and DM1 in transparent reactor under sunlight for photo-dark fermentation, and PM2 and DM2 in
Available online 20 May 2014
non-transparent reactor for dark fermentation. DM2 had the best cumulative biogas production (CBP)
of 15,447.5 mL, followed by PM1 (15,020 mL) with stable pH and low total ammonium nitrogen (TAN)
Keywords:
concentration (1384.99 mg/L), and DM1 and PM2. The CBP of DM2 was 5.77 times as much as PM2. The
Anaerobic digestion
Pig manure
relationship between CBP and four factors including volatile fatty acid (VFA), TAN, total alkalinity and
Dairy manure pH was analyzed. pH gained the maximum determination coefficient with the CBP among all sets and
Photo-dark fermentation total alkalinity showed negative correlation with CBP of PM1 and DM1.
Path analysis Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction
Along with the development of intensive feeding in China, the
livestock manure increases fast with an annual production rate of
Abbreviations: AD, anaerobic digestion; CBP, cumulative biogas production; C/N,
carbon-to-nitrogen ratio; DBP, daily biogas production; DM, dairy manure; DM1,
dairy manure with photo-dark fermentation condition; DM2, dairy manure under
dark fermentation condition; PM, pig manure; PM1, pig manure with photo-dark
fermentation condition; PM2, pig manure under dark fermentation condition; TAN,
total ammonium nitrogen; TN, total nitrogen; TS, total solids; VFA, volatile fatty
acid; VS, volatile solid.
⇑ Corresponding author at: College of Agronomy, No. 95 Mailbox, North Campus
of Northwest A&F University, Yangling, Shaanxi 712100, China. Tel.: +86
13709129773; fax: +86 029 87092265.
E-mail address: ygh@nwsuaf.edu.cn (G. Yang).
406 million tons in 2013 (Jin et al., 2013). Livestock manure waste
is a significant source of fecal pollution and serious sanitary prob-
lems due to its high chemical oxygen demand, high concentrations
of suspended solids, and nitrogen and phosphorus compounds
(Song et al., 2010). Hence, efficient disposal of manure has been a
key factor limiting expansion of the livestock industry in China
(Shi et al., 2011).
Anaerobic digestion (AD) with high biological transformation is
a well-established technology to deal with animal manure, and it
can be used for electricity generation, residential heating, and
cooking, etc. to save energy expenditure and produce renewable
energy (Wang et al., 2012). Thus, it has fundamental significance
to utilize pig manure (PM) and dairy manure (DM) to produce
clean renewable energy in order to protect environment and
mitigate the energy demand (Yang and Zhang, 2008). Recently,

