During acid hydrolysis of lignocellulosics, aliphatic acids (acetic, formic, and lev — ulinic acid), furan derivatives, and phenolic compounds are formed in addition to the sugars. Furfural and 5-hydroxymethyl furfural (HMF) are the most important furans, formed by decomposition of pentoses and hexoses respectively [24]. Acetic acid has been reported in the hydrolysis of the acetyl groups into hemicellulose as a consequence of deacetylation of acetylated pentosan [25]. Multiple phenolic com­pounds are derived from lignin, including vanillin, vanillic acid, vanillyl alcohol, 4-hydroxybenzoic acid, 4-hydroxybenzaldehyde, coumaric acid, syringaldehyde, syringic acid, cinnamaldehyde, dihydroconiferyl alcohol, hydroquinone, catechol, veratrole, acetoguaiacetone, homovanillic acid, and Hibbert’s ketones [25]. HMF is converted at a lower rate than furfural, which may be due to lower membrane permeability and cause a longer lag-phase in the growth of microorganisms [26]. The phenolic compounds penetrate biological membranes and cause them to lose integrity, thereby affecting the membranes’ ability to serve as selective barriers. The microbial growth was found to be inhibited in the presence of acetic acid (>3.5 g/l) in hemicellulosic hydrolysates, this phenomenon may occur due to the inflow of undissociated acid into cytosol [26].

From an economic, social and environmental perspective, lignocellulosic biomass wastes are good feedstocks for methane production through AD. Due to the slow hydrolysis of lignocellulose, however, methane production is slow, and a long reten­tion time and large digester volumes are required to produce enough methane biogas for cost-effective recovery. In the case of livestock manure, 40-50% of the solid passes through mesophilic AD undigested [8]. Two-stage AD processes can improve solid reduction and stability by separating the more robust hydrolysis and acidogenesis from the less robust syntrophic acetogenesis and methanogenesis [13]. TPAD digesters are promising two-stage designs, with the hydrolysis being enhanced in the first digester operated at an elevated temperature (typically at 55°C) and syntrophic acetogenesis and methanogenesis being enhanced and stabilized in the second digester operated at a mesophilic temperature (typically 35°C) [46, 78]. Indeed, significant increases in hydrolysis, TS reduction, and methane production resulted from the co-digestion of a primary sludge and OFMSW in a TPAD sys­tem [76]. Additionally, TPAD enhances sanitation of waste streams [73], reducing potential risks associated with certain types of feedstocks (e. g., municipal sludge and animal manures). Furthermore, TPAD processes eliminate the AD inhibition caused by the self-heating of mesophilic AD of high-energy feedstocks (e. g., energy crop and OFMSW) [49, 52]. The higher energy input required to operate TPAD is more than offset by the increased biogas and heat produced therefrom [22]. The TAPD technology will probably be applied more commonly in the near future when more lignocellulosic feedstocks (e. g., energy crops, animal manure, crop residues, and OFMSW) are subjected to AD.

Size reduction can dramatically enhance the AD of certain feedstocks, such as crop residues, OFMSW, and energy crops. Physical and chemical pretreatments can further enhance AD of these feedstocks [45, 53], but currently they may not be cost — effective, especially for those feedstocks that contain high water contents and for wastewaters. Low cost and efficient pretreatments need to be developed.

The entire AD process is often limited by three of the four steps of the AD process: hydrolysis, syntrophic acetogenesis, and methanogenesis. Hydrolysis of biomass polymers is typically the rate-limiting step of the entire AD process of lignocellulolytic feedstocks. Single or mixed cultures of lignocellulolytic microbes may be used to augment the capability of hydrolysis in digesters as exemplified by enhanced AD of cattle manure [62] and municipal sludge [30]. Methanogenesis can become the rate-limiting step when feedstocks containing large amounts of read­ily fermentable substrates (e. g., starch) are digested. In this scenario, acid-tolerant methanogen (e. g., Methanobrevibacter acididurans) cultures may be prepared and used to enhance the entire AD process or remediate upset AD operation. Bioaugmentation can also enhance the AD of feedstocks containing high concen­trations of particular substances, such as lipids [19]. As more and more digesters are put into operation, there will be increasing needs for such specialty cultures to enhance existing digesters, start up new digesters, and prevent AD failures.

Community DNA was obtained from samples by using a kit for extracting DNA from soil samples, e. g. UltraClean Soil DNA (Mo Bio Laboratories, USA), which includes different scales, and their use is recommend depending on the sample condition.

