Comparison of Different Mechanical Pretreatment Methods for the Anaerobic Digestion of Landscape Management Grass

Abstract

The aim of this study was to use landscape grass from species-rich orchards for biogas production, thus preserving these very valuable areas for future generations. Since these grass clippings have high lignocellulose content, the substrate has to be pretreated before being fed into the biogas digester. In this study, three different mechanical treatment processes (cross-flow grinder, ball mill and a mounted mower) were investigated and compared with untreated grass clippings. Chemical composition, specific methane yield, degradation kinetics and microscopic images were analyzed. In order to derive recommendations, the harvesting and pretreatment processes were examined in terms of energy demand, additional methane yield, and suitability of the substrate for use in biogas plants, taking into account conservation aspects. Within the pretreatment process, ball milling leads to the highest significant increase in specific methane yield of up to 5.8% and the fastest gas formation kinetics (lag time λ_BM: 0.01 ± 0.0 d; duration to reach half of total gas production ½M(x)_BM: 5.4 ± 0.2 d) compared to the untreated variant (λ_UT: 1.02 ± 0.2 d; ½M(x)_UT: 6.5 ± 0.2 d). A comparison of the energy required for the mechanical disintegration of the substrates with the increased yield of methane during the digestion process shows that the mechanical processing of these substrates appears to be useful. A positive energy balance was achieved for the cross-flow grinder (12.3 kWh t_VS⁻¹) and the ball mill (21.4 kWh t_VS⁻¹), while the Amazone Grasshopper left a negative balance (−18.3 kWh t_VS⁻¹), requiring more energy for substrate pretreatment than was generated as methane surplus. In summary, the pretreatment of landscape management grass is a suitable approach for utilizing agricultural residues efficiently in a biogas plant and thus contributing to sustainable energy production.

1. Introduction

The increasing demand for renewable energy sources and the urgent need to reduce greenhouse gas emissions have led to significant research efforts with regard to sustainable solutions for energy generation from waste. In this context, anaerobic digestion (AD) has emerged as a promising technology for converting organic wastes into valuable biogas, which can be converted into both electricity and heat in a combined heat and power plant or can be purified and fed into natural gas grids [1]. Biogas is a renewable energy that is well developed in Germany, with more than 9000 biogas plants (BGPs) [2]. However, a large amount of organic waste and organic residues, such as landscape management grass, is not yet utilized because they often have high contents of lignocellulose and lignin, some of which are not or are only slowly anaerobically degradable [3,4,5,6]. Grass biomass represents a significant organic feedstock potential for AD due to its widespread availability and seasonal abundance [7,8].

Landscape management grass is defined as grass biomass produced by activities that primarily and predominantly pursue the goals of nature conservation and landscape management, within the scope of the German Federal Nature Conservation Act [9]; are not selectively cultivated; and have a maximum of two cuts per year [5,9,10]. According to an estimate by Meyer et al. 2018 [11], between 20 and 110 million tons of excess grass could be available in the European Union by 2030, with meadows and permanent grassland accounting for the largest share.

The use of landscape management grass offers many opportunities, such as the valorization of agricultural residues [12] for biogas production from biomass, without land competition with food production [5,6,13,14,15]. Additionally, the maintenance and conservation of these areas increases the biodiversity of wildlife and plants by preserving their habitats [9,10,16,17,18]. The German Federal Nature Conservation Act goes even further in this regard, addressing the importance of the long-term conservation of diversity, distinctiveness, and beauty. In addition, the act also takes into account the recreational value of nature and landscape, the protection of which includes the maintenance, development, and, if necessary, restoration of nature and landscape as a general principle [9,10]. In this context, farmers act as both landscape managers and energy producers in order to maintain the viability of these areas and simultaneously use the harvested biomass for sustainable energy production from agricultural residues.

Open grassland biotopes offer these unique habitats due to a high number of fruit trees, a high groundwater level, or very shallow soils, in combination with the long-term extensive cultivation of the grassland almost without the use of fertilizers. As a result, the harvesting and AD of landscape management grass encounters certain challenges. Compared to conventional energy crops, harvesting is more difficult and involves higher costs for the salvage and transport of cuttings, as the areas are mostly small-parcel biotopes that are often decentralized and difficult to access [5,10]. In addition, grass residues are often legally classified as biowaste, which further complicates their use in the biogas process [19]. The delayed mowing necessary to protect ground-nesting birds and late-flowering plants results in high lignocellulosic and lignin contents of the harvested biomass, which increases resistance to enzymatic degradation [20,21,22]. These characteristics limit the accessibility of microbial consortia to the degradable components of the substrate, resulting in slower degradation rates and lower biogas yields [6,10,23]. As many studies show that pretreatment is required to efficiently degrade such materials in a biogas plant [19,24,25,26]. The main pretreatment methods include a range of techniques, such as physical methods involving milling and grinding; physicochemical methods, including thermolysis, steam-based methods, and oxidation techniques; chemical methods involving the use of acids, alkalis, and oxidants; and biological methods involving enzymes or microorganisms [12,27,28,29,30,31].

In the area of full-scale biogas production, especially mechanical disintegration processes have received considerable attention for breaking down the lignocellulosic structure, reducing particle size, increasing particle surface area, and improving biodegradability, thus leading to increased biogas yields compared to untreated substrates [13,24,32]. It is particularly important to find a balance between optimal particle size distribution and process efficiency.

In the review study by Arce and Kratky (2022) [33], the milling of lignocellulose-rich substrates is discussed in detail [33]. However, there is no universal standard for optimal particle size distribution, as it depends on the particular substrate and the technology used for biogas production [34]. In the study by Tsapekos et al. (2018) [35], two different mechanical treatment methods were investigated in more detail. One pretreatment method consisted of a road-sweeping brush and a steel mesh conditioner. In the second experiment, the shredder was partially modified to increase the damage to the biomass surface. In this context, a maximum increase in specific methane yield from 297 to 376 mL g_VS⁻¹ was observed for meadow grass, although not all variants analyzed led to a statistically significant increase in methane yield [35].

Previous studies by Mönch-Tegeder et al. (2014) [36] using the cross-flow grinder have shown that different substrates, including those with very high lignocellulose content, like horse manure, require different specific energy demands at different treatment durations (15 and 30 s), but can also achieve positive energy balances due to the additional methane yield. A positive energy balance could only be achieved at a pretreatment time of 15 s for lignin-rich substrates. For silage based on energy crops, the mechanical pretreatment did not show any positive effects, probably due to the already high anaerobic degradability of the substrates. In addition, these silages contain high amounts of volatile organic acids, which may evaporate at the high temperatures generated during grinding [36]. In the study by Fernandez et al. (2022) [37], cattle manure, maize, and triticale silage as well as a mixture of cattle manure and silage were mechanically pretreated for anaerobic digestion, and it was shown that the specific methane yield was not increased by the mechanical treatment. However, effects on kinetics were demonstrated, leading to a faster methane production rate due to the release of soluble fractions.

The objective of the present study is to evaluate the impact of mechanical pretreatment processes on landscape management grass from orchard meadows in terms of degradation kinetics and specific methane yield during anaerobic digestion, energy requirements, and the resulting economics of the studied disintegration methods. In order to evaluate all pretreatment methods in terms of their energy requirements during harvesting and treatment, a working time study was conducted. The investigations regarding the valorization of these agricultural residues through mechanical treatment prior to the biogas process in conjunction with the working time study can be considered a novelty in this field of research. For this purpose, the cross-flow grinder (CFG), ball mill (BM), and Amazone Grasshopper mounted mower (AG) pretreatment techniques, which have already been tested on a practical scale, were used.

