Next Article in Journal
Analyzing the Eucalyptus Timber Value Chain in Thailand: Profit Distribution and Opportunities for Improvement
Next Article in Special Issue
Timber Extraction by Farm Tractors in Low-Removal-Intensity Continuous Cover Forestry: A Simulation of Operational Performance and Fuel Consumption
Previous Article in Journal
Virulence and Pathological Characteristics of a New Metarhizium anisopliae Strain against Asian Long-Horn Beetle Anoplophora glabripennis Larvae
Previous Article in Special Issue
Assessing the Residual Stand Damage after Thinning with Different Levels of Mechanization
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Forests
Volume 15
Issue 6
10.3390/f15061046
forests-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

A Multi-Criterion Evaluation Process for Determining Cost-Effective Harvesting Systems in Fragmented Boreal Forests

by
Léo Painchaud
1 and
Luc LeBel
2,*
1
FORAC Research Consortium, Université Laval, Québec, QC G1V 0A6, Canada
2
Department of Wood Science and Forest, Université Laval, Québec, QC G1V 0A6, Canada
*
Author to whom correspondence should be addressed.
Forests 2024, 15(6), 1046; https://doi.org/10.3390/f15061046
Submission received: 3 May 2024 / Revised: 5 June 2024 / Accepted: 13 June 2024 / Published: 17 June 2024
(This article belongs to the Special Issue Sustainable Forest Operations Planning and Management)
Browse Figures
Versions Notes

Abstract

:
Nordic forests, like those found in Canada, used to consist of large and relatively homogeneous mature stands. Such a spatial pattern allows for harvest operations to be highly concentrated, minimizing procurement costs. However, the growing fragmentation of these forests makes planning difficult and increases the costs of road building and machinery relocation. While operational solutions have been developed in regions with small harvest areas, their transferability to different settings is unknown. Finding the most suitable combination of equipment for a given context is challenging considering the multitude of possibilities. The objective of this study is to identify, from all possible options, a subset of harvest systems expected to perform well in fragmented boreal forests. The results from this research are two-fold. First, a comprehensive review of forest machines and harvest systems is provided. Second, a multi-criteria decision analysis (MCDA) methodology is proposed to evaluate the alternatives. In a boreal forest context, the conventional harvester-forwarder system (CTL) was ranked among the best solutions, along with mild adaptations of the usual configurations. Several whole-tree (WT) system configurations were also highly ranked. While the results are specific to the case studied, the review and selection methodology can serve in different operational contexts.

1. Introduction

Harvest operations in the boreal forests of eastern Canada were traditionally planned within vast uninterrupted and homogeneous areas. This provided the opportunity to maximize harvested volume per kilometer of road built and to minimize relocation of machines and workers. Though this approach may often lead to emulate spatial patterns of natural disturbances [1], it may not be entirely compatible with wildlife habitat protection and community expectations for lower-impact forestry. Considerations of these social and environmental issues lead to smaller and more dispersed cut-block areas, which reduced wood volumes available for harvest [2]. In effect, this leads to an increased dispersion of harvest areas across the landscape. Dispersion negatively impacts the efficiency of timber procurement activities, mainly because of frequent relocation and the lower volume available for harvest for given investments in roads [3,4,5,6], thus increasing the harvesting cost [7]. Dispersed patch harvesting also increases the complexity of operational supervision, logistics, and planning.
Specific means to counter the negative financial impacts of dispersion mainly fall into two categories: (i) cut-block design and (ii) production systems. The cut-block design approach attempts to optimize cut-blocks, landing, and road layouts. Harvest area planning and cut-block design have been studied thoroughly over the years [8,9,10,11,12,13]. Though most studies do not focus on fragmented forests, the methods they propose remain relevant.
Scientific papers rarely describe complete harvest systems but rather individual machines as alternatives for specific tasks. For this reason, our review identifies and describes all machine types and work methods used in forest operations to generate potential solutions that are currently unknown in eastern Canada.
The production system approach aims at selecting or designing machine systems used to best execute the operations and develop work methods and logistic strategies. There are two main production methods in eastern Canada: cut-to-length (CTL) (harvester and forwarder), and whole-tree (WT) (feller buncher, grapple skidders and stroke-boom delimbers) [14]. Timber is piled at the roadside then loaded onto trucks by a knuckle boom loader or self-loading trucks. Transportation to the mills is decoupled from harvesting and occurs weeks or months after trees are cut [15] because of the seasonality of road access. Few contractors diverge from these methods.
The issue of increased dispersion and tract size reduction is not confined to eastern Canada. Parcelization of private land in the United States (US) has similarly been a source of concern [16,17]. There, experts have suggested that forestry contractors should adapt production systems to stay competitive [18]. More labor-intensive harvest systems or those with lower capital investments have been proposed, although the lack of available workforce is limiting their applicability [19,20]. Likewise, the use of older equipment may result in a financially beneficial trade-off between decreased productivity and capital investments [21]. In South Carolina, contractors have adjusted to tract size reduction by decreasing the number of employees or machines required by the harvest system [22].
Solutions developed in other contexts but also focused on issues such as increased road construction and relocation costs or decreased revenues can be applicable to the boreal region. An example is the case of biomass recovery operations in the US, where economic viability is challenged by low product value [23]. Potential solutions from biomass harvesting include the use of highly maneuverable trucks to haul woody residues to central chipping landings [24,25]. As chip vans have limited adaptability on off-highway roads [26], this approach increases accessibility to residues [27], reduces road construction costs [28], and limits machine relocation [29]. These benefits must be sufficient to counterbalance the additional handling costs and lower efficiency of smaller trucks [30]. In New Zealand, a similar centralization approach is referred to as two staging and is employed in roundwood harvest operations to solve the issue of high road and landing construction costs on steep ground [31]. The centralization of processing limits infrastructure construction to strategic locations, lowering associated costs [32].
In Scandinavia, low revenues in early thinning operations have led to a growing interest in harwarders, which are machines capable of both felling and extracting timber [33,34]. Their anticipated use for thinning, as well as dispersed cut-block harvest, is motivated by the lower-relocation costs than when multiple machines are used [35,36,37]. A drawback of this approach is the lower production and higher operating costs of combined machines, especially during extraction [38,39].
Although all of the above solutions demonstrated potential in their respective contexts, the suitability of their transposition to a different environment is unknown. Compared to regions of the world where most of these solutions were developed, the forest industry of eastern Canada is facing additional financial challenges. The first is poor forest productivity, with average annual volume growth of barely 1 m3/ha/year [40]. The average harvested tree volume is only 0.113 m3, and the average harvested volume is 116 m3/ha [41]. In such conditions, operations are costly, and forest products tend to have a low value. Other important challenges are the requirement for investment in road access and the long transport distance from cut-blocks to mills, with an average of 151 km for the Province of Quebec [41]. Transportation costs, as well as the construction and maintenance of the forest road network, represent almost 50% of operational costs [42].
The combined effect of these factors raises questions as to whether solutions developed elsewhere are suitable to eastern Canada. In any case, the potential solutions question the current practice of using the same harvest system for both clustered and dispersed cut-blocks. There is a need to investigate systems that could reduce the harvest costs of dispersed cut-blocks in the boreal forests of eastern Canada. To that end, a systematic review of available harvest systems is required. Also, planning procurement activities in small and dispersed cut-blocks is challenging and selecting the harvesting system best adapted for this context is complex. Solutions exist to optimize the planning of forest operations and can be applied to a wide range of contexts. Some studies can guide the selection of the best harvesting equipment [43,44,45], but not in fragmented conditions like those found in the eastern Canadian boreal forest, and none, to our knowledge, adopt a complete system approach.
The objective of this paper is to identify the most cost-effective harvest systems to operate in fragmented boreal forests of eastern Canada. Achieving this objective involves (1) systematically reviewing existing harvest systems and (2) providing a method for selecting harvest systems for fragmented boreal forests. The underlying hypothesis is that harvest systems currently used in eastern Canada represent only a fraction of the complete set of potential alternatives. This paper therefore provides an updated description of practices in ground-based mechanized harvest and suggests a structured evaluation framework. These results should be adaptable to a wide array of environments.
The remainder of this paper is structured as follows. First, the research methodology details the development of a framework to identify available options, starting with a review of fully mechanized ground-based harvest systems. The multi-criteria decision analysis (MCDA) approach used to identify relevant evaluation criteria, evaluate the system performance, and assess the weight of the criteria is then described. The results of the review and selection procedure for fragmented operations are subsequently presented. Lastly, the results of the study and possible future research are discussed.

2. Material and Methods

The scope of this study was limited to fully mechanized harvest systems. Also, only ground-based systems for slope grades under 40% were considered, since this is the upper threshold for harvesting operations in eastern Canada [14]. Terrain roughness and snow accumulation in the region do not allow for the effective use of drive-to-tree machines, leading to their exclusion from the analysis. Also, because the regional demand for woody biomass is still limited, integrated systems were not considered for this study. Should this evolve in the future, biomass recovery could represent an opportunity for harvesting dispersed cut-blocks, as it would increase the total value extracted from a given tract and reduce fixed costs per unit. Finally, analyses were limited to in-wood operations, excluding transport operations and potential solutions involving permanent log yards.
The following section describes the methodology developed to review existing harvest systems included in the scope of this study. The procedure used to evaluate and rank the alternatives is described in the subsequent section.

2.1. Review of Existing Systems

To consider all possibilities included in the scope of this study, a thorough review was conducted, inspired by [46,47]. As a starting point, forestry equipment in manufacturers’ catalogues was reviewed to list machine types currently available on the market. These were then used as key words in a search through the Google Scholar database as well as several national forest research organizations’ publication catalogues to document associated work methods, capabilities, and drawbacks (FPInnovations, New Zealand Forest Growers, Skogforsk and USDA Forest Service). Non-peer-reviewed sources of information were researched in addition to academic papers to seek practices that have seemingly not yet been formally studied. Apart from brochures of machine manufacturers, these included videos of forest operations presented on social media platforms.
A systematic procedure was developed to generate all combinations of harvest systems from a list of machine types and work methods. Given that generating all potential combinations leads to many incoherent systems, a set of filtering rules was integrated in the script, taking into consideration the specific environment of eastern Canada’s boreal forests. The procedure was first developed for machines and work methods currently used in Canada, resulting in the initial list of systems. It was then adapted in a second iteration to include the machines retrieved in the international review. The resulting list of harvest systems is considered as the complete set of alternatives for operations in eastern Canada.

2.2. Systems Evaluation

Simulation is a proven way to identify the most cost-effective harvest system among potential alternatives (e.g., [48,49,50]. However, when considering a large set of alternatives, this approach becomes inefficient and cumbersome. MCDA has successfully been used as an alternative to simulation approaches for the selection of the optimal harvest system. For example, ref. [51] used a subjective score grid for unweighted performance criteria to evaluate the environmental performance of different harvest systems. Ref. [52] consulted experts to evaluate the performance of alternatives about relevant criteria and ranked them according to weights provided by practitioners. Ref. [53] developed a spatial decision model integrating an initial filtering of systems according to their limitations, followed by a multi-criterion utility evaluation based on direct weighting of criteria and numeric performance indicators. An MCDA was then developed for a preliminary evaluation of harvest systems. Key steps were the selection of evaluation criteria, the identification of alternatives, their evaluation for each criterion and finally, the weighting of criteria and ranking of alternatives. The methodology employed to achieve these steps is schematized in Figure 1 and described in the following sections.

2.2.1. Evaluation Criteria

Key criteria to evaluate the performance evaluation of harvest systems for fragmented operations include road investments, mobility (or ease of machine relocation), productivity, and machinery operating costs. Road investments and machine relocation represent the main cost elements impacted by dispersion [3,5,6]. Productivity and machinery operating costs are important parameters for any harvest system in any environment. Industry experts familiar with fragmented cut-blocks were consulted using semi-structured interviews to validate the relevance of the identified criteria. Six wood procurement superintendents for softwood lumber sawmills in north-central Quebec, Canada, were selected with a purposive sampling approach used by [54]. They were questioned on their respective industrial context, level of forest fragmentation in their area, and its impacts on wood procurement. Experts were also invited to share solutions that they deemed interesting.

2.2.2. Alternatives

The complete list of harvest systems generated from reviewed machine types and work methods was not used as an input for the evaluation procedure because of its size (1448). Instead, considering the validated criteria, an approach inspired by [47] served to select solutions for further considerations. First, the two harvest systems used by consulted companies were chosen as base-case systems. These are (1) the harvester/forwarder CTL system and (2) the feller buncher, grapple skidder and stroke-boom delimber WT system, both with a knuckle boom loader and standard trucks for cold transport. Second, all machines and work methods that could constitute potential adaptations of base-case systems for fragmented operations were identified. This resulted in a set of 28 alternatives to be further considered in the selection process.

2.2.3. Performance Evaluation

System performance evaluation was based on a quantitative indicator chosen for each criterion, an approach used by [53]. The performance indicator and evaluation method for each criterion are presented as follows:
(a)
Road investments—Extraction distance (m)
As forest road density is mainly a function of extraction distance, the latter was chosen as an indicator [55,56,57]. The performance value was based on theoretical maximum distances for extraction machines presented by [58].
(b)
Mobility—number of machines
The number of machines to be moved was chosen as a performance indicator for mobility. Machines operating at a central landing were assigned the value of 0.5 machine because they are not relocated as frequently. For this purpose, it was assumed that systems involving a central landing required a knuckle boom loader and a processor. All truck types were excluded as their relocation is not considered. Since wheeled machines in eastern Canada are usually equipped with chain tracks, their travelling speed is greatly limited, making this factor less relevant.
(c)
Productivity—System productivity (m3/PMH)
Hourly production of machines was estimated for the average stem size in eastern Canada, rounded to 0.1 m3. Moreover, the productivity for extraction machines was estimated based on the mean extraction distance of half their theoretical maximum used as a road investment indicator. For machines used in eastern Canada, the estimations were taken from productivity models developed by FPInnovations. The most recent references available for other machines were selected based on their similarity to the studied context regarding stem size, species, harvest type and extraction distance. From these numbers, the systems were balanced when required and the resulting bottleneck operations determined total productivity.
(d)
Machinery operating costs—Unit cost (USD/m3)
The unit cost of each system was calculated from the sum of unit costs of individual machines using the COST model [59]. Values for several input parameters were retrieved from Procalc and kept constant for all machines (Table 1). Likewise, variable parameters for machines implemented in the software were used in the COST model, such as purchase price and fuel consumption. Again, machines not used in eastern Canada required data from studies conducted elsewhere. Purchase prices retrieved from past studies were converted to US dollars using historical Official Exchange Rates [60] and updated to the reference year (2019) using the “production price index for construction machinery” [61], following the method used by [62]. Annual maintenance and repair costs as well as utilization rates were estimated based on the machine used in eastern Canada considered as the most similar.
Retrieved and calculated indicators were classified according to the scales presented in Table 2. Based on concerns raised by the experts, these rates were adjusted using performance modification factors, as inspired by [58]. This allowed for fully capturing the performance of systems for which the indicators are not entirely representative.
(e)
Weighting and ranking
The PAPRIKA method, developed by [63], was used as the criteria weighting technique to rank the systems. It relies on experts to indicate their relative preferences for pairs of alternatives with performance trade-offs for two criteria at a time, while considering all other criteria as equal. This results in a system of inequalities between alternatives, which is solved through linear programming to obtain the criteria weights and final ranking. The procedure was conducted with the 1000minds online software, allowing the distribution of a survey through which the experts were consulted once again. This approach was chosen for its simplicity and ranking accuracy. Weights and rank frequencies of alternatives resulting from the survey were directly retrieved from 1000minds. The final rank of a system was determined based on the median rank given by the different experts.
To evaluate the reliability of the ranking, Kendall’s coefficient of concordance (W) was calculated and corrected for tied ranks (0: no agreement; 1: full agreement). The permutation test of Kendall’s W, described by [64], was performed with 1000 permutations to test for the null hypothesis that the ranking by the experts was all independent (no agreement).

3. Results

3.1. Review of Existing Systems

Forestry machine types and work methods retrieved in the review of fully mechanized ground-based harvest systems are highlighted in Table 3. They are briefly described in the following sections, grouped by operation phase.

3.1.1. Felling

The harwarder, or combi, is the combination of a harvester head with a grapple or clambunk skidder carrier or, more commonly, a harvester head with a forwarder carrier. Figure 2 shows the Komatsu X19 prototype (Komatsu Forest, Umeå, Sweden). Other existing harwarder models include the Timberpro 830/840D (TimberPro Inc, Shawano, WI, USA), and, and the Usewood Combi Master (Usewood Eesti OÜ, Pärnu maakond, Estonia).
The harwarder relies on a harvester head specifically designed to also be used as a grapple for log handling [36,65]. An advantage of this machine is the possibility to directly load logs into the bunk while processing, which is facilitated by purpose-built rotating bunks [38,66]. Otherwise, direct loading requires the complicated maneuver of processing trees from the back end of the bunk, justifying the use of work methods involving independent loading [67,68]. Innovative work methods have been researched for the use of direct loading with a harwarder, such as the autonomous load-changing system described by [48], where the harwarder switches loads with an empty forwarder, which then proceeds to extraction. While combi machines have lower relocation costs, their potential is limited to small tracts with short extraction distances because of their high operating cost in comparison with forwarders [69].
For the scope of this review, a feller director, sometimes referred to as a directional feller, is a swing machine with a dangle bar saw felling head. It is normally intended to fell large timber. With their high log handling capacity, feller directors can also be used as a loader for extraction, processing, or loading.
Like the combi concept, the dual is a convertible machine that first proceeds as a feller before converting to an extraction machine [36]. As for the combi, any combination is possible. Harvester forwarders are, however, the most common option offered by manufacturers (Ponsse Buffalo Dual, Timberpro 830/840 D, Entracon EC75/60/45, Vimek 404 Duo). Compared to the combi, a dual may be used for longer extraction distances [70,71], and their production in either forwarder or harvester mode is similar to specialized machines [72,73]. They are swing machines with a dangle bar saw felling head. Normally intended to cut large trees, such a machine would usually not be considered as a viable option in the boreal forests of eastern Canada. However, in small and dispersed tracts, the benefit of its versatility, limiting the number of machines to be relocated, might counterbalance its low productivity in stands with small trees.

