. u . ting

5 . ! Wk» .1 .... “$3..

g 4 I I, Z I . ‘ . V 9 v VAC”
95”“?! 11.. it”: ”It 2; . . . . V H 5 ... ”cut... .3
kn mm ixMWkp Uﬁﬂhur . ‘ . «munuﬁuw. mu .. .

3 . ... , . . . ﬁuﬁﬁiﬁ
, . .3 _ u.
t

..nﬂk .Rm' {afar

Shaw 1.... 32$...)
F I
f

v.
r

{JP}
. .1

... ...».
.
gag... ......
OIFHJNR‘IE
‘
. .

knits}... . . . . .

Laid?» .

i. I ..u Ir
, I'lyo (.....stili‘ll

 

.. ...:hauii‘ «:

...; . ..u

. 1m?! 3.3... s.
:

.mrvv
3

.gavhil‘v‘tn

. 2131‘ 3.. :VIoILKi‘siiis

.I‘.\Ivurl.uv7t IVA-.xolx‘ttnvt
‘ y Igllfx

Sci .

311115.»..‘Iﬁnct n... . rvuiiun

...Ii.lf¢\.§.li’.l~‘llv
...:

 

X: vf-tilQVJinaullftlﬁt
9|; l :lt‘IalofvﬁIl
zillivc‘xlltonl.

 

f, .u . in”); . Wm. 1hr ..u ..
.. a. . .31, .1 1.4.43.
gks ...,aﬁﬁ: Ema .,

 

 

- LIBRARY
MlChig‘al l crate
Un hair‘s a}; :3 y

.. ______J

 

 

 

This is to certify that the
dissertation entitled

IMPACT OF SAND MANURE SEPARATION ON
ANAEROBIC DIGESTION

presented by

DANA M KIRK

has been accepted towards fulﬁllment
of the requirements for the

Ph.D. degree in Biosystems Engineering

 

 

Ax.» X. gig/o..—

Major Professor’s/[Signature

11/2/2007

Date

MSU is an Afﬁrmative Action/Equal Opportunity Employer

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

APR 0 4’20l2

H.011- 13

NOV 3 O 2014

11121

 

 

 

4:

 

 

 

 

 

 

 

 

 

 

 

5/08 K:IProjIAcc&Pres/ClRC/Daleom.hdd

Impact of Sand Manure Separation on Anaerobic Digestion
By

Dana M Kirk

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
- for a degree of

DOCTOR OF PHILOSIPHY
BIOSYSTEM ENGINEERING

2009

ABSTRACT

Impact of Sand Manure Separation on Anaerobic Digestion
By
Dana M Kirk

Bedding dairy cows on sand improves animal health resulting in higher
milk production and overall farm proﬁtability. The resulting sand-laden dairy
manure (SLDM), however, complicates manure management, causing premature
equipment wear, clogging pipes and settling during storage. Sand separation
systems (SSS) remove and reclaim sand from SLDM. The increased handling of
manure and the addition of dilution water associated with the removal of sand
alter the manure characteristics. Efﬁciency of sand separation and the resulting
impact on anaerobic digestion (AD) has not been extensively evaluated.

The objective of this research was to determine the impact of sand
manure separation on AD. To accomplish this objective, a technique to estimate
separation efﬁciency ﬁrst needed to be developed. The separation efﬁciency of
sand, in combination with an understanding of the residual sand characteristics
and the loss of volatile solids (VS), allows for solid’s balances to be determined
across the entire sand separation system ($88) and AD. This balance can then
be used to predict impact on AD performance. To verify the predictions,
comparison to a full-scale, operating digester was conducted.

Mass balance was found not found to be possible due to the unstable ﬂow

rates of several SSS inputs and outputs. Consequently, a semi-empirical

evaluation technique was developed that required a combination of industry
standards and on-farrn measurements. For the test farm, Green Meadow Farms,
the overall ﬁxed solids (FS) separation efﬁciency of 91 to 99% was estimated.

The average sand particle size remaining in the manure following the
SSS, residual sand, was determined to be between 0.18 mm and 0.21 mm.
Installed mixer power, theoretically could achieve the scour velocity for the
residual sand average particle size, indicating that settling should be minimal.
This was conﬁrmed when one AD tank was emptied after ﬁfteen months of
operation revealing only 25 to 50 mm of sludge (sand and manure solids)
accumulation.

During the sand separation process, a loss of VS from the manure stream
was observed, however the change was not found to be a statistically signiﬁcant
treatment effect. The observed cumulative change in the mass of VS determined
using the semi-empirical mass balance ranged from 33 to 53%. Changes in VS
are important due to the direct correlation between V8 and biogas potential. The
theoretical electrical energy potential of the full-scale AD at the case study farm,
which utilized SSS efﬂuent as the feedstock, was 5,890 kWh/d. In 2008, the
maximum electrical output of the full-scale system was achieved in July, when
5,505 kWh/d was produced or 93% of theoretical potential. The lost electrical
generation revenue due to loss of VS throughout the SSS, assuming $0.08 kWh,

was $123,200 per year.

Copyright by
Dana M Kirk
2009

Dedication

To My Wife Andrea and My Family

Acknowledgements

I would like to express my gratitude to my major professor, Dr. Steven
Safferrnan, for his knowledge and support throughout my studies at Michigan
State University.

I am also grateful to my committee members for their prolonged support. I
would like to acknowledge Dr. Bill Bickert for the years of support and mentoring.
It is largely due to your guidance that l have been able to complete this project.
Thank you to Dr. Wolf, Dr. Harrigan and Dave Wallace for your support and
interest in this project and for serving as members of my guidance committee. I
would also like to thank the numerous undergraduate research assistants who
participated in this project and others over the course of my research, with out

you much of this work may have not been completed.

vi

TABLE OF CONTENTS

LIST OF TABLES ................................................................................................ xi
LIST OF FIGURES ............................................................................................ xvi
ABBREVIATIONS ........................................................................................... xviii
CHAPTER 1: INTRODUCTION ............................................................................ 1
1.1 Development of sand manure separation technologies ............................... 1
1.2 Sand-free manure ....................................................................................... 2
1.3 Research farms ........................................................................................... 5
1.3.1 Green Meadow Farms .............................................................................. 5
1.3.2 Minnis Dairy Farm .................................................................................... 7
1.4 Research objectives .................................................................................... 7
CHAPTER 2: Review of Literature ..................................................................... 9
2.1 Manure production and characteristics ........................................................ 9
2.2 Freestall bedding and cow comfort ............................................................ 10
2.3 Sand separation technology development ................................................ 13
2.3.1 Factors affecting sand separation ................................................... 14
2.3.1.1 Fluid viscosity .................................................................... 14

2.3.1.2 Sand particle size .............................................................. 16

2.3.2 Equations governing sand separation system design ..................... 19
2.3.2.1 Terminal settling velocity ................................................... 19

2.3.2.2 Scour velocity .................................................................... 24

2.3.3 Early sand separation research ....................................................... 26

2.3.4 Sand separation goals ..................................................................... 27

2.3.5 Fundamentals of sand separation technology ................................. 29

2.4 Sand separation technologies ................................................................... 30
2.4.1 SLDM collection and conveyance ................................................... 30

vii

2.4.2 Mechanical sand separation systems .............................................. 32

2.4.2.1 Counter-current upﬂow separators .................................... 32

2.4.2.2 Centrifugal separations ...................................................... 35

2.4.3 Passive sand separation systems ................................................... 38
2.4.4 Integration of technologies to create a sand separation system ...... 42

2.5 Determination of separation efﬁciency ...................................................... 44
2.6 Anaerobic digestion ................................................................................... 46
2.6.1 Biogas potential of dairy manure ..................................................... 49
2.6.2 Mixing .............................................................................................. 52
2.6.3 Heating requirements ...................................................................... 56
2.6.4 Sand bedding and anaerobic digestion ........................................... 57
CHAPTER 3: EXPERIMENTAL METHODS ...................................................... 59
3.1 Solids Analysis and characterization procedures ...................................... 59
3.2 Particle size distribution ............................................................................. 61
3.3 Laboratory quality assurance .................................................................... 62
3.4 Statistics analysis ...................................................................................... 62
3.5 Research farms ......................................................................................... 65
3.5.1 Farm descriptions ............................................................................ 65
3.5.1.1 Green Meadow Farms ....................................................... 65

3.5.1.2 Minnis Dairy Farm ............................................................. 68

3.5.2 Manure sample collection ................................................................ 70
3.5.2.1 GMF sample collection ...................................................... 72

3.5.2.2 MDF sample collection ...................................................... 73

3.5.3 Flow measurement .......................................................................... 74
3.5.4 Density measurement ..................................................................... 75

3.6 Fixed solid separation efﬁciency evaluation .............................................. 75
3.6.1 Mass balance approach .................................................................. 76
3.6.2 Semi-empirical mass balance ......................................................... 76

3.7 Particle size distribution ............................................................................. 78
3.8 Volatile solids loss during sand separation ................................................ 79

viii

3.9 Anaerobic digester design considerations ................................................. 79

CHAPTER 4: RESULTS AND DISCUSSION ..................................................... 81
4.1 Technique for quantiﬁcation of sand separation efﬁciency ........................ 81
4.1.1 Total solids characteristics of sand separation system products ..... 82

4.1.2 Fixed solids characteristics of sand separation system products ....84

4.1.3 Mass balance separation efﬁciency ................................................ 85
4.1.3.1 Flow rate determination ..................................................... 85

4.1.3.2 Separation efﬁciency ......................................................... 89

4.2 Semi-empirical mass balance separation efﬁciency .................................. 93
4.2.1 Semi-empirical mass balance parameter speciﬁcation .................... 94

4.2.2 Sand separation efﬁciency .............................................................. 99

4.3 Particle size distribution of sand separation products .............................. 104
4.3.1 Settling and scour velocity of sand ................................................ 107

4.3.2 Residual sand impact on anaerobic digester mixing system desigr1ro8

4.4 Volatile solids loss due to sand separation .............................................. 112
4.4.1 Volatile solids characteristics of sand separation products ........... 112

4.4.2 Volatile solids changes due to sand separation ............................ 114

4.5 Impact of SSS on anaerobic digester design .......................................... 118
4.5.1 Heating requirements of sand separation system efﬂuent ............. 119

4.5.2 Biogas potential from sand separation products ........................... 120
CHAPTER 5: SUMMARY AND CONCLUSIONS ............................................. 128
5.1 Summary ................................................................................................. 128
5.2 General conclusions ................................................................................ 130
CHAPTER 6: RECOMMENDATIONS FOR FUTURE WORK .......................... 132

 

APPENDICIES

APPENDIX A .................................................................................................... 134
APPENDIX B .................................................................................................... 138
APPENDIX C .................................................................................................... 149
APPENDIX D .................................................................................................... 155
APPENDIX E .................................................................................................... 157
APPENDIX F .................................................................................................... 160
REFERENCES ................................................................................................. 161
x

 

 

 

LIST OF TABLES

Table 2.1: Parameters impacting sand separation efﬁciency .............................. 14
Table 2.2: Dairy manure dynamic viscosity range (Keener t al., 2006) ............... 16
Table 2.3: Comparison of sand particle size range from three common
classiﬁcation systems ......................................................................................... 17
Table 2.4: Standard particle size distributions .................................................... 19
Table 2.5: Theoretical settling velocity of sand and manure particles in slurry
(Camp, 1945) ...................................................................................................... 23
Table 2.6: Theoretical scour velocity of sand and manure particles (sand: f=0.03
and s=1.76, manure: f=0.03 and s=1.04) ............................................................ 25
Table 2.7: US. Farm Based Anaerobic Digester and Electrical Generation
Capacity .............................................................................................................. 52
Table 3.1: Standard procedures for manure and sand characterization ............. 61
Table 3.2: SSS sample locations and description - GMF ................................... 68
Table 3.3: SSS sample locations and description — MDF ................................... 70
Table 3.4: Sample location data source — GMF .................................................. 75
Table 4.1: Total solid concentration — GMF and MDF ......................................... 83
Table 4.2: Fixed solid concentration - GMF and MDF ........................................ 84
Table 4.3: Sand separation system ﬂow rate — MDF .......................................... 86
Table 4.4: Sand separation system material density - MDF ............................... 88
Table 4.5: Sand separation system mass ﬂow rate - MDF ................................. 88
Table 4.6: Sand separation system ﬁxed solids mass ﬂow rate - MDF .............. 89
Table 4.7: Fixed solids separation efﬁciency - MDF ........................................... 90
Table 4.8: Standards used in the semi-empirical mass balance evaluation ........ 95

xi

Table 4.9: Mass and ﬂow rate data — GMF ......................................................... 96

Table 4.10: Overall FS mass by treatment step - GMF ....................................... 96
Table 4.11: Daily ﬁxed solids mass data — GMF ............................................... 100
Table 4.12: Cumulative ﬁxed solid separation efﬁciency - GMF ....................... 100
Table 4.13: Treatment level ﬁxed solid separation efﬁciency — GMF ................ 102
Table 4.14: Mean sand particle size distribution - GMF ................................... 105
Table 4.15: Residual sand scour velocity — GMF .............................................. 107
Table 4.16: Power required to achieve scour velocity of sand - GMF .............. 110
Table 4.17 Sand separation system volatile solid concentration - GMF ........... 112
Table 4.18: Overall VS mass by treatment step - GMF .................................... 113
Table 4.19: Daily volatile solids mass data — GMF ........................................... 115
Table 4.20: Cumulative change in volatile solids - GMF .................................. 116
Table 4.21: Treatment level change in volatile solids - GMF ............................ 117

Table 4.22: Heating required achieve AD operating temperature - GMF ......... 119
Table 4.23: Anaerobic digester biogas potential - GMF ................................... 121
Table 4.24: Anaerobic digester energy potential — GMF ................................... 121
Table 4.25: AD energy potential compared to heating requirement - GMF ...... 122

Table 4.26: AD energy potential compared to heating requirement assuming a
combustion efﬁciency of 80% — GMF ................................................................ 123

Table 4.27: Anaerobic digester heating requirement from waste heat - GMF .. 124

Table 4.28: Electrical generation potential of SSS products — GMF ................. 125
Table 4.29: Operational data from the electrical generator - GMF ................... 126
Table A1: Solids concentration data, MDF ........................................................ 134

xii

Table A1: Solids concentration data, MDF continued ....................................... 135
Table A1: Solids concentration data, MDF continued ....................................... 136
Table A2: TS concentration data, MDF ............................................................. 137
Table A3: FS concentration data, MDF ............................................................. 137
Table A4: VS concentration data, MDF ............................................................. 137
Table B1: Solids concentration data from GMF ................................................ 138
Table 32: Total solids concentration data, GMF ............................................... 144
Table B3: Fixed solids concentration data, GMF .............................................. 145
Table B4: Volatile solids concentration data, GMF ........................................... 145
Table 85: Individual FS, Type III tests of ﬁxed effects (ANOVA), GMF ............. 145
Table 36: Individual FS, trt least squares means, GMF .................................... 145
Table B7: Individual FS, trt least squares means conﬁdence interval, GMF ..... 146
Table B8: Individual FS, Differences of trt least squares means, GMF ............. 146
Table 39: Individual VS, Type III tests of ﬁxed effect (ANOVA), GMF .............. 146
Table B10: Individual VS, trt least squares means ........................................... 146
Table B11: Individual VS, trt least squares means conﬁdence interval, GMF... 147
Table 812: Individual VS, Differences of trt least squares means, GMF ........... 147
Table B13: HC FS, Type III tests of ﬁxed effects (ANOVA), GMF ............... I ...... 147
Table B14: HC FS, trt least squares means ...................................................... 147
Table B15: HC FS, trt least squares means conﬁdence interval, GMF ............. 147
Table B16: HC FS, Differences of trt least squares means, GMF ..................... 148
Table B17: HC VS, Type III tests of ﬁxed effects (ANOVA), GMF .................... 148
Table B18: HC VS, trt least squares means ..................................................... 148

xiii

Table B19: HC VS, trt least squares means conﬁdence interval, GMF ............. 148

Table B20: HC VS, Differences of trt least squares means, GMF .................... 148
Table C1: SSS ﬂow rate (Us), MDF .................................................................. 149
Table 02: $88 ﬂow rate (m3/hr), MDF .............................................................. 149
Table C3: Material density, MDF ....................................................................... 150
Table C4: SSS mass ﬂow rate, MDF ................................................................ 150
Table C5: TS concentration, MDF ..................................................................... 150
Table C6: SSS TS mass ﬂow rate, MDF ........................................................... 151
Table C7: FS concentration, MDF ..................................................................... 151
Table 08: SSS FS mass ﬂow rate, MDF ........................................................... 152
Table C9: VS concentration, MDF .................................................................... 152
Table C10: SSS VS mass ﬂow rate, MDF ......................................................... 153
Table C11: SSS TS separation efﬁciency, MDF ............................................... 153
Table C12: SSS FS separation efﬁciency, MDF ............................................... 153
Table C13: Piston pump speed measurements ................................................ 153
Table C14: SMS efﬂuent rate, 45 minutes sample period ................................. 154
Table D1: New sand PSD ................................................................................. 155
Table D2: Reclaimed sand PSD ....................................................................... 155
Table D3: HC underﬂow PSD ........................................................................... 155
Table D4: HC overﬂow PSD ............................................................................. 156
Table 05: Tank sludge PSD ............................................................................. 156
Table E1: Anaerobic Digester Efﬂuent .............................................................. 157
Table E2: HC Underﬂow ................................................................................... 157

xiv

Table E3: Post AD Equalization Tank Sludge ................................................... 158

Table E4: New Sand ......................................................................................... 158
Table E5: Reclaimed Sand ............................................................................... 159
Table F1: Operational Data from Several US. Based Anaerobic Digesters ..... 160

LIST OF FIGURES

Figure 2.1: USDA soil classiﬁcation triangle (USDA, 1993) ................................ 18
Figure 2.2 Force balance on a particle settling in a quiescent ﬂuid (Wedel, 19962)0
Figure 2.3: Diagram of column style counter-current upﬂow separator (Kim, 200:;
Figure 2.4 McLanahan Sand Manure Separator (Inglis et al., 2006) ................... 33
Figure 2.5: Typical centrifugal separator cross-section ....................................... 36
Figure 2.6: Hydrocyclone ﬂow pattern (Metcalf and Eddy, 1991) ........................ 36
Figure 2.7: Force balance about a particle settling in a centrifugal separator ..... 37
Figure 2.8: Typical settling basin conﬁguration ................................................... 39
Figure 2.9: Force balance about a particle settling in a passive gravity separatozo
Figure 2.10: Integration of conveyance and sand separation technologies ........ 43
Figure 2.11: Schematic diagram of a separator (Svarovsky, 1990) .................... 44
Figure 2.12: Biology of anaerobic digestion (Barker, 2001) ................................ 47
Figure 2.13: Appropriate manure characteristics for anaerobic digestion (US.

Environmental Protection Agency, 2002) ............................................................ 48
Figure 3.1: GMF SSS schematic ......................................................................... 67
Figure 3.2: SSS process ﬂow diagram - GMF .................................................... 68
Figure 3.3 SSS process ﬂow diagram — MDF ..................................................... 70
Figure 4.1: Sand manure separator efﬂuent ﬂow rate — MDF ............................. 92
Figure 4.2: Cumulative ﬁxed solid separation efﬁciency - GMF ....................... 101
Figure 4.3: Treatment level ﬁxed solid separation efﬁciency - GMF ................. 102

xvi

Figure 4.4: Average sand particle size distribution - GMF ................................ 106

Figure 4.5: Anaerobic digester mixer conﬁguration - GMF ............................... 109
Figure 4.6: Cumulative change in volatile solids - GMF ................................... 116
Figure 4.7: Treatment level change in volatile solids — GMF ............................ 118

Acc
AD
AU

BMP

BP

Btu
CCUS
Cd

CD

CD.

pr

cP

D

d

dF(x)
de(x)/dx
ch(x)/dx

e

ABBREVIATIONS

Area

Signiﬁcance level

Acceleration

Anaerobic digestion or anaerobic digester
Animal unit

Constant describing condition of sand in channel
Biochemical methane potential

Biogas potential

British thermal unit

Counter-current upﬂow separator

Drag coefﬁcient

Speciﬁc heat

Speciﬁc heat of sand

Speciﬁc heat of water

Centipoise

Depth

Diameter

Mass ﬂow of the inﬂuent

Mass ﬂow of the ﬁne efﬂuent

Mass ﬂow of the coarse efﬂuent

Residual term

xviii

ET

FS

GMF

HC

Mlm
mash
Mc
mar,
Mr
mm”.
Ms

Mw
Machine
MC
MDF

Overall separation efﬁciency

Shape factor

Darcy Weisbach friction factor

Force

Fixed solids

Gravitational acceleration

Green Meadow Farms, Inc

total heat

Hydrocyclone or cyclone

Horsepower

Liter

Dynamic I absolute viscosity

Mass

Sample mass after igniting

Mass of the. coarse particles contained in the underﬂow
Sample mass after drying

Mass of the ﬁne particles contained in the overﬂow
Initial (wet) sample mass

Mass of sand

Mass of water

Individual machine at a treatment level
Moisture content

Minnis Dairy Farm
xix

mgt
MINI
min
MSU
mx

NR

PD
PGS
PSD

Px
r
rpm

S
2:intluents

Zefﬂuents
SAS
SLDM
SMS

Sn, aoc

SSS

Farm management

Settling basins at GMF that follow SMS
Minute

Michigan State University

Daily manure production per cow
Reynolds number

Overﬂow ﬂow rate

Percent difference

Passive gravity separator

Particle size distribution

Power

Flow rate

Density of material

Radius

Revolutions per minute

Speciﬁc gravity

Sum of all material entering the device
Sum of all material exiting the device
Statistical analysis software
Sand-laden dairy manure

Sand manure separation

Mass of sand sample accumulated on the nth sieve

Sand separation system
xx

AT
trt

TS

Unit

v

VS
VSConversion
Vh

Vs

Vx

W

wb

xn, aoc

Time

Temperature

Temperature difference
Treatment

Total solids

Overall mean

Unferﬂow ﬂow rate

Level or machine within a sand separation system
Mean ﬂuid velocity

Volatile solids

Volatile solids destroyed in AD
Scour velocity

Settling velocity

Volume of a material

Width

Wet basis

Mass of total sand sample accumulated on the sieve of interest and

all larger sieves

Response variable

xxi

CHAPTER 1: INTRODUCTION

Sand is the preferred bedding material at many dairy farms in Michigan
and around the United Sates. As an inorganic bedding option, sand drains well,
is less likely to harbor mastitis-causing organisms, reduces lameness, increases
milk quality and production, provides traction in alleys and freely adjusts allowing
even distribution of the cow’s weight (Stowell and Inglis, 2000; Cook and
Nordland 2004b).

Sand bedding mixes with manure resulting in sand laden dairy manure
(SLDM). SLDM is difﬁcult to manage as it clogs pipes, causes premature
equipment wear, settles during storage and limits manure management options
(Gooch, et al., 2002). Overcoming these limitations is essential for dairy farms to

successfully manage manure.

1.1 Development of sand manure separation technologies
Prior to the development of sand separation technologies, SLDM was

typically handled by daily scrape and haul (land application). Due to the volume
of manure produced by an average dairy cow, 68 LJd (American Society of
Agricultural and Biological Engineers, 2005), reliance on daily scrape and haul is
difﬁcult and inconvenient as it is dependent on weather conditions and land
availability during the growing season. This lead many dairy farmers to construct
long-ten'n manure storages to provide ﬂexibility in its management. However,

during long-terrn storage, with the addition of dilution water from the milking

center or precipitation, sand settles, accumulating as grit. Settled sand is not
easily re-suspended into the manure slurry by agitation and often requires
excavation. To facilitate excavation, long-tenn storages were constructed with
concrete access ramps and floors, adding considerably to capital costs. Over
time, operators developed a skim and haul technique, where the liquid fraction of
the stored manure is pumped off allowing a longer interval between excavations.
However, time and costs led producers to seek alternatives.

One such alternative is the sand manure separator (SMS), developed in
the mid 1990’s (patent number 5950839). Grit separation technology from the
mining and wastewater industries provided the development platform (Wedel,
1995). Based on aerated grit chamber principals, sand is separated in the SMS
in four steps: metering, agitation or turbulence, sedimentation and grit removal
(Wedel and Bickert, 1996). Metering and dilution free sand from manure allowing
the dense sand particles to settle. The lighter manure solids remain suspended
in the liquid manure. Other separation technologies include the passive gravity-
settling basin (PGS), sand lane and hydrocyclone (HC) (Wedel, 1995), further

details in section 2.4.

1.2 Sand-free manure
Social and environmental concerns with modern dairy farms are leading to

the development of innovative manure treatment technologies to improve nutrient
utilization, diminish the potential for water pollution, reduce odor and emissions,
decrease the cost and time associated with manure land application and

2

potentially create discharge quality water from manure. However, the complex,
mechanical nature of these technologies, requires manure entering the system to
be virtually sand free.

Anaerobic digestion (AD), as part of an integrated manure management
system, addresses many of these concerns including decreasing odor and
emissions, lowering of water pollution potential by reducing the biological oxygen
demand (BOD) and pathogen load, converting manure nutrients to more plant
available forms and creating biogas, a source of renewable energy. Between
1981 and 1985 six-digester system were installed in Michigan (Rozdilsky, 1997).
However, for technical, economic and managerial reasons, by 1990 only one of
the original six AD was operating. Technological improvements and interest in
odor control, energy generation and bio-ﬁber bedding production from digested
manure solids is leading to renewed interest in digestion. According to the
AgStar Program in 2008, there were 121 farm based AD operating in the United
States with six in Michigan (US. Environmental Protection Agency, 2009). The
majority of the US systems are on dairy farms that use organic bedding
(shavings, sawdust or bio-ﬁber).

Operating data indicates that the performance of commercial anaerobic
digestion systems, biogas yield as well as generator size and output, are highly
variable and site-speciﬁc (Cornell, 2009). Seldom are changes in the VS content
of manure due to manure collection, conveyance and pretreatment identiﬁed.
Speciﬁcally, the impact of sand manure separation on the mass of volatile solids

in the manure stream has not been well evaluated with Inglis et al. (2006)
3

completing the ﬁrst document literature evaluation. Currently, only three
digesters systems are operating on farms using sand bedding; Green Meadow
Farms, Fair Oaks Dairy and Bridgewater Dairy. As greenhouse gas emission
and renewable energy become increasingly important for regulatory and
economic reasons, fully understanding the impact of bedding and manure
management prior to anaerobic digestion will become crucial.

According to the American Society of Agricultural and Biological Engineers
(2005), manure production from a lactating dairy cow is 68 L/cow/d. Midwest
Plan Service (2000), reports that sand-bedding usage averages 22.3 kg/cow/d,
equivalent to 13.1 L/d, which is signiﬁcant when compared to manure.
Accumulation of residual sand in the AD tanks is a primary reason why digesters
are not more prevalent on sand bedded dairy farms (Inglis et al., 2006). If AD is
used, the goal of sand manure separation shifts from removing enough sand to
alleviate settling downstream and reducing wear on equipment. Sand-free
manure is deﬁned in this research as containing only residual sand that will
remain suspended as it passes through units associated with AD systems.

The importance of sand-free manure to AD drives the need for techniques
to quantify sand separation efﬁciency and to characterize the residual sand that
passes through. This residual then needs to be correlated to the performance of

the AD.

1.3 Research farms
Research was conducted at two commercial dairy farms, Green Meadow

Farms, Inc., (GMF) and Minnis Dairy Farm (MDF). Both use similar sand

separation equipment.

1.3.1 Green Meadow Farms
Green Meadow Farms operated one of the original six Michigan AD from

1983 until 1990 (Rozdilsky 1997). In 1990, the original AD was decommissioned
due to mechanical issues with heating, sand accumulation and building decay.
Removal of it, sand, which accumulated in the plug ﬂow digester, required a
complete system shut down resulting in signiﬁcant labor cost and lost revenue
due to a lack of biogas production.

In 1998, GMF installed four of the ﬁrst commercially available SMS
(Wedel, 2009) at their new 2,000-cow Farm 2. The sand separation system
(SSS), and associated SMS, was installed to minimize equipment wear and
prevent clogging and settling in downstream units. Each SMS has an integrated
passive settling basin (MINI) to capture residual sand in the SMS efﬂuent,
resulting in a two level SSS. At this farm, all manure (SSS efﬂuent) and milking
center wastewater was transferred via single force main to a solid-liquid
separator, which removed coarse manure solids from the slurry stream.
Separated manure solids were composted or land applied while liquid manure
was contained in long-term storage (180+ days) until it was land applied by

ﬁngaﬁon.

In 2001, Farm 3 was constructed at GMF. To manage the manure from
the additional 1,200 lactating and dry two additional SMS-MINI combinations
were added. Shortly thereafter, a chemical phosphorus separation system was
installed which used chemical precipitation and coagulation followed by
ﬂocculation and a belt ﬁlter press to remove phosphorus and manure solids from
the manure stream. This resulted in a liquid fraction low in phosphorus such that
it is irrigated growing crops and used as dilution water for the SSS. The solid
material contains high phosphorus levels and is land applied.

Over the ﬁrst seven years of operation, 1998 to 2005, the SSS performed
well. However, sand did cause premature wear on the transfer and chemical
phosphorus pumps, settled in pump chambers requiring regular excavation,
occasionally clogged force-main lines and accumulated in long-term storages
(Green, 2008). Quantitative measurements of the separation efﬁciency of the
original SSS were not conducted however; Wedel and Bickert (1998) reported a
separation efﬁciency for this general type of system in the range of 80% and
90%.

Because of the lack of data and experience in digesting manure from a
farm that uses sand beddings, MSU researchers partnered with GMF in 2005 to
develop a research/demonstration AD system to treat liquid manure from Farms
2 and 3. The project was largely funded by the Michigan Public Service
Commission (MPSC). GMF was interested in AD because for odor control,
potential of chemical cost reductions in their chemical phosphorus separation

system, and revenues associated with the production of renewable energy,
6

including carbon credits. Following the decision to install an AD, the goal of the
SSS changed to creating sand-free manure. Consequently, a third level of
separation technology was added to the system, a hydrocyclone (HC). The HC
size and conﬁguration was determined by the manufacturer and installed by GMF

staff in 2006.

1.3.2 Minnis Dairy Farm
Minnis Dairy Farm is a 600-cow dairy that uses sand bedding. A two-level

SSS that consisted of a single SMS and HC was installed in 2005. Unlike GMF,
the SSS at MDF was installed with the goal of removing and reclaiming sand for
reuse as bedding. MDF does not employ an AD. Since the installation of the
SSS, residual sand has not accumulated in downstream units, however, the
stainless steel dewatering screen and screw of the solid-liquid separator shows

premature wear.

1.4 Research objectives
Dairy farms, in general, are unique facilities with varying management

styles which combine to create site-speciﬁc manure management systems.
Operation of these systems results in site-speciﬁc manure properties which
should be considered when evaluating advance manure management systems,
such as AD. Quantifying the site-speciﬁc impact of manure pretreatment
(management) is important during the design of AD systems as the size of the

tank(s), heating system and biogas utilization may be impacted. Improperly

7

sized equipment can impact the system performance and cost the operator
signiﬁcant time and capital to remedy. Green Meadow Farms and Minnis Dairy
Farm were used as case studies for testing techniques to quantify sand
separation efﬁciency. Additionally, data from the operation of the AD at GMF
was used to determine important design and performance considerations for
dairy farms using sand bedding.
The speciﬁc research objectives follow.
1. Develop a technique to quantify the efﬁciency of sand separation systems.
2. Evaluate the residual sand particle size distribution to determine the
potential for settling in anaerobic digester tanks.
3. Quantify manure volatile solids changes resulting from sand separation.
4. Determine the impact of sand separation on anaerobic digester

performance (mixing, heating and biogas potential) and revenue potential.

Solids data from GMF was originally collected with the intent of determining
the efﬁciency of the SSS as plans were being developed to install an AD that
would utilize SSS efﬂuent as feedstock. Performance of the SSS was critical to
predicting sand accumulation in the AD tanks. Fixed solid (FS) were used to
track SSS efﬁciency. During sample analysis for F8, volatile solids (VS) are also
determined. While evaluating the FS data, it was observed that the VS data also
provided insight into changes in the mass of VS throughout the SSS. This
observation led to the development of objectives 3‘and 4 and a shift in the focus

toward determining the impact of sand separation on anaerobic digestion.
8

CHAPTER 2: Review of Literature

The following literature review examines the characteristics of
manure, freestall-bedding options, the development of SSS, methods for
determining SSS efﬁciency and the impact on the theoretical design of an AD

system.

2.1 Manure production and characteristics
Manure production for lactating dairy cows is predicted to be between 67

and 68 kg/cow/d by American Society of Agricultural and Biological Engineers
(2005) and Midwest Plan Service (2000), respectively. Dry cow manure
production was estimated to be between 38 and 52 kg/cowld (American Society
of Agricultural and Biological Engineers, 2005; Midwest Plan Service, 2000).
American Society of Agricultural and Biological Engineers (2005) estimates for
manure and nutrient excretion were derived from the combination of multiple data
sets from Washington State University, University of California - Davis, The Ohio
State University, and Pennsylvania State University. Both sources deﬁne
manure as the combination of feces and urine with no addition of water or
bedding. Midwest Plan Service (2000) indicated that actual characteristics of
manure could vary 130%.

Moisture content of lactating and dry cow manure was similar for both
references, ranging from 87% to 88% (American Society of Agricultural and
Biological Engineers, 2005; Midwest Plan Service, 2000). American Society of

9

Agricultural and Biological Engineers (2005) determined the total solids (TS)
excretion for a lactating and dry cow to be 8.9 and 4.9 kg/d/cow, respectively.
Volatile solids (VS) excretion was found to be 7.5 and 4.2 kg/d/cow, for a
lactating and dry cow respectively (American Society of Agricultural and
Biological Engineers, 2005). The difference in TS and VS is the ﬁxed solids (FS)
(Wedel, 1995).

Nennich et al. (2006) and Bannink et al. (1999) reported that the mass of
urine was approximately one-third of the total mass of manure excreted daily.
The TS concentration of urine ranges from 3% to 4.5% (American Society of
Agricultural and Biological Engineers, 2005; Bannink et al., 1999). Urine volume
and composition is variable depending on ration, mineral supplement, lactation
stage and environmental conditions (American Society of Agricultural and

Biological Engineers, 2005).

2.2 Freestall bedding and cow comfort
Animal health, cow comfort and milk production are advantages of sand

bedding, compared to organic bedding (Inglis et al., 2006; Wedel, 2001).
However, the production of SLDM limits the options available for manure
collection, conveyance, treatment, storage and utilization (Wedel and Bickert,
1996). Advance manure treatment technologies are particularly susceptible to
operating problems if sand is in the manure stream (Wilkie, 2005; Rozdilsky,

1997). Separation of sand from manure, resulting in an efﬂuent that is essentially

10

sand free, is crucial for the successful adoption of AD and other advance
treatment (Inglis et al., 2006).

Cow comfort is an important management component of production and
overall animal health (Wagner-Starch et al., 2003). Studies show that cow
comfort impacts milk yield and quality as well as animal health and longevity
(Linn, 2001). Bedding type is an important factor contributing to cow comfort.
Freestall bedding material should provide a clean, dry surface (Bewley et al.,
2001). Karszes (2003) added that properly designed and maintained freestalls
minimize the potential for mastitis, reduce hock abrasions and limit injuries to
animals. Freestall bedding materials are generally categorized as organic or
inorganic.

Common organic freestall bedding materials include crop residues (straw
and corn stalks), wood biomass (sawdust and shavings) and recycled material
(separated manure solids and newspaper). Cost and availability generally
determine which organic bedding is used on organically bed dairy farms. To
reduce bedding usage and costs, concrete, rubber mattresses or rubber pillows
(ﬁlled with a variety of materials) are sometimes used to create the freestall base
on farms using organic bedding (Cook et al., 2004; Bewley, et al. 2001). Another
advantage of manure containing organic bedding is that it requires little to no
pretreatment prior to treatment or storage. Manure collection and conveyance on
dairies using organic bedding consists of scrape or ﬂush collection with gravity
ﬂow or pump conveyance. Conventional clay-lined, concrete or steel manure

storages are used for manure containing organic bedding.
1 1

Disadvantages of organic bedding include harboring mastitis causing
microorganisms, absorption of water, urine and milk, and slippery freestall alleys.
Rubber mattresses and hard surfaces also have disadvantages. Cook et al.
(2004) found that dairy cows spend signiﬁcantly more time standing in freestalls
with rubber mattresses than cows housed in barns that use sand bedding.
Extended standing time and ﬁrm or hard freestall-bases increased the number of
lameness cases, compared to sand (Cook, 2003). Cook (2001) estimated that
the average cost to treat lameness on dairy farms using organic bedding was
$82.50/cow (2001 dollars) compared to sand based dairy farms.

Sand is the most common inorganic bedding. Others options include
crushed limestone and byproducts from industrial manufacturing. Sand has been
promoted as the “gold standard” bedding material because it is non-hygroscopic,
drains well, provides traction in alleys, is less likely to harbor mastitis-causing
organisms and moves freely allowing even distribution of the cow’s weight
(Stowell, 2000;Bernard and Bray, 2004).

Inorganic sand bedding results in an increase of 1.4 to 1.8 kg/d of milk
production compared to organic bedding materials (Stone, 2003). Likewise,
herds on sand bedding typically have milk somatic cell counts (SCC) 50,000 cells
per milliliter less than comparable herds on organic bedding (Stone, 2003). Cook
and Nordlund, (2004) determined that sand bedding created a $152 cow/yr (2004
dollars) advantage over organic bedding materials.

Positive beneﬁts of sand bedding are balanced by the difﬁculties of

managing sand in the manure stream. SLDM is abrasive, increasing wear and
12

shortening the life of manure handling equipment (Stowell and Bickert, 1995).
According to the Midwest Plan Service (2000), a typical mature dairy cow
requires 8,130 kg/yr of sand bedding. In addition, SLDM is typically not
stackable or pumpable (Wedel and Bickert, 1994). Bedding sand tends to settle
out of suspension during conveyance, treatment and storage resulting in clogging
systems and reduced capacity (Inglis, 2006). Sand that settles during storage is
difﬁcult to re-suspend and generally requires physical removal (excavation) with
a loader tractor. To accommodate removal, manure storages are generally
constructed with a concrete ﬂoor and access ramp. In order to reduce issues
with SLDM, sand separation technologies have been developed to allow for the

removal of sand prior to treatment or storage.

2.3 Sand separation technology development
Over the past two decades, signiﬁcant advances have been made in the

development of technologies for separating sand from manure (Wedel and
Bickert, 1994; Wedel and Bickert, 1996; Wedel and Bickert, 1998; Wedel, 2001).
Separation technologies from the mining industry and municipal and industrial
wastewater treatment have been adapted (Wedel, 1995) to SLDM. Several
separation techniques are used including screening, sedimentation, centrifugal
force (hydrocyclone), dissolved air ﬂoatation and belt ﬁlter press with polymer

addition (Wedel, 1995).

13

2.3.1 Factors affecting sand separation
All sand separation technologies rely on the basic principles of

sedimentation (Wedel, 2001). Sedimentation is the separation of grit (heavy
particles) from water by gravitational settling (Metcalf and Eddy, 1991).
Performance of individual sand separation technologies and the overall sand
separation system (SSS) is affected by several characteristics of SLDM including
speciﬁc gravity, particle size distribution, viscosity and sand quantity (Wedel,
2001; Wedel and Bickert, 1996). Table 2.1 summarizes range of the various
parameters impacting sand removal.

Table 2.1: Parameters sand

     

manure 6 36 et
beddi sand size 0.076 2.01 mm 1991

sand 1.4 5 1995 &
of manure 0.4 1. Glover 1995 & Wed 2000

2.3.1.1 Fluid viscosity
Viscosity, the resistance of a ﬂuid to deformation under shear stress

(Steffe, 1996), is commonly used to describe the internal resistance to ﬂow.
Viscosity in terms of sand manure separation is important because high viscosity
increases friction, slowing the rate of settling. A basic understanding of the
rheological properties, physical and ﬂow properties (Landry et al., 2003; Steffe,
1996), of dairy manure is needed to determining settling characteristics. Manure
is a mixture of water and solids in a matrix of long chain organic molecules

described as mucus (Wedel and Bickert, 1996). Mucus, a weak viscoelastic gel

14

(Allen et al., 1984), is commonly found in the gastrointestinal tract of dairy cows
(Wedel, 2001). Manure exhibits non-Newtonian, shear-thinning properties
(Wedel, 1995), where increasing the shear rate decrease the apparent viscosity
(Steffe, 1996). Separation of sand from manure requires disruption of the mucus
molecule (Wedel, 2001), this is accomplished by agitation and turbulence (Wedel
and Bickert, 1996) which thins the material while the addition of dilution water is
disperses the particles.

Mixing, time, temperature and pressure inﬂuence the viscosity of a ﬂuid
(Steffe, 1996). Kumar et al. (1972) reported that the viscosity of dairy manure
decreased with increasing temperature. Viscosity is also inﬂuenced by the TS
concentration of the slurry inside the SSS. Landry et al. (2003) found that
apparent (dynamic) viscosity of manure is well correlated to the TS
concentration. Keener et al. (2006) conﬁrmed this ﬁnding and found that
viscosity decreased exponentially as the moisture level increased. On dairy
farms, TS concentration can be lowered through dilution (Kenner et al., 2006;
Landry et al., 2003). For dairy manure as excreted, Keener et al. (2006) studied
the change in viscosity and developed Equations 2.1 and 2.2, to predict dynamic
viscosity based on moisture content and rotational velocity of the spindle (spindle

speed).

[1 = 863.289—0.6211*MC (2_1)

15

u = 858.218-0.570*MC (2.2)

MC = moisture content, %

II = dynamic viscosity, cp

Equation 2.1 is based upon a spindle speed of 30 rpm, while Equation 2.2 uses a
spindle speed of 60 rpm. Keener et al., (2006) used a rotary viscometer to
measure viscosity. The rotary viscometer used a ﬁxed cup and a spindle to
measure the resistance of a ﬂuid to ﬂow. Trials were carried out at two
temperature proﬁles 20°C and 23°C and over a range of moisture contents, on a
weight basis (wb). The range of dairy manure apparent viscosity typically found

on farms is shown in Table 2.2 for both spindle speed equations.

Table 2.2: Dairy manure dynamic viscosity range (Keener t al., 2006)

 

 

 

 

Moisture Dymanic Viscosity (p) Average
Content 30 rpm @ 23°C 60 rpm @ 23°C I1
(wb) (GP) (69) (en)
85 36,152 17566 26,809
99 6.1 6.0 6.0

 

 

 

 

 

 

2.3.1.2 Sand particle size
Sand particle size characteristics vary by location, mineral type and mining

or manufacturing process (Gooch and Inglis, 2007). Bedding sand is a
composite material comprised of particles of varying size, density and shape.

Table 2.3 compares the soil particle size distribution of three common

16

classiﬁcation systems; United States Department of Agriculture (USDA), Uniﬁed

and American Association of State Highway and Transportation Ofﬁcials

(AASHTO). All three-classiﬁcation systems overlap with a sand particle size

range of 0.08 mm to 2.0 mm.

Based on the USDA soil classiﬁcation triangle, shown in Figure 2.1, to be

classiﬁed as sand the material can contain a combination silt and clay that is no

more than 10% of the total mass of the material.

Table 2.3: Comparison of sand particle size range from three common

classiﬁcation s

rstems

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

USDA (2008) Uniﬁed (ASTM, 2006) AASHTO (1991)
Description Min Max Min Max Min Max
(mm) (mm) (mm) (mm) (mm) (mm)
Very coarse sand 1.00 2.00
Coarse sand 0.50 1.00 2.00 4.83 0.43 2.00
Medium sand 0.25 0.50 0.43 2.00
Fine sand 0.10 0.25 0.08 0.43 0.08 0.43
Very ﬁne sand 0.05 0.10
Silt 0.03 0.05 0.08 0.08
Clay 0.03

 

17

 

100

 

 

 

 

 

     

We?
70
S so 97:;
s a 4:
s ‘0
so .
dy O
53"
40 cl A . Q
30 V
\ /\ /\ A A 7\

 

 

sand clay loam

fAV“A\WAVMV_ i
sadn \'°as'§'Id\ VY V‘_ V VV V

 

\ \
a, \e \e \e \oa \v‘a \e e
l greent sand

Figure 2.1: USDA soil classiﬁcation triangle (USDA, 1993)

Concrete and Mason sand are construction categories by the American
Society for Test and Materials (ASTM, 2006) as standards C-33 and C-144,
respectively. These are common bedding sands used on dairy farms around the
Midwest. Table 2.4 shows the high and low particle size limits established by
ASTM for Concrete and Mason sand. Also, 2NS is a Michigan Department of
Transportation (2003) standard that is commonly used for bedding sand on dairy

farms in Michigan.

18

Table 2.4: Standard rticle size distributions

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

. Percent PassirLL
US Sreve
Standard Opening Concrete Sand Mason Sand 2N8 Sand
Sieve # (ASTM C-33, 2006) (ASTM C-144, 2006)
(mm) Low Limit Eh Limit Low Limit High Limit (MDOT, 2003)

4 4.75 100% 95% 100% 100% 98%

8 2.36 100% 80% 100% 95% 80%
16 1.18 85% 50% 100% 70% 55%
30 0.6 60% 25% 75% 40% 38%
50 0.3 30% 5% 35% 10% 20%
100 0.15 10% 0% 15%. 2% 5%

 

Understanding the sand particle size distribution is critical to the

development of a SSS because the diameter of the sand is needed to determine

the settling and scour velocity. American Society of Civil Engineers (1975)

indicates that particle size is the most important parameter, related to the sand

grain, for predicting sedimentation. In addition, Zimmels (1984) stated that

separation technologies are less efﬁcient for wide distributions of particle size.

Therefore, different separation technologies are effective over only a portion of

the sand range.

2.3.2 Equations governing sand separation system design
Settling and scour velocity are the two most important parameters in the

design of hydraulic conveyance and sand manure separation systems. The

subsections below discuss each.

2.3.2.1 Terminal settling velocity
When a particle is released in a still ﬂuid, it accelerate until the sum of the

drag (upward) and the buoyant (upward) force equal the weight of the particle

19

 

(downward) and the buoyant force (downward) (Liu, 2001), this phenomenon is
known as terminal settling velocity (Wedel, 1995). Figure 2.2 depicts the force

balance in a still ﬂuid.

DRAG BUOYANT
FORCE FORCE

 

V
WEIGHT

Figure 2.2 Force balance on a particle settling in a quiescent ﬂuid (Wedel,
1996)

Particle size and density are two critical parameters used in calculating the
terminal settling velocity (Vs) of a spherical sand grain using Stokes Law,

Equation 2.3 (Lamb, 1993; Wedel and Bickert, 1996).

 

V = (Pp-pf)*g*d2
5 18m

(2.3)
V. = settling velocity, m/s
p9 = particle density, kglm3

p. = ﬂuid density, kg/m3
20

g = gravitational acceleration, m/s2
d = particle diameter, m
l1 = dynamic (apparent) viscosity, kg/m - s

18 = particle area and drag coefﬁcient correction factor

Stokes law holds true for ﬂow ﬁelds with Reynolds numbers less than 0.5
(Camp, 1945; Metcalf and Eddy, 1991). The Reynolds number can be

determined using Equation 2.4 (Metcalf and Eddy, 1991).

ads
NR = 37-91 (2.4)

NR = Reynolds number, dimensionless
v = mean ﬂuid velocity, m/s

d = diameter, m

In addition, Stokes law assumes that settling particles are spherical.
Settling velocity of non-spherical particles can be determined by a modiﬁcation of
Newton’s Law (Gregory et al., 1999), which includes a terms for the particle drag
coefﬁcient and shape. Newton’s Law is expressed in equation 2.5.

1* is(pp—pf)sd

VS = (3 Cd ¢ * pf )0.5 (2.5)

21

Cd = drag coefﬁcient

II = shape factor

The drag coefﬁcient as expressed in equation 2.6 (Concha and Almendra,

1979).

9.0
Cd = 0.28 .. (1 +53; 2 (2.6)

Shape, the combination of sphericity and roundness, impacts the particle
friction; angular particles are subject to more friction then rounded particles
(Alshibli et al., 2004). Angular sand particles generate more friction during
sedimentation processes than rounded grains. The result of the increase in
friction is slower settling velocity for angular particles. Shape factor ranges from
0 to 1 for sand. Natural sand on average has a shape factor of 0.7 (Vanoni,
2006).

Table 2.5 compares the terminal settling velocity of sand and manure
particles in liquid manure. Equation 2.5 was used to calculate the settling
velocity of sand particles with a shape factor of 0.7 in a laminar ﬂow (R=2,300) .
Due to a lack of shape information, the settling velocity of manure was calculated
using Equation 2.3 with a moisture content of 95%. For both, ﬂuid density of

manure was assumed to be 1,000 kglm3 (Table 2.1).

22

Based on the predicted values in Table 2.5, sand particles have a settling
velocity at least 13 times greater than that of equal size manure particles.
However, smaller sand particles have settling velocities similar to larger manure
particles. For example, in Table 2.5, it can be noted that a sand grain with a
diameter of 0.15 mm has a settling velocity similar to manure solids with a
particle size of 4.75 mm. This indicates that coarse manure particles may settle

with smaller diameter sand grains.

Table 2.5: Theoretical settling velocity of sand and manure particles in
slurry (Camp, 1945)

 

 

 

 

 

 

 

 

 

 

Particle Diameter swim V°'°°“y
Sand Manure
(mm) (W3) (W3)
4.75 4.1501 3.1502
2.35 2.9501 7.7503
1.13 2.1501 1.9503
0.50 1.5501 5.0504
0.30 1.0501 1.2504
0.15 7.3502 3.1505
0.07 5.2502 7.5505

 

 

 

 

 

Settling velocity is a useful tool in determining sand removal because it
account for particle size, density and ﬂuid density. Simpliﬁed sediment transport
models assume complete removal of all particles with settling velocities greater
than the overﬂow rate of the sand separation device (Jin, et al., 2000). The
determination of settling velocity is important when selecting bedding sand
particle size. Based on Stokes Law (Equation 2.3), the ideal sand particle in
terms of size and density should be a medium to coarse grain with a high

23

density. Understanding the settling velocity of a particle is important, however,
the energy required to resuspend (scour) a particle is greater and should be used

to size mixing systems used on dairy farms with sand bedding.

2.3.2.2 Scour velocity
Scour velocity is the mean horizontal velocity necessary to impart motion

on a particle at rest (Wedel, 2000). Similar to settling velocity, particle size and
speciﬁc gravity of sand is critical to determining scour velocity. Shields’ equation
(Equation 2.7), as described by Camp (1945) and Crites and Tchobaboglous

(1998), is used to determine the horizontal scour velocity of particles.

 

1
80321: ands —1 _
VH = (( 9f (S ))2 (2.7)

V" = scour velocity, m/s

B = constant describing condition of sand in channel
9 = speciﬁc gravity

f= Darcy-Weisbach friction factor

8 = Darcy—Weisbach correction factor

The minimum scour velocity (initiation) will move particles by saltation, the
fonlvard movement of particles by bouncing along a surface (Wedel, 2000), while

the complete scour velocity fully re-suspends particles. According to Wedel

24

(2000), the bed constant, B, ranges from 0.04 for scour initiation to 0.8 for full
scour. Typical Darcy-Weisbach friction factors range from 0.02 to 0.03 (Metcalf
and Eddy, 1991). The Darcy-Weisbach friction factor depends on the surface
characteristics over which the material ﬂows. Table 2.6 summarizes the range of
scour velocities for a range of sand and manure particle sizes.

Similar to the settling velocity relationship, the initiation and complete
scour velocities of manure solids is roughly one fourth that of an equal size sand
grain (T able 2.6). The difference in the scour velocities of sand and manure is
largely attributed to the difference in speciﬁc gravity, 1.76 for sand compared to

1.04 for manure.

Table 2.6: Theoretical scour velocity of sand and manure particles (sand:
f=0.03 and s=1.76, manure: f=0.03 and s=1.04)

 

 

 

 

 

 

 

 

 

 

 

Particle . Scour Velocity
Diameter Initiation Com lets
Sand Manure Sand Manure

(mm) (W8) (W8) MS) (tn/8)
4.75 0.68 0.14 3.02 0.63
2.36 0.48 0.10 2.13 0.44
1.18 0.34 0.07 1.51 0.31
0.60 0.24 0.05 1 .07 0.22
0.30 0.17 0.04 0.76 0.16
0.15 0.12 0.02 0.54 0.11
0.07 0.08 0.02 0.38 0.08

 

 

 

 

 

 

 

Interestingly, the velocity required for complete scour of manure particles is

nearly equal to the initiation velocity of similar sized sand grains.

25

Camp (1945) suggested that the mean velocity within the SSS should not
exceed the settling velocity of the largest sand grain to be recovered. Ideally, the
velocity of the separation device would not exceed the settling velocity of the
target sand particle, but would surpass the scour velocity of a majority of the
manure particles. However, due to the similarity in sand grain settling velocity
and manure solid initiation velocity it may not be possible to design SSS that can
practically achieve the optimal velocity. For evaluation and design purposes,

scour velocity is used to size mixing systems to minimize sedimentation.

2.3.3 Early sand separation research
Research conducted by Wedel (1995) found that sand does not settle out

of suspension in undiluted raw dairy manure due to the high viscosity created by
mucus and the variable and irregular shapes of the particles contained in the
SLDM matrix. Wedel (1995) discovered that by diluting the manure with a little
as 0.5 parts water to 1 part SLDM followed by agitation was sufﬁcient to initiate
the separation of sand, manure solids and water. Dilution dispersed the solids
and mucus while agitation enhanced shear-thinning behavior, thus reducing the
overall slurry viscosity.

Wedel (1995) conducted settling experiments in clear columns, at a
dilution ratio ranging from 0.5 to 5 parts water to 1 part SLDM. The research
found that at dilution ratios as low as 1:1 caused distinct layers of sand manure

solids and liquid began to form (Wedel, 1995). Increasing the dilution ratio from

26

1 to 1 up to 5 to 1 signiﬁcantly reduced the time required for sand to settle out of
the SLDM mixture.

Hinder settling is the predominate type occurring in diluted SLDM. The
other settling types are simultaneously occurring to a lesser extents (Inglis,
2006). As the dilution ratio increases, the SLDM mixture viscosity decreases
transitioning the settling from hindered to discrete.

Hindered and discrete settling are two of the four particle settling classes,
which include (I) discrete (or free), (ll) ﬂocculent, (lll) hindered and (IV)
compression settling (Metcalf and Eddy, 1991). Discrete settling occurs when
individual particles settling without interaction or ﬂocculation. Flocculent settling
occurs when particles in a dilute suspension interact and aggregate, creating
larger heavier particles. Hindered settling occurs in solutions with intermediate
particle concentration. The particles density causes interparticle forces to ﬁx
each particles relative position, causing the mass of particles to settle at a
constant rate. Suspensions with high particle concentrations result in particles

touching and settling by compaction of the mass.

2.3.4 Sand separation goals
SSS is intended to remove sand from the manure to achieve different

goals, as discussed below (Bickert and Kirk, 2007).
1. To remove most, but not all, of the sand from the manure stream with no
intention of using the removed sand for bedding. Removal is just enough
to reduce downstream problems.

27

2. To reclaim sand clean enough for reuse as freestall bedding.

3. To create a sand-free manure stream for downstream treatment as well as
reclaim sand for reuse as bedding. Sand-free manure is critical for farms
implementing advanced manure treatment systems such as anaerobic

digestion.

Kappe and Neighbor (1951) reported that grit removal systems used to
treat municipal wastewater captured a sufﬁcient quantity of particles with a
diameter of 0.2 mm. (U.S. Sieve No. 70) to effectively protect pumps from heavy
wear and prevent deposits in downstream treatment units. Wedel and Bickert
(1998) asserted that mechanical SMS remove between 80% and 90% of the
sand contained in SLDM. Fulhage (2003) reported that settling basins removed
between 71% and 75% of bedding sand. Both demonstrate that goal 1 is
achievable.

Findings by Hamer et al. (2005) found no difference in the bacterial
concentrations of new (fresh) and 7 to 10 day old reclaimed sand, indicating that
SSS could achieve goal 2.

Sand-free manure, goal 3, is deﬁned as containing residual sand that does
not settle during storage. Sedimentation chambers (settling basins) operating at
municipal wastewater treatment plants have recovered up to 99% of grit with a

diameter of 0.003 mm (US Sieve No. 200) (Wedel and Bickert, 1996).

28

2.3.5 Fundamentals of sand separation technology
For all three goals, successful separation of sand and manure is based on

four key steps; metering, mixing, which includes agitation and turbulence,
sedimentation and sediment (sand) removal (Wedel and Bickert, 1996). The
intent of metering is to balance the input of raw SLDM in to the SSS so that the
dilution and agitation capabilities are optimized. Improper metering can lead to
poor sedimentation and reclaimed sand containing a high concentration of VS.
As discussed earlier, dilution is necessary to reduce the viscosity of the SLDM
mixture and reduce the hindrance (Zimmels, 1984).

In combination, or just after the addition of dilution water, the manure
slurry is agitated to wash sand grains free of manure. In mechanical systems,
agitation is achieved by the turbulent addition of dilution water near the base of
the separation unit and by the sand removal auger. Passive systems achieve
agitation by the ﬂush conveyance system. Sedimentation, the third step, results
from a quiescent condition or the application of centrifugal force. In both
mechanical and passive separation devices, sedimentation occurs by differential
settling, where the settling velocity of the sand grain is greater than that of
manure particles (Kim and Stolzenbach, 2003). Passive settling devices also
have a horizontal component to the ﬂow velocity, causing saltation of settled
particles. Saltation is the movement of particles by bouncing along the channel
bottom, occurs when the incomplete scour velocity has been achieved (Liu,
2001). Sand removal, the ﬁnal step, is the mining of separated sand from the
separation devices.

29

A beneﬁt of mechanical SSS is that all four critical steps in the sand
removal process are package in a single machine. Mechanical SSS can operate
with any manure conveyance system. Passive SSS require that dilution and
agitation occur in part of the manure conveyance system. Sand removal in
passive systems is a manual process requiring excavation using a front-end

loader and operator.

2.4 Sand separation technologies
Dairy producers interested in separating sand from manure have several

technologies to select from depending on their goals for separation and
conveyance system employed at the dairy farm. First SLDM is collected using
either a scrape or ﬂush system. There after the SLDM must be conveyed to a
central treatment location, typically using mechanical or hydraulic conveyance.
Sand removal is achieved by either a mechanical or passive separation unit.
Then the sand is moved into storage and the liquid slurry travels to the next
treatment step in the manure management system. Each step is described in

detail below.

2.4.1 SLDM collection and conveyance
Two types of manure conveyance exist for dairy farms, mechanical and

hydraulic (Kirk, 2005). Mechanical conveyance includes both scrape and
vacuum-scrape, while hydraulic systems are categorized as scrape-ﬂush and

ﬂush (Kirk, 2005).

30

Scrape and vacuum-scrape system use a device to push manure from the
freestall alley to a collection/treatment point near the building. Included are
traditional alley scrapers, tire scrapers mounted on skid loaders, and the
vacuum-scrapers (vacuum-scrape). Vacuum-scrapers use a scraper bar to

collect the manure which is transferred into a self-propelled or trailer mounted
tank using vacuum (similar to a vacuum cleaner), allowing it to be transported to
the SSS. The product of these systems are manure as excreted mixed with
bedding and urine.

Flush collection systems create a wave of water that collects manure from
the freestall alley and moves it to the point of treatment. A water release rate of
‘ at least 0.63 mals is recommended for SLDM (Hamer et al., 2003). Scrape-ﬂush
systems combine physical scrape collection of manure from the freestall alleys
with hydraulic conveyance of manure from the barn. Manure collection by the
scraper is deposited into a ﬂush channel that conveys manure from the barn.
Hydraulic collection and conveyance systems are typically designed to meet or
exceed the mean scour velocity of the largest particle, generally in the range of
2.4 to 3 m/s for SLDM (Harner et al., 2003). Wedel (2000) suggested that the
mean ﬂow velocity should be between 1.5 to 2.5 m/s. This ﬂow velocity range
achieves the initiation scour velocity for the typical distribution of sand particles
used for bedding. Achieving the complete scour velocity is desirable because it

assures that particles will not settle in the conveyance system.

31

2.4.2 Mechanical sand separation systems
Counter-current upﬂow separators (CCUS) and centrifugal separators

[hydrocyclones (HC)] are the two mechanical SSS technologies. Each is

discussed below.

2.4.2.1 Counter-current upﬂow separators
Counter-current upﬂow separators combine the four key steps of sand

separation into a compact continuous ﬂow machine. Concepts used in
sedimentation basins, aerated grit chambers, and hydrocyclones are all used
(Krou et al., 2006; Wedel and Bickert, 1996). Figure 2.3 shows the conﬁguration
of a column style CCUS. The sand manure separator (SMS) employed at GMF

and MDF and shown in Figure 2.4 is a commercial version of the CCUS.

32

Input

I n- 411
Overﬂow

H

i
Particles
I

 

 

 

 

 

Water
lili

Dilution =>
Water

Underflow

Figure 2.3: Diagram of column style counter-current upﬂow separator (Kim,
2003)

Reclaimed
Sand Auger Reclaimed

Sand
/ Outlet

Rear Auger
Weir Weir

 

 

 

Figure 2.4 McLanahan Sand Manure Separator (Inglis et al., 2006)

Counter-current upﬂow separators use a pool of fresh or recycled water to

dilute the SLDM input. Pool depth is determined by an overﬂow weir that is set
33

based on the input rate of dilution water and SLDM, generally speciﬁed by the
manufacturer. Dilution water is injected into all variations of the CCUS so that it
creates a rising current. SLDM is metered into the unit at a point near the
surface of the pool of dilution water. The difference in the elevation of the inputs
(dilution water and SLDM) results in the counter-current effect of settling sand
and a rising current of dilution water.

Figure 2.2 shows the forces exerted on a sand particle in the pool of a
CCUS. Under ideal conditions, the buoyant and drag forces are less than the
weight force for the smallest sand particle to be separated, but greater than the
gravity force for organic particles. Otherwise, settling will not occur and the sand
particle will be carried out of the CCUS with the overﬂow ﬂuid. Due to the particle
size, ﬁne sand, silt and clay are often washed out of the CCUS with organic
matter (Tables 2.5 and 2.6). The proportionally large surface area of the small
particles increases the drag force exerted on the grains. The opposite condition,
insufﬁcient buoyant and drag forces, will result in sand and organic particles
settling together.

Counter-current upﬂow separators operate with the four major types of
SLDM conveyance and capable of achieving Goals #1 and #2 (discussed in
Section 2.3.4). In most cases, CCUS require a secondary removal step to

achieve Goal #3, sand-free manure.

34

2.4.2.2 Centrifugal separations
Centrifugal separators, or hydrocyclones (HC), separate solid material

from slurries by centrifugal sedimentation. Hydrocyclones operate on the theory
that suspended particles subject to centrifugal acceleration force denser particles
to the cyclone wall where they settle by gravity (Svarovsky, 1990). Classiﬁcation
of sand in the mining industry is a common application of HC technology.
Hydrocyclones consist of a cylindrical body and a cone section, as shown
in Figure 2.5. The cylinder section includes the inlet, which introduces ﬂow to
enter tangentially. Tangential entry creates a swirling action inside the cylinder.
The swirling action continues and the velocity increases as the slurry moves
down the cone section, shown in Figure 2.6. Dense, coarse material exits the
cone section through the concentrated suspension outlet, also known as
underﬂow. Liquid is siphoned up the center of the cone and cylinder, exiting the
HC through the diluted suspension outlet, commonly referred to as the vortex

ﬁnder or overﬂow.

35

 

 

 

I ovenrtowj
'Wi
—%\

 

 

 

 

I UNDERFLOW I

Figure 2.5: Typical centrifugal separator cross-section

 

Figure 2.6: Hydrocyclone ﬂow pattern (Metcalf and Eddy, 1991)

36

To initiate the swirling ﬂow pattern in the HC, the inlet is connected at a
right angle to the cylinder. Liquid manure is pumped into the HC inlet. As the
ﬂuid is accelerated inside the cyclone, particles are subjected to three forces:
external and internal acceleration and drag, due to ﬂow (Svarovsky, 1990),
Figure 2.7. The effect of gravity in HC is generally neglected. Velocity is
greatest near the center of the HC, below the overﬂow, and decreases
proportionally as the radius increases until the cone radius is less than the
overﬂow radius. When the overﬂow radius exceeds the cone radius, a siphon is

created at the core drawing liquid and ﬁne solids out the overﬂow.

 

 

DRAG BUOYANT
FORCE FORCE
DRAG ACCELERATION
FORCE
WEIGHT
l (GRAVITY)
I

 

Figure 2.7: Force balance about a particle settling in a centrifugal separator

Flow velocity in the HC can be resolved to three components: axial,
tangential and radial (Svarovsky, 1990). Axial velocity results in the downward
ﬂow along the outer wall of HC and an upward ﬂow near the core of the cyclone.

Because of axial ﬂow, HC are created with an underﬂow oriﬁce pointed

37

downward. Sand moving downward is collected near the bottom and discharge
through the underﬂow at the bottom of the cone. The tangential velocity is
responsible for the movement of dense particles to the cylinder and cone walls.
Radial velocity, the weakest velocity component, occurs near the outer wall of the
HC and is directed inward, its magnitude decreases with decreasing radius.

The necessity of the pump to initiate ﬂow and create pressure in the HC
limits the applicability. Hydrocyclones can achieve goals 1 and 2 for sand
separation. However, due to the inﬂuent pump are most often used as a second

level of sand separation to achieve sand-free manure, goal 3.

2.4.3 Passive sand separation systems
Passive gravity separators (PGS) are adapted from grit (type)

sedimentation tank design used in municipal and industrial wastewater treatment
(Wedel, 1995). These systems operate in the realm of discrete settling where
particles settle individually with minimal interaction with other particles (Metcalf
and Eddy, 1991).

Early PGS systems used in municipal wastewater treatment were
designed to remove a speciﬁc percentage of grit based on the tank overﬂow rate
based on isoremoval plots (Swamee and Tyagi, 1996 ; Jin, et al., 2000). The
discrete settling conditions typically modeled in wastewater treatment are caused
by low total solids concentration (TS). Common PGS used to for separating
sand from dairy manure include the settling basin and sand lane. Figure 2.8

. shows the basic layout of a PGS used to settle sand from manure.

38

 

 
   
 

Settling Zone

Figure 2.8: Typical settling basin conﬁguration

Q = ﬂow rate, m3/s

Passive gravity separators are generally coupled with a hydraulic
conveyance system (Fulhage, 2003). Hydraulic (ﬂush) conveyance creates a
condition similar to grit sedimentation tanks by using large quantities of dilution
water.

The success of PGS depends on the ability to slow the ﬂuid velocity to
between 0.3 to 0.6 m/s (Harner et al., 2003). According to Shields’ equation
(Equation 2.4), that velocity range will allow settling of some sand particles, but
some particles with a diameter of 2.3 mm or less may pass through the system.
In theory, manure particles will remain suspended in that velocity range.

Forces exerted on a particle in a PGS include both vertical and horizontal.
Vertical forces are buoyant, drag, and weight, similar to the CCUS. The
difference in the vertical force of the PGS, compared with the CCUS is the
magnitude of the buoyant forces, as there is no rising current. Two horizontal
forces are exerted in a PGS, ﬂow (momentum) and drag (friction), Figure 2.9.

The ﬂow force is created by the momentum of hydraulic conveyance system.

39

DRAG BUOYANT

 

 

FORCE FORCE
DRAG FLow
FORCE
wEIGHT
(GRAVITY)
l P

 

Figure 2.9: Force balance about a particle settling in a passive gravity
separator

Settling basins are relatively deep storages where settled material
accumulates for long periods (weeks to months). In agriculture, settling basins
are designed to accumulate solids to a predetermined level. The time required to
reach that level is the accumulation period. To dissipate the hydraulic
conveyance energy, settling basins operate full of ﬂuid. The sudden decrease in
velocity of the ﬂush water allows material entering the settling basin to settle by
gravity. Gravity settling, as discussed in Section 2.3.5, is inﬂuenced by particle
size and density. Because sand and manure solids accumulate over time in a
settling basin, sand removed is generally not clean enough for reuse as bedding.
Settling basins are designed based on the settling velocity of the smallest particle
to be separated (Wedel and Bickert, 1996). Using the settling velocity (Equation
2.3) and ﬂow rate of the conveyance system, the surface area of a settling basin

is determined using Equations 2.8.

40

A = 9- (2.3)

A = plan view area of basin, m2

Flow rate of the conveyance system is used because that is assumed to be the
settling basin inlet ﬂow rate. The depth of the settling basin can then be
determined using the conveyance ﬂow rate, basin width and scour velocity in

Equation 2.9.

 

_ Q
D — Vmw (2.9)

D = chamber depth, m

W = basin width, m

F ulhage (2003) reported that settled solids accumulated at a rate of 0.06 to 0.07
m3/cow/d. Sand accounts for approximately for a quarter of the accumulation or
0.015 m3lcow/d.

Sand lanes were developed to separate sand of sufﬁcient quality for reuse
while removing enough sand to minimize downstream problems. Unlike settling
basins, sand lanes are shallow and drain completely between manure ﬂushes.

The design of sand lanes is such that the ﬂow from the conveyance device is

41

dissipated quickly, creating a shallow even ﬂow over the entire width of the lane.
Settling and scour velocities, Equations 2.3 and 2.7, of the largest particle size to
be removed should serve as the design parameters for controlling ﬂow rate in the
PGS (Harner et al., 2003). Equations 2.8 and 2.9 are used to determine the
dimensions (Wedel and Bickert, 1996). Sand lanes are sometimes constructed
with a gradual slope to facilitate drainage. Sand is removed from PGS manually
using a loader. Dilution water volume and sand excavation vary from farm to

farm depending on management.

2.4.4 Integration of technologies to create a sand separation system
Determining the goals of sand separation (Section 2.3.4) and the preferred

manner for manure collection and conveyance (Section 2.4.1) limit sand
separation options. Similarly, not all sand separation technologies can achieve
each goal for sand separation. Figure 2.10 is a simple decision ﬂow diagram of

conveyance options with sand removal technologies and sand separation goals.

42

Mechanical Hydraulic
220191939! (Scrape 8: Vacuum-Scrape) (Flush & Scrape-F lush)

 

5.5mm Counter-Current Passive
193110250591 Upﬂow """ 9 Gravity ""9 °°""'"“9"
\ / r
\ / '
/ I
w 9
Walter: Goals 1 8r 2 Goal 3

1Goal 3 of sand separation requires a multiple step sand separation system, denobd by dashed line

Figure 2.10: Integration of conveyance and sand separation technologies

Both CCUS and PSG can achieve goals 1 and 2 with a single treatment
unit or level. To achieve sand separation goal 3, sand-free manure, multiple
sand separation technologies are generally integrated in series into a system,
similar to the multi-Ievel system at GMF. Multiple technologies increase the
range of sand particle which can be effectively removed, compared to an
individual separation unit.

Compared to dairy farms using organic bedding, sand bedded dairy farms
require several additional pretreatment steps to remove sufﬁcient sand such that
downstream processes, anaerobic digestion, are not negatively impacted (goal
3). As discussed previously, manure with organic bedding requires little to no
pretreatment. For organically bedded dairy farms using AD, pump conveyance is
the simplest pretreatment or pre AD management system. In comparison, sand
bedded dairy farms require a minimum of two level of sand separation to ensure

43

that sedimentation will not negatively impact AD performance. Short—term
storage and pumping equipment are typically associated with each level of the
SSS, providing opportunities for changes in the manure characteristics due to

aging and aeration.

2.5 Determination of separation efﬁciency

Sand separation efﬁciency is the quantitative technique used to determine
the effectiveness of the SSS. Mass balance techniques are used to achieve the
ﬁrst research objective, determination of sand removal efﬁciency by the SSS.

Svarovsky (1990) shows the mass balance of solid liquid separation device by

Figure 2.11.

 

Feed—b —> Overﬂow

M, dF(x), O M,, dF,(x)/dx, O
Separator

 

 

 

Underﬂow
MC, ch(x)/dx, U

Figure 2.11: Schematic diagram of a separator (Svarovsky, 1990)

M = mass of particles contained in the inﬂuent

Me = mass of the coarse particles contained in the underﬂow
44

M, = mass of the ﬁne particles contained in the overﬂow
dF(x) = mass ﬂow of the inﬂuent

dF(x)/dx = mass ﬂow of the ﬁne efﬂuent

ch(x)Idx = mass ﬂow of the coarse efﬂuent

Q = inﬂuent ﬂow rate

0 = overﬂow ﬂow rate

U = underﬂow ﬂow rate
Assuming that material does not accumulate in a separation device, the
mass of the sand contained in the inﬂuent must equal the total mass of the sand

contained in the system products, overﬂow and underﬂow (Svarovsky, 1990).

The governing mass balance equation is shown as Equation 2.10.
M = MC + Mf (2.10)

Overall separation efﬁciency is described as the ratio of the mass of coarse

particles removed to the mass of the feed in Equation 2.11.
Mc
ET = -M— * 100 (2.11)

ET = overall separation efﬁciency, %

45

The overall separation efﬁciency can also be calculated using the mass ﬂow of

the ﬁne particles contained in the efﬂuent as shown in Equation 2.12.

ET = 1 ig- =1 100 (2.12)

Mass balance closure is the comparison of the mass of the inﬂuent
material to the mass of the efﬂuent of a system or unit. Manure and sand in
theory do not accumulate in SSS operating at equilibrium. Ideally the mass of
inﬂuents and efﬂuents would sum to 100%, however do to the dynamic operation
of commercial SSS closures to within i10%systems were deemed acceptable.
Percent difference, the technique identiﬁed by Gooch (2007) for determining the

accuracy of mass balance closure, is shown in Equation 2.11 (Gooch, 2007).

 

PD = (zinfluents'zeffluents) * 100 (2.13)

2lnt‘luents

PD = Percent difference
2m...“ = sum of all material entering the device

Zemuent, = sum of all the material entering the device

2.6 Anaerobic digestion
Anaerobic digestion is the biological conversion or degradation of biomass

into biogas and digestate (slurry exiting the digester). Digestion technology, an
46

advance treatment technology used around the world, has been in use on
livestock farms in the United States for over thirty-ﬁve years. The ﬁrst known
farm application was on a swine farm in Iowa in 1972 (Lusk, 1995).

During anaerobic digestion, carbon based material (biomass) is degraded
biologically in an atmosphere devoid of oxygen (Bracmort, et al., 2008) by
multiple microbial communities in a symbiotic relationship. Included are the acid
forming (hydrolytic, fennentatative, acidogenic) and methane-fanning
(methanogenic) (Rozdilsky, 1997) microorganisms. Figure 2.12 shows the

multiple step process and products.

Acid-Forrning Methane-Forming

 

 

Organic . > Simple . > Biogas
Matter Bactena Organic Bactena
Acids
Carbohydrates Acetic Methane
Proteins Propionic Carbon Dioxide
Fats Butyric WaterVapor
Fonnic Ammonia

Hydrogen Sulﬁde

Figure 2.12: Biology of anaerobic digestion (Barker, 2001)

Biogas, considered a low-grade form of natural gas, is a mixture of
methane (CH4), carbon dioxide (002) and other trace gases including hydrogen
sulﬁde (H28). The energy density of biogas ranges from 16,750 to 23,450 kJ/m3
(MWPS, 2000). Digestate, AD efﬂuent, is a mixture of undigested and partially

digested biomass and water.

47

Anaerobic digestion is beneﬁcial to livestock producers for numerous
reasons including the stabilization of waste, reduced odor and pathogens,
decreased emissions and the production of renewable energy (Wright et al.,
2003; US Environmental Protection Agency, 2002). Wright et al. (2003) reported
a 3-log reduction in fecal coliform due to anaerobic digestion.

There are several common designs of anaerobic digesters including
covered lagoons, plug ﬂow, complete mixed and ﬁxed ﬁlm (Wilkie, 2005). The
digester design for an individual farm is dependent on the solids concentration of
the slurry, space constraints and the farm management preference. Figure 2.13
is a decision support aid to identify appropriate conditions for anaerobic

digestion.

TotalSolid8(%)
--.-.- 0 .. 5 10 . 15 ., 20 25 30

Manure E Water Added 3 I: Bow
E: As Excreted :3

Classriicalron[ Liquid sruny _ samrsoud C3000 >

WOW E: Pump I Scrape__ I ScrapeSStack >

m- “...: ....qu

WW I: RecOmmended I Not Recommended -_>
m WW I l

DWTW. §Covered Complete Plug
2.1-359°" Mix ... Flew

 

 

 

Figure 2.13: Appropriate manure characteristics for anaerobic digestion
(US. Environmental Protection Agency, 2002)

48

2.6.1 Biogas potential of dairy manure
Prediction of biogas production from AD systems is normally based on the

VS mass or the chemical oxygen demand of an organic material. Before
estimating biogas production, it is important to consider the digestibility of the
biomass. One method to determine the potential to produce biogas, anaerobic
digestibility, is the biochemical methane potential test, also known as serum
bottles (Chynoweth et al., 1993; Owen et al., 1979). Serum bottle tests identify
unexpected results associated with site-speciﬁc constituents. Serum bottles use
relatively small quantities of sample, less than 250 ml, to predict anaerobic
digestibility and total biogas potential (Owen, et al., 1979). Using serum bottle
techniques, for manure, biogas production, has been reported to be in the range
of 0.18 and 0.39 m3/kg of VS destroyed (US Department of Agriculture, 2007;
Steffen et al., 1998; Morris, 1979). To address concerns with sample size and
data collection, anaerobic respirometer techniques have been developed which
use large samples sizes and automated gas measurement (Szczegielniak,
2008). However, sources of variability during the laboratory prediction of biogas
potential still occur, due to a number of factors including nutrient limitation,
bacterial acclimation, feedstock characteristics (VS) and experimental or
sampling error.

While the biogas production determined during the BMP does give an
indication of the biogas potential, it is not intended to be used for design and
equipment selection. It is recommend that pilot-scale test or actual operational
data be determined prior to the sizing and selection of biogas utilization

49

equipment. Using the range of predicted biogas production values, the biogas -
yield from a commercial system can be predicted using equation 2.14 (U.S.

Department of Agriculture, 2007).

Biogas = VS * VSConvm-on * MC * BP (2.14)

Biogas = volume of biogas produced, m3

VS = mass of volatile solids, kg

VSCOM....°.. = volatile solids destroyed in AD, %
MC = manure collected, %

BP = biogas potential, malkg of VS destroyed
According to U.S. Department of Agriculture (2007), AD systems are

expected to produce 1.9 m3 of biogaslcowld. Energy potential can be

determined once the biogas production is known using equation 2.15.

Energy Potential = Biogas * Energy Density (2.15)

Energy Potential = theoretical energy available, kJ

Energy Density = 16,750 to 23,450 kJ/m3

50

The energy potential an be converted to the theoretical electrical energy
output using the conversions factor of 3,600 kJ per kilowatt-hour (Glover, 1995).
According to US. Department of Agriculture (2007), using the basic
stoichiometric calculation for chemical oxygen demand (COD), the manure from
a single cow can generate approximately 42,000 kJ/d or 11.6 kWh/d.
Stoichiometrically, for every kilogram of COD destroyed, 0.395 cubic meters of
methane are produced (Speece, 1996). Wright et al. (2003) provided a general
prediction that seven mature dairy cows are required to support one kilowatt of
generation capacity. However, actual biogas yield will vary based on a number
of site-speciﬁc inﬂuences including feedstock, management, toxic substances
and system design.

Operating data from existing AD indicates a signiﬁcant amount of
variability in the installed electrical generation capacity. Table 2.7 summarizes
the key operational data from a majority of the operating AD in the United States,

compiled by the United States Environmental Protection Agency (2009).

51

Table 2.7: US. Farm Based Anaerobic Digester and Electrical Generation

 

 

 

 

 

 

 

 

 

 

 

 

 

Capacity
Total Mean Installed Generation Capacity
Anaerobic Digester Number Number
Type of of Number Mean 32:350.:
Systems Animals of
Systems (kW) (animalsIkW) (animals/kW)
Complete Mix 26 1,628 22 415 4.7 2.6
Covered Lagoon 10 1,778 8 247 9.2 6.3
Fixed Film 1 250 1 30 8.3
Horizontal Plug Flow 32 1,621 30 330 7.2 3.8
Induced Blanket Reactor 2 775 2 100 7.5
Mixed Ply Flow 33 2,878 27 589 4.2 1.1

 

 

 

 

Table F1 in Appendix G, evaluates the performance of AD, including the

bedding material and feedstocks. Similar to laboratory data, ﬁeld data also

indicated that biogas production and the subsequent utilization was highly

variable, often with the standard deviation exceeding 50% of the mean for

installed generation capacity per animal. The variability of biogas data available

from both laboratory experiments and operating commercial systems indicated

how important it is that system planners understand site-speciﬁc characteristics

to deal with uncertainty during the design of the biogas utilization system.

2.6.2 Mixing

Mixing in an AD is important for introducing new substrate to the viable

bacterial populations, heat transfer, reducing particle size and for releasing

biogas from the slurry (Karim et al., 2005a). For digester using SSS efﬂuent,

mixing is also needed to minimize the settling of residual sand in the digester

tanks.

 

Mixing options for anaerobic digesters included mechanical mixers, slurry
recirculation or biogas recirculation (Karim et al., 2005a). Mechanical mixers
have been identiﬁed as being the most efﬁcient, however servicing of internal
mechanical mixing systems in closed digester vessels is problematic (Brade and
Noone, 1981). Slurry and biogas recirculation mixing uses external components
to recycle material for mixing, simplifying maintenance and operation. Several
resources have identiﬁed recirculation as the most efﬁcient mode of mixing AD
(Karim, et al., 2005a). Design of a mixing system should maximize biogas
production while minimize the parasitic energy load of the mixing system and grit
accumulation. The mixing pattern, intensity and duration are believed to impact
biogas production but the body of literature is contradictory (Karim et al. 2005b).

Traditionally, acceleration, force and power equations have been used to
determine the bulk mixing energy requirements of anaerobic digesters (Smith,
2008). If the mixing time and scour velocity of the particle of interest are known,
Newton’s Second Law can be applied to determine the required acceleration,

Equation 2.16.

Acc = VH * t (2-16)

Acc = acceleration, m/s2

t = time, s

53

act

de

Using Equation 2.16 to determine the acceleration, the required force to

achieve the acceleration can be determined by Equation 2.17.

F = Acc * m (2.17)

F = force, N

m = mass of the material being accelerated, kg

Once the force is known, the mixer power required to achieve the force is

determined by Equation 2.18.

P=F*m mm)

P = power, kW

In recent years, computational ﬂuid dynamics (CF D) software has been
used to improve mixing systems design, predict the overall ﬂow pattern, location
of circulation cells and stagnant regions, trends of liquid velocity proﬁles and
volume of dead zones (Vesvikar and Al-Dahhan, 2004). Dead zones are deﬁned
as an area where the velocity was less than 5% of the maximum tank velocity
(Wu and Chen 2007) and can reduce the effective volume of a digester tank by

70% (Wu and Chen, 2007). The 3-Dimensional Multiphase CFD model prepared

54

by Vesvikar and Al-Dahhan (2005) indicated that the volume of dead zones in
typical an AD ranged from 11% to 60%, depending on mixer and tank
conﬁguration. For their research, dead zones were deﬁned as having a ﬂuid
velocity less than 5% of the maximum, 17-27 cm/s. Wu and Chen (2007) found
that increasing viscosity (total solids concentration) decreased high velocity
zones while having little impact on the percentage of low velocity zones that lead
to dead zones.

Even with improved modeling techniques, the only ﬁrm recommendation
concerning the power input for anaerobic digester agitation was made in an EPA
manual published in 1979. The US. Environmental Protection Agency (1979)
manual suggested a mixing power input n the range of 5.3 to 7.9 kW/1000 m3 for
anaerobic digester tanks.

Karim et al. (2005a) tested six biogas recirculation mixing regimes, which
the varied the recirculation rate from 0 to 3 L/min and the draft tube height from
the tank bottom from 13 to 40 mm. No signiﬁcant difference in biogas production
was identiﬁed. The low solids content of the substrate and the long retention
time in the AD were cited as causes for results. In a follow up study, Karim et al.
(2005b) conﬁrmed that mixing did not improve gas production for dilute
feedstocks (<5% TS). However, increasing the feedstock TS concentration to
10% did produce differences in biogas production based on mixing and mixer
type. Biogas production improved by 15% to 29% for mixed digester with high
solids compared to unmixed conditions (Karim et al., 2005b). Hoffman, et al.

(2008) found that mixing intensity had no impact on biogas production, while
55

operating four continuously stirred digester with mixing intensity ranging from 50
to 1,500 revolutions per minute. The contradictory data on mixing and biogas
production demonstrates how much uncertainty exists regarding AD mixing and

the impact of performance.

2.6.3 Heating requirements
Anaerobic digesters typically operate in one of two temperature ranges,

mesophilic (35°C to 41°C) or thermophilic (52°C to 57°C) (Pennsylvania State
University, 2009). The elevated operating temperature is intended provide the
optimum environment for the microbial consortium responsible for the anaerobic
degradation. Dilution water added to the manure stream during SSS inﬂuences
the design of the heating system. Additional mass of dilution water increases the
energy needed to achieve the target operating temperature of a digester (Inglis,
2006). The formula for used to determine the heat requirement of digester

inﬂuent is Equation 2.19.

H = CW * MW * AT (2.19)

H = total heat, kJ
CM = speciﬁc heat of water
M... = mass of water, kg

AT = temperature difference, °C

56

ter

As shown in Equation 2.20, Equation 2.19 can be expanded by adding

terms to account for other components in the slurry, such as sand.

H = (an .. MW ... AT) + (Cm * M, at AT) (2.20)

C” = speciﬁc heat of sand

M. = mass of sand, kg

The speciﬁc heat of water and sand is 4.18 kJ/kg °C and 0.76 kJ/kg °C,

respectively (Inglis, Gooch and Timmons, 2006).

2.6.4 Sand bedding and anaerobic digestion
Currently, there are only three anaerobic digesters operating on sand

bedded dairy farms. Green Meadow Farms (GMF) near Elsie, Ml operates
complete mixed anaerobic digester with SSS effluent serving as the feedstock.
The other systems are at the Fair Oaks Dairy, near Fair Oaks, IN and
Bridgewater Dairy near Bridgewater, OH. Anaerobic digestion technology has
not been more widely deployed on dairy farms using sand bedding due to the
history of system failures caused by grit accumulation and clogging (Rozdilsky,
1997; Wilkie, 2005; US. Department of Agriculture, 2007). A lack of data on the
efﬁciency of SSS has fueled a debate on the cost/beneﬁts of AD use on dairy
farms using sand bedding (Gooch and Inglis, 2007; Inglis, 2006). Important

issues include the possibility of sand settling and rapidly ﬁlling tanks, excessive
57

wear and tear of equipment and reduced energy production due to the need to
heat the dilution water required to remove sand from the manure, as discussed in

the subsections below.

58

CHAPTER 3: EXPERIMENTAL METHODS

This Chapter summarizes the methods and procedures used to conduct

the research associated with the objectives outlined in Chapter 1.

3.1 Solids Analysis and characterization procedures
Table 3.1 summarizes the standard procedures used for the

characterization of TS, FS, V8 and PSD of sand and manure samples with the
following modiﬁcation. Weights were taken on a “hot basis” instead of “cold
basis.” To measure on a “hot basis,” weights were taken immediately after
removing the sample from the oven at 105°C; this technique was preferable
because of the elimination of the potential for a faulty desiccant allowing moisture
to accumulate and alter weights (Wedel, 1995).

Total solids are the sum of dissolved and insoluble organic and inorganic
solids contained in the sample. Percent total solids was deﬁned as the ratio of
the mass of the dried sample to the mass of the original (wet) sample and was

calculated using Equation 3.1.

TS = 4°11— * 100 (3.1)

msample

T8 = total solids, %

mm = sample mass after drying, g

59

mum... = initial (wet) sample mass, 9

The ash remaining after ignition of the sample constituted the FS, or inorganic
material, Equation 3.2 was used to calculate the percent FS. Fixed solids are the
dissolved and insoluble inorganic material remaining after the sample has been

combusted at 550°C for at least one hour.

FS = 33$"- * 100 (3.2)

FS = ﬁxed solids, %

mm = sample mass after igniting, g

The portion of the sample vaporized during ignition is the VS; the percent VS are

calculated using Equation 3.3.

VS = W ... 100 (3.3)

mdry
VS = volatile solids, %

The percent total solids are equal to the sum of the ﬁxed solids (FS) and volatile

solids (VS), Equation 3.4.

60

 

 

 

 

 

 

 

TS = VS + F5 (3.4)

Table 3.1: Standard procedures for manure and sand characterization
Procedure Abbreviation Standard Source
Total solid* TS 2540-B APHA, 2008
Fixed solid* FS 2540-E APHA, 2008
Volatile solid* VS 2540-E APHA, 2008
Particle size distribution PSD D422-63 ASTM, 2002
*Modiﬁed to use hot weight measurement

 

 

 

 

 

3.2 Particle size distribution
Particle size results are often reported as the percent retained or passing

a speciﬁed sieve, using Equation 3.5.

 

”3 “Sn c
Percent retained = ( “mp" '“ ‘) (3.5)
”sample

Mum”. = mass, hot basis, of the initial sand sample

8", m = mass of sand sample accumulated on the nth sieve
The percent passing for a given sieve was calculated using Equation 3.6.

(M sample 'Xn,acc) (3 . 6)

 

Percent passing = M
sample

X", m = mass of total sand sample accumulated on the sieve of interest and all

61

larger sieves

3.3 Laboratory quality assurance
Samples for TS, FS, VS and PSD were refrigerated at 4 to 6°C until the

analysis was conducted and were tested within 48 hours of collection. Duplicate
analytical evaluations were conducted on all samples. Similarly, compromised
samples were re-tested when additional samples were available. Samples were
saved until analysis was completed and a preliminary evaluation of the data

performed.

3.4 Statistics analysis
Data generated from research at GMF and MDF was evaluated using

descriptive statistical techniques including mean, standard deviation, coefﬁcient
of variation, median and count. Coefﬁcient of variation was used to compare the
variability of data collected at different sample locations. Count refers to the
number of samples included in the data set. Data analysis tools in Microsoft
Excel were used to conduct the descriptive statistical analysis.

Total and volatile solids data from GMF was also evaluated to determine if
the management (Farm 2 or Farm 3), treatment level (SMS, MINI or H0) or the
machine (SMS 1-6 and MINI 1-6) contributed to the difference in PS and VS
results. The management effect consider differences caused by management
style, Farm 2 is operated as a production facility while Farm 3 management is
focused on treatment and special needs. Treatment evaluated the changes

caused by the different levels of the SSS, while machine considered differences
62

in the data attributed to the six different SMS and MINI that make of the SSS.
Data for the different sample locations (feces, SLDM, SMS and MINI); was
evaluated as individual samples for signiﬁcant changes in the FS and VS mass
due to management, treatment level and machine differences. To evaluate the
impact of the HC on F8 and VS, the data from the six different machines at the
MINI level was combined for each sample event or day. Similar to evaluation for
the individual sample locations, the combined data was also evaluated to identify
signiﬁcant changes in the FS and VS mass caused by management, treatment
level and machine. SAS 9.2 was used to perform the analysis of the statistical
signiﬁcance FS and VS data from GMF.

To statistically evaluate the data, normality was ﬁrst determined. Data that
was not normally distributed was transformed to ﬁt into a normal distribution. A
generalized linear mixed model (GLMM) in PROC GLIMMIX (SAS Inc., 2006)
was used to evaluate the normally distributed data. The mixed model contained
both ﬁxed and random effects. Fixed effect groups included management and
treatment, while the machine was considered a random effect.

Equation 3.7 was the statistical model used to evaluate the individual

samples (feces, SLDM, SMS and MINI) using FS and VS data.

y = u + mgt + machine(mgt) + trt + trt * mgt + machine a: trt + e (3.7)

y = response variable

u= overall mean
63

mgt (management) = ﬁxed factor about the management factor which had two
levels (lactating and special needs)

machine (mgt) = random factor about the sample location which had six levels
(1—6 shown in Figure 3.1). Management interacted with the
machine as machines 1—4 operated at Farm 2 and machines 5—
6 operated at Farm 3.

trt(treatment) = ﬁxed factor about the sample identiﬁcation which had four
levels for F8 (feces, SLDM, SMS and MINI) and three levels for
VS (SLDM, SMS and MINI)

a-t'mgt= interaction between trt and mgt

e= residual term

Due to its single input, the HC (average MINI) had only one ﬁxed effect,

treatment, modifying equation 3.7. The resulting statistical model is shown in

Equation 3.8.

y= u+trt+e (3.8)

trt = ﬁxed factor about the SSS averaged sample identiﬁcation which has two

levels (MINI and HC overﬂow)

3.5 Research farms
3.5.1 Farm descriptions
Two commercial dairy farms operating similar SSS were used to collect

data associated with the research objectives of this project, GMF and MDF.

3.5.1.1 Green Meadow Farms
GMF houses approximately 2,900 lactating and 300 dry cows at two

adjacent facilities, Farms 2 and 3. The three level SSS at GMF, installed 1998 to
2001, includes six SMS, each followed by a passive gravity-settling basin (MINI).
Four of the SMS are located at Farm 2 and two are located at Farm 3, as shown
Figure 3.1. Farm 2 houses only lactating cows while farm 3 maintains a
combination of lactating, dry and fresh cows, referred to as special needs
animals.

Feces, urine, sand bedding and water from the drinkers is collected from
the freestall barns using a skid loader equipped with a tire scraper. The loader
deposits SLDM into reception pits located at the end of each barn. Sand laden
dairy manure is metered from the reception pit into the adjacent SMS using an
auger or positive displacement pump. Efﬂuent recycled from the phosphorus
separation system is used for dilution water. Agitation is achieved by injecting
compressed air into the dilution water pool near the bases of the SMS and by the
turning motion of the sand removal auger (Figure 2.4). Settled sand in the SMS
is removed by the internal SMS screw conveyor and discharge on to a concrete

stacking pad. Liquid efﬂuent ﬂows by gravity from the separators into the MINI
65

pits adjoining each SMS. Solids accumulated in the MINI’s are excavated and
land applied daily.

It should be noted that the MINI’s are not a common component of the
mechanical SSS. The MINl’s were included because the system installed at
GMF was the ﬁrst of its kind and a level of uncertainty existed with the level of
sand separation efﬁciency. Installation of the settling basins provided a second
level of sand removal. An attempt to remove the MINI’s from the SSS in 2007
resulted in a transfer line from Farm 3 to the transfer station becoming clogged.
It is believed that the clog occurred because the residual sand in the SMS
efﬂuent and the fact that the transfer line is unlevel, allowing settling sand to pool
in the low spots. As a result of the failed attempt in 2007, the MlNl’s remain an
integral part of the SSS at GMF.

After the MINI, liquid manure is pumped to tank 2 (Figure 3.1) at the
transfer station, it mixes with liquid manure from the other SMS-MINI systems
prior to being pumped to the HC (Figures 3.1 and 3.2). Before entering the HC,
manure passes through a macerator to reduce all particles to 12.5 mm diameter
or smaller. After passing through the HC, this underﬂow is directed into MINI #5.
HC Overﬂow, sand-free liquid manure, is pumped into the AD. Digester efﬂuent
is pumped to the AD equalization tank before being transferred to the

phosphorus separation system.

66

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

E E FAR 2 E E

0 (U (U n:

‘2 ‘2 E ‘2

1% g g Milking g g NM
u. u. Center u. u. ~SMS - Sand Manure Separator
I e I— e _ — e -- e l I — e ‘ °N0tdl’awn b sca'e

MS 1 SMS 2 Treatment [ISMs Msms 4

f 4 Barn 9 4

 

Luann

- - '> Platform Flush

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

P
—-> Transfer Station : ' I
........ e suslnﬂuent E - I E E
I I g “a '3: Milk' 8 8
- - -> SMS Effluent ' = g'z C It": 3 =
— - °> ManureTreatmenl Pipeline | i g m an r g g
Anaerobic
Digester

Figure 3.1: GMF SSS schematic

Figure 3.2 shows the general SSS process ﬂow for GMF. Sample points
in Figure 3.2 are identiﬁed by numbers. The sample numbers in Table 3.2

correspond with Figure 3.2 and offer a description of the sample and collection

location.

67

Un’ne 8. water
from drinkers

1. New Sand

3. Sand-Laden I

 

  

 

 

 

 

Water

 

l Sand Manure -

 

 

 

 

 

 

 

 

 

 

 

Parlor wastewater

Recycled / 8. belt press effluent

0"” Manure Racgftuon .——p Separator - J- - — - - - - - -|- - - '1
(SLDM) |_ - I (SMS) ; 5 R 7 . ed
. . . ecarm
1 I l e — a — I I - I - I SBDd
2. F . 4. SMS _
i Settling Basln ------------ :
(MINI) , . . _ i MINI
- 6. MlNll ' SM”
Main FIW E..............V ........... ..E I
- - - - Product "3"??ng — 4 -> 3. HC Underﬂow '
— - - - . SMS Boundary ' : !
...... .'... HC Boundary ..IIIIIIIIIOI use-essence... i
— . - - SSS Boundary - . C - . . — U . . — . . — . . - — ' .
7. HC Overﬂow

Figure 3.2: SSS process ﬂow diagram - GMF

Table 3.2: SSS sample locations and description - GMF

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sample it Sample ID Description Sample Location
1 New Sand New bedding sand Sand stockpile
2 Feces Feces only Freestall alley
3 SLDM Manure, urine, water and sand in a slurry Metering system discharge_
4 SMS Effluent from SMS SMS liquid outlet
5 Reclaimed Sand Reclaimed sand from SMS SMS sand dischage
6 MINI Efﬂuent from MINI MINI overﬂow efﬂuent
7 HC Overﬂow Fine effluent (overflow) from the HO HO overﬂow efﬂuent
8 HC Underﬂow Coarse efﬂuent @nderﬂow) from the HO HO underflow effluent

 

3.5.1.2 Minnis Dairy Farm
MDF houses approximately 600 mature dairy cows. The two-level SSS

was installed in 2005 and consists of a SMS and HC, operated in series (Figure

3.3). MDF scrapes manure from the freestall alleys using a skid loader to two

reception pits. A piston pump in each reception pit meters manure in to the SMS.

68

The piston pump operates like a syringe feeding material into the SMS during the
down stroke and ﬁlling the pump chamber on the up stroke. Milking center
wastewater and efﬂuent from the solid-liquid separator are used as dilution water.
Efﬂuent from the SMS ﬂows by gravity into a sump from which it is pumped to the
HC. Reclaimed sand is removed from the SMS by the built-in screw conveyor.
Periodically, the low limit switch is triggered, deactivating the feed pump and
consequently, turning off the HC. Overﬂow from the HC ﬂows into a reception
and then pumped to the solid-liquid separator. Hydrocyclone underﬂow is
discharged into the SMS dilution pool. The solid-liquid separator removes coarse
manure solids and a portion of the liquid efﬂuent is used as dilution water for the
SMS. Excess efﬂuent is transferred to the long-term manure storage. Because
this SMS uses a closed-loop conﬁguration, the dilution water has a relatively high
solids concentration and larger ratios, greater than a 1-part SMS to 1-part
dilution water, is required. Typically, the SSS at MDF operates for eight to twelve
hours each day. Table 3.3 summarizes the sample collection location and

description; the numbers correspond with those shown in Figure 3.3.

69

Urine 8. water

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

. Parlor
from drinkers 2. Recycled
Sand Water 4--( thstewater
— I I — I — I I — I I \\
: - I - i 1. \
I 3. Sand-Laden I Sand Manure ' I- X
j 3 Dairy Manure —7—> Separator - + - - - \
I (SLDM) I I l (SMS) ) I v \
, ’ (_ . _ ' I 4. Reclaimed \
' I, ssM-s -.-.-' sand '
Manure I I a eeeeee .aaaaaaaa sssssssssss so. . I
7. HC Underflow 4 - -:- - V '°°V° °"° E Z I
I 5 (HC) ‘ 5 I r
I :IIIIII IIIIIIIIIIIII IIIIIII; . '
L 6. HC Overﬂ - l
I I — I I _ I I — I I — I I — I — I I d '
. I
— Marn Flow Solid-liquid _ _ _ d. Separated (
Separator Liquid
- - - Product I \
- - - SMS Boundary | i
........ HC Boundary I Liquid
Se arated
— - - SSS Boundary golids Storage
Figure 3.3 SSS process ﬂow diagram - MDF
Table 3.3: SSS sample locations and description - MDF
Sample # Sample ID Description Sample Location
1 Fresh Water Fresh water used to rinse reclaimed sand Spray barn on SMS
2 Recycle Water Dilution water for SMS Inlet to SMS
3 SLDM Manure, urine, water and sand in a slmy Meterinﬁystem discharg_e_
4 Reclaimed Sand Reclaimed sand from SMS SMS sand discharge
5 SMS Effluent from SMS SMS liquid outlet
6 HC Overﬂow Fine efﬂuent (overﬂow) from the HO HO overﬂow efﬂuent
7 HC Underflow Coarse efﬂuent (mderﬂow) from the HO HO underliow effluent

 

 

 

 

 

 

3.5.2 Manure sample collection

Fixed and volatile solid concentration results, described in Section 3.1, of

samples from the SSS at GMF and MDF were used to estimate the FS

separation efﬁciency and the change of VS attributed to sand removal. Solids

data for the separation efﬁciency was based on manure and sand samples

70

collected from the SSS at GMF and MDF by personnel from MSU between May
of 2006 and May of 2007. Samples were collected by three employees of the
Biosystems and Agricultural Engineering Department.

Feces samples at both dairy farms were collected as random samples
from the freestall barn alleys. The intent of the feces samples was to
characterize raw manure with no bedding. Sand-laden dairy manure samples
were collected from the reception pit where scrape manure was deposited. At
both GMF and MDF, SLDM manure samples were collected from the outlet of the
metering device feeding manure from the reception pit into the SMS. Samples
were collected when the pits were 50 to 100% full.

Efﬂuent samples from the different levels, SMS, MINI and HC, were
collected directly from the discharge, efﬂuent, of each machine. Sand separation
systems samples were collected when the machines were fully operational with
discharge from both the coarse and ﬁne ﬂows. Operational status was
determined by visual observation of the person conducting the sampling.

Sample volume or mass was generally four-times greater than what was
required for laboratory analysis. Sub-samples from the farm sample were sued
for analysis.

Differences in the layout and operation of the SSS at each case-study
farm resulted in the farm-speciﬁc sampling protocols described in the following

sections.

71

3.5.2.1 GMF sample collection
Samples were collected and analyzed from the locations shown in Figure

3.2 and described in Table 3.2 between May of 2006 and January of 2009.
Analysis of the samples included solids characterization and particle size
distribution of sand samples.

At GMF, grab samples for solids analysis were collected of feces, SLDM
and from the efﬂuent of each machine. To avoid contamination with bedding
sand, feces samples were collected from freshly excreted manure patties in the
freestall barn alleys. The sample protocol for feces excluded urine. Urine
accounts for 1/3 of the total manure excreted by a dairy cow, feces makes up the
balance (Nennich et al. 2006; Nennich et al., 2005). To account for the volume
and solids contributed by urine, it is assumed to have a moisture content of
95.5% to 97%, with half of the TS contributed by F8 (American Society of
Agricultural and Biological Engineers, 2005; Bannink et al., 1999). Using
American Society of Agricultural and Biological Engineers (2005) manure
production values, the FS baseline was determined by multiplying two-thirds of
the manure production (feces) by the measured FS concentration and one-third
of the manure (urine) by the predicted urine FS concentration of 1.5%. Due to
the operational variability of the SSS, this estimation is adequate for the objective
of determining the impact of sand on AD.

Sand laden dairy manure samples at GMF were collected from the

scraped manure at the end of the freestall alley, just prior to the reception pit.

72

Feces, urine, bedding sand and water from the drinkers is mixed during the
scraping process, creating a homogenous sample at the end of the freestall alley.

Efﬂuent samples from the different levels and machines in the SSS were
collected when the units were observed to be operating at a steady-state
condition. \ﬁsual observation veriﬁed that all components were operating and
ﬂow was not obstructed.

Additional samples were collected from the new sand, reclaimed sand, HC
underﬂow and HC overﬂow (including tank sludge) for sand particle size analysis.
Samples, 500 gram, of reclaimed sand and HC underﬂow were collected as grab
samples in sample bags. For new and reclaimed sand samples, material was
collected from several locations on the sand pile. HC underﬁow sand samples
were collected from the discharge of the pipe carrying the full underﬂow stream.
HC overﬂow samples were collected three ways; by drying 10 gallons of HC
overﬂow liquid, from residue in the heat exchangers of the AD and from sludge
accumulated in the AD tank. Sludge accumulation in the MlNl’s was not sampled

due to the inability to collect representative samples.

3.5.2.2 MDF sample collection
Samples were collected from MDF between June of 2006 and May of

2007. At MDF, samples for solids analysis were collected from each location

shown in Figure 3.3 and described in Table 3.3.

73

Similar to GMF, 100 mL vials were used to collect samples for solids
evaluation. Sample collection occurred when operation was stable, as

determined by visual observation.

3.5.3 Flow measurement
At GMF, HC overﬂow was the only measured ﬂow rate. Flow meters

installed at the phosphorus separation system and the AD provided the ﬂow rate
of the HC overﬂow. Other ﬂow rates were unable to be measured due to a lack
of ﬂow meters and inaccessibility for direct measurement.

Manure production and bedding sand usage at GMF was estimated using
data from the American Society of Agricultural and Biological Engineers (2005);
Midwest Plan Services (2000). To estimate the efﬂuent ﬂow rates of the SMS
and MINI at GMF, it was assumed that the inﬂuent to a unit was 10% greater
than the ﬁne (liquid) efﬂuent stream. For example, sine the MINI efﬂuent is the
inﬂuent to the HC overﬂow, it was assumed that the MINI efﬂuent ﬂow rate was
110% of the HC overﬂow (ﬁne efﬂuent) ﬂow rate. The ﬂow rate change of 10%
was predicted using data from MDF, where the ﬁne efﬂuent (liquid) ﬂow rates of
the SMS and HC decreased by 9% to 12% compared to the input, the change is
the result of the coarse material separation.

At MDF, the ﬂow rate of each sample location was determined by direct
measurement using a 19 L bucket to collect the entire ﬂow for a measured period
of time. Direct measurement was not possible for all sample locations at GMF.

Table 3.4 summaries the source of ﬂow rate data for GMF. Similar to the sample

74

collection process for solids analysis, measurements were made when the

system operation was observed to be stable.

Table 3.4: Sample location data source - GMF

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sample Sample Parameter
if ID FS VS Density Flow Rate Particle Slze
1 New Sand Measured Measured Glover, 1995 MWSP, 2000 Measured
2 Feces Measured Measured ASABE, 2005 ASABE, 2005 -
Gooch & hglis, ASABE, 2005 &
3 SLDM Measured Measued 2007 MWPS, 2000 -
4 SMS Measured Measured ASABE, 2005 Apprixmatedz -
5 Reclaimed Sand Measured Measured Glover, 1995 - Meaered
6 MINI Measured Measured ASABE, 2005 Approximatedz -
7 HC Overﬂow Measured Measured - Measued1 Measu‘ed
8 HC Underﬂow Measured Measured - - Measued
1Measwe by the GMF staff usinginline flow meter
2Used HC Overﬂow flow rate and increased the ﬂow rate 10% for each sss level ___________

 

3.5.4 Density measurement

The densities of samples from each location at GMF were assumed to be

similar industry data presented in American Society of Agricultural and Biological

Engineers (2005), Midwest Plan Service (2000) and Gooch and Inglis (2007). At

MDF, the densities were determined by weighing the bucket containing the

known volume of sample used to determine ﬂow.

3.6 Fixed solid separation efﬁciency evaluation

The mass balance technique described in Section 2.5 provided the basis

for the FS separation efﬁciency at both MDF and GMF. The approach is

described below.

75

3.6.1 Mass balance approach
A traditional mass balance evaluation, as illustrated in Figure 2.8, was

completed at MDF since all of the SSS inputs and outputs could be measured
(solids concentration and density). This enabled the determination of separation
efﬁciency for each component of the MDF’s SSS including the SMS, HC and

SSS. Equation 2.10 was used to predict the overall separation efﬁciency.

3.6.2 Semi-empirical mass balance
Manure management, speciﬁcally sand separation, involves complex

systems with site-speciﬁc conditions. The semi-empirical mass balance was
intended to be a simple, yet robust, method for evaluating the performance of
such systems operating at commercial facilities. The GMF SSS is an example of
a complex SSS, consisting of several levels (SMS, MINI and HC) with multiple
machines at two levels with limited access to the input and output ﬂows of each
treatment level. Utilizing a semi-empirical approach, system planners can
efﬁciently gather important design data.

Consequently, a semi-empirical mass balance approach was developed.
This approach uses standard industry values from the American Society of
Agricultural and Biological Engineers (2005) and the Midwest Plan Service
(2000) to estimate manure production and bedding usage to predict the ﬂow rate
of manure and SLDM. Daily manure and bedding production, along with
measure FS concentration is used to establish the baseline mass of FS

contributed by the manure and sand bedding (SLDM). Flow rate and FS data

76

from the ﬁne outlet (liquid efﬂuent) of each SSS level were used to track the FS
change, the separation efﬁciency. Fixed solids remaining in the efﬂuent of the
ﬁnal level of the SSS were considered residual FS or residual sand. However, it
is important to understand that not all of the residual F8 are contributed by
bedding sand. Approximately 15%, of the manure as excreted from a dairy cow
is FS (American Society of Agricultural and Biological Engineers, 2005).
Regardless of source, FS entering the AD are inorganic and could contribute to
sedimentation and sludge accumulation, as they are not degraded biologically.

As described in Section 3.5.3, to estimate the ﬂow rate of the SMS and
MINI efﬂuents at GMF, effluent was assumed to be less than 10% of the inﬂuent.
Increasing the HC overﬂow ﬂow rate by 10% provided the MINI efﬂuent ﬂow rate.
Similarly, increasing the MINI efﬂuent ﬂow rate by 10% approximated the SMS
efﬂuent ﬂow rate. An inline ﬂow meter at the phosphorus separation system and
AD provided the ﬂow rate of the HC overﬂow.

Equation 2.10 was used to determine the cumulative SSS and unit
separation efﬁciency, based on the ﬁne material ﬂow (overﬂow or effluent) of both
the SMS and HC. Unit separation efﬁciency compared the change in the inﬂuent
and efﬂuent mass for each level of the SSS. Cumulative efﬁciency compared the
mass of FS contributed by sand bedding to the FS contained in the efﬂuent of

each SSS level.

77

3.7 Particle size distribution
Sand samples (grab samples from stockpiles) for particle size distribution

(PSD) were collected from GMF to achieve research objective two, determination
of the residual sand characteristics. Unused (new) sand samples were collected
from GMF and provided the baseline PSD to compare against sand samples
collected throughout the SSS and AD. Reclaimed sand was collected from the
discharge of SMS for evaluation. Hydrocyclone underﬂow samples were
collected at the discharge of the HC.

HC overﬂow samples were collected by two means. Large samples of
liquid HC overﬂow, 40 L, were collected and using a drying oven, the water was
evaporated leaving only the solid residuals. 'As the HC overﬂow sample dried,
residual sand accumulated near the bottom of the drying pan with a crust of
manure ﬁbers forming above the sand. Due to the inability to precisely separate
the sand and manure ﬁbers, this method was unreliable. During routine
maintenance, it was discovered that small quantities of sand accumulated on the
bottom of the heat pipes in the AD heat exchanger. The heat exchanger
receives HC overﬂow prior to entry into the AD system. Residual sand samples
were collected from the heat exchanger in January and May of 2008 when the
heat exchanger was taken ofﬂine for service. During the service, 3 to 5 mm of
sand had accumulated in the pipes.

Additional samples for PSD analysis were collected at GMF from the
efﬂuent equalization tank and AD tank #3 (tank sludge samples). These samples
represent the characteristics of sand accumulation downstream of the HC. A set

78

of samples were collected from the digester efﬂuent equalization tank in May of
2007. The farm had begun to use the equalization tank prior to completion of the
AD system. In October of 2008, AD tank #3 was taken ofﬂine and drained to
allow the farm to service mixers that had failed. Only 150 mm of liquid manure
was remaining. Prior to the mixer service, tank #3 was receiving half of the
manure production directly from the farms each day, tank #1 received the other
half of the daily manure production. Both tanks overﬂowed into tank #2.
Qualitative measurements indicated that a thin sand/sludge layer averaging
approximately 25 mm (1 in.) thick blanketed the tank bottom. Samples of the

material were collected for PSD analysis from two locations.

3.8 Volatile solids loss during sand separation
Volatile solids loss during the sand separation process was determined

using the semi-empirical technique developed for predicting FS separation
efﬁciency at GMF (Equation 2.10). Sand laden dairy manure, containing feces,
urine, bedding and water, was used as the baseline for the VS loss
determination. The contribution of VS from new bedding sand was assumed to

be negligible. Both the cumulative and unit separation VS loss were determined.

3.9 Anaerobic digester design considerations
Sand separation changes the composition of manure. Using data from

GMF, design implications of anaerobic digestion were evaluated, including

mixing, heating and biogas potential.

79

Bulk mixing power to achieve the initiation and complete scour velocity of
the mean sand particle size for each sample location were calculated using
Equations 2.12, 2.13 and 2.14. Each GMF digester tank has three mixers, 2 — 13
kW horizontal and 1 —10 kW vertical. Currently, GMF operates AD mixers
according to the supplier recommendations, ﬁve-minutes of mixing per hour.

Heat requirements for the daily manure mass for each sample location
were calculated using Equation 2.15 over a range of initial temperatures
(ambient) to with the typical target (ﬁnal) of 35°C.

Biogas potential was determined using Equations 2.16 and 2.17.
Equation 2.16 used the daily mass of VS from each sample location to predict
the gross biogas potential. Assuming biogas was approximately 60% methane

(NRCS, 2007), the energy potential was determined using Equation 2.17.

80

CHAPTER 4: RESULTS AND DISCUSSION
In order for advanced manure treatment systems like anaerobic digestion
to be successfully integrated on dairy farms using sand bedding, the impact of
sand must be understood. Included is the amount of sand that can be removed
using current sand separation technologies and the particle distribution of
residual sand in the efﬂuent. This research both modeled the impact and veriﬁed

using an actual AD, as discussed in the subsections below.

4.1 Technique for quantiﬁcation of sand separation efﬁciency
Grit accumulation and clogging has been cited as a leading cause of past

AD failures. Consequently, determining the SSS efﬁciency and efﬂuent
composition are crucial for the planning and design of downstream systems.
During planning for the AD at GMF, the lack of information on the efﬁciency of
SSS and the characteristics of the residual sand limited the ability to optimize the
design of the mixing system. Based on this lack of data, research objective 1 is
to quantify the efﬁciency of SSS. The generally accepted approach is by mass
balance (Svarovsky, 1990). Flow rate, density and solids data collected from
MDF was used to determine if the mass balance was a practical tool to meet
objective 1, measurement of the separation efﬁciency. However, measurement
of the mass ﬂow rate was not practical at GMF because of the inability to
measure all the ﬂow rates and operational variability, thus an semi-empirical
method was developed that used a combination of measured data and industry
standards. Both approaches are presented below.

81

Analyses of samples from the 888’s at MDF and GMF were evaluated for
TS, VS and FS over an eighteen-month period. More samples were collected
from the SSS at GMF than MDF due to the number of machines, six SMS and
MINI, in the system. Due to the design, construction and start-up of the AD, the
sampling period was also longer at GMF compared to MDF. The process ﬂow
diagram for both dairy farms was shown in Figures 3.2 and 3.3. The complete
characterization data sets are included in Appendices A and B for MDF and

GMF, respectively.

4.1.1 Total solids characteristics of sand separation system products
Determination of the TS is the ﬁrst step in the processes of quantifying the

FS and VS of a sample. The procedure for measuring solids was described in
Chapter 3. Table 4.1 summarizes the TS results from the SSS at both GMF and
MDF. To allow for comparison between data sources/treatment conditions,
solids concentration data is summarized as a percentage in the text, equivalent
to the grams of solids over the grams of the wet sample. Total solids

concentration was computed using equation 3.1.

82

Table 4.1: Total solid concentration - GMF and MDF

 

 

 

 

 

 

 

 

 

 

 

 

Mean Standard
Data Source Sample TS Deviation Median Count
Locauon
(%L (7.) (7.)
Feces 14.9 2.1 14.9 68
SLDM 28.3 11.0 30.3 69
GMF SMS 6.1 2.0 6.0 54
MINI 4.9 1.3 5.0 70
HO Overﬂow 4.1 1.2 4.4 40
Feces 15.2 5.0 16.9 6
SLDM 19.0 9.4 18.8 14
MDF SMS 4.8 2.2 4.0 14
HO Overﬂow 4.7 1.9 4.3 14

 

 

 

 

 

 

 

 

Average TS concentration of feces, as excreted, was similar for both GMF
and MDF. Feces samples collected for this research excluded urine and
bedding. Industry standards predict the TS of dairy cow manure, including urine,
is between 12 and 13% (Midwest Plan Service, 2000; American Society of
Agricultural and Biological Engineering, 2005) and consequently, the feces
samples for both farms listed in Table 4.1 have slightly higher TS concentration
due to the exclusion of urine.

Inclusion of bedding sand with manure resulted in an increase in T8
concentration of the SLDM sample compared to the feces. SLDM from GMF had
a much higher TS concentration than MDF indicating that more sand bedding
may have been used at GMF. New bedding was added to the freestalls at both
farms on a random schedule that ranged from days to weeks. The random
bedding schedule contributed to the large TS standard deviation for SLDM.

Visual observations indicated that the quantity of sand mixed with manure

83

peaked immediately following the addition of bedding and then diminished until
the next addition. The T8 concentration of the samples decreased throughout
the SSS (SMS, MINI, and the HC overﬂow), this change was caused by the
addition of dilution water and the removal of FS by the SSS components.

The count is the number of sampling events for each sample location.
Variability in the count is attributed to daily operational differences and the intent
of the sampling event (not all sample locations or treatment levels were

evaluated during each sample event).

4.1.2 Fixed solids characteristics of sand separation system products
Fixed solids results for the SSS at GMF and MDF are presented in Tables

4.2. As described in Chapter 3, the FS concentration of manure (feces + urine)
was used to establish a baseline for the determination of sand separation

efﬁciency. Equation 3.2 was used to calculate the FS concentration.

Table 4.2: Fixed solid concentration - GMF and MDF

 

 

 

 

 

 

 

 

 

 

 

 

Sample Mean Standard
Data Source Location FS Deviation Median Count
(°/.) (%) (%)
Feces 2.1 1.0 1.9 67
SLDM 20.0 9.7 21.6 67
GMF SMS 2.2 1 .1 2.0 53
MINI 1.5 0.4 1.4 70
HO Overﬂow 1.1 0.5 1.2 40
Feces 3.0 1.1 2.8 6
SLDM 12.1 7.0 10.6 14
MDF SMS 2.1 1.4 1.7 14
HO Overﬂow 1.6 1.0 1.3 14

 

 

 

 

 

 

 

84

 

Differences in feed ration and water intake are the likely causes of
variation in the FS concentration of feces between the two farms (Nennich, et al.
2006). Mean FS concentrations followed similar trends to TS with an increase in
F8 from manure to SLDM due to the addition of bedding sand and a decrease in
FS with each progressive step in the SSS. The decline in F8 concentration is
attributed to the addition of dilution water and the removal of sand in the SSS.
Similar to the TS, the high concentration of SLDM FS at GMF, compared to MDF,

is attributed to differences in the quantity of bedding sand used.

4.1.3 Mass balance separation efﬁciency
Using the mass balance, described by Svarovsky (1990) in Equations 2.10

through 2.12, and sample collection techniques, described by Gooch (2007), the
efﬁciency of the SSS at MDF was calculated during the spring of 2007 when the
SSS was believed to be operating at equilibrium. To make these calculations,
the ﬂow rates were ﬁrst determined. Included were calculations of closures to
assess the quality of the measurements. Based on the densities of removed
materials, the mass ﬂow rate and closures were then calculated. This then

allowed the efﬁciencies to be determined.

4.1.3.1 Flow rate determination
Table 4.3 summarizes the SSS measured ﬂow rate data from MDF. To

account for the intermittent operation of the piston pump feeding SLDM into the
SMS and the HC, ﬂow rates were normalized to one-hour increments shown in

Table 4.3 using the measured on time of the piston-pump and HC. Sample
85

events for the mass balance evaluation at MDF occurred on 03/05/07 (1) and

03/17/09 (2 & 3). Appendix A contains the complete data set for MDF, including

FS data for the sample events.

Mass balance calculations were completed using hourly ﬂow rate data

contained in Table 4.3. The percent difference was used to compare the volume

or mass of the inputs to the mass or volume of the efﬂuents so that the mass

balance closure can be determined, Equation 2.13. Input and efﬂuent stream for

the SMS, HC and SSS at MDF are shown in Figure 3.3. A mass balance closure

of 210% was established as the target based experiences with other biological

and mechanical processes (Schell, et al. 2002).

Table 4.3: Sand separation system ﬂow rate — MDF

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sample Sample Event Mean Standard Coe fﬂclen t
Point °°'°""°°" 1 2 3 OWN“ of Variation
(m’nrr) (m’lhr) (111% r) (mslhr) (ma/hr)
1 Fresh water 0.58 0.71 0.69 0.7 0.1 10%
2 Recycled water 13.37 13.36 16.07 14.3 1.6 11%
3 SLDM 2.75 2.89 3.12 2.9 0.2 6%
4 SSS reclaimed sand 0.18 0.22 0.39 0.3 0.1 42%
5 SMS liquid efﬂuent 15.33 14.08 19.07 16.2 2.6 16%
5a HC imut 24.23 23.72 26.43 24.8 1.4 6%
6 HC overﬂow 21.86 21.36 22.16 21.8 0.4 2%
7 HC mderﬂow 0.34 1.14 0.48 0.7 0.4 65%
SMS percent cloere: 9% 21% 4% 1 1% 9%
HC percent closue: 8% 5% 14% 9% 5% ___
SSS percent closue: -32% -27% -13% -24% 10% W __-_._.__-,
SMS dilution ratio: 4.9 4.6 5.2 4.9 0.3

 

 

 

 

 

 

 

 

tSLDM ﬂow rate based was based on the daily measured piston pump cycle time (down and Ip-stroke)
minus the up-stroke.

 

 

 

2Cyclone flow rate was based on a measured airtime of 46 minute per hour.

 

 

86

Flow rate measurements presented in Table 4.3 indicated the ﬁne efﬂuent
streams of the SMS (SMS liquid efﬂuent) and HO (HO overﬂow) were reduced by
9% and 12%, respectively, compared to the inﬂuent ﬂow rate.

Successful mass balance closures were achieved for two out three
attempts for each level of the SSS, SMS and MINI. The successful mass
balance closures all exceeded 100%, indicated that the inﬂuent ﬂow rate
exceeded efﬂuent. This could be attributed to variations in the ﬂow rate of the
coarse material separated in the SMS and HC. Coarse material ﬂow rates from
the SMS (SSS reclaimed sand) and HC (HC underﬂow) had the largest
coefﬁcients of variation.

None of the SSS mass balance attempts closed within the target of 110%.
All the SSS closure attempts indicated that efﬂuent ﬂow rates exceeded the
inﬂuent (<100%), a result opposite of the machine level. lnterrnittent operation of
the HC, resulting in collection and temporary storage of the SMS efﬂuent, is the
likely reason for the efﬂuent ﬂow rate exceeding inﬂuent.

The dilution ratio, recycled water to SLDM, is included in Table 4.3. MDF
used an average dilution ratio nearly ﬁve times that of the 1 to 1 dilution ratio
suggested by Wedel (1995) as a minimum for sand separation. Due to the
closed loop recycle of solid-liquid separator efﬂuent, dilution water at MDF had a
relatively high mean TS concentration, 4.3% (Appendix A), causing the
equipment supplier to recommend operating the system at a higher dilution ratio.

Tables 4.4 and 4.5 summarize the material density and mass ﬂow rate for

each sample location.
87

Table 4.4: Sand separation system material density - MDF

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

E
Sample 1 Sampl; vent 3 M08" Sargon Coefficient
Point Description 3 3 3 av n of Variation
(kc/m) (kg/m) (kg/m) (kg/m“) (kahn’)
1 Fresh water 997 1 ,005 1 ,000 1 ,000 4 0%
2 Recycled water 1,015 1,037 1,026
3 SLDM' 1,092 1,003 1,048
4 SSS reclaimed sand 2,038 1,708 1,617 1,787 221 12%
5 SMS liqu’d efﬂuent 1.024 1,059 1,035 1,039 18 2%
5a HC imutz 1,024 1,059 1,035 1,039 13 2%
6 HC overﬂow 1,026 1,004 1,024 1,018 12 1%
7 HC underﬂow 1,653 1,149 1,371 1,391 253 18%
Table 4.5: Sand separation system mass ﬂow rate - MDF
Sample Salee Event Mean Standard Coefﬁcient
Point Description 1 2 3 Deviation of Variation
(kg/hr) (kﬂr) (kg/hr) (kglhr) (kg/hr)
1 Fresh water 580 710 691 660 70 1 1%
2 Recycled water 1 3,573 1 0,987 16,669 1 3143 2,845 21 %
3 SLDM‘ 3,007 2,694 3,127 2,943 223 3%
4 SSS reclaimed sand 361 384 629 458 148 32%
5 SMS liquid efﬂuent 15,695 14,913 19,730 16,779 2,585 15%
5a HC input2 24,813 25,125 27,344 25,761 1,380 5%
6 HC overﬂow 22,443 21,449 22,689 22,194 656 3%
7 HC underﬁow 564 1 L306 654 841 405 48%
SMS percent closue: 9% 3% 4% 5% 4%
HC percent closue: 7% 9% 15% 10% 4%
SSS percent closu'e: -33% -52% -14% -33% 19%
1SLDM flow rate based was based on the daily measured piston pump cycle time (down and up-stroke)
minus the up-stroke.
2Cyclone flow rate was based on a measured runtime of 46 minute per hour. ,,,, _

 

 

Similar to the ﬂow rate results, all of the SMS and two of the HC mass ﬂow
rate closure attempts shown in Table 4.5 closed to within 110% while the SSS
mass ﬂow rate measurements did not.

Table 4.6 summarizes the FS mass ﬂow rate for the various sample

locations associated with the SSS at MDF. The F8 mass ﬂow rate was

88

 

determined by multiplying the mass ﬂow rate (Table 4.5) by the FS concentration
(Appendix A).

Table 4.6: Sand separation system ﬁxed solids mass ﬂow rate — MDF

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sample Event Standard
833:? Descriptlon 1 2 3 Mean Deviation ﬁvxgz;
(k Ih (kglhr) (kglhr) (kglh r) (Iblh r)
1 Fresh water -0.2 1 .1 0.2 0.4 0.7 174%
2 Recycled water 188.5 183.2 292.6 221 62 28%
3 SLDM1 250.0 375.1 234.7 287 77 27%
4 SSS reclaimed sand 209.3 268.3 469.9 316 137 43%
5 SMS liquid effluent 238.6 411.6 452.2 367 113 31%
5a HC input2 377.2 693.5 626.7 566 167 29%
6 HC overflow 240.3 381.8 324.2 315 71 23%
7 HC underﬂow 229.1 445.1 337
SMS percent clause: 33% 32% -75% -3% 62%
HO percent closue: -24% -19% -22% W -
SSS percent closu‘e: -3% -16% -51% -23% 25%
1SLDM flow rate based on the measued piston punp downstroke, Lpstroke deducted
2Cyclone flow rate based on measured ru'rtime of 46 minute per hour

 

 

Only one successful FS mass balance closure was achieved for sample
event 1. The poor FS mass ﬂow rate closures were assumed to be caused by
the intermittent and unstable operation of SSS components, primarily the piston

pump and HC.

4.1.3.2 Separation efﬁciency
Equation 2.11 compares the mass of the coarse FS material removed by

the SSS to the mass of the inﬂuent PS to determine separation efﬁciency. Table
4.7 summarizes the results. Fixed solid concentration for the HC sample event 3

was not reported clue to a laboratory error. Mass balance separation efﬁciency

89

calculations assumed that the change in F8 throughout the SSS is caused by the

removal of sand.

Table 4.7: Fixed solids separation efﬁcienc — MDF

 

 

 

 

 

 

 

 

 

 

Sample Event Standard
Component 1 2 3 Mean Deviation
SMS: 31% 27% 89% 49% 35%
HC: 61% 64% 62%
SSS: 48% 48% 89% 62% 24%

 

 

 

Mean separation efﬁciency of the SSS, at only 62%, was less than
reported by Wedel and Bickert (1998). If the efﬁciency results were accurate,
accumulation of sand in the storage would be anticipated. However, discussions
with the owner/operator of the dairy indicated that sand accumulation
downstream of the SSS did not occur (Minnis, 2008). The contrast of the poor
separation efﬁciency to the lack of downstream sand accumulation evidence
indicates that the direct mass balance approach may have limitations when it
comes to predicting separation efﬁciency. Unstable operation of the SSS at MDF
is believed to be the major limitation impacting the ability to conduct direct mass
balance measurements of separation efﬁciency. Consequently, an in depth
study of MDF operations was conducted.

An improperly sized feed pump caused the HC to cycle on and off several
times each hour. Time measurements collected during the sampling events and
discussions with the operator indicate that the HC operated approximately 45 min

of each hour of SSS operation. The HC underﬂow discharged into the dilution

90

pool of the SMS, contributing approximately 18.9 Umin with a FS mass of 18.2
kg or 6% of the average SMS inﬂuent mass. When the HC cycled off, the
contents within the unit emptied completely through the underﬂow discharging
between 132 and 151 L (131 to 150 kg) of manure slurry instantaneously,
creating a short, but substantial increase in the mass ﬂow entering the SMS,
upwards of 45% of the normal SMS inﬂuent FS mass. This instantaneous
increase in HC underﬂow could potentially interrupt the ﬂow of reclaimed sand
from the SMS temporarily. This could occurred because the auger used to
excavate sand from the SMS rides on a bed of sand, approximately one inch
thick, that forms between the auger ﬂighting and the SMS body, eliminating wear
due to metal on metal contact. The velocity of the sudden HC underﬂow
discharge when the unit cycled off had the potential to erode the sand bed in the
SMS, interrupting the ﬂow of reclaimed sand until the bed was reestablished.

At MDF, operation of the piston pump feed system was another SMS input
with a variable ﬂow rate. Manure was metered into the SMS only during the
down stroke of the piston pump. During the up stroke, the piston reﬁlled. The
SLDM transferred into the SMS by the piston pump contributed about 6.4 L/min.
with a mass of 2.9 kg, equivalent to 33% of the input FS into the SMS.
Temporarily interrupting the ﬂow of SLDM directly impacts the ﬂow of reclaimed
sand from the system at any instance.

To further verify that the SSS never reaches static equilibrium, the efﬂuent
ﬂow rate of the SMS was measured every 1.5 to 2 minutes for a period of 45

minutes. Each time the ﬂow rate was measured, the density and solids
91

concentration of the efﬂuent was also determined. Four measurements were

made while the piston pump was on the up-stroke and 3 with the HC not

operating. No measurements were taken when both the piston pump and

cyclone were not operating. The SMS effluent ﬂow rate data is contained in

Appendix A. Figure 4.1 graphically displays the ﬂow range (5.7 to 9.9 Us during

the 45—minute sample period). No pattern or consistency relating to the operation

of the piston pump or HC was observed.

 

 

10

Flow rate (Us)

l

 

o Piston pump and HC on
A Piston pump off, HC on

I HC off, piston pump on

 

 

 

 

 

5

 

T

 

T T

0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 2600 2800 3000

Time (sec)

 

Figure 4.1: Sand manure separator efﬂuent ﬂow rate - MDF

Changes in the recycle water ﬂow rate contribute to variations in the SMS

efﬂuent ﬂow rate. Recycle water ﬂow rate at MDS was controlled using a bypass

system with a valve restricting ﬂow in the line leading to the SMS, excess water

92

 

ﬂow freely through the bypass line. The position of the valves was established
by the SSS operator, however at times valve was known to become partially
clogged with manure ﬁbers and debris, causing changes in the ﬂow rate.

The mass balance approach to determining FS separation efﬁciency was
unsuccessful because the SSS never operated consistently. On/off cycles of
several system components and variability of the ﬂow rate contributed to
unsteady operation. The inability to successful use the mass balance technique
for even a simple system consisting of two technology levels with only one SSS

operating at each level and warrants investigation of an alternative approach.

4.2 Semi-empirical mass balance separation efﬁciency
Limitations with the mass balance approach, as demonstrated at MDF, led

to the development of a semi-empirical mass balance approach that used a
combination of measured and established industry data to predict the separation
efﬁciency. The semi—empirical technique used the daily efﬂuent mass ﬂow rate
and the FS concentration for each component of the SSS to determine the mass
of FS remaining in the efﬂuent (residual sand) contributed by bedding sand and
the FS separation efﬁciency. Because of the complexity at GMF, this technique
is believed to be the only valid technique to estimate of the sand separation
efﬁciency.

Data collected at GMF during the design, construction and startup of the
AD were used for the semi-empirical mass balance. As discussed in Chapter 3,
the SSS at GMF consisted of six SMS-MINI combinations feeding into a single

93

HC. Flow was only measured at a single location, HC overﬂow entering the

digester and phosphorus separation process.

4.2.1 Semi-empirical mass balance parameter speciﬁcation
To complete the semi-empirical mass balance, several key parameters

were either assumed or determined by measurements.

Due to the inability to accurately measure manure production and bedding
usage, both were assumed to be similar to industry standards (American Society
of Agricultural and Biological Engineers, 2005; Midwest Plan Service, 2000). The
standards are developed by industry experts using data collected from numerous
research trials around the United States. Both standards are used in Michigan
for technical and regulatory guidance (Michigan Department of Environmental
Quality, 2006; US. Department of Agriculture, 2009). Table 4.8 summaries the
standard production values on a per cow basis and the bulk density of different
system inputs and products. Table 3.4 indicates which data was measured and
which was based on industry standards at GMF, the case study farm used for the

semi-empirical mass balance.

94

Table 4.8: Standards used in the semi-empirical mass balance evaluation
Parameter Value Unit Source
Lactating cows 2,900
Dry cows 300
Lacatinggw manure 68 kg/d/cow ASABE, 2005
Dry cow manure 38 kg/d/cow ASABE, 2005
Bedding 22 kg/d/cow MWPS-18, 2000

 

 

 

 

 

 

 

 

Bulk density of manure 1,001 kg/m3 MWPS-18, 2000
Bulk density of SLDM 1L163 kg/m3 Gooch and Inglis, 2007
Bulk density of sand 1,696 kg/m3 MWPS-18, 2000

 

 

 

 

 

 

 

 

The primary component of FS above the base line level in feces was
assumed to be sand. Also, changes in F8 throughout the sand separation
process were assumed to be sand. This assumption was based on the measure
difference in the FS concentration of feces compared to SLDM (Table 4.2).

F8 from the dilution water was assumed to be negligible. The basis was
that the TS concentration in the recycled dilution water used at GMF ranged from
1.05% to 1.95% (Green, 2005), with a FS concentration range of 0.16% to 0.3%.

Efﬂuent ﬂow rate for each component of the SSS was assumed to be 10%
less than the inﬂuent ﬂow rate. Section 3.5.3 provides a detailed discussion of
the ﬂow rate determination for the SMS and MINI. This assumption was based
on ﬁndings from MDF which indicated that the SMS and HC efﬂuent (ﬁne
material) ﬂow rates were between 9% and 12% less, respectively, than the
inﬂuent ﬂow rates (Table 4.3).

Using the industry data in Table 4.8, the mass and volume of manure at

each GMF sample location were calculated (Table 4.9). Table 4.9 summarizes

95

the theoretical and measured volumetric and mass ﬂow rate of each sample

location at GMF.

Table 4.9: Mass and ﬂow rate data — GMF

 

 

 

 

 

 

 

 

 

F low Rate
Sample Volume Mass
Locatlon 3
(m Id) (kg/d) ,

Manure 208 208,600
Bedding sand 42 71,273
SLDM 241 279,873
SMS 504 504,497
MNI 458 458,634
HC overﬂow 416 416,940

 

 

 

 

 

Based on the mass ﬂow data in Table 4.9 and the FS data in Table 4.2,
the mean mass of FS contained in the efﬂuent of each step in the treatment
process is presented in Table 4.10. The 95% conﬁdence interval for the FS

mass is also included in Table 4.10.

Table 4.10: Overall FS mass by treatment step - GMF

 

 

 

 

 

 

 

 

 

 

 

 

Mean 3!”.99,” 95% Conﬁdence Limit
Treatment Error Lower Upper Count
(kg/d) (kgldj (kgld) Wkgld)
Manure 3,978 1,455 1,926 8,217 68
SLDM 43,887 7,975 28,821 66,828 64
SMS 10,199 2,373 6,299 16,515 50
MINI . 6,402 1,728 3,709 11,049 67

 

 

 

Mean mass ﬂow rates of the overall data in Table 4.10 for manure, SLDM,

SMS and MINI represent samples from the six sample locations (machines) at

96

SMS and MINI level of the SSS. Whenever possible, samples were collected
from all SSS levels and machines during a sampling event. However, due to
system management and sample event goals, not all of the SSS components
were evaluated during all of the sampling events.

For all sample locations, except manure, the mass of the F8 in Table 4.10
was determined by multiplying the mass ﬂow rate in Table 4.9 by the FS
concentrations (Table 4.2). To compute the FS contributed by manure in Table
4.10, two-thirds of the daily manure excretion was assumed to be feces and one-
third urine (Nennich et al., 2006; Bannink et al. 1999). The F8 concentration of
feces was measured (Table 4.2), while the FS concentration of the urine was
assumed to be 1.5% (Bannink, 1999).

Factors contributing to the change in ﬁxed solids of samples collected
from GMF were evaluated using a SAS mixed model with both ﬁxed and random
effects. Fixed effects included the following.

0 Management: operational differences between Farms 2 and 3.
0 Treatment: compared feces, SLDM, SMS, MINI and HC.
0 Interaction of management and treatment.

Machine was considered a random effect, which represented the six SMS
and MINI operating as part of the SSS (Figure 3.1). Section 3.4 provides
additional details on the statistical evaluation of the change in PS.

The mass of F8 for feces, SLDM, SMS and MINI from GMF were not
normally distributed. Consequently, a logarithmic transformation was used to ﬁt

the data to a normal distribution curve. Outliers, values distant from the critical
97

mass of data, including SLDM from Farm 3, Machine 6 on 07/05/06 and
03/23l07, Feces from Farm 2, Machine 4 on 07/05/06 and SMS from Farm 2,
Machine 4 on 03/23/07 were deleted. A test of the ﬁxed effects found treatment
to be the only statistically signiﬁcant factor contributing to the change in PS at
a=0.05 (p50.0001). Management and the combination of management and
treatment were not signiﬁcant at a=0.05, resulting in a p=0.7464 and p=0.7912,
respectively. This supported the assumption that changes in the mass of FS
throughout the SSS are attributed to treatment.

Comparing the least square means for the different treatment levels
indicated that the mass of F8 in feces compared to SLDM (p50.0001) and SLDM
compared to SMS (p50.0001) were statistically different. However, FS mass of
the SMS compared to MINI (p=0.1768) was not statistically different. The result
indicated that the SMS as a system was the only statistically important level of
treatment. ANOVA and least square mean tables are included in Appendix B.

The large increase in the mass of F8 from manure to SLDM results from
the addition of bedding sand. Similarly, the decreasing mass of F8 from SLDM
to SMS to MINI is from the removal FS. Differences in the management (Farms
2 and 3) and the sample location (machine) were not considered in the empirical
mass balance because those factors were not found to be statistically important
in explaining the change in PS throughout the SSS, as discussed above.

Hydrocyclone overﬂow data resulted from samples collected from the
single machine (Figure 3.1). To evaluate the impact of the HC on the mass of FS

remaining in the manure, data from the other treatment levels was averaged for
' 98

sample events when the HC was operational. Averaging results by sample event
was done to allow for direct comparison with the HC and because the effects of
machine and management were not signiﬁcant for changes in PS. Table 4.11
summarizes the mean and 95% conﬁdence intervals for the daily MINI and HC
data, as well as the other treatment levels. A check of the residuals found that
the data was normally distributed. A test of the single ﬁxed effect, treatment,
found a statistically signiﬁcant difference in the FS of the HC overﬂow compared
to the MINI at or=0.05 (p- =0.0017), indicating that the HC had a statistically
important effect on the mass of F8 in the manure. ANOVA and least square

mean tables for F8 mass are contained in Appendix B.

4.2.2 Sand separation efﬁciency
Svarovsky’s (1990) cumulative separation efﬁciency equation for ﬁne

particulate (Equation 2.12) was used to determine cumulative FS separation of
the SSS. Since management and machine effects were not found to have a
signiﬁcant effect on the change of FS mass, the mean of the treatment levels for
a given sample event (sample day) was used to determine the separation
efﬁciency. Table 4.11 summarizes the FS mass data for the twelve sample
events where all components were sampled. Using the daily mean of the FS
mass for each sample location allowed for a direct comparison of HC to the other

treatment levels with multiple machines.

99

Table 4.11: Daily ﬁxed solids mass data — GMF

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Treatment Standard
Level _WMean Deviation Median Count
(kg/d) (kg/d) (kg/d)
Manure 3,705 571 3,847 12
SLDM 52,716 11,240 56,003 12
SMS 9,878 2,173 10,018 12
MINI 6,695 787 6,504 12
HO Overﬂow 5,334 621 5,094 12

 

The difference in the FS mass of SLDM and Manure provided the baseline
mass of FS contributed by the sand bedding. That baseline was used to
determine the cumulative FS separation efﬁciency. Table 4.12 summarizes the

mean FS separation efﬁciency at each treatment level. The 95% conﬁdence

interval is also shown in Table 4.12.

Table 4.12: Cumulative ﬁxed solid separation efﬁciency — GMF

 

 

 

 

 

 

 

 

 

 

95% Conﬁdence Limits
Treahnent
Level Mean Lower Upper
1%) Clo) (7.)
SMS 87 83 90
MINI 94 92 95
HO Overﬂow 97 96 98

 

Figure 4.2 graphically displays the mean cumulative FS separation
efﬁciency with the 95% conﬁdence interval for the three levels of the SSS. The

95% conﬁdence interval of the SMS was within the range predicted by Wedel

and Bickert (1998).

100

 

 

 

 

 

 

 

 

 

A 100
32
; [l1
g 95 [I]
.3 Cl Mean
E 90 — 95%
c [I] Conﬁdence
3 85 Limit
8 l
a
‘13 80
If SMS MINI HC Overflow

Sand Separation System Component

 

 

 

Figure 4.2: Cumulative ﬁxed solid separation efﬁciency — GMF

Data in Table 4.13 compares the separation efﬁciency for each level of the
SSS. Table 4.13 was constructed with the data shown in Table 4.11. Similar to
the cumulative separation efﬁciency, treatment level efﬁciency was calculated
using Equation 2.10. To determine the treatment level efﬁciency, the FS mass of
the component inﬂuent (efﬂuent from the previous unit) was used. Compared to
the SMS, the separation efﬁciencies of the MINI and HC were signiﬁcantly less.
Figure 4.3 depicts the mean and conﬁdence interval for the FS separation

efﬁciency of each SSS treatment level.

101

Table 4.13: Treatment level ﬁxed solid separation efﬁciency — GMF

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Treatment 95% Conﬁdence Limits
Level "9"“ Lower Upper
(7°) Clo) (%)
MINI 45 36 55
HO Overﬂow 47 33 60
A 100
.\°
5. 90 [I]
c 80
.9
g 70 Cl Mean
lg 6° * —95%
3 50 I I Conﬁdence
E [I3 [F Limit
3 40 l I
0
g 30
it SMS MINI HC Overflow
Sand Separation System Component

 

Figure 4.3: Treatment level ﬁxed solid separation efﬁciency - GMF

The semi-empirical mass balance approach estimated sand separation

efﬁciency values similar to those reported by Wedel and Bickert (1998).

Speciﬁcally, the SMS separation efﬁciency had a 95% conﬁdence interval of 83%

and 90% separation efﬁciency of the bedding sand from SLDM. Incremental

improvements in the separation efﬁciency were observed for each component of

102

 

the SSS, resulting in a cumulative FS separation efﬁciency with a 95%
conﬁdence interval of 96% to 98%.

Based on the conﬁdence interval of the FS separation efﬁciency
determined by the semi-empirical evaluation, the mass of residual sand in the
manure after the SSS was in the range of 1,564 mm to 2,994 kg/d. This
assumed the GMF uses 71,273 kg/d of bedding sand (Midwest Plan Service,
2000). Determining the FS separation efﬁciency was an important factor for
predicting sand accumulation potential (quantity) in the digester at GMF.
However, the PSD of this residual FS must be examined to determine if the
mixing system was capable of keeping residual sand suspended in the AD at
GMF.

Accuracy of the semi-empirical technique was inﬂuenced by the manure
production and sand usage data sets used to estimate the baseline FS
contributed by manure and bedding sand, the ability to only track the separation
of FS contributed by sand and the management of the SSS. Manure production
and sand usage are highly variable due to differences in environmental
conditions and management, resulting in an error range of 130% (Midwest Plan
Service, 2000). Similarly, urine production and composition is variable
depending on water intake, ration, feed quantity and milk production (Nennich et
al., 2006).

Fixed solids provide the simplest most logical method for tracking sand
changes in manure and the semi-empirical mass balance method provided useful

information for system planners. However, simply tracking the FS change does
103

not provide sufﬁcient information for planners of AD or other advance treatment
systems. For that reason particle size data of new sand, sand removed and
residual sand was also investigated to provide additional measure for tracking
the fate of bedding sand. Particle size data also provides important information

to system planners sizing mixing systems to address sedimentation.

4.3 Particle size distribution of sand separation products
Understanding that the HC overﬂow did contain residual sand,

determination of the PSD of residual sand was needed to evaluate the potential
for sedimentation in the downstream processes, in particular the AD tanks. This
data will enable the determination if the AD mixing system is sufﬁcient to achieve
the scour velocity of the mean particle size remaining the HC overﬂow so that
sedimentation does not occur.

As described in Chapter 3, samples were collected from new sand,
reclaimed sand, HC Underﬂow, HC overﬂow and sludge accumulation in the
digester tanks at GMF for PSD evaluation. Figure 3.2 and Table 3.4 indicate the
locations samples were collected to analyze for PSD. New sand samples served
as the baseline.

Table 4.14 summarizes the results of the PSD analysis as the mean
percent passing of particles through an array of sieves with decreasing opening
sizes (the complete PSD data set including statistical analysis is contained in
Appendix D). The sieve opening size in Table 4.14 represents the maximum
particle diameter passing the given sieve calculated using equation 3.6.

104

Table 4.14: Mean sand particle size distribution - GMF

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sieve Mean Percent PassingW
Sieve # Opening New Reclaimed HC HC Tank
4mm) Sand Sand Underﬂow Overﬂow Sludg;

4 4.75 100 99 93 98 96

8 2.36 88 83 87 96 92

16 1 .18 69 60 77 94 86

30 0.60 45 33 61 89 78

50 0.30 12 4 37 72 63
100 0.149 2 0 10 39 37
200 0.074 1 0 2 10 9
Pan 0 0 1 -1 -1
Number of Samples 6 7 5 5 4

 

 

 

 

 

 

 

Based on the PSD, new sand from GMF was classiﬁed as concrete sand
according to ASTM (2006) Standard, Table 2.3. According to the ASTM (2006),
AASHO (1991) and USDA (1993) standards, the sand remaining in the manure
stream after the HC was classiﬁed as ﬁne sand to very ﬁne sand.

Figure 4.4 graphically displays the PSD data presented in Table 4.14.
The X-axis represents the particle size (diameter) and the Y-axis indicates the
percentage of the sample that passed through that size sieve. Using new sand
as the baseline, curves to the right represent coarser material and curves to the
left, material that is ﬁner than new sand. Reclaimed sand from the SMS was
coarser than the new sand, with a lower percent pass at all of the measured
particle sizes. Speciﬁcally, sand reclaimed by the SMS contained less than 4%
(by mass) particles with a diameter of 0.3 mm or less, compared to new sand

which contained 12%. This signiﬁes that the SMS was more effective at

105

removing coarse sand particles leaving ﬁner sand particle in the manure stream

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(SMS efﬂuent).
100%
. »-X-"" " -'
.X ' —. ‘1 W l— ,5".
m ' A ,- w .' iW/tr ‘
,3 PI" “ﬂ...
f I ’ ,’ ,I
00% _ . , ,4 /.
.’ ’0 A ...I’
(I ,' J’ /'
70% , .4 I r ".1
,’ ’ ’ / I"
,r’ ”I {I ."I
g 6“ IV I! ’r I!
I 4 r
I I ,4
3 ... I, 7:,
I I "
E 40% ' ’ ” /
0' a ll / ,1"
l/ I H
I" ’ / . +NewSand
m I), ’7 / [‘1‘ _—
7) I / ,4’ "I'" Reclaimed Sand
1 ' J".
20% l, I / ,2 -s-HCUndarflow "i”
4 ,4
10% ‘V’ J, -*-' HCOverflow __
’ a ‘ / ,
. r +. -.. - Tank Sludge
r .......
0% F... I 1 l I 1 I
0.01 0.1 ‘I 10

Particle Size (mm)

 

 

 

Figure 4.4: Average sand particle size distribution - GMF

The PSD of HC underﬂow was visibly ﬁner than both new and reclaimed
sand. These ﬁndings for the SMS and HC support the statement in Zimmels
(1984) that individual separation technologies are only effective over a fraction of
the sand particle size range. By design, the HC separator at GMF was intended
to separate ﬁner (smaller) sand particles than the SMS and MINI in order to

reduce the residual sand in the manure stream entering the AD.

106

lnterpolating the mean particle size of residual sand in the HC overﬂow
and in the tank sludge resulted in diameters of 0.18 mm and 0.21 mm,
respectively. This resulted in an average PSD of residual sand of 0.195 mm. In
comparison, the mean particle size of new sand was 0.6 mm. This reduction
indicated that the SSS was effective at removing coarse sand particles from the

manure.

4.3.1 Settling and scour velocity of sand
The mean particle size (50% passing) for each sample location was

interpolated from the graph shown in Figure 4.5. Scour velocity (Equation 2.7)
was calculated for new and residual sand using the mean particle size, results
are shown in Table 4.15. Data for new sand is included for comparison

purposes.

Table 4.15: Residual sand scour velocity- GMF

 

 

 

 

 

 

 

 

 

 

 

Mean Scour
Sample Particle Velocity
Location Size Initiation Complete
(mm) (mls) (mls)
New Sand 0.69 0.26 1.15
HC Overﬂow 0.18 0.13 0.59
Tank Sludﬁ 0.21 0.14 0.64

 

 

Compared to new sand to the mean particle size of residual sand in the
HC overﬂow and found in the tank sludge was reduced by three quarters. This
particle size reduction reduced the settling velocity by an order of magnitude and

the scour velocity (initiation and complete) by half compared to new sand.
1 07

4.3.2 Residual sand impact on anaerobic digester mixing system design
Anaerobic digester mixing systems are typically intended to bring viable

bacteria into contact with substrate, minimize heat and solids gradients and to
release biogas from the slurry (Hoffman et al., 2008). Additionally, the mixing
system at GMF, as with any dairy farm using sand bedding, also needs to
prevent grit accumulation.

Annually, GMF used an estimated 26,000 metric tons of bedding sand,
occupying a volume equivalent to 15,300 m3lyr. Theoretically, if no sand
separation was in place at GMF, bedding sand could ﬁll the 10,200 m3 anaerobic
digester in less than ﬁve months. Based on the cumulative separation efﬁciency
presented in Table 4.12, the mass and volume of residual sand in the HC
overﬂow is between 1,564 kg/d to 2,994 kg/d and 0.92 m3/d and 1.77 m3/d,
respectively. In theory, if all the residual sand accumulated in AD tanks it would
ﬁll the tanks in the range of 16 to 30 years. To prevent this sedimentation, the
power in the AD mixing system must maintain produce a minimum velocity that
exceeds the initiation scour velocity of residual grit.

Each digester tank at GMF was constructed with three submersible
propeller type mixers, based on the German technology suppliers experience
with systems in Europe. Two 13 kW mixers were positioned to provide 26 kW of
total horizontal mixing, while one 10 kW mixer was aligned vertically (Figure 4.5).
Scour velocity as determined using Shields’ (Equation 2.5) only considers the
horizontal velocity (Wedel, 2000).

108

DIGESTER SUBSTRATE

/ VERTICAL MIXER

I
O HORIZONTAL MIXER —> :

 

Figure 4.5: Anaerobic digester mixer conﬁguration - GMF

For comparison purposes, Table 4.16 summarizes the minimum
theoretical power required to initiate and develop complete scour for the mean
sand particle size found in HC overﬂow and tank sludge. Again, new sand was
included for comparison purposes. The bulk mixing power required was
determined using Equations 2.16 through 2.18, with the scour velocity of the
average particle size for each sample location (Table 4.15) substituted as the
velocity in Equation 2.17. A mixing time of ﬁve minutes per hour, typical at GMF,
was used, as recommended by the technology supplier and supported by
Hoffmann et al. (2008). Annual operating cost of the mixing operation was
calculated assuming 365 days of operation and an electricity purchase price of

$0.10 per kWh.

109

Table 4.16: Power required to achieve scour velocity of sand - GMF

 

 

 

 

 

 

 

 

 

 

 

 

Sample Mean Power Operatin Cost
Location Particle Size Initiation Full Initiation Full
(mm) (kW) W) (In) (er
New Sand 0.69 0.75 15 $55 $1 ,090
HO Overﬂow 0.18 0.19 4 $14 $284
Tank Sludge 0.21 0.23 5 $17 $332

 

 

The power required to achieve full scour ranged from 4 kW for the mean

sand particle contained in H0 overﬂow to 15 kW for new sand. Based on the

EPA (1979) recommendation, the AD tanks at GMF should have between 18 and

27 kW of mixing power to suspended residual sand contained in the inﬂuent.

The, actual horizontal mixer power was GMF 26 kW. Due to the change in the

mean particle size resulting from sand separation, the power required for full

scour of residual sand in H0 overﬂow is one fourth that of new sand. It should be

noted that while the mixer power installed at GMF is adequate for the mean

particle size of new sand, it does not have sufﬁcient power to completely scour

sand grains greater than 1.18 mm. New sand used for bedding at GMF

contained over 30% particles greater than 1.18 mm, compared to only 6% of the

residual sand in HC overﬂow. Consequently, the mixing system installed in the

AD system at GMF is capable of fully scouring 94% of the residual sand. The

cost to operate the mixing system was less than $1,000 annually. Even with

sufﬁcient power, improper mixer selection, conﬁguration and Operation may still

result in sedimentation in the AD tanks as low velocity (dead) zones can account

for a large portion of the tank volume (Wu and Chen, 2007).

110

4.3.3 Veriﬁcation of sand accumulation predictions

Consequently, the semi empirical mass balance approach used to
determine the SSS efﬁciency combined with the modeling of mixing indicated
that sand accumulation in the GMF AD tanks should not occur. In October of
2008, after ﬁfteen months of operation, AD tank #3 at GMF was emptied for
maintenance on the mixers (a detailed explanation of the maintenance event is
contained in Chapter 3). This provided the opportunity to verify the mixing
system predictions in regards to sand accumulation.

Prior to the maintenance, tank #3 had received half of the digester
feedstock each day, resulting in a load of F8 from residual sand of between 713
kgld to 1,425 kgld (0.46 m3/d to 0.89 m3ld). In theory, if no mixing were provided
and all residual sand settled, a layer of residual sand 160 to 310 mm thick could
have accumulated over the ﬁfteen-month operating period.

During the mixers repairs, the tank volume was lowered so that
approximately 150 mm of liquid remained in the tank. By probing accumulated
sediment, it was approximated that sludge accumulation (sand and manure) was
approximately 25 to 50 mm thick (Green, 2008). For safety reasons, exact
measurements could not be taken. Grab samples of the sediment were collected
from two locations for PSD analysis, results are presented as tank sludge in
Figure 4.4 and Table 4.14. This veriﬁed that the combination of SSS and mixing
had achieved the goal of minimizing sand settling in the tanks and the modeling

approach was accurate.
1 1 1

4.4 Volatile solids loss due to sand separation
The mass of VS contained in feedstock directly correlates to the potential

biogas production under anaerobic conditions (U.S. Department of Agriculture,
2007). Consideration of the impact of pretreatment of manure, sand separation,
on the mass of VS is essential for predicting the impact on biogas production.
Volatile solids loss is an unintended consequence of sand separation. Removal
of VS with FS and increased aeration are two effects that contribute to any

reduction in the manure VS mass.

4.4.1 Volatile solids characteristics of sand separation products
Volatile solids concentrations for each SSS sample location at GMF are

summarized in Table 4.17. Equation 3.3 describes how the VS concentration is
determined. The feces sample excluded urine, spilled water and sand bedding
so consequently, using this material for comparison is not warranted. Therefore,
SLDM was used as the baseline for comparing the change in VS throughout the

$88 at GMF.

Table 4.17 Sand separation system volatile solid concentration - GMF

 

 

 

 

 

 

 

 

Samph p M98" smﬂdﬂt‘L
D“ s°"'°° Locaﬁon vs Deviation Median Count
(94) (Va) (°/.)
Feces 12.6 1.7 12.5 68
SLDM 7.6 1.4 8.0 69
GMF SMS 3.7 0.9 3.8 54
MINI 3.4 1.0 3.5 70
HC Overﬂow 3.0 1.1 3,2 40

 

 

 

 

 

 

 

 

112

Unlike FS, manure is the primary source of VS in manure. The mass of
volatile solids in manure is inﬂuenced mainly by the ration fed to the cattle.
Rations generally remain stable for long periods (months), which was the case
during the research. The addition of recycled dilution water during the sand
separation also contribute to changes in the mass of VS. Dilution water used in
the SMS contained less than 2 % TS and was consistent over long time intervals
at GMF due to the operation of the phosphorus separation system.

VS data from GMF was evaluated to determine which factors contributed
to changes in the mass of VS in the manure and if it was statistically signiﬁcant.
Fixed and random effects, as described in Section 4.1.2, were evaluated using a
SAS mixed model. Treatment, management and combination of treatment and
management were not found to be statistically signiﬁcant effects causing
changes in the VS of the manure stream at a=0.05 with values of p=0.2413,
p=0.1995 and p=0.7022, respectively. Comparing the treatment least square
means indicated no signiﬁcant difference in the mass of VS of SMS compared to
SLDM (p=0.3187) or MINI compared to SMS (p=0.2821). Table 4.18 shows the

mean mass of VS at each treatment level in addition to the 95% conﬁdence limit.

Table 4.18: Overall VS mass by treatment step - GMF

 

 

 

 

 

 

 

 

 

Mean Standard 95% Conﬁdence Limit
Treatment Error Lower Upper Count
(kgld) (kg/g (kgld) (kgld)
SLDM 21,259 1,326 4,410 38,109 64
SMS 18,738 1 .347 1 .626 35,851 50
MINI 15,846 1 .321 -937 32,629 67

 

 

 

 

 

113

Similarly, a test of the single ﬁxed treatment effect, HC overflow compared
to the MINI, did not contribute to a statistically signiﬁcant difference in the mass
of VS at a=0.05 (p=0.2568). ANOVA and least square mean tables for VS are
contained in Appendix B. While changes in the VS mass throughout the SSS
were not found to be statistically signiﬁcant, the data does indicate that smaller

mass of VS remained in the manure following sand separation.

4.4.2 Volatile solids changes due to sand separation
Techniques described in Section 4.2.2 for use in determining the

separation efﬁciency of FS were used to track the change in the mass of VS.
Again, since management and machine were not found to have a signiﬁcant
effect on the change of VS mass, the mean of individual machine samples at
each treatment level for a given sample event was used to determine the change
in VS mass. Table 4.19 summarizes the mean and conﬁdence intervals of the
daily averaged MINI data and the HC overﬂow. While Table 4.18 evaluates all
the data for SLDM, SMS and MINI, Table 4.19 contains the only the data for
sample events that included the HC. Compared to F8, VS had one less sample
event due to an isolated laboratory problem. Using the daily mean of the VS
mass for each sample location allowed for a direct comparison of HC to the other
treatment levels with multiple machines.

Techniques described in Section 4.2.2 for use in determining the

separation efﬁciency of FS were used to track the change in the mass of VS.

114

Again, since management and machine were not found to have a signiﬁcant
effect on the change of VS mass, the mean of individual machine samples at
each treatment level for a given sample event was used to determine the change
in VS mass. Table 4.19 summarizes the VS mass data for the eleven sample
events used to evaluate change in VS. Compared to F8, VS had one less
sample event due to sample corruption during laboratory analysis. Using the
daily mean of the VS mass for each sample location allowed for a direct

comparison of HC to the other treatment levels with multiple machines.

Table 4.19: Daily volatile solids mass data — GMF

 

 

 

 

 

 

 

 

 

 

 

 

 

Treatment 3‘39“?” .
Level Mean Devuatlon Median Count
(kgld) (kgldl (kgld)
SLDM 22,050 1,765 22,254 11
SMS 17,467 2,304 18,420 1 1
MINI 15,154 3,177 15,983 11
H0 Overﬂow 12,765 1,960 13,418 1 1

 

Sand laden dairy manure provided the baseline mass of VS, which was
used to determine the cumulative change in VS. Table 4.20 summarizes the
mean percent change in the VS mass at each treatment level. The 95%

conﬁdence interval is also shown in Table 4.20.

115

Table 4.20: Cumulative change in volatile solids — GMF

 

 

 

 

 

 

 

 

 

 

 

Treatment 95% Conﬁdence Limits
Level _’ ”£9.99”--. Lower Upper
0%) (%I (%I
SMS 20 12 27
MINI 29 21 38
H0 Overﬂow 42 36 48

 

 

Figure 4.6 graphically displays the cumulative VS change with the 95%

conﬁdence interval for the three levels of the SSS.

 

 

50
45
A 40
t 35 DMean
a 30 — 95%
3') :3 Confidence
Limit
> 15
10
5
SMS MINI HC Overflow
Sand Separation System Component

 

 

 

Figure 4.6: Cumulative change in volatile solids - GMF

The change in VS during sand separation is likely caused by the removal
of VS with FS. Coarse material, sand, removed by the SMS and HC (underﬂow)

had higher concentrations of VS compared to new sand, 3.3% and 4.2%

116

compared to 2.2%, respectively. Sand sample data is included in Appendix F.
Samples from the MINI were not evaluated, as a representative sample of could
not be collected.

Biodegradation of VS may have also contributed to the VS reduction. At
multiple locations in the SSS, signiﬁcant aeration could occur due to agitation
and turbulence. Both short-ten'n intense aeration and long-term low intensity
aeration has been shown to lower chemical oxygen demand manure, an indicator
of waste strength similar to VS, by as much as 60% (Classen and Liehr, 2005;
Zhang et al., 1997).

Data in Table 4.21 compares the change in VS for each level of the SSS,
constructed using the data set shown in Table 4.19. Similar to the cumulative
change, treatment level changes were calculated using Equation 2.12. The VS
mass of the component inﬂuent (efﬂuent from the previous unit) was used
instead of the VS mass of SLDM. Each level of the SSS saw similar changes in
the mean VS mass remaining in the manure. Figure 4.7 depicts the mean and

conﬁdence interval for the VS change for each SSS treatment level.

Table 4.21: Treatment level change in volatile solids - GMF

 

 

 

 

 

 

95% Conﬁdence Limits
Treatment
Level Mean Lower Upper
(%) (%) (%)
SMS 20 12 27
MINI 13 5 21
HC Overﬂow 13 3 24

 

 

 

 

 

 

117

 

 

35

 

 

 

 

 

 

 

. 30
i 25
3 El Mean
3 20 ﬁ —95%
‘9 15 I , Confidence
[In [F Limit
10 I
5

 

SMS MINI HC Overflow
Sand Separation System Component

 

 

 

Figure 4.7: Treatment level change in volatile solids - GMF

Uncertainty or design imprecision is the uncontrolled stochastic variation
in variable values (Antonsson, 2001). Based on the discussion in section 2.6.1,
biogas production under both laboratory and ﬁeld conditions is naturally
uncertain. Utilizing site-speciﬁc data, such as a change in the mass VS due to
sand separation observed at GMF, even if not statistically signiﬁcant, provides
system planners with information to help cope with the uncertainty of biogas

production.

4.5 Impact of SSS on anaerobic digester design
Based on the data evaluated to determine the FS separation and change

in the VS, it is obvious that SSS efﬂuent has characteristics very different from

118

that of SLDM. The addition of dilution water and changes in the mass of VS due
to the operation of the SSS at GMF impact the amount of energy needed to heat

the inﬂuent to the operating temperature of 35°C and the biogas potential.

4.5.1 Heating requirements of sand separation system efﬂuent
Energy required to heat SLDM and the various SSS products from

ambient temperature to the operating temperature of the AD, typically 35°C, was
determined using Equation 2.19. Table 4.22 summarizes the requirements of
SLDM and the SSS products at GMF over a range of ambient air temperatures.

Manure mass ﬂow rates from Table 4.9 were used for the predictions.

Table 4.22: Heatin required achieve AD operating temperature - GMF

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Ambient Enerﬂ Required
Temperature Manure SLDM SMS MINI HC overﬂow
(°C) (kJId) (kJId) (kJId) (kJId) (kJId)
0 28,934,420 30,647,565 73,836,251 67,123,865 61,021,695
5 24,800,932 26,269,342 63,288,215 57,534,741 52,304,310
10 20,667,443 21,891,118 52,740,179 47,945,618 43,586,925
15 16,533,955 17,512,894 42,192,143 38,356,494 34,869,540
20 12,400,466 13,134,671 31,644,108 28,767,371 26,152,155
25 8,266,977 8,756,447 21,096,072 19,178,247 17,434,770
30 4,133,489 4,378,224 10,548,036 9,589,124 8,717,385
35 0‘ 0 O 0 0

 

 

Dilution water, added during SSS, approximately doubled the energy
needed to raise the ambient temperature of SMS, MINI and HC overﬂow to the
temperature. Increasing the need for energy to heat the AD inﬂuent (with an
ambient temperature less than the operating temperature) results in less energy

available for offset elsewhere on the farm or for sale as renewable energy,
1 19

unless waste heat from a combined-heat and power biogas utilization system is

available.

4.5.2 Biogas potential from sand separation products
The USDA (2007) indicated that digesters using dairy manure as a

feedstock can produce approximately 1.9 m3 of biogas/cow/day with an average
energy content of 20,900 kJ/m3 (NRCS, 2007). This is equivalent to 39,700
kJ/cow/day. Biogas production is the ultimate measure of economic viability of
an AD and is based largely on the mass of VS entering the system daily, as
evident by the above conversions. Sand bedding and the subsequent separation
does statistically impact the mass of VS contained in the manure stream, as
previously demonstrated.

Biogas and energy potential of the various sample locations at GMF,
summarized in Tables 4.23 and 4.24, were predicted using Equations 2.14 and
2.15, which rely on the VS concentration of AD feedstock. Mean and standard
deviation data from the daily average VS mass data presented in Table 4.19 was
used to calculate the low (mean - standard deviation), high (mean + standard
deviation and mean biogas potential in Table 4.23. A biogas yield of 0.27 m3/kg
VS was used to calculate the biogas potential. The biogas yield was the mean of
literature values (U.S. Department of Agriculture (2007); Steffen et al., 1998;

Morris, 1976), which ranged from 0.18 m3/kg VS to 0.39 m3/kg VS.

120

Table 4.23: Anaerobic digester biogas potential - GMF

 

 

 

 

 

 

 

 

 

 

 

 

VS Biogas Yield
Treatment Mass

Level Low Mean High
(kgld) (msld) (m3ld) (ma/d)

SLDM 22,050 3,984 5,907 8,655
SMS 17,467 3,156 4,680 6,856
MINI 15,154 2,738 4,060 5,948
HC Overﬂow 12,765 2,307 3,420 5,011

 

 

The theoretical biogas yield values in Table 4.23 was determined using
the mass of VS in the manure at each treatment level, Table 4.20. Table 4.23
indicates that on average, biogas potential was reduced by 42%, due to the
change in the mass of VS in the manure during sand separation. Theoretical

energy potential based on biogas potential is summarized in Table 4.24.

Table 4.24: Anaerobic diggter energy potential - GMF

 

 

 

 

 

 

 

 

 

 

 

 

Treatment Energy Potential .
Level Low Mean High
(kJId) (kJId) (kJId)
SLDM 80,060,494 118,710,388 173,924,522
SMS 63,419,254 94,035,445 137,772,861
MINI 55,021,966 81,584,294 119,530,478
HC Overﬂow 46,349,048 68,724,450 100,689,311

 

Comparing the energy potential of HC overﬂow as a AD feedstock (Table
4.24) to the energy needed to heat the inﬂuent (Table 4.22) indicates that biogas

can produce sufﬁcient energy to heat the inﬂuent from 0 to 35°C (Table 4.25).

121

 

 

 

 

 

 

 

 

 

 

Table 4.25: AD energy otential compared to heatigg requirement - GMF
Ambient Temperature Energy potential required to heat inﬂuent, %
(°C) SLDM sus MINI HC Overﬂow

0 26 75 81 89
5 23 64 70 76
10 19 53 58 64
15 15 43 46 51
20 1 1 32 35 38
25 8 21 23 25
30 4 11 12 13
35 0 0 0 0

 

 

 

 

 

 

On average, HC overﬂow compared to SLDM required over three times

the energy potential to achieve the target temperature. According to the

theoretical data in Table 4.25, at freezing only 26% of the energy potential of

SLDM was needed to heat the AD inﬂuent to the operating temperature

compared to 89% of the HC overﬂow. lnefﬂciencies with direct combustion are

between 65% and 85% (Wilkie, 2008), potential resulting in a negative energy

balance for dilute inﬂuent feedstocks. Table 4.26 evaluates the energy balance

assuming all biogas is used for heating, with an average combustion efﬁciency of

80%.

122

 

Table 4.26: AD energy potential compared to heating requirement
assuming a combustion efﬁciency of 80% - GMF

 

 

 

 

 

 

 

 

 

 

 

 

Energy potential required to heat lnﬂuent
Ambient Temperature assuming a combustion efﬁciency of 80%
(°C) SLDM SMS MINI HC Overﬂow
0 32 98 1 1 1 1 1 1
5 28 84 95 95
10 23 70 79 79
1 5 1 8 56 63 63
20 14 42 48 48
25 9 28 32 32
30 5 14 16 16
35 0 0 0 0

 

 

 

 

 

 

Table 4.26 indicates that at 0°C, MINI and HC overﬂow cannot produce
enough energy from biogas to support the heating needs of the inﬂuent. Due to
the potential energy deﬁciency, alternative heat sources maybe required to
maintain temperature during extreme cold. Alternative heat options included
standby heat or waste heat recovery from AD efﬂuent. In addition, maximizing
biogas production, as shown in Tables 4.23 and 4.24 could increase energy yield
by nearly 50%, eliminating energy deﬁcient periods.

Waste heat recovered from an electrical generator (combined heat and
power unit) using biogas as its fuel is commonly used to heat anaerobic
digesters. Assuming that 40% of the biogas energy potential is recoverable as
waste heat through the cooling of the engine and exhaust, Table 4.27 was
generated to demonstrate the percent of the heat required that could be provided
by waste heat. Sand laden dairy manure has the potential to generate sufﬁcient

waste heat from combined heat and power to maintain the AD operating

123

temperature at ambient temperature of 0°C. In comparison, the AD heat

requirement could only be supported by waste heat from the SMS, MINI and HC

overﬂow to an ambient temperature of 20°C.

Table 4.27: Anaerobic dilester heating requirement from waste heat — GMF

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Amblent Waste heat required to maintain AD operating
Temperature temperature, %1
(°C) SLDM sms MINI HC Overﬂow
0 66
5 56
10 47
15 38
20 28 80 87 96
25 19 53 58 64
30 9 27 29 32
35 0 0 0 0
1Assuminga 40% heat recovery efﬁciency of the total biogas energy potential

 

In addition to the heat energy balance, the loss of energy revenue due to a

decreased amount of VS is critical to the economic viability of AD systems.

Table 4.28 summarizes the annual electrical energy potential from biogas

produced by an AD system using manure from the different sample locations

available at GMF. Table 4.28 is based on the mean mass of VS shown in Table

4.19. The change in VS during sand separation amounted to a reduction of

approximately $128,000 in electrical revenue when comparing SLDM to HC

overﬂow.

124

 

Table 4.28: Electrical generation potential of SSS products - GMF

 

 

 

 

 

 

 

 

 

 

 

 

AD Inﬂuent . ma" .
Source ..--.E'BF’mFFQ' POWEPEL-_§_'§9E'§§! 3991192-..
(kWhlyr)* (_yr)“

SLDM 3,791,313 $303,305

SMS 3,003,257 $240,261

MINI 2,605,598 $208,448

HC Overﬂow 2,194,887 $175,591
*Assuming 35% ﬂywheel efﬁciency and 90% on-time efﬁciency
”Assumes electrical value of $0.08 per kWh

 

Utilizing the data in Table 4.28, along with FS separation efﬁciency and
residual particle size data, system planners can value engineer the integration of
sand separation and AD systems. Economic and risk factors should be
considered when balancing between the level of sand separation, mixing costs,
the risk of sedimentation and impacts on biogas potential for new systems.
Similar, at existing facilities, an economic evaluation can provide guidance for the
allocation of funds for improving or abandoning SSS components. An example
of this is in regards to the operation of the MINI pits at GMF. Early in the AD
project, the MlNl's were identiﬁed as a cause of lost VS, however due to a
management decision, the MlNl’s were not abandoned. A system wide

economic evaluation may have provided different guidance to the management.

4.5.3 Veriﬁcation of energy analyses
Actual operational data from the electrical generator at GMF allowed for
the veriﬁcation of the energy analyses presented previously. Table 4.29 contains

electrical generation data using HC overﬂow as the AD feedstock from GMF.
125

 

Table 4.29: Operational data from the electricaﬁnerator - GMF

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Electrical production Generator Size

Year Month

MWhImonth kWh/month kWh/d kwnryr’ kW

March 108 107,590 3,471 1,266,785 145
April 81 80,825 2,694 983,371 112
May 110 110,221 3,556 1,297,763 148
June 50 49,801 1,660 605,912 69
2008 July 171 170,669 5,505 2,009,490 229
gust 65 64,501 2,081 759,447 87
September 24 23,806 794 289,640 33
October 15 15,497 500 182,465 21
November 58 57,960 1,932 705,180 81
December 57 56,860 1,834 669,481 76
January 129 129,443 4,176 1,524,087 174
20093 February 235 235,331 8,405 3,067,708 350
March 318 318,369 10,270 3,748,538 428
2008 mean 74 73,773 2,403 876,953 100
2009 mean 228 227, 714 7,617 2,780,111 317
Overall mean 109 109,298 3,606 1,316,144 150

 

 

 

 

 

 

 

 

‘Provided by North American Biofuels (Nan/in 2009)
2Extrapolation of monthly electrical production to annual value I I

 

I°Starting in mid-January 2009, 240,000 lblweek of ethanol syrup was added to the AD inﬂuent

 

 

The digester at GMF began receiving manure in July of 2007, but the
biogas utilization system, an 800 kW electrical generation, did not become
operational until March of 2008. Electrical production during 2008 was highly
variable due to common startup issues. Generation peaked in July of 2008 when
170,669 kWh were produced for the month, equivalent to an annual electrical
output of 2.01 million kWh/yr. Comparing the actual peak electrical output of the
AD system at GMF to the theoretical electrical potential of the HC overﬂow,

shown in Table 4.28, the actual operating system achieved 92% of the predicted

126

potential. The basis for Table 4.28 is the estimated mean mass of VS contained
in Table 4.19.

The combination of an electrical storm in August and the mixer failure in
September 2008 negatively impacted biogas production and electrical generation
for the remainder of 2008. By January of 2009, the system was restored to full
capacity and operating at the target temperature of 35°C.

To improve biogas production, condensed distillers soluble (syrup) from
ethanol production was added to the digester feedstock beginning in January of
2009. The target addition of syrup was 109,000 kg/week, but the actual mass
and timing of addition varied week to week. Based on the data in Table 4.28, it is

clear that the addition of syrup resulted in increased biogas production.

127

CHAPTER 5: SUMMARY AND CONCLUSIONS

5.1 Summary
The objectives of the research were to quantify separation efﬁciency of the

sand separation system, evaluate the particle size distribution of residual sand in
the manure stream to determine the potential for accumulation in the AD tanks,
predict the loss of volatile solids during sand separation and determine the
design implication of sand separation on anaerobic digester.

Quantiﬁcation of sand separation efﬁciency was attempted using
traditional mass balance, however the unstable ﬂow rate of several inputs and
outputs from a simple SSS did not produce results within the target closure of
:I:10%. To address difﬁculties with the traditional mass balance, a semi-empirical
mass balance technique was developed to evaluate the separation efﬁciency of a
complex SSS with multiple units and treatment levels, similar to GMF. This
technique used a combination of industry standards and farm-speciﬁc data to
predict daily manure production, bedding usage and manure and solids ﬂow
rates throughout a sand separation system. For the semi-empirical mass
balance the efﬂuent ﬂow rate of the system, at a minimum, should be measured
on a daily basis. In addition, samples should be collected and evaluated to
determine the solids concentration of the efﬂuent from each step of the SSS.
Usage of industry standard data was limited to predicting the baseline mass of
manure and bedding generated by the dairy cows on a daily basis. The manure
and solids ﬂow information was used to predict the FS separation efﬁciency of

128

 

the SSS, in addition to the change in VS resulting from the removal of FS. Mean
cumulative F S separation efﬁciency of the SSS was found to be 97% with a 95%
conﬁdence interval of 96% to 98%. Sand manure separation accounted for the
majority of the FS separation, with a conﬁdence interval of 83% to 90%.

Effectiveness of the semi-empirical mass balance compared to the
traditional mass balance was improved by using measured daily ﬂow rates and a
standardized data set to predict baseline ﬂow rates for which direct measurement
is not practical. While direct ﬂow rate measurement is desirable and would
reduce variability, safety and cost often are prohibitive on commercial operations.
Thus, the semi-empirical mass balance provides a cost effective and efﬁcient
method for qualitatively determining the separation efﬁciency on a variety of
commercial dairy farms.

For planning and design purposes, understanding the change in particle
size of new sand compared to residual sand provides the basis for predicting
scour velocity and sizing mixing systems. The mean particle size of the residual
sand was 0.195 mm compared to the average particle size of new sand, 0.69
mm. During system maintenance, one of the anaerobic digester tanks was
emptied after approximately ﬁfteen months of Operation, inspection of the tank
ﬂoor indicated that only 25 to 50 mm of sludge (sand and manure) had
accumulated, verifying the modeling technique. Theoretical mixing power
required to initiate and completely scour residual sand entering the anaerobic
digester was determined using the mean particle size of the residual sand. The

mixing system installed and operating in the anaerobic digester at GMF has
129

 

sufﬁcient power to achieve complete scour of the mean sand particle remaining
in the manure used as feedstock. Consequently, the model and veriﬁcation
steps result in a conclusion that the SSS is adequate to prevent AD system
operating difﬁculties.

Sand separation does, however, come with a cost in regards to
downstream treatment. Dilution water added during the sand separation process
more than doubled the energy needed to heat the HC overﬂow to the target
operating temperature. Further, the change in the VS mass throughout the SSS
reduced the theoretical biogas potential of the system efﬂuent compared to
SLDM. The treatment effect of the SSS on mass VS in manure was not found to
be statistically signiﬁcant. However, system planners should be aware of

potential changes in biogas production due to changes in VS mass.

5.2 General conclusions
Major research ﬁndings from the research are summarized below.

1. Quantifying sand separation efﬁciency using mass balance was not
effective due to large operational variations over the sample collection
interval. However, a semi-empirical mass balance approach was
developed to provide an effective method for acquiring necessary FS
separation efﬁciency design information over a wide range of dynamic
commercial SSS. Using the semi-empirical approach, treatment was

found to have a statistically signiﬁcant effect on the change in F8 mass

130

 

throughout the SSS, resulting in a FS separation efﬁciency conﬁdence
interval (95%) of 96% to 98%.

. Residual sand particles found in the sand separation system efﬂuent and
sludge of downstream units had a mean diameter of 0.195 mm compared
to new sand which had a mean diameter of 0.69 m. This resulted in
greater than a 50% decrease in the mean scour velocity. Observational
measurements conﬁrmed theoretical prediction, that the propeller mixing
system used in the AD at GMF (case study farm) achieved the scour
velocity of the mean residual sand particle size.

. While not found to have produced statistically signiﬁcant results on the
case study farm, the semi-empirical mass balance approach applied to
track changes in the mass of VS remaining in the manure after sand
separation, provided valuable information to system planners coping with
uncertain biogas production potential.

. Sand separation can remove sufficient sand to minimize negative impacts
on AD, however the addition of dilution water and the potential loss of VS
could potentially reduce the amount of biogas-derived energy available for
use elsewhere on the farm or as a salable product. System planners need
to consider changes in the AD feedstock due to sand separation when

selection heating and biogas utilization equipment.

131

 

CHAPTER 6: RECOMMENDATIONS FOR FUTURE WORK

The following recommendations are made with respect to areas that need

further research.

1.

Evaluate deleterious effect on microbial activity or biogas pr0duction
during anaerobic digestion of manure containing sand bedding or
residual sand. Compare other AD system to see if any common
problems exist other than the addition of dilution water and the
reduced mass of VS.

Investigate methods for improving biogas production from AD systems
using dilute materials as feedstocks. Included is the development of
hybrid systems that combine existing technologies for treating both
high and low solids feedstocks. Also, bolt on technologies such as
solid-liquid separation and membrane ﬁlters could be used to
decouple the solid and liquid retention times, thus increasing the
residence time of the organic matter in the AD.

Investigate methods for improving the heat energy balance for
systems using dilute feedstock by exchanging heat from the digester
efﬂuent with the cooler inﬂuent. Based on experiences at GMF, the
use of traditional shell and tube heat exchangers for recovering waste
heat from the digester efﬂuent may be impractical due to clogging
caused by the buildup of scale containing animal hair, organic matter

and crystalline formations believed to be struvite.

132

 

4. Explore the usefulness of computer-based models to optimize the
digester mixing system design to effectively control sedimentation,
reduce low velocity (dead) zones while maximizing biogas production.
Computational ﬂuid dynamic software could be used to model and
optimize anaerobic digester mixing for the feedstock characteristics.

5. Alternative uses of biogas should be evaluated to improve the revenue
potential of AD systems. Such alternatives may include operation of

the combined heat and power unit as a peak demand plants to

 

maximize electrical revenue or biogas upgrading to natural gas
standards and storage for on farm heating during cooler periods or as

transportation fuel in short range vehicles.

133

APPENDIX A
MDF Solids Data

Table A1: Solids concentration data, MDF

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sample Sample Moisture TS FS VS
ID Date (%) 1%) (%) (%)
FAN efﬂuent 06/27/06 95.2 4.83 1.72 3.11
HC underﬂow 06/27/06 93.0 7.03 2.54 4.49
Manure 06/27/06 83.8 16.18 1.94 14.23
Reclaimed
sand 06/27/06 8.7 91 .27 89.23 2.04
SLDM 06/27/06 79.6 20.38 12.01 8.36
SMS 06/27/06 93.8 6.20 2.49 3.70
FAN efﬂuent 07/12/06 97.3 2.72 0.83 1.89
HC underﬂow 07/12/06 97.5 2.50 0.74 1.76
Manure 07/12/06 81.7 18.27 3.23 15.04
New sand 07/12/06 17.65 82.35 78.66 3.69
New sand 07/12/06 5.05 94.95 93.73 1.22
SLDM 07/12/06 95.9 4.14 0.92 3.23
SMS 07/12/06 97.5 2.54 0.78 1.76
FAN efﬂuent 12/18/06 94.73 5.27 2.25 3.02
HC underﬂow 12/18/06 93.57 6.43 2.85 3.58
Manure 12/18/06 82.31 17.69 2.51 15.18
Reclaimed
sand 12/18/06 14.26 85.74 84.32 1.42
SLDM 12/18/06 82.68 17.32 9.08 8.24
SMS 12/18/06 92.61 7.39 3.74 3.64
FAN efﬂuent 12l21/06 96.29 3.71 1.36 2.35
HC underﬂow 12/21/06 95.61 4.39 1.54 2.85
Manure 12/21/06 83.47 16.53 3.04 13.49
Reclaimed
sand 12/21/06 8.66 91.34 89.41 1.93
SLDM 12/21/06 84.43 ' 15.57 8.40 7.17
SMS 12/21/06 94.32 5.68 2.60 3.08
FAN efﬂuent 01/03/07 96.24 3.76 1.25 2.51

134

 

Table A1: Solids concentration data, MDF continued

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

HC underﬂow 01/03/07 96.47 3.53 1.04 2.49
Manure 01/03/07 82.64 17.36 2.44 14.93
Reclaimed

sand 01/03/07 7.55 92.45 90.65 1 .80
SLDM 01/03/07 76.36 23.64 14.20 9.44
SMS 01/03/07 95.10 4.90 1.85 3.05
FRESH H20 03/09/07 99.91 0.09 -0.04 0.12
HC overﬂow 03/09/07 96.09 3.91 1.07 2.83
HC underﬁow 03/09/07 46.83 53.17 50.42 2.75
HC underﬂow 03/09/07 66.50 33.50 31.82 1.68
Recycle H20 03/09/07 95.84 4.16 1.39 2.77
Reclaimed

sand 03/09/07 34.34 65.66 64.56 1.10
Reclaimed

sand 03/09/07 71.51 28.49 27.29 1.20
SLDM 03/09/07 86.55 13.45 7.56 5.89
SLDM 03/09/07 85.00 15.00 9.26 5.74
SMS 03/09/07 95.94 4.06 1.63 2.43
SMS 03/09/07 96.94 3.06 1.01 2.05
FRESH H20 03/16/07 99.88 0.12 0.11 0.01
FRESH H20 03/16/07 99.74 0.26 0.24 0.02
FRESH H20 03/16/07 99.92 0.08 0.04 0.04
FRESH H20 03/16/07 99.96 0.04 0.03 0.02
HC overﬂow 03/16/07 96.89 3.11 0.86 2.26
HC overﬂow 03/16/07 95.30 4.70 1.29 3.41
HC overﬂow 03/16/07 91.25 8.75 4.21 4.54
HC overﬂow 03/16/07 97.95 2.05 0.69 1.36
HC overﬂow 03/16/07 95.35 4.65 1.34 3.31
HC overﬂow 03/16/07 93.67 6.33 2.32 4.01
HC underﬂow 03/16/07 64.17 35.83 34.07 1.75
Recycle H20 03/16/07 95.73 4.27 1.66 2.61
RecycleHZO 03/16/07 95.78 4.22 1 .68 2.55
Recycle H20 03/16/07 95.60 4.40 1.68 2.72 .
Recycle H20 03/16/07 95.46 4.54 1.84 2.70
Reclaimed

sand 03/16/07 18.25 81.75 78.25 3.49
Reclaimed

sand 03/16/07 43.87 56.13 54.98 1.15

 

 

 

135

 

 

 

Table A1: Solids concentration data, MDF continued

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Reclaimed

sand 03/16/07 18.37 81.63 78.86 2.77
Reclaimed

sand 03/16/07 16.48 83.52 80.32 3.20
Reclaimed

sand 03/16/07 18.90 81.10 79.59 1.51
SLDM 03/16/07 97.44 2.56 1.36 1.20
SLDM 03/16/07 74.92 25.08 16.94 8.14
SLDM 03/16/07 67.74 32.26 23.38 8.87
SLDM 03/16/07 89.34 10.66 7.50 3.16
SMS 03/16/07 97.60 2.40 0.54 1.86
SMS 03/16/07 90.62 9.38 5.30 4.08
SMS 03/16/07 97.02 2.98 0.76 2.22
SMS 03/16/07 91.93 8.07 3.79 4.28
FRESH H2O 05/16/07 99.92 0.08 0.05 0.03
FRESH H20 05/16/07 99.90 0.10 0.04 0.06
HC overﬂow 05/16/07 95.97 4.03 1.15 2.88
HC overﬂow 05/16/07 95.89 4.1 1 1.22 2.89
HC underﬂow 05/16/07 38.50 61.50 59.83 1.67
HC underﬂow 05/16/07 36.29 63.71 59.65 4.06
Recycle H20 05/16/07 95.68 4.32 1.47 2.85
Recycle H20 05/16/07 96.1 1 3.89 1.14 2.75
Reclaimed

sand 05/16/07 31.31 68.69 65.16 3.53
Reclaimed

sand 05/16/07 27.46 72.54 68.47 4.07
SLDM 05/16/07 72.15 27.85 18.34 9.51
SLDM 05/16/07 70.48 29.52 20.89 8.63
SMS 05/16/07 96.07 3.93 3.02 0.91
SMS 05/16/07 96.46 3.54 1.04 2.50
HC underﬂow 05/31/07 43.75 56.25 48.16 8.09
New sand 05/31/07 2.11 97.89 95.85 2.04
Reclaimed '

sand 05/31/07 18.24 81.76 79.66 2.10
SLDM 05/31/07 70.84 29.16 19.08 10.08
SMS 05/31/07 96.35 3.65 1.05 2.60

 

136

 

 

Table A2: TS concentration data, MDF

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Mean Standard Coefﬁcient
50:27:); TS Deviation of Variation Median Ngxnbgzzf
(%) (%) (%) (%)
Manure 15.2 5.0 33.2 16.9 6
SLDM 19.0 9.4 49.4 18.8 14
SMS 4.8 2.2 45.3 4.0 14
H0 Overﬂow 4.7 1.9 39.7 4.3 14
Table A3: FS concentration data, MDF
Sample Mean Standard Coefﬁcient Number of
Location FS Devration of VariatIon Median Samples
(%) (%) (%) (%)
Manure 3.0 1.1 35.3 2.8 6
SLDM 12.1 7.0 57.8 10.6 14
SMS 2.1 1.4 67.8 1.7 14
HC Overﬂow 1.6 1.0 61.2 1.3 14
Table A4: VS concentration data, MDF
Sample ma" SW?” “and?“ . Number of
Location VS Devration of VariatIon Median Samples
(%) (%) (%) (%)
Manure 13.0 4.0 30.5 14.6 6
SLDM 7.0 2.7 39.4 8.2 14
SMS 2.7 1.0 35.4 2.6 14
H0 Overﬂow 3.0 0.9 30.6 2.9 14

 

 

 

 

 

 

 

137

 

 

 

 

APPENDIX B

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

GMF Solids Data
Table B1: Solids concentration data from, GMF
Sample Sample Herd Sample Moisture TS FS VS
ID Location Management Date (%) (%) (%) (%)

Mantre Farm 3 SMS 5 Special Needs 05/09/06 83.5 16.48 5.14 11.34
MINI Farm 3 SMS 5 Special Needs 05/09/06 96.7 3.28 1.10 2.18
SLDM Farm 3 SMS 5 Special Needs 05/09/06 80.0 19.98 11.27 8.71
SMS Farm 3 SMS 5 Special Needs 05/09/06 96.2 3.82 1.31 2.51
Manure Farm 3 SMS 5 Special Needs 05/15/06 83.7 16.29 2.50 13.79
MINI Farm 3 SMS 5 Special Needs 05/15/06 96.4 3.63 1.10 2.53
SLDM Farm 3 SMS 5 Special Needs 05/15/06

SMS Farm 3 SMS 5 Special Needs 05/15/06 95.8 4.23 1.33 2.89
Mantre Farm 2 SMS 4 Lactating 05/17/06 86.6 13.36 2.32 11.04
MINI Farm 2 SMS 4 Lactating 05/17/06 95.0 5.03 1.44 3.59
SLDM Farm 2 SMS 4 Lactating 05/17/06 72.3 27.71 19.25 8.46
SMS Farm 2 SMS 4 Lactating 05/17/06 94.0 5.96 2.03 3.93
Mantre Farm 2 SMS 3 Lactating 05/19/06 86.8 13.21 1.13 12.08
MINI ’ Farm 2 SMS 3 Lactating 05/19/06 95.1 4.94 1.38 3.55
SLDM Farm 2 SMS 3 Lactating 05/19/06 76.3 23.73 15.80 7.93
SMS Farm 2 SMS 3 Lactating 05/19/06 94.5 5.51 1.84 3.67
Manure Farm 2 SMS 3 Lactating 05/22/06 87.2 12.80 1.32 11.48
Manure Farm 3 SMS 5 Special Needs 05/22/06 86.5 13.47 1.64 11.83
Manure Farm 3 SMS 6 Special Needs 05/22/06 87.3 12.67 1.70 10.97
MINI Farm 2 SMS 3 Lactating 05l22/06 95.4 4.58 1.25 3.32
MINI Farm 3 SMS 5 Special Needs 05/22/06 94.4 5.61 1.79 3.82
MINI Farm 3 SMS 6 Special Needs 05/22/06 91.2 8.82 2.28. 6.54
SLDM Farm 2 SMS 3 Lactating 05/22/06 71.9 28.10 19.21 8.89
SLDM Farm 3 SMS 5 Special Needs 05/22/06 71.1 28.86 20.86 8.00
SLDM Farm 3 SMS 6 Special Needs 05/22/06 72.5 27.54 18.85 8.69
SMS Farm 2 SMS 3 Lactating 05/22/06 94.2 5.77 2.05 3.72
SMS Farm 3 SMS 5 Special Needs 05/22/06 93.8 6.18 2.48 3.69
SMS Farm 3 SMS 6 Special Needs 05/22/06 95.2 4.76 1.71 3.06
Manure Farm 2 SMS 2 Lactating 05/26/06 86.7 13.34 1.52 11.82
Manure Farm 2 SMS 3 Lactating 05/26/06 87.0 13.01 1.38 11.64
Manure Farm 2 SMS 4 Lactating 05/26/06 87.9 12.11 1.46 10.65
Manure Farm 3 SMS 5 Special Needs 05/26/06 82.8 17.22 4.81 12.41
Mantre Farm 3 SMS 6 Special Needs 05/26/06 85.1 14.91 1.88 13.03

 

138

 

 

 

Table B1: Solids concentration data from, GMF continued

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MINI Farm 2 SMS 2 Lactating 05/26/06 95.7 4.32 1.08 3.23
MINI Farm 2 SMS 3 Lactatirg 05/26/06 95.7 4.32 1.26 3.06
MINI Farm 2 SMS 4 Lactating 05/26/06 96.8 3.17 0.88 2.28
MINI Farm 3 SMS 5 Special Needs 05/26/06 98.2 1.83 0.69 1.14
MINI Farm 3 SMS 6 Special Needs 05/26/06 93.8 6.23 2.22 4.01
SLDM Farm 2 SMS 2 Lactating 05/26/06 83.4 16.56 8.53 8.04
SLDM Farm 2 SMS 3 Lactating 05/26/06 71.4 28.62 20.43 8.20
SLDM Farm 2 SMS 4 Lactating 05/26/06 70.1 29.86 25.32 4.54
SLDM Farm 3 SMS 6 Special Needs 05/26/06 90.1 9.86 4.74 5.11
Manure Farm 2 SMS 1 Lactating 06/02/06 86.7 13.30 1.35 11.94
Manure Farm 2 SMS 2 Lactating 06/02/06 85.5 14.51 1.65 12.86
Manure Farm 2 SMS 3 Lactating 06/02/06 86.0 13.96 1.29 12.66
Manure Farm 2 SMS 4 Lactating 06/02/06 87.0 12.99 1.29 11.69
Manure Farm 3 SMS 5 Special Needs 06/02/06 85.6 14.38

Manure Farm 3 SMS 6 Special Needs 06/02/06 85.9 14.11 1.39 12.72
MINI Farm 2 SMS 1 Lactati_ng 06/02/06 94.9 5.10 1.38 3.72
MNI Farm 2 SMS 2 Lactating 06/02/06 97.1 2.95 0.94 2.01
MINI Farm 2 SMS 3 Lactating 06/02/06 93.4 6.59 1.93 4.66
MINI Farm 2 SMS 4 Lactatng 06/02/06 93.8 6.19 1.45 4.75
MINI Farm 3 SMS 5 Special Needs 06/02/06 95.3 4.74 1.29 3.44
SLDM Farm 2 SMS 1 Lactatng 06/02/06 55.5 44.54 35.32 9.22
SLDM Farm 2 SMS 2 Lactating 06/02/06 90.3 9.74 3.92 5.82
SLDM Farm 2 SMS 3 Lactating 06/02/06 67.4 32.58 23.97 8.60
SLDM Farm 2 SMS 4 Lactatirg 06/02/06 62.6 37.42 28.45 8.97
SLDM Farm 3 SMS 5 Special Needs 06/02/06 74.8 25.21 16.74 8.48
Manure Farm 2 SMS 3 Lactating 06/16/06 87.1 12.87 1.45 11.41
Manure Farm 2 SMS 4 Lactating 06/16/06 87.7 12.27 1.22 11.06
Manure Farm 3 SMS 5 Special Needs 06/16/06 85.0 14.99 1.99 13.01
Manure Farm 3 SMS 6 Special Needs 06/16/06 94.4 5.59 1.90 3.69
MINI Farm 2 SMS 3 Lactating 06/16/06 95.2 4.79 1.43 3.37
MINI Farm 2 SMS 4 Lactating 06/16/06 95.6 4.42 1.16 3.26
MINI Farm 3 SMS 5 Special Needs 06/16/06 97.3 2.67 1.05 1.62
MINI Farm 3 SMS 6 Special Needs 06/16/06 94.3 5.69 1.79 3.90
SLDM Farm 2 SMS 3 Lactating 06/16/06 69.6 30.36 21.85 8.51
SLDM Farm 2 SMS 4 Lactating 06/16/06 60.2 39.82 31.61 8.21
SLDM Farm 3 SMS 5 Special Needs 06/16/06 50.2 49.83 42.32 7.51
SLDM Farm 3 SMS 6 Special Needs 06/16/06 89.2 10.80 6.67 4.12
Manure Farm 2 SMS 1 Lactating 06/20/06 83.2 16.83 1.94 14.89

 

139

 

Table B1: Solids concentration data from, GMF continued

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Manure Farm 2 SMS 2 Lactating 06/20/06 81.0 19.04 4.72 14.32
Manure Farm 2 SMS 3 Lactating 06/20/06 85.7 14.28 1.88 12.39
Manure Farm 2 SMS 4 Lactating 06/20/06 87.0 13.03 1.53 11.50
Manure Farm 3 SMS 5 Special Needs 06/20/06 80.9 19.06 1.47 17.60
Manure Farm 3 SMS 6 Special Needs 06/20/06 85.2 14.84 2.11 12.73
MINI Farm 2 SMS 1 Lactating 06/20/06 92.6 7.39 2.18 5.22
MINI Farm 2 SMS 2 Lactating 06/20/06 95.1 4.87 1.26 3.61
MINI Farm 2 SMS 3 Lactating 06/20/06 95.1 4.89 1.53 3.35
MINI Farm 2 SMS 4 Lactating 06/20/06 94.6 5.37 1.46 3.91
MINI Farm 3 SMS 5 Special Needs 06/20/06 96.0 4.02 1.40 2.63
MINI Farm 3 SMS 6 Special Needs 06/20/06 94.5 5.54 1.87 3.68
SLDM Farm 2 SMS 1 Lactating 06/20/06 69.7 30.26 20.89 9.38
SLDM Farm 2 SMS 2 Lactating 06/20/06 75.3 24.70 15.33 9.37
SLDM Farm 2 SMS 3 Lactating 06/20/06 55.6 44.41 36.18 8.23
SLDM Farm 2 SMS 4 Lactating 06/20/06 69.6 30.37 21.59 8.78
SLDM Farm 3 SMS 5 Special Needs 06/20/06 67.0 32.95 24.40 8.55
SLDM Farm 3 SMS 6 Special Needs 06/20/06 89.6 10.45 5.30 5.14
SMS Farm 2 SMS 1 Lactating 06/20/06 93.5 6.49 0.85 5.64
SMS Farm 2 SMS 2 Lactatigg 06/20/06 93.2 6.82 2.25 4.57
SMS Farm 2 SMS 3 Lactating 06/20/06 96.1 3.95 1.50 2.45
SMS Farm 2 SMS 4 Lactating 06/20/06 94.0 5.98 2.18 3.80
SMS Farm 3 SMS 5 Special Needs 06/20/06 95.3 4.66 1.72 2.94
SMS Farm 3 SMS 6 Special Needs 06/20/06 94.7 5.33 1.87 3.46
Manure Farm 2 SMS 1 Lactating 06/29/06 84.2 15.81 1.72 14.09
Manure Farm 2 SMS 2 Lactating 06/29/06 85.1 14.94

Manure Farm 2 SMS 3 LactatinL 06/29/06 86.3 13.73 1.50 12.23
Manure Farm 2 SMS 4 Lactating 06/29/06 86.3 13.74 1.35 12.39
Manure Farm 3 SMS 5 Special Needs 06/29/06 84.1 15.90 2.51 13.40
Manure Farm 3 SMS 6 Special Needs 06/29/06 84.5 15.52 2.00 13.52
MINI Farm 2 SMS 1 Lactating 06/29/06 94.2 5.82 1.76 4.07
MINI Farm 2 SMS 2 Lactating 06/29/06 94.2 5.79 1.63 4.16
MINI Farm 2 SMS 3 Lactating 06/29/06 94.5 5.47 1.47 3.99
MINI Farm 2 SMS 4 Lactating 06/29/06 94.6 5.45 1.64 3.80
MINI Farm 3 SMS 5 Special Needs 06/29/06 95.2 4.83 1.46 3.37
MINI Farm 3 SMS 6 Special Needs 06/29/06 94.7 5.29 0.89 4.40
SLDM Farm 2 SMS 2 Lactating 06/29/06 68.3 31.66 23.40 8.27
SLDM Farm 2 SMS 3 Lactating 06/29/06 64.6 35.38 27.13 8.25
SLDM Farm 2 SMS 4 Lactating 06/29/06 57.2 42.83 34.40 8.43
SLDM Farm 3 SMS 5 Special Needs 06/29/06 76.4 23.57 15.14 8.42

 

 

140

 

 

Table B1: Solids concentration data from, GMF continued

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

SLDM Farm 3 SMS 6 Special Needs 06/29/06 88.9 11.06 5.95 5.11
SMS Farm 2 SMS 1 Lactating 06/29/06

SMS Farm 2 SMS 2 Lactating 06/29/06 92.4 7.62 3.00 4.62
SMS Farm 2 SMS 3 Lactating 06/29/06 93.7 6.31 2.16 4.15
SMS Farm 2 SMS 4 Lactating 06/29/06 94.1 5.90 2.22 3.68
SMS Farm 3 SMS 5 Special Needs 06/29/06 96.7 3.28 1.28 2.00
SMS Farm 3 SMS 6 Special Needs 06/29/06 92.8 7.20 3.08 4.12
Manure Farm 2 SMS 1 Lactating 07/03/06 85.7 14.29 1.88 12.41
Manure Farm 2 SMS 2 Lactating 07/03/06 86.5 13.49 2.24 11.25
Manure Farm 2 SMS 3 Lactating 07/03/06 86.4 13.61 1.41 12.21
Manure Farm 2 SMS 4 Lactating 07/03/06 84.4 15.55 2.50 13.06
Manure Farm 3 SMS 5 Special Needs 07/03/06 84.1 15.89 3.32 12.57
Mantre Farm 3 SMS 6 Special Needs 07/03/06 85.0 15.02 2.68 12.34
MINI Farm 2 SMS 1 Lactating 07/03/06 93.9 6.07 1.80 4.27
MINI Farm 2 SMS 2 Lactating 07/03/06 93.5 6.47 2.01 4.46
MINI Farm 2 SMS 3 Lactating 07/03/06 95.2 4.84 1.35 3.49
MINI Farm 2 SMS 4 Lactating 07/03/06 94.7 5.31 1.64 3.67
MINI Farm 3 SMS 5 Special Needs 07/03/06 97.2 2.79 1.15 1.63
MINI Farm 3 SMS 6 Special Needs 07/03/06 94.7 5.27 1.88 3.39
SLDM Farm 2 SMS 1 Lactaﬁgg 07/03/06 69.4 30.57 22.53 8.04
SLDM Farm 2 SMS 2 Lactating 07/03/06 69.6 30.43 22.06 8.37
SLDM Farm 2 SMS 3 Lactating 07/03/06 63.3 36.72 28.17 8.55
SLDM Farm 2 SMS 4 Lactating 07/03/06 64.9 35.11 27.10 8.01
SLDM Farm 3 SMS 5 Special Needs 07/03/06 66.5 33.53 25.86 7.67
SLDM Farm 3 SMS 6 Smecial Needs 07/03/06 95.3 4.66 1.88 2.79
SMS Farm 2 SMS 1 Lactating 07/03/06 92.1 7.87 3.20 4.66
SMS Farm 2 SMS 2 Lactating 07/03/06 91.4 8.59 3.94 4.65
SMS Farm 2 SMS 3 Lactating 07/03/06 93.2 6.78 2.74 4.04
SMS Farm 2 SMS 4 Lactatirg 07/03/06 93.3 6.67 2.76 3.90
SMS Farm 3 SMS 5 Sgcial Needs 07/03/06 97.2 2.79 1.17 1.63
SMS Farm 3 SMS 6 Syecial Needs 07/03/06 93.4 6.64 2.80 3.83
Manure Farm 2 SMS 1 Lactatipg 07/05/06 86.1 13.85 1.84 12.01
Manure Farm 2 SMS 2 Lactating 07/05/06 83.5 16.55 1.93 14.61
Manure Farm 2 SMS 3 Lactaﬂg 07/05/06 86.4 13.65 1.38 12.27
Manure Farm 2 SMS 4 Lactating 07/05/06 81.4 18.60 7.62 10.99
Manure Farm 3 SMS 5 Special Needs 07/05/06 79.2 20.79 6.74 14.05
Manure Farm 3 SMS 6 Special Needs 07/05/06 85.1 14.95 2.55 12.40
MINI Farm 2 SMS 1 Lactating 07/05/06 95.0 5.03 1.42 3.61
MINI Farm 2 SMS 2 Lactatigg 07/05/06 95.0 4.99 1.32 3.67
MINI Farm 2 SMS 3 Lactating 07/05/06 95.7 4.35 1.31 3.03
MINI Farm 2 SMS 4 Lactating 07/05/06 96.2 3.77 1.09 2.68
MINI Farm 3 SMS 5 Special Needs 07/05/06 94.8 5.22 1.53 3.69
MINI Farm 3 SMS 6 Special Needs 07/05/06 97.5 2.55 0.80 1.75

 

141

 

 

Table B1: Solids concentration data from, GMF continued

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

SLDM Farm 2 SMS 1 Lactating 07/05/06 66.1 33.86 25.54 8.32
SLDM Farm 2 SMS 2 Lactating 07/05/06 63.8 36.24 28.00 8.24
SLDM Farm 2 SMS 3 Lactaﬂig 07/05/06 61.4 38.63 30.80 7.83
SLDM Farm 2 SMS 4 Lactating 07/05/06 59.5 40.51 33.15 7.36
SLDM Farm 3 SMS 5 Special Needs 07/05/06 63.7 36.32 27.48 8.84
SLDM Farm 3 SMS 6 Special Needs 07/05/06 50.8 49.18 41.93 7.25
SMS Farm 2 SMS 1 Lactating 07/05/06 93.2 6.84 2.53 4.31
SMS Farm 2 SMS 2 Lactating 07/05/06 93.9 6.05 1.88 4.17
SMS Farm 2 SMS 3 Lactating 07/05/06 92.7 7.27 3.08 4.19
SMS Farm 2 SMS 4 Lactating 07/05/06 95.9 4.15 1.00 3.14
SMS Farm 3 SMS 5 Special Needs 07l05/06 93.4 6.62 2.50 4.12
SMS Farm 3 SMS 6 Sgecial Needs 07/05/06 97.2 2.78 0.91 1.87
Manure Farm 2 SMS 1 Lactating 07/28/06 82.3 17.70 1.72 15.97
Manure Farm 2 SMS 2 Lactating 07/28/06 85.1 14.91 2.04 12.87
Manure Farm 2 SMS 3 Lactatirg 07/28/06 86.1 13.92 1.50 12.42
Manure Farm 2 SMS 4 Lactating 07/28/06 81.7 18.26 1.83 16.43
Manure Farm 3 SMS 5 Special Needs 07/28/06 83.8 16.17 2.27 13.90
Manure Farm 3 SMS 6 Special Needs 07/28/06 85.1 14.88 1.77 13.11
MINI Farm 2 SMS 1 Lactating 07/28/06 95.8 4.24 1.20 3.04
MNI Farm 2 SMS 2 Lactating 07/28/06 96.4 3.61 1.08 2.53
MNI Farm 2 SMS 3 Lactating 07/28/06 95.0 5.05 1.51 3.53
MINI Farm 2 SMS 4 Lactating 07/28/06 96.2 3.83 1.16 2.66
MINI Farm 3 SMS 5 Special Needs 07/28/06 98.3 1.66 0.74 0.92
MINI Farm 3 SMS 6 Special Needs 07/28/06 95.3 4.75 1.57 3.18
SLDM Farm 2 SMS 1 Lactating 07/28/06 78.0 22.00 14.41 7.59
SLDM Farm 2 SMS 2 Lactating 07/28/06 78.5 21.55 13.87 7.68
SLDM Farm 2 SMS 3 Lactating 07/28/06 70.1 29.86 22.77 7.09
SLDM Farm 2 SMS 4 Lactating 07/28/06 58.5 41.54 34.37 7.17
SLDM Farm 3 SMS 5 Special Needs 07/28/06 76.0 23.97 16.49 7.48
SLDM Farm 3 SMS 6 Special Needs 07/28/06 92.7 7.30 2.78 4.52
SMS Farm 2 SMS 1 Lactating 07/28/06 95.7 4.33 1.25 3.08
SMS Farm 2 SMS 2 Lactating ' 07/28/06 95.8 4.24

SMS Farm 2 SMS 3 Lactating 07/28/06 94.3 5.67 1.92 3.76
SMS Farm 2 SMS 4 Lactating 07/28/06 94.4 5.63 1.37 4.26
SMS Farm 3 SMS 5 Special Needs 07/28/06 97.9 2.10 0.84 1.25
SMS Farm 3 SMS 6 Special Needs 07/28/06 94.7 5.34 1.74 3.60
Manure Farm 2 SMS 1 Lactating 08/07/06 86.0 14.00 1.92 12.08
Manure Farm 2 SMS 2 Lactating 08/07/06 84.4 15.62 2.27 13.34
Manure Farm 2 SMS 3 Lactating 08/07/06 85.3 14.68 1.82 12.87
Manure Farm 2 SMS 4 Lactating 08/07/06 84.4 15.62 2.36 13.27
Mantle Farm 3 SMS 5 Special Needs 08/07/06 84.7 15.33 2.66 12.67

 

142

 

Table B1: Solids concentration data from, GMF continued

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Mantre Farm 3 SMS 6 Special Needs 08/07/06 85.6 14.42 2.53 11.89
MINI Farm 2 SMS 1 Lactatirg 08/07/06 96.2 3.78 1.17 2.61
MINI Farm 2 SMS 2 Lactating 08/07/06 97.0 2.98 0.97 2.01
MINI Farm 2 SMS 3 Lactating 08/07/06 95.0 5.01 1.45 3.57
MINI Farm 2 SMS 4 Lactating 08/07/06 95.3 4.69 1.38 3.32
MNI Farm 3 SMS 5 Special Needs 08/07/06 95.1 4.88 1.61 3.27
MINI Farm 3 SMS 6 Special Needs 08/07/06 94.2 5.82 1.74 4.08
SLDM Farm 2 SMS 1 Lactating 08/07/06 85.2 14.85 6.79 8.06
SLDM Farm 2 SMS 2 LactatinL 08/07/06 86.0 14.00 6.29 7.71
SLDM Farm 2 SMS 3 Lactating 08/07/06 76.1 23.93 15.75 8.18
SLDM Farm 2 SMS 4 Lactating 08/07/06 68.8 31.17 23.50 7.67
SLDM Farm 3 SMS 5 Special Needs 08/07/06 62.3 37.71 30.34 7.37
SLDM Farm 3 SMS 6 Special Needs 08/07/06 87.4 12.57 7.91 4.66
SMS Farm 2 SMS 1 Lactating 08/07/06 95.7 4.29 1.22 3.07
SMS Farm 2 SMS 2 Lactating 08/07/06 97.2 2.83 0.95 1.88
SMS Farm 2 SMS 3 Lactating 08/07/06 94.4 5.56 1.69 3.87
SMS Farm 2 SMS 4 Lactating 08/07/06 95.1 4.90 1.50 3.40
SMS Farm 3 SMS 5 Special Needs 08/07/06 95.1 4.85 1.70 3.15
SMS Farm 3 SMS 6 Special Needs 08/07/06 95.3 4.71

Manu‘e Farm 2 SMS 1 Lactating 08/21/06 85.0 15.02 2.27 12.74
Manu'e Farm 2 SMS 2 Lactating 08/21/06 85.2 14.83 2.37 12.46
Manure Farm 2 SMS 3 Lactatirg 08/21/06 85.9 14.11 1.98 12.13
Manure Farm 2 SMS 4 Lactating 08/21/06 84.9 15.09 1.95 13.14
Manure Farm 3 SMS 5 Special Needs 08/21/06 83.4 16.63 2.79 13.84
Manure Farm 3 SMS 6 Special Needs 08/21/06 85.3 14.66 1.71 12.95
MINI Farm 2 SMS 1 Lactating 08/21/06 94.4 5.61 1.56 4.05
MNI Farm 2 SMS 2 Lactating 08/21/06 94.3 5.67 1.55 4.11
MINI Farm 2 SMS 3 Lactating 08/21/06 95.7 4.28

MINI Farm 3 SMS 6 Special Needs 08/21/06 94.4 5.55 1.73 3.82
MINI 08/21/06 95.9 4.11 1.30 2.81
SLDM Farm 2 SMS 1 Lactatipg 08/21/06 77.7 22.33 13.17 9.17
SLDM Farm 2 SMS 2 Lactating 08/21/06 81.9 18.07 8.82 9.25
SLDM Farm 2 SMS 3 Lactating 08/21/06 74.5 25.51 16.46 9.05
SLDM Farm 3 SMS 6 Special Needs 08/21/06 90.4 9.65 4.47 5.18
SMS Farm 2 SMS 1 Lactating 08/21/06 94.1 5.86 1.69 4.18
SMS Farm 2 SMS 2 Lactating 08/21/06 93.4 6.55 2.06 4.50
SMS Farm 2 SMS 3 Lactating 08/21/06 95.2 4.83 1.50 3.33
SMS Farm 3 SMS 6 Special Needs 08/21/06 92.5 7.47 2.70 4.77
SMS 08/21/06 94.0 6.04 2.43 3.61
SLDM 09/26/06 65.6 34.35 26.84 7.51
SLDM 09/28/06 69.5 30.45 24.08 6.37

 

143

 

 

Table B1: Solids concentration data from, GMF continued

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Manure Farm 2 SMS 1 Lactating 03/23/07 85.6 14.43 1.93 12.50
Manure Farm 2 SMS 2 Lactating 03/23/07 84.6 15.40 2.31 13.09
Manure Farm 2 SMS 3 Lactating 03/23/07 84.2 15.77 2.63 13.13
Manure Farm 2 SMS 4 Lactating 03/23/07 85.0 14.98 2.56 12.41
Manure Farm 3 SMS 5 Special Needs 03/23/07 83.8 16.25 2.39 13.86
Manure Farm 3 SMS 6 Special Needs 03/23/07 83.5 16.47 2.77 13.70
MINI Farm 2 SMS 1 Lactating 03/23/07 93.6 6.41 2.17 4.23
MINI Farm 2 SMS 2 Lactating 03/23/07 92.8 7.16 2.82 4.34
MINI Farm 2 SMS 3 Lactating 03/23/07 93.6 6.41 1.70 4.71
MINI Farm 2 SMS 4 Lactating 03/23/07 92.8 7.25 2.67 4.58
MINI Farm 3 SMS 5 Special Needs 03/23/07 95.3 4.71 1.64 3.07
MINI Farm 3 SMS 6 Special Needs 03/23/07 97.0 2.99 1.05 . 1.95
SLDM Farm 2 SMS 1 Lactating 03/23/07 78.0 21.97 14.09 7.87
SLDM Farm 2 SMS 2 Lactating 03/23/07 85.4 14.58 7.04 7.54
SLDM Farm 2 SMS 3 Lactating 03/23/07 63.9 36.11 28.55 7.56
SLDM Farm 2 SMS 4 Lactating 03/23/07 82.2 17.78 10.23 7.55
SLDM Farm 3 SMS 5 Special Needs 03/23/07 64.8 35.20 28.12 7.09
SLDM Farm 3 SMS 6 Special Needs 03/23/07 46.4 53.64 47.82 5.82
SMS Farm 2 SMS 1 Lactaﬁgg 03/23/07 90.4 9.61 4.97 4.64
SMS Farm 2 SMS 2 Lactating 03/23/07 91.9 8.12 2.95 5.17
SMS Farm 2 SMS 3 Lactating 03/23/07 88.9 11.12 6.36 4.76
SMS Farm 2 SMS 4 Lactating 03/23/07 86.6 13.42 7.78 5.64
SMS Farm 3 SMS 5 Special Needs 03/23/07 93.4 6.61 2.66 3.95
SMS Farm 3 SMS 6 Special Needs 03/23/07 93.2 6.85 3.25 3.60
MINI Farm 2 SMS 4 Lactating 06/26/07 95.0 5.03 1.69 3.33
SLDM Farm 2 SMS 2 Lactating 06/26/07 64.9 35.15 27.36 7.79
SMS Farm 2 SMS 3 Lactating 06/26/07 93.1 6.91 3.14 3.78
MINI 07/10/07 94.5 5.53 1.73 3.80
SLDM 07/10/07 62.3 37.75 30.25 7.50
SMS 07/10/07 92.9 7.10 3.20 3.90
MINI Farm 2 SMS 3 Lactating 07/20/07 94.1 5.85 1.84 4.01
SLDM Farm 2 SMS 1 Lactating 07/20/07 65.9 34.07 25.11 8.96
SMS Farm 2 SMS 2 Lactating 07I20/07 91.3 8.73 4.26 4.46
Table 32: Total solids concentration data, GMF
Mean Standard Coefﬁcient
Sample . . . . Number of
Location TS DeVIatIon of VariatIon Median Samples
(%) (%) (%) (%)

Feces 14.9 2.0 13.7 14.9 70

SLDM 28.3 11.0 36.4 30.3 69

SMS 6.0 2.0 33.5 6.0 56

MINI 4.9 1.3 26.0 5.0 71

HC Overﬂow 4.1 1.2 27.6 4.4 40

 

 

 

 

 

 

 

144

 

 

Table B3: Fixed solids concentration data, GMF

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sample - Mean S@Q_Q§r9_____c_9§fﬁ.9igﬂt__ Number of
Location FS DeVIation of VanatIon Median Samples
(%) (%) (%) (%)
Feces 2.2 1 .2 60.1 1 .9 68
SLDM 20.7 10.4 47.8 21.9 69
SMS 2.3 1.3 63.0 2.1 54
MINI 1.5 0.4 28.8 1.4 70
H0 Overﬂow 1.1 0.5 42.6 1.2 40
Table B4: Volatile solids concentration data, GMF
Sample Mean Standard Coefﬁcient Number of
Location VS Deviation of Variation Median Samples
(%) (%) (%) (%)
Feces 12.6 1.7 13.3 12.5 68
SLDM 7.6 1.4 17.6 8.0 69
SMS 3.7 0.9 24.4 3.8 54
MINI 3.4 1.0 27.0 3.5 70
HC Overﬂow 3.0 1.1 34.1 3.2 40

 

 

 

 

 

 

 

Table B5: Individual FS, Type III tests of ﬁxed effects (ANOVA), GMF

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Degrees Denominator ,
Effect of Degrees of F Value Pr > F
Freedom Freedom
trt 3 31.43 23.39 <0.0001
_gIgt 1 49.55 0.1 1 0.7464
trt*mgt 3 31.43 0.35 0.7912
Table 86: Individual FS, trt least squares means, GMF
Level Estimate Standard Degrees of t Pr > I'll
Error Freedom value
Manure 8.2886 0.3658 104.8 22.66 <0.0001
SLDM 8.7643 0.27 39.63 32.47 <0.0001
SMS 10.6894 0.1817 7.841 58.83 <0.0001
MINI 9.2301 0.2326 22.4 39.68 <0.0001

 

 

 

 

 

 

 

145

 

 

 

 

Table B7: Individual FS, trt least squares means conﬁdence interval, GMF

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Treatment Mean Standard 95% Conﬁdence Interval
Level Error Lower Upper Count
(kg/d) (kg/d) (kg/d) (kg/d)
Manure 3,978 1 .455 1 .926 8,217 68
SLDM 43,887 7,975 28,821 66,828 64
SMS 10,199 2,373 6,299 16,515 50
MINI 6.402 1 ,728 3.709 1 1 .049 67
Table 88: Individual FS, Differences of trt least squares means, GMF
Treatment Level Estimate “award “9”” °f t Pr > |t|
Error Freedom value
Manure vs. SLDM -2.4008 0.3923 62.08 -6.12 <0.0001
SLDM vs. SMS 1.4593 0.2728 16.57 5.35 <0.0001
SMS vs. MINI -0.4657 0.3383 37.59 -1.38 0.1768
Table B9: Individual VS, Type III tests of ﬁxed effect (ANOVA), GMF
Degrees Denominator
Effect of DeLees of F Value Pr > F
Freedom Freedom
trt 2 1 8.09 0.2413
ﬁt 1 1 9.52 0.1995
trt*mgt 2 1 0.51 0.7022
Table B10: Individual VS, trt least sguares means
Triatment Estimate Standard Degrees of t Pr > I 11
evel Error Freedom value
SLDM 21 .259 1 .326 1 16.03 0.0397
SMS 18,738 1,347 1 13.91 0.0457
MINI 15.846 1,321 1 12 0.0529

 

 

 

 

 

 

 

146

 

 

 

 

Table 811: Individual VS, trt least squares means conﬁdence interval, GMF

 

 

 

 

 

 

 

 

 

 

 

 

Treatment Mean Standard 95% Conﬁdence Interval
Level Error Lower Upper Count
(Lg/d) (kgld) (kg/d) (kg/d)
SLDM 21,259 1,326 4.410 38.109 64
SMS 18.738 1.347 1 .626 35,851 50
MINI 15,846 1 .321 -937 32,629 67

 

Table B12: Individual VS, Differences of trt least squares means, GMF

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Treatment Standard De rees of t
Level Estimate Error Fgeedom value Pr > "I
SLDM vs. SMS 2,520 1.379 1 1.83 0.3187
SMS vs. MINI -2.892 1.373 1 -2.11 0.2821
Table B13: HC FS, Type III tests of ﬁxed effects (ANOVA), GMF
Degrees Denominator
Effect LN.-- MQL_____-999'99§ 9L F Value Pr > F
Freedom Freedom
trt 1 20 13.21 0.0017
Table B14: H0 PS, trt least squares means
Treatment Estimate Standard Degrees of t Pr > N
Level Error Freedom value
MINI 6,395 269 20 23.79 <0.0001
HC overﬂow 5,014 269 20 18.65 <0.0001
Table B15: HC FS, trt least squares means conﬁdence interval, GMF
Standard 95% Conﬁdence Interval
Treatment Mean —-—
Love. -____.____--___._ _, -EI'IQL____..._-ML2_W€‘L-AL._QRPSL_. 00"”
(kg/d) (kg/d) (kgld) ikg/d)
MINI 6.395 269 5,835 6,956 12
HC overﬂow 5.014 269 4.453 5,574 12

 

 

 

 

 

 

 

147

 

 

 

 

 

Table B16: HC FS. Differences of trt least squares means, GMF

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Treatment . Standard Degrees of t *
Level Estimate Error Freedom value Pr > |t|
HC vs. MINI -1,381 .380 20 -3.63 0.0017 I
Table B17: HC VS, Type III tests. of ﬁxed effects (ANOVA), GMF
Degrees Denominator
Effect of Degrees of F Value Pr > F
Freedom Freedom
trt 1 20 1.36 0.2568
Table B18: HC VS, trt least squares means
Treatment . Standard Degrees of t
Level Estimate Error Freedom value Pr > m
MINI 15.154 895 20 16.94 <0.0001
HC overﬂow 13.677 895 20 15.29 <0.0001

 

 

 

 

 

 

 

 

Table B19: HC VS, trt least squares means conﬁdence interval, GMF

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Standard 95% Conﬁdence Interval
Treatment Mean - —» — _-
Level Error Lower Upper Count
(kgld) (kgld) (kg/d) (kg/d)
MINI 15.154 895 13.288 17.020 11
HC overﬂow 13,677 895 11,81 1 15.543 11
Table 320: HC VS, Differences of trt least squares means, GMF ,
Treatment Estimate Standard Degrees of t Pr > III
Level Error Freedom value
HC vs. MINI -1,477 1,265 20 -1.17 0.2568 |

 

 

 

 

 

 

 

148

APPENDIX C

MDF Sand Separation Mass Balance Data

Table C1: SSS ﬂow rate (le), MDF

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sample Sample Event Mean Standard Coefficient
Point Description 1 2 3 Deviation of Variation
(Hal M) (U8) (U8) (U8)
1 Fresh water 0.16 0.20 0.19 0.18 0.02 10%
2 Recycled water 3.71 3.71 4.46 3.96 0.43 11%
3 SLDM 0.91 0.92 0.97 0.93 0.03 ' 4%
4 SSS reclaimed sand 0.05 0.06 0.11 0.07 0.03 42%
5 SMS quu‘d efﬂuent 4.26 3.91 5.30 4.49 0.72 16%
53 HC input 8.77 8.58 9.56 8.97 0.52 8%
6 HC overﬂow 7.91 7.73 8.02 7.89 0.15 2%
7 HC underﬂow 0.12 0.41 0.17 0.24 0.15 65%
SMS dilution ratio: 4.1 4.0 4.6 4.2 0.3
Table c2: sss ﬂow rate (m3lhr), MDF
Sample Sample Event Mean Standard Coefﬁcient
Point Description "m -1...--.- 2 3 ‘ Deviation of Variation
(m’m r) (ma/hr) (ma/hr) (ms/hr) (ma/hr)
1 Fresh water 0.58 0.71 0.69 0.7 0.1 10%
2 Recycled water 13.37 13.36 16.07 14.3 1.6 11%
3 SLDM 2.75 2.89 3.12 2.9 0.2 6%
4 SSS reclaimed sand 0.18 0.22 0.39 0.3 0.1 42%
5 SMS liguid efﬂuent 15.33 14.08 19.07 16.2 2.6 16%
5a HC imut 24.23 23.72 26.43 24.8 1.4 6%
6 HC overﬂow 21.86 21.36 22.16 21.8 0.4 2%
7 HC underﬂow 0.34 1.14 0.48 0.7 0.4 65%
SMS percent closu'e: 9% 21% 4% 1 1% 9%
HC percent clostre: 8% 5% 14% 9% 5% _
SSS percent closu'e: -32% 27% -13% -24% 10% __
SMS dilution ratio: 4.9 4.6 5.2 4.9 0.3 _ _ _’
1SLDM ﬂow rate based was based on the daily measured piston pimp cycle time (down and Ip-stroke)
prime the up-stroke. .-.-..
LZCyclone ﬂow rate was based on a measued mrtime of 46 mintﬂe per hour.___ ___ 2---- _ 2

 

 

 

 

149

Table C3: Material density, MDF

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sample 1 Sampl; EN.“ 3 Mean gutted Coefﬁcient
Point °°°°"'°°°" 3 3 3 °" 3 " ofVarlatlon
Ike/m (kglm I IkoIm’I Ike/m I Ike/m I
1 Fresh water 997 1,005 1.000 1.000 4 0%
2 Recycled water 1,015 1.037 1,028
3 SLDM1 1.092 1.003 1,048
4 sss reclaimed sand 2,038 1.708 1,617 1,787 221 12%
5 SMS liquid eﬁluent 1,024 1,059 1,035 1,039 18 2%
5a HC input2 1,024 1,059 1.035 1,039 18 2%
6 HC overﬂow 1.026 1,004 1.024 1.018 12 1%
7 HC mderﬂow 1,653 1,149 1.371 1,391 253 18%

 

 

 

 

Table C4: SSS mass ﬂow rate, MDF

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sample Sample Event Mean Standard Coefﬁcient
Point Description ..-h_,1____2____- 2 3 Deviation of Variation
(kg/hrI Ikﬂhq Ikg/hrI (kg/MI (kg/hr)
1 Fresh water 580 710 691 660 70 1 1%
2 Recycled water 13,573 10,987 16,669 13,743 2,845 21%
3 SLDM1 .1007 2,694 3,127 2.943 223 8%
4 SSS reclaimed sand 361 384 629 458 148 32%
5 SMS liquid eﬂ'luent 15.695 14.913 19.730 16.779 2.585 15%
5a HC input2 24.813 25,125 27.344 25.761 1,380 5%
6 HC overﬂow 22,443 21,449 224689 22,194 656 3%
7 HC underﬂow 564 1,306 654 841 405 48%
SMS percent closue: 9% 3% 4% 5% 4%
HC percent closue: 7% 9% 15% 10% 4%
SSS percent closue: -33% -52% -14% -33% 19%
1SLDM ﬂow rate based was based on the daily measued piston pump cycle time (down and tp-stroke)
minus the up—stroke.
2Cyclone ﬂow ratewas based on a measured rmtime of46 minute per hour, m

Table C5: TS concentration MDF

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sample Sample Event Mean Standard
Point Description 1 2 3 Deviation

(%) (%) L711 (%) 1%)

1 Fresh water 0.1 0.2 0.1 0.1 0.1

2 Recycled water 4.2 4.2 4.5 4.3 0.2

3 SLOM1 14.1 20.0 10.7 14.9 4.7

4 SSS reclaimed sand 59.1 72.5 77.2 69.6 9.4

5 SMS liquid efﬂuent 3.9 5.7 5.6 5.0 1.0

5a HC input2 3.9 5.7 5.6 5.0 1.0

6 HC overﬂow 3.9 5.0 4.3 4.4 0.6

 

 

150

Table C6: SSS TS mass ﬂow rate, MDF

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sample Event Standard
333:? Description 1 2 3 Mean Deviation 3273::ng
(kg/hrI Ikﬂhrz (kg/hrI Ikg/hrI (kg/hr)
1 Fresh water 0.5 1 .2 0.4 0.7 0.4 59%
2 Recycled water 565 467 745 592 141 24%
3 SLDM1 425 539 333 432 103 24%
4 SSS reclaimed sand 213 279 485 326 142 44%
5 SMS liquid eﬂ‘luert 610 844 1095 850 243 29%
5a HC input2 964 1422 1518 1.301 296 23%
6 HC overﬂow 876 1073 974 975 98 10%
7 HC urlderﬂow 241 468 355
SMS percent closue: 133% 124% 53% 104% 44%
HC percent clostre: 86% 92% 89%
SSS percent closue: 90% 66% 65% 73% 14%
1SLDM ﬂow rate based on the measurgd piston punp downstroke, upstroke deducted
2Cyclone ﬂow rate based on measued rtntime of 46 mintte per hour
Table C7: FS concentration MDF
Sample Sample Event Mean Standard
Point Description 1 2 3 _ DeVIatlon
t% L (%l (%) (%) (%)
1 Fresh water 0.0 0.2 0.0 0.1 0.1
2 Recycled water 1.4 1.7 1.8 1.6 0.2
3 SLDM1 8.3 13.9 7.5 9.9 3.5
4 SSS reclaimed sand 57.9 69.9 74.7 67.5 8.6
5 SMS liquid efﬂuent 1.5 2.8 2.3 2.2 0.6
5a HC input2 1.5 2.8 2.3 2.2 0.6
6 HC overﬂow 1.1 1.8 1.4 1.4 0.4
7 HC underﬂow 40.6 34.1 37.4 4.6

 

 

 

 

 

 

 

 

 

151

Table CS: SSS FS mass ﬂow rate, MDF

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sample Event Standard
8:32? Descriptlon 1 2 3 “a" Deviation vamt
(km (kg/hr) (1921;) _(gg/hr) Alb/hr)
1 Fresh water -0.2 1 .1 0.2 0.4 0.7 174%
2 Ragged water 188.5 183.2 292.6 221 62 28%
3 SLDM' 250.0 375.1 234.7 287 77 27%
4 SSS reclaimed sand 209.3 268.3 469.9 316 137 43%
5 SMS liquid efﬂuent 238.6 411.6 452.2 367 113 31%
5a HC input2 377.2 693.5 626.7 566 167 29%
6 HC overﬂow 240.3 381.8 324.2 315 71 23%
7 HC tnderﬂow 229.1 445.1 337
SMS percent closue: 33% 32% -75% -3% 62%
H0 percent closure: -24% -19% -22%
666 percent closure: -3% -16% -51% -23% 25% _ _ 7y
1SLDM ﬂow rate based on the measu'ed piston ptmp downstroke, Ipstroke deducted
2Cyclone ﬂow rate based on measured runtime of 46 minute per hour 2.- 7,
Table C9: VS concentration. MDF
Sample Event Standard
533:? Description _ 1 __2 3 ___ M”: Deviatipnmq
(%) (%) (%) (%) (%)
1 Fresh water 0.1 0.0 0.0 0.1 0.1
2 Recycled water 2.8 2.6 2.7 2.7 0.1
3 SLDM 5.8 6.1 3.2 5.0 1.6
4 SSS reclaimed sand 1.1 2.6 2.4 2.1 0.8
5 SMS liquid efﬂuent 2.4 2.9 3.3 2.8 0.4
5a HC input - 2.4 2.9 3.3 2.8 0.4
6 HC overﬂow 2.8 3.2 2.9 3.0 0.2
7 HC underﬂow 2.2 1.8 2.0 0.3

 

 

 

 

 

 

 

 

 

152

 

1 2 -L In 1. I p p - T.I ~ ~ ~ ~lnl~ s
h—IbmrrF_—_L_I— FhFI—IL--._ TIP

 

 

Table C10: SSS VS mass ﬂow rate, MDF

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sample Event Standard
833:? Description 1 2 3 “m" Deviation 333:3:
(km Ikgan (kg/ii!) _(llglhr) (lb/hr)
1 Fresh water 0.7 0.1 0.2 0.3 0.3 97%
2 Recycled water 376.6 283.3 452.0 371 84 23%
3 SLDM1 175.1 163.6 98.8 146 41 28%
4 $58 reclaimed sand 4.0 10.2 15.4 10 6 58%
5 SMS liquid efﬂuent 371.3 432.3 642.9 482 142 30%
5a HC input2 587.0 728.3 891.0 735 152 21%
6 HC overﬂow 636.2 691.2 650.1 659 29 4%
7 HC underﬂow 12.3 22.9 18
SMS percent clostre: 34% 6% -19% 7% 27%
HC percent clostre: -10% 2% -4%
SSS percent closue: -16% ~57% -21% -31% 22%
1529819»: rate begonbsmeasueq pistonpune smegma-arm clogs-2mm e e .e .. . _
2Cyclone ﬂow rate based on measued ruitime of 46 minute per hotr
Table C11: SSS TS separation efﬁciency, MDF
Sample Event Standard
Component 1 2 3 Mean Deviation
SMS: 17% 19% 45% 27% 16%
HC: 25% 33% 29%
SSS: 22% 28% 45% 31 % 12%
Table C12: SSS FS separation efﬁciency. MDF
Component Sample Event Mean Standard
1 2 3 Deviation
SMS: 31 % 27% 89% 49% 35%
HC: 61 % 64% 62%
SSS: 48% 48% 89% 62% 24%
Table C13: Piston pump speed measurements
Component Sample Event Mean Standard Coefﬁcient
1 2 3 Deviation of Variation
Down-stroke. s: 157.6 170.8 211.5 180.0 28.1 16%
Up-stroke. s: 29.4 25.8 23.7 26.3 2.9 11%
Cycle time, s: 187 196.6 235.2 206 25.5 12%
Pump cycles per hr. 19.3 18.3 15.3 17.6 2.1 12%
On-time, min/hr: 50.6 52.1 54.0 52.2 1.7 3%

 

 

 

 

 

 

 

 

153

 

 

Table C14: SMS efﬂuent rate, 45 minutes sample period

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1 Time Weig ht Volume Density Flow Rate Mass Flow F8 F8 Flow
sampk’ # (min) (kg) (m3) (kg/m3) (m3/min) (kg/min) (%) (k Imln

1 2 10 0.01 919.10 0.5 480 0.62 2.99
2 4 13.409 0.01 924.32 0.4 374 1.62 3.92
3 6 14.773 0.02 928.09 0.5 457 2.62 4.86
4 8 12.955 0.01 931.81 0.5 508 3.62 3.74
5 10 12.727 0.01 915.46 0.5 468 4.62 4.35
6 12 15.682 0.02 937.72 0.4 390 5.62 3.98
7 14 16.591 0.02 935.71 0.6 ‘ 524 6.62 8.11
8 16 16.818 0.02 937.87 0.5 476 7.62 5.85
9 18 15.455 0.02 935.40 0.5 444 8.62 9.37
10 20 16.591 0.02 914.92 0.3 312 9.62 7.20
11 22 13.864 0.02 905.35 0.6 520 10.62 7.01
12 24 12.5 0.01 939.98 0.6 521 11.62 11.80
13 26 17.727 0.02 935.98 0.5 467 12.62 7.27
14 28 17.045 0.02 939.98 0.5 424 13.62 6.78
15 30 15.455 0.02 913.13 0.5 444 14.62 9.73
16 32 15.227 0.02 933.02 0.5 431 15.62 9.27
17 34 17.955 0.02 938.01 0.5 472 16.62 7.78
18 35.5 15.227 0.02 921.64 0.6 513 17.62 7.45
19 37 15.227 0.02 933.02 0.4 412 18.62 7.91
20 38.5 13.182 0.01 934.61 0.4 390 19.62 7.59
21 40 15.682 0.02 937.72 0.5 485 20.62 7.82
22 41.5 12.5 0.01 912.34 0.4 408 21.62 5.99
23 43 15 0.02 942.36 0.4 393 22.62 6.00
24 44.5 13.636 0.01 914.58 0.4 401 23.62 8.15

 

 

'Bold 91820111992 .oﬂeitﬂiqydmcmone 0”

 

 

 

 

154

 

 

..H... --_ ...U I

 

ﬁr

APPENDIX D

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

GMF Particle Size Data
Table D1: New sand PSD
-§.iov1-_-+w-2__ _ _ _ ___ Persegtﬂagintl _ﬁ-_*-__._

Openlng Sample Event Mean Sara:
nnml 01105107 06116107 06126107 07120107 10129107 11109107 11130107 °"
4.75 100 100 99 99 99 100 100 100 1
2.36 87 88 87 85 85 91 90 88 2
1.18 66 69 69 68 64 78 70 69 4
0.60 41 45 49 48 43 53 39 45 5
0.30 7 12 17 16 16 8 7 12 5
0.149 1 2 2 3 3 1 1 2 1
0.074 0 1 0 1 0 1 0 0 1

Pan 0 0 0 0 0 0 0 0 0
Table DZ: Reclaimed sand PSD
Sieve y , y , y 7 Percent Passing y ,, . y .222--- __

Opening Sample Event Mean 3:11:50:
(mm) 01105107 06116107 06126107 07120107 10129107 11109107 11130107
4.75 100 100 100 99 99 98 100 99 1
2.36 87 85 83 85 72 80 90 83 6
1.18 64 63 60 65 43 55 73 60 9
0.60 36 35 32 39 16 28 44 33 9
0.30 3 4 4 6 1 4 3 4 2
0.149 0 0 1 0 0 0 0 0 0
0.074 0 0 0 0 0 0 0 0 0

Pan 0 0 1 0 0 0 0 0 0
Table D3: HC underﬂow PSD 3
Sieve Percent Passing j Standard

Opening Sample Event I Mean Deviation
(mm) 0611 8I07 06I26I07 07/20I07 10I29/07 11I09/07 L
4.75 99 1 00 100 86 79 93 10
2.36 94 96 97 79 71 87 12

1 .18 80 85 88 68 64 77 1 1
0.60 61 62 75 52 57 61 9
0.30 35 36 47 32 34 37 6
0.149 13 13 1 9 4 8 5
0.074 5 3 0 1 1 2 2

Pan 4 1 0 0 0 1 2

 

 

 

 

 

 

 

 

 

155

 

Table D4: HC overﬂow PSD

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

jigs £9.r¢9_n_tPassi09 Standard
Opening Sample Event Mean Deviation
(mm) 01I05IO7 01128I08 01128l08 04/11/08 04/11/08
4.75 1 00 93 97 99 98 97 3
2.36 99 90 96 98 96 96 3
1 .1 8 98 88 94 96 94 94 4
0.60 91 84 90 92 89 89 3
0.30 51 68 72 85 83 72 14
0.149 13 28 28 61 65 39 23
0.074 0 7 7 16 21 1 0 8
Pan -1 0 0 -5 -1 -1 2
Table D5: Tank sludggPSD
Sieve 3 lo E Pergent Passing
amp ven
02:11:?“ 05123107 10I02/08 Mean 3:35:
1 2 Mean 1 2 Mean
4.75 98 99 99 92 94 93 96 3
2.36 92 96 94 89 91 90 92 3
1 .18 83 90 87 85 88 87 87 3
0.60 68 81 75 78 83 81 78 7
0.30 47 62 55 69 75 72 63 12
0.149 20 29 25 45 53 49 37 15
0.074 3 5 4 1 1 15 13 9 6
Pan 0 0 0 -2 0 -1 -1 1
* 5/23/07 post digester equalization tanke sludge. 10/2/2008 AD tartr 3 sludge g

 

 

 

 

156

 

GMF Solids Data from Other Sample Locations

APPENDIX E

Table E1: Anaerobic Digester Efﬂuent

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sample Sample Moisture TS FS VS
ID Date (%) (%) (%) (%)
Digesterefﬂluent 1012512007 98.83 1.17 1.16 0.01
Digesterefﬂluent 1012512007 93.48 6.52 1.11 5.41
Digester efﬂluent “1012912007 7 '“ “ 97.29 ” ””271“ A 7 1.20 ””7115"?
Digesterefﬂluent 11/11/2007 97.49 2.51 1.08 1.43]
Digestereﬂlluent 1111312007 97.37 2.63 1.17 1.46]
Digesterefﬂluent 1113012007 97.11 2.89 1.00 1.89]
Digestereiﬂiuent 111301200? 97.54 2.46 0.84 1.62]
Tar—9331575516551 121312007 97.33 2.67 1.02 1.64|
Digester eﬂlluent 12/5/2007 98.66 1.34 0.64 0.70I
Digesterefﬂluent 12/9/2007 98.07 1.93 0.87 1.07
Moisture TS FS vs
(%) (%) (%) (%)
Average 97.3 2.7 1.0 1.7
Standard deviation 1.5 1.5 0.2 1.4
Median 97.4 2.6 1.1 1.5
Count 10 10 10 1g
Table E2: HC Underﬂow ,
Sample Sample Epicure TS FS VS I
In Date (%) (%) (%) (%) 1
HC underﬂow 312312007 77.29 22.71 15.45 7.25]
HC underﬂow 612612007 89.60 10.40 4.38 6.02]
HC underﬂow 7/9/2007 96.64 3.36 0.92 2.44I
HC underﬂow 711012007 93.07 6.93 2.55 4.38|
HC underﬂow 712012007 81.10 18.90 11.74 7.16|
Wit—naérﬂ‘JwW‘z‘lmom 98.14 1.86 0.57 1.29|
HC underﬂow 121912007 98.45 1.55 0.45 1.09
Moisture 1's FS vs
(%) (%) (%) (%)
Average 90.6 9.4 5.2 4.2
Standard deviation 8.5 8.5 6.0 2.7
Median 93.1 6.9 2.6 4.4
Count 7 7 7 7

 

 

 

 

 

 

 

157

I... .

 

Table E3: Post AD Equalization Tank Sludgg

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sample Sample Moisture TS FS VS I
In Date 1%) (%) (%I (%) |
Tank Sludge I 5/23/2007 53.05 46.95 36.61 1035'
Tank Sludge I 5/23/2007 59.15 40.85 29.72 11.13I
Tank Sludge ll 5/23/2007 44.69 55.31 47.16 8.15
Tank Sludge ll 5/23/2007 51.08 48.92 38.99 9.94
Moisture TS FS VS
(%) (%) (%) (%)
Average 52.0 48.0 38.1 9.9]
Standard deviation 6.0 6.0 7.2 1.3I
Median 52.1 47. 9 37. 8 10.1I
Count 4 4 4 4|
Table E4: New Sand
Sample ngple Moisture TS FS VS |
ID Date (%) (%) (%) (%) |
New Sand 6/26/2007 3.61 96.39 94.66 1.73]
New Sand 7/9/2007 2.43 97.57 95.37 2.19]
New Sand 7/20/2007 3.90 96.10 94.29 1.82I
NewSand 1219/2007 4.20 95.80 92.81 2.99
Moisture TS FS VS
(%) (%) (%) (%)
Average 3.5 96.5 94.3 2.2
Standard deviation 0.8 0.8 1.1 0.6
Median 3.8 96.2 94.5 2.0]
Count 4 4 4 4|

 

158

 

 

Table E5: Reclaimed Sand
Sa Sa Herd
ID Locaan Mana

arm 2 1
arm

Farm 2 SMS
arm 2
arm S
arm

Farm 2

Farm 2 S
arm 2 SMS 1
arm 2 2
arm

Farm 2
arm 2
arm S
arm 2
arm
arm 2
arm 2
arm

 

arm 2
arm
arm
arm 3

 

159

 

APPENDIX F

Anaerobic Digester Systems — Operation Data

Table F1: Operational Data from Several US. Based Anaerobic Digesters

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Total Mean Installed Generation Capacity
Anaerobic Digester Number Number
Type of of Number Mean 3:11:53
Systems Animals Of
' Systems (kl/V) (animals/kW) (animals/kW)
Complete Mix 26 1,628 22 415 4.7 2.6
Covered Lagoon 10 1,778 8 247 9.2 6.3
Fixed Film 1 250 1 30 8.3
Horizontal Pﬁg Flow 32 1,621 30 330 7.2 3.8
Induced Blanket Reactor 2 775 2 100 7.5
Mixed Pig Flow 33 2.878 27 589 4.2 1.1

 

160

 

 

A1C\

 

REFERENCES

Allen, A., D.A. Hutton, J.P. Pearson, and LA. Sellers. 1984. Mucus glycoprotein
Structure, Gel Formation and Gastrointestinal Mucus Function. In Mucus and
Mucosa. London: Pitman. 137-156.

Alshibli, K.A. and MI. Alsaleh. 2004. Characterizing Surface Roughness and
Shape of Sands Using Digital Microscopy. JOURNAL OF COMPUTING IN CIVIL
ENGINEERING. 18: 3645.

American Association of State Highway and Transportation Ofﬁcials (AASHTO).
1991. AASHTO M 145 Standard Speciﬁcations for Classiﬁcation of Soils and
Soil-Aggregate Mixtures for Highway Construction Purposes.

American Society of Agricultural and Biological Engineers (ASABE). 2005.
Manure Production and Characteristics. Standard ASAE 0384.2 MAR2005. St.
Joseph, MI.

American Society for Civil Engineers (ASCE). 1975. Sedimentation
Engineering. ASCE. New York, NY.

American Society of Testing and Materials (ASTM). 2006. Annual Book of
Standards.

American Public Health Association (APHA). 2008. Standard Methods for the
Examination of Water and Wastewater. Washington, DC. American Public
Health Association.

Antonsson, E.K. 2001. merecision in Engineering Design. Division of
Engineering & Applied Science, California Institute of Technology. Pasadena,
California, USA.

Bannink, A., Valk, H. and AM. Van Vuuren. 1999. Intake and Excretion of
Sodium, Potassium and Nitrogen and the Effects on Urine Production by
Lactating Dairy Cows. Journal of Dairy Science. 82: 1008-1018.

161

 

 

Barker, JC. 2001. Methane Fuel Gas from Livestock Wastes: A Summary.
North Carolina Cooperative Extension Service. EBAE 071-80. Raleigh, NC.

Berge, N.D., D.R. Reinhart, and T.G. Townsend. 2005. The Fate of Nitrogen in
Bioreactor Landfills. Environmental Science and Technology. 35:365-399.

Bernard, J.K. and DR. Bray. 2004. How to live with sand bedding. Hoards
Dairyman. August, 24‘“, 2004. Page 600.

Bernard, J.K., D.R. Bray, and J.W. West. 2003. Bacteria concentrations and
sand usage in freestalls bedded with fresh or recycled sand. 5‘" International
Dairy Housing Conference. ASABE, St. Joseph, MI.

Bewely, J, R.W. Palmer, and DB. Jackson-Smith. 2001. A Comparison of Free-
Stall Barns Used by Modernized Wisconsin Dairies. Journal of Dairy Science.
84: 528—541

Bickert, W.G. and 0M. Kirk. 2007. Sand is Great for Cows, but other
Challenges Exist. Hoards Dairyman. May 10‘", 2007: 361.

Bracmort, K, R.T. Burns, J. Beddoes, and W. Lazarus. 2008. An Analysis of
Anaerobic Digestion System Costs on US. Livestock Production Systems. 2008
ASBAE Annual lntemational Meeting, Providence, RI. ASBAE Paper No.
084643

Brade, CE. and Noone, GP. 1981. Anaerobic digestion—need it be
expensive? Water Pollution Control. 80: 70—76.

Camp, TR. 1945. Sedimentation and the Design of Settling Tanks.
Transactions of ASCE. 1 1 1: 895-952.

Chicago Climate Exchange (CCX). 2009.
h_ttg:l/www.chicagoclimgtexcom/incjgtﬁf

162

Chynoweth, D.P., C.E. Turick, J.M. Owens, D.E. Jerger, and MW. Peck. 1993.
Biochemical Methane Potential of Biomass and Waste Feedstocks. Biomass
and Bioenergy. 5: 95—111.

Classen, J.J. and SK. Liehr. 2005. Alternative Natural Technologies
Sequencing Batch Reactor Performance Verification. North Carolina State
University Animal and Poultry Waste Management Center.

Concha, F. and ER. Almendra. 1979. Settling Velocities of Particulate Systems,
1. Settling Velocities of Individual Spherical Particles. International Journal of
Mineral Processing. 5: 349-357.

Cook, N.B., T.B. Bennett, and K.V. Nordlund. 2004. Effect of Free Stall Surface
on Daily Activity Patterns in Dairy Cows with Relevance to Lameness
Prevalence. Journal of Dairy Science. 87: 2912-2922.

Cook, MB. and K.V. Nordlund. 2004b. Why it’s worth it to mess with sand.
Hoards Dairyman. Page 166.

Cook. MB. 2003. Prevalence of Lameness among Dairy Cattle in Wisconsin as
a Function of Housing Type and Stall Surface. JAVMA. 223: 1324-1328.

Cook, MB. 2001. How good is sand bedding for your cows? Hoards Dairyman.
October. Page 667.

Cornell Manure Management Program. 2008. Anaerobic Digester Case Study
Papers, Series. Authors: P. Wright, K. Graf, J. Ma, C. Gooch, J. Pronto, S. Inglis.

Crites, R.W. and G. Tchobanoglous. 1998. Small and decentralized wastewater
management systems. McGraw Hill Companies. New York, NY. 1084 pp.

El-Mashad, H.M., W.K.P. van Loon, G. Zeeman, and GPA. Bot. 2005.
Rheological Properties of Dairy Cattle Manure. Bioresource Technology. 96:
531-535.

163

 

Fronshell, G. 2000. Settling Basin Design. Western Regional Aquaculture
Center. WRAC-106.

Fulhage, C. 2003. Sizing and Management Considerations for Settling Basins
Receiving Sand-Laden Flushed Dairy Manure. Animal, Agricultural and Food
Processing Wastes, Proceedings of the Ninth lntemational Symposium. Raleigh,
NC. 456-462.

Glover, T.J. 1995. Pocket Ref. Sequoia Publishing, Littleton, CO.

Gloy, BA. 2008. Creating Renewable Energy from Livestock: Overcoming
Barriers to Adoption. E.B. 2008-02

Gooch, CA. and SF. Inglis. 2007. Sand for Bedding Dairy Cow Stalls. Cornell
Pro-Dairy. h_ttp://www.a_nsci.cornell.edu/grodai[yl

Gooch, CA. 2007. Sand Reclamation System Evaluation. Unpublished
research.

Gooch, C.A, A.W. Wedel, and J. Karszes. 2002. Economic Analysis for A Dairy
Waste Treatment System that Employs Mechanical Separation of Bedding Sand
from Scraped Sand-Laden Dairy Manure. ASAE Paper No. 02-4208, St. Joseph,
MI.

Gould, C. 2008. Personal communication regarding manure ﬁber plastic
composites.

Green, C. 2008. Personal communication regarding operation of sand
separation and anaerobic digester systems.

Green, C. 2005. Manure analysis results for the manure press.

Gregory, R., T.F. Zabel, and J.K. Edzwald. 1999. Water Quality and Treatment.
5th ed. American Water Works Association, Denver, Co.

164

 

Hamer, J.P., M.J. Brouk, and J.F. Smith. 2005. Sand Quality in Free Stalls.
2005 ASAE Annual International Meeting, Tampa, FL. ASAE Paper No. 054109.

Hamer, J.F., J.P. Murphy, J.F. Smith, M. Brouk, and TD. Strahm. 2003.
Experiences with Handling Sand Laden Manure Using Gravity. 5th lntemational
Dairy Housing Conference. ASABE, St. Joseph, MI.

Hashimoto, AG. and Y.R. Chen. 1976. Rheology of Livestock Waste Slurries.
Transactions of ASAE. 19(5):930-934.

Hayes, S. 2003. A lot of info to digest — Producers are focused on anaerobic
digesters. Hoard’s Dairyman. 388.

Hoffmann, R.A., M.L. Garcia, M.Veskivar, K. Karim, M.H. Al-Dahhan, L.T.
Angenent. 2008. Effect of Shear on Performance and Microbial Ecology of
Continuously Stirred Anaerobic Digesters Treating Animal Manure.
Biotechnology and Bioengineering. 100(1):3848.

Holtz, RD. and WD. Kovacs. 1981. An Introduction to Geotechnical
Engineering. New Jersey: Prentice Hall.

Inglis, S., C. Gooch, and M. Timmons. 2006. Engineering Analysis of Digesting
Sand-Laden Dairy Manure. Presented at the 2006 Annual ASABE Meeting.
Paper No. 064196. ASABE, St. Joseph, MI.

Inglis, SF. 2006. Settling Characteristics of Sands Contained in Anaerobically
Digested Dairy Manures A Project. Master Thesis. Cornell University.

Jacobs, J. 2009. Phone conversation discussing the operation and performance
of the Green Valley Dairy anaerobic digester.

Jin, Y.C., F. Lu, and M. Badruzzaman. 2005. Simpliﬁed Model for Class I
Settling Tank Design. Journal of Environmental Engineering. 131 (12): 1755-
1759.

165

 

 

 

Jin, Y.C., Q.C. Guo, and T. Viraraghaven. 2000. Modeling of Class I Settling
Tanks. Journal of Environmental Engineering. 126(8): 754-760.

Jonsson, O. 2005. Biogas upgrading and use as transport fuel.

http://wwwbiogasmaxco.uk/media/biggas upgrading and use 2004 0629442
00 1011 _g4042007.gf

Kappe, E. and JB Neighbor. 1951. Some New Developments in Aeration.
Sewage and Industrial Wastes. 23: 833-838.

Karim, K., K.T. Klasson, R. Hoffmann, S.R. Drescher, D.W. DePaoli and M.H. Al-
Dahhan. 2005a. Anaerobic digestion of animal waste: Effect of mixing.
Bioresource Technology. 96: 1607—1612.

Karim, K., R. Hoffmann, K.T. Klasson and M.H. AI-Dahhan. 2005b. Anaerobic
digestion of animal waste: Waste strength versus impact of mixing. Bioresource
Technology. 96: 1771-1781.

Karszes, J. 2003. The high cost of poor cow comfort. Northeast Dairy
Business. 34-35.

Kearney, T.E., M.J. Larkin, J.P. Frost’ and RN. Levett. 1993. Survival of
pathogenic bacteria during mesophilic anaerobic digestion of animal waste.
Journal of Applied Bacteriology. 75: 2 1 5 - 2 1 9

Keener, H.M., J.J. Hoon'nan, and M.H. Klingman. 2006. Rheology and
F lowability Properties of Liquid Dairy and Swine Waste. 2006 ASABE Annual
lntemational Meeting, Portland, OR. ASABE Paper No. 064072.

Kirk, OM. 2005. Selecting a Sand Laden Dairy Manure (SLDM) Handling
System. NRAES-176.

Kim, EH 2003. Modeling of hindered-settling column separations. Unpublished
PhD Diss. The Pennsylvania State University.

166

 

Kim, AS. and KD. Stolzenback. 2003. Aggregate formation and collision
efﬁciency in differential settling. Journal of Colloid and Interface Science. 271:
1 10-1 19

Kramer, J.M. 2004. Agricultural Biogas Casebook-2004 Update. Resource
Strategies, Inc. Madison, WI.

Krou, J.F., J.F. Fox, and S. Safferman. 2006. Experimental Data and
Theoretical Analysis of Particle Removal Efﬁciency in a Novel Hydraulic
Separation Unit. Journal of Environmental Engineering. 132(10): 1307-1313.

Kumar, M., H. D. Bartlett, and N. N. Mohsenin. 1972. Flow properties of animal
waste slurries. Transactions of the ASAE. 15(4): 718-722.

Landry, H., C. Lague, and M. Roberge. 2003. Physical and Rheological
Properties of Manure Products. Applied Engineering in Agriculture. 20 (3): 277-
288.

Linn, D. 2001. Bedding plays a big role in milk quality. Hoard's Dairyman. 680.

Liu, Z. 2001. Sediment Transport. Laboratoriet for Hydraulik og Havnebygning
Instituttet for Vand, Jord og Miljoteknik Aalborg Universitet

Lusk, P. 1995. Methane Recovers from Animal Manure. Regional Biomass
Energy Program, United States Department of Energy. Washington, DC.

Lusk, P. 1994. Methane Recovery from Animal Manures: A Current
Opportunities Casebook. National Renewable Energy Laboratory. Golden, CO.

Mangino, J., D. Bartram and A. Brazy. 2002. Development of a Methane
Conversion Factor to Estimate Emissions from Animal. Washington, DC, US.
Environmental Protection Agency.

167

 

 

Marvin, B. May, 2009. Personal communication concerning electrical output
from the GMF anaerobic digester.

Metcalf and Eddy, Inc. 1991. Wastewater Engineering Treatment, Disposal, and
Reuse, 3rd Edition. Burr Ridge, IL: McGraw-Hill.

Michigan Department of Environmental Quality (MDEQ). 2005. Large
Concentrated Animal Feeding Operation General Permit. Permit No.
MIGO19000. Lansing, MI.

Michigan Department of Transportation (MDOT). 2003. Standard Speciﬁcations
for Construction. mpzl/mgotwa_51.mdot.state.mi.us/public/specbook/

Midwest Plan Services (MWPS). 2000. Manure Characteristics, Section 1.
Manure Management Series. Ames, IA.

Morris, GR. 1976. Anaerobic Fermentation of Animal Wastes: A Kinetic and
Empirical Design Evaluation. Unpublished MS. Thesis. Cornell University,
Ithaca, New York.

Nelson, C. and J. Lamb. 2002. Final Report: Haubenschild Farms Anaerobic
Digester, Updated. The Minnesota Project. St. Paul, MN.

Nennich, T.D., J.H. Harrison, L.M. VanWieringen, D. Meyer, A.J. Heinrichs, W.P.
Weiss, N.R. St-Pierre, R.L. Kincaid, D.L. Davidson, and E. Block. 2005.
Prediction of Manure and Nutrient Excretion from Dairy Cattle. Journal of Dairy
Science. 88: 3721-3733. ‘

Nennich, T.D., J.H. Harrison, L.M. VanWieringen, N.R. St-Pierre, R.L. Kincaid,
M.A. Wattlaux, D.L. Davidson, and E. Block. 2006. Prediction and Evaluation of
Urine and Urinary Nitrogen and Mineral Excretion from Dairy Cattle. Journal of
Dairy Science. 89: 353-364.

Neter, J., W. Wasserman and GA. Whitmore. 1993. Applied Statistics, Fourth
Edition. Allyn and Bacon. Boston, MA.

168

Owen, W.F., D.C. Stuckey, J.B. Healy, Jr., L.Y. Young, P.L. McCarty. 1979.
Bioassay for Monitoring Biochemical Methane Potential and Anaerobic Toxicity.
Water Research. 13:485-492.

Pennsylvania State University. 2009. Digester Terminology, Abbreviations and
Units. http://www.biogas.psu.eduherminoLogv.html

Powers, W. J., H. H. Van Horn, A. C. VVrIkie, C. J. Wilcox and R. A. Nordstedt.
1999. Effects of anaerobic digestion and additives to efﬂuent or cattle feed on
odor and odorant concentrations. Journal of Animal Science. 77: 1412-1421.

Rozdilsky, J. 1997. Farm Based Anaerobic Digestion in Michigan: History,
Current Status, and Future Outlook, 1997. Michigan Biomass Energy Program,
Awa-8.

SAS Institute. 2006. SAS user's guide. Version 9.2. SAS |nst., Cary, NC.

Schell, D.J., J.C. Saez, J. Hamilton, A. Thuludur and JD. McMillan. 2002. Use
of Measurement Uncertainty Analysis to Assess Accuracy of Carbon Mass
Balance Closure for a Cellulase Production Process. Applied Biochemistry and
Biotechnology. 98: 509-523.

Shrift, R. 2008. Personal communication.

Speece, RE. 1996. Anaerobic Biotechnolpgy for Industrial Wastewaters.
Archae Press. TN.,

Steffe, J.F. 1996. Rheolpgical Methods in Food Process Engineering, 2""
Edition, East Lansing, MI: Freeman.

Steffen, R. Szolar, O, and Braun, R. 1998. Feedstocks for Anaerobic Digestion.
Institute for Agrobiotechnology Tulln University of Agricultural Sciences Vienna.

Stone, B. 2003. They Made the Sand Switch. Northeast Dairy Business. 30-
31.

169

Stowell, RR. and C. Henry. 2004. Making a manure digester pay. What are the
economic impacts of various public-policy scenarios? Hoard’s Dairyman. 523.

Stowell, R. R., and S. Inglis. 2000. Sand for bedding. In: Dairy Housing and
Equipment Systems: Managing and Planning for Proﬁtability. NRAES-129. 226-
234.

Stowell, RR. and WG. Bickert. 1995. Storing and handling sand-laden dairy
manure. Extension Bulletin E-2561. Michigan State University, East Lansing,
MI.

Svarovsky, L. 1990. Solid-Liquid Separation, 3'd Edition. Toronto, CA:
Butterworths.

Swamee, PK. and A. Tyagi. 1996. Design of Class-l Sedimentation Tanks.
Journal of Environmental Engineering. 122 (1): 71-73.

Szczegielniak, EH. 2008. Development of an Anaerobic Treatment Screening
Protocol for Fruit and Vegetable Processing Wastewater. Unpublished MS.
Thesis. Michigan State University, East Lansing, Michigan.

US. Department of Agriculture. 2009. Electronic Field Ofﬁce Technical Guide -
Michigan. http:/lwww.nrcs.usda.gov/t§phnical_lefo;q[. Natural Resources
Conservation Service. East Lansing, MI.

U. S. Department of Agriculture. 2008. Tech Reg website.
http: lltechreg. sc. egov. usda. gov/NTEITSPNTEZIAnima__IUnits. htrnl. Natural
Resources Conservation Service. Washington, D. C.

US. Department of Agriculture. 2007. An Analysis of Energy Production Costs
from Anaerobic Digestion Systems on US. Livestock Production Facilities.
Technical Note 1. Natural Resources Conservation Service. Washington, DC.

US. Department of Agriculture. 1993. Soil Survey Manual, 3'“ Edition. Natural
Resources Conservation Service. Washington, DC.

170

US. Department of Energy. 2009. http:/Iwww.eia.doe.gov/. Energy Information
Administration, Washing DC

US. Environmental Protection Agency. 2009. ht_tp:l/www.epa.gov/agstarl.
AgStar Program, Washington DC.

US Environmental Protection Agency. 2008. AgStar Digest, Winter 2006.

http://www.epa.govlagstar/pdf/2006digestpdf. AgStar Program, Washington
DC.

US. Environmental Protection Agency. 2008. 40 CFR 355.32. Final Rule, 73
Fed. Reg. 76948. December 18, 2008. Washington, DC.

US. Environmental Protection Agency. 2005. Inventory of US Greenhouse
Gas Emissions and Sinks: 1990-2003. EPA 430-R-05-003. Ofﬁce of
Atmospheric Programs, Washington, DC.

US. Environmental Protection Agency. 2002. Managing Manure with Biogas
Recovery Systems Improved Performance at Competitive Costs. EPA-430-F-02-
004. AgStar Program, Washington, DC.

US. Environmental Protection Agency. 1999. National Ambient Air Quality
Standards. Ofﬁce of Air and Radiation. Washington, DC

US. Environmental Protection Agency . 1979. Process Design Manual for
Sludge Treatment and Disposal. EPA 625/1 -79-01 1. Cincinnati, OH.

Vanoni, VA. 2006. Sedimentation Engineering, Edition 2. American Society of
Civil Engineers.

Vesvikar, MS. and M. AI-Dahhan. 2005. Flow Pattern Wsualization in a Mimic
Anaerobic Digester Using CFD. Biotechnology and Bioengineering. 89 (6): 719-
723

171

Wagner-Storch, A.M., R.W. Palmer, and D.W. Kammel. 2003. Factors Affecting
Stall Use for Different Freestall Bases. Journal of Dairy Science. 86: 2253-2266.

Wedel, AW. 2009. Personal communication regarding the sand separation
system at Green Meadow Farms.

Wedel, AW. 2001. Passive and Mechanical Sand-Manure Separators. NRAES
Dairy Manure Systems Conference. Rochester, NY. March 20-22, 2001.

Wedel, AW. 2000. Hydraulic Conveyance of Sand-Laden Dairy Manure in
Collection Channels. 2000 ASAE Annual lntemational Meeting, Milwaukee, WI.
ASAE Paper No. 004106.

Wedel, AW. 1995. Evaluation of Liquid/Solid Separation Techniques Applied to
Sand-Laden Dairy Manure. Master Thesis. Michigan State University.

Wedel, AW. and W.G. Bickert. 1996. Separating sand from sand-laden
manure: factors affecting the process. Paper No.: 19964016. ASAE, St. Joseph,
MI.

Wedel, AW. and W.G. Bickert. 1994. Handling and storage systems for sand-
Iaden dairy manure from freestall barns. 3rd lntemational Dairy Housing
Conference Dairy Systems for the 21st Century. ASAE. St. Joseph, MI.

Wilkie, AC. 2008. http:/lbiogas.ifas.uﬂ.eduluses.htm

Wilkie, AC. 2005. Anaerobic Digestion of Dairy Manure: Design and Process
Considerations. NRAES Dairy Manure Management: Treatment, Handling and
Community Relations. NRAES-176: 301-312.

Wong, K., I. Xagoraraki, J. Wallace, W. Bickert, S. Srinivasan, and J.B.Rose.
2008. Removal of Viruses and Indicators by Anaerobic Membrane Bioreactor
Treating Animal Waste.

172

Wright, P., S. Inglis, J. Ma, C. Gooch, B. Aldrich, A. Meister and N. Scott. 2003.
Preliminary Comparison of Five Anaerobic Digestion Systems on Dairy Farms in
New York State. Paper No.: 044032. ASAE, St. Joseph, MI.

Wu, B and S. Chen. 2007. CFD Simulation of Non-Newtonian Fluid Flow in
Anaerobic Digesters. Biotechnology and Bioengineering. 99(3): 700-711.

Zimmels, Y. 1984. Theory of Density Separation of Particulate Systems.
Powder Technology. 43: 127

Zhang, R. H., P. N. Dugba, and D. S. Bundy. 1997. Laboratory Study of Surface
Aeration of Anaerobic Lagoons for Odor Control of Swine Manure. Transactions
of ASAE. 40: 1 85-1 90.

173

LB

 

”7111111111115“!IINIIIVIIEIIIIIIIII WI "WES
3 1 2 9 3 0 3 0 6 3 1

117