Life cycle greenhouse gas emissions of electricity generation from corn cobs in Ontario, Canada

Correspondence to: David Sanscartier, Environment Division, Saskatchewan Research Council, 125 – 15 Innovation Boulevard,

Saskatoon, Saskatchewan, Canada S7N 2X8. E-mail: [email protected], [email protected]

*Both authors contributed equally to the study and manuscript.

568 © 2014 Society of Chemical Industry and John Wiley & Sons, Ltd

Modeling and Analysis

Life cycle greenhouse gas emissions of electricity generation from corn cobs in Ontario, CanadaDavid Sanscartier,* Environment Division, Saskatchewan Research Council, Saskatoon, Saskatchewan, Canada Goretty Dias,* Faculty of Environment, University of Waterloo, Ontario, CanadaBill Deen and Humaira Dadfar, Department of Plant Agriculture, University of Guelph, Ontario, CanadaIan McDonald, Ontario Ministry of Agriculture, Food and Rural Affairs, University of Toronto, Ontario, CanadaHeather L. MacLean, Department of Civil Engineering, University of Toronto, Ontario, Canada

Received September 13, 2013; revised February 27, 2014, and accepted February 28, 2014 View online April 9, 2014 at Wiley Online Library (wileyonlinelibrary.com); DOI: 10.1002/bbb.1485; Biofuels, Bioprod. Bioref. 8:568–578 (2014)

Abstract: Policy initiatives have motivated a search for environmentally sustainable alternatives to fossil-fuel-based electricity generation. Agricultural residues such as corn cobs may be a suitable feedstock. A life cycle approach was used to estimate the greenhouse gas (GHG) emission impacts associated with the use of pellets produced from corn cobs as the sole fuel for the generation of electricity at a hypothetically retrofi tted coal-fi red generating station in Ontario, Canada. Pellets are compared with current coal and hypothetical natural gas combined cycle (NGCC) facilities. A life cycle model and soil carbon model calibrated for the agricultural region of interest were combined to quantify the GHG emissions of the biomass product system. The corn cob product system’s life cycle emissions (240 g CO2eq kWh–1) are 40% and 80% lower than those of the NGCC and coal product systems, respectively. If corn cobs are left in the fi eld to decompose, some carbon is sequestered in the soil, thus their removal from the fi eld and combustion at the generation station represents a net GHG emission, accounting for 60% of life cycle emissions. In addition to the GHG impacts of com-bustion, removing agricultural residues from fi elds may reduce soil health, increase erosion and affect soil fertility through loss of soil organic carbon and nutrients. Their sustainable use should therefore consider the maintenance of soil fertility over the long-term. Nevertheless, the use of the feedstock in place of coal may provide substantial GHG emissions mitigation. © 2014 Society of Chemical Industry and John Wiley & Sons, Ltd

Keywords: LCA; biomass; coal; climate change mitigation; soil carbon

© 2014 Society of Chemical Industry and John Wiley & Sons, Ltd | Biofuels, Bioprod. Bioref. 8:568–578 (2014); DOI: 10.1002/bbb 569

Modeling and Analysis: GHG Emissions of Electricity Generation from Corn Cobs D Sanscartier et al.

the commercial-scale conversion of lignocellulosic feed-stock, such as corn cobs, into bioethanol in a cost-eff ective and effi cient manner, but research and development eff orts are addressing these constraints.6 In contrast, there have been successful conversions of coal GS to use 100% bio-mass (e.g. the Schiller Station, NH, USA combusts wood chips) and biomass co-fi ring (simultaneously combusting coal and biomass) is a current practice at some GS (e.g. the Drax Power Station, UK combusts a variety of feed-stocks).7 Ontario Power Generation (OPG) is currently retrofi tting its Atikokan, Ontario coal GS to utilize wood pellets, and has conducted assessments that suggest that combustion of agricultural biomass pellets as a sole fuel in its coal GS is technically feasible.

Biomass-based electricity systems may provide GHG emissions reductions relative to fossil-based systems, but do not result in 100% reduction because the production of biomass and its processing into useful forms of energy requires fuel and material inputs that result in emissions. Life cycle assessment (LCA) is a tool that can be used to estimate the GHG emissions of an entire system, from pro-duction of the fuel to its combustion for electricity genera-tion. LCAs of electricity generation with a variety of bio-mass feedstock in retrofi tted coal GS have been published (e.g. Zhang et al.8 modeled electricity generation from wood pellets in Ontario GS), but none have examined corn cobs as a feedstock. Th ere have been several LCAs that have studied GHG emissions and on-farm environmental implications (e.g. soil carbon and nutrient losses, soil ero-sion) of using corn stover as a bioenergy feedstock.9–12 Kim et al.10 used a 17% corn stover removal scenario in DAYCENT to model corn cobs, but modeled the cobs as a fraction of corn stover (rather than cobs directly). Although cobs are a portion of corn stover, the properties of cobs diff er from those of the remainder of the stover,3 which may aff ect the life cycle of bioenergy systems. Th eir study also excluded the end-use of the biomass, and the geographical location of the study (Corn Belt states, USA) diff ered from the location of interest in the current study (i.e. Ontario, Canada). Cherubini and Ulgiati11 included biomass end-use in their study boundaries, but did not specifi cally examine corn cobs as a feedstock. To date, electricity generation from corn cobs has not been exam-ined on a life cycle basis.

Goal

Th is study investigates, on a life cycle basis, the GHG emissions of the production of corn cob pellets and their combustion in a coal GS retrofi tted to use biomass as its

Introduction

The government of Ontario, Canada, has committed to eliminating the use of coal for electricity genera-tion by December 31, 2014, to make progress toward

achieving the provincial greenhouse gas (GHG) emission reduction goal. As a separate initiative, the Canadian government recently regulated coal-based electricity gen-eration with GHG emission performance standards for new units and those that reach the end of their useful life, promoting a shift to lower emitting electricity generation. With the new regulation, coal-fi red generating stations (GS) will be required to achieve the performance standard of 420 g CO2eq/kWh (direct emissions at the generating station).1 Th ese initiatives have led to a search for lower carbon fuels for electricity generation.