2. Materials and Methods

2.1. Description of the Harvest Area

A lowland hay meadow was selected for harvesting the landscape management grass, which is protected by the Natura 2000 network according to the Habitats Directive (Annex I; habitat type 6510) [38]. The meadow had a medium conservation status and is located at the foot of the Swabian Alb in Eningen unter Achalm, in south-western Germany, at 48°29′25.7″ N 9°16′49.2″ E, 540 m above sea level. The entire area is part of the UNESCO Swabian Alb Biosphere Reserve. The harvest of the landscape management grass took place in mid-June 2022 and was based on the optimal phenological cutting time for the conservation of biodiversity. It was a 6124 m² extensively used orchard meadow with an old apple tree population. The meadow itself is located on a west-facing slope near the valley floor, resulting in relatively nutrient-rich and moist conditions. The species composition is an Arrhenatheretum plant community with a high proportion of Geranium pratense, which is typical for such sites. At the same time, intensive grassland species such as Anthriscus sylvestris and Dactylis glomerata, as well as lean areas with, e.g., Bromus erectus, Primula veris and Luzula campestris, occur locally in this area [39].

2.2. Harvesting

As recommended by the local authorities, the meadow was mowed during the flowering period of the stand-forming grasses [40]. First, the tractor equipped with the Amazone Grasshopper mounted mower (Grasshopper GHS Drive 2100, Amazonen-Werke H. Dreyer SE & Co. KG, Hasbergen, Germany) drove off a part of the meadow, harvested it, pretreated the grass internally, and unloaded it at the biogas plant. For all other analyzed variants of mechanical disintegration, the grass was harvested with a tractor to which a front-mounted disc mower (Krone easy cut F320 CU, Maschinenfabrik Bernard KRONE GmbH & Co. KG, Spelle, Germany) with a working width of 3.2 m was attached. The cuttings were then swathed up and taken to the biogas plant by a loader wagon.

There, a sample was first taken from the untreated crop. In the same way, a sample was taken from the emptied container of the Amazone Grasshopper. All samples were immediately deep-frozen for further analysis.

The cutting height was set at 10 cm for all mowing operations. Increasing the cutting height to 10 cm, leaving at least 10% of the area untreated, and the use of species-specific mowing techniques are important measures to counteract species extinction [41,42]. At the time of harvest, the grass had a growth height of up to 1.3 m.

Usually, these extensive areas are mowed twice a year. However, only the first cut described above was included in the analyses.

2.3. Pretreatment Methods

2.3.1. Cross-Flow Grinder

The MEBA cross-flow grinder, further described as CFG (Bio-QZ, MEBA, Nördlingen, Germany) at the research biogas plant “Unterer Lindenhof” (Eningen unter Achalm, Germany) of the University of Hohenheim (Stuttgart, Germany), is a device for the mechanical pretreatment of landscape management grass. The device shown in Figure 1 is patented as a decomposing device for various materials and was designed specifically for the recycling industry [43]. The CFG mainly consists of a cylindrical working chamber with a fast-turning rotor equipped with two opposing steel chains at the bottom. The motor has an installed power of 55 kW. As soon as the rotor head starts turning, the lignocellulosic biomass is conveyed from the solid feeder via screw conveyors into the working chamber, where the biomass is comminuted. The rotating chains cause the biomass to flow both radially and vertically (cross-flow), resulting in particle collisions that cause the biomass to be shredded and defibrated [32].

Two different operating modes are possible: batch and continuous. The working chamber is automatically filled with a portion of material during batch operation, and the treatment runs for a configurable period of time before the material is released. For continuous operation, the substrate release is opened, and the material is continuously passed through. To meet large-scale conditions and to better compare the processes of the different mechanical pretreatment devices, continuous operation was selected for the CFG. This increases the output rate while decreasing the disintegration effect [36].

Before starting the pretreatment process in the CFG, 0.5 t of grass biomass was weighed and placed into the solid feeder.

2.3.2. Ball Mill

The ball mill used in this work, further described as BM (Biokraft Kugelmühle, Biokraft Energietechnik GmbH, Stuttgart, Germany), serves as a comminution device specifically for the mechanical pretreatment of fiber-rich biomass for biogas production. The device consists of a cylindrical rotating drum with a diameter of 2 m and a length of 3 m, which is mounted on a truck tandem axle. It is located at the research biogas plant “Unterer Lindenhof” of the University of Hohenheim and is integrated into the biogas process as a substrate pretreatment system. The BM is powered by a 45 kW electric motor. The drum is filled with grinding balls, each with an 80 mm diameter and made of forged steel. The collisions and interactions between the grinding balls and the substrate create frictional and impact forces that enable the grinding and comminution process, reducing the particle size and increasing the particle surface area.

Figure 2 shows the prototype ball mill, the solid feeder, and the screw conveyor from the rear view.

The substrate is treated in the drum in a continuous flow and then conveyed to the discharge disc, where it is pushed through. The pretreated substrate is then transported directly to the biogas plant via screw conveyors.

The comminution device is described in detail by Heller et al. (2023) [32].

2.3.3. Amazone Grasshopper

The Amazone Grasshopper mounted mower, further described as AG (Grasshopper GHS Drive 2100, Amazonen-Werke H. Dreyer SE & Co. KG, Hasbergen, Germany), has a working width of 2.1 m and a hopper volume of 3.5 m³. This tractor-mounted mower is attached via a drawbar and is driven by the tractor’s power take-off shaft.

It enables mowing and chopping with the help of a fine-flail mower, in which the wing blades are individually suspended from the rotor in an oscillating manner, and simultaneously collects the cuttings in one operation. With this approach, the use of the mounted mower represents harvesting and pretreatment processes at the same time (Figure 3).

2.4. Batch Digestion Test

2.4.1. Digester

Fifteen horizontal, continuous, stirred tank reactors with a total volume of 20 liters (L) each (working volume of 17 L; headspace volume of 3 L) were used to investigate biogas production and determine the specific methane yield (SMY) of landscape management grass (Figure 4b). The operation of the reactors was changed to batch mode. For this purpose, the outlet pipe was replaced by a plastic plate with a sealing ring. The inlet pipe was sealed with a gas-tight silicone lid. A water jacket connected to a hot-water bath, set to 42 ± 0.5 °C, was used for temperature regulation (Figure 4a). Gas was vented from the top of the reactors and then passed through a gas wash bottle (Figure 4c) before being collected in sealed gas bags (Figure 4d). The electrically operated agitators homogenized the digester contents twice per hour for five minutes each time. A detailed description of the system is given by Haag et al. (2015) [44].

2.4.2. Inoculum

A standardized inoculum for batch digestion tests at the State Institute of Agricultural Engineering and Bioenergy (University of Hohenheim, Stuttgart, Germany) was taken from a 400 L laboratory reactor. The inoculum reactor was operated at a mesophilic temperature of 42 °C and fed daily with substrate from several practice biogas plants and a balanced nutrient mixture of carbohydrate, fat, and protein at a low organic loading rate of 0.3 kg_VS m⁻³ d⁻¹, in order to obtain a wide spectrum of methanogenic microbes and ensure the low gas production of the inoculum for optimal experimental conditions [45,46]. A detailed description of the inoculum is given by Hülsemann et al. (2020) [46].

In the laboratory of the State Institute of Agricultural Engineering and Bioenergy, the samples of fresh inoculum were analyzed for pH and volatile fatty acids (VFA), namely acetic acid, propionic acid, and n-butyric acid, using gas chromatography (GC; 785 DMP Titrino, Metrohm, Herisau, Switzerland). In addition, samples were analyzed for pH value, total solids (TS), and volatile solids (VS), and the ratio of volatile fatty acids (VFA) to total alkalinity (TA) (also known as FOS/TAC) was determined.

2.4.3. Experimental Setup

The substrate-to-inoculum ratio was set to 0.33 based on volatile solids (VS) according to guideline VDI 4630 [47]. On the day of the start of the experiment, 15 kg (0.783 kg_VS) of inoculum and 1 kg (0.260 kg_VS) of landscape management grass, based on fresh mass, were added to the digesters through the inlet pipe. The inlet pipe was then sealed, and the agitators were started to homogenize the substrate and inoculum. The same procedure was performed for the untreated variant (UT). All variants were conducted in triplicate, providing a total of 15 reactors to detect outliers and ensure statistical analysis.