3.1.2. Extraction

Shovel loggers, which are essentially knuckle boom loaders, are versatile machines which can move timber from the stump to roadside and load trucks [74,75]. Their use may be advantageous regarding relocation costs, as fewer machines are required. However, as they grab logs or trees in a repetitive fashion, they tend to be less productive as distance increases when compared to other ground-based machines [76]. Their use for extraction is therefore limited to small tracts. They are also well suited to limiting soil disturbance, both on low bearing capacity and steep ground [76,77,78].
The addition of a purpose-built trailer to a shortwood forwarder (see Figure 3) increases its load capacity by 60% but is ill suited to a rough terrain [79]. This raises productivity and fuel consumption efficiency. However, this solution is financially beneficial only over long extraction distances of 600 m or more because of the long loading and unloading times [80].

3.1.3. Processing

Some machines used for extraction can also contribute to processing operations by delimbing trees, at least partly. To achieve this, grapple skidders break limbs by backing up their load through a delimbing gate (see Figure 4), which is a sturdy metal grid that can be assembled by any metalworker [81].
Although it decreases skidder productivity and results in poor quality delimbing, skidding is seldom the bottleneck of whole-tree systems and delimbing is of sufficient quality for pulpwood [81,83,84]. Grapple skidders may be more efficient when using a drive-through delimber. These are static devices equipped with hydraulic delimbing knives closing on the stems as the skidder pulls them through on its way to the landing. Only one or two stems may be processed with such a machine, which still allows for attaining good productivity for timber with large dimensions [85], but would be too unproductive for smaller dimensions. Accordingly, this alternative was not considered for further analysis.
If operated by a knuckle boom loader at the landing, a pull-through delimber can be advantageously used in association with a skidder and a delimbing gate [81]. Following the same principle as the skidder drive-through delimber, pull-through delimbers usually have a rotating base and are mounted on a trailer (see Figure 5). The loader then delimbs and optionally tops or slashes trees while either piling processed stems before a truck arrives at the landing or loading trucks directly as stems are processed [84,86].

3.1.4. Loading

There are many alternatives other than knuckle boom loaders to accomplish loading, starting with self-loading trucks, which are widespread in North America. What is much less common are detachable crane consoles for self-loading trucks. This equipment allows the driver to load the truck, decouple the crane, and leave it in the woods for the next trip (see Figure 6).
The reduced tare weight increases payload capacity, which is profitable when hauling over long distances [87,88]. However, there are situations, such as consecutive trips from different tracts, or when a backhauling opportunity requires the truck to be self-loading, which require the crane to be kept on [89,90]. Front-end loaders are also an alternative to knuckle boom loaders in various conditions (e.g., [91,92]. However, they require more space to work effectively, making them unsuitable for roadside or small landing operations [93,94]. They were not considered as a possible option for extraction in further analyses because of the terrain roughness and snow accumulation in eastern Canada. Forwarders are occasionally used for loading trailers (e.g., [91,95,96]), which may be preferable to mobilizing an independent loader [97]. The loading operation conducted by a forwarder may be integrated with the extraction phase by merging the log handling tasks of unloading itself and directly loading the trailers. Alternatively, loading may be executed immediately when a truck arrives at the landing, or only after all logs are extracted, with the forwarder then acting as an independent loader [98].
Processing machines at roadside such as stroke-boom delimbers and processors may also perform truck loading. This task may be integrated in a processing work cycle by placing trees on the truck rather than piling them on the ground. Truck loading may also be decoupled from processing, with the machine acting as a knuckle boom loader until loading is done [99]. Processing heads have been adapted for enhanced grapple capacity, making processors capable of truck loading (see Figure 7, [100]). However, the same cannot be said of stroke-boom delimbers, which have poor multiple stem handling efficiency, potentially resulting in unacceptable truck turn times when loading small diameter timber.

3.1.5. Transport

The use of adapted all-wheel-drive trucks on rough trails has the potential to reduce road investments [31,102]. Indeed, a two-stage approach involving small roadside landings and a centralized one for processing and the loading of highway trucks makes it possible to build steeper roads with tighter curves [91,94,103]. Furthermore, the forwarding of unprocessed stems to a centralized landing serving multiple tracts would also reduce required relocations of processing and truck loading equipment. There are different logistical configurations when implementing a two-stage approach depending on the equipment used for secondary extraction. First, the off-road capabilities vary between potential solutions, with the most capable being purpose-built extraction machines. Other solutions, referred to as ‘super-forwarders’, include articulated trucks adapted for timber transport and all-wheel-drive off-highway trucks, like the Bell T403 and Tatra TERRNo1 (see [32]). Second, these machines may have self-loading capabilities. In addition to conventional cranes, hook-lift systems with preloaded roll-on/off bins or preloaded trailers can be used as set-out trailers.
Compared to cranes, hook-lift trucks allow the decoupling of trucking from loading and reducing truck turn time [24,29]. Although this solution is more readily suited to woody debris or chips [27,30], it can be adapted to roundwood transport trucks with roll-on/off bunks. These may either be loaded by a loading machine at the roadside or by a forwarder equipped with a hook-lift system in forests [104,105].
The use of set-out trailers presents similar advantages, except that it may also be used for long-distance trucking with regular trailers. The reason for this is that, in comparison, roll-on/off bins are limited in capacity, as the hook-lift system must be powerful enough to pull the load from the ground onto the truck. A common feature of both alternatives is the possible avoidance of multiple handlings of logs by merging forwarder or processing machine unloading and trailer loading (e.g., [95,106,107]. Set-out trailers equally reduce truck turn time by decoupling transport from loading [25,28,108]. However, this solution requires sufficient space to park multiple trailers and allow loading while not interfering with active transport.

3.1.6. Complete Harvesting Systems

The procedure for generating a combination of machine types and work methods, coded in R script, was first applied to current practices in eastern Canada, resulting in 72 harvest systems. Partly inspired by the ‘Harvest network’ of [47], applied a second time to include alternatives from the review, the procedure resulted in a total of 1448 coherent harvest systems, synthesized in Figure 8. The chart tracks the system configurations resulting from the combination procedure.

3.2. Systems Evaluation

3.2.1. Evaluation Criteria

Four aspects or criteria were discussed with six operations managers to estimate the performance of each candidate system in the context of fragmented operations: (1) the required investments in road construction and maintenance, (2) machine relocation costs, (3) overall productivity, and (4) machinery operating costs. The managers were interviewed to further define their operational context and to validate the relevance of the chosen criteria. Table 4 summarizes the information retrieved during these interviews. The table presents for each manager an overview of their operating context, how often they consider to be operating in fragmented forests, and solutions they applied for these situations. In the context of this research, dispersion was defined qualitatively following discussions with logging managers. By consensus, it was agreed that dispersion was a concern if teams had to be moved using trailers (flatbed trucks) once or more per week and whenever less than 1500 m3 was harvested per km of road to be built or maintained.
The experts reported a few solutions that they currently applied for fragmented forests. They attempted to minimize road construction costs and machine movements by extending skidding distances (up to a maximum of 1 km) and paying financial compensation to contractors for wood beyond 500–600 m. They would usually schedule fragmented areas for the winter months, when roads are cheaper to build. Investment in road construction and maintenance was critical for all experts, confirming its relevance as an evaluation criterion.
Companies using both WT and CTL systems preferred to use CTL teams for fragmented operations. The benefits identified from this practice were not only that relocation costs were minimized, but also that the lower productivity (m3/day) of smaller teams facilitated logistical planning and supervision. Scheduling lowbed trucks appeared to become an issue as relocation frequency increased. Alternatively, self-relocation (under own power) was used for short distances, with a reported absolute maximum of 3 km for harvesters and 5 km for forwarders. Hence, machine relocation was confirmed to be an important criterion when choosing the best harvest system for fragmented operations.
Though most experts reported that their contractors used recent equipment providing high productivity, expert #4 disagreed. With virtually all his operations in fragmented forests, he considered that contractors using old, fully paid equipment were less impacted by low utilization rates resulting from frequent relocations. Other experts thought that high productivity was more important and were not willing to trade it for lower machine operating costs or fewer relocations. For example, though most experts mainly used larger-capacity trucks for transport, they all had access to self-loaders. However, they only favored their use as an alternative to mobilizing a loader over short distances (less than 17 km according to expert #3). Productivity is therefore a valid performance evaluation criterion for system selection according to all experts.
Experts generally lacked enthusiasm towards implementing new harvest systems. For example, none considered the possibility of using hot loading, especially during winter, as transport operations would not be concentrated, increasing costs of snow removal and sand spreading. Furthermore, those who were operating CTL systems believed a conversion of mills to receive tree length products to be unworkable.

3.2.2. Alternatives and Performance Evaluation

System productivity and machine operating cost indicators require the calculation of intermediary results for individual machines. All such parameters are presented in Table 5, along with the reference used to identify them.
The two chosen base-case systems along with their potential adaptations for fragmented operations are presented in Table 6. Performance rating resulting from the indicator value level (see Table 2 for scale levels) and relevant modification factors for all criteria are indicated. Modification factors (+1 or −1) for road investments include the use of secondary extraction (+1) and allowing a decrease in road construction costs from the central landing to each block. Also, the use of hot loading has a negative impact (−1), as it increases road maintenance costs during winter. For hourly costs, the use of set-out trailers, which involve the acquisition of supplemental trailers, has a negative impact (−1). Table 6 shows that while CTL systems are more mobile, WT systems are more productive, except when they involve some form of shovel logging. Their ratings pattern is similar to CTL systems involving a harwarder, with low productivity and high mobility. These results show that a rigorous weighting scheme is needed to further assist system selection.

3.2.3. Weighting and Ranking

The criteria weights resulting from the expert’s PAPRIKA survey are presented in Figure 9. The median weight of the machinery cost criterion was the highest with 31.9% followed by roading and relocation, at 24.7% and 24.3%, respectively. The productivity criterion was rated less important, with 16.5%.
The rankings provided by the experts are presented in Table 7. The median ranks of the top alternatives indicate their dominance over the other systems. The CTL system using removable crane self-loaders for transport came out as the best alternative. The CTL base-case system is the 4th best system according to these results, while the WT base-case system is in the 8th position. Kendall’s W, calculated to assess the level of agreement of experts’ ranking, had a value of 0.690. The permutation test rejected the null hypothesis (p = 1.0 × 10−3), which along the linear aspect taken by the ranking frequencies presented in Table 7 confirms a certain level of agreement between the experts [64].

4. Discussion

According to the median rank in the evaluation procedure, a few harvest systems seem to outperform the other alternatives. Indeed, median ranks follow a steady progression from the first to the seventh position, after which the increase is slightly larger, with several equivalent solutions. The concordance of experts’ ranking provides credibility to these results. However, definite conclusions should not be made solely according to the scores of close alternatives because of the inevitable uncertainties associated with a multicriteria analysis. The top seven alternatives could therefore be subjected to closer scrutiny, such as detailed simulations or field testing. These include the CTL base-case system, comprising a harvester, a forwarder and a knuckle boom loader. It is reassuring that the base cases are among the better systems. This could explain why some experts had difficulty suggesting which new systems would be better adapted. It also suggests that the gains obtained from a new system might be marginal and not worth the financial risks.
Nevertheless, the analysis yielded new systems for eastern Canada. Some appear relatively simple to implement. For example, the use of an auxiliary forwarder trailer is a small adaptation of the CTL base-case system, yet it yields the best-ranked alternative. The option that is ranked third involves the use of self-loaders with a removable crane for transport. This also appears like a minor adaptation for contractors.
However, some of the top-ranked WT systems (second, seventh, and eighth) may not be as simple to implement. Indeed, they may differ from current practices such as the use of small teams. However, two of them consist of machines currently used in eastern Canada, namely the forwarder and the clambunk (ranked second and seventh). The clambunk system with processors used for loading trucks in addition to processing is untested. Grapple processor heads specially designed for effective log handling have not yet been imported to eastern Canada to our knowledge. Otherwise, the main adaptation resides in the logistics necessary to merge processing and loading.
The sixth highest ranked system uses the forwarder for loading in addition to extraction. Wood must be loaded on trucks as it is harvested, implying a drastic adaptation in inventory management and transportation. As previously stated, trucking is usually performed weeks after harvest in eastern Canada. However, no investment for a base-case CTL team is required to implement this solution, making it low risk and easy to test.
A counterintuitive result is the poor performance of harwarder systems, with the first one only ranking in 16th position. While harwarders have been tested in Scandinavian countries for harvesting small, dispersed tracts [35,37,38], their limited effective extraction distance greatly reduces their potential in eastern Canada, where road access is more limited. The low productivity of a single machine system also penalizes harwarders according to our results. However, this solution has an advantage regarding workforce efficiency, which our analysis does not account for. Indeed, comparing productivity/workforce ratios of a harwarder system with a heavily mechanized one could favor the harwarder. For instance, a harvester and forwarder team producing 12.8 m3/PMH requires two operators, resulting in a 6.4 m3/PMH/man ratio, while a harwarder outputs 7.7 m3/PMH with a single operator.

5. Conclusions

Harvesting conditions have evolved as forestry policies suggest or impose different silvicultural treatments such as constant covert management, cut-block patterns, and landscape-level dispersion patterns. For practitioners, it can be difficult to choose from the large selection of existing forestry machines and assemble a system that meets financial and operational constraints. In Canada, one important change to have occurred in recent decades is related to the requirement for more dispersed operations over the large forest landscape.
To help companies better face this challenge, an MCDA was developed to compare the performance of fully mechanized ground-based harvest systems from around the globe. Candidate systems for fragmented boreal forests were established through a thorough literature review. The latter allowed cataloguing 1448 potential machine combinations, among which only 72 currently exist in eastern Canada. The synthesis and contextualization of up-to-date information regarding existing harvesting systems are additional contributions of this study. It could assist companies in adapting their practices in other operational environments. For the specific case of the boreal forest of Canada, this study offers new alternatives for conducting forest operations. From a methodological perspective, this paper presents a detailed evaluation process that can be replicated in different regions and different selection criteria.
The high number of alternatives generated does not allow for a proper comparison of harvesting costs. A gradual selection approach was therefore developed. Key evaluation criteria were retrieved from the literature and validated by six experts. The performance of harvest systems was determined from numeric indicators. Harvest systems were ranked using a rigorous method never used before in forestry (PAPRIKA) involving a survey of experts’ preferences among alternatives regarding identified criteria. Among the seven systems that topped the ranking, four are not currently used in eastern Canada. The results confirm the high suitability of the conventional CTL system, ranked as the fourth-best solution. Other dominant solutions include minor adaptations of the latter and WT systems. The experts’ rankings were relatively similar, which bolsters their credibility. Further analysis is needed to measure the potential gain and risks of implementing these innovative solutions. For this, we suggest first developing detailed stochastic simulations of the best-performing systems. Then, based on these results, field experiments should be conducted.

Author Contributions

Conceptualization, methodology, L.P. and L.L.; software, L.P.; validation, formal analysis, investigation, L.P. and L.L.; resources, L.L.; data curation, writing—original draft preparation, L.P.; writing—review and editing, L.L.; supervision, project administration, funding acquisition, L.L. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by NSERC, grant number IRCPJ 545469-18 and NSERC Discovery grant program.

Data Availability Statement

Key data used for this paper are presented in the included tables and figures. All codes are available on demand (corresponding author).

