Printer Friendly

Productivity and costs of an integrated mechanical forest fuel reduction operation in southwest Oregon.

Abstract

Mechanical forest fuel reduction treatments that harvest and extract small, non-merchantable trees are often integrated into commercial thinning operations. Harvesting system feasibility and/or costs from such operations has been sparsely reported in the literature. To broaden the knowledge of mechanical approaches of harvesting and utilizing small trees, this study assessed the productivity and cost from an integrated forest harvesting/mechanical forest fuel reduction operation in southwest Oregon. The study was conducted in a fuel reduction thinning of a 20-acre mixed conifer stand on gentle terrain. A tracked swing-boom feller-buncher, two rubber-tired grapple skidders, a swing-boom grapple processor, an in-woods chipper, and a tub grinder were used to fell, extract, and process non-merchantable stems and limbs and tops from felled merchantable trees into fuel (energy-wood) chips. Thinned merchantable trees were also extracted and processed into log lengths. Results indicate that harvesting and processing non-merchantable trees increased total costs by $1,193.43 per acre. From a biomass harvesting perspective, removing only the non-merchantable portion of the stand would have resulted in a net cost of $968.96 per acre. Thinning merchantable trees added value to the operation, subsidized costs, and decreased the net loss by $872.00 per acre, resulting in a net cost of $96.96 per acre.

**********

Forested landscapes, specifically in the Western United States, have changed greatly since European settlement. Many presettlement forests were characterized by a large mature overstory with a sparse and mostly open understory (Weaver 1959). It has been reported that these conditions occurred largely due to frequent burning by indigenous people (Kimmerer and Lake 2001). In the years since European settlement, human impact on forest structure and composition is evident (Harvey et al. 1995, Brose et al. 2001). Anthropogenic influence, such as fire exclusion, has caused public lands in the United States to become severely overstocked with small stagnant trees. The U.S. Forest Service (2005) reported that at least 28 of a total 236 million acres of forestland in the 15 western states could benefit from fuel reduction treatments. O'Laughlin and Cook (2003) indicated that National Forests in the United States are on average 50 percent denser than forests in any other ownership. Overall, forests in the Pacific Northwest region contain more live-tree biomass than any other U.S. region (Woodall et al. 2006).

Small trees, tightly spaced in the understory of mature forests, inherently increase wildfire hazards. The small trees provide a ladder for surface fires to reach the overstory crown and can result in stand-replacement wildfires by promoting fire spread. Laverty and Williams (2000) report that fire suppression activities have caused public lands to over-accumulate shrubs and small trees which "reduce ecosystem diversity, health, resiliency, and fuel conditions for unnaturally intense fires." These effects are particularly important for stands with histories of frequent low intensity fire. The over-accumulation of small trees and understory vegetation has changed some fire regimes from low intensity to high severity allowing crown fires to occur in forest types not historically prone to such occurrences (Mutch et al. 1993, Brown et al. 2004).

Given the current forest health and wildfire hazard situation on many forests in the Western United States, research is needed to address methods of managing vegetation structure for the purpose of producing more fire resilient landscapes. There is ample opportunity and much interest in employing precommercial thinning that could alleviate the overstocking problem along with wildfire hazard. Such a proactive approach to fuel management may produce less monetary and environmental cost than fire suppression and stand replacement (Lynch 2004). With these observations, it is imperative that fuel reduction activities be investigated in attempts to protect the forest resource along with its associated assets, both market and non-market (Mason et al. 2006).

Commercial fuel reduction activities can be defined as operations with an end goal of changing forest fuel structure while extracting fiber in hopes of producing utilizable wood products that can be sold to finance fuel reduction treatments (Bolding 2006). Hollenstein et al. (2001) explained that mechanical harvesting differs from other methods of fuel reduction, especially prescribed fire, in that "removal is immediately effective, does not result in air pollution or escaping fires, and may be economically self-sustaining." They also report that reducing fuel loads through thinning could slow stand-replacement fire occurrences.

As the need and justification for vegetation removal through thinning of overstocked stands has grown, commercial systems have been used to harvest small-diameter, nonmerchantable trees. This approach to forest fuel reduction poses many challenges to the forest manager. The majority of mechanized logging equipment is designed to fell and extract trees large enough to produce revenue. These machines are typically not configured to productively fell, process, and extract small trees (Bolding 2002). In economic terms, mechanical thinning is problematic due to the fact that harvesting small stems is expensive and the resulting wood product has low value, producing high harvesting costs per unit or area (Watson et al. 1986, Bolding and Lanford 2005). Also, few cost and productivity estimates have been developed for mechanical fuel reduction/small wood harvesting systems (Bolding and Lanford 2001, Coulter et al. 2002). STHarvest (Hartsough et al. 2001, Fight et al. 2003) and the Fuel Reduction Cost Simulator (FRCS) (Fight et al. 2006) are cost models that have been used to predict relative costs for harvesting small trees. These models, however, are driven by production equations from older machine studies of conventional equipment and are not specifically geared toward treatments that often require harvesting small trees to diameters of 1 inch. Extrapolation beyond the range of reliable data can often lead to extreme variation in cost estimates. This observation highlights the need for additional research on systems that specifically target small, non-merchantable trees commonly extracted in fuel reduction applications.

Producing energy from woody biomass

Rising energy prices and volatile supplies of crude oil have forced policy makers in the United States to explore alternative sources of energy. One of the most commonly explored alternatives is that of energy generation from renewable sources. In most cases, small trees, limbs, and tops generated from harvesting activities or fuel reduction treatments have little monetary value and are, therefore, left on-site and not utilized (Bolding et al. 2003). Energy produced from biomass is usually from "dirty" chips, including bark and foliage, processed by in-woods chippers. Stokes (1988) reported that "the advantages of in-woods chipping systems include the ability to recover fiber from limbs, tops, and unmerchantable wood, high productivity, and advanced site preparation." In response to competitiveness, some forest products companies are interested in purchasing chips for bioenergy, whether dirty or clean, directly from suppliers. More industries, including forest products, are beginning to utilize the energy contained in woody biomass. Biomass is being used to fuel boilers and in some cases generate electrical power (Kutscha 1999).

The possibility of utilizing woody biomass for energy has great potential throughout the nation (Zerbe 2006). A report by Perlack et al. (2005) indicated that forestlands in the United States have the potential to provide approximately 370 million dry tons of biomass per year. Due to economic and technological barriers, however, the utilization of wood for energy has been slow to gain wide implementation (Bain and Overend 2002). With technology advancing daily, uses for wood fiber as an alternative energy source are expected to expand.

This study was initiated to investigate productivity and cost from an integrated mechanical forest fuel reduction operation on a 20-acre mixed conifer stand in southwest Oregon.

Methods

Study site, prescription, and harvesting equipment

This study was conducted in a fuel reduction thinning of a 20-acre mixed conifer stand on gentle terrain with average slopes of 12 percent (min 5%, max 17%) in southwest Oregon. Tree species consisted predominately of incense-cedar (Calocedrus decurrens), Douglas-fir (Pseudotsuga menziesii), and white fir (Abies concolor).