According to the guideline VDI 4630 for biogas tests, the test duration has to be set to a minimum of 25 days and has to be stopped as soon as the daily biogas production for three consecutive days is less than 0.5% of the total biogas formation up to that point. In addition, all parallel tests have to be stopped at the same time and the criterion is considered to be achieved when all digesters meet the criterion [47]. In the presented study, the experiment was terminated after 50 days because the daily biogas rate for three consecutive days was less than 0.5% of the total biogas volume produced up to that point. All parallel experiments were terminated at the same time.

In the same way as the tested variants, a zero variant was examined. The zero variant was performed with inoculum only, necessary to calculate the SMY of the landscape management grass. The gas production of the inoculum showed 16.8 ± 0.5 L_CH4 kg_VS⁻¹.

2.4.4. Gas Measurements

The gas measurements were started manually, depending on the amount of gas produced, and then performed by a fully automated gas measurement system at the State Institute of Agricultural Engineering and Bioenergy [44,48].

At the beginning of the measurement, the gas was sucked from the bags via a pump into two mass flow meters (EL-FLOW Select F-112AC, Bronkhorst Deutschland Nord GmbH, Kamen, Germany) connected in series, to measure the gas flow. The second mass flow meter is used to control the measurements of the first device. The gas was then fed into gas analysis sensors, connected in series. In these sensors, the contents of CH₄ (Gasmitter CH₄, Sensors Europe GmbH, Erkrath, Germany), CO₂ (Gasmitter CO₂, Sensors Europe GmbH, Erkrath, Germany), H₂ (AGM22, Sensors Europe GmbH, Erkrath, Germany) and H₂S (H2S/S-5000-S, Membrapor AG, Wallisellen, Switzerland) were measured in triplicate.

2.5. Microscopy

Mechanical pretreatments like grinding and milling reduce the particle size of the substrates, which also affects the appearance and the surface morphology of the material, thus significantly influencing anaerobic digestion.

The surface morphology of landscape management grass was studied using a Zoom-Microscope (Zeiss Axio Zoom. V16, Carl Zeiss Microscopy GmbH, Jena, Germany) to investigate the surface changes caused by different mechanical treatments compared to the untreated reference. Therefore, samples treated with CFG, BM, and AG and the untreated reference were dried at 60 °C and then examined under the microscope.

The microscope used a 12 megapixel microscope camera (Zeiss Axiocam 712 color, Carl Zeiss Microscopy GmbH, Jena, Germany) and was operated at different magnification levels. To improve the depth of field at higher magnification levels, Z-stacking was integrated into the ZEN core analysis software (ZEN core Version 3.3.92.00000, Carl Zeiss Microscopy GmbH, Jena, Germany).

2.6. Chemical Composition Analysis

2.6.1. Total Solids and Volatile Solid

Prior to batch digestion, total solids (TS) and volatile solids (VS) content was measured for each sample. The TS content was determined by drying the samples at 105 °C using a drying oven (UF450, Memmert GmbH, Schwabach, Germany) until a constant weight was reached. The determination was carried out according to the guideline DIN EN 12880:2000 [49]. After the determination of the TS content, the samples were incinerated in a muffle furnace at 550 °C for 6 h to determine the VS content, according to the guideline DIN EN 12879:2000 standard [50].

2.6.2. Ingredient Analysis

Samples of all variants were dried in a drying chamber at 60 °C for 48 h to preserve the volatile components. After drying, the substrates were ground to a particle size of 1 mm for chemical composition analyses using a cutting mill (Pulverisette 19, Fritsch, Idar Oberstein, Germany) to obtain homogenous samples. The samples were then analyzed by ISF Schaumann Forschung (ISF GmbH, Wahlstedt, Germany). The feed analysis was performed according to the guidelines of the Federation of German Agriculture Investigation and Research Institutes (VDLUFA) [51]. Acid detergent lignin (ADL) content was analyzed by wet chemistry according to VDLUFA method 6.5.3 due to low residual moisture.

The other ingredients, such as crude protein (XP), crude fat (XL), crude fiber (XF), neutral detergent fiber (aNDF), and acid detergent fiber (ADF), were analyzed by near-infrared spectroscopy according to VDLUFA method 31.2.

For energy content and potential utilization pathways, nitrogen-free extracts (NfE), gross energy (GE), metabolizable energy (ME), and net energy lactation (NEL) were determined. The analyses were also performed by ISF Schaumann Forschung.

In the laboratory of the State Institute of Agricultural Engineering and Bioenergy, the defrosted samples underwent analysis for volatile fatty acids, including acetic acid, propionic acid, and n-butyric acid, using gas chromatography (GC; 785 DMP Titrino, Metrohm, Herisau, Switzerland).

2.7. Calculations and Statistical Analysis

2.7.1. Modified Gompertz Model

The modified Gompertz kinetic model has been widely used to mathematically describe biogas production [52,53,54,55,56,57,58], assuming that the biogas production rate under batch conditions is a function of the bacterial growth rate of the predominant bacteria in the biogas plant:

$$\mathrm{M}=\mathrm{P} \times \exp \left\{-\exp \left[\frac{{\mathrm{R}}_{\mathrm{m}} \times \exp \left(1\right)}{\mathrm{P}}\left(\mathsf{\lambda }-\mathrm{t}\right)+1\right]\right\}$$

(1)

$${\mathrm{R}}_{\mathrm{m}}\left(\mathrm{x}\right)={\mathrm{R}}_{\mathrm{m}} \times \mathrm{e} \times \exp \left\{-\exp \left[\left(\frac{{\mathrm{R}}_{\mathrm{m} }\times \exp \left(\mathsf{\lambda } -\mathrm{x}\right)}{\mathrm{P}}\right)+1\right]+\left[\left(\frac{{\mathrm{R}}_{\mathrm{m} }\times \exp \left(\mathsf{\lambda } -\mathrm{x}\right)}{\mathrm{P}}\right)+1\right]\right\}$$

(2)

where M is the cumulative methane production (L_CH4 kg_VS⁻¹) after the digestion time t (d), P is the methane production potential (L_CH4 kg_VS⁻¹), R_m is the maximum daily methane production rate (L_CH4 kg_VS⁻¹ d⁻¹), λ is the lag phase duration (d), and e is exp(1), or 2.7182818.

To calculate the day of the maximum methane production rate, the second derivative (2) was used, which was created with the program Maxima (Maxima, Version 5.47.0, 2023) using the graphical user interface wxmaxima, an open-source computer algebra system [59]. In addition, the duration (days) in which half of the total gas ½M(x) was produced was determined.

The kinetic constants were determined using non-linear regression analysis (Microsoft Excel Solver Add-In, Microsoft Corporation, Redmond, WA, USA). A modified Gompertz fit was performed for each sample in the batch digestion experiment, and then the mean and standard deviation were calculated.

2.7.2. Specific Energy Demand for Harvesting and Pretreatment

For the calculation of energy demand at harvesting, a working time measurement was performed at the first cut on 72 plots. The average plot size was 1085 m². Half of the plots were harvested with the help of the AG, and the other plots were harvested conventionally, as described in Section 2.3. A Fendt 210 Vario tractor (AGCO GmbH, Marktoberdorf, Germany) with a rated power of 73 kW was used, to which the various attachments such as a disc mower, swather, or Amazone Grasshopper mounted mower were attached.

After harvesting with the AG, the cuttings were unloaded into a provided container. The container was transported to the biogas plant by a John Deere 6150R tractor (Deere & Company, Moline, IL, USA) with a rated power of 110 kW. This tractor also pulled the loader wagon for conventional harvesting followed by the transport of the cuttings to the BGP. However, the route to the BGP was not included in the energy demand because it is different for each study due to location.