Agricultural biomass feedstock, either crop residues or dedicated energy crops, are potential alternative fuel sources for generating electricity. Dedicated cellulosic energy crops have yet to be produced at a large commercial scale and their production would require displacement of crops currently used for food production. On the other hand, agricultural residues are currently and potentially available and would have minimal impact on land acre-age used for food production. Corn cobs, which are by-products of existing commercial corn production, are generally left in the fi eld at harvest of the grain (along with the stalks and leaves, the cobs comprise the corn stover). Cobs have an energy density comparable to other biomass feedstocks (e.g. switchgrass, wood), and off er some relative advantages. Compared to corn leaves and stalks, corn cobs are generally a more uniform and cleaner material (e.g. limited soil material is collected with the cobs at harvest),2 and have a lower ash content.3 Th ese properties are advan-tageous for combustion applications. While the removal of corn cobs impacts nutrient availability, soil carbon, and soil erosion, it appears to have relatively minor eff ects on soil properties compared to removal of corn leaves and stalks.4,5 Th is is, in part, due to lower removal rates of corn cobs; the mass of the leaves and stalks is approximately fi ve times greater than that of the corn cobs.

Use of corn cobs as an energy feedstock is not novel; for example, historically, corn cobs were combusted for heating purposes in the Central USA.4 Th e main current and future end-use competing with electricity generation for corn cob feedstock may be the production of second-generation liquid biofuels. Th ere are few such plants in North America; for example, a bioethanol pilot plant facility uses corn cobs as their main feedstock.6 Techno-economic challenges (e.g. high cost of enzymes) persist in

570 © 2014 Society of Chemical Industry and John Wiley & Sons, Ltd | Biofuels, Bioprod. Bioref. 8:568–578 (2014); DOI: 10.1002/bbb

D Sanscartier et al. Modeling and Analysis: GHG Emissions of Electricity Generation from Corn Cobs

combustion in one unit of an existing coal-fi red electricity GS (Nanticoke GS), hypothetically retrofi tted to handle 100% biomass. Th e boundaries of the bioenergy system include upstream activities (e.g. production of fuels, fer-tilizers), harvest of corn cobs, fertilizer inputs, eff ects of harvest on soil carbon, transportation of the cobs and pel-lets, densifi cation of the cobs into pellets, and combustion of the pellets (Fig. 1). A soil carbon model is integrated into the life cycle model to include the soil carbon eff ect of corn cob removal in the analysis.

Corn cob harvest and impacts of removal

Only the incremental eff ects of harvesting corn cobs are considered in the life cycle model as in Kim et al.10 Th ese eff ects include fuel consumption for harvesting residues, additional nutrient requirements, changes in soil carbon, and emissions of N2O.

It is assumed that cobs are harvested in a one-pass sys-tem with a modifi ed combine that collects cobs instead of discarding them in the fi eld, a harvest system that is commercially available.15 Th e additional fuel use result-ing from a slower harvesting process, and additional tractors/trucks and trailers for the harvest of the cobs are attributed to the system. Emissions associated with corn production (e.g. planting, application of pesticides) are assumed to be associated with the grain only (the eff ect of this assumption is examined using sensitivity analysis in the Discussion section). Corn cobs lost or not collected during harvest, lost during transport and spoiled during storage are considered (Table 1).

Field experiments suggest that, with the exception of potassium, the nutrients in corn cobs would have limited availability to the crop planted in the subsequent year.5 However, as a conservative approach, it is assumed that the nutrients in the corn cobs are replaced with generic synthetic NPK fertilizers (data from GHGenius13) through the system expansion approach (a common approach in LCA and used in similar studies – e.g. Kim et al.10). Replacement rates are based on elemental composition of nutrients in the corn cobs (Table 1). We account for net GHG (CO2 and N2O) emissions by considering the reduc-tion of N2O emissions due to corn cob removal, as well as GHG emissions related to synthetic fertilizer produc-tion,13 fertilizer application (based on Nagy16), and N2O emissions from the soil. Th e changes in direct and indirect GHG emissions due to corn cob removal and replacement with synthetic fertilizers are estimated based on the IPCC country-specifi c approach (Table 1).17

Although earlier LCA studies (e.g. Searcy and Flynn18) did not consider the GHG emission impact of removal

sole fuel for the generation of electricity. Th e study focuses on the Ontario context, and attention is placed on the GHG emissions implications of corn cobs removal. Th e life cycle GHG emissions of cob-pellet-based electricity generation are compared to those resulting from coal and natural gas electricity generation, to quantify emissions mitigation potential of the pellet-based generation.

Methods

Scope of the study

Th e study focuses on the quantifi cation of selected GHGs (CO2, CH4, and N2O) on a life cycle basis for the genera-tion of electricity with three product systems: corn cob pellets, coal, and natural gas. Results are presented based on the functional unit of one kWh of electricity generated for all activities up to the delivery of the electricity to the grid. Two intermediate indicators are presented for the corn cob product system: (i) one oven dry tonne (ODT) of corn cob biomass (ODTbiomass) for all activities up to the farm gate, and (ii) one ODT of corn cob pellets (ODTpellet) for all activities up to, and including, the delivery of pellets to the GS.

GHGs are reported as CO2eq based on the Intergovernmental Panel on Climate Change (IPCC) 100-year global warming potentials (GWP: CO2=1; CH4=25; N2O=298).12 Fossil CO2 and loss of soil carbon due to corn cob removal are reported, but biogenic CO2 emitted dur-ing the combustion of the pellets is not counted in calcula-tions as it does not have a net contribution to the global warming eff ect. However, emissions of biogenic CH4 are counted, consistent with IPCC.14

A life cycle model is developed to quantify the GHG emissions associated with electricity generation from corn cobs. Th e fossil-fuel-based reference electricity generation product systems (coal and natural gas combined cycle) are based on published models. Cradle-to-gate modules are developed or obtained from existing databases.13,14 Inputs and emissions from the construction of infrastructure and the manufacture of equipment generally have much smaller impacts than the operation of the systems and are therefore not considered. Th e product systems are described below. Table 1 presents the main parameters and data sources for the corn cob product system.

Corn cobs electricity product system description

Th e corn cob product system is based on a hypotheti-cal agricultural biomass supply chain in Ontario and



Tab le 1. Parameters and assumptions for the corn cob product system.

Parameter Value Comment and Data Source

Corn cob properties

Moisture content at harvest (% weight) 25% Moisture content of corn cob at harvest varies from 20 to 35%.34,35 It is assumed that moisture content remains constant until delivered to pellet plant.

Higher Heating Value (HHV) (MJ kg–1) 18.8 (dry)18.0 (5% moisture)

Based on various sources.3, 36–38 Moisture content of the delivered pellets is assumed to be 5%.

Cob yield (dry t ha–1) 1.2 Based on four-year (2004–2008) weighted average grain yield in Ontario.27 Assume a cob:grain ratio of 0.16.34, 35 Corn cob yield ranges from 0.9 to 1.4 dry t ha–1 in Ontario.

Cob elemental composition CNPK

48%0.50%0.05%0.51%

Percentages are reported by dry weight based on various sources.3,39

On-farm and transport

Additional fuel for harvesting corn cob (L ha–1)

4.8 Additional fuel use resulting from slower harvest, and additional tractors/trucks and trailers for cob harvest.40

Losses during harvest, storage and transport (% weight)

6.5% Lewandowski and Heinz 41

Farm gate to pellet plant by truck (km) 90 Authors’ assumption.

Pellet plant to port/GS by truck (km) 90 It is assumed that 10% of the pellet supply is delivered to the GS by truck due to proximity based on realistic truck traffi c limit in the vicinity of the GS.

Port to GS by vessel (km) 230 Weighted average based on cob production in Ontario and distances between ports along the St Lawrence seaway and Nanticoke GS.

Direct emissions of N2O from synthetic fertilizer (kg N2O kg–1 N applied)

0.0165 Regional fertilizer-induced emissions for Ontario.42 All other emission factors (indirect N2O emission and CO2 from urea fertilizers) are IPCC generic values43

Pellet plant

Electricity use (kWh ODTpellet–1) 144 Includes initial grinding, drying, milling, compression, cooling and sieving.8

Electricity is provided by the Ontario electricity grid.44

Raw biomass for drying (ODTbiomass ODTpellet

–1)0.13 Drying heat is provided by combustion of raw biomass feedstock, modifi ed for

feedstock moisture content.8

Moisture content pellets 5% Zhang et al.8

ICBM parameter values

Initial soil carbon (t C ha –1) 39 Value estimated for typical agricultural soil in Ontario45

Stover yield (dry t ha–1) 7.6 Based on four-year (2004–2008) weighted average grain yield in Ontario27 and harvest index for corn.29 Stover comprises the stalks, leaves and cobs.

External control factor (re) (dimensionless) 1.305 Average for Ontario 21

Decomposition rates for ‘young’ and ‘old’ soil carbon (1/year)

Young: 0.8Old:0.006

Values from19

Humifi cation factor for crop residues (dimensionless)

0.12 Values from19

Number of years modeled (yr) 20 Soil carbon changes were estimated over 20 years

of agricultural residues, we consider the GHG implica-tions of cob removal because of the role they play in the carbon cycle. During plant growth, corn cobs sequester CO2 from the atmosphere. When they are left on the fi eld at harvest, they decompose through microbial processes,

which release a portion of the stored CO2 into the atmos-phere, and sequester a portion of it as soil organic carbon (i.e. a carbon sink). However, when the corn cobs are removed for use as bioenergy feedstock for a combustion process, all the carbon stored in the cobs is released to the



pellet facility in the northeastern USA. Modifi cations are made to the process in Zhang et al.8 to refl ect the diff er-ent moisture content of cobs (25% moisture) compared to the wood (34%). Pellets are delivered to the GS using two modes of transportation: trucks (10% of the pellets – authors’ assumption) and vessel (Nanticoke GS is located on Lake Erie, within the Great Lakes St Lawrence Seaway System, a deep draft waterway).

Electricity generation

Th e generation of electricity with corn cob pellets is assumed to occur in one unit at the GS hypothetically retrofi tted to combust 100% biomass. Th is is modeled based on Zhang et al.8 who examined the combustion of pelletized woody biomass at the same facility. Based on test burns with wood pellets, the unit’s capacity is reduced from 500 to 250 MW and its net effi ciency is reduced by 10% (from 35% of 32%) when combusting wood pellets due to technical limitations.8 As there are no data available on combusting corn cob pellets in GS, it is assumed that these values are applicable to these pellets and that there are no restrictions on the consumption of biomass within the unit due to operational limitations such as fouling or slagging (e.g. from the higher alkali content of agricul-tural feedstock relative to wood). Biogenic emissions of CO2 due to the removal of corn cobs (i.e. the carbon sink loss) are accounted at the combustion of pellets in the GS. Emissions of biogenic CO2 from combustion of pellets are not counted and reported in the results as discussed above. Methane and N2O emissions resulting from pellet com-bustion are based on emission factors for the combustion of biomass in an industrial boiler in GHGenius.13

Reference product systems description

Th e corn-cob-based electricity is compared with two reference fossil fuel electricity product systems: coal and natural gas. For both reference product systems, upstream emissions (extraction of the fossil fuels, processing and transport to the GS) and combustion at the GS are counted. Th e coal product system is based on one unit of the Ontario Nanticoke GS. Th is GS is equipped with wall-fi red natural circulation pulverized coal boilers with low-NOx burners. Th e GS has a net electricity generation effi ciency of 35%. A blend of sub-bituminous coal (84% by weight, Southern Powder River Basin) and bituminous low sulfur coal (16% by weight, Central Appalachian) are used at the GS (HHV of the coal blend = 29 MJ kg–1 [dry]). Th e natural gas product system is based on a hypothetical natural gas combined cycle (NGCC) GS (net electricity

atmosphere during combustion. Because the corn cobs contribute to a carbon sink when left in the fi eld, we refer to their removal and subsequent impacts as a ‘carbon sink loss’. Th e carbon sink loss (converted to CO2 emissions) due to corn cob removal is allocated to the life cycle as an emission at the combustion life cycle stage. Th is approach of accounting for GHG emissions due to agricultural resi-due removal has been applied previously by others.9–11

Th e eff ect of corn cob removal on soil carbon is modeled using the Introductory Carbon Balance Model (ICBM).19 Th e model was chosen because it has been shown to pre-dict soil carbon as reliably as other soil carbon models (e.g. CENTURY), it is well suited to modeling soils in cooler climates such as Ontario, and it has been successfully applied to Eastern Canadian agricultural regions.19–21 It is also a relatively simple model to use. Th e ICBM estimates changes in soil C content based on initial soil carbon, annual input of fresh biomass, biomass mineralization and humifi cation rates. Th e ICBM model has been initialized to be at equilibrium as described in previous studies19,22,23 (i.e. initial distribution between young and old soil carbon was set to steady-state values for modeling), and the val-ues for the parameters are selected to represent Ontario soils (Table 1). Th e soil carbon content is determined for a 20-year period for two conditions: (i) when stover (stalks, leaves and cobs) are left in the fi eld; and (ii) when stalks and leaves only is left in the fi eld. Th e diff erence in soil carbon content for the two conditions represents the eff ect of corn cob removal. Th e initial soil carbon conditions should not aff ect the outcome because (i) ICBM is a linear model and (ii) we are considering the relative diff erence between the two residue removal conditions.24

Transportation and pellet production

Cobs are transported directly from the fi eld to the pellet facility. Th e densifi cation process is based on Zhang et al.8 who modeled an industrial-scale state-of-the-art wood

Replace-ment of lost

nutrients

Transport cob to pellet

plant

Pelletproduction

Transport pellet to GS


Corn cob harvest

Transport cob to on-

farm storage

Soilcarbon losses

Figure 1. Process fl ow diagram for the corn cob product system. GS= generating station.



80% lower than those of the NGCC (390 g CO2eq kWh–1) and coal product systems (1130 g CO2eq kWh–1), respec-tively, illustrating substantial GHG emissions mitigation through substituting coal with the cob pellets. Note that biogenic emissions of CO2 from combustion of biomass and pellets are not counted, as discussed in the Methods section. To our knowledge, the current study is the fi rst to consider electricity generation from corn cobs and therefore, there are no life cycle results that are directly comparable to those obtained in the study. Excluding the ‘carbon sink loss’ due to cob removal, our value (90 g CO2eq kWh–1) is slightly higher than the life cycle emis-sions estimated for direct combustion of wheat straw and corn stover in Searcy and Flynn18 (56 g CO2eq kWh–1), who assumed that there was no GHG emissions impact of removing residues. Methodological choices and the dif-ferent technologies analyzed (a purpose-built wood-fi red power plant in the case of Searcy and Flynn) may explain the diff erences in the result.

Th e majority of life cycle GHG emissions associated with the coal and NGCC product systems result from the com-bustion of the fuel in the GS, while upstream emissions (fuel production and transport) represent small portions of the product systems’ emissions. In contrast, for the cob product system, upstream emissions contribute a relatively larger portion of the product system’s emissions. Th e ‘carbon sink loss’ contributes 150 g CO2eq kWh–1(Fig. 2), 60% of the total life cycle emissions. Cherubini and Ulgiati11 used a similar approach, and found that the carbon sink loss repre-sented 50% of the emissions in the life cycle of corn stover used in a biorefi nery to produce various products.

Corn cob harvest and pellet production

Th e GHG emissions associated with corn cob harvest up to the farm gate (all on-farm activities and emissions in the fi eld) are 42 kg CO2eq ODTbiomass

–1 (Fig. 3(a)). Th is value falls within the range found by Kim et al.10 (i.e. 30 to 80 kg CO2eq dry t–1 of corn cobs harvested) for the Corn Belt states in the USA.

Th e largest contributors to on-farm GHG emissions associated with corn cob production are those resulting from fossil fuel use for harvest (36% of emissions up to the farm gate – Fig. 3(a)), and production and transport of synthetic fertilizers required to replace nutrients in the removed cobs (34% of emissions up to the farm gate). Th e fossil fuel used to apply fertilizers and the net soil N2O emissions related to N-fertilizer use contribute 16% and 15% of emissions up to the farm gate, respectively.

Th e production of the pellets and their transporta-tion to the GS results in GHG emissions of 115 kg CO2eq

generation effi ciency=53%; natural gas from Alberta, Canada). Details on both reference product systems can be found in Zhang et al.8 and Sanscartier et al. 24

Results

Impact of corn cob removal on soil carbon content

Based on the analysis using the ICBM the removal of corn cobs results in a diff erence in soil carbon content of 1.4 t C ha–1 aft er 20 years compared to if the cobs had been left in the fi eld to decompose. Th is corresponds to a reduction in soil carbon of 0.06 t C ha–1yr–1 per tonne of cobs removed. Th is value is consistent with those reported by other stud-ies who found that removal of agricultural residues could reduce soil carbon by 0.04 to 0.09 t C ha–1 yr–1 per tonne of residue removed (the studies cited did not specifi cally examine corn cobs).11,25 Leaving the agricultural resi-dues on the fi eld would result in sequestered CO2, but by removing and combusting the cobs for electricity genera-tion, this carbon sequestration does not occur. Th e portion of CO2 that would have been sequestered in the fi eld must therefore be treated as a GHG emission of the system (i.e. the carbon sink loss).

Life cycle GHG emissions


Th e life cycle GHG emissions for electricity generation for the coal, NGCC and corn cob pellets are disaggregated by major life cycle stage in Fig. 2. Th e corn cob product sys-tem’s life cycle emissions (240 g CO2eq kWh–1) are 40 and

050

100150200250300350400450500

Coal

Carbon sink loss

Direct emissions at GS

Transport of the fuel to GS

Fuel extraction,production and processing

1130

GH

G e

mis

sion

s (g

CO

2eq

kW

h-1)

NGCC Corn cobs

Figure 2. Life cycle GHG emissions per kWh generated for the two fossil fuel-based reference product systems and the corn cob product system. NGCC: natural gas combined cycle; GS: generating station.



Sensitivity analysis

Sensitivity analysis is conducted to determine the impact on life cycle GHG emissions of key corn cob life cycle parameters and the approach used to quantify soil carbon impacts. Most parameters are varied based on the vari-ability of data found in the literature (Table 1). Th e ranges of values used are shown with the results of the sensitivity analysis (Fig. 4).

Th e life cycle GHG emissions of the cob product system remain below those of the reference fossil product systems for the ranges of all parameters examined in the sensitiv-ity analysis. Life cycle GHG emissions for the cob product system are most sensitive to the values for soil carbon sink losses used in the life cycle model. Th is is mainly due to the fact that a small change in the soil carbon level (the carbon sink loss) results in a considerable change in emis-sions of CO2 to the atmosphere. Two types of carbon sink loss values are tested in the sensitivity analysis: (i) car-bon sink loss value (0.06 t C ha–1 yr–1 per tonne of cobs removed) estimated with the ICBM model using Ontario-specifi c parameters (Table 1), and (ii) values found in the literature (0.04 and 0.09 t C ha–1 yr–1 per tonne of agri-cultural residues removed).11,25 Using the ICBM allowed the consideration of site-specifi c conditions and factors aff ecting soil carbon dynamics, and more accurately refl ected the eff ect of agricultural residue removal on the life cycle of the biomass product system in the region of interest.

ODT pellet–1 (Fig. 3(b)), and is comparable to that of pellets

produced from roundwood harvested in the Great Lakes St Lawrence forest region of the province (i.e. 133 kg CO2eq ODTpellet

–1).8 On-farm processes, transportation to the pel-let plant and the GS, and the densifi cation process contrib-ute 42, 36, and 23% of the total emissions related to pellet production and delivery, respectively (Fig. 3(b)).

On an energy basis, the GHG emissions associated with the delivered solid fuels are 6.4 and 5.2 kg CO2eq GJ–1 (all fossil-based emissions) for corn cob pellets and coal, respectively (based on HHV and excluding the carbon sink loss). In addition to the lower energy density of corn cobs (18.8 MJ kg–1 [dry]) compared to coal (29 MJ kg–1 [dry]), the cob supply chain has several more emission-intensive steps than does the coal supply chain to produce a useable form of solid fuel (e.g. harvest, fertilizer produc-tion, densifi cation).

Discussion

Emission reduction potential

Even when accounting for the carbon sink loss due to the removal of corn cobs, the use of pelletized corn cobs for electricity generation as the sole fuel in a retrofi tted coal GS has the potential to mitigate up to 900 g CO2 kWh–1 generated (an 80% reduction in life cycle emissions). Furthermore, the corn cob product system’s emissions are 40% lower than those of the NGCC product system, the fossil-fuel-based electricity generation option with the lowest GHG emissions (assuming no carbon capture and storage) and that achieves the Government of Canada performance emission standard of 420 g CO2 kWh–1 (note that the standard is based on direct emissions at the gener-ating station not life cycle emissions).1

0

5

10

15

20

25

30

35

40

45

GH

G e

mis

sion

s(k

g C

O2e

q O

DT

biom

ass-

1)

Harvest

Fertilizerapplication

Fertilizerproduction

0

20

40

60

80

100

120

140

GH

G e

mis

sion

s(k

g C

O2e

q O

DT

pelle

t-1) Transport

Densi-fication

On-farm

(b)(a)

N2Oemissions

Figure 3. GHG emissions per (a) one ODTbiomass at the farm-gate, and (b) one ODTpellet delivered to the generating sta-tion. Note the different Y-axis scales and legends. ODT = Oven dry tonne.

150 200 250 300 350

Cob:grain dry mass ratio (15, 16, 20%)

Distance by vessel (160, 230, 290 km)

Cob yield (0.9, 1.2, 1.4 dry t/ha)

Distance farm to pellet plant (60, 90, 120 km)

Corn cob N content (0.33, 0.50, 0.69%)

Moisture content at harvest (20, 25, 35%)

Allocation of grain production to cob (none, mass)

Electricity generation efficiency (29, 32, 34%)

Carbon sink loss (0.04, 0.06, 0.09 t C/ha)

GHG emissions (g CO2eq kWh-1)

Figure 4. Results of sensitivity analysis for life cycle GHG emissions for corn cob system (g CO2eq kWh–1 generated). Parameters’ values used in sensitivity analysis are shown in parenthesis: low, base-case (i.e., those used in the main model), and high values. Carbon sink loss values used were either estimated with ICBM (0.06 – our value) or found in the literature (0.04 and 0.09). Allocation of the grain produc-tion to cobs: not allocated (indicated as none), and portion allocated based on mass of grain:cob harvest ratio (indi-cated as mass).



on soil quality parameters, such as aggregate stability and water holding capacity.

Impacts of agricultural residue removal

Assuming ‘carbon neutrality’ of biomass is a common approach in LCA of agricultural-residue-based bioen-ergy systems. In contrast, the current analysis allocates a carbon sink loss due to removal of residue from the fi eld to the product system as an emission (at combustion). Excluding the carbon sink loss due to cob removal, the life cycle emissions estimate is 90 g CO2eq kWh–1, com-pared to 240 g CO2eq kWh–1 when the carbon sink loss is included. Th e substantial contribution of the carbon sink loss (60% of the total life cycle emissions) illustrates the potential implications associated with removing agri-cultural residues from fi elds and questions the carbon neutrality assumption of agricultural residues. It also sup-ports a recommendation made by Cherubini and Ulgiati (p. 56)13 that these emissions ‘should always be taken into consideration when estimating the GHG balance of a biorefi nery system’.

Based on the sensitivity of life cycle GHG emissions to the estimate of soil carbon rates, to improve the accuracy of life cycle studies of agricultural residues, a priority should be placed on ensuring accurate estimation of resi-due removal impacts on soil carbon. Estimates of agri-cultural residue removal impact on soil carbon that are derived from the literature may be based on assumptions of mineralization and soil organic matter humifi cation rates that do not accurately refl ect actual rates for a study region. Collection of site-specifi c data, along with soil carbon models calibrated to a region may be required to ensure accuracy of life cycle studies.

Leaving agricultural residues in fi elds is an eff ective practice for erosion control, maintaining soil organic content, soil carbon sequestration, as well as recycling nutrients.30 Harvesting corn cobs for bioenergy may therefore negatively aff ect soil fertility and reduce soil carbon content, with several implications. Johnson et al.31 observed little evidence for short-term impacts on crop yield (aft er three cycles of corn stover removal), but detected subtle changes that may indicate that repeated harvest could have negative consequences. According to soil carbon modeling results, long-term (>10 years) corn stover removal in the Corn Belt of the USA resulted in soil carbon levels dropping to unacceptable levels.10 Field-to-fi eld and regional variation in soil type, topography, tillage practice, cropping system, and climate dictate crop yield, residue production, and amount of residue that can be sus-tainably removed.25,29,32 Overall estimates of sustainable

Life cycle GHG emissions are also sensitive to the net electricity generation effi ciency. During the test burns with pelletized woody biomass as sole feedstock at Nanticoke GS, the net effi ciency of the unit decreased from 35% with coal to 32% with biomass.8 Th e net effi ciency could be even lower with pelletized agricultural residues due to chal-lenges such as slagging and fouling (impact not estimated in the current study due to a lack of reliable data).

Methodological choices can impact results. Th e corn cob product system did not consider the production of the cob as it was assumed that 100% of the production of the corn grain and cob is allocated to the grain. Th is assumption was examined by allocating emissions related to the pro-duction of corn grain (and cob) to the cob on a mass basis based on the life cycle emissions for the production of corn grain in Ontario (282 kg CO2eq per dry t of grain)26 and the grain:cob mass ratio of 0.16 (Table 1). Th is increased the corn cob product system’s life cycle emissions by 6% to 255 g CO2eq/kWh, a relatively small impact compared to those of the parameters above.

Life cycle GHG emissions are relatively insensitive to the other parameters examined, suggesting robust results. Variation in cob moisture content, which may result from harvest date and weather variation, has little impact on life cycle results. However, high moisture levels (>35%) may be a factor for long-term storage of cobs and approaches exist to mitigate moisture content in cob piles such as forced ventilation (not estimated in the current study).2

Electricity generation potential

While on a per kWh basis, there is considerable potential for emissions reductions with corn cobs, the emissions mitigation should be considered in light of the amount of cobs potentially available in Ontario. Based on the 2004–2008 Ontario corn grain harvest statistics27 and a cob:grain ratio of 0.16, approximately 0.8 million ODT of cobs are available annually. Th erefore, up to 1.1 TWh of electricity could potentially be generated with corn cobs as sole fuel, which is equivalent to only 0.8% of the total annual electricity demand in Ontario.28 Realistically, even lower amounts of electricity are likely to be generated from this feedstock because: (i) the estimate is based on combusting all potentially harvestable corn cobs in the province, (ii) there is likely to be competition for this feed-stock with other renewable energy product systems (e.g. second-generation biofuels), (iii) soil carbon maintenance objectives may limit or restrict crop residue removal rates at a regional or fi eld level,29 and (iv) producers may leave residues in the fi eld if returns to removal are valued lower than economic benefi ts associated with residue impacts



and minimized or mitigated based on local conditions. However, these impacts should not discourage sustainable use of agricultural residues, which may provide a biomass source until large-scale commercial production of dedi-cated bioenergy feedstock can be established.

Acknowledgments

Th is research was fi nancially supported by Ontario Power Generation (OPG) and the Natural Sciences and Engineering Research Council of Canada. We thank Robert Lyng, Les Marshall, Tammy Wong and Sandy Drysdale, Ontario Power Generation; and Martin Bolinder, Soils and Agrifood Engineering, Université Laval for data and insights.

References1. Environment Canada, Reduction of Carbon Dioxide Emissions

from Coal-Fired Generation of Electricity Regulations. [Online]. (2011). Available at: http://ec.gc.ca/lcpe-cepa/eng/regulations/detailReg.cfm?intReg=209 [May 10, 2013].

2. Zych D, The viability of corn cobs as a bioenergy feedstock. West Central Research and Outreach Center, University of Minnesota, Minneapolis, MN (2008).

3. Mullen CA, Boateng AA, Goldberg NM, Lima IM, Laird DA and Hicks KB, Bio-oil and bio-char production from corn cobs and stover by fast pyrolysis. Biomass Bioenerg 34:67–74 (2010).

4. Avila-Segura M, Barak P, Hedtcke JL and Posner JL, Nutrient and alkalinity removal by corn grain, stover and cob harvest in Upper Midwest USA. Biomass Bioenerg 35:1190–1195 (2011).

5. Wienhold BJ, Varvel GE and Jin VL, Corn cob residue car-bon and nutrient dynamics during decomposition. Agron J 103(4):1192–1197 (2011).

6. Menon V and Rao M, Trends in bioconversion of lignocellu-lose: Biofuels, platform chemical & biorefi nery concept. Prog Energ Combust 38:522–550 (2012).

7. Drax. Biomass. [Online]. (2013). Available at: http://www.drax.com/biomass/ [August 1, 2013].

8. Zhang Y, McKechnie J, Cormier D, Lyng R, Mabee W, Ogino A et al., Life cycle emissions and cost of producing electricity from coal, natural gas, and wood pellets in Ontario, Canada. Environ Sci Technol 44:538–544 (2010).

9. Sheehan J, Aden A, Paustian K, Killian K, Brenner J, Walsh M et al., Energy and environmental aspects of using corn stover for fuel ethanol. J Ind Ecol 7:117–146 (2004).

10. Kim S, Dale BE and Jenkins R, Life cycle assessment of corn grain and corn stover in the United States. Int J Life Cycle Assess 14:160–174 (2009).

11. Cherubini F and Ulgiati S, Crop residues as rax materials for biorefi nery systems - A LCA case study. Appl Energ 87:47–57 (2010).

12. Intergovernmental Panel on Climate Change, Technical Summary - Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, UK (2007).

residue removal for Ontario vary from no removal to 4.7 ODTbiomass ha–1 yr–1 due to ranges in assumptions for soil organic matter mineralization and humifi cation rates.29 However, based on average literature values for soil organic matter mineralization and humifi cation rates, estimated sustainable removal of residue from the major crops (corn, soybean, and winter wheat) in Ontario is 1.9 million ODT per year,29 in comparison to 0.8 million ODT per year of cobs potentially available.

Corn cob removal rates that result in reduced soil quality could result in lower yields relative to the baseline situa-tion examined, which would then aff ect life cycle GHGs. Where removal exceeds rates required to maintain soil carbon, management practices can be implemented to help mitigate the soil carbon losses: complex crop rotations, inclusion of winter cover crops and semi-perennial crop in crop rotations, moving to no-tillage systems, organic soil amendments (e.g. manure and compost), and application of biofuel production co-products (e.g. biochar, ash).11,32

When left in the fi eld, the corn cob decomposes. Th rough this process, a portion of the carbon in the cob is emit-ted to the atmosphere as CO2, while the remainder of the carbon is stored in the soil and contributes to soil carbon sequestration. If corn cobs are removed for electricity generation, the carbon stored in the cobs is emitted to the atmosphere during combustion, except for some unburned carbon residue in the ash.33 To better understand the soil carbon impacts and GHG implications of cob removal, it is important to determine how the ash will be handled to help manage this carbon loss. A common practice for coal-based ash is its use as an additive in concrete, but it is unclear whether this approach is valid for biomass-based ash. Ash could be used as a soil amendment, thus returning some carbon to soil in the form of unburned carbon, but may impact soil quality due to the presence of inorganic compounds in the ash (e.g. silicon, chlorine, metals).33

Conclusion

Biomass-based electricity may complement intermittent renewable electricity systems (e.g. solar, wind) because it is dispatchable and can therefore more easily follow demand. Th e use of corn cobs as a source of pellets to replace coal for electricity generation may provide sub-stantial GHG emission reductions on a per kWh basis but there are limitations to the eff ectiveness of this option to reduce emissions, due to the limited quantities of corn cobs available. Additionally, site-specifi c dynamics and impacts of cob removal on soils (e.g. soil carbon losses and associated fertility implications) need to be considered,



13. Natural Resources Canada. GHGenius 3.17 - A model for the lifecycle assessment of transportation fuels. [Online]. (2010). Available at: http://www.ghgenius.ca/ [May 12, 2010].

14. Nagy CN, Energy Coeffi cients for Agriculture Inputs in Western Canada. Canadian Agricultural Energy End-Use Data Analysis Centre, Ottawa, ON, Canada (1999).

15. Vermeer, CCX770 Cob Harvester. [Online]. (2013). Available at: http://www2.vermeer.com/vermeer/AP/en/N/equipment/cob_harvester/ccx770 [August 26, 2013].

16. Nagy C, Energy Coeffi cients for Agriculture Inputs in Western Canada. Canadian Agricultural Energy End-Use Data Analysis Centre, University of Saskatchewan, Saskatoon, SK, Canada (1999).

17. Rochette P, Worth DE, Lemke RL, McConkey BG, Pennock DJ, Wagner-Riddle C et al., Estimation of N2O emissions from agricultural soils in Canada. I. Development of a country-spe-cifi c method. Can J Soil Sci 88:641–654 (2008).

18. Searcy E and Flynn PC, Processing of straw/corn stover: Comparison of life cycle emissions. Int J Green Energ 5(6):423–437 (2008).

19. Andren O and Katterer T, ICBM: The introductory carbon bal-ance model for exploration of soil carbon balances. Ecol Appl 7:1226–1236 (1997).

20. Vanden Bygaart AJ, Gregorich EG, Angers DA, Bolinder MA, Janzen HH and Campbell CA, Modeling soil organic carbon change in Canadian agroecosystems: Testing the introductory carbon balance model. Soil Carbon Sequestration and the Greenhouse Effect. Soil Science Society of America, Madison, WI, Special Publication 57, pp. 13–28 (2009).

21. Bolinder MA, Andrén O, Kätterer T and Parent LE, Soil organic carbon sequestration potential for Canadian agricultural ecoregions calculated using the introductory carbon balance model. Can J Soil Sci 88:451–460 (2008).

22. Katterer T, Andersson L, Andren O and Persson J, Long-term impact of chronosequential land use change on soil carbon stocks on a Swedish farm. Nutr Cycl Agroecosyst 81(2):145–155 (2008).

23. Bolinder MA, Katterer T, Andren O and Parent LE, Estimating carbon inputs to soil in forage-based crop rotations and mod-eling the effects on soil carbon dynamics in a Swedish long-term fi eld experiment. Can J Soil Sci 92:821–833 (2012).

24. Sanscartier D, Deen B, Dias G, Dadfar H, MacLean HL, McDonald I et al., Implications of land class and environmen-tal factors on life cycle GHG emissions of Miscanthus as a bioenergy feedstock. GCB Bioenergy, DOI:10.1111/gcbb.12063 (2013).

25. Gregg JS and Izaurralde RC, Effect of crop residue harvest on long-term crop yield, soil erosion and nutrient balance: trade-offs for a sustainable bioenergy feedstock. Biofuels 1(1):69–83 (2010).

26. Jayasundara S, Wagner-Riddle C, Dias G and Kariyapperuma KA, Energy and greenhouse gas intensity of corn (Zea mays L.) production in Ontario: A regional assessment. Can J Soil Sci 94:77–95 (2014).

27. Ontario Ministry of Agriculture and Rural Affairs. Field Crop Statistics. [Online]. (2009). Available at: http://www.omafra.gov.on.ca/english/stats/crops/index.html [April 12, 2010].

28. Independent Electricity System Operator. Independent Electricity System Operator Releases 2013 Ontario Electricity Data. [Online]. (2014). Available at: http://www.newswire.ca/

en/story/1286799/independent-electricity-system-operator-releases-2013-ontario-electricity-data [March 24, 2014].

29. Kludze H, Deen B, Weersink A, Acker Rv, Janovicek K, Laporte AD et al., Estimating sustainable crop residue removal rates and costs based on soil organic matter dynamics and rotational complexity. Biomass Bioenerg 56:607–618 (2013).

30. Blanco-Canqui H and Lal R, Crop residue removal impacts on soil productivity and environmental quality. Crit Rev Plant Sci 28:139–163 (2009).

31. Johnson JMF, Acosta-Martinez V, Cambardella CA and Barbour NW, Crop and soil responses to using corn stover as a bioenergy feedstock: Observations from the Northern US corn belt. Agr Ecosyst Environ 3:72–89 (2013).

32. Blanco-Canqui H, Crop residue removal for bioenergy reduces soil carbon pools: How can we offset carbon losses? Bioenerg Res 6:358–371 (2013).

33. James AK, Thring RW, Helle S and Ghuman HS, Applications for biomass bottom ash. Energies. Ash Manage Rev 5:3856–3873 (2012).

34. Varvel GE and Wilhelm WW, Cob biomass production in the western corn belt. BioEnerg Res 1:228–233 (2008).

35. Halvorson AD and Johnson JMF, Corn cob characteristics in irrigated central great plains studies. Agron J 101:390–399 (2009).

36. Azeez AM, Meier D, Odermatt J and Willner T, Fast pyrolysis of African and European lignocellulosic biomasses using Py-GC/MS and fl uidized bed reactor. Energ Fuel 24:2078–2085 (2010).

37. Dermibas A, Calculation of higher heating values of biomass fuels. Fuel 76:431–434 (1997).

38. Dermibas A, Effects of moisture and hydroben content on the heating value of fuels. Energ Source 29:649–655 (2007).

39. Sawyer J and Mallarino A, Nutrient Removal when Harvesting Corn Stover. [Online]. Integrated Crop Management. (2007). Available at: http://www.ipm.iastate.edu/ipm/icm/2007/8-6/nutrients.html [January 5, 2011].

40. Shinners KJ, Boettcher GC, Hoffman DS, Munk JT, Muck RE and Weimer PJ, Single-pass harvest of corn grain and stover: Performance of three harvester confi gurations. Trans ASABE 52(1):51–60 (2009).

41. Lewandowski I and Heinz A, Delayed harvest of Miscanthus-infl uences on biomass quantity and quality and environ-mental impacts of energy production. Eur J Agron 19:45–63 (2003).

42. Rochette P, Worth DE, Lemke R, McConkey B, Pennock DJ, Wagner-Riddle C et al., Estimation of N2O emissions from agricultural soils in Canada. I. Development of a country spe-cifi c methodology. Can J Soil Sci 88:641–645 (2008).

43. Intergovernmental Panel on Climate Change (IPCC). Chapter 11: N2O Emissions from Managed Soils, and CO2 Emissions from Lime and Urea Application, in Intergovernmental Panel on Climate Change National Greenhouse Gas Inventories Programme. Institute for Global Environmental Strategies (IGES), Hayama, Japan (2006).

44. Ontario Ministry of Energy, Ontario’s Long-Term Energy Plan. [Online]. (2013). Available at: http://www.energy.gov.on.ca/en/ltep/ [March 24, 2014].

45. Vanden Bygaart AJ, Gregorich EG and Angers DA, Infl uence of agricultural management on soil organic carbon: A com-pendium and assessment of Canadian Studies. Can J Soil Sci 83:363–380 (2003).



David Sanscartier

David Sanscartier is a Research Engineer in the Environment Division, Saskatchewan Research Council (Canada). His recent research focuses on the application of life cycle as-sessment and life cycle cost analysis to bioenergy systems and to the remediation of contaminated sites as

decision-support tool s.

Goretty Dias

Goretty Dias is an Assistant Professor in the Faculty of Environment at the University of Waterloo, Canada. Her expertise is in greenhouse gas meas-urement, agricultural systems, and the development of life cycle sustainability assessment approaches for biobased products and food. She holds a PhD

from the University of Guelph, Canada.

Bill Deen

Bill Deen is an Associate Professor in the Department of Plant Agriculture, University of Guelph. His expertise is in temperate agroecosystems, particu-larly the impact of management and crop species on soil quality, nutrient dynamics, and productivity. He holds a PhD from the University of Guelph.

Ian McDonald

Ian McDonald investigates crops and agricultural-based feedstock for ap-propriate opportunities in the emerging bioeconomy including types, agron-omy, economics, and environmental benefits. He is involved in the coordi-nation of research and demonstration activities within the Ontario Ministry of

Agriculture and Rural Affairs ADB Field Crops unit and is involved in scientific review, protocol and project devel-opment, data analysis, and reporting.

Heather MacLean

Heather MacLean is a Professor in the Departments of Civil Engineering and Chemical Engineering & Applied Chemistry at University of Toronto. Her expertise is in the development/appli-cation of life cycle-based approaches for evaluating energy systems, particu-larly bioenergy systems. She holds a

PhD from Carnegie Mellon University, Pittsburgh, PA.

Humaira Dadfar

Humaira Dadfar’s expertise is in characterizing and modeling solute transport and reaction in soil, prefer-ential flow, and nutrient management. Her recent research interest includes application of life cycle assessment to agricultural systems and bioenergy sector. She holds a PhD from the

University of Guelph.

Life cycle greenhouse gas emissions of electricity generation from corn cobs in Ontario, Canada

Documents