Nutrient Value of Digestate from Farm-Based Biogas Plants in ...

Nutrient Value of Digestate from Farm-Based Biogas Plants in Scotland. Report for Scottish Executive Environment and Rural Affairs Department - ADA/009/06.

ADAS UK Ltd Woodthorne Wergs Road Wolverhampton WV6 8TQ

SAC Commercial Ltd Kings Buildings West Mains Road Edinburgh EH9 3JG

July 2007

Nutrient Value of Digestate from Farm-Based Biogas Plants in Scotland

2

Contents 1 Executive Summary.......................................................................................................... 3

2 Introduction ....................................................................................................................... 4 2.1 Project purpose .................................................................................................................. 4 2.2 Project background............................................................................................................. 4 2.3 Technical background ........................................................................................................ 4 2.4 Nutrient benefit ................................................................................................................... 5 2.5 Project objectives ............................................................................................................... 6

3 Methodology and Approaches.............................................................................................. 7 3.1 Review of existing research ............................................................................................... 7 3.2 Chemical analysis of input slurry and digestate output ...................................................... 7 3.3 Further investigation of the effectiveness of plant nutrients in digested slurry.................... 8 4 Review of existing research .................................................................................................. 9 4.1 Digestate nutrient content .................................................................................................. 9 4.2 Nitrogen emissions during storage and following land application .................................. 12 4.3 Nitrogen fertiliser replacement value (following land application) .................................... 14 5 Chemical analysis of input slurry and digestate output....................................................... 17 5.1 Site 1. Ryes Farm............................................................................................................. 17 5.2 Site 2. Corsock Farm........................................................................................................ 18 5.3 Digester operation ............................................................................................................ 20 5.4 Analysis and nutrient content of digester feedstock and digestate .................................. 21 5.5 Conclusions on chemical analysis of slurry and digestate ............................................... 26 6 Further investigation of the effectiveness of plant nutrients in digested slurry ................... 28 6.1 Proposals for field assessment of digestate nutrient value.............................................. 28 6.2 Proposed modelling appraisal of nutrient fluxes following land application of digestate. 31 6.3 Other suggestions for action ............................................................................................ 31

Acknowledgements ................................................................................................................. 32

7 References...................................................................................................................... 33

Annex A Anaerobic Digestion and Digestate Analysis..................................................... 36

Annex B Glossary of Terms ............................................................................................. 37

Annex C Nutrient content of livestock slurries before and after anaerobic digestion....... 39


3

1 Executive Summary There is a strong body of opinion that, among the claimed benefits of anaerobic digestion, there are improvements in the effluent (digestate) quality as a result of the digestion process. This project considered this aspect of anaerobic digestion via:

• a detailed technical review of published and unpublished research data, and

• a short-term study of two farm-scale digesters in SW Scotland. Results of the detailed study of the farm-scale plant were reviewed in the context of the main findings within the technical review. During anaerobic digestion (AD), organic compounds are broken down by bacteria resulting in the production of methane and carbon dioxide. As a result of the digestion process a number of changes in slurry analysis can be expected. These include a substantial reduction (up to 25%) in solids content and a consequential increase in ash content, due to the conservation of minerals and reduced slurry carbon (and organic matter content). Increases in slurry pH (up to 0.5 pH units) and ammonium nitrogen (N) content (up to 25%) may also occur, though these changes are less consistent than the reduction in solids content and organic matter content, and may be transient or dependent on digester operating conditions and the analysis of the feedstock slurries. Because of the increase in slurry ammonium-N content, usually with increased pH and reduced solids content, there is a risk of increased emissions of ammonia during post-digestion storage. Such increased emissions have been confirmed by Danish research but have been shown to be effectively controlled by a range of store coverings. Although the increased pH and ammonium-N content might be expected to increase risk of ammonia emissions following application of slurry to the land, the reduced solids content would be expected to improve surface infiltration of the slurry which should help to conserve slurry N. Low emission application techniques are recommended for AD treated slurries. Increased ammonium-N content of slurries, even with reduced ammonia emissions, does not guarantee improved crop recovery and utilisation of slurry N and increased savings in fertiliser N. The limited research covering agronomic assessments has generated mixed results with small, short term, or inconsistent benefits. On the basis of available evidence, it is recommended that farmers with AD slurries should at least have an occasional laboratory analysis of digestate quality; this should include dry matter content, total and ammonium-N content, for which rapid field assessment techniques are also available. There is strong evidence, from the literature and from other recent research, to suggest that an increased availability factor for the phosphate (P) content of AD slurries should be considered, although the current study failed to show any increase in the water soluble P content of the digestate. It is recommended that carefully designed field experiments should be undertaken to assess the likely impact of AD on crop response to slurry N and the potential for fertiliser savings. Depending on the location of suitable P responsive field sites, slurry P availability should also be included within the proposed field experiments. Ammonia emission measurements should also be undertaken and a modelling assessment of the wider implications of AD on agriculture and the environment. This report was written by Ken Smith, John Grylls and Phil Metcalfe of ADAS; Bill Jeffrey and Alex Sinclair of SAC.


4

2 Introduction

2.1 Project purpose The purpose of this project is to identify whether there is any greater nutrient benefit derived from farm slurry that has been processed through anaerobic digestion (as some empirical farm work in South West Scotland and work undertaken by the Danish Agricultural Advisory Service has suggested), and to assess the likely costs of further research work on its use as a fertiliser.

2.2 Project background As part of a strategy1 to improve the quality of Scottish bathing waters, the Scottish Executive funded a number of projects to pilot innovative approaches to reducing the impact on bathing waters from diffuse agricultural pollution. One of these projects has been the installation of anaerobic digestion (biogas) plants on farms in the Sandyhills and Saltcoats catchments, both in the South West of Scotland2. Its purpose has been to examine the potential of anaerobic digestion as a tool to reduce the bacterial content of slurry prior to it being applied to land. Initial results have shown significant reductions. This would indicate that spreading the resulting digestate on the land would present a reduced risk of Faecal Indicator Organism (FIO) contamination of bathing waters. This study is to seek further clarification on whether there are other environmental benefits that may accrue through changes in the chemical composition of the major plant nutrients. The aim of this project is to review existing relevant research, to analyse the chemical characteristics of slurry before and after it has been through an anaerobic digestion plant; and to advise on the means to measure the effectiveness (benefits and risks to the environment) of anaerobically digested farm slurry as a fertiliser. This research will complement earlier farm biogas pilot studies undertaken by the Executive to examine the potential of digesting livestock slurry to reduce the risk of bacterial pollution to bathing waters.

2.3 Technical background During the anaerobic digestion process, organic compounds are broken down, firstly via acetogenic bacteria to methane precursors, largely volatile fatty acids (VFAs) and then to methane and other products via methanogenic bacteria. Under anaerobic conditions, organic forms of nitrogen (N) are converted into ammonium-N (NH4-N), i.e. readily available nitrogen. The readily available nitrogen (RAN) content of cattle slurry is typically 50% and pig slurry c. 60% of total-N (Anon, 2000). It might be anticipated that a measurable increase in the proportion of readily available N would occur in these materials, as a result of the digestion process. In addition to nutrient impacts, a number of benefits are claimed to accrue as a result of AD, including a reduced risk of odour nuisance and a reduction in viable pathogenic organisms (Sood, 2006).

1 ‘Scotlands Bathing Waters: A strategy for improvement’, available via the following link: http://www.scotland.gov.uk/Resource/Doc/46905/0031395.pdf 2 Anon (2006). ‘Farm Scale Biogas and Composting to improve Bathing Waters – a report for the Scottish Executive by Enviros/Greenfinch report (Feb 2006)’, available via the link: http://www.scotland.gov.uk/Resource/Doc/1057/0048383.pdf


5

A number of studies have demonstrated apparently significant changes in slurry composition following anaerobic digestion (Hobson et al., 1974; Baldwin 1993). The results of some early research in Germany (Vetter et al., 1987) showed a small reduction in slurry solids content, a decline in organic N content and an increase in NH4-N content (from 50% to about 60% of total N). However, without detailed information about the representative nature of the sampling (doubts about balancing of daily slurry input, sedimentation within store and whether the slurry input can be considered comparable to the digester output) such analyses can be misleading. The current project provided an opportunity for the study of nutrient transformations following anaerobic digestion of cattle slurries, with samples collected according to a rigorous protocol. The aim was to ensure comparability of raw slurry with the digestate and to contribute a robust and reliable set of results to the critical review process. Nitrogen can be taken up directly by plants as NH4-N or, more rapidly, as nitrate-N (NO3-N) following nitrification, a process which occurs very rapidly in fertile soils under favourable conditions. The plant uptake of N from digested (readily available N “enriched”) manures might therefore be expected to be closer to that from commercial fertilisers, as a result of the digestion process and may be regarded as a more predictable source of N than raw slurry of lower RAN content. However, it must also be remembered that in the NH4-N form, slurry N is more vulnerable to environmental losses. Substantial losses may occur to the atmosphere, as NH3 gas, both during slurry storage and, especially following land application. Furthermore, following the rapid conversion of NH4-N to NO3-N in the soil, further losses to surface and ground waters can readily occur through nitrate leaching and, to the atmosphere, as nitrous oxide gas (N2O) following denitrification. Changes in slurry P availability may also occur as a result of the release of P from organic forms during digestion, leading to an increase in the water-soluble P fraction. This may increase the vulnerability of slurry P to losses by surface run-off or via by-pass flow through field drainage systems, unless application practices are carefully managed and controlled. This study will highlight the extent of nutrient transformations within cattle slurries during anaerobic digestion, taking account of other evidence within the published literature and in unpublished research reports accessed via national and international contacts. The findings of this work will inform proposals on the need for, and the structure of, potential future research on the nutrient benefits and dis-benefits of AD in Scotland.

2.4 Nutrient benefit The nutrient parameters included in the study were dry matter (DM) (solids content), organic matter, pH, total N, NH4-N, NO3-N, total P, bio-available P (water soluble), total K, total S, total Mg and total Na. Where possible, data on nutrient transformations would be related to the digestion process, in particular, flow rate and digester retention time. Continuous flow digesters normally have a retention time (RT) of c. 20 days and sampling should relate to normal digester operation and include records of flow volumes during the monitoring period, so that nutrient balances can be constructed. Assessment of the likely effects of the digestion process on digestate quality should include consideration of potential for:

• Improved uptake of nutrients by growing crops and reduced losses to the water and air environments;


6

• Improved crop responses, increased yield and savings in inorganic fertiliser nutrient inputs;

• Improved predictability of manure nutrients for utilisation by crops;

• Reduced risk of scorch to growing crops. Assessment of these potential benefits will be based upon the available evidence in the scientific and research literature and on the monitoring undertaken within the project. However, nutrient benefit will ultimately depend upon overall manure management on the producing unit; for example:

• Nature of slurry, e.g. dilute slurry of low solids content and high NH4-N content, will be changed to a lesser extent by digestion than a high DM slurry;

• Losses of ammonia following land application may be impacted by other components of slurry analysis, e.g. DM content, pH; and by timing and method of application;

• Availability of adequate slurry storage on the unit (influencing timeliness of application);

• Post-digestion physical treatment, such as solids-liquids separation;

• Range and extent of cropping on the farm;

• Crop growth stage at the time of application;

• Soil type and land accessibility. The research outputs will include recommendations on the best way to maximise potential benefits and on the need for future research.

2.5 Project objectives The overall objectives of the project were, thus, to provide: (1) review of existing research on the environmental benefits and the nutrient value of farm slurry digestate from anaerobic biogas systems; (2) comparative chemical analysis of farm slurry and the digestate resulting from the anaerobic digestion of that slurry; and (3) proposals for field trials to evaluate crop response to farm slurry and biogas digestate and, thus, to determine the potential chemical fertiliser replacement value of digestate compared to untreated slurry; also for a modelling approach to undertake a range of scenario analyses on the nutrient benefit and the likely wider environmental impacts of the N content of raw slurry and biogas digestate across typical Scottish farming systems and environmental conditions.