http://dx.doi.org/10.1016/j.biortech.2014.05.037
0960-8524/Ó 2014 Elsevier Ltd. All rights reserved.
374 D. Yin et al. / Bioresource Technology 166 (2014) 373–380
biogas production by AD has been drawing increasing attention,
since it is an efficient substitute for traditional energy. However,
the biogas production during AD is influenced by many factors
such as temperature, pH, VFA, feed composition and C/N ratio.
Previous works have studied the effects of different factors on
biogas production. Hobson et al. (1980) suggested that the biogas
production was linearly correlated with temperature from 25 to
44 °C. Chae et al. (2008) also showed that digestion temperature
had influence on the ultimate methane yield and higher tempera-
ture from 25 to 35 °C increased the methane yield. However, differ-
ent results were reported by Hansen et al. (1999) who suggested
that increase of temperature reduced biogas yield because of the
increased inhibition of free ammonia (NH3) when temperature
was increasing. Van Haandel and Lettinga (1994) suggested that
the rate of methane production declined at pH-values <6.3 or
>7.8 and Van Lier (1995) showed pH impacted enzymatic activity.
According to EI-Mashad et al. (2004), the stability of fermentation
process largely depended on the chemical equilibrium between
VFA, bicarbonate and ammonia. In addition, the researches on feed
composition and C/N ratio showed that co-digestion using various
materials could optimize the digestion process and improve biogas
yield, and a C/N ratio of 25:1 had the best digestion performance
(Burton and Turner, 2003; Wang et al., 2012). Among these factors,
very few studies have investigated the effects of sunlight as an
external artificial factor on fermentation. It was suggested that
sequential dark and light anaerobic fermentations could enhance
bio-hydrogen production from carbohydrate rich biomass or waste
materials (Argun and Kargi, 2010). However, the effect of photo-
dark condition on biogas yield is unclear to the date.
This study investigated the effects of photo-dark and total dark
conditions on fermentation of PM and DM. Moreover, the correla-
tion level between VFA, Total alkalinity, TAN and pH, and CBP under
photo and dark conditions was examined by path analysis and
analysis of decisive degrees to confirm which factor has the decisive
effect. Finally, to improve biogas production these factors were
artificially controlled under different fermentation conditions.
2. Methods
2.1. Origin and characterization of substrates
In the study, PM and DM were collected from a livestock farm
located in Yangling, China. The inoculum was obtained from
household biogas digesters in a local biogas demonstration village
in Yangling, China. The substrates and inoculum were separately
homogenized and subsequently stored at 4 °C for further use
(Wang et al., 2012). The chemical characterization of DM is given
in Table 1. All samples were collected in triplicates, and the aver-
ages of the three measurements are presented.
2.2. Feeding of digesters
Anaerobic fermentation of PM and DM was carried out under
mesophilic (T = 35 ± 2 °C) with the total solid concentration of 8%
for 53 days according to the method described in Wang et al.
(2012). The 1 L glass reactor with 700 g total liquid, including
140 g inoculum, was conducted by controlled and constant
Table 1
Chemical characterization of substrates used in the digestion experiments.
Material Total solid Volatile Organic carbon
(%) solid (%) (g/kg VS)
PM 27.7 79.2 78.3
DM 14.8 78.6 65.3
a
Dry basis.
temperature anaerobic fermentation device (Fig. 1). The experi-
ment included four sets as following:
1) Pig manure (PM1) and dairy manure (DM1) in 1 L transparent
glass reactor placed in laboratory with nature sunlight for
photo-dark continuous fermentation.
2) Pig manure (PM2) and dairy manure (DM2) in 1 L non-
transparent glass reactor wrapped with black foils and
placed in laboratory for dark continuous fermentation.
The trial was carried out from September 17 to November 11 in
2012. The monthly sunlight durations of the three months in Yan-
gling, Shaanxi Province, China were 148.3, 146.7 and 154.1 h,
respectively. For the photo-dark continuous fermentation of PM1
and DM1, the averaged photo fermentation time was 4.9 h whereas
the dark fermentation time was 19.1 h per day. The data on sun-
light duration was gathered from the Yangling Meteorological
Information Network (http://www.ylqx.gov.cn/).
Three blank samples containing 140 g inoculum and 560 g dis-
tilled water were carried out to determine the biogas production of
inoculum. The gas volume was measured every day. Furthermore,
samples were drawn periodically to measure VFA, TAN, total alka-
linity and pH. All reactors were tightly closed with rubber septa
and screw caps. The head space of each reactor was flushed with
nitrogen gas for about 3 min to assure anaerobic conditions prior
to the start of the digestion tests. To assure the mixing in the reac-
tor, all reactors were shaken manually for about 1 min once a day
prior to measurement of biogas volume (Wang et al., 2012).
2.3. Analysis methods
Total organic carbon was determined by the method described
in Cuetos et al. (2011). The volume of biogas was measured by dis-
placement of water. Methane content in the obtained biogas was
analyzed using the fast methane analyzer (Model DLGA-1000,
Infrared Analyzer, Dafang, and Beijing, China). Volatile fatty acid
(VFA) concentration was analyzed using a UV detector (754P).
Determination of total solid (TS), volatile solid (VS), and total
ammonium nitrogen (TAN) was performed according to APHA
Standard Methods (1995). Total alkalinity analysis was conducted
using a titration with 0.02 M H2SO4 and all titrations were per-
formed in duplicates (Ward et al., 2011).
2.4. Path analysis and analysis of decisive degree
Fig. 2 showed the path network of four independent variables
included in path analysis and analysis of decisive degree.
Path analysis can determine whether the decisive effect of Xi on
Y is significant or not and can identify the indirect effect of Xi on Y
through Xj (Xi  Xj  Y, i – j). Thereby, the correlation coefficient
(riy) contains the direct path coefficient (bi; Xi  Y) and the indirect
path coefficient (rij bj; Xi  Xj  Y, i – j) (Eq. (2)). However, a large
direct effect between Xi and Y does not always imply a strong
correlation between them. Therefore, path analysis in the path net-
work could not identify which independent variable has the deci-
sive effect on the dependent variable. Thus, further analysis using
the decisive coefficient should be conducted (R2ðiÞ ) to reflect the
TKNa C/N pH Total alkalinity Total ammonium Volatile fatty
(g/kg VS) (mg/L) nitrogen (mg/L) acid (mg/L)
6.1 12.8 6.4 5093.0 1328.7 5569.5
3.1 30.1 7.0 5914.0 1220.8 5818.5
 D. Yin et al. / Bioresource Technology 166 (2014) 373–380 375

Fig. 1. Controlled and constant temperature anaerobic fermentation device. (1)

Temperature controlling box; (2) temperature sensor; (3) insulated cover; (4)
thermostatic water tank; (5) strip heater; (6) none transparent digester; (7)
transparent digester; (8) taking sampling; (9) airway tube; (10) taking biogas; (11)
aqueduct; (12) air pipe; (13) biogas collecting bottle; (14) water collecting bottle.

Fig. 2. Network of path analysis. X1 – VFA; X2 – TAN; X3 – total alkalinity; X4 – pH.

