Professional Documents
Culture Documents
Plants in Scotland.
Report for Scottish Executive Environment and
Rural Affairs Department - ADA/009/06.
ADAS UK Ltd
Woodthorne
Wergs Road
Wolverhampton
WV6 8TQ
July 2007
Contents
1
2
2.1
2.2
2.3
2.4
2.5
References...................................................................................................................... 33
Annex A
Annex B
Annex C
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:
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 postdigestion 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.
Introduction
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
Improved uptake of nutrients by growing crops and reduced losses to the water and
air environments;
Improved crop responses, increased yield and savings in inorganic fertiliser nutrient
inputs;
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;
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.
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).
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.
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; Mller, 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
9
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
A1
A2
A3
A4
A5
A8
Suffolk
Yorks
Kent
N Ireland
Scotland
Denmark
Denmark
Denmark
Denmark
Denmark
Denmark
USA (NY)
USA (Wisc.)
USA (NY)
USA (NY)
USA (NY)
USA (NY)
USA (NY)
A10
A11
A13
Mean
Median
Observations
1
DM
N-total
NH4-N
P2O5
pH
COD
-10.0
-21.0
-29.5
-26.1
-19.2
-25.2
-35.4
-27.3
-25.1
-60.3
-11.1
-16.4
13.0
-11.5
-14.3
-7.0
0
0
0
0
3.0
10.4
-6.6
6.7
0.9
-4.6
3.5
-5.5
15.0
40.0
-12.4
8.7
10.2
32.0
14.0
13.0
42.0
45.0
52.0
33.3
24.9
36.5
27.7
11.3
37.7
31.1
18.0
-6.2
-12.7
3.2
-8.4
2.1
0
-6.2
5.9
10.9
0.45
0.5
-0.09
0.4
0.5
0.6
0.7
0.18
0.3
0.29
0.22
-38
-41
-33.3
-24.4
-17.2
-41.9
-38.5
-30.3
-9.3
-61.3
-9.0
-14.3
-25.6
-25.15
12
-0.8
0
16
25.7
29.4
18
0.7
1.05
10
0.4
0.4
11
-29.9
-31.8
12
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,
10
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.
DM input
5
DM output
2.5
NH4-N input
Slurry DM content %
1.5
0.5
Ammonium-N g/l
NH4-N output
0
3 days
5 days
7 days
10 days
(+21%)
(+2%)
(-1.7%)
(-16%)
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
11
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
3
Holsworthy1
(England)
Ribe average
1992-96 (Denmark)
Ribe2 1992
(Denmark)
Cattle slurry
(Denmark)
Pig slurry (Denmark)
NH4-N
3
NH4-N/N
P2O5
K2O
3
DM
3
kg/m
kg/m
% total
kg/m
kg/m
6.6
5.0
75.8
3.3
4.5
5.8
4.9
3.2
65.3
2.4
4.2
5.8
4.6
3.1
67.4
2.1
4.2
6.4
4.7
2.7
57.4
1.4
5.3
8.5
5.3
3.7
69.8
3.4
2.8
6.0
Feedstocks 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 - OSullivan, 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.
12
Table 3: Analysis of pig slurry used in storage and application method experiments
Year
Slurry type
2002
2002
2002
Undigested
Digested
Digested
separated
Undigested
Undigested
separated
Digested
Digested
separated
2003
2003
2003
2003
Dry matter
%
3.4
3.2
2.1
pH
NH4-N
7.4
8.1
8.3
Total N
kg/m3
4.3
5.2
4.8
3.1
3.7
3.6
NH4-N
% of total
72
71
75
3.3
1.5
7.2
8.6
3.7
4.9
2.4
3.9
65
80
2.8
2.2
8.1
8.2
4.3
4.2
2.9
3.4
67
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
digested
slurry
0.9
Separated
undigested
slurry
-
09/0101/05/2002
20/0306/05/2003
Covered*
0.8
Uncovered
2.5
4.4
6.1
4.4
-0.1
* Slurry stores for each slurry type covered with a 15 cm layer of Leca (lightweight-expanded
clay aggregates)
13
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)
2002
2003
27
46
22
34
23
Separated
digested slurry
(liquid fraction)
26
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 Perl, 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 (Perl 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.