The energy demand of the two different tractors was recorded by measuring the hours of operation and diesel consumption per hour. The measured diesel consumption was 8.02 L h⁻¹ for the Fendt 210 Vario tractor and 8.35 L h⁻¹ for the John Deere 6150 R. To calculate the specific energy demand in the field, the diesel consumption was multiplied by the calorific value of diesel, which is 9.96 kWh L⁻¹ [60]. Based on the yield per hectare, measured in t ha⁻¹, the energy consumption per ton was calculated in kWh t⁻¹. The treatment duration, fresh mass used and energy requirement of the CFG and the BM were logged in a database during pretreatment. The specific energy demand was then related to kilowatt hours per ton of volatile solids (kWh t_VS⁻¹).

2.7.3. Energy Balance

Determining the energy demand during mechanical pretreatment and the specific methane yield surplus resulting from the treatment are crucial parameters for determining the overall energy balance. Key parameters were recorded and analyzed to evaluate the overall energy balance. Specific energy demand (SED) per ton of fresh mass was calculated in kilowatt-hours per ton of volatile solids (kWh t_VS⁻¹). Additional specific methane yield (SMY_Add) was determined by quantifying the difference in methane yield between the pretreated samples and the untreated sample (L_CH4 t_VS⁻¹). The additional energy surplus (ES) resulting from the increased methane yield due to the pretreatment process was calculated using the equation:

$$\mathrm{ES}={\mathrm{SMY}}_{\mathrm{Add}}\times \mathrm{LHV}\times {\mathsf{\eta }}_{\mathrm{C}\mathrm{P}}$$

(3)

Here, LHV represents the lower heating value of methane, which is 35.894 MJ m⁻³ or 9.971 kWh m⁻³ [61], and η_CP denotes the assumed electrical energy conversion efficiency (38%) of the cogeneration plant for converting methane to electrical energy [36].

The energy balance (EB) was established by the following equation:

$$\mathrm{EB}=\mathrm{ES}-\mathrm{SED}$$

(4)

This equation includes the additional energy yield obtained by subtracting the specific energy demand required for the pretreatment process (SED) from the additional energy surplus (ES) obtained by pretreating landscape management grass.

Methane energy (ME) was calculated by multiplying SMY by the LHV of methane:

$$\mathrm{ME}=\mathrm{SMY}\times \mathrm{LHV}$$

(5)

The energy recovery rate (ER) was calculated as the ratio of methane energy (ME) recovered from digestion to the gross energy (GE) content of the substrate:

$$\mathrm{ER}=\frac{{\mathrm{ME}}_{\mathrm{digestion}}}{{\mathrm{GE}}_{\mathrm{substrate}}}$$

(6)

The gross energy content (MJ kg_TS⁻¹) of the substrates can be estimated using a predictive equation based on the results of the Weender feed analysis [62]:

$$\mathrm{GE}=0.0239 \mathrm{MJ} {\mathrm{g}}^{-1}\times \mathrm{XP}+0.0398 \mathrm{MJ} {\mathrm{g}}^{-1}\times \mathrm{XL}+0.0201 \mathrm{MJ} {\mathrm{g}}^{-1}\times \mathrm{XF}+0.0175 \mathrm{MJ} {\mathrm{g}}^{-1}\times \mathrm{NfE}$$

(7)

Nitrogen-free extractives (NfE) are calculated by subtracting crude ash (XA; not shown in

Table 1. Chemical composition of untreated and pretreated landscape management grass.

Determination Method	Unit	Treatment Variant
		UT	CFG	BM	AG
Total solids (TS)	% FM	26.9	25.9	25.8	25.2
Volatile solids (VS)	% TS	89.3	84.3	85.3	83.8
Acetic acid (AA)	g kgFM−1	0.09	0.14	0.35	0.10
Sugar (S)	g kgVS−1	75.5	42.5	27.0	30.5
Crude protein (XP)	g kgVS−1	51.9	55.9	55.8	58.9
Crude fat (XL)	g kgVS−1	21.2	22.3	20.2	22.4
Crude fiber (XF)	g kgVS−1	370.4	360.9	380.5	376.4
Neutral detergent fiber (aNDF)	g kgVS−1	659.7	567.9	622.7	645.9
Acid detergent fiber (ADF)	g kgVS−1	423.6	402.9	434.3	440.6
Acid detergent lignin (ADL)	g kgVS−1	81.2	91.0	96.1	98.3
Nitrogen free extracts (NfE)	g kgVS−1	556.5	560.9	543.5	542.3
Gross energy (GE)	MJ kgTS−1	17.7	16.8	16.8	17.2
Metabolizable energy (ME)	MJ kgTS−1	7.5	7.2	6.6	6.9
Net energy lactation (NEL)	MJ kgTS−1	4.2	4.1	3.7	3.9

Abbreviations: UT = untreated, CFG = cross-flow grinder, BM = ball mill, AG = Amazone Grasshopper, FM = fresh mass, TS = total solids, VS = volatile solids (TS, VS n = 3, AA = 2, others n = 1), g = gram, kg = kilogram, MJ = megajoules.

), crude protein (XP), crude fat (XL) and crude fiber (XF) from the content of total solids.

It should be noted that the coefficients in this equation are predetermined and represent the specific energy content of each component parameter [32].

2.7.4. Statistical Evaluation

Statistical significance (p < 0.05) was tested by one-way ANOVA (Excel, Microsoft Corp., Redmond, WA, USA) followed by Tukey–Kramer post-hoc test using Real Statistics Resource Pack software for Excel XRealStats (Release 8.8.2) [63], for the statistical analyses of kinetic constants and specific methane yields.

A generalized linear model was used to compare the measured values of the different treatments at different measurement times, and Bonferroni Sequential was selected as an appropriate method to fit for multiple comparisons in scenarios with multiple contrasts. This approach was chosen to reduce the accumulation of alpha errors and thus reduce the likelihood that results would be misclassified as significant. The statistical program SPSS−IBM (International Business Machines Corp., Armonk, NY, USA) was used for this purpose. The significance level was p < 0.05.

The graphs were created using OriginPro 2022 (OriginLab, Northampton, MA, USA).

3. Results and Discussion

3.1. Biomass Yields

The biomass yield was 3.10 t_VS ha⁻¹ (12.91 t_FM ha⁻¹ 3.47 t_TS ha⁻¹). Normally, the plots are mowed a second time, but here only the first cut was considered. The second cut accumulates less biomass on the land, which would further increase the energy requirement for harvesting. To compare harvesting techniques, the biomass produced per hectare must be recorded.

The understory biomass of orchard meadows as an agroforestry system is less than that of protected lean lowland meadows due to the additional tree cover. In the study by Boob et al. (2019) [4], more than 4 t_TS ha⁻¹ were reached at the time of seed maturity during the first cut, and the studies were conducted in the same area at the foot of the Swabian Alb. In general, calcareous soils, such as those found in the Swabian Alb, have the highest species diversity in Europe [64]. The existing tree population with a trunk height of over 1.80 m creates further niches [65]. However, the calcareous soil has the property of poorly retaining rainwater, so the biomass produced can be classified as low [66].

When considering the biomass potential of other landscape management areas, such as the Nuthe Nieplitz Niederung Nature Park (Brandenburg, Germany), it was found that meadow foxtail, purple moor grass, and acute sedge vegetation achieved biomass yields of 4.1, 5.1, and 5.4 t_TS ha⁻¹, respectively [18,67,68,69]. Higher biomass yields were obtained in Narew National Park (Poland), where wetland species in protected landscape areas such as reed and tufted sedge achieved biomass yields of 9.3 and 7.4 t_TS ha⁻¹ when harvested in late summer [70]. However, the early harvesting of these areas is usually not compatible with the requirements of landscape management areas [10].

In addition to late mowing timing influencing biodegradability by increasing lignin content, the amount of biomass plays a major role in determining the methane yield potential of wetlands [3].

3.2. Chemical Composition of Raw Materials

The pH of the initial inoculum was 8.2, and the FOS/TAC value was 0.16, both of which are in a range suitable for later use in batch digestion tests [47]. The TS content was 5.3% and the VS content was 3.5% (66.0%_TS) on a fresh mass basis.

