MSU
LIBRARIES
_;_..

RETURNING MATERIALS:

 

 

Place in book drop to
remove this checkout from
your record. FINES wilI
be charged if book is
returned after the date
stamped below.

 

1w 2:

 

 

IMPACT OF A CORN COB GASIFIER

ON THE FUEL CONSUMPTION OF A DUAL FUELED DIESEL ENGINE

by

DAVID BALLINGER JACKSON

A THESIS

Submitted to
Michigan State University in partial fullfillment

of the requirements for the degree of

Master of Science
Agricultural Engineering Technology

Department of Agricultural Engineering

1987

ABSTRACT

IMPACT OF A CORN COB GASIFIER ON THE FUEL CONSUMPTION

OF A DUAL FUELED DIESEL ENGINE

BY

DAVID BALLINGER JACKSON

Corn cob fuel was used to operate an open top,
stratified down draft gasifier, powering a 3.7 kw slow
speed diesel engine. Three cob moisture levels; 5%, 15%,
18% wet basis, were tested at 3 engine speeds; 400, 500,
and 600 rpm; in a randomized complete block experimental
design. Engine speed and power were held constant while
diesel fuel consumption was monitored to determine the
combination that saved the most diesel fuel, for the given
speed and power level. At all speeds the lowest moisture
cobs produced the greatest savings in diesel fuel.

To assess the potential financial benefits from the
substitution of corn cobs for diesel fuel, partial budgets,
incremental net benefit streams, and financial internal
rate of return calculations were done for various cob
yields, cob prices, and diesel fuel prices; for the Sahel

region of West Africa.

ACKNOWLEDGEMENTS

The author would like to express his sincere
gratitude to the following people for their contributions
and assistance toward completing this task:

Dr. Robert Wilkinson for his guidance, encouragement, humor
and persistance in tackling this research problem.

Dr. Merle Esmay for his confidence; enthusiasm; advice and
experience on the issues and problems facing developing
nations.

Richard Wolthuis and Dennis Welch for their invaluable
assistance suggestions and tips on constructing the
gasification system and using the research lab equipment.

Graduate students - Pascal Kaumbutho, Abbas Etigani, Steve
Ferns, Luke Reese, Maria Adam, Alex Akor, and Ibrahim
Mohammad for the many hours of theraputic input required to
maintain a positive attitude.

Gary Conner - the technician extrodinaire, for the numerous
hours of unfailing dedication to the conception and
realization of the instrumentation and data acquisition
system.

Dr. Robert Stevens for providing insight into the
importance of economics in the success of this research.

Isabel Gabashane for her unselfish assistance and willing
support at all times and under any circumstances.

Friends and neighbors too numerous to mention, for their
perennial support and encouragement even when they hadn't
the slightest idea what it was I was doing.

Most of all to my parents, Clarence and Bobette Jackson for
the many years of confidence, sympathy, pain and sacrifice
that made this undertaking possible.

TABLE OF CONTENTS

LIST OF TABLES
LIST OF FIGURES

CHAPTER -INTRODUCTIONOOOOOOOOOOOOQ ..... O. ....... 0.0... 1

CHAPTER - OJECTIVES ...... .............. .... ....... .... 6
LITERATURE REVIEW . ........................ .. 9
Gasifier Designs.............. ....... ..... 9
Processes ... .......................... . 10
Wood Fueled Downdraft Gasifiers ......... 13
Biomass Gasification in Developing
Nations.................................. 18
Corn Cobs as a Gasifier Feedstock ....... 22
Economics of Biomass Gasification ....... 28
Applications Meriting the Gasifier
Alternative ............................. 28
Financial Cost ............ ...... ........ 31
System Cost ........................... 31
Maintance and Labor Cost ... ....... .... 32
Diesel Fuel Cost ........ .......... .... 33
Biomass Feedstock Cost ..... . ...... .... 33

CHAPTER

HFJHFJI
O O O O

U"! DMNH

O
H

NNNNN

NNNNN NNH
pump

QUIPMENT .................................. 38
Gasification System Description ......... 38
Engine Characteristics .................. 40
Analog to Digital Data Conversion
Hardware .......................... ...... 40
Engine Horespower Measurement . ...... .... 44
Gas Temperature Measurement
Instrumentation ......................... 49
Data Acquistion Software ................ 53
Sensor and Instrument Calibration ....... 55
Engine Speed ............................ 58
Load Cell Calibration ........ ...... ..... 59
Thermistor Calibration .................. 62
Thermocouple Calibration .............. .. 64

CHAPTER

Hrahtl
O I 0

ON U1"h (AND-JP]

l
2
3
3.
3.
3.
3.
3
3
3
3
3
3
3
3
4
4
4
4

O 0
HH
0 0

PP

NNNNNH
IFUJNH

- METHODS AND PROCEDURE .................... .. 68
.1 Experimental Design . ................ .... 68
1.1 Considerations ..... ..................... 68

CHAPTER

(flLflUI a>pu>¢>pu>

5.1.2 Treatment Selection ..................... 69
5.1.3 Experimental Error ...................... 70
5.1.4 Replication and Blocking ..... ........... 70
5.2 Experimental Procedure .................. 71
5.2.1 Treatment Preparation ...... ..... ........ 71
5.2.2 Treatment Application .................. . 73
CHAPTER 6 - RESULTS AND DISCUSSION ... ................. . 77
6.1 Overview ............... ..... . ..... ...... 77
6.2 Grate Temperature Data ......... ...... ... 80
6.2.1 Temperature Profiles for High
Moisture Cobs ................. ......... 82
6.2.2 Temperature Profiles for Medium

Maisture CObs ...OOOOOOOOOOOOOOOI0.0.0... 86
6.2.3 Temperature Profiles for Low

Moisture Cobs ........ ...... ........... 87
6.2.4 Summary of Temperature Results

at 400 rpm ........ ...................... 87
6.2.5 Summary of Temperature Results

at 500 rpm ............ ....... . ..... ..... 88
6.2.6 Summary of Temperature Results

at 600 rpm ..... ...... .............. 89
6.3 Gas Cooler Inlet Temperature Results .... 90
6.3.1 Cooler Inlet Temperature Results

at 400 rpm .............................. 96
6.3.2 Cooler Inlet Temperature Results

at 500 rpm .............................. 96
6.3.3 Cooler Inlet Temperature Results

at 600 rpm .............................. 98
6.4 Gas Cooler Exit Temperature Results .... 100
6.4.1 Gas Cooler Exit Temperature Results

at 400 rpm ............................. 104
6.4.2 Gas Cooler Exit Temperature Results

at 500 rpm ............................. 105
6.4.3 Gas Cooler Exit Temperature Results

at 600 rpm ............. ..... . ..... ..... 106
6.5 Fuel Consumption Test Results .......... 107
6.5.1 Diesel Fuel Mode ...... ....... .......... 108
6.5.2 Dual Fueled Mode ......... .. ........ 110
6.5.3 Regression Curve Fit to Fuel, Power,

and Speed Means ...... .. ........... 116
6.6 Hypothetical Irrigated Farm Model

Using Gasification ..... ........ . ....... 119
6.6.1 Engine Power Derating and Pump

Discharge Capacity ............. ........ 120
6.6.2 Water Requirement for Corn ..... ....... . 121
6.6.3 Quantity of Cobs Available for

Gasification ..................... ...... 122
6.6.4 Corn Cob Consumption ................... 124
6.6.5 Theoretical Corn Cob Consumption

on the Model Farm ...................... 125
6.6.6 Diesel Fuel Consumption ................ 125
6.6.7 Diesel Fuel Saved By Using Gasification 126

V

O O O O O O
a: cnaaawq~q
u: hora rd

CHAPTER

\lflfl ON mmmmm

I.
NHI

Financial Cost — Benefit Analysis ......

Inclusion of Cob Costs

senSitiVity AnaIYSiS ......OOOOIOOIOOOOO

Increased Corn Yield ..

Increased Corn Yield Combined with

Cob Cost

Increased Diesel Fuel Price

CONCLUSIONS

Experimental Aspects ... ...... ..........
Financial Aspects ....

APPENDIX - Slow Speed Diesel Manufacturers... ......... .

BIBLIOGRAPHY

vi

127
132
136
136

141
145

150
150
152
154

160

Table

Table

Table

Table

Table
Table

Table

Table
Table
Table
Table

Table

Table

Table

Table

Table

3.3

L I ST OF TABLES

Energy Content of Fuel Gases and Their Uses.. 11
Comparative Performance of a 3.7 kw (Shp)

Diesel Engine Run in (A) Dual Fuel Mode and

(B) Single Fuel Mode (Diesel) Mode.

source: crUZ’ 1980............OOOOOOOOOOOOOO. 19
Land Area Under Irrigation With Full Water
Control in the Senegal River Valley.

source: Boutillier' 1980.........IOOOOCOOOOOO 30
Behavior Ranking of Various Fuel Forms

source: Jenkins, 1980.00.00.0000000000000.... 36
Engine Speed Calibration Data................ 59
Load Cell Calibration Data - Increasing

Force OI....OOOOOOOOOOOIOOOOOOOO0.0.0.0000... 61
Load Cell Calibration Data - Decreasing
Force.......OOOOCOOOOOOOO 000000000 0.0.0.0.... 62
Thermistor Calibration Data and Averages ..... 64
Treatment Randomization Reasults... ..... ..... 75
Wet Basis Cob Moisture Data.. ..... ........... 78
Engine Speed Treatment Data.... ........ . ..... 79
Engine Speed Means Across Blocks and
Treatments.........OOOOOOOOO. ..... 0.0.0.0.... 79
Engine Power Treatment Data in kilowatts ..... 81
Engine Power Means Across Blocks and

Treatments (in kilowatts)......... ........ . . 81
Engine Speed ANOVA Table.................... 109
Engine Power ANOVA Table ................... . 109

vfi

Table
Table
Table
Table
Table
Table

Table

Table

Table

Table

Table

Table

Table

6.8 Diesel Fuel Consumption Treatment Data......

6.9 Diesel Consumption ANOVA Table..............

6.10
6.11
6.12
6.13
6.14

6.15

6.16

6.17

6.18

6.19

6.20

Diesel Fuel Consumption in ML/KWH..........
Corn Cob Consumption Data and Statistics...
Model Farm Pump Discharge Rates............
Corn Cob Produced on the Model Farm........

Effective Seasonal Working Days Using
Producer Gas in the Farm Model.............

Partial Budget for Irrigation Costs, With
and Without Gasification Fue1,for a Model
Farm of 8.77ha Low Moisture Cobs at Zero
Cost - Diesel Fuel Cost of $0.65 per liter
Grain Yield of 3 tons per hectare..........

Partial Budget for Irrigation Costs, With
and Without Gasification Fue1,for a Model
Farm of 8.77ha Medium Moisture Cobs at Zero
Cost - Diesel Fuel Cost of $0.65 per liter
Grain Yield of 3 tons per hectare..........

Partial Budget for Irrigation Costs, With
and Without Gasification Fue1,for a Model
Farm of 8.77ha High Moisture Cobs at Zero
Cost - Diesel Fuel Cost of $0.65 per liter
Grain Yield of 3 tons per hectare..........

Incremental Net Benefit and Financial
Internal Rate of Return for an 8.77 ha
Model Farm, Low Moisture Cobs at Zero
Cost - Diesel Fuel Cost of $0.65 per liter
Grain Yield of 3 tons per hectare..........

Incremental Net Benefit and Financial
Internal Rate of Return for an 8.77 ha
Model Farm, Medium Moisture Cobs at Zero
Cost - Diesel Fuel Cost of $0.65 per liter
Grain Yield of 3 tons per hectare

Incremental Net Benefit and Financial
Internal Rate of Return for an 8.77 ha
Model Farm, High Moisture Cobs at Zero
Cost - Diesel Fuel Cost of $0.65 per liter
Grain Yield of 3 tons per hectare ..........

WH

112
112
118
119
121
124

125

129

129

130

130

131

131

Table

Table

Table

Table

Table

Table

Table

Table

6.21

6.22

6.23

6.24

6.25

6.26

6.27

6.28

Partial Budget for Irrigation Costs, With
and Without Gasification Fue1,for a Model
Farm of 8.77ha Low Moisture Cobs at cost

of Sl/m - Diesel Fuel Cost of $0.65 per
liter - Grain Yield of 3 tons per hectare..

Partial Budget for Irrigation Costs, With
and Without Gasification Fuel,for a Model
Farm of 8.77ha Low Moisture Cobs at cost

of $3/m3- Diesel Fuel Cost of $0.65 per
liter - Grain Yield of 3 tons per hectare..

Partial Budget for Irrigation Costs, With
and Without Gasification Fuel,for a Model
Farm of 8.77ha Low Moisture Cobs at cost

of $5/m3- Diesel Fuel Cost of $0.65 per
liter - Grain Yield of 3 tons per hectare..

Incremental Net Benefit and Financial
Internal Rate of Return for an 8.77 ha
Model Farm, Low Moisture Cobs at cost of
Sl/m - Diesel Fuel Cost of $0.65 per liter
Grain Yield of 3 tons per hectare..........

Incremental Net Benefit and Financial
Internal Rate of Return for an 8.77 ha

Mode Farm, Low Moisture Cobs at cost of
$3/m - Diesel Fuel Cost of $0.65 per liter
Grain Yield of 3 tons per hectare..........

Incremental Net Benefit and Financial
Internal Rate of Return for an 8.77 ha
Model Farm, Low Moisture Cobs at cost of
$5/m - Diesel Fuel Cost of $0.65 per liter
Grain Yield of 3 tons per hectare..........

Partial Budget for Irrigation Costs, With
and Without Gasification Fuel,for a Model
Farm of 8.77ha Low Moisture Cobs at Zero
Cost - Diesel Fuel Cost of $0.65 per liter
Grain Yield of 5.65 tons per hectare.......

Partial Budget for Irrigation Costs, With
and Without Gasification Fue1,for a Model
Farm of 8.77ha Medium Moisture Cobs at Zero
Cost - Diesel Fuel Cost of $0.65 per liter
Grain Yield of 5.65 tons per hectare.......

u

133

133

134

134

135

135

138

139

Table 6.29 Partial Budget for Irrigation Costs, With
and Without Gasification Fuel,for a Model
Farm of 8.77ha High Moisture Cobs at Zero
Cost - Diesel Fuel Cost of $0.65 per liter
Grain Yield of 5.65 tons per hectare....... 139

Table 6.30 Incremental Net Benefit and Financial 140
Internal Rate of Return for an 8.77 ha
Model Farm, Low Moisture Cobs at Zero
Cost - Diesel Fuel Cost of $0.65 per liter
Grain Yield of 5.65 tons per hectare....... 140

Table 6.31 Incremental Net Benefit and Financial
Internal Rate of Return for an 8.77 ha
Model Farm, Med Moisture Cobs at Zero
Cost - Diesel Fuel Cost of $0.65 per liter
Grain Yield of 5.65 tons per hectare....... 140

Table 6.32 Incremental Net Benefit and Financial
Internal Rate of Return for an 8.77 ha
Model Farm, High Moisture Cobs at Zero
Cost - Diesel Fuel Cost of $0.65 per liter
Grain Yield of 5.65 tons per hectare....... 141

Table 6.33 Partial Budget for Irrigation Costs, With
and Without Gasification Fuel,for a Model
Farm of 8.77ha Low Moisture Cobs at cost
of Sl/m - Diesel Fuel Cost of $0.65 per
liter - Grain Yield of 5.65 tons/hectare... 142

Table 6.34 Partial Budget for Irrigation Costs, With
and Without Gasification Fuel,for a Model
Farm of 8.77ha Low Moisture Cobs at cost
of $3/m3- Diesel Fuel Cost of $0.65 per
liter - Grain Yield of 5.65 tons/hectare... 143

Table 6.35 Partial Budget for Irrigation Costs, With
and Without Gasification Fuel,for a Model
Farm of 8.77ha Low Moisture Cobs at cost
of $5/m3- Diesel Fuel Cost of $0.65 per
liter - Grain Yield of 5.65 tons/hectare... 143

Table 6.36 Incremental Net Benefit and Financial
Internal Rate of Return for an 8.77 ha
Mode Farm, Low Moisture Cobs at cost of
$l/m - Diesel Fuel Cost of $0.65 per liter
Grain Yield of 5.65 tons per hectare....... 144

Table 6.37

Table 6.38

Table 6.39

Table 6.40

Table 6.41

Table 6.42

Table 6.43

Incremental Net Benefit and Financial
Internal Rate of Return for an 8.77 ha

Mode Farm, Low Moisture Cobs at cost of
$3/m - Diesel Fuel Cost of $0.65 per liter
Grain Yield of 5.65 tons per hectare.......

Incremental Net Benefit and Financial
Internal Rate of Return for an 8.77 ha

Mode Farm, Low Moisture Cobs at cost of
$5/m - Diesel Fuel Cost of $0.65 per liter
Grain Yield of 5.65 tons per hectare.......

Partial Budget for Irrigation Costs, With
and Without Gasification Fuel,for a Model
Farm of 8.77ha Low Moisture Cobs at Zero
Cost - Diesel Fuel Cost of $0.78 per liter
Grain Yield of 3 tons per hectare..........

Partial Budget for Irrigation Costs, With
and Without Gasification Fuel,for a Model
Farm of 8.77ha Medium Moisture Cobs at Zero
Cost - Diesel Fuel Cost of $0.78 per liter
Grain Yield of 3 tons per hectare..........

Incremental Net Benefit and Financial
Internal Rate of Return for an 8.77 ha
Model Farm, Low Moisture Cobs at Zero
Cost - Diesel Fuel Cost of $0.78 per liter
Grain Yield of 3.0 tons per hectare........

Incremental Net Benefit and Financial
Internal Rate of Return for an 8.77 ha
Model Farm, Med Moisture Cobs at Zero

Cost - Diesel Fuel Cost of $0.78 per liter
Grain Yield of 3.0 tons per hectare........

Summary of the Assumed Values of Important
Parameters and Financial Internal Rates of
Return for Corn Cob Gasification, on an
8.77 ha Model Farm Under Surface Irrigation
Using a 3.13 kw Slow Speed Diesel Pumpset..

xi

144

145

146

147

147

148

149

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure
Figure
Figure

Figure

1.1

3.2

3.3

3.5

3.6

LIST OF FIGURES

Map of the African Continent with an outline
of the Sahel Zone. Source: CILSS, 1979 ..... .. 3

Various Gasifier Types; Updraft, downdraft
and Crossdraft. Source Breag and Chittenden
1979........................................ 12
Maximum Power Performance at Diesel and Dual
Fuel Operations. Diesel Engine, 3.5 liters
cyl. displ., mounted in a tractor. Diesel
fuel for ignition: 29g/Nm3 producer gas
Source: Johansson, 1980............. ....... .

Maximum Power Performance at Diesel and Dual
Fuel Operations. Diesel Engine, 6.2 liters

cyl. displ., mounted in a truck. Diesel

fuel for ignition: 12-19g/Nm producer

Gas. Source: Johansson, 1980................ 16

Wet Basis Desorption Isotherms from
Modified Henderson Equation When Predicting

EMC. Source: White et al, 1984........ ..... . 25
Difference in Wet Basis Desorption and
Adsorption Isotherms at 30°C When

Predicting EMC With Modified Henderson
Equation. Source: White et al, 1984......... 25
Tar From Corn Cob Gasification.

Source: Silva, 1984 ..... . ..... ...... ........ 26
Effect of Diesel Fuel Price on Gasifier
Economics. Source: Barnard, 1984.... ........ 34
Open Top Gasifier Components......... ....... 39
Gasifier System/Engine Arrangement.. ........ 41
Digital Data Acquistion System.... ........ .. 45
The Prony Brake Mechanism ......... . ........ . 45

xfi

Figure

Figure

Figure

Figure
Figure
Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

Figure

4.8
4.9
4.10
4.11

4.12

6.1

6.2

6.3

6.5

6.6

6.7

Reed Relay Pick-Up Position......... ..... ... 45

Flowchart for A/D Converter Channel
Sampling Parameters......................... 56

Flowchart of the Downloading Program
to transfer Data from the A/D Converter
to the Portable Computer.................... 57

Engine Speed Calibration Curve.............. 60
Load Cell Calibration Curve................. 60
Thermistor Calibration Curve............... 63

High Temperature Thermocouple Calibration
CurveOOOOOOOOO ....... O ...... .....OOOOOQOOOO 67

Low Temperature Calibration Curve... ...... . 67

Gasifier grate temperature profiles at 400

rpm engine speed. The curves represent
averages over the 3 blocks for the 10

minute length of the test................... 83

Gasifier grate temperature profiles at 500

rpm engine speed. the curves represent

averages over the 3 blocks for the 10

minute length of the test ...... ............. 84

Gasifier grate temperature profiles at 600

rpm engine speed. The curves represent

averages over the 3 blocks for the 10

minute length of the test............. ...... 85

Gas temperature profiles at the cooler
entrance for 400 rpm. Averaged across 3
blocks for the 10 minute length of the
test........................................ 91

Gas temperature profiles at the cooler
entrance for 500 rpm. Averaged across 3
blocks for the 10 minute length of the

testOOOOOOOOOOOOI00.0.0.0.........OOOOOOOOOO 92

Gas temperature profiles at the cooler
entrance for 600 rpm. Averaged across 3
blocks for the 10 minute length of the
test........................................ 93

Gas temperature profiles at the cooler

exit for 400 rpm. Averaged across 3 blocks
for the 10 minute length of the test.... 101

“U

Figure

Figure

Figure

Figure

Figure

Figure

6.8

6.9

6.11

6.13

Gas temperature profiles at the cooler
exit for 500 rpm. Averaged across 3 blocks
for the 10 minute length of the test....... 102

Gas temperature profiles at the cooler
exit for 600 rpm. Averaged across 3 blocks
for the 10 minute length of the test....... 103

Diesel fuel consumption at low, medium

and high cob moisture levels. Engine

speed points are 373 rpm, 496 rpm, and

585 rpm.......................... ......... 114

Diesel fuel consumption at low, medium
and high engine speed levels. Moisture
points (wet basis) are 5%, 13%, and 18%... 115

Linear regression plots of diesel fuel
consumption in ML/KWH. Curves 1,2, and 3

are the low medium and high moisture levels
wet basis. Curve 4 is pure diesel fuel.... 117

Mean Relative Humidity for Africa for the
month of October. Source: FAO, 1980....... 123

“v

CHAPTER 1

INTRODUCTION
1.1 Background

In recent years interest in alternative energy
technologies has proliferated in developing countries
(George,1980;Stassen,1980; Barnard,1982). Among the more
promising techniques. reviewed is the realm of biomass
energy conversion. This appears to be a logical approach
since the majority of rural populations in developing areas
utilize some form of biomass (such as firewood or charcoal)
as a major energy source. Indeed it is one of the oldest
sources of energy known to mankind.

Even though petroleum prices have recently declined on
the world market, consumers in many developing nations have
continued to experience price increases. In 1985 the
government of the Republic of Mali increased the diesel
fuel price 16 percent (MFC, 1985) over the 1984 level.
Government intervention by price setting is one reason for
these distorted prices. Governments often obtain tax
revenues on fuel, and while this is not a free market
inflationary factor on pricing, the end result is the same
for the consumer, inflated prices. There are also problems
of irregular supplies and high transport cost to remote
areas (OFDA,1978). These factors can push price levels far
beyond the official (government) price.

1

2

In the Sahel region of Africa (figure 1.1), many
consumers are located in areas that have only seasonal
access by road. It is not unusual for these areas to be
among the more agriculturaly productive in the country due
to controlled water application by irrigation. The
principal source of power for water lifting in these areas
is the diesel engine. Surface water is pumped mainly from
rivers like the Niger river (passing through Guinea, Mali,
Niger and Nigeria) and the Senegal river bordering on
Senegal and Mauritania (Figure 1.1). Diesel powered
irrigation has been the quickest and surest approach for
offsetting declines in crop production. Salinger and
Stryker (1983) compared returns to irrigated versus dry
land agriculture in the Senegal river basin, and found
irrigated production to provide superior rates of return in
Senegal and Mauritania.

Surface irrigation is the predominant technique used
in the area. Two schemes are employed: one is referred to
as small perimeters, the other, large perimeters. Small
perimeters are operated in a cooperative manner by the
members of a village. The average surface area is around 40
hectares,and is managed with practically no mechanised
inputs other than the diesel powered irrigation pump. Large
perimeters are owned and operated by the governments. The
minimum size of these schemes is typically 500 hectares.
They are characterized by a high level of mechanized

inputs, and heavy bureaucracy.

DJ

Africa political

3:: ‘7? E ‘ ::::::.::::.:~°

...... 1%“ l.”

IOUAIOMRI
Ion-IA
“0 tone I m 0
CO ‘ . “""‘ (0-00
noon

...“. ......”

 

 

         

 

 

 

 

0—.
" “T,” thou-II
avenue menu “—-
nun-u .-
. o a
. l-
e' .. ‘
IUCOLA !
I
3.“... 0|“ §.~
0—1.
.,\' f
\.__ _' .. OJ
0
...—.-
I...“
.u.

 

Figure 1.1 Map of the African Continent with an outline
of the Sahel Zone. Source: CILSS, 1979

4

Diesel powered irrigation in the Senegal River Valley,
had expanded from 2,000 hectares in 1977 to 20,000 hectares
in 1982 (CILSS,1982). Future growth is projected at 2—3,000
hectares per year through the year 2000. This large
increase in area irrigated has proportionally increased the
demand for diesel fuel. Experience thus far indicate that
fuel is the highest production cost of diesel oowered
irrigation systems in the Senegal river basin (Fieloux,
1980).

Since many nations are not producing their own
petroleum fuel they must import it. Foreign exchange
payments can be high to those nations that consume large
quantities of fuel. In Mauritania, imports quadrupled over
a 12 year period from 1959 to 1972. Petroleum products
comprised a significant amount of the increase and the
trend is continuing (OFDA,1978). Reducing the demand for
petroleum by substituting alternative fuels would free
foreign exchange for other purchases.

Gas producers are a potential response to the energy
problems experienced by rural inhabitants in some of these
countries. Gas producers or gasifiers, convert biomass into
energy in the form of gas. The process is called
gasification. Gasification occurs when biomass is burned
with a limited supply of air. The products are carbon
monoxide, carbon dioxide, hydrogen, and methane gases. The
energy available from this gas mixture is suitable as a

supplementary fuel for diesel engines.

5

There is substantial information available on using
gas generators to power gasoline and diesel engines. The
bulk of this information however, refers to the use of wood
and/or charcoal as the biomass feedstock for the gas
generator. This is suitable for countries with large forest
resources. In other countries, forest resources are so
limited that harvesting of wood on a large scale is not
practical. Environmental transformation would be
significant and most likely destructive. There is still the
possibility that other sources of biomass, like crop
residues, could furnish significant quantities of energy.

Gas generators are not without drawbacks and
shortcomings. While it is a relatively inexpensive unit to
build, the supply of biomass must be substantial to operate
on a continuous basis. Considerable fuel preparation may be
required for it to perform adequately ie. drying, sizing
etc. The human element plays a major role in operational
efficiency of the unit. High output by the gas generator
unit demands that attention is paid to details. Leaks must
be kept to a minimum, air/gas mixture precisely maintained,
and cleaning of filters and piping done frequently.

Until recently, all of the low cost gas generators
were of the closed batch type design. The open top
stratified design may offer some new advantages due to

its design simplicity.

CHAPTER 2

OBJECTIVES

The intent of this investigation is to obtain more
specific information on the relationship of moisture in
biomass fuel, to the power produced with biomass under
gasification.

In this particular case the biomass is corn cobs.
Three moisture levels were selected for analysis. They were
in the ranges of 5, 15, and 25% wet basis. This range of
moisture contents were selected based on the work of White
et a1 (1984) and weather data for West Africa taken from
the Food and Agriculture Organization (1980). At
temperatures of 30-50 degrees celsius the equilibrium
moisture content wet basis, of corn cobs peaks at just
above 25% with 90 percent equilibrium relative humidity.
The lowest moisture content is around 5% moisture wet
basis, at a relative humidity of 30 percent.

Data from the FAO indicate that the mean daily
relative humidity in humid zones of Africa is in the 70-80
percent range. In the more arid areas such as the Sahel,
the mean daily relative humidity can drop as low as 30
percent during the dry season. On the basis of this data
the range 5 to 25 percent moisture content wet basis was
chosen for the experiment.

The method of determining power is the diesel engine.

6

7
The productivity of biomass fuel was measured by monitoring
the amount of diesel fuel consumed, at a given speed and
power level. This was done when the engine was operated in
the dual fueled mode, running on both gas from biomass and
diesel fuel. The hypothesis is that the energy provided by
the biomass gas as moisture content is lowered, will
increasingly replace and reduce diesel fuel consumption at
a given speed and power level.
The experimental objectives then were to:

1. Determine which of the conn cob moisture levels
ie. 5, 15, 25 percent nominal moisture content
wet basis, produces the greatest reduction in
diesel fuel consumption.

2. Determine if there is a significant difference
between the amount of diesel fuel consumed at
the 3 corn cob moisture levels.

3. Evaluate the relationship between the moisture
content of the corn cobs and the temperature at
the grate of the gasifier, and at the cooler
entrance and exit.

4. Evaluate the feasibility of using corn cobs for
energy in the Sahel zone of West Africa, by

means of financial cost benefit analysis.

The engine speeds selected for the experiment were
400 500 and 600 rpm. This speed range is selected because

in preliminary tests of the gasification system,

8
satisfactory performance was unattainable at higher speeds
with the 15 and 25 percent cob moisture levels. Also
engines of the type used in these studies are typically
slow speed diesels operating at less than 1,000 rpm. Such
models are manufactured in India and the Peoples Repubic of

China. Specifications of some models are available in

the appendix.*

* Note: the mention of a particular manufacturers or
suppliers products in this document does not constitute an

endorsement of the product(s).

CHAPTER 3
LITERATURE REVIEW

Gasifiers have been used as sources of fuel since he
1800's. The earliest ones were made of fire brick. They
were fueled with coal, charcoal, or wood and were used to
heat blast furnaces in the steel industry (Donkin,l905).
During World War II gasifiers or producer gas generators,
as they were sometimes called, gained popularity as a fuel
source for transportation in Europe (Griffen,l944;
Nowakowska et al,l945). Many trucks,buses, and tractors
were equipped with gasifiers and ran on wood or charcoal
since fossil fuels were scarce during World War II. Once
cheaper fossil fuels became available in the post war
period, attention to gasification technology was largely
abandoned, until the mid 1970's when higher fossil fuel

cost generated renewed interest in alternative energy.
3.1.1 Gasifier Designs

Gasifiers, or producer gas generators, are available

in 3 principal configurations (Reed, 1981) :

Updaft
Downdraft
Crossdraft
These terms refer to the direction in which the air
stream flows through the gasifier as it feeds the

9

10
combustion process and breaks the fuel down into the
desired gases. Downdraft indicates that the air flow is in
a downward direction, through the gasifier (fig 3.1).
Crossdraft refers to a sideways or lateral path of air
flow, and updraft refers to an upward direction of air
flow.

Downdraft models are popular for applications
involving internal combustion engines because they produce
a cleaner gas. This eliminates the need for frequent
cleaning of the filters, coolers and piping between the
gasifier and the engine. It also permits the use of simpler
and less costly filtering equipment. The disadvantage is
that the gas produced has a slightly lower energy content

than that of the other models.
3.1.2 Processes

Air gasifiers are those which combust fuel with a
limited amount of air. The gas produced has an energy
content of 5,600-7500 kj/m3 (150-200 btu/scf). By contrast
oxygen, hydrogen, and pyrolysis gasifiers produce gas with
energy contents of 11,000-18,500 kj/m3 (300-500 btu/scf,see
Table 3.1). These higher energy system are more
sophisticated and expensive than air gasifiers.(Reed 1981).
For this reason the air gasification process is selected
for study as a potential candidate for developing areas due
to its simplicity and low cost.

The gasification processes yields a combustible gas

11

.HmmH .xmoHocnumE cam mmHQLocLLd i coflumuwuwmmo mmmEoLm “mousom

mm:

xuwu Ucm.uwzod.ummc
Hmpmcwm LOW mmcfia
imdwm mocmumwc mcoq

Amcfinnauum\3v mafia
immwa .umw; mmmooLm

mLcoEEm 0cm maosu Lo
mwmonucam.mmcwamawd
Hmwuumaccfi Hmcowmmm

mumaflon HL0\mmm
.mcmwuc douu.mwcflmcm
xLMQm cam Hmmmflo

oooa

oomuoom

oomioom

oomioma

mHHmE mmD\HLO

cofibmmmfio ownomec<

cofiumuwufimmw
cmmwxo

coflumoMmMmmu Edd

mmo xmumcm nmwm

mmmowm

Aomzv
mmu amuwcm ancmz

Aomqv
mow hmumcm 30¢

AmUm\DBmv
m02<m wummzm

mom: Lfimce cam momma Hmsm mo ucmbcou xmcmcm

H.m mHDmE

12

Oily-m 1
Up -- thought unit

Drying zone

Dittiiiotioo tone

 

 

 

Reduction to».

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

L Heath tone

 

i. ______ ... _ Gale

__ _J
wgm Ash tone
Diagram 2 H". PF\L
Down - dtauqiit unit “A x

 

 

 

 

 

\\ ‘\ \_\ \ .va-I _ D'v."° IO".
\ '.

 

Dittiiietion tone

...- .. ‘ ________ iioouh lone

AU-—ljiiifliil_iliiii§:__m.

('30 0|.

 

 

 

 

Ath pit

”imam 3
Cut” ‘- “taught unit

Drying ton.

Dittimtuon zone

noduclion Inn.
Ail

homing that

Mouth tom Gut.

Ad! pit

 

Figure 3.1 Various Gasifier types; Updraft, downdraft

and Crossdraft. Source: Breag and Chittenden,
1979

13
from solid fuels such as coal, charcoal wood and crop
residues. The fuel is ignited and fed with a stream of air
to perpetuate combustion. The fuel reacts with air at at
temperatures of 1,000°C to produce a gas contaianing
approximatly 20-30 percent carbon monoxide (CO),5-15
percent hydrogen (H), and up to 3 percent methane (CH4)

(Kaupp et a1 l98l,Reed 1981, Sakai et a1 1978).
3.1.3 Wood Fueled Downdraft Gasifiers

There is substantial information on downdraft
gasification of wood and charcoal fuels, but comparitively
little on corn cobs, especially where the operation of IC
engines is concerned. As such, much of the information that
follows is related to wood or charcoal fueled gas producers
coupled to IC engines.

Payne (1978) reports that the downdraft type gasifier
design first appeared in 1924. It was developed by the
Imbert company of West Germany. Downdraft units produce a
cleaner gas product and thus are more suitable to power
internal combustion engines. Downdraft refers to the
direction of flow of the combustion air through the
gasifier. Titl (1981) evaluated saw dust, bark and wood
waste in gasifier diesel cogeneration system for electric
power and process heat. Montgomery (1980) sucessfully
operated an air blown gasifier to fire a boiler and
generated process steam. McGowan and Jape (1981) did

comprehensive studies on wood fueled gasifiers, and

14

documented the performance of numerous gasifier types. From
these studies data were obtained on pressure drops,
flammability limits of producer gas, and combustion
characteristics of selected fuel gases. Eoff and Post
(1982) converted a six cylinder Chevrolet truck engine for
operation on producer gas as a demonstration unit. Power
output from the spark ignition engine was reported to be
50% that of pure gasoline.

Russell (1980) used a similar approach and converted a
Chevrolet Malibu station wagon engine to run on producer
gas from a downdraft gasifier. The unit was fueled with
wood blocks of 15-25% moisture content wet basis. The
vehicle was driven for 12,320 KM (7,700 miles) and no
unusual wear or abnormalities were found in the engine.

Johansson (1980) tested spark ignition and compression
ignition engines' in trucks and tractors in Sweden. He
obtained performance levels of 96% that of pure diesel
fuel, when operating a truck mounted gasifier on wood with
moisture contents of 12% wet basis. The tractor mounted
unit acheived a diesel fuel substitution of 80%. These
results are depicted in figures 3.2 and 3.3.

Generally, improved indicated thermal efficierncy
occurs with producer gas due to more complete combustion
and a lower flame temperature (Ogunlowo 1979, Held and
Koenig 1982). Kaupp (1980) revealed that low speed engines
having a large inertial mass, large piston displacement and

large combustion space have significant advantages over

15

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

50
S: 1 /
o2
é’ 45 DIESEL /
o. /
... i
m
u, 40 1‘
x
E‘ / /
a / /
35 //
.30 ‘ ’////<:\
/ \DUAL FUEL
25 /
20.
90-7. PERCENT OF DIESEL PERFORMAMEE
60
70
l
1000 1200 1400 1600
ENGINE SPEED. RPM

Figure 3.2 Maximum Power Performance at Diesel and Dual
Fuel Operations. Diesel Engine, 3.5 liters

cyl. displ.,
fuel for ignition:
Source: Johansson,

mounted in a tractor.
29g/Nm3 producer gas
1980.

Diesel

l6

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1001
o.
I IESEL /
a? 90 x
33 '/
<3
IL
u:
g; 1’,/”’T’f””’
g; 00 44,
~ I
x
3‘; DUAL FUEL

\CL

“' 70 /,

60 /'

50‘#/

40

PERCEN -

1000,]. TIOF DIESEL PERFORMANCE

90

60

1000 1500 2000
ENGINE SPEED, RPM
Figure 3.3 Maximum Power Performance at Diesel and Dual

Fuel Operations. Diesel Engine, 6.2 liters
cyl. displ., mounted in a truck. Diesel
fuel for ignition: 12-19g/Nm producer
Gas. Source: Johansson, 1980.

17
compact, light, high speed engines. These advantages
combined with large intake valves, appropriate opening
timing and aerodynamic induction system make the difference
between a poorly functioning engine and one with a power
output very close to that of the normal fuel.

Nordstrom (1962) conducted thorough studies on various
diesel engines operated with producer gas. He concentrated
on gasifiers for mobile applications on trucks, tractors
and buses. Significant among his work were the effects of
increasing the pilot amount of diesel fuel injection for
igniting the producer gas. He revealed that the gain in
power increased at an increasing rate proportionally with
engine speed. Also the effect of doubling the amount of the
pilot injection only increases the power output over a
range of 1-7% from low speed to high speed operation.

A reduction in diesel fuel consumption of 75% was
reported by Shaw et al (1983) using a downdraft gasifier
fueled with citrus wood, powering a six cylinder naturally
aspirated 5.8 liter Perkins diesel. The engine developed
73% of its rated full power on diesel fuel when operated on
the duel fuel mode. Johansson (1980) did similar studies
and reported reductions in diesel fuel comsumption of 85-
90% of the normal quantity, with tractor and truck engines
of 3.5 liters and 6.2 liters displacement. Dennetiere et a1
(1980) worked with a super charged diesel connected to a
1000 KVA generating set at Duvant Moteurs in ‘France. He

obtained results showing an 80% reduction in diesel fuel

18

consumpiton.
3.1.4 Biomass Gasification in Developing Nations

Producer gas powered internal combustion engines have
seen growing applications in developing nations. In the
Philippines Cruz (1980) conducted studies on single and
multiple cylinder dual fueled diesel engines. His aim was
to test the practicality of producer gas powered diesels
for the Philippines agricultural environment. The gasifier
design was a convertible type that would operate in either
the updraft or downdraft configuration. Cruz found that the
updraft mode was unmanageable due to the excessive tar
production in this mode. Tests were conducted on charcoal,
coal, coconut shell, wastewood, rice hulls, and corn cobs.
A single cylinder 5 HP engine was used for these tests. He
cited all fuels other than charcoal as posing a gas
cleaning problem for the scrubber and filter system. This
necessitated that the engine be disassembled and cleaned
after 50 hours of operation. Table 3.2 contains the results
of the trials performed with the 5 HP single cylinder
engine.

Breag and Chittenden (1979) developed plans for a
low cost crossdraft gasifier at the Tropical Products
Institue in London, Egland. The unit had a power output of
3 KW and was intended for stationary applications. They
concluded that biomass gasification systems were a serious

alternative to steam for mechanical power in rural areas of

19

00
Ln
0"
(x

m m mmm.

«3
<3
u»
ox

O

N hmm.

'0
O
m
I‘

N mom.

'0
r-i
\D
N

H mom.

LON
<1"'<i‘

d'm
0
mm

v—ir—i
Q'd‘

(\N
O
Q‘Q‘

NNNH
mmma

QBHH
¢HNH

nmma
HNNH

moma
NHNH

mama
mama

oooa
mwoa

«cm «on 4211]

4111)

4:13 «cm

Hmmmma
mnou :Lou

memwo
naasm mowm

Hmmmwo
Uoo3mumm3

Hmmwmo
Hamnm uncoooo

Hmmmwo
Hmou

Hmmmwo
Hmoopmzu

Uw>mm

Hmnnfin w

mmuamm

cam

mcwmcm

omma

.NDLU "muusom

.wnoz Aammmwov Hmzm mamcwm Amv cam .mp0: Hmsm Hmso A<V cm

memwn Adm mv 3x n.m o no mucmELOLme w>wumLmQEou

N.m manmfi

20
developing nations. Also they contended that the
introduction of gas producers must only occur where there
are trained personnel to operate and service the units.
They emphasized that each application must be considered
individually in order to determine the suitability of a gas
producer for that case.

Breag and Chittenden compared producer gas and steam
power at small scale (<20 kw) outputs. At or below these
levels producer gas had a higher efficiency than steam, 18%
vs 3-5% per brake horsepower hour. Moreover, additional
disadvantages of steam power on a small scale were high
capital outlay, added cost of water treatment devices, and
poor infrastructure for testing and maintaining high
pressure vessels in developing nations.

Stassen and Zjip (1980) designed a downdraft gasifier
at Twente University of Technology in the Netherlands. This
unit was sent to Tanzania for testing under local
conditions. The purpose of the exercise was (1) to evaluate
the feasibility of gasification technology, for diesel
driven corn shellers in rural villages of the Arusha
district, and (2) derive a gasifier design suitable for
manufacture in Tanzania.

The original prototype was designed, built and tested
in the Netherlands. After its arrival in Tanzania, the gas
producer was connected to a Petters PH-Z diesel engine and
used to operate a corn sheller. The unit operated on

charcoal for 200 hours, and achieved a reduction in diesel

21
fuel consumption of 80%. The optimal engine thermal
efficiency was obtained at 1200 rpm. Stassen and Zjip found
that fuel consumption was dependent on the air—gas mixture
ratio, the optimum being 1.05:1. They report that the
engine produced more than adequate power to run the sheller
even at 1,400 meters above sea level.

A tar accumulation problem was attributed to low
quality charcoal. Better charcoal produced a tar free gas
at engine speeds between 1,000 and 1,450 rpm.

The original design underwent modifications to create
simpler and cheaper gasifiers for (l) easier operation by
villagers and (2) easy local fabrication. All necessary
materials were locally available except for refractory
cement. Clay was substituted with partial success.

After extensive operation under village conditions
Zjip and Stassen (1981) implemented further modifications
to the gasification system. Improved gas cleaning equipment
was employed to reduce the high amount of soot reaching the
engine. Cob flow in the producer was enhanced with a
stirring rod and by increasing the bunker diameter, to
eliminate the deterioration of gas quality caused by
bridging. Other changes such as a diesel fuel indicator,
cast iron throat section, large capacity fuel hopper,
elinination of gaskets at the reduction zone, addition of
an ignition pipe for easy ignition of the biomass fuel, and
elinination of the primary air heat exchanger due to

manufacturing complexity. Five villages were scheduled to

22

recieve and operate the modified units for one year.

3.1.5 Corn Cobs as a Gasifier Feedstock

Initial efforts to study corn cobs as a gasifier fuel
were applied to grain drying energy concerns. Payne (1978)
evaluated pyrolysis, gasification, and combustion of corn
cobs as energy substitutes for propane in grain drying.
Gasification was identified as the least costly of the
three methods, though hardware needs improvement for ash
and char elimination. Peart et a1 (1980) built a prototype
gasifier- grain dryer and used the results of tests to
develop an economic model for predicting drying costs of
gasifier powered and conventional grain drying. The
analysis endoresd the corn cob gasifier dryer as economical
if cob handling equipment on the combine and the gasifier
cost less than $30,000. Doering (1978) developed an
elaborate computer simulation model of corn cob
gasification for grain drying, and concluded that the
decision in favor of cob gasification systems would depend
on system cost, the value of the advantage of controlling a
critical fuel supply, in addition to the extra management
required along with safety and convenience considerations.

Richey (1983) fabricated and sucessfully tested an
automatic cob feeding system on a pressurized downdraft
gasifier. Methods for regulating the cob level, eliminating

gas leakage and metering cob flow into the gasifier were

23
incorporated into the design. This system was used to fuel

a dryer for corn.

The particulate emissions of corn cobs in a
rectangular downdraft gasifier were studied by Kutz (1983).
He found a linear relationship between the gasification
rate and the primary air flow rate. The moisture content of
the cobs was identified as an important factor affecting
the gasification rate. Increasing the moisture content
caused a decrease in gasification rate for a constant air
flow rate. This occured as a result of the energy consumed
to evaporate water in the fuel. The range of moisture
contents evaluated were 8.1%, 23.2%, and 32.0% wet basis.

Riggins et al (1980) compared the energy content of
corn cobs and corn stalks at varying maoisture contents for
the cobs, stalks, and kernels. This work showed that the
lower the moisture in the cobs, the higher their energy
content. A cob moisture prediction equation was developed
based on the kernel moisture content. Equilibrium moisture
characteristics of corn cobs are important due to the
dynamic nature of the climate, and the need to keep cob
moisture at acceptable levels. High humidity will cause
adsorption of atmospheric water, and low humidity will
cause desorption. These phenomena were the subject of a
study by White et a1 (1984) on equilibrium moisture content
and equilibrium relative humidity of corn cobs. The

experiments were conducted at tempertaures of 10, 30, and

24

50°C, while the moisture contents involved five levels in a
range from 5.7%-27.5% wet basis. White compared the
Henderson and Chung equations for their predictive behavior
in regard to EMC and ERH. Both methods were found to
adequately represent the experimental data, though standard
errors were variable with both methods. White recommended
selecting the method that best fits an application and had
the lowest standard error. White derived desorption
isotherms for corn cobs and that data is presented in
figure 3.4 and 3.5.

Silva et al (1984) researched tar formation in
downdraft and updraft gasifiers using corn cobs as a fuel
source. A major conclusion of this research was that tar
formation increased with increasing moisture content. The
fuel was evaluated at 3 moisture levels; 8, 16, and 24
percent wet basis. Secondly, 3 air flow rates were analyzed
along with moisture content. They were 16, 32, and 48
liters per minute. The outcome of this study is graphically
displayed in figure 3.6.

The gasification characteristics of 30 residue derived
fuels used in downdraft gasifiers were presented in Jenkins
(1980). His work focused on the variability in fuel
quality that a gas producer would tolerate and still
perform satisfactorily. The parameters of variability
involved fuel form, fuel ash content, air flow rate, and
fuel moisture content. The results he obtained stated that

corn cobs of a 37 mm (1.5 inch) nominal size with low

25

  
 
  

  

 

 

I00 1—

a i-

3 90 - DESORPTION VI ABSORPTION ISOTIIEIIMS

¢ _ MODIFIED IIENDEI‘ISON EQUATION

n PREDICYEO EMC IWET BASISI

5 '° ‘ TEMPenAtunE— 30-1:

3 .

w

I-

2

B CO ‘

g i.

It.) 50 -

m i-

E

g.) 40 -

O

I

2 30 -

2

n:

m oesonPnon

Q 20 -

:3

8
'0- / Ansonrnou
0- LI #1 l :4 I_1I_1i_gi._1_.l_1_i_-_1_l

0 IO 20 30 40 5") GI) 70 Ii!) 90 ")0

EQUILIBRIUM RELATIVE HUMIDITY (7.)

Figure 3.5 Difference in Net Basis Desorption and Adsnrpiinn
Isotherms at 30 C When Predictinu FHC with Hudilivd
Henderson Equation. Source: White Pt nl, thh

$5)

90‘ PnenunEo nEsnnrrnniIsnrnrnms
MOUHHED HENDERSON Ennninni
WET OASIS MOISIUIIE couIEnr

Q
A

00-

70-

(1'. war a

IT

Vt. I --

60'

-gn-‘l A‘._
I 93": C

40-

30-

20-

UILIBRIUM .‘ciCIS

q
d

E

 

 

011.‘|I|4_l;1111_111,_41,_1__i

O . i0 20 30 40 50 60 70 nn 90 '00
EQUILIBRIUM RELATIVE HUMIDITY (7.1

Figure 3.4 Not Basis Desorption Isotherms [nrm Mndifind Henderan
Equation When Predicting EMC. Source: White et ni, i984

«mod .m>HHm "muuaom .cowumUHMHmno LOUCuou Eoum umH o.m wuswfim

nouDCmE\uLoum~u sofim LMO

 

26

 

 

 

 

 

 

 

 

 

 

 

av mm mm a
— d u d _ a aaoa
I avfiu
.....w ,
9.5.5.2... .............. ....... 1 8a
4 ........ fig. ..
..... ......o. no. . ... a... Q I SN.—
6000 0..
I an.”
I and“
Im— .
El
6! L
I mad"
4. ton? . . . 1
m3 .
.o‘NvN4 I aNm
tool name 1
Z— 3‘nt I anfi
COPU‘WC Omh<n=um Dang ..
. — . _ . _

 

as .v

ooomogq omoas gar/Jog omoqs

27

percentage of fines was a reasonably good fuel form for
gasification. This judgement was made on a criteria that
considers the tendency to "bridge” in the gasifier (poor
flow characteristics that can cause large voids in the fuel
hopper, and inhibit the flow of fuel into the firebox), the
magnitude of pressure drops across the fuel bed (low being
desirable), the quality of the gas over the duration of the
test, and the bulk density refueling interval associated
with the fuel type. Wood chips less than or equal to 5 cm
square were rated excellent, as were fruit pits. Fine
particles, small pellets, walnut shell and low density
cubes and pellets of barley straw, cotton gin trash, and
cotton stalks were rated poor.

Ash content was important because the ash contains
mineral matter (such as silicon) which could melt and then
form slag (fused ash) within the firebox of the gasifier.
This occurence restricted fuel flow and caused a decrease
in gas quality. Jenkins reports the fuels with ash contents
lower than 6.0% by weight, did not form slag in the
gasifier. Corn cob tests indicated an ash content of only
1.5% and no evidence of slagging.Input air flow rate tests
were only reported for peach pits. The results however are
applicable to any fuel used in a gasifier. Jenkins found
that varying the air flow rate did not alter the air to
fuel ratio (AFR) significantly. Increasing air flow rate by
125% increased AFR by 6.7%. Hot and cold gas efficiency

increased slightly and reaction temperatures in the

28
oxidation and reduction zones increased. The net heating
value of the gas dropped slightly at increased flow rates
while the power density (power output of cold gas divided
by the fire box volume) increased substantially.

In regard to fuel moisture content, Jenkins work
indicated that increasing fuel moisture content lead to an
increase in the gas moisture content in addition to a drop
in thermal efficiency, reduced firebox temperature, and
reduced gas quality. The decrease in thermal efficiency
occured due to the energy expended to vaporize the higher
volume of water in the fuel. Since this energy is not
recovered, efficiency decreases compared to lower moisture
conditions. Water vaporization and poor ignition of wet
fuel were responsible for the firebox temperature drop.
Lower temperatures in the gasifier altered the gas

composition and a less "burnable" gas resulted.
3.2 Economics of Biomass Gasification
3.2.1 Applications Meriting the Gasifier Alternative

From their conception through WW II, gas producers
were widely used to operate stationary and mobile engines
(Hiscox, 1919; Levin, 1910). In developing nations, the
often numerous stationary engines used for grain
processing, electrical power generation, and irrigation
water pumping, promote interest in gasification as a

potential fuel system. Barnard (1984) conducted an

29

assessment for the World Bank and the United Nations
Develpoment Program (UNDP). to identify applications in
develpoing nations where gasification technology is
feasible economic and suited to local needs. He concluded
that enterprises and agencies in rural areas already
possessing diesel engines, are the principal candidates for
the reception of gasification technology. Such a system
would be particularly suited to remote areas where
electrical grids are non—existant and diesel and gasoline
power are the typical options. Barnard reports that remote
locations that experience interruptions in fossil fuel
supplies might warrant high cost equipment to achieve fuel
security.

The principal requirements cited for determining the

suitability of a gas generation system were:

1. That the power demand is fairly steady

2. Qualified operators are available

3. A supply of biomass is available
Irrigation was one of the most frequent applications of
diesel power in the Sahel. Boutillier (1980) reported that
nearly 24,000 hectares of land (table 3.3) are irrigated in
the Senegal river valley. This river supplies water to
Senegal, Mali, and Mauritania. About 4,000 hectares (10,325
acres) of the total are small irrigation systems of 40
hectares or less. All of the area was fed with water

lifted by diesel driven pumps.

30

Hmuoe canto

mumuwewnmm

SHOE tantuflm

mumumeflumd HHmEm mwxmx

HHmEm cam mmuma Homuoo
Lmumefiumd “coax
mumumefiuwm Macaw mnmom
meow aux.m
mumbmewomm HHmEm muawo
mflm20d.z

Hoxmm
Emumz
Locom
mmcmfiz
......Lmuwefiumd mammma
.mesm Hmmmcmm

.....OOOOOOOOOO

o o o o o o o o o o .Whmuméwhmm MUHwQ

Hmpoe ham

Hmuoe 22m

mwcmbfiusmz

ammocmm

on»

:1

Houucou

ome

.Loflaafluaom
Looms Hasm :uw3 cofiummLLLH Lona: mmu< ccmq

"muuaom

.mmdam> Lm>wm Hmmmcmm
m.

m manme

31
Groenveld and Westerterp (1980) monitored the
performance of a producer gas powered corn sheller and mill
in Tanzania. Corn cobs were part of the output of the
sheller and were used as an energy input to the engine that
operates the sheller/mill. Mill operators were business
people who owned their equipment and provided a service to

the public.
3.2.2 Financial Cost
3.2.2.1 System Cost

The capital cost of the gas generation system is an
important variable affecting the outcome of the financial
analysis. Van den Aarsen (1984) evaluated the effects of
capital and operating costs on the economical feasibility,
of small power plants using gasification as a fuel system.
His study revealed that large systems have favorable
economies of scale over small systems. The higher the power
generating capcity the lower the capital cost per unit of
power. That data included the cost of the engine.

The problem of comparability between different systems
and the lack of a broad set of cost data handicaps any
general assessment of economic feasibility of gas
generators. In spite of this, progress has been made in
obtaining a broad view of the suitability of gasifier
technology in developing nations. Barnard (et a1) collected

information from manufacturers on system costs.

32
Considerable variation in capital cost was evident, and for
this reason the data was divided into 3 cost ranges; low,
medium and high. From these he selected an intermediate
value in each range and defined this as the "typical" or

average cost for the range. The following are his results:

~Case 1. Low Cost Gasifier $75/KW
Case 2. Medium Cost Gasifier $200/KW
Case 3. High Cost Gasifier $800/KW
The low cost option represents a scenario in which a
nation has significant manufacturing and servicing
resources. Mass production of equipment would lower costs
to the indicated level. Nations in this category would be
India, Brazil, and the Philippines. The medium cost option
is considered to be typical of custom made systems for wood
or charcoal. The high cost option refers to the lowest cost
wood/charcoal gasifiers manufactured in industrialized
countries and imported to developing nations. These costs

do not include the diesel engine.
3.2.2.2 Maintenance and Labor Cost

Barnard (et a1) employs a figure of 10 percent of the
capital cost as maintenance cost and adds a lubricant cost
of twice that of a diesel. Van den Aarsen (1984) uses a
rate of 5-6 percent of capital cost for maintenance. No
provision was made for lubricant costs in his work.

Groenveld et a1 indicates maintenance cost are 14 percent

33
higher for the gasification system than without it, for the
Tanzanian system.

Additional labor cost of $1,000/year for operating
gasifier systems was cited by Barnard. Van den Aarsen shows
a higher figure of $4,330/year for systems in the range 20
kva to 150 kva. Groenvelds' study reports that labor costs

are the same with gasification and without.
3.2.2.3 Diesel Fuel Cost

The price of fossil fuels has a direct impact on the
economics of energy systems. The OPEC oil price increases
of the mid 1970's inspired the interest in alternative
fuels that followed. Barnard (et a1) had performed
sensitivity analysis on diesel fuel price versus overall
power cost. His findings are presented in figure 3.7.

The 1985 price of diesel fuel in Mali was 213 francs
per liter ($1 = 325 francs). Prices for petroleum products
in Senegal, Mali, and Mauritania were controlled by their

respective governments.
3.2.2.4 Biomass Feedstock Costs

The biomass fuel source under consideration as a
gasifier feedstock was corn cobs. Corn was grown under
irrigation in the Senegal river valley area. Manteuffel and
Tyner (1980) have reported yields of 3 tons per hectare.
Wheat and rice are also grown but rice is the predominent

crop of the three though it is grown in the wet season,

34

 

40'-

   
  

 

 

 

? Cepiiei Coet

a 30 P M Gaeiiiet

3 (s/uw) -——

z — “ --—*

0 d“’ ‘0

o - ”...-fl

3 2°. -‘,-«' ’,-«*’

‘3 400 —“' " -..-w"

8 ”’

200 - ‘
75
I0 -
Oeeiiiet Syeiem _ - - ..
Dieeei Syetem —-—--
20 40 60 60
Diesel Price (Millie)
Figure 3.7 Effect of Diesel Fuel Price on Unsiiivr hcnnumics

Source: Barnard, 1984

35
while corn and wheat are grown in the dry season. Corn cobs
rank as a higher quality feedstock for gasification than
either wheat or rice residues (Jenkins, 1980). Table 3.4
characterizes fuel suitability for power.

Based on the authors experience in the area, corn
residues were employed as a feed input to livestock
production, but not in a formal manner. That is, cobs were
not gathered, stored, and fed to livestock to any
significant degree. Rather the cobs were thrown out after
the shelling operation and grazing animals may eat them
when a better source of forage is not immediately
available. Livestock typically graze on grass, low growing
vegetation and standing crop residue remaining in the
fields after harvest. Given this situation, the cost of
corn cobs is assumed to be zero since little if any
transportation is necessary and alternative uses were not
significant.

A recent development in gasifier designs is the open
top or stratified gasifier. As the name implies, the top of
the gasifier is open, exposing the fuel bed. Air for the
reaction in the firebox is drawn through the top of the
unit and through the fuel bed. The advantages of this
design are that (l) the simpler design permits lower
construction cost, (2) the open top allows continous
feeding of biomass into the furnace, (3) monitoring of the
fuel level and condition are easier for the operator.

Tests conducted by Eoff (1985) produced data which

36

Table 3.4 Behavior Ranking of Various Fuel Forms
Source: Jenkins, 1980.

 

Rank

Fuel Form

 

Excellent

Fruit pits

khole log chips of uniform size
distribution 5 2 inches square

 

Good

Blocks (cut up trim ends from
saw boards. etc.) approxi-
mately 1" cubes

Hard. durable cubes (cubed alf-
alfa seed straw, corn stalks)
approximately 1 1/4" x 1 1/4"
x I III"

Cracked nut shells (walnut,
almond shells)

Hammernilled corn cobs (i ll!"
nominal size and low fraction
of fines)

Hard durable pellets 2 3/8"
diameter (rice hulls)

 

Poor

 

Loose, low density cubes or
pellets (barley straw, cotton
gin trash. cotton stalks, pel-
leted RDF)

Fine pellets (1/4" walnut pel-
lets)

Fine particles
flogged wood chips with length

to diameter ratio 2 2 or large
fraction of fines.

 

37
described the minimum and maximum power outputs for a range
of open tOp gas generator dimensions, using wood as a

feedstock.

CHAPTER 4

4.1 EQUIPMENT
4.1.1 Gasfication System Description

The open top stratified downdraft gasifier was of a
cylindrical shape. It measured 20 cm outside diameter by 50
cm high. The interior walls were lined by 2 cm thick
asbestos insulation. The fuel hopper was 10 cm in diameter,
had a grate area of 45.6 cm2, that was positioned
vertically in the interior chamber of the unit (figure
4.1).

The gas exit was at the side of the unit through a
6.2 cm round mild steel pipe, and into a 7.5 x 3.5 cm
rectangular mild steel section. The gas then entered a
cyclone separator which removed large dust and ash
particles from the gas stream. The next stage of gas
conditioning was via a polyester cloth filter that removed
the smaller particles from the gas stream. The gas then
proceded to a cooler which was a 7.5 cm diameter X 12 meter
nylon hose, immersed in a water bath. A 55 gallon oil drum
contained the water for the bath. The hose ultimaetly was
connected to the engine via a tee-type mixing valve, which

had a throttle plate for metering air.

38

39

C088 AND AIR IN

111

FUEL HOPPER
INNER CASING

OUTER CASING

INSULATION

FIRE BOX

..., GAS our
CRATE

    
 

ASH BIN

Figure 4.1 Open Top Gasifier Components

40.

4.1.2 Engine Characteristics

The engine used in the experiment was a single
cylinder Shp Stover pre combustion type diesel. It was
water cooled, had a manual start and a maximum speed of
1,000 rpm. Stock performance specifications are cited in
the Appendix. The Bosch injection pump was not modified to
control the pilot amount of fuel injected for ignition of
the gas. Instead the governing mechanism was used to
control the quantity of fuel delivered to the combustion
chamber. This permitted a quantity of fuel sufficient for
idle speed operation and ignition of the producer gas, to
be delivered to the combustion chamber at all times. When
the engine was running at speed and producer gas was
introduced to the engine, the combined energy of the two
fuels caused the engine speed to increase. The governor
reacted to the higher speed by diminishing the amount of
diesel fuel injected into the combustion chamber, reducing
the engine speed to the previous level. Figure 4.2 shows

the layout of the gasifier and engine system.
4.1.3 Analog to Digital Data Conversion Hardware

The above sensors fed analog voltage signals to the
Starbuck 8232 data acquisition and control unit. The unit
had 8 analog inputs, 8 digital inputs, and 8 digital

outputs. This study uses 5 analog inputs as follows:

41

unoEuwcouu< unam:m\amumzm uuwuumoo N.¢ uuowfim

       

   

qumJOUOEENIL.

t \ /

Q
Some 33

a.

    

        
 
    

mgrUZm
m3 34> QSXSQ # a
m<0\ We? Mud 4000 _
W<U Mullen—“Nut MMEGYU

WT

 

Mob 9.2 twig.

 

42

Channel 1 - magnetic pickup

" 2 - Load cell

" 3 - Linear thermistor 0-100°C
" 4 — K- type thermocouple

" 5 - 2nd K-type thermocouple

The unit required a power supply of 9 VDC at 100 MA. This
was provided by a 115 VAC to 9 VDC transformer.

The Starbuck 8232 analog to digital (A/D) and digital
to analog (D/A) converter, produced an 8 bit digital signal
from an analog voltage. This process was called
quantization.

The quantization operation involved replacing the
exact value of the analog signal with one of the quantized
values. A signal was quantized by dividing the amplitude
axis into levels which were assigned "N" numbers 1e 0-255.
When the interval between levels was defined, quantization
could be carried out. The interval between levels was
called a quantum (q). If the amplitude of an input signal
vector lies within the interval (nq,(n+1)q) its value was
obtained by the quantity nq.

Accuracy of the quantized signal depended on the size
of the quantum. The smaller the quantum the higher the
accuracy. Error was introduced in the quantization process
since an exact input signal value was replaced by an
approximate one. This occured because a signal range could

theoretically divided into an infinite number of values,

43
while quantization imposes a finite number of values for
the same signal range. For digital information given in n
bits to the base b, and a maximum input signal amplitude

given by E, the quantum level is:

The full scale analog input votltage was 5.33 volts
dc. With n = 8 bits and b = base 2 (digital signals) the
peak digital signal was equivalent to a decimal value of
255 (bn). Thus the 5.33 volt range was divided (quantized)
into 256 equal parts. Each part represents 0.02 volts dc
(quantum) of the 5.33 volt range. This value was also
referred to as the resolution of the converter since 0.02
volts dc was the smallest analog signal difference the
converter can detect. The quantization error for the
Starbuck 8232 was plus or minus one least significant bit
(LSB).

Initially data points were held in the on board Random
Access Memory (RAM) of the A/D converter. The memory
capacity of the unit was 1,975 data points. Calibration
data and raw treatment data were downloaded to the Radio

Shack Model 100 upon completion of each treatment run. The

44
data was then transferred to magnetic cassette tape for
permanent storage. The Model 100 had 32 kilobytes of RAM,
27 of which were available for temporary data storage.

The Model 100 contains a built in modem with software
selectable communications parameters. It was connected to
the A/D converter via serial type RS—232c connectors and
standard ribbon type cable. The A/D connverter was capable
of communicating to Data Terminal Equipment (DTE) or Data
Communications Equipment (DCE). The DTE mode was used with
the Model 100. No special hardware or equipment was
necessary to interface the Model 100 to the A/D converter.

The data acquisition hardware is represented in figure 4.3.
4.1.4 Engine Horsepower Measurement

Engine horespower is a function of enigine speed and
applied load. When a prony brake was used to measure
horespower, the length of the lever arm multiplies the load
and was considered in calculating horespower. The

expression for horsepower is:

33,000 ft-lb/min

where: L = the length of the prony brake arm (feet)
F = the force measured at the tip of the prony
brake (lbs)
N = the speed of the engine in rpm

45

 

 

 

REED RELAY

:3 54.7. l__‘

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

LOAD CELL. [
.I:l AMP ANALOG [:3
THERMIST'OR

CONVERTER I._____I

 

 

 

 

 

 

 

 

 

 

H

103 l
l‘fll
WERMOCZUPLE 1 Egg——
THER ; PL ”m PORTABLE CASSETTE
mac u E z 0,,
5r @ I COMPUTER RECORDER

 

 

 

 

 

 

Figure 4.3 Digital Data Acquisition Syscem

BRAKE PADS

J l

invert ARM A

O

 

 

 

LOAD CELL—*- L
=— BRAKE DRUM

Figure 4.4 The Prony Brake Mechanism

 

 

 

) [—REED RELAY PICK'UP
ENGINE atom—.- l £51K”:

 

 

é—FLYWHEEL

 

CRANKSHAFT—v

 

 

 

 

 

 

 

 

 

 

Figure 4.5 Reed Relay Pick-up Pncirinn

46

In these experiments a reed relay pick-up and triggering
magnet was used to obtain readings on engine speed. The
relay switch was rigidly mounted to the engine block and a
single 12 mm diameter permanent magnet was glued to the
inside face of the engine flywheel (figure 4.5). A voltage
source was connected to the leads of the normally open
relay. As the flywheel rotated the magnet passed the relay
causing it to momentarily close. This closure allowed
current to flow in the relay circuit. The rapid Opening and
closing of the relay created countable pulses that
correspond to a frequency in cycles per second (Hertz). One
pulse was generated per revolution of the crankshaft with
one magnet. If for example 10 revolutions occured per
second 10 pulses were generated each second. The engine
speed in revolutions per minute (RPM) was:

10 revs 60 sec
Engine Speed = ------- X ---f-- = 600 RPM

The frequency was then translated into engine speed
electronically via a frequency to voltage converter.
Basically, the frequency to voltage converter was a signal
processing circuit that produces an analog DC voltage
output proportional to the input frequency. The higher the
frequency the higher the output voltage. The frequency
voltage circuit for this experiment produced 90-100 mv per
100 rpm change in speed in the range 400 to 600 rpm.

The overall accuracy of the engine speed detection

47
system was a function of the accuracy of the calibration
instrument and the accuracy of the measurement system. The
A/D converter had an accuracy of :1 least significant bit.
With a full scale (FS) input range of 5.33 volts dc, 1 bit
was equivalent to 0.0208 volts. Thus the accuracy of the
A/D converter was :0.0208 volts at 5.33 volts PS. The
strobe tachometer had scale graduations of 20 rpm. Readings
could be made to the nearest 10 rpm ie 1/2 graduation. The
highest calibration factor encountered was 174.9 rpm/volt.

The maximum accuracy then:

(5.33 + 0.0208) x (174.9 + 10) - (5.33 x 174.9)
----------------------------------------------- = 0.06136
5.33 x 174.9
or 6.136%.

The prony brake was made of a cast iron drum 35 cm in
diameter with a contact surface was 12.7 cm wide. The lever
arm was 75 cm long and was attached to a brake! shoe
composed of a steel strap carrier with 19 brake pads. The
brake pads were simple wooden blocks 4.0 x 2.0 x 11.0 cm. A
threaded T-handle rod clamped the arm and shoe assembly to
the drum. As the engine turned the drum, the drum tried to
turn the lever arm. The brake pads permitted some slippage
allowing the lever arm to remain stationary, but transmited
some force from the rotating drum. A weighing scale or
other force measuring device was placed under the tip of

the lever arm, to obtain the magnitude of the force

transmitted by the prony brake mechanism. The brake

48

mechanism is diagramed in figure 4.4. A Daytronics model
152A-25 1b LVDT load cell was employed to sense the load
transmitted by the prony brake. The LVDT (linear varibale
differential transformer) load cell was a force transducer.
This device produced an electrical signal (voltage)
proportional to the applied force. It functioned on the
principles of cantilever beam deflection. When the beam
within the device was loaded, a deflection of the beam
occured which was proportional to the load (force). The
deflection of the beam displaced the armature of the
variable differential transformer. As the armatures postion
changed, the distribution of the magnetic flux field
changed. This change in magnetic flux altered the voltage
induced in the transformer. In this manner each load had an
associated deflection and output voltage (Seippel 1983).
The LVDT was supplied with an input AC voltage as a source
of elecrical energy.

The load cell was connected to a Daytronics Model 300
signal transformer/amplifier. The amplifier amplified the
analog voltage originating from the load cell. This signal
was calibrated to produce an output equal to 0.5 volts DC
per pound. The output of the amplifier/transformer was
directed to the analog to digital converter for futher
processing. These readings were required to determine the
horspower produced by the engine.

The accuracy of the force measuring system was

accounted for by the accuracy of the A/D converter and the

49
accuracy of the weights used as reference loads. The
accuray for the A/D converter was :0.0208 volts in 5.33
volts F8. The reference weights had an error of 12 grams in
4,540 grams. The calibration factor for the load cell was
0.84 kg/volt. The maximum accuracy for the force measuring

system was equal to:

(5.3508 x (0.84 + 0.012)) - (5.33 x 0.84)
----------------------------------------- = 0.0182 or 1.82%

5.33 x 0.84

4.1.5 Gas Temperature Measurement Instrumentation

Temperature information was needed from the fire box
zone of the gasifier to monitor the effect fuel moisture
content had on the reaction temperature. Temperatures
inside the gasifier unit typically reached peak of around
1,000°C. A sensor for this application must tolerate
oxidizing and reducing environments, and withstand high
temperatures for extended periods of time. Response time
was not a critical factor because the temperature levels
were relatively stable for all treatments.

In the gas circuit between the gasifier and the gas
cooler, temperatures in the 300-400°C range occured. A
sensor for this application must have characteristics
similar to those indicated above though toleration of
oxidizing and reducing environments was not important here.

For both of the above applications temperature sensors of

50
the thermocouple type K were suitable.

A thermocouple was a thermo-electrical device that
converts temperature to a corresponding electrical signal.
The principal of operation behind the thermocouple involves
the Seebeck Effect (discovered by Thomas J. Seebeck). Two
wires, each composed of dissimilar metals, were fused
together on both ends forming a loop. When one of the
connections (junction) was heated, an electrical current
flows form one wire to the other (Sippel 1983).
Thermocouples produce electrical signals in millivolts. The
higher the temperature the larger the millivolt signal.
Downstream from the gas cooler temperatures will be much
lower, in the 30-70°C range. The sensor must withstand
temperature peaks up to 100°C and be capable of functioning
at 70°C for extended periods. In addition the sensor must
penetrate a closed pipe to sense the temperature of the gas
within. It must therefore be designed in a way that will
facilitate mounting and positioning it in a pipe, and also
be resistant to the chemical composition of the gas. Given
these requirements a thermistor type sensor with stainless
steel sheath and a compression type fitting were most
appropriate.

A thermistor was a thermoelectric device fabricated
from semiconductor materials such as oxides of cobalt,
manganese, and nickel. This device had a characteristic
behavior such that as temperature increased, its resistance

(to electric current) decreased. This was referred to as

51

the Faraday Effect (Seippel,l983), and the thermistor had a
negative temperature coefficient since resistance decreased
as temperature increased. Temperature information was
obtained by exciting the device with a power supply such as
a small battery and measuring the voltage that exits the
thermistor at various temperatures. Since the thermistor
had a negative temperature coefficient, the voltage at
relatively low temperatures was high, and was lower at
higher temperatures.

Temperature sensors were used at 3 points in the

gasifier system (figure 4.2):

1. At the firebox within the gasifier
2. At the gas cooler inlet

3. At the gas cooler outlet

At the firebox and the gas cooler inlet Omega Type R
thermocouples were used. The firebox was fitted with a 14
gage K type element with an Omegatite 300 1/2" O.D. X 5/16"
I.D. X 12" long ceramic protection tube. The thermocuple
assembly was inserted through the bottom of the gasifier
and protrudes upward stopping just below the grate of the
firebox. This permitted the sensing of the temperature of
the incandesent fuel in the fire box.

The cooler inlet was fitted with a stainless steel
protected K-type thermocouple for extruders No. CF-090-K-4-
60-1 also from Omega emgineering. The compression fitting

on the thermocouple assembly attached to a corresponding

52
riser on the inlet pipe. The depth was adjusted to place

the tip of the thermocouple in the center of the gas
stream.

Each thermocouple was connected to an Omega MCJ-K
electronic ice point with a temperature reference setting
of 0°C. K-type thermocouple wire was employed to extend the
connections to the electronic ice points. The electronic
ice point effectively replaced the external reference
junction needed for measuring the output voltage of
thermocouples and therefore temperature. The electronic ice
point converted thermocouple outputs to 0°C referenced
voltages. In this way the voltage reading from the
thermocouple corresponded to a temperature based on 0°C as
a starting point. The electronic ice point was fully
compensated so copper wire could be used to connect the ice
point to the data acquisition unit. Slightly downstream
from the cooler a linear output thermistor was used to
sense the cooled gas temperature just prior to entry into
the engine.

As with the other instruments, the accuracy of these
devices depended on that of the reference instruments and
of the A/D converter. The referfence tables were used for
the thermocouple temperatures. These tables were provided
by the manufacturer of the devices, and correspond to
standrads established by the American National Standards
Institute (ANSI). The error limit cited for type K

thermocouples was i 2.2°C in the range 0 - 1250°C. The

53
maximum calibration factor for the high temperature
thermocouple was 176.98°C/volt. Maximum error for this
device then was 1.64% using the same procedure as above.

The accuracy for the low temperature thermocouple
(channel 5) was likewise calculated. The calibration factor
in this instance was 89.24°C/volt. The resulting maximum
error value was :2.865%.

In the case of the thermistor (channel 3), a mercury
filled glass thermometer was used as a reference. The
thermometer had graduations of 1°C. Readings were possible
to 0.5°C. The largest calibration factor attained was

6.62°C/volt. The maximum accuracy in this case was 7.97%
4.1.6 Data Acquisition Software

The Starbuck 8232 A/D and D/A converter contained
instruction sets for: (1) selecting sampling rates, (2)
allocating RAM to each channel, (3) starting the
acquisition of data, (4) displaying the acquired data, (5)
displaying the channel pre-set status and (6) clearing the
channel settings. These instructions were combined with
BASIC language commands from the Radio Shack Model 100 , in
this way, instructions were sent to the A/D converter using
a BASIC language program in the Model 100.

When the system was operating, the sensors,
transducers and their signal conditioning circuitry,
provided continuous dc voltage input signals to the

Starbuck 8232 A/D connverter. This input signal was not

54

continuously recorded by the A/D converter, samples were
taken at regular intervals, and the average of these
samples was taken as the reading for the treatment. The
sampling rate was an important factor in the precision of
data acquisition. If the dimension being sensed behaved or
responded in a non-steady state manner, a low sampling rate
may miss important data points during the interval period
between samples. At the other extreme, very high sampling
rates generated large data sets, much of which was
redundant and required massive memory space in the
equipment that holds the data. The ideal rate was one which
provides the necessary information and was compatible with
the available data storage devices.

There were two channels in the system that exhibited
signal fluctuation warranting relatively high sampling
rates; channel 1 (reed relay pick-up) and channel 2 (load
cell). Trials were run at rates of 2 Hz, 1 Hz, 0.5 Hz, and
0.25 Hz. These trials indicated that 0.5 Hz provided
sufficient information.

The sampling rate for channels 1 and 2 was 2 seconds.
The interval for channels 3 and 4 was 15 seconds, and
channel 5 had an interval of 5 seconds.The treatment run
period was 10 minutes giving 300 data points for each of
the first two channels, 40 data points per channel for
channels 3 and 4, and 120 data points for channel 5. The
total number of data points collected per treatment was

800.

55

Configuration of the A/D converter for the sampling
rate, channel memory, and the "start acquisition” command
was accomplished with a BASIC language program. This
initiated data acquisition within the A/D converter and
acquisition ceased when the channel memory area was full.
When all channel memories were full another routine
downloads the raw data from the A/D converter to the Model
100. The raw data was held in a RAM data file for
subsequent storage on magnetic cassette tape. The raw data
was then routed to a subroutine that determined the mean
and the standard deviation for the treatment. These were
also held in a RAM data file for later storage on magnetic
tape.

When all treatments had been executed an analysis of
variance program was used to calculate the test statistic
for the treatment classes. Flow charts of the configuration
,program, and downloading program are presented in figures

4.6 and 4.7.
4.2 Sensor and Instrument Calibration

Calibration was the process of indexing the output of
measurement devices to physical quantities that were of
interest to the investigator. Without calibration one
cannot be certain of the magnitude of a particular
dimension, the sensors output represents.

Since all measurement devices in this experiment

produced direct current voltage signals in the range 0-5.0

56

/ SET CHANNEL
/ MEMORYSIZE

SET—ACQUISITION /
/ RATE ’CH (€43--.

SEE/450019770117
[ RATE CH3'5

[.3 Mktacguié 2 110M

 

 

 

 

 

 

 

 

Figure 4.6 Flowchart for A/D Converter Channel Sampling
Parameters

57

* OPEN INPU T/OO TPUT
COMMUNICATION LINES

 

 

 

 

F/X ARRAYS/25
FOR CHANNEL DATA

_ l
. GET DATA POINT
FROM A/D CONVERT.

1
MOVE DATA INTO ARRAY

 

 

 

 

 

 

 

    
     
 

NO LAST
DATA POINT
?

 

 

. C‘

 

GET DAT/l POINT
FRO/VI DAT/l ARR/l7

I
SEND DATA TO
RAM DATA FILE

 

 

 

 

 

 

 

 

 

 

 

CLOSE I/O LINES

Figure 4.7 Flowchart of the Downloading Program to Transfer

Data from the A/D Converter to the Portable
Computer

 

 

58
volts, the real world physical dimensions corresponded to a

voltage in this range.
4.2.1 Engine Speed

The engine speed in revolutions per minute (RPM) was
provided by a reed relay pick-up and triggering magnet. The
output pulse signal was conditioned, counted and converted
to a voltage by electronic circuitry. The circuit was
referred to as a frequency to voltage converter.

To determine the output voltage at various speeds, an
analog strobe type tachometer was employed. This device
operated on reflected light signals recieved from
reflective tape placed on the rotating flywheel of the
engine. In essence the number of reflections recieved in a
given time period were directly converted to speed.

Two pieces of reflective tape were placed equidistant
on the flywheel to provide a more accurate reading of the
engine speed. With this arrangement the strobe tach reading
was divided by 2 to obtain the actual engine speed.Readings
were taken in steps of 100 rpm beginning with 400 rpm and
extending up to 600 rpm. This process was repeated three
times and the values averaged to obtain an output voltage
at the given speed.

Table 4.1 shows the data points taken during the
calibration runs. Readings were repeated three times and

averages calculated for each speed.

59

TABLE 4.1 ENGINE SPEED CALIBRATION DATA

ENGINE SPEED REP l REP 2 REP 3 MEAN
400 3 25 3.23 3.215 3.23
500 3 35 3.35 3 33 3.34
600 3.44 3.43 3.42 3 43

A linear regression equation was derived from the
averages and used to produce a calibration curve for engine
speed. The curve is plotted in figure 4.8. The regression
equation was employed in the data conversion and analysis
computer program for horespower determination. The equation
had a correlation coefficient (r2) of 0.997 indicating

strong linear relationships.
4.2.2 Load Cell Calibration

The load cell senses the force exerted by the engine
through the prony brake. This device was of the linear
variable differential transformer type (LVDT). The output
was linear for loads within its operating range of 0—11.34
kg (0-25 lbs).

Calibration was performed using weights in 454 gram
increments in the range 454 grams to 4,540 grams (1 lb to
10 lbs). The load cell and amplifier were allowed to warm
up for 30 minutes prior to calibration. The load cell was
fitted with a rubbing block for the prony brake arm, before
being adjusted to zero output. The load cell was

subsequently loaded to 4,540 grams in steps of 454 grams.

FF.)

(R

”:3

ENGINE SP

3C5

na- Ina-—
L... f"—

' JCS

Pf. b
we.

60

 

'mL
Gau— '
5...; ~ /
«aa-

300 L-

 

l l

—I

 

 

l
3. 00 3. l0 3. 20 3. 30 3. 40

UUII’UU VUL1 AGE (UL‘)

Figure 4.8 Engine Speed Calibration Curve

Data points from Table 4.1

 

lawn , , , , , . r ,___'

0. 00

0. 00 -

 

l ___l .1 l l l l l

 

 

T—

l

3. 6‘,

M,---——

 

i

 

.0" ‘
0 0.00 .50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50

OUTPUT VULIAGE (UC VULlS)

P

J.

EU

5. 50 U. m!

Figure 4.9 Load Cell Calibration Curve - using

data points from Table 4.1

Y=o+bX

a= -.QH?E*HU4
b= 9.9U7E4UHZ

.997

R Square”

Y"(l'bx

n” U.HllL-UHI
b- 1.731Lvuun

N Shrunrv” LUUU

61
Then it was unloaded in the same manner to reveal if any
hysterisis effects were present. The process was repeated 2
times for loading and unloading the load cell.

The values obtained were averaged and used to fit a
linear regression curve. The raw data points from the
calibration exercise are in tables 4.2 and 4.3. The
calibration curve is presented in figure 4.9. The
regression equation for the load cell data was contained in
the computer program for horse power determination. As
expected the data exhibited extremely high linearity as

judged by the correlation coefficient of 1.000.

TABLE 4.2 LOAD CELL CALIBRATION DATA - INCREASING FORCE

FORCE REP l REP 2 MEAN

GRAMS
454 0 22 0.23 0 225
908 0 75 0.77 0 760
1362 1.32 - 1.34 1.330
1816 1.88 1.92 1 900
2270 2.50 2.50 2 500
2724 3 08 3.10 3 090
3178 3.67 3.68 3 675
3632 4.26 4.26 4 260
4086 4.84 4 85 4.845

62

TABLE 4.3 LOAD CELL CALIBRATION DATA - DECREASING FORCE

FORCE REP 1 REP 2 MEAN
IN GRAMS

4540 5.42 5.42 5 420
4086 4.84 4.83 4 835
3632 4.25 4 25 4.250
3178 3.66 3 66 3 660
2724 3.08 3.07 3 075
2270 2.49 2.48 2.485
1816 1.92 1.90 1.910
1362 1.34 1.32 1.330
908 0.77 0.75 0 760
454 0 24 0.22 0 230

4.2.3 Thermistor Calibration

The thermistor was a temperature sensing device. In
this application it has sensed gas temperatures in the 30-
50°C range as the gas exits the cooler. The thermistor
probe was calibrated using a mercury filled thermometer
with 1°C graduations.

The method used involved immersing the thermistor and
the thermometer into cold water and taking readings as the
water warmed up to room temperature. The same procedure was
used with hot water allowed to cool to room temperature.
The process was replcated three times and averages
calculated for the levels of readings acquired. As was done
previously, the averages were used to derive a calibration
curve using linear regression. Table 4.4 contains the data
readings and the averages while figure 4.10 is a plot of

the calibration curve.

63

«.3 magma aoum mump
maaamm wcwm: m>u=o :oflumunwamo poumwsuwza o~.q gunman

AUG tho>v wu<h40> z<mz

au.v am.m mm.m on.m mm.m nm.m av.m

4 4 4 J a 5.”:

 

.. / - new

Nam. HwLoawm m

Nnm+m~MB.a um
Nnm+mmmm.N no

xg.ou>
1 n afifi

T 1 as:

 

 

 

b b in b L a 05m

w>m2u onh<mmH4<u mohmmzmwzh

$015133 $338330 NI BEOLVHBdHEL

64

TABLE 4.4 THERMISTOR CALIBRATION DATA AND AVERAGES

TEMPOC REP 1 REP 2 REP 3 MEAN
16 3.86 3.85 3 84 3.85
25 3.75 3.75 3.75 3.75
28 3.73 3.73 3.73 3.73
40 3.54 3.56 3 55 3.55
41 3.53 3.55 3.54 3.54
42 3.52 3.52 3.53 3.52
43 3.50 3.51 3.49 3.50
44 3.48 3.49 3.47 3.48
45 3.46 3.46 3.46 3.46

4.2.4 Thermocouple Calibration

The thermocouples detected the highest temperatures
encountered in the experiment. As thermocouples produce
very small output signals (millivolts), the output must be
amplified. The amplified signal will correspond to the
temperature sensed by the thermocouple.

The calibration of the thermocouple was accomplished
using a propane torch as a heat source, and using a digital
voltmeter to read the unamplified output. The A/D converter
was used to obtain the amplified output signal level.

Readings were taken in increments of 0.5 mv (12°C) in
the range 367-673°C for the high temperature thermocouple,
and from 98-221.5°C for the low temperature thermocouple.
Sample observations were replicated at least three times.
The means of the observed output levels are shown in Tables
4.5 and 4.6. These mean values were used to obtain linear

regression equations which are plotted in figures 4.11 and

65

4.12. Both sets of data displayed very strong linear
relationships as noted by correlation coefficients of
0.999897 for the high temperature thermocouple,and 0.9998
for the low temperature thermocouple at the cooler inlet.

The high temperature thermocouple was outfitted with a
ceramic protection tube to protect it from contamination.
The protection tube had a slight insulating effect on the
the thermocouple and thus the output signal level was lower
than an unprotected thermocouple. The magnitude of the
difference was measured using two thermocouples. One with
ceramic protection and one without the protection. The two
thermocouples were simultaeneously heated with a propane
torch and the output signal observed. The result of the
observations indicated that there was a difference of 6 mv
between the two, with the unprotected unit being higher.
After amplification this represents 0.71 volts. This
difference was allowed for in the data conversion program

for the high temperature thermocouple.

66

Table 4.5 Hi Temperature Thermocouple Calibration Data

REP 1 REP 2 REP 3 REP 4 TEMP VAVG
voc VDC VDC VDC

1.958 1.958 1.958 1 958 367 1.958
2.0208 2.0208 2.0208 ----- 379 2.0208
2.0833 2 0833 2.0833 2 0833 391 2.0833
2 1667 2 1667 2.1667 ------ 402 2.1667
2.2292 2 2292 2.2292 ------ 414 2.2292
2.2917 2 2917 ------------ 426 2.2917
2 375 2 3333 ------------ 438 2.3541
2 4167 2 4167 2.4167 2.4167 450 2.4167
2.4792 2.4792 2.4792 2.4792 461 2.4792
2.5417 2 5417 2.5417 ------ 473 2.5417
2 5833 2.6042 2.5833 ------ 485 2.5903
2 6667 2.6667 2.6875 ------ 497 2.6736
2 750 2.750 2.750 ------ 508 2.750
2.8125 2.8125 ----------- 520 2.8125
2.875 2.875 2.875 2.875 532 2.875
2.9375 2.9375 ---------- 544 2.9375
3.0 3.0 3.0 3.0 555 3.0
3.0833 3.0625 ---------- 567 3.0729
3.125 3.125 ---------- 579 3.125
3.1875 3 2083 3.2083 ----- 591 3.2014
3.25 3.25 3.25 3.2708 602 3.2569
3.3333 3.3333 3.3333 3.3125 614 3.3250
3.3958 3.3958 3.3958 3.3958 626 3.3958
3.4583 ------------------ 638 3.4583
3.5208 3 5208 3.5208 3.5208 649 3.5208
3.5833 3 5833 3.6042 3.6042 661 3.5937
3 6667 3 6667 3.6667 ------ 673 3.6667

TEMP REPl REP 2 REP 3 REP 4 REP 5 REP 6 REP 7 VAVG

oC VDC VDC VDC VDC VDC VDC VDC
98 1.125 1.1042 1.125 -------------------- 1.1181
110 1.25 1.271 1.25 1.271 --------------- 1.2605
122 1.39 1.437 1.39 1.375 --------------- 1.3935
134 1.521 1.542 1.542 -------------------- 1.5350
147 1.688 1.667 1.688 1.688 1.688 ---------- 1.6833
159 1.792 1.813 1.792 1.833 1.792 1.792 1.813 1.8036
172 1.958 1.917 1.958 1.958 1.917 ---------- 1.9417
184 2.083 2.104 2.063 -------------------- 2.0833
197 2.229 2.229 2.229 2.229 --------------- 2.2292
209 2.354 2.333 2.354 -------------------- 2.3472
222 2.520 2.479 2.500 -------------------- 2.4997

TEMPERATURE IN DEGREES CELSI'JS

ZUS

ELS

IN DEGREES C

TEMPERATUR

67

 

 

 

 

 

 

 

 

 

I". I" I 1 I T
m ..
NHL-
ma 1- Y=a+bX
a” 1.412E+HMI
5"” ' be 1.8829882
593 L. R Square": 1.140”
m - -I
X
l 1 L l l
2.00 2.50 3.00 3.50 4.00 4.50 5.00
OUTPUT VOLTAGE (DC)
Figure 4.11 High Temperature Thermocouple
Calibration Curve - data from
Table 4.5
400 r r I T I —_1
3mlb 1
r Y=o*bX
//
am 1
r (,9 a= -.394E*IIJHI
b= 9.025E+le
R Square: LUUU
IBHF 1
a l l I I l
1.00 1.25 l. 50 I. 75 2.00 2 25 2 50

OUTPUT VOLTAGE (DC)

Figure 4.12 Low Temperature Thermocouple (Gas cooler

inlet) Calibration Curve - data from Table
4.6

CHAPTER 5

METHODS AND PROCEDURE

5.1 Experimental Design

5.1.1 Considerations

The objective of the investigations undertaken in the
experiment was to assess the effect of fuel moisture
content when gasification was used to power a diesel

engine. Specifically, the task was to :

1. Measure the degree to which the cob moisture
levels reduces diesel fuel consumption.

2. Deduce if there was a significant difference
in diesel fuel consumption between the 3 fuel
moisture contents.

3. Determine to what degree cob moisture affected the
equilibrium grate temperature.

4. Evaluate the financial implications of

gasification with the aid of cost benefit analysis.

The preceeding objectives involve two independent
variables; moisture content and engine speed, each tested
at 3 different levels. The moisture levels were 5%, 15%,
and 25% wet basis. The engine speeds were 400, 500 and 600
rpm. The optimum level of each variable was sought, in

68

69
addition to the relationship between the two independent
variables and the dependent variable, diesel fuel
consumption.

Statistical principles provide for an experimental
design that accommodates the information desired from this
experiment. The design was called a factorial experiment
(Steel and Torre, 1980). Factorial designs were valuable
for exploratory work where optimum levels of factors were
sought or where the objective was to determine which
factors were important.

A factor was simply one of the independent variables.
However, each factor supplies several treatments to the
experimental design. Each treatment represents a given

level of the factor.
5.1.2 Treatment Selection

This experiment contains two factors; fuel moisture
content and engine speed. Both factors had three levels or
treatments. Three levels of each factor were chosen since
it was important to determine if the nature of the response
to the treatments was curvilinear.

Factorial experiments consist of all possible
combinations of the three levels of the two factors. This
combination yields 3 X 3 or nine treatments.

The term treatment refers to the procedure whos'
effect was measured and compared with other treatments. A

treatment was applied to an experimental unit, which was

70
the material under evaluation in the experiment. For this

case the experimental units were corn cobs.
5.1.3 Experimental Error

Variation that exists among observations on
experimental units treated alike was called experimental
error (Steel and Torre, 1980; Cochran and Cox, 1976). It
had two main sources: (1) inherent variability that exits
in experimental material to which treatments were applied,
and (2) variation resulting from lack of uniformity in the
physical implementation of the experiment. Such variation,
if large, weakens the strength of the inferences made about
the experiment. The weakness was manifested by long
confidence intervals about the mean (and or variance) and
low power of the test. The principle methods of controlling

experimental error are:

1. Handling the experimental material in a way that
reduces the effects of inherent variability.

2. Employing refined experimental technique
A detailed discussion of the use of these controls is
given in subsequent sections.
5.1.4 Replication and Blocking

Replication implies repetition of something. In

statistical procedures, replication involves repeated

71
appearances of a treatment (Steel and Torre,l980).
Replication was a means of increasing the sensitivity of an
experiment by, reducing the standard deviation of the
treatment mean. When more replicates were used, estimates
of population means were more precise, since more
observations were used to obtain that mean. Experimental
error and confidence interval information were obtained
from replication. Experimental error in the form of the
error mean square, was necessary for tests of significance.
Without replication one cannot determine if observed
differences were due to the difference between treatments
(real differences) or the difference between experimental

units (inherent variation).
5.2 Experimental Procedure
5.2.1 Treatment Preparation

In July 1986, approximately 70 KG of corn cobs were
obtained from an elevator in Weberville, Michigan. The cobs
came from the 1985 crop. The variety of the corn was
unknown. After delivery to the Agricultural Engineering
Department on the Michigan State University campus, the
cobs were cleaned. The cleaning process involved the
removal of corn husks, and stalk pieces. Following this the
cobs were placed in a sieve with 8 mm X 8 mm openings, to
remove corn kernels, cob fines, and dust. Finally the cobs

were graded using a 30 mm x 30 mm wire mesh screen. The

72

grading yielded 2 classes of cobs. One class was of whole
cobs 53 mm in diameter and 75 to 125 mm long. The second
class contained slightly smaller cobs, 30 mm in diameter
and 30 to 125 mm long. This class also contained split and
broken cobs plus small cob pieces. The latter class of cobs
were used in the experiment because there was less void
space.

Three moisture levels were prescribed in the factorial
experimental design. The initial corn cob moisture level
was determined to be 4.8 to 5.2 percent wet basis. Thus to
achieve the higher moisture levels of 15% and 25%, moisture
was added to the corn cobs. A problem arises in obtaining
and maintaining a precise moisture level by adding water.
The volume of material was considerable, and some moisture
loss could occur over time. Measures taken to reduce this
variation involved (1) preparing a volume (batch) that was
adequate for completing one replication and (2) keeping the
cobs in moisture proof containers. Three batches of cobs
were pepared for each replication. Each batch represents 1
of the 3 moisture levels in the experiment. Additionally,
the moisture added must be absorbed by the cobs.

Moisture was added according to the procedure
described by White et al (1984). With this method the cobs
were repeatedly sprayed with water during a 48 hour period
and stored at 4-5°C. After each spraying the cobs were
mixed and stored again. The appropriate moisture level was

determined by drying a sample to obtain the initial

73
moisture content wet basis. Using this information, the
weight of the water required to achieve the specified
moisture level was derived. The weight of the water plus
the initial weight of the batch of cobs gave the total
weight of the treatment. Water was sprayed on the cobs
until this treatment weight was reached. Once the moisture
addition was complete the cobs were held in a 4°C cold
storage for 4 days, allowing them to reach equilibrium.
Prior to using the cobs in the experiment, the cobs were
removed from the cold storage and allowed to warm up to
ambient air temperature. At this point three samples were
taken from the treatments to preciesly obtain the cob

moisture content at the time of use.
5.2.2 Treatment Application

The diesel engine was warmed up for approximately 30
minutes before any treatments were applied. Start up of the
gasifier involves igniting a small quantity of corn cobs
with the help of charcoal lighter fluid. The cobs were
allowed to burn until the lighter fluid was exhausted , at
which time the cobs had acquired a glowing red color
similar to charcoal in a barbeque grill.

The delivery hose was then attached to the engine
intake manifold creating a vaccum through the gasifier
system. The smoldering cobs were poured into the fuel
hopper and regular cobs were added on top of the smoldering

ones. The air/gas mixing valve was adjusted so that

74
combustion of the cobs was promoted. Once the cobs were
burning well enough to produce a suitable gas the air/gas
mixing valve was again adjusted to draw the maximum gas to
the engine without choking for air. After 15 minutes of
operation at this level the system achieves a pseudo
eqilibrium.

The speed of the engine was adjusted, as was the load
on the prony brake, for the requirements of that particular
treatment. For the purpose of comparing the amount of
diesel fuel consumed under each treatment, the same load
was applied to the engine for a given engine speed,ie 2.7
kgs of load was applied to the engine for all tests at 400
rpm, 3.2 kgs at 500 rpm, and 3.9 kgs at 600 rpm. After the
correct speed and load were obtained, more cobs were added
to the fuel hopper and data acquisition was subsequently
initiated.

Diesel fuel consumption and ambient air temperature
were monitored simultaeneously. A 250 ml graduated cylinder
was filled with diesel fuel and a siphon hose was attached
via a tee fitting to the engines fuel supply line. When
data acquisition was started, the main fuel tank was shut
off and the siphon was opened. Visual readings were taken
at the beginning and the end of the runs.

All treatments were applied in a random manner. Each
block was randomized separately. Ramdomization was
accomplished by writing the nine treatment combinations on

nine pieces of paper. The papers were folded and put into a

75
cup. One paper at a time was drawn without replacement.
Treatment combinations were noted in the order drawn, and
excuted as such beginning with block 1. The order of the
treatments were shown in table 5.1. The randomization
process was done three times, once for each of the three
blocks.

Treatment combinations were designated by two mumbers
separated with a hyphen. The first number was precent
moisture content, the second was engine speed in rpm.

Experimental conditions were kept as constant as
possible for the duration of the experiment. Gas filtering
material was changed once every three treatments, in
addition to emptying ashes from the ash pan and char from
the cyclone gas cleaner. Water vapor condensate was also
emptied from the gas cooler unit after every three
treatments. The gasifier throat was emptied of residual
material after each treatment run. The grate was also

cleaned of any slag.

TABLE 5.1 TREATMENT RAMDOMIZATION RESULTS

ORDER BLOCK 1 BLOCK 2 BLOCK 3
1 25-600 5-600 5-500
2 15-500 5-400 25-400
3 25-400 15-400 15-600
4 5-400 25-400 15-400
5 15-500 5-500 25-500
6 5-600 25-600 5-600
7 5-600 15-600 25-600
8 5-500 25-500 5-400
9 25-600 15-500 15-500

76
At the engine, motor oil changes were done and the
injector nozzle cleaned before each block of treatments
were run. Intake and exhaust valve lash was also checked.
The engine was cooled using a thermo siphon raditor System.

Radiator leaks requierd frequent filling of the radiator to

maintain the water level. Engine temperature was not

monitored in this experiment.

CHAPTER 6
RESULTS AND DISCUSSION
6.1 Overview

The execution of the experiment was accomplished
without any major difficulties, though there were few
conditions that required attention as far as the analysis
was concerned. The conditions were the folowing: l)
variation in moisture content of the corn cobs, 2) engine
speed variation, and 3) horsepower variation. Inspite of
measures taken to control these parameters variations did
occur. Taking this into account, additional analytical
procedures were employed to obtain unbiased estimates of
the statistics.

The variation in corn cob moisture content was the
most significant of the variations encountered. The
average moisture content wet basis was shown in table 6.1
for the range of engine speeds and the "nominal" moisture
contents of 5,15, and 25%. The highest moisture loss
occurred with the 25% material. Losses occurred during
handling while running the experiment and for the period of

time the material was kept at ambient air temperature.

77

78

Table 6.1 Wet Basis Cob Moisture Data

Nominal Moisture Actual Moisture
"ESEEESEZJIE """" £132?Iméiééi'éméiééfi """"" (4;;-
5% 4.7 5.14 5.145 4.995
15% 14.38 13.1 12.25 13.24
25% 18.54 13.45 22 0 17.996

The overall average moisture contents wet basis were
low for the 15% nominal value by 1.75 points at 13.24.
The 25% nominal level was lower by 7.0 points. Thus plots
of temperature profiles found in this chapter involve the
moisture content means found in table 6.1.

The engine speed variation occurred as a result of l)
varying load at the prony brake drum, 2) changes in the
sensitivity responsiveness and calibration of the engine
speed detection instrumentation. The speed sensor
required 3 calibration curves by the end of the experiment,
due to changes in the instruments output signal for a given
speed.

Engine speed data for the three blocks of the
experiment were presented in Table 6.2. The 400 rpm
nominal tests were consistently below four hundred rpm.
Some of this was due to the prediction equation which under
predicts speed at the low end of the range. At the 500
rpm nominal speed, the prediction equations had bettter

correlation and these results were closer to the nominal

79

Table 6.2 Engine Speed Treatment Data in RPM

lst Letter : L - Low Speed M - Medium Speed H - High Speed

2nd Letter L - Low Moisture M - Medium Moisture H 8 High Moisture
"""""" ££"""£§""‘L§'""&£'""§i""'§§""’fi£’"'§§""’§§""'%;Z;I;’
SISER’I"'3§5"‘"’3£3”"333"";5S""£E§""§IS”"§§5"'£8§""855""Z553"’
Block 2 364 378 364 516 496 511 610 590 606 4435
Block 3 395 382 380 517 493 492 571 564 576 4370

Totals 1131 1102 1123 1518 1434 1515 1739 1719 1810 13091
Means 377 367 374 506 478 505 579 573 603 484.67

Table 6.3 Engine Speed Means Across Blocks and Treatments

Low Speed Medium Speed High Speed
"L;;‘§;E;£;;;"""'S§§ """"""""" £6; ””””””””” é}; """"
Med Moisture 367 478 573
Hi Moisture 374 505 603

Overall Means 373 496 585

80

value. The 600 rpm nominal speed had results above and
below the nominal speed, differing by as much as 42 rpm.
The engine speed results of the three blocks were averaged
across blocks for like treatments. The block/treatment
averages are contained in Table 6.3. These mean engine
speed figures were included in a linear regression curve
fitting. These results were elaborated elsewhere in this
chapter.

The final component of variation was engine power.
This component had the smallest variation of those
mentioned thus far. It was necessary to minimize power
variations to obtain fuel consumption results that were
directly comparable, between moisture levels. Table 6.4
contains the power results for blocks 1,2 and 3 of the
experiment. Variation in power output originated from the
same sources as it did for engine speed, namely, the speed
detection system and the force measuring apparatus.

The block means for engine power are contained in
Table 6.5. These means were used to obtain fuel
consumption figures per unit of power in a subsequent

section of this chapter.
6.2 Grate Temperature Data

The high temperature thermocouple on channel 4 sensed
the temperature just below the gasifier grate. Forty data

points were collected for each treatment run. Temperature

81

Table 6.4 Engine Power Treatment Data in Kilowatts

lst Letter : L a Low Speed M - Medium Speed H a High Speed
2nd Letter : L . Low Moisture M - Medium Moisture H - High Moisture

Block 1 0.613 0.553 0.626 0.902 0.872 0.984 1.34 1.31 1.55 8.75
Block 2 0.581 0.611 0.585 0.962 0.962 0.984 1.47 1.45 1.46 9.065
Block 3 0.626 0.619 0.619 0.947 0.980 0.902 1.387 1.36 1.39 8.83

Totals 1.82 1.783 1.830 2.811 2.814 2.870 4.197 3.94 4.40 26.465
Means 0.607 0.594 0.610 0.937 0.938 0.957 1.399 1.313 1.467 0.98

Table 6.5 Engine Power Means Across Blocks and Treatments (in kilowatts)

Low Speed Medium Speed High Speed
LETTERS?”""635; """""""" 6:335""""""'""ET3§§ """"""
Med Moisture 0.594 0.938 1.313
Hi Moisture 0.610 0.957 1.467
432;1;"'"""""T511 """"""""" 5:535 """"""""" 2:17; """"""

Overall Means 0.604 0.944 1.393

82
data for like treatments were averaged across blocks
(replicates). The averages are plotted in figures 6.1 to
6.3. In general the temperature profiles were lower as cob
moisture content increases.

High moisture cobs however, were considerably more
difficult to gasify than lower moisture material. Wet cobs
tended to bridge easily inside the fuel hOpper, thus
producing erratic flow conditions into the fire box zone.
This characteristic necessitated frequent monitoring of the
fuel in the hopper, in order to manually assist cob flow.

Igniting the wet fuel proved to have its difficulty.
Wet cobs were slow burning and easily extinguished
themselves before reaching a sustainable level of
combustion. To overcome this problem the drier 5% moisture
cobs were used to start the combustion reaction. The heat
produced by the dry fuel was used to ignite the wetter cabs
and build up a sustainable incandesent mass capable of
consuming the wet cobs. The high moisture cobs were slowly
added, and once they had decended into the fire box data

acquistion was started.

6.2.1 Temperature Profiles for high Moisture Cobs

The initially high temperatures depicted for the high
moisture profiles in figures 6.1 to 6.3, were likely due to
the use of the drier cobs as starter fuel for the wet cobs.
The temperature profile decreases over time reaching an

equilibrium toward the end of the treatment test. This was

CRATE TEMP IN DEGREES CELSIUS

83

 

can

mmmmssm
asa- FIRM“
8UP

 

 

 

 

 

 

 

 

awk-
-————-um<sornumnscms
—— "emu: (Ian mime was
558 t HIGH (18:) MOISTURE cues
5m 1 l l 1 l l l l l l l
asamaxsazmzsaamasawavsasaassn
ELAPSED TIME IN SECONDS
Figure 6.1 Gasifier grate temperature profiles at 400

rpm engine speed. The curves represent
averages over the 3 blocks for the 10
minute length of the test.

800

84

 

CRATE TE)? IN DEGREES CELSIUS
i

 

mmmmmssm
mason

-———-uw1501nnmm£aus
———IenIIu (13:) mm
IuaIamoInmmnsius

 

 

I I I I I I I I I I

 

53 .-
a I
I E!
Figure 6.2

In 150 2” an an 33 4' 45¢ 54! 53 am

am TIDE IN SECIIIJS

Gasifier grate temperature profiles at 500
rpm engine speed. the curves represent
averages over the 3 blocks for the 10
minute length of the test.

85

 

 

 

 

 

 

 

 

an
nanmnuunensnomasanas
fininlmm
mar
mn-
§ ‘————-uu<monnmmn£cus
3, —--— MEDIuI (13:) 1mm (ms
(a 75. 4 HIGH (18:) MOISTURE cues
E; \"\.x.
g . \\\\\\
a nn
5. \ \<r _ ~—
31 ~ \\ '\.\.\
.- 3" I- .~ .
E \\\ \-\_0-\.\.~.-——
g uni. .\\\\\\\\\L
53 " \
m L g I l I L I l l L l
assimiazmzsaamammaasmssam
EJHEDTHEINSHDMS
Figure 6.3 Gasifier grate temperature profiles at 600

rpm engine speed. The curves represent
averages over the 3 blocks for the 10
minute length of the test.

86
the case at all of the engine speeds examined and was,
consistent with expectations and results of other

investigators.
6.2.2 Temperature Profiles for the Medium Moisture Cobs

The medium moisture fuel behaved similar to the high
moisture fuel in that dry cobs were required to get the
combustion process started. When the cobs reached full
combustion, the medium moisture cobs were added on top of
the burning mass.

The results of the medium moisture cob temperature
data place this moisture level at a near steady state
temperature higher than that of the high moisture cobs, as
expected. Steady state conditions were indicated by a
flatened curve displaying nearly constant temperatures.
Psuedo steady state conditions appeared at the very start
of the test for the 400 rpm treatments, 300 seconds into
the run for the 500 rpm treatments and 425 seconds into
the run for the 600 rpm treatments. Prior to the steady
state conditions temperature was decreasing due to the
consumption of the dry cobs, and they're replacement with
the wetter medium moisture cobs.

At all engine speeds, the medium moisture cobs show
steady state temperatures higher than the high moisture
cobs and lower than that of the 5% moisture cobs. Oven
dried samples of the medium moisture cobs averaged 13.2%

moisture wet basis.

87

6.2.3 Temperature Profiles for the Low Moisture Cobs

The low moisture corn cobs were gasified easily.
Igniting the cobs and obtaining a self-sustaining
combustible mass was achieved without complications.

Psuedo steady state conditions were rapidly achieved
and continued for the duration of the tests. At steady
state conditions the temperature of the 5% moisture cobs
was in all cases higher than both the medium and the high
moisture tests. This outcome was confirmed by literature
which relates that the wetter the biomass fuel the lower
the reaction temperatures in the gasifier fire box. This is
due to the endothermic process of vaporizing the water
contained in the cobs (biomass). The result was a lowering
of the reaction temperature and, as is explained later, a
decline in gas quality - ultimately producing less energy

from the gas.
6.2.4 Summary of Temperature results at 400 RPM

Low Moisture Cob Averages (5%) - Initial temperatures were

 

at the 650°C level and rose to 675° at the end of the test.
The temperature rise was probably due to starting data
acquisition prior to achieveing steady state conditions. In
the steady state range, these temperatures were 25° higher
than the medium moisture test profile and 30° higher than
the high moisture test profile. These were the smallest

differences between the various moisture levels of the

88

three engine speeds tested. The reason for this was
attirbuted to the likelyhood of some moisture loss from the
medium and high moisture material, and a lower air velocity
through the gasifier given the slow engine speed.

Medium Moisture Averages (13%) - This higher moisture
material also began with temperatures of 650°C but
maintained this level for the duration of the test. This
profile was 5-100 higher than the high moisture cobs.

High Moisture Averages (18%) - This treatment average had

 

the highest inital temperature of the three moisture
levels. This again, was due to the initiation of data
acquisition while some residual 5% moisture material
remained in the gasifier. The profile however showed
continous decline in temperature until a psuedo steady
state was achieved at 635°C. This temperature level was
very close to that of the medium moisture cobs. A much

lower temperature level was anticipated for this treatment.
6.2.5 Summary of Temperature Results at 500 RPM

Low Moisture Cobs (5%) - This was another test which shows

 

steady state conditions for the entire test. Grate
temperatures were in the 700-705°C area, highest of the
three moisture levels. This temperature was also very close
to that of the 5% moisture test at 400 rpm.

Medium Moisture Cobs (13%) - This treatment had initial

 

temperatures of 680°C. Decreasing temperatures occurred in

a continuous manner for 300 seconds. Thereafter the

89
temperature stabilized in the 650-6600 range. This level
was 40-50° lower than that of the 5% moisture cobs tested
at this speed and roughly the same as the medium moisture
test at 400 rpm.

High Moisture Cob Averages (18%) - The high initial

 

temperatures were again due to the employment of dry cobs
to start combustion and beginning data acquisition before
this material was exhausted. The profile shows continuous
decline in temperature throughout the duration of the test.
The slope of the curve remains negative at the end of the
test indicating that further decreases in temperature were
probable. The final temperature was slightly below 550°C.
This was 85° lower than the high moisture test at 400 rpm.
Comparing it with the other moisture treatments of 500 rpm
the final temperature was 150° lower than the 5% moisture

test and 100° lower than the medium moisture test.
6.2.6 Temperature Results at 600 RPM Engine Speed

Low Moisture Cobs (5%) - Rising temperatures characterize

 

this test, with an initial level of 675°C that moved up to
700° at the mid point of the test. Steady state conditions
arrived at the 700° level, though the profile of this
treatment was a few degrees lower than the 5% moisture test
at 500 rpm but equaled the 5% moisture results at 400 rpm.

Medium Moisture Content Cobs (13%) - The highest initial

 

temperature level occured with this treatment average, again

due to factors mentioned earlier. A constant decline was

90

obvious to about 425 seconds into the test. The temperature
stabilized slightly below 650°C, and stayed in this area
for the remainder of the test. This level was practically
equal to that of the medium moisture cobs examined at 500
rpm.

High Moisture Content Cobs (18%) - Distinguishing this
treatment average from the others at this moisture level
was the low initial temperature of 700°(figure 6.3). This
indicated that the data acquisition began later in the
cooling down period than it did in treatments at other
speeds. The profile shows an extremely constant negative
slope for 400 seconds. At this point the curve flattens
somewhat but continued in a negative direction. This
treatment average did not achieve a steady state before the
end of the test. The temperature at the end was near 530°,
the lowest temperature of any test at any of the engine

speeds.
6.3 Gas Cooler Inlet Temperature Results

The temperature data for the thermocouple at the
cooler inlet (channel 5) was averaged across blocks for
like treatments. These averages were plotted in figures
6.4, 6.5, and 6.6. Raw data used for the averages was
listed in Appendix 2. In general the outcome of the
results for this sensor mimicked those of the grate
temperature sensor covered in previous sections of this

document. However these averages exhibited a moderate

91

 

GASENTRYTEN’TOCEDLERCC)

 

 

mmmusmmsm
mum

——-LOI (51) MISTURE ans
—---— E0113 (13!) mm CW5
---—- 111131 (181) MOIST“ C(BS

 

 

Figure 6.4

Gas temperature profiles at the cooler
entrance for 400 rpm. Averaged across 3

blocks for the 10 minute length of the
test.

GAS ENTRY TEMP TO COOLER ( C)

50

1-
,.

92

 

I

 

“NKMWREEWBAUEBBUDB
HREEIHN

———m (5:) mm cues
—-— IEDIlIl (13:) IDISTIRE
HIGH (18:) mm

 

    

'R‘

 

. \\‘
./OV -— .
L I L I I
a ll! an an an in an
aAfifinTnEINSHnMB
Figure 6.5 Gas temperature profiles at the cooler

entrance for 500 rpm. Averaged across 3

blocks for the 10 minute length of the
test.

 

93

 

GAS ENTRY TEMP TO COOLER ( C)

un-

 

nanmnuuwcnsImmssamos
flRifllflM

—--—L0' (51) 1mm ans
—’-— IE“! (13!) 1015“]! LIBS
HIGH (181) 101m 1135

 

 

 

Figure

6.6

ELAPSED TIDE IN SECONDS

Gas temperature profiles at the cooler
entrance for 600 rpm. Averaged across 3
blocks for the 10 minute length of the
test.

 

94
amount of fluctuation. This was especially evident in the
400 rpm test for the medium moisture cobs (figure 6.4) and
in the 600 rpm test for the low moisture cobs (figure 6.6).

Much of this fluctuating tendency was attributable to
the sensor circuitry employed to amplify the signal.
Preliminary calibration tests revealed that the
thermocouples' output would randomly fluctuate above and
below the actual temperature level. The reasons for this
were not precisely known but were believed to be due to
faulty components in the amplification circuitry. Since
this was a secondary measuring device, it was decided to
implement the experiment without corrrecting the fault, in
an effort to complete the experiment prior to the program
deadline.

Other potential sources of disturbance were the
reversion pulses of the single cylinder diesel engine. the
lack of insulation on equipment downstream from the gas
producer and backfiring occurring in the gas producer.

Reversion was a reversal of the direction of flow of
the incoming gas and air mixture. This reversal occurred
because the suction produced by the engine was intermittent
rather than constant. Suction only occurs when the intake
valve was open, and the valve only opened once for every
two revolutions of the crankshaft. This condition permitted
cooler gas downstream from the sensor to reverse direction
and flowed backward toward the sensor.

The gasifier furnace was insulated with asbestos on

95

the inside of the outer casing. No other parts of the
apparatus were insulated, and except for the cooling hose
and delivery hose, all components were mild steel. The
tests were conducted indoors but with the doors open to
ventilate the area. In this environment gusts of wind and
irregular breezes might have impacted the output of the
sensor on occasion. Since the temperture gradient was high
(18-24° vs ISO-250°) some fluctuation in temperature could
occur.

Finally, backfiring within the gasifier could also be
a factor. The author does not believe it was a significant
one since no backfiring was observed during data
acquisition. However, strong bacfiring was evident on most
tests prior to initaiting data acquisition. Backfiring
occurred when pockets of combustible gases were produced
above the firebox, and were drawn through or pushed into
the burning mass of fuel, causing them to detonate. This
detonation monetarily increased pressure inside the system
pushing hot gases in all directions of least resistance.
There were two such directions in this gasifier system and
one of them leads to the thermocouple in question, the
other leads out of the gasifier and into the surrounding
atmosphere. Such detonation could momentarily raise
temperatures in the vicinity of the thermocouple at the
cooler inlet causing a spike in temperature. This spike
would not be detected at the grate thermocouple because

temperatures there were much higher and the effect of

96

detonation on temperature was less obvious if at all.
6.3.1 Cooler Inlet Temperature Results at 400 RPM

The moisture test averages yielded a steadily
increasing temperature profile. This was also the highest
temperature regime of all tests at 400 rpm. This treatment
average never truly achieved a near steady state condition.
End of run temperatures were in 240°-250°C range.

Very erratic behavior and an uncharacteristically low
temperature profile was the result for the medium moisture
test. Temperatures were steady for the first 2 minutes of
the test, but then repeatedly drop and rise in 50° swings
for the duration of the experiment. This outcome was likely
due to one of the disturbances discussed earlier in this
section. Review of the grate temperature profiles show that
nothing in the gasifier temperature profile could explain
this occurrance at the cooler inlet. The end of test
temperature was in the 190° area only 5° lower than that of
the medium cob moisture tests.

Except for the medium moisture profile the profiles of
the cooler inlet temperature sensor were consistent with

those of the grate temperature sensor.
6.3.2 Cooler Inlet Temperature Results for 500 RPM

The tests at this speed provided outcomes that the
author considered typical for the experiment. All tests had

relatively steady profiles, with the low cob moisture level

97
being the warmest and the high cob moisture level the
coolest.

In this treatment averge the low cob moisture level
displayed an increasing temperature profile, but reached a
fairly steady state 400 seconds into the test. At this
point in the test it had the highest temperature of all
moisture levels, as expected, though the medium moisture
profile was only a few degrees lower (figure 6.5). The
pseudo steady state temperature at the end of the test was
220°,roughly 30° lower than the 400 rpm test.

A fairly steady state profile was also evident in the
medium cob moisture level test (figure 6.5). After a steep
decline in the first few seconds of the test, the profile
closely followed that of the 5% cob moisture level. It
hovered in the 200°C area for much of the test before
rising to the 220° just as the 5% cob moisture averages
did. At the end of the profile the temperature was 215°,a
scant 4° lower than that of the 5% cob moisture averages
and 25° higher than the high cob moisture averages.

The high cob moisture averages began at a high level
(235°C) and gradually decreased continually thereafter.
During the final four minutes the temperature dropped below
those of the low and medium moisture averages. At the end
of the test the temperature was 190°, 25° lower than the
medium cob moisture averages and 300 lower than the 5% cob
moisture averages. The slope of the profile maintained a

negative direction throughout the run, though some

98
flattening was apparent at the end. Interestingly, the
temperature at the end of the high cob moisture profile at
400 rpm was only 7.0°C higher than this one.

All of the temperature average profiles at the cooler
inlet thermocouple at 500 rpm were consistent with those of
the grate temperature profiles (see figures 6.2 and
6.5). This was the only speed for which all cooler inlet
profiles were consistent with all grate temperature

profiles.
6.3.3 Cooler Inlet Temperature Results for 600 RPM

One of the profiles at 600 rpm displayed wide
fluctuations in temperature levels which were assumed to
have been caused by disturbances described earlier, namely:
reversion, lack of insulation down stream for the gas
producer, and backfiring in the gas producer. The profile
in question was the 5% cob moisture average. Noticable in
figure 6.6, the temperature for this treatment begins at
about 175°C. Immediately it dropped by 25°, then rose by
75° and again dropped by about 50°. This behavior continues
for the duration of the test, though the oscillations have
a lower amplitude during the final 2.5 minutes of the test.
The temperature was rising for most of the first half of
the test and declined somewhat in the second half. The test
ended with a temperature of 225°, 50° higher than the

initial temperature of 175°. The profile had a steep

postive slope in the final seconds indicating that the

99

oscillating tendency was continuing. The best commentary
possible for this particular test was that it ultimately
was higher in temperature than the others. Futhermore, it
was lower than the 5% cob moisture level average at 400
rpm and greater than the 5% cob moisture level average at
500 rpm by a couple of degrees. This profile never achieved
a steady state condition.

The profile of the medium cob moisture test at 600 rpm
contained some oscillation in the first 3 minutes of test,
but settled down to a gradual temperature decrease
thereafter. Initial temperatures were at the 250° level and
finished in the 200° area at the profiles end. The final
four minutes showed reasonably constant temperatures in the
ZOO-220° range. This was very similar to the final
temperatures for the medium cob moisture profile at 500
rpm, but somewhat higher than the result at 400 rpm.

The behavior of the high cob moisture level profile
was close to that of the medium cob moisture profile.
Temperatures were initially high and gradually declined to
the 175-210° range. the profile of this moisture level was
between 5 and 10 degrees lower than that of the medium
moisture level. The slopes of the two curves were also
quite similar. The final temperature of this test was near
200 degrees. This was practically the same as for the high
moisture profile at 500 rpm and about 10 degrees less than
the high moisture average at 400 rpm.

Making comparisons with the grate temperature profiles

100

(figures 6.3 and 6.6 ), it was apparent that the 13 and 18
percent cob moisture profiles were consistent for the two
sensors. In both cases the the profiles started at high
levels and decreased stedily until the final minutes of the
test. The 13 percent cob moisture averaged were higher in
temperature on profiles of both sensors (grate and cooler
inlet).

Even the 5 percent moisture profile follows a similar
overall trend for both sensors. However the wide
oscillations at the cooler inlet sensor were not evident in

the grate temperature profile.
6.4 Gas Cooler Exit Temperature Results

A thermistor was used to detect the temperature of
the gas at the cooler exit. The purpose of this sensor was
to verify that the temperature of the gas leaving the
gasification system was as cool as was economically
possible. The advantages of cool gas were: 1) water vapor
and tar are condensed and removed from the gas stream, and
2) the gas is denser and contains more energy per unit
volume as it enters the engine. Water at ambient air
temperature was used as the cooling medium, thus the gas
temperatures were expected to be in the vicinity of the
ambient air temperature. Indeed, the results in figures
6.7, 6.8, and 6.9 show that this was the case.

Ambient air temperatures during the experiment ranged

from 17 degrees to 23 degrees. The gas temperatures

101

 

 

 

 

 

 

 

 

 

 

 

5
TDPERATIRE ENS amass am
PM 4. m
a I-
'—UJI (5!) 101m
""'—_ $1111.“ (13!) 10151!!!
E HID! (1813 mlsrme CW5
6
8
a h-
E A / V
2 23. . _
35 w.__/ v \f
m
3 N
23 L
a I_ I I I L
0 1m 2m 3m 4m 5m

ELAPSED TIE IN SECGIJS

Figure 6.7 Gas temperature profiles at the cooler
exit for 400 rpm. Averaged across 3 blocks
for the 10 minute length of the test.

GAS EXIT TEMP FROM COOLER ( C)

Figure 6.8

102

 

TDI’ERATIIEIEANSACHJSSEJIJCS
ENSURE!

——LOV €51) mISTIRE CIBS
——-—IEILII (131) 1015115 com
HIGH (181) IIIISTIRE CW5

 

 

, ,/ ‘\/”‘\ ./\.3/ ‘C/—._"

 

 

 

 

ELAPSED TIIE 1N saunas

Gas temperature profiles at the cooler
exit for 500 rpm. Averaged across 3 blocks
for the 10 minute length of the test.

 

GAS EXIT TEFP FRO! COOLER (C)

103

 

Iaoauneuawsmmmsaunm:
Rhianna

-——LOI (59 101511]!
—-- all! (139 1015115 UB5
HIGH (189 lIJISTIRE ems

 

 

._/ K-Jr—Mv “\,"\\/—-

 

w ._._.\/,_._

 

 

ELAPSED THE IN SECWOS

Figure 6.9 Gas temperature profiles at the cooler

exit for 600 rpm. Averaged across 3 blocks
for the 10 minute length of the test.

 

104
detected by the thermistor (channel 3) range from 23
degrees to 28 degrees. At the maximum, only 4 degrees
separated the warmest and coolest treatment averaged at any
given engine speed. Suprisingly, the high cob moisture
averages were the warmest of all moisture levels and the
low cob moisture averages were the coolest. This was the
case at all engine speeds. This outcome was possibly due to
three principal conditions: 1) the higher cob moisture
profiles started at higher temperatures at the gasifier
grate than the low cob moisture profiles. This heat was
transferred to the water in the cooler, raising the water
temperature. The heat stored in the water was not rapidly
dissipated, and thus the temperature was maintained at a
higher level than would be the case with a low moisture
fuel that did not raise the gasifier grate temperature as
high, 2) the water was not stirred, agitated, or pumped to
a heat exchanger to remove heat from the water or
distribute the heat through the mass of water, 3) the
cooling hose was positioned at the top of the gas cooler
unit such that it was two thirds submerged. Since heat
rises, the warmest water would have remained in the
vicinity of the hose, and maintained the temperature at

equilibrium.
6.4.1 Gas Cooler Exit temperature Results at 400 RPM

The low percent cob moisture average was steady from

the beginning of the run, rising about 0.5 degrees to 24.5

105
degrees celsius at the mid point of the test. This was the
lowest temperature at 400 rpm.

The 13 percent cob moisture averages were confined to
the 25.5 to 26.0 degree range for practically the entire
test, moving up to 26.5 degrees at the very end.

Up at the 18 percent moisture level temperatures were
steady at 28 degrees for more than half the duration of the
test. An increase of 0.5 degrees occurred during the final

3 minutes.
6.4.2 Gas Cooler Exit Temperature Results at 500 RPM

In this series of profiles the low percent moisture
averages began at slightly less than 24 degrees and
remained at this level until the mid point of the test.
From there it increased 1 degree in a step like fashion.
The test expired with the temperature steady at 25 degrees.
This was half a degree higher than the 5 percent moisture
profile at 400 rpm.

Initial temperature levels for the medium cob moisture
averages were around 25.5 degrees. There was a jump to 26.5
degrees about 1.5 minutes into the test. After a period of
choppy behavior the temperature settled to a level between
26 and 26.5 degrees. This was the same level as the medium
moisture profile at 400 rpm, and 1.5 degrees higher than
the low percent moisture level at 500 rpm.

For the high cob moisture profile, temperatures were

confined to a 0.5 degree range between 26.5 and 27.0

106
degrees for the entire test. This moisture level yielded
the highest temperatures at this engine speed. However this
was 1.5 degrees lower than the same moisture profile at 400

rpm.
6.4.3 Gas Cooler Temperature Results at 600 RPM

Steadily rising temperatures characterized the low
percent cob moisture profile (figure 6.9) for three
quarters of the test. Subsequently, temperatures stabilized
in the 24.5 to 25 degree range. This was the same area that
the 400 rpm and the 500 rpm profiles attained.

The only real decrease in temperature occurred with
the medium cob moisture level in this group of tests.
Beginning at 26 degrees the temperature rose to nearly 27
degrees after 3.5 minutes. This was followed by a steep
decline to just above 25 degrees, where it remained for the
rest of the test. This was about 1 degree lower than the
outcome of the 400 and 500 rpm profiles, and 0.5 degrees
above the 5 percent moisture profile at 600 rpm.

At 25 percent cob moisture temperatures at 600 rpm
were the most stable of all tests at this speed (figure
6.9). Varying between 27 and 28 degrees, the profile stayed
within these limits for the entire test. Again, this
profile contains the highest temperatures recorded at this
speed, but this was only three degrees above the lowest
profile which was the low (5 percent) cob moisture

temperature profile.

107

6.5 Fuel Comsumption Test Results

Given the variations in engine speed and engine power
encountered in the experiment, tests of significance were
performed on both of these parameters. The factorial
design was employed to accomplish this since it was capable
of testing significance for the factors involved, and test
for significant differences between blocks. The tests of
significance for speed and power were considered over the
three levels of the moisture factor, and for blocks. It
was these two conditions that determined if the differences
in the observations were large enough to be attributed to
causes other than chance.

The engine speed data of Table 6.2 was used in the
factorial analysis of variance program to test for
significant differences between moisture levels and between
blocks. The ANOVA (Analysis of Variance) Table is table
6.6. To determine significance at the 5% level the
calculated P value in the ANOVA table was compared with the
tabular F value with 2 degrees of freedom for the numerator
and 16 degrees of freedom for the denominator. This P
value was 3.63, and was larger than the calculated F values
in Table 6.6. Therefore the observed difference between
moisture levels and between blocks were not significant at
the 5% level. However at the 10% level, speed differences

between moisture contents do become significant.

108

The data obtained on engine power was compiled in
Table 6.4. That data comprised observations on all nine
treatments of each block. These values were used to
obtain the ANOVA results of Table 6.7. Comparing the
calculated F value, with the tabular F value, revealed no
significant differences between blocks or between moisture
levels. This was true at both the 5 and 10 percent
confidence levels.

The two analyses mentioned above confirm that the
engine speed and power data were not sufficiently different
at the 5% level to void the objectives of the experiment,
although engine speed was barely significant at the 10%
level. With this in mind the ANOVA for diesel fuel

consumption follows.
6.5.1 Diesel Fuel Mode

The reduced power levels at 400, 500 and 600 RPM were
duplicated to measure the amount of diesel fuel consumed at
each of these speeds. Diesel fuel consumption was
measured using a 250ml graduated cylinder (graduations of
2ml) with a siphon tube attached to the engines' fuel line.
The power output of the engine was measured using the load
cell-prony brake system described earlier in this document.
The digital data acquisition system accumulated readings
for the 5 minute long tests at a rate of one point per
second for force and engine speed. A digital multimeter

was alternately connected to the speed and force sensors to

109

Engine Speed ANOVA Table

1235.67
204867.0
2155.56
1010.77
6271.33

617.835
102434.0
1077.78
252.693
391.958

261.339**
2.74973

0.64469

Table 6.6

Source df
Blocks 2
A (Speed) 2
B (Moisture) 2
AB 4
Error 16
Total 26

Table 6.7

Source df

Blocks 2

A (Speed) 2

B (Moisture) 2

AB 4

215540.33

Engine Power ANOVA Table

0.005925

2.97252
0.008633
0.006413
0.043505

0.0029625
1.48626
0.004317
0.001603
0.002719

1.08953
546.608**
1.58756
0.58967

110

monitor the speed and force magnitudes during the test.
Minor adjustments to the speed and the prony brake were
made necessary.

The power levels employed were 0.63 KW (0.85HP) at 400
RPM 0.87 KW (1.17 HP) at 500 RPM and 1.3 KW (1.80 HP) at
600 RPM. Once the power level was adjusted, fuel
consumption tests were run for 5 minutes. The 5 minutes
consumption figures were 46.00 ml at 380 RPM, 67.5 ml at
500 RPM, and 81.5 ml at 600 RPM. These values served as the

basis for comparison with the dual fueled results.
6.5.2 Dual Fueled Mode

The diesel fuel tests performed in this experiment
used both fuel and producer gas from corn cobs to power the
engine, since diesel fuel was necessary to initiate
combustion. In essence the more energy the producer gas
contained, the less diesel fuel was consumed. This will
occur to the point at which the amount of diesel fuel
injected into the cylinder was insufficient for combustion.
The engine then slowed until the governor permitted the
injection of enough diesel fuel for ignition again.

The tests spanned a period of 3 weeks. Five
treatments were re-applied during this period because of
incorrect adjustments and/or calibration of instruments
and/or equipments. A major difficulty was maintaining the
moisture levels of the high moisture cobs employed in the

test. During the execution of the block 1 treatments,

111

moisture loss from the 15 and 25% moisture cob was evident.
Modification in the conduct of the experiment diminished
the moisture loss but did not eliminate it altogether. It
was most pronounced in the 25% moisture material, loosing
up to 7% points wet basis in one day. To a lesser degree
engine speed and power also posed some difficulty in
adhering to desired levels. Details of these conditions
were provided in previous sections of this chapter.

The diesel fuel consumption data for the experiment
are in Table 6.8. This was the absolute diesel
consumption for blocks(replicates) l, 2, and 3. The
general trend in that data follows expectations, which
were; A) increasing diesel fuel consumption with increasing
engine speed and load, and 8) increasing diesel fuel
consumption with increasing cob moisture content. There
were 2 exceptions out of the 29 data points. One at the
5% nominal moisture level at 400 and 500 rpm, the other at
500 rpm between the 15 and 25% moisture levels.

The objectives of the experiment were to 1) determine
which of the moisture levels most reduced the consumption
of diesel fuel, and 2) determine whether the differences in
diesel consumption due to moisture were significant. The
data in Table 6.8 contains the observations on diesel
fuel consumed and related statistics. The values in the
table have been converted from ml per 10 minutes of test to
ml per kilowatt hour. The data forms the basis of the

analysis of variance shown in Table 6.9.

112

Table 6.8 Diesel Fuel Comsumption Treatment Data - Dual Fueled Mode

lst Letter : L - Low Speed M - Medium Speed H a High Speed
2nd Letter : L - Low Moisture M . Medium Moisture H - High Moisture

LL LM LH ML MM MH HL HM HH Totals
Block 1 22 40 41 16 48 50 36 66 85 404
Block 2 17 18 23 36 56 57 44 60 83 394
Block 3 23 35 42 26 48 46 47 73.5 80.5 421
ESZQI; '''' ES """ 53“"163“"‘“§é""'155""'I§3""IE§"'I§§f§"3£éT§-'i5i§"
Means 20.67 31 35.33 26 50.67 51 42.33 66.5 82.83 45.15

Table 6.9 Diesel Consumption ANOVA Table

Source df SS Mean Square F
‘éiééi; """" 5 """ ii?$"""‘26?$ """ 6:3é6365'
A (Speed) 2 5568.29 2784.15 47.1291**
8 (Moisture) 2 3456.07 1728.04 29.2516**
AB 4 609.04 152.26 2.5774
Error 16 945 20 59 075

Total 26 10620.0

113

The test of significance was performed using the F-
test. For both factors the numerator degrees of freedom
was 2 and the denominator degrees of freedom was 16.
Dividing the factor mean squares by the error mean square
resulted in significant outcomes for both factors at the 5%
level. Asteriks adjacent to figures in the F column
indicate significance. The interaction, line AB in the
ANOVA table, was not significant but it was close to being
significant at the 10% level. More information on
interaction was revealed by constructing response scales
for each factor. Interaction was important because it
revealed if the response of one factor to the other (or
others) was of a dependent or independent nature.

Figure 6.10 shows the response curve for factor B at
factor A engine speed. Interaction was likely when the
curves displayed change in magnitude (lepe) or a change in
direction on the interval between any two levels of the
factor represented by the X-axis. From these curves it
appears that interaction was at hand. The slope of the
medium moisture content curve (factor B) was different than
for the low and high moisture curves below and above it.
However since the curve in question stayed within the
bounds established by the other two curves, its response
was not radical enough to cause outright significance.

In figure 6.11, the response curves for factor A at
factor B, cob moisture content, were diagrammed. The

evidence for interaction in this case was diminished

114

 

 

 

 

 

1.
II LOW (51) IOISTLRE UB3
8' r- ‘ Islam: (131) MIST“
0 HID! (182) 10151115
8
8 an»
i
I
g u h
m
z. ..
I L 1 L
an 4U 8U

FACT!!! A - 848115 S’EED

Figure 6.10 Diesel fuel consumption at low, medium
and high cob moisture levels. Engine
speed points are 373 rpm, 496 rpm, and
585 rpm.

115

 

 

 

 

 

 

1-
FACTUIA
II MMMIES’EED
‘ m. C I
U- o
g
8
3 an-
i
I
3 "L
3
23(-
a l l l 1
fl 5 ll 15 23

Pm B - IDISTIRE (I.D.)

Figure 6.11 Diesel fuel consumption at low, medium
and high engine speed levels. Moisture
points (wet basis) are 5%, 13%, and 18%.

116
considerably compared to the previous one, though there
were some hints of interaction. Again it was the middle
curve, 496 rpm that showed magnitude changes different from
the other two. As was the case in the previous response
curve, the apparent interaction was not radical enough to
affect the test of significance. In fact, there was a
strong possibility that these manifestations were not
interaction at all, but variance attributable to chance, or
bias in the experimental conduct i.e. variation in moisture
level of the cobs. It was impossible to tell what the
source of these conditions were without further

experimentation focusing on these effects.
6.5.3 Regression Curve Fit to Fuel, Power and Speed Means

These means of table 6.8 were converted to fuel
consumption in ml/KWH (milliliters per kilowatt hour). The
results were included in Table 6.10. The data in Table
6.10, were combined with engine speed data of Table 6.3, to
perform four linear regressions, one for the diesel fuel
consumption in diesel only mode and one for each of the 3
moisture levels. The composite plots of these fitted
regression curves are in figure 6.12. This figure shows
the relationship of fuel consumption at the various
moisture levels, to each other, and to fuel consumption on
diesel fuel alone. It was evident in this representation
that low moisture fuel produced the highest diesel fuel

substitution effect.

117

 

 

 

 

 

-
I LOVGnIIJISTIREGJRVEI)
‘ Elmclmmlmma
O HIGIQBDIIJIS‘MIEGJRVES)
u_ 9 DIESELFIELGLYmv
S o
E o
2
H 4"-
; .
O 4
k 0 *3
‘ 6
3 an_- ‘ y‘_ 2
d
§i
D
an ‘5
a
' 1
in: L . .
a 4' SE a
“Hem

Figure 6.12

Linear regression plots of diesel fuel
consumption in ML/KWH. Curves 1,2, and 3
are the low medium and high moisture levels
wet basis. Curve 4 is pure diesel fuel.

 

118

The consumption of corn cobs was recorded though not
as precisely as the diesel fuel consumption. Only 2
replicates of data were acquired, one for block one and one
for block 2. The measurement method was a simple pan type
container with a volume of 1,375m1. The number of
containers used during a treatment was noted, and this
value was multiplied by the volume of the container to get
the volume of cobs consumed. The results are included in
Table 6.11. Though the process was crude the trends in
the data show that as moisture content increased the volume
of cobs comsumed decreased. This was consistent with the
diesel fuel consumption results explained elsewhere.
Naturally, the drier and easier burning cobs were consumed
more rapidly than wet ones which don't burn as easy. Thus
the proportion of energy supplied by each fuel varied with
the moisture content of the corn cobs. The drier the cobs

the more cobs and less diesel fuel consumed and vice versa.

Table 6.10 Diesel Fuel Consumption in ML/KWH

Low Speed Medium Speed High Speed
Low H20 205.32 166 5 181.49
Med H20 312.96 323.57 290.8

119

Table 6.11 Corn Cob Comsumption Data and Statistics

Block 1 Block 2 Total Mean

Low | 1.71 2.275 3.985 1.99
Speed IM 1.422 2.275 3.697 1.848
1.479 1.71 3.189 1.59
Med I 2.559 2.275 4 834 2.417
Speed IM 1.71 0.85 2.56 l 28
H 1.71 0.85 2.56 1.28

Hi I 2.275 2.559 4 834 2.417
Speed IM 1.71 2.275 3.985 1.99
H 1 422 1 422 2.844 1 422

L = low moisture, M = medium moisture cobs, H =' high

moisture cobs.

6.6 Hypothetical Irrigated Farm Model Using Gasification

The experimental results indicated an average diesel
fuel savings of 55 percent at 600 rpm, and 56 percent at
400 rpm when low moisture corn cobs (5% wet basis) were
used. The value of such savings in a typical production
situation is the theme of this section.

The fuel consumption results were applied to a
hypothetical farm in the Sahel region of Africa. The power
of the engine was scaled up to 4.47 kw instead of 1.39 kw.
This was done to take advantage of the most of the optimum
power for use in the farm model. The engine chosen for the
model was the 4.47 kw — 650 rpm Eicher Goodearth model 88W-

1. This unit had a maximum speed near that studied in the

120
research experiment, and a low power output. The engine was
a product of India and the same firm also manufactured
pumps for their engines. The 125 mm X 125 mm, 1500 rpm pump
was selected for this application, since data available
from Jenkins (1984) show the static head to be less than 10
meters in northern Mali. The technical specifications for
the engine and pump were available in the Eicher Goodearth

Limited brochure provided in the appendix.
6.6.1 Engine Power Derating and Pump Discharge Capcity

Since producer gas has much less energy per unit
volume than diesel fuel, the power produced by an engine
will be lower than its normal rating at a given engine
speed (Cruz, 1980; Johansson, 1980). Derating the power
output of the engine by 30 percent provided the probable
power when operating on producer gas. This was 3.13 kw for
the engine used in the farm model. This power was employed
in the financial cost - benefit analysis and used to
determine the capacity of the surface irrigation system and
the level of fuel consumption for both diesel and biomass
fuels. Table 6.12 shows the pump discharge rates at 6, 8,
and 10 meter heads. Since engine power was reduced by 30

percent so were the discharge rates for the pump.

121

Table 6.12 Model Farm Pump Discharge Rates

Head in Flaw in Flow Reduced
Meters m /hr by 30 %

6 126 88.2

8 111.6 78.1

10 90 63

6.6.2 Water Requirement for Corn

The water requirement for corn in arid areas was 0.774
cm per day peak consumptive use for a yield of 5.65 tons/ha

(90 bu/acre) in the Texas plains (Turner and Anderson,

3 per hectare per day.

This value was increased by 30 percent to 100.6 m3 per day

1971). This was equivalent to 77.4 m

to account for losses due to surface irrigation. Carruthers
and Clark (1983) cites surface irrgation application
efficiency as being around 70 percent. It was further
assumed that the pumping plant and equipment operate a
maximum of 10 hours per day. This was typically the daily
pumping period observed in the Timbuctu region of Mali.
Given these operational parameters the land area irrigable
at each of the specified heads mentioned in table 6.13 were
8.77 hectares at 6 meters, 7.77 hectares at 8 meters and

6.26 hectares at 10 meters.

122

6.6.3 Quantity of Corn Cobs Available for Gasification

Mantuffel and Tyner (1980) cited yields of 3.0 tons
(3,000 kg) per hectare for corn in the Senegal River
Valley. This value was assumed for the corn yield in the
farm model. A higher yield of 5.65 tons per hectare was
employed in a sensitivity analysis included in subsequent
sections of this chapter. Payne (1979) reported that
corn cobs equal approximately 18.6 percent of the grain
yield. Corn cobs at harvest were usually high in moisture.
rFor this model the moisture content at harvest was assumed
to be 30 percent wet basis. Relative humidity levels in the
Sahel zone of Africa were low during the irrigation season
from October to March. Figure 6.13 shows the mean humidity
levels for Africa in October. The Sahel region averaged 30-
50 percent relative humidity for October. In conjunction
with the data provided by White (1984) on equilibrium
moisture content of corn cobs (figure 3.4 and 3.5) it was
evident that equilibrium corn cob moisture levels will fall
to the 5 percent wet basis range. Thus the weight of the
cobs was reduced by 25 points to the 5 percent moisture
level, for the farm model. These weights are listed in
table 6.13 for the three surface areas irrigable by the

system.

123

 

 

cu
Q5
‘1

gsfi

.
.
2‘:

 

 

 

 

 

 

 

 

 

 

  

 

“3 mmmmEtva F3
naxmm mmumw ‘
FEM! mmmw
.1 ’5 OCTOBER . . (’5
"3 x
19w R" S o S no IS % 25 45 E

 

 

Figure 6.13 Mean Relative Humidity for Africa for the
month of October. Source: FAO, 1980

76

124

Table 6.13 Corn Cob Produced on the Model Farm

Area (ha) Weight at 30% Weight at 5%
moisture content moisture content
8.77 4,893.7 kg 3,610.0 kg
7.77 4,335.7 " 3,195.0 ”
6.26 3,493.1 " 2,574.0 "

The experimental data for corn cob consumption were
expressed in terms of volume in liters per 10 minutes of
test. The experimental values indicated that there were
165.5 grams per liter of cobs. Therefore 3,610 kgs of cobs

3 of cobs. Corresponding values

were equivalent to 21.8 m
for the remaining two field sizes were 19.3 m3 and 15.55

m3 for the 7.77 ha and 6.26 ha fields respectively.
6.6.4 Corn Cob Consumption

The experimental data showed that 2.42 liters fo cobs
were consumed during a 10 minute test period at 600 rpm and
1.4 kw of power. This translates to 14.52 liters/hour. This
value was inflated arbitrarily by 30% to be conservative in
deriving the financial benefit as a result of employing the
gasification system. The inflated corn cob consumption
figure was 18.88 liters/hour. At 1.4 kw the specific fuel

consumption became 13.48 liters/kw-hr.

125

6.6.5 Theoretical Corn Cob Consumption on the Model Farm

The engine power on the hypothetical farm was 3.13 kw.
Given the experimental cob consumption of 13.48 liters per
kw-hr,the 3.13 kw engine consumed 42.19 liters of corn cobs
per hour. If the pump was operated for 10 hours a day, the
daily cob consumption would have been 421.9 liters or
0.4219 m3. Table 6.14 shows the number of days and the

percent of the season covered by gasifier operation.

Table 6.14 Effective Seasonal Working Days Using
Producer Gas in the Farm Model

Area (ha) Number of days % of 105 day
operation w/gas season covered

8.77 51.7 49.24

7.77 45.75 43.57

6.26 36.86 35.1

6.6.6 Diesel Fuel Consumption

Fuel consumption figures in grams per kw—hr ranged
from a high af .326 kg/kw-hr to a low of .265 kg/kw-hr
(Eicher Goodearth Ltd.,1984; Overseas Ltd., 1984),
according to specifications recieved from manufacturers.
The higher figure was used in the analysis, so the
predicted benefit would be conservative.

An engine producing 3.13 kw at 650 rpm required 1.02

126
kg of diesel fuel per hour of operation. Taking 1 gram of
diesel to be approximately equal to 1 ml in volume, diesel
fuel consumed became 1,020 ml per hour or 1.02 liters per
hour.

The seasonal diesel fuel consumption was obtained by
multiplying the hourly consumption by the number of hours
operated per day (10), and multiplying the result by the
number of days of irrigation per season. The outcome was

1,071 liters per 105 day season.
6.6.7 Diesel Fuel Saved by Using Gasification

Gasification operated the engine - pumpset for 49.24 %
of the season for the largest field (8.77 ha). Diesel fuel
consumption was reduced by 55 % while producer gas was
used. Thus the amount of diesel fuel saved in a typical

season is:

Low Moisture 1,071 l/sn X 0.4924 X 0.55 290.05 liters/sn

Med " 1,071 l/sn X 0.4924 X 0.20 105.47 " "

High " 1,071 1/sn X 0.4924 X 0.113= 59.59 " "

therefore 290.05 liters of diesel fuel were saved each
season using gasification to operate a diesel powered pump,
pumping at 6 meters of head to irrigate 8.77 ha of corn
field. This was equivalent to reducing seasonal diesel fuel
consumption by 27.08%. At medium cob moisture content

seasonal diesel fuel consumption is reduced by 9.85%, and

127
at high cob moisture diesel fuel consumption is reduced by

only 5.56%.

6.7 Financial Cost - Benefit Analysis

The financial analysis procedure used was the
financial internal rate of return, based on that described
by Gittinger (1982) in his work on project analsis. With
this method the incremental net benefit stream or
incremental cash flow is used to derive the internal rate
of return. The incremental net benefit stream is obtained
by subtracting the "with" gasification systems costs from
the "without” gasification system costs. The "with"
gasification costs include the cost of the gasification
equipment, the added cost of maintenance, and the costs of
fuel both diesel and corn cobs. The financial internal
rate of return is a measure of investment worth. It is a
measure of the rate of return on capital outstanding per
period, during the time that capital is invested. The
financial internal rate of return is the discount rate that
yields a net present worth of the incremental net benefit
stream equal to zero.

Initially three scenarios were considered, all of
which concern corn cob moisture content. The effect on
financial attractiveness of the gasification system was
evaluated, as the corn cob moisture content increased.

In these scenarios it was assumed that the cost of

128

corn cobs was zero, the cost of diesel fuel was $0.65 per
liter, and that the total cost of the gasification system
was $626.00. The cost of the pump set including the engine
was $1,000.00, 2.99 times the f.o.b. price ($355.00) at the
shipping point from the country of origin (pump price taken
from Eicher Good Earth Ltd. 1984 price list). Maintenance
cost for a normal diesel was 20% of the purchase price per
year. Maintenance for a diesel gasification system was
calculated_ as that of a regular diesel unit plus 14%
(Groenveld, et al). The cost of a gasifier system operator
was excluded because the irrigation pumps in the Sahel
typically had an attendant on hand to refuel the engine,
check the motor oil, start and stop the engine, and deter
theives that would otherwise steal fuel and parts from the
pump set. It was assumed that this person would handle
loading of corn cobs into the gasifier and related duties.
The life of the system was assumed to be 5 years. The
varying factor in these scenarios is cob moisture content,
which imply higher diesel fuel consumption as moisture
increases. Otherwise, the above assumptions are equal for
each of these scenarios.

Tables 6.15 to 6.17 contain the partial budgets for
the three scenarios with and without gasification. The
incremental net benefit and financial rates of return that
correspond to the partial budgets are shown in tables 6.18

to 6.20.

129

Table 6.15 Partial Budget for Irrigation Costs, With

and Without Gasification Fuel,for a Model Farm of 8.77ha
Low Moisture Cobs at Zero Cost - Diesel Fuel Cost of
$0.65 per liter - Grain Yield of 3 tons per hectare

Item w/o gas Y1 Y2 Y3 Y4 Y5
Gasifier Cost 0 626 - - - -
Maintenance 200 228 228 228 228 228
Diesel Fuel 696 508 508 508 508 508
Biomass Fuel 0 O 0 0 0 0
Total 896 1362 736 736 736 736

Table 6.16 Partial Budget for Irrigation Costs, With

and Without Gasification Fuel,for a Model Farm of 8.77ha
Medium Moisture Cobs at Zero Cost - Diesel Fuel Cost of
$0.65 per liter - Grain Yield of 3 tons per hectare

Item w/o gas Y1 Y2 Y3 Y4 Y5
Gasifier Cost 0 626 - - - -
Maintenance 200 228 228 228 228 228
Diesel Fuel 696 635 635 635 635 635
Biomass Fuel 0 0 0 0 0 0

130

Table 6.17 Partial Budget for Irrigation Costs, With

and Without Gasification Fuel,for a Model Farm of 8.77ha
High Moisture Cobs at Zero Cost - Diesel Fuel Cost of
$0.65 per liter - Grain Yield of 3 tons per hectare

Item w/o gas Y1 Y2 Y3 Y4 Y5
Gasifier Cost 0 626 - - - -
Maintenance 200 228 228 228 228 228
Diesel Fuel 696 669 669 669 669 669
Biomass Fuel 0 0 0 0 0 0
$32; """""" 5'92""753 """ 5;§""§§§""§§§""§§§"

Table 6.18 Incremental Net Benefit and Financial Internal
Rate of Return for an 8.77 ha Model Farm, Low Moisture
Cobs at Zero Cost - Diesel Fuel Cost of $0.65 per liter
Grain Yield of 3 tons per hectare

Year Cost W/O Gas Cost W/Gas Increm Net NPW
Benefit 14%
l 896 1362 -466 ~409
2 896 736 160 123
3 896 736 160 108
4 896 736 160 95
5 896 736 160 83
--6-

Financial Internal Rate of Return = 14%

131

Table 6.19 Incremental Net Benefit and Financial Internal
Rate of Return for an 8.77 ha Model Farm, Medium Moisture
Cobs at Zero Cost - Diesel Fuel Cost of $0.65 per liter
Grain Yield of 3 tons per hectare

Year Cost W/O Gas Cost W/Gas Increm Net NPW
Benefit 0%

1 896 1489 -593 -593

2 896 863 33 33

3 896 863 33 33

4 896 863 33 33

5 896 863 33 33
£81"

The Financial Internal Rate of Return is negative since the
total Net Present Worth at 0% discount rate is negative.

Table 6.20 Incremental Net Benefit and Financial Internal
Rate of Return for an 8.77 ha Model Farm, High Moisture
Cobs at Zero Cost - Diesel Fuel Cost of $0.65 per liter
Grain Yield of 3 tons per hectare

Year Cost W/O Gas Cost W/Gas Increm Net
Benefit
1 896 1523 -627
2 896 897 -l
3 896 897 -l
4 896 897 -l
5 896 897 -l

The Financial Internal Rate of Return is negative since the
entire net benefit stream is negative.

132
Of the above three scenarios only that of the low cob
moisture has a positive internal rate of return. Thus under
the given assumptions, low moisture cobs are necessary for
the gasification system to operate more profitably than a

pure diesel system.
6.7.1 Inclusion of Cob Costs

Corn cob costs were considered in 3 additional
scenarios. These involved costs of $1.00 per m3, $3.00 per
m3, and $5.00 per m3. These cob costs were only analysed
with the low moisture cobs since that was the only out come
with a positive financial internal rate of return.

In previous sections of this chapter it was shown that
a grain yield of 3 tons per hectare will provide 22 m3 of
corn cobs from an 8.77 ha farm. This quantity was used as
the seasonal consumption of cobs, and thus was the basis
for determining the cost of cobs in the partial budgets.
Tables 6.21 to 6.23 contain the partial budgets for these
scenarios. Following these are the corresponding

incremental net benefit streams and the financial internal

rates of return, in tables 6.24 to 6.26.

133

Table 6.21 Partial Budget for Irrigation Costs, With

and Without Gasification Fuel,for a odel Farm of 8.77ha
Low Moisture Cobs at cost of $1.00/m - Diesel Fuel Cost
of $0.65 per liter - Grain Yield of 3 tons per hectare

Item w/o gas Y1 Y2 Y3 Y4 Y5
Gasifier Cost 0 626 - - - -
Maintenance 200 228 228 228 228 228
Diesel Fuel 696 508 508 508 508 508
Biomass Fuel 0 22 22 22 22 22
$3231 """""" 393""735; """ 3%;""5éé'm'7'é'5'miéé"

Table 6.22 Partial Budget for Irrigation Costs, With

and Without Gasification Fuel,for a Model Farm of 8.77ha
Low Moisture Cobs at cost of $3/m3- Diesel Fuel Cost of
$0.65 per liter - Grain Yield of 3 tons per hectare

Item w/o gas Y1 Y2 Y3 Y4 Y5
Gasifier Cost 0 626 - - - -
Maintenance 200 228 228 228 228 228
Diesel Fuel 696 508 508 508 508 508
Biomass Fuel 0 66 66 66 66 66

134

Table 6.23 Partial Budget for Irrigation Costs, With

and Without Gasification Fuel,for a Model Farm of 8.77ha
Low Moisture Cobs at cost of $5/m3- Diesel Fuel Cost of
$0.65 per liter - Grain Yield of 3 tons per hectare

Item w/o gas Y1 Y2 Y3 Y4 Y5
Gasifier Cost 0 626 - - - -
Maintenance 200 228 228 228 228 228
Diesel Fuel 696 508 508 508 508 508
Biomass Fuel 0 110 110 110 110 110
ESEQI """""" 353""‘IZ35 """ 523’"'QZE""§ZE""§ZE"

Table 6.24 Incremental Net Benefit and Financial Internal
Rate of Return for an 8.77 ha Model Farm, Low Moisture
Cobs at cost of Sl/m - Diesel Fuel Cost of $0.65 per
liter - Grain Yield of 3 tons per hectare

Year Cost W/O Gas Cost W/Gas Increm Net NPW
Benefit 5% 6%

1 896 1384 -488 -465 -460
2 896 758 138 125 123
3 896 758 138 119 116
4 896 758 138 114 109
5 896 758 138 108 103

"i'_':§

The Financial Internal Rate of Return is for this scenario
is 5.10%

135

Table 6.25 Incremental Net Benefit and Financial Internal
Rate of Return for a3 8.77 ha Model Farm, Low Moisture
Cobs at cost of $3/m - Diesel Fuel Cost of $0.65/liter
Grain Yield of 3 tons per hectare

Year Cost W/O Gas Cost W/Gas Increm Net NPW
Benefit 0%

l 896 1428 -532 -532

2 896 802 94 94

3 896 802 94 94

4 896 802 94 94

5 896 802 94 94
-iéé'

The Financial Internal Rate of Return is negative since the
total Net Present Worth at 0% discount rate is negative.

Table 6.26 Incremental Net Benefit and Financial Internal
Rate of Return fog an 8.77 ha Model Farm, Low Moisture
Cobs costing $5/m - Diesel Fuel Cost of $0.65 per liter
Grain Yield of 3 tons per hectare

Year Cost W/O Gas Cost W/Gas Increm Net NPW
Benefit 0%

1 896 1472 -576 —576

2 896 846 50 50

3 896 846 50 50

4 896 846 50 50

5 896 846 50 50
—§§E’

The Financial Internal Rate of Return is negative since the
total Net Present Worth at 0% discount rate is negative.

136
Of the three cob cost scenarios, only one had a
positive financial internal rate of return (IRR). The
outcome with low moisture cobs at a cost of $1.00 per m3,
and diesel fuel cost of $0.65 per liter had a financial of
5.1 percent. The other outcomes had negative financial
IRR's, indicating that raising cob prices above the $1.00

per m3 level had a negative financial effect on the

gasification system.
6.8 Sensitivity Analysis

Certain assumptions in the scenarios of the two
preivious sections, such as water use, grain yield, and
diesel fuel price, might have been too conservative. In the
following sections of this chapter, these assumptions were
modified by increasing them and recalculating the outcomes.
This process revealed how sensitive the attractiveness of
the gasification system is to changes in important

assumptions.
6.8.1 Increased Corn Yield

The grain yield was increased from 3.0 tons to 5.65
tons per hectare. The higher figure corresponds to the
yield cited by Turner and Anderson (1972) for a water use
rate of 0.774 cm per day (see section 6.6.2). The higher
yield would make more cobs available for operating the
gasification system. The more cobs are used the less diesel

fuel is used, thus increasing the benefits of gasification.

137
The higher corn yield provided more corn cobs. A yield of

3 3 of cobs

5.65 tons per hectare provided 41 m versus 22 m
from a corn yield of 3.0 tons per hectare. From section
6.6.5 it was calculated that the gasification system would

3 of low moisture

consume 0.422 m3 per day. Therefore 41 m
cobs are sufficient for 97.24 days of operation. The
irrigation season covered 105 days, thus 41 m3 of cobs
permitted the gasification system to operate for 92.61
percent of the season. Diesel fuel consumption without
gasification was 1,071 liters (section 6.6.6). The diesel

fuel saved by the gasification system at the higher yield

of 5.65 tons of grain per hectare is as follows:

Low Moisture Cobs 1,071 liters X 0.9261 X 0.55 = 545.5 1

Medium " " " " X " X 0.20 198.37 1

112.081

High " " " " x " x 0.113

Where 0.9261 is the percent of the irrigation season
that gasification was operable, and 0.55, 0.20, and 0.113
are the percents of diesel fuel saved at low, medium, and
high cob moisture contents respectively, when gasification
was employed. These savings translated to a seasonal
reduction in diesel fuel consumption of 50.9%, 18.5%, and
10.5% for low, medium, and high cob moistures respectively.

Taking these savings into account, the seasonal cost

of diesel fuel for the 8.77 hectare model farm was:

138

Low Moisture Cobs (1071 l - 545.5 l)X $0.65 = $341.58
Med " " (1071 1 - 198.371)X $0.65 = $567.21
High " " (1071 1 - 112.081)X $0.65 = $623.30

These are the values contained in the partial budgets of
tables 6.27 to 6.29.

The financial internal rate of return, incremental net
benefit, and discountet net present worth for the three

outcomes cited above, are contained in tables 6.30 to 6.32.

Table 6.27 Partial Budget for Irrigation Costs, With

and Without Gasification Fuel,for a Model Farm of 8.77ha
Low Moisture Cobs at Zero Cost - Diesel Fuel Cost

of $0.65 per liter - Grain Yield of 5.65 tons per hectare

Item w/o gas Y1 Y2 Y3 Y4 Y5
Gasifier Cost 0 626 - . - - -
Maintenance 200 228 228 228 228 228
Diesel Fuel 696 342 342 342 342 342
Biomass Fuel 0 0 0 0 0 0

139

Table 6.28 Partial Budget for Irrigation Costs, With
and Without Gasification Fuel,for a Model Farm of 8.77ha
Medium Moisture Cobs at Zero Cost - Diesel Fuel Cost
of $0.65 per liter - Grain Yield of 5.65 tons per hectare

Item w/o gas Y1 Y2 Y3 Y4 Y5
Gasifier Cost 0 626 - - - -
Maintenance 200 228 228 228 228 228
Diesel Fuel 696 567 567 567 567 567
Biomass Fuel 0 0 O 0 0 0
«E52; """""" 553”“‘IZETWEQM'Tzéé""}'§§""§§%"

Table 6.29 Partial Budget for Irrigation Costs, With

and Without Gasification Fuel,for a Model Farm of 8.77ha
High Moisture Cobs at Zero Cost - Diesel Fuel Cost

of $0.65 per liter - Grain Yield of 5.65 tons per hectare

Item w/o gas Y1 Y2 Y3 Y4 Y5
Gasifier Cost 0 626 - - - -
Maintenance 200 228 228 228 228 228
Diesel Fuel 696 623 623 623 623 623
Biomass Fuel 0 0 0 0 0 0

140

Table 6.30 Incremental Net Benefit and Financial Internal
Rate of Return for an 8.77 ha Model Farm, Low Moisture
Cobs at Zero Cost - Diesel Fuel Cost of $0.65 per liter
Grain Yield of 5.65 tons per hectare

Year Cost W/O Gas Cost W/Gas Increm Net NPW
Benefit 90%

1 896 1196 -300 -158
2 896 570 326 90
3 896 570 326 48
4 896 570 326 25
5 896 570 326 13
-1;-

The Financial Internal Rate of Return is greater than 90%.

Table 6.31 Incremental Net Benefit and Financial Internal
Rate of Return for an 8.77 ha Model Farm, Med Moisture
Cobs at Zero Cost - Diesel Fuel Cost of $0.65 per liter
Grain Yield of 5.65 tons per hectare

Year Cost W/O Gas Cost W/Gas Increm Net NPW
Benefit 0%

1 896 1421 -525 -525

2 896 795 101 101

3 896 795 101 101

4 896 795 101 101

5 896 795 101 101
-121-

The Financial Internal Rate of Return is negative since the
total NPW is negative at 0% discount rate.

141

Table 6.32 Incremental Net Benefit and Financial Internal
Rate of Return for an 8.77 ha Model Farm, High Moisture
Cobs at Zero Cost - Diesel Fuel Cost of $0.65 per liter
Grain Yield of 5.65 tons per hectare

Year Cost W/O Gas Cost W/Gas Increm Net NPW
Benefit 0%

1 896 1477 -581 -581

2 896 851 45 45

3 896 851 45 45

4 896 851 45 45

5 896 851 45 45
-261-

The Financial Internal Rate of Return is negative since the
total NPW is negative at 0% discount rate.

At low cob moisture content, the financial internal
rate of return was extremely high (>90%) as can be seen
from table 6.30. However, for cobs at high moisture
contents, even the better yield does not produce a postive
financial IRR (tables 6.31 and 6.32). This resulted from
the relatively ismall amount of diesel fuel saved when
medium and high moisture cobs were gasified. At medium
moisture only 20% of the diesel fuel was saved and at high

moisture only 11.3% of the diesel fuel was saved.
6.8.2 Increased Corn Yield combined with Cob Cost

This analysis was done only for cobs at low moisture

content, because positive rates of return were not

142
possible at medium and high moisture contents. The cost of
cobs were considered at three levels; $1.00, $3.00, and
$5.00 per cubic meter.

At the higher yield of 41 m3 per season, the cost of
the corn cobs was $41.00, $123.00 and $205.00. These values
are used in the partial budgets of tables 6.33, 6.34 and
6.35.

The outcomes for the financial internal rate of return
are given in tables 6.36, 6.37 and 6.38. Of the 3 outcomes,
2 have positive financial IRR's. The cob price of $1.00 per
m3 showed a return of 75%. The cob price of $3.00 per m3

showed a return of 32.24%. Only the cob price of $5.00 per

m3 gave a negative rate of return.

Table 6.33 Partial Budget for Irrigation Costs, With
and Without Gasification Fuel,fgr a Model Farm of 8.77ha
Low Moisture Cobs costing $l/m - Diesel Fuel Cost
of $0.65 per liter — Grain Yield of 5.65 tons per hectare

Item w/o gas Y1 Y2 Y3 Y4 Y5
Gasifier Cost 0 626 - - - -
Maintenance 200 228 228 228 228 228
Diesel Fuel 696 342 342 342 342 342
Biomass Fuel 0 41 41 41 41 41

143

Table 6.34 Partial Budget for Irrigation Costs, With

and Without Gasification Fuel,fgr a Model Farm of 8.77ha
Low Moisture Cobs costing $3/m -Diesel Fuel Cost

of $0.65 per liter - Grain Yield of 5.65 tons per hectare

Item w/o gas Y1 Y2 Y3 Y4 Y5
Gasifier Cost 0 626 - - - -
Maintenance 200 228 228 228 228 228
Diesel Fuel 696 342 342 342 342 342
Biomass Fuel 0 123 123 123 123 123
$32.11 """""" £98"mi31§""'333""353""853""353"

Table 6.35 Partial Budget for Irrigation Costs, With

and Without Gasification Fuel,fgr a Model Farm of 8.77ha
Low Moisture Cobs costing $5/m -Diesel Fuel Cost

of $0.65 per liter - Grain Yield of 5.65 tons per hectare

Item w/o gas Y1 Y2 Y3 Y4 Y5
Gasifier Cost 0 626 — - - -
Maintenance 200 228 228 228 228 228
Diesel Fuel 696 342 342 342 342 342
Biomass Fuel 0 205 205 205 205 205

144

Table 6.36 Incremental Net Benefit and Financial Internal
Rate of Return fog an 8.77 ha Model Farm, Low Moisture
Cobs costing $1/m - Diesel Fuel Cost of $0.65 per liter
Grain Yield of 5.65 tons per hectare

Year Cost W/O Gas Cost W/Gas Increm Net NPW
Benefit 70% 80%

1 896 1237 -341 -200 -190
2 896 611 285 99 88
3 896 611 285 58 49
4 896 611 285 34 27
5 896 611 285 20 15

"II--311

The Financial Internal Rate of Return is 75%

Table 6.37 Incremental Net Benefit and Financial Internal
Rate of Return £05 an 8.77 ha Model Farm, Low Moisture
Cobs costing $3/m - Diesel Fuel Cost of $0.65 per liter
Grain Yield of 5.65 tons per hectare

Year Cost W/O Gas Cost W/Gas Increm Net NPW
Benefit 30% 35%

--—-----—-—-----*--__———_-—----—-_—-———-_---—_-__---—_---—-

1 896 1319 ~423 -325 -313
2 896 693 203 120 111
3 896 693 203 92 82
4 896 693 203 71 61
5 896 693 203 55 45

'13'-312

The Financial Internal Rate of Return is 32.24%.

145

Table 6.38 Incremental Net Benefit and Financial Internal
Rate of Return £05 an 8.77 ha Model Farm, Low Moisture
Cobs costing $5/m - Diesel Fuel Cost of $0.65 per liter
Grain Yield of 5.65 tons per hectare

Year Cost W/O Gas Cost W/Gas Increm Net NPW
Benefit 0%
1 896 1401 -505 -505
2 896 775 121 121
3 896 775 121 121
4 896 775 121 121
5 896 775 121 121
. ---:31-_

The Financial Internal Rate of Return is negative since the
total NPW is negative at 0% discount rate.

The outcome for the $1.00 per m3 cob price had a
financial IRR of 75 percent. The $3.00 per m3 cob price
gave a return of 32.24 percent. In this case increased cob
yield permitted higher cob prices to be accomodated by the
gasification system. the $5.00 per m3 cob price gave a

negative rate of return even at the higher yield.
6.8.3 Increased Diesel Fuel Price

The price of diesel fuel was increased by 20 percent
to evaluate the financial effects on the gasification
systems viability. The initial diesel price was $0.65 per
liter. Adding 20 percent to this resulted in a price level

of $0.78 per liter. The seasonal diesel fuel consumption

146
remained at 1,071 liters. The grain yield used was 3 tons
per hectare, as was the case in the initial assumptions.
The partial budgets and outcomes for low and medium
moisture cobs are presented in tables 6.39 to 6.42. The
high moisture outcome was excluded because the medium

moisture results were negative.

Table 6.39 Partial Budget for Irrigation Costs, With
and Without Gasification Fuel,for a Model Farm of 8.77ha
Low Moisture Cobs at Zero Cost - Diesel Fuel Cost of
$0.78 per liter - Grain Yield of 3.0 tons per hectare

Item w/o gas Y1 Y2 Y3 Y4 Y5
Gasifier Cost 0 626 - - - -
Maintenance 200 228 228 228 228 228
Diesel Fuel 835 609 609 609 609 609
Biomass Fuel 0 0 0 0 0 0

147

Table 6.40 Partial Budget for Irrigation Costs, With
and Without Gasification Fuel,for a Model Farm of 8.77ha
Medium Moisture Cobs at Zero Cost - Diesel Fuel Cost

of $0.78 per liter - Grain Yield of 3.0 tons per hectare

Item w/o gas Y1 Y2 Y3 Y4 Y5
Gasifier Cost 0 626 - - - -
Maintenance 200 228 228 228 228 228
Diesel Fuel 835 753 753 753 753 753
Biomass Fuel 0 0 0 0 0 0
;;;;1"“"“"153;""‘1ga;“"‘;g;""551“";g1'“";;;"

Table 6.41 Incremental Net Benefit and Financial Internal
Rate of Return for an 8.77 ha Model Farm, Low Moisture
Cobs at Zero Cost - Diesel Fuel Cost of $0.78 per liter
Grain Yield of 3.0 tons per hectare

Year Cost W/O Gas Cost W/Gas Increm Net NPW
Benefit 30%
1 1035 1463 -428 -329
2 1035 837 198 117
3 1035 837 198 90
4 1035 837 198 69
5 1035 837 198 53
_ ____ 6--

The Financial Internal Rate of Return is for this scenario
is 30%.

148

Table 6.42 Incremental Net Benefit and Financial Internal
Rate of Return for an 8.77 ha Model Farm, Med Moisture
Cobs at Zero Cost - Diesel Fuel Cost of $0.78 per liter
Grain Yield of 3.0 tons per hectare

Year Cost W/O Gas Cost W/Gas Increm Net NPW
Benefit 0%
1 1035 1607 -572 -572
2 1035 981 54 54
3 1035 981 54 54
4 1035 981 54 54
5 1035 981 54 54
"33%;"

The Financial Internal Rate of Return is negative since the
total NPW is negative at 0% discount rate.

Higher diesel fuel prices had a positive effect on
the low moisture scenario. The financial IRR rose 16 points
to 30.0 percent compared to the lower price (see table
6.18). In effect the diesel fuel price increase of 20% more
than doubled the financial IRR. At higher moisture levels
however the return remains negative though better than at

the lower diesel fuel price.

A summary of the results of the financial sensitivity
analysis is presented in table 6.43. As can be seen from
the table positive rates of return were possible only for

the low cob moisture content.

149

Table 6.43 Summary of the Assumed Values of Important
Parameters and Financial Internal Rates of Return for Corn
cob Gasification, on an 8.77 ha Model Farm Under Surface
Irrigation Using a 3.13 kw Slow Speed Diesel Pumpset

Cob Cob Cast Diesel Fuel Grain Yield Financial
M01sture $/m Cost S/l tons/ha IRR %
‘E;;"""""6 """""" Bié§"""m3""m"mIZ'B """
Med 0 0.65 3 0 <0
High 0 0.65 3 0 <0

Low 1 0 65 3 0 5 1
Low 3 0.65 3 0 <0

Low 5 0.65 3.0 <0

Low 0 0.65 5 65 >90.0
Med 0 0.65 5.65 <0
High 0 0.65 5.65 <0

Low 1 0.65 5.65 75.0
Low 3 0.65 5.65 32.2
Low 5 0 65 5 65 <0

Low 0 0.78 3.0 30.0

CHAPTER 7

CONCLUSION
7.1 Experimental Aspects

The ability of a producer gas - diesel engine
combination to reduce diesel fuel consumption was evaluated
at the laboratory level. Corn cobs of three differing
moisture contents was the biomass feedstock. Given the
findings of this study the following statements can be

made:

1. Dry corn cobs yield the greatest reduction in diesel
fuel consumption, in a range of engine speeds between 400
and 600 rpm. At the same power output level, the driest
cobs (5% w.b.) saved from 55% to 56% of the fuel normally

consumed in full diesel operation (measured in ml/kwh).

2. Higher moisture corn cobs when gasified, also reduce
diesel fuel consumption but to a lesser degree than low
moisture cobs. As cob moisture increases so does diesel
fuel consumption. Medium (13% w.b.,15% d.b.) moisture cobs
only saved between 20 and 29% of the diesel fuel consumed
on full diesel operation. High moisture cobs (18% w.b.)

only reduce diesel fuel consumption by 11.3 - 24.0 percent.

3. Temperatures in the gasification system do not vary
appreciably with engine speed or power output, in the range

150

151
of engine speeds examined, when the corn cob (biomass)

moisture content is relatively steady.

4. The moisture in the corn cobs cools the reaction inside
the gas producer. Increasing moisture content increases the
temperature drop at the grate. High cob moisture levels
cause lower furnace operating temperatures in all cases.
The effect is more prevalent at 500 and 600 rpm than at 400
rpm.

At 500 rpm the high cob moisture level steady state
temperature mean (552.23OC) was 90.44 degrees less than
the medium cob moisture mean. The medium cob moisture mean
is 57 degrees less than the low cob moisture temperature
mean (699.7°C).

The situation is practically identical at 600 rpm the
means for the low, medium, and high cob moisture levels are
685.3, 626.0, and 536.3 degrees respectively. The high
moisture level is 89.7 degrees less than the medium
moisture mean. The medium moisture mean is 59.3 degrees
less than the low moisture one.

The means at 400 rpm for low, medium, and high
moisture cobs are 681.5, 642.7, and 632.1 respectively.
Only 10.6 degrees separates the medium and high moisture

means .

152

7.2 Financial Aspects
It was clear from the financial analyses that the

highest returns came from employing low moisture cobs. Only
the low moisture scenarios had positive financial rates of
return in all cases analysed. This occurred because low
moisture cobs saved the greatest amount of diesel fuel; 55%
as compared to 20% for medium and 11.3% for high moisture
cobs. Medium moisture cobs never had a positive rate of
return in any scenario and neither did high moisture cobs.

At low moisture, the greatest benefit was obtained by
increasing the grain yeild. A yield of 5.65 tons per
hectare produced a return of over 90%. This was due to the
fact that at the higher yield there were enough cobs to
operate the gasifier more than 90% of the irrigation
season.

When the diesel fuel price was raised 20% to $0.78 per
liter, from $0.65 per liter, the financial IRR more than
doubled, going from 14% to 30%. Thus higher diesel fuel
prices had a favorable financial effect on the gasification
system.

Corn cob costs in general could be tolerated in the
low cob moisture scenarios at a cost of $1.00 per m3. This
price level gave positive rates of return in the two cases
studied. Rates of return at medium and high cob moisture
were negative without exception when cob costs were

introduced. At high yield the $3.00 per m3 price had a

153
return of 32.24% with low moisture cobs. In all other
scenarios with cob prices higher than $1.00 per m3 rates of
return were negative.

Given these outcomes, gasification is feasible if the
climate in the area allows the equilibrium wet basis
moisture content to remain in the 5% range, during the
period in which the gasifier operates. In addition, costs
of corn cobs could not exceed $1.00 per m3 if yields are
low. If this was possible, the system could pay returns
when yields were low and even if there was a cost, albeit
low, for the cobs.

Beyond this the benefit of gasification was derived
from the provision of security in the event that a fuel
shortage occurred. These circumstances occur frequently in
the more isolated areas of the Senegal River Valley and
along parts of the Niger river as was indicated elsewhere
in this study. Furthermore, given the likelihood of fuel
price increases in the future, the benefit of a gasifier

would increase accordingly.

APPENDIX

154

APPENDIX

Slow Speed Diesel Manufacturers

l_«__s.'1' 03v; 5,7,.

    

EXHAUSY
OUYLE‘I’

Alfl lNTAKE—>‘

INJEC'IOII PUMP

GOVERNOR
LEVER

Descriptive Illustration of 6 and 10 H. P. Stover
Dielel Engines Indicating Location 0! Operating
Parts Not Shown in Illustration on Page II

No. 1206 and 1210 Series

Specifications for 5 H.P. Stover Diesel
Engine— No. 1205

Speed—000 III 1250 |I.I'.M.

Rating if ll.|'. pvr lllll |I.I'.I\I.

Maximum Horsepower .013 |I.I'. per |00 II.I’.M.

Maximum Torque~— :III II) II

Fuel Rec ommended—I Iv III (Itc- -l|IIwing iucl nil—
lhuune (ianly 28 III

Fuel Gomumptlon—.- IX III .51 lb. per ll.l’. hour, on
lomle IIIIIIIIIIII: [mm 00 In 00% of Maximum

Lubrication—~ l'l‘cKNlII‘I‘ l’lunncr ()il l'IIIIIp iII Sump.

Cooling System I 'III Illllil” III Wnlcr lIIIIk—l25 unis.

or llndiniur. "llII‘I'IIIIhNiphuII SysLeII.

WIN!“ Evaporation lhIpI-IIIIM "|qu IIIIIII, lI-IIIIII'I'II-
Inn: and altitude upproximntcly 3% uIIIIoIIII per hour
on 5 . ’. on III. eon level with a 70° room tem-
pcrntum nI'Icr vanr rnnchm I20“ In I40".

Rotation I'IIIIIIII-r-I'lm-IIIvin- III' III I.III- III". \I'Iwu [III-.-
inu I'IIwIIr 'l'IIkII-IIII‘ HIIIIII.

Cylinder—line removable Sleeve—Ground to Size'

Cylinder Head—Cant lmn ltemnvnhlc.

Pluto" I'mII ll‘llll. III-III 'I'I'IIIIII-Il IIIIII linIIIIIII III Niven

Pinon Diaplacement- (52 cu. inches.

(5) 4| l'lltill“ I (III.
Piston Pin Diameter I? .; in.

Plato!) Pin Boatinc—Ilronw—lfia. 191; in., length

I in.

Connecting Rod llrupl"mw-d. II-wvlinn.

Crank Pin Bearing —HI.ecl Muck IIubbinIl. ”in.
2% in. —l.cngth 2/filn ‘

Flywheel—Din. 22 In. ——l~‘IIco 'I‘.’. in. ~wI-iIdIt 230 lbs.

Crank Sh nit limp II‘IIIw‘II It}; in. in III-IIriIIgs,
2V 5 in. nt- l’ uell
Keyway—‘A' In. wide x' i.ln deep.

Clmlhlit ”mp I'IIIw'II (' IIIIIK l'Dl'Ifl‘Il IIIII-urnl.

Camahait Bearing: 2.

Timing Oearr-(‘m-I, lrun—Autunudicnlly llIIIIlII-II.

Crank Shaft Main Bearings
lulled. l)iI.I 2"- -_: in. ;lI-,III:IIIZ

GOV vernorr llv lIdl 'l'jpt‘ ( 'IIIIIpli.-Il‘ly I'IIII'IIIsI-d —
[\IIIUIIIINIL‘I Illy II.IIIIIIIIIJI‘I.

onPump~\t'llli1ll"l'llIIILl'rTVIK‘.

vel—SpI-Iiul III III III-«Ming .i\l'lny SII-I'l luI :III-II III

III'IIII llin. V :lIdVI-l ‘.' III I'

Rocker Arma- (' mu Imn.

Air Cleaner and Silencers SuppliI-d :u- ('uInlnu
l'ApIipInIIII niIIIIIIII ( IIIII um

Crank Shaft Extension Beyond Frame < 7:1. in.

Oil Requlred in Sump~ ~7 pIs.

Fuel Tlnk~IiIdvuIIizIIl SIIIII SIII-l lIw-IIIIII in "usa-
I-IIpnrin~Il gills. lll'lllllil‘ IIInk I‘ll" be nlsu used.

Beech Acro-Combuation Chamber.

Bosch Fuel Filter Injection Pump and Nozzle.

A. 0. Fuel Pump.

Starting—Ilium CI'IIIIkcd—Stnriinu (‘rIIIIk Furnished.

Clutch- Twin DlICNIPIII‘IIIMllNl III I-‘(IrII muI mm
or willmnI. (‘llllt‘l‘ V or |IlIIt illcl‘ pulley. l’rice de«
ponds upmf size and ly

Pulley— (‘IIIIIIIg I‘IplipIuoIiI does not include pulleys
Id nny IINI'rII IIiIIII. AddiIiIIIml cunt governed by
KIM IIIIII I.\'K|

Silencer tEx he uet-
depending upon [.pr

III-Imm- llnrk liIdI-

I'IIIIIIIIIIIIII III IIIldiIiIIIIIIl price

 

 

a; it;
‘I ~ .* I.I:

.Iv ’ié’fiifiifi “k";
I a???
‘ I.

v
I
l
I
..
'.
L‘ .
.
I.

i’

" I‘ ’ ' gil- i ;' iv
III‘SIIIISIII “I.I. “W M?
13‘2

fisw- {a Cnb'l
:. :Fvigxtt“

{1 '_'-00Icie {2|
. ZFII

 

 

156

3%? P1
*‘ :Ath" g"?

,.

.i‘(
.V'(

 

 

157

 

 

 

"‘14", splil
uuing «war

II) p wilh
loose and [all
pulley, pump
but plan
and

ptdtslnk

158

83.5

 

Model ESW
Doublc Cvlind';
I2 H P. 650m
um cooled"

(.uinq luv
pump, uilh
inqlr 1.1 l

[IIIHI‘L nu um
h“. «- plnl!

n:r:ulnl1

 

 

 

Eicher Goodearth Limited '

ol tractors in India. in colla

ed the manulacture

Gmbh. West Germany. Currently Eicher
over 15.000 tractors per annurn.

ration with Gebr Eicher.
practices

Today. Eicher is more than iust a tractor company.
Its mmulacturing activity extends. besides tractors to
‘ nerating sets. agricultural

l engines.

implements special purpose machines.
Eicher ha successfully exported its tructors. tiesel

159

engineering items like sewing machines. bicycle parts.
pumps sets. and hand tools to severd countries all
over the world.

Eicher is geared to meet export rewirements at all
types at engineering items with a solid inlrastructure
to ensure strict quality control and adherence to
delivery schedules. It can also assist in setting up
assembly and manufacturin plants lor tractors.
engines and agicultural im ements in other

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

m agricultural implements and a host ol other countries.
‘I’ype Totaly enclosed. slow speed. water cooled. single cylinder.
tour stroke. compress‘un tuition type.vertical diesel engne.
PM Woe is rated on eflective HP at crankshalt as a manmum
ENGINE SPECIF'CAT'Ofls load lor conlimrous running It can he used to give rated HP
lor any elevation upto 210 m above sea level. with luel oil ol
not lea Iii-i lm BTU per I).
"0°“- esw.t esw-z eswa 55"“ csw-s 55“” Lubrication Sat-ii system at lubricatbn is prowded‘ lor .r critical pan.
" " “fl" ‘ ' '° 6;: ”3: 2° arm The mu end bearings and cam shalt has...” are hi my.
I P M 650 “0' 'm ’ 1m quaity gun metd. The C R hearings are at high quality white
NONCM I l I 2 2 3 MIMWUCMlmlmIWNMW‘
Bore turn.) I“) "0.3 I!!! lit! "0.3 120 ,
Stroke lrnrn.) 139.1 139.1 tan 139.1 no.1 1391 Gum 3th I: I'd 0' a “fififid m. 9M m be
src (WW) we use m m 2l0 no , ”" “"V "'" "'9'" " "’"M’
Cooling System Water Wat. Water Water Water Water Vives Both the enhaust and inlet valves are mnde nt heat resisting
Nett Wt. or.) 350 350 350 520 520 520 “9‘ W” “m" “"1
Gross Wt. "(9.) 470 ‘70 ‘70 6” 6.0 W W The W is fitted with two wel halnnred cast irrin
Packed Vobme .. .U .U l.“ I.I I.” fluvheels which ensure that speed vanatinn is kept at a
mirirmm even under fluctuating load cnndmnns.
sum-Jud Foil-mm items are minded‘ free with the emine F nel Yank.
Equipment Starting hande, Foundation Bolts. Silencer. Spanner Set.
Oi Cm. Screw hives.
PUMP SPECIFICATIONS
_ Carng & Both casing rind impeller are made ol
5'" d N" E 3 TOTAL HEAD 0" “E1135 ENGlNE lmpeler hiqii quality rinse rpm" rnst irm
‘——_ 8 “tr- lW‘PuQ' is shrouded tyne. well
.I s s to 12 ii is is 20 MW...
E g H P. it e M
run Inches D'SCHARGE (”IRES/SEC) Shall Shall is made at h-qh rarhnn steel
Bal Bearing The hall hennng is single row. deep
”I I) 3 It 3 15m 2” - — l9 5 l7 5 l5 5 l2.5 - - 6 HP 650 gooved MR.
|00I|00 ‘ - 4 1500 2'5 - 1‘ 5 ’2 7 '° 7 ‘7 ' ° ' 5 ”5’ 55° Stutrmg 80: k ril ample depth mi. Cl ”land and
ensures ellr-t tn'r- u'.\l-'r srnl
I25: RS 5 I 5 ISM I” § 3] 75 - - - - - 6 ”P 650 Gun metal sleeve is also mouded
New 313 1500 2:!) - - -265 255 24 23 ll? - Bll‘ Mn
WARRANTY
IN! IN 4 I I tam 230 - - JD 32 22 5 - - - 8 ”I" I‘ll These products rarw mutant;- .vimnst .tiiv
mnmlnrturmq delerts .msinq mt nl normal mane tor a
125: I25 5 I 5 lm 2|!) 4? 39 .16 32 - - - 8 HI" nan Mimi ol lull 50‘ months
A‘ P" (mum's pn’rr y "l t utilitilv'uix
m I m J I 3 I5“) 2]) - - I) 27 25 '9 5 I7 5 In Mr mt“ imprrwer-uts the qui IIKJIIHV‘ mum in the lr‘nflct
.m- suhiert tn (hJD'It' nr u-iiliilrm-nl it mu. nine
IMIIN 4 I 4 mm 250 - - - 1') II 25 2“ In ”I“ "1‘“ without no'ire and the ("infinity m rrl‘ls' mi
inn-isthdtlv lrir new rlm rnum ur-s uln. li ring m it"
'25‘ '25 5 . 5 I“ 2'0 _ a y. M 32 n , m ”P ""1 lithe-er" .\r tu.\l sever the Mums and the di er nptnn Iii
ths pul‘buit‘on.

 

 

 

EIGHER GOODEARTH LIMITED

Deepak. 3rd Floor. l3. Nehru Place. New Delhi llOOl9.(|ndial
Phone: 681612. 681432 Cable: TRACEICHER Telex: 031-4937

 

 

 

 

B I BL I OGRAPHY

160

B I BLOGRAPHY

 

Allen,H. 1908. "Gas and Oil Engines". The Scientific
Publishing Company. Manchester, England

Barnard, G.W. 1984. Small Scale Gasification in Practice in
Developing Countries. Bio Energy 84 Symposium
Proceedings

Bhattacharyya, G.K. and R.A. Johnson 1977. "Statisical
Concepts and Methods". John Wiley and Sons, New York

Boutillier, J.L. 1980. Irrigated Farming in the Senegal
River Valley. Presented at the Workshop on Sahelian
Agriculture - Department of Agriculture Economics,
Purdue University

Carruthers,I and C. Clark. l983."The Ecomics of Irrigation"
Liverpool University Press. Press Building, Grove St.
Liverpool, England L77AF

Club Des Amis Du Sahel. 1977. L'Irrigation en Mauritania:
Situation Actuelle - Proposition De Programmation
1977-2000. Club Des Amis Du Sahel - Equipe Cultures
Irriguees

Cruz, I.E. 1980. Studies on the Practical Application of
Producer Gas from Agricultural Residues as
Supplementary Fuel for Diesel Engines. American
Chemical Society Symposium Series N 130

Dennetiere, A.A., F. Leorat, and G.F. Bonnici 1980. Power
Generation form Biomass Residues Using the Gasifier
Dual Fuel Engine Technique. American Chemical Society
Symposium Series N130

161

Doering, O.C., T.J. O'Hare, and R.M. Peart. 1978. Small
Scale Gasification of Biomass - the Case of Corn Cob
Gasifiers. Journal Paper No. 7582 of the Prudue
University Experiment Station

Donkin,B. 1905. "Gas,Oil and Air Engines". Charles Griffen
and Co. LTD London, England

Eoff, K.M. and D.M. Post. 1982. How to Power a Gasoline
Engine with Wood. Florida Cooperative Extension
Service Bulletin No.9-15M80

FAO 1980. On Farm Maize Drying and Storage in the Humid
Tropics. Agriculture Service Bulletin No. 40. Rome,
Italy

Fieloux, M. 1980. Socio-economic Study of a Toucouleur
Village - Bow, Matam Republic of Senegal. Presented
at the Workshop on Sahelian Agriculture - Department
of Agriculture Economics, Purdue University

Gittenger,J.P. 1982. "Economic Analysis of Agricultural
Projects”. Second Edition Johns Hopkins University
Press. Baltimore, Maryland

Groenveld, M.J.,and Westerterp. 1980. Thermal Conversion of
Solid Waste and Biomass - Social and Economic Aspects
of Gasification Technology In Rural Areas of
Developing Countries, Tanzania. American Chemical
Society Symposium Series N 130 p705

Hall,D.O.; Barnard,G.W. and P.A. Moss. 1982. "Biomass For
Energy in the Developing Countries". Pergamon Press
Maxwell House, Fairview Park - Elmsford, New York

Held, W. and A. Koenig. 1982. Gas Generation from Biomass
For Use In Internal Combustion Engines. Proceedings of
2nd International Seminar on Energy Conservation and
the Use of Renewable Energies in the Bio Industries.
Trinity College - Oxford, England

Hiscox, G.D. 1919. "Gas, Gasoline and OiI Engines". Norman
W. Henly Publishing Co. New York

162

Johansson, Eric 1980. Swedish Tests of the Otto and Diesel
Engines Operated on Producer Gas. Proceedings -
International Gas Research Conference, Chicago,IL.
p902

Kutz,L.J., J.R. Barrett, C.B. Richey, R.B. Jacko. 1983.
Downdraft Channel Gasifier Operation and Particulate
Emissions. Transactions of the ASAE 1979

Manteuffel, A.M..and W.E. Tyner. 1980. Benefit Cost Analysis
of Small Irrigated Perimeters - A Sahelian Case Study.
Paper presented at the Workshop on Sahelian
Agriculture Department of Agriculture Economics -
Purdue University

McGowan, T.F. and A.D. Jape. 1981. Air Blown Wood
Gasification. Proceedings - International Gas Research
Conference. Los Angeles, California

MFC. 1985. Prix Officiels Des Carburants Applicables Au
Mali A Compter Du 3 Aout 1985. Ministere de Finance et
Commerce. Republique du Mali.

Montgomery, R.B. 1980. Production of Low BTU Gas from
Biomass in Commercial Air-Blown Gasifiers. Biomass
Energy and Wastes

Nordstrom,O. 1960. Aktuelle Arbeiten auf den Gebiet der
Ersatztriebstoffe in Schweden. Motorlastwagon Nrn
18,19, 20 und 22

OFDA, 1978. Economic Effects of Drought in Mauritania.
Office of Foreign Disaster Assistance - USAID Dept.
of State; Washington,D.C.

Payne, Fred 1978. Potential of Biomass Conversion
Processes for Grain Drying in Kentucky. Kentucky State
University Cooperative Extension Service Bulletin
3M-9-78

Peart, R.M., H.R. Zink, O.C. Doering, and M.R. Ladisch.
1980. Corn Cob Gasification for Corn Drying. Presented
at ASAE Energy Symposium, Kansas City, Missouri

163

Peter, D., I. LeBloas, N. Kehmeier, M. Raymond. 1979.
Developpement Des Cultures Irrigues En Mauritanie.
Bilan et Perspectives - Proposition pour un Second
Programme 1980-1985. Comite Inter Etats De Lutte
Contre La Secheresse Au Sahel

Posselius, J., C. Meyers, B. Stout, and J. Sakai. 1979. An

Updraft Producer Gas Generator. Michigan State
University Cooperative Extension Service Bulletin,

AEIS - 394

Reed, T.B., B. Levie, and A. Das. 1984. Understanding,
Operating, and Testing Fixed Bed Gasifiers. Bio Energy

84 Symposium Proceedings

Richey, C.B. and J.R. Barrett. 1983. Automatic Corn Cob
Feeding into a Gasifier Furnace. Presented at ASAE
Summer meeting - Bozeman, Montana. Paper No.83-3071

Riggins, J.K., D.B. Vaughan, and A.J. Lambert. 1980. Corn
Cobs versus Corn Stalks: A Comparison of Energy Values

for Drying Corn. Energy for Agriculture Vol 1.

Russell, Ben. 1980. The Evaluation of a Biomass
Gasification Experiment for Vehicle Power. Institute

of Gas Technology - Energy from Biomass and Wastes
Symposium No.IV. p 819

Sakai, J. and Y. Shibata. 1982. Alternative Gas Fuel
Producer for Driving Gasoline Engine Part 1.
Agricultural Mechanization in Asia,Africa, and Latin

America. Winter, 1982

Salinger, Lynn and J.D. Stryker. 1983. Comparative Returns
to Irrigated vs. Dryland Agriculture in the Senegal

River Basin: Model Results. Associates for
International Resources and Development. Sommerville,

Massachusetts

Shaw, L.N., D.M. Post, J.D. Whitney, S.L. Hedden, and D.B.
Churchill. 1983. Energizing An Irrigation Pump Engine
with Citrus Wood. ASAE Paper No.83-3076, Presented at
ASAE Summer meeting in Bozeman, Montanna

164

Silva, J.N., R.M. Peart, M.R. Ladisch, H.G. Gibson and V.L.
Anderson. 1984. Tar Formation in Corn Cob
Gasification. Presented at the ASAE Summer Meeting in
Knoxville, Tennessee. Paper No.84-3076

Srivastava, A.K., J.A. Posselius. 1981. Dimensional
Analysis and Similitude Modeling of the Gasification
Process. ASAE paper No.81-3538 Presented at the ASAE
Winter Meeting - Chicago, Illinois

Stassen, H.E.M., and T. Zijp. 1980. Progress Report - The
Gasification by Partial Combustion Project in Tanzania.
Twente University of Technology. Department of Chemical
Tecnology

Steel, R.G.D. and J.H. Torrie. 1980. "Principles and
Procedures of Staistics". Second Edition. McGraw Hill
Book Company - New York

Titl, A.J. 1981. Utilazation of Biomass for Cogenaeration
of Heat and Power. Proceedings - International Gas
Research Conference. Los Angeles, California

Turner, C. and R. Anderson. 1971. "Irrigation Principles
and Practices". AAVM Publishers - Boulder, Colorado

Van den Arsen, F.G. 1984. The Economics of Thermochemical
Biomass Conversion Systems for Small Scale Power
Generation. Bio Energy 84 Symposium Proceedings

White, G.M., T.C. Bridges, S.G. McNeill, and D.G. Overhults.
1984. Equilibrium Moisture Properties of Corn Cobs.
Presented at ASAE Summer Meeting in Knoxville,
Tennessee. Paper No. 84-3044

Zijp, T. and H.E.M. Stassen. 1981. The Gasification by
Partial Combustion Project in Tanzania - A Case Study.
Twente University of Technology - Department of
Chemical Technology.