1.2.1 16S rDNA Clone Analysis

Construction of 16S rDNA Clone Library

Community DNA was used as the template DNA. Bacterial partial 16S rDNA (about 1,500 bp) was amplified by PCR with a forward primer B27F (5′- AGAGTTTGATCCTGGCTCAG, Escherichia coli position 9-27) and a reverse primer U1492RM (5′-GGYTACCTTGTTACGACTT, E. coli position 1512-1492) [17]. Amplification by PCR comprised 25 cycles of 30 s at 94°C, 30 s at 60°C, 1.5 min at 72°C, and a final extension of 5 min at 72°C using Ex Taq DNA polymerase (Takara Bio, Japan). PCR products were purified using a DNA purifi­cation kit (e. g. GFX PCR DNA and Gel Band Purification Kit, GE Healthcare, UK) and cloned into the plasmid vector (e. g. pT7Blue T-Vector, Novagen, Germany).

Transformation of E. coli and Sequencing of 16S rDNA Clone Library

E. coli strain DH5 alpha was transformed with this plasmid library. Plasmid DNAs were prepared from the cultures. The 16S rDNAs were sequenced using an appropriate DNA sequencer (e. g. DNA Analyzer 3730 xl, Applied Biosystems, USA).

Homology Search and Estimation of Phylogenetic Affiliations

A homology search was conducted using BLASTN database (BLAST, http://www. ncbi. nlm. nih. gov/BLAST/) [18]. Checks for chimeric sequences were conducted using the software Pintail [19] which is available from the Ribosomal Database Project, followed by NCBI BLASTN database [18].

Analysis of Homologous Coverage

Coverage of the clone library describes the extent to which the sequences in the library represent the total population. In order to calculate the coverage of a library, the criterion for what constitutes a unique sequence must first be decided. Various studies have used different criteria, generally based on sequence similarity (e. g. 97 or 99% similarity) or evolutionary distance (e. g. < 0.01). These values can then be used to plot a coverage curve (C vs D) that describes how well the library represents the total community given varying criteria of uniqueness. The homologous coverage (Cx) is calculated by the following formula Cx = 1-(Nx/n), where Nx is the number of unique sequences in the sample and n is the total number of sequences [20].

Construction of a Phylogenetic Tree

Sequences are aligned using the Clustal W program version 1.7 [21, 22], and all sites with gaps in any sequences and the regions of PCR primers are removed from the alignment. The phylogenetic trees are constructed by the neighbor-joining method [23] or the maximum-likelihood method [24] using the PHYLIP (Phylogeny Inference Package) program version 3.5c (http://evolution. genetics. washington. edu/phylip. html) [25]. The stability of relationships was assessed by performing bootstrap analyze of the neighbor-joining data based on 1,000 resamplings.

The first TGER prototype (Fig. 3) was built as a part of a Phase IISTTR (Small busi­ness Technology Transfer Research) program and demonstrated proof of principle,

Fig. 4 TGER after retrofit

but was not rugged enough to deploy to an OCONUS (outside the continental United States) site for field testing and validation. The initial function of the follow-on effort was to upgrade the existing prototype with better, more advanced equip­ment that could withstand the stresses of a three month OCONUS deployment in an operationally harsh environment (Fig. 4).

Three of the key improvements identified during testing of the Phase II TGER and applied during the retrofit and fabrication are highlighted below.

(1) First stage materials preparation (Industrial shredder and separations system). This component combines several key tasks which currently are done on the original prototype with separately acquired and integrated third-party compo­nents. Tasks include shredding, rinsing, auguring and compacting bioreactor residuals. The Industrial shredder performs these functions as a single com­ponent with half of the electrical power required by the original TGER. The new Industrial shredder was retrofitted onto the original prototype and included during fabrication of the second prototype.

(2) Second stage pelletizer. Testing demonstrated that the size and shape of the pel­lets were the most critical qualities of gasifier feed-stock, followed by pellet density and then proportions of waste content (plastic vs. cellulosic, other). Our original view of the feedstock had focused on the latter, i. e. waste content pro­portions, and had used a less expensive compaction channel for gasifier pellets. Subsequent off-line testing with pellets made with equipment demonstrated a marked improvement in gasifier performance and subsequent engine output. The pelletizer, shown in Fig. 5, was included in the second TGER design and was a retrofitted improvement to the original prototype.

(3) Stainless steel commercial grade distilling column. The stainless steel distilling column was upgraded from standard steel to stainless to prevent the introduction of rust into the distilling apparatus [9].