The total solids content ranged from 25.2% in AG to 26.9% in UT. The highest VS concentration was found in UT at 89.3%. The lowest VS content was found to be 83% in AG, a difference of more than 5%. While the lowest ash content was found in the AG sample, this can be explained by the harvesting method itself. With the traditional harvesting method, the rake tines are positioned very close to the ground, and due to the unevenness of the meadow and the somewhat more humid conditions, soil may have been introduced into the substrate, which could explain the higher ash content.

The pH values of landscape management grass ranged from 5.9 for UT to 6.3 for AG.

Pre-batch digestion analysis revealed differences in acetic acids in the treatment variants. Acetic acid in BM was 0.35 g kg_FM⁻¹, followed by CFG, AG and UT with 0.14, 0.10 and 0.09 g kg_FM⁻¹, respectively (Table 1).

A different result is obtained for sugar content. UT has the highest sugar content with 75.5 g kg_VS⁻¹, followed by CFG, AG and BM with 42.5, 30.5 and 27 g kg_VS⁻¹, respectively. XP content ranged from 51.9 g kg_VS⁻¹ in UT to 58.9 g kg_VS⁻¹ in AG. XL content was about 2% in all variants, and XF content also varied only slightly within variants from 360.9 for CFG and 380.5 g kg_VS⁻¹ for BM. The late cutting date resulted in protein content below 8% and higher fiber content; therefore, the substrate has poor forage quality as both usable energy content and protein content decrease with the increasing maturity of grassland biomass, while fiber content increases [4]. Therefore, landscape management grass is not suitable for extensive livestock grazing or other feed uses [4,10,22,71,72], so strategies for energy use are well justified.

Among cell wall components, the greatest difference between UT, CFG, BM and AG was found in the content of aNDF. UT had the highest content with 659.7 g kg_VS⁻¹, followed by 645.9 g kg_VS⁻¹ in AG, 622.7 g kg_VS⁻¹ in BM, and almost 10% less, 567.9 g kg_VS⁻¹, in CFG. The ADF content ranged from 402.9 g kg_VS⁻¹ in CFG to 440.6 g kg_VS⁻¹ in AG, and ADL content ranged from 81.2 g kg_VS⁻¹ in UT to 98.3 g kg_VS⁻¹ in AG. Since the range of the contents of ADF and ADL are not so far apart in percentage, it can be concluded that the hemicellulose content is different between the variants and that there was significantly less hemicellulose in CFG than in UT. Hemicellulose is more easily degraded than cellulose, but cellulose has a higher biomethane potential [73]. The higher aNDF and sugar content could be a reason for the high specific methane yields in UT, so this study may overestimate this variant. However, the difference in hemicellulose in CFG is probably due to the heterogeneity of the meadow, which is also reflected in the subsequent sampling. NfE was calculated to determine GE. Since XP, XL and XF differed only slightly between variants, no significant differences in NfE could be detected. The values of NfE varied in a range from 542.3 in AG to 560.9 g kg_VS⁻¹ for CFG, and they averaged 55%. GE ranged from 16.75 to 17.69 MJ kg_TS⁻¹.

The chemical constituents differed to some extent among the variants, which was not due to the treatment but also reflected the heterogeneity of the substrate. The reason for this could be the sampling, which reflects the heterogeneity of the substrate due to the different composition of the plants in the meadow [74].

The ME for all cattle, except dairy cows, and NEL for ruminant milk production are used as benchmarks for the energy evaluation of the substrate [75]. ME ranged from 6.60 to 7.51 MJ kg_TS⁻¹ and NEL ranged from 3.68 to 4.24 MJ kg_TS⁻¹. These values must be considered too low if these areas are to be used for extensive livestock farming [76]. In addition, sheep grazing should be avoided where mowing can be continued, especially in species-rich meadows [40].

After the experimental period of 50 days, a sample of the liquid digestate was taken from each reactor. The results of the analysis show that all samples had undetectable concentrations (0 ppm or 0 g kg⁻¹) of examined VFAs. This absence of volatile fatty acids suggests that the samples underwent a complete degradation of these compounds and thus an efficient anaerobic digestion [77].

Due to the design of the laboratory digesters, it was not possible to take a representative sample of the solid fraction after the fermentation process, especially since the grass fibers (except treatment variant BM) were wrapped around the agitators, which made sampling even more difficult. Investigating the VS degradation would have been an interesting approach to better evaluate and understand the degradation process, but this was infeasible as it was not possible to take a homogeneous sample consisting of liquid and solid fractions of the digestate. The rheological flow properties and viscosity of the different treatment variants as well as analysis before and after the fermentation process are also parameters that play an important role in practice when it comes to saving agitation and pumping energy. Unfortunately, the laboratory digesters could not be reconstructed for the reasons mentioned.

3.3. Effects of Pretreatment on Surface Morphology

The color of the substrate changes due to the heat generated during mechanical processing, but also due to the squeezing effect of the grinding media in the BM or the chain elements in the CFG (Figure 5).

Grinding leads to an apparent increase in the surface area of the grass (Figure 6: BM-1-2) compared to fresh and untreated material (Figure 6: UT-1-2). In addition, there may be changes in the surface roughness (Figure 6: CFG-1-3, BM-1-3), microstructure and fragmentation of grass fibers (Figure 6: CFG-3-4, BM-3-4). The increased surface area is particularly beneficial for anaerobic digestion as it allows better contact between the grass particles and the microbial consortia responsible for methane production. The improved accessibility of anaerobic bacteria to the grass can lead to the faster hydrolysis and degradation of complex organic compounds during anaerobic digestion, accelerating biogas production and increasing yield.

Broken particle ends are also visible in the UT and AG variants but are largely limited to straight-cut edges caused by the mowers (Figure 6: UT-2-4, AG-1, AG-4). Nevertheless, the AG variant also shows broken fiber ends due to internal pretreatment, but these are significantly less numerous (Figure 6: AG-2).

The total particle size is reduced by the mechanical pretreatment. In particular, the variant treated in the BM shows a distinct color change (Figure 5: Ball Mill) and a strong defibration (Figure 6: BM-2) as well as small and homogeneous particles compared to the other variants.

Mechanical pretreatment has a profound effect on the surface morphology of grass biomass and is therefore a crucial factor for the success of aerobic digestion processes of lignocellulosic biomass, such as landscape management grass. These pretreatment-induced changes are crucial for evaluating the suitability of alternative substrates for mechanical pretreatment and highlight the importance of considering surface morphology among other parameters when developing and optimizing pretreatment processes.

3.4. Harvest and Pretreatment Process Data

Since one of the three processing methods already takes place in the field, the energy balance of the harvesting process was included in the analysis. The results of labor time requirements are measured in s ha⁻¹ and are shown in

Table 2. Work requirement of the 2 different harvesting methods in the field for the first cut. Hand-guided machines used under the tree cover and the energy demand required to get to the BGP go unnoticed.

Activity and Determination Method	Unit	Treatment Variant
		UT	CFG	BM	AG
Setup Time	s ha−1	951	-	-	190
Mowing	s ha−1	4739	-	-	22,731
Swathing	s ha−1	4396	-	-	-
Loading	s ha−1	3086	-	-	-
Total Harvesting Time	s ha−1	13,172	13,172	13,172	22,921
Total Harvesting Time	s tVS−1	4244	4244	4244	7384
Specific Energy Demand (Harvesting)	kWh tVS−1	95.2	95.2	95.2	163.9
Specific Energy Demand (Treatment)	kWh tVS−1	-	32.4	36.2	-
Total Energy Demand (Harvesting + Treatment)	kWh tVS−1	95.2	127.6	131.4	163.9

Abbreviations: UT = untreated, CFG = cross-flow grinder, BM = ball mill, AG = Amazone Grasshopper, s = seconds, ha = hectare, VS = volatile solids, t = tons, kWh = kilowatt hour, t = ton, 1 L Diesel = 35.87 MJ = 9.96 kWh [60].