7

3 Methodology and Approaches ADAS and SAC Commercial Ltd met representatives of SEERAD at a Project Inception Meeting to establish working links and agree project approaches, including the selection of preferred farm-scale digester sites for sampling and monitoring and the detailed sampling protocol.

3.1 Review of existing research In order to identify suitable information sources, some preliminary networking and initial scoping of known reference material was undertaken, to identify further key reference data. This also included an outline internet search of known research organisations in Europe and USA, for example, the FAO RAMIRAN network conference proceedings and research database (www.ramiran.net). Follow-up requests for papers and reports were made, initially largely via existing relevant contacts; RAMIRAN network, N European network of specialists (Danish Agricultural Advisory Centre), EU-AGRO-BIOGAS STREP project (T Amon, University of Natural Resources and Applied Life Sciences, Vienna). Relevant analytical data and technical information were also collected from recent projects in the UK (e.g. the Holsworthy project, Devon).

3.2 Chemical analysis of input slurry and digestate output Careful site selection of two representative farm-scale digesters was agreed using local knowledge (SAC) and in consultation with the installing company (Greenfinch Ltd, Bishops Castle, Shrops) and the project Steering Group. Selection criteria included consideration of: Range and type of livestock; Livestock feeding system; Match of digester with livestock slurry production and calculated retention times; Potential for homogeneous and consistent feedstock and representative sampling; Location, management and capacity of the farm to accommodate sampling visits.

Sampling and analysis costs The detailed work plan included provision for sampling of the two farm sites on two occasions per week over a four-week period. Two samples were collected on each occasion (input and outlet samples), i.e. a total of 18 samples per site and 38 samples in total, including separate samples from each of the digestate stores.

Sampling methodology The proposed sampling methodology was designed to facilitate the comparative analysis of feedstock slurry and digestate. Obstacles for obtaining comparable input and output data stem from the digester retention time, and the consequential time lag between the material passing into the digester and its appearance as digestate. To accommodate this “system inertia”, samples were taken on a twice-weekly basis, spanning four full weeks of digester operation; giving operational coverage over a c. 30 day period, thus covering, with some margin, the likely operating digester retention time. Sampling on Monday and Thursday in each week, starting in week 1 and finishing with final samples on Monday, week 5, yielded a total of 18 samples from each of the two plants (36 samples in total).


8

Analyses included: DM (solids content), organic matter, pH, total N, NH4-N, NO3-N, total P, bio-available P (water soluble), total K, total S, total Mg, total Na. The sampling was carried out by experienced scientific staff from SAC, using standard operating procedures and within the agreed protocol. On collection, the samples were cooled and refrigerated, then submitted for analysis within 24 hours to SAC, Analytical Services Department, a designated UKAS accredited laboratory.

3.3 Further investigation of the effectiveness of plant nutrients in digested slurry The analysis dataset was considered in relation to information on digester operating conditions. The potential implications of these results were evaluated in the context of other research results and published information and further research needs considered.


9

4 Review of existing research The anaerobic degradation of organic substances to the most reduced form of methane (CH4) is a microbial process. The energy released in the process is mostly recovered in the methane. The degradation of organic substances is a complex process, involving (i) (slow) enzymatic hydrolysis and the formation of sugars, amino acids and fatty acids; (ii) (fast) acetogenesis of volatile fatty acids (VFAs) and (iii) methane (and CO2) formation. A number of groups of bacteria are involved in the various stages. Details of the process are available from a number of sources (Hobson et al., 1974; Møller, H.B., 2001; Burton and Turner, 2003) and an appreciation of at least part of the biochemistry will assist in understanding the nutrient transformations occurring during digestion and the nutrient content of the final digestate product. pH and buffer capacity – the equilibrium of CO2 and bicarbonate (HCO3

-) with ammonium cations (NH4

+), exerts substantial buffering on digestate pH, the breakdown of organic acids generating CO2 and, hence, carbonic acid in solution: CO2 + H2O → H2CO3 → HCO3

- + H+ VFAs decrease the buffering capacity of the bicarbonate ions: RCOO-H + NH4HCO3 → RCOO-NH4 + H2CO3 While the addition of NH3 will increase bicarbonate in balancing the generation of CO2: CO2 + H2O + NH3 → NH+

4 + HCO-3

The higher the bicarbonate concentration, the greater the buffering in solution and resistance to changes in pH. The optimum pH varies according to the stage in the degradation process. Volatile fatty acids – the effect of VFA levels on the micro-organisms involved in the process is complicated by their impact on pH; with near neutral pH, the VFAs have no toxic effect on the methanogenic bacteria at concentrations < 10,000 mg/l. Ammonia is formed during the breakdown of proteins and, where free NH3 is formed, can act as a potent inhibitor of methanogenesis. Thus, it can be seen that pH and temperature (via its effect on pH) can have a strong effect on the NH3 concentrations and the stability of CH4 generation. It is reported that up to 1500 mg/l as NH4

+ can be tolerated though, with acclimatization, stable operation has been demonstrated at NH4-N concentrations of up to 8000 mg/l (van Velsen, 1979). Against this background, the evidence available from the research community and in the literature and, also, the supporting analytical data collected from the farm plant monitoring, all need to be considered in assessing the likely nutrient benefits and associated environmental impacts of the anaerobic digestion process.

4.1 Digestate nutrient content Analytical data to allow comparison of results for anaerobic digested and undigested animal slurries have been compiled from some UK (Tables C1-C5), European (Tables C6-C9) and USA (Table C10-C13) sources. These research data span a period from 1979 until 2007. Data derived from any controlled comparison of digester input (substrate) and resultant digester output (digestate) were limited, not least because the main focus of much of the research has been on digester performance in terms of energy balance. In the main, the results present the mean analyses of the animal manure digester feedstock and of the resultant digestate products. In some cases, there have


10

been mixed feedstocks (e.g. slurry from both cattle and pigs) and this is particularly the case in centralised anaerobic digesters (CADs) where feedstock materials such as abattoir and food processing wastes have been widely used. These materials will have a substantial impact on digestate analysis. In some cases, mean results from farm digester sites were for different numbers of digested and undigested slurry samples. Where it is thought that meaningful and reliable comparisons of digester input and output analyses can be made, the changes in analyses have been summarised in Table 1, with reference to further information in the relevant appendix tables. Table 1: Change in nutrient content as a result of anaerobic digestion (comparison between digester input and output expressed as % except for pH units)

Table1 Location Substrate DM N-total NH4-N P2O5 pH COD

A1 Suffolk Cattle & pig -10.0 13.0 15.0 18.0 0.45 -38A2 Yorks Pig -21.0 - 40.0 -6.2 0.5 -41A3 Kent Dairy cattle -29.5 -11.5 -12.4 -12.7 -0.09 -33.3A4 N Ireland Beef cattle -26.1 -14.3 8.7 - 0.4 -24.4A5 Scotland Dairy cattle -19.2 - 10.2 - - -17.2A8 Denmark Cattle - -7.0 32.0 - - - Denmark Pig - 0 14.0 - - - Denmark Pig - 0 13.0 - - - Denmark Pig - 0 42.0 - - - Denmark Sep. solids - 0 45.0 - - - Denmark Sep. solids - 3.0 52.0 - - -A10 USA (NY) Dairy cattle -25.2 10.4 33.3 3.2 0.5 -41.9A11 USA (Wisc.) Dairy cattle -35.4 -6.6 24.9 -8.4 0.6 -38.5A13 USA (NY) Dairy cattle -27.3 6.7 36.5 2.1 0.7 -30.3 USA (NY) Dairy cattle -25.1 0.9 27.7 0 0.18 -9.3 USA (NY) Dairy cattle -60.3 -4.6 11.3 -6.2 0.3 -61.3 USA (NY) Dairy cattle -11.1 3.5 37.7 5.9 0.29 -9.0 USA (NY) Dairy cattle -16.4 -5.5 31.1 10.9 0.22 -14.3

Mean -25.6 -0.8 25.7 0.7 0.4 -29.9 Median -25.15 0 29.4 1.05 0.4 -31.8 Observations 12 16 18 10 11 12

1 – Note source data from each of these sites presented in the appendix tables listed. Data from all of these research sites showed a reduction in slurry dry matter (DM) content as a result of anaerobic digestion with, overall, a difference of c. 25% between input and output slurry DM content (Table 1). This reflects the breakdown of organic matter and loss of carbon from the substrate, with the generation of CH4 and CO2. The substantial reduction in COD of c. 30% is also as anticipated and, whilst a much larger reduction in BOD of c. 70%, was also observed, these data were available from four of the UK research sites only and have not been included in the summary table. Increases in effluent NH4-N content and pH are also anticipated as a result of the generation of NH4-N (resulting from the degradation of proteins) and the production of CO2. Such changes were recorded in most of the studies and averaged a c. 26% increase in NH4-N and 0.4 unit rise in pH (Table 1). Although some of the data presented in the appendix tables show small and inconsistent changes in total N, P2O5 and K2O content, such changes would not be anticipated since these elements should be conserved during the digestion process. Moreover, any such apparent differences are thought to fall within typical sampling and analytical error and,


11

when averaged across the range of the more reliable data, they disappear (Table 1). The consistency in total content of N, P2O5 and K2O (confirmation of the anticipated lack of change), in fact, gives greater confidence in the reliability of the changes observed in DM, NH4-N and in pH. While much of the recent research data relate to performance of farm scale digester plants, some much earlier pilot-scale research has also provided valuable insight, e.g. studies on the impact of factors on the efficiency of digester performance included retention times (Summers and Bousfield, 1978). In these experiments at the Rowett Research Institute, optimum retention time for pig slurry digestion proved to be 10 days. Although this work showed generally increasing reduction in slurry DM content, BOD and COD, with increasing retention time, in contrast to much of the other research reported above, a short retention time resulted in increased slurry NH4-N content, with the opposite effect apparent with longer retention times (Fig. 1). Although this result at first appears contradictory to other evidence, this reflects the complexity of the process, with many different bacteria demanding N as well as energy from the mix of substrate materials available. In general, livestock manures supply a surplus of N, so there will usually be an increase in digestate NH4-N content as proteins are broken down in digestion. However, this state of flux will also depend upon the balance of nutrients including carbon supply, C:N ratio and the extent of bacterial growth and N utilisation.

