Productivity of a Small Cut-to-Length Harvester in Northern Idaho

i
PRODUCTIVITY OF A SMALL CUT-TO-LENGTH HARVESTER IN NORTHERN
IDAHO
A Thesis
Presented in Partial Fulfillment of the Requirements for the
Degree of Master of Science
With a
Major in Forest Products (Timber Harvesting)
in the
College of Graduate Studies
University of Idaho
by
Douglas R. Turner
January 2004
Major Professor: Harry W. Lee
ii
AUTHORISATION TO SUBMIT
THESIS
This thesis, of Douglas R. Turner, submitted for the degree of Master of Natural Resources
with a major in Forest Products (Timber Harvesting) and titled “Productivity of a Small Cutto-Length Harvester in Northern Idaho”, has been reviewed in final form. Permission, as
indicated by the signatures and dates given below, is now granted to submit final copies to
the College of Graduate Studies for Approval.
Major Professor
______________________________ Date _____________
Harry W. Lee, Ph.D.
Committee Members ______________________________ Date _____________
Steven R. Shook, Ph.D.
______________________________ Date _____________
Harold L. Osborne, M.F.
Department
Administrator
______________________________ Date _____________
Thomas M. Gorman, Ph.D.
Discipline’s
College Dean
______________________________ Date _____________
Steven B. Daley Laursen, Ph.D.
Final Approval and Acceptance by the College of Graduate Studies
______________________________ Date _____________
Katherine Aiken, Ph.D
iii
Productivity of a Small Cut-to-Length Harvester in Northern Idaho
Abstract
The use of small cut-to-length harvesters in forest stands composed of small-diameter trees is
of increasing interest to natural resource managers. This interest is due to the efficiency and
cost effectiveness of using small harvesters in the harvest of small diameter trees. This paper
presents research regarding the productivity of a Neuson 11002 HV harvester in a clearcut
with reserves and a thinning treatment of a Douglas-fir tussock moth damaged stand, as well
as the investigation of the influence of tree branch characteristics upon harvester
productivity. Tree heights, diameter at breast height (DBH), species, branch size and branch
interval data were collected, to investigate the effect of these characteristics on productivity.
An elemental time study was conducted using video footage of harvest operations to
determine machine productivity. Production costs are estimated at $5.54 /m3 in the clearcut,
and $6.22 /m3 for the thinning, on a scheduled machine hour and mean tree volume basis.
Statistical analysis revealed that brush density, machine travel distance, tree DBH and branch
size are influential upon tree harvest time in the clearcut, and brush density, machine travel
distance and tree DBH were influential upon tree harvest time in the thinning. Branch size
and branch interval are influential upon tree processing time, with branch size of high
statistical significance. The Neuson 11002 HV harvester was found to be cost effective in
the small diameter stand in which it was employed.
Keywords: small harvester, brush density, branch size, branch interval, branch
characteristics, tree characteristics, Neuson 11002 HV, Logmax 3000, clearcut, thinning.
iv
Acknowledgements
Many people have helped me during my master’s studies, and during the research project on
which this thesis is based, and I would like to take this opportunity to express my thanks and
appreciation to them.
I would like to thank Dr. Harry Lee, my major professor, for his patience in answering the
many questions, for his help in the set up and conduct of the project, and for his support
throughout my time at the University of Idaho. This support and counsel were instrumental
to the completion of both my masters and this research.
I would also like to thank Harold Osborne, committee member and University of Idaho
Experimental Forest Manager (retired), for his efforts in the setting up of the field trial which
is the subject of the this masters thesis, for his advice and practical perspective of the project,
and for time spent serving as a committee member.
I would also like to thank Dr. Steven Shook, committee member, for his help and assistance
in maintaining objectivity in this research, his assistance with the statistical analysis of this
work, and for time spent serving as a committee member.
I would also like to thank Dave Ehrmantrout (Blondin Inc.) for all of his co-operation and
assistance in this study.
I must thank Jan Pitkin (Department of Forest Products secretary), for all her help and
assistance with the many trials and tribulations that occur on the way to obtaining a master’s.
I would like to thank Ross Applegren (current University of Idaho Experimental Forest
Manager) for his help with my various enquiries.
I would like to thank Dr. Andrew Robinson Department of Forest Resources, University of
Idaho) for his help in the field of forest mensuration.
Thanks also to Karl Froese, Damon Hartley, Jeff Halbrook, Andres Soria, graduate students
at the University of Idaho, who helped in the collection of field data used in this research.
Thanks must be made to both the Inland-Northwest Forest Products Research Consortium,
and the USDA Forest Service for their financial support for this research.
Finally, I would like to thank my family and friends, and in particular I would like to thank
my parents, Ian and Mary Turner, for their continued support in the development of my
education and career.
v
TABLE OF CONTENTS
Page
AUTHORIZATION TO SUBMIT THESIS……………………………………………… ii
ABSTRACT……………………………………………………………………………….. iii
ACKNOWLEDGEMENTS……………………………………………………………….. iv
TABLE OF CONTENTS………………………………………………………………….. v
LIST OF TABLES…………………………………………………………………………. vi
LIST OF FIGURES………………………………………………………………………... vi
LIST OF APPENDICES…………………………………………………………………… vii
INTRODUCTION ………………………………………………………………….…....... 1
Research Question…………………………………………………………………. 3
Research Objectives………………………………………………………………... 4
SITE DESCRIPTION……………………………………………………………… ……... 5
MACHINE DESCRIPTION ………………………………………………………. ……... 7
STUDY METHODS ………………………………………………………………. ……... 8
RESULTS – Clearcut with reserves……………………………………………………….. 12
RESULTS – Thinning………………………………………………………………………16
DISCUSSION………………………………………………………………………………20
Influences of branch characteristics……………………………………….……… 27
Observations on harvesting methodology………………………………………… 33
Observations on Neuson 11002 HV harvester…………………………………….. 34
Observations on research methodology. …………………………………………... 35
CONCLUSIONS…………………………………………………………………………... 37
LITERATURE CITED…………………………………………………………………….. 38
APPENDICES……………………………………………………………………………... 42
vi
LIST OF TABLES
Table 1:
Page
Predictive production equation for Neuson 11002HV harvester in
Bear Pine clearcut. ………………………………………………………… 15
Table 1a:
Analysis of Variance, Neuson 11002 HV harvester in clearcut
treatment unit. ……………………………………………………………... 15
Table 1b:
Parameter Estimates, Neuson 11002 HV harvester in clearcut
treatment unit. ……………………………………………………………... 15
Table 2:
Predictive production equation for Neuson 11002HV harvester in
Bear Pine thinning. …………………………………………………………19
Table 2a:
Analysis of Variance, Neuson 11002 HV harvester in thinning
treatment unit. ……………………………………………………………... 19
Table 2b:
Parameter Estimates, Neuson 11002 HV harvester in thinning
treatment unit. ……………………………………………………………... 19
LIST OF FIGURES
Figure 1:
Location of study site in State of Idaho. ………………………………….. 5
Figure 2:
Clearcut unit prior to treatment. …………………………………………... 6
Figure 3:
Harvester branch assessment key. ………………………………………….9
Figure 4:
Harvester cycle time elements. ………………………………………….… 10
Figure 5:
Harvester variables captured. ……………………………………………… 11
Figure 6:
Clearcut with reserves unit, post treatment. ………………………………12
Figure 7:
Neuson harvester CTL clearcut trial – Delay free activity summary. …….. 13
Figure 8:
Neuson harvester CTL clearcut Trial – Delay summary. …………………. 14
Figure 9:
Thinning unit, post treatment. ……………………………………………... 16
Figure 10:
Neuson harvester CTL thinning trial – Delay free activity summary. .……. 17
Figure 11:
Neuson harvester CTL thinning trial – Delay summary. ………………….. 18
Figure 12:
Influence of brush density upon harvester cycle time. ……………………. 22
Figure 13:
Influence of brush density upon harvester volume production per PMH. … 23
vii
Figure 14:
Influence of branch size upon harvester cycle time. ………………………. 29
Figure 15:
Influence of branch size upon harvester volume production per PMH. …... 29
Note: All images are credited to the author, unless a credit noted below the image states
otherwise.
LIST OF APPENDICES
Page
Appendix I: Treatment unit stand summary. ……………………………………………... 42
Appendix II: Machinery specifications
Appendix II.a – Neuson 11002 HV harvester specifications. ……………………... 43
Appendix II.b – LogMax 3000 harvester head specifications. ………………….… 45
Appendix III: Harvester cutting specifications. …………………………………………. 46
Appendix IV: Harvester machine cost analysis. ………………………………………… 48
Appendix V: Neuson harvester research statistics: summaries by treatment
Table 1: Neuson harvester clearcut summary statistics. ………………………….. 50
Table 2: Neuson harvester thinning summary statistics. …………………………. 50
Appendix VI: Research data collection forms
Form 1: Tree data sheet. …………………………………………………………...51
Form 2: Harvester data collection form. ………………………………………….. 52
1
Productivity of a Small Cut-to-Length Harvester in Northern Idaho
Introduction
While small harvesters have been in use in the forests of Europe and Eastern Canada and the
eastern states of the United States of America (USA) for some time, their use in the western
USA is relatively new. The interest in the application of such machines has been initiated by
the need to thin second growth forests (Kellogg and Bettinger 1994) and by the identification
of 12 million hectares of high density/high stem count forest stands as high priority for fuel
reduction treatment in the western states of the USA (Vissage and Miles 2003).
Small harvesters, designed to harvest smaller diameter trees, offer the natural resource
manager a machine with lower initial capital cost. The work of Ewing (2001) on a series of
small, tracked harvesters confirmed that these machines offered reasonable efficiency, lower
capital cost, and cost effective operation. Ewing observed a Neuson 11002 HV in operation
that achieved a production rate of 67 trees per productive machine hour (PMH), with a
volume production of 14.1 m3/PMH, with direct operating cost of CAN$96 /PMH, and unit
production cost of CAN$7.00/m3 in a 34 percent volume removal thinning of a mixed-wood
forest stand with 1340 stems/ha, 161 m3/ha and 16 cm mean diameter at breast height (DBH).