14
90
Digested slurry
Pig slurry
Cattle slurry
Slurry N utilization %
80
70
60
50
40
30
20
10
0
trailing hose
injection
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).
Schrder 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
15
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
NFRVs for both types of slurry (Fig. 3).
70
60
Untreated
50
NFRV %
Digested
40
30
20
10
0
Year 1
Year 2
Year 3
Year 4
16
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 months 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.
17
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
18
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.
(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.
19
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.
temp
50
43
40
42
30
41
20
40
10
39
38
12-Feb
(a)
15-Feb
19-Feb
22-Feb
26-Feb
01-Mar
05-Mar
08-Mar
12-Mar
Slurry volume m3
40
40
30
35
20
30
10
25
Dig temp
(b)
Temperature deg C
44
20
12-Feb
15-Feb
19-Feb
22-Feb
26-Feb
01-Mar
05-Mar
08-Mar
Temperature deg C
Slurry volume m3
60
12-Mar
20
Input
Output
Difference
%1
P value2
Dry matter %
pH
Ash %DM
Total N %
NH4-N %
NO3-N %
Total P2O5 %
Water sol P %DM
Total K2O %
8.1
7.68
34.84
0.29
0.11
0.03
0.111
0.012
0.43
6.23
7.84
40.29
0.29
0.13
0.03
0.108
0.012
0.42
-21.2
0.16
16.7
-0.87
17.7
-2.49
-1.9
0.005
0.264
0.017
0.608
0.092
0.347
0.299
0.383
Dry matter %
pH
Ash %DM
Total N %
NH4-N %
NO3-N %
Total P2O5 %
Water sol P %DM
Total K2O %
7.68
7.35
16.32
0.29
0.11
0.02
0.10
0.025
0.23
6.76
7.58
19.41
0.30
0.12
0.03
0.11
0.017
0.23
-11.8
0.22
19.1
4.17
4.5
8.2
-32.0
3.4
0.002
0.008
<0.001
0.107
0.296
0.065
<0.001
0.262
Corsock Farm
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.
21
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
Mean
Digester input
St.dev.
Store1
cv %
Dry matter %
9.1
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
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
Dry matter %
5.8
5.9
6.1
6.3
6.2
6.1
7.7
6.23
0.570
9.15
10.1
pH
7.77
8.07
7.86
7.87
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
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
Digester output
22
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
Mean
Digester input
St.dev.
Store1
cv %
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
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
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
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
Digester output
23
0.4
0.35
0.35
8
7
0.3
0.1
0.05
2
1
0
12-Feb 15-Feb 19-Feb 22-Feb 26-Feb 01-Mar 05-Mar 08-Mar 12-Mar
44
50
43
40
42
30
41
20
40
10
39
38
temp
40
Slurry volume
60
Temperature deg C
30
35
20
30
10
25
Dig temp
40
Temperature
0.05
5
0.2
0.15
0.1
pH-in
pH
Slurry N content %
pH
pH-out
0.25
0.15
Total N-out
NH4-N-out
pH-in
pH-out
0.2
Total N-in
NH4-N-in
Total N-out
0.25
Slurry volume m3
0.4
NH4-N-in
NH4-N-out
0.3
Slurry N content %
Total N-in
20
12-Feb 15-Feb 19-Feb 22-Feb 26-Feb 01-Mar 05-Mar 08-Mar 12-Mar
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.
24
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 NH4N, 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
25
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.
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.
27
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:
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.
28
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).
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.
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.
29
Proposed treatments
Key:
SUDS = surface applied undigested slurry
SADS = surface applied digested slurry
IUDS = shallow injected undigested slurry
IADS = shallow injected digested slurry
Treatment
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
2nd Cut
1st Cut
Control Nil fert N
30 kg/ha fert N
60 kg/ha fert N
90 kg/ha fert N
120 kg/ha fert N
150 kg/ha fert N
180 kg/ha fert N
Control Nil fert N plus SUDS1
Control Nil fert N plus SADS1
60 kg/ha fert N, nil slurry
60 kg/ha fert N, nil slurry
Control Nil fert N plus IUDS1
Control Nil fert N plus IADS1
60 kg/ha fert N, nil slurry
60 kg/ha fert N, nil slurry
*150 kg/ha fert N plus SUDS1
*150 kg/ha fert N plus SADS1
Control Nil fert N, Nil P plus SUDS1
Control Nil fert N, Nil P plus SADS1
Control Nil fert N, Nil P, nil 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.
30
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 cooperation 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.
32
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 0-11243058-9).
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., Ptsch, E., Amon, B., Milovanovic, D., Mayr, H.,
Weichselbaum, W. (2007). Endbericht Analyse und Optimierung neuer Biogasanlagen.
33
34
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 Schrpel, 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.
35
Annex A
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.
36
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
AD
Anaerobic digestion
ASH
BIOLOGICAL OXYGEN
DEMAND (BOD)
CAD
CH4
CHEMICAL OXYGEN
DEMAND (COD)
C:N RATIO
DSS
FIOs
37
HCO3-
pH
MANNER
Mg
Ntotal
Nmin/ Ntotal
Na
NH3
Ammonia (gas)
NH4+
NO3-
N2O
PLANET
RB209
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
TOTAL KJELDAHL
NITROGEN (TKN)
TOTAL AMMONIACAL
NITROGEN (TAN)
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.
38
Annex C
Table C1: Comparison of analysis results for undigested and digested dairy cattle/pig slurry mix1, Suffolk, 1979-80
(Nielsen, 1980).
Feedstock1
Digestate
Change %2
Total N
kg/m3
3.0 (8)
3.4 (10)
+13
NH4-N
kg/m3
2.0 (8)
2.3 10)
+15
NH4-N
% of total
66.7
67.6
-
P2O5
kg/m3
1.4 (7)
1.6 (8)
+18
K2O
kg/m3
3.5 (7)
3.2 (9)
-7
DM
%
4.7 (14)
4.2 (16)
-10
pH
7.30 (17)
7.75 (20)
+0.452
COD
mg/I
49,800 (17)
30,400 (19)
-38
BOD
%
4.7 (14)
4.2 (16)
-10
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)
Digestate (3 samples)
Change %
Total N
kg/m3
7.6
nr
-
NH4-N
kg/m3
3.5
4.9
+40
NH4-N
% of total
46
-
P2O5
kg/m3
0.65
0.61
-6.2
K2O
kg/m3
1.3
nr
DM
%
2.33
1.84
-21
pH
7.6 (13)
8.1 (7)
+0.52
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.
39
COD
mg/I
36,200 (3)
21,400 (2)
-40.9
BOD
%
nd
nd
Table C3: Comparison of analysis results for undigested and digested dairy cattle slurry1, Kent, January February 1982
(Friman, 1982).
Total N
kg/m3
4.45
3.94
-11.5
NH4-N
kg/m3
1.69
1.48
-12.4
NH4-N
% of total
38.0
37.6
P2O5
kg/m3
1.57
1.37
-12.7
K2O
kg/m3
5.86
5.07
-13.5
DM
%
10.5
7.4
-29.5
pH
COD
mg/I
72,000
48,000
-33.3
8.1
8.01
-0.092
BOD
%
8,630
5,138
-67.5
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).
Total N
NH4-N
NH4-N
P2O5
K2O
DM
pH
COD
BOD
3
3
3
3
kg/m
kg/m
% of total
kg/m
kg/m
%
mg/I
%
1
Feedstock
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
2
Change %2
-14.3
+8.7
-26.1
+0.4
-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
Total N
kg/m3
-
NH4-N
kg/m3
1.04
NH4-N
% of total
-
P2O5
kg/m3
-
K2O
kg/m3
-
DM
%
7.3
Digestate
1.15
5.9
Change %
+10.2
-19.2
pH
-
COD
mg/I
80,280
BOD
%
18,470
66,490
3,840
-17.2
-79.2
40
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
3
3
3
3
kg/m
kg/m
% of total
kg/m
kg/m
%
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
1
Digestate (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).
Pre-digestion
Post digestion
Change %2
Total N
kg/m3
4.85
4.24
-12.6
NH4-N
kg/m3
2.18
2.69
+23.4
NH4-N
% of total
44.9
63.5
P2O5
kg/m3
K2O
kg/m3
DM
%
11.3
5.58
-50.6
pH
6.71
7.90
+1.192
41
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
(%)
Cattle slurry
Pre- dig
3.0
Post-dig
2.8
Change
-7
Pre- dig
1.6
Post-dig
2.1
Change
+32
Pre- dig
53
Post-dig
75
4.4
4.4
3.6
4.1
+14
82
93
5.6
5.6
4.4
5.0
+13
79
89
4.0
4.0
2.2
3.1
+42
55
78
12.0
12.0
5.0
7.3
+45
42
60
7.6
7.8
+3
4.0
6.3
+52
53
81
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
4.75
kg/m3
3.67
kg/m3
2.11
kg/m3
2.75
%
4.49
% of total
82
3.56
2.49
1.67
3.10
5.26
70
4.33
3.07
1.90
2.99
5.15
71
3.62
2.07
1.69
3.67
7.36
57
4.06
3.19
1.99
2.96
4.26
79
4.27
3.37
2.29
2.89
4.55
79
Ref: Data provided by Torkild Birkmose, Danish Advisory Service personal communication.
42
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.
Feedstock
Digestate
Change %2
1
2
Total N
kg/m3
4.63
5.11
+10.4
NH4-N
kg/m3
2.16
2.88
+33.3
NH4-N
% of total
46.7
56.4
P2O5 total
kg/m3
1.86
1.92
+3.2
P2O5 ortho
kg/m3
1.05
1.29
+22.3
P2O5 Ortho
% of total
56.5
67.2
pH
7.4
7.9
+0.52
COD
mg/I
153,496
89,144
-41.9
DM
%
11.32
8.47
-25.2
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
Digestate
Change %2
1
2
Total N
kg/m3
3.48
3.25
NH4-N
kg/m3
1.70
2.12
-6.6
+24.9
NH4-N
% of total
48.9
65.2
P2O5 total
kg/m3
1.79
1.64
P2O5 ortho
kg/m3
0.017
0.011
-8.4
-35.3
P2O5 Ortho
% of total
0.9
0.7
pH
7.6
8.2
COD
mg/I
69,923
43,000
DM
%
8.81
5.69
+0.62
-38.5
-35.4
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.
Total N
NH4-N
NH4-N
P2O5 total
P2O5 ortho
P2O5 Ortho
pH
COD
DM
3
3
3
3
kg/m
kg/m
% of total
kg/m
kg/m
% of total
mg/I
%
2
Farrowing house
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.
43
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.
Digester
Total N
kg/m3
NH4-N
kg/m3
NH4-N
% of total
P2O5 total
kg/m3
44
P2O5 ortho
kg/m3
P2O5 Ortho
% of total
pH
COD
mg/I
DM
%
1.08
1.26
+16.7
56.3
64.3
7.21
7.92
+0.71
134,695
94,148
-30.3
11.42
8.30
-27.3
0.59
0.67
+13.6
50.4
57.3
7.45
7.63
+0.18
121,987
110,658
-9.3
9.01
6.75
-25.1
0.88
0.95
+8.0
60.7
69.9
7.45
7.75
+0.3
109,723
42,416
-61.3
9.58
3.80
-60.3
0.92
0.59
0.80
-
59.7
46.6
59.7
5.64
4.15
7.61
-
137,547
271,945
63,996
-
12.46
17.60
5.50
-
0.41
0.66
0.60
+61.1
40.2
61.7
53.6
7.45
7.74
7.67
+0.29
72,100
65,627
61,823
-9.0
8.99
7.99
7.52
-11.1
+0.22
-14.3
-16.4
+46.3