0

1

2

3

4

5

3 days 5 days 7 days 10 days

Digester retention time

Slu

rry D

M c

onte

nt %

0

0.5

1

1.5

2

2.5

Am

mon

ium

-N g

/l

DM inputDM outputNH4-N inputNH4-N output

(+2%) (-1.7%) (-16%)(+21%)

Figure 1: Impact of digester retention time on DM and NH4-N content of pig slurry input and digestate; figures in brackets represent % change in NH4-N. (Summers & Bousfield, 1978). It is of interest to note that the US studies (Tables C10-C13) included orthophosphate analysis and that, in each case, an increase in orthophosphate content of the digestate compared to the influent slurry was recorded, in contrast to the unchanged total P


12

content. This reflects the solubilisation of some of the organic P as a result of the digestion process. In recent years, there has been increasing interest in centralised anaerobic digestion (CAD) plants. While comparative data on the analysis of feedstock and the digestate are generally unavailable, it is of interest to note the often high nutrient content of the output. Data from two CAD plants are presented in Table 2. Table 2: Comparison of digestate analysis for two centralised anaerobic digestion (CAD) plants CAD Plant Total-N

kg/m3

NH4-N

kg/m3

NH4-N/N

% total

P2O5

kg/m3

K2O

kg/m3

DM

%

Holsworthy1 (England)

6.6 5.0 75.8 3.3 4.5 5.8

Ribe – average 1992-96 (Denmark)

4.9 3.2 65.3 2.4 4.2 5.8

Ribe2 – 1992 (Denmark)

4.6 3.1 67.4 2.1 4.2 6.4

Cattle slurry (Denmark)

4.7 2.7 57.4 1.4 5.3 8.5

Pig slurry (Denmark) 5.3 3.7 69.8 3.4 2.8 6.0 1Feedstocks by volume 57% dairy cow slurry, 19% blood, 11% food waste, 8% chicken manure, 5% other non-farm waste. Results relate to May 2004. 2 Feedstocks (1992) by volume 84% from 71 farms (56 dairy, 7 pig, 3 mixed, 5 mink or poultry), 16% from industry (mostly from an abattoir) Sources: Holsworthy - O’Sullivan, C.M. and Cumby, T.R. (2004). Ribe - Holm-Nielsen et al., 1997 The relatively high total N and P2O5 content of the Holsworthy CAD digestate is likely to be due to the blood used as a feedstock. Blood has a high total N content (>15 kg/m3) compared to dairy cow slurry (c. 4 kg/m3 undiluted). For the Ribe CAD in 1992, 14% of the total N in the feedstock came from industry. An estimate of total N from blood for the Holsworthy CAD is >40%.

4.2 Nitrogen emissions during storage and following land application One of the possible consequences of the increase in slurry pH and NH4-N content following anaerobic digestion is an increased risk of NH3 losses during storage and after land application. Also the reduction in slurry solids content may reduce the likelihood for natural crust formation in stored slurry and this, too, may increase the risk of NH3 losses during storage (Smith et al., 2007). Some recent Danish research has studied the environmental effects of anaerobic digestion (Hansen et al., 2004). Anaerobic digestion and, especially, separation reduced slurry DM content (Table 3); digestion also increased slurry pH.


13

Table 3: Analysis of pig slurry used in storage and application method experiments Dry matter Total N NH4-N NH4-N Year Slurry type %

pH kg/m3 % of total

2002 Undigested 3.4 7.4 4.3 3.1 72 2002 Digested 3.2 8.1 5.2 3.7 71 2002 Digested

separated 2.1 8.3 4.8 3.6 75

2003 Undigested 3.3 7.2 3.7 2.4 65 2003 Undigested

separated 1.5 8.6 4.9 3.9 80

2003 Digested 2.8 8.1 4.3 2.9 67 2003 Digested

separated 2.2 8.2 4.2 3.4 81

In the first storage season, the slurry stores were covered with a 15 cm layer of Leca (lightweight-expanded clay aggregates) which resulted in low nitrogen losses from all slurry types. In 2003, however, the stores were left uncovered and, as anticipated, NH3 losses increased from digested and separated slurries, with the greatest loss being from separated undigested slurry (Table 4). These results were attributed to the elevated pH and low DM content in these slurries (Table 3). Table 4: Monthly relative loss of nitrogen from covered and non-covered stores with the four slurry types indicated as percentage of the initial nitrogen content.

Storage period

Cover treatment

Undigested slurry

Digested slurry

Separated undigested slurry

Separated digested slurry

09/01-01/05/2002

Covered* 0.8 0.9 - -0.1

20/03-06/05/2003

Uncovered 2.5 4.4 6.1 4.4

* Slurry stores for each slurry type covered with a 15 cm layer of Leca (lightweight-expanded clay aggregates) Ammonia losses were measured following application of 30 m3/ha of the slurries by trailing hoses to spring barley. The lowest losses following application were from the digested and separated slurries particularly in 2003 (Table 5). The reduced loss from digested and/or separated slurries reflects the likely quicker infiltration into the soil as a result of the low DM content.


14

Table 5: Ammonia loss following land application via trailing hoses to spring barley, losses expressed as a % of NH4-N applied

Year Undigested slurry Digested slurry Separated undigested slurry (liquid fraction)

Separated digested slurry (liquid fraction)

2002 27 22 - 26 2003 46 34 23 18

Some recent work in Finland has compared NH3 and greenhouse gas (N2O & CH4) emissions following undigested and digested pig and cattle slurry applications (Regina and Perälä, 2006). In field experiments in 2005-06 pig slurry was applied on barley with target soluble-N application rates of 100 kg/ha. With injected (undigested or digested) pig slurry, NH3 emissions were undetectable. Where the slurries were band spread before sowing the barley crop NH3 emissions continued until the slurry was incorporated (one hour after application). There were no statistical differences between emissions from the different slurries. When the slurries were band spread into the growing crop two weeks after sowing, NH3 emissions were higher than those for band spreading on the day of sowing. The digested slurry gave higher NH3 emissions than the undigested slurry, probably as a result of the high pH. The digested slurry was also separated and emissions from solid fraction gave higher emissions than the liquid fraction because of the lack of infiltration into the soil. Considering nitrous oxide (N2O) emissions over the first month after sowing, emissions were lower from the solid fraction of digested slurry than from the liquid slurries. Because the solid fraction could not infiltrate the soil, denitrification from this fraction was not likely. There were no clear effects of slurry digestion on the annual emissions of N2O. Digestion seemed to lower emissions compared to undigested slurry one month after injection, but later there were no marked differences between treatments. Digestion appeared to reduce CH4 emissions from slurry spreading. In cattle slurry experiments on grass, NH3 emissions were higher from digested than from undigested slurry (Perälä and Regina, 2006). Slurry injection decreased NH3 emissions but less for digested than for undigested slurry. Considering nitrous oxide emissions in the cattle slurry experiment, cumulative total emissions over the first four months were lowest from band spread digested slurry. Emissions from both digested and undigested slurry were much higher when the slurry was injected into the soil, with undigested being the highest. More CH4 was emitted from injected digested slurry than from band spread digested slurry, possibly indicating that there was some CH4 production in the soil in addition to the release of dissolved CH4 from the slurry. Cumulative CH4 emissions after both four and eleven months showed highest emissions from digested injected and lowest from digested band spread. Undigested gave emissions between these with band spread undigested being higher than injected undigested.

4.3 Nitrogen fertiliser replacement value (following land application) Claims of the beneficial impacts of anaerobic digestion on slurry analysis are of no real significance if these are not reflected by a positive benefit in terms of nutrient recovery by crops and of crop yield response. Whilst there have been few well controlled comparisons of the impact of digestion on slurry analysis, there are even fewer data where crop response has been assessed.


15

In Denmark, field assessments on the utilisation of slurries following a range of treatments, in particular including anaerobic digestion and separation, have been carried out for a number of years. In fact, it is claimed that the utilisation of N in manure has increased dramatically and the use of mineral fertiliser N has decreased by 50% (Sommer and Birkmose, 2007). These authors presented results from several years of research at a national crop production seminar including data from 11 trials with digested slurry and 15 trials with pig and cattle slurry (Fig. 2) and the Danish Advisory Service are now actively promoting the benefits of increased NH4-N content and improved utilisation of fertiliser N in digested slurries.

0

10

20

30

40

50

60

70

80

90

trailing hose injection

Slu

rry N

util

izat

ion

%

Digested slurry Pig slurry Cattle slurry

Figure 2: Utilisation of N in digested slurry compared with pig and cattle slurry in field trials with the Danish Advisory Service. (Sommer and Birkmose, 2007). These Danish results (Fig. 2) suggest an overall 15-30% increase in slurry N efficiency, depending on slurry type and application technique. However, the results of individual experiments are not always consistent, with sometimes only marginal benefit apparent from digestion, or higher efficiencies following slurry separation treatment (Pedersen, 2002). Schröder and Uenk (2006) studied the nitrogen fertiliser replacement value (NFRV) or N efficiency of undigested and digested cattle slurry. Efficiency was determined from grass DM yields and apparent N recoveries in a replicated field trial running from 2002-05. In each year a total of 300 kg/ha N was applied either as mineral N fertiliser, undigested slurry (50% of total N present as NH4-N) or digested slurry (58% of total N present as NH4-N). Application of the total amount of N as either mineral N fertiliser or slurries was split between the start of the growing season, after the first cut and after the second cut in 2002, 2003, 2004 and 2005. The slurries were applied by injection. The yields from mineral N fertiliser, undigested and digested cow slurry treatments were compared with control plots receiving no mineral N fertiliser or slurry. The relative NFRV of both slurries was calculated as the ratio of the apparent N recoveries (N uptake increase, kg per kg total N applied) or N efficiencies (DM yield increase, kg per total N applied) of slurry and mineral N fertiliser. In the first year the NFRV of digested slurry exceeded that of


16

undigested slurry by 5%. However this initial advantage was completely offset when residual N effects in years 2, 3 & 4 were taken into account, yielding similar long term NFRV’s for both types of slurry (Fig. 3).

0

10

20

30

40

50

60

70

Year 1 Year 2 Year 3 Year 4

NFR

V %

UntreatedDigested

Figure 3: Impact of anaerobic digestion on the nitrogen fertiliser replacement value of cattle slurry over four years following application (Dutch experiments) (Schröder and Uenk, 2006).


17

5 Chemical analysis of input slurry and digestate output As a result of the consultation process, two farm sites from within the Sandyhills catchment in Dumfries and Galloway, were selected for the detailed sampling and monitoring work, which after an introductory letter from the Scottish Executive, involved the full cooperation of the host farmers. The basic layout of the biogas plant, comprising holding tank for raw slurry, digester and digestate storage tank, is the same at both farms. Other relevant details of the two sites are presented below.

5.1 Site 1. Ryes Farm Slurry is sourced from 110 dairy cows plus dairy waste, but the digester was scaled to take slurry from young stock also. Currently, the digester (Fig. 4) is working below capacity, with a c. 40 day retention period (the design retention period is 21 days). The raw slurry holding tank of 84m3 has sufficient capacity for once a week loading by tractor pump. The digester volume is 251m3 and the storage tank volume, 1000m3, or around 3 month’s capacity. A small open yard area collects rainwater via a slatted tank, but the slurry cellars are known to admit groundwater. Maintaining a high level of slurry in the cellars minimises water ingress.

Figure 4: General view of digester facility at Ryes Farm. Operating regime: slurry is transferred weekly (usually at weekends) from the slatted tanks to the digester holding tank, by tractor driven pump. Both digester feed and discharge functions are achieved by identical helical screw pumps, both operating simultaneously from a timer. The run time is around 1 min 20 sec once an hour, but run time can be adjusted by the farmer. Five minutes before the transfer pumps run, the holding tank is agitated by recirculation (by high volume centrifugal pump), also once an


18

hour. The reactor tank is agitated by gas recirculation. Reactor temperature, pump run times and tank contents (holding, digester and storage) can all be read from the control panel. In addition, a diary is provided for the farmer to log events such as loading/unloading, changes in control settings and power failures. Gas production: in addition to maintaining digestion temperature, surplus gas is piped to a domestic heating boiler in the farmhouse. In general, gas production has exceeded demand and, rather than allowing surplus gas to vent to atmosphere the digester has been operating at a higher than normal temperature, around 42oC. This was simply in order to burn off the excess gas. Sampling procedures: Both feed and delivery screw pumps have sample ports accessible from inside the control room. It was not feasible to sample both during the hourly timed run periods due to the limited run time. Therefore the pumps were operated manually while sampling, keeping the time down to a minimum. The aim was for the recirculation pump to be run for a short period prior to sampling, with the reactor tank (digestate) sampled first, to avoid mixing with raw slurry, then the holding tank (feedstock). Two bucketfuls were taken, the first (to purge pipework) discarded, with samples drawn from the second. On the basis that further activity is known to occur in storage facilities post digestion, an attempt was made to also sample the storage tank to evaluate any possible further impact on slurry nutrient content. Provision was made for this using a weighted bucket on a rope lowered from the access ladder, using the bucket to give some local mixing – no agitation is provided. Unloading: normal farm practice is for the vacuum tanker to be coupled direct to pipework connected from the base of the storage tank.