Fig. 3. The dynamic changes of VFA (a) and pH (b) values in anaerobic digestion of
comprehensive decisive effect of Xi on Y by X1, X2, . . ., Xp and deter-
PM and PM.
mine the decisive factors. The availability of R2ðiÞ sorts the compre-
hensive effect of Xi on Y. Xi with the largest or smallest value is the
decisive or limiting variable, respectively (Eq. (4)) (Yuan et al., the last week of experiment, which is consistent with the theory
2001). of anaerobic fermentation (Yin et al., 2012). In the early fermenta-
8 tion stage, VFA of PM1 and PM2 had a sharp decrease, after that VFA
> b1 þ r 12 b2 þ . . . þ r1p bp ¼ r1y
>
> of PM1 steeply decreased until the end of fermentation with the
>
< r 21 b1 þ b2 þ...þ r2p bp ¼ r2y concentration of 1418.5 mg/L while VFA of PM2 kept tapering
.. .. .. ð1Þ
>
> down. From the beginning to the end of fermentation, VFA con-
>
> . . .
: tents of PM1 and PM2 decreased from 5467.5–1418.5 mg/L, and
r p1 b1 þ r p2 b2 þ . . . þ bp ¼ r py 5671.5–1633.5 mg/L, respectively. The difference of initiative
X contents of VFA in DM1 and DM2 was not significant. VFA contents
r iy ¼ bi þ bj rij ð2Þ of DM1 and DM2 reached the peak on the 8th day during the
j–1 fermentation, with concentrations of 8488.5 and 8088.5 mg/L,
respectively. After the peak, the content of VFA in DM2 decreased
2
R2i ¼ bi ; R2ij ¼ 2bi r ij bj ð3Þ rapidly to 4608.5 mg/L by the middle of the fermentation. Finally,
the VFA of DM1 and DM2 decreased to 2760 mg/L and 4855 mg/L
2
X 2 than the previous, respectively.
R2ðiÞ ¼ bi þ 2 bi r ij bj ¼ 2bi riy  bi ð4Þ
Fig. 3(b) shows, pH value of PM1 and PM2 decreased to 6.4 and
j–i
6.2, respectively on the 7th day of digestion, which was probably
where bi is the direct path coefficient, rij is the correlation coeffi- due to the hydrolysis acidification resulted from the fact that PM
cient between Xi and Xj, riy is the correlation coefficient between had large amounts of protein and carbohydrates and small
Xi and Y, i, j = 1, 2, . . ., p, and R2ðiÞ is the decisive coefficient. amounts of lipids (Astals et al., 2012). With the advance of fermen-
Moreover, all the results were presented as mean values ± stan- tation, methanogens rapidly consumed VFA increasing pH and sta-
dard deviations (SD). The significance of the difference between bilizing digester performance (Wang et al., 2013). At the end of
mean values was determined by one-way ANOVA flowed by fermentation, pH value of each group was in the range of 6.5–7.5
Tukey’s multiple range test using SPSS19.0 software. p val- and was higher than that of the beginning. The change of DM
ues < 0.05 were considered as statistically significant. Correlation was consistent with PM. The pH value of DM1 decreased from 6.4
analysis was carried out using SPSS 19.0 software and path analysis to 6.0 in the first week because of the high initial accumulation
was performed using DPS 2006 software. of VFA (from 5583.5 to 8488.5 mg/L), and then increased to 7.3 fol-
lowed by a stable period. The pH value of DM2 rose sharply from
5.9 to 7.4 in the first 15 day, and then the trend was stable. The
3. Results and discussion
ending pH values of DM1 and DM2 were 7.5 and 7.6, respectively,
which were higher than the beginning.
3.1. Dynamic changes of VFA, pH, total alkalinity and TAN