(All costs for the Ewing study are expressed in Canadian dollars, denoted CAN$). Rummer
(2002) investigated the use of the Neuson 11002 HV in a lodgepole pine thinning trial. The
results were a productivity of 4.2 m3/ scheduled machine hour (SMH), with a cost per unit
production of US$16.60/m3 (calculated from cost per ton, based on lodgepole pine density
tables: Briggs 1994) in a stand of 10 cm mean DBH (Rummer 2002). These results
demonstrate a considerable range in productivity and cost, and reflect the variation in the
stands and harvest prescription that the harvester was employed.
It must be noted that the Rummer study (2002) was conducted to compare the productivity of
the small CTL harvester with a motor manual thinning operation in a very dense stand of
lodgepole pine (4950 stems/ha) where the products merchandised from the small diameter
trees were post, rail, and fencing material. While the operator of the harvester in this study
2
was experienced in forest harvest operations and harvesting technology, he was not a full
time harvester operator, nor was he familiar with the production of fencing materials. The
great majority of tress being harvested in the Rummer study would also fall below the critical
value of 0.15 m3, identified by Richardson and Makkonen (1994) as the breakeven tree
volume for profitable CTL operations. Thus the work reported by Rummer is representative
of the lower percentiles of CTL harvester production, and in harvesting trees below the
recognized critical tree size, might be regarded as a pre-commercial thinning operation.
Rummer asserts that, while his work saw a greater cost efficiency and lower level of residual
stand damage with a crew of five employed in motor-manual operations than that of the CTL
operation, the harvester matched the five fallers in productivity. Further, given the huge area
to be treated, if the western states of the USA are to mitigate the forest fire hazard condition
present in many ecosystems, the number of fallers that would be required to carry out the
task would be considerable (Rummer 2002). As such, the employment of such machines
such as the Neuson, and others of the class investigated by Ewing (2001), may be the only
viable option for the natural resource manager.
To place the significance of the cost per unit production in context, even the relatively
competitive production cost reported in Ewing’s 2001 study must be tempered by the
assessment of the overall value of the material cut. The observation is made by Rummer and
Klepac (2002, p.1) that the products typically cut with such small CTL machines are
typically of lower value, while “cost of most physical operations per unit volume handled
increases exponentially as diameter decreases”. Thus a greater number of stems are required
to achieve 1 m3 of volume, all at greater cost per unit production.
A number of researchers have noted that branches are of influence upon the productivity of
CTL harvesters. One of the best examples was Gingras (1994, p.6), who observed in his
research “The productivity of the harvester was higher in the jack pine blocks mainly because
of easier delimbing and a longer length of merchantable stem. In the mixed wood blocks,
reduced visibility and more difficult delimbing conditions resulted in lower productivity”.
Observations such as those made by Gingras (1994), and also by Mellgren (1990), and
3
Richardson and Makkonen (1994), support the observation that tree branch characteristics are
highly influential upon tree processing time that was made by the researchers and the
participating harvester operator during the set up of this research.
The influence of branch size and interval were identified by Drolet et al. (1971) in their
investigation of branch characteristics of jack pine (Pinus banksiana), black spruce (Picea
mariana) and balsam fir (Abies balsamea) as having an effect on mechanised delimbing.
Drolet et al. (1971) found that delimbing of 95 percent of all trees studied would be possible
with a machine capable of removing branches with a branch stub area of 51.6 cm2 (8.1 cm
diameter branch) at a 7.6 cm interval, and a 77.4 cm2 branch stub area (9.9 cm diameter
branch) at 30.5 cm interval.
In contrast to the substantial machine suggested by the results of the research by Drolet et al
(1971), the specifications of a small harvester impose design limitations on dimensions,
machine weight and engine power, resulting in reduced power at the harvester head,
diminishing delimbing capabilities and decreased ability to handle large trees. Thus, it might
be hypothesised that the contributing effect of branch size and interval upon the processing
time of the harvester will increase with decrease in harvester size (in terms of available
horsepower, hydraulic systems etc.).
Research Question
The requirement for CTL harvesting equipment that is small and agile enough to enter a
forest stand, whether for commercial thinning treatment, fuel management, or ecosystem
restoration treatment, has been the cause of the development and introduction of small
harvesters and forwarders into the western USA. As the CTL system is relatively new to
northern Idaho, and small machines very new and novel, there is very little information
available to the logging contractor and machine owner, or to the natural resource manager
concerning the productivity, cost and effectiveness of these systems. Therefore, there is
significant need to study smaller CTL harvesters in operation and to identify those factors of
4
machine operation and forest and stand conditions that influence the productivity and
profitability of small scale CTL technology in the forests of the Interior Northwest.
Research Objectives
The objectives of this research were to:
Estimate the productivity of the Neuson 11002 HV harvester.
Estimate the cost of operation and unit volume production of the Neuson 11002 HV
harvester.
Identify the elements of machine operation, stand condition and silvicultural system
influential upon machine operation and production.
Construct regression equations for the operation of the Neuson 11002 HV harvester in
the respective stand harvest treatments (clearcut with reserves, and thinning).
Observe the harvest operations to identify issues in system application and future
areas of research.
Investigate the influence of tree branch characteristics upon harvester processing
performance.
5
Site Description
The Neuson 11002 HV Harvester was studied in operation in the University of Idaho
Experimental Forest, in the Flat Creek Unit (Land Description: NE ¼, Section 31, Township
41N, Range 3W, Boise Meridian), located northwest of Deary, Idaho. (Relative location
within the state of Idaho is shown in Figure 1.)
Figure 1: Location of study site in the State of Idaho
Source: USGS
Both the clearcut and thinning treatment units occupied a low ridge running in an East-West
direction. Soils present are principally Santa silt loam, with some small areas of Spokane silt
loam, and are deep and relatively well drained. Soil parent material is loess.
The stand to be clearcut covered approximately 3.25 ha, and was composed of 66 percent
Douglas-fir (Pseudotsuga menziesii var. glauca), 16 percent grand fir (Abies grandis), 16
percent ponderosa pine (Pinus ponderosa), and 2 percent western larch (Larix occidentalis).
6
(Figure 2: Clearcut treatment unit, pre-harvest condition.) Harvest was scheduled to salvage
trees severely defoliated by Douglas-fir tussock moth (Orgyia pseudotsugata). Stand volume
was estimated as 253 m3/ ha, 23 cm estimated mean DBH, and a harvest DBH range of 12.5
to 45 cm. (Harvest DBH range was defined at the lower diameter limit by timber recovery,
and at the upper diameter limit defined by harvester head capacity.) Stand density was
approximately 710 stems/ha of 12.5 cm DBH and greater (Appendix 1: treatment unit
summary table). Slopes on the clearcut site varied from 5 to 36 percent, with a mean of 18
percent. The silvicultural prescription for the clearcut was the harvest of the unit, retaining
approximately 16 stems per hectare of ponderosa pine and western larch for aesthetic/stand
structural reasons.
Figure 2: Clearcut treatment unit pre-harvest (September 2002).
The thinning treatment was conducted on an adjacent stand, composed of 83 percent
Douglas-fir, 16 percent ponderosa pine and 1 percent western larch. The exact intent of the
7
thinning was timber harvest and also to shift the species composition of the treatment unit in
a stand improvement cut. Stand volume was estimated as 223 m3/ ha, 24 cm mean DBH, and
a harvested DBH range of 12.5 to 45 cm. Stand density was approximately 575 stems/ha of
12.5 cm DBH or greater. The thinning treatment prescription was to harvest severely
defoliated trees as part of a low thinning of the stand (33 percent volume removal), while
favouring those ponderosa pine and western larch present (Appendix I: treatment unit
summary table). The stand had previously been harvested in 1984/5, with systematic trail
location (estimated 12 m trail spacing). This previous treatment was to convert a two age
stand to a single age stand by removal of the older cohort trees in the area between the trails.
A substantial proportion of the understorey trees (suppressed and sub-commercial size) had
been removed or otherwise eliminated from the stand during this previous harvest operation.
This removal was due in part to issues of faller access and safety, and through damage during
the extraction process. The thinning trial area was approximately 4.5 ha. Slopes encountered
in the thinning treatment varied from 3 to 45 percent, with a mean slope of 14 percent.
Trees to be retained post harvest in both treatment units were pre-marked by Experimental
Forest staff with a ring of orange paint at approximately chest height for both the thinning
and harvest units.
Machine Description
The Neuson 11002 HV is a purpose built, tracked-based harvester, manufactured in Austria.
The undercarriage is derived from a D4 Caterpillar carriage, on which is mounted the
leveling harvester body and 76 kW (102 hp) engine. The harvester is equipped with a Patu
parallel action crane (9.1 m reach), and a Logmax 3000 harvesting head capable of
harvesting trees up to 50 cm diameter at the base. Machine slope capabilities are specified
by the manufacturer up to 50 percent (Rocan 2003). (Appendix II: machine specifications.)
8
Study Methods
Prior to the start of harvesting operation 500 trees within each silvicultural treatment were
assessed for tree species, DBH, tree height, proportion of clean bole, proportion of dead
crown, proportion of live crown, branch size, branch interval (Figure 3), and tree form
(sweep, crook, fork etc.). (For example of data sheet, see Appendix VI, Form 1.) All trees
were marked with a unique number, for reference by the observer. Tree details were entered
into a spreadsheet, and analysed in that format. Tree volumes were calculated using
formulae prescribed by Wykoff et al. (1982).
To examine the effect of tree branches upon harvester productivity, a branch classification
key was developed that would allow rapid assessment of branch characteristics (Figure 3).
The branch classification for each tree was made on the basis of the highest score for branch
interval and branch size observed for the tree, in a conservative assessment of the branch
characteristics of each tree, as they would be encountered during in harvesting.
The harvester was to be employed in cutting 7 log specifications: 16’6”, 18’6” and 20’6”
mixed conifer species logs of minimum 6” top diameter (inside bark) for Bennett Forest
Products Inc. sawmill at Princeton, Idaho. 9’6” and 16’6” logs, of minimum 4” top diameter
(inside bark) in mixed conifer species for Plummer Forest Products at Plummer, Idaho. 10’
and 16’6” logs, of minimum 6” top diameter for Regulus Stud Mills Inc. at St. Maries, Idaho
(Appendix III: harvester cutting specifications).
Harvest operations were initiated on September 9th, and continued over ten working days,
being completed on 20th September 2002, and were conducted in late summer conditions.
9
Figure 3: Harvester branch assessment key.
Branch
size
code
0
1
2
3
Description
Fine branches
< 1.2cm (½
inch) diameter
(Includes
brittle dry
branches and
fine live
branches.
Light branches
1.2 to 2.5 cm
(½ to 1 inch)
diameter
Sturdy
branches
2.5 to 5 cm
(1 to 2 inch)
diameter
Heavy
branches
> 5cm (2 inch)
diameter
Appearance
Branch
interval
code
1
2
3
Description
Appearance
0.9m (3 feet)
or greater
branch
interval
0.45 to 0.9 m
(1 ½ to 3 feet)
branch
interval
< 0.45m (1 ½
feet) branch
interval.
How to use branch assessment key:
Branch size code and branch interval
code are added to give total branch
score; e.g. branch size code 2 and a
branch interval code 1, total branch
score of 3. (Exception: branch size
code 0, default total branch score of
0.)
Logging operations were video taped using a video camera equipped with a telescopic lens,
mounted on a light weight tripod. Tree numbers were reported to the camera operator/
observer by the harvester operator via a two way radio equipped with a voice activated
microphone, the number then being repeated into the video camera microphone by the
observer. A time study was then conducted on the recorded clearcut operation using a Husky
feX21 field computer equipped with a time study program (Wang et al. 2003), and using an
analogue stop watch for the thinning operation. (Figure 4, cycle elements; Figure 5 cycle
10
variables captured; Appendix VI, Form 2, example of data sheet.) Data were then entered
into an Excel spreadsheet for initial analysis and data management purposes.
Figure 4: Harvester cycle time elements
Machine Travel and Locate Harvester Head: Begins with the completion of the last
element of the previous tree harvest cycle (typically when the tree top is released from the
harvester head), and comprises all machine movement (crane and machine travel) related to
the location of the harvester head on the tree preparatory to the felling of the tree, or the
harvester undertakes a different task.
Brushing: Removal of undergrowth, saplings, and unmerchantable trees.
Felling: Begins when the harvesting head initiates the cut, and ends when the tree falls or
processing begins.
Process: Begins when the tree has completed its fall (following cutting), or when the
harvesting head begins the operation of processing (delimbing and bucking), and ends when
the last part of the tree being harvested is released from the harvester head.
Delays: Operational/ management, mechanical or personal delays that interrupt the normal
work activity of the harvester were noted, with description being made of the cause of the
delay.
• Pile: Activity outside the processing element of the harvester’s activities, where the
operator is sorting or piling the products/ logs in the forest.
• Personal: Time spent by the operator on personal needs, such as meals, etc.
• Maintenance: Time spent on routine activities related to the function of the machine
such as fueling, lubrication and greasing, cleaning of debris, cleaning of cab
windows.
• Breakdown: Time lost to damage or malfunction of the machine.
• Management: Activities relating to the harvest of the sale, such as consultation with
supervisors, log specification checks etc.
• Unidentified: Those delays that could not be classified from the video evidence,
falling into one of the above delay categories. Most probably to either personal,
maintenance, as they generally feature the operator staying in the cab.
11
Figure 5: Harvester variables captured.
Distance: Assessment of distance harvester has traveled in maneuvering to harvest a
specific tree, visually assessed from harvester dimensions and position.
Slope: Assessment of slope on which the harvester is operating.
Brush Code: Assessment of brush density treated for that tree cycle.
Product Cut: Count of logs cut per tree by assessed product specification.
During the time study, density of brush cut by the harvester at each tree was assessed, being
described as nil, light, moderate or heavy, and a nominal brush code allocated (0 for nil, 1 for
light brush, 2 for moderate brush, up to a score of 3 for heavy brush).
Machine owning and operating costs were calculated using standard machine rate analysis
(Miyata 1980); assumptions being $290,000 delivered cost, 2000 scheduled machine hours
(SMH), estimated salvage price of 20 percent after five years, fuel consumption of 13.25
litres/hour, maintenance and repair at 70 percent of depreciation, interest rate of 11 percent,
insurance cost of 4 percent, taxes estimated at 3 percent, fuel cost estimate $0.27 /litre, labour
cost of $25 /hour with benefits of 40 percent.
Nonparametric statistical analyses were conducted to investigate the effect of various tree
branch characteristics upon process time. A Kruskal-Wallace statistical test was conducted
over the full range of branch size code, branch interval code, and total branch score,
respectively. A Mann-Whitney statistical test was conducted between categories for each
variable type with respect to processing time, to analyze for differences between the
respective categorical values of branch size code, branch interval code, and total branch score
on tree process time.
A Pearson pair wise analysis was conducted to assess potential interactions between
independent and response variables. Following initial analysis, a stepwise analysis of the
12
dependent (travel time and harvester head placement, brushing, felling, and processing) and
independent variables (travel distance, brush density, DBH, tree height, tree volume, branch
size code, branch interval code, total branch score) was conducted to assess the significance
of the independent variables in the operation, as well as to construct a predictive equation for
tree harvest time.
Results – Clearcut with Reserves Treatment Unit
Figure 6: Clearcut with reserves treatment unit, post harvest (June 2003).
Mean productivity of the Neuson harvester in the clearcut (Figure 6) was 68.5 trees/PMH or
19.17m3/PMH. Mean tree volume harvested was 0.28 m3. Cost of machine operation was
$96.65/SMH (Appendix IV: machine cost analysis for the clear cut unit). Estimated
production rate of 17.45 m3/SMH (for the utilization rate observed) resulted in a calculated
production cost of $5.54/m3.
13
Inspection of the relative proportion of delay free cycle time spent on various activities
(Figure 7) reveals that the greatest proportion of time is spent in the processing (delimbing
and bucking) of the harvested stems (47%), with the next most significant time element being
machine travel and harvester head location (31%). Brushing was noted to take a significant
proportion of the harvester’s operational time (11%).
Figure 7: Neuson harvester CTL clearcut trial – Delay free activity summary.
Felling - 11%
Brushing - 11%
Processing - 47%
Machine Travel &
Harvester Head
Locate - 31%
Utilization Rate: 91%
The utilization rate was observed to be 91 percent over the duration of the study. Of the 9
percent time that was lost to delays, the majority of the delays were attributable to
breakdowns (46%); while this is not a large amount in terms of overall machine time, it is
indicative of the potential causes of loss of productive time for the harvester. Also of note is
the proportion of time spent by the harvester in piling logs (21%). Those delays described as
unidentified are those where it was not possible for the researcher to ascertain a particular
cause of delay due to lack of a view of the operator. These delays might be best classified as
breakdown (harvester computer), management (amendment of log bill within harvest
14
computer) or personal delays, as they were generally unidentifiable due to the operator being
in the harvester cab (Figure 8: summary of the delay structure of the clearcut operation).
Figure 8: Neuson 11002 HV harvester clearcut trial – Delay summary.
Maintenance - 9%
Unidentified - 11%
Breakdown - 46%
Pile - 21%
Personal - 4%
Management - 9%
Average delay free cycle time to harvest a tree was 52.6 seconds, with mean values of 2.3 m
travel distance, 21.6 cm DBH, 0.3 brush code, and 1.4 branch size code (Appendix V, Table
1: statistical summary of delay free harvest operation in the clear cut harvest unit).
Stepwise analysis of the variables collected in the study allowed the construction of a
predictive regression equation for the clear cut operation (Table 1). The effect of travel
distance, brush code, DBH and branch size code were all found to be highly significant (α =
0.05, p < 0.05). To eliminate some observed structure in the residual analysis, a
transformation on the response variable was conducted, using log10 transformation, at the cost
of some decline in the observed adjusted R2 value. It must be noted that it is not typical to
15
include categorical independent variables in a regression equation, as they are not normally
and independently distributed about (0, 1).
Table 1: Predictive production equation for Neuson 11002 HV harvester in clearcut
treatment unit.
Log{delay free cycle time} (seconds) = 1.2614 + 0.0153 Travel distance (m)
+ 0.1326 Brush Code + 0.0125 DBH (cm)
+ 0.0431 Branch size code
Adjusted R2 = 0.496, n = 397, α = 0.05, p < 0.05
Table 1a: Analysis of Variance, Neuson 11002 HV harvester in clearcut treatment unit.
Source
Degrees of
Freedom
Sum of
Squares
Mean
Square
Model
4
8.8562
2.2140
Error
389
8.8133
0.0227
Corrected Total
393
17.6695
F Statistic
97.72
Pr > F
<0.0001
Table 1b: Parameter Estimates, Neuson 11002 HV harvester in clearcut treatment unit.
Variable
(β)
Degrees of
Freedom
Parameter
Estimate
Standard
Error
t Statistic
Pr >|t|
Intercept (βo)
1
1.2614
0.0327
38.62
<0.0001
Distance (β1)
1
0.0153
0.0014
11.27
<0.0001
Brush Code (β2)
1
0.1326
0.0130
10.20
<0.0001
DBH (β3)
1
0.0125
0.0018
7.08
<0.0001
Branch Size (β4)
1
0.0431
0.0117
3.68
0.0003
The results of the Kruskal-Wallace non-parametric test revealed that branch size had high
statistical significance upon process time, while branch interval showed no significant effect
on process time. Total branch score, an additive combined score designed to account for
branch size and branch interval, demonstrated marginal significance; this might be attributed
to insignificance of branch interval upon process time being influenced by the positive
16
relationship between branch size and tree processing. A Mann-Whitney test to investigate
the difference between categories for the branch size was found to be significant, and
suggested a non linear relationship in which increasing branch size may have an exponential
effect upon tree processing.
A retrospective construction of a multiplicative combination of branch size and branch score
with respect to processing time was analyzed using Chi square. This analysis revealed a
significant effect. This result may be attributable to the influence of branch size upon
processing time, with the multiplicative effect magnifying the observed significance of the
branch size.
Results - Thinning Treatment Unit
Figure 9: Thinning treatment unit, post harvest (June 2003).
Mean productivity of the Neuson harvester in the thinning was 67.4 trees/PMH or 18.86
m3/PMH (Figure 9). Mean tree size harvested was 0.29 m3. Cost of machine operation was
17
$96.21/SMH. Estimated production of 15.47 m3/SMH (for the utilization rate observed)
resulted in a calculated production cost of $6.22 /m3 (Appendix IV: machine cost analysis
for the thinning unit). Utilization was noted as 82 percent over the duration of the study.
Review of the structure of the cycle elements forming delay free tree harvest cycle (Figure
10) reveals that the greatest proportion of harvester time is spent in processing (45%), with
machine travel and harvester head location the next greatest component of machine time
(36%). Brushing is also noted as being a significant use of machine time (9%) despite the
relatively low amount of brush encountered in the thinning trial area.
Figure 10: Neuson 11002 HV harvester thinning trial – Delay free activity summary
Felling - 9%
Brushing - 10%
Processing - 45%
Machine Travel and
Harvester Head Locate
- 36%
Utilization Rate: 82%
Utilization rate was observed to be 82 percent over the duration of the study. Of the 18
percent time that was lost to delays, the substantial majority of the delays were attributable to
breakdowns (74%); for the level of utilization observed, this is a significant loss of the
scheduled machine operating time, and therefore potential productivity of the harvester. As
18
with the clear cut operation, there were a number of unidentified delays, all of which might
be attributed to one of three delay classes: breakdown (harvester computer), management
(amendment of log bill within harvest computer) or personal delays, as they were
unidentifiable due to the operator being in the harvester cab (Figure 11: summary of the
delay structure of the thinning operation).
Figure 11: Neuson 11002 HV harvester thinning trial - Delay summary.
Management - 7%
Breakdown -74%
Unidentified - 5%
Pile - 5%
Personal - 1%
Maintenance - 8%
Stepwise analysis of the variables collected in the study allowed the construction of a
predictive regression equation for the thinning operation (Table 2). Average delay free cycle
time to harvest a tree was 53.4 seconds, with mean values of 4.0 m travel distance, 21.4 cm
DBH, and 0.4 brush code (Appendix V, Table 2: statistical summary of delay free harvest
operation in the thinning treatment unit.). The effect of travel distance, brush code and DBH
were all found to be highly significant (α = 0.05, p < 0.0001). To remain consistent with the
equation format for the clearcut, a transformation on the response variable was conducted,
using log10 transformation, at the cost of some decline in the observed adjusted R2 value. As
19
with the regression equation constructed for the clearcut unit, it must be noted that it is not
typical to include categorical independent variables in a regression equation, as they are not
normally and independently distributed about (0, 1).
Table 2: Predictive production equation for Neuson 11002 HV harvester in thinning
treatment unit.
Log{delay free cycle time} (seconds) = 1.2124 + 0.0096 travel distance (m)
+ 0.1578 Brush Code + 0.0151 DBH (cm)
Adjusted R2 = 0 .521, n = 356, α = 0.05, p < 0.05
Table 2a: Analysis of Variance, Neuson 11002 HV harvester in thinning treatment unit.
Source
Degrees of
Freedom
Sum of
Squares
Mean
Square
Model
3
9.7358
3.2452
Error
354
8.7956
0.0249
Corrected Total
357
18.5314
F Statistic
130.61
Pr > F
<0.0001
Table 2b: Parameter Estimates, Neuson 11002 HV harvester in thinning treatment unit.
Variable
(β)
Degrees of
Freedom
Parameter
Estimate
Standard
Error
t Statistic
Pr >|t|
Intercept (βo)
1
1.2124
0.03372
35.96
<0.0001
Distance (β1)
1
0.0096
0.00091
10.63
<0.0001
Brush Code (β2)
1
0.1578
0.01252
12.60
<0.0001
DBH (β3)
1
0.0151
0.00150
10.07
<0.0001
The results of the Kruskal-Wallace non parametric test revealed that, for the thinning, both
branch size and branch interval showed high statistical significance with respect to tree
processing time. The additive combined score, as might be expected, also showed statistical
significance. A Mann-Whitney test, which was used to investigate the difference between
categories for branch size and branch interval, was also found to be significant, with the
results suggesting a non-linear relationship, in which the influence of branch size and branch
20
interval has an exponential effect upon tree process time (for increasing branch size and
decreasing branch interval).
A retrospective construction of a multiplicative combination of branch size and branch score
with respect to processing time was analysed using Chi square, and revealed a significant
effect, as was found with a similar investigation for the clearcut treatment unit.
Discussion
The observed production costs for the harvester operation in both treatment units were
relatively competitive when compared to those seen in other studies. Data for a large
harvester (Valmet 500T) in a commercial thinning estimated production as 21.1 m3/SMH,
with $104.54/SMH owning and operating cost. Cost per unit production was estimated as
$5.47/m3 in a thinning with 0.63 m3 mean tree volume (Turner 2003). When these costs are
compared with the data generated by this study, the costs for the operation of the Neuson are
quite favourable, given the smaller trees that the Neuson was harvesting. When compared
with the production cost of $16.60/m3 calculated from the study conducted by Rummer
(2002), the calculated production costs of this study are notably low. This might be entirely
attributable to the larger tree sizes that were encountered by the harvester in both the thinning
and clearcut treatments of this research; the 10 cm mean DBH (4.04 cm standard deviation)
reported by Rummer (2002) contrasts sharply with the 23 and 24 cm mean DBH reported for
the clearcut and thinning, respectively. From this, the observation made by Richardson and
Makkonen (1994) that the critical tree size of 0.15 m3 for commercially viable CTL
operations appears to hold true for single tree processing.
While a direct comparison between the two treatments in which the harvester worked is not
statistically defensible due to differing stand structure and terrain conditions, a general
comparison is merited. The rate of production was higher for the harvester in the clearcut
operation (19.17 m3/PMH) relative to that observed in the thinning (18.86 m3/PMH), though
not demonstrating as great a difference as might initially be expected. This might be
21
attributed to the clearcut capturing all tree classes, from suppressed through dominant trees,
with consequently greater variation in the volume production per tree and per unit time. In
contrast, the thinning removed some suppressed trees, but harvest was mainly focused upon
intermediate and co-dominant trees. Thus, in the thinning, the smaller trees and the larger
trees were not part of the harvest, so the mean tree volume encountered (0.29 m3) represents
a far narrower degree of variation in tree size and a consistently larger tree than that
harvested in the clearcut (0.28 m3) than the marginal difference in size suggests. This
effective difference in tree size would serve to offset the fewer number of trees harvested per
hour.
The discrepancy in the number of trees harvested between the two treatment units might be
attributed to the amount of time spent traveling and locating the harvester head (36% in the
thinning, 31% in the clearcut). The difference in proportion of time spent traveling may be
apportioned to a relatively lower density of stems to be cut in the thinning, and the greater
amount of manoeuvering required in the thinning to locate, fell and process the trees.
The potential advantage of the clearcut, with its simpler operation and greater density of take
trees was ameliorated by a greater amount of time spent brushing. The cause of the greater
amount of brushing in the clearcut may be attributed to two causes – the lack of previous
intervention in this stand, and the requirement to leave the stand in a state suitable for site
preparation for replanting. In both treatments, a significant amount of time was spent on
brushing, with 11 percent for the clearcut, and 9 percent in the thinning. This suggests that
there might be significant production increase if the brushing was eliminated from the
harvest operation. Ignoring of brush by the harvester operator might be suggested, however
previous experience of this practice led to the discovery that the hydraulic hoses which
operate the crane and harvester head are frequently damaged when reaching through brush,
resulting in loss of machine time to replace blown hoses, and additional issues of lubricant
leaks causing environmental contamination. There is a general consensus amongst harvester
operators and logging contractors as to the wisdom of removing brush as a means of reducing
machine repair time. An alternative is to conduct a brushing operation (by either motor
22
manual or mechanized means) prior to harvester entry, thus eliminating the need to spend 9
to 10 percent of machine time on brushing.
The potential impact of the brush density merited further investigation. A sensitivity analysis
was conducted, using mean values for travel distance, tree DBH, branch size (for the clearcut
only) and varying the brush code for both the thinning and the clearcut. As might be seen
from Figure 12, there is a significant effect of brush density upon harvester cycle time.
If this cycle time is then extrapolated into a number of cycles per hour, at a mean tree
volume, a variation in volume production can be observed, assuming uniform brush density
for each tree cycle (Figure 13).
Figure 12: Influence of brush density upon harvester cycle time.
Harvester Cycle Time per Tree with Varying Brush Density
120.0
110.0
Cycle Time (seconds)
100.0
90.0
80.0
70.0
60.0
50.0
40.0
30.0
20.0
0
1
2
3
Clearcut
42.4
57.5
78.0
105.9
Thin
37.5
53.9
77.5
111.5
Brush Code (0 = nil, 1 = light, 2 = moderate, 3 = heavy)
23
Figure 13: Influence of brush density upon harvester volume production per PMH.
Volume Production per Productive Machine Hour with Variation in Brush Density
30.0
Volume per PMH (m^3)
25.0
20.0
15.0
10.0
5.0
0
1
2
3
Clearcut
23.8
17.5
12.9
9.5
Thin
26.9
18.7
13.0
9.0
Brush Code
From these graphs (Figures 12 and 13), it might be seen that further calculations to assess the
merit of a pre-harvest brushing treatment would be worthwhile. Given the mean brush code
values seen in this research (0.3 and 0.4 for clear cut and thinning respectively) and the
relatively low value of the material produced, it would not appear to merit the cost of a preharvest treatment. However, in circumstances where the value of the product is higher, and
the productivity level greater (such as with a larger harvester in larger timber), the influence
of brush and its impact upon productive machine time may merit detailed cost/benefit
analysis into the appropriateness of a pre-harvest brush treatment.
The large amount of variation in the ‘machine travel and locate harvester head’ element of
the machine cycle might be attributed to the operators active use of the 360o rotation of the
harvester, and in the full use of the long reach of the crane (further enabled by active use of
the extension or “squirt” boom). There was no data collected to account for the amount that
the harvester operator slewed the body of the machine around, nor how far the crane was
moved away from the machine. Given the 9.1 m reach of the crane, and the 2.4 m width of
24
the machine, there is a potential 20.6 m width of operating zone for the harvester. The
degree of movement observed within that zone is highly dependent on the location of the
harvester head at the end of the previous cycle, and the residual stem density of the unit in
which the harvester is operating (the harvester having a direct route available to the next tree
in a clearcut, the vector of the harvester head to the crane becoming more complex with
increasing residual stem count).
The technique of the operator also influenced the time recorded for the travel; wherever
possible, the operator tried to complete the processing of the previous tree close to the next
tree to cut. An additional factor in the move element was the scanning that was required to
locate the next tree, and analyze the best way to position the harvester head and machine to
fell and process this tree. The degree to which this affected the timing of the element was
strongly influenced by the number of residual stems that the harvester operator was working
among. Within the clearcut area the number of residual stems was generally low, and was
considerably higher within the thinning unit. However, within the thinning unit considerable
variation in the number of residual stems was noted (as a result of marking policy), and so
this influence might be expected to have varied throughout the thinning in effect. It should
also be noted that while there was no statistical significance observed between machine
travel time and slope in the construction of the regression equation, inspection of the Pearson
analysis showed a significant effect of slope upon the time taken in travel and harvester head
location for both the thinning and the clearcut.
The researchers observed a large amount of variation in processing method during the time
study, and this might be suggested as a cause of variance in tree processing and total harvest
cycle time. As with the travel and placement of the harvester head, this variation was caused
by the operator’s full use of the harvester’s rotation and reach of the parallel action crane in
the handling of unprocessed and processed trees. The amount of movement was observed to
be greatest with smaller, lighter, trees, and diminished with increase in tree size. This also
varied with the trail arrangement and topography, as the operator would make use of the
crane where appropriate, and if tree mass allowed, to manoeuvre the tree into a position to
allow optimal location of the cut product for forwarder extraction. This feature of the
25
operator’s technique varied with residual stem density, topography, and forwarder trail
configuration throughout the operation, and was thus a source of unexplainable variance
within the results and model.
Taper and tree length, combined with a desire to maximize the product recovery from the
operation, were observed to be of considerable influence on the time taken to process the
tree. Where the tree was of such dimensions that the programmed log lengths could be easily
cut tree processing time was relatively speedy. Where the diameter to length ratio combined
to make the optimizing of the cut problematic, the result was repeated movement of the
harvester head to assess diameter and length, with consequent impact upon processing time.
Thus, slight differences in tree dimensions might cause considerable variation in tree
processing time. The result of the repeated harvester head movement thus impacts the
relationship described between tree characteristics (DBH, tree height, branch features etc.)
and processing time acquires a degree of statistical noise from a factor that is difficult to
describe in terms of a variable of influence. This observation is supported by the findings of
Richardson and Makkonen (1994), who found that precise log lengths adversely affected the
productivity of a harvester.
The number of residual stems was of influence in other elements of the harvesting cycle also.
Of particular influence was the felling stage, where the maneuvering of the harvester head in
the effort to bring the tree down was affected by the number of residual trees that the felled
tree and head must be moved through. The residual trees also caused a degree of resistance
to the movement of the tree as it was felled and maneuvered down, and during processing;
the branches on the felled tree become entangled with the crowns of the residual and uncut
crop trees. This effect varied with the residual stem density, but was also seen to vary with
tree species (as an influence of branch habit, branch angle, branch coarseness, and other tree
characteristics), the proportion of live and dead crown, and previous stand silvicultural
treatments (growth density, prior stand treatments such a pre-commercial thinning etc.).
The processing of the tree is also significantly affected by the residual stand density, given
the care that is required to avoid damage to the residual stand varies with residual density and
26
with the type of stand entry. A far lower degree of concern is necessary when conducting a
clearcut operation with only a limited number of reserve stems, and the degree of care must
be expected to increase with the number of trees remaining. This would also be shown in the
placement of the processed product for later extraction, with residual stems being of greater
influence at greater densities and with increase in the number of different specification
products harvested. The objective of an effective harvester operator is to sort the cut product
during processing by the positioning of the harvester head, thus improving the efficiency of
the forwarding operation.
The level of utilization observed in the clearcut element of this study is notably high (91%),
and contrasts sharply with that observed in the thinning (82%). It must be stressed that the
utilization data gathered was based on the cycles of the trees pre-assessed for the productivity
research. As such, the research failed to capture much of the scheduled daily maintenance
cycle, or refueling, nor any of the delays that occurred in relation to trees excluded from the
study by their non-selection for the research. Information that might have been of use
relative to the harvesters own computer output of productivity was sought, but this
information was found to be of little value as a result of some difficulties with the system.
Some concept of the level of utilization might have been obtained if the researchers had
gathered shift level data, whereby the machine operator would have recorded the days
activities in terms of hours of operation, time spent on maintenance, breakdowns, personal
time, etc. and a more generalized long term assessment of machine activity could be made.
However, it was thought that there might be only limited merit in the collection of shift level
data given the short duration of the research (10 days of operation) in which the harvest site
was subject to a large number of visits from University of Idaho Experimental Forest staff,
timber harvesting students from the University, local logging contractors and local forest and
harvest managers.
In consideration, the data from the utilization might be best reviewed as an erratic activity
sampling, and, lacking the time structure that such a study would require (Olsen and Kellogg
1983) is impossible to develop the data further. It must be concluded that the utilization
levels are not necessarily representative of those that might be encountered in the long term.
27
It must be noted that the harvester operator in this study was not employed full time in the
operation of this machine, though he was familiar with the operation of the study machine
and mechanized harvesting operations in general. The training and experience of a harvester
operator were identified by Richardson and Makkonen (1998) and Gellerstadt (2002) as
being of great significance in the performance and productivity of a machine, and suggested
as being related to machine breakdown by Mitchell and von der Gönna (1994). As such, the
observed productivity for this study might be considered to be an underestimate for what a
fully trained and experienced full time operator might achieve, an observation supported by
the findings of Ewing (2002) in his study of small harvesters. In consideration of this
research, given the increasing familiarity with the operation of the harvester due to a series of
consecutive days of work, the operator’s performance may have varied over the course of the
research, with his familiarity improving performance and being most definitely expressed in
the last treatment unit, the thinning. This observation does not account for the variation in
human performance over the period of a working day, nor how this would impact alertness,
dexterity etc. (Gellerstadt 2002). This later factor would not be possible to reconstruct
without detailed data on the operators rest periods, and other activities, which it would not be
practical to attempt to deduce from the video coverage of the harvest operation.
Influence of Branch Characteristics
The statistical significance of the branch characteristics was firmly established by the
analysis conducted. The Kruskal-Wallace and Mann-Whitney tests indicate that branch size
was significant for both treatment units in which the harvester was observed, and that branch
interval was significant in the thinning treatment. This significance is further suggested by
the inclusion of branch size in the predictive equation developed from the data for the
clearcut operation. However, some caution must be made in interpreting the regression
results given that a categorical variable is included in the regression equation. This
relationship is also seen to exist despite many potential causes of statistical noise in the
processing phase, as identified in the discussion section.
28
The lack of significance in the total branch score is somewhat disappointing, but might be
attributed to the lower proportion of certain factor combinations, especially the relatively few
long branch interval (branch interval score 1) trees that were observed in the clearcut, and
absence of long branch interval trees in the thinning, where the better growing trees were
selected for the final crop. A limited number of trees of large branch size and short branch
interval were recorded, and this lack of interaction would also cause difficulties in describing
a relation. Hind sight would also suggest that a multiplicative approach (in which the branch
interval code is multiplied by the branch size code, magnifying the potential influence of
branch size) might offer a greater chance of correlation between these two factors and tree
processing time.
The non linear relationship suggested by the non parametric statistical analysis of branch size
for the clearcut, and branch size and interval in the thinning suggests that there may be an
exponential effect of branch characteristics upon tree processing time. To hypothesise
further, the relationship between tree characteristics and tree silvical characteristics and
silvicultural treatments is of influence in the processing time for harvested trees, given
uniform values for harvester, harvest operator, terrain etc. This was investigated in a
sensitivity analysis for mean values of travel distance, brush code, and DBH for the
predictive equation developed for the clearcut stand (Figure 14).
The influence upon tree cycle time that is recognized by the largest branch size observed by
the researchers is significant. If this condition is extended over the period of a productive
machine hour, then the results are readily seen (Figure 15).
29
Figure 14: Influence of branch size upon harvester cycle time.
Harvester Cycle Time per Tree with Variation in Branch Size
65.0
Harvester Cycle Time (seconds)
60.0
55.0
50.0
45.0
40.0
35.0
30.0
Clearcut
0
1
2
3
41.7
46.0
50.8
56.1
Branch Size Code
Figure 15: Influence of branch size upon harvester production per PMH.
Volume Production per Productive Machine Hour with Variation in Branch Size
25.0
Volume Production per PMH (m^3)
24.0
23.0
22.0
21.0
20.0
19.0
18.0
17.0
16.0
15.0
Clearcut
0
1
2
3
24.2
21.9
19.8
18.0
Branch Size Code
30
Both Figure 14 and 15 demonstrate the strong influence of branch size upon the productivity
of the Neuson harvester. The full impact of such branch influence over an entire year of
operation can only be speculated upon. However, the potential losses in annual productivity
due to coarse branches that Figure 15 suggests would be considered highly significant by
many logging contractors when assessing a timber sale unit prior to bid for harvest payment.
Review of the Pearson pair-wise correlation analysis demonstrated a highly significant
relationship between tree height, tree volume, tree DBH, and branch size and branch interval.
While the relationship to volume is expected, given the method of volume calculation used,
the relationship between the branch characteristics is of great interest. This would suggest
that, for a given uniform stand or cohort, the size of the tree is directly related to canopy
position, growth pattern, and crown form, all with consequent effects on branch
characteristics. Thus, it might be suggested that, for the Douglas-fir harvested in this
research, a suppressed or sub-dominant tree might have smaller branches at shorter intervals.
Similarly, a more open grown tree, deprived of competition, would have longer branch
interval but coarser branches. To describe a co-dominant tree, a longer branch interval might
be expected due to inter tree competition, with branch senescence defined by the relative
density of the stand. Therefore, it can be hypothesised that stand characteristics define the
branch characteristics that are encountered by the harvester, and that these characteristics can
be influenced by the forester in his silvicultural activities throughout the rotation of the stand.
In an observational review of the effect of branches upon harvester productivity, the operator
technique and anticipation of the branch condition of the tree were noted as being of
influence. An example of this would be that the operator observes some heavier branching
above the anticipated cutting point of the log currently being processed, and continues to
delimb beyond the cutting point to keep the harvester head momentum and improve
delimbing speed (as opposed to making the cut, and having the harvester head stall against a
concentration of branches once processing has resumed). Thus experience, operator
anticipation of processing problems and operating style are of influence; Gellerstadt (2002)
estimates a learning period of up to three years to fully develop these anticipatory skills.
There is also the chance that future developments in processing computer systems may
31
eclipse the effect of operator style; Murphy’s (2003) investigation of influence of scanning
technique found that advancing the harvester head beyond the predicted cutting point gave
better stem analysis and greater value recovery. The increase in the movement of the
harvester head was justifiable in financial terms, but might also be predicted to result in
improved harvester head momentum, resulting in superior delimbing performance.
It was observed that the more open grown the trees were, the greater the proportion of live
crown and substantial branches. As one might expect, the generally observed condition was
that these trees offered the greatest difficulties to the harvester in the process of delimbing.
An additional difficulty was also noted, in that the natural senescence of the lower branches
was delayed (when compared to trees forming part of a denser cohort) with the additional
effect upon the harvester of making the secure location of the harvester head upon the stem
preparatory to making the felling cut more difficult. In the most extreme cases, the harvester
head was being used to delimb or knock off branches prior to head placement, significantly
increasing the cycle time per tree, with some negative impact on the overall productivity of
the harvester anticipated. Where the author has previously encountered such trees in
plantation conditions in the past, where the outer trees of a stand have grown with 40 percent
or more of the crown developing in open grown conditions. These coarse branched trees
have necessitated the use of fallers to fell and process the stems in a motor manual operation,
such has been the difficulty of effective operation of the harvester in such conditions.
It might be speculated that the resistance to cutting for branches would vary with the season,
live branches being most supple in the spring and summer months due to wood moisture and
sap flow, and most brittle in the cold winter months. Thus, the resistance to delimbing might
be expected to be least when temperatures are below freezing, and greatest in the spring and
early summer. To follow this logic further, the greatest issue in delimbing (besides the knife
profile and configuration) is the harvester head’s feed force and feed speed, both of which are
dependent upon tractive force applied to the stem via the feed rollers. At those times of the
year where the bark is loose about the stem, particularly in the spring of the year when the
sap flow has begun, the chance of the feed system losing traction is greatest. Thus, the issues
of traction loss due to loose bark and resistance to delimbing combine to make the spring
32
time the most difficult scenario in which to process trees. In the remainder of the year, the
bark may be found to ‘tighter’ on the stem, and provide a greater degree of traction. The
degree to which the bark and traction interact would also be expected to vary with species,
for example the bark characteristics of grand fir is significantly different to that of ponderosa
pine or western larch.
The effectiveness of the harvester head investigated must be seen only as a true reflection of
the relatively small harvester and harvester head involved in this study. Some effect of
branch characteristics must also be expected with other harvester heads and harvester units,
though the degree to which branch characteristics impact the process can only be speculated
upon. A series of heads that have been developed in the Southern hemisphere, in particular
those developed in Australia and New Zealand, have been introduced into the North
American CTL technology market. These heads have been developed for use in radiata pine
(Pinus radiata) plantations, and so have been designed to meet the branch characteristics of
this rapidly growing species. Discussion with the users of such equipment in North America,
especially by those harvesting the heavier branched pine species such as ponderosa pine,
suggest that these heads are better suited to dealing with the branch characteristics
encountered.
A second factor that may influence the impact that branches have upon the processing
operation is the mass of the tree contained in the crown. The greater the proportion of live
crown, and the wider spreading the trees branches, the greater the mass that might be
expected to add to the overall mass of the tree. The greatest proportion of crown can be
anticipated in dominant and more open grown trees. This is of particular significance when a
harvester is felling and processing a tree that is at the upper end of its size capabilities,
typically a co-dominant or dominant tree. In this situation, the additional mass of the crown
can make the tree both awkward to handle in terms of its larger dimensions and its greater
mass. The difficulty of processing might be further increased if the tree is a ‘wolf’ – an
especially vigorous tree of poor form and coarse branches. This then is the worst situation
that can be met in the processing element of a CTL harvester’s operation. To counter this
effect, emphasis might be best placed on an operator adopting techniques where the greatest
33
proportion of the tree as possible is delimbed and processed while the tree is still falling. The
elimination of a log length and potentially some branch mass can considerably reduce the
overall mass of the tree being handled by the harvester.
The branches were also observed to be of impact to timber quality. In those situations where
the branches resisted easy delimbing by the harvester, and a number of passes of the
harvester head were required to remove limbs, there was evidence of rotation of the harvester
feed rollers on the stem, causing debarking of the tree, and some damage to fibre. This
damage is less likely to occur in the lower part of the stem due to natural branch senescence,
and so less likely to damage the most valuable portion of the stem. However, there is still
some concern with such damage, especially as it provides an entry point for wood degrading
and devaluing fungi such as blue stain (Ophiostoma, Leptographium, and Sphaeropsis
species of fungi), the financial significance of the damage and blue stain depending on the
specification of the product cut.
Observations of Harvesting Methodology
Harvester operators who have previously worked as forwarder operators for a significant
length of time appear to take greater interest in the separation and piling of the different sorts
during the processing element of the harvest cycle. This results in better presentation of the
different specification of the cut product for the forwarder, and also a tendency to concentrate
the volume in fewer piles than those generally seen with less experienced machine operators.
This is of great significance in the loading time of the forwarder, and therefore in overall
machine productivity, and so aids in the reduction of total harvesting cost (Gellerstadt 2002).
It might be observed that the relative time to harvest a small tree in comparison to a larger
tree is not proportional to the effort, a factor observed by many forest harvesting researchers
(Rummer et al. 2003, Rummer 2002, Holtzscher and Lanford 1997, Kellogg and Bettinger
1994, Richardson and Makkonen 1994 and more). This has been regarded by many as one of
the deficits of the CTL system, especially when compared with a feller buncher. Indeed, the
34
smaller trees allow the feller-buncher to make use of the accumulator facilities included in
the majority of felling heads, and collect a number of smaller trees before placing them in a
bunch. This ability mitigates the influence of smaller trees upon the overall productivity of
the feller buncher. Smaller trees having been identified as a limiting factor in the operation
of CTL systems, notably by Favreau and Gingras (1998), there is a development process
underway to allow multiple stem handling in CTL systems. Examples are in development or
in use in New Zealand (Waratah), Sweden (Silvatec) and Finland (Timberjack), with
investigations being conducted by Skogforsk (Bergkvist 2003) into the effectiveness of such
systems. The use for such multiple stem capable heads is initially anticipated as being the
smaller stem, lower value, stands that currently act as a lower limit to the economically
successful application of the CTL system. In these cases, the drive would appear to be to
increase volume production per hour, with respect to reducing effective cost per unit volume
production. The work done by Bergkvist (2003) found that there was an 18 percent increase
in productivity when 54 percent of the trees in a stand were treated in a multi-tree processing
operation, resulting in an improvement of 31 percent on the mean harvest cycle time per tree.
Bergkvist observed that a previous study conducted by Skogforsk in which 84 percent of the
trees had been subject to multi-tree processing saw a 40 percent increase in productivity.
Observations on the Neuson Harvester
The Neuson was equipped with a LogMax 3000 harvester head, a head developed in Sweden,
and in widespread use, including on Neuson harvesters in Eastern Canada with an excess of
9,000 operating hours. It was noted during the observation of the Neuson that the saw chain
in the harvester head was a frequent cause of stoppages. The tensioning device in the
LogMax head is manually adjusted, and the loss of correct tension in the saw chain was the
cause of chain loss or misalignment on the saw bar, requiring operator attention. The
researchers perceived this to be a cause of delay, when compared with the Valmet 500T/
Valmet 960 harvester head automatic tensioning system. In the Valmet 960 system, the
chain bar is connected to the harvester head hydraulics, so that as soon as the harvester
engine is started, the hydraulic pressure tensions the bar to the chain. While the Valmet
35
system is more complicated, the total number of stoppages was smaller, and the process of
replacing a chain was far faster, and could be completed without tools (though at the expense
of a more complex harvester head).
Slower harvester travel time was also evident, when compared to that previously observed by
the researchers with rubber tired/ wheeled harvesters. While this was not evident in the
results, especially as the travel time to the harvest site from the roadside was not captured,
this may have some influence on harvest operations where travel forms a large component,
particularly evident in thinning operations (36% of active harvester operation was spent
traveling in the thinning operation, as seen Figure 10).
Observations on Research Methodology
The use of a voice activated microphone seemed the ideal solution to the safety concerns and
research requirements of the Neuson harvester study. However, a number of concerns arose
as a result of the use of this equipment. The reporting of the tree numbers via a relay from
the harvester operator to the camera operator/ observer, and then repeated into the camera
microphone, was a cause of mis-reporting of tree numbers. The author also wonders if there
was not some influence in the technique of the operator, such that his approach to the marked
trees differed from that of unmarked trees, so that he might clearly see and report the
research tree number. In addition, the marking of trees as those pre-assessed for the study
may contribute to other unidentifiable forms of bias in the operation e.g. making these trees
easier to spot, relative to the orange marked leave trees. An additional consideration might
be the realization that the trees so marked were of commercially saleable size, as opposed to
many of the under sized trees that were felled and bucked to waste by the operator.
On one occasion during the thinning phase of the research, it was noted that the operator
amended his initial plan of approach to accommodate the requirement to read and report the
numbers painted on the trees to be harvested. Besides the 4 ½ minute delay in operations
that this caused, while the operator walked the area and rethought his plan of approach, this
36
offers some indication that the operator may have modified his operational technique to
accommodate the research effort, and this may also have occurred in the clearcut area. While
the researchers endeavoured to clearly mark the trees in anticipation of the direction of
approach of the harvester, this was not always entirely effective, and may have been an
influence upon the productivity of the operation.
Tree numbers painted on the smallest trees were hardest for the harvester operator to read,
and were most commonly missed in the reporting of the trees harvested. It might be
speculated that the harvesting productivity estimates were impacted by this problem. This
was also an influence in the total trees timed during the study, together with the miss
identification of tree numbers, and with the transposing of numbers due to communications
problems and static interference on the radio system. The misidentification did cause a
number of trees that were timed to be discounted from the production study (a reduction from
the original 500 pre-assessed trees in the order of 20.5 percent and 29.1 percent for the
clearcut and thinning respectively), due to lack of corresponding data for the tree volume and
branches. (The data from incorrectly identified trees was still used to analyse the machine
activity.)
The camera operator/ observer was of influence on the operation of the harvester on a
number of occasions in the thinning operation. In the majority of these occasions, the camera
operator was located within the danger area of the harvester, especially with respect to tree
felling direction, and also in the direction in which the trees were being processed. On these
occasions, the researchers felt that the operation of the harvester had been impacted by the
camera operator’s location, not so much as a result of the harvester operator being aware of
his presence, but modifying his operation to prevent injury to the camera operator. There
were also a number of occasions when the camera operator was a cause of direct interference
in the operation, when he communicated with the harvester operator to fell trees that had
been marked to leave, and elected to leave a designated take tree in its place. This did not
interfere with the time study on any occasion, but would be a consideration in the
development of future research using a similar format.
37
Conclusion
Analysis of the data generated in this study suggests that the productivity of the Neuson
harvester is significantly affected by the harvester travel distance (as a function of harvest
stem density), tree DBH, and brush density (as an effect on time spent in brush clearance).
Where these factors combine to adversely affect harvester productivity, such as the study
conducted by Rummer (2002), productivity of the harvester is diminished, and financial
viability of the operation is impacted. However, the rate of production observed with the
Neuson 11002 HV harvester in this study suggests that this machine can be a competitive and
financially viable harvester when used in stands composed of smaller trees. This would
allow the conclusion that similar forest stands in North America might be successfully
harvested using this machine.
The investigation into the effect of branch characteristics upon tree processing and total
harvest time strongly suggests that branch size is of significant influence upon tree process
and harvest time, and branch interval showed some statistically significant effect upon tree
process time. The significance of branch characteristics with respect to other combinations
of harvester and harvester head, feed system and knife configuration is more difficult to
assess. Given the numerous permutations of harvester and harvester head, further
investigation of the influence of branch characteristics is merited, to investigate whether the
effect of branch size is of universal significance in harvesting operations.
In conclusion, it might be noted that the data collected in an observational study of a
harvesting operation will never be able to sufficiently identify and describe the independent
variables that impact the productivity of a CTL harvester to generate a predictive equation
that would have an R-square to generate accurate production estimates. This case study
expands the data available for the logging contractor, harvest supervisor and natural resource
manager to make informed decisions. However, without investigating the elements of the
harvester cycle with all the independent variables and co-variables held to a minimum,
further investigation of the factors influencing the productive cycle of CTL machinery is
unlikely to progress further.
38
Literature Cited
Bergkvist, I. 2003. Multitree-handling increases productivity in smallwood thinning.
Skogforsk Results, No. 3, 2003. 4p.
Briggs, D. 1994. Forest products methods and conversion factors, with special emphasis on
the US Pacific Northwest. College of Forest Resources, University of Washington.
161 p.
Drolet, J.C., R.M. Newnham and T.B. Tsay. 1971: Branchiness of jack pine, black spruce,
and balsam fir in relation to mechanized delimbing. Inform, FMR-X-34, Report to
Forest Management Institute, Ottawa. 34 p.
Ewing, R.H. 2001. Four evaluations of compact tracked harvesters and forwarders in
commercial thinning. Advantage, 2(37), Forest Engineering Research Institute of
Canada, Pointe Claire, Quebec, Canada. 8 p.
Favreau, J., and J.-F. Gingras. 1998. An analysis of harvesting costs in eastern Canada.
Forest Engineering Research Institute of Canada Special Report No. SR-129. 8p.
Gellerstadt, S. 2002. Operation of the single-grip harvester: Motor-sensory and cognitive
work. International Journal of Forest Engineering, Vol. 13, No. 2. 35-47.
Gingras, J.-F. 1994. A comparison of full tree versus cut-to-length systems in the Manitoba
model forest. Forest Engineering Research Institute of Canada Special Report No.
SR-92. 16p.
Holtzscher, M.A., and R.L. Lanford. 1997. Tree diameter effects on cost and productivity of
cut-to-length systems. Forest Products Journal, Vol. 47, No. 3. 25-30.
39
Kellogg, L.D. and P. Bettinger. 1994. Thinning productivity and cost for a mechanized cutto-length system in the Northwest Pacific Coast Region of the USA. Journal of
Forest Engineering, 5(2): 43-54.
Mellgren, P.G. 1990. Predicting the performance of harvesting systems in different
operating conditions. Forest Engineering Research Institute of Canada Special
Report No. SR-67. 22p.
Mitchell, J.L. and M.A. von der Gönna. 1994. At-the-stump processing and roadside log
processing: costs and impacts of harvesting and forest renewal. FERIC Special
Report No. SR-93. 17p.
Miyata, E.S. 1980. Determining fixed and operating costs of logging equipment. USDA
Forest Service General Technical Report NC-55, USDA Forest Service, St. Paul,
Minnesota. 16 p.
Murphy, G. 2003. Procedures for scanning radiata pine stem dimensions and quality on
mechanised processors. Journal of Forest Engineering. Vol. 14 No.2, July 2003.
91-101.
Olsen, E.D., and L.D. Kellogg. 1983. Comparison of time study techniques for evaluating
logging production. American Society of Agricultural Engineers Transactions 00012351/83/2606. 1665-1668, and 1672.
Richardson, R., and I. Makkonen. 1994. The performance of cut-to-length systems in
eastern Canada. Forest Engineering Research Institute of Canada Technical Report
TR-109. 16p.
Rummer, R. 2002. Smallwood logging production and cost study summary; Medicine Bow
– Routt National Forest. Draft document, Southern Research Station, Auburn,
Alabama, USDA Forest Service. 17 p.
40
Rummer, R., and J. Klepac. 2002. Mechanized or hand operations: Which is the less
expensive for small timber? Proceedings of Small Diameter Timber: Resource
Management, Manufacturing and Markets Conference. Spokane, Washington,
February 2002. 7p.
Rummer, R., J Prestemon, D. May, P. Miles, J. Vissage, R. McRoberts, G. Liknes,
W.D. Shepherd, D. Ferguson, W. Elliott, S. Miller, S. Reutebuch, J. Barbour, J Fried,
B. Stokes, E. Bilek, K. Skog. 2003. A strategic assessment of forest biomass and
fuel reduction treatments in western states. USDA Forest Service Research and
Development. 18p.
Turner, D.R. 2003. An investigation of the productivity of cut-to-length harvesting in a
commercial thinning. Draft document. Department of Forest Products, University of
Idaho. 26p.
Vissage, J. S. and P.D. Miles. 2003. Fuel reduction treatment: A west-wide assessment of
opportunities, from Focus on Wildfire. Journal of Forestry 101(2): 5-6.
Wang, J., J. McNeel & J. Baumgrass. 2003: A computer based time study system for
timber harvesting operations. Forest Products Journal, 53(3): 47-53.
Wykoff, W.R., N.L. Crookston and A.R. Stage. 1982: Users guide to the Stand Prognosis
Model. General Technical Report INT-133, Intermountain Forest and Range
Experiment Station, Ogden, Utah, 84401, USDA Forest Service. 114 p.
41
Literature Cited (Cont.) - Technical Data Sources
24 November 2003 Neuson Website / Neuson Baumaschinen GmbH
http://www.neuson.com/default.htm
http://www.neuson.com/ASP/frameset.asp?MMARK=HARVESTER_EN&LID=EN
&companyID=BMI
http://www.neuson.com/ASP/frameset.asp?MMARK=HARVESTER_EN&LID=ENc
ompanyID=BMI
Rocan 24 May 2003. Technical sales information, Neuson 11002 HV harvester.
www.rocan.com/Neuson3.html
42
Appendix I: Treatment unit stand summary
Silviculture
Treatment Unit
Clearcut with
Reserves
Thinning
(Pre-Harvest)
Thinning
(Post Harvest)
Area (ha)
3.25
4.50
4.50
Stems/ha
710
575
193
Tree Volume (m3/ha)
253
223
150
Mean Tree Volume
0.36
0.59
0.78
(> 12.5 cm DBH)
3
0.39 (harvested)
(m )
Basal Area (m2/ha)
24.1
20.7
10.7
23
24
34
50
50
50
Douglas-fir
66
83
69
Grand fir
16
0
0
Ponderosa pine
16
16
29
Western larch
2
1
3
Quadratic Mean
Diameter (cm)
Stand Age (years)
Species Composition
(%)
43
Appendix II: Machinery specifications
Appendix II.a - Neuson 11002 HV Harvester
(Photo Credit: Neuson GmbH)
Machine weight: 11,600 kg (25,575 lbs)
Width: 2.40 m (7.9’)
Length: 3.35 m (11’) Track length
Ground clearance: 0.56 m (1.9’)
Engine: Deutz BF4M 1012 EC 4 cylinder turbo diesel
Power: 75 Kw (102 hp) @ 2300 rpm
Drawbar pull: 11900 kg (26235 lbs)
Transmission: Hydrostatic
Fuel tank: 170 l (45 gallons)
Undercarriage: Neuson D4
Tracks: 0.5 m (19.5”) width; 0.387 kg/cm2 (5.5 lbs/ sq. in.). (Rubber or steel tracks. Study
machine equipped with North American standard fitting of steel tracks.)
44
Appendix II.a (Continued).
Crane: Patu parallel action.
Reach: 9.1 m useful reach from crane pivot.
Slew: 30o to either side of crane pivot
Harvester head: Logmax 3000 (See Appendix II.b)
Configuration: Engine up leveling
Level: 4 way level, controlled by machine operator.
25o forward tilt / 25o rearward tilt.
15o lateral tilt.
Slope capability: Rated to operate on slopes to 50% by manufacturer
Data from and Ewing (2001), Rocan (2003) and Neuson (2003).
45
Appendix II.b – LogMax 3000 harvester head specifications
(Photo credit: LogMax AB)
Head specification
Weight (Including Rotator): 525kg (1155 lbs)
Maximum head width (open): 1.05 m (42”)
Minimum head width (closed): 0.87 m (35”)
Height to upper knife: 1.13 m (45”)
Height to upper rotator pin: 1.34 m (54”)
Saw specifications:
Maximum cutting diameter: 49 cm (19”)
Maximum cutting diameter (extended saw bar option): 59 cm (23”)
Saw chain: 0.404”-69 DL
Maximum chain speed: 41 m/s (8365 ft/min)
Computer: 15 log lengths + 1 pulp length, with three species specification (3 x 5 sorts)
46
Appendix III – Harvester cutting specifications
Bennett Forest Products Inc. specifications
Douglas-fir, western larch, grand fir:
10’6”, 12’6”, 14’6”, 16’6” (preferred length), 18’6”, 20’6”.
Ponderosa pine, lodgepole pine:
10’6”, 12’6”, 14’6”, 16’6” (preferred length).
Log diameter
Top diameter: 6” top diameter inside bark
Log quality
No burned or charred logs, forks, offsets, crooks or sweep.
Plummer Forest Products specifications
Acceptable species:
Douglas-fir, western larch, grand fir, western hemlock, lodgepole pine, Engelmann spruce,
sub-alpine fir, and western red-cedar.
Log lengths:
Preferred (all species except western red-cedar): 18’6”
Acceptable (all species except western red-cedar): 9’6”, 16’6”, 17’6”.
(Western red-cedar log lengths not reported as there was no cedar component in either
treatment unit.)
Log Diameter:
Minimum top diameter: 4” inside bark
Maximum butt diameter: 10” under bark. 7 to 8” preferred.
47
Appendix III (Cont.)
Log quality:
No spiral check (grain), no rot, all limbs, pig ears and burls to be cut flush with stem.
No burned or charred logs, forks, offsets, or crooks.
Sweep – for each 8-9” log segment, a tape must remain aligned with the centre.
Regulus Stud Mills Inc. specifications
Acceptable species
Spruce, hemlock, lodgepole pine, ponderosa pine, spruce, Douglas-fir and grand fir.
Log lengths
10’, 16’6” (preferred), 25’, and 33’.
Log diameter
6” top diameter inside bark.
Log quality:
No foreign objects, no rot, all limbs, pig ears and burls to be cut flush with stem.
No burned or charred logs, forks, offsets, or crooks.
No stump shot/stump pull, bias cut, under cut, or broken ends.
48
Appendix IV. - Harvester machine cost analysis
Manufacturer
Neuson
Machine Model
Machine Purchase Details
Initial Investment
Salvage Value ( % )
Salvage Value
Economic Life
Depreciation
Average Value of
Yearly Investment
11002 HV Harvester
Labour Cost
$290,000.00
20%
$ 58,000.00
5
$ 46,400.00
$
$
25.00
40%
10.00
35.00
$197,200.00
Fuel & Lubricant Cost
Fuel Cost
$
Lubricant Cost
$
0.27
0.10
/litre
/litre
Financial Details
Interest
Insurance
Taxes
Interest
Insurance
Taxes
11%
4%
3%
$ 21,692.00
$ 7,888.00
$ 5,916.00
/year
/year
/year
Machine Characteristics
Horsepower
Fuel Consumption
Fuel Cost
$
Lubricant Cost
$
$
Total Fuel & Lube Cost
75
13.25
3.58
1.32
4.90
kW
l/H
/hour
/hour
/hour
Total IIT
IIT
$ 35,496.00
$
17.75
/year
/SMH
Scheduled Operating Time
Weeks per Year
40
Days per Week
5
Hours per Day
10
Scheduled Machine
Hours (SMH)
2000
Years
/year
Wages
Benefits
Benefits
Total Labour Cost
Maintenance &
Repair Rate
Maintenance &
Repair
$
/hour
of Wages
/SMH
/SMH
Proportion
of
70%
$32,480.00
Depreciation
/year
weeks
days
hours
hours
Productive Machine Time: CLEARCUT
Productive Machine Time: THINNING*
Scheduled Machine
Scheduled Machine
Hours (SMH)
Utilization
2000
91%
hrs
Hours (SMH)
Utilization
1820
hrs
19.17
m3/hour
Hours (PMH)
Predicted Machine
Output (PMH)
Productive Machine
Hours (PMH)
Predicted Machine
Output (PMH)
Productive Machine
Predicted Machine
Output (SMH)
2000 hrs
82%
1640 hrs
18.86 m3/hour
Predicted Machine
17.45
3
m /hour
* All thinning items reported in italics.
Output (SMH)
15.47 m3/hour
49
Appendix IV. (Cont.)
Cost Elements
Scheduled Machine
Productive Machine
Hour (SMH)
Clearcut
Thinning*
Hour (PMH)
Thinning*
Clearcut
Depreciation
$ 23.20
$ 23.20
$ 25.49
Insurance,
Interest
& Taxes
$ 17.75
$ 17.75
TOTAL Fixed
Costs
$ 40.95
Maintenance &
Repairs
Per Annum
Clearcut
Thinning*
$ 28.29
$ 46,400.00
$ 46,400.00
$ 19.50
$ 21.64
$ 35,496.00
$ 35,496.00
$ 40.95
$ 45.00
$ 49.94
$ 81,896.00
$ 81,896.00
$ 16.24
$ 16.24
$ 17.85
$ 19.80
$ 32,480.00
$ 32,480.00
Fuel & Lubricants
$
$
$
$
4.90
$ 8,920.14
$ 8,037.93
TOTAL
Operating
Cost
$ 20.70
$ 20.26
$ 22.75
$ 24.71
$ 41,400.14
$ 40,517.93
Wages
$ 25.00
$ 25.00
$ 27.47
$
30.49
$ 50,000.00
$ 50,000.00
Benefits
$ 10.00
$ 10.00
$ 10.99
$ 12.20
$ 20,000.00
$ 20,000.00
TOTAL Labour
Cost
$ 35.00
$ 35.00
$ 38.46
$ 42.68
$ 70,000.00
$ 70,000.00
$ 96.65
/hour
$ 96.21
/hour
$ 106.21
/hour
$ 117.33
/hour
$193,296.14
/year
$192,413.93
/year
34,889.40
m3/year
30,930.4
m3/year
Fixed Costs
Operating Costs
4.46
4.02
4.90
Labour
TOTAL MACHINE
COST
Production Rate
Cost Per Unit
Production
17.45
m3/hr
$
5.54
/m3
15.47
m3/hr
$
6.22
/m3
* All thinning items reported in italics.
19.17
m3/hr
18.86
m3/hr
50
Appendix V. – Neuson harvester research statistics: summaries by treatment
Table 1: Neuson harvester clearcut summary statistics
Variable
Travel & Harvester Head Locate (sec)
Distance (m)
Slope(%)
Brushing (sec)
Brush code
Falling (sec)
Processing (sec)
Total Cycle Time (sec)
DBH(cm)
Tree Height (m)
Tree Volume (m3)
Branch Size
Branch Interval
Total Branch Score
N
397
397
397
397
397
397
397
397
397
397
397
397
397
397
Mean
16.87
2.32
13.27
6.24
0.30
5.84
25.26
53.82
21.63
18.66
0.28
1.39
2.32
3.47
Std.
Dev
13.68
5.67
8.41
16.01
0.59
3.10
22.47
34.24
5.41
3.68
0.20
0.81
0.49
1.46
Max.
Value
119.76
50.00
38.00
167.30
3.00
26.00
245.81
302.00
35.00
28.80
0.99
3.00
3.00
6.00
Min.
Value
2.36
0.00
0.00
0.00
0.00
1.57
3.83
14.10
10.00
10.20
0.01
0.00
1.00
0.00
Table 2: Neuson harvester thinning summary statistics
Variable
Travel & Harvester Head Locate (sec)
Distance (m)
Slope(%)
Brushing (sec)
Brush code
Falling (sec)
Processing (sec)
Total Cycle Time (sec)
DBH(cm)
Tree Height (m)
Tree Volume (m3)
Branch Size
Branch Interval
Total Branch Score
N
358
358
358
358
358
358
358
358
358
358
358
358
358
358
Mean
18.53
4.00
9.38
5.59
0.38
4.90
20.85
49.86
21.36
18.16
0.29
1.39
2.60
3.56
Std.
Dev
16.74
9.29
6.40
12.15
0.67
2.95
19.72
32.68
5.59
4.08
0.18
0.90
0.49
1.68
Max.
Value
205.80
140.00
28.00
82.80
3.00
29.40
181.20
240.60
35.00
30.40
0.93
3.00
3.00
6.00
Min.
Value
4.20
0.00
2.00
0.00
0.00
0.60
4.20
13.80
12.00
7.10
0.05
0.00
2.00
0.00
51
Appendix VI. – Research data collection forms
Form 1: Tree data sheet
Tree
Number
Tree
DBH
(cm)
Tree
Height
(M)
Crown Proportion ( % )
Clean
Dead
Live
Bole
Tree
Form
Branch
Size
Score
Branch
Spacing
Score
Total
Branch
Score
Date
#
# Posn.
(m)
(%)
Cycle Tree Trav. & Dist. Slope Swamp
Proj.
Code
Brsh
Obs Start
Felling Process L Logs S Logs
Pile
Obs End
Prsnl Un ID. Maint. Brkdwn Comments
Obsrvr
52
Appendix VI (Cont.) – Form 2: Harvester data collection form