Factors affecting utilization of woody residues for bioenergy production in the southern United States

Biomass and Bioenergy 105 (2017) 278e287

Contents lists available at ScienceDirect

Biomass and Bioenergy journal homepage: http://www.elsevier.com/locate/biombioe

Research paper

Factors affecting utilization of woody residues for bioenergy production in the southern United States Raju Pokharel*, Robert K. Grala, Donald L. Grebner, Stephen C. Grado Department of Forestry, College of Forest Resources, Forest and Wildlife Research Center, Mississippi State University, Box 9681, Mississippi State, MS 397629681, United States

a r t i c l e i n f o

a b s t r a c t

Article history: Received 6 December 2016 Received in revised form 28 June 2017 Accepted 3 July 2017

Biomass is the renewable resource with the potential to serve as a substitute for fossil fuels. A form of biomass, woody residues, includes woody byproducts such as mill residues, logging residues, and other wood waste. This study estimated the impact of woody residue processing capacity, mill willingness to utilize additional logging residues for electricity production, mill equipment upgrades to facilitate electricity production, and different factors limiting the utilization of logging residues on the utilization of woody residues by mills. Data on the utilization of woody residues were collected via a census mail survey of mills in the southern United States. Data were analyzed using a two-stage least square (2SLS) regression model. Results indicated that average utilization was 91% of available woody residue processing capacity for bioenergy. Woody residue utilization increased with an increase in the processing capacity of woody residues, equipment upgrades, and lower transportation costs. In addition, the study found that there were only a few mills with a large enough processing capacity to utilize woody residues for bioenergy production. Thus, an increased utilization of woody residues will be highly dependent on mills' ability to increase woody residues processing capacity, implement necessary equipment upgrades, and lower transportation and processing costs. Logging residue utilization was relatively limited and had less of an impact on overall woody residue utilization. Findings will assist policy makers and mill managers in making informed decisions about woody residue utilization and identify specific mill factors that might help improve the economic feasibility of using woody biomass for bioenergy purposes. © 2017 Elsevier Ltd. All rights reserved.

Keywords: Biomass Logging residues Mill processing capacity Mill residues Transportation costs Two-stage least squares

1. Introduction In 2012, the United States produced 8% of its energy from renewable sources and approximately 50% of renewable energy was produced from biomass [1]. Biomass is utilized to produce energy through direct combustion and gasification processes to produce heat and electricity, and extraction of liquid fuels and chemicals [2e4]. A study done by U.S. Department of Energy (DOE) called the Billion-Ton Update of 2011 estimated that 907 million tonnes (Mt) [one billion U.S. tons] of dry biomass was available in the United States from forest and agricultural lands, and could potentially replace 30% of the national petroleum consumption [5]. Therefore, utilization of woody biomass for bioenergy purposes can

* Corresponding author. Current address: Department of Natural Resources and Society, College of Natural Resources, Policy Analysis Group, University of Idaho, 875 Perimeter Drive MS 1142, Moscow, ID 83844-1142, United States. E-mail address: [email protected] (R. Pokharel). http://dx.doi.org/10.1016/j.biombioe.2017.07.002 0961-9534/© 2017 Elsevier Ltd. All rights reserved.

help lower fossil fuel usage and reduce carbon dioxide (CO2) emissions. Depending on its source, woody residues utilized for bioenergy purposes are categorized into mill residues, logging residues, and other woody wastes such as construction wood waste, urban wood waste, and solid waste. Mill residues are residues produced as a byproduct during manufacturing activities at a mill such as the production of lumber, pulp, veneer, and plywood [6]. Mills include primary forest products manufacturers representing processing facilities classified under codes 312 and 322 of the North American Industry Classification System (NAICS) and they manufacture their products using raw material from forests [7]. Primary forest products manufacturers differ from secondary forest products manufacturers, classified under NAICS code 337, that produce their final products, such as furniture, by remanufacturing products obtained from primary manufacturers [7]. Logging residues are the commercially unwanted woody materials left at the logging site and include forest residues, coarse woody debris (down wood and standing snags), and fine woody debris (snags, cut limbs, and

R. Pokharel et al. / Biomass and Bioenergy 105 (2017) 278e287

leaves) [8]. Other woody wastes such as construction wood waste, urban wood waste, and solid waste are byproducts of construction activities and human consumption [9]. Urban wood wastes were not available in sufficiently large quantities to serve as a single feedstock for producing bioenergy although they were identified as a potential source [9]. According to the Billion-Ton Update study, 117 Mt of dry forest biomass and 77 Mt of dry agricultural biomass were utilized in 2010, and these quantities were projected to increase to 205 and 93 Mt by 2030, respectively [5]. Skog et al. [10] estimated that from 75 to 115 Mt of dry forest biomass could be supplied from all forest lands in the United States. Approximately 50% (87 Mt) of biomass energy consumption was contributed by woody residues and pulping liquors from the mills in the United States [11]. By 2031, utilization of woody residues is expected to account for 18e36% of the inputs for bioenergy production [12]. Therefore, a substantially large quantity of woody biomass is potentially available in the United States as a source of renewable energy and forest products industry can play an important role in increasing bioenergy production. Among all types of woody residues, mill residues are the least costly fiber residues that can be utilized to produce bioenergy at a mill because they are generated at the site or obtained from nearby wood processing facilities, thus avoiding transportation costs or lowering them substantially [13]. It is logistically more efficient and financially less costly to trade or give away mill residues to nearby facilities as compared to procuring logging residues or urban wood wastes [13]. However, the Billion-Ton Update study reported that 98% of mill residues were already utilized for bioenergy and nonbioenergy purposes such as animal bedding and mulch [5]. On the other hand, some studies indicated that a substantial quantity of mill residues was sold or used for purposes other than bioenergy, and this quantity could potentially be diverted to bioenergy production if there was a cost advantage [14] and a market for bioenergy [9]. Therefore, it is possible that mills will first utilize mill residues currently used for purposes other than bioenergy before considering other forms of residues as alternative feedstocks. While mill residues were already being used for bioenergy and other purposes, logging residues are being advocated as a potential alternative feedstock that can be utilized for bioenergy production [15]. Substantial quantities of logging residues are typically left unused on harvest sites after logging operations, increasing the likelihood of fire, pest outbreaks, and diseases [16]. On the other hand, logging residues offer many benefits such as maintaining soil productivity [16e18] and mitigating soil compaction as well as providing habitat for small mammals, birds, and arthropods, while maintaining soil hydrology through regulated evaporation, transpiration, and runoff control [19]. Although discussion on the impacts related to removal of logging residues is still unsettled, some studies have shown that a certain portion of logging residues (from 30 to 70%) can be removed for bioenergy purposes without affecting the soil productivity [5,16,20]. Forest certification standards can be used to ensure that best forest management practices include leaving a sufficient quantity of logging residues on a harvest site to maintain soil productivity and provide these other benefits [3]. Nationwide, the quantity of recoverable logging residues was approximately 36.2 Mt (dry weight) and this amount of residues has a potential of producing 67.5 TW h
1 of electricity [21]. Additionally, Gan and Smith [21] estimated that recoverable logging residues could displace 17.6 Mt of CO2 in the United States. Similarly, Hall [22] reported that 6 Mkm2 of biomass plantations having an average yield of 1200 t km
2 could offset 50% of CO2 emissions annually by 2050. Based on these studies, it seems that utilization of additional logging residues might play an important role, both in