Hydraulic retention time (HRT) influences the H2 generation process significantly. Reducing HRT from 18 to 12 h has improved H2 yield without affecting substrate removal efficiency [57]. Maximum H2 yield was reported between 0 and 14 h in all the experimental variations studied in batch mode during dairy, chemical and distillery wastewater treatment [22, 25, 31, 38]. Longer fermentation periods induce a metabolic shift from the acidogenic process to the methanogenic pro­cess which is unfavorable for H2 production. Shorter HRT’s have been shown to dilute out slow growing MB [86]. However, in continuous operation mode, H2 production was observed at long HRTs of 3 days (pH 6.4) without encountering problems with methanogenesis [87]. Optimal HRT mostly depends on the nature and composition of the substrate, function of biocatalyst, loading rate and fermen­tation pH employed. HRT can be considered as a readily manipulated variable for process control. Optimum HRTs from 8.0 to 14 h were reported for effective H2 production [57].

Animal manures represent a huge methane biogas potential. As estimated, 106 million dry tons of animal manures are produced each year in the USA, with approx­imately 87 million dry tons being available for methane biogas production [69]. Given a BMP of 200-400 m3 CH4/dry ton [8], the amount of animal manures available for AD provides a potential of 17-35 billion m3 of CH4 per year in the USA. The animal manures produced from confined animal feeding operations (CAFOs) offer one of the most abundant single feedstocks available for large-scale methane biogas productions. The composition and physical features (e. g., water contents) of animal manures vary widely from species to species and from oper­ation to operation [58]. In general, animals manures have relatively high water contents, ranging from 75% (poultry manure) to 92% (beef cattle manure). Most of the animal manure is organic matter, with VS contents ranging from 72% (poul­try manure) to 93% (beef cattle manure) of TS. Inorganic nutrients, including N,

P and K, are rich in animal manures, especially poultry manure. Because most of the readily degradable substances, especially carbohydrates, have been digested and absorbed by the animals, animal manures have very little readily fermentable sub­strates. Additionally, animal manures have high concentrations of amino nitrogen such as urea and ammonia and a large pH buffering capacity against acids. Thus, the fermentative acidogenesis during AD of animal manures typically does not result in significant pH decline, but high concentrations of ammonia can result, causing toxicity to methanogens, especially in thermophilic digesters where methanogens are very susceptible [43]. Furthermore, animal manures contain large amounts of microbial biomass, including bacteria and methanogens. Consequently, AD reac­tors digesting animal manures, especially livestock manures, can be started without the addition of external digested sludge as a start culture or inoculum.

Because of the relatively low contents of readily degradable substances, the methane biogas production from animal manures is generally slow. Thus, when digested alone a long retention time is needed. Co-digestion with nitrogen-poor yet carbohydrate-rich feedstocks, such as food-processing wastes and OFMSW, can substantially enhance CH4 production and stabilize the AD process of animal manures [59, 94]. Some animal manures, especially dairy cattle manure, contain sand from the sand bedding [42], which settles in AD reactors and can cause oper­ational problems if not dealt with properly. Due to the large differences in many physicochemical characteristics and degradability, different manures may require different AD technologies for efficient and cost-effective AD. Here the AD tech­nologies suitable for beef manure, diary manure, swine manure, and poultry litter will be discussed.

The influence of self-immobilization of enriched acidogenic mixed consortia on fermentative H2 production was studied on different supporting materials [SBA — 15 (mesoporous) and activated carbon (granular; GAC and powder; PAC)] [126].

2.5

>

03

T3

2.0

E

E

c

о

1.5

О

£

с

О)

О)

о

1.0

>

X

ф

>

го

0.5

Е

3

О

0.0

Suspended growth (SG-control) of cultures showed inhibition in terms of both H2 production and substrate degradation, especially at higher loading rates. On the other hand, attached growth showed marked improvement in both H2 yields and substrate degradation efficiency, particularly at higher loading rates. Self­immobilization on SBA-15 showed nine times higher H2 production than the non-attached (SG) operation. Attached growth on GAC and PAC also showed marked improvement in the process performance at higher OLRs compared to SG operation. Immobilization of microflora on the support medium as biofilm results in high biomass hold up, which enabled the operation of the process at signifi­cantly higher liquid throughputs and OLRs. Immobilization protects the cells from environmental/chemical toxicity and from predation by other organisms and may enhance survival under extreme environments with relatively high survival rates even after prolonged storage [26, 92, 126]. Immobilized cells survive even at high temperatures.

In order to enhance the efficiency of hydrolysate fermentation, several detoxification methods have been employed, including chemical, physical, and biological methods [25]. These methods include neutralization, overliming, use of ion exchange resins, adsorption onto activated charcoal or tin oxides, and treatments with enzymes such as peroxidase and laccase [3, 25]. Since detoxification increases the cost of the pro­cess, it is important to either overcome the need for detoxification steps or develop cheap and efficient detoxification methods. Overliming with CaO or Ca(OH)2 is a classical chemical detoxification method. It efficiently removes furans and phe — nolics with marginal loss of sugars [24]. Organic solvents such as ether or ethyl acetate have also been applied to extract most of the inhibitors, such as phenolics, weak acids, and furans [25].

Activated charcoal treatment is an efficient and economical method of remov­ing phenolic compounds, acetic acid, aromatic compounds, furfural, and HMF by adsorption [25]. Biological detoxification is another method that enhances the fermentability of hydrolysates, substantially eliminating phenolic compounds. An enzymatic method using laccase was developed to eliminate the impurities of phe­nolic monomers and phenolic acids from hemicellulosic hydrolysates of sugarcane bagasse [24].

AD process control on current digesters is still relying on input and output data: primarily biogas yield and composition, and pH. When the output data suggest any abnormality in performance, it is often too late to intervene, leading to severe disrup­tion of normal operation. Thus, there is an urgent need for research and development of on-line systems that can monitor important parameters of the actual AD pro­cess. Some of the key parameters of AD and their modeling have been reported [35, 71], which can guide the research effort to develop online monitoring systems. Propionate was recently identified to be an important indicator of AD performance [12, 63], and online monitoring of this important SCFA using gas chromatogra­phy seems promising [12, 70]. Further understanding of the microbial communities involved in AD processes may also allow for the development of biosensors that can achieve microbe-based continuous online monitoring. Such real-time monitor­ing can directly link to automated digester controls, such as loading and mixing. Advanced understanding of the microbial community structure, population dynam­ics, metabolic kinetics, and online monitoring in digesters will also improve the modeling of AD processes [6, 35].

Fabrication of the second TGER prototype began in early March 2008 and was completed in three weeks. During fabrication, additional modifications were applied to the second prototype that could not be applied to the first. These modifications are discussed in more detail below.

a. Water circulation system. The material rinsing water was routed away from the main system through an intermediate sump pump and into a 500 gallon tank (see Fig. 6), and then routed back into the wash tank on the system using a sump pump. There were several reasons for this modification. First, the intermediate sump pump broke up any large debris (e. g. food slop and paper material) that passed through the sieve. This ensured that the re-circulated liquids would not cause any clogging of the plumbing. Using the large 500 gallon tank at ground level also made it easier and more efficient for the operators to monitor the fermentation process and add the necessary biocatalysts.

b. Rubber/flexible plumbing. The plumbing on the first TGER prototype was fabricated using standard two inch PVC pipe. When operating in freezing tem­peratures, water would collect in the pipes after operation, freeze overnight and cause the pipes to burst, causing significant delays in operation due to the time required to repair the pipes. The second TGER prototype therefore used a flex­ible rubber hose with quick disconnect fittings instead of pipes, allowing the water to be drained from the hoses after operation in order to prevent the pipes from freezing. Flexible hosing also eliminated the possibility of pipes breaking

due to excessive vibration of the TGER either while in operation or during transport.

c. Chiller. During testing of the first prototype, a chiller was needed to efficiently and quickly condense the distilled ethanol into a liquid state and collect it in the ethanol fuel tank. Due to design issues, the chiller could not be retrofitted on the first prototype but was included on the second. The chiller cooled a mixture of 50% water and 50% antifreeze and circulated it into a heat exchanger (condenser) where the ethanol vapor would condense into liquid ethanol, allowing the TGER to operate efficiently in hotter climates.

d. Reflux valve. The reflux valve is a programmable valve that automatically redi­rects condensed ethanol from the condenser to either the ethanol storage tank or back to the distillation column at a 5:2 time ratio. By redirecting condensed ethanol back into the distillation column at a 5:2 time ratio, the ethanol purity improved from 80% to 85%.

e. Pellet auger/elevator. An external pellet elevator was purchased in order to automate the process of supplying waste-derived pellet fuel into the downdraft gasifier (Fig. 7). On the original prototype, a technician was required to climb onto the top of the TGER in order to pour waste pellets from a bucket into the gasifier, a time consuming and unsafe process. The pellet elevator allowed the technician to dump the pellets into a large collection bin at ground level and the pellet elevator would automatically deliver the correct quantity of pellets into the gasifer based on data received from an infrared sensor suspended over the gasification chamber.

f. Centrifuge pump and basket filter configuration. On the original prototype, the centrifuge pump and basket filter had to be installed on their side. In order

Fig. 7 Pellet auger/elevator

to achieve optimal performance from the pump and filter it is necessary to install them upright. The frame on the second prototype was redesigned to accommodate an upright installation of both the pump and filter.