. Labor time per hectare varied greatly due to plot size, tree cover, and slope, so the average value per harvesting method was calculated. The small meadow plots result in high setup times per hectare in the field, which were 951 s ha⁻¹ for UT and 190 s ha⁻¹ for AG. However, the harvesting method in AG takes almost 1.75 times longer than the front-mounted disc mower used for the untreated grass. This is due to the fact that, for practical reasons, it was used on areas that could not be accessed with a large tractor plus loader wagon because the plots were narrow or the tree cover was high, so damage would have occurred. This increased the amount of turning and shunting in the areas, which was not recorded separately. Therefore, due to their maneuverability, these plots could only be harvested using the smaller tractor with the attached AG mounted mower. Other factors included the smaller working width, slower travel over the area under certain conditions, and the time required to unload the hopper volume, which took more time with the AG method. Since harvesting with the AG method involves mowing, mechanical pretreatment, and the collection of the mowed material in one operation, the total travel distance between the BGP and the harvest area is reduced. This has a positive effect on the overall energy balance of the AG method.

The specific energy requirement for harvesting is 95.2 kWh t_VS⁻¹ for UT and 163.87 kWh t_VS⁻¹ for AG. However, only part of the energy balance was considered in this study; in addition to the transport routes, the handling of the substrate during operation was not included in the energy balance. The study by Meyer et al. (2015) [16] dealt in detail with the energy balance of nature conservation-relevant meadow grass for biogas production and provides an overview of the individual process steps and the values that can be used for them. In this study, 1315 MJ ha⁻¹ (365.2 kWh ha⁻¹; 3.6 MJ = 1 kWh [60]) was assumed for harvesting with a forage harvester, which also reflects the additional effort required for smaller plot sizes (3–6.5 ha) and is aligned with studies by Smyth et al. (2009) [78] and Dubgaard (2012) [79]. Based on the harvested amount of the considered area, the value is 117.8 kWh t_VS⁻¹. Thus, the results are in a similar range of values between the two harvesting methods.

The mechanical treatment of CFG and BM was associated with specific energy requirements of 7.79 and 8.71 kWh t_FM⁻¹, respectively. For CFG, 771 ± 5 kg of fresh and untreated landscape management grass was fed. The treatment time was 0.63 h. Thus, the throughput was 1.22 t h⁻¹. For the BM, 533 ± 5 kg_FM was fed. The duration of the process was 0.53 h, corresponding to a throughput of 1.00 t h⁻¹. CFG consumed 6 ± 1 kWh and BM consumed 4.64 ± 0.5 kWh during the treatment process. The total energy demands for UT, CFG, BM and AG are 95, 128, 131, and 164 kWh t_VS⁻¹, respectively. In the studies by Leible et al. (2015) [24] on the treatment of landscape maintenance grass with CFG, the energy demand for 1 t_FM⁻¹ was between 21 and 27 kWh t_FM⁻¹, about three times higher than in the case investigated in the present study. Since the BM runs in continuous mode, the CFG was also operated in this mode. The batch operation of the CFG results in higher energy demand but also higher treatment efficiencies. Nevertheless, in contrast to this case, no significant additional methane yield could be determined in the previously mentioned study [24].

3.5. Effects of Pretreatment on Specific Methane Yield and Degradation Kinetics

The untreated variant yielded 263.9 ± 7.9 L_CH4 kg_VS⁻¹. Similar results were reported by other authors. Cossel et al. (2019) [80] determined 289.3 ± 4.0 L_CH4 kg_VS⁻¹ for a permanent grassland. It should be noted that the authors ground the substrate to less than 1 mm and digested it in 100 mL glass bottles for 35 days. In general, the specific methane yields are in the upper range of values found in the literature (80–315 L_CH4 kg_VS⁻¹) [7,18,67,81,82].

Across all digesters and over the entire duration of the experiment, the average quality of the biogas was 50.93% methane.

The review by Prochnow et al. (2009) [7] provides a good and general overview of the potential of extensively and intensively used grassland for biogas production. Intensively used grassland reaches area-specific methane yields of 1872 m³_CH4 ha⁻¹ with a late first cut [83], in our case yields range from 820 to 865 m³_CH4 ha⁻¹ depending on the degree of treatment, i.e., not even half compared to the intensively used land. Later cutting dates are negatively correlated with SMY [84]. Since the main purpose of harvesting is to maintain or enhance biodiversity, the timing of harvesting should be planned accordingly [16].

The mechanical pretreatment methods BM and AG showed higher specific methane yields of 5.8% and 5.1% respectively, compared to the variant UT (

Table 3. Cumulative specific methane yields of untreated and pretreated landscape management grass.

Determination Method	Unit	Treatment Variant
		UT	CFG	BM	AG
Specific methane yield	LCH4 kgVS−1	263.9 ± 7.9 a	275.7 ± 4.0 ab	279.1 ± 2.9 b	277.2 ± 2.0 b
Additional methane yield	%	-	4.5 ± 3.0	5.8 ± 2.9	5.1 ± 2.8
Methane energy	MJ kgVS−1	9.5	9.9	10.0	10.0
Energy recovery	%	58.3	68.1	68.4	65.0

Abbreviations: UT = untreated, CFG = cross-flow grinder, BM = ball mill, AG = Amazone Grasshopper, L = liter, kg = kilogram, VS = volatile solids, MJ = megajoule, 1 m³ CH₄ = 35.894 MJ = 9.971 kWh [61], ^ab = significance.

).

The CFG variant in continuous flow was subjected to mechanical pretreatment with rotating steel chains and had a specific methane yield of 275.7 ± 4.0 L_CH4 kg_VS⁻¹ which was not statistically significantly different from UT (Figure 7). No significant differences were found when comparing CFG to the other two treatment methods. Due to the long grass blades, the fibers in CFG were only struck but not shredded. Mönch-Tegeder et al. (2014) [36] additionally mention that batch operation instead of continuous operation is recommended for intensive pretreatment. Due to the shorter retention time in continuous operation, the disintegration effect decreases while the output efficiency increases.

A notably higher specific methane yield of 279.1 ± 2.9 L_CH4 kg_VS⁻¹ was exhibited by the variant of the BM. This difference was statistically significant compared to the UT variant. This result can be attributed to the intensive disintegration process of the BM and the longer retention time in the drum compared to the CFG. In addition, the particle size was clearly smaller compared to the other pretreatment variants, and the color of the grass particles changed from light to dark green during the process (Figure 5 and Figure 6: BM-1), indicating an intensive impact on the substrate.

The AG mounted mower, which was subjected to mechanical pretreatment during the harvest of the lowland hay meadow, had a specific methane yield of 277.2 ± 2.0 L_CH4 kg_VS⁻¹, and this result was also statistically significant compared to UT. In the studies by Tsapekos et al. (2017) [85,86], different harvesting methods were compared with classic disc mowers, and an increase in methane yield of 10 to 20% was found.

When the four variants are examined in more detail across the fifteen measurement time points (MTP), it can be seen that significant differences can be found between day 36 and day 50 for CFG, BM and AG. For UT, on the other hand, there are no significant differences between the 30th, 36th, and 50th MTP. Thus, all three treatment techniques still had an effect on methane production and still produced a measurable additional specific methane yield towards the end of the experiment. UT is significantly different from the other variants in almost all MTPs. Only at the last MTP is there no significant difference between UT and CFG. This strongly suggests that CFG also has a detectable effect on the substrate. When CFG was considered in relation to BM and AG, significant differences were initially found at the beginning of the experiment. When considering the variants of each MTP, significant differences were found between CFG and BM until day 23 and between CFG and AG at the first measurement time point (first day). Thereafter, the variants converged. There were significant differences between BM and AG up to day 13 of the experiment and also at day 23. Toward the end of the experiment, no significant differences were found between CFG, BM and AG. This indicates that mechanical treatment has a major impact on kinetics and degradation processes and has the potential to improve the methane yield of landscape management grass.