3.1.1. VFA and pH 3.1.2. Total alkalinity and TAN

Fig. 3(a) presents the changes in VFA of PM and DM. VFA Fig. 4(a) shows that the average total alkalinity concentrations in
decreased gradually at the beginning and tended to be stable at PM1 (6684.8 mg/L) was higher than that of in PM2 (5891 mg/L). The
376 D. Yin et al. / Bioresource Technology 166 (2014) 373–380
Fig. 4. The dynamic changes of total alkalinity (a) and TAN (b) in anaerobic
digestion of PM and PM.
amount of total alkalinity in PM1 grew from 5308 mg/L to 7746
mg/L in the first 29 day of fermentation, and then this number
remained largely unchanged until on the 43rd day, after that, it
reduced gradually to 6246 mg/L. On the 43rd day, total alkalinity
of PM2 reached the peak of 6866 mg/L, which was less than that
of PM1 (7766 mg/L). Then the number fell to 6386 mg/L. By the
end of the experiment, the amount of total alkalinity in each group
nearly kept the similar value as the beginning.
The changing trend of total alkalinity values in DM1 and DM2
was very similar (Fig. 4(a)) with bimodal curve at the beginning,
which was quite different from PM. DM1 dropped from 7954
mg/L to 4634 mg/L at the 22nd day with a large fluctuation, then
raised gradually until the 43rd day. The average total alkalinity
content in DM2 was always higher than that in DM1 and the final
value of total alkalinity in DM2 (7134 mg/L) was 1.14 times as
much as in DM1 (6274 mg/L), which explained why the CBP of
DM1 was lower than that of DM2.
Besides, TAN is also an important factor affecting the methane
production because the increase of ammonia/ammonium concen-
tration provides better buffering capacity (Walker et al., 2011).
But ammonia inhibition usually happens when pH is above 7.4
and total ammonia nitrogen is 1500–3000 mg/L, whereas it will
occur toxic irrespective of pH when total ammonia nitrogen con-
centration exceed 3000 mg/L (Calli et al., 2005). In addition, 100%
inhibition occurs between the range of 8000–13000 mg/L depend-
ing on the condition of acclimatization and the pH in fermentation
system (Sung and Liu, 2003).
According to Fig. 4(b), the mount of TAN in PM1 grew from
1289.5 to 1397.2 mg/L on the 15th day, and then fell to
1214.1 mg/L on the 22th day, after when it rebounded
to 1602.7 mg/L on 43rd day. TAN in PM2 increased from 1367.8 to
1706.3 mg/L in the first 15 day, whose increasing rate was faster
than that in PM1. TAN in PM2 reached the maximum of
1866.5 mg/L on the 22nd day, which exceeded the range of
1500 mg/L and pH value reached 7.6. Thus, biogas production of
PM2 (2675 ml) was significantly less than that of PM1 (15020 ml).
Compared with PM, the fluctuation ranges of TAN in DM were
smaller. The amount of TAN in DM1 reached the peak
(1402.24 mg/L) until at the middle of fermentation. However on
the 8th day, the content of TAN in DM2 reached 1393.3 mg/L and
obtained the second peak with 1433.6 mg/L on the 43rd day. At
the end of fermentation, the TAN of DM1 decreased to slightly less
than the beginning, while the TAN in DM2 (1323.84 mg/L) was
higher than the beginning (1199.5 mg/L).
3.2. Correlation of VFA, pH, total alkalinity and TAN with biogas
production
Anaerobic digestion consists of acid fermentation and methane
production, and these two steps must be balanced. The accumula-
tion of VFA leads to the decrease of pH value, thus affects the
growth of methanogens (Baba et al., 2013). pH value is the pivotal
factor influencing methane production efficiency and it has been
proven that the optimal range of pH to obtain maximal biogas yield
in anaerobic digestion is 6.5–7.5 (Liu et al., 2008). Therefore, VFA
and pH are usually measured to assess the state of pilot plants
(Baba et al., 2013). The process of anaerobic digestion is unstable,
particularly when subjected to changes in the fermentation envi-
ronment (Nain and Jawed, 2006). Such variations often have the
effect of reducing methane production in the digester, so
Fernandez et al. (2008) suggested alkalinity as an ideal parameter
to monitor the anaerobic digestion process. Besides, TAN is the also
an important factor affecting the methane production because the
increase of ammonia/ammonium concentration provides buffering
capacity (Walker et al., 2011). Thus, this work analyzed the effects
of photo-dark fermentation on VFA, TAN, total alkalinity and pH,
and then analyzed the influence of dynamic change of four factors
on biogas production.
3.2.1. Path analysis
Path analysis was carried out to interpret the correlation
between the CBP of PM and the different factors (i.e., VFA, TAN,
total alkalinity, and pH). Table 2 shows the results of path analysis
for PM1. Total alkalinity (X3PM1, 0.6327) had directly negative effect
on CBP, while the other three factors played a positive direct role in
CBP under photo-dark condition. X3PM1 showed the maximum
negative direct effect (0.6327) in all factors, but with the minimum
correlation coefficient (0.219; p < 0.05) to the CBP because it’s neg-
ative direct effect was counterbalanced by the positive indirect effect
of X1PM1, X2PM1 and X4PM1. Whereas, under dark condition (Table 2),
X1PM2 had negative direct effect on CBP (0.5013). X2PM2, with the
maximum positive and direct effect (0.8335), had correlation coeffi-
cient of 0.4714 to the CBP, which was lower than that of X4PM2
(0.9373) with the positive direct effect of 0.7247, because positive
direct effect of X2PM2 on CBP was weakened by negative indirect
effect of X1PM2, X3PM2 and X4PM2.
Table 3 shows the path analysis between DM and the four
factors. The other three factors had positive direct effect on CBP
of DM1 except X3DM1 (0.7327) under the photo-dark condition.
There is a close analogy between PM1 and DM1. X3DM1 also showed
the maximum negative direct effect (0.7327) and the minimum
correlation coefficient (0.319) with the CBP. Its maximum negative
direct effect was also offset by the positive indirect effect of X1DM1,
X2DM1 and X4DM1. X4DM1 had the maximum correlation coefficient
(0.9388, p < 0.05) with CBP. Under dark condition, X2DM2 had the only
negative direct effect (0.4913) and the minimum correlation
coefficient (0.4523) with CBP. Moreover, X4DM2 had the positive direct
effect of 0.7246 and the maximum correlation coefficient with the
CBP.
 D. Yin et al. / Bioresource Technology 166 (2014) 373–380 377

Table 2
Path analysis between the CBP and VFA, TAN, total alkalinity and pH of PM1 and PM2.

Factors p-Value Correlation coefficients (riy) Direct path coefficients (bi) Indirect path coefficients (rijbj)
Total ?X1PM1 ?X2PM1 ?X3PM1 ?X4PM1
X1PM1 0.0253 *
0.5679 0.2227 0.3452 0.2766 0.7613 0.1395
X2PM1 0.0056** 0.6236 0.3417 0.2819 0.0344 0.3355 0.583
X3PM1 0.0341* 0.219 0.6327 0.4137 0.3529 0.2879 0.3487
X4PM1 0.0032** 0.8387 0.5163 0.3224 0.1286 0.2749 0.1761
?X1PM2 ?X2PM2 ?X3PM2 ?X4PM2
X1PM2 0.0456* 0.3412 0.5013 0.8425 0.1766 0.6796 0.3395
X2PM2 0.0371* 0.4714 0.8335 0.3621 0.0265 0.2373 0.1513
X3PM2 0.0044** 0.8951 0.4764 0.4187 0.0344 0.2355 0.6198
X4PM2 0.0026** 0.9373 0.7247 0.2126 0.0286 0.0651 0.1761
Correlation coefficients: X4PM1 > X2PM1 > X1PM1 > X3PM1 X4PM2 > X3PM2 > X2PM2 > X1PM2

Note: X1 – VFA; X2 – TAN; X3 – total alkalinity; X4 – pH.


P < 0.05.

P < 0.01. Absolute values of path coefficients (direct and indirect path coefficients) were used for analysis.

Table 3
Path analysis between the CBP and VFA, TAN, total alkalinity and pH of DM1 and DM2.

Factors p-Value Correlation coefficients (riy) Direct path coefficients (bi) Indirect path coefficients (rijbj)
Total ?X1DM1 ?X2DM1 ?X3DM1 ?X4DM1
X1DM1 0.0300 *
0.8683 0.5127 0.3556 0.2666 0.7613 0.1391
X2DM1 0.0376* 0.7347 0.4418 0.2929 0.0344 0.3255 0.584
X3DM1 0.0641 0.319 0.7327 0.4135 0.3528 0.2879 0.3486
X4DM1 0.0112* 0.9388 0.5663 0.3725 0.1286 0.2750 0.2261
?X1DM2 ?X2DM2 ?X3DM2 ?X4DM2
X1DM2 0.0244* 0.8948 0.4762 0.4186 0.0343 0.2354 0.6197
X2DM2 0.0656 0.4523 0.4913 0.8536 0.1666 0.6806 0.3396
X3DM2 0.0342* 0.7636 0.9637 0.2001 0.1465 0.2253 0.1411
X4DM2 0.0126* 0.9492 0.7246 0.2246 0.0286 0.0749 0.1783
Correlation coefficients: X4DM1 > X1DM1 > X2DM1 > X3DM1 X4DM2 > X1DM2 > X3DM2 > X2DM2

Note: X1 – VFA; X2 – TAN; X3 – total alkalinity; X4 – pH.

*
P < 0.05;
**
P < 0.01. Absolute values of path coefficients (direct and indirect path coefficients) were used for analysis.

Obviously, large direct effect between the factors and the CBP
does not always imply strong correlation between them. Four
factors and CBP formed a complicated path network. Thus, it is
difficult to determine which factors have the decisive effect on
CBP by path analysis. Therefore, further analyses were conducted
to find out the decisive factors and to what extent they determined
the CBP.
3.2.2. Analyses of decisive degree
Under photo-dark condition (Table 4), the sort of decision
coefficients is R4PM1 > R2PM1 > R1PM1 > R3PM1, which indicated pH (R4PM1)
with the largest value became the main decision variable on CBP while
total alkalinity (R3PM1) with the smallest value had limitation decision
on CBP. Whereas, under dark condition, the maximum value of deci-
sion coefficient in PM2 was also obtained by pH (0.8333), however
the limitation factor for CBP with the minimum decision coefficient
value (0.5934) was VFA. The maximum decision coefficient values
in DM did accord with PM, which meant pH became the primary deci-
sion variable with the maximum decision coefficient of 0.5663 and
0.7246 respectively on CBP under the two treatment, while the limita-
tion factors was total alkalinity (R3DM1 = 0.0691) under photo-dark
condition and TAN (R2DM2 = 0.5974) under dark condition.
Above results showed the decision coefficients between the
four factors and CBP, indicating which factor is the decisive
variable under different conditions. According to Eq. (4), factors
with the largest or smallest R2ðiÞ values are the main decisive or lim-
iting variables. Therefore, the work provides guiding for promoting
biogas production by artificial regulation the content of factors in
the process of fermentation according to their effect of R2ðiÞ value
on CBP.
3.3. Biogas production
The different responses of PM and DM to photo-dark fermenta-
tion determined the variation of daily biogas production (DBP) and
time to reach the peak. Fig. 5(a) shows that DBPs of PM2 was lower
than that of PM1 at the initial fermentation stage, which is likely
due to the less reactiveness of methanogen and lower biogas pro-
duction caused by the accumulation of VFA with low pH of PM2
(6.3). Along with the fermentation, DBPs of PM1 and PM2 increased
gradually and achieved the peaks. The maximum DBP of PM1 was
740 mL/d on the 21st day and maximum DBP of PM2 was 419 mL/d
on the 14th day. The pH values were 6.9 and 6.8 when the
maximum yields were obtained, which has been reported that
the maximal biogas yield occurs when pH values was 6.5–7.5
(Liu et al., 2008). After the peak, DBP began to decrease. The
fermentation period of PM2 just lasted for 20 days, which perhaps
because average value of VFA in PM2 (3386.38 mg/L) was higher
than that of in PM1 (3119.62 mg/L) and amount of total ammo-
nium nitrogen in PM2 reached the maximum (1866.5 mg/L) with
a pH of 7.4 on the 22th day. Calli et al. (2005) explained that
ammonia inhibition usually happens when pH is above 7.4 and
total ammonia nitrogen was 1500–3000 mg/L. Moreover, the lower
average total alkalinity in PM2 (5891 mg/L) meant lower buffer
capacity and worse ability to prevent acidification of fermentation,
which may be other feasible reason for the short fermentation time
and lower biogas production than that of in PM1.
378 D. Yin et al. / Bioresource Technology 166 (2014) 373–380

Table 4 (5234 mg/L) failed to provide buffering capacity for digester. On

Analysis of decisive degree for PM and DM. the 15th day, DBP of DM2 reached 654 mL/d, and then DBP gradu-
Factors Determination coefficients R(i) ally decreased. At the end of fermentation, DBP of DM2 showed an
X
2
R2ðiÞ ¼ bi þ 2
2
bi r ij bj ¼ 2bi r iy  bi increasing trend after the 27th day.
j–i
As shown in Fig. 5(b), the CBPs of PM1, PM2, DM1 and DM2 were
X1PM1 R1PM1 = 2  (0.2227)  (0.5679)  (0.2227)2 = 0.2033 15020, 2675, 10788, and 15447 mL, respectively, which illustrates
X2PM1 R2PM1 = 2  (0.3417)  (0.6236)  (0.3417)2 = 0.3094 that PM and DM were different in response to photo-dark
X3PM1 R3PM1 = 2  (0.6327)  (0.2190)  (0.6327)2 = 0.1232
condition for fermentation. Analysis of decisive degrees showed
X4PM1 R4PM1 = 2  (0.5163)  (0.8387)  (0.5163)2 = 0.5995
X1PM2 R1PM2 = 2  (0.5013)  (0.3412)  (0.5013)2 = 0.5934 pH had the maximum determination coefficient among the four
X2PM2 R2PM2 = 2  (0.8335)  (0.4714)  (0.8335)2 = 0.0911 sets. However, as can be seen from Fig. 3(b), the change interval
X3PM2 R3PM2 = 2  (0.4764)  (0.8951)  (0.4764)2 = 0.6259 and trends of pH were similar in the four sets, which illustrates
X4PM2 R4PM2 = 2  (0.7247)  (0.9373)  (0.7247)2 = 0.8333 that the significant difference of CBP between PM and DM was
X1DM1 R1DM1 = 2  (0.5127)  (0.8683)  (0.5127)2 = 0.6274
X2DM1 R2DM1 = 2  (0.4418)  (0.7347)  (0.4418)2 = 0.4540
caused by other three factors.
X3DM1 R3DM1 = 2  (0.7327)  (0.3192)  (0.7327)2 = 0.0691 PM1 obtained 15020 mL CBP that was 5.6 times as much as PM2,
X4DM1 R4DM1 = 2  (0.5663)  (0.9388)  (0.5663)2 = 0.7426 which indicated the photo-dark condition promoted the fermenta-
X1DM2 R1DM2 = 2  (0.4762)  (0.8948)  (0.4762)2 = 0.6254 tion for PM. PM easily occurs hydrolysis acidification resulted from
X2DM2 R2DM2 = 2  (0.4913)  (0.3623)  (0.4913)2 = 0.5974
the fact that pig manure had large amounts of protein and
X3DM2 R3DM2 = 2  (0.9637)  (0.7636)  (0.9637)2 = 0.5430
X4DM2 R4DM2 = 2  (0.7246)  (0.9492)  (0.7246)2 = 0.8505 carbohydrates, while the lipids content was small (Chen, 2005).
So, the total alkalinity with ability to prevent acidification has
R4PM1 > R2PM1 R4PM2 > R3PM2 > R2PM2 > R1PM2
> R1PM1 > R3PM1 the important impact on CBP. The average total alkalinity in PM1
R4DM1 > R1DM1 R4DM2 > R1DM2 > R3DM2 > R2DM2 (6684.8 mg/L) was significantly higher than in PM2 (5891 mg/L),
> R2DM1 > R3DM1 which meant PM1 has the strong buffer capacity to prevent acidi-
Note: X1 – VFA; X2 – TAN; X3 – total alkalinity; X4 – pH; i = 1, 2, . . ., p; bi was direct fication and promote biogas production. In addition, Calli et al.
path coefficient; riy was correlation coefficient between Xi and Y. (2005) explained that ammonia inhibition usually happens when
pH is above 7.4 and total ammonia nitrogen was 1500–3000
mg/L. The Fig. 4(b) showed the amount of TAN in PM2 reached
1515.2 mg/L on 8th day, then reached the peak (1866.5 mg/L) with
a pH of 7.4 on the 22nd day. So, the accumulation of TAN inhibited
the CBP of PM2 under the dark condition and led to the low CBP.
For DM, dark treatment was a more suitable fermentation con-
dition because CBP of DM2 was 1.43 times as much as DM1. Wang
et al. (1999) and Siegert and Banks (2005) suggested that high VFA
concentration in lead the inhibition of methanogen and caused
decrease in pH. While Fig. 3(a) showed that the concentration of
VFA in DM1 lasted more than 8300 mg/L from 8 to 15 day in fer-
mentation and peaked 8378.5 mg/L on 15th day, which led to the
DBP of DM1 was lower than in DM2 during the same period and
the CBP of DM1 (1759 mL) was lower than DM2 (2712 mL) as well.
Furthermore, the same as PM, the lower average total alkalinity
with weaken buffer capacity in DM1 (6024 mg/L) than in DM2
(7249 mg/L) resulted in lower biogas production in DM1.
Therefore, the results of the work provided the reference for the
improvements of biogas digesters in physical structure and mate-
rials, because digesters constructed by transparent materials are
more favorable for CBP of PM and more conductive for observing
the internal change of digester, then timely control fermentation
system when the raw materials occur the phenomena of acidificat-
ion, crust or floating.

3.4. Economic benefit evaluation

The photo- dark fermentation had significant effect on CBP of

Fig. 5. Daily biogas production (a) and cumulative biogas production (b) of PM and
DM.
The trend of DBPs of DM1 and DM2 (Fig. 5(a)) was different from
that of PM. The time to reach the peak rates of both DM1 and DM2
were longer compared to PM1 and PM2 because of the high accu-
mulation of VFA in DM1 (4726.6 mg/L) and DM2 (3932.3 mg/L) lead
to the decrease of pH value, thus affecting the growth of methano-
gens (Baba et al., 2013). Fluctuating DBP of DM1 showed three
peaks with values of 406, 530 and 470 mL/d on the 19th, 26th
and 50th day, respectively. Compared with DM1, DM2 produced
biogas earlier and the initiative DBP achieved 210 mL/d which
was probably due to the low initiative total alkalinity of DM1
PM rather than DM. Thus, only the economic impact on the CBP
of PM was evaluated.
This work set 8 m3 in size, a commonly used in household
biogas, as an example to analyze the photo-dark fermentation
impact on the CBP of PM. The difference of digester reformed from
concrete structure to hard plastic or fiberglass structure were listed
in Table 5 that showed the digester made by transparent materials
has significant advantages in terms of useful life, construction time
and weight, although the price of fiberglass digester is slight higher
than the traditional brick and concrete digester (Li, 2008; Zhang
et al., 2006).
In accordance with Section 2.2, 1 L glass reactor filled with
700 g row material including inoculum, PM and water produced
15.020  103 m3 biogas under photo-dark condition and
 D. Yin et al. / Bioresource Technology 166 (2014) 373–380
Table 5
Comparison of three biogas digesters made of different materials.
Materials The thickness Digester
of digester (cm) capacity (m3)
Bricks, cement, and sand 10 8 15–20
Plexiglass and fiberglass 8–10 8 20–25
Hard plastic 5–8 8 10–15
2.675  103 m3 under dark condition in 53 days, which indicates
CBP was increased by 12.345  103 m3 in a 1 L digester by
photo-dark fermentation. Thus, the inferred increase of CBP is
98.4 m3 in a 8 m3 household biogas digester under photo-dark
condition, which not only prolongs approximate one month of
biogas using time for a rural family with five people, but also
increases 39.4 yuan for farmers in whole fermentation period
(the price of biogas is 0.4 yuan/m3) (Shi, 2001).
Therefore, it is significant to study the effects of photo-dark
fermentation on biogas production. Further work may also focus
on the effects of photo-dark condition on fermentation by
photoheterotrophic bacteria. Yu et al. (2008) proposed a kind of
non-methanogens, photoheterotrophic bacteria that can synthetic
organic materials using carbon source and light energy in the
process of fermentation. In addition, other raw materials, such as
other livestock manure, straw, food waste and sludge, could be
studied since they all have large production rates and their biogas
production rates could be potentially increased through
photo-dark condition.
4. Conclusion
This study suggested that the fermentation of PM and DM can
be carried out under both photo-dark and dark conditions. CBP
values of PM1, PM2, DM1 and DM2 were 15020 mL, 2675 mL,
10788 mL and 15447.5 mL, respectively. PM1 achieved higher
biogas production than PM2, which illustrates that appropriate
light can promote anaerobic fermentation of PM, while CBP of
DM2 was 1.43 times as much as DM1 shows dark fermentation
was more suitable for DM. Furthermore, pH value had the maxi-
mum determination coefficient with the CBP among all four sets
and total alkalinity presented negative correlation with CBP of
PM1 and DM1.
Acknowledgements
This work was supported by Science and Technology Support
Projects (2011 BAD15B03) from Ministry of Science and Technol-
ogy Department of the People’s Republic of China and the Basic
Scientific Fund of Northwest A&F University (QM2012002).
References
Apha, A., 1995. Wpcf, Standard Methods for the Examination of Water and
Wastewater. American Public Health Association/American Water Works
Association/Water Environment Federation, Washington, DC, USA.
Argun, H., Kargi, F., 2010. Effects of light source, intensity and lighting regime on
bio-hydrogen production from ground wheat starch by combined dark and
photo-fermentations. Int. J. Hydrogen Energy 35, 1604–1612.
Astals, S., Nolla-Ardèvol, V., Mata-Alvarez, J., 2012. Anaerobic co-digestion of pig
manure and crude glycerol at mesophilic conditions: biogas and digestate.
Bioresour. Technol. 110, 63–70.
Baba, Y., Tada, C., Watanabe, R., Fukuda, Y., Chida, N., Nakai, Y., 2013. Anaerobic
digestion of crude glycerol from biodiesel manufacturing using a large-scale
pilot plant: methane production and application of digested sludge as fertilizer.
Bioresour. Technol. 140, 342–348.
Burton, C.H., Turner, C., 2003. Manure Management: Treatment Strategies for
Sustainable Agriculture. Editions Quae, Silsoe Research Institute, UK.
379
Useful Time for Price (yuan) Weight (kg)
life (years) construction (days)
15–20 1700 2000
5–10 1800 200
3–5 1200 146
Calli, B., Mertoglu, B., Inanc, B., Yenigun, O., 2005. Effects of high free ammonia
concentrations on the performances of anaerobic bioreactors. Process Biochem.
40, 1285–1292.
Chae, K.J., Jang, A., Yim, S.K., Kim, I.S., 2008. The effects of digestion temperature and
temperature shock on the biogas yields from the mesophilic anaerobic
digestion of swine manure. Bioresour. Technol. 99, 1–6.
Chen, S.Y., 2005. Multivariate Statistical Analysis. Hwatai Books Co, Taipei.
Cuetos, M.J., Fernández, C., Gómez, X., Morán, A., 2011. Anaerobic co-digestion of
swine manure with energy crop residues. Biotechnol. Bioprocess Eng. 16, 1044–
1052.
EI-Mashad, H.M., Zeeman, G., Van Loon, W.K., Bot, G., Lettinga, G., 2004. Effect of
temperature and temperature fluctuation on thermophilic anaerobic digestion
of cattle manure. Bioresour. Technol. 95, 191–201.
Fernandez, N., Montalvo, S., Borja, R., Guerrero, L., Sanchez, E., Cortes, I.,
Colmenarejo, M., Travieso, L., Raposo, F., 2008. Performance evaluation of an
anaerobic fluidized bed reactor with natural zeolite as support material when
treating high-strength distillery wastewater. Renew. Energy 33, 2458–2466.
Hansen, K.H., Angelidaki, I., Ahring, B.K., 1999. Improving thermophilic anaerobic
digestion of swine manure. Water Res. 33, 1805–1810.
Hobson, P., Bousfield, S., Summers, R., Mills, P., 1980. Anaerobic digestion of piggery
and poultry wastes. In: Stafford, D.A., Wheatley, B.I., Hughes, DE (Eds.). Anaerobic
Digestion: [Proceedings of the first International Symposium on Anaerobic
Digestion, held at University College, Cardiff, Wales, September 1979].
Jin, H.M., Fu, G.Q., Chang, Z.Z., Ye, X.M., 2013. Distribution of pb and its chemical
fractions in liquid and solid phases of digested pig and dairy slurries. Trans.
Chin. Soc. Agric. Eng. 29, 218–225.
Li, D.S., 2008. Study on the status of frp biogas digester. J. Liaoning Agric Coll. 10,
36–37.
Liu, C.F., Yuan, X.Z., Zeng, G.M., Li, W.W., Li, J., 2008. Prediction of methane yield at
optimum ph for anaerobic digestion of organic fraction of municipal solid
waste. Bioresour. Technol. 99, 882–888.
Nain, M., Jawed, M., 2006. Performance of anaerobic reactors at low organic load
subjected to sudden change in feed substrate types. J. Chem. Technol.
Biotechnol. 81, 958–965.
Shi, J., 2001. Refixxing of price of marsh gas in countryside. Acta Agric. Jiangxi 13,
55–57.
Shi, J.C., Liao, X.D., Wu, Y.B., Liang, J.B., 2011. Effect of antibiotics on methane arising
from anaerobic digestion of pig manure. Anim. Feed Sci. Technol. 166, 457–463.
Siegert, I., Banks, C., 2005. The effect of volatile fatty acid additions on the anaerobic
digestion of cellulose and glucose in batch reactors. Process Biochem. 40, 3412–
3418.
Song, M., Shin, S.G., Hwang, S., 2010. Methanogenic population dynamics assessed
by real-time quantitative pcr in sludge granule in upflow anaerobic sludge
blanket treating swine wastewater. Bioresour. Technol. 101, S23–S28.
Sung, S., Liu, T., 2003. Ammonia inhibition on thermophilic anaerobic digestion.
Chemosphere 53, 43–52.
Van Haandel, A.C., Lettinga, G., 1994. Anaerobic Sewage Treatment: A Practical
Guide for Regions with a Hot Climate. John Wiley & Sons, Chichester.
Van Lier, J.B., 1995. Thermophilic Anaerobic Wastewater Treatment, Temperature
Aspects and Process Stability (Ph.D. thesis). Department of Environmental
Technology, Wageningen University, The Netherlands.
Walker, M., Iyer, K., Heaven, S., Banks, C., 2011. Ammonia removal in anaerobic
digestion by biogas stripping: an evaluation of process alternatives using a first
order rate model based on experimental findings. Chem. Eng. J. 178, 138–145.
Wang, Q.H., Kuninobu, M., Ogawa, H.I., Kato, Y., 1999. Degradation of volatile fatty
acids in highly efficient anaerobic digestion. Biomass Bioenergy 16, 407–416.
Wang, X.J., Yang, G.H., Feng, Y.Z., Ren, G.X., Han, X.H., 2012. Optimizing feeding
composition and carbon–nitrogen ratios for improved methane yield during
anaerobic co-digestion of dairy, chicken manure and wheat straw. Bioresour.
Technol. 120, 78–83.
Wang, X.J., Yang, G.H., Li, F., Feng, Y.Z., Ren, G.X., Han, X.H., 2013. Evaluation of two
statistical methods for optimizing the feeding composition in anaerobic co-
digestion: mixture design and central composite design. Bioresour. Technol.
131, 172–178.
Ward, A.J., Hobbs, P.J., Holliman, P.J., Jones, D.L., 2011. Evaluation of near infrared
spectroscopy and software sensor methods for determination of total alkalinity
in anaerobic digesters. Bioresour. Technol. 102, 4083–4090.
Yang, Z., Zhang, H.L., 2008. Strategies for development of clean energy in China. Pet.
Sci. 5, 183–188.
Yin, D.X., Liu, J.J., Ji, Y.M., Liu, W., Yang, G.H., Ren, G.X., Feng, Y.Z., Zhou, X.W., 2012.
Study on effect of illumination intensity on pig manure, cow dung’ anaerobic
fermentation. J. Agro-Environ. Sci. 31, 428–434.
380 D. Yin et al. / Bioresource Technology 166 (2014) 373–380

Yu, F.B., Luo, X.P., Guan, L.B., Zhang, M.X., Shan, S.D., 2008. Research advances in Zhang, W.J., Xie, J., Zhang, Z.Y., Song, H.C., Xia, C.F., 2006. Research and application of
biogas fermentation microorganism. J. Anhui Agric. Sci. 36, 15658–15660. mechanical massive production of plastic digester: new progress. Trans. CSAE
Yuan, Z.F., Zhou, J.Y., Guo, M.C., Lei, X.Q., Xie, X.L., 2001. Decision coefficient-the 22, 80–85.
decision index of path analysis. J. Northwest A&F Univ. 29, 131–133.