(a) (b)

Figure 5: Ryes Farm (a) digester and gas holder; (b) feed pumps and controls

5.2 Site 2. Corsock Farm Corsock is a low maintenance beef fattening business. Around 185 cattle, average age 18 months, are kept in a variety of sheds over winter. Some sheds have slatted cellars; others are old cubicle buildings, scraped to small holding tanks. This inevitably means a considerable amount of rainwater addition to the slurry. As the farmer depends entirely on the biogas for domestic heating, he has sourced glycerol (a by-product of biodiesel production) to add to the slurry as a means of increasing gas production, particularly when significant dilution by rainwater has occurred. He adjusts the amount of glycerol


19

according to rainfall; this is added once a week along with slurry when loading the holding tank. The slurry transfer rate is calculated to retain sufficient raw material to continue gas production through the summer.

Figure 6: General view of digester facility at Corsock Farm

(a) (b)

Figure 7: Corsock Farm (a) raw slurry holding tank and yards and buildings for stock; (b) digester, gas holder and digestate storage tank. The raw slurry holding tank is of 37m3, sufficient for once a week loading by vacuum tanker. Digester volume is 78m3 and digestate storage tank volume, 455m3. Operating regime: slurry is transferred weekly on Saturday or Sunday from the various collecting tanks to the digester holding tank, by vacuum tanker. The system is otherwise identical to that at Ryes, even using the same types of pump for digester feed and discharge, although the run times are necessarily shorter at around 40 seconds once an hour.


20

Sampling procedures: identical to Ryes in every respect. The storage tank level sensor proved unreliable, so an attempt was made to assess the volume of contents by counting internal panel bolt heads visible from the access ladder. A similar event log is kept in the control room.

5.3 Digester operation An attempt was made to record mass flow of digester input and output. This was based on farm diary entries for amounts added and removed, plus tank levels recorded at each visit. The control panels included a read-out of levels in each tank (reception, digester, storage), although there was a fault in the storage tank level sensor at Corsock Farm, so that level was manually assessed by counting rivet heads visible from the top of the tank. Subsequently it was apparent that digester levels are affected by the heat exchanger circulation pumps, such that the level falls when the pumps cut in. As this was not known in advance, it was not possible to take account of this in assessing tank levels at Corsock Farm and, consequently, the recorded volumes were unreliable. Digester temperatures for both sites and estimated digester output volumes for Ryes Farm are shown in Fig. 8.

0

10

20

30

40

50

60

12-Feb 15-Feb 19-Feb 22-Feb 26-Feb 01-Mar 05-Mar 08-Mar 12-Mar

Slu

rry v

olum

e m

3

38

39

40

41

42

43

44

Tem

pera

ture

deg

C

temp

0

10

20

30

40


Slu

rry v

olum

e m

3

20

25

30

35

40Te

mpe

ratu

re d

eg C

Dig temp

Figure 8: Digester temperature and estimated output volumes of digestate at (a) Ryes Farm and (b) Corsock Farm during the monitoring period. At Corsock Farm, where the biogas was the sole source of heat for the farm, glycerol was occasionally added to the digester to ensure a consistent and reliable generation of biogas. This was to cover for the sometimes rather dilute slurry output from the farm as a result of the large volumes of rainfall and ground water accessing the system. Otherwise, a diesel-fired boiler was used to provide heat for both digester and farm. During the monitoring period no glycerol was added after Feb 18th. Problems with the diesel boiler meant that the digester temperature fell to 29 °C by Feb 26th and digester operating efficiency will no doubt have been impeded during this period. Repairs to the

(a)

(b)


21

diesel boiler restored digester temperature to 35 °C on March 5th, and to 39 °C on March 8th. Average operating temperature at Corsock Farm was 35.3 °C, which is typical for mesophilic anaerobic digestion. Because of the surplus gas production at Ryes Farm, the digester was operated at an elevated temperature, averaging 41.6 °C.

5.4 Analysis and nutrient content of digester feedstock and digestate Samples were collected for analysis from the two digesters on two occasions each week, commencing February 12th and finishing March 12th (9 sampling dates), thus covering by some margin the estimated hydraulic retention time of the digesters. Mean input and output analyses, with overall differences for the two sites are presented in Table 6. The details of these analyses, for both the input feedstock and the digestate output, are presented in Tables 7 and 8. Table 6: Mean digester input and digestate analyses with overall differences and estimated statistical significance.

Difference P value2

Ryes Farm Input Output %1

Dry matter % 8.1 6.23 -21.2 0.005 pH 7.68 7.84 0.16 0.264 Ash %DM 34.84 40.29 16.7 0.017 Total N % 0.29 0.29 -0.87 0.608 NH4-N % 0.11 0.13 17.7 0.092 NO3-N % 0.03 0.03 - - Total P2O5 % 0.111 0.108 -2.49 0.347 Water sol P %DM 0.012 0.012 - 0.299 Total K2O % 0.43 0.42 -1.9 0.383 Corsock Farm Dry matter % 7.68 6.76 -11.8 0.002 pH 7.35 7.58 0.22 0.008 Ash %DM 16.32 19.41 19.1 <0.001 Total N % 0.29 0.30 4.17 0.107 NH4-N % 0.11 0.12 4.5 0.296 NO3-N % 0.02 0.03 - - Total P2O5 % 0.10 0.11 8.2 0.065 Water sol P %DM 0.025 0.017 -32.0 <0.001 Total K2O % 0.23 0.23 3.4 0.262 1 Difference between output and input analyses expressed as % of input, except pH which is expressed as units; -ve values indicating a reduction between input and output level. 2 Indication of significance from paired t-test between input and output variables.


22

Table 7: Digester input and output analyses with sampling date – Ryes Farm

Date 12-Feb 15-Feb 19-Feb 22-Feb 26-Feb 01-Mar 05-Mar 08-Mar 12-Mar 08-Mar

Digester input Mean St.dev. cv % Store1

Dry matter % 9.1 9 9.1 7.9 8.6 7.9 7.7 7.5 6.1 8.1 0.979 12.08 -

pH 7.41 8.12 7.82 7.62 7.66 7.66 7.3 7.51 8.05 7.68 0.274 3.57 -

Ash %DM 34.5 34.9 35.3 33.7 34 32.9 33.7 32.1 42.5 34.84 3.033 8.70 -

Total N % 0.31 0.3 0.32 0.27 0.29 0.29 0.28 0.27 0.29 0.29 0.017 5.81 -

NH4-N % 0.104 0.111 0.083 0.106 0.115 0.104 0.115 0.111 0.157 0.11 0.020 17.47 -

NO3-N % 0.028 < 0.010 0.051 0.013 < 0.010 0.022 0.018 0.018 < 0.010 0.03 - - -

Total P2O5 % 0.12 0.11 0.12 0.1 0.12 0.11 0.1 0.11 0.11 0.11 0.008 7.04 -

Water sol P %DM 0.015 0.011 0.012 0.012 0.013 0.012 0.013 0.013 0. 009 0.012 0.002 13.43 -

Total K2O % 0.46 0.43 0.41 0.39 0.43 0.41 0.45 0.45 0.47 0.43 0.026 6.11 -

Digester output

Dry matter % 5.8 5.9 6.1 6.3 6.2 6 6 6.1 7.7 6.23 0.570 9.15 10.1

pH 7.77 8.07 7.86 7.87 8 7.88 8.05 7.78 7.32 7.84 0.225 2.87 7.44

Ash %DM 40.4 40.7 41 42 41.5 40.8 42.3 40.2 33.7 40.29 2.569 6.38 39.4

Total N % 0.27 0.29 0.3 0.28 0.29 0.3 0.29 0.29 0.28 0.29 0.010 3.38 0.30

NH4-N % 0.138 0.114 0.095 0.147 0.15 0.149 0.132 0.122 0.114 0.13 0.019 14.78 0.111

NO3-N % 0.015 0.035 0.056 < 0.010 0.011 < 0.010 0.018 0.036 0.015 0.03 - - < 0.010

Total P2O5 % 0.1 0.11 0.11 0.11 0.11 0.11 0.11 0.11 0.1 0.11 0.004 4.09 0.22

Water sol P %DM 0.012 0.012 0. 01 0.011 0.011 0.011 0.013 0.012 0.012 0.012 0.001 7.63 0.024

Total K2O % 0.41 0.37 0.41 0.41 0.44 0.43 0.45 0.45 0.45 0.42 0.027 6.36 0.3 1 Large digestate store sampled only on one occasion, 8th March.


23

Table 8: Digester input and output analyses with sampling date – Corsock Farm

Date 12-Feb 15-Feb 19-Feb 22-Feb 26-Feb 01-Mar 05-Mar 08-Mar 12-Mar 08-Mar

Digester input Mean St.dev. cv % Store1

Dry matter % 7.2 7.5 8.3 8.3 8.5 7.8 7.3 7.5 6.7 7.68 0.579 7.78 -

pH 7.07 7.32 7.55 7.37 7.36 7.45 7.39 7.31 7.3 7.35 0.130 1.77 -

Ash %DM 16.6 16.2 16.2 16.6 16 18.4 16 15.4 15.5 16.32 0.883 5.41 -

Total N % 0.3 0.3 0.29 0.3 0.32 0.31 0.27 0.29 0.25 0.29 0.021 7.21 -

NH4-N % 0.118 0.095 0.136 0.124 0.128 0.13 0.095 0.103 0.095 0.11 0.017 14.76 -

NO3-N % 0.022 0.048 < 0.010 0.021 0.015 <0.010 0.018 0.034 0.01 0.02 0.013 53.85 -

Total P2O5 % 0.1 0.12 0.11 0.11 0.11 0.1 0.1 0.1 0.09 0.10 0.009 8.44 -

Water sol P %DM 0.023 0.022 0.025 0.024 0.025 0.024 0.026 0.027 0.028 0.025 0.002 7.64 -

Total K2O % 0.24 0.22 0.23 0.24 0.25 0.23 0.22 0.23 0.19 0.23 0.017 7.53 -

Digester output

Dry matter % 6.3 6.3 6.5 6.5 8.2 6.8 6.9 6.8 6.5 6.76 0.583 8.63 4.0

pH 7.51 7.6 7.68 7.5 7.5 7.41 7.46 7.73 7.87 7.58 0.149 1.97 7.51

Ash %DM 20.4 19.8 19.5 20.4 17.8 19.6 18.9 19.3 19 19.41 0.805 4.15 22.8

Total N % 0.31 0.32 0.28 0.3 0.31 0.32 0.29 0.30 0.3 0.30 0.013 4.36 0.21

NH4-N % 0.111 0.103 0.156 0.129 0.112 0.126 0.1 0.112 0.115 0.12 0.017 14.37 0.095

NO3-N % 0.041 0.048 < 0.010 0.014 0.029 0.032 0.03 0.017 0.021 0.03 0.012 40.13 < 0.010

Total P2O5 % 0.11 0.11 0.11 0.11 0.11 0.11 0.12 0.12 0.11 0.11 0.004 3.93 0.09

Water sol P %DM 0.015 0.014 0.013 0.016 0.026 0.019 0.021 0.018 0.014 0.017 0.004 24.13 0.015

Total K2O % 0.25 0.23 0.22 0.23 0.24 0.23 0.24 0.24 0.23 0.23 0.009 3.76 0.22 1 Large digestate store sampled only on one occasion, 8th March.


24

(a) Ryes Farm (b) Corsock Farm

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4


Slu

rry N

con

tent

%

0

1

2

3

4

5

6

7

8

9

pH

Total N-inNH4-N-inTotal N-outNH4-N-out

pH-inpH-out

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4


Slu

rry N

con

tent

%

0

1

2

3

4

5

6

7

8

9

pH

Total N-inNH4-N-inTotal N-outNH4-N-out

pH-outpH-in

0102030405060


Slu

rry v

olum

e m

3

38394041424344

Tem

pera

ture

deg

C

temp

0

10

20

30

40


Slu

rry v

olum

e

20

25

30

35

40

Tem

pera

ture

Dig temp

Figure 9: Changes in slurry pH, total N and NH4-N over the monitoring period in digester input and output at (a) Ryes and (b) Corsock Farms; digester temperature for both sites and output volumes for Ryes, also shown.


25

In general the results followed the pattern of those identified within the technical review, with a substantial and highly significant reduction (P<0.01) in slurry DM content of 21% at Ryes and 12% at Corsock. Similarly, the highly significant increase in ash content at both sites reflects the breakdown of organic matter and removal of carbon as CH4 and CO2 against the background of minerals retention confirmed, for example, by the consistent levels or small and non-significant changes in total N, total phosphate and total potash. Although an 18% increase in NH4-N in the digestate was recorded at Ryes Farm, which is in line with the levels of performance reported in the literature, this increase failed to reach statistical significance (Table 6). A rather small increase in NH4-N (4.5%) was observed at Corsock, also non-significant. The variability in the NH4-N data can be seen with time in Figure 9 (a and b) and is reflected by the relatively high coefficient of variability (cv%) for both slurry input and digestate in Tables 7 and 8. Small increases in pH also occurred at both sites but again were inconsistent and failed to reach statistical significance. The cv is a useful indicator of data variability, but reflects not only the consistency or otherwise of the data, but the magnitude of the observations. Thus, although slurry NO3 levels were measured and as expected were consistently very low, the mean was associated with a high cv. In general, the cvs recorded here were in line with those often reported in the results of designed field experiments, e.g. <10% in a well controlled field experiment on grass. The cvs recorded of between c. 2% and 9%, for the major nutrients N, P2O5, K2O and pH, at both sites, indicate a satisfactory level of variability within the sampling and analysis in a study of this nature and bearing in mind the difficulties associated with representative sampling of livestock manures (Chambers, 2005). The higher cvs were associated with the more ephemeral parameters like NH4-N and DM content, which are subject to rapid change, particularly in dilute slurries as a result of solids settlement and, in the case of NH4-N, ionic buffering within the slurry; also, of course, as a result of microbial activity within the digestion process. As the results of Summers and Bousfield (1978) showed (Fig. 1), NH4-N content can change quite rapidly according to retention time, bacterial growth and C:N ratio of the substrate mix. These authors also identified what appeared to be an optimum range of slurry solids content of between 4 and 8% DM. Moreover, 2.0% DM was considered to be the minimum acceptable, below which ‘wash-out’ of digester bacteria was said to occur, and breakdown of the digestion process (Summers and Bousfield, 1978). The results of the current study, overall, show a consistency with experience elsewhere in terms of the anticipated changes in DM and, hence, ash content and in the conservation of total N and mineral content between digester input and output digestate. Although what appeared a substantial increase in NH4-N occurred at Ryes, the large variability in both NH4-N and in slurry pH reflected the state of flux within the digestion process. An indication of the variability of these parameters can be seen in Fig. 9 and, although it remains unclear why such variability has occurred, there is a suggestion that a wider and more consistent difference in NH4-N and pH between input and output coincides with the periods of higher digester temperature (also the converse) and, possibly, more efficient digester operation. The initial increase in digestate NH4-N at Corsock, appeared to be cut short by a decline in performance associated with the loss of temperature following the boiler breakdown during late February (Fig. 9b). It is clear that slurry NH4-N content and pH level are


26

subject to rapid transformation during digestion, as a result of microbial activity and ionic balance. Water soluble P content was very low in the slurry at both sites (1.5 – 4.2% of total P) with no difference between input and digestate at Ryes, but an apparent decrease in water soluble P in the digestate at Corsock. It must be noted, however, that the laboratory detection limit for this particular determination is at 0.0065%DM and differences between observations only 2-3x the detection limit must be regarded as unreliable and probably within assessment variability. The analysis of the slurry on the single occasion (March 8th) from the large storage tank, was closely similar, in both cases, to the overall mean of the digestate analyses, except for a reduced DM content at Corsock, in this case reflecting the likely extra dilution by rainfall within the store and the normal in-store settlement of solids. Unfortunately, the data on digester input and output volumes were not considered sufficiently reliable to allow calculation of a nutrient mass balance for the monitoring period.

5.5 Conclusions on chemical analysis of slurry and digestate There is a significant body of opinion that, among the claimed benefits of anaerobic digestion, are improvements in the effluent (digestate) quality, as a result of the digestion process. Based on both the findings of the technical review and the results of the farm studies, a number of observations and conclusions can be drawn concerning the impact of digestion on slurry nutrient content.

• As a result of the digestion process a number of changes in slurry analysis can be expected. These include a substantial reduction (up to 25%) in solids content and a consequential increase in ash content, due to the conservation of minerals against a background of reducing slurry carbon (and organic matter content).

• Increases in slurry pH (up to 0.5 pH units) and NH4-N content (up to 25%) may also occur, though these changes are less consistent than the reductions in solids content and BOD and may be transient, or dependent on digester operating conditions and the analysis of the feedstock slurries. Thus, although an increase in Nmin/Ntotal is expected, AD treated slurries should not be regarded as ‘mineral fertiliser solutions’, as has sometimes been reported.

• To address another occasional misconception about AD, although the reductions in solids content and BOD are significant, there is no reduction in either the volume or nutrient load of effluent for land application, since the total N and P2O5 content remains the same as in the digester input.

• Because of the increase in slurry NH4-N content, usually with an associated increase in pH and reduced solids content, there is a risk of significantly increased emissions of NH3 during post-digestion storage. Such increased emissions have been confirmed by Danish research but NH3 emissions and, also, odour nuisance have been shown to be effectively controlled by a range of store coverings.

• Although the increased pH and slurry NH4-N content might be expected to increase risk of NH3 emissions following surface application of slurry to the land,


27

the reduced solids content would be expected to improve surface infiltration of the slurry, which should help to conserve slurry N. Recent Danish research has shown reduced NH3 emissions from AD slurries when band applied via trailing hose application. Low emission application techniques are recommended for AD treated slurries.

• Increased mineral N content of slurries, even with reduced NH3 emissions, does not guarantee improved crop recovery of slurry N content and savings in fertiliser N, with maintained crop yields. The limited research covering agronomic assessments presented in the review have generated mixed results with Dutch experiments showing small and short term benefits only; Danish research has produced more encouraging results, though with not always consistent benefits.

• There is strong evidence, from the literature and from other recent research, to suggest that an increased availability factor for the phosphate content of AD slurries should be considered. Although the digester monitoring study failed to show an increase in the water soluble P content of the digestate, several other recent studies have indicated significant potential. Depending on the location of suitable P responsive field sites, this aspect could be included within the proposed field experiments (see section 6.2).

• Following on from these conclusions and, from element (3) of section 2.5 Project Objectives, a number of recommendations for further action and for further research are drawn together within Chapter 6 of this report.


28

6 Further investigation of the effectiveness of plant nutrients in digested slurry In addition to the significant changes in slurry analysis with digestion highlighted by the technical review and site monitoring data, benefits in terms of reduced risk of crop scorch would be anticipated as a result of reduced solids content and VFA concentrations in the digestate (Smith and Chambers, 1992; Smith et al., 1995). Thus, although the potential nutrient benefits of slurry digestate are unproven, these are of sufficient number and magnitude to justify further research (i) on emissions from digestate storage and following land application; and (ii) on agronomic impacts and nutrient recycling.

6.1 Proposals for field assessment of digestate nutrient value Based on the findings of the literature review and the monitoring of farm-based biogas plants, it is recommended that carefully designed, replicated field experiments are undertaken to assess the real potential of digested slurries under practical farm conditions. It is well known that it is often difficult to demonstrate measurable agronomic effects arising from differential treatment or application techniques for organic manures. This is because the potential magnitude of such treatment effects is likely to be relatively small, particularly when considered against the likely background of substantial in-field soil and sward variability. Thus, it is proposed that sites are very carefully selected for uniformity and a robust experimental design, with good replication and a large number of treatment degrees of freedom is adopted. It is proposed that the trial design should include the following elements and careful consideration of the following guidelines:

• An N response curve with a minimum of 7 N levels, including ‘nil’ N and ranging up to a maximum ensuring the optimum N is exceeded. The levels might be up to 330 kg/ha N, split between applications on 1st and 2nd cuts (up to say 180 kg/ha on 1st cut and 150 kg/ha on 2nd cut), though final decision will depend on site, soil type and grass growth potential.

• Slurry treatments will include digested and undigested slurry at a rate aimed at supplying a significant rate of mineral N (say minimum of 80 kg/ha NH4-N, which will mean a total N rate of c. 160 kg/ha for cattle slurry), so that the chances of measuring treatment differences are enhanced.

• The slurry treatment evaluations should include applications timed at both 1st cut and 2nd cut (i.e. application post 1st cut) on separate plots since the response to slurry N is likely to be different from these timings. Farmers have often been reluctant to apply slurry to silage aftermaths.

• In view of the concern about increased risk of NH3 emissions from AD slurry, both surface broadcast application and shallow injection should be included within the experimental design.

• Ammonia emissions following application of digested and untreated slurries should also be measured, in view of the importance of emissions arising from land applied manures. Moreover, whilst the agronomic effects of relatively


29

small differences in slurry N supply are often difficult to measure, significant differences in emissions from land application practices are more often observed (e.g. Smith et al., 2000).

• The residual N effects of the slurry are of interest because of the different analysis of digested and undigested slurries. It is therefore proposed that a 3rd cut should be taken to allow an indication of the shorter term residual effects following the slurry applications before 1st and 2nd cuts.

• To ensure that crop responses are to N only, plots should receive phosphate (as triple superphosphate) and potash (as sulphate of potash) according to soil analysis. Sulphate of potash to be used as potash source rather than muriate of potash in view of the likely response of grass to sulphur across much of Scotland. This will apply to all treatments except for the possible P evaluation treatments outlined below.

• If a low soil P status site can be located, a set of additional treatments should be included to evaluate the potential difference in P availability between AD and raw slurry. This would require only surface applied digested and raw slurry treatments (at the same rate as the main experimental treatments) for 1st cut (when greatest response to fresh P is likely).

• Treatments should also be included to evaluate the proposed reduced risk of scorch associated with AD slurry – this should include surface broadcast digested and undigested slurries only, applied before 1st cut in association with a high rate of fertiliser N, say 150 kg/ha, which would remove the grass response to slurry N, leaving the potential scorch effect to be examined against the equivalent fertiliser N treatment.

• The selected site(s) should be on a uniform soil type, on a short term ryegrass ley (i.e. excluding clover), with a known history, which should exclude intensive grazing or heavy slurry/manure applications and with none applied since the previous spring. An early site, with good grass growth potential to be preferred.

• An experiment of this type is unsuited to the application of treatments using field-scale equipment. For the slurry treatments, the ADAS purpose-built slurry plot applicator will allow careful control of the slurry treatments within a small-plot design (Basford et al, 1996).

• Consideration should be given to running an associated, pot-based experiment with matched key treatments but ensuring uniformity of soil and sward and hence the sensitivity of the measurements to treatment effects which may not otherwise be isolated within a field experiment.

Therefore, taking account of the above concerns, guidelines and caveats, the following treatments and experimental design, outline monitoring programme and reporting targets are proposed. The detail of any final experimental design may vary, however, according to site, slurry analysis, cropping and seasonal factors.


30

Proposed treatments Key: SUDS = surface applied undigested slurry SADS = surface applied digested slurry IUDS = shallow injected undigested slurry IADS = shallow injected digested slurry

Treatment 1st Cut 2nd Cut 1 Control Nil fert N Control Nil fert N 2 30 kg/ha fert N 25 kg/ha fert n 3 60 kg/ha fert N 50 kg/ha fert N 4 90 kg/ha fert N 75 kg/ha fert N 5 120 kg/ha fert N 100 kg/ha fert N 6 150 kg/ha fert N 125 kg/ha fert N 7 180 kg/ha fert N 150 kg/ha fert N 8 Control Nil fert N plus SUDS1 Nil fert N, nil slurry

9 Control Nil fert N plus SADS1 Nil fert N, nil slurry

10 60 kg/ha fert N, nil slurry Control Nil fert N plus SUDS1

11 60 kg/ha fert N, nil slurry Control Nil fert N plus SADS1

12 Control Nil fert N plus IUDS1 Nil fert N, nil slurry

13 Control Nil fert N plus IADS1 Nil fert N, nil slurry

14 60 kg/ha fert N, nil slurry Control Nil fert N plus IUDS1

15 60 kg/ha fert N, nil slurry Control Nil fert N plus IADS1

16 *150 kg/ha fert N plus SUDS1 Nil fert N, nil slurry

17 *150 kg/ha fert N plus SADS1 Nil fert N, nil slurry

18 Control Nil fert N, Nil P plus SUDS1 Control Nil fert N, Nil P, nil slurry 19 Control Nil fert N, Nil P plus SADS1 Control Nil fert N, Nil P, nil slurry 20 Control Nil fert N, Nil P, nil slurry Control Nil fert N, Nil P, nil slurry 1 Rate to supply c.160 kg/ha total N * Treatments 16 & 17 designed to evaluate the proposed reduced risk of scorch associated with AD slurry Design Fully randomised block, minimum 4 replicates. Monitoring Site details - Geology, soil series, soil texture (hand and lab particle size distribution for the site), field history (when grass established, previous cropping, stocking and fertiliser and manure application history). Soil analysis - Before establishment of treatments in the early spring, for each block analysis of top 0-15 cm for pH, available P, K & Mg, organic matter, total N, K2H2PO4 extractable sulphate, 0-30 cm, 30-60 cm & 60-90 cm increments for soil mineral nitrogen (NO3 plus NH4-N, DM). Slurry analysis - Dry matter, ash content, pH, total N, ammonium-N, nitrate-N, total and water soluble phosphate, total potassium and magnesium, total sulphur, total sulphate-S.


31

Ammonia emissions – Ammonia emissions over a period of 5-6 days following application of digested and untreated slurries, from both injection and surface application treatments, using small wind tunnels (Lockyer, 1984; Smith et al., 2000). Grass recording - Yield of fresh and DM; - Analysis for dry matter, total-N, total P, K, Mg & S; - Assessment of any visual herbage scorch effects following surface application before 1st cut, with comparison of digested and undigested slurry and against equivalent control (no slurry) plots. Weather records Particularly for the day and period after slurry application. Report Including statistical analysis of grass yields and analytical data. Assessment of slurry N use efficiency to be based on (1) fertiliser N equivalent of slurry by interpolation against the fitted fertiliser N response curves for 1st and 2nd cuts and (2) grass N recovery from slurry treatments relative to control (nil fertiliser N treatment).

6.2 Proposed modelling appraisal of nutrient fluxes following land application of digestate. Following on from the findings of this report it is proposed that a modelling assessment of nutrient fluxes following land application be carried out using the recently developed MANNER-PSM (MANure Nitrogen Evaluation Routine- Policy Support Model) software. A range of slurry management scenarios relevant to farmers in the south-west of Scotland should be evaluated. For example the impact of increasing slurry NH4-N content, but decreasing slurry solids content, which will tend to act in opposing directions on NH3 emissions following surface applications of slurry to land. Also the impact of digestate application to land across a range of timings and climatic conditions via different application techniques, on gaseous N emissions and nitrate leaching losses. This approach will allow the potentially beneficial contribution of digestate to be evaluated more widely, in the context of Scottish farming conditions and practical issues. The modelling work will also help to identify the likely optimum range of digestate quality, in terms of improved N utilisation efficiency and reduced environmental emissions. This work offers scope for a detailed analysis of possible impacts in the SW region of Scotland and depending upon the quality of information available, includes the possibility for spatially disaggregated data on emissions (nitrate leaching or gaseous emissions) presented as mapping outputs. This work may require the application of the ADAS developed “NARSES” mass flow model (Webb and Misselbrook, 2004), the “MANNER-PSM” decision support software and other catchment based modelling tools, according to the specific requirements and priorities that may be identified within such a study.

6.3 Other suggestions for action • On the basis of available evidence, it is recommended that farmers with AD

slurries should at least have an occasional laboratory analysis check on digestate quality; this should include DM, total N and NH4-N content, for which rapid field assessment techniques have also been successfully used.


32

• Carefully designed, replicated, field experiments should be undertaken to compare crop response to anaerobically digested and untreated slurries and the potential for fertiliser savings. The results should be used to inform and update current advice on N availability of livestock slurries and other effluent/waste streams, e.g. output of mixed digestate materials from CAD plants. Evidence from this work would contribute to the revision of published advice, including RB209; and also current manure nutrient DSSs such as MANNER and PLANET. Recommendations for such experiments are included in section 5.2.

• In view of the lack of well documented research data in the scientific literature on the impact of anaerobic digestion on slurry/effluent nutrient content, it is proposed that an edited version of the results should be prepared for publication in a suitable peer reviewed scientific journal.

• The main highlights and conclusions of this research should be promoted via appropriate industry/farm events and via the agricultural press.

Acknowledgements The authors of this report and the Scottish Executive gratefully acknowledge the co-operation and assistance of the host farmers in the monitoring and sampling of the farm digesters used in this study: Mr Wesley Millar, Ryes Farm; Mr Brian Smallwood, Corsock Farm; We also acknowledge the advice and technical assistance provided by Mr Jamie Gascoigne, field engineer, Greenfinch Ltd.


33

7 References Anon (1981). ETSU Report ETSU B 1052 (1987). Report on Anaerobic Digestion of Dairy Wastes to October 1981. Microbiology Department, Rowett Research Institute Engineering Division, North of Scotland College of Agriculture.

Anon. (2000). Fertiliser recommendations for Agricultural and Horticultural Crops. MAFF Reference Book RB209, 7th Edition, Dec 2000. The Stationery Office, Norwich (ISBN 0-11-243058-9).

Anon. (2006). Farm Scale Biogas and Composting to improve Bathing Waters – a report for the Scottish Executive by Enviros/Greenfinch. (Feb 2006). Available via the link: http://www.scotland.gov.uk/Resource/Doc/1057/0048383.pdf

Basford, W.D, Briggs, W, Jackson, D.R and Smith, K.A. (l996). Equipment for the improved precision application of animal waste slurries and liquid effluents to field plots. Proceedings of the International Conference on Mechanisation of Field Experiments (IAMFE ‘96), Versailles, France, June l996 pp 161-166.

Baldwin, D.J.(1993) Anaerobic digestion in the UK. A review of current practice ETSU B/FW/00239REP. DTI Renewable Energy Research Programme 1993.

Bauermeister, U., Spindler, H. and Wild, A. (2006). Treatment of Digested Substrates for Nitrogen Removal and Emission Decrease. In “Use of Bioenergy in the Baltic Sea Region” – Proceedings of the 2nd IBBC 2006, Stralsund, Germany. GNS – Gesellschaft fur Nachhaltige Stoffnutzung mbH Weinbergweg 23, D-06120 Halle/Saale, Germany.

Burton, C.H. and Turner, C. (2003). Manure management. Treatment strategies for sustainable agriculture. 2nd Edition. Silsoe Research Institute, Bedford, UK. 2003.

Chadwick, D.R., Brookman, S.K.E, Williams, J.R., Smith, K.A., Chambers, B.J., Scotford, I.M and Cumby,T.R. (2004). On-farm quick tests for manure. In “Advanced Silage Corn Management” Editors. S Bittman and C.G. Kowalenko. Pacific Field Corn Association, PO Box. 1000, Agassiz, BC V0M 1A0, Canada. pp.53-55.

Chambers, B.J. (2005) Developing improved sampling guidelines for liquid and solid manures. Defra Contract NT2009, Final Project Report, 2005. Available on Defra website: www.defra.gov.uk.

Cheng, J et al. (1999). What to Expect When You Clean Out a Plug Flow Digester. In: Proceedings of the North Carolina State University Animal Waste Management Symposium, Raleigh, North Carolina, January 27-28, 99. Evaluation System For Swine Waste Treatment and Energy Recovery.

Clarkson, C.R. (1990). Long Term Performance of Anaerobic Digester at Bethlehem Abbey, Portglenone, Northern Ireland. ADAS Research Report, 1990.

Hansen, M.H., Birkmose, T, Mortensen, B and Skaaning, K. (2004). Effects of Separation and Anaerobic Digestion of Slurry on Odour and Ammonia Emission during Subsequent Storage and Land Application. Vol I. Proceedings of the 11th International Conference of the FAO ESCORENA Network on Recycling of Agricultural, Municipal and Industrial Residues in Agriculture, RAMIRAN 2004, Murcia Spain, pp 265-268.

Hobson, P.N., Bousfield, S. and Summers, R. (1974). The anaerobic digestion of organic matter. Critical Reviews in Environmental Control, 4, pp. 131-191.

Holm-Nielsen, J.B., Halberg,N., Huntingford, S & Al Seadi, T. (1997). Joint Biogas Plant, Agricultural Advantages – Circulation of N, P and K. Report made for The Danish Energy Agency 2. Edition August 1997.

Hopfner-Sixt, K., Amon, T., Walla, Ch., Pötsch, E., Amon, B., Milovanovic, D., Mayr, H., Weichselbaum, W. (2007). Endbericht “Analyse und Optimierung neuer Biogasanlagen”.


34

Forschungsprojekt Nr. 811941/8539 SCK/SAI, Auftraggeber: Österreichische Forschungsförderungsgesellschaft mbH, Wien. In press.

Lockyer, D.R. (1984). A system for the measurement in the field of losses of ammonia, through volatilisation. Journal of the Science of Food and Agriculture, 35, pp 837-848.

Martin, J.H.Jr. (2004). A Comparison of Dairy Cattle Manure Management with and without Anaerobic Digestion and Biogas Utilization. Report to U.S. Environmental Protection Agency by Eastern Research Group Inc. 17 June 2004. EPA Contract No. 68-W7-0068, Task Order No. 400.

Martin, J.H.Jr. (2005). An Evaluation of a Mesophilic, Modified Plug Flow Anaerobic Digester for Dairy Cattle Manure. Submitted to Kurt Roos AgSTAR Program U.S. Environmental Protection Agency by Eastern Research Group Inc. 20 July 2005. EPA Contract No. GS 10F-0036K Work Assignment/Task Order No. 9.

Møller, H.B. (2001). Anaerobic digestion and separation of livestock slurry – Danish experiences. Report to MATRESA 2nd edition, Danish Institute of Agricultural Sciences, Dept Agricultural Engineering, Research Centre, Bygholm, Horsens, Denmark.

Møller, H.B. (2006). Kvaelstofomsaetning I biogasanlaeg (Turnover of nitrogen in AD plants). Forskning i Bioenergi. Nr 17. December 2006.

O’Sullivan, C.M. and Cumby, T.R. (2004). Analytical assessment of digestate samples from Holsworthy Biogas PLC. Silsoe Research Institute Contract Report CR/1534/04/3516.

Pedersen, C.A. (2002). Annual Report of the National Field Trials – Animal Manure. Extract of the fertiliser chapter, pp 25-29. Dansk Landbrugsraadgivning Landscentret. www.lr.dk

Perälä, P. and Regina, K. (2006). The effect of cow slurry fermentation and application technique on greenhouse gas and ammonia emissions from a grass field. In 12th RAMIRAN International Conference, “Technology for Recycling of Manure and Organic Residues in a Whole Farm Perspective” Vol. II. Paper P-306 pp245-247.

Regina, K. and Perälä, P. (2006). Ammonia and greenhouse gas emissions from pig slurry – the effect of slurry fermentation, separation of the fermentation product and application technique. In 12th RAMIRAN International Conference, “Technology for Recycling of Manure and Organic Residues in a Whole Farm Perspective” Vol. II. Paper P-305, pp241-243.

Schröder, J and Uenk, D. (2006). Cattle slurry digestion does not improve the long term nitrogen use efficiency of farms. In 12th RAMIRAN International Conference, “Technology for Recycling of Manure and Organic Residues in a Whole Farm Perspective”, RAMIRAN 2006, Aarhus, Denmark. Vol. II, pp 9-11.

Smith, K. A. & Chambers, B. J. (1992). Improved utilisation of slurry nitrogen for arable cropping. “In Nitrate and Farming Systems”. Aspects of Applied Biology 30, 1992 p127-134.

Smith, K. A. Jackson, D. R., Unwin, R. J., Bailey G & Hodgson I (1995). Negative effects of winter and spring applied cattle slurry on the yield of herbage at simulated early grazing and first-cut silage. Grass and Forage Science 50 pp 124-131.

Smith, K.A, Jackson, D.R., Misselbrook T.H., Pain, B.F. and Johnson, R.A. (2000) Reduction of ammonia emission by slurry application techniques. Journal of Agricultural Engineering Research, 2000, 77 (3), 277-287.

Smith, K.A., Cumby, T., Lapworth, J, Misselbrook, T.H, Nigro, E. and Williams, A.G. (2007). Natural crusting of slurry storage as an abatement measure for ammonia emissions on dairy farms. Biosystems Engineering, in press.

Sommer, S.G. and Birkmose, T. (2007). Valuable fertilizer from animal manure. Danish Crop Production Seminar 2007, Agromek 2007, Danish Agricultural Advisory Service.


35

Sood, D. (2006). Waste to Watts: Anaerobic Digestion of Livestock Manures (World Bank Funded Study) Workshop Monitoring nutrient related pollution reduction from diffuse agricultural sources and from agro-industrial point sources, Moldova September 2006.

Summers, R. and Bousfield, S. (1978). Anaerobic digestion of farm wastes – experimental experiences. (Rowett Research Institute, Bucksburn, Aberdeen). In “Anaerobic Digestion of Farm Wastes” Seminar, 18-20 October, 1978. pp. 30-42. ADAS report.

Van Velsen, A.F.M. (1979). Anaerobic digestion of wet piggery waste. In Hawkins,J.C. (ed.) Engineering problems with effluent from livestock. CEC, Luxembourg, pp. 476 – 489.

Vetter, H., Steffens, G. and Schröpel, R. (1987). The influence of different processing methods for slurry upon its fertiliser value on grassland. In HG van der Meer, et al (eds). Animal Manure on Grassland and Fodder Crops. Fertiliser or waste ? Martinus Nijhoff, Dordrecht, pp 73-86.

Webb, J. and Misselbrook, T.H. (2004). A mass-flow model of ammonia emissions from UK livestock production. Atmospheric Environment 38, pp 399-406.

Wright, P et al. (2004). Preliminary Comparison of Five Anaerobic Digestion Systems on Dairy Farms in New York State. Department of Biological and Environmental Engineering, Cornell University, Ithaca, NY 14853. 2004 ASAE/CSAE Annual International Meeting.


36

Annex A Anaerobic Digestion and Digestate Analysis Anaerobic digestion (AD) is a microbial process via which organic substances are broken down, the major ultimate products of which include carbon dioxide (CO2), methane (CH4) and water (H2O). Significant by-products include ammonia (NH4

+/NH3) and hydrogen sulphide (H2S). Although the degradation of organic compounds is a complex process, involving many groups of bacteria, the process comprises three main steps: (i) enzymatic hydrolysis of organic compounds with the formation of sugars and amino acids (slow); (ii) conversion of sugars and amino acids to volatile fatty acids (VFAs) (acetogenesis) (rapid); (iii) formation of CH4 + CO2 + H2O (methanogenesis). The resulting biogas is a mixture of the gases CH4 and CO2, with smaller concentrations of other gases, in particular NH3 and H2S, the latter resulting from the breakdown of proteins. Anaerobic digestion is a dynamic process in which the analysis of the digestate will depend on the dominant phase at the time of sampling and a number of buffering reactions in solution. The equilibrium of CO2 and bicarbonate (HCO3

-) with ammonium cations (NH4+), exerts substantial buffering on

digestate pH, the breakdown of organic acids generating CO2 and, hence, carbonic acid in solution: CO2 + H2O → H2CO3 → HCO3

- + H+ VFAs decrease the buffering capacity of the bicarbonate ions: RCOO-H + NH4HCO3 → RCOO-NH4 + H2CO3 The formation of NH3 will increase bicarbonate in balancing the generation of CO2: CO2 + H2O + NH3 → NH+

4 + HCO-3

The higher the bicarbonate concentration, the greater the buffering in solution and resistance to changes in pH. The optimum pH varies according to the stage in the degradation process. Overall, the breakdown of organic substances in AD will result in a reduction in organic matter (OM) content and, hence, a reduction in BOD, COD and in solids (DM). Minerals are retained and, thus, ash content (expressed on DM basis) will increase following digestion. Generally, the surplus N content in manures leads to an increase in NH4-N content in digestate, but C:N ratio and bacterial growth (and N retention in bacterial protein) are important. Digestate pH impacts on the balance between NH4

+ in solution and the generation of NH3 and, similarly, between H2S and S2-, sulphide in solution.


37

Annex B Glossary of Terms To facilitate understanding and reduce the risk of ambiguity or confusion over terms which have been used in this report, a number of technical terms and acronyms have been listed below with simple definitions or explanations.

ANAEROBIC Containing no free oxygen (or not requiring free oxygen such as

ANAEROBIC BACTERIA) or chemically bound oxygen such as nitrates (NO3).

AD Anaerobic digestion

ASH Product remaining after incineration in laboratory combustion.

BIOLOGICAL OXYGEN DEMAND (BOD)

Together with the COD, BOD is the measure of the pollution potential in water bodies and organic wastes. A laboratory test is used to measure the amount of dissolved oxygen consumed by chemical and biological action when a sample is incubated at 20 C°

for a given number of days (five for BOD5).

CAD Centralised anaerobic digester: plant designed to receive organic substrates from several sources (e.g. SLURRIES from neighbouring farms, wastes from abattoirs, food processing factories etc.), so offering economies of scale in investment and operating costs.

CH4 Methane; a greenhouse gas produced during anaerobic fermentation of organic matter, especially from enteric fermentation in ruminants and storage of liquid manure. A constituent of biogas

CHEMICAL OXYGEN DEMAND (COD)

A measure of the amount of oxygen consumed in the microbial oxidation of decomposable and inert organic matter and the oxidation of reduced substances in water. The COD is always higher than the BOD, but measurements can be made in a few hours while BOD measurements take five days.

C:N RATIO The amount of total carbon divided by the amount of total nitrogen contained in livestock manures. Manures with a high C:N RATIO such as FARMYARD MANURE usually take longer to break down, or mineralise, in the soil than those such as slurry with a lower C:N RATIO.

DRY MATTER (DM) The residue remaining following heating under standard conditions (usually around 105 °C to constant weight) to drive off water. Often expressed as a percentage of the weight of original material.

DSS Decision Support Software: computer software or programme developed to carry out difficult or complex calculations rapidly, as an aid to decision making.

FIOs Faecal indicator organisms; bacteria normally present in the lower digestive tract (e.g. Eschericia coli) and, hence, in the faeces; when found in water, are indicative of faecal contamination.


38

HCO3- Bicarbonate ion; from the dissolution of CO2 in solution.

K Chemical symbol for potassium.

ORGANIC MATTER (OM) Residues derived from plants, animals and micro-organisms in various stages of decomposition.

pH A measure of the hydrogen ion concentration of a solution and an indication of its` acidity or alkalinity. Expressed on a scale from 0 to 14, 7 is neutral, higher values more alkaline, lower values more acid.

MANNER MANure Nutrient Evaluation Routine: DSS tool designed to assess the fate of manure nutrients following application to land.

Mg Chemical symbol for magnesium.

Ntotal Total nitrogen content; also Total Kjeldahl Nitrogen.

Nmin/ Ntotal The proportion of mineral nitrogen (usually ammonium-N) content of total nitrogen; provides an indication of the readily available N content of manures.

Na Chemical symbol for sodium

NH3 Ammonia (gas)

NH4+ Ammonium; ionic form following dissolution of ammonia gas in

aqueous solution.

NO3- Nitrate; oxidised form of nitrogen in solution, readily available for

uptake by plant roots, but vulnerable to loss via leaching in drainage water.

N2O Nitrous oxide; powerful greenhouse gas.

P Chemical symbol for phosphorus.

PLANET Nutrient management software to facilitate the planning of fertiliser and manure nutrient inputs

RB209 Reference Book 209; Fertiliser recommendations for Agricultural and Horticultural Crops. MAFF Reference Book RB209, 7th Edition, Dec 2000. Published by The Stationery Office, Norwich.

RCOO-H Generic formula for short chain fatty acid, where the radical ‘R’ may represent an alkyl radical containing from one to four carbon atoms, e.g. methyl (CH3), ethyl (C2H5) etc. (see also VFA below)

RT Retention time: the time for which a substrate e.g. slurry is retained in a treatment vessel or REACTOR.

S Chemical symbol for sulphur.

TOTAL KJELDAHL NITROGEN (TKN)

Total amount of organic and reduced forms of nitrogen contained in e.g. LIVESTOCK MANURES, excluding nitrate.

TOTAL AMMONIACAL NITROGEN (TAN)

The total amount of ammonium and AMMONIA nitrogen contained in slurries and manures.

VOLATILE FATTY ACID (VFA) Short chain fatty acids containing two to five carbon atoms that are produced as end products of microbial FERMENTATION in the digestive tract or in anaerobic digestion.


39

Annex C Nutrient content of livestock slurries before and after anaerobic digestion. Table C1: Comparison of analysis results for undigested and digested dairy cattle/pig slurry mix1, Suffolk, 1979-80 (Nielsen, 1980).

Total N NH4-N NH4-N P2O5 K2O DM pH COD BOD kg/m3 kg/m3 % of total kg/m3 kg/m3 % mg/I %

Feedstock1 3.0 (8) 2.0 (8) 66.7 1.4 (7) 3.5 (7) 4.7 (14) 7.30 (17) 49,800 (17) 4.7 (14) Digestate 3.4 (10) 2.3 10) 67.6 1.6 (8) 3.2 (9) 4.2 (16) 7.75 (20) 30,400 (19) 4.2 (16) Change %2 +13 +15 - +18 -7 -10 +0.452 -38 -10 1 The results relate to digestion of slurry from 200 dairy cows and 3,000 fattening pigs. ( ) Figure in brackets = number of samples from which mean derived 2 All changes expressed as % except for pH units. Source: Nielsen, V.C. (1980) Internal ADAS R&D Report, ADAS Farm Waste Unit. Table C2: Comparison of analysis results for undigested and digested pig slurry1, Yorkshire, 1981 (Friman, 1981).


Feedstock1 (5 samples) 7.6 3.5 46 0.65 1.3 2.33 7.6 (13) 36,200 (3) nd Digestate (3 samples) nr 4.9 - 0.61 nr 1.84 8.1 (7) 21,400 (2) nd Change % - +40 -6.2 -21 +0.52 -40.9 1 Slurry from 900 sows with progeny to bacon weight; includes cleaning water. Slurry separated and liquid fraction digested. ( ) Figure in brackets = number of samples from which mean derived; nr – not reported (unreliable data); nd – not determined. 2 All changes expressed as % except for pH units. Source: Friman, R. (1981) Internal ADAS R&D Report, ADAS Farm Waste Unit.


40

Table C3: Comparison of analysis results for undigested and digested dairy cattle slurry1, Kent, January – February 1982 (Friman, 1982).


Feedstock1 (16 samples) 4.45 1.69 38.0 1.57 5.86 10.5 8.1 72,000 8,630 Digestate (28 samples) 3.94 1.48 37.6 1.37 5.07 7.4 8.01 48,000 5,138 Change %2 -11.5 -12.4 -12.7 -13.5 -29.5 -0.092 -33.3 -67.5 1 Slurry from 900 sows with progeny to bacon weight; includes cleaning water. Slurry separated and liquid fraction digested. ( ) Figure in brackets = number of samples from which mean derived; nr – not reported (unreliable data); nd – not determined. 2 All changes expressed as % except for pH units. Source: Friman, R. (1982) Internal ADAS R&D Report, ADAS Farm Waste Unit. Table C4: Comparison of analysis results for undigested and digested beef cattle slurry1, Northern Ireland, 1989-90 (Clarkson, 1990).


Feedstock1 4.9 2.3 46.9 nd nd 8.8 7.2-7.5 82,000 12,600 Digestate 4.2 2.5 59.5 nd nd 6.5 7.7-7.8 62,000 4,100 Change %2 -14.3 +8.7 - - - -26.1 +0.42 -24.4 -67.5 1 The results relate to digestion of slurry from beef cattle housed on slats. nd – not determined Source: Clarkson, C.R. (1990) Long Term Performance of Anaerobic Digester at Bethlehem Abbey, Portglenone, Northern Ireland. ADAS Research Report, 1990. Table C5: Comparison of analysis results for undigested and digested dairy cattle slurry1, Scotland, 1981 (Anon, 1981).


Feedstock1 - 1.04 - - - 7.3 - 80,280 18,470

Digestate - 1.15 - - - 5.9 - 66,490 3,840

Change % +10.2 -19.2 -17.2 -79.2 1 Dairy cattle slurry. Source: Anon (1981). ETSU Report ETSU B 1052 (1987). Report on Anaerobic Digestion of Dairy Wastes to October 1981. Microbiology Department, Rowett Research Institute Engineering Division, North of Scotland College of Agriculture.


41

Table C6: Comparison of analysis results for undigested livestock slurries and mixed digestate, Austria (Hopfner-Sixt, et al., 2007).

Total N NH4-N NH4-N P2O5 K2O DM pH kg/m3 kg/m3 % of total kg/m3 kg/m3 %

Cattle slurry (47) 2.63 1.27 48.3 0.98 3.21 4.18 7.96 Pig slurry (16) 3.72 2.20 59.1 1.37 2.28 3.03 7.95 Mixed slurry (11) 3.19 1.83 57.4 1.19 2.88 3.34 8.14 Digestate1 (6) 4.12 2.24 54.4 1.90 3.31 4.87 7.79 1 Digester feedstock average 62% cow/pig slurry. Source: Hopfner-Sixt, et al., (2007). Data from Biogas Forum Austria, University of Natural Resources and Applied Life Sciences, Vienna. Table C7: Comparison of analysis results before and after anaerobic digestion, from studies in Germany1 (Bauermeister et al., 2006).

Total N NH4-N NH4-N P2O5 K2O DM pH kg/m3 kg/m3 % of total kg/m3 kg/m3 %

Pre-digestion 4.85 2.18 44.9 11.3 6.71 Post digestion 4.24 2.69 63.5 5.58 7.90 Change %2 -12.6 +23.4 - - -50.6 +1.192

1 Data represent average from 43 biogas plants in Thuringia. 2 All changes expressed as % except for pH units. Source: Bauermeister et al., (2006). (Reference for this data given as Reinhold, G. , Eigenschaften und Einsatz der Gärreste in Pflanzenproduction; Vortrag zum ZONARO-Fachgespräch am 26.10.05 an der LLG in Bernburg).


42

Table C8: Comparison of nitrogen in different types of biomass before and after anaerobic digestion, from studies in Denmark (Moller, 2006).

Feedstock Total N kg/m3 % NH4-N kg/m3 % NH4-N/tot N (%)

Pre- dig Post-dig Change Pre- dig Post-dig Change Pre- dig Post-dig Cattle slurry 3.0 2.8 -7 1.6 2.1 +32 53 75

Pig slurry (thermophilic AD) 4.4 4.4 0 3.6 4.1 +14 82 93

Pig slurry (thermophilic AD) 5.6 5.6 0 4.4 5.0 +13 79 89

Pig slurry (mesophilic AD) 4.0 4.0 0 2.2 3.1 +42 55 78

Solids from decanter centrifuge 12.0 12.0 0 5.0 7.3 +45 42 60

Solids from separation1 7.6 7.8 +3 4.0 6.3 +52 53 81 1 60% solid fraction from separation (Kemira) and 40% untreated slurry. Source: Moller, H.B. (2006). Kvaelstofomsaetning I biogasanlaeg (Turnover of nitrogen in AD plants). Forskning i Bioenergi. Nr 17. December 2006. Table C9: Nutrient concentrations in animal manure before land application (sampled by the Danish Advisory Service)1.

N Total

NH4-N P2O5 K2O DM NH4-N

kg/m3 kg/m3 kg/m3 kg/m3 % % of total Slurry – after anaerobic digestion (mean 44 samples)

4.75 3.67 2.11 2.75 4.49 82

Slurry (means 228-238 samples depending on determination)

3.56 2.49 1.67 3.10 5.26 70

Acidified slurry (mean 10 samples)

4.33 3.07 1.90 2.99 5.15 71

Cattle slurry (mean 104 samples)

3.62 2.07 1.69 3.67 7.36 57

Pig slurry – finishing pigs (mean 24 samples)

4.06 3.19 1.99 2.96 4.26 79

Pig slurry (mean 107 samples)

4.27 3.37 2.29 2.89 4.55 79

1 Ref: Data provided by Torkild Birkmose, Danish Advisory Service – personal communication.


43

Table C10: Comparison of digested (digestate) and undigested (feedstock) cow slurry analysis results in New York State, USA (mean results for semi-monthly sampling late May 2001 to early June 2002)1.

Total N NH4-N NH4-N P2O5 total P2O5 ortho P2O5 Ortho pH COD DM kg/m3 kg/m3 % of total kg/m3 kg/m3 % of total mg/I %

Feedstock 4.63 2.16 46.7 1.86 1.05 56.5 7.4 153,496 11.32 Digestate 5.11 2.88 56.4 1.92 1.29 67.2 7.9 89,144 8.47 Change %2 +10.4 +33.3 +3.2 +22.3 +0.52 -41.9 -25.2 1 source: Martin (2004). 2 All changes expressed as % except for pH units.

Table C11: Comparison of digested (digestate) and undigested (feedstock) cow slurry analysis results in Wisconsin, USA (mean results for semi-monthly sampling late January to December 2004)1.


Feedstock 3.48 1.70 48.9 1.79 0.017 0.9 7.6 69,923 8.81

Digestate 3.25 2.12 65.2 1.64 0.011 0.7 8.2 43,000 5.69

Change %2 -6.6 +24.9 -8.4 -35.3 +0.62 -38.5 -35.4 1 source: Martin (2005). 2 All changes expressed as % except for pH units.

Table C12: Comparison of digested (digestate) and undigested waste water from pig housing (farrowing and gestation) analysis results from sampling in N Carolina, USA (mean results 1998)1.


Farrowing house2 1.31 0.79 60.3 0.85 0.43 50.6 6.88 14,847 0.94 Gestation2 1.42 0.85 59.9 1.09 0.51 46.8 7.21 15,621 1.10 Digestate 0.92 0.78 84.8 0.24 0.20 83.3 7.48 897 0.24 Change %3 -32.9 -5.3 -75.6 -57.9 +0.44 -94.1 -76.7 1 source: Cheng et al. (1999). 2 The authors report that reduction in N and P2O5 in digestate compared to farrowing and gestation wastes likely to be due to precipitation in the covered digestion lagoon. 3 Change based on the waste water digested being composed of 43.8% farrowing and 56.2% gestation. 4 All changes expressed as % except for pH units.


44

Table C13: Comparison of digested (digestate) and undigested (feedstock) cow slurry analysis results for five digester systems in New York State, USA (mean monthly sampling 2002-03)1.

Total N NH4-N NH4-N P2O5 total P2O5 ortho P2O5 Ortho pH COD DM Digester kg/m3 kg/m3 % of total kg/m3 kg/m3 % of total mg/I %

Site 1 (63 samples)

Feedstock 4.96 1.92 38.7 1.92 1.08 56.3 7.21 134,695 11.42

Digestate 5.29 2.62 49.5 1.96 1.26 64.3 7.92 94,148 8.30

Change %3 +6.7 +36.5 +2.1 +16.7 +0.71 -30.3 -27.3

Site 2 (16 samples)

Feedstock 3.43 1.73 50.4 1.17 0.59 50.4 7.45 121,987 9.01

Digestate 3.46 2.21 63.9 1.17 0.67 57.3 7.63 110,658 6.75

Change %3 +0.9 +27.7 0 +13.6 +0.18 -9.3 -25.1

Site 3 (12 samples)

Feedstock 3.89 2.22 57.1 1.45 0.88 60.7 7.45 109,723 9.58

Digestate 3.71 2.47 66.6 1.36 0.95 69.9 7.75 42,416 3.80

Change %3 -4.6 +11.3 -6.2 +8.0 +0.3 -61.3 -60.3

Site 4 (12 samples)

Feedstock 3.38 1.35 39.9 1.54 0.92 59.7 5.64 137,547 12.46

Food waste 2.59 0.72 27.8 1.19 0.59 46.6 4.15 271,945 17.60

Digestate 3.27 1.47 45.0 1.34 0.80 59.7 7.61 63,996 5.50

Change %2 - - - - - - -

Site 5 (9 samples)

Feedstock 4.01 1.67 41.6 1.01 0.41 40.2 7.45 72,100 8.99

Digestate 1 4.15 2.30 55.4 1.07 0.66 61.7 7.74 65,627 7.99

Digestate 2 3.79 2.19 57.8 1.12 0.60 53.6 7.67 61,823 7.52

Change 1 %3 +3.5 +37.7 +5.9 +61.1 +0.29 -9.0 -11.1

Change 2 %3 -5.5 +31.1 +10.9 +46.3 +0.22 -14.3 -16.4 1 source: Wright, et al. (2004). 2 Change not calculated since proportion of cattle slurry/food waste not known. 3 All changes expressed as % except for pH units.

Date post:	27-Jan-2017
Category:	Documents
Upload:	trinhtruc
View:	218 times
Download:	0 times

Nutrient Value of Digestate from Farm-Based Biogas Plants in ...

Documents