279

bioenergy expansion based on woody biomass and lowering CO2 emissions. It should be noted that utilization of logging residues as a bioenergy feedstock is restricted by high delivery costs that include price paid to the owner, transportation costs, and logger charges [6]. Although, logging residues are widely available and have been promoted as an energy feedstock, mills have utilized logging residues in small quantities because their recovery and transportation costs were not feasible [23,24]. Furthermore, recovery of highervalue products might be more cost-effective to many logging operators rather than collecting logging residues. Logging residues require more space compared to logs and chips for the same weight to be transported and stored [25,26]. Thus, infrastructure is necessary to maintain a supply of logging residues and accommodate additional space and costs [27]. Lack of equipment to handle logging residues is another factor limiting their utilization [28]. Most equipment was designed for processing saw logs and pulpwood chips rather than small-diameter trees and logging residues. Purchasing specialized equipment to extract, transport, and process logging residues increases costs and reduces its economic feasibility [29]. The delivery cost of logging residues can be reduced and energy content improved if these residues are dried for 4e8 weeks as demonstrated after harvests of southern pine stands [30,31]. However, it is not a common practice to dry logging residues and deliver them to mills in the southern United States, because logging operators in the region are paid per unit of green weight [30,31]. Similarly, examples from other countries (e.g. Finland) indicated that most logging residues produced after harvesting were not used because of drying and storage problems [32]. Therefore, the necessity to dry logging residues and limited storage space at the mill increase the cost of logging residue utilization. While it might be easier to dry residues in warmer climates, such as in the southern United States, storage limitation remains a concern for many mill owners. However, some of these limitations might be easier to overcome in the southern United States, because the region has established logging, timber, and pulpwood supply chains [2,33]; has half of its land under forests [34,35]; and hosts large clusters of forest product manufacturers which might facilitate more efficient residue recovery [7,36]. This study determined the relationship between mill's bioenergy production characteristics, different factors limiting utilization of logging residues, and the quantity of woody residues utilized at mills in the southern United States. Although Gan and Smith [21] and the Billion-Ton Update study of DOE [5] indicated the availability of logging residues for bioenergy production, their utilization has been limited. A better understanding of bioenergy production potential based on woody biomass, consisting of mill and logging residues, will be useful in developing more effective programs and policies facilitating bioenergy production, and determining potential implications of proposed bioenergy policies. 2. Material and methods 2.1. Study area This study was conducted in 12 states in the southern United States, which includes Alabama, Arkansas, Florida, Georgia, Kentucky, Louisiana, Mississippi, North Carolina, Oklahoma, South Carolina, Tennessee, and Virginia (Fig. 1). The southern region is considered a wood basket of the United States because forests cover about 33% (3.3 Mkm2) of the United States land area and in the southern region of the country they cover approximately 50% (1.08 Mkm2) [34,37]. The majority of forests in the region are classified as timberlands, which are forests capable of producing at least

280

R. Pokharel et al. / Biomass and Bioenergy 105 (2017) 278e287

1500 m3 km
2 of industrial wood annually [35,38]. Most forest land (81%) in the southeastern United States is owned by nonindustrial private forest (NIPF) landowners [35]. In 2006, the United States Department of Agriculture (USDA) reported 376.6 Mm3 of timber growth and 178.4 Mm3 of timber removal in the southern United States, compared to 756.1 Mm3 of timber growth, and 438.9 Mm3 of timber removal nationwide [39]. Forestry was also one of the major industries in the southern regional economy, which provided, in 2009, 470 449 jobs and produced 133 billion $ in gross output [40e42]. A forest products supply chain in the region is wellestablished and includes large clusters of forest products manufacturers, logging contractors, and abundant woody biomass feedstocks which can be used not only for traditional wood product manufacturing but also for bioenergy production [30,34,42]. Approximately 16.5 Mt of dry logging residues was estimated as recoverable from the study area [21]. 2.2. Data sources A mail survey was conducted to collect data on woody residue utilization from all 2138 mills in 12 states of the southern United States. Mill mailing addresses were obtained from a database maintained by USDA Forest Service Southern Research Station [43]. The survey was implemented following an adaptation of Dillman's Tailored Design Method and consisted of four mail contacts: an introductory letter, a letter with a survey questionnaire, and two additional follow-up letters with questionnaires sent to nonresponding mills [44]. The survey questionnaire collected data related to existing capacity to process woody residues for bioenergy production, the quantity of woody residues utilized for bioenergy purposes, mill willingness to utilize additional quantities of logging residues to produce electricity, logging residue gate prices and hauling distances, and factors limiting utilization of additional logging residues. 2.3. Non-response bias test A non-response bias test was conducted to detect potential differences between first and last 50 responding mills in terms of mill characteristics. The last 50 mills served as a proxy for nonresponding mills. Additional tests were conducted to compare a proportion of questionnaires returned from each state with a proportion of questionnaires sent to a respective state and the proportion of questionnaires received from each manufacturer type with a proportion of questionnaires sent to a respective manufacturer type. A Wilcoxon-Mann-Whitney (WMW) two-sample test was used to determine if a non-response bias was present in data. This non-parametric test was selected because it produces consistent and efficient estimators when the distribution of the observed data is not known [45]. 2.4. Model specification

yi ¼ Xb þ y þ ε*

(2)

An insignificant parameter estimate of y indicates an absence of endogeneity, such that ε ¼ ε* . However, if the parameter estimate of y is significant, the model suffers from endogeneity. The y was obtained from regressing capacity on all independent variables in the OLS model including an instrumental variable, disposal, as specified in Equation (3) below. A variable disposal represents the quantity of mill residues or woody waste generated as a byproduct at a mill site and/or disposed of by reusing, selling, giving away, or landfilling. Since, 95% of woody residues used in the southern United States to produce bioenergy were mill residues according to the mill responses, the amount of disposal might be related to available capacity to produce bioenergy. The 2SLS model with two simultaneous equations was developed to address the endogeneity problem in the model. A model, 1st stage OLS regression, was developed where disposal (Z) served as an instrumental variable along with other exogenous variables to estimate the capacity as specified in Equation (3) [47]. The R-squared of this model was observed to determine how the selected instrumental variable b * from performed. Predicted capacity to process woody residues X Equation (3) was then used as an explanatory variable in the 2nd stage IV regression model instead of the observed capacity, as specified in Equation (4), to correct the endogeneity problem. 1st stage OLS regression:

b * ¼ Za þ Xd þ y X

(3)

2nd stage IV regression:

An ordinary least square (OLS) regression model was first constructed to quantify the relationship of woody residue utilization by a mill to produce bioenergy with bioenergy production profile characteristics of a mill and factors limiting residue utilization. A general model was expressed as a linear function of selected explanatory variables (X) as specified in Equation (1) [46e48]:

yi ¼ Xb þ ε

harvest site (utilize); X represents selected bioenergy production characteristics of a mill and utilization limiting factors; b stands for parameter estimates; and 3 is the error term. The dependent variable, utilize, and one of the independent variables, representing a capacity to process woody residues at a mill to produce bioenergy (capacity), were suspected of having a causal relationship with each other. This is due to the fact that the processing capacity to utilize woody residues is one of the major factors affecting quantity utilized to produce bioenergy [47,48]. Such relationship usually generates endogeneity in an OLS model, and the model does not produce consistent estimators following Markov's law of the best linear unbiased estimator (BLUE) due to an additional endogenous component in the error term [47]. Therefore, the Hausman endogeneity test was implemented to test for endogeneity by introducing an instrumental variable (IV) to determine if there was any causation between utilize and capacity in the model. The null hypothesis of Hausman endogeneity test was represented as plim {b2sls
b ¼ 0} where b, parameter estimates of OLS from Equation (1), are consistent and efficient when compared to b2sls, parameter estimates of the two-stage least squares (2SLS) regression from Equation (4) below [47]. The OLS model was re-specified as in Equation (2), where the error term (3) was replaced with ε* þ y, and a significance level of the endogenous component (y) was tested.

(1)

where yi indicates a dependent variable representing the quantity of woody residues utilized for bioenergy, including both mill residues produced at a mill site and logging residues recovered from a

b * g þ ε* yi ¼ Xb2sls þ X

(4)

where X represents all exogenous independent variables with a row of a constant; Z is an instrumental variable (IV) represented by b * is a predicted capacity to correct for endogeneity in the disposal; X model; b2sls, a, d, and g represent parameter estimates; and ε* and y are the error terms [47]. In addition, the Hausman specification test was conducted to identify consistent parameter estimators between OLS and 2SLS models as specified in Equation (5). Under the null hypothesis of the Hausman specification test, b2sls are consistent estimators (i.e., plim{X0 3/n ¼ 0}) [47].

R. Pokharel et al. / Biomass and Bioenergy 105 (2017) 278e287

281

Fig. 1. Study area in the southern United States for determining the utilization of woody residues for bioenergy production based on a 2012 mail survey of mills.

H ¼ d0 fEstimated asymptotic variance of dg d ¼ d0



0

b* X b* X


1

and; d ¼ ðX0 XÞ


1


ðX0 XÞ


1 . d s2


1 0

Xε

(5)

(6)

(7)

where s2 is the variance estimator of the variance in Equation (1). The 2SLS is the most common method for estimating simultaneous equations using an instrumental variable to correct for endogeneity and produces the most consistent estimators if 2SLS errors do not have heteroskedasticity and autocorrelation [47]. To ensure consistency and efficiency of parameter estimates in all models, heteroscedasticity-consistent standard errors (HCSE) or robust errors were used for hypothesis testing. Multicollinearity was tested using variance inflation factor (VIF), whereas serial autocorrelation in the errors was tested using the Durbin-Watson (DW) test [47]. In addition, an analysis was conducted to determine if capacity and utilize were normally distributed because data skewness produces too wide or too narrow confidence intervals, results in a disproportionate effect on the parameter estimates, and might influence inferences about the model and parameter estimates with an incorrect prediction error. The usage of a two-stage regression model, where capacity was estimated from the first model and used in the second stage as predicted capacity with a normal distribution, eliminated the issue of non-Gaussian distribution and the impact of skewness on the regression model. 2.5. Model and variable definitions A dependent variable representing the quantity of woody residues utilized to produce bioenergy (utilize), to a large extent, is affected by the capacity to process woody residues at a mill [47,48].

However, an actual utilization of woody residues can be smaller than available processing capacity due to other factors such as the price of woody residues, transportation costs [49], lack of storage space [32], lack of equipment to handle woody residues [28], and availability of woody residues [15]. Therefore, independent variables used to model utilize included a capacity to process woody residues at a mill to produce bioenergy (capacity); an amount of mill residues or wood waste generated on a mill site as a byproduct and/or disposed of by reusing, selling, giving away, or landfilling (disposal); past equipment upgrades to produce electricity (pupgrade); anticipated future equipment upgrades to produce electricity (fupgrade); and mill willingness to utilize additional logging residues to produce electricity (willing). A variable pupgrade, while not necessarily related to processing capacity, increases mill ability to produce bioenergy from residues and, thus, impacts a woody residue utilization [6]. If utilization of woody residues for bioenergy production is limited by a processing capacity or deficiency in the infrastructure necessary to produce bioenergy, mills might decide to upgrade their equipment to facilitate increased bioenergy production. The importance of fupgrade to mills can indicate whether current infrastructure to produce bioenergy limits woody residue utilization. Official statistics indicate that only 1.5% of the mill residues were not utilized [5]. Thus, procurement of additional woody feedstocks, such as mill residues from other facilities and logging residues from harvest sites, might be necessary to increase utilization of woody residues in the mill. Mill willingness to utilize additional logging residues was a key issue in their utilization because it involves additional costs and requires specialized equipment to handle logging residues [28]. Therefore, it was assumed that willing would increase utilize because mills would procure additional logging residues to produce bioenergy if this activity was cost-effective [50]. However, it is also possible that a mill would first utilize its own residues currently diverted to uses other than bioenergy production because they are already available and their recovery and processing would be less costly than a

282

R. Pokharel et al. / Biomass and Bioenergy 105 (2017) 278e287

procurement of logging residues. Variables representing bioenergy production profile characteristics of a mill (capacity, utilize, and disposal) were reported for a month in green t (weight reflecting an approximate 50% moisture content), whereas pupgrade, fupgrade, and willing were coded as binary variables. There are numerous challenging aspects in logging residue utilization as a feedstock for bioenergy production. High transportation costs (tcost) to procure logging residues were reported as an important factor limiting its utilization [5,51]. For example, Perez-Verdin et al. [6], quantified that transportation accounted for 41% of woody biomass feedstock recovery costs, whereas harvest and stumpage accounted only for 22% and 15%, respectively. Similarly, lack of, or limited storage space (lstorage), often limits mill ability to process logging residues [32]. Another limitation relates to a lack of specialized equipment (lequip) necessary to handle and process logging residues. This is equipment that is different from conventional equipment used by mills, and often mills lack or do not have sufficient quantities of logging residues to afford such equipment [28]. Other studies reported that availability of logging residues (lresid) might be an important concern for mills interested in electricity production or other forms of bioenergy [15]. Variables representing factors limiting additional utilization of logging residues were originally measured on a Likert scale from one to six: 1 e unimportant, 2 e of little importance, 3 e moderately important, 4 e important, 5 e very important, and 6 e unsure. Unsure responses were removed from further analysis. To facilitate econometric analysis, the original Likert scale categories were recoded into a binary variable, in which original categories 3, 4, and 5 were coded as important and 1 and 2 coded as not important. The final empirical model for utilization of woody residues for bioenergy production was specified as a vector of bioenergy production profile characteristics of a mill and factors limiting utilization (Equation (8)): utilize ¼ f (capacity, capacity_sq, pupgrade, fupgrade, willing, tcost, lstorage, lequip, lresid) (8) where the square of the independent variable, capacity, was included along with capacity to allow for a direction of a quadratic relationship because utilize might not change linearly with varying capacity. The analysis was carried out in SAS 9.4 software. 3. Results

to other mills or a landfill. Approximately 36% of mill residues was recycled or reused by a mill for bioenergy purposes, 39% was used for purposes other than bioenergy, 19% was sold to other mills, 1% was given away free, 1% was disposed of, whereas utilization method of remaining 4% was unknown. Mills reused woody residues for a variety of bioenergy purposes including heat production (35%), electricity production (4%), pellets (3%), chemical extraction (0.01%), and others (9%). The reuse of a remaining portion of mill residues (49%) was not reported by mills. In terms of physical quantities, the largest monthly usage accounted for heat and electricity production at 2596 green t (n ¼ 69) and 14 005 green t (n ¼ 9), respectively of woody residues. Although several mills reported using residues for extraction of chemicals, they did not report quantities. Additionally, mills used 590 green t (n ¼ 17) of woody residues for purposes other than bioenergy production such as mulch and livestock bedding. About half of mills (52%) did not report how they utilized their woody residues although they reported woody residues available at the mill. In terms of implemented equipment upgrades to facilitate bioenergy production, only a small number of mills (2%) upgraded their equipment to produce electricity, whereas 8% were planning such upgrades in the near future. Similarly, only a small proportion of mills, about 11%, indicated that they were interested in utilizing additional quantities of logging residues for electricity production. These mills were willing to pay 11.58 $ green t
1 (median: 14.58 $ green t
1) for logging residues at the mill gate. Mills also reported that their average hauling distance for logging residues was 79 km (median: 72 km) and indicated that, on average, it was economically feasible to haul logging residues over 93 km (median: 80 km). Mills reported that utilization of logging residues, as an alternative feedstock for bioenergy production, was associated with many challenges. Approximately 78% of mills reported that high transportation costs were the most important factor limiting the utilization of additional logging residues for electricity production. For 75% of mills, a lack of sufficient capacity for processing woody residues was the most limiting factor, whereas a lack of proper equipment to handle logging residues and a lack of storage space was most limiting for 70% and 68% of mills, respectively. Furthermore, 53% of mills reported other limiting factors such as the price of logging residues, policy regulations concerning biomass utilization and harvesting, investment in bioenergy equipment, quality of logging residues, and the lack of economic feasibility tied to logging residue utilization. However, only 48% of mills reported the availability of logging residues as the most important limiting factor.

3.1. Descriptive results 3.2. Econometric analysis results In total, 258 usable mail questionnaires were returned by mills resulting in an adjusted response rate of 20%. Other similar studies using a survey approach in the southern United States received comparable response rates [13,52]. A non-response bias test did not detect any significant difference between early and late respondents for selected variables (p > 0.10): willing, capacity, disposal, pupgrade, fupgrade, gprice, haul, and ehaul. Similarly, there were no differences in a proportion of questionnaires sent to and received from each individual state (p ¼ 0.4679) and each individual mill type (p ¼ 0.8629). Approximately 70% of mills indicated they utilized woody residues for bioenergy purposes. The capacity to process woody residues in a month at a mill to produce bioenergy was 3698 green t whereas a median capacity was 486 green t. However, on average, mills utilized 3363 green t (median: 777 green t) of woody residues to produce bioenergy in a month indicating approximately 91% usage of their available processing capacity (Table 1). On average, a mill generated and disposed of 2605 green t (median: 729 green t) of mill residues and/or wood waste, which was reused, sold or sent

The Hausman endogeneity test indicated that the endogenous part of the error from the initial OLS regression model, y, was different from zero (n ¼ 61, t-stat ¼ 0.26983, HCSE ¼ 0.13209, p ¼ 0.046) indicating the presence of endogeneity in OLS estimations. Endogeneity confirmed the causality between capacity and utilize. Furthermore, we failed to reject Hausman specification test (null hypothesis: OLS produces consistent estimators compared to 2SLS, n ¼ 71, H-stat ¼ 4.73, df ¼ 6, p ¼ 0.5790) indicating that parameter estimates produced by the 2SLS model were consistent when compared to OLS. Therefore, to correct for endogeneity and obtain consistent estimators, the second order 2SLS model was used. R-squared for the 1st stage OLS regression was 69%, indicating that disposal was a good instrumental variable in explaining capacity. In addition, 96% of mills used only mill residues to produce bioenergy, whereas the remaining 4% also used logging residues. This indicated that most of the available capacity was used to process mill residues rather than other types of woody residues. Therefore, disposal was an important factor in explaining available

R. Pokharel et al. / Biomass and Bioenergy 105 (2017) 278e287

283

Table 1 Factors affecting the utilization of woody residues to produce bioenergy based on a 2012 mail survey of mills in 12 states in the southern United States. Variable name Variable description

N

Mean SD

Median Max

a. Bioenergy production characteristics of a mill capacity utilize disposal pupgrade fupgrade willing

Capacity to process woody residues for production of bioenergy in a month at a mill in green weight ca. 50% moisture mass fraction (t). Quantity of woody residues utilized in a month at a mill in green weight (t). Quantity of woody residues and waste generated and disposed of (sold, reused, or given away) in a month at a mill in green weight (t). Past equipment upgrades to produce electricity from woody residues: 1 if mill upgraded equipment in the past, 0 if not. Planned future equipment upgrades to produce electricity from woody residues: 1 if mill planned future upgrades, 0 if not. Willingness to utilize additional quantities of logging residues to produce electricity: 1 if the mill was willing to utilize additional logging residues, 0 if not.

176 3698

10 143 486

63 170

208 3363 182 2605

7908 5868

777 729

59 283 50 439

228 0.02

0.13

0.00

239 0.08

0.26

0.00

227 0.11

0.31

0.00

153 0.78

0.42

1.00

153 0.68

0.47

1.00

156 0.70

0.46

1.00

145 0.48

0.50

0.00

150 0.75 36 0.53

0.43 0.51

1.00 1.00

b. Factors limiting utilization of additional logging residuesa tcost lstorage lequip lresid lmillcap other

Importance of high transportation cost in limiting utilization of additional quantities of logging residues: 1 if important, 0 if not. Importance of lack of storage space in limiting utilization of additional quantities of logging residues: 1 if important, 0 if not. Importance of lack of equipment to handle logging residues at a mill in limiting utilization of additional quantities of logging residues: 1 if important, 0 otherwise. Importance of limited availability of logging residues in limiting utilization of additional quantities of logging residues: 1 if important, 0 otherwise. Importance of lack of a capacity to process woody residues at a mill: 1 if important, 0 if not. Importance of other factors in limiting utilization of additional quantities of logging residues; 1 if important, 0 not.

a Originally measured on a Likert scale from one to six: 1 e unimportant, 2 e of little importance, 3 e moderately important, 4 e important, 5 e very important, and 6 e unsure. Unsure responses were removed from further analysis. The Likert scale categories were recoded into a binary variable, in which original categories 3, 4, and 5 were coded as important and 1 and 2 coded as not important.

capacity and was selected to serve as an instrumental variable. Variables including capacity, pupgrade, fupgrade, tcost, and lequip were significant at a 5% level in explaining the quantity of woody residues utilized by mills to produce bioenergy (Table 2). Overall, the R-squared was 95% indicating a good model fit. VIF estimates were smaller than 10, as reported in Table 2, and indicated a lack of multicollinearity between explanatory variables. Insignificant serial autocorrelation of the first order (DW-stat ¼ 0.23, p ¼ 0.99) indicated that the errors were not generated by a first-order autoregressive process. An increase in capacity increased utilize at an increasing rate (p ¼ 0.0480, p < 0.0001). At a monthly processing capacity of 500 green t, the marginal rate of utilize was 0.42 green t for one additional t of increased capacity. At a processing capacity of 3600 green t (an average capacity for all surveyed mills), the marginal rate of utilize was 0.53 green t, whereas for a larger mill with a processing capacity of 15 000 green t, it was 0.94 green t. However, for a mill with a processing capacity of 65 000 green t, the marginal rate of utilize was substantially higher at 2.76 green t for one green t of added capacity. A variable pupgrade had a positive relationship with utilize (p ¼ 0.0002). Mills that implemented equipment upgrades in the past to facilitate electricity production utilized additional 17 710 green t of woody residues per month when compared to mills that did not implement any upgrades. However, fupgrade had a negative relationship with utilize (p ¼ 0.0424). Mills that planned to implement equipment upgrades in the near future utilized 842 green t less woody residues per month compared to mills that did not plan equipment upgrades. The most limiting factor in the utilization of additional quantities of logging residues, tcost, had a significant association with utilize (p ¼ 0.0145). Mills reporting a high transportation cost as an important factor limiting utilization of additional logging residues utilized 1890 green t less woody residues per month than mills reporting this factor as unimportant. However, leqip was positively related with woody residue utilization (p ¼ 0.0218). Mills stating a

lack of equipment to handle logging residues as an important limiting factor utilized additional 1707 green t of woody residues per month compared to mills for which this factor was unimportant.

4. Discussion Mills in the southern United States are often aggregated in specific geographical areas commonly referred to as forest products industry clusters [7,36,41]. Mills located in industry clusters might be better positioned to utilize larger quantities of woody residues due to information and technology transfers and resource pooling within the cluster [7,33,53]. In such clusters, mills with larger processing capacities to produce bioenergy will first exhaust their mill residues, which is the least costly feedstock source. Once, large mills fully utilize their mill residues, they will seek additional feedstock sources and possibly increase usage of mill residues obtained from nearby manufacturers that cannot fully utilize their woody residues. After all mill residues in a forest industry cluster are utilized, mills might start procuring logging residues which are abundant in the southern United States but are also costlier than mill residues. On the other hand, with continuing technological advancements and improvements in timber processing, mills will manufacture their products with better efficiency and produce smaller quantities of mill residues [54e56]. Thus, a reduction of disposable mill residues might shift demand towards logging residues as an alternative feedstock for bioenergy production. Logging residues are often mentioned as an alternative feedstock for bioenergy production in addition to mill residues [15]. They are less costly than short-rotation woody crops (SRWCs) and other energy crops [23]; however, compared to mill residues, utilization of logging residues is associated with many challenges. Results indicated that the high transportation costs had a negative association with current utilization of logging residues. Gan and Smith [21], Galik et al. [15], and Guo et al. [57] also indicated that

284

R. Pokharel et al. / Biomass and Bioenergy 105 (2017) 278e287

Table 2 Two-stage least squares (2SLS) estimation results of factors related to utilization of woody residues to produce bioenergy by mills in the southern United States based on a 2012 mail survey. Variable

Coefficient (HCSEa) constant capacity capacity_sqc pupgrade fupgrade willing tcost lstorage lequip lresid disposal (instrumental variable)

2nd stage IV regression Estimation of UTILIZE (n ¼ 61)

1st stage OLS regression Estimation of CAPACITY (n ¼ 69) VIFb

2664.59** (1122.24)

397.61 (1122.24) 1267.37 (1362.98) 900.40 (6953.41) 1998.77 (1187.57) 302.92 (1242.51)
4330.53** (1872.27)
748.89 (1156.10) 0.87*** (0.20)

R-squared Serial autocorrelation (Durbin-Watson First-Order Autocorrelation)

2.19 1.28 1.30 1.50 1.31 1.55 1.31 2.21 0.69 0.46 [p ¼ 1.00]

Coefficient (HCSEa)

VIFb


191.65 (716.12) 0.40** (0.20) 0.000036*** (0.0000039) 17,710.00*** (451.67)
842.00** (404.43) 119.19 (414.63)
1389.88*** (548.88) 398.23 (369.69) 1706.67** (721.46)
132.15(397.28)

5.71 4.50 3.59 1.18 1.21 1.59 1.24 2.11 1.37

0.95 0.23 [p ¼ 0.99]

***

p < 0.01. p < 0.05. * p < 0.10. a Heteroscedasticity Consistent Standard Error (HCSE). b Variance Inflation Factor (VIF). c The SQ suffix indicates the quadratic term of the original variable. **

high transportation cost for logging residues was an important limitation. Perez-Verdin et al. [6] reported that cost of harvesting and transporting logging residues from a harvest site to a processing facility accounted for two-thirds of the overall cost of producing energy from logging residues. Furthermore, purchasing specialized equipment to extract, transport, and process logging residues increases the cost to both logging contractors and mill owners, and thus reduces the economic viability of utilizing logging residues [29]. Therefore, reducing the procurement cost of logging residues will be a key component in facilitating additional utilization of logging residues to produce electricity and other forms of bioenergy. On the other hand, increase in demand for renewable feedstocks due to policies requiring minimum amounts of renewable energy in the production of electricity might help mitigate the effect of high transportation costs and increase bioenergy production [58,59]. For example, policies related to required shares of renewable energy in utility companies' energy portfolios have not only increased the transatlantic pellet market but also made it economically viable. In a recent decade, the pellet market in Europe has grown as a result of increased efforts to reduce the CO2 emissions as mandated by European Union (EU) regulations [58,59]. In 2014, the USA exported four Mt of wood pellets of which about 92% were shipped to the European countries [60,61]. The southeastern United States experienced growth in wood pellet industry due to a large availability of forest-based feedstock at competitive prices, well-established supply chain, less costly transportation compared to other regions, and sustainably managed forests [58]. Although high transportation costs have limited electricity production from logging residues, they had a minimal impact on a pellet industry [62]. However, while tree logs and chips have been predominantly used for pellet production, few mills reported selling mill residues to pellet production facilities. Wood pellets seem to be most competitive and viable bioenergy product that can be produced from logging residues [59,62]. In addition, a technological potential to utilize logging residues to produce biodiesel and jet fuels has successfully been demonstrated by Northwest Advanced

Renewables Alliance (NARA) [63,64]. However, the cost was relatively high compared to the conventional fossil fuels due to a high transportation cost [65,66]. Other studies have also indicated that logging residues can potentially be used as a feedstock for bioenergy purposes [6,21]. However, mills did not utilize logging residues in substantial quantities because of high transportation costs and other economic limitations such as high procurement cost [15,21,57], need of specialized equipment to handle logging residues at mill and logging sites [28], and high investment costs related to bioenergy production [67,68]. Another factor that can potentially limit future utilization of logging residues is related to available processing capacity to produce bioenergy. The actual average monthly utilization of woody residues (3363 green t) accounted for approximately 91% of available average woody residue processing capacity (3698 green t) indicating that, on average, mills used most of their available processing capacity to produce bioenergy. Therefore, additional utilization of woody residues will depend largely on their ability to increase processing capacity to produce bioenergy and implement equipment upgrades that will facilitate handling, processing, and utilizing woody residues. Although 70% of mills reported utilizing woody residues for bioenergy, there were only a few mills with a relatively large processing capacity to utilize woody residues. A possible explanation for this trend is that only mills with large woody residue processing capacity had a suitable infrastructure and equipment to utilize woody residues at higher levels and thus had the ability to produce bioenergy [53,69]. In contrast, mills with small capacity might face various challenges, such as a limited ability to finance costly equipment upgrades necessary for facility expansion and limited storage space. Consequently, for many mills, it might seem more practical to dispose of woody residues than to use them to produce bioenergy. While a processing capacity increase has been determined as a major factor facilitating a future increase in wood-based bioenergy production, the results also indicated that capacity increase in a mill with larger woody residue processing capacity was associated with a greater impact on the utilization of woody residues than a

R. Pokharel et al. / Biomass and Bioenergy 105 (2017) 278e287

similar increase in a mill with smaller processing capacity. An increased marginal rate of utilization was observed at higher capacities. For example, the marginal rates of utilization were 53%, 94% and 276% at mills with woody residue processing capacities of 3600, 15 000, and 65 000 green t, respectively, per month. The marginal rates of utilization represent the rate at which mills increase their utilization for a unit change in processing capacity at that capacity level. For example, for a mill with a processing capacity of 15 000 green t, the utilization of woody residues was 94% for one added green t of processing capacity. Furthermore, the quantity of utilized woody residues was 276% for one added t of processing capacity at a processing capacity of 65 000 green t. This might seem counterintuitive because woody residue utilization cannot be greater than a processing capacity. However, mills with a large capacity to utilize woody residues are able to start utilizing portions of processing capacity that typically are unused at smallcapacity levels. This is due to equipment efficiency, better infrastructure, and associated economies of scale that lead to improved utilization of woody residues at larger capacity levels. However, the ability to increase overall utilization of woody residues in the region, based on large-capacity mills, might be limited because there are only a few such mills. Therefore, the study does not discourage smaller capacity upgrades because the availability of internally produced residues may be limited to justify large-scale upgrades and mills may lack equipment or interest to procure additional feedstocks such as logging residues. Discouraging smaller upgrades may restrict utilization of internally produced mill residues, which may end up in a landfill or be used for other purposes. Although the marginal rate of utilization for added capacity will be smaller, upgrading woody residues processing capacity in many smallcapacity mills might be an effective approach to increase the number of mills utilizing woody residues and discouraging the use of residues for purposes other than bioenergy. Related to capacity increases are equipment upgrades facilitating bioenergy production. Past upgrades for electricity production were positively associated with the utilization of woody residues indicating that these upgrades led to an increased utilization. Such upgrades can result not only in an increased processing capacity but also can increase the efficiency of bioenergy production at the same processing capacity level. A current utilization of woody residues already accounts for approximately 91% of existing woody residue processing capacity and, therefore, equipment upgrades to increase processing capacity will be critical in increasing woody residue utilization for bioenergy purposes. However, results also indicated that mills planning future upgrades utilized smaller quantities of woody residues compared to mills that did not plan upgrades. A possible explanation for this trend is that mills planned to implement future upgrades because they had available woody residues, but not enough processing capacity or equipment to utilize them. Thus, these mills were utilizing only a portion of their woody residues that was not sufficient to fulfill their bioenergy needs and, thus, anticipated upgrades were negatively associated with utilization of woody residues. In addition, mill's decision to implement an upgrade might be constrained due to associated costs. Equipment upgrades can be costly for a mill to install and operate a system to produce bioenergy using woody residues. For example, the initial capital cost of producing electricity from biomass was estimated at 140 to 4260 $ kW
1 h
1 (2010 $), and the operational costs were 0.21 to 0.40 $ kW
1 h
1 [67,68]. Most mills could not afford such costs and, therefore, state and federal policies and incentives such as startup support, tax incentives, renewable energy credits, and low-interest rate loans will be needed to facilitate utilization of more woody residues for bioenergy production. Results also indicated that a lack of equipment to handle logging residues at a mill had a positive association with current woody

285

residue utilization. This might seem contradictory, as previous studies indicated that a lack of equipment to handle logging residues was a factor limiting utilization of logging residues [28]. A possible explanation might be that only 4% of mills, mostly manufacturing pulp, paper, paperboard, and composite products, reported utilizing logging residues in addition to mill residues. Thus, it was possible that survey responses included mills that never used logging residues and were not aware of special equipment required to handle them. On the other hand, logging residues are usually chipped or grounded and transported for bioenergy purposes [55], similar to chips used for pulp, paper, paperboard, and composite wood products. Pulp, paper, and paperboard, as well as composite wood products mills, are also the major wood-based bioenergy producers [53] with most suitable infrastructure for logging residue utilization. Thus, these mills reported a lack of equipment to handle logging residues as not limiting current utilization because they can do so without substantial modifications to their existing equipment. Additionally, although statistical tests suggested that the bias between responding and non-responding mills was not present, it is possible that a bias might have occurred in answering this particular question due to a lack of required information and experience with handling logging residues. Therefore, further research in this area is recommended to better understand an impact of available equipment and infrastructure on the utilization of mill and logging residues. A decision to utilize additional logging residues by a mill is often affected by additional external factors such as transportation cost or lack of equipment. Additionally, low energy prices, when compared to the cost of producing electricity from logging residues, also limited mill willingness to utilize logging residues because the cost of industrial electricity was substantially lower than electricity produced from logging residues [21,68]. However, availability of logging residues was not significant in woody residue utilization to produce bioenergy. A preliminary geospatial analysis on the availability of logging residues completed by the authors indicated that almost all logging residues in the southern United States could potentially be recovered within a 56 km procurement zone around each mill. Survey responses indicated that mills, on average, were already hauling logging residues over a 72 km. This indicated that an actual economically feasible distance for hauling logging residues was longer and mills did not perceive acquiring logging residues as a limitation. Findings are expected to help decision makers and mill managers identify the factors important to increase woody residue utilization and make informed decisions related to infrastructure facilitating woody residue utilization for bioenergy production. Government policies and programs will be critical to increasing wood-based bioenergy production and policies have been formulated and implemented to support biomass-based bioenergy in the United States. For example, Agricultural Acts of 2008 and 2014, commonly known as Farm Bills of 2008 and 2014, have emphasized bioenergy production and established funding mechanisms aiming to facilitate bioenergy production [70]. The Biomass Crop Assistance Program (BCAP), authorized under the 2008 Farm Bill, established a mechanism to incentivize agricultural producers to establish, cultivate, and harvest biomass for bioenergy. Collection, harvest, storage, and transportation (CHST) matching payment values were established in BCAP for eligible material owners to help support biomass-based bioenergy program. To ensure a continuation of biomass-based bioenergy production, the Agricultural Act of 2014 [71] reauthorized and provided funding of 880 million $ for energy programs established in the 2008 Farm Bill, and expanded Biorefinery Assistance Program and Bio-Preferred Program to include forest products. However, most of these programs are focused on feedstock producers rather than mills. Therefore,

286

R. Pokharel et al. / Biomass and Bioenergy 105 (2017) 278e287

additional federal and state programs will be needed to provide mechanisms incentivizing bioenergy-related capacity and equipment upgrades at mills. Such programs should focus on aspects such as disseminating relevant information, providing technical assistance, offering investment incentives (such as startup incentives, and low-interest rate loans), providing tax incentives, and developing standards for bioenergy production. The development of bioenergy-focused policies addressing efforts to reduce high transportation costs will be pivotal for improving woody residue utilization and increasing bioenergy production. Important parts of forest-based bioenergy supply chains are logging operators and truckers who usually are not included in bioenergy incentive programs. Thus, we also recommend extending incentives and tax benefits to logging operators and truckers to help offset high procurement and transportation costs of recovering logging residues for bioenergy purposes. This study is not without limitations due to missing data. The mail survey collected information about procurement factors such as gate prices and hauling distances for logging residues; however, the number of observations on prices and hauling distances was too small to facilitate an analysis and consequently procurement factors were not included in this study. The inclusion of procurement factors would produce more precise estimates and provide better insights on how procurement factors affected the utilization of woody residues for bioenergy purposes. In addition to economic feasibility, environmental sustainability of procuring logging residues for bioenergy is important. Increased utilization of logging residues is associated with concerns related to soil productivity [16e18]. The benefit of removing logging residues is associated with nutrient losses and accelerated soil compaction and should be examined before procuring logging residues to determine the sustainability and environmental consequences of utilizing logging residues.

5. Conclusions With the increasing need to reduce fossil fuel usage, woody residues have gained a momentum in terms of their potential usability for energy purposes. Historically, mills have utilized woody residues to produce heat, electricity, and chemicals to substitute their energy needs and for sale. Mill residues produced at a mill site have been the most common form of woody residues used to produce bioenergy by mills. However, almost all mill residues have already been utilized for bioenergy and non-energy purposes raising the need to identify viable alternative woody feedstocks, such as logging residues. This study determined the potential for woody residue utilization by measuring the relationship between their utilization at a mill for bioenergy purposes, processing capacity to utilize woody residues, and factors limiting utilization of additional logging residues. Mill processing capacity was one of the main factors driving residue utilization. Higher utilization rates were observed while increasing capacity levels, indicating that large-capacity upgrades will be more effective. On the other hand, high transportation cost of logging residues was negatively related to utilization of woody residues. Moreover, utilization of logging residues to produce electricity was relatively limited and not related with overall utilization of woody residues at the mill. Therefore, biomass policies should focus on assisting mills in increasing their processing capacity, installing upgrades facilitating utilization of woody feedstocks, and lowering transportation cost of logging residues. A further research on factors affecting economic viability of using logging residues to produce bioenergy by mills will help identify strategies facilitating a more cost-effective recovery and economic availability of this feedstock.

Acknowledgement Approved as a publication FO465 of Forest and Wildlife Research Center, Mississippi State University. This material is based upon work that is supported by the National Institute of Food and Agriculture, U.S. Department of Agriculture, McIntire-Stennis project under accession number 1006988. The completed research was also funded by the USDA Cooperative State Research, Education and Extension Service (CSREES) Special Grant for Wood Utilization Research and College of Forest Resources, Mississippi State University.

References [1] U.S. Energy Information Administration (EIA), Annual Energy Outlook 2014 with Projections to 2040, U.S. Energy Information Administration Office of Integrated and International Energy Analysis, Washington, DC, 2014. http:// www.eia.gov/forecasts/aeo/pdf/0383(2014).pdf (Accessed April 9, 2015). [2] J.L. Conrad, M.C. Bolding, Virginia's woody biomass market: opportunities and implications, South. J. Appl. 35 (2) (2011) 67e72. [3] P. Dargusch, T.N. Maraseni, P. Schmidt, A review of research on forest-related environmental markets (including certification schemes, bioenergy, carbon markets and other ecosystem services), 2010, Plant Sci. Rev. 5 (4) (2011) 125. [4] M. Ruth, T. Mai, E. Newes, A. Aden, E. Warner, C. Uriarte, D. Inman, T. Simpkins, A. Argo, Transportation Energy Futures Series: Projected Biomass Utilization for Fuels and Power in a Mature Market, National Renewable Energy Laboratory, Golden, Colorado, 2013. http://www.osti.gov/scitech/biblio/1219930 (Accessed April 9, 2015). [5] U.S. Department of Energy, US Billion-ton Update: Biomass Supply for a Bioenergy and Bioproducts Industry, ORNL/TM-2011/224, Oak Ridge National Laboratory, Oak Ridge, Oak Ridge, Tennessee, 2011. http://www.osti.gov/ scitech/biblio/1023318 (Accessed January 12, 2014). [6] G. Perez-Verdin, D.L. Grebner, C. Sun, I.A. Munn, E.B. Schultz, T.G. Matney, Woody biomass availability for bioethanol conversion in Mississippi, Biomass Bioener 33 (3) (2009) 492e503. [7] F. Aguilar, S. Bratkovich, K. Fernholz, A. Garrard, R. Grala, L. Leightley, W. Martin, I. Munn, The Status of and Opportunities for Business Clustering within the Forest Products Sector in the U.S, U.S. Endowment for Forestry and Communities, Inc, Greenville, South Carolina, 2009. http://www. usendowment.org/images/BS_Cluster_Full_report.pdf (Accessed April 9, 2015). [8] S. Riffell, J. Verschuyl, D. Miller, T.B. Wigley, Biofuel harvests, coarse woody debris, and biodiversity e a meta-analysis, For. Ecol. Manage 261 (4) (2011) 878e887. [9] O. Joshi, D.L. Grebner, P.N. Khanal, Status of urban wood-waste and their potential use for sustainable bioenergy use in Mississippi, Resour. Conserv. Recycl 102 (2015) 20e26. [10] K. Skog, J. Barbour, M. Buford, D. Dykstra, P. Lebow, P. Miles, B. Perlack, B. Stokes, Forest-based biomass supply curves for the United States, J. Sustain. 32 (1/2) (2013) 14e27. [11] R.D. Perlack, L.L. Wright, A.F. Turhollow, R.L. Graham, B.J. Stokes, D.C. Erbach, Biomass as Feedstock for a Bioenergy and Bioproducts Industry: the Technical Feasibility of a Billion-ton Annual Supply, USDA Forest Service, Oak Ridge National Laboratory, Oak Ridge, Tennessee, 2005. http://energy.gov/sites/ prod/files/2014/04/f14/final_billionton_vision_report2.pdf (Accessed July 1, 2013). [12] R.C. Abt, K.L. Abt, Potential impact of bioenergy demand on the sustainability of the southern forest resource, J. Sustain. 32 (1/2) (2013) 175e194. [13] O. Joshi, D.L. Grebner, I.A. Munn, S.C. Grado, R.K. Grala, A. Hussain, Factors influencing utilization of woody biomass from wood processing facilities in Mississippi, For. Prod. J. 64 (1/2) (2014) 64e71. [14] A.W. Garrard, L. Leightley, Characterizing Wood Waste from Wood Products Companies in North Mississippi, Research Report, Forest and Wildlife Research Center, Mississippi State University, 2005. Mississippi State, MS. [15] C.S. Galik, R. Abt, Y. Wu, Forest biomass supply in the southeastern United States - implications for industrial roundwood and bioenergy production, J. For. 107 (2) (2009) 69e77. [16] A.M. Evans, R.T. Perschel, B.A. Kittler, Overview of forest biomass harvesting guidelines, J. Sustain. 32 (1/2) (2013) 89e107. [17] J. Gan, Supply of biomass, bioenergy, and carbon mitigation: method and application, Energy Policy 35 (12) (2007) 6003e6009. [18] L. Johnson, B. Lippke, E. Oneil, Modeling biomass collection and woods processing life-cycle analysis, For. Prod. J. 62 (4) (2012) 258. [19] D. Abbas, D. Current, M. Phillips, R. Rossman, H. Hoganson, K.N. Brooks, Guidelines for harvesting forest biomass for energy: a synthesis of environmental considerations, Biomass Bioener 35 (11) (2011) 4538e4546. [20] D.L. Grebner, R.K. Grala, O. Joshi, G. Perez-Verdin, Physical and economic aspects to assessing woody biomass availability for bioenergy production and related supply constraints, in: S.D. Eksioglu, S. Rebennack, P.M. Pardalos (Eds.), Handbook of Bioenergy: Bioenergy Supply Chain - Models and Applications,

R. Pokharel et al. / Biomass and Bioenergy 105 (2017) 278e287 Springer International Publishing, Cham, 2015, pp. 299e321. [21] J. Gan, C.T. Smith, Availability of logging residues and potential for electricity production and carbon displacement in the USA, Biomass Bioener 30 (12) (2006) 1011e1020. [22] D.O. Hall, Cooling the greenhouse with bioenergy, Nature 353 (1991) 11e12. [23] J. Cook, J. Beyea, Bioenergy in the United States: progress and possibilities, Biomass Bioener 18 (6) (2000) 441e455. [24] N. Shabani, S. Akhtari, T. Sowlati, Value chain optimization of forest biomass for bioenergy production: a review, Renew. Sustain. Energy Rev. 23 (2013) 299e311. [25] C. Zygarlicke, J. Hurley, T. Auich, B. Folkedahl, J. Strege, N. Patel, M. Swanson, C. Martin, E. Olson, B. Oster, EERC center for biomass utilization 2008e2010: phases IeIII, Energy & Environmental Research Center, University of North Dakota, Grand Forks, North Dakota, 2015. http://www.osti.gov/scitech/biblio/ 1206635 (Accessed March 5, 2016). [26] T. Tahvanainen, P. Anttila, Supply chain cost analysis of long-distance transportation of energy wood in Finland, Biomass Bioener 35 (8) (2011) 3360e3375. [27] J. Wu, J.X. Wang, J. McNeel, Economic modeling of woody biomass utilization for bioenergy and its application in central Appalachia, USA, Can. J. For. Res. 41 (1) (2011) 165e179. [28] G. Shivan, K. Potter-Witter, An examination of Michigan's logging sector in the emerging bioenergy market, For. Prod. J. 61 (6) (2011) 459e465.
rez, M.P. Morales, P. Mun ~ oz, M.A. Mendívil, Biomass logistics: [29] J.A. Ruiz, M.C. Jua financial & environmental costs. Case study: 2 MW electrical power plants, Biomass Bioener 56 (2013) 260e267. [30] J.B. Cutshall, W.D. Greene, S.A. Baker, Transpirational drying effects on energy and ash content from whole-tree southern pine plantation chipping operations, South. J. Appl. 37 (2013) 133e139. [31] W.D. Greene, J.B. Cutshall, C.C. Dukes, S.A. Baker, Improving woody biomass feedstock logistics by reducing ash and moisture content, BioEnergy Res. 7 (2014) 816e823. [32] J. Nurmi, The storage of logging residue for fuel, Biomass Bioener 17 (1) (1999) 41e47. [33] J.L. Conrad, R.L. Smith, W.M. Aust, M.C. Bolding, Wood-to-energy expansion, forest ownership changes, and mill closure: consequences for U.S. South's wood supply chain, For. Policy Econ 12 (6) (2010) 399e406. [34] M. Alvarez, The State of America's Forests, Bethesda, Maryland, 2007. [35] S.N. Oswalt, W.B. Smith, US Forest Resource Facts and Historical Trends, US Dept. of Agriculture, Forest Service, Washington, DC, 2014. [36] T.A. Hagadone, R.K. Grala, Business clusters in Mississippi's forest products industry, For. Policy Econ 20 (2012) 16e24. [37] M. Howell, T.G. Johnson, J.W. Bentley, Forest Inventory and Analysis: Fiscal Year 2012 Business Report, Washington, DC, 2012. [38] S.N. Oswalt, Mississippi's Forests, 2013, Asheville, North Carolina, 2015, http:// www.srs.fs.usda.gov/pubs/48754 (Accessed April 9, 2015). [39] USDA National Agricultural Statistics Service, Agricultural conservation and forestry statistics, in: 2012 Agric. Stat. Annu, United States Department Of Agriculture National Agricultural Statistics Service, Washington, DC, 2012. XII1 to XII-30. [40] J. Poudel, I.A. Munn, J.E. Henderson, Economic contributions of wildlife watching recreation expenditures (2006 & 2011) across the U.S. south: an input-output analysis, J. Outdoor Recreat. Tour. 17 (2017) 93e99. [41] J. Poudel, J.E. Henderson, I.A. Munn, Economic contribution of hunting expenditure to the southern United States of America, Int. J. Environ. Stud. 73 (2) (2016) 236e254. [42] R.P. Dahal, J.E. Henderson, I.A. Munn, Forest products industry size and economic multipliers in the U.S. South, For. Prod. J. 65 (7/8) (2015) 372e380. [43] J. Prestemon, J. Pye, J. Barbour, G.R. Smith, P. Ince, C. Steppleton, W. Xu, U.S. Wood-Using Mill Locations - 2005, Data Tools, Washington, DC, 2010, http:// www.srs.fs.usda.gov/econ/data/mills/ (Accessed January 1, 2013). [44] D.A. Dillman, J.D. Smyth, L.M. Christian, Internet, Mail, and Mixed-mode Surveys: the Tailored Design Method, third ed., John Wiley & Sons, New Jersey, NJ, 2009. [45] W.W. Daniel, Applied Nonparametric Statistics, first ed., PWS-Kent Pub, Boston, MA, 1978. [46] C. Heij, P. De Boer, P.H. Franses, T. Kloek, H.K. Van Dijk, Econometric Methods with Applications in Business and Economics, first ed., University Press, Oxford, UK, 2004. [47] W.H. Greene, Econometric Analysis, sixth ed., Prentice hall, New Jersey, 2008. [48] D.N. Gujarati, D.C. Porter, Basic Econometrics, fifth ed., The McGraw-Hill Companies, New York City, NY, 2008. [49] S.R. Gruchy, D.L. Grebner, I.A. Munn, O. Joshi, A. Hussain, An assessment of nonindustrial private forest landowner willingness to harvest woody biomass

[50] [51]

[52]

[53]

[54]

[55]

[56]

[57]

[58]

[59]

[60]

[61]

[62]

[63]

[64]

[65]

[66]

[67]

[68]

[69]

[70] [71]

287

in support of bioenergy production in Mississippi: a contingent rating approach, For. Policy Econ 15 (2012) 140e145. M.F. Smidt, Y. Zhang, B. Zheng, Financial assessment of logging residue collection in the southeast United States, For. Prod. J. 62 (1) (2012) 10e17. S.T. Grushecky, D.W. McGill, J. Wang, Influence of site characteristics and costs of extraction and trucking on logging residue utilization in southern West Virginia, For. Prod. J. 57 (7/8) (2007) 63e67. P.N. Khanal, D.L. Grebner, I.A. Munn, S.C. Grado, R.K. Grala, J.E. Henderson, M.K. Measells, Nonindustrial private forest landowner beliefs toward climate change and carbon sequestration in the Southern United States, J. For. 114 (5) (2016) 524e531. F.X. Aguilar, N. Song, S. Shifley, Review of consumption trends and public policies promoting woody biomass as an energy feedstock in the US, Biomass Bioenergy 35 (2011) 3708e3718. R. Zamora-Cristales, J. Sessions, G. Murphy, K. Boston, Economic impact of truck-machine interference in forest biomass recovery operations on steep terrain, For. Prod. J 63 (5/6) (2013) 162. R. Zamora-Cristales, J. Sessions, K. Boston, G. Murphy, Economic optimization of forest biomass processing and transport in the Pacific Northwest USA, For. Sci. 61 (2) (2015) 220e234. €ki, G. Mery, M. Holopainen, J. Hyyppa €, L.-M. Vaario, K. Yrj€ €, ImL. Hetema ala plications of technological development to forestry, in: For. Soc. e Responding to Glob. Drivers Chang, IUFRO (International Union of Forestry Research Organizations) Secretariat, Vienna, Austria, 2010, pp. 157e181. Z. Guo, D.G. Hodges, T.M. Young, Woody biomass policies and location decisions of the woody bioenergy industry in the southern United States, Biomass Bioener 56 (2013) 268e273. €, C.S. Goh, M. Junginger, M. Cocchi, D. Marchal, D. Thr€ an, C. Hennig, J. Heinimo L. Nikolaisen, P.-P. Schouwenberg, D. Bradley, R. Hess, J. Jacobson, L. Ovard, M. Deutmeyer, Wood pellet market and trade: a global perspective, Biofuels, Bioprod. Biorefining 7 (2013) 24e42. P. Dwivedi, E. Johnson, D. Greene, S. Baker, Tracking economic and environmental indicators of exported wood pellets to the United Kingdom from the southern United States: lessons for policy? BioEnergy Res. (2016) 1e10. U.S. Energy Information Administration (EIA), UK's Renewable Energy Targets Drive Increases in U.S. Wood Pellet Exports, Washington, DC, 2015, http:// www.eia.gov/todayinenergy/detail.php?id¼20912 (Accessed January 31, 2017). U.S. Industrial Pellet Association (USIPA), Wood Supply Market Trends in the US South 1995-2015: Overview, Richmond, Virginia, 2015, http://www. theusipa.org/MarketTrends/overview.htm (Accessed January 31, 2017). P. Dwivedi, R. Bailis, M. Khanna, Is use of both pulpwood and logging residues instead of only logging residues for bioenergy development a viable carbon mitigation strategy? BioEnergy Res. 7 (2014) 217e231. U.S. Department of Agriculture, Partners Celebrate First Wood-to-jet-fuel Commercial Flight, Washington, DC, 2016. https://www.usda.gov/wps/ portal/usda/usdahome?contentid¼2016/11/0241.xml&contentidonly¼true (Accessed February 1, 2017). Northwest Advanced Renewables Alliance (NARA) and Washington State University, First Flight Using Biojet Made from Wood, Pullman, Washington, 2016, https://nararenewables.org/first-flight-using-biojet-made-from-wood/ (Accessed February 1, 2017). A. Milbrandt, C. Kinchin, R. McCormick, The Feasibility of Producing and Using Biomass-based Diesel and Jet Fuel in the United States, U.S. Department of Energy, Washington, DC, 2013. http://www.nrel.gov/docs/fy14osti/58015.pdf (Accessed February 1, 2017). G. Marrs, R. Zamora-Cristales, J. Sessions, Forest biomass feedstock cost sensitivity to grinding parameters for bio-jet fuel production, Renew. Energy 99 (2016) 1082e1091. Federal Energy Management Program, Biomass for Electricity Generation, Biomass Electr, Gener, Washington, DC, 2011. http://www.wbdg.org/ resources/biomasselectric.php (Accessed April 8, 2015). International Renewable Energy Agency (IRENA), Renewable Energy Technologies: Cost Analysis Series, Biomass Power Gener, Abu Dhabi, UAE, 2012, https://www.irena.org/DocumentDownloads/Publications/RE_Technologies_ Cost_Analysis-SOLAR_PV.pdf (Accessed April 9, 2015). S. Radhakrishnan, J.O. Paz, F. Yu, S. Eksioglu, D.L. Grebner, Assessment of potential capacity increases at combined heat and power facilities based on available corn stover and forest logging residues, Energies 6 (9) (2013) 4418e4428. N.D. Hamilton, 2014 Farm Bill: lessons in patience, politics, and persuasion, Drake J. Agric. Law 19 (2014) 1e37. Agricultural Act of 2014, H.R.2642, Pub. L. 113e79, 128 STAT. 649, 133th Cong. xx 113-179, 2014.