Since an effect of pretreatment on kinetics was assumed, the specific methane yield values from the batch digestion experiments were fitted using the modified Gompertz model. The results of the modified Gompertz fit are shown in

Table 4. Results of modified Gompertz fitting.

Determination Method	Unit	Treatment Variant
		UT	CFG	BM	AG
Rm	LCH4 kgVS−1 d−1	24.4 ± 0.9 a	25.2 ± 0.6 ab	25.8 ± 0.5 ab	26.2 ± 0.4 b
λ	d	1.02 ± 0.2 a	0.56 ± 0.5 ab	0.01 ± 0.0 b	0.43 ± 0.0 ab
½M(x)	d	6.5 ± 0.2 a	6.1 ± 0.5 ab	5.4 ± 0.2 b	5.7 ± 0.1 ab
Rm(xmax)	d	5.0 ± 0.1 a	4.6 ± 0.5 ab	4.0 ± 0.1 b	4.3 ± 0.1 ab
Adj. R2		0.978 ± 0.003	0.968 ± 0.007	0.970 ± 0.002	0.972 ± 0.002

Abbreviations: UT = untreated, CFG = cross-flow grinder, BM = ball mill, AG = Amazone Grasshopper, L = liter, kg = kilogram, d = day, VS = volatile solids, R_m = maximum daily methane production rate, λ = lag time, ½M(x) = duration to reach half of total gas production, R_m(x_max) = day when the highest gas production rate is reached (n = 3, mean value ± standard deviation), ^ab = significance.

. Statistical analysis by a Tukey–Kramer post-hoc test (significance level p < 0.05) revealed significant differences, and homogeneous groups are indicated with the same lowercase letter.

All treatment variants had higher maximum daily methane production rates (R_m) compared to the untreated variant. The differences were statistically significant between the untreated variant and the AG treatment. The highest R_m was observed in the AG treatment with 26.2 ± 0.4 L_CH4 kg_VS⁻¹ d⁻¹. For the pretreated BM and CFG variants, R_m was present in both homogeneous subgroups and thus did not show much variation. In general, it was found that the values obtained for the different variants differed significantly, especially between BM and AG, with the only exception being the value for R_m, where AG differed significantly from UT.

The BM treatment had the shortest lag phase of 0.01 days, meaning that gas production started about half a day earlier compared to the other treatment variants and more than a day earlier than the untreated reference. It had the lowest value for the time to reach half of the total gas production (½M(x)), at 5.45 ± 0.2 days, followed by the AG treatment at 5.75 ± 0.1 days. The untreated variant had the longest duration at 6.47 ± 0.2 days.

The BM treatment reached the earliest time for maximum daily methane production rate at 3.99 ± 0.1 days, followed by the AG treatment at 4.32 ± 0.1 days. The untreated variant had the latest time for R_m(x_max) at 5.01 ± 0.1 days.

All treatment variants, as well as the untreated variant, showed high adjusted R² values, indicating a good fit of the model to the data and a reliable and accurate prediction of the methane yield [87].

Based on the data comparisons, all mechanical pretreatment variants (CFG, BM, and AG) resulted in significant improvements in the methane production potential and degradation kinetics of the landscape management grass compared to the untreated variant. The BM treatment was found to be most effective as it has the highest methane yield, the shortest lag phase, and the shortest time to reach half of the total gas production. The AG and CFG treatments showed similar degradation kinetics.

The results suggest that mechanical pretreatment methods can positively affect specific methane yields, as well as the kinetics of methane production from landscape management grass. The factors that cannot be controlled and could influence the SMY are the time of harvest, the weather conditions during the harvesting process, the duration of harvesting and pretreatment, and the period of time between harvesting and processing, during which the substrate is already exposed to aerobic degradation. These influencing factors, from the point of harvest to utilization in the biogas plant, differ for all pretreatment variants and cannot be kept the same. Another challenge is the relatively low biomass yields compared to intensively managed grasslands [88]. In addition, the areas often consist of small, decentralized parcels, which, as already mentioned, results in higher costs for the harvest, salvage, and transportation of the grass cuttings to the biogas plant. When utilizing these substrates in biogas plants, problems arise when they are used directly in an unprocessed state. Due to the described properties of the lignocellulosic substrates, they must be mechanically processed first so that they can be fed into the biogas plant at all. As mentioned before, grass residues are legally classified as biowaste, which can make their use in biogas plants even more difficult [19].

The integration of biogas production into energy grids and markets requires the ability of a facility to respond to fluctuations in electricity demand and prices. Faster kinetics allow biogas plants to instantly adjust their operations to meet peak electricity demand or grid requirements. By increasing or decreasing their production more quickly, biogas plants can contribute to load control and thus to the stability of the electricity grid. This not only optimizes energy use but also improves the economic efficiency of the biogas production process.

For more sustainable biogas production, flexible plants should be able to process different raw and residual materials depending on availability, which, however, are subject to strong seasonal fluctuations. In the summer months, substrates from landscape maintenance measures are increasingly produced. Due to the increased workload caused by small-scale parceling, small quantities are continuously produced, so ensiling the substrate is not worthwhile [72]. In addition, there seem to be conflicting opinions on the suitability of landscape management grass for ensiling as the inhomogeneity of the substrate [6] and late harvest dates [65] can negatively affect the fermentation process, and a size reduction is necessary prior to the ensiling procedure [71]. The substrate is therefore well suited for direct utilization in biogas plants that are to be operated in a seasonally flexible mode, as the energy demand in the summer months is lower than in the winter months, and thus energy production can be adapted to the seasonally fluctuating demand.

By promoting flexible plant operation due to faster degradation kinetics through substrate pretreatment and the use of seasonally available organic feedstocks, biogas plants can play a key role in meeting energy demand, integrating into energy markets and promoting the transition to sustainable energy systems.

3.6. Energy Balance of the Pretreatment and Consideration of Profitability

After the investigation of the specific energy demand during the harvesting and pretreatment of the landscape management grass described in Section 3.4, the additional methane surplus, which is generated during batch digestion of the grass, is presented in Section 3.5. Finally, these two parameters are offset against each other in the energy balance shown in

Table 5. Energy balance of untreated and pretreated landscape management grass.

Treatment Variant	Specific Energy Demand kWh tVS−1	Additional Methane Surplus kWh tVS−1	Energy Balance kWh tVS−1
CFG	32.4	44.7	12.3
BM	36.2	57.6	21.4
AG (on the field)	68.7	50.4	−18.3

Abbreviations: CFG = cross-flow grinder, BM = ball mill, AG = Amazone Grasshopper, kWh = kilowatt hour, t = ton, VS = volatile solids.

.

The untreated landscape management grass, as found in its natural state in the field, has a specific energy demand of 95.2 kWh t_VS⁻¹. However, due to the lack of pretreatment, this variant serves as a baseline for comparison with the treated substrates.

As described in Section 2.3.1, the continuous mode was chosen to improve comparability between the pretreatment variants. CFG had a notably low specific energy demand of 32.4 kWh t_VS⁻¹. Furthermore, CFG provided an additional methane yield surplus of 44.7 kWh t_VS⁻¹. The energy balance for CFG was thus 12.3 kWh t_VS⁻¹, indicating that this substrate treatment not only requires minimal energy input but also produces an additional energy surplus in the form of methane.

BM-treated grass had a slightly higher specific energy requirement of 36.2 kWh t_VS⁻¹ compared to CFG, but this was significantly lower than the energy requirement observed for AG. Importantly, BM had an impressive additional methane surplus of 57.6 kWh t_VS⁻¹, which contributed to a substantially positive energy balance of 21.4 kWh t_VS⁻¹. This result highlights the benefits of ball milling as a pretreatment method for landscape management grass, which allows for increased methane production, and at the same time, higher energy efficiency in substrate treatment, resulting in a positive energy balance and thus an economical benefit.

The AG mounted mower chops the mowed material with the rotating fine-flail mower with a simultaneous collection function and has a considerably higher specific energy demand of 68.7 kWh t_VS⁻¹ compared to the untreated substrates. Furthermore, although AG provided an additional methane surplus of 50.4 kWh t_VS⁻¹, it still resulted in a negative energy balance of −18.3 kWh t_VS⁻¹. These results suggest that landscape management grass, when treated with AG, may not be a favorable method for biogas production. However, the AG harvest method has limited comparability to the traditional method because this method allows grass to be harvested where large tractors and loader wagons cannot be used due to tree cover, which are common in orchard meadows.

An increase in pretreatment efficiency with the AG mounted mower can be achieved by lowering the distance to the ground. However, in this case, a higher ground clearance was selected, and an attempt was made to reduce the air flow when driving over the fields, which sucks insects into the mower, in order to make the process as species-friendly as possible. The blades used were also adapted to species-rich mowing meadows. The goal was not only to achieve maximum biomass for a higher-area methane yield with a high level of mechanical cultivation but above all to meet the requirements of nature conservation. A cost-effective alternative would be to simply mulch the areas and leave the substrate in the meadow, which is also possible with the AG mounted mower. This method is very likely to have a negative effect on the diversity of native species, as it has a greater impact on the fine-root features in the upper soil layer and also slightly increases the nutrient content in the soil, establishing a more productive plant community [89].

The additional electricity generated by the energetic use of the landscape management grass in a biogas plant, even if it is mechanically pretreated, only compensates for a small part of the effort for transport, mowing, and harvesting activities. However, it should be noted that the energy in the form of heat, generated during the conversion of biogas into electricity in the cogeneration plant, is not included in the balance. Nevertheless, it should also be noted that the use of landscape management grass offers further opportunities beyond the energy balance. The use of a digestate from the biogas process, which can be used as fertilizer on agricultural land, would be one way of expanding the value chain. Fiber production from the very fiber-rich landscape management grass could also be another possible use.

Furthermore, there will be social and political demands in the future for agricultural biogas production to be an integral part of the circular bioeconomy. The agricultural biogas plant of the future will be therefore primarily based on residuals [90].

The German Renewable Energy Sources Act 2023 (EEG 2023) sets the highest compensation amounts at 16.07 cents kWh⁻¹ for new plants and 18.03 cents kWh⁻¹ for existing plants in the tender procedure, so that electricity revenues of EUR 161–191 t_VS⁻¹ can be achieved for the substrate in Germany, depending on the processing method [91]. Consideration of the energy balance shows that 12.3 or 21.4 kWh t_VS⁻¹ only represents a small monetary value added of EUR 2–4 t_VS⁻¹. This is not sufficient to operate the treatment technology economically, since, in addition to the acquisition costs, maintenance costs as well as labor are needed to operate the plant. Nevertheless, pretreatment is necessary. This is the only way to introduce the straw-like biomass into a biogas plant, which without pretreatment leads to the clogging of plant components [37,92,93,94] or the formation of floating layers in the digester [92,95]. Since the pretreatment process is an energy-intensive operation, a reduction in energy demand would positively affect the economics of the entire process [31]. In addition, substrate pretreatment can also have a positive effect on the mixing and pumpability of the digester contents and the prevention of floating layers, which can lead to a reduction in stirring and pumping efforts and thus savings in operating costs [25,96]. However, these side effects are very difficult to prove in laboratory experiments and would need to be tested in field trials.

The total amount of energy of the substrate generated was 1000–1057 kWh t_VS⁻¹. The research conducted by Meyer et al. (2015) [16] demonstrated that utilizing nature conservation meadow grass for biogas production results in a positive energy balance, considering the energy needed for harvesting and processing the grass [16].

However, according to Blokhina et al. (2011) [22], profitability depends on factors such as low supply and investment costs; appropriate biogas production strategies; diversified revenue sources, including sales of electricity and heat; and subsidies for land use. In contrast to the mentioned study, our supply costs for the harvested material were three times higher than in their study. In their study, the supply costs were categorized as moderate. Thus, lower supply costs than in our case are required for economic profitability and use in a BGP. Further, they state that the most important factors for the profitability of biogas production from landscape conservation grass within the current German legislative framework are biomass supply costs, which must not be significantly higher than those of intensive grassland; low investment costs for the BGP; and the sale of heat.

Various subsidy programs, such as the Landscape Maintenance Guidelines (LPR) or the Climate Protection and Animal Welfare (FAKT II) in Baden-Württemberg (Germany), offer grants to financially compensate for the increased effort required to maintain these species-rich areas [97]. Particularly encouraging is the increase in compensation for the additional effort in grassland maintenance from EUR 2.50 per tree to EUR 5.00 per tree (eligible up to 100 trees per hectare) for traditional orchard meadows [98].

Adequate funding should be provided to ensure biodiversity conservation and ecosystem adaptability and development, as the benefits of these aspects are shared by the general public but are difficult to quantify. In this context, it has already been shown in the project Natural Capital Germany that the loss of natural capital also has economic impacts and ultimately proves detrimental to society [99].

In this project, nature conservation was accounted for with a monetary value between EUR 300 and 1000 ha⁻¹ a⁻¹. The net benefit of grassland conservation for High Nature Value Grassland is between EUR 440 and 2990 ha⁻¹ a⁻¹ [100]. In particular, traditional orchard meadows have other important characteristics, as they store 15% more carbon than permanent grassland [101]. Thus, they can reduce climate impact costs to a not inconsiderable extent, regardless of whether they are newly planted or the stand is maintained [102].

Unfortunately, these economic relationships are not yet sufficiently recognized in Germany, so the EU Commission has taken Germany to the European Court of Justice because the species-rich meadows in the NATURA 2000 protected areas are not managed in a species-appropriate manner, so their condition has deteriorated and 18,000 hectares have been lost [103]. As long as the use of this substrate is not profitable for farms and operators of biogas plants, it will not be used, which is why it is imperative that policy-makers make improvements at this point.

4. Conclusions

Based on the research results on different mechanical pretreatment methods for the anaerobic digestion of landscape management grass as potential organic feedstock for biogas plants, several key conclusions can be drawn.

First, landscape management grass shows promising potential with wide availability and seasonal occurrence as feedstock for biogas production without competing with land for food or feed production. However, the complexities associated with harvesting and pretreatment lead to higher costs.

In addition, the late harvest date of the grass results in high lignocellulosic and lignin content, which increases its resistance to enzymatic degradation. This in turn leads to slower degradation rates and lower biogas yields. Consequently, pretreatment methods are required to improve its suitability for biogas production.

In the context of pretreatment, the study shows that mechanical pretreatment, especially the use of a ball mill and the Amazone Grasshopper mounted mower, with its combined harvest and treatment directly in the field, allows for faster degradation (λ_BM: 0.01 ± 0.0 d; ½M(x)_BM: 5.4 ± 0.2 d; λ_UT: 1.02 ± 0.2 d; ½M(x)_UT: 6.5 ± 0.2 d) and leads to higher specific methane yields (SMY_BM: 279.1 ± 2.9 L_CH4 kg_VS⁻¹; SMY_AG: 277.2 ± 2.0 L_CH4 kg_VS⁻¹) compared to the untreated variant (SMY_UT: 263.9 ± 7.9 L_CH4 kg_VS⁻¹). This highlights the importance of appropriate pretreatment techniques in maximizing the energy yield of landscape management grass. In addition, the overall energy balance analysis shows the positive effects of using the cross-flow grinder (+12.3 kWh t_VS⁻¹), and in particular the ball mill (+21.4 kWh t_VS⁻¹), compared to the Amazone Grasshopper (−18.3 kWh t_VS⁻¹), highlighting their energy efficiency in pretreating lignocellulosic substrates for sustainable biogas production.

Overall, while landscape management grass has considerable potential as an organic feedstock for biogas production, its effective utilization requires a comprehensive understanding of the challenges associated with its harvest, its natural lignocellulosic composition, and the application of appropriate pretreatment techniques to maximize biogas yields.

Further research and experimentation are needed to optimize pretreatment strategies related to sustainable energy production through the anaerobic digestion of lignocellulosic biomass.