The stand consisted of approximately 694 trees per acre with a quadratic mean diameter of 6.4 inches. Detailed pre- and post-treatment stand density and biomass characteristics are shown in Table 1. Residual trees were thinned to a 20- by 20-ft spacing using a thinning from below approach. Merchantable leave trees were those greater than or equal to 5 inches in diameter at a height of 17 feet (approximately 7 in. in diameter at breast height [DBH]). Small trees 3 inches or greater but less than 7 inches DBH were considered non-merchantable within local merchantability standards; these trees were harvested and transported to landings to meet forest fuel reduction objectives. The resulting landing slash and extracted small, non-merchantable trees were then processed by an in-woods chipper and tub grinder into fuel or "dirty" chips. No trees less than 3 inches DBH were intentionally harvested. This constraint was imposed by the landowner and harvesting contractor for operational feasibility. The harvesting contractor had approximately 25 years of experience implementing ground-based harvesting treatments which included extensive experience with thinning prescriptions.

Harvesting equipment used during the study were as follows: (1)

Feller-buncher.--The TigerCat L830 with a 5400 series single-post felling saw was a tracked swing-to-tree excavator with 24-inch-wide single grouser tracks and a 280 horsepower (hp) Cummins diesel engine. The felling head had a maximum 22 inch DBH felling capacity.

Rubber-tired skidders.--Two rubber-tired grapple skidders were used during the study: a Caterpillar 518C (154 hp) and a John Deere 548E (127 hp).

Processor and loader.--Both the processor and the loader were Madill 2800 swing-boom excavators. Each machine had a Cummins 260 hp diesel engine and a 36 foot boom reach. The processing head was a Waratah HTH 624 Super with a 31 inch diameter cutting capacity and a maximum 25 inch delimbing diameter. Once trees were delivered to the landing, the processor performed the delimbing, topping, and bucking functions of the operation.

In-woods chipper and tub grinder.--The in-woods chipper was a Morbark 27RXL (650 hp) with an 83-inch rotating disc and three chipping knives. The chipper predominantly processed whole trees (bole, limbs, tops, and foliage) of non-merchantable size (> 3 and < 7 in. DBH). Large limbs and tops from felled merchantable stems were also processed.

The tub grinder was a Morbark 1200 series with a 630 hp Cummins diesel engine. The addition of the tub grinder was necessary in order to process material not effectively handled by the chipper, such as slash (small limbs and tops) produced by the processor. The combination of the chipper and tub grinder resulted in clean landing areas after operations were complete.

Experimental design and data collection

To access the individual machine and overall system productivity of the fuel reduction treatment, each function of the operation was studied as harvesting occurred. Functions included:

1. felling and bunching,

2. skidding,

3. processing,

4. loading,

5. chipping, and

6. tub-grinding.

Detailed time studies were used to determine feller-buncher and skidder productivity, whereas activity sampling was used to assess system productivity per unit volume and area for processing and loading. Activity sampling measures the proportion of a workday that people or machines spend performing a series of activities (Olsen and Kellogg 1983). This procedure builds information on interactions between machines and the percentages of time spent on each activity. Chipping and tub grinding were studied using shift-level information including processing times and van load weights provided by contractors. For all of the functions, the effects on system productivity of harvesting the non-merchantable portion of the stand were assessed. This reduction in productivity and projected increase in cost was attributed to the extra time required to harvest small trees and meet fuel reduction objectives. Developing productivity estimates for harvesting both the non-merchantable and merchantable portions provided information on the effect of tree size and merchantability class on overall operation profitability. Felling, bunching, and skidding were conducted without treating non-merchantable material separately from merchantable stems. For example, stems of all sizes were included in bunches and skidded to the landing as such. The study design, however, attempted to recover the time spent treating each portion (non-merchantable or merchantable) separately; therefore, merchantability class was determined for each tree felled and skidded.

To determine feller-buncher productivity, 10 hours of videotape were collected as the machine moved throughout the stand performing the silvicultural prescription. The intent was to determine the amount of time spent felling and bunching non-merchantable and merchantable sized trees separately. Tree size, whether non-merchantable (> 3 in. and < 7 in. DBH) or merchantable ([greater than or equal to] 7 in. DBH), was estimated visually and recorded by the video camera. To avoid tagging each tree in the stand, a 1-acre area was used to train researchers to estimate non-merchantable and merchantable trees from a distance. A bunching cycle included the time required to both fell and bunch a group of trees beginning and ending when the machine bunched and dropped a group of trees. A cycle included traveling, swinging, felling, bunching, and dropping. During videotape analysis, 753 bunching cycles were observed.

To determine skidding productivity, the Caterpillar 518C rubber-tired grapple skidder was studied. Before skidding, four skid trails were identified and distances from landings were measured. A total of 120 skidding turns were observed during the study. For each turn, the following variables were recorded. Time variables were recorded to the nearest 0.01 second and subsequently converted to minutes:

1. outhaul time,

2. grapple time,

3. inhaul time,

4. delay time and reason,

5. skidding distance,

6. ground slope, and

7. the number and merchantability category of pieces skidded.

To identify bottlenecks and determine landing efficiency, fixed-interval activity sampling information was collected for each machine during the productivity study. During the skidding, processing, and loading phase of the study, activity samples were recorded at 30-second intervals. The loader was observed for 260.5 minutes or 521 observations. The skidders and processor were observed for 335.5 minutes or 671 observations.

Following the conventional harvesting operation, an in-woods chipper and tub grinder were used to process landing slash into fuel chips to be utilized for the production of electricity. This operation occurred approximately 2 months after harvesting had been completed. The chipper processed slash piles consisting of both non-merchantable trees and limbs and tops from felled merchantable trees. Resulting material not effectively processed by the chipper was fed into the tub grinder for further utilization. Time required to chip or grind each vanload was recorded along with gross, tare, and net weights (green tons) reported by the processing facility.

Data analysis

To determine feller-buncher productivity, the amount of time spent treating non-merchantable and merchantable stems was estimated separately by analyzing various factors predicted to affect efficiency such as stems per bunch and their merchantability class. Predictor variables were examined for statistical importance using a stepwise regression procedure in Number Cruncher Statistical Systems (NCSS) (Hintze 2004).

Because all of the trees were felled and bunched together, felling and bunching time per tree (min.) was distributed evenly across all of the trees in the cycle using the following formula:

Felling and bunching time per tree (min.) = (F & B/[Pieces.sub.Total])

where:

F & B = delay-free felling and bunching time per cycle (min.)

[Pieces.sub.Total] = total number of pieces per cycle (non-merchantable and merchantable)

Individual felling and bunching times for each tree were characterized by merchantability class. The procedure was used to build information (variables) necessary to establish productivity equations for felling and bunching the non-merchantable and merchantable portions of the stand separately. A manual stepwise variable selection procedure was conducted to build a linear regression model that described each of the dependent variables. The model was fit to the data using NCSS. The criteria used for letting a variable enter and stay in the model was [alpha] = 0.05.

Similar to feller-buncher analysis, productivity of the skidding function was determined by estimating the amount of time spent extracting non-merchantable and merchantable trees separately. From the 120 skidding cycles observed during time studies, least squares stepwise regression was used to analyze and model variables affecting productivity. As for felling and bunching, all of the trees were skidded together irrespective of merchantability class. Therefore, skidding time per non-merchantable and/or merchantable tree (min.) was calculated using the following formula:

Skidding time per tree (min.) = (SK/[Pieces.sub.Total])

where:

SK = delay-free skidding time per cycle (min.)

[Pieces.sub.total] = total number of pieces per cycle (nonmerchantable and merchantable)

The formula distributes total skidding time per cycle evenly between each tree recorded during a cycle.

Landing efficiency was assessed through simultaneous activity sampling of the loader, processor, and two rubber-tired skidders. Percent of observations out of total observations approximated the percentage of time spent on one activity and identified landing bottlenecks, delays, and the time spent processing and separating non-merchantable and merchantable trees. Detailed productivity information was not collected for the loader and processor. Productivity of chipping and tub grinding was predicted by computing mean chipping and grinding times per vanload of chips and net load weights from scale tickets.

Based on information collected during the productivity study of individual machines, combined system productivity and costs of the fuel reduction operation were predicted using the Auburn Harvesting Analyzer (AHA) spreadsheet model (Tufts et al. 1985). Two AHA spreadsheets were constructed:

1. one spreadsheet calculated the system productivity and cost for the entire fuel reduction operation including harvesting and processing both non-merchantable and merchantable trees as observed in the field, and

2. a second spreadsheet determined the productivity and cost of harvesting/thinning only the merchantable portion of the stand.

The difference in the output of the two spreadsheets resulted in the added time and resulting increased cost due to harvesting and processing the non-merchantable portion and meeting fuel reduction objectives. Input assumptions for each model are listed in Tables 2 and 3.

Results and discussion

The difference between pre- and post-harvest stand measurements determined the number of trees and volume per acre removed during treatment. Since trees < 3 inches DBH were not removed during treatment, 309 residual non-merchantable trees per acre remained (Table 1). Tree weights (green tons) were estimated by calculating non-merchantable and merchantable components of each tree measured. Non-merchantable tons include: 1) foliage, limbs, and tops from a 6-inch diameter outside bark (DOB) non-merchantable top of merchantable sized trees, and 2) total weights for non-merchantable sized trees (foliage, limbs, tops, and bole). Total aboveground biomass was estimated with equations from Jenkins et al. (2004), whereas residue weight (foliage, limbs, and tops) was calculated using weight tables from Snell and Brown (1980). The difference between total weight and residue weight determined merchantable weight. Additionally, tons were predicted for merchantable-sized trees to a 6 inch (DOB) merchantability limit.

Felle-buncher productivity

Three linear regression models were constructed to estimate the following dependent variables:

1. delay-free felling and bunching time per cycle,

2. felling and bunching time per non-merchantable tree, and

3. felling and bunching time per merchantable tree.

Descriptive statistics for analyzed dependent and independent variables are listed in Table 4. During analysis of residual histograms, 13 observations were removed as outliers from the dataset, yielding 740 felling and bunching cycles for analysis. The residuals of each dependent variable were normally distributed and exhibited constant variance after transformations were applied.

Non-merchantable pieces per bunch (NMPPB) was chosen as the best predictor of productivity and merchantable pieces per bunch (MPPB) was the second best predictor. When combined, both terms were statistically significant ([NMPPB.sub.(t = 13.607, p < 0.0001)] and [MPPB.sub.(t = 8.825, p < 0.000l)]) and stayed in the model. The best model for estimating delay-free time per felling and bunching cycle included both analyzed independent variables (Model 1). It predicts the amount of time required to fell and bunch non-merchantable stems along with merchantable stems when piece counts are known or can be estimated.

Delay-free time per F&B cycle (min.) = [[0.70 + (0.10 x MPPB) + (0.056 x NMPPB)].sup.3] [R.sup.2] = 0.20, F-ratio = 92.74, RMSE = 0.129, p - value < 0.0001 [1]

The best model for estimating the felling and bunching time per non-merchantable tree included only the independent variable [NMPPB.sub.(t = -9.278, p < 0.0001)] (Model 2). Only 396 of the 740 felling and bunching cycles included at least one non-merchantable tree. These cycles were used to construct the regression to predict delay-free time to fell and bunch non-merchantable stems which was used to complete the cost analysis.

F&B time per non-merchantable tree (min.) = [[0.72 - (0.03 x NMPpB)].sup.3] [R.sup.2] = 0.18, F-ratio = 86.09, RMSE = 0.102, p-value < 0.0001 [2]

The best model for estimating the felling and bunching time per merchantable tree included both analyzed independent variables (Model 3). NMPPB was chosen as the best predictor and MPPB was the second best predictor. When combined, both terms were statistically significant (NMPPB(t = 12.216, p < 0.0001) and [MPPB.sub.(t = 6.806, p < 0.0001)]). Only 496 of the felling and bunching cycles included at least one merchantable tree.

F&B time per merchantable tree (min.) = [[1.06 + (0.22 x MPPB) + (0.14 x NMPPB)].sup.-3] [R.sup.2] = 0.27, F-ratio = 89.93, RMSE = 0.205, p-value < 0.0001 [3]

Skidding productivity

Three linear regression models were constructed to estimate the dependent variables of:

1. delay-free skidding cycle time,

2. skidding cycle time per non-merchantable tree, and

3. skidding cycle time per merchantable tree.

Descriptive statistics for analyzed dependent and independent variables are listed in Table 5. From field data collection, 120 skidding cycles were observed. During analysis of residual histograms, three observations were removed as outliers from the dataset, yielding 117 remaining skidding cycles for analysis.

The model constructed for estimating delay-free time per skidding cycle included the independent variables skidding distance (SD) and MPPC (Model 4). When combined, both terms were statistically significant ([SD.sub.(t = 15.061, p < 0.000l)] and [MPPC.sub.(t = 2.197, p = 0.0301)]). This model predicts the amount of time required to skid non-merchantable stems along with merchantable stems when merchantable piece counts and skidding distance (ft) are known or can be estimated.

Delay-free time per skidding cycle (min.) = [1.47 + (0.07 x MPPC) + (0.004 x SD)] [R.sup.2] = 0.67, F-ratio = 115.28, RMSE = 0.540, p-value < 0.0001 [4]

The model chosen for estimating skidding time per non-merchantable tree included the independent variables non-merchantable pieces per cycle ([NMPPC.sub.(t = 21.067, p < 0.0001)]), skidding distance ([SD.sub.(t = -11.019, p < 0.000l)]), and merchantable pieces per cycle ([MPPC.sub.(t = 7.676, p < 0.0001)]) (Model 5). Of the 117 initial skidding observations, only 105 included at least one non-merchantable tree. These 105 were used to construct the regression model.

Skidding time per non-merchantable tree (min.) = [[1.30 + (0.08 x MPPC) + (0.08 X NMPPC) -(0.0009 X SD)].sup.-2] [R.sup.2] = 0.82, F-ratio = 155.98, RMSE = 0.163, p-value < 0.0001 [5]

The regression model for estimating skidding time per merchantable tree included the independent variables non-merchantable pieces per cycle ([NMPPC.sub.(t = 23.168, p < 0.0001)]), skidding distance ([SD.sub.(t = -11.093, p < 0.0001)]), and merchantable pieces per cycle ([MPPC.sub.(t = 8.430, p < 0.0001)]) (Model 6). Although similar to model [5], this model included all 117 observations and was fit to give the best possible prediction of skidding time per merchantable tree.

Skidding time per merchantable tree (min.) = [[1.15 + (0.05 x MPPC) + (0.05 x NMPPC) -(0.0005 X SD)].sup.-3] [R.sup.2] = 0.83, F-ratio = 182.97, RMSE = 0.093, p-value < 0.0001 [6]

Time elements of a skidding cycle included outhaul, grapple, inhaul, and delay times. Outhaul and inhaul times were comparable with values of 1.19 and 1.44 minutes per cycle, respectively. Since the skidder brought turns to the same landing that it left from, this result implies that the skidder traveled slightly slower when loaded. Average grapple time per cycle was 0.86 minutes and is largely a function of bunch integrity which can be attributed to proper skid trail placement and felling pattern. Grapple time ranged from 0.25 to 3.27 minutes per cycle with few observations of multiple stops and was not highly correlated with any of the independent variables. Delay-free cycle time averaged 3.5 minutes and was used to model the linear regressions associated with productivity. Fifty-two delays averaging 2.19 minutes per occurrence were recorded. Distributing the delays over all 117 cycles resulted in an average delay time per cycle of 0.98 minutes or 22 percent. As a result, the total time per skidding cycle including delays was 4.47 minutes. Of the 52 delays recorded during skidding, 68.54 percent of the time was a result of waiting on the processor. This was due to congestion on the landing resulting from the extra time required to sort and handle non-merchantable trees. Interaction and waiting on the other skidder accounted for 8.26 percent of the delay time with research discussions totaling 8 percent. A detailed distribution of delays is shown in Figure 1.

Landing equipment productivity and interaction

Skidding activity sampling results indicate that harvesting the additional small non-merchantable trees congested the landing. Of 671 observations between the two skidders and the processor, waiting on the processor accounted for 19.23 percent and 14.61 percent of the sample time for the John Deere and Caterpillar skidders, respectively. These results imply that the processing function of the operation reduced skidding productivity.

Processing merchantable logs comprised the majority of the 335.5 minutes of scheduled time (57.82%) for the processor. Due to harvesting non-merchantable trees, piling non-merchantable trees and separating non-merchantable and merchantable trees comprised 14.46 percent and 2.38 percent, respectively, of observations. This result implies that the processor's productive capacity was reduced due to the influx of non-merchantable trees brought to the landing by skidders.

Since processor productivity per tree or volume was not directly measured, the following formula, using activity sampling results, was used to estimate its productive capability:

Processor productivity (tons/PMH) = SKP - [SKP x (WOP + WOS/100)]

where:

PMH = productive machine hour

SKP = skidding productivity (tons/PMH) for both skidders combined

WOP = average percent of activity sampling observations that a skidder was waiting on the processor

WOS = percent of activity sampling observations that the processor was waiting on a skidder

This formula reduces the processor's productivity by the total percentage of skidder and processor waiting time recorded during activity sampling. This calculation was necessary to adjust processor productivity because its capability is directly related to skidding productivity.

Harvesting the non-merchantable portion of the stand had little effect on loader productivity since this machine is primarily designed and used to sort and load merchantable logs. From activity sampling, 83.88 percent of observations were recorded for adjusting log sorts (53.36%) and loading merchantable logs (30.52%). The non-merchantable harvesting effects (separating non-merchantable and merchantable trees) had only two observations or 0.38 percent of the total. The following formula was used to estimate loader productivity based on average truck loading times and activity samples recorded during the study:

Loader productivity (tons/PMH) = (LS/MTL/60) x (PL + PD - PNM/100)

where:

PMH = productive machine hour

LS = average payload per truckload of roundwood (tons)

MTL = actual loading time per truckload of roundwood (min.)

PL = percent of activity sampling observations recorded as loading roundwood

PD = percent of activity sampling observations recorded for a delay

PNM = percent of activity sampling observations recorded as separating non-merchantable and merchantable trees

This formula calculates loader productivity based on two factors: 1) maximum productivity if loading occurred for 100 percent of the scheduled time, and 2) the percent of time the loader was available to actually load trucks. The result is the number of tons per PMH that the machine can load given the other activities that are required. For example, the loader has responsibility for adjusting log sorts which consisted of 53.36 percent of the scheduled time during this study.

During the chipping and tub-grinding phase of the operation, 11 vanloads of chips were produced, 7 from the chipper and 4 from the tub-grinder. These operations occurred approximately 2 months after conventional harvesting and were, therefore, not included in simultaneous activity sampling. Processing times per load and scale tickets, however, were recorded for determining productivity. Average productivity rates between the machines were very similar, 24.65 and 24.68 green tons per hour for the chipper and tub-grinder, respectively.

Harvesting system productivity and costs

Whole system.--The Auburn Harvesting Analyzer (AHA) analysis of the whole system as observed included the feller-buncher, two rubber-tired skidders, processor, loader, chipper, tub-grinder, and hauling. Results from the individual machine productivity analysis were used as inputs into the model. Trees and tons per acre inputs were based on removal density and volume.

Feller-buncher productivity was determined with regression models [2] and [3]. Using average non-merchantable (1.25) and merchantable (0.73) trees per cycle, productivity was determined to be 0.32 minutes per non-merchantable tree and 0.37 minutes per merchantable tree. When multiplied by the number of trees per acre removed in each merchantability class, hours per acre for the non-merchantable portion was 1.21 and 0.40 for the merchantable portion. The total tons per acre (61.29) divided by the total productive hours per acre (1.61) resulted in a productivity rate of 38.09 tons per PMH for the feller-buncher. Availability was set at 90 percent and production per SMH was 34.28 tons. After combining all of the machines in the system, cost per ton for felling and bunching was $4.94.

Productivity of the skidding function was estimated with regression models [5] and [6]. Using average non-merchantable (5.26) and merchantable (3.87) trees per cycle along with average skidding distance (483.89 ft.), productivity was determined to be 0.38 minutes per non-merchantable tree and 0.40 minutes per merchantable tree. When multiplied by the number of trees per acre removed in each merchantability class, hours per acre for the non-merchantable portion was 1.45 and 0.44 for the merchantable portion. The total tons per acre (61.29) divided by the total hours per acre (1.89) resulted in a productivity rate of 16.23 tons per PMH per skidder or 32.46 combined. Availability was set at 85 percent, production per SMH was 27.59 tons, and the resulting cost per ton for the skidding function was $6.21.

Since processor capability was directly related to skidding productivity, processor production was based on the maximum volume brought to the landing (32.46 tons per PMH). Due to long waiting times for both the skidders and processor, maximum productivity was reduced by the percent of activity samples recorded as waiting (22.73%). Using the formula previously described, resulting processor productivity was found to be 25.08 tons per PMH. With availability at 85 percent, production per SMH was 21.32 tons, and the resulting cost per ton for the processing function was $6.54.

A similar approach was taken to estimate loader productivity based on the average loading time per truckload (24.5 min.) and load size (26.84 tons). Using the formula previously constructed, the percentage of productive time available to load trucks was found to be 36.09 percent and resulted in a productivity rate of 23.72 tons per PMH (roundwood). Availability was set at 90 percent and production per SMH was 21.35 tons. After combining all of the machines in the system, cost per ton for the loading function was $5.16.

AHA inputs for chipper and tub-grinder productivity were estimated from average load processing times and weights. With availability at 90 percent, productivity per SMH was 22.19 and 22.22 tons for the chipper and tub-grinder, respectively. Resulting costs were $7.15 per ton for the chipping function and $6.89 per ton for tub-grinding.

Hauling productivity for both roundwood and chips was combined using the inputs:

1. haul distance--40 miles,

2. average speed--45 mph,

3. average load size--25.98 tons,

4. average load time--39.04 minutes, and

5. average unload time--30 minutes.

Round trip time to and from White City, Oregon was estimated to be 2.93 hours and productivity was 8.87 tons per PMH. With an availability of 90 percent, productivity per SMH was 7.98 tons with a unit hauling cost of $3.85.

After combining all of the machines in the operation, the AHA calculated the system rate (the productivity of the least productive function) to be 21.32 tons per SMH which is limited by the processor. The system rate was used to determine utilization and operating cost for each function. Actual utilization for the functions was: felling and bunching 56 percent, skidding 66 percent, processing 85 percent, loading 90 percent, chipping 86 percent, tub-grinding 86 percent, and hauling 90 percent.

Detailed AHA inputs and results for the whole system are outlined by Bolding (2006). In summary, on-board truck cost for harvesting both non-merchantable and merchantable trees was $39.83 per ton (chips and roundwood combined). After including the hauling cost of $3.85 per ton, the cut and haul costs were $43.68 per ton.

Merchantable portion.--The AHA analysis of removing only the merchantable portion included the feller-buncher, two rubber-tired skidders, processor, loader, and hauling. Chipping and tub-grinding were excluded from analysis because no non-merchantable material would be processed in such an operation. In addition, this analysis does not consider the cost of slash piling or backhauling into the stand that may be necessary in situations where biomass utilization is not feasible. Results from the individual machine productivity analysis were used as inputs into the AHA spreadsheet. Trees and tons per acre inputs were based on removal density and volume of the merchantable sized portion of the stand.

Feller-buncher productivity was determined with regression model [3]. Using average merchantable trees per cycle observed when no non-merchantable trees were present (1.11), productivity was determined to be 0.45 minutes per merchantable tree. When multiplied by the number of merchantable trees per acre removed, there were 0.49 hours per acre for the merchantable portion. The total harvested merchantable tree tons per acre (54.82) divided by the total hours per acre (0.49) resulted in a productivity rate of 110.71 tons per PMH for the feller-buncher. At 90 percent availability, production per SMH was 99.64 tons and the cost per ton was $2.10.

Productivity of the skidding function was estimated with regression model [6]. Using average merchantable trees per cycle observed when no non-merchantable trees were present (4.25) and average skidding distance of 483.89 feet, productivity was determined to be 0.71 minutes per merchantable tree. When multiplied by the number of merchantable trees per acre removed, hours per acre for the merchantable portion was 0.77. The total merchantable tree tons per acre (54.82) divided by the total hours per acre resulted in a productivity rate of 35.42 tons per PMH per skidder or 70.85 combined. Production per SMH (85% availability) was 60.22 tons and the cost per ton for the skidding function was $2.85.

Processor capability was based on maximum skidding productivity of 70.85 tons per PMH. This productivity was reduced by the same percentage of waiting time used in whole system analysis (22.73%). Using the formula described previously, processor productivity was found to be 54.75 tons per PMH. With an availability of 85 percent, production per SMH was 46.53 tons and the total costs per ton for the processing function was $3.00.

Maximum loader productivity was estimated using average loading time per truckload (24.5 min.) and load size (26.84 tons). Using the formula constructed previously, productive time to load trucks was found to be 85.95 percent and resulted in a productivity rate of 56.50 tons per PMH (roundwood). Production per SMH was 50.85 tons at 90 percent availability. After combining all of the machines in the system, cost per ton for the loading function was $2.28.

Hauling productivity for the merchantable portion was estimated using the inputs:

1. haul distance--40 miles,

2. average speed--45 mph,

3. average load size--26.84 tons,

4. average load time--24.5 minutes, and

5. average unload time--30 minutes.

Round trip time was estimated to be 2.69 hours and productivity was 9.99 tons per PMH, 8.99 tons per SMH, with a resulting unit cost of $3.73 per ton.

After combining all of the machines in the operation, the AHA calculated the system rate to be 46.53 tons per SMH. The processor was the limiting function. Actual utilization for the functions was: felling and bunching 42 percent, skidding 66 percent, processing 85 percent, loading 82 percent, and hauling 90 percent. After including support ($0.55) and moving ($1.26) costs, on-board truck cost for harvesting merchantable trees was $12.05 per ton. Once the hauling cost of $3.73 per ton was added, cut and haul costs were $15.77 per ton.

Non-merchantable portion.--Based on this analysis, the difference between the whole system and merchantable portion AHA spreadsheet outputs is attributed to the additional time necessary to harvest and process the non-merchantable portion of the stand and meet fuel reduction objectives. Figure 2 shows the breakdown of cost estimates per green ton for each function of the operation for harvesting: 1) only the merchantable portion, and 2) non-merchantable and merchantable trees as observed.

As expected, the most substantial increase in cost per ton due to harvesting trees < 7 inches DBH was for tub-grinding ($6.89) and chipping ($7.15). These functions were necessary to recover and transport the harvested biomass. It is also important to note that new machine prices were used for all of the machines. The chipper and tub-grinder have new purchase prices of $443,000 and $421,000, respectively. It is common for such fuel-wood operations to employ older equipment with subsequently lower capital expense required. For example, if the chipper and tub-grinder were purchased for $150,000 each, the added cost of the machines would be $9.12 per ton vs. $14.04 when using new machines. A reduction in availability and increase in maintenance and repair costs, however, may occur for used machinery.

No other harvesting functions increased in cost more than $4.00 per ton. Of the other functions, processing incurred the largest increase in cost of $3.54 per ton and was the limiting function in both AHA models. This result is likely due to the added time necessary to separate and pile non-merchantable stems which accounted for 16.84 percent of the activity sampling observations. Skidding costs increased by $3.36 per ton. This increase is a function of waiting times incurred at the landing due to processor interaction and a decrease in the number of tons skidded per SMH (60.22 vs. 27.59) due to the small volume contained in non-merchantable trees. Analysis of all harvesting functions combined resulted in a total increase in cost of $27.78 per ton onboard truck and $27.91 delivered. Hauling costs were nearly equal for each scenario and only increased by $0.12 per ton. This is due to similar average payload sizes for both chips and roundwood and equal hauling distances, since products were transported to the same facility.

Cost vs. revenue.--During the study, 16 truckloads or 429.37 tons of roundwood and 11 vanloads of fuel chips or 272.09 green tons were also sold. Total revenue generated was $24,211.07 for roundwood and $4,489.49 for fuel chips or $1,210.55 and $224.47 per acre, respectively. Using estimated harvesting costs per ton, cost per acre for harvesting both the non-merchantable and merchantable portions was $1,531.99 with total revenue of $1,435.03 and a resulting net treatment cost of $96.96 per acre. The increase in revenue from the sale of fuel chips was not enough to offset the additional costs incurred from harvesting and processing the non-merchantable portion of the stand. Harvesting more merchantable trees per acre would have increased overall revenue and resulted in a lower per acre treatment cost.

From a biomass harvesting and fuel reduction perspective, removing only the non-merchantable portion of the stand would result in a per acre cost of $1,193.43, revenue of $224.47, and a net negative margin of $968.96 (Fig. 3). Due to merchantable tree removal, however, operation costs were increased by $338.56 per acre, but revenue was also increased by $1,210.55, resulting in treatment cost reduction of $872.00 per acre. By integrating the harvesting system to include both non-merchantable and merchantable tree removal, fuel reduction benefits were gained and the resulting treatment cost was lessened from $968.96 to $96.96 per acre.

Summary and conclusions

Harvesting productivity, defined by the system rate, declined from 46.53 tons per SMH for harvesting the merchantable portion only to 21.32 tons per SMH when trees > 3 and < 7 inches DBH were processed. The system rate, or limiting function, for each scenario was produced by the processor. When small trees were brought to the landing, bottlenecks occurred as waiting time accounted for almost 17 percent of skidder activity samples. This identifies an opportunity for improvement through system balancing by adding another processor or removing a skidder from the operation. A more optimally balanced system would allow wood to flow through the landing more efficiently and likely increase machine utilization and subsequently reduce costs. Based on the volume of material produced, harvesting the non-merchantable portion increased both costs and revenue per acre. The increase in revenue through the sale of fuel chips, however, did not offset the increase in costs. Possible approaches to realize a positive margin when harvesting small trees include harvesting more merchantable trees per acre, better utilizing chipped material through valued-added products, and improving system balancing. Through sensitivity analysis, revenue per ton of chips would have to be $23.63 per green ton to break-even as opposed to the $16.50 per green ton realized during the study.

These results should be used cautiously and applied to similar stands, treatment types, and machine configurations. Productivity and cost information was derived from a combination of detailed and gross level samples. Without obtaining detailed productivity information for each machine, assumptions based on interaction are necessary. Costs, revenues, and profits per acre are controlled by silvicultural prescription, machine productivity, fuel loading, and product prices at the time of sale. As these metrics change, operation feasibility and profitability will change. This study provides information helpful for understanding the effects of harvesting different merchantability classes when small trees are removed in integrated harvests. Given the lack of available productivity and cost information concerning fuel reduction thinning with commercial harvesting systems, there is a great opportunity for further research in the area. Much research is needed to explore appropriate harvesting systems as well as processing and consumption possibilities for converting woody biomass into energy. The outcome of any fuel treatment is a function of the amount of merchantable material removed, the utilization success of the harvested biomass component, and the productivity of the harvesting system. Currently, most energy generated from biomass comes from utilizing landing slash or fuel reduction biomass and is processed with in-woods chippers. This process has been proven but its success is at the mercy of available markets. The expansion of the woody biomass market along with harvesting and transporting technology is necessary for renewable energywood sources to become economically competitive. In addition, long-term stand management practices after an initial treatment must be better defined to identify and justify initial stand prescriptions. Silvicultural strategies should be better linked to fire behavior and control objectives while also accounting for site disturbance, environmental factors, and harvesting system performance. Strategies should also recognize the importance of long-term stand trajectory, including structural and species diversity. This integrated approach to forest fuel reduction/biomass utilization will create more viable and sustainable options for forest management decision makers.

Literature cited

Bain, R.L. and R.P. Overend. 2002. Biomass for heat and power. Forest Prod. J. 52(2):12-19.

Bolding, M.C. 2002. Forest fuel reduction and energywood production using a CTL/small chipper harvesting system. MS thesis, Auburn Univ., Auburn, AL. 111 pp.

--. 2006. An integrated study of mechanical forest fuel reduction: Quantifying multiple factors at the stand level. PhD dissertation, Oregon State Univ., Corvallis, OR. 368 pp. Available at: http://hdl.handle.net/1957/2270.

--and B.L. Lanford. 2001. Forest fuel reduction through energywood production using a small chipper/CTL harvesting system. In: Proc. of the 24th Annual Council on Forest Engineering Meeting, Snowshoe, WV. 5 pp.

--and--. 2005. Wildfire fuel harvesting and resultant biomass utilization using a cut-to-length/small chipper system. Forest Prod. J. 55(12): 181-189.

--, --, and L.D. Kellogg. 2003. Forest fuel reduction: Current methods and future possibilities. In: Proc. of the 26th Annual Council on Forest Engineering Meeting, Bar Harbor, ME. 5pp.

Brinker, R.W., J.S. Kinard, R.B. Rummer, and B.L. Lanford. 2002. Machine rates for selected forest harvesting machines. Circular 296. Alabama Agric. Exp. Sta., Auburn Univ., Auburn, AL. 23 pp.

Brose, P., T. Schuler, D.V. Lear, and J. Berst. 2001. Bringing fire back: The changing regimes of the Appalachian mixed-oak forests. J. of Forestry 99(11):30-35.

Brown, R.T., J.K. Agee, and J.F. Franklin. 2004. Forest restoration and fire: Principles in the context of place. Conserv. Biol. 18(4):903-912.

Coulter, E.D., K. Coulter, and T. Mason. 2002. Dry forest mechanized fuels treatment trials project. In: Proc. of the 25th Annual Council on Forest Engineering Meeting, Auburn, AL. 3 pp.

Fight, R.D., B.R. Hartsough, and P. Noordijk. 2006. Users guide for FRCS: Fuel reduction cost simulator software. Gen. Tech. Rept. PNW-GTR-668. USDA Forest Serv., Pacific Northwest Res. Sta., Portland, OR. 23 p.

--, X. Zhang, and B.R. Hartsough. 2003. User's guide for STHARVEST, software to estimate the cost of harvesting small timber. Gen. Tech. Rept. PNW-GTR-582. USDA Forest Serv., Pacific Northwest Res. Sta., Portland, OR. 12 pp.

Hartsough, B.R., X. Zhang, and R.D. Fight. 2001. Harvesting cost model for small trees in natural stands in the interior Northwest. Forest Prod. J. 51(4):54-61.

Harvey, A.E., P.F. Hessburg, J.W. Byler, G.I. McDonald, J.C. Weatherby, and B.E. Wickman. 1995. Health declines in western interior forests: Symptoms and solutions. In: Proc. Ecosystem Management in Western Interior Forests. Washington State Univ. Cooperative Extension, Pullman, WA. pp. 163-170.

Hintze, J. 2004. NCSS and PASS. Number Cruncher Statistical Systems (NCSS). Kaysville, UT.

Hollenstein, K., R.L. Graham, and W.D. Shepperd. 2001. Biomass flow in western forests: Simulating the effects of fuel reduction and presettlement restoration treatments. J. of Forestry 99(10):12-19.

Jenkins, J.C., D.C. Chojnacky, L.S. Heath, and R.A. Birdsey. 2004. Comprehensive database of diameter-based regressions for North American tree species. Gen. Tech. Rept. NE-319. USDA Forest Serv., Northeastern Res. Sta., Newton Square, PA. 45 pp. [1 CD-ROM].

Kimmerer, R.W. and F.K. Lake. 2001. The role of indigenous burning in land management. J. of Forestry 99(11):36-41.

Kutscha, N. 1999. Trends in wood research and utilization. Forest Prod. J. 49(7/8):12-17.

Laverty, L. and J. Williams. 2000. Protecting people and sustaining resources in fire-adapted ecosystems: A cohesive strategy. USDA Forest Serv. 85 pp. Available at: www.fs.fed.us/publications/2000/cohesive_strategy10132000.pdf. (Last accessed Nov. 24, 2008).

Lynch, D.L. 2004. What do forest fires really cost? J. of Forestry 102(6):42-49.

Mason, C.L., B.R. Lippke, K.W. Zobrist, T.D. Bloxton, K.R. Ceder, J.M. Comnick, J.B. McCarter, and H.K. Rogers. 2006. Investments in fuel removals to avoid forest fires result in substantial benefits. J. of Forestry 104(1):27-31.

Mutch, R.W., S.F. Arno, J.K. Brown, C.E. Carlson, R.D. Ottmar, and J.L. Peterson. 1993. Forest health in the Blue Mountains: A management strategy for fire adapted ecosystems. Gen. Tech. Rept. PNW-GTR-310. USDA Forest Serv., Pacific Northwest Res. Sta., Portland, OR. 14 pp.

O'Laughlin, J. and P.S. Cook. 2003. Inventory-based forest health indicators: Implications for national forest management. J. of Forestry 101(2): 11-17.

Olsen, E.D. and L.D. Kellogg. 1983. Comparison of time-study techniques for evaluating logging production. T. ASAE 26(6):1665-1668.

Perlack, R.D., L.L. Wright, A.F. Turhollow, R.L. Graham, B.J. Stokes, and D.C. Erbach. 2005. Biomass as feedstock for a bioenergy and bioproducts industry: The technical feasibility of a billion-ton annual supply. DOE/GO-102995-2135. USDE and USDA. Oak Ridge National Lab., Oak Ridge, TN. 59 pp.

Snell, J.A.K. and J.K. Brown. 1980. Handbook for predicting residue weights of pacific northwest conifers. Gen. Tech. Rept. PNW-103.

USDA Forest Serv., Pacific Northwest Forest and Range Expt. Sta., Portland, OR. 43 pp.

Stokes, B.J. 1988. Timber harvesting systems in the southern United States. Publication no. 89-P-3. Forest Resources Assoc., Rockville, MD. pp. 1-10.

Tufts, R.A., B.L. Lanford, W.D. Greene, and J.A. Burrows. 1985. Auburn harvesting analyzer. Compiler. 3(2): 14-15.

USDA Forest Service. 2005. A strategic assessment of forest biomass and fuel reduction treatments in western states. Oen. Tech. Rept. RMRS-GTR-149. USDA Forest Serv., Rocky Mountain Res. Sta., Fort Collins, CO. 17 pp.

Watson, W.F., B.J. Stokes, and L.W. Savelle. 1986. Comparisons of two methods of harvesting biomass for energy. Forest Prod. J. 36(4):63-68.

Weaver, H. 1959. Ecological changes in the ponderosa pine forests of the Warm Springs Indian Reservation. J. of Forestry 57:15-20.

Woodall, C.W., C.H. Perry, and P.D. Miles. 2006. The relative density of forests in the United States. Forest Ecol. Manage. 226:368-372.

Zerbe, J.I. 2006. Thermal energy, electricity, and transportation fuels from wood. Forest Prod. J. 56(1):6-14.

M. Chad Bolding *

Loren D. Kellogg

Chad T. Davis

The authors are, respectively, Assistant Professor, Forest Operations/Engineering, Virginia Tech Dept. of Forestry, Blacksburg, Virginia (bolding@vt.edu); Lematta Professor of Forest Engineering, Dept. of Forest Engineering, Resources and Management, Oregon State Univ., Corvallis, Oregon (loren.kellogg@oregonstate.edu); and Program Manager, Sustainable Northwest, Portland, Oregon (chad.t.davis@gmail.com). This paper was received for publication in October 2007. Article No. 10408.

* Forest Products Society Member.

(1) The use of brand or model names is for reader convenience only and does not represent an endorsement by the authors, Virginia Tech, Oregon State University, or Sustainable Northwest.
Table 1.--Stand density and biomass statistics.

 Pre-
 harvest (a) Harvested (a) Residual (a)

Trees per acre
 Non-merchantable (b) 536 227 309
 Merchantable (c) 158 65 93
 Total 694 293 401
Green tons per acre
 Non-merchantable (b) 57.18 23.91 33.27
 Small tree 12.75 6.52 6.23
 Merchantable tree 44.43 17.37 27.06
 (limbs and tops) (d)
 Merchantable (c) 94.65 37.45 57.20
 Total 151.83 61.29 90.54
Basal area per acre
 ([ft.sup.2]) 144.37 59.30 85.07

(a) Number of observations = 15.

(b) Trees < 7 inches DBH.

(c) Trees [greater than or equal to] 7 inches DBH.

(d) Biomass above a 6-inch top (DOB).

Table 2.--Auburn Harvesting Analyzer input
assumptions for the whole system (as observed).

General information

 Hours/day 10
 Days/week 5
 Weeks/year 50
 Tract size 20 acres
 Move-to-tract 2 hours
 Move rate $3.00/mile
 Move distance 50 miles
 Distance home 15 miles
Support
 Pickups 2 @ $0.50/mile
 Foreman $2,500/month
 Overhead $2,500/month
 Chain saws 1 @ $700

Machine productivity (a)

Feller-buncher (b)
 NMPPB 1.25
 MPPB 0.73
 F&B time per NM tree (min.) [[0.7238 -(0.0326 X
 NMPPB)].sup.3]
 F&B time per Merch tree (min.) [[1.0578 + (0.2190 X
 MPPB) + (0.1413 X
 NMPPB].sup.-3]
Skidder (c)
 NMPPC 5.26
 MPPC 3.87
 Skidding distance (SD) 483.89 feet
 SK time per NM tree (min.) [[1.2995 + (0.0845 X
 MPPC) + (0.0810 X
 NMPPC) -2-(0.0009
 X SD)].sup.-3]
 SK time per Merch tree (min.) [[1.1458 + (0.0510 X
 MPPC) + (0.0479 X
 NMPPC) -3-(0.0005
 X SD)].sup.-3]
Processor (d)
 SK WOP 16.92%
 + PR WOS 5.81%
 = Total productivity reduction 22.73%
Loader
 Loading time/truckload 24.5 minutes
 Load size 26.84 green tons
 % of hour loading 30.52
 + % of hour in delays 5.95
 - % of hour spent on NM 0.38
 = total available % of hour 36.09
Chipper
 Chipping time/truckload 60 minutes
 Load size 24.65 green tons
Tub-grinder
 Grinding time/truckload 60.5 minutes
 Load size 24.89 green tons
Hauling
 Haul distance 40 miles
 Average speed 45 miles/hour
 Load size 25.98 green tons
 Load time 39.04 minutes
 Unload time 30 minutes
 Round trip time 2.93 hours
 Haul rate $2.50/mile

Machine cost

 Feller-
 buncher Skidder Processor

Initial cost ($) $433,000 $200,000 $560,000
Machine life (c) (yr) 5 5 5
Insurance and taxes (c)
 (% of initial) 3.5 5 2
Fuel and lubrication (c) ($/PMH) 20.17 11.80 15.40
Maintenance and repair (c) ($/PMH) 34.64 16.81 48.25
Labor (f) ($/SMH) 17.50 15.34 17.01
Interest rate (%) 10 10 10
Salvage value (% of initial) 20 20 30
Labor overhead (%) 40 40 40
Availability (%) 90 85 85
% of work day 100 100 100
Number of machines 1 2 1

 Tub-
 Loader Chipper grinder

Initial cost ($) $385,000 $443,000 $421,000
Machine life (c) (yr) 5 5 5
Insurance and taxes (c)
 (% of initial) 2 2.5 2.5
Fuel and lubrication (c) ($/PMH) 15.40 53.34 51.70
Maintenance and repair (c) ($/PMH) 33.17 37.80 35.93
Labor (f) ($/SMH) 17.12 17.12 17.12
Interest rate (%) 10 10 10
Salvage value (% of initial) 30 20 20
Labor overhead (%) 40 40 40
Availability (%) 90 90 90
% of work day 100 100 100
Number of machines 1 1 1

(a) Production equations were generated during productivity
analysis. NM = non-merchantable (< 7 in. DBH) and Merch =
merchantable ([greater than or equal to] 7 in. DBH).

(b) NMPPB = non-merchantable pieces per bunch, MPPB =
merchantable pieces per bunch, and F&B = felling and bunching.

(c) NMPPC = non-merchantable pieces per cycle, MPPC =
merchantable pieces per cycle, SD = average skidding distance
(ft), and SK = skidding.

(d) SK = skidder, PR = processor, WOP = waiting on processor,
and WOS = waiting on skidder.

(e) Brinker et al. 2002.

(f) Associated Oregon Loggers 2005 Wage Survey.

Table 3.--Auburn Harvesting Analyzer input
assumptions for the merchantable portion.

General information

 Hours/day 10
 Days/week 5
 Weeks/year 50
 Tract size 20 acres
 Move-to-tract 2 hours
 Move rate $3.00/mile
 Move distance 50 miles
 Distance home 15 miles
Support
 Pickups 2 @ $0.50/mile
 Foreman $2,500/month
 Overhead $2,500/month
 Chain saws 1 @ $700

Machine productivity (a)

Feller-buncher
 NMPPB 0
 MPPB 1.11
 F&B time per Merch tree (min.) [[1.0578 + (0.2190 X MPPB) +
 (0.1413 X NMPPB].sup.-3]
Skidder (c)
 NMPPC 0
 MPPC 4.25
 SD 483.89 feet
 SK time per Merch tree (min.) [[1.1458 + (0.0510 X MPPC) +
 (0.0479 X NMPPC) - (0.0005
 X SD)].sup.-3]
Processor (d)
 SK WOP 16.92%
 + PR WOS 5.81%
 = total productivity reduction 22.73%
Loader
 Loading time/truckload 24.5 minutes
 Load size 26.84 green tons
 % of hour loading 80
 + % of hour in delays 5.95
 - % of hour spent on NM 0
 = total available % of hour 85.95
Hauling
 Haul distance 40 miles
 Average speed 45 miles/hour
 Load size 26.84 green tons
 Load time 24.5 minutes
 Unload time 30 minutes
 Round trip time 2.69 hours
 Haul rate $2.50/mile

Machine cost

 Feller-
 buncher Skidder

Initial cost ($) $433,000 $200,000
Machine life (c) (yr) 5 5
Insurance and taxes
 (c) (% of initial) 3.5 5
Fuel and lubrication (c) ($/PMH) 20.17 11.80
Maintenance and repair (c) ($/PMH) 34.64 16.81
Labor (f) ($/SMH) 17.50 15.34
Interest rate (%) 10 10
Salvage value (% of initial) 20 20
Labor overhead (%) 40 40
Availability (%) 90 85
of work day 100 100
Number of machines 1 2

 Processor Loader

 $560,000 $385,000
Initial cost ($) 5 5
Machine life (c) (yr)
Insurance and taxes 2 2
 (c) (% of initial) 15.40 15.40
Fuel and lubrication (c) ($/PMH) 48.25 33.17
Maintenance and repair (c) ($/PMH) 17.01 17.12
Labor (f) ($/SMH) 10 10
Interest rate (%) 30 30
Salvage value (% of initial) 40 40
Labor overhead (%) 85 90
Availability (%) 100 100
of work day 1 1
Number of machines

(a) Production equations were generated during productivity
analysis. NM = non-merchantable (< 7 in. DBH) and Merch =
merchantable ([greater than or equal to] 7 in. DBH).

(b) NMPPB = non-merchantable pieces per bunch, MPPB =
merchantable pieces per bunch, and F&B = felling and bunching.

(c) NMPPC = non-merchantable pieces per cycle, MPPC =
merchantable pieces per cycle, SD = average skidding distance
(ft), and SK = skidding.

(d) SK = skidder, PR = processor, WOP = waiting on processor,
and WOS = waiting on skidder.

(e) Brinker et al. 2002.

(f) Associated Oregon Loggers 2005 Wage Survey.

Table 4.--Feller-buncher analysis descriptive statistics.

 Mean SD (a) Min. Max.

Dependent variables (min) (a,b)
 Total time per F&B cycle (c) 0.66 0.35 0.17 2.18
 F&B time per NM pieced 0.30 0.16 0.04 1.09
 F&B time per Merch piece (d) 0.48 0.26 0.09 1.48
Independent variables
(per F&B cycle) (a,c)
 NM pieces (NMPPB) 1.25 1.58 0 7
 Merch pieces (MPPB) 0.73 0.56 0 2

(a) NM = Non-merchantable, Merch = merchantable, F&B = felling
and bunching, NMPPB = non-merchantable pieces per bunch, MPPB =
merchantable pieces per bunch, and SD = standard deviation.

(b) Delay-free.

(c) Number of observations = 740.

(d) Number of observations = 396.

(e) Number of observations = 496.

Table 5.--Skidding analysis descriptive statistics.

 Count Mean SD (a) Min. Max.

Dependent variables
(min.) (a,b)
 Total time per SK cycle 117 3.50 0.93 1.19 5.90
 SK time per NM tree 105 0.44 0.22 0.14 1.24
 SK time per Merch tree 117 0.47 0.24 0.14 1.26

Independent variables
(per SK cycle) (a)
 NM pieces 117 5.26 4.66 0 21
 Merch pieces 117 3.87 1.48 1 8
 Average SK distance (ft) 117 483.89 210.80 10 830
 Average ground slope (%) 117 7.57 3.48 2 16

(a) SK = skidding, NM = non-merchantable,
Merch = merchantable, and SD = standard deviation.

(b) Delay-free.

Figure 1.--Breakdown of delay time incurred during
52 skidding delays.

Waiting on Processor 68.5%
Waiting on Skidder 8.3%
Research 8.0%
Personal 5.4%
Mechanical 9.8%

Note: Table made from pie chart.

Figure 2.--Merchantable portion and whole system harvesting
cost estimates.

 Harvesting Cost (S/green ton)

 Merch Portion Whole System

Fell & Bunch $2.10 $4.94
Skidding $2.85 $6.21
Processing $3.00 $6.54
Loading $2.28 $5.16
Chipping $0.00 $7.15
Tub-grinding $0.00 $6.89
Support $0.55 $1.33
Moving $1.26 $1.61
Total (Onboard Truck) $12.05 $39.83
Hauling $3.73 $3.85
Total (Cut-and-Haul) $15.77 $43.68

Note: Table made from bar graph.

Figure 3.--Effects on system cost, revenue, and margin ($/acre) of
harvesing the merchantable (Merch) and non-merchantable (NM) portions
of the stand.

 Cost Revenue Margin

NM & Merch $1,531.99 $1,435.03 -$96.96
NM Only $1,193.43 $224.47 -$968.96
Change $338.56 $1,210.55 $872.00

Note: Table made from bar graph.
COPYRIGHT 2009 Forest Products Society
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2009 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Bolding, M. Chad; Kellogg, Loren D.; Davis, Chad T.
Publication:Forest Products Journal
Article Type:Statistical data
Geographic Code:1USA
Date:Mar 1, 2009
Words:9978
Previous Article:Management styles of lumber mill managers in the northern United States.
Next Article:Changes in Appalachian hardwood lumber exporter practices, 1989-2002.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters