Biogas from Fresh Spring and Summer Grass: Effect of the Harvesting Period

: Yard trimmings, landscape management and agricultural practices determine the collection of biomass currently destined mainly to the production of a valuable soil amendant by composting. While composting requires energy, especially for the turning/aeration phases and for air treatment (i.e., bioﬁlters in the case of enclosed systems), anaerobic digestion represents an energy positive process that results in production of biogas and digestate, which can be used as fuel and fertilizer, respectively. The focus of the present research was the evaluation of biogas and methane potential of grass collected in two different periods of the year (spring and summer) from riverbanks located in Northern Italy. The conversion to biogas of feedstocks is greatly inﬂuenced by the composition of the organic matter, content of cellulose, and lignin in particular. The production of biomass per hectare and the consequent biogas production were also evaluated. The experimental tests were performed on both samples of fresh grass in laboratory scale batch reactors, characterized by 4.0 L of volume and operated in mesophilic conditions (38 ◦ C), for 40 days per cycle. The anaerobic digestion process was performed on a mixture of inoculum and grass, characterized by inoculum:substrate VS (volatile solids) ratio equal to 2. The inoculum was represented by digestate from a full-scale anaerobic digestion plant fed with dairy cow manure. The results in terms of biogas production, biogas quality (CH 4 , CO 2 , H 2 S), and emissions from digestates (NH 3 , CO 2 and CH 4 ) are presented in the paper. Total solids (TS), volatile solids (VS), pH, volatile fatty acids (VFA), alkalinity, acidity vs. alkalinity ratio, ﬁbers (cellulose, lignin), and total Kjieldahl nitrogen (TKN) were determined both on input and output of the process. The biogas yield obtained from grass resulted higher than expected, quite similar to the yield obtained from energy crops, with Biomethane Potential (BMP) of 340.2 NL · kg − 1 VS and of 307.7 NL · kg − 1 VS, respectively, for spring and summer grass. Biogas quality was slightly lower for summer grass, perhaps in relation to the higher content of ﬁbers (lignin). Alternatively, the yield of grass per surface was signiﬁcantly different between spring and summer with the highest production in the summer. In fact, the results revealed a methane yield of 263 Nm 3 · ha − 1 and of 1181 Nm 3 · ha − 1 , respectively for spring and summer grass.


Introduction
Sustainable development is a current notion strictly related to the concept of a circular economy. In this optic, the production of renewable energy by the valorization of wastes or by-products is considered as one of the most dominant future renewable energy sources [1][2][3][4][5].

Lab Scale Anaerobic Digestion System and Experimental Setup
The lab scale AD system was composed of six digesters, each with a volume of 4 L and equipped with propeller-type mixers ( Figure 2). Temperature control was achieved by a thermostatic bath heated by an electric resistance controlled by a digital thermostat with a water recirculation pump to maintain homogeneous conditions in the tank.
The biogas line was composed by condensation traps and biogas meters (MilliGascounter ® , Ritter, Bochum, Germany) that continuously registered the volume of produced biogas, normalizing it at the normal conditions of temperature and pressure (273.15 K; 101.33 kPa). Biogas quality (CH4, and CO2) was determined by Siemens Ultramat IR analyzer (Siemens Automation Group, Karlsruhe, Germany) and ProTec gas pump with Kitagawa (Kitagawa, Japan) detection tubes (H2S).
Digestate from a full scale biogas plant (65 kWe, 30 days Hydraulic Retention Time (HRT), mesophilic -39 °C, single stage), was chosen as inoculum, considering its relatively stable characteristics deriving from the use of a single feedstock (cow manure) as input (Table 1).

Lab Scale Anaerobic Digestion System and Experimental Setup
The lab scale AD system was composed of six digesters, each with a volume of 4 L and equipped with propeller-type mixers ( Figure 2). Temperature control was achieved by a thermostatic bath heated by an electric resistance controlled by a digital thermostat with a water recirculation pump to maintain homogeneous conditions in the tank.

Lab Scale Anaerobic Digestion System and Experimental Setup
The lab scale AD system was composed of six digesters, each with a volume of 4 L and equipped with propeller-type mixers ( Figure 2). Temperature control was achieved by a thermostatic bath heated by an electric resistance controlled by a digital thermostat with a water recirculation pump to maintain homogeneous conditions in the tank.
The biogas line was composed by condensation traps and biogas meters (MilliGascounter ® , Ritter, Bochum, Germany) that continuously registered the volume of produced biogas, normalizing it at the normal conditions of temperature and pressure (273.15 K; 101.33 kPa). Biogas quality (CH4, and CO2) was determined by Siemens Ultramat IR analyzer (Siemens Automation Group, Karlsruhe, Germany) and ProTec gas pump with Kitagawa (Kitagawa, Japan) detection tubes (H2S).
Digestate from a full scale biogas plant (65 kWe, 30 days Hydraulic Retention Time (HRT), mesophilic -39 °C, single stage), was chosen as inoculum, considering its relatively stable characteristics deriving from the use of a single feedstock (cow manure) as input (Table 1).   The biogas line was composed by condensation traps and biogas meters (MilliGascounter ® , Ritter, Bochum, Germany) that continuously registered the volume of produced biogas, normalizing it at the normal conditions of temperature and pressure (273.15 K; 101.33 kPa). Biogas quality (CH 4 , and CO 2 ) was determined by Siemens Ultramat IR analyzer (Siemens Automation Group, Karlsruhe, Germany) and ProTec gas pump with Kitagawa (Kitagawa, Japan) detection tubes (H 2 S).
Digestate from a full scale biogas plant (65 kWe, 30 days Hydraulic Retention Time (HRT), mesophilic −39 • C, single stage), was chosen as inoculum, considering its relatively stable characteristics deriving from the use of a single feedstock (cow manure) as input (Table 1). Digesters n. 1, 2, and 3 were filled only with the inoculum (2.5 kg) in order to determine the residual production from this substrate. Digesters n. 4, 5 and 6 were filled with a mixture of inoculum and fresh grass (2.5 kg of inoculum, 0.5 kg of water, 0.3 kg of grass) to attain an inoculum:substrate VS ratio equal to 2:1), as suggested by standard Biomethane Potential (BMP) tests [29].
Two cycles of tests were conducted in mesophilic conditions (38 • C) for 40 days, the first with spring grass and the second with summer grass. Net biogas production was determined as the difference between the production of digesters n. 4, 5 and 6 and the residual production measured in reactors n. 1, 2, and 3. BMP, referred to the mass unit of organic matter (VS) contained in input grass (L CH 4 ·kgVS −1 or mL CH 4 ·gVS −1 ), was calculated from the net production of biogas, considering CH 4 concentration. The results are reported as average of the three reactors treating spring grass (referred to as Spring) of the three treating summer grass (referred to as Summer).
The digestates from the reactors were subject to the determination of the emissions of NH 3 , CO 2 , CH 4 by means of Bruel and Kjiaer 1302 multi-gas analyzer with the static chamber method [30][31][32]. This method is based on the determination of the increase trend of the concentration of a gas in an enclosed chamber, as consequence of the emission from a sample, until the saturation is reached. A continuous measurement for approximately 20-30 min is performed, depending on the time in which the gas concentration increase linearly, before reaching saturation. Redox potential, pH, feedstock temperature, and ambient temperature were also measured at the beginning of each measurement run.
The specific flow of the monitored gas (F gas ), also defined as emissivity (mg·m −2 ·h −1 g −1 ), is obtained by the following equation: These analyses were performed on both the inoculum and on digestates from AD of fresh spring and summer grass. To perform these analyses, samples of the feedstocks (inoculum and digestates) were collected in aluminium containers (approximately 250-450 mg per container) immediately after the opening of the digesters. This decision was intended to operate on samples at temperatures close to the 38 • C of the AD process, minimizing the effect of ambient temperature (lower temperatures determine reduced emissions) and of time, as the emissivity diminishes with time for effect of loss of gaseous compounds. This would be a "worst case scenario" and emissions would be expected to be lower in the field.

Characteristics of the Feedstocks
Grass was composed mainly by Poaceae, with prevalence of Poa spp. and Festuca spp. in the spring and Poa spp., Festuca spp., Sorghum spp., and Phragmites spp. for the summer samples. Minor percentages of other species, such as, Asteraceae spp., Equisetaceae spp. and Polygonaceae spp. were also detected [33].
The characteristics of spring grass, summer grass, of the inoculum, and of the final digestate obtained at the end of each test are reported in Table 1. Spring grass samples and summer grass samples presented similar TS content, 32.43% and 33.30%, respectively; VS resulted 79.21%TS for the spring grass samples and manifested 90.21%TS for the summer grass samples.
Relevant differences were detected in terms of content of fibers: spring grass presented lignin, cellulose, and hemicellulose concentrations of 3.93%TS, 21.24%TS and 23.07%TS, respectively, whereas summer grass revealed concentrations of 8.65%TS, 34.72%TS and 28.12%TS. This difference is related to the natural growth of the plants, but has relevance in terms of degradability, taking into account that fibers, lignin especially, have limited degradability in the AD process and in acceptable period of time (HRT) [34].
The inoculum presented typical characteristics of a digestate from a mesophilic, complete mix digester operating wet fermentation. Considering that the biogas plant is fed with a single feedstock (manure from dairy cows) the characteristics of the samples of the spring and summer tests resulted analogous. The TS of the inoculum resulted 7.85% in the spring test and of 7.28% in the summer test, typical concentration of solids for a wet AD process [35,36]. Furthermore, the VS content did not vary significantly from 80.80%TS for the spring test and 78.45%TS for the summer.
The input mix for the spring test presented a TS concentration of 8.90% while the summer mix presented a TS concentration of 8.55%. In both cases, this is typical of a wet AD process and is adequate to achieve an efficient mixing/stirring of the substrates [36].
Digestates collected at the completion of the tests presented a significant reduction in terms of TS, resulting in 6.28% and 6.06% respectively for spring and summer, as effect of the degradation of the organic matter by AD populations.
VS reduction, in fact, resulted of 38.6% and of 35.0% for spring and summer, respectively. These findings demonstrate appropriate degradation of the organic matter but exhibit a slightly lower degradation for summer grass, presumably as effect of the higher concentration of fibers, lignin in particular.
Nitrogen concentration of the inoculum resulted quite constant and typical for a biogas plant fed with dairy manure, with TKN of 4. 19  These values indicate that the appropriate mixing of the inoculum with grasses did not determine a significant increase of Nitrogen concentration in the mixture subject to AD process, maintaining NH + 4 concentration below toxicity values [33]. Furthermore, TKN was not reduced by the AD process while NH + 4 was slightly increased as effect of the degradation of proteins.
Redox potential represents a valid indication of the reductive or oxidative conditions in a substrate. The inoculum presented a Redox potential of −400 mV and −395 mV, while digestates of −391 mV and −455 mV for spring and summer grass respectively, to indicate appropriate anaerobic conditions [33].
The acidity/alkalinity ratio of the inoculum resulted in 0.24 for the spring and 0.19 for the summer. These values are close to the lower end of the optimal range of 0.20-0.50 [37] indicating a favourable Energies 2018, 11, 1466 6 of 13 degradation of acids by the methanogenic bacteria, not only underlining a successful process, but also that the organic load was quite low. In this specific case the performance of the full scale digester was optimal and these value could be a reference of the standard operating conditions of this specific plant. The identical parameter applied to the analysis of digestate from the lab scale tests, conducted in discontinuous conditions, clearly indicates a satisfactory degradation of the feedstock with optimal conversion of acids into methane. In fact, digestate from the spring test, presented a ratio of 0.15, while digestate from the summer test a ratio of 0.19.

Biogas Quality
Biogas quality, in terms of CH 4 , CO 2 and H 2 S concentration, is reported in Figure 3. Each plot represents the average of the three digesters for each test cycle. The reactors treating spring grass produced a biogas with constant composition for the entire 40-days of the test. CH 4 content varied from a minimum of 46.0%, recorded at the beginning of the process (day n. 3), to a maximum of 55.2%, recorded three days after (day n. 6), and subsequently remained quite constant and close to the average value of 51.5 ± 1.4%, with maximum of 55.2% and minimum of and CO 2 concentration of 46.3 ± 1.4%. Average CO 2 concentration resulted in 46.3 ± 1.4%, with maximum of 53.0% recorded at day n. 3 when the methanogenic production was not fully developed, and with a minimum of 43.6%.
Energies 2018, 11, x FOR PEER REVIEW 6 of 13 feedstock with optimal conversion of acids into methane. In fact, digestate from the spring test, presented a ratio of 0.15, while digestate from the summer test a ratio of 0.19.

Biogas Quality
Biogas quality, in terms of CH4, CO2 and H2S concentration, is reported in Figure 3. Each plot represents the average of the three digesters for each test cycle. The reactors treating spring grass produced a biogas with constant composition for the entire 40-days of the test. CH4 content varied from a minimum of 46.0%, recorded at the beginning of the process (day n. 3), to a maximum of 55.2%, recorded three days after (day n. 6), and subsequently remained quite constant and close to the average value of 51.5 ± 1.4%, with maximum of 55.2% and minimum of and CO2 concentration of 46.3 ± 1.4%. Average CO2 concentration resulted in 46.3 ± 1.4%, with maximum of 53.0% recorded at day n. 3 when the methanogenic production was not fully developed, and with a minimum of 43.6%. Spring grass revealed a methane concentration quite typical for co-digestion [1,16,34,36]. The concentration of H2S resulted significantly variable with time, ranging from a minimum below the detecting threshold (25 ppm) to a peak of 295 ppm corresponding to day n. 13, after the production peak of day n. 6. The evolution of this parameter is not very clear, but the average concentration is well below the limit of 300 ppm referred as safe threshold for Combined Heat and Power (CHP) units, and also the peak is acceptable without the potential need of H2S removal systems.
Summer grass also showed a relatively constant biogas quality, in reference to CH4 and CO2, with variations from a minimum of 45.6% (day n. 6) to a maximum of 55.2% (day n. 36). CH4 concentration exceeded 50% only after day n. 21, exhibiting a slower methanogenic activity compared to spring grass, that reached the maximum around day n. 6. Average CH4 concentration resulted 49.7 ± 1.2%, lower than spring grass and of typical co-digestion values.
H2S concentration in biogas from summer grass demonstrated higher variations compared to spring grass, with the highest registered value of 1016.7 ppm at day n. 3, followed by a lower peak of 556.7 ppm at day n. 13. These peaks exceed the threshold of 300 ppm considered as admissible for CHP use of biogas and suggest treatments for H2S removal. The average value resulted acceptable, 295.1 ppm. Nevertheless, H2S concentration recorded in both tests could be effectively reduced even with biological removal by Thiobacillus under micro-aeration conditions: Mulbry et al. [38] determined desulfurization efficiencies ranging from 74 to >99% treating biogases that, when untreated, presented H2S concentration from 800 to 7500 mg·m −3 .
Overall, biogas quality was acceptable in both tests, but resulted higher for spring grass. This could be related to the different composition of the two grasses, mainly in terms of fiber content, which was higher for summer grass. Spring grass revealed a methane concentration quite typical for co-digestion [1,16,34,36]. The concentration of H 2 S resulted significantly variable with time, ranging from a minimum below the detecting threshold (25 ppm) to a peak of 295 ppm corresponding to day n. 13, after the production peak of day n. 6. The evolution of this parameter is not very clear, but the average concentration is well below the limit of 300 ppm referred as safe threshold for Combined Heat and Power (CHP) units, and also the peak is acceptable without the potential need of H 2 S removal systems.
Summer grass also showed a relatively constant biogas quality, in reference to CH 4 and CO 2 , with variations from a minimum of 45.6% (day n. 6) to a maximum of 55.2% (day n. 36). CH 4 concentration exceeded 50% only after day n. 21, exhibiting a slower methanogenic activity compared to spring grass, that reached the maximum around day n. 6. Average CH 4 concentration resulted 49.7 ± 1.2%, lower than spring grass and of typical co-digestion values. CO 2 concentration ranged from a minimum of 45.4% (day n. 36 corresponding to maximum CH 4 concentration) to a maximum of 47.5% (day n. 21, minimum CH 4 ), with average 46.4 ± 1.4%.
H 2 S concentration in biogas from summer grass demonstrated higher variations compared to spring grass, with the highest registered value of 1016.7 ppm at day n. 3, followed by a lower peak of 556.7 ppm at day n. 13. These peaks exceed the threshold of 300 ppm considered as admissible for CHP use of biogas and suggest treatments for H 2 S removal. The average value resulted acceptable, 295.1 ppm. Nevertheless, H 2 S concentration recorded in both tests could be effectively reduced even with biological removal by Thiobacillus under micro-aeration conditions: Mulbry et al. [38] Energies 2018, 11, 1466 7 of 13 determined desulfurization efficiencies ranging from 74 to >99% treating biogases that, when untreated, presented H 2 S concentration from 800 to 7500 mg·m −3 .
Overall, biogas quality was acceptable in both tests, but resulted higher for spring grass. This could be related to the different composition of the two grasses, mainly in terms of fiber content, which was higher for summer grass.
Yu and colleagues [39] report higher concentration of methane (average 71%) in biogas from grass, in a two-phase AD process, more similar to a dry AD process with recirculation of leachate [36] but simulating degradation times of a landfill. This result could be explained by the different chemical composition of grass and different AD processes. Other studies determined that the concentration of CH 4 from AD of grass can vary depending on the AD system, showing values of 71% for Upflow Anaerobic Sludge Blanket digesters, 51-54% for smaller scale BMP units, and 52% for continuous-flow stirred tank reactor (CSTR) digesters [40].

Biogas Production and Yield
Biogas production of the single reactors, for both spring and summer tests, is reported in Figure 4. Yu and colleagues [39] report higher concentration of methane (average 71%) in biogas from grass, in a two-phase AD process, more similar to a dry AD process with recirculation of leachate [36] but simulating degradation times of a landfill. This result could be explained by the different chemical composition of grass and different AD processes. Other studies determined that the concentration of CH4 from AD of grass can vary depending on the AD system, showing values of 71% for Upflow Anaerobic Sludge Blanket digesters, 51-54% for smaller scale BMP units, and 52% for continuousflow stirred tank reactor (CSTR) digesters [40].

Biogas Production and Yield
Biogas production of the single reactors, for both spring and summer tests, is reported in Figure 4. Total cumulated biogas production of the two inoculums resulted similar in the two test runs. Residual production of biogas resulted 30.6 NL for the spring test and of 27.0 NL for the summer test. It is notable how digestate from the cow farm (HRT 30 days) still presented some biogas potential after 40 days of anaerobic digestion. A reactor with increased volume (higher HRT) would probably allow the farm to obtain more energy from manure. This also can indicate that the standard BMP procedure of allowing the inoculum degasify for a week may be not sufficient to completely eliminate the residual production.
The curves of biogas production depict the evolution of the comprehensive production, inclusive also of the residual production of the inoculum. For spring grass, in particular, the three reactors presented a continuous production that showed higher intensity in the first days, as highlighted by a peak in the daily production corresponding to day n. 3. The average peak resulted in 9.3 NL·day −1 ± 0.9. The average final biogas production resulted in with average production of 79.9 NL ± 7.5.
The biogas production curves for summer grass demonstrated a similar but less steep trend. Also in this case a peak in the daily production was detected for all the digesters corresponding to day n. 3, with average peak of 5.2 NL·day −1 ± 0.9. Obtained values resulted lower than the peaks recorded for spring grass (44% lower with respect to the average values). This result did not affect the final production, which resulted notably analogous to that from spring grass, with average of 80.1 NL ± 9.8. Figure 4 shows that the daily production curves of summer grass, despite a lower peak compared to spring grass, presented a lower decrease after the maximum resulting in a more constant production, reaching a similar final volume of biogas. Total cumulated biogas production of the two inoculums resulted similar in the two test runs. Residual production of biogas resulted 30.6 NL for the spring test and of 27.0 NL for the summer test. It is notable how digestate from the cow farm (HRT 30 days) still presented some biogas potential after 40 days of anaerobic digestion. A reactor with increased volume (higher HRT) would probably allow the farm to obtain more energy from manure. This also can indicate that the standard BMP procedure of allowing the inoculum degasify for a week may be not sufficient to completely eliminate the residual production.
The curves of biogas production depict the evolution of the comprehensive production, inclusive also of the residual production of the inoculum.
For spring grass, in particular, the three reactors presented a continuous production that showed higher intensity in the first days, as highlighted by a peak in the daily production corresponding to day n. 3. The average peak resulted in 9.3 NL·day −1 ± 0.9. The average final biogas production resulted in with average production of 79.9 NL ± 7.5.
The biogas production curves for summer grass demonstrated a similar but less steep trend. Also in this case a peak in the daily production was detected for all the digesters corresponding to day n. 3, with average peak of 5.2 NL·day −1 ± 0.9. Obtained values resulted lower than the peaks recorded for spring grass (44% lower with respect to the average values). This result did not affect the final production, which resulted notably analogous to that from spring grass, with average of 80.1 NL ± 9.8. Figure 4 shows that the daily production curves of summer grass, despite a lower peak compared to spring grass, presented a lower decrease after the maximum resulting in a more constant production, reaching a similar final volume of biogas.
Biogas production from spring grass reached 50% in correspondence to day n. 5 and 80% at day n. 15, while biogas production from summer grass reached 50% in correspondence to day n. 9 and 80% at day n. 21.
This difference could be explained by taking into account the higher content of fibers of summer grass, with consequent slower production rate. On the other hand, spring grass presented an initial higher production rate that diminished when the easily degradable organic matter was transformed into biogas. Other studies performed in a leach bed upflow anaerobic sludge blanket (UASB) process, with higher HRT, showed that 58% of the acid soluble lignin was solubilized within the 49 days of process, whereas lignin was most recalcitrant [27].
The net production of biogas from both spring and summer grass, was calculated by subtracting the recorded values of the residual production of the inoculums and was referred to mass units of raw product, of TS and of VS ( Figure 5). The trends remained quite comparable to those depicted in Figure 4, with highest peaks reported for spring grass but steadier production for summer grass. After day n.8, the daily production of summer grass surpassed the daily production of spring grass. Biogas production from spring grass reached 50% in correspondence to day n. 5 and 80% at day n. 15, while biogas production from summer grass reached 50% in correspondence to day n. 9 and 80% at day n. 21.
This difference could be explained by taking into account the higher content of fibers of summer grass, with consequent slower production rate. On the other hand, spring grass presented an initial higher production rate that diminished when the easily degradable organic matter was transformed into biogas. Other studies performed in a leach bed upflow anaerobic sludge blanket (UASB) process, with higher HRT, showed that 58% of the acid soluble lignin was solubilized within the 49 days of process, whereas lignin was most recalcitrant [27].
The net production of biogas from both spring and summer grass, was calculated by subtracting the recorded values of the residual production of the inoculums and was referred to mass units of raw product, of TS and of VS ( Figure 5). The trends remained quite comparable to those depicted in Figure 4, with highest peaks reported for spring grass but steadier production for summer grass. After day n.8, the daily production of summer grass surpassed the daily production of spring grass.  (Figure 6). Results can be considered comparable, despite the higher production that was achieved, in reference to wet mass and TS, for summer grass and for spring grass in reference to VS.
Methane yield resulted in 87.4 NL·kg −1 (wet mass), 269.5 NL·kg −1 TS with BMP of 340.2 NL·kg −1 VS for spring grass and 92.5 NL·kg −1 (wet mass), 277.7 NL·kg −1 TS with BMP of 307.7 NL·kg 1 VS for summer grass.  (Figure 6). Results can be considered comparable, despite the higher production that was achieved, in reference to wet mass and TS, for summer grass and for spring grass in reference to VS.
Methane yield resulted in 87.4 NL·kg −1 (wet mass), 269.5 NL·kg −1 TS with BMP of 340.2 NL·kg −1 VS for spring grass and 92.5 NL·kg −1 (wet mass), 277.7 NL·kg −1 TS with BMP of 307.7 NL·kg 1 VS for summer grass.  The biogas yield of grass was higher than expected, similar to the yield from energy crops, especially in terms of NL·kg −1 VS [33,34]. Other authors reported BMP values of 337 NL·kg −1 VS for a mixture of switchgrass and swine manure without assessing the specific yield from grass [26].
Methane potential varying from 0.141 to 0.204 m 3 ·CH4 kg −1 added volatile solids was obtained by Lehtomäki and collegues [27], while other authors report a BMP from ensiled grass varying from 350 to 493 L CH4·kg −1 VS for three different AD processes, with 451 L CH4·kg −1 VS for a CSTR system over a 50-day retention period [40]. Grass characteristics, eventual conservation, and AD process are essential parameters influencing BMP of grass.
Grass production per surface of land resulted in 3.01 t·ha −1 , wet mass, and 0.98 t·ha −1 , dry mass, in the spring 12.78 t·ha −1 wet mass, and 3.89 t·ha −1 , dry mass, in the summer. These results are related to the fact that the harvest was performed in two different areas rather than in two subsequent cuts in the same area.
The production of biogas and methane per surface of grassland was calculated intersecting biogas and methane yields and grass production (wet mass). The results revealed a biogas yield of 496 Nm 3 ·ha −1 for spring grass and 2377 Nm 3 ·ha −1 for summer grass, while methane yield resulted in 263 Nm 3 ·ha −1 and 1181 Nm 3 ·ha −1 , respectively (Figure 7). Other authors report a methane yield of 2700 and 3500 Nm 3 ·ha −1 for Alpine grassland [2]. Clearly this result is strictly related to the investigated areas and influenced by several factors including grass The biogas yield of grass was higher than expected, similar to the yield from energy crops, especially in terms of NL·kg −1 VS [33,34]. Other authors reported BMP values of 337 NL·kg −1 VS for a mixture of switchgrass and swine manure without assessing the specific yield from grass [26].
Methane potential varying from 0.141 to 0.204 m 3 ·CH 4 kg −1 added volatile solids was obtained by Lehtomäki and collegues [27], while other authors report a BMP from ensiled grass varying from 350 to 493 L CH 4 ·kg −1 VS for three different AD processes, with 451 L CH 4 ·kg −1 VS for a CSTR system over a 50-day retention period [40]. Grass characteristics, eventual conservation, and AD process are essential parameters influencing BMP of grass.
Grass production per surface of land resulted in 3.01 t·ha −1 , wet mass, and 0.98 t·ha −1 , dry mass, in the spring 12.78 t·ha −1 wet mass, and 3.89 t·ha −1 , dry mass, in the summer. These results are related to the fact that the harvest was performed in two different areas rather than in two subsequent cuts in the same area.
The production of biogas and methane per surface of grassland was calculated intersecting biogas and methane yields and grass production (wet mass). The results revealed a biogas yield of 496 Nm 3 ·ha −1 for spring grass and 2377 Nm 3 ·ha −1 for summer grass, while methane yield resulted in 263 Nm 3 ·ha −1 and 1181 Nm 3 ·ha −1 , respectively ( Figure 7).  The biogas yield of grass was higher than expected, similar to the yield from energy crops, especially in terms of NL·kg −1 VS [33,34]. Other authors reported BMP values of 337 NL·kg −1 VS for a mixture of switchgrass and swine manure without assessing the specific yield from grass [26].
Methane potential varying from 0.141 to 0.204 m 3 ·CH4 kg −1 added volatile solids was obtained by Lehtomäki and collegues [27], while other authors report a BMP from ensiled grass varying from 350 to 493 L CH4·kg −1 VS for three different AD processes, with 451 L CH4·kg −1 VS for a CSTR system over a 50-day retention period [40]. Grass characteristics, eventual conservation, and AD process are essential parameters influencing BMP of grass.
Grass production per surface of land resulted in 3.01 t·ha −1 , wet mass, and 0.98 t·ha −1 , dry mass, in the spring 12.78 t·ha −1 wet mass, and 3.89 t·ha −1 , dry mass, in the summer. These results are related to the fact that the harvest was performed in two different areas rather than in two subsequent cuts in the same area.
The production of biogas and methane per surface of grassland was calculated intersecting biogas and methane yields and grass production (wet mass). The results revealed a biogas yield of 496 Nm 3 ·ha −1 for spring grass and 2377 Nm 3 ·ha −1 for summer grass, while methane yield resulted in 263 Nm 3 ·ha −1 and 1181 Nm 3 ·ha −1 , respectively ( Figure 7). Other authors report a methane yield of 2700 and 3500 Nm 3 ·ha −1 for Alpine grassland [2]. Clearly this result is strictly related to the investigated areas and influenced by several factors including grass Other authors report a methane yield of 2700 and 3500 Nm 3 ·ha −1 for Alpine grassland [2]. Clearly this result is strictly related to the investigated areas and influenced by several factors including grass species, biomass production and harvesting period as result of different climate conditions and soil characteristics. In terms of energy, considering a specific heating value of 37.7 MJ·m −3 , the results of the present study demonstrate a potential energy yield of 9915 MJ·ha −1 for spring grass and 44,523 MJ·ha −1 for summer grass.

Gaseous Emissions
Gaseous emissions from the inoculum and from digestates from anaerobic digestion of fresh spring and summer grass, reported as specific emissivity per unit of emitting surface and mass unit of sample (mg·h −1 ·m −2 g −1 for NH 3 and CH 4 and g·h −1 ·m −2 g −1 for CO 2 ), are reported in Figure 8.

Gaseous Emissions
Gaseous emissions from the inoculum and from digestates from anaerobic digestion of fresh spring and summer grass, reported as specific emissivity per unit of emitting surface and mass unit of sample (mg·h −1 ·m −2 g −1 for NH3 and CH4 and g·h −1 ·m −2 g −1 for CO2), are reported in Figure 8. Ammonia emissions revealed 0.310 mg·h −1 ·m −2 g −1 for the inoculum, 0.293 and 0.170 mg·h −1 ·m −2 g −1 for anaerobically digested fresh spring and summer grass, respectively. Whereas spring grass seems to release a similar quantity of ammonia compared to digestate, summer grass appears to emit less. This result is inconsistent with the concentration of N of grass, higher for summer grass.
Methane emissions resulted in 0.049 mg·h −1 ·m −2 g −1 for the inoculum, of 0.129 and 0.043 mg·h −1 ·m −2 g −1 for fresh spring and summer grass, respectively. Furthermore, carbon dioxide, revealed specific emission of 5.124 g·h −1 ·m −2 g −1 for the inoculum, 6.858 and 3.696 g·h −1 ·m −2 g −1 for anaerobically digested fresh spring and summer grass, respectively. These results can be considered as a reference more than providing indication of different emission rates between the feedstocks. In fact, a certain variability is related to the variations in the atmospheric conditions, temperature of air, in particular, between the different measurement campaigns. Alternatively, it would be meaningful to evaluate the emissions of greenhouse and acidifying gases from anaerobic digestion in comparison with traditional management practices, like composting of grass.
Nevertheless, the results are not intended as simulation of real conditions in which digestate is conveyed to a storage tank, in many cases equipped with airtight cover, where is stored for months before being applied on cropland.

Conclusions
The biogas yield obtained from grass resulted higher than expected, quite similar to the yield obtained from energy crops, especially in terms of volume per mass unit of volatile solids (NL·kg −1 VS).
This promising result can be related to the fact that grass was loaded in the digesters just few hours after being collected, with the maximum availability of organic matter. However, iunder actual operative conditions, storage of feedstocks is required, with possible loss of VS and consequently lower biogas potential. Ammonia emissions revealed 0.310 mg·h −1 ·m −2 g −1 for the inoculum, 0.293 and 0.170 mg·h −1 ·m −2 g −1 for anaerobically digested fresh spring and summer grass, respectively. Whereas spring grass seems to release a similar quantity of ammonia compared to digestate, summer grass appears to emit less. This result is inconsistent with the concentration of N of grass, higher for summer grass.
Methane emissions resulted in 0.049 mg·h −1 ·m −2 g −1 for the inoculum, of 0.129 and 0.043 mg·h −1 ·m −2 g −1 for fresh spring and summer grass, respectively. Furthermore, carbon dioxide, revealed specific emission of 5.124 g·h −1 ·m −2 g −1 for the inoculum, 6.858 and 3.696 g·h −1 ·m −2 g −1 for anaerobically digested fresh spring and summer grass, respectively. These results can be considered as a reference more than providing indication of different emission rates between the feedstocks. In fact, a certain variability is related to the variations in the atmospheric conditions, temperature of air, in particular, between the different measurement campaigns. Alternatively, it would be meaningful to evaluate the emissions of greenhouse and acidifying gases from anaerobic digestion in comparison with traditional management practices, like composting of grass.
Nevertheless, the results are not intended as simulation of real conditions in which digestate is conveyed to a storage tank, in many cases equipped with airtight cover, where is stored for months before being applied on cropland.

Conclusions
The biogas yield obtained from grass resulted higher than expected, quite similar to the yield obtained from energy crops, especially in terms of volume per mass unit of volatile solids (NL·kg −1 VS).
This promising result can be related to the fact that grass was loaded in the digesters just few hours after being collected, with the maximum availability of organic matter. However, iunder actual operative conditions, storage of feedstocks is required, with possible loss of VS and consequently lower biogas potential.
The specific yield of spring and summer grass, both in terms of biogas and methane, revealed similarities between the two types of feedstocks. Biogas quality resulted slightly lower for summer grass, perhaps in relation to the higher content of fibers (lignin in particular). Alternatively, the yield of grass per surface was significantly different between spring and summer, with the highest production in the summer. This difference also affects the biogas and methane yield per hectare resulting more than 4 times higher for the summer grass. It is clear that the results in terms of grass yield, characteristics, and BMP are influenced by the specific conditions (e.g., climate, soil properties) of the investigated area and may result different in other geographic areas.
The present research indicates that grass can be successfully utilized as feedstock in the anaerobic digestion process, resulting in the production of biogas or biomethane, with acceptable HRT. This result opens the prospective of recovering this feedstock from various areas, including parks, riverbanks, or public areas etc. In theory, lower HRT could be taken in consideration for spring grass, but in practice this would be impractical and could result in an incorrect sizing of the reactors in case of employing of other feedstocks, such as summer grass itself.
For a practical exploitation of this resource, appropriate conservation of the feedstock is necessary, by storing grass in order to enable the daily load of digesters while reducing the loss of organic matter (e.g., biogas potential). To simulate operative conditions in relation to the most common conservation practices, spring and summer grass were air dried and ensiled, and BMP was assessed. The results of the last anaerobic digestion tests will be object of a further paper.