BioEnergy Research

, Volume 7, Issue 1, pp 217–231

Is Use of Both Pulpwood and Logging Residues Instead of Only Logging Residues for Bioenergy Development a Viable Carbon Mitigation Strategy?

Authors

    • Energy Biosciences InstituteUniversity of Illinois at Urbana-Champaign
  • Robert Bailis
    • School of Forestry & Environmental StudiesYale University
  • Madhu Khanna
    • Energy Biosciences InstituteUniversity of Illinois at Urbana-Champaign
Article

DOI: 10.1007/s12155-013-9362-z

Cite this article as:
Dwivedi, P., Bailis, R. & Khanna, M. Bioenerg. Res. (2014) 7: 217. doi:10.1007/s12155-013-9362-z
  • 296 Views

Abstract

This study adopts an integrated life-cycle approach to assess overall carbon saving related with the utilization of wood pellets manufactured using pulpwood and logging residues for electricity generation. Carbon sequestered in wood products and wood present in landfills and avoided carbon emissions due to substitution of grid electricity with the electricity generated using wood pellets are considered part of overall carbon savings. Estimated value of overall carbon saving is compared with the overall carbon saving related to the current use of pulpwood and logging residues. The unit of analysis is a hectare of slash pine (Pinus elliottii) plantation in southern USA. All carbon flows are considered starting from forest management to the decay of wood products in landfills. Exponential decay function is used to ascertain carbon sequestered in wood products and wood present in landfills. Non-biogenic carbon emissions due to burning of wood waste at manufacturing facilities, wood pellets at a power plant, and logging residues on forestlands are also considered. Impacts of harvest age and forest management intensity on overall carbon saving are analyzed as well. The use of pulpwood for bioenergy development reduces carbon sequestered in wood products and wood present in landfills (up to 1.6 metric tons/ha) relative to a baseline when pulpwood is used for paper making and logging residues are used for manufacturing wood pellets. Avoided carbon emissions because of displacement of grid electricity from the electricity generated using wood pellets derived from pulpwood fully compensate the loss of carbon sequestered in wood products and wood present in landfills. The use of both pulpwood and logging residues for bioenergy development is beneficial from carbon perspective. Harvest age is more important in determining overall carbon saving than forest management intensity.

Keywords

Bioenergy developmentCarbon sequestration in wood productsElectricity generationPine plantationsSouthern USAWood pellets

Introduction

The Energy Independence and Security Act of 2007 has set a production target of 136.3 billion liters of biofuels by 2022 out of which 60.5 billion liters will be produced in the form of cellulosic biofuels [1]. It is also projected that the peak capacity of biomass-based electricity generation will increase from 11.5 to 49.3 billion kilowatt hour between 2010 and 2035 countrywide because of incentives announced by federal and state governments [2]. It is quite likely that the biomass obtained from forestlands will play a critical role in meeting the projected demand for biomass feedstocks towards bioenergy development nationwide [3].

Several studies have analyzed carbon benefits related with the use of forest biomass for bioenergy development. Dwivedi et al. [4] reported that the use of wood pellets for electricity generation in Florida could reduce greenhouse gas (GHG) emissions by 82 % relative to a unit of electricity produced from coal. In another study, Dwivedi et al. [5] found that the use of ethanol derived from slash pine (Pinus elliottii) wood could save 76 % of GHG emissions relative to gasoline over the average life-span of a small passenger car in the USA. Steele et al. [6] estimated that the use of bio-oil, manufactured using fast pyrolysis technology from southern pine (Pinus taeda) biomass, could reduce GHG emissions by 70 % over residual fuel oil. Katers et al. [7] found that the use of wood pellets derived from forest mill residues for domestic heating in Wisconsin could save about 27 % of GHG emissions relative to natural gas. A review of existing studies reveals two methodological similarities: (a) overall saving in GHG emissions due to use of bioenergy products is based on wood obtained from only one harvest cycle and (b) simultaneous impacts of a change in harvest age and forest management intensity on GHG emissions are not considered. It is important to analyze impacts of harvest age and forest management intensity over a long planning horizon as these parameters significantly determine total wood availability, and therefore availability of different bioenergy products.

A few existing studies consider utilization of logging residues for bioenergy development [811]. Logging residues are mostly comprised of non-merchantable portion of a harvested tree along with branches and tops. Currently, logging residues are either burned or left in open fields by forestland owners as markets for logging residues are practically absent. It is generally thought that logging residues could be used for bioenergy development to reduce GHG emissions [3] and increase profitability of forestland owners [12]. Additionally, there is an implicit assumption that the use of logging residues for bioenergy development will not affect traditional forest-based industries which are dependent on sawtimber, chip-n-saw, and pulpwood for manufacturing various wood products like lumber, plywood, oriented structural board (OSB), and paper.

However, evidence suggests that the production of wood pellets from pulpwood is increasing in the USA [13]. It is likely that the present trend of utilizing pulpwood for bioenergy development will continue in the foreseeable future driven by rising demand for different bioenergy products at regional, national, and global levels [14]. Utilizing pulpwood for bioenergy development could affect long-term carbon sequestered in wood products and wood present in landfills [15]. Existing studies which quantify carbon benefits of forest biomass-based bioenergy development ignore this critical aspect as they mostly focus on carbon emissions avoided due to displacement of energy products derived from fossil fuels [47]. A few studies which do consider carbon sequestered in wood products and wood present in landfills do not consider impact of multiple harvest cycles on the dynamics of carbon sequestered in wood products and wood present in landfills [16, 17]. Some studies that do consider multiple harvest cycles only report carbon sequestered in wood products and wood present in landfills for the first 100 to 200 years [18, 19]. Furthermore, no study, to the best of our knowledge, has yet analyzed the impact of diversion of pulpwood for bioenergy development, instead for paper production, on carbon sequestered in wood products and wood present in landfills. An understanding about this alternate usage is critical as carbon stored in wood products and wood present in landfills is a major percentage of total carbon sequestered by the forestry sector. For example, the forestry sector sequestered about 159 million metric tons of carbon in 2005, of which 36.5 % (58 million metric tons) was sequestered in wood products and wood present in landfills [20].

This study adopts an integrated life-cycle approach to assess the impact of utilizing wood pellets derived from only logging residues or both logging residues and pulpwood, as a feedstock for electricity generation on overall carbon savings—sum of carbon sequestered in wood products and wood present in landfills and avoided carbon emissions due to displacement of grid electricity with the electricity generated using wood pellets. The focus of this study is on the southern USA as this region supplied about 62 % (275 million cubic meters) of nation’s total forest removals in 2006 [21]. Moreover, about 30 % of national wood pellet production capacity in 2009 was located in this region [13]. Slash pine was selected as a representative species. Slash pine is a major commercial softwood tree species of the southern USA. The longleaf slash pine forest-type group covered 5.3 million hectares in 2007 [21]. We estimated overall carbon savings for four different cases (Table 1). Cases LEFT-LR and BURN-LR were selected as two separate baselines as forestland owners in southern region opt for different practices to remove logging residues. Cases ENE-LR and ENE-LR&PW were compared with cases LEFT-LR and BURN-LR individually to ascertain differences in overall carbon savings. Case ENE-LR&PW was also compared with case ENE-LR to assess the impact of diversion of pulpwood for electricity generation instead of paper manufacturing on overall carbon savings. Two scenarios of forest management (intensive and non-intensive) and 45 different harvest ages (starting from age 6 to age 50) were also selected to assess impacts of forest management intensity and harvest age on overall carbon savings. Unlike non-intensive forest management, intensive forest management includes herbicide (at the time of site preparation) and fertilizer (2nd and 12th year of plantation) applications.
Table 1

Details of selected cases. First and second cases are considered as two separate baselines

Case name

Sawtimber

Chip-n-saw

Pulpwood

Logging residues

LEFT-LR

Lumber

OSB

Paper

Left on the ground

BURN-LR

Lumber

OSB

Paper

Burn on the ground

ENE-LR

Lumber

OSB

Paper

Wood pellets

ENE-LR&PW

Lumber

OSB

Wood pellets

Wood pellets

OSB oriented strand board

Methods

Overall carbon flows are based on the schematic shown in Fig. 1. Parameters used to operationalize each case were derived from different sources (Table 2). We used a popular growth and yield model of slash pine [22] to ascertain availability of four timber products (logging residues, pulpwood, chip-n-saw, and sawtimber) under both forest management scenarios at different harvest ages. The availability of logging residues at any harvest age was calculated as the difference between total merchantable biomass available and biomass present in sawtimber, chip-n-saw, and pulpwood plus 20 % of all biomass present in sawtimber, chip-n-saw, and pulpwood [23]. Additional 20 % biomass was added as a proxy for biomass available in branches and tops [23]. We assumed that a hectare of slash pine plantation will be harvested multiple times at a constant harvest age for the next 500 years starting from the first harvest. This simulation period was selected to better understand overall carbon dynamics. The model was operationalized in MS Excel 2010© using VBA.
https://static-content.springer.com/image/art%3A10.1007%2Fs12155-013-9362-z/MediaObjects/12155_2013_9362_Fig1_HTML.gif
Fig. 1

Selected systemic boundary for case ENE-LR under intensive forest management. Sources of carbon emissions are highlighted. We have opted for a closed system boundary as direct and indirect impacts of regional timber markets are not considered in this study

Table 2

Parameters used for modeling carbon flow. Non-biogenic carbon emissions due to burning of logging residues on forestlands contain N2O and CH4 (100-year global warming potential). We used same value for ascertaining quantities of non-biogenic GHG emission released at the time of burning green wood (wood waste and logging residues) due to lack of sufficient information. All values of carbon emissions can be converted into CO2e by multiplying 44/12

General assumption

 Carbon present in harvested timber products (by mass) (percent)—25

Parameters used for plantation management

 Carbon emissions with intensive forest management before age 12 (metric tons/hectare)—0.66 [4]

 Carbon emissions with intensive forest management after age 12 (metric tons/hectare)—1.3 [4]

 Carbon emissions with non-intensive forest management (metric tons/hectare)—0.60 [4]

 Non-biogenic carbon emissions due to burning of logging residues on forestlands (gram/kilogram)—143.12 [26]

 Half-life of logging residues (years)—5.7 [27]

Parameters used for converting timber products into wood products

 Conversion efficiency of lumber production from sawtimber (percent)—64.5 [24]

 Conversion efficiency of OSB production from chip-n-saw (percent)—64.5 [24]

 Conversion efficiency of paper production from pulpwood (percent)—58 [24]

 Conversion efficiency of wood pellet production (percent)—80 [24]

 Carbon emissions related with finished lumber production (gram/kilogram)—25.83 [28]

 Carbon emissions related with OSB production (gram/kilogram)—115 [29]

 Carbon emissions related with paper production (gram/kilogram)—369 [30]

 Carbon emissions related with wood pellet production (gram/kilogram)—42.46 [4]

 Non-biogenic carbon emissions due to wood waste burning in controlled conditions (gram/kilogram)—9.38 [31]

Parameters used for modeling carbon flow during the use phase of wood products

 Half-life of lumber (years)—100 [24]

 Half-life of OSB (years)—30 [24]

 Half-life of paper (years)—2.6 [24]

 Recycling rate of lumber (%)—25 [24]

 Recycling rate of OSB (%)—25 [24]

 Recycling rate of paper (%)—50 [24]

 Carbon intensity of a unit of grid electricity (kilogram/kilowatt hour)—0.212 [28]

 Conversion efficiency of a 100 MW power plant (percent)—31.6 [32]

 Calorific value of wood pellet (megajoule/kilogram)—18.5

Parameters used for modeling carbon flow in the landfill

 Lumber to landfill (percent)—76 [24]

 OSB to landfill (percent)—67 [24]

 Paper to landfill (percent)—34 [24]

 Lumber changing to non-degradable portion in landfill (percent)—77[24]

 OSB changing to non-degradable portion in landfill (percent)—77 [24]

 Paper changing to non-degradable portion in landfill (percent)—44 [24]

 Lumber changing to degradable portion in landfill (percent)—23 [24]

 OSB changing to degradable portion in landfill (percent)—23 [24]

 Paper changing to degradable portion in landfill (percent)—56 [24]

 Half-life of degradable portion in landfill (years)—14 [24]

The following steps were used to ascertain quantity of carbon sequestered in wood products and wood present in landfills under the case LEFT-LR. For the first harvest cycle at the very first year of simulation period, we allocated carbon emissions from plantation management to timber products based on their mass percentage. We did not allocate any plantation management-related carbon emissions to logging residues under LEFT-LR and BURN-LR cases. We ascertained carbon emissions related with manufacturing of different wood products—lumber, OSB, and paper (Table 2). We assumed that wood waste produced while manufacturing different wood products was burned within the manufacturing facility itself. We estimated non-biogenic carbon emissions due to burning of wood waste at manufacturing facilities (Table 2). We subtracted carbon emissions from plantation management, manufacturing of wood product, and burning of wood waste from the carbon sequestered in different wood products to estimate net carbon present in different wood products before entering the use phase.

We used an exponential decay function [Eq. 1] to estimate carbon present in a given wood product (CWoodProt) after t years of use [24]. We subtracted carbon present in a wood product for any two consecutive years (CWoodProt and CWoodProt − 1) to estimate availability of carbon for subsequent phases. This available carbon was sequentially channeled to different phases using parameters reported in Table 2. The difference between available carbon and the carbon moved into the next phase was considered as an emission to the atmosphere. Carbon released due to the decay of logging residues between any two consecutive years was not channeled to the next phase. It was considered as an emission to the atmosphere. We used Eq. 1 to ascertain carbon emissions (in the form of methane) from the degradable portion of carbon sequestered in landfills as well. We repeated these steps for other harvest cycles present within the simulation period. Finally, we added carbon sequestered for all harvest cycles to estimate quantities of carbon sequestered in wood products and wood present in landfills for each year of simulation period at a given harvest age.
$$ {C}_{\mathrm{WoodPro}}^t={e}^{\left.\hbox{--} t\times \ln (2)/\mathrm{half}-\mathrm{life}\ \mathrm{of}\ \mathrm{the}\ \mathrm{wood}\ \mathrm{product}\right)} $$
(1)

We adjusted the above procedure for each case. For the case BURN-LR, we did not consider carbon sequestered in logging residues at all. We first estimated carbon emissions related with the burning of logging residues by multiplying the quantity of available logging residues and a suitable emission factor (Table 2). We divided total carbon emissions with the harvest age to ascertain average annual carbon emissions. Then, we calculated cumulative average carbon emissions for every year present in the simulation period. Finally, we subtracted cumulative average carbon emissions related with the burning of logging residues from the carbon sequestered in wood products and wood present in landfills to estimate overall carbon savings for all years present within the simulation period.

For cases ENE-LR and ENE-LR&PW, we determined avoided carbon emissions (EWP-avoid) due to the use of wood pellets for electricity generation and supply of this electricity to the grid using Eq. 2:
$$ {E}_{\mathrm{WP}\hbox{--} \mathrm{avoid}}\kern0.5em =\kern0.5em \left[{B}_{\mathrm{WP}}\times \mathrm{CV}\times \mathrm{EF}\times \mathrm{CI}\times \left(1/3.6\right)\times 0.001\right]\kern0.5em -\kern0.5em \left[\left[{E}_{\mathrm{WP}-\mathrm{pro}}\right]+\left[{E}_{\mathrm{WP}-\mathrm{burn}}\right]\right]\times {Q}_{\mathrm{WP}}\kern0.5em -{E}_{\mathrm{WW}-\mathrm{burn}}\kern0.5em -\kern0.5em {E}_{\mathrm{WP}-\mathrm{bio}} $$
(2)
where BWP is wood pellet availability (in kilogram per hectare) at 5 % moisture content at a given harvest age, CV is calorific value of wood pellets in megajoule per kilogram, EF is efficiency of converting heat obtained from burning wood pellets to electricity, and CI is carbon intensity of grid electricity (in kilogram C per kilowatt hour). The unit of EWP-avoid is metric tons of avoided carbon emissions per hectare of forestland at a given harvest age. EWP-pro is carbon emissions related with the production of wood pellets and EWP-burn is total quantity of non-biogenic carbon emission generated due to burning of wood pellets in controlled settings. The units of EWP-pro and EWP-burn were metric tons of carbon emissions per metric ton of wood pellets. Parameter QWP is the total quantity of wood pellets produced in metric tons per hectare depending upon the biomass availability at a given harvest age and biomass-to-wood pellet conversion efficiency. EWW-burn is the total quantity of non-biogenic carbon emissions (metric tons) generated due to burning of wood waste in controlled settings. EWP-bio reflects carbon emissions allocated to logging residues at the plantation site (metric tons/hectare). We divided total avoided carbon emissions with the corresponding harvest age to ascertain average annual avoided carbon emissions. Then, we calculated cumulative average avoided carbon emissions for every year present in the simulation period at a given harvest age. We did not consider carbon sequestered in logging residues for cases ENE-LR and ENE-LR&PW. Similarly, we did not consider carbon sequestered in pulpwood for the case ENE-LR&PW. Finally, we added cumulative average avoided carbon emissions to the carbon sequestered in wood products and wood present in landfills to estimate overall carbon savings for all years during the simulation period.

Results

The availability of large-diameter (sawtimber and chip-n-saw) timber products was smaller in initial plantation years relative to small-diameter (pulpwood and logging residues) timber products. However, availability of large-diameter timber products increased as trees gained girth and height with time (Fig. 2). The combined availability of pulpwood and logging residues reached to a maximum value (179.6 and 172.8 metric tons/ha at plantation ages 26 and 27 years under intensive and non-intensive forest management scenarios, respectively) and then started to decrease as large-diameter classes replaced small-diameter classes. Total biomass availability was higher under intensive than non-intensive forest management scenario for all years when only logging residues or both pulpwood and logging residues were considered as a feedstock. Figure S1 shows percentage contribution of different timber products towards the combined weight of all timber products. These percentages were used for allocating carbon emissions related to plantation management to different timber products.
https://static-content.springer.com/image/art%3A10.1007%2Fs12155-013-9362-z/MediaObjects/12155_2013_9362_Fig2_HTML.gif
Fig. 2

Availability of different timber products under intensive and non-intensive forest management scenarios. B(d,t) = timber availability with diameter-at-breast height ≥d centimeters to a merchantable diameter t centimeters for sawtimber, chip-n-saw, and pulpwood is taken as B(25.4,20.32), B(20.32,15.24), and B(10.16,5.08), respectively. Site index is about 21 m at 25th year of a slash pine plantation with initial plantation density as 1,235 seedlings/ha

Carbon emissions due to burning of logging residues (under case BURN-LR) or displacement of grid electricity (under cases ENE-LR and ENE-LR&PW) were dependent on harvest age (Fig. 3). Net carbon emissions were greater under intensive than non-intensive forest management scenario due to higher feedstock availability under intensive forest management. The use of both pulpwood and logging residues for wood pellet production under both forest management scenarios avoided more carbon emissions than the use of logging residues only because of higher availability of biomass under ENE-LR&PW than ENE-LR case. Net carbon emissions showed an increasing trend with respect to harvest age and then became almost constant (or decreased very little) at higher harvest ages. However, average annual net carbon emissions were high at initial harvest ages and began to decline with a rise in harvest age. Figure 4 shows cumulative carbon emissions for 20-, 30-, and 50-year harvest cycles due to burning of logging residues or displacement of grid electricity. As expected, the values of cumulative carbon emissions rise linearly with time.
https://static-content.springer.com/image/art%3A10.1007%2Fs12155-013-9362-z/MediaObjects/12155_2013_9362_Fig3_HTML.gif
Fig. 3

Distribution of net carbon and average annual net carbon emissions. Emissions are positive whereas savings are negative. Carbon emissions are due to burning of logging residues on forestlands. Carbon savings are due to substitution of grid electricity after accounting for carbon emissions due to logging residues/pulpwood production, wood pellet manufacturing, and release of non-biogenic carbon due to burning of wood waste at manufacturing facilities and wood pellets at the power plant. Intensive forest management (I), non-intensive forest management (NI)

https://static-content.springer.com/image/art%3A10.1007%2Fs12155-013-9362-z/MediaObjects/12155_2013_9362_Fig4_HTML.gif
Fig. 4

Distribution of cumulative carbon emissions with respect to harvest age. Carbon emissions are positive whereas savings are negative. Simulation period is 500 years

Figure 5 shows trajectory of carbon sequestered in wood products and wood present in landfills for 20-, 30-, and 50-year harvest cycles under both scenarios of forest management. The carbon sequestered in wood products and wood present in landfills was cyclical, with a period that matched harvest age. The carbon sequestered in wood products and wood present in landfills almost reached a steady state by the end of simulation period. Figure 6 shows distribution of average carbon sequestered in wood products and wood present in landfills at the end of simulation period for all cases under both forest management scenarios along with corresponding harvest ages. Method of moving average was used to estimate quantities of carbon sequestered in wood products and wood present in landfills at the end of simulation period. The period of moving average was equal to the harvest age itself. Carbon sequestered in wood products and wood present in landfills started to increase with a rise in harvest age, reached to a maximum value, and then decreased because of changes in the availability of different timber products with respect to plantation age coupled with their manufacturing and use characteristics. This shows that an increase in harvest age will not necessarily lead to sequestration of more carbon in wood products and wood present in landfills. Carbon sequestered in wood products and wood present in landfills under intensive forest management reached a maximum value at a lower harvest age than non-intensive forest management. The carbon sequestered in wood products and wood present in landfills was always higher under intensive forest management relative to non-intensive forest management. This implies that higher biomass production obtained through the use of herbicides and fertilizers helps in sequestering more carbon in wood products and wood present in landfills even after accounting for carbon emissions related to their production and application.
https://static-content.springer.com/image/art%3A10.1007%2Fs12155-013-9362-z/MediaObjects/12155_2013_9362_Fig5_HTML.gif
Fig. 5

Distribution of carbon sequestered in wood products and wood present in landfills with respect to time. Harvest age (years) is shown in legend. Simulation period is 500 years

https://static-content.springer.com/image/art%3A10.1007%2Fs12155-013-9362-z/MediaObjects/12155_2013_9362_Fig6_HTML.gif
Fig. 6

Average carbon sequestered at the end of the simulation period in wood products and wood present in landfills. Reported values are calculated based on moving averages. Simulation period is 500 years

Carbon sequestered under case LEFT-LR at the end of simulation period was more than cases ENE-LR and ENE-LR&PW for both scenarios of forest management at all harvest ages (Fig. S2). This suggests that leaving logging residues on forestlands increases quantities of carbon sequestered in wood products and wood present in landfills. Carbon sequestered in wood products and wood present in landfills at the end of simulation period under case ENE-LR was higher relative to case BURN-LR starting from 10th harvest age onwards (Fig. S3). However, this was not true under case ENE-LR&PW for all harvest ages. Carbon sequestered in wood products and wood present in landfills at the end of the simulation period under case ENE-LR&PW was lower than case ENE-LR (Fig. 7). This implies that diversion of pulpwood for wood pellet production instead of paper production decreased carbon sequestered in wood products and wood present in landfills. However, the magnitude of this difference went down with an increase in harvest age and became almost negligible at higher harvest ages because of an increase in the availability of large-diameter timber products at higher harvest ages.
https://static-content.springer.com/image/art%3A10.1007%2Fs12155-013-9362-z/MediaObjects/12155_2013_9362_Fig7_HTML.gif
Fig. 7

Difference of carbon sequestered at the end of simulation period in wood products and wood present in landfills between cases ENE-LR&PW and ENE-LR (baseline). Carbon emissions are positive whereas savings are negative. Simulation period is 500 years

Figure 8 shows distribution of overall carbon savings for all cases for harvest ages 20, 30, and 50 years. Overall carbon saved under cases ENE-LR and ENE-LR&PW did not reach a steady state within the simulation period under both forest management scenarios. This suggests that avoided carbon emissions significantly affect magnitude and trajectory of overall carbon savings. The same was true for overall carbon emissions related to case BURN-LR. Figure 9 shows distribution of overall carbon savings for selected harvest ages. Overall carbon saving was higher at lower harvest ages under intensive forest management for case ENE-LR&PW relative to other cases.
https://static-content.springer.com/image/art%3A10.1007%2Fs12155-013-9362-z/MediaObjects/12155_2013_9362_Fig8_HTML.gif
Fig. 8

Distribution of overall carbon savings with respect to time. Harvest age (years) is shown in legend. Simulation period is 500 years

https://static-content.springer.com/image/art%3A10.1007%2Fs12155-013-9362-z/MediaObjects/12155_2013_9362_Fig9_HTML.gif
Fig. 9

Overall carbon savings at different harvest ages at the end of simulation period. Carbon emissions are positive whereas savings are negative. Simulation period is 500 years

Figures S4 and S5 show differences in overall carbon savings for cases ENE-LR&PW and ENE-LR with respect to cases LEFT-LR and BURN-LR, respectively. The overall carbon saving under cases ENE-LR and ENE-LR&PW was always greater than cases LEFT-LR and BURN-LR indicating that the overall carbon saving is always higher when logging residues are utilized for electricity generation than leaving or burning them on forestland. Figure 10 shows the difference in overall carbon saving between cases ENE-LR&PW and ENE-LR. More carbon is saved under case ENE-LR&PW than case ENE-LR especially under intensive forest management scenario. However, this difference decreases with an increase in harvest age. It can be inferred that avoided carbon emissions due to supply of electricity to the grid obtained from wood pellet manufactured from both pulpwood and logging residues instead of just logging residues fully compensates the carbon lost in wood products and wood present in landfills due to the diversion of pulpwood for bioenergy development instead of paper manufacturing. This shows that the diversion of pulpwood for bioenergy development is a reliable source of carbon reduction.
https://static-content.springer.com/image/art%3A10.1007%2Fs12155-013-9362-z/MediaObjects/12155_2013_9362_Fig10_HTML.gif
Fig. 10

Difference of overall carbon savings at the end of simulation period between cases ENE-LR&PW and ENE-LR (baseline). Carbon emissions are positive whereas savings are negative. Simulation period is 500 years

We selected five different simulation periods (100, 200, 300, 400, and 500 years) to ascertain impact of length of simulation period on overall carbon savings. Figure 11 shows the difference in overall carbon savings between cases ENE-LR&PW and ENE-LR for different simulation periods. Overall saving in carbon emissions was proportional to the length of simulation period because quantity of avoided carbon emissions linearly increases with respect to time. This also implies that at as the length of simulation period decreases, net carbon sequestered in wood products and wood present in landfills becomes a significant determinant of overall carbon saving instead of avoided carbon emissions (Fig. 12).
https://static-content.springer.com/image/art%3A10.1007%2Fs12155-013-9362-z/MediaObjects/12155_2013_9362_Fig11_HTML.gif
Fig. 11

Difference of overall carbon savings at the end of different simulation periods between cases ENE-LR&PW and ENE-LR (baseline). Carbon emissions are positive whereas savings are negative

https://static-content.springer.com/image/art%3A10.1007%2Fs12155-013-9362-z/MediaObjects/12155_2013_9362_Fig12_HTML.gif
Fig. 12

Percentage contribution of carbon sequestered in wood products and wood present in landfills towards overall carbon savings at the end of different simulation periods

Discussions and Conclusions

This study assesses the impact of diversion of pulpwood for bioenergy development instead for paper manufacturing on overall carbon savings. We estimated carbon dynamics for two additional cases and compared our original results with them to obtain a better understanding of overall carbon savings. For estimating overall carbon savings, we included carbon sequestered in wood products and wood present in landfills along with avoided carbon emissions due to the use of logging residues (with and without pulpwood) as a feedstock for manufacturing wood pellets, which were then used to generate electricity. We also assessed impact of changes in forest management intensity, harvest age, and simulation period on overall carbon savings. This study clearly indicates that use of pulpwood and logging residues for bioenergy development helps in significant amount of carbon savings under realistic baselines depending upon the length of simulation period.

We acknowledge that validation of our results on carbon sequestered in wood products and wood present in landfills using real data is not possible considering the peculiarity of the problem analyzed. However, we have followed standard guidelines to account for carbon sequestered in wood products and wood present in landfills [24]. Moreover, reported carbon sequestered values are well within ranges reported by other similar studies over first 100 or 200 years of simulation [18, 19]. Additionally, trajectory of carbon sequestered in wood products and wood present in landfills reported is very similar to the existing work [18, 19].

The important findings of this study are summarized below:
  1. (a)

    Intensive forest management sequesters more carbon in wood products and wood present in landfills relative to non-intensive forest management.

     
  2. (b)

    Leaving logging residues will sequester highest amount of carbon in wood products and wood present in landfills relative to burning them without displacing fossil fuels.

     
  3. (c)

    Carbon sequestered in wood products and wood present in landfills reaches to a steady state around 500 years under multiple harvest cycles at a constant harvest age.

     
  4. (d)

    An increase in harvest age will not always lead to more carbon sequestration in wood products and wood present in landfills.

     
  5. (e)

    Harvest age and intensity of forest management play an important role in determining quantities of carbon sequestered in wood products and wood present in landfills.

     
  6. (f)

    Availability of large-diameter timber products is a significant determinant of carbon sequestered in wood products and wood present in landfills.

     
  7. (g)

    Carbon sequestered in wood products and wood present in landfills decreases when pulpwood is diverted to bioenergy development instead when it is used for paper manufacturing. However, this decrease is insignificant over time.

     
  8. (h)

    Avoided carbon emissions play a significant role in determining overall carbon savings especially under a long planning horizon.

     
  9. (i)

    Carbon credits generated due to use of wood pellets derived from pulpwood for electricity generation fully compensates the loss of carbon sequestered in wood products and wood present in landfills due to diversion of pulpwood for wood pellets instead of paper manu-facturing.

     
  10. (j)

    Harvest age is more important in determining overall carbon savings than forest management intensity.

     
  11. (k)

    Simulation period is a key determinant of overall carbon savings. However, trajectory of overall carbon savings does not change a lot with simulation period but magnitude does.

     
  12. (l)

    Overall carbon benefit varies significantly with respect to the selected baseline. Therefore, a caution should be exercised in selecting a baseline while promoting use of either logging residues, pulpwood, or both as a feedstock for bioenergy development, in general or electricity generation, in particular.

     

We have not considered any carbon sequestered in above- or belowground in this study. Aboveground carbon shows a cycling trend with respect to various silvicultural practices and carbon in forest soils attains an equilibrium value in a relatively short time on reforested forestlands [25]. This suggests that net carbon sequestered on those lands which are continuously being used for raising forest plantations might be negligible relative to carbon sequestered in timber products. However, we strongly feel that more information is needed to objectively define the role of carbon sequestered in above- or belowground while calculating carbon intensity of wood-based bioenergy products. Additionally, we have assumed that all logging residues will be collected at the time of harvest. Collection efficiency could be varied along with several other parameters used in this study to assess changes in trajectories and quantities of carbon sequestered in wood products and wood present in landfills.

We only considered major wood products for ascertaining impact of use of small-diameter timber products for electricity generation on carbon sequestered in wood products and wood present in landfills. However, there are many more wood products available in the market having different manufacturing and use characteristics. The developed model in this study can be extended to include additional wood products to get a more precise estimate of carbon sequestered in wood products and wood present in landfills. Additionally, a need exists to evaluate other forest species growing in different regions of the country for understanding the broader impact of utilizing timber products on carbon sequestered in wood products and wood present in landfills. This is particularly true as yields and wood characteristics vary based on management practices adopted by forestland owners, local weather, and soil conditions. Furthermore, we have assumed that all parameters (yields, wood product use characteristics, and GHG intensities of different wood products) will remain constant over the length of selected simulation period. This model can be further improved by including time-dependent values of these and other parameters. Finally, we have not included any transport-related carbon emissions in this study. Including transport-related emissions will further improve the model. We hope that the present study should be able to guide future research in a suitable manner. It is expected that the findings of this study will suitably guide policy deliberations over carbon benefits of utilizing logging residues and pulpwood for bioenergy development in the USA and elsewhere.

Acknowledgments

The authors appreciate the funding support provided by Yale Climate and Energy Institute and Energy Biosciences Institute, University of Illinois at Urbana-Champaign. The authors are also thankful to Drs. Jianbang Gan (TAMU), Wendell Cropper (UF), Douglas Carter (UF), Timothy Martin (UF), and Gary Peter (UF) for their helpful suggestions.

Supplementary material

12155_2013_9362_MOESM1_ESM.pdf (179 kb)
Fig. S1Percentage contribution of timber products towards total weight at different harvest ages. (PDF 178 kb)
12155_2013_9362_MOESM2_ESM.pdf (175 kb)
Fig. S2Difference of carbon sequestered at the end of simulation period in wood products and wood present in landfills for cases ENE-LR and ENE-LR&PW with respect to LEFT-LR (baseline). Carbon emissions are positive whereas savings are negative. Simulation period is 500 years. (PDF 175 kb)
12155_2013_9362_MOESM3_ESM.pdf (174 kb)
Fig. S3Difference of carbon sequestered at the end of simulation period in wood products and wood present in landfills for cases ENE-LR and ENE-LR&PW with respect to BURN-LR (baseline). Carbon emissions are positive whereas savings are negative. Simulation period is 500 years. (PDF 174 kb)
12155_2013_9362_MOESM4_ESM.pdf (174 kb)
Fig. S4Difference of overall carbon savings at the end of simulation period for cases ENE-LR and ENE-LR&PW with respect to LEFT-LR (baseline). Carbon emissions are positive whereas savings are negative. Simulation period is 500 years. (PDF 174 kb)
12155_2013_9362_MOESM5_ESM.pdf (175 kb)
Fig. S5Difference of overall carbon savings at the end of simulation period for cases ENE-LR and ENE-LR&PW with respect to BURN-LR (baseline). Carbon emissions are positive whereas savings are negative. Simulation period is 500 years. (PDF 174 kb)

Copyright information

© Springer Science+Business Media New York 2013