Acknowledgments

The authors would like to thank the FORAC research consortium and its partners for their financial support of this project. The authors are also grateful for the editorial support from Shuva Gautam and François Sarrazin. Skogforsk (Lars Eliasson, Rikard Lundqvist) and Ola Lindros generously shared pictures for this article.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Belleau, A.; Bergeron, Y.; Leduc, A.; Gauthier, S.; Fall, A. Using spatially explicit simulations to explore size distribution and spacing of regenerating areas produced by wildfires: Recommendations for designing harvest agglomerations for the Canadian boreal forest. For. Chron. 2007, 83, 72–83. [Google Scholar] [CrossRef]
  2. Dhital, N.; Raulier, F.; Asselin, H.; Imbeau, L.; Valeria, O.; Bergeron, Y. Emulating boreal forest disturbance dynamics: Can we maintain timber supply, aboriginal land use, and woodland caribou habitat? For. Chron. 2013, 89, 54–65. [Google Scholar] [CrossRef]
  3. Ketcheson, D.E. A Study of the Cost of Strip Cutting Black Spruce Stands in Northern Ontario; Canadian Forest Service, Great Lakes Forest Research Centre: Sault Ste. Marie, ON, Canada, 1979; 29p. [Google Scholar]
  4. Cubbage, F.W. Economies of Forest Tract Size in Southern Pine Harvesting; United States Department of Agriculture: New Orleans, LA, USA, 1982; 27p. [Google Scholar]
  5. Greene, W.D.; Harris, T.G.; DeForest, C.E.; Wang, J. Harvesting Cost Implications of Changes in the Size of Timber Sales in Georgia. South. J. Appl. For. 1997, 21, 193–198. [Google Scholar] [CrossRef]
  6. Favreau, J.; Gingras, J.-F. Facteurs Susceptibles D’influencer le Coût de Récolte par Coupe Mosaïque; FERIC: Pointe-Claire, QC, Canada, 2000; 5p. [Google Scholar]
  7. Eliasson, L.; Grönlud, Ö.; Lundström, H.; Sonesson, J. Harvester and Forwarder productivity and net revenues in patch cutting. Int. J. For. Eng. 2021, 32, 3–10. [Google Scholar] [CrossRef]
  8. Sundberg, U.; Silversides, C.R. Operational Efficiency in Forestry: Vol. 1: Analysis; Springer: Dordrecht, The Netherlands, 1988; 221p. [Google Scholar]
  9. Öhman, K.; Eriksson, L. Aggregating harvest activities in long term forest planning by minimizing harvest area perimeters. Silva Fenn. 2010, 44, 77–89. [Google Scholar] [CrossRef]
  10. Smaltschinski, T.; Seeling, U.; Becker, G. Clustering forest harvest stands on spatial networks for optimised harvest scheduling. Ann. For. Sci. 2012, 69, 651–657. [Google Scholar] [CrossRef]
  11. Heinonen, T.; Mäkinen, A.; Rasinmäki, J.; Pukkala, T. Aggregating microsegments into harvest blocks by using spatial optimization and proximity objectives. Can. J. For. Res. 2018, 48, 1184–1193. [Google Scholar] [CrossRef]
  12. Mobtaker, A.; Montecinos, J.; Ouhimmou, M.; Rönnqvist, M.; Paquet, M. Minimizing spatial dispersion of forest harvest areas using spectral clustering and set covering modelling. Can. J. For. Res. 2018, 48, 1563–1576. [Google Scholar] [CrossRef]
  13. Naderializadeh, N.; Crowe, K.; Rouhafza, M. Solving the integrated forest harvest scheduling model using metaheuristic algorithms. Oper. Res. 2022, 22, 2437–2463. [Google Scholar] [CrossRef]
  14. Fortin, Y. Procédés de Récolte des Bois Utilisés Dans les Forêts du Domaine de l’État—Rapport Annuel 2004–2005; Gouvernement du Québec: Quebec City, QC, Canada, 2006; 40p. [Google Scholar]
  15. Audy, J.-F.; Pinotti Moreira, M.; Westlund, K.; D’Amours, S.; LeBel, L.; Rönnqvist, M. Alternative Logistics Concepts Fitting Different Wood Supply Situations and Markets; CIRRELT Report; CIRRELT: Montréal, QC, Canada, 2012; 347p. [Google Scholar]
  16. Decoster, A. The Boom in Forest Owners—A Bust for Forestry? J. For. 1998, 96, 25–28. [Google Scholar] [CrossRef]
  17. Hatcher, J.; Straka, T.; Greene, J. The Size of Forest Holding/Parcelization Problem in Forestry: A Literature Review. Resources 2013, 2, 39–57. [Google Scholar] [CrossRef]
  18. Conrad, J.L. Adapting to a Changing Landscape: How Wisconsin Loggers Persist in an Era of Parcelization. For. Prod. J. 2014, 64, 273–280. [Google Scholar] [CrossRef]
  19. Cubbage, F.W.; Granskog, J.E. Harvesting systems and costs for southern pine in the 1980s. For. Prod. J. 1982, 32, 37–43. [Google Scholar]
  20. DeLasaux, M.J.; Hartsough, B.R.; Spinelli, R.; Magagnotti, N. Small Parcel Fuel Reduction with a Low-Investment, High-Mobility Operation. West. J. Appl. For. 2009, 24, 205–213. [Google Scholar] [CrossRef]
  21. Abbas, D.; Di Fulvio, F.; Marchi, E. A Proposal for an Integrated Methodological and Scientific Approach to Cost Used Forestry Machines. Crojfe J. For. Eng. 2021, 42, 63–75. [Google Scholar] [CrossRef]
  22. Moldenhauer, M.C.; Bolding, C. Parcelization of South Carolina’s private forestland: Loggers’ reactions to a growing threat. For. Prod. J. 2009, 59, 37–43. [Google Scholar]
  23. Han-Sup, H.; Lee, H.; Johnson, L. Economic feasibility of an integrated harvesting system for small-diameter trees in southwest Idaho. For. Prod. J. 2004, 54, 21–27. [Google Scholar]
  24. Rawlings, C.; Seeley, C.; Thomas, C.E.; Morrison, D.; Han, H.-S.; Levi Cheff Atkins, D.; Graham, D.; Windell, K. A Study of How to Decrease the Costs of Collecting, Processing and Transporting Slash; Montana Community Development Corporation: Missoula, MT, USA, 2005; 21p. [Google Scholar]
  25. Bisson, J.A.; Han, S.-K.; Han, H.-S. Evaluating the System Logistics of a Centralized Biomass Recovery Operation in Northern California. For. Prod. J. 2016, 66, 88–96. [Google Scholar] [CrossRef]
  26. Sessions, J.; Wimer, J.; Costales, F.; Wing, M.G. Engineering Considerations in Road Assessment for Biomass Operations in Steep Terrain. West. J. Appl. For. 2010, 25, 144–153. [Google Scholar] [CrossRef]
  27. Daugherty, B.; Sessions, J.; Zamora-Cristales, R.; Wing, M.G. Improving large trailer access for biomass recovery in steep terrain. For. Sci. 2018, 64, 429–441. [Google Scholar] [CrossRef]
  28. Zamora-Cristales, R.; Sessions, J.; Boston, K.D.; Murphy, G. Economic Optimization of Forest Biomass Processing and Transport in the Pacific Northwest USA. For. Sci. 2015, 61, 220–234. [Google Scholar] [CrossRef]
  29. Han, H.-S.; Halbrook, J.; Pan, F.; Salazar, L. Economic evaluation of a roll-off trucking system removing forest biomass resulting from shaded fuelbreak treatments. Biomass Bioenergy 2010, 34, 1006–1016. [Google Scholar] [CrossRef]
  30. Harrill, H.; Han, H.-S. Application of hook-lift trucks in centralized logging slash grinding operations. Biofuels 2010, 1, 399–408. [Google Scholar] [CrossRef]
  31. Hall, P. Two-Staging to Super-Skids with Bell Off-Higway Trucks; Liro: Rotorua, New Zealand, 1998; Volume 23, pp. 1–8. [Google Scholar]
  32. Rendell, T. A Case Study of Two-Stage Operations in the Kenderdine Forest and Survey Assessment of Current Two-Stage Operations in Industry; University of Canterburry: Christchurch, New Zealand, 2019; 22p. [Google Scholar]
  33. Sirén, M.; Aaltio, H. Productivity and Costs of Thinning Harvesters and Harvester-Forwarders. Int. J. For. Eng. 2003, 14, 39–48. [Google Scholar] [CrossRef]
  34. Affenzeller, G.; Stampfer, K. Comparison of integrated with conventional harvester-forwarder-concepts in thinning operations. In Proceedings of the International Mountain Logging and 13th Pacific Northwest Skyline Symposium, Corvallis, OR, USA, 1–6 April 2007; Sessions, J., Havill, Y., Eds.; International Union of Forest Research Organizations: Corvallis, OR, USA, 2007; pp. 32–38. [Google Scholar]
  35. Talbot, B.; Nordfjell, T.; Suadicani, K. Assessing the Utility of Two Integrated Harvester-Forwarder Machine Concepts Through Stand-Level Simulation. Int. J. For. Eng. 2003, 14, 31–43. [Google Scholar] [CrossRef]
  36. Asikainen, A. Integration of Work Tasks and Supply Chains in Wood Harvesting—Cost Savings or Complex Solutions? Int. J. For. Eng. 2004, 15, 11–17. [Google Scholar] [CrossRef]
  37. Väätäinen, K.; Liiri, H.; Röser, D. Cost-Competitiveness of Harwarders in CTL-Logging Conditions in Finland—A Discrete-Event Simulation Study at the Contractor Level. In Proceedings of the International Precision Forestry Symposium, Stellenbosch, South Africa, 5–10 March 2006; pp. 451–463. [Google Scholar]
  38. Wester, F.; Eliasson, L. Productivity in Final Felling and Thinning for a Combined Harvester-Forwarder (Harwarder). Int. J. For. Eng. 2003, 14, 45–51. [Google Scholar] [CrossRef]
  39. Jonsson, R.; Rönnqvist, M.; Flisberg, P.; Jönsson, P.; Lindroos, O. Comparison of modelling approaches for evaluation of machine fleets in central Sweden forest operations. Int. J. For. Eng. 2023, 34, 42–53. [Google Scholar]
  40. MFFP. Ressources et Industries Forestières du Québec: Portrait Statistique (Forest Resources and Industry of Quebec: Statistics), 2017 ed.; MFFP: Quebec City, QC, Canada, 2018. [Google Scholar]
  41. Gouvernement du Québec. Budget 2016–2017: Compétitivité de L’industrie Forestière au Québec; Gouvernement du Québec: Quebec City, QC, Canada, 2016.
  42. Groupe DDM. Enquête sur les Couts D’opération Forestière dans les Forêts du Domaine de L’État Ainsi que sur les Couts et Revenus de L’industrie du Sciage du Québec 2019; Report; Groupe DDM: Quebec City, QC, Canada, 2020; 23p. [Google Scholar]
  43. Diniz, F.; Sessions, J. ESPDS: A Support Tool to Assist Forest Equipment Purchase Decisions. Croat. J. For. Eng. 2022, 43, 99–108. [Google Scholar] [CrossRef]
  44. Jonsson, R.; Rönnqvist, M.; Flisberg, P.; Jönsson, P.; Lindroos, O. Country-Wide analysis of the potential use of harwarders for final felling in Sweden. Scand. J. For. Res. 2023, 38, 105–120. [Google Scholar] [CrossRef]
  45. Yoshimura, T.; Suzuki, Y.; Sato, N. Assessing the productivity of Forest Harvesting Systems Using a Combination of Forestry Machines in Steep Terrain. Forests 2023, 14, 1430–1446. [Google Scholar] [CrossRef]
  46. Kellogg, L.; Bettinger, P.; Robe, S.; Steffert, A. Mechanized Harvesting: A Compendium of Research; Forest Research Laboratory, College of Forestry, Oregon State University: Corvallis, OR, USA, 1992; 401p. [Google Scholar]
  47. Dempster, P.; Gallo, N.; Hartsough, B.; Jenkins, B.; Tittmann, P. Sponsorship-Equipment Review Agreement Number 8CA05704—Final Report to State of California Department of Forestry and Fire Protection; University of California: Davis, CA, USA, 2008. [Google Scholar]
  48. Ringdahl, O.; Hellström, T.; Lindroos, O. Potentials of possible machine systems for directly loading logs in cut-to-length harvesting. Can. J. For. Res. 2012, 42, 970–985. [Google Scholar] [CrossRef]
  49. Lindroos, O. Evaluation of Technical and Organizational Approaches for Directly Loading Logs in Mechanized Cut-to-Length Harvesting. For. Sci. 2012, 58, 326–341. [Google Scholar] [CrossRef]
  50. Lundbäck, M.; Hägstrom, C.; Fjeld, D.; Lindroos, O.; Nordjfell, T. The Economic Potential of semi-automated tele-extraction of roundwood in Sweden. Int. J. For. Eng. 2022, 33, 271–288. [Google Scholar] [CrossRef]
  51. Wästerlund, I.; Hassan, A.E. Forest Harvesting Systems Friendly to the Environment. In Proceedings of the ASME Winter Annual Meeting, Atlanta, GA, USA, 13–16 December 1994. 30p. [Google Scholar]
  52. Dimou, V.; Malivitsi, Z. Strategic decision model for the evaluation of timber harvesting systems. Int. J. For. Eng. 2015, 26, 146–157. [Google Scholar] [CrossRef]
  53. Kühmaier, M.; Stampfer, K. Development of a Multi-Attribute Spatial Decision Support System in Selecting Timber Harvesting Systems. Croat. J. For. Eng. 2010, 31, 75–88. [Google Scholar]
  54. Martínez-Mesa, J.; González-Chica, D.A.; Duquia, R.P.; Bonamigo, R.R.; Bastos, J.L. Sampling: How to select participants in my research study? An. Bras. Dermatol. 2016, 91, 326–330. [Google Scholar] [CrossRef] [PubMed]
  55. Matthews, D.M. (Ed.) Cost Control in the Logging Industry; McGraw-Hill: New York, NY, USA, 1942; 374p. [Google Scholar]
  56. Thompson, M.A. Considering overhead costs in road and landing spacing models. J. For. Eng. 1992, 3, 13–19. [Google Scholar] [CrossRef]
  57. Plamondon, J.A. Établir la Distance de Débardage Optimale en Fonction du Coût des Chemins; FERIC Report; FERIC: Pointe-Claire, QC, Canada, 1994; 8p. [Google Scholar]
  58. Macdonald, A.J. Harvesting Systems and Equipment in British Columbia; Ministry of Forests: Vancouver, BC, Canada, 1999; 56p. [Google Scholar]
  59. Ackerman, P.; Belbo, H.; Eliasson, L.; de Jong, A.; Lazdins, A.; Lyons, J. The COST model for calculation of forest operations costs. Int. J. For. Eng. 2014, 25, 75–81. [Google Scholar] [CrossRef]
  60. The World Bank. International Financial Statistics, Official Exchange Rate (LCU per US$, Period Average). 2020. Available online: http://data.worldbank.org/ (accessed on 29 May 2020).
  61. BLS. United States Department of Labour Statistics, Bureau of Labour Statistics. 2020. Available online: http://www.bls.gov/ (accessed on 29 May 2020).
  62. Di Fulvio, F.; Abbas, D.; Spinelli, R.; Acuna, M.; Ackerman, P.; Lindroos, O. Benchmarking technical and cost factors in forest felling and processing operations in different global regions during the period 2013–2014. Int. J. For. Eng. 2017, 28, 94–105. [Google Scholar] [CrossRef]
  63. Hansen, P.; Ombler, F. A new method for scoring additive multi-attribute value models using pairwise rankings of alternatives. J. Multi-Criteria Decis. Anal. 2008, 15, 87–107. [Google Scholar] [CrossRef]
  64. Legendre, P. Species associations: The Kendall coefficient of concordance revisited. J. Agric. Biol. Environ. Stat. 2005, 10, 226–245. [Google Scholar] [CrossRef]
  65. Lilleberg, R. Harvester-Forwarder for Logging in First-Thinning Stands; Metsäteho Report 28; Metsäteho Oy: Helsinki, Finland, 1997; 24p. [Google Scholar]
  66. Manner, J.; Jonsson, R.; Jönsson, P.; Björheden, R.; Lundström, H. Productivity and Logging Costs of the Harwarder Prototype Komatsu X19 and a Conventional CTL System; Skogforsk: Uppsala, Sweden, 2016; 40p. [Google Scholar]
  67. Kärhä, K. Whole-tree harvesting in young stands in Finland. For. Stud. Metsanduslikud Uurim. 2006, 45, 118–134. [Google Scholar]
  68. Laitila, J. Harvesting technology and the cost of fuel chips from early thinnings. Silva Fenn. 2008, 42, 267–283. [Google Scholar] [CrossRef]
  69. Bergkvist, I.; Hallonbord, U.; Nordén, B. Valmet 801 Combi with Rotating Load Space in Thinnings and Final Cuttings; Skogforsk: Uppsala, Sweden, 2003. [Google Scholar]
  70. Väätäinen, K.; Asikainen, A.; Sikanen, L.; Ala-Fossi, A. The cost effect of forest machine relocations on logging costs in Finland. For. Stud. Metsanduslikud Uurim. 2006, 45, 135–141. [Google Scholar]
  71. Kärhä, K.; Poikela, A.; Palander, T. Productivity and Costs of Harwarder Systems in Industrial Roundwood Thinning. Croat. J. For. Eng. 2018, 39, 23–33. [Google Scholar]
  72. Plamondon, J.A. Use of Convertible Machines “Combo” and “Dual” in Cut-to-Lenght Harvest; FPInnovations: Pointe-Claire, QC, Canada, 2009. [Google Scholar]
  73. Di Fulvio, F.; Bergström, D. Analyses of a single-machine system for harvesting pulpwood and/or energy-wood in early thinnings. Int. J. For. Eng. 2013, 24, 2–15. [Google Scholar] [CrossRef]
  74. Fisher, J. Shovel logging: Cost-effective systems gains ground. In Proceedings of the International Mountain Logging 10th Pacific Northwest Skyline Symposium, Corvalis, OR, Cananda, 28 March–1 April 1999; pp. 61–67. [Google Scholar]
  75. Sessions, J.; Boston, K. Optimization of Road Spacing for Log Length Shovel Logging on Gentle Terrain. Int. J. For. Eng. 2006, 17, 67–75. [Google Scholar] [CrossRef]
  76. Floch, R.F. Shovel Logging and Soil Compaction: A Case Study; Oregon State University: Benton County, OR, USA, 1988; 75p. [Google Scholar]
  77. McNeel, J.F.; Andersson, B. Shovel Logging with Different Retrieval Patterns in Second-Growth Timber. In Proceedings of the COFE 16th Annual Meeting, Savannah, GA, USA, 8–11 August 1993; pp. 109–122. [Google Scholar]
  78. Egan, A.; Hicks, R.; Waldron, K.; Skousen, J. Effects of Shovel Logging and Rubber-tired Skidding on Surface Soil Attributes in a Selectively Harvested Central Hardwood Stand. Int. J. For. Eng. 2002, 13, 27–32. [Google Scholar] [CrossRef]
  79. Eriksson, P. Påhängsvagn för Skotare (Semi-Trailers for Forwarders); SkogForsk Arbetsrapport Nr 404:14; SkogForsk: Uppsala, Sweden, 1998; 21p. [Google Scholar]
  80. Lindroos, O.; Wästerlund, I. Utvärdering av Skotning med Påhängsvagn (Evaluation of Forwarding with a Trailer); Swedish University of Agricultural Sciences: Umeå, Sweden, 2014; 48p. [Google Scholar]
  81. Miller, C.F.; Greene, W.D. Delimbing southern pines: Gate versus the CTR Model 314 De-Limber. For. Prod. J. 1992, 42, 11–16. [Google Scholar]
  82. Cutting Systems Inc. FDT-6000-COMBO Flail/Pull-Thru Delimber. Union Grove (NC). Available online: http://www.cuttingsys.com/ (accessed on 12 February 2020).
  83. Klepac, J.; Rummer, B. Productivity and Cost Comparison, of Two Different-Sized Skidders. In Proceedings of the 2000 ASAE Annual International Meeting, Milwaukee, WI, USA, 9–12 July 2000; pp. 1–10. [Google Scholar]
  84. Cass, R.D.; Baker, S.A.; Greene, W.D. Cost and Productivity Impacts of Product Sorting on Conventional Ground-Based Timber Harvesting Operations. For. Prod. J. 2009, 59, 108–114. [Google Scholar] [CrossRef]
  85. Hill, S.; Evanson, T. The Trinder Static Delimber in a Ground-Bsed Clearfell Operation; Report 20; Liro: Rotorua, New Zealand, 1995; 6p. [Google Scholar]
  86. Visser, R.; Stampfer, K. Tree-Length System Evaluation of Second Thinning in a Loblolly Pine Plantation. South. J. Appl. For. 2003, 27, 77–82. [Google Scholar] [CrossRef]
  87. Fogdestam, N.; Lundström, H. Studier av Offset Crane Concept, OCC hos Kjellbergs Logistik & Teknik i Hällefors (Studies of Offset Crane Concept, OCC at Kjellbergs Logistik & Teknik in Hällefors); Skogforsk: Uppsala, Sweden, 2011. [Google Scholar]
  88. Jakobsson, M.; Nilsson, N. Faktorer som Påverkar Valet av Kranavställning vid Rundvirkestransporter hos Holmen Skog AB (Factors Affecting the Removal of the Self-Loader Crane during Roundwood-Transports for Holmen Skog AB); Sveriges Lantbruksuniversitet: Uppsala, Sweden, 2015. [Google Scholar]
  89. Erlandsson, E. Framgångsfaktorer för Rundvirkesåkerier i Mellansverige (Success Factors for Roundwood Truck Transport Companies in Mid-Sweden); Sveriges Lantbruksuniversitet: Uppsala, Sweden, 2008. [Google Scholar]
  90. Andersson, E. Kartläggning av Typfordon för Rundvirkestransport vid Holmen Skogs Regioner (A Mapping of Timber Truck Types on Holmen Skog’s Regions); Sveriges Lantbruksuniversitet: Uppsala, Sweden, 2014; 44p. [Google Scholar]
  91. Spinelli, R.; Hartsough, B.R. Harvesting SRF popular for pulpwood: Experience in the Pacific Northwest. Biomass Bioenergy 2006, 30, 439–445. [Google Scholar] [CrossRef]
  92. Mousavi, R.; Naghdi, R. Time consumption and productivity analysis of timber trucking using two kinds of trucks in northern Iran. J. For. Sci. 2013, 59, 211–221. [Google Scholar] [CrossRef]
  93. Kellogg, L. Small Landing Operation: A Study with A Mobile Hauler & Hydraulic Knuckle Boom Loader; Liro: Rotorua, New Zealand, 1987. [Google Scholar]
  94. Visser, R.; Spinelli, R.; Magagnotti, N. Landing Characteristics for Harvesting Operations in New Zealand. Int. J. For. Eng. 2011, 22, 23–27. [Google Scholar] [CrossRef]
  95. Lanford, B.L.; Stokes, B.J. Comparison of Two Thinning Systems. Part 2. Productivity and Costs. For. Prod. J. 1996, 46, 47–53. [Google Scholar]
  96. Akay, A.E.; Oǧuz, H.; Karas, I.R.; Aruga, K. Using LiDAR technology in forestry activities. Environ. Monit. Assess. 2009, 151, 117–125. [Google Scholar] [CrossRef] [PubMed]
  97. Adebayo, A.B.; Han, H.-S.; Johnson, L.R. Productivity and cost of cut-to-length and whole tree harvesting in a mixed-conifer stand. For. Prod. J. 2007, 57, 59–69. [Google Scholar]
  98. Baek, K. Productivity and Cost of a Cut-to-Length Commercial Thinning Operation in a Northern California Redwood Forest. Master’s Thesis, Humbolt State University, Arcata, CA, USA, 2018; 41p. [Google Scholar]
  99. Scheve, G. Stroker Loading a Log Truck. Youtube Video. 2012. Available online: https://youtu.be/rZrPwD_1Xsw (accessed on 15 June 2020).
  100. Shukla, M. Pierce GP—Grapple Processor HD. Youtube Video. 2014. Available online: https://youtu.be/2eA1BKEwQCU (accessed on 15 June 2020).
  101. Pierce Pacific. Pierce Grapple Processor. 8. Available online: https://piercepacific.com/docs/catalogs/pdf/forestry-gp/pierce-forestry-gp.pdf (accessed on 31 March 2020).
  102. Stokes, B.J.; Schilling, A. Improved harvesting systems for wet sites. For. Ecol. Manag. 1997, 90, 155–160. [Google Scholar] [CrossRef]
  103. McKerchar, A.; Twaddle, A. Central Log Yards—An Option to Consider; LIRA Technical Release: Rotorua, New Zealand, 1987; 8p. [Google Scholar]
  104. Atkins, D.; Rummer, B.; Dodson, B.; Thomas, C.E.; Horcher, A.; Messerlie, E.; Rawlings, C.; Haston, D. A Report on Conceptual Advances in Roll On/Off Technology in Forestry; USDA Forest Service: Washington, DC, USA, 2007. [Google Scholar]
  105. Kash, A.; Dodson, E. Production of a Roll-off Equipped Harwarder for Forest Biomass Utilization. Int. J. For. Eng. 2010, 21, 48–56. [Google Scholar] [CrossRef]
  106. Granskog, J.E. Harvesting Costs for Mechanized Thinning Systems in Slash Pine Plantations; Res paper. SO-141; U.S. Department of Agriculture, Forest Service: New Orleans, LA, USA, 1978. [Google Scholar]
  107. Akay, A.E.; Erdas, O.; Sessions, J. Determining Productivity of Mechanized Harvesting Machines. J. Appl. Sci. 2004, 4, 100–105. [Google Scholar] [CrossRef]
  108. Jernigan, P.; Gallagher, T.; Aulakh, J.; Tufts, R.; McDonald, T. Implementing residue chippers on harvesting operations in the southeastern US for biomass recovery. Int. J. For. Eng. 2013, 24, 129–136. [Google Scholar] [CrossRef]
  109. Brinker, R.W.; Kinard, J.; Rummer, B.; Lanford, B. Machine Rates for Selected Forest Harvesting Machines; Circular 296; Alabama Agriculture Experiment Station: Auburn, AL, USA, 2002; 32p. [Google Scholar]
  110. Väätäinen, K.; Laitila, J.; Anttila, P.; Kilpeläinen, A.; Asikainen, A. The influence of gross vehicle weight (GVW) and transport distance on timber trucking performance indicators—Discrete event simulation case study in Central Finland. Int. J. For. Eng. 2020, 31, 156–170. [Google Scholar] [CrossRef]
Forests 15 01046 g001
Figure 1. Flowchart of the MCDA for the selection of harvest systems.
Figure 1. Flowchart of the MCDA for the selection of harvest systems.
Forests 15 01046 g001
Forests 15 01046 g002
Figure 2. Komatsu X19 prototype—Combi (Photo: Skogforsk).
Figure 2. Komatsu X19 prototype—Combi (Photo: Skogforsk).
Forests 15 01046 g002
Forests 15 01046 g003
Figure 3. Forwarder equipped with an auxiliary trailer (Photo: Ola Lindroos).
Figure 3. Forwarder equipped with an auxiliary trailer (Photo: Ola Lindroos).
Forests 15 01046 g003
Forests 15 01046 g004
Figure 4. Delimbing gate [82] (Cutting Systems, Inc., Union Grove, WI, USA).
Figure 4. Delimbing gate [82] (Cutting Systems, Inc., Union Grove, WI, USA).
Forests 15 01046 g004
Forests 15 01046 g005
Figure 5. Cutting Systems FDT 6000 Combo pull-through/flail delimber [82] (Cutting Systems, Inc., Union Grove, WI, USA).
Figure 5. Cutting Systems FDT 6000 Combo pull-through/flail delimber [82] (Cutting Systems, Inc., Union Grove, WI, USA).
Forests 15 01046 g005
Forests 15 01046 g006
Figure 6. Removable crane for self-loading timber trucks (Photo: L. Eliasson, Skogforsk).
Figure 6. Removable crane for self-loading timber trucks (Photo: L. Eliasson, Skogforsk).
Forests 15 01046 g006
Forests 15 01046 g007
Figure 7. Pierce grapple processor [101] (Pierce Pacific Manufacturing, Inc. Portland, OR, USA).
Figure 7. Pierce grapple processor [101] (Pierce Pacific Manufacturing, Inc. Portland, OR, USA).
Forests 15 01046 g007
Forests 15 01046 g008
Figure 8. Chart of possible harvest systems to be used in forest operations in the boreal forests of eastern Canada.
Figure 8. Chart of possible harvest systems to be used in forest operations in the boreal forests of eastern Canada.
Forests 15 01046 g008
Forests 15 01046 g009
Figure 9. PAPRIKA criteria weights of expert’s survey.
Figure 9. PAPRIKA criteria weights of expert’s survey.
Forests 15 01046 g009
Table 1. COST model input parameters kept constant for all machines based on FPI Procalc default values.
Table 1. COST model input parameters kept constant for all machines based on FPI Procalc default values.
ParameterUnitValue
Salvage value% of purchase price10
Expected economic lifeYears5
Interest rate%10
Insurance% of the purchase price2.2
Fuel costUSD/l1.00
Operator wage with benefitsUSD/SMH30
Scheduled days per yeardays/y168.75
Scheduled shifts per dayshifts/day2
Scheduled hours per shiftSMH/shift12
Profit margin% gross cost and overhead10
Table 2. Performance indicator scale levels.
Table 2. Performance indicator scale levels.
CriteriaProductivityMobilityRoad InvestmentMachinery Operating Cost
IndicatorSystem ProductivityNb of MachinesMaximum Extraction DistanceUnit Cost
Unitm3/PMHNbm$/m3
Levels1<10≥5<225≥30
210–203–5225–50028–30
320–302–3500–100026–28
4≥30<2≥1000<26
Table 3. Reviewed machine types and work method for each operation phase.
Table 3. Reviewed machine types and work method for each operation phase.
FellingExtractionProcessingLoadingTransport
CombiShovel loggerGate delimberRemovable crane truckSecondary extraction
DualForwarder trailerPull-through delimberFront-end loaderSuper forwarder
Feller director Flail delimberExtraction machineSet-out trailers
Processing machineHook-lift
Table 4. Highlights from the interviews with forest operations managers.
Table 4. Highlights from the interviews with forest operations managers.
Expert (#)123456
Context of
operations		Approximate harvested volume (m3/y)500,0001,000,000400,000200,000200,0001,000,000
Harvest systemCTLCTLCTLCTLWT and CTLWT and CTL
Productivity of contractorsHighHighHighLowHighHigh
Work scheduleDay and nightDay and nightDay and nightDay onlyDay onlyDay and night
Main transport typeMegaloadMegaloadStandardStandard/self-loaderStandardStandard/Megaload
FragmentationPerceived proportion of annual volume10%10%50%100%50%33%
SolutionsPrioritized during winterxx xx
Longer extraction distancesxxxxxx
Prioritization of smaller teamsxxxxxx
Table 5. Cost and productivity estimations for individual machines.
Table 5. Cost and productivity estimations for individual machines.
ParameterPurchase PriceFuelOil and LubricantsMaintenance and RepairUtilizationHourly CostProductivityUnit Cost
MachineReference
(Purchase Price and Fuel Consumption)		$l/PMH% Fuel Cost% Purchase Price% PMH/
SMH		$/
PMH		m3/PMH$/m3
HarvesterFPInnovations650,000301511575198.7012.815.52
Feller buncherFPInnovations550,000451510080185.1930.86.01
Feller directorEstimated equal to feller buncher550,000451510080185.1923.37.95
Combi[66]750,000201511580192.857.725.05
Dual[73]550,000201511575168.097.721.83
Combi hook lift[66,104]800,000201511580201.277.726.14
Dual hook lift[73,104]600,000201511575177.077.723.00
ForwarderFPInnovations450,000301010080148.74197.83
Clambunk skidderTanguay, personal communications 22 July 2019900,000301510080219.9425.68.59
Grapple skidderFPInnovations350,000401010080145.12169.07
Shovel logger[109]450,000401010080160.8416.59.75
Trailer for forwarder[80]35,0002101005011.2814.20.79
Hook-lift forwarderFPInnovations; [104]500,000301010080156.65198.24
Gate delimber for skidder[4]400001065800.5413.80.04
Stroke boom delimberFPInnovations430,000251010080139.53168.72
ProcessorFPInnovations600,000301511080178.2017.410.24
Trailer loader/Pull-through delimberPro Pac personal communications300,000351010080131.0612.210.74
LoaderFPInnovations410,000351010075155.5478.21.99
Hook-lift truck[30]180,0002601009091.8113.56.80
Self-loading crane[110]82,000351010021248.9278.23.18
Table 6. Performance rating of base-case systems’ adaptations for fragmented harvest operations and dominated alternatives (1 to 4, darker gray is for higher score).
Table 6. Performance rating of base-case systems’ adaptations for fragmented harvest operations and dominated alternatives (1 to 4, darker gray is for higher score).
CriteriaProductivityMobilityRoad InvestmentsMachinery Operating Costs
CTL Base case2234
CTL Harwarder felling/extraction1323
CTL Forwarder w\trailer extraction2343
CTL Self-loader removable crane transport2333
CTL TL harvester, shovel logger extraction/loading2313
CTL Forwarder extraction/loading2324
CTL Forwarder loading set-out trailers2323
CTL Articulated self-loader secondary extraction2341
CTL Hook-lift forwarder and trucks secondary extraction2341
CTL Harwarder felling/extraction, self-loader removable crane transport1421
CTL Harwarder felling/extraction/loading1413
CTL Harwarder felling/extraction/loading set-out trailers1413
CTL Harwarder felling/extraction, articulated self-loader secondary extraction1431
CTL Hook-lift harwarder felling/extraction, hook-lift trucks secondary extraction1431
WT Base case4124
WT Clambunk extraction3134
WT 2 forwarders extraction4134
WT Gate delimber skidders, 2 loaders/pull-through delimber processing/loading3113
WT Self-loader removable crane transport4123
WT 2 processors processing/loading3123
WT Articulated self-loader secondary extraction, centralized processing4231
WT Clambunk extraction, 2 processors processing/loading3233
WT 2 forwarders extraction, 2 processors processing/loading 3133
WT Clambunk extraction, articulated self-loader secondary extraction, centralized processing3241
WT 2 forwarders extraction, articulated self-loader secondary extraction, centralized processing4241
WT Shovel logger extraction/processing with pull-through delimber/loading1312
WT Feller director felling/extraction/processing with pull-through delimber/loading1411
WT Feller director felling/extraction, articulated self-loader secondary extraction, centralized processing1321
Table 7. PAPRIKA alternatives rank medians and frequencies.
Table 7. PAPRIKA alternatives rank medians and frequencies.
#Candidate Harvest SystemsMedian12345678910111213141516171819202122232425262728
1CTL Forwarder w\trailer extraction1.532 1
2WT 2 forwarders extraction2.5121 1 1
3CTL Self-loader removable crane transport3.0 13 1 1
4CTL Base case4.02 1 1 1 1
5WT Clambunk extraction, 2 processors processing/loading5.0 2211
6CTL Forwarder extraction/loading6.0 21 1 11
7WT Clambunk extraction6.5 11 12 1
8WT 2 forwarders extraction, 2 processors processing/loading8.5 1 21 11
8WT Base case8.5 1 1 111 1
10CTL Forwarder loading set-out trailers9.0 13 2
11WT 2 forwarders extraction, articulated self-loader secondary extraction, centralized processing12.5 11 11 1 1
12WT Clambunk extraction, articulated self-loader secondary extraction, centralized processing15.0 2 2 1 1
13CTL Articulated self-loader secondary extraction15.5 1 1 1 2 1
13CTL Hook-lift forwarder and trucks secondary extraction15.5 1 1 1 2 1
15WT Articulated self-loader secondary extraction, centralized processing16.0 2 1 11 1
16CTL Harwarder felling/extraction16.5 1 1 12 1
17CTL Harwarder felling/extraction, articulated self-loader secondary extraction17.0 1 12 1 1
17CTL Hook-lift harwarder felling/extraction, hook-lift trucks secondary extraction17.0 1 12 1 1
17WT Self-loader removable crane transport17.0 1 11 2 1
20CTL TL harvester, shovel logger extraction/loading18.5 11 1 11 1
21CTL Harwarder felling/extraction/loading20.5 1 1 121
21CTL Harwarder felling/extraction/loading set-out trailers20.5 1 1 121
21WT 2 processors processing/loading20.5 1 1 111 1
24CTL Harwarder felling/extraction, self-loader removable crane transport24.0 1 3 2
25WT Gate delimber skidders, 2 loaders/pull-through delimber processing/loading25.0 1 1 21 1
26WT Shovel logger extraction/processing with pull-through delimber/loading26.0 111 3
27WT Feller director felling/extraction, articulated self-loader secondary extraction27.0 1 11 3
27WT Feller director felling/extraction/processing with pull-through delimber/loading27.0 132
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Painchaud, L.; LeBel, L. A Multi-Criterion Evaluation Process for Determining Cost-Effective Harvesting Systems in Fragmented Boreal Forests. Forests 2024, 15, 1046. https://doi.org/10.3390/f15061046

AMA Style

Painchaud L, LeBel L. A Multi-Criterion Evaluation Process for Determining Cost-Effective Harvesting Systems in Fragmented Boreal Forests. Forests. 2024; 15(6):1046. https://doi.org/10.3390/f15061046

Chicago/Turabian Style

Painchaud, Léo, and Luc LeBel. 2024. "A Multi-Criterion Evaluation Process for Determining Cost-Effective Harvesting Systems in Fragmented Boreal Forests" Forests 15, no. 6: 1046. https://doi.org/10.3390/f15061046

APA Style

Painchaud, L., & LeBel, L. (2024). A Multi-Criterion Evaluation Process for Determining Cost-Effective Harvesting Systems in Fragmented Boreal Forests. Forests, 15(6), 1046. https://doi.org/10.3390/f15061046

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop