 

 

LIBRARY
Michigan State

I lniuorcd’y

'V'UI

 

 

 

This is to certify that the
dissertation entitled

ESSAYS ON NUTRIENT MANAGEMENT RISK IN
LIVESTOCK PRODUCTION: CITIZEN ENVIRONMENTAL
COMPLAINTS, MANURE HAULING SYSTEM COSTS, AND
ANIMAL EMISSION TAXES

presented by

JOLEEN CHRISTINE HADRICH

has been accepted towards fuiﬁllment
of the requirements for the

degree In Aricultural Economics

 

 

PhD/M

Major Professor's SignatuU

At?“ [7/ 2.00?

 

Date

MSU is an Afﬁrmative ActiorVEqual 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

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K:IProj/Acc&Pres/CIRCIDateDue_indd

 

ESSAYS ON NUTRIENT MANAGEMENT RISK IN LIVESTOCK PRODUCTION:
CITIZEN ENVIRONMENTAL COMPLAINTS,
MANURE HAULING SYSTEM COSTS,
AND ANIMAL EMISSION TAXES

By

Joleen Christine Hadrich

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Agricultural Economics

2009

ABSTRACT
ESSAYS ON NUTRIENT MANAGEMENT RISK IN LIVESTOCK PRODUCTION:
CITIZEN ENVIRONMENTAL COMPLAINTS,
MANURE HAULING SYSTEM COSTS,
AND ANIMAL EMISSION TAXES
By

Joleen Christine Hadrich

Nutrient management on livestock operations must be environmentally friendly,
labor efﬁcient, and cost effective. Poor nutrient management practices may lead to
potential citizen complaints resulting in mitigation costs, ﬁnes or lawsuits. An
econometric analysis of citizen complaints regarding surface water, groundwater, and
odor concerns was completed to analyze farm characteristics affecting the probability of a
veriﬁed citizen complaint. Farm compliance with environmental regulations involves
policy uncertainty and sunk cost investments. A Spreadsheet-based manure transport and
land application decision tool, MANURESHAUL, was developed to provide farmers,
custom applicators, and others involved with the manure management a manure hauling
capacity, time and cost calculator for liquid manure hauling systems using tractor-drawn
tank Spreaders and truck-drawn nurse tanks used in parallel with tractor-drawn tank
spreaders. An optimal control theory model was used to model the uncertainty regarding
the Size of an animal air emission tax and its effects on a farmer’s investment in

emission-reducing technology.

To Mom, Dad, Dean, Daryl, Joelle, and Joylynn. Thanks for supporting me.

iii

ACKNOWLEDGEMENTS

I would like to thank my dissertation committee composed of, Dr. Christopher A.
Wolf, Dr. J. Roy Black, Dr. Stephen Harsh, and Dr. Timothy Harrigan for their guidance
throughout my dissertation research. My major professor, Chris Wolf, has been a major
resource for me during my graduate career at MSU. His advice, encouragement, and
friendship have helped me become a better researcher and academic professional—for
this I am grateful. Roy Black’s wealth of knowledge about “everything” has been
invaluable in my research endeavors at MSU. I want to thank Steve Harsh for his
feedback on my dissertation work and his expertise in budgeting. Tim Harrigan’s
extensive knowledge in manure hauling and agricultural engineering practices and
constant feedback on my research has always lead to higher quality outputs.

I would like to thank the graduate students and other faculty members I have
worked with at Michigan State University. In particular, I would like to thank Nicole
Olynk for her willingness to help with whatever was needed. My non-academic
experience at MSU would never have been as enjoyable without the company of Bill and
Cassie Burke.

Finally, I would like to thank my family for their support throughout my graduate
career. While they did not always understand what I was doing, they were always willing

to listen.

vi

TABLE OF CONTENTS

LIST OF TABLES ........................................................................................................... iii
LIST OF FIGURES ......................................................................................................... ix
CHAPTER 1: GENERAL INTRODUCTION ............................................................... 1
CHAPTER 2: CITIZEN COMPLAINTS AND ENVIRONMENTAL
COMPLIANCE ON MICHIGAN LIVESTOCK OPERATIONS ............................... 5
2.1 Introduction ............................................................................................................... 5
2.2 Data ........................................................................................................................... 8
2.3 Veriﬁed Citizen Complaints ................................................................................... 10
2.4 Corrective practices implemented ........................................................................... 17
2.5 Conclusions ............................................................................................................. 24
REFERENCES ............................................................................................................. 26
CHAPTER 3: ECONOMIC COMPARISON OF LIQUID MANURE TRANSPORT
AND LAND APPLICATION ........................................................................................ 29
3.1 Introduction ............................................................................................................. 29
3.2 Objectives ............................................................................................................... 32
3.3 Model Development ............................................................................................... 32
3.3.1 Manure Hauling Rate ....................................................................................... 33
3.3.2 Equipment costs ............................................................................................... 34
3.3.3 Ownership costs ............................................................................................... 37
3.3.4 Operating costs ................................................................................................. 38
3.4 Nutrient value of manure ........................................................................................ 40
3.4.1 N volatilization losses ...................................................................................... 40
3.4.2 Fertilizer recommendations ............................................................................. 42
3.5 Procedure ................................................................................................................ 43
3.6 Representative dairy farm ....................................................................................... 49
3.7 Results Discussion .................................................................................................. 53
3.7.1 Labor Requirements ......................................................................................... 53
3.7.2 Hauling costs .................................................................................................... 56
3.7.3 Manure pumping and land application on four 1,400 cow dairies .................. 57
3.7.4. Broadcast application for four 1,400-cow dairies ........................................... 60
3.7.5 Subsurface injection application results for four 1,400-cow dairy farms ........ 64
3.7.6 Return on fertilizer value of manure ................................................................ 66
3.8 Model Validation .................................................................................................... 67
3.8.1 Swine producer ................................................................................................ 67
3.8.2 Cash crop producer .......................................................................................... 69
3.9 Hauling cost model ................................................................................................. 69
3.9.1 Hauling cost coefﬁcients ...................................................................................... 71
3.10 Conclusion ............................................................................................................ 75

REFERENCES ............................................................................................................. 78

CHAPTER 4: IMPLICATINS OF AN AIR EMISSIONS TAX ON LIVESTOCK

PRODUCER’S INVESTMENT IN ABATEMENT TECHNOLOGY ...................... 81
4.1 Introduction ............................................................................................................. 81
4.2 Basic Emission Tax Model .................................................................................... 87
4.3 Numerical Example: Basic Emissions Tax Model ................................................. 93
4.4 An Uncertain Emissions Tax Increase at a known Future Date ........................... 100
4.5 Empirical analysis of uncertain tax ....................................................................... 105
4.6 Conclusions ........................................................................................................... 110
REFERENCES ........................................................................................................... 112

vi

LIST OF TABLES

Table 2a. Deﬁnition and Summary Statistics of Explanatory Variables, All Complaints 12

Table 2b. Probability of a Veriﬁed Complaint ................................................................. 14
Table 2c. Corrective Practices to Mitigate Veriﬁed Complaints ...................................... 18
Table 2d. Average Costs to Implement Corrective Practices ........................................... 20

Table 2e. Explaining Corrective Practice Costs Required to Mitigate Veriﬁed

Complaints ................................................................................................................ 23
Table 3a. Manure hauling system equipment .................................................................. 34
Table 3b. Estimated equipment list price functions ......................................................... 37
Table 3c. Repair factors for trucks, manure Spreaders, and agitators and pumps ............ 39
Table 3d. Estimated N volatilization losses by manure applicatin method ..................... 41

Table 3e. Bray P1 soil test results, fertilizer recommendations, manure application rates,
and nutrient credit guidelines used in MANURESHAUL ........................................ 43

Table 3f. MANURESHAUL user inputs, defaults, and override values .................... 45-46

Table 3g. Fertilizer recommendations and soil test results for corn grain, corn silage, and
alfalfa crops .............................................................................................................. 51

Table 3h. Farm characteristics for 175-cow, 350-cow, 700-cow, and 1,400-cow dairy . 52

Table 3i. Equipment used, size, and estimated purhcase price for 175-cow, 350-cow,
700-cow, and 1,400-cow dairy ............................................................................. 52-53

Table 3j. Hauling time and costs for 175-cow, 350-cow, 700-cow, and 1,400-cow dairy
using broadcast and injection application ........................................................... 54-56

Table 3k. 1,400 cow dairy farm characteristics and equipment used for broadcast and
injection application ............................................................................................. 58-60

Table 31. Fertilizer recommendations, manure nutrients applied, and nutrient credit
values for broadcast application ............................................................................... 64

Table 3m. Nutrient value of manure for injection, broadcast with incorporation, and
broadcast application as a function of soil test results .............................................. 67

vii

Table 3n. Liquid manure machinery system speciﬁc coefﬁcients for estimating cost for
agitation, pumping, transport and land applicaiton .............................................. 73-75

Table 4a. Parameter descriptions, values, and sources .................................................... 95

viii

LIST OF FIGURES

Figure 2a. Complaint type by livestock enterpise .............................................................. 9
Figure 2b. Complaint classiﬁcation by livestock enterprise ............................................ 10

Figure 3a. Hauling capacity of two 26,495 L (7,000 gal) tractor-drawn tank spreaders
working in parallel, one 13,248 L (3,500 gal) tractor-drawn spreader in parallel with
two 26,495 L (7,000 gal) nurse trucks, and two 26,495 L (7,000 gal) tractor-drawn
spreaders in parallel with two 26,495 (7,000 gal) nurse trucks over 16.1 km (10.1)
mile hauling distance ................................................................................................ 47

Figure 3b. Manure hauling cost for a 700-cow dairy using two 34,065 L (9,000 gal)
tractor-drawn spreaders working in parallel and one 34,065 L (9,000 gal) tractor-
drawn spreader in parallel with two 34,065 L (9,000 gal) nurse tanks over 12.9 km

(8 miles) .................................................................................................................... 49
Figure 3c. Calculated manure agitation, pumping, transport, and spreading versus

predicted cost using machinery speciﬁc coefﬁcients ................................................ 72
Figure 4a. Optimal investment path for t=$1.92/1b ammonia emission ........................... 97

Figure 4b. Optimal investment path for t=$1.92/lb and t=$0.48/lb ammonia emission . 99

Figure 4c. Optimal investment path for tL=$O.48/lb and tH=$1 .92/lb ammonia
emissions at time t=T .............................................................................................. 100

Figure 4c. Emission reducing capital stock per cow levels as tax rate increases .......... 102

Figure 4d. Optimal investment path for an uncertain tax increase at time t=T with a low
initial tax rate, IL =$O.48/lb .................................................................................... 108

Figure 4e. Optimal investment path for an uncertain tax increase at time t=T with t=0
for (=0 to t=T .......................................................................................................... 109

ix

CHAPTER 1: GENERAL INTRODUCTION

Animal production levels and average herd sizes have been trending upwards on
US agricultural farms for decades. Manure hauling systems must accommodate the ever
increasing volume of manure while considering more stringent environmental
regulations. Poor nutrient management practices could lead to manure leaks, spills, and
run-off from ﬁeld application or manure storage resulting in ﬁnes or lawsuits. In
addition to managing the farming operation, farmers must also address increased public
scrutiny regarding farming practices with urban areas expanding closer and closer to rural
agricultural settings. In order to remain proﬁtable agricultural producers must be
cognizant of these challenges and their implications on future management decisions.

In response to the increased interaction between the urban areas and production
agriculture, past legislation regarding environmental pollution from non-point pollution
sources, such as the 1972 Clean Water Act and 1990 Clean Air Act, many states have
created “Right to Farm” Programs. These programs provide legal protection for livestock
producers against nuisance lawsuits and citizen complaints while also providing a set of
accepted management practices to be in compliance with environmental guidelines.
Nutrient management practices vary by state but provide an environmental compliance
benchmark for livestock producers. In Michigan, the Right to Farm program was
initiated in 1981 which developed a set of Generally Accepted Agricultural Management
Practices (GAAMPS). This program is voluntary, but participating in the program

provides producers with a form of legal protection against nuisance lawsuits and citizen

complaints which may outweigh the potential future costs of ﬁnes and legal actions if the
farm is not in compliance with current environmental regulations.

Environmental regulations continue to evolve over time with knowledge gained
through research and evaluating the end result of current situations. Many times these
updates and changes in environmental regulations cause livestock producers to delay
investment in potential abatement technology in fear that it may be outdated before the
useful life of the equipment has expired. This uncertainty must be accounted for when
assessing the potentially large capital investments made by farmers in abatement '
technology.

Manure nutrient management considering environmental regulations requires an
understanding of the manure hauling system components and its associated costs.
Choosing a less than optimal manure hauling system may increase manure hauling time
and cause potential delays in crop tillage and planting. Therefore, individual farms must
consider each of these components for their farm before making potential capital
investments in manure hauling systems. Current options in manure hauling systems vary
greatly in labor, machinery requirements, ownership and operating costs, and
compatibility with environmental regulations.

In addition to manure nutrient management, agricultural producers are facing

increased public scrutiny for animal air emissions (ammonia (NH 3), methane, and

particulate matter) from their farm. Odor management is an area of growing concern for
agricultural producers. While odor is not regulated in 2009, per se, agricultural producers
must be vigilant in adopting practices to limit odor from their farm in an attempt to

decrease future citizen complaints and be pro-active about potential future air emission

regulations. Examples of air emission reduction practices include incorporating manure
into the soil immediately after application, injecting manure, or building a long-term
manure storage facility with appropriate abatement technology to limit air emissions
(manures storage covers, bioﬁlters, etc.).

This dissertation consists of three essays which each address particular elements
of nutrient management risk on livestock operations in Michigan. The ﬁrst essay
provides an understanding of the interaction between production agriculture and urban
areas through an analysis of Right to Farm program citizen complaints regarding surface
water, groundwater, and odor concerns. Environmental citizen complaint data was
collected and used to determine farm and county level factors influencing the probability
of a veriﬁed environmental citizen complaint issued against livestock producers in
Michigan. Costs of implementing corrective practices required to mitigate a veriﬁed
citizen complaint were estimated which were used in a two-stage Heckman procedure to
determine the individual farm characteristics inﬂuencing the cost of the corrective
practices required to mitigate environmental citizen complaints. The second essay
develops an excel spreadsheet-based manure transport and land application decision tool,
MANURESHAUL, to evaluate comparisons of cost-effective alternative manure hauling
systems. MANURESHAUL provides an accurate estimate of time needed for manure
pumping, transport, and land application as a function of hauling distance, spreader
capacity, manure equipment cost, labor, and the nutrient value of manure. The ﬁnal
essay considers the investment policy in air-emission abatement for uncertain

environmental taxes on animal air emissions. An optimal control theory model is used to

model the uncertainty regarding the size of the emission tax and its effect on a farmer’s

investment policy.

CHAPTER 2: CITIZEN COMPLAINTS AND ENVIRONMENTAL
COMPLIANCE ON MICHIGAN LIVESTOCK OPERATIONS

2.1 Introduction

Interaction between urban areas and production agriculture often results in citizen
complaints regarding manure management, air quality, and water quality concerns.
Recognizing that livestock operations must be able to collect and dispose of manure
while being sensitive to environmental consequences, the 1981 Michigan Right to Farm
Act deﬁned a set of generally accepted manure management practices (GAAMP) that, if
followed, ensure farm protection from nuisance complaints and lawsuits (Michigan
Department of Agriculture 2008b). The GAAMP standards are reviewed and updated
annually to address current environmental concerns by a committee of industry, state and
university personnel. A response program was initiated in 1986 to address citizen
environmental complaints received by the Michigan Department of Agriculture (MDA),
Right to Farm Program (Michigan Department of Agriculture 2008b). When an
environmental complaint is ﬁled against a farm an inspection is scheduled within seven
business days. Common examples of potential GAAMP standard violations evaluated
during inspections include livestock in streams and rivers, surface applied manure not
incorporated within forty-eight hours of application, and manure application on frozen or
snow-covered soil.

Following an initial inspection each complaint is categorized as non-veriﬁed,
veriﬁed, or transferred to an enforcement agency. If an inspected farm is complying with
all relevant GAAMP standards, the complaint is classiﬁed as non-veriﬁed by the Right to

Farm inspector. While non-veriﬁed complaints may have been caused by practices or

events that legitimately irritated the complainant, they were determined to require no
corrective action and, thus, do not require mitigation or otherwise alter producer
behavior. If the inspected farm is out of compliance, the complaint is classiﬁed as
veriﬁed. Farms with veriﬁed complaints must correct the environmental issue on their
farm in a timely manner to regain Right to Farm protection. Progress towards completion
of corrective practices is assessed by follow-up inspections. Should the farm fail to make
adequate progress to correct environmental concerns, the Right to Farm inspector may
close the case leaving no protection or forward the case to the Michigan Department of
Environmental Quality for enforcement. In situations where the original complaint
violation is not under Michigan Department of Agriculture jurisdiction, such as a direct
manure discharge into public waters, the complaint is transferred to the Michigan
Department of Environmental Quality for enforcement action. This distinction of
complaint classiﬁcation allows for an examination of the factors related to complaint
status.

Past research has recognized the importance of relationships between
environmental regulations and citizen complaints (Cohen; Eckert; Helland; and Heyes).
Few studies have evaluated the effects of citizen complaints and consequences on
behavior. Dasgupta and Wheeler assessed factors affecting citizen environmental
complaints using Chinese provincial data and determined that complaints provided useful
information but consumed a large Share of inspection resources making them relatively
costly. Huang and Miller evaluated the relationship between citizen complaints, swine
production, and county characteristics using swine farm inspection data for Illinois. They

concluded that citizen issued complaints were a more efﬁcient source of monitoring

information than regularly scheduled inspections. Huang and Miller also found that
building type and swine production intensity were the factors that most influenced the
probability of a regulatory violation.

Citizen complaints are potentially a source of low-cost monitoring of
environmental violations (Huang and Miller; Dasgupta and Wheeler). Neighbors and
passersby observe livestock facilities on a daily basis and may witness environmental
issues, whereas regulatory agencies usually do not have the resources to monitor a large
number of farms on a regular basis. On the other hand, complainants may not be able to
identify legitimate environmental concerns as opposed to acceptable management
practices. In some cases, individuals or groups may have a high propensity to complain.
These instances often occur in areas where there are concentrated animal feeding
operations (CAFOs) located near the rural-urban fringe. With little basis to evaluate
practices used on the farm, complaints instead may be ﬁled because that person or group

disapproves of the location, Size, or production practices of the farm particularly as these

I . . . . . .
relate to odor. ThIs research exammes how IndIVIdual farm production and county level
characteristics influence the probability of a veriﬁed complaint. We also examine the
interdependence between farm production characteristics and costs associated with

corrective practices required to mitigate veriﬁed complaints.

 

l Odor is not regulated in Michigan (or in most other states). However, the underlying
issue(s) causing odor may be regulated. Air quality issues are typically handled through
corrective measures such as incorporating manure into soil within forty-eight hours of
application, limiting manure application on the weekends, or developing a manure
management system plan in accordance with Michigan GAAMP standards.

2.2 Data

Environmental citizen complaint data were collected from the Michigan Department of
Agriculture for the period from October 1998 through December 2007. The reports
detailed individual characteristics of the farm inspected including: zip code and county of
both complainant and livestock operation, type of livestock enterprise, herd Size in animal
units (AU), type of manure storage, current manure analysis, soil tests, existence of
comprehensive nutrient management plan (CNMP) or manure management system plan
(MMSP) and whether either plan was under development or updating, manure

incorporation, corrective practices implemented to respond to veriﬁed complaints, and

days required to implement corrective practices.2

Environmental citizen complaints were categorized as relating to air, ground
water, surface water, combination, or “other” complaints which include flies, dust, and
pro-active complaints. Pro-active complaints were those requested by the farmer to
ensure GAAMP standards were followed. Over the approximately ten year period
examined, the most common complaint types were air and surface water which together
accounted for 75% of all complaints (Figure 2a). Ground water, combination, and other

complaints were less common. Dairy producers (32%), beef producers (16%), and horse

 

2 An “animal unit” is a metric of manure generation used to assess the size of operations
across animal species. One animal unit was deﬁned as: one feeder calf, heifer, or steer;
0.7 mature dairy cows (whether a milking or dry cow); 25 pigs weighing over 55 pounds;
0.5 horses; 10 sheep or lambs; 55 turkeys; 100 laying hens or broilers when the facility
has unlimited continuous ﬂow watering systems; 30 laying hens or broilers when facility
has liquid manure handling system (MDA, 2008b).

faCIlItIes (16%) received the largest share of complamts. SImIlarly daIry, beef, and
equine enterprises were the focus of the majority of surface water complaints while dairy

and swine operations received the largest number of odor complaints.

 

180
160 -
140 ~
120 -
100 A
80 -
60 a
40 _
20 ~

0 L

 

Number of Complaints

 

 

 

K c, -d . (g, . 0° , c s
62.9 069 or» $9 (20$ 9““ 60$ OS”
(c

Livestock Enterprise

 

[[1] Odor H Groundwater I Surface Water I Combination B Other]

 

 

 

 

Figure 2a. Complaint type by livestock enterprise

By complaint status 45% were classiﬁed as non-veriﬁed and 55% were classiﬁed
as veriﬁed (including enforcement level complaints). Figure 2b. presents the number of
complaints by complaint status classiﬁcation and livestock enterprise. Dairy, beef and
equine farms received more veriﬁed complaints whereas as the opposite held for poultry

and swine farms.

 

3 The remaining livestock enterprises included poultry, swine, crops, combination
livestock, and other livestock. Crops referred to fertilizer practices, soil erosion, and crop
production practices. Other livestock include goats, sheep, deer, elk, bees, and by-
product utilization.

 

 

 

 

 

 

 

 

Number of Complaints

 

Type of Livestock

 

 

in All SI Veriﬁed EI Non-veriﬁed]

 

 

Figure 2b. Complaint classification by livestock enterprise

2.3 Veriﬁed Citizen Complaints

In order to understand the factors affecting the likelihood that a complaint was veriﬁed,
we used a probit model to estimate probability of a veriﬁed complaint as deﬁned by
complaint type, farm characteristics, county characteristics, and seasonal factors. Using

this model, the probability of a veriﬁed complaint can be expressed as:
a]:
(1) Y] = X 77 + e]

y>l< __ 1, if verified
0, if nonveriﬁed

where Y] is the a binary variable equal to one for veriﬁed complaints and zero for non-

veriﬁed complaints, X denotes an array of variables that are hypothesized to affect the

probability of a veriﬁed complaint, 7] is a vector of parameters, 61 is the error term, and l

indexes farm. We assumed that e] was normally distributed which allowed us to estimate

a probit model from equation (1) using maximum likelihood techniques (Wooldridge,
2003)

Explanatory variables for the analysis were divided into four categories:
complaint type, farm characteristics, county characteristics, and seasonal factors.
Summary statistics of the explanatory variables are presented in Table 2a. Complaint
type signiﬁcance is likely to be related to how recognizable the potential violation is to a
typical citizen. For example, surface water related complaints may be more likely to be
visible concerns such as waste run-off.

Farm characteristics included livestock enterprise, manure handling system,
animal units, and distance between complainant and farm. Livestock enterprise types
were beef, crops, dairy, equine, poultry, swine, a combination of two or more groups, and
“other” livestock. Crop complaints referred to fertilizer practices, soil erosion, and crop
production practices. The “other” livestock category included complaints concerning by-
products from fruit and vegetable processing, sheep, goats, deer, and elk.

Manure storage was categorized into three groups. No storage meant the farm did
not have manure storage requiring, in the case of dairy farms, hauling manure on a daily
basis. Short-term storage was deﬁned as manure storage for less than six months and
included stockpiling on dirt and cement as well as manure stored in barns and lots. Long-
term manure storage was deﬁned as adequate for six months or more. Earthen and
concrete manure pits as well as composting were examples of long-term storage for beef,
dairy, swine, and poultry operations. Long-terrn manure storage for equine operations

included stockpiling of manure. A manure storage structure is not required for equine

ll

Table 2a. Deﬁnition and Summary Statistics of Explanatory Variables, All

 

 

Complaints
Variable Obs. Mean Std Deﬁnition
Value Dev.
Dependent Variable
Verified Complaint 1307 0.554 -- Verified complaint (0/ 1)
Complaint Type
Odor 1297 0.396 -- Odor complaint (0/1)
Groundwater 1297 0.094 -- Groundwater complaint (0/ 1)
Surface water 1297 0.352 -- Surface water complaint (0/1)
Combination 1297 0.1 13 -- More than one environmental concern issued
complaint in the complaint (0/1)
Other complaint 1297 0.045 -- Other complaints-ﬂies, noise, dust (0/ 1)
Farm Characteristics
Distance 1310 0.498 -- Zip code between complainant and farm is
different (0/ 1)
AU 1097 548.4 1 182 Animal units on farm (AU)
Days 646 172.4 167 Days used to implement corrective practices
Manure Storage
No Storage 1029 0.080 -- No manure storage (0/1)
Short-term 1029 0.245 -- Short-term manure storage (0/1)
Long-term 1029 0.490 -- Long-term manure storage (0/1)
Livestock Enterprise
Beef 1310 0.157 -- Beef cattle (0/1)
Dairy 1310 0.320 -- Dairy cattle (0/ 1)
Swine 1310 0.116 -- Swine (0/1)
Equine 1310 0.158 -- Equine (0/1)
Poultry 1310 0.057 -- Poultry (0/ 1)
Crop 1310 0.075 -- Crops (0/1)
Other Livestock 1310 0.062 -- Goat, sheep, other livestock types (0/1)
Combination 1310 0.055 -- More than one livestock type (0/1)
Livestock
Seasonal factors
Spring 1310 0.340 -- Complaint issued in April, May, June (0/1)
Summer 1310 0.309 -- Complaint issued July, August, September
(0/ 1)
Fall 1310 0.175 -- Complaint issued in Oct., Nov., Dec. (0/ 1)
Winter 1310 0.175 -- Complaint issued in Jan., Feb., March (0/1)
Year 1310 2003 2.6 Time trend (years)
County Characteristics
AU density 1309 42.3 36.9 County animal unit density (au/milez)
Population density 1309 227.9 307 County population density (pop/milez)
Farms (county) 1309 1026 355 Number of farms in the county
Median household 1309 4239 7212 County level median household income (S)
Income 5
HS education 1309 83.5 3.6 County level residents with high school

diploma or higher (%)

12

facilities under GAAMP standards due to low nutrient content and amount of manure
produced.

Distance between complainant and farm was represented by a dummy variable
coded as one for those complainants that resided at a different zip code than the farm in
question. The null hypothesis was that complaints from other zip codes would be more
likely veriﬁed as those complainants would be less likely bothered by nuisance issues.

County characteristics included animal unit density and number of farms (United
States Department of Agriculture, 2007), median household income (Michigan
Information Center), and percent of population with high school education level or higher
(United States Census Bureau). County level animal density and number of farms in the
county captured farming intensity. It was hypothesized that higher county animal unit
densities were more likely to have veriﬁed complaints due to a higher proportion of
farmers and familiarity with agriculture. County level education and income variables
were included to capture the characteristics of communities around and near the farm.
Seasonal factors were addressed using dummy variables. Complaint year was also
included.

Table 2b presents regression results for the probability that a ﬁled complaint was
veriﬁed. The omitted base set of characteristics for the categorical variables were an
odor complaint ﬁled against a dairy operation with long-term storage in the Spring
season. This base case represented the most common type of complaint, operation,
manure storage, and season. Results revealed that the probability of a veriﬁed complaint
was affected by complaint type. Surface water and combination complaints were 18 and

16% more likely to be veriﬁed relative to odor complaints, respectively. A combination

l3

Table 2b. Probability of a Verified Complaint

Variable Coefﬁcient Standard Error ‘ Marginal effects

 

Complaint Type 1

Groundwater -0.0065 (0.1632) -0.0025

Surface water 0.4766 (0.1231)*** 0.1757

Combination complaint 0.4552 (0.153 l)*** 0.1615

Other complaint -0.3883 (0.2398) -0.1528

F arm Characteristics

AU '0.0002 (0.0001)*** -0.0001

Distance 0.1767 (0.1009)* 0.0672

2

Manure Storage

No Storage 0.1666 (0.1766) 0.0621

Short-term -0.0466 (0.1389) -0.0178
. . 3

Livestock Enterprise

Beef 0.1 180 (0.1638) 0.0444

Swine -0.5166 (0.1622)*** -0.2028

Equine -0.1932 (0.1520) -0.0748

Poultry -0.6926 (0.2796)*** -0.2708

Crop -0.3366 (0.2660) -0.1322

Other livestock -0.21 19 (0.2252) -0.0826

Combination livestock 0.0079 (0.2089) 0.0030

Seasonal factors 4

Summer 0.0223 (0.1 160) 0.0085

Fall -0.4306 (0.1335)*** -0.1684

Winter 0.2032 (0.1369) 0.0758

Year -0.0239 (0.0202) -0.0091

County Characteristics

AU density -0.0011 (0.0015) -0.0004

Population density 0.0003 (0.0002) 0.0001

Farms (county) 0.0001 (0.0002) 4.33E-05

Median household income 9.16E-06 (1.04E—05) 3.49E-06

HS education -0.0489 (0.0184)*** -0.0186

Constant 51.7860 (40.5349)

Chi-square 138.96

Probability>Chi-square 0.000

Log-likelihood -51 1.15

Pseudo R-Square 0.1197

Predicted probability at mean 0.6174

Sample size 867

 

T .

Base complaint type = odor *** Signiﬁcant at 1% level
2

Base livestock type = dairy
3

Base manure storage type = long-term

4 .
Base season = sprIng

** Signiﬁcant at 5% level

* Signiﬁcant at 10% level

14

complaint addressed more than one issue on a farm, for example odor and surface water
concerns. Thus, the complaint types that could be visually observed in the form of, for
example, manure run-off were more likely to be veriﬁed than odor complaints. We
suspect that most Michigan citizens were unaware that there were no explicit odor
regulations pertaining to livestock operations.

Complaints issued against swine and poultry operations had a 20 and 27% lower
probability of a veriﬁed complaint relative to dairy operations, respectively. This may be
related to odor as conﬁnement swine and poultry operations following standard practices
often produce odor that people ﬁnd more objectionable than cattle or horse operations.
Thus, even though type of complaint is controlled for and odor itself is not regulated, the
objectionable odor from swine and poultry farms may contribute to a higher level of
nuisance complaints.

The probability of a veriﬁed complaint was not dependent on manure storage
type. We found this result surprising given the focus on avoiding manure spreading on
frozen ground in Michigan (Michigan Department of Agriculture, 2008b). With many
operations lacking long-term storage, we expected more veriﬁed complaints would be
associated with short-terrn storage.

Animal units (AU) present on farm, which measures herd size, was found to be
negative and signiﬁcant. As the number of animals units increased, probability of a
veriﬁed complaint declined. This may be surprising since large animal operations seem
to be the focus of many environmentally related controversies. However, large
operations are often newer facilities with modern manure handling technologies which

have completed thorough and intensive site selection review. Site selection involves an

15

extensive inspection of buildings and waste storage facilities on a farm and practices used
in order for the farm to be incompliance with GAAMP standards for their day-to-day
farming operations (Michigan Department of Agriculture, 2008a). These results suggest
that large operations were signiﬁcantly more likely to receive non-veriﬁed nuisance
complaints perhaps in part caused by perceptions and press related to operation size and
production practices. This indicates a need for conﬁned animal feeding operations to be
pro-active in public and neighbor relations that past research has found to produce
positive results with respect to complaints (Hadley, Harsh and Wolf).

The probability of a veriﬁed complaint increased by 7% when the complainant
and farm were not located in the same zip code. This may indicate that people passing by
are more likely to call only when noticing a potentially serious violation. It may also
indicate the effect of citizen groups who actively and aggressively monitor large livestock
Operations in some parts of Michigan (Sierra Club; Environmentally Concerned Citizens
of South Central Michigan). Finally, it may indicate a hesitation on the part of neighbors
to report others in close proximity with whom they are likely tohave future interaction.

A complaint issued in the Fall had a decreased probability of veriﬁcation relative
to Spring complaints. People tend to be more active during the Spring creating
opportunities for complaints. During Fall months farmers are harvesting crops and often
incorporating manure Shortly after harvest, a practice which would decrease the
likelihood of a veriﬁed complaint.

The negative and signiﬁcant marginal effect for the percent of the county
population with a high school level education or higher indicated that more educated

people were less likely to make veriﬁed complaints. Population density, the number of

16

farms, AU density, and median household income at the county level did not signiﬁcantly
inﬂuence the probability of veriﬁed complaints.

2.4 Corrective practices implemented

For veriﬁed complaints, mitigating practices aligned with the GAAMP standards were
required. Corrective practices included developing a manure management system plan
(MMSP) or a more formal comprehensive nutrient management plan (CNMP), soil
analysis, manure analysis, incorporating applied manure, manure stockpile utilization,
installing stream bank fencing, and controlling waste run-off. Completing and ﬁling an

MMSP or CNMP entails submitting an ofﬁcial document outlining manure production,

utilization, and application on the farm.4 Manure stockpile utilization required the farm
to remove manure stockpiles either through manure application or disposal through other
arrangements, such as potentially giving it away to neighboring farms. Installing stream
bank fencing included controlling water access for livestock near lakes, rivers, and
streams. Controlling waste run-off required the farmer to install appropriate waste
storage for manure as well as milk waste water for dairy operations.

Table 2c displays the corrective practices implemented to mitigate veriﬁed
complaints across livestock enterprises. Dairy and swine operations were most often
required to develop a MMSP whereas equine and “other” livestock operations were
frequently required to remove stockpiled manure. “Other” livestock groups were

typically small farms (less than 10 acres) with goats or sheep who typically did not have a

 

4 A MMSP must be ﬁled with the Right to Farm Program for AF Os. Soil and manure
analysis are needed as well as a formal document outlining manure management. CNMP
are a requirement for the National Pollution Discharge Elimination System for CAFOs.
CNMP must be certiﬁed whereas MMSP do not require certiﬁcation.

17

Table 2c. Corrective Practices to Mitigate Veriﬁed Complaints
1
Beef Crops Dairy Equine Poultry Swine Comb. Other

 

 

Corrective Percent (%)

Practice

Soil analysis 11.54 37.21 14.22 17.86 31.82 22.92 0.00 30.30
MMSP 19.23 4.65 47.25 18.75 31.82 43.75 28.57 9.09
CNMP 0.77 2.33 3.21 0.89 9.09 4.17 2.38 0.00
Manure 2.31 6.98 14.68 4.46 13.64 18.75 4.76 12.12
incorporation

Stockpile 4.62 13.95 1.38 22.32 9.09 0.00 16.67 39.39
utilization

Stream bank 53.08 0.00 10.09 16.07 0.00 4.17 30.95 3.03
fencing

Vegetative 3.85 32.56 1.83 13.39 0.00 2.08 9.52 3.03
buffer

Control run-off 4.62 2.33 7.34 6.25 4.55 4.17 7.14 3.03
structure

1 . . .
Comb.=CombInatIon livestock

large land base on which to dispose of manure. In Michigan, beef cow and feeder
operations typically use a pasture-based system. Over ﬁfty percent of beef operations
with veriﬁed complaints were required to install stream bank fencing indicating
Michigan’s increased efforts to exclude livestock from waterways. Cropping operations
were most commonly required to provide soil analysis and install vegetative buffers to
prevent waste run-off.

A second objective was to understand how farm characteristics inﬂuenced cost of
implementing corrective practices required to abate environmental problems on farms
receiving a veriﬁed complaint. While the costs to mitigate complaints were not collected,
we were able to estimate mitigation costs for each operation using farm and complaint
information. A manure management plan was assumed to cost $1,498 per farm under
1,000 animal units (APO) and $3,382 per farm with more than 1,000 animal units

(CAFO) (Vollmer-Sanders, Batie, and Wolf). Soil samples cost $15 per sample with one

18

sample taken for every ﬁve acres (Bundy et al.). Manure analysis cost $25 per sample
(Michigan State University, 2008).

Hadrich, Harrigan and Wolf estimated typical incorporation costs for Michigan
livestock operations at $5.94 per acre. Manure stockpile utilization was calculated based
on livestock enterprise and manure production levels. Equine manure disposal costs at
$200/horse were calculated since land was not available for manure disposal requiring
spreading on neighboring land and large stockpiles—often multiple years worth of
manure—were typically present (Murphy and Nicholson). Beef and dairy operation
disposal costs were calculated as a function of manure produced. Using the tool
developed by Hadrich, Harrigan, and Wolf manure stockpile utilization cost was
converted to a per acre cost at $37.43/acre for beef and $149.46/acre for dairy. Poultry
manure stockpile utilization cost $42.77 per ton (Young et al.).

. To prohibit livestock from waterways, a common fence consisting of barbed wire,
steel t-posts, wooden posts, and t-post clips with one post every twelve feet and a wood
post between every four steel posts was assumed. Using this fence, for example, the
average amount of stream bank fencing installed was 1,695 feet at a cost of $945.
Controlling run-off involved installing vegetative buffers. Run-off control for dairy and
beef farms with greater than 1,000 animal units was valued at $1.42 per AU and $4.69
per AU for all other farms (V ollmer-Sanders, Batie, and Wolf). Following Marlado a
vegetative buffer installed for equine operations was estimated to cost $1,300. In some
instances a farm was required to install a manure storage facility or milk house-water

facility to contain run-off. The cost of controlling waste run-off by building a storage

19

facility was calculated using Harrigan’s results where the cost was for example, $582 per
AU for a 100 AU herd and $259 per AU for a 1,000 AU herd.

Cost to implement corrective practices to mitigate veriﬁed complaints varied by
livestock enterprise and farm Size (Table 2d). Across all farms, dairy operations resulted
in the highest average cost of $16,502. The average dairy CAFO corrective practice cost
of $27,657 was almost twice the amount for dairy AFOS at $14,117. Beef AFOS average
corrective practice cost was higher than beef CAFO cost since the majority of beef AFOS
were required to install stream bank fencing. Poultry AFOS were most commonly
required to implement MMSPS which resulted in a higher average cost than poultry
CAFOS who were required to provide soil analysis and incorporate manure. Swine
operations had, the lowest average cost for corrective practices demonstrating their
awareness of environmental complaints and ability to be pro-active in prevention of
complaints.

Table 2d. AveragCosts to Implement Corrective Practices

 

 

 

All Farms CAFO‘ AFO2
Standard Standard Standard
Enterprise Mean Deviation Mean Deviation Mean Deviation
($0

Beef 4,809 8,322 1,847 1,691 4,983 8,525

Dairy 16,592 55,680 27,657 95,724 14,1 17 41,977

Swine 1,421 1,020 1,822 1,559 1,296 783

Equine 3,124 7,089 -- -- 3,124 7,089

Poultry 5,477 9,665 2,070 1,297 8,883 13,176

Crop 980 1,076 -- -- 980 1,076

Combination ‘

Livestock 3,771 7,568 -- -- 3,782 7,678

Other

Livestock 748 693 -- -- 748 693

 

rCAFOS (Concentrated Animal Feeding Operations) are livestock operations with
greater than 1,000 AU.

AF Os (Animal Feeding Operations) are livestock Operations with less than 1,000 AU.

20

We examined a log-level regression on those observations with positive
corrective practices costs to determine individual farm characteristics influencing these

costs expressed as
(2) log(C,°) = Xﬂ + e,-
where C i is cost of implementing corrective practices required to mitigate a veriﬁed

complaint, X denotes an array of variables that are hypothesized to affect cost Of

implementing corrective practices, ,8 is a vector of parameters, 5i is the error term, and I'

farm Index. We assumed that 5i 15 normally dIstrIbuted WhICh allows estImatIon of an

OLS regression from equation (2) using maximum likelihood techniques (Wooldridge
2002; 2003).

As in the probit estimation, the base scenario was an odor complaint issued in the
Spring for a dairy farm using long-term storage. Results are presented in Table 2e.
Receiving a surface water complaint was predicted to cost 45% more than receiving an
odor complaint. This resulted in an estimated average cost of $7,326 for implementing

the necessary corrective practices to mitigate a surface water complaint compared to an

 

5 We initially examined the probability of a veriﬁed complaint followed by a conditional
cost of mitigating the complaint using the Heckman two-step procedure where the ﬁrst
step determined the factors affecting the probability of a veriﬁed complaint (equation 1).
The results from estimating the probability of veriﬁed complaints (step 1) excluding
county characteristics were used to estimate the inverse mills ratio to test for selection
bias in the second step. The inverse mills ratio included in the estimation of equation (2)
was not signiﬁcant indicating no evidence of selection bias and allowing for independent
evaluation of equation (2).

21

odor complaint.6 Surface water complaints required the most expensive corrective
practices implemented on farms.

Increasing the number of animal units on the farm increased corrective practice
costs by 0.03%. This translated into a $3.87 cost increase at the mean for each additional
animal unit. The probit analysis above revealed that as the number of animal units
increased on a farm the probability of a veriﬁed complaint decreased. However, when a
veriﬁed complaint was realized, the costs were higher for larger livestock operations.
Large farms must have adequate manure storage and apply this manure within a short
period of time, leading to potential veriﬁed complaints regarding incorporating manure
and waste run-off. Also, larger dairy farms were often required to install run-off control
structures, which had declining cost on a per animal unit basis, but resulted in a higher
total cost than smaller operations.

Corrective practice costs for swine and equine operations were 77% and 96% less
than dairy operations costs, respectively. Equine operations often stockpile manure near
property lines or wooded areas without containment walls which could lead to potential
waste run-off and potential high corrective practice costs. The average equine herd size
with a veriﬁed complaint was 32 animal units (16 horses) on 29 acres compared to 360
animal units on 576 acres across all livestock enterprises. While the corrective practice
costs for equine operations were lower than dairy operations, they are still signiﬁcant

when holding all other factors constant.

 

6 The estimated average cost of the surface water complaint was adjusted as outlined in
Wooldridge (2003) by regressing the coefﬁcient vector of corrective practice costs on the
predicted corrective practice costs estimated in equation (2) with no constant at the data
means for the explanatory variables.

22

Corrective practices costs increased by 0.1% for each additional day needed to
implement the required corrective practices. The total cost increased by $15.01 at the
mean for each additional day it took to mitigate a veriﬁed complaint. Incorporating
manure or taking soil or manure samples takes little time, which may result in a lower
corrective practice costs. However, installing stream bank fencing and controlling run-
off require longer implementation time.

Table 2e. Explaining Corrective Practice Costs Required to Mitigate Veriﬁed
Complaints

 

Standard
Variable Coefﬁcient Error P-value
Complaint Type 1
Groundwater 0.4245 0.3524 0.2290
Surface water 0.4549 * 0.2454 0.0650
Combination complaint -0.1202 0.2992 0.6880
Other complaint 0.3295 0.5662 0.5610
Farm Characteristics
AU 0.0003 ‘ *** 0.0001 0.0140
Distance -0.0165 0.1962 0.9330
Days to implement 0.0012 *** 0.0006 0.0400
Manure Storage
No storage 0.1568 0.3248 0.6300
Short-term -0.0153 0.2445 0.9500
. . 3
Livestock Enterprise
Beef -0.4203 0.2863 0.1430
Swine -0.7661 ** 0.3531 0.0310
Equine -0.9636 *** 0.2621 0.0000
Poultry -0.6408 0.6695 0.3390
Crop -l.0849 0.7759 0.1630
Other livestock -1.3726 0.9932 0.1680
Combination livestock -0.3328 0.3904 0.3950
4
Seasonal factors
Summer 0.0003 0.2224 0.9990
Fall 0.0962 0.2837 0.7350
Winter -0.2200 0.2541 0.3870
Year -0.01 1 1 0.0449 0.8050
Constant 29.4330 89.9102 0.7440

23

Table 2e cont.

 

 

Prob F(20,321) 0.0001

Pseudo R-square 0.1441

Predicted probability at mean 0.0908

Sample size 342

rBase complaint type = odor *** Signiﬁcant at 1% level
2Base livestock type = dairy ** Signiﬁcant at 5% level
3Base manure storage type = long-term * Signiﬁcant at 10% level
4

Base season = spring

.5 Conclusions

We explored the relationship between citizen complaints, livestock production
characteristics, county level characteristics, and costs associated with corrective practices
implemented on Michigan livestock farms. Farms that received surface water and
combination complaints as compared to odor were more likely to have a veriﬁed
complaint. In contrast an increase in the number of animal units decreased the
probability of a veriﬁed complaint. Swine and poultry operations were found to have a
decreased probability of receiving a veriﬁed complaint. The implication is that poultry
and hog farms may be justiﬁed in higher expenditures to control odor even though it is
not currently a legal environmental compliance issue.

Veriﬁed complaints and corrective practices were required for the majority of
complaints issued. The corrective practices required to mitigate surface water complaints
resulted in the highest costs. Surface water control is a necessity in a state surrounded by
four of the ﬁve Great Lakes. Surface water complaints were received by all livestock
groups and indicate the importance of education and assistance to ensure livestock

operations are controlling potential run-off.

24

Dairy Operations realized the highest costs to implement corrective practices to
mitigate veriﬁed complaints. This may be due to the manure and milk-house wastewater
handling technology set common to dairy farms built prior to the recent stringent
regulations. In addition, swine and poultry farms have a history of being relatively pro-
active regarding environmental concerns.

Manure management for equine facilities is becoming increasingly important due
to number of horse facilities with a limited land base to properly store and dispose of
manure. Horse farms, even though they had less than four horses on average, often were
required to dispose of manure stockpiles and control run-off (United States Department
of Agriculture, 2008). These practices can become very costly for operations with a
small number of animals.

The results identiﬁed potential areas of improvement for voluntary programs,
such as Michigan’s Right to Farm program. Voluntary programs are designed to help
producers follow environmental guidelines with the objective of avoiding ﬁnes and
possible legal actions. The results support continued programs for producer and public
education as well as the continued support of cost share programs which provide partial
funding for operations to update their manure storage facilities and install stream bank
fencing, the two most common and expensive capital investment corrective practices

implemented on farms.

25

REFERENCES

Bundy, L., D. Wolkowski, J. Peters, and C. Laboski. “Avoiding Soil Sampling and
Testing Pitfalls in Qualifying For Government Programs Requiring a Nutrient
Management Plan.” Wisconsin Crop Manager. July 2006. Online. Available at
httg//ipcm.wisc.edu/ WC MNews/tabid/53/Entryld/l 03/Avoiding-Soil-Sampling-
and-Testing-Pitfalls.aspx. [Accessed March 2009.]

Cohen, M.A. “Empirical Research on the Deterrent Effect of Environmental Monitoring
and Enforcement.” Environmental Law Rep. 30,4(2000): 10245-10252.

Dasgupta, S. and D. Wheeler. “Citizen Complaints as Environmental Indicators:
Evidence from China.” Working Papers-Environment. Pollution, Biodiversity, Air
Quality. 1704. The World Bank, Washington DC, 1997.

Eckert, H. “Inspection, Warnings, and Compliance: The Case of Petroleum Storage
Regulation.” Journal of Environmental Economics and Management
47,2(2004):232-259.

Environmentally Concerned Citizens of South Central Michigan. Online. Available at
http://nocafos.orR/. [Accessed March 2009.]

Hadley, G., S. Harsh, and C. Wolf. “Managerial and Financial Implications of Major
Dairy Farm Expansions in Michigan and Wisconsin.” Journal of Dairy Science
85,8(August 2002):2053-2064.

Hadrich, .I.C., T.M. Harrigan, and CA. Wolf. “Economic Analysis of Livestock Manure
Handling Systems in the Great Lakes Region.” Working Paper. 2009.

Hadrich, .I.C., C.A. Wolf, J.R. Black, and SB. Harsh. “Incorporating Environmentally
Complaint Manure Nutrient Disposal Costs into Least-Cost Ration Formulation.”
Journal of Agricultural and Applied Economics 40, 1 (April 2008):287-300.

Harrigan, T.M. “Simulation of Dairy Manure Management and Tillage Systems.”
Unpublished Ph.D. Dissertation, Department of Agricultural Engineering,
Michigan State University, 1995.

Heyes, A. “Implementing Environmental Regulation: Enforcement and Compliance.”
Journal of Regulatory Economics 17,2(2000): 107-129.

Huang, H. and G.Y. Miller. “Citizen Complaints, Regulatory Violations, and Their
Implications for Swine Operations in Illinois.” Review of Agricultural Economics
28,1(2006):89-110.

26

Maraldo, D. “A Pollution Reduction Optimization Analysis for the Lower Fox River
Basin and Green Bay Area of Concern.” Online. Available at
http://www.uvgb.edu/watershed/REPORTS/Svmposium/2008/Poster-
I_Cadmus_Final.pdf. [Accessed March 2009.]

 

Michigan Information Center, State Budget Ofﬁce. “Estimates Population of Michigan
Counties, 1990-1999.” April 2000.

Michigan Department of Agriculture. Michigan Commission of Agriculture.
“Generally Accepted Agricultural and Management Practices for Site Selection

and Odor Control for New and Expanding Livestock Production Facilities.”
March 2008a.

. Michigan Commission of Agriculture. “Generally Accepted Agricultural and
Management Practices for Manure Management and Utilization.” March 2008b.

Murphy, M. and D. Nicholson. “Manure Management Practices of 20 Horse Ranches in
Marin County.” Marin Resource Conservation District and Marin County
Storrnwater Pollution Prevention Program. 2002.

Sierra Club. “Living a Nightmare: Animal Factories in Michigan.” Online. Available at
http://wwwsierraclub.org/factoryfarms/nightmarejdocumentaryasp [Accessed
March 2009.]

United States Census Bureau. State Population Datasets-County Population Datasets.
1997, 2002. Online. Available at http://www.census.gov/_popest/datasets.html.
[Accessed December 2007.]

 

United States Department of Agriculture. US. Census of Agriculture. State and County
Reports. 1997, 2002. Online. Available at hjpz//ww1.agensus.usda.gov/.
[Accessed December 2007.]

United States Department of Agriculture-National Agricultural Statistics Service (USDA-
NASS). 2007 Equine Survey Summary. NR-08-04, Lansing, MI, January 2008.

Vollmer-Sanders, C., S. Batie, and CA. Wolf. “Implementing a voluntary environmental
assurance program in Michigan” Under Review.

Wooldridge, J. M. Introductory Econometrics, A Modern Approach. 2nd ed. Mason,
Ohio: South-Westem College Publishing, 2003.

Wooldridge, J. M. Econometric Analysis of Cross Section and Panel Data. Cambridge,
Massachusetts: The MIT Press, 2002.

27

Young, K., C]. Carreira, H.L. Goodwin, and E. Wailes. “Economics of Transporting
Poultry Litter from Northwest Arkansas to Eastern Arkansas Cropland.”
Selectedpaper for the Southern Agricultural Economics Association, February
2005.

28

CHAPTER 3: ECONOMIC COMPARISON OF LIQUID MANURE TRANSPORT
AND LAND APPLICATION
3.1 Introduction

On many livestock operations farm managers have transitioned from daily manure
hauling to long-term manure storage. Increased herd sizes on a smaller number of farms
have resulted in increased manure with greater hauling distance to ﬁelds for land
application. Applying manure on a limited land base creates management challenges for
protecting surface and groundwater quality. Many states have adopted best management
practices for manure use as a condition of Right to Farm Act protection (Michigan
Department of Agriculture, 2008) which speciﬁcally address waste run-off control and
management. Such practices include long-term manure storage, manure application rates
based on soil test levels and limited winter-spreading. The need for cost effective options
has caused farmers to evaluate and change their manure hauling systems.

Suitable working days are the days available in a scheduled period during which
ﬁeld operations can be performed (Harrigan et a1, 1996; Rotz and Harrigan, 2005). If
manure hauling delays crop tillage and planting, the number of suitable days for planting
will decrease and decrease crop yield. Current options in manure transport and land
application vary greatly in labor, machinery requirements, ownership and operating costs,
and compatibility with the environment. Labor hours are a function of the machinery set,
hauling distance, and spreader capacity. If a farm manager has a poorly designed manure
hauling system, delays in manure transport and land application may cause delays in crop
tillage and planting. Harrigan et al. (1996) evaluated the effect of manure hauling in

livestock-based cropping systems for a 150 and 400-cow dairies in Michigan. Labor

29

availability was one 10 hr-person-day/day with the 150-cow herd and three 10 hr-person-
days/day with the 400-cow herd. Increased labor on the 400-cow dairy allowed parallel
manure application, tillage and planting operations and allowed ﬁeld work to be
completed within the time available.

Decision support systems to evaluate livestock manure as a nutrient source have
been developed. Koehler and Lazarus (2009) developed a spreadsheet-based decision
tool to calculate the value of manure in Minnesota. Farm inputs included livestock
enterprise (beef, dairy, swine and poultry), type of manure (solid or liquid manure),
volume of manure, manure analysis, acres available for land application, manure
application method (broadcast or injection), planned application rate, crop nutrient needs,
and commercial fertilizer cost. Manure application rates were calculated for nitrogen
availability and phosphorus limiting application rates. Commercial fertilizer prices were
used to calculate the value of manure. Machinery costs were not included.

Leibold and Olsen (2007) developed a spreadsheet-based cost calculator for swine
manure to evaluate swine manure application with three different Iowa crop rotations.
User inputs included the number and Size of hogs, average manure analysis, 5-year
average crop yield, planned application rate, and fertilizer prices. A base hauling cost of
0.26 ¢/L (1¢/gal) for liquid manure was assumed with a surcharge of 0.026 ¢/L-mile
(0.1¢/gal-mile) for hauling. The cost calculator did not restrict manure application rates
based on nitrogen availability or phosphorous limitations.

Whole-farm simulation models have been developed to evaluate farming systems
with manure management as a sub-model (Borton et al., 1995; Harrigan et al, 1996; Rotz

et al., 2008). The Integrated Farm System Model (IFSM) uses simulation to evaluate

30

production costs, incomes, and economic returns of the farming operation based on local
weather data (Rotz et al., 2008). The model simulates forage production on dairy farms
with sub-models evaluating cropping systems, manure production, and return of nutrients
back to the land. Borton et al. (1995) expanded DAFOSYM, an earlier version of IFSM,
to compare the performance and economics of manure hauling systems and interaction
with feed production on dairy farms. Net return over feed and manure costs was $25/cow
at distances greater than 5 km (3 mile) for truck-drawn nurse tanks for over-the-road
transport. Harrigan et al. (1996) used DAFOSYM to evaluate the effect of manure
hauling systems on timeliness of tillage and planting.

Koelsch et al. (2007) developed a decision support system that included the
feeding, cropping system, and costs associated with the manure hauling system. User
inputs included animal numbers, body weight, ration formulation, housing, manure
analysis, crop yield and fertilizer needs, manure application method (spreader tank, towed
hose, and big gun), and average distance to ﬁeld. Equipment size and travel speed was
deﬁned by the user and used to calculate ﬁeld application time, road travel time, set-up
time, and manure application rate.

Accounting for manure hauling system costs must include the major components
of the manure hauling operation while providing a ﬂexible tool for evaluating fann-
speciﬁc manure hauling system decisions. Timeliness of tillage and planting is
dependent on manure hauling time. When planning equipment purchases, farmers must
be able to estimate manure hauling time as a function of spreader capacity and hauling
distances. Farmers must also select manure hauling systems in a cost effective and

environmentally compliant manner. There is a need for a user-friendly decision support

31

system that provides an accurate estimate of time needed for manure pumping, transport,
and land application as a function of hauling distance, spreader capacity, manure
equipment cost, labor, and the nutrient value of manure. Accurate cost estimates and
hauling time will facilitate efﬁcient comparison of cost-effective alternative systems to
aid manure movement between livestock and crop producers.
3.2 Objectives

There is a need for a user-friendly decision support system that provides an accurate
estimate of cost, time needed for manure pumping, transport, and land application, and
the nutrient value of the manure. Speciﬁc objectives of this work were to:

1. Develop a spreadsheet-based model (MANURE$HAUL) to estimate manure
hauling cost, labor requirements and nutrient value for commonly used top-
loading tank spreader systems.

2. Validate the model by comparing estimated hauling costs and time with those

reported by two Michigan farms

3. Develop a model in equation for estimating the cost of commonly used tank

spreader system as a function of tank volume and transport distance

3.3 Model Development

A spreadsheet-based model, MANURE$HAUL, was developed to evaluate the
effect of machinery set and hauling distance on hauling capacity, time, and cost.
MANURE$HAUL estimates the manure hauling rate as a function of hauling distance
and spreader capacity. Ownership and operating costs were calculated for tractors,

trucks, manure Spreaders, nurse tanks, agitation and pit pumps, and tillage equipment.

32

Nutrient use was based on the Tri-state fertilizer recommendations (Vitosh, Johnson, and
Mengel, 1995).
3.3.1 Manure Hauling Rate

Liquid manure production for beef, dairy, and swine was based on livestock
enterprise, animal size and number of animals on the farm (MWPS, 2004). The manure
hauling rate was calculated for top-loading tractor-drawn spreader tanks and spreader
tanks with truck-drawn nurse tanks for over-the-road transport to a tractor-drawn
spreader in the ﬁeld. Machinery system-speciﬁc coefﬁcients were used to estimate
hauling capacity as a function of spreader volume, material ﬂow rates, transport distance,
and support time for loading and unloading spreaders (Harrigan, 2009).

Three manure hauling systems were used: (1) tractor-drawn spreader tank, (2)
truck-mounted spreader tank, and (3) truck-drawn nurse tanks for over-the-road transport
to a tractor-drawn spreader tank for ﬁeld spreading. The truck-drawn nurse tanks were
equal to or twice the volume of the tractor-drawn spreader tank in the ﬁeld. Hauling rates
were estimated for standard and high speed tractors. Hauling capacity decreased as
hauling distance increased. The hauling capacity of truck-mounted spreaders was similar
to tractor-drawn spreaders when hauling near the storage structure. Truck-mounted
spreaders had faster over-the-road transport speeds than tractor-drawn spreaders which
provided an advantage with longer hauling distances. Truck-drawn nurse tanks for over-
road-transport to a tractor-drawn spreader in the ﬁeld provided greater hauling capacity
for longer distances. When hauling near storage, truck-drawn nurse tanks had idle time

in waiting to transfer manure to the tractor-drawn spreader. Two truck-drawn nurse tanks

33

hauling to a tractor-drawn spreader required three operators compared to a one operator
for a tractor-drawn spreader tank or truck-mounted spreader.

MANURE$HAUL users can deﬁne the type and number of machinery sets to be
used with each farm. For example, a farm may choose to use two tractor-drawn spreader
tanks in parallel rather than two truck-drawn nurse tanks with a tractor-drawn spreader.
Broadcast application, broadcast application with tillage incorporation, and injection
were the three manure application methods. Broadcast application with tillage
incorporation required an additional tractor and tillage equipment for manure
incorporation. Slurry injection decreased the manure hauling capacity compared to .
broadcast application and resulted in greater downtime for repair and maintenance.

3.3.2 Equipment costs

Equipment ownership costs included depreciation, interest, taxes, housing, and
insurance. Operating costs included repairs and maintenance, fuel, lubrication, and labor.
Total hauling cost is the sum of ownership and operating costs. Purchase prices were
collected for four equipment categories: tractors, tank spreaders, agitators and pumps, and
tillage equipment for manure incorporation. Annual and hourly ownership and operating

costs were calculated for all items speciﬁed in Table 3a.

Table 3a. Manure Hauling System Equipment

 

Tractor/T ruck (fog Manure Spreader Agitator and Purm) Incorporation

 

- Manure spreader - Slurry - Small pit pump - Injector System
- Agitation - Truck-mounted - Medium pit pump - Tandem Disk
- Incorporation - Nurse tank - Large pit pump - Field Cultivator
- Truck-mounted - Small lagoon pump - Combination
Spreader Tool
- Truck for nurse tank - Large Lagoon

pump

 

34

Purchase price was collected for a range of tractor power (pto-kilowatt, pto-pto-
hp) using the on-line “build your own” tractor utility for Case 1H and John Deere
equipment. Prices were calculated for 36 diesel-powered, ﬁxed frame (non-articulating)
wheeled tractors ranging from 56-205 pto-kW (75-275 pto-pto-hp). All list prices were
for new tractors effective October 2008. Estimated price functions for new tractors based
on the data collected are presented in table 3b.

The purchase price for used tractors was based on auction data for AGCO-Allis,
Case-1H, and John Deere tractors (Iron Solutions, 2006). The average used price for ten
to twenty year old tractors was collected for 116 diesel-powered, ﬁxed frame (non-
articulating) wheeled tractors ranging from 56-194 pto-kW (75-260 pto-pto-hp). The
used tractor price functions based on a linear regression of tractor power on sale price are
presented in table 3b.

Tank spreader purchase prices were collected from three manufacturers. Tank
capacities ranged from 6,813-35,957 L (1,800-9,500 gal). Company representatives
indicated the most common options selected for the spreader tanks. Estimates for truck-
mounted and truck-drawn tanks were supplied by custom applicators. In Michigan,
refurbished gasoline tankers are typically used as nurse tanks for manure transport from
the manure storage facility to a tractor-drawn tank spreader in the ﬁeld. Custom
applicators reported that the purchase price of a used gasoline tank was approximately
$10,000 plus an additional $15,000 investment for hydraulics, pumps and other
modiﬁcations. The purchase price for truck-mounted Spreader tanks were provided by
one manufacturer for tanks ranging from 11,923-18,168 L (3,150-4,800 gal). Truck-

mounted purchase price functions are presented in table 3b.

35

Semi-tractors were used with truck-drawn nurse tank systems. Based on
discussions with three custom manure applicators, an average age of semi-tractors ranged
from 5-15 years with an average age of 10 years. A representative purchase price was
$10,000 with an additional $10,000 investment for a hydraulic (wet) kit for the truck. A
truck-drawn nurse tank list price was $45,000--$25,000 for a nurse tank and $20,000 for
the truck.

Farmers generally have three options for liquid manure application: 1) inject
manure directly into the soil during application, 2) surface broadcast, and 3) surface
broadcast with incorporation. Injection reduces the number of passes over the ﬁeld and
improves nitrogen recovery, but reduces the manure hauling rate (Harrigan, 2009). Cost
information was collected for injection equipment from three custom manure applicators
and two Spreader tank manufacturers. Mounting a toolbar for injectors and other
alterations to a manure spreader were valued at $9,000 with an additional $1,600 per
injector installed on the toolbar. In MANURE$HAUL, a toolbar with 6 injectors was
used for equipment comparisons.

Manure incorporation requires the farmer to use a tractor and tillage tool to
incorporate manure into the soil. Purchase price data was collected for tandem disks,
ﬁeld cultivators, and combination tillage tools using on-line “build your own” equipment
options for Case-1H and John Deere tillage equipment (Case-1H, 2009; Deere and Co,
2009). All purchase prices were effective February 2009 for 44 tandem disks ranging in
ﬁeld width from 3.5-10.36 m (11.5-34 ft.), and 39 ﬁeld cultivators ranging 5.5-18.3 m
(18-60 ft.). Estimated tillage equipment purchase price functions are presented in table

3b.

36

Table 3b. Estimated equipment list price functions

 

 

 

 

Equipment Price AY) Intercept Slope X1 R2 X variable
Tractors
New tractor PriceNT -32,582 884 pt0-hpNT 0.90 pIO-hpNT
pto— pto-kWNT
kWNT
Used tractor Priceur 7,470 183 PIO'hPUT 0.43 ptO-hPUT
Pto' ptO-kWUT
kWNT
Semi-tractor 20,000
Truck-mounted 50000
truck
Spreaders
Slurry tank Prices -3,786 2.91 Ls 0.87 LS = liters
('3’786) (l ]) (gals) (gals = gallons)
Truck- PriceTM 18,219 0.63 LTM 0-97 LTM = liters
mounted tank (18,219) (2.4) (gale) (gaITM = gallons)
Nurse tank 25,000
Injectors & Price] 9,000 1,600 Injectors -- Injectors=number
toolbar of injectors
Tillage
Equipment
Tandem disk PriceTD -884 5,1 17 ﬁeldmm 0.89 ﬁeldmm =ﬁe|d
(-884) (1.560) (ﬁeldﬁm width (m)
(ﬁeldﬁTD =ﬁeld
width (ft))
Field cultivator Pricepc -2.534 3,391 Fieldmpc 0.76 FICldeC =ﬁe1d
(-3,068) (1,125) (ﬁeldftpc) width (m)
(ﬁeldﬁpc =ﬁeld
width (ft))
Combination PriceCT 2,577 5,691 ﬁeldthT 0.84 ﬁeldmCT =ﬁe1d
tool (2,404) (1,735) (ﬁeldﬁCT width (m)
(ﬁeldﬁCT =ﬁeld
width (ft))
3.3.3 Ownership costs

Straight line depreciation was calculated as the difference between the beginning

and ending value for a 10 year economic life. Remaining values were based on list price,

tractor age, and annual hourly use (ASABE, 2007). Remaining value coefﬁcients for

37

agitators and pumps are not provided in the ASABE Standard (EP496), therefore
coefﬁcients for the nearest equipment in size and use was used, which was
“miscellaneous farm equipment”.

New tractors were assumed to have a base annual hourly use of 500 hours and
350 hours for used tractors. Calculated manure hauling hours were added to the base use
hours for new and used tractors. Manure spreader base hours were calculated in
MANURE$HAUL as time needed for manure transport and land application. Time for
pumping and agitation was estimated as the manure pumping time plus eight hours for
the initial set-up and agitation for each of two pumping events each year. The salvage

value of equipment was estimated as:

(1) SVn =(I-RV,,)*LP

where S V" was salvage value of equipment in year n, RV" was remaining value of

equipment in year n, and LP was equipment list price.

MANURE$HAUL valued taxes, housing, and insurance at 1%, 0.75%, and 0.25%
of the list price of equipment, respectively (ASABE, 2007). The real interest rate was set
at 5% (Edwards, 2005).

3.3.4 Operating costs

Annual operating costs included repairs and maintenance, fuel, lubrication, and
labor. Repair and maintenance costs were based on accumulated use (ASABE, 2007).
When repair factors were not listed in the ASABE standard, a composite of repair factors
was used to best reﬂect repair and maintenance costs provided by custom manure
applicators. Repair factors were not provided for trucks, manure spreaders and agitator

pumps. Repair factors for manure spreaders, agitators, pumps and other equipment are

38

listed in table 3c. Repair and maintenance cost factors for injectors were estimated based
on information from three custom manure applicators. Repair and maintenance costs
were assumed to be $240/injector point for every 405 ha (1000 ac) of use in loam or

sandy-loam soil.

Table 3c. Repair factors for trucks, manure spreaders, and agitators and pumps

 

 

Equipment RFl RF2 Similar machinery

Small Tractors , <60 kW (<80 hp) 0.007 2.0 --

Medium Tractors, 60-112 kW (80-150 hp) 0.007 2.0 --

Large Tractors, >112 kW (>150 hp) 0.007 2.0 --

Truck* 0.007 2 2 wheel drive and
stationary tractors

Manure Spreader* 0.16 1.6 Forage wagons and
fertilizer spreaders

Agitators and pumps* 0.22 1.8 Forage blowers

Disk 0.180 1.7 --

Field Cultivator 0.270 1.4 --

Chisel Plow 0.280 1.4 --

 

*Composite repair factors

Fuel use was estimated as 0.22 L/pto-kW-h (0.044 gal/pto-hp-h) for tractors
(ASABE, 2007) and 0.086 L/pto-kW-h (0.0170 gal/pto-hp-hr) for trucks (Harrigan,
2001). Lubrication was estimated as 15% of the fuel cost (ASABE, 2007). Labor was
valued at $12/hr (Black et al, 2008; Koelsch et al., 2007). Labor hours for each operation
were those calculated by MANURE$HAUL plus 10% for set-up and scheduled
maintenance. Agitation and pumping hours were estimated based on a pumping rate of
7,192 L/min (1,900 gal/minute) plus an additional 16 hours (eight hours, two times per
year) for set-up and agitation. Tillage hours were based on machine width, a travel

speed of 8 km/hr (5 mph), and tillage implement ﬁeld efﬁciency of 85%.

39

3.4 Nutrient value of manure

The value of manure nutrients applied to the land is a function of the nutrient
content of manure; quantity of manure applied, and method of application. Tillage
incorporation or injection conserve volatile nitrogen and prevents run-off. Injection
reduces the odor associated with land application but results in greater downtime for
repairs and maintenance for injection equipment. A broadcast application with
immediate incorporation is generally faster than injection, but nitrogen losses can be
Signiﬁcant if there is a time lag between manure application and incorporation.
3.4.1 N volatilization losses

The best way to recover costs associated with manure storage and handling is to
apply the manure at an agronomic rate, account for manure nutrients, and reduce
commercial fertilizer purchases. Non-mobile nutrients such as potassium (K) and
phosphorus (P) are easy to account for, but calculating nitrogen (N) credits is a challenge.

Manure contains nitrogen in inorganic and organic forms. Organic N is not available for

crop growth untIl It Is mIneralIzed to ammonIum (N H4 ). Ammomum N IS faIrly stable

and available for plant uptake, but a portion is immobilized by microbial biomass, and

. . . . + . - . . . .
nItrIfyIng bacteria convert NH4 to nItrate (N03 ) which 15 subject to loss by leaching or
denitriﬁcation and subsequent loss to the atmosphere. Volatile ammonia (NH3) is

+ . .
transformed from NH4 and can be lost to the atmosphere after land application.

Nitrogen lost to the atmosphere is not available for crop production. Injecting the

slurry into the soil or incorporating it with tillage is the most effective ways to reduce

40

NH; losses. Ammonia emissions increase with an increase in temperature and wind
speed, and decrease with an increase in relative humidity. Organic N becomes available

for crop growth over time as it is mineralized to the ammonium form. Available organic

N is deﬁned as:
(2) Available Organic N = (Total N- NH4-N) *m

where Total N is total nitrogen, NH4-N is ammonium, and m is the mineralization factor.
The mineralization factor, m, describes the fraction of organic N available for plant use
in the ﬁrst season following manure application (MWPS, 1993).

Plant available N (PAN) is a function of total soil N, the N available in soil for crop

use, amount of organic N mineralized, and the amount ofNH4—N in the soil. Jacobs

(1995b) estimated NH4-N volatilization losses for surface broadcast and manure injection

in Michigan as a function of the time delay between manure application and

incorporation (Table 3d).

Table 3d. Estimated N volatilization losses by manure application method

 

 

Days ‘0 , NH4-N NH4-N
Incorporation Retained (%) Lost (%)
Injection 100 0
0-1 day 70 30
2-3 days 40 60
4-7 days 20 80
>7 days 10 90

 

*Source: Jacobs (1995b ), Table 3.

41

3.4.2 Fertilizer recommendations

Fertilizer recommendations for ﬁeld crops are a function of the crop grown,
expected crop yield, soil type, and soil test levels. Field crop fertilizer recommendations
for Michigan, Ohio and Indiana are published in the Tri-State Fertilizer
Recommendations (1995) and follow a “build-up”, “maintenance” and “draw-down”
approach to managing soil phosphorus. Soil test results below a critical level indicate a
nutrient deﬁcit and a need to “build-up” or raise soil test levels for optimal crop yield.
Critical and maintenance limits vary by crop, soil type, and state (Vitsosh, Johnson, and
Mengel, 1995). Soil test results at the maintenance level result in a level of nutrients to
provide optimal crop yield. Soil test values greater than the “maintenance” level indicate
a surplus and the need to “draw-down” or reduce nutrients to “maintenance” levels. For
example, for loam soil in Michigan, Bray Pl soil tests results with less than 167 kg/ha (74
lb/ac) allow a “build-up” of soil phosphorus whereby manure can be applied at N-

removal application rates. Because manure application rates based on N typically exceed

crop P205 removal, the soil P level increases. Fields testing 167-336 kg/ha (75-299

lbs/acre) P205are in the “maintenance” zone and manure or commercial nutrients can be

applied at crop removal rates. Phosphorus generally limits manure application at the crop

“maintenance” level. Fields testing 337 kg/ha (300 lb/ac) of P205require a “draw-down”

of soil phosphorus and manure application is not allowed until soil P2051evels drop

below 337 kg/ha (300 lb/ac).
Fertilizer recommendations in MANURE$HAUL are based on input by the user

for crop grown and expected yield using crop nutrient removal guidelines for Michigan

42

ﬁeld crops (Wamcke et al., 2004). The nutrient content of the manure can be estimated
based on typical values for livestock enterprises (MWPS, 2001) or can be provided by the
user based on manure analysis results. The quantity of manure nutrients applied was
based on the manure application rate manure analysis. Table 3e presents a summary of
Bray P1 soil test results, fertilizer recommendations, application rates, and nutrient credit
guidelines used in MANURE$HAUL.

Table 3e. Bray Pl soil test results, fertilizer recommendations, manure application
rates, and nutrient credit guidelines used in MANURE$HAUL

 

Bray P1 Soil Units Soil test Application Nutrient credits
Test classiﬁcation rate
0-167 kg/ha “build-up” Nitrogen N, p205 and K
(0-149 ) (lbs/acre) removal
P205
168-336 kg/ha “maintenance” Phosphorus N, p205 and K up
(1 50-299) (lbs/acre) removal
to crop P205
P205 removal
337+ kg/ha “draw-down” No None
(300+) (lbs/acre) application
P205

 

3.5 Procedure

The objective of this work was to develop a ﬂexible, easy-to-use model to
describe, evaluate, and compare a range of liquid manure transport and land application
systems. The model includes beef, dairy, and swine operations using tractor-drawn tank
spreaders, truck-mounted and truck-drawn tank spreaders, and tractor-drawn tank
spreader with truck-drawn nurse tanks for over-the-road transport. Manure hauling rates

were a function of spreader capacity, distance, and the manure hauling system chosen

43

(Harrigan, 2009). Manure was applied to ﬁelds with injection, surface broadcast, or
surface broadcast with incorporation.

Table 3f lists MANURE$HAUL user inputs and default values. Required inputs
are: livestock type, number of animals (or volume of manure for land application, L, gal),

tractor size (pto-pto-kW, pto-pto-hp), spreader volume (L, gal), crop area (ha, acres), crop

yield (kg/ha, ton/acre), soil test results (N, P205and K20) and hauling distance to ﬁeld

zones (km, mi).

Manure production and nutrient content are based on the user input for the
livestock type, size, and number of animals on the farm. Users can accept the default
values or override the calculated manure production and nutrient levels with results of a
manure analysis. Tractor, spreader tank, and tillage equipment ownership and operating
costs are based on user inputs for tractor size (pto-pto-kW, pto-pto-hp), spreader capacity
(L, gal), and equipment width (m, ft), respectively. Manure injector ownership and
operating costs are based on user input for the number of injectors used.

The default values for the economic parameters in MANURE$HAUL are listed in

Table 3f. Users can change fuel price ($/L, $/gal), labor wage rate ($/h), fertilizer prices
($/kg, $/lb N, PzOSand K20), and the economic life of equipment (5-10 years).

Economic parameters that are ﬁxed are fuel use (L/pto-kW-h, gal/pto-hp-h), annual

tractor use (hours), real interest rate, taxes, housing, and insurance.

44

Table 3f. MAN URE$HAUL user inputs, defaults, and override values

 

 

 

Parameter User Input Default Value Override
Animals and Equipment Yes No
Beef, dairy, swine“ Number of animals -- X
Manure Analysis N-P205-K20 -~ X
Manure production -- X
Tractor for spreader* pto-kW (pto-hp) -- X
Tractor for agitator“ pto-kW (pto-hp) -- X
Tractor for tillage* pto-kW (pto-hp) -- X
Truck for nurse tanks* pto-kW (pto-hp) -- X
Truck for truck- pto-kW (pto-hp) -- X
mounted spreader*
Manure spreader* capacity, L (gal) —- X
Nurse tanks* capacity, L (gal) -- X
Injectors* Number 6 X
Tillage equipment* Width, m (ft) -- X
Field Zones (1-4)
Crop acres* Yield, unit/ha --
(unit/acre)
hauling distance, --
km (mi)
soil test results --
Economic Parameters
Diesel fuel price $/ L ($/ gal) $0.46/L X
(S ‘1 .75/ gal)
Tractor fuel usage L (gal) 0.22 per pto- X
kW-h (0.044
per pto-hp-h)
Truck fuel usage L (gal) 0.086 per pto- X
kW—h (0.0170
per pto-hp)
Labor wage rate $/hr $12/hr X
Fertilizer prices
N $/kg ($/lb) $1 .43/kg X
($0.65/ lb)
P205 $/kg ($/1b) $2.03/kg X
($0.92/ lb)
K20 $/kg ($/lb) $1.65/kg X
($0.75/lb)
Economic life Years 5-10 X

45

Table 3f cont.

 

Tractor annual use Hours 500 X

Used tractor annual use Hours 350 X

Real Interest rate % 5% X

Taxes % of machinery 1% X
list price

Housing % of machinery 0.75% X
list price

Insurance % of machinery 0.25% X
list price

 

 

*Input values required for MANURE$HAUL to calculate manure hauling costs

AS farms consolidate and increase in size they acquire a land base with varying
distance for manure application. Delays in manure application in the spring can delay
crop planting and reduce crop yield. A well-designed manure hauling system will
prevent delays in crop planting in most years (Rotz and Harrigan, 2005). The manure
hauling cycle includes time required for loading the spreader, transporting the spreader to
the ﬁeld, unloading the spreader, and transporting the spreader back to the storage
structure. Manure hauling rates vary with machinery sets, hauling distance, spreader
capacity and other factors (Harrigan, 1997; 2009). A tractor-drawn spreader tank uses
one tractor, one spreader tank and one operator, and is an efﬁcient system when hauling
within a few mi of storage. An alternative is to use truck-drawn nurse tanks for over-the-
road transport to a tractor-drawn spreader in the ﬁeld. Compared to a tractor-drawn
spreader alone, this machinery set requires additional equipment and three operators, but
is more cost and labor efﬁcient for greater hauling distances. There is a need for a
decision tool to help manure managers evaluate, compare and select machinery systems

suitable for a range of hauling distances.

46

Tractor-drawn spreaders have an advantage when the ﬁelds are close to storage
because there is no need for in-ﬁeld nurse tank-to-spreader transfer but the hauling
capacity diminishes rapidly as hauling distance increases (Fig 3a). The hauling capacity
with a 4.8 km (3 mi) haul is less than one-half the capacity when hauling near storage.
Truck-mounted spreaders and tank spreaders working in parallel with nurse trucks for
over-the-road transport have an advantage with longer hauls because of their greater road

travel Speed.

 

0.1 .1 2 3 4 5 6 7 8 9 10
240000 ‘ ' ' ' A I A 1 I I 1
Transport distance, miles __ 60000
220000 ‘.
. ‘ - 55000
200000 J ................. _
I- 50000
180000 - .—; 2-26495 L tractor-drawn spreaders __ ”000
I. 160000 --— — - — H 13248 L spreader, 2-26495 L nurse — ,_
S ‘ +—+ 20495 L spreader, 2-26495 L nurse ‘" 400”" 8
.: 140000 - a
I- - 35000 E.-
“120000 —— —\. ——————————————— 3
a. '- 30000 '”

 

 

 

 

°IIIIIITTIII 0

0.161.6 3.2 4.8 6.4 8 9.6 11.3 12.914.516.1
Transport distance, kilometers

Figure 3a. Hauling capacity of two 26,495 L (7,000 gal) tractor-drawn tank spreaders
working in parallel, one 13,248 L (3,500 gal) tractor-drawn spreader in parallel with two
26,495 L (7,000 gal) nurse trucks, and two 26,495 L (7,000 gal) tractor-drawn spreaders in
parallel with two 26,495 L (7,000 gal) nurse trucks over 16.1 km (10.1 mile) hauling distance

47

MANURE$HAUL was used to estimate the costs and labor requirements for two
manure transport and land application systems for a representative, 700-cow dairy using
(1) two 26,495 L (7,000 gal) tractor-drawn spreaders, and (2) one 26,495 L (7,000 gal)
tractor-drawn spreader in the ﬁeld with two 26,496 L (7,000 gal) truck-drawn nurse tanks
for over-the-road transport. The hauling distance was varied from 0.5-8 miles when
hauling 6.1 million gallons for broadcast application. Two tractor-drawn Spreaders had a
lower cost 0.35-0.53¢/L (1.3 to 2 ¢/gal) than one tractor-drawn spreader in the ﬁeld with
two over-the-road transport nurse tanks, 0.53-0.58¢/L (2.0 to 2.2 ¢/gal), when land
application was within 3 miles of storage (Fig. 3b). Hauling time ranged from 126 h to
210 h when hauling up to two and one-half miles with the tractor-drawn Spreaders and
210 h to 211 h with the tractor-drawn/nurse truck system. Beyond 3 miles the cost for the
two tractor-drawn spreaders increased from 0.53 to 1 ¢/L (2.2 to 3.8 ¢/gal) with an eight
mile haul while the cost for the tractor-drawn/nurse truck system increased to 3.2 ¢/gal
with an eight mile haul. Hauling time was 458 h (22.9 days) with two tractor-drawn
Spreaders and 365 h (36.5 days) with the nurse trucks with eight-mile hauls. Based on
cost and labor requirements for manure transport and land application, nurse truck-based
systems had an advantage when the hauling distance was three miles or more. This result
was consistent with the experience of custom applicators in the Great Lakes Region of

when to switch from tractor-drawn Spreaders to nurse truck/spreader-tank systems.

48

 

 

0.6 I l l l l j I 1
Transport distance, miles ‘ f’ 2-2
0.55 -I
i” — 2
0.5 P—-— ............ J, ._ . - __
L 1.8
0.45 - ..,..
0 ‘1 H
E _ — 1.6-.2
3 0 4 —- ------- 3T - _________ 30
E / g
8 r— 1.4 Q
035 - U
e- 1.2
0.3 —"_' ’ . ________________
0.25 .. H 2-34065 L tractor-drawn " I
H 1-34065 L tractor-drawn w/ 2 nurse tanks
r- 0.8

 

 

“ II I I I I I II

0.81.6 3.2 4.8 6.4 8 9.7 11.3 12.9
Transport distance, kilometers

Figure 3b. Manure hauling cost for a 700-cow dairy using two 34,065 L (9,000 gal)
tractor-drawn spreaders working in parallel and one 34,065 L (9,000 gal) tractor-drawn
spreader in parallel with two 34,065 L (9,000 gal) nurse tanks over 12.9 km (8 miles).

3.6 Representative dairy farm

Many of the questions that manure managers have at the systems engineering
level relate to capacity, cost and labor requirements of the manure hauling system. An
Objective in developing MANURE$HAUL was to create a ﬂexible model that could be
use to describe, evaluate and compare a range of manure transport and land application
methods. Land application methods include surface broadcast or subsurface injection

with tractor-drawn tank spreaders, truck-mounted or truck-drawn tank Spreaders, or a

49

tractor-drawn spreader in parallel with nurse trucks for over-the-road transport. TO
illustrate the ability of the model to describe, evaluate and compare a range of manure
transport and land application Options, four systems were compared on four
representative dairy farms with 175-, 350-, 700- and l400-cow herds.

The land available for each herd was based on 1.2 ha (3 acres) per cow and a
cropping program of corn grain, corn silage and alfalfa on loam soil (Wittenberg and
Wolf, 2005). The area allocation for corn grain, alfalfa, and corn silage was typical for
Michigan farms with 50% of the area in alfalfa and the remaining land divided between
corn grain and corn Silage. Sixty percent of the corn ground was planted to corn grain
with the remaining land in corn silage. Corn silage was assumed to be grown in ﬁelds
closest to the farm to facilitate corn silage harvest. Soil test results report the current
nutrients available in the soil before the crop is planted. Soil test results allow the farmer
to determine if the soil is in a “build-up”, “maintenance”, or “draw-down” zone following
the Tri-state fertilizer guidelines. Fertilizer recommendations as determined by crop
nutrient removal for optimal crop yield were determined by estimated yield goals for corn
grain and corn silage as listed in Table 3g (Wamcke et al., 2004). An average hauling
distance was 1.6 km (1 mi) for the 175-cow herd and 2.4, 3.2, 4.8 km (1.5, 2 and 3 mi)

for the 350-, 700- and 1400-cow herds, respectively.

50

Table 3g. Fertilizer recommendations and soil test results for corn grain and corn

silage

 

 

 

Yield N P205 K20
Fertilizer kg/ha
recommendation (lbs/acre)
Corn Grain 8.1 mg/ha 131 54 39
(130 bu/acre) (117) (48) (35)
Corn Silage 33.6 mg/ha 158 56 134
(15 ton/acre) (141) (50) (120)
Alfalfa Hay 13.4 mg/ha 302 87 336
(6 ton/acre) (270) (78) (300)
Soil tests results
Corn grain 45 90 112
(40) (80) (100)
Corn silage 78 90 157
(70) (80) (140)

 

 

Machinery sets were selected to complete manure hauling in approximately
twenty 10-hour calendar days or less. Tank size and equipment complements were
changed to accommodate greater volumes of manure as herd size and hauling distance
increased. Farms have numerous ﬁelds at varying hauling distances on their farm. To
decrease the number of inputs for ﬁelds in MANURE$HAUL, ﬁeld zones were created.
For example, a farm with four ﬁelds of varying area planted in com grain within one mi
of the manure storage facility are categorized as one ﬁeld zone with an average hauling
distance of 1.6 km (1 mi). (Table 3h) Tractor power was increased by 15 pto-pto-kW (20
pto-pto-hp) compared to a broadcast application when manure was injected. The
machinery sets selected were not necessarily optimal or least-cost systems, rather
machinery sets that would likely be used with herds of that size in the Great Lakes

Region. Purchase prices for the machinery sets chosen are listed in Table 3i.

51

Table 3h. Farm characteristics for l75-cow, 350-cow, 700-cow, and 1,400 cow dairy

 

 

Animals 175-cow dairy 350-cow daig 700-cow dairy 1,400-cow dairy
Dairy Cows 175 350 700 1,400
Dry Cows‘ 35 70 140 280
Heifers 88 175 350 700
Manure prod, L 5,781,822 11,554,038 23,108,079 46,216,155
(gal) (1,527,562) (3,052,586) (6,105,173) (12,210,345)
N, kg 36,100 144,300 288,601
(lbs) (79,515) 72,150 (158,921) (317,842) (635,684)
P205, kg 17,408 34,799 69,598 139,196
(lbs) (38,343) (76,650) (153,300) (306,600)
K20,kg 22,964 91,754 183,507
(lbs) (50,582) 45,877 (101,050) (202,101) (404,201)
Cropping System ha, km (acres, miles)
Field Zone 1: 30.8 ha, 2.4 km 61.5 ha 2.4 km 123 ha, 3.4 km 246.1 ha, 6.8 km
Corn grain (76 ac, 1.5 mi) (152 ac, 1.5 mi ) (304 ac, 2.1 mi) (608 ac, 4.2 mi)
Field Zone 2: 30.8ha, 3.2 km 61.5 ha, 5km 123 ha, 5 km 246.1 ha, 5 km,
Corn grain (76 ac, 2 mi) (152 ac, 3.1 mi) (304 ac, 3.1 mi) (608 ac, 3.1 mi)
Field Zone 3: 20.6 ha, 0.8 km 40.9 ha, 1.6 km 81.8 ha, 1.2 km 163.5ha 2.4 km
Com silage (51 acr, 0.5 mi) (101 ac, 1 mi) (202 ac, .75 mi) (404 ac, 1.5 mi)
Field Zone 4: 20.6 ha, .40 km 40.9 ha, 0.8 km 81.8 ha, 1.9 km 163.5 ha, 1.6 km
Corn silage ( 51 ac 0.25 mi) (101 ac,0.5 mi) (202 ac, 1.2 mi) (404 ac, 1 mi)
Average hauling
distance 1.6 km (1 mi) 2.4] km (1.5 mi) 3.2 km (2 mi) 4.8 km (3 mi)

 

Table 3i. Equipment used and size, and estimated purchase price for l75-, 350-,
700-, and 1,400 cow dairy herds

 

175-cow dairy 350-cow dairy 700-cow dairy 1,400 cow dairy

 

 

Broadcast Size Purch. Size Purchas Size Purchas Size Purch.

application Price (8) e Price e Price Price ($)
(3) (3)

Agitator 75 25,668 127 38,173 127 38,173 127 38,173

tractor, pto- (100) (170) (170 ) (170 )

kW (pto-hp)

Spreader 89 73,540 164 161,976 179 179,663 179 179,663

tractor, pto- (120) (220) (240 ) (240 )

kW (pto-hp)

Tillage 104 91,227 134 126,601 134 126,601 134 126,601

tractor,pto- (140) (180) (180) (180)

kW (pto-hp)

Nurse tank -- -- -- -- —- -- 298 50,000

truck, pto- (400 )

kW (pto-hp)

Spreader 11,355 25,978 28,388 71,195 34,065 86,268 34,065 86,268

tank, L (gal) (3,000) (7500) (9000) (9000)

Nurse tank, -- -- -- -- -- -- 2-34,065 25,000

pto-kW (2-9000)

(PIG-hp)

52

Table 3i cont.

 

Lagoon Med. 16,500 Large 30,000 Large 30,000 Large 30,000

Pump ‘

Tandem 5.5 27,817 7.6 38,104 9.8 49,021 9.8 49,021

disk, m (it) (18) (25) (32) (32)

Injection

Application

Spreader 1 12 100,071 179 179,663 194 197,350 194 197,3 50
tractor, pto- (150) (240 ) (260 ) (260 )

kW (pto-hp)

Toolbar 6 18,600 6 I 8,600 6 18,600 6 18,600
w/injectors

 

3.7 Results Discussion

MANURE$HAUL was used to estimate manure hauling costs, time, and the value
of land applied manure nutrients for representative 175-, 350-, 700-, and 1,400-cow
dairy herds. Broadcast, broadcast with immediate incorporation, and injection
application were used to compare nutrient recovery and costs.

3.7.1 Labor Requirements

The labor requirement for manure agitation, pumping, transport and application
was 16.6, 20.0, 18.7, and 34.6 days for the l75-, 350-, 700-, and 1,400-cow herds,
respectively (table 3 j). Additional time for manure incorporation with a tandem disk
ranged from 3.1 days with the 175-cow herd to 13.6 days with the l400-cow herd. One
11,365 L (3,000 gal) tractor-drawn spreader tank was able to complete the land
application within the time available for the 175-herd, and one 28388 L (7,500 gal)
spreader tank was adequate with the 350—cow herd when the average hauling distance
was 2.4 km (1.5 mi). Although two nurse trucks were used to increase hauling capacity
to ﬁelds greater than 4.8 km (3 mi) from storage, one 34,065 L (9,000 gal) spreader
working with two 34065 L (9000 gal) nurse trucks was unable to complete manure

hauling within the time available for the 1400-cow herd. In such a case the manure

53

manager may choose to purchase additional equipment, work longer days, or custom hire
manure hauling and land application services to increase hauling capacity. An alternative
management approach would be to include wheat or a small grain in the crop rotation to
expand the window of opportunity for land application.

When subsurface injection was used for manure application, time needed for
agitation, pumping, transport and land application increased 14%, 18.5%, 19.3%, and
6.8% for the 175-, 350-, 700-, and 1,400-cow herds, respectively, compared to a
broadcast application. Injection had less impact on timeliness with the 1400-cow herd
because 60% of the manure slurry was hauled more than 4.8 km (3 mi) with truck-drawn
nurse tanks. At this hauling distance the tractor-drawn spreader experienced idle time in
the ﬁeld waiting for a nurse truck to arrive. Because of this idle time there was little

advantage for broadcast application compared to subsurface injection.

Table 3j. Hauling time and costs for l75-cow, 350-cow, 700-cow dairy, and 1,400-
cow dairy using broadcast and injection application

 

175-cow dairy 350-cow dairy 700-cow dairy 1,400-cow dairy

 

Manure volume, L (gal) 5,781,822 11,554,038 23,108,079 46,216,155
(1,527,562) (3,052,586) (6,105,173) (12,210,345)
Hauling distance, km
(mi)
Field Zone 1: Corn
silage 0.40 (0.25) 0.8 (0.5) 1.9 (1.2) 1.6 (1.0)
Field Zone 2: Corn
silage 0.8 (0.5) 1.6 (1.0) 1.2 (0.75) 2.4 (1.5)
Field Zone 3: Corn
grain 2.4 (1.5) 2.4 (1.5) 3.4 (2.1) 5.0 (3.1)
Field Zone 4: Corn
grain 3.2 (2.0) 5.0 (3.1) 5 (3.1) 6.8 (4.2)
Average hauling 4.0 (2.5)
distance 1.6 (1.) 2.4 (1.5) 3.2 (2.0)
Manure machinery set 1 tractor-drawn . 1 tractor-drawn 2-tractor drawn 2 tractor-drawn
spreader spreader spreaders spreaders <3 mi,

1 spreader and 2
nurse tanks >3 mi

54

 

Table 3j cont. l75-cow dairy 350-cow dairy 700-cow dairy 1,400-cow dairy
Broadcast Application
Equipment
Pumping and agitation
Tractor, pto-kW (pto-
hp) 75 (100) 127(170) 127(170) 127(170)
medium lagoon large lagoon 2 large lagoon 2 Large lagoon
Pump (size) pump pump pumps pump
Manure Hauling
Spreader tractor,pto-
kW (pto-hp) 89 (120) 164 (220) 179 (240) 179 (240)
Spreader tank, L (gal) 11,355 (3,000) 28,388 (7,500) 34,065 (9,000) 34,065 (9,000)
Truck , pto-kW (pto-
hp) -- -- -- 298 (400)
Nurse tank, L (gal) -- -- -- 34,065 (9,000)
Incorporation
Tractor for tillage,
pto-kW (pto-hp) 104 (140) 134 (180) 134 (180) 134 (180)
Tillage equipment m 5.5 (18) 7.6 (25) tandem 9.8 (32) tandem 9.8 (32) tandem
(ft) tandem disk disk disk disk
Results 175-cow dairy 350-cow dairy 700-cow dairy 1,400-cow dairy
Hauling rate, L/hr
(gal/hr)
Field zone 1: com grain 32,744 (8,651) 62,528 (16,520) 121,910 (32,209) 92,231 (24,368)
Field zone 2: corn grain 28,795 (7,608) 41,447 (10,950) 94,281 (24,909) 102,83l(27,168)
Field zone 3: corn 42,339
silage (11,186) 71,102 (18,785) 172,471 (45,567) 142,235 (37,579)
Field zone 4: corn 45,149
silage (11,928) 80,851 (21,361) 153,635 (40,591) 161,738 (42,731)
Average hauling rate 37,257 (9,843) 63,982 (16,904) 135,574 (35,819) 124,759 (32,961)
Labor
Pumping, agitation,
transport and
application, hours
(days) 166 (16.6) 200 (20) 187 (18.7) 407 (34.6)
Tillage incorporation,
hours (days) 31 (3.1) 44 (4.4) 68 (6.8) 136 (13.6)
Cost
Agitation, pumping,
transport, application;
$/h 155.22 259.94 526.51 586.36
¢/L (¢/gal) 0.24 (0.93) 0.29 (1.09) 0.33 (1.24) 0.53 (2.02)
$lha (S/acre) 22.69 (56.07) 26.62 (65.75) 27.47 (67.88) 36.24 (85.55)
Tillage incorporation,
$/h 119.82 134.03 124.97 99.71
¢/L (gt/gal) 0.06 (0.25) 0.05 (0.19) 0.04 (0.14) 0.03 (0.11)
S/ha (S/acre) 5.98 (14.77) 4.72 (11.66) 3.40 (8.40) 2.72 (6.71)
Total cost ¢/L (¢/gal) 0.31 (1.18) 0.34 (1.28) 0.36 (1.38) 0.56 (2.13)

55

Table 3j cont.

Injection Application

Equipment

Manure Hauling
Spreader Tractor, pto-

 

kW (pto-hp) 112 (150) 179 (240) 194 (260) 194 (260)
Spreader tank, L (gal) 11,355 (3,000) 28,388 (7,500) 34,065 (9,000) 34,065 (9,000)
Truck , pto-kW(pto-
hp) -- -- -- 298 (400)
Nurse tank , L(gal) -- -- -- 34,065 (9,000)
Injector 6-point 6-point 6-point 6-point
Results
Hauling rate, L/hr
(gal/hr)
Field Zone 1: Corn
grain 29,092 (7,686) 52,343 (13,829) 102,313 (27,031) 90,737 (23,973)
Field Zone 2: Corn
grain 25,893 (6,841) 36,054 (9,525) 81,047 (21,413) 100,843 (26,643)
Field Zone 3: Corn
silage 36,726 (9,703) 58,810 (15,538) 140,132 (37,023) 121,516 (32,105)
Field Zone 4: Corn 38,928
silage (10,285) 66,077 (17,458) 126,183 (33,338) 134,296 (35,481)
Average hauling rate 32,660 (8,629) 14,087 112,419 (29,701) 111,848 (29,550)
Labor 175-cow dairy 350-cow dairy 700-cow dairy 1,400-cow dairy
Pumping, agitation,
transport and
application, hours
(days) 188 (18.8) 237 (23.7) 223 (11.35) 435 (36.3)
C 0st
Agitation, pumping,
transport, application;
$/h 167.81 265.03 533.07 605.43
¢/L (gt/gal) 0.29 (1.09) 0.33 (1.25) 0.38 (1.45) 0.60 (2.23)
$/ha ($lacre) 29.35 (72.52) 31.66 (78.23) 19.50 (48.16) 38.69 (95.59)
3.7.2 Hauling costs

The ownership and operating costs for manure agitation, pumping, transport and

land application ranged from about $166/h for the tractor-drawn spreader with broadcast

application on the 175-cow herd to more than $5 86/h with the 1400-cow herd using two

tractor-drawn spreaders for hauling distances less than 4.8 km (3 mi) and one tractor-

drawn spreader with two nurse tanks of equivalent size for over-the-road transport (Table

3j). On a per liter (gallon) basis the cost for agitation, pumping, transport and broadcast

56

application ranged from 0.33 ¢/L (1.24¢/gal) when using two tractor-drawn spreaders
with an average hauling distance of 3.2 km, (2 mi) with the 700-cow herd to 0.53 ¢/L
(2.02 ¢/gal) when using nurse trucks and hauling up to 4.8 km (3 mi) with the 1,400-cow
herd. Additional costs for tillage incorporation of manure with a tandem disk ranged
from 0.03-0.06 ¢/L (0.11 to 0.25¢/gal). Manure injection increased agitation, pumping,
transport and land application costs by 15% for the 700-cow dairy using two sets of

tractor-drawn manure spreaders.

3.7.3 Manure pumping and land application on four 1,400 cow dairies

Machinery sets and an average transport distance were varied for four
representative 1400-cow dairies to evaluate the cost for labor and cost for manure
agitation, pumping, transport and land application (Table 3k). The average transport
distance for farms 1 and 2 was 4 km (2.5 mi). Farm 1 used four 34,065 L (9,000 gal)
tractor-drawn spreaders working in parallel. Farm 2 used two 34,065 L (9,000 gal)
tractor-drawn spreaders for distances less than 4.8 km (3 mi), and a 34,065 L (9,000 gal)
tractor-drawn spreader with two 34,065 L (9,000 gal) nurse tanks for distances greater
than 4.8 km (3 mi). The average transport distance was increased by 50% to 6.4 km (4
mi) for farms 3 and 4. Similar to farm 2, farm 3 used two 34,065 L (9,000 gal) tractor-
drawn spreaders for distances less than 4.8 km (3 mi), and a 34,065 L (9,000 gal) tractor-
drawn spreader with two 34,065 L (9,000 gal) nurse tanks for distances greater than 4.8
km (3 mi). Farm 4 used two 34,065 L (9,000 gal) tractor-drawn spreaders with two
34,065 L (9,000 gal) nurse tanks. A list of farm parameters and results are provided in

Table 3k.

57

Table 3k. 1,400-cow dairy farm characteristics and equipment used for broadcast
and injection application

 

 

Farm 1 Farm 2 Farm 3 Farm 4

Manure volume, L 46,216,155 46,216,155 46,216,155 46,216,155
(gal) (12,210,345) (12,210,345) (12,210,345) (12,210,345)
Hauling distance,
km (mi)
Field Zone 1: Corn

silage 1.6 (1.0) 1.6 (1.0) 2.4 (1.5) 2.4 (1.5)
Field Zone 2: Corn

silage 2.4 (1.5) 2.4 (1.5) 3.2 (2.0) 3.2 (2.0)
Field Zone 3: Corn

grain 5.0 (3.1) 5.0 (3.1) 7.6 (4.7) 7.6 (4.7)
Field Zone 4: Corn

grain 6.8 (4.2) 6.8 (4.2) 10.1 (6.3) 10.1 (6.3)

Average hauling 4.0 (2.5) 4.0 (2.5) 6.4 (4.0) 6.4 (4.0)

Broadcast
Application
Equipment

Machinery set
Pumping and
agitation
Tractor, pto-kW
(PIG-hp)

Pump (size)

Manure Hauling
Spreader
tractor,pto-kW
(pto-hp)
Spreader tank, L
(gal)
Truck , pto-kW
(mo-hp)
Nurse tank, L
(gal)
Incorporation
Tractor for
tillage, pto-kW
(pto-hp)
Tillage
equipment m (ft)
Results
Hauling rate, L/hr
(gal/hr)
Field zone 4: corn
silage
Field zone 3: corn
silage

4 sets - tractor—
drawn spreader

127 (170)
2 Large lagoon

pump

4-179 (4-240)
4—34,065 (4-
9,000)

134 (180)
9.8 (32) tandem
disk

323,477 (85,463)

284,470 (75,157)

2 tractor-drawn
spreaders <3 mi,
1 spreader and 2

nurse tanks >3 mi

127 (1 70)
2 Large lagoon

pump

2-179 (2240)
234,065 (2-
9,000)

2-298 (2400)

234,065 (2-
9,000)

134 (180)
9.8 (32) tandem
disk

161,738 (42,731)

142,235 (37,579)

58

2 tractor-drawn
spreaders <3 mi,
1 spreader and 2

nurse tanks >3 mi

127(170)
2 Large lagoon

pump

2-179 (2-240)
2—34,065 (2-
9,000)

2-298 (2-400)

2-34,065 (2-
9,000)

134 (180)
9.8 (32) tandem
disk

142,235 (37,579)

125,083 (33,047)

2 sets- tractor-
drawn spreaders
with 2 nurse tanks

127 (170)
2 Large lagoon

pump

2-179 (2-240)
2-34,065 (2-9,000)
4-298 (4-400)

4-34,065 (49,000)

134 (180)
9.8 (32) tandem
disk

220,201 (58,177)

220,201 (58,177)

Table 3k cont.

 

Field zone 2: corn
grain

Field zone 1: corn
grain

Average hauling
rate

Labor
Pumping,
agitation, transport
and application,
hours (days)
Tillage
incorporation,
hours (days)

Cost
Agitation,
Pumping,
transport,
application; $/h
¢/L (¢/gal)
$/ha ($/acre)
Tillage
incorporation, $/h
¢/L (¢/gal)
S/ha ($/acre)

Total cost ¢/L
(It/gal)

Nutrient credit,
¢/L (If/gal)

Net return

over

hauﬂng

costs, ¢/L (It/gal)

Injection Application

Equipment
Spreader Tractor,
pto-kW (pto-hp)
Spreader tank, L
(gal)
Truck , pto-kW
(pto-hp)
Nurse tank, L
(gal)

Results

188,562 (49,818)
154,947 (40,937)

237,864 (62,844)

224 (22.4)

136 (13.6)

787.71
0.35 (1.31)
29.88 (73.84)
99.71

0.03 (0.11)
2.72 (6.71)

0.38 (1.42)

-0.64 (-2.41)

0.26 (0.99)

194(260)
4-34,065 (4-
9,000)

Hauling rate, L/hr (gal/hr)

Field zone 4: corn
silage
Field zone 3: corn
silage
Field zone 2: corn
grain

264,404 (69,856)
235,328 (62,174)

162,094 (42,825)

102,83 1(27,168)

92,231 (24,368)
124,759 (32,961)

407 (34.6)

136(13.6)

586.36
0.40 (1.51)
36.24 (89.55)
99.71
0.03 (0.11)
2.72 (6.71)
0.43 (1.62)

-0.64 (-2.41)

0.21 (0.79)

2194 (2260)
234,065 (2
9,000)
2298 (2400)
234,065 (2
9,000)

134,296 (35,481)
121,516 (32,105)

100,843 (26,643)

59

87,782 (23,192)
74,934 (19,798)

107,508 (28,404)

482 (41.3)

136 (13.6)

578.57
0.47 (1.79)
42.08 (103.97)
99.71
0.03 (0.11)
2.72 (6.71)
0.50 (1.90)

-0.64 (-241)

0.l4(0.51)

2194 (2260)
234,065 (2
9,000)
2298 (2400)
2-34,065 (2
9,000)

121,516 (32,105)
109,952 (29,049)

86,484 (22,849)

175,563 (46,384)
149,868 (39,595)

191,459 (50,584)

255 (12.7)

136 (13.6)

796.76
0.39 (1.49)
34.23 (84.59)
99.71
0.03 (0.11)
2.72 (6.71)
0.42 (1.49)

-0.64 (-241)

0.22 (0.81)

2194 (2260)
234,065 (29,000)
4-298 (2400)

4-34,065 (2-9,000)

186,028 (49,149)
186,028 (49,149)

172,969 (45,698)

Table 3k cont.

 

Field zone 1: corn

grain
Average hauling
rate

Labor
Pumping,

agitation, transport
and application,
hours (days)

Cost
Agitation,
pumping,
transport,
application; $/h
¢/L (¢/gal)
$/ha ($/acre)
Nutrient credit,
¢/L (¢/gal)
Net return
over
hauling
costs, ¢/L (¢/gal)

137,139 (36,232)

199,741 (52,772)

261 (26.1)

Farm 1

809.83
0.55 (1.54)
35.60 (87.98)

-0.70 (-2.65)

0.15(1.ll)

90,737 (23,973)

1 1 1,848 (29,550)

435 (36.3)

Farm 2

605.43
0.46 (1.73)
38.70 (95.59)

-0.70 (-2.65)

0.24(0.92)

74,170 (19,596)

98,031 (25,900)

508 (42.8)

Farm 3

597.75
0.52 (1.79)
45.48 (112.38)

-070 (-2.65)

0.12 (0.69)

148,341 (39,192)

173,341 (45,797)

273 (27.3)

Farm 4

808.91
0.42 (1.60)
37.11 (91.69)
-070 (-2.65)

0.28 (1.05)

 

3.7.4. Broadcast application for four 1,400-cow dairies

Selecting two tractor-drawn spreaders for transporting less than 4.8 km (3 mi) and

a tractor-drawn spreader with 2 nurse tanks for distances greater than 4.8 km (3 mi; farm

2) increased labor hours by 82% compared to farm 1 (4 tractor-drawn spreaders). Forty

percent of the crop area for Farm 2 was within a 4.8 km (3 mi) hauling distance which

resulted in an average hauling rate of 64,609 L/hr (40,155 gal/hr) to complete manure

agitation, pumping, transport and spreading within 122 hours. The remaining 60% of the

crop area was located at hauling distances greater than 4.8 km (3 mi) which decreased the

average hauling rate to 43,070 L/hr (25,768 gal/hr) and required 285 hours to complete

manure agitation, pumping, transport and spreading. Farm 2 had lower hourly costs

compared to Farm 1, but the increased hours required for further hauling distances

resulted in 15% higher ¢/L (¢/gal) costs.

60

When transporting manure greater than 4.8 km (3 mi), systems using nurse trucks
are generally more efﬁcient than tractor drawn spreader tanks of the same volume (ﬁg
3b). Sixty percent of the acreage on farms 1-4required hauling distances greater than 4.8
km (3 mi). Farm 2 and 3 used equivalent machinery sets to transport and spread manure.
Increasing the hauling distances to all four ﬁeld zones by 50% resulted in farm 3. The
increased hauling distances using equivalent equipment increased hourly labor
requirements by 18.2 % and manure pumping, agitation, transport and spreading costs by
18.6% for broadcast application.

Hauling distances for farm 4 were equivalent to those used by farm 3. Farm 4
used two tractor-drawn spreaders, each with two nurse tanks in parallel, rather than a split
system at 4.8 km (3 mi) used by farm 3. Labor requirements for farm 4 were 255 hours,
46% lower than farm 3. Using two sets of tractor-drawn spreaders with two nurse tanks
in parallel allows for two tractors to spread manure, while four nurse trucks are
transporting manure over-the-road to the tractor in the ﬁeld with limited to no idle time.
Farm 3 used 2 tractor-drawn spreaders for manure application within 4.8 km (3 mi) and
only 1 tractor-drawn spreader with 2 nurse trucks for over-the-road transport for hauling
distances greater than 4.8 km (3 mi) which increases potential idle time since 60% of the
acreage available for manure application is at hauling distances greater than 4.8 km (3
mi).

Hauling distance has an impact on the timeliness of manure agitation, pumping,
transport and Spreading based on the machinery set chosen. Compared to Farm 1, labor
for Farm 4 increased by 14% but hauling costs decreased 20% with the truck-drawn nurse

tank-based systems even though the average transport distance increased by 50%. This

61

illustrates the advantage of nurse trucks for over-the-road transport compared to tractor-
drawn spreaders when the hauling distance exceeds 4.8 km (3 mi) as was shown earlier in
Figure 2.

Crediting manure nutrient value allows the farmer to reduce commercial fertilizer
purchase and reduce total manure hauling system cost. Fertilizer recommendations are
based on crop nutrient removal requirements for crop yield as shown in Table 3g. The
nutrient value of manure was calculated based on the soil test restuls as outlined by the
Tri-state fertilizer recommendations as shown previously in Table 3e. If crop soil test
results indicate nutrient deﬁcit soil (“build-up zone”), additional nutrients above crop
removal needs can be applied up to nitrogen removal rates for the crop. Crop soil test
results in the “maintenance zone” indicate that adequate nutrients are available in the soil
for the crop to obtain an optimal yield. At a “maintenance zone” soil test result level, the
farmer must apply manure nutrients at the P205 removal needs of the crop. Over-
application of the crop nutrient needs will lead to increased nutrients in the soil not used
by the crop, and require a nutrient “draw-down” as more nutrients are applied over time.

Soil test results for farm 1-4 were assumed to be in the “maintenance” zone as shown in

table 3 g. Therefore, manure application rates must not exceed the P205 nutrient removal

for the crop. MANURE$HAUL does not limit the manure application rate based on crop

nutrient removal rates and soil test results. Rather, MANURE$HAUL assigns a N,
P205, and K20 nutrient value to manure up to the crop P205 nutrient removal for soil test

results in the “maintenance zone” (Table 3e). Table 12 lists the fertilizer
recommendations for farm 1-4 based on nutrient removal for corn grain and corn silage.

Manure production for farm 1-4 was 46,216,155 L (12,210,345 gal) of manure with 6.24

62

kg/1000 L (52 lbs/1000 gal) ofN, 3.01 kg/1000L (25 lbs/1000 gal) P205, and 3.97

kg/1000L (33 lbs/ 1000 gal) K20. Manure nutrients applied are listed in Table 31 and

compared to the fertilizer recommendations for corn grain and corn silage. N removal

needs were not met with the manure application rate; therefore supplement commercial N
fertilizer will be needed for optimal crop yields. P205and K20 applied through manure
application exceeded the crop removal needs. Since soil test results were in the

“maintenance zone” a nutrient value of manure was assigned to N, P205, and K20 up to

the crop P205nutrient removal as listed in Table 31. This resulted in a nutrient value of

manure of 0.63 ¢/L (2.41 ¢/gal) for broadcast incorporation within one day of manure
application. Injection application reduced N losses to increase manure nutrients applied
to 105 kg/ha (94 lbs/ac). Injection incorporation resulted in 0.70 ¢/L (2.65 ¢/gal) nutrient

value of manure.

63

Table 31. Fertilizer recommendations, manure nutrients applied, and nutrient
credit values for broadcast application

 

 

 

 

 

Fertilizer recommendationI Nutrient Credits3
Corn grain Corn silage Manure Corn grain Corn silage
Nutrients
applied2
N, kg/ha 131 (117) 158 (141) 80 (72) 80 (72) 80 (72)
(lbs/acre)
P205, kg/ha 54 (48) 56 (50) 122 (109) 54 (48) 56 (50)
(lbs/acre)
K20 kg/ha 39 (35) 154 (120) 196 (175) 39 (35) 154 (120)
(lbs/acre)
1 . .
EstImated In Table 3g

2 . .
Manure nutrrent content presented In Table 3h.

3Nutrient credits are calculated up to crop removal for P205 and K20 since soil test results are in
the maintenance zone as shown in Table 3g.
3.7.5 Subsurface injection application results for four 1,400-cow dairy farms
Manure injection increased labor requirements for farms 1-4 (Table 3k). Using
four sets of tractor-drawn spreaders for subsurface injection application (farm 1) with an
average hauling distance of 4 km (2.5 mi) increased labor requirements by 17%
compared to broadcast application with an 18% increase in manure agitation, pumping,
transport and spreading costs. However, injection application resulted in a nutrient value
of manure of 0.70 ¢lL (2.65 ¢/gal) compared to 0.63 ¢/L (2.41 ¢/gal) for broadcast
application. Applying the nutrient value of manure resulted in manure agitation,
pumping, transport and spreading costs 12% higher for subsurface injection compared to
broadcast application for farm 1. Manure hauling costs decreased by 6% when a nutrient

value for manure was assigned for farm 1.

64

Manure agitation, pumping, transport, and spreading cost was 0.46 ¢/L (1.73
¢/gal) and 0.52 ¢/L (1.96 ¢/gal) for farm 2 and 3 using two tractor-drawn spreaders for
hauling distances less than 4.8 km (3 mi) and two truck-drawn nurse tanks for over-the-
road transport to a tractor-drawn spreader in the ﬁeld using injection application for
hauling distances greater than 4.8 km (3 mi). Crediting the nutrient value of manure
resulted in a net return over manure agitation, pumping, transport, and application costs
of 0.24 ¢/L (0.92 ¢/gal) for farm 1 and 0.12 ¢/L (0.69 ¢/gal) for farm 2, which was 16 and
36% higher than that recognized for broadcast application, respectively .

Farm 4 using two sets of 2 truck-drawn nurse tanks for over-the-road transport to
a tractor-drawn spreader in the ﬁeld in parallel had the highest net return over hauling
costs of 0.28 ¢/L (1.05 ¢/gal) for injection application with the second lowest labor
requirement (25.5 days) across the four 1,400-cow dairy farms.

The return over hauling cost was positive for farms 1-4 indicating a cost savings
when considering the fertilizer value of manure for both injection and broadcast
application. Injection application allows farmers to decrease nutrient losses, which are
valued at fertilizer prices. However, injection application increases the labor
requirements needed for manure agitation, pumping, transport and spreading. Farmers
must evaluate the trade-off for increased labor requirements and costs for injection
application compared to broadcast application. If manure application will cause delays in
crop tillage and planting, a farmer may want to evaluate the use of a different manure

hauling system to better accommodate the needs of their farm.

65

3.7.6 Return on fertilizer value of manure

The fertilizer value of manure depends on the manure application method used
(subsurface injection, broadcast with incorporation, and broadcast) as well as the soil test
results. Using the results from the 1,400-cow dairy as an example, manure is valued at
1.3 ¢/L (4.8 ¢/gal) based on manure analysis. If the farmer uses subsurface injection and
soil test results are in the “build-up zone” the farmer will assign a nutrient value
equivalent to the manure analysis results (Table 3m). Using broadcast incorporation
resulted in 70% retention of nitrogen resulting in a 4% loss of manure nutrient value
compared to subsurface injection whereas failure to incorporate manure resulted in a 15%
loss in nutrient value compared to subsurface injection for manure applied on soil in the
“build-up zone”.

Nutrient value of manure is lower for soil test results in the maintenance zone

since soil tests results are at a level for optimal crop yield. Therefore only nutrients

applied up to P205 crop removal needs are used to calculate the nutrient value of manure.

Applying manure at levels greater than P205 crop removal needs causes nutrient build-

up which may lead to the farmer limiting manure application at later dates to “draw-
down” soil nutrient levels. Broadcast with incorporation and broadcast application
resulted in a nutrient values of manure 9% and 18% lower than subsurface injection,
respectively.

If soil test results are in the “draw-down zone” the nutrient value of manure is zero.
Applying manure on soil in the “draw-down zone” adds additional nutrients above
recommended levels which further exacerbates the nutrient problem in the soil.

Assigning a nutrient value of zero to manure demonstrates that for a farmer to fully

66

recognize the nutrient value of manure, it must be applied following recommended

fertilizer guidelines.

Table 3m. Nutrient value of manure for injection, broadcast with incorporation,
and broadcast application as a function of soil test results

 

Soil test result zone

 

 

 

Application Method N-Nutrient Build- Maintenance Draw-down
retention up
M (#890
Injection 100% 1.27 0.70 0.0
(4.8) (2.65) (0.0)

Broadcast with 70% 1.22 0.64 0.0
incorporation (4.6) (2.41) (0.0)
Broadcast 10% 1.08 0.51 0.0

(4.1) (1.94) @0)
3.8 Model Validation

The ownership and operating costs calculated by MANURE$HAUL were
compared with costs reported by two Michigan livestock producers. Each farm used a
tractor-drawn spreader tank. One cooperator was a swine producer handling about 22.7
million L (6 million gallons) of manure per year. The other cooperator was a crop
producer who hauled approximately 11.4 million L (3 million gal) per year from a nearby
dairy. Each of the livestock managers had current records of costs and labor
requirements for their manure hauling operations.

3.8.1 Swine producer

The swine producer raised 9,600 ﬁnishing pigs to 136 kg (300 lbs) each year.

The volume of manure hauled in 2008, 22,839,670 L (6,034,259 gal) was within one

percent of that calculated by MANURE$HAUL (23,740,125 L (6,272,160 gal)). The

67

swine manure was stored at two locations and an average hauling distance from each
storage pit was one mi. Manure was transported and applied with a 205 pto-kW (275
pto-pto-hp) tractor with a 37,850 L (10,000 gal) tank and injected with a 6-point injector.
The fuel price in 2008 was $0.85/L ($3.20/ga1). I

MANURE$HAUL calculated a hauling cost of 0.29 ¢/L (1.12 ¢/gal) for agitation,
pumping, transport and land application using the default values for depreciation (IO-yr,
straight line) and repair and maintenance costs based on accumulated use. The producer
calculated his cost as 0.32 ¢/L (1.23 ¢/gal). Annual repair and maintenance cost was 15%
of the equipment purchase price, and he used a 5-yr rather than a 10-yr depreciation
schedule. Manure pumping and agitating time was estimated by the producer to be 25%
of the total hauling time and was valued at $17/h of total hauling time (based on $68/h of
continuous use). Labor was valued at $15/h.

A change in the depreciation schedule from 10 to 5 years increased the
MANURE$HAUL calculated hauling cost by 16%. Increasing the labor wage rate from
$12 to $15/h increased the calculated hauling costs by 1%. Estimating the annual repair
and maintenance costs as 15% of the list price rather than basing the cost on accumulated
use lowered the MANURE$HAUL calculated cost by 3%. Decreasing the hourly charge
on the agitation and pump tractor decreased hauling costs by 1.8%. When the default and
calculated values for depreciation, repair and maintenance, labor wage rate and
pump/agitation in MANURE$HAUL were aligned with those of the swine producer,
MANURE$HAUL calculated a cost for pumping, agitation, transport and land
application of 0.31 ¢/L (1.20 ¢/gal) which was within 1% of the cost reported by the

swine producer 0.32 ¢/L (1.23 ¢/gal)..

68

3.8.2 Cash crop producer

The cash-crop farmer had an agreement with a neighboring dairy farmer to take
1 1.4 million L (3 million gal) of manure as a soil amendment and source of crop
nutrients. The dairy farmer provided the agitation and pumping and the crop farmer
provided a 179 pto-kW (240 pto-pto-hp) tractor and 26,495 L (7,000 gal) tractor-drawn
spreader for broadcast application with tillage incorporation. The average hauling
distance was 4.8 km (3 mi). Costs for tillage incorporation of the manure were allocated
to the cropping program and were not included in the calculation of hourly costs. Fuel
costs in 2008 were $0.86/L ($3.25/ga1). Labor was valued at $20/h. When the labor and
fuel costs reported by the crop producer were used with MANURE$HAUL the
calculated hauling cost was $156/h. The crop farmer’s reported hourly cost for transport
and broadcast application was $155/hr.

When the standard default values in MANURE$HAUL were adjusted to reﬂect
the speciﬁc parameters reported by two Michigan livestock and crop producers,
MANURE$HAUL cost estimates were within 1% of those reported by the cooperating
producers. MANURE$HAUL calculated costs were within 9% of the reported costs
when using the standard default values. MANURE$HAUL is a decision support tool
suitable for estimating costs for a speciﬁc farm or comparing alternative manure transport
and land application methods across a range of transport distances.

3.9 Hauling cost model

Manure managers may ﬁnd it useful to estimate manure hauling costs in

aggregate form as a base cost plus a mileage differential. Leibold and Olsen (2007)

reported a base cost of 0.26 ¢/L (1 ¢/gal) plus a mileage differential of 0.26 ¢/L-km

69

.(O.1¢/gal-mi). MANURE$HAUL was used to estimate costs for agitation, pumping,
transport and land application for six machinery sets using top-loading tank spreaders: (1)
standard tractor-drawn spreader, (2) high-speed tractor-drawn spreader, (3) truck-
mounted tank Spreader, (4) truck-drawn tank spreader, (5) two nurse tanks for over-the
road transport to a tractor-drawn spreader in the ﬁeld with the nurse tank volume equal to
spreader tank volume, and (6) two nurse tanks for over-the road transport to a tractor-
drawn spreader in the ﬁeld with the nurse tank volume two times the spreader tank
volume. Cost estimates for subsurface injection were included with tractor-drawn tank
spreaders, and estimates for tillage incorporation with a tractor and tandem disk were
included with surface broadcast systems.

The hauling cost (¢/L, ¢/gal) was calculated for each machinery set such that the
manure was applied within 175 hours with transport distances ranging from 0.16 to 16
km (0.1 to 10 mi). A multiple linear regression of the calculated costs was completed
with cost as the dependant variable and transport distance and tank volume as the
independent variables. A linear relationship among variables was achieved by regressing
distance and volume on the reciprocal of the square root of the calculated cost. The
proposed equation is a composite of machinery system-speciﬁc parameters and

coefﬁcients:
2 -l
C = ((A+BD+EV) ) (3)
Where:
C is cost for manure agitation, pumping, transport or application; ¢/L (¢/gal)

A, B and E are dimensionless machinery system-speciﬁc regression coefﬁcients

(Table 3n)

70

D is transport distance; 0.16 to 16 km (0.1 to 10 mi)

V is spreader tank volume; 9464 to 37850 L (2500-10 000 gal).

3.9.1 Hauling cost coefﬁcients

Representative machinery and labor costs were used to model the cost of manure
agitation, pumping, transport and land application for tank spreader systems as a function
of spreader tank volume and transport distance. Simulated hauling rates were ﬁt to a
general model to develop machinery system-speciﬁc coefﬁcients to predict costs for
tractor-drawn and truck-drawn spreader tanks, and hauling systems using truck-drawn
nurse tanks for over-the-road transport to tractor-drawn spreader tanks for ﬁeld
spreading. The machinery-system speciﬁc coefﬁcients are presented in a reference table
and can be used to estimate liquid manure hauling costs over a range of tank volume and
travel distance (Table 3n).

The proposed equation and machinery-speciﬁc coefﬁcients is a reliable predictor
of the calculated costs. The correlation coefﬁcients (R2) of the predicted costs ranged
from 96.8 to 99.7% (Table 3n). Costs for two tractor-drawn (22710 L, 6000 gal) top-
loading spreader tanks were calculated for transport distances ranging from 0.16 to 6.4
km (0.1to 4 mi). Calculated costs for a standard tractor with slurry injection including
agitation and pumping ranged from 0.26 to 0.63 ¢/L (0.98 to 2.39 ¢/gal) over the 6.4 km
(4 mi) hauling distance (Fig. 3d). Predicted costs ranged from 0.27 to 0.67 ¢/L (1.02 to
2.53 ¢/gal). Lower costs were calculated for a high-speed tractor with broadcast
application. In each case the predicted costs provided a close approximation of the

calculated costs.

71

The proposed model with machinery-speciﬁc coefﬁcients provides a convenient
option for comparing alternative hauling systems. Representative costs can be predicted
for commonly used spreader tank systems with tank volumes ranging from 9,460 to
37,850 L (2,500 to 10,000 gal) and hauling distances of 0.16 to 16 km (0.1 to 10 mi).
Scarborough et al. (1978) evaluated spreader tank costs and reported that selection of a
larger than optimal spreader tank was generally more economical than a smaller than
optimal tank, and optimal tank volume increased as transport distance increased. The

proposed model will facilitate such detailed comparisons and Optimization procedures.

 

    

 

 

 

0.1 0.5 1 2 3 4
0 8 L l I l l l 3
Transport distance, miles
0.7 ~ . . .
H Std. trac.. Inject, predrcted __ 15
:—-:. Std. trac, inject.
0.6 H Hi-speed trac., brdcst.
+—+ Hi-speed trac., brdcst, predicted _ 2
I; ' ' i ' _ I 5
a- '5
55 E 1.5 g
o =
U ('3
— 1
l" 0.5
0.1 -
0 I I I I T I 0
0.16 0.8 1.6 3.2 4.8 6.4

Transport distance, kilometers

Figure 3c. Calculated manure agitation, pumping, transport and spreading versus
predicted costs using machinery speciﬁc coefﬁcients

72

Table 3n. Liquid manure machinery system-speciﬁc regression coefﬁcients for

estimating cost for agitation, pumping, transport and land application.

 

Cost coefﬁcients

 

 

Machinery Set A B E R2
Tractor-drawn tank
spreaders
Standard tractor, A,P,T, S
Broadcast, ¢/L 2.20 -0.138 0.000002 99.2
¢/gal (1.13) (-0.1 14) (0.000004) (99.2)
Injection, ¢/L 0.964 -0.0927 0.000006 98.7
¢/gal (1.87) (-0.1 12) (0.000003) (98.7)
Standard tractor, T,S
Broadcast, ¢/L 2.62 -0. 162 -0.000004 99. 1
¢/gal (1.35) (-0.134) (-0.000008) (99.1)
Injection, ¢/L 1.1 1 -0.106 -0.000001 98.8
¢/ga1 (2.16) (-0. 128) (-0.000007) (98.8)
Standard tractor, T,S, I
Broadcast, ¢/L 2.28 -0.123 -0.000003 99.3
¢/gal (1.17) (-0. 101 ) -0.000006 (99.3)
Standard tractor, A,P
Broadcast, ¢/L 3.73 -0.242 0.000052 98.6
¢/ gal (1.92) (-0.200) (0.000101) (98.6)
Injection, ¢/L 1.82 -0.179 0.000091 98.2
¢/gal (3.54) (-0.217) (0.000047) (98.2)
High-speed tractor, A,P,T,S
Broadcast, ¢/L 1.16 -0.104 0.000002 99.1
¢/gal (2.20) (-0.138) (0.000002) 99.1
Injection, ¢/L 0.973 -0.0801 0.000004 97.8
93/ gal (1.89) (-0.097) (0.000002) (97.8)
Hi gh-speed tractor, T,S
Broadcast, ¢/L 1 .38 -0.122 -0.000010 99.0
¢/ gal (2.62) (-0.162) (-0.000004) (99.0)
Injection, ¢/L 1.12 -0.0915 -0.000004 98.2
¢/gal (2.18) (-0.1 1 1) (-0.000002) (98.2)
Hi gh-speed tractor, T,S,l
Broadcast, ¢/L 1 .19 -0.0901 -0.000006 99.1
¢/gal (2.27) (-0.123) (-0.000008) (99.1)
High-speed tractor, A,P
Broadcast, ¢/L 1.94 -0.175 0.000103 98.9
¢/gal (3.74) (-0.242) (0.00052) (98.9)
Injection, ¢/L 1.82 -0.152 0.000089 97.8
¢/ gal (3 .54) (-0.184) (0.000046) (97.8)

73

Table 3n continued

 

Machinery Set
Tractor-drawn/nurse tank
systems
Spreader = 1X nurse tank,
A,P,T,S
Broadcast, ¢/L
¢/gal
Injection, ¢/L
¢/gal
Spreader = 1X nurse tank, T,S
Broadcast, ¢/L
¢/gal
Injection, ¢/L
¢/gal
Spreader = 1X nurse tank,
T,S,I
Broadcast, ¢/L
¢/ga1
Spreader = 1X nurse tank, A,P
Broadcast, ¢/L
¢/gal
Injection, ¢/L
¢/gal

Nurse tank = 2X spreader,
A,P,T,S
Broadcast, ¢/L
¢/gal
Injection, ¢/L
¢/gal
Nurse tank = 2X spreader, T,S
Broadcast, ¢/L
¢/gal
Injection, ¢/L
¢/gal
Nurse tank = 2X spreader, T,S,
1
Broadcast, ¢/L
¢/gal
Nurse tank = 2X spreader, A,P
Broadcast, ¢/L
¢/gal
Injection, ¢/L
¢/gal

0.665
(1.293)
0.620
(1.208)

0.727
(1.418)
0.677
(1.318)

0.691
(1.346)

1.61
(3.14)
1.56
(3.03)

0.666
(1.300)
0.621
(1.211)

0.700
(1.362)
0.650
(1.270)

0.669
(1.300)

2.07
(4.037)
2.14
(4.16)

74

-0.0312
(0.0375)
-0.0279
(0.0339)

-0.0351
(-0.0426)
-0.0316
(0.0381)

-0.0307
(0.037)

-0.0577
(0.070)
-0.0534
(-0.064)

-0.0259
(0.0314)
-0.0228
(-0.0281)

-0.0276
(0.0335)
0.0243
(0.0294)

-0.0241
(0.0291)

-0.0592
(0.0712)
-0.0633
(0.0761)

0.000025
(0.000013)
0.000025
(0.000013)

0.000027
(0.000014)
0.000027
(0.000014)

0.000023
(0.000012)

0.000062
(0.000032)
0.000064
(0.000033)

0.000046
(0.000023)
0.000046
(0.000024)

0.000048
(0.000025)
0.000048
(0.000024)

0.000042
(0.00022)

0.000161
(0.000083)
0.000149
(0.000076)

99.3
(99.3)
99.5
(99.5)

99.2
(99.2)
99.5
(99.5)

99.6
(99.6)

99.5
(99.5)
99.5
(99.5)

99.6
(99.6)
99.7
(99.7)

99.4
(99.4)
99.7
(99.7)

99.7
(99.7)

97.1
(97.1)
98.8
(98.7)

Table 3n continued

 

Machinery Set A B E R2
Truck-mounted/drawn
systems
Truck-mounted, A,P,T,S
Broadcast, ¢/L 0.878 -0.0513 0.000053 97.3
¢/gal (1.710) (-0.0623) (0.000027) (97.3)
Truck-mounted, T,S
Broadcast, ¢/L 1 .01 -0.0613 0.00006 96.9
¢/gal (1.97) (-0.0741) (0.000031) (96.9)
Truck-mounted, T,S,I
Broadcast, ¢/L 0.923 -0.0467 0.000045 98.0
¢/ gal (1.797) (-0.0567) (0.00023) (98.0)
Truck-mounted, A,P
Broadcast, ¢/L 1.73 -0.0904 0.0001 15 96.8
¢/gal (3.363) (-0.109) (0.00060) (96.8)
Semi-tractor drawn, A,P,T,S
Broadcast, ¢/L 0.987 -0.0524 0.000034 97.6
¢/gal (1 .92) (-0.063) (0.000018) (97.6)
Semi-tractor drawn, T,S,
Broadcast, ¢/L 1.15 -0.0634 0.000038 97.4
¢/gal (2.23) (-0.0768) (0.00002) (97.4)
Semi-tractor drawn, T,S,I
Broadcast, ¢/L 1.02 -0.0458 0.000028 98.5
¢/gal (1.98) (-0.0554) (0.000014) (98.5)
Semi-tractor drawn, A,P
Broadcast, ¢/L 1.92 -0.0879 0.000079 97.8
¢/gal (3.73) (-0.1063) (0.000041) (97.8)

 

3.10 Conclusion

MANURE$HAUL provides a ﬂexible decision tool for comparing cost-effective
alternative manure hauling systems. MANURE$HAUL provides an accurate estimate of
time needed for manure pumping, transport, and land application as a function of hauling
distance, spreader capacity, manure equipment cost, labor, and nutrient value of manure.
MANURE$HAUL was used to estimate costs for agitation, pumping, transport and land

application for two machinery sets using top-loading tank spreaders: (1) tractor-drawn

75

spreader and (2) two nurse tanks for over-the road transport to a tractor—drawn Spreader in

the ﬁeld with the nurse tank volume equal to spreader tank volume for a 175-, 350-, 700-,

and 1,400-cow dairy. Manure hauling cost estimates using MANURE$HAUL for the

representative dairy farms illustrated the following:

Ownership and operating costs for manure agitation, pumping, transport and land
application ranged from $166/hr for a tractor-drawn spreader with broadcast
application on a 175-cow dairy to more than $586/hr with the 1,400-cow dairy
using two large tractor-drawn spreaders and two nurse trucks.

Manure tillage costs ranged from $99/hr for the 1,400-cow dairy to $134/hr for
the 350-cow dairy. Manure tillage costs were dependent on the acreage used for
manure application and equipment size.

Ownership and operating costs for manure agitation, pumping, transport and land
application ranged from 0.40-0.52¢/L (1.54-1.96 ¢/gal) for injection application,
which was 7-18% higher than broadcast application.

Using two sets of two nurse tanks for over-the road transport to a tractor-drawn
spreader in the ﬁeld with an average hauling distance of 6.4 km (4 mi) for the
1,400-cow dairy resulted in a 14% increase in hauling time compared to using 4
tractor-drawn spreaders with an average hauling distance of 4.8 km (3 mi).
Labor requirements for subsurface injection increased by 14%, 18.5%, 19.3%,
and 6.8% for the 175-, 350-, 700-, and 1,400-cow herds, respectively, compared
to a broadcast application

The nutrient value of manure using subsurface injection on soil in the

“maintenance zone” was 0.70 (UL (2.65 ¢/gal) compared to 0.63¢/L (2.41 ¢/gal)

76

for broadcast application with incorporation and 0.51¢/L (1.94 ¢/gal) for
broadcast application.

0 The manure hauling cost was most sensitive to tank spreader capacity and manure
hauling distance. Increasing hauling distance by 50% for a 1,400-cow dairy using
two 34,065 L (9,000 gal) tractor drawn spreaders for hauling distances less than
4.8 km (3 mi) and two 34,065 L (9,000 gal) truck-drawn nurse tanks for hauling
distances greater than 4.8 km (3 mi) increased labor requirements by 18.2% and
manure pumping, agitation, transport and spreading costs by 18.6% for broadcast
application.

0 Machinery speciﬁc coefﬁcients were estimated for commonly used tank spreader
systems as a function of tank volume and transport distance. Machinery speciﬁc

coefﬁcients are presented in a convenient reference table.

77

REFERENCES

ASAE Standards, 50th ed. 2003. D497.4: Agricultural Machinery Management Data. St.
Joseph, Mich.: ASAE.

ASAE Standards, 53rd ed. 2006. EP496.3: Agricultural Machinery Management Data. St.
Joseph, Mich.: ASAE.

Black, J .B., 2008. Working Paper. Tart Cherry Production.

Borton, L.R., C.A. Rotz, H.L. Person, T.M. Harrigan, and W.G. Bickert. 1995.
Simulation to Evaluate Dairy Manure Systems. Applied Eng. Agric. 11(2):301-
310.

Case-1H. “Machine Builder.” Accessed October 2008 and February 2009:
http://caseih.cnhbuilder.com

Harrigan, T.M. 1997. Manure hauling rate of spreader tank systems. Applied Eng. A gric.
13(4):465-472.

Harrigan, T.M. 2001. Manure Transport Rates and Land Application Costs for Tank
Spreader Systems. Michigan State University Extension Bulletin E-2767. East
Lansing, MI.

Harrigan, T.M, W.G. Bickert, and CA. Rotz. 1996. Simulation of dairy manure

management and cropping systems. ASAE Paper No. 95-4509. St. Joseph, MI:
ASAE.

Harrigan, T.M. 2009. Liquid Manure Hauling Capacity of Custom Applicators using
Tank Spreader Systems in the Great Lakes Region. In Submission.

Iowa State University. 2005. Machinery Management—Estimating Farm Machinery
Costs. Iowa State University Extension Bulletin PM 710. Ames, Iowa.

IRON Solutions. 2006. Farm Equipment Ofﬁcial Guide. Fenton, Missouri: IRON
Solutions Equipment Market Intelligence.

Jacobs, L.W. 1995a. Manure Management—Utilization of Animal Manure for Crop
Production. Part 1. Management of Manure Nutrients and Water Quality.
Michigan State University Extension Bulletin MM-l. East Lansing, MI.

Jacbos, L.W. 1995b. Manure Management—Utilization of Animal Manure for Crop

Production. Part II. Manure Application to Cropland. Michigan State University
Extension Bulletin MM-2. East Lansing, MI.

78

 

John Deere. “Build Your Own John Deere.” Accessed October 2008 and February 2009:
http://conﬁguraiordeere.com

Koehler, B., B. Lazarus, and W. Meland. 2009. “What’s Manure Worth” Spreadsheet.
University of Minnesota, St. Paul, Minnesota. Available at:

http://wwwapecumn.edu/faculty/wlazarus/interests_manureworthhtml Accessed
on 4 May 2009.

Koelsch, R., G. Erickson, and R. Massey. 2007. “Feed Nutrient Planning Economics
(FNMP$).” University of Nebraska-Lincoln, Lincoln, Nebraska. Available at:
http://manure.unl.edu/computer.html Accessed on 9 February 2009.

Leibold, K., and T. Olsen. 2007. “Value of Manure Nutrients” Spreadsheet. Iowa State
University, Ames, Iowa. Available at:
http://www.extension.iastateedu/AgDM/livestock/html/b1-65.html Accessed on
9 February 2009.

Michigan Department of Agriculture. 2008. Generally accepted agricultural and
management practices for manure management and utilization. Michigan
Commission of A griculture.

MWPS. 1993. Livestock Waste Facilities Handbook. 2nd ed. Ames, Iowa: Midwest
Plan Service.

Midwest Plan Services. “Manure Characteristics.” Manure Management Systems Series.
MWPS-18, Sect 1, 2nd Edition. 2001.

Rotz, C.A. M.C. Corson, D.C. Chianese, and C.U. Coiner. 2008. “The Integrated Farm
System Model—Reference Manual 3.0.” Pasture Systems and Watershed
Management Research Unit. Agricultural Research Service. United States
Department of Agriculture. Available at:
http://ars.usda.gov/SP2UserFileS/Place/ 1 9020000/ifsmreference._pdf Accessed on
9 February 2009.

 

Rotz, C.A., and T.M. Harrigan. 2005. Predicting Suitable Days For Field Machinery
Operations in a Whole Farm Simulation. Applied Eng. Agric. 21(4): 563-517.

Vitosh, ML, J.W. Johnson, and DB. Mengel. 1995 “Tri-State Fertilizer
Recommendations for Corn, Soybeans, Wheat, and Alfalfa.” The Ohio State
University Extension Bulletin: E-2567.

Wamcke, D., .l. Dahl, L. Jacobs, and C Laboski. Nutrient Recommendations for Field
Crops in Michigan. MSU Extension Bulletin E-2904, May 2004.

79

Wittenberg E., and CA. Wolf. “2004 Michigan Dairy Farm Business Analysis
Summary.” Staff Paper 2005-10, Sept. 2005.

80

CHAPTER 4: IMPLICATINS OF AN AIR EMISSIONS TAX ON LIVESTOCK
PRODUCER’S INVESTMENT IN ABATEMENT TECHNOLOGY
4.1 Introduction

Environmental regulation of livestock production continues to become more
stringent as knowledge increases about production processes generating emissions and
appropriate abatement technology required to control these emissions. In order to remain
proﬁtable agricultural producers must be cognizant of these changes and their
implications on future emission-reducing technology investment decisions.

Increasing attention on regulating greenhouse gas emissions under the Clean Air
Act from non-point pollution sources, such as the agricultural sector, has been evident
with the introduction of the Environmental Protection Agency’s 2005 Air Compliance
Agreement (Environmental Protection Agency, 2009) and potential climate legislation.
Livestock producers had an option to sign this agreement and pay a fee to EPA for
unknown past emissions that would be used to collect data on animal air emissions for
various livestock enterprises and management practices. In return for signing the
agreement, producers were not held liable for any emissions in the interim period prior to
implementation of air emission policies—perhaps in 2011 or 2012 and agreed to abide
by the decision. By declining to sign the agreement, those producers were left open to
penalties for emissions during the interim period. This agreement foreshadows the reality
that policy instruments in the form of carbon-equivalent air emission standards, taxes, or
some combination thereof will be part of agricultural producer management decisions in

the near future.

81

Air emissions from livestock production are a function of livestock species,
housing, and manure storage and application methods. Estimates of livestock air

emissions vary greatly as a function of these factors and the study area (Koelsch and

Stowell, 2009, USDA/ARS, 2008; Gay et al., 2003). For example, daily ammonia (NH 3)

air emissions estimates for a dairy cow based on a Texas study vary from 0.025-0.25
lbs/day (USDA/ARS, 2008) whereas air emissions range from 0.10-1.02 lbs/day based
on a Nebraska emissions estimate (Koelsch and Stowell, 2009). If a dairy farm is located
in an area where a study has not been completed (as is the case for many livestock
operations) the large variation in estimated air emissions makes it difﬁcult for livestock
producers to know if they are exceeding air emission limits.

Livestock producers can adopt abatement technologies to reduce animal emission
levels. These exist for housing, manure storage, and manure application practices.
Housing abatement technologies include bio-ﬁltration system and urine-feces separation.
Building long-term manure storage facilities with manure storage covers are examples of
storage abatement technologies. Many times these abatement technologies involve
irreversible investments with sunk costs. For example, installing a long-term manure
storage facility involves a large capital investment which is speciﬁc to the livestock
enterprise and farm size. There is limited ability for the farmer to sell the manure storage
facility if they decide to exit the industry or later learn that a different abatement
technology would better suit the needs of their farm based on changing emission
regulations. Therefore livestock producers must evaluate the trade-offs of various
abatement technologies while considering the uncertainty surrounding future air emission

regulations.

82

A farmer’s incentive to invest in emission-reducing technology is inﬂuenced by
the environmental regulation chosen by the governmental agency. The policy instrument
can be market based (emission taxes or tradable pollution permits) or take on the form of
a command-and-control policy (performance and technology standards). It has been
argued that taxes encourage ﬁrms to invest in more efﬁcient pollution abatement
technologies than other market-based pollution control methods (Caswell, Lichtenberg,
and Zilberman, 1990; Farzin and Kort, 2000; Millman and Prince, 1989; Tarui and
Polasky, 2005). Indeed much of the discussion regarding the Air Compliance Agreement
has focused on emission taxes as the likely policy instrument. The American Farm
Bureau Foundation estimated yearly emission taxes for livestock operations at $175/dairy
cow, $87.50/beef cow, and $21.87/hog. These emissions taxes were projected to apply to
any agricultural operation with more than 25 dairy cows, 50 beef cattle, 200 hogs, or 500
acres of corn (Dairy Herd Management, 2009). If emission taxes are chosen as the
environmental policy control instrument, these estimated animal levels indicate that
almost all farms would face some form of an emission tax. To decrease future potential
emission tax burdens, livestock producers can adopt abatement technologies. However,
the level at which they adopt these technologies is dependent on the “estimated” tax rate.

Past literature has evaluated the behavior of a ﬁrm subject to environmental
regulation. Xepapadeas (1992) used an inﬁnite planning horizon dynamic game
framework to develop incentive schemes for investment which accounted for the
dynamics of non-point source pollution problems. Xepadadeas concluded that an
increase in an emission tax always resulted in a larger stock of abatement capital for the

ﬁrm compared to an emission standard. A static incentive scheme was solved and

83

compared to the dynamic solution which suggested that static incentives schemes were
suboptimal in the long run.

Kort (1996) extended the theoretical framework of Xepapadeas (1992) assuming
that abatement technology would be required to reduce pollution. He evaluated the effect
of a pollution tax and marketable permits on ﬁrm investment in abatement technology.
Productive capital stock was the single input used in the production process, which
generated a by-product, emissions. Emissions were generated as function of two forms
of capital stock, productive and non-productive, where non-productive capital stock
cleaned pollution generated by productive capital stock. Deﬁning the emissions function
is the manner extends the work of Xepapadeas to account for the fact that it is more
difficult to reduce emissions with abatement technology when emissions are already at a
low level. Using an optimal control theory model Kort determined that an increase in a
pollution tax does not always result in a decrease of productive capital stock and increase
in non-productive capital stock as was found previously by Xepapadeas. Kort also found
that in the long-run ﬁrm investment behavior was equivalent whether a pollution tax or
marketable permit was imposed.

Hartl and Kort (1996) evaluated switching to cleaner inputs in the production
process of a ﬁrm when an emissions tax was imposed. Emissions were assumed to be
generated through a production process deﬁned by a single input, capital stock. Hartl and
Kort included an investment grant in the decision that was found to induce investment at
a relatively earlier date.

F arzin and Kort (2000) extend Hartl and Kort’s model to consider two forms of

uncertainty for the optimal investment policy of a ﬁrm facing environmental regulation:

84

(1) an increase of unknown size in the future pollution tax rate and (2) an unknown
timing of the tax increase. Secondly, F arzin and Kort assumed that emissions were a
function of the production process as deﬁned by a single variable input rather than capital
stock as speciﬁed in Hartl and Kort. This assumption allows the ﬁrm to decrease
emissions by decreasing the variable input, rather than the capital stock level. F arzin and
Kort determined that abatement investment rates were lower than the certainty case when
uncertainty existed about the magnitude of the future tax increase. Uncertainty
surrounding emission tax increase timing resulted in increased under-investment in
abatement capital.

Investment in emission reducing technologies is a dynamic and potentially
irreversible investment for many agricultural producers. Livestock producers are aware
that an environmental policy instrument is scheduled to be imposed for air emissions
December 31, 2011. However, uncertainty exists surrounding the stringency of new
environmental regulations regarding air emissions for livestock operations. Producers
must evaluate tradeoffs between investing in emission-reducing technology today versus
waiting to invest at a later date when additional information regarding new emission
reducing technologies could become available.

A study that combines the theoretical framework developed by Kort (1996) which
speciﬁed a emissions production function dependent on productive and non-productive
capital stock and uncertainty surrounding the size of an emissions tax developed by
Farzin and Kort (2000) is absent in current literature. Combining these two topics allows
for an analysis of the current situation faced by livestock producers where animals are

productive capital stock and abatement technology is non-productive capital stock.

85

Secondly, uncertainty is included with an uncertain emission tax increase imposed on
December 31, 2011. Determining how a livestock producer’s investment path changes
based on current estimates of an emission tax rate allows us to better understand how the
dairy industry will respond to different levels of emission taxes.

This analysis adapts the model of Farzin and Kort (2000) to evaluate the effect of
an emission tax policy on farm investment in emission-reducing technology. This
analysis differs from Farzin and Kort’s in three ways. First, the production process is
deﬁned as a function of productive capital rather than a variable input, since animals are a
form of capital on livestock operations. Second, in addition to a productive capital stock,
non-productive capital stock is introduced in the model to reduce emissions. Therefore a
functional form for the emissions function is deﬁned to address the interaction between
productive and non-productive capital stock as outlined by Kort (1996), rather than a
pollution function proportional to output level used by Farzin and Kort (2000). Finally,
an empirical analysis is implemented at the aggregate level which has been absent in
previous analysis. Functional forms for milk production and emissions functions, and
numeric values for price and tax parameters are deﬁned to provide a tractable analysis
which can be used in a policy context. The objective of this analysis is to determine the
optimal investment path for (1) a certain emission tax at time t=0 and (2) an uncertain
emission tax increase at time T which considers the potential uncertainty regarding
environmental regulation faced by agricultural producers.

The paper proceeds as follows. In section 4.2 an analytical model of an emission
tax that is known with certainty and does not change over time is presented for the dairy

industry which is followed by an empirical analysis of the basic tax model for a known

86

low and high tax rate in Section 4.3. Section 4.4 presents an analytical analysis of an
unknown tax rate at a known time T with the empirical analysis in Section 4.5.

Discussion, policy implications, and conclusions are presented in Section 4.6.

4.2 Basic Emission Tax Model

The basic model adapted from Farzin and Kort (2000) is an emission tax that is
known with certainty and does not change over time. The basic model is applied to the
dairy industry herd population at an aggregate level but could be directly applied to other
livestock enterprises. The results of this basic model are used as a benchmark to analyze
uncertain tax policy in later sections.

Consider a risk-neutral farmer which has the opportunity to invest in two types of
capital, productive and non—productive. The productive capital is an input (dairy cows,
(9)) used to produce a homogenous output (milk) according to a simple production
process

(1) "1:”!(9).
where m(0)=0, m'(Q)>0, and m"(Q)SO. Jointly with milk production, emissions are
generated as a function of the input level of productive capital stock, cows (Q). The
second type of capital (K) used by the farm is non-productive, but reduces emissions
generated by the productive capital stock, 0. Examples of non-productive capital
include animal housing, manure storage, and manure application methods used to
minimize emissions (Gay et al., 2003; Koelsch and Stowell, 2009). The animal

emissions function is given by

(2) A = A(Q,K),

87

where A is total emissions generated by productive and non-productive capital. The
emissions function must satisfy the following conditions (Kort, 1996):

(28) A(Q,K) > 0 for all Q > 0 and K 2 0,

(2b) AQ(Q,K) > O and AQQ(Q,K) > 0,

(2c) AK(Q,K) < 0 for all Q > 0 and AKK(Q,K) > O for all Q > 0

(2d) AQK(Q,K) = AKQ(Q, K) < 0.
Condition (2a) implies that emissions are positive as long as cows are on the farm.
Condition (2b) shows that emissions increase in a convex way with an increasing number
of cows for a given level of emission-reducing capital. Diminishing returns for emission-
reducing technologies is shown with condition (2c) which states that emission output is
smaller for larger amounts of emission-reducing technologies for a given level of cows
(Q). Condition (2d) implies that an increase in emissions due to one additional cow is

smaller for larger stocks of emission-reducing capital. Therefore, emission-reducing

technologies are more effective for reducing emissions than reducing the number of cows
on the farm. The emissions function is not separable in Q and K which implies that

increased investments in emission-reducing capital stock is required to reduce emissions
with some ﬁxed level of productive capital, Q.

Following Chavas and Klemme (1987) aggregate dairy herd population dynamics
are a function of the current level of cows in the female population, a survival rate, and a
net birth rate. The survival rate is deﬁned as the proportion of animals still in the dairy
population after one time period. The net birth rate is the difference between the birth

rate and a constant natural death rate for offspring. The dynamics of the adult female

88

population can be written as Q = aQS , where a is the net birth rate and S is the
Survival rate of cows. We assume that the birth rate and survival rate are independent of
the size of the dairy cow population. Since cows produce emissions as a by-product of
the milk production process, decreasing the number of cows through slaughter (which
changes the survival rate of the cows) decreases the amount of emissions generated.

Emission abatement capital stock can be increased by making an investment, I, in
emission-reducing technology. The total investment cost, C(1), is assumed to be a convex
increasing function of the investment level such that,

(3) (3(0) = 0.
where C '(1)>0 and C "(I)>0. It is assumed that investment in emission abatement
technology is irreversible such that I _>_ 0. Without investment, 1, emission abatement
capital stock is assumed to depreciate at a constant proportional rate of 5.

An emission tax, 2' > 0 , is deﬁned as the pollution tax per unit of emissions. The
total farm emissions tax payment at any point in time is rA(cows, K).

The management decision for the farms is to choose the survival rate, S, (or

consequently the cull rate (1-S)) and emission-reducing technology investment, I, to

maximize the present value of its cash ﬂows over an inﬁnite planning period,

(4) max j[p,,,m(r2) + pSQ(1 — d)(1 — s) — wQ — C(1) — rA(O,K)]e""dr,

5,10
s.t. Q=aQS
K=I—6K,
120.

89

where Q is the productive capital stock (cows), m(Q) is the milk production function,

pm is the market milk price, pS is the market slaughter price, d is the death loss among

cows, S is the survival rate, w is the input price of milk production (ie. feed for cows), K
is the non-productive capital stock (ie. manure storage), I is the emission-reducing

technology investment, A(Q, K) is the emissions production function, a: is the net birth

. rate, 2' is the per anima emissions tax, 5 is the depreciation rate for non-productive
capital stock, K and r is the constant discount rate.

Equation (5) Simply states that the returns at time t are equal to the revenue from
milk production plus the revenue from an animal leaving the population (ie. slaughter
value) less the input costs for milk production, investment costs for emission-reducing
technology, and the emission tax liability faced by the industry. The current value
Hamiltonian for the optimal control problem is deﬁned as,

(5)
H = pmm(Q)+ pSQ(l—d)(1—S)-— wQ —C(I)—z'A(Q,K)—/1(aQS)—77(I - 5K)

where ,1 is the shadow price of the productive capital, cows, and r] is the shadow price of

non-productive capital, emission reducing technology. The necessary conditions for the

optimal policy are,

(6) 22539112,
62
(7) n=C'(1).

(8) 2 =(r —a)2 — pmm'(Q)-— p5(1—- d)(1—S)+ w+ TAQ(Q,K)

(9) r) =(r + (5)7; + rAK(Q,K).

90

Equation (6) shows the marginal impact of cow sales on the Hamiltonian where the
shadow price of cows must equal the Slaughter price adjusted by the net birth rate and
natural death loss along the optimal slaughter path. Equation (7) shows that the shadow
price of the emission-reducing technology must equal its marginal cost along the optimal
investment path. Equation (8) is the adjoint condition which states that an additional cow
slaughtered is equal to the net revenue from that cow. Equation (9) is the adjoint
condition which states that an additional unit of investment is equal to the savings on the
emissions tax payment. The adjoint equations must hold at each point in time and can be
expressed as “golden rule” equations (typically found in resource management literature)
by taking the time derivative of the shadow price equation and setting it equal to the

adjoint condition for A and 77. Taking the time derivative of equation (6) and setting it

equal to equation (8) and solving for, r, results in,

r : pmm'tn» _ 3 _ amok) + a.

10
() xi ,1 /1

 

Equation (10) equates the return from holding dairy cows (not slaughtering) to its
opportunity cost, r. The ﬁrst and second RHS terms are the marginal revenue and cost
from keeping the dairy cow in the milking population, respectively. The third term is the
tax cost of investing in a larger dairy population at the margin. The fourth term is the
marginal impact of cows on reproduction.

Taking the time derivative of equation (7) and setting it equal to equation (9) and

solving for r results in,

 

(11) r=[C—H(D—j-6]—MK(Q’K).

C'(1) C'U)

91

Equation (11) equates the return from not investing in emission-reducing

technology to its opportunity cost, r. The ﬁrst RHS term in the bracket is the capital

garns to emISSIOn reducmg technology less deprec1at10n. The second RHS term IS the

marginal impact of taxes on investing in new emission-reducing technology.

PsIl-d)
a

From equation (6) we know that A = , which results in a singular

solution for the survival rate control variable since there are no control variables in
equation (10). Therefore, we can solve for the number of cows in the herd as a function

of capital (K), rather than S. The number of cows in the dairy herd changes as the

emission-reducing capital stock changes. We can solve for the I = O and K = 0

isoclines to analyze the phase diagram for the optimal investment path in the (K, I)-plane
rather than the (K,S)-plane. The K = 0 isocline is a positively sloped straight line where

I = OK. The isocline for I = 0 is deﬁned by solving for l in equation (1 1) such that,

(r + 6)C'(I) + rAK(Q,K)

12 i:
( ) C"(I)

 

From equation (12), a unique saddle point exists where I = K = 77 = 0 and

:1:
I = 5K such that,

(13) — rAK(O*,K*) = (r + 5)C'(5K*).

 

7 The ﬁrst RHS term in equation (1 1) can also be represented as 1 where 77 = C" (1)]
77

and 77=C'(I).

92

4.3 Numerical Example: Basic Emissions Tax Model

A numerical application of the basic emissions tax model was implemented to
analyze the numerical phase plane diagram in the (K,1)-plane rather than theoretical
diagrams completed in previous analysis (F arzin and Kort, 2000). Data used to
parameterize the model are provided in Table 4a. Prices parameters used in the model
were based on dairy industry average values. Slaughter price was valued at $650/cwt
based on a ﬁve year average for dairy cow slaughter prices from USDA-NASS (2001-
2006). Feed cost was assumed to be $4.68/cow/day for purchased, homegrown, and
grazing feed which resulted in a yearly cost of $1,709/cow (ARMS-ERS, 2005).

The American Farm Bureau estimated a potential emissions tax for dairy of

$175/cow. The emissions tax rate in the model is speciﬁed based on units of emissions.

The per animal emissions tax rate is converted to pounds of ammonia (NH 3) emissions
assuming that a dairy cow produces 0.25 pounds of ammonia emissions on a daily basis
or 91.25 pounds annually (USDA/ARS, 2009; Gay et al., 2003).8 The emission tax rate

was calculated as $1 .92/1b NH 3 emissions.

The depreciation rate for capital, 5, was assumed to be 0.05 which assumes a 10
year useful life of the emission-reducing technology investment. A discount rate of 0.09
was used (Wolf et al., 2002).

A constant returns to scale milk production function was implemented to estimate

milk production such that, m(Q) = 210Q. It was assumed that the average yearly milk

 

8 . . . . .
0.25 pounds of ammonIa emISSIons IS the most common value used to estImate
emissions and was used to adjust a per cow tax to per pound ammonia (NH3) emissions.

93

yield was 21,000 pounds per cow. Annual yield was converted to hundredweight (cwt)
basis to be in equivalent unit terms for milk price.

The milk price pm is endogenous in the model since the demand for milk is
downward sloping. An inverse demand function for milk was used to determine the price
of milk within the model as a function of the number of cows. The inverse demand
function was deﬁned assuming aggregate milk production was 190 billion pounds of milk
with average milk production per cow of 21 ,000 pounds. A CES production function was
assumed for quantity demanded with the simplifying assumption that cows are the only

input in the milk production process. The inverse demand function was,

QD

= 5978 — —- , where the uant't demanded for
pm 304,237 q ' y

milk, QD = 2100IO‘0-91‘1/09.

Dairy herd population dynamics is a function of the natural death rate, survival
rate, and net birth rate. The natural death rate, d, for dairy cows was assumed to be
4.91% (McDonald et al., 2007). The net birth rate was valued at 1.8 (McDonald et al.,

2007)

The functional form for the emissions production was speciﬁed using conditions

QO'5K_O'5

(2a)-(2d) which resulted in A(Q, K) = . The negative coefﬁcient on the

non-productive capital, K, means that as the amount of emission-reducing capital stock

increases through emission-reducing technology investment, animal emissions decrease.

94

Total investment cost was assumed to be a convex increasing function of the

investment level in the emission-reducing technology. This resulted in the following

1
function form of, C(1) = —2- 12 for investment cost.

Table 4a. Parameter description, values, and sources

 

 

Parameters Description Value Source
p Slaughter price $650/cow USDA-NASS (2001-
S
2006)
w Feed cost $1,709/yr/cow ARMS-ERS (2005)
($4.68/day/cow)
r Tax rate $1 .92/emission unit American Farm Bureau
($175/cow) (2009)
6 Depreciation rate 0.05 Assumption
for non-productive
capital
r Discount rate 0.09 Wolf et a1. 2002
a Net birth rate 1.8 Assumption
d Death rate 0.491 McDonald et a1. 2007
m(Q) Milk production 210 Q Assumption
function A
A(Q, K) Air emissions 0.5 -—O, 5 Assumption
production function 365 Q K
C(1) Investment cost 0512 Kort and Hartl (1996)

- function

 

 

Using the assumptions in Table 4a, the numerical equivalents for equation (6)-(9)

for the optimal control theory problem with a certain tax rate r = 1.92 /lb ammonia

emissions are the following,

(14) x1 = 343,

(16) 77:1.

95

 

(17) 2L=0.09/1+1091+0.150—210[5978— 210“] 359“

—---— +
304,237 91/2 K1/ 2

350.452“2
K3/2

 

(18) 77 20.1977—

Taking the time derivative of (14) and setting it equal to (17) we can solve for Q in terms

 

of emission-reducing capital since we have a singular solution for the survival rate '2"
control variable. Substituting Q in terms of K into (17) allows us to plot the isoclines for

l = 0 and K = 0 for a tax rate r = $1 .92/lb ammonia as a numerical result for a singular

solution with respect to the survival rate control variable (Figure 4a). The thin line in i

Figure 4a is the saddle path leading to the stable steady state equilibrium where I=45and
K=900. Along the saddle path investment falls as emission-reducing capital stock
increases. This is a result of the functional form for the emissions production function
including both cows (productive) and emission-reducing (non-productive) capital stock.
The phase-plane diagram shows that with low levels of aggregate of emission-reducing
capital stock (i.e., K=300) and investment level of I=100 is needed for the dairy industry
to reach the saddle path for investment and move towards the steady state equilibrium
level of investment in emission-reducing technology and non-productive capital stock.
Along the optimal investment path (saddle path) the Shadow price must always

equal the marginal cost such that,

(20) j— rAK(Q,K)e‘(r+5)(-H)ds = C'(I(t)) = 77
t

where the LHS expression of the equation is the reduction in emission tax payments

resulting from an additional unit of emission-reducing technology investment at time t.

96

 

When aggregate emission-reducing capital stock is relatively low (i.e., K =300) the
optimal investment rate is high and decreases over time as it reaches the steady state at
equilibrium at point A. High levels of emission-reducing capital stock (K>900) requires
lower investment levels to reach the investment saddle path and the steady state
equilibrium for investment in emission-reducing technology and non-productive capital

stock as compared to low initial capital stock levels.

 

600

I

[721.92 = 0

 

 

 

I l T j

500

T fﬁ]

400

  
 

I f T—Tﬁ I 1

300
........................ Saddle path

200

T r T

 

 

 

 

100~ A

ll 4—1
,. .
I. / ‘M

I .1 l J I L #1 1| 1 l I J l

500 1000 1500 2000

 

 

Figure 4a. Optimal investment path for r = $1 .92 /lb ammonia emissions

The tax rate level inﬂuences investment decisions made by farmers. In the model

we assumed 2' = $1.92 per lb of ammonia emissions ($175/cow) which is a high tax rate.

97

For example an individual dairy farm with 100 cows would incur $17,500 in emission

taxes. If the milk price is $14/cwt and a dairy cows produces 70 pounds of milk per day
per cow, milk production revenue for approximately 18 days would be needed to pay the
emission tax. A second tax rate was included in the analysis to compare how investment

decisions change with a lower tax rate. We assumed the lower tax rate

was, t' L = $0.48 per lb ammonia emission, which was 75% lower than the tax rate )1-

reported by the American Farm Bureau (2009).

Figure 4b presents the phase-plane diagram for ‘l' H = $1.92 and

 

TL = $0.48 /lb ammonia emissions. The K = O isocline did not change with a new b

tax rate. I 2- 1 =0.48 = 0 shifted downward with the lower tax rate (red line in Figure

4b). The steady state equilibrium for 1' L = $0.48 is represented by point B with I=25

and K=516. Increasing the tax rate by 75% (T L to 2' H ) increased the optimal

investment rate by 80% and emission reducing capital stock by 74%. With the tax

increase from T L to 2' H the productive capital stock, cows, in the dairy industry

remained constant while the non—productive capital stock, K, increased. It was more
efﬁcient for the industry to add emission-reducing technologies on farms to decrease

emission tax payments rather than decreasing the aggregate herd population. Implicitly,

Ql/ZK—l/Z

the tax payment faced by an individual farm is T365 . Taking the

derivative of the tax payment with respect to emission-reducing capital, K, results in

— 1831Q1 / 2K —3 / 2 < 0 which shows that the emission tax payment decreases as

emission-reducing capital stock increases.

98

 

 

 

 

 

600— lr=0.48 =0
5001
4001
300L .

t 11:1.92 = 0

l
2001

. K7:1.92 = Kr=0.48 = 0 J
100—

. 1 1 . 1 l . 4 K
500 1000 1500 2000

Figure 4b. Optimal investment path for r” = $1.92 and TL = $0.48 /lb ammonia emission

Figures 4a and 4b demonstrate that as the emissions tax rate increases investment
levels must also increase in order to decrease or avoid the potential tax liability. The
optimization problem presented in equation (5) was solved for a series of emission tax
rates to determine how the amount of emission-reducing capital stock changed per cow as
the tax rate increased. Due to the speciﬁcation of this model, the number of dairy cows
remained constant at its current industry level. As shown previously, it was more
efﬁcient for the dairy industry to increase the emission-reducing capital stock rather than

decrease the aggregate herd size. Figure 4c demonstrates that as the tax rate increases,

99

the equilibrium level of capital stock per cow must increase. Therefore the investment

level per cows must also increase to adjust for the increased levels of emission-reducing

capital stock.

Capital/Cow
0002 .0003 .0004
I 1

.0001
l

 

I— I I
0 .5 1 1.5
Tax rate

—

N-I

Figure 4c. Emission-reducing capital stock per cow as tax rate increases

4.4 An Uncertain Emissions Tax Increase at a known Future Date

We now consider the case where an emissions tax will be imposed at a known
future date, T, but the magnitude of the emission tax is uncertain. For this problem the
dairy industry (all dairy farmers) consider potential tax rates that may be imposed at time

T, to adjust investment rates in emission-reducing technology from time t=0 to t=T. At

100

time T, the actual tax rate iS revealed and ajump in the investment rate may occur to
adjust to the desired saddle path.

Before we solve the case with an uncertain tax increase, it is useful to analyze the

case where the tax increase is known with certainty. Suppose a low tax rate, 1' L , is

imposed and at time T, the tax rate increases to T H which results in the following

maximization problem,

T
(20) max j[p,,,m(t2) + p590 — d)(1 — S) — wQ — C(1) — rLAm, K)]e’"dr
s,1 0

+ j[pmm(Q) + pSQ(I - d)(1 — S) — wO — C(1) — 2' HA(Q,K)]e_rt dt
T

s.t Owns, 520(0)
K=1—5K, K0(O)

Using the example presented in the previous section and Figure 4b, the

investment rate for the aggregate dairy industry must adjust to account for the higher tax

rate imposed at time T. With an increase to T H , the farmer’s investment rate will
deviate away from the saddle path for I 2' L = 0 and move towards the saddle path for
l t' H = Oas shown by the dashed line in Figure 4d. The new investment rate in

emission-reducing technology changes such that at time T when 2' H is imposed, the
farmers are on the saddle path for 1' H and moving towards (or at) the steady state

equilibrium (point A) for 1' H .

101

 

 

 

 

 

 

600 —

500 —

4001

3001

2001

: 1Kr=192 =Kr=0.48 =0

1005 A
I NT“ K
500 1000 1500 2000

Figure 4d. Optimal investment path for a tax increase from TL = $0.48to TH = $1 .92 at
time T.

With an uncertain emission tax increase, there is uncertainty regarding how to
adjust the investment rate in emission-reducing technology by time T. The new problem
considering this uncertainty can be set-up as a two—stage optimal control problem where
in the second stage the expected present value of cash ﬂows is maximized Since the tax
increase is unknown (D051 and Moretto, 1990). In the ﬁrst stage the present value of cash
ﬂows is maximized with the constraint that the cash ﬂows at time T must be equal to the

present value cash ﬂows calculated in stage two of the optimal control problem. This

102

leads to the farmers choosing the survival rate, S, and emission-reducing technology
investment, I, to maximize the present value of its cash ﬂows over an inﬁnite planning

period such that,

T
(21 ) max ﬂpmm(Q) + pSQ(I — d)(1— S) — wQ — C(1) - TLA(Q, K)]e"’dr
s 1
. 0

CD
+ E Hpmm(Q) + p590 — d)(1— S) — wO — C(1) — rA(O,K)]e‘”dr
T

s.t ('2 = 0205, (20(0)

K = 1 — 6K, K0(0)

We assume all dairy farmers are risk neutral and the proﬁt function is linear in

taxes such that 7tE(1') = E (72(7)) which implies that the comparison of optimal

investment paths with full certainty versus uncertainty depends on the tax rate at time T
and expected tax rate after time T. To solve the two stage optimal control problem we
ﬁrst solve the second term of equation (21),

(22)

max e‘” E ﬂpmm(Q) + pSQ(1 — d)(1— S) — wO — C(1) — rA(O,K)]e"("T)dr
s,1 T

s.t O = 0:95, 90(0)

K = I — 6K, K0(0)

103

 

Following this, the numerical solution to the second stage,

e—rtﬂ(K(TT ), Q(TT ), TT) , is included in the ﬁrst stage problem,

T
(23) max [[pmmm) + pSQ(1— d)(1— S) — wQ - C(I) — rLA(O,K)]e""dr
s,1 0

+e‘"rr(1<(rr).n(rr),rr)

s.t Owns, (20(0), K=I—5K, K0(0)

The necessary conditions for the optimal investment policy must include,

2 : ps(1-d) : 68(K(TT).Q(TT)JT)
(z (662

58(K(TT),Q(TT)JT)
6K '

 

(24)

(25) n = 0(1) =

 

If conditions (24) and (25) do not hold, the present value cash ﬂows from stage one will
not be equal to the present value calculated in stage two.
The numerical solution for the expected proﬁt in the second stage is dependent on

the tax rate. The optimal investment path was calculated for each tax rate
(1'1, 2'2 , and 1'3 ) for a given level of emission-reducing capital stock. The expected

proﬁt is,

(26) E (It K0(1)) = 12(71 W11) + p(72)7r(12) + p(T3)7r(r3)

= e‘"n<K<rr).cows(rr).rr).
The results of equation (26) are then included in the ﬁrst stage regression to solve the

optimal control theory problem for an uncertain tax increase at time T.

104

 

4.5 Empirical analysis of uncertain tax

The two-Stage optimal control theory model shows that the dairy industry takes
into account future tax rate increases to adjust (or not adjust) investment in emission-
reducing capital stock even though the tax rate is not imposed until time T. There are two
cases that can be analyzed with the two-stage optimal control modelf The ﬁrst case is
when the dairy industry is faced with a low tax rate from time t=0 and knows a tax
increase will be imposed at time T. This two stage problem is equivalent to equation
(23).

In the second case, an emission tax is not imposed in the ﬁrst stage of the
problem. Therefore, the adjustment to the optimal investment rate is dependent on the
expected tax rate imposed at time T and its effect on the proﬁt function for time Tto

inﬁnity. The two-stage optimal control theory problem for case 2 is,

T
(27) max j[p,,,m(t2) + p500 — d)(1 — S) — wQ — C(1)]e"”dr
s,1 0

+ e-” ”(K (1T). QITT ). TT)
s.t Owns, 520(0), K :1 —6K, K0(0).
It may be hypothesized that without a tax imposed at time i=0, there exists little incentive
to invest in emission-reducing technology. However, the possibility of future emissions
taxes may create incentive to increase investment in emission-redUcing technologies up
to time t=T.
A numerical application of an uncertain tax increase at time Twas implemented

using the data and parameters presented in Table 4a for the case where there is no initial

emissions tax in the ﬁrst stage of the problem. Three tax rates were assumed with equal

105

 

probability to estimate the expected proﬁt function for stage two at a given level of
emission-reducing capital stock. The three tax rates were: 1'1 = $0 as the minimum tax,
2'2 = $0.96 as an average tax, and a maximum tax rate of T3 = $1 .92 . An equal
probability was assigned to each tax rate where P( T] = $0 )=0.33, .

P( 1'2 = $0.96 )=0.33, and P( 1'3 2 $1 .92 )=0.33. The expected proﬁt function for a

given level of emission reducing capital stock was estimated as,

15(7: K0(,-)) = 0.337471 = O|K0(,-)) + 0.337r(72 = 0.96 K0(,-))

+ O.33)7r(r3 = 1.92|K0(,-)).

and included in the ﬁrst stage problem to solve for the optimal investment rate

considering an uncertain tax increase at time T.

Case 1: Low tax rate imposed in stage I

First, the optimal investment path for an uncertain tax increase at time t=Twith a

low tax rate 2' L = $0.48 /lb ammonia emissions imposed from time t=0 to t=T was
solved and presented in Figure 4e. First, if no tax rate increase was imposed, the

management decision for investment would continue on the saddle path for I 2' L towards

the steady state equilibrium level of investment in emission-reducing technology and
capital for the low tax rate. However, we know that an emission tax increase will occur

at time t=T, so the management decision for the investment rate level in emission-

reducing technology must deviate away from the I 2- L and move towards 113(1). The

106

 

rate at of this change in the investment path towards I E“) is determined by the

transversality conditions holding (equation (24) and (25)).

The investment rate for small levels of emission-reducing capital stock does not
deviate from the investment path for the low tax rate, 2' L = $0.48 , since at these points

the present value cash flows in stage one are equal to stage two. Uncertainty regarding
the increase in the tax rate when a low tax rate is imposed causes a U-shaped investment
path for small levels of initial emission-reducing capital stock, such that at small levels of
initial emission reducing capital stock, the investment path will not deviate from the low
tax rate investment path. However, at an emission-reducing capital stock level of K=252,
it is optimal to deviate away from the investment path for the low tax rate and move
toward the investment for the expected tax rate. By time T, the dairy industry has
adjusted to the saddle path for the expected tax rate. If the actual tax rate is equivalent to
the expected tax rate at time T, the dairy industry as a whole is already on the optimal
investment path and will move towards the steady state equilibrium level of emission-
reducing capital stock through increased investment in emission-reducing technology. If
the actual tax rate is higher than the expected rate, there will be an immediate jump
upward in the investment rate to the investment saddle path for the actual tax rate. A
realized tax rate at time t=Twhich is less than the expected rate results in investment

rates less than depreciation (K (K ) to reach the new steady state equilibrium.

107

 

 

 

 

 

 

 

 

 

 

 

 

 

6005 . L
5001
: 1' Em = 0
400 j
i Saddle path for 1' Em = 0 l
300 —
Saddle path for I}, = 0 I
2001
Saddle path for Fri = 0 l
100~
' KE(T)=KTL=0.48=(H
* ‘ . - K
500 1000 1500 2000

Figure 4e. Optimal investment path for uncertain tax increase at t=Twith a low
initial tax rate TL = $0.48

Case 2: No tax rate imposed in stage 1

In the second case, the expected future emissions tax at time t=T is the ﬁrst time a
tax has been imposed. Investment rates have not been adjusted previously as was the
case when a tax rate was already imposed before the announcement of a tax increase.
With the expected future tax, the investment rate in emission-reducing capital stock
increases up to time t=T(Figure 4f) for all levels of initial emission-reducing capital

stock up the equilibrium for the expected tax rate. To reach the saddle paths for

I E0") = 0 high investment rates must be incurred for low levels of emission-reducing

108

capital without knowing if an actual tax rate will be imposed since there is a positive

probability that the tax rate could be zero at time t=T. Without prior investment

adjustments, I EU) = 0 all farmers in the industry will need to increase investment rates

based on the deviation saddle path, j'r=0 = 0 rather than moving directly

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

toIE(z.) = 0.
l
600~ ,
:1 112(1) =0
500;
4001 '
300: ,Saddle path for I Eff) = 0
I
200;
Z Saddle path for i'Tzo = o
100; KE(T)=O.96 = K1=0 = O
F
A 500 A A A ‘lOOO‘ L A #15001 A ‘2000

Figure 4f. Optimal investment path for an uncertain tax increase at time t=Twith
r = O for t=0 to t=T.

Comparing the change in investment rates for the two cases shows that when the
uncertain tax increase is the ﬁrst time a tax is imposed, adjustment in investment rates

occurs for all levels of emission-reducing capital. When an initial low tax rate was

109

 

imposed in stage one, aggregate emission-reducing capital levels greater than K =252

adjusted the investment paths towards the expected tax rate. The difference in area

between the saddle paths for the optimal investment rate between I 2' L = Oand

I 13(7) = 0 were smaller for lower levels of emission-reducing capital compared to the

saddle path for 1.1-=0 = Oand I 13(7) = 0. Since the tax increase, of uncertain scale, in

time t=T is the ﬁrst introduction of an emissions tax, there are an increased number of
possible outcomes for investment rates and greater uncertainty surrounding it compared
to the case where there is a positive initial tax rate in the ﬁrst stage, in which fewer

possible outcomes exist.

4.6 Conclusions

The incentive to invest in emission-reducing technology is inﬂuenced by the size
and type of environmental policy instrument imposed. In this paper, we extended F arzin
and Kort’s (2000) theoretical model to include a functional form for animal air emissions
as a function of productive (cows) and non-productive (emission-reducing capital stock)
capital stock. Further, numerical solutions for an introduction of a certain tax rate at time
t=0 and an uncertain tax increase at time t=Twere estimated for an aggregate dairy
population model.

A certain tax rate imposed at time t=0 resulted in immediate increased investment

in emission-reducing technology. An uncertain tax increase imposed at time t=Tfor an

initial low tax, 2' L = $0.48 lb/ ammonia emissions, demonstrated farmers adjusted their

investment rates in emission reducing technologies towards the optimal investment level

110

 

in anticipation of an expected tax increase imposed at time t=T. When farmers did not
face an initial tax rate at time t=0 the amount of investment necessary to achieve the
optimal path was increased compared to those farmers who had faced a positive initial tax
rate. In other words, because farmers who had not faced any tax rate prior to time t=T
had not responded to potential tax increases, greater levels of investment were necessary
approaching time t=T relative to those facing positive initial tax rates. Finally, as the tax
rate increased, the amount of emission-reducing capital stock per cow increased to
decrease the amount of tax liability faced by the industry.

The numerical solution for the optimal investment policy allows policy makers to
evaluate how investment rates changed as the tax rate and uncertainty surrounding it
changed. This numerical model can be further extended to consider such items as
emission standards and tradable permits rather than emission taxes as well as the
inclusion of investment grants through cost-share programs to decrease investment costs

associated with emission-reducing technologies.

111

REFERENCES

Caswell, M., E. Lichtenberg, and D. Zilberman. “The effects of pricing policies on water
conservation and drainage.” American Journal of Agricultural Economics.
72(1990):883-890.

Dairy Herd Management. “Headed Down a Slippery Slope.” April 21, 2009. Available
at: http://www.daigherd.com/directories.asﬂngD=675&ed_id=8475.

Dosi, C., and M. Moretto. 1990. “Nonpoint —source pollution control, information
asymmetry, and the choice of time proﬁle for environmental fees.”

F arzin, Y.H., and RM Kort. “Pollution Abatement Investment when Environmental
Regulation is Uncertain.” Journal of Public Economic Theory. 2(2000)2183-212.

Gay, S.W., D.R. Schmidt, C.J. Clanton, KA. Janni, LD. Jacobson, and S. Weisberg.
“Odor, Total Reduced Sulfur, and Ammonia Emissions from Animal Housing
Facilities and Manure Storage Units in Minnestoa.” Applied Engineering in
Agriculture. l9,3(2003):347-360.

Hartl, R.F. and RM Kort. “Capital Accumulation of a Firm Facing an Emissions Tax.”
Journal of Economics. 63(1996): 1-23.

Kort. P.M. “Pollution Control and the Dynamics of the Firm: The Effects of Market-
Based Instruments on Optimal Firm Investments.” Optimal Control Application
and Methods. 17(1 996):267-279.

Koelsch, R., and R. Stowell. Ammonia Emissions Estimator (Daily Version). University
of Nebraska-Lincoln. January 2009. Available at:
http://water.unl.edu/c/document_library/get_ﬁle?folderld=67759&name=DLFE-
4647.pdf

McDonald, J .A., E.J. O’Donahue, W.D. McBride, R.F. Nehring, C.L. Sandretto, and R.
Musheim. Sept. 2007. Proﬁts, Costs, and the Changing Structure of Dairy
Farming. ERS Report-47.

Millman, SR. and R. Prince. “Firm incentives to promote technological innovation in
pollution control. Journal of Environmental Economics and Management.
17(1989):247-265.

Paulson, Jim. Feeding Dairy Cattle in 2009: Feed Costs and Proﬁtability—“What’s in
Your Wallet?” University of Minnesota. Available at:

http://www.extension.umn.edu/forages/pdfs/paulson__iefac_feed_costs_nov__2008.
pdf

112

Rosen, S. “Dynamic Animal Economics.” American Journal of Agricultural Economics.
69(1987):547-557.

Tarui N., and S. Polasky. “Environmental regulation with technology adoption, learning,
and strategic behavior.” Journal of Environmental Economics and Management.
50(2005):447-467.

United States Environmental Protection Agency (EPA). “Animal Feeding Operations Air
Quality Compliance Agreement Information Sheet.” March 2009/ Available at:
http://www.epa.gov/compliance/resources/agreements/caa/cafo-infosht-O309.html

United States Department of Agriculture (NASS). Agricultural Statistics Board—
National Agricultural Statistics Service. Livestock Slaughter Data. January
2008.

United States Department of Agriculture-Economic Research Services (NASS/ERS).
2007. Proﬁts, Costs, and the Changing Structure of Dairy Farming. ERR-47.
Available at: http://www.ers.usd2_1.gov/publications/err47/err47d.pdf.

United States Department of Agriculture-Agricultural Marketing Services
(NASS/ARMS). 2007. Market News and Transportation Data. Dairy Milk Price
Data. Available at:
http://www.ams.usda.gov/AMSvl .O/ams.fetchTemplateData.do?template=Templ
ateA&navID=MarketNewsAndTransportationData&leftNav=MarketNewsAndTr
ansportationData&page=DairyCommodityAveragesMain

United States Department of Agriculture-Agricultural Research Service (USDA/ARS),
Texas Agrilife Research, Texas Agrilife Extension, Texas A&M University, and
West Texas A&M University. 2003-2008. “Air Quality: Reducing Emissions
from Cattle Feedlots and Dairies.” USDA-CSREES funded projected.

Wolf, C., S. Hanson, E. Wittenberg, and S. Harsh. 2002. Discount rates for dairy farms.
Michigan State University Agricultural Economics Staff Paper 02-08.

Xepapadeas, A.P. “Environmental Policy Design and Dynamic Nonpoint-Source
Pollution.” Journal of Environmental Economics and Management. 23(1992):22-
39.

113