Biofuels from Crop Residue: Soil Organic Carbon and Climate Impacts in the US and India

Transcription

1 University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln Adam Liska Papers Biological Systems Engineering Biofuels from Crop Residue: Soil Organic Carbon and Climate Impacts in the US and India Adam Liska University of Nebraska - Lincoln, aliska2@unl.edu Follow this and additional works at: Part of the Bioresource and Agricultural Engineering Commons, Environmental Engineering Commons, Environmental Indicators and Impact Assessment Commons, Oil, Gas, and Energy Commons, and the Other Earth Sciences Commons Liska, Adam, "Biofuels from Crop Residue: Soil Organic Carbon and Climate Impacts in the US and India" (217). Adam Liska Papers This Article is brought to you for free and open access by the Biological Systems Engineering at DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion in Adam Liska Papers by an authorized administrator of DigitalCommons@University of Nebraska - Lincoln.

2 Biofuels from Crop Residue: Soil Organic Carbon and Climate Impacts in the US and India Adam J. Liska Associate Professor George Dempster Smith Chair of Industrial Ecology Departments of Biological Systems Engineering, and Agronomy & Horticulture Program Coordinator, Energy Science Minor Daugherty Water for Food Global Institute Faculty Fellow University of Nebraska-Lincoln, ASA Crop Residues for Advanced Biofuels Workshop: Exploring Soil Carbon Effects April 15, 217, Sacramento, CA

3 Soil organic carbon (SOC) levels are at equilibria determined by Below is world soil organic carbon density carbon (tons/hectare, inputs from as above). plant material (+I c ) and USDA: Natural Resource Conservation Service, 2. SOC tends to decrease as atmospheric temperature increase loss from oxidation to CO 2 (-kc soc-co2 ): SOC = I C - kc soc-co2 Fraction of plant material remaining after 1 year = UK Australia Batjes, N. H Harmonized soil property values for broad-scale 35% 2% modelling (WISE3sec) with estimates of global soil carbon stocks. Geoderma, 269: Austria 4% Nigeria Costa Rica 2% 3% Tropical soils have higher C input rates due to higher levels of plant growth but higher rates of oxidation due to higher temperatures. USA 4% via Source: Jenkinson et al., 1991, Nature. usoils_docs/other/eur25225.pdf, p. 57 Sources: Batjes, 216, Geoderma; Lal & Sanchez, 1992, Myth & Science of Soils in the Tropics, Greenland et al. p.17-33; Kutsch et al, 29, Soil Carbon Dynamics, Cambridge.

4 Most simple SOC model SOC = I C - kc o Converted to integral form of equation + daily temp + slower over time (S) K s and S s parameter values based on 36 data sets from 36 studies from Europe, North America, Asia, Africa, Australia, New Zealand, S. America Batjes, N. H Harmonized soil property values for broad-scale modelling (WISE3sec) with estimates of global soil carbon stocks. Geoderma, 269: Tropical soils have higher C input rates due to higher levels of plant growth but higher rates of oxidation due to higher temperatures. via usoils_docs/other/eur25225.pdf, p. 57 Source: Yang HS & Janssen BH. Eur. J. Soil Sci. 51, , 2; Yang HS & Janssen BH. Plant Soil 239, , 22; Liska et al. Nature Climate Change 4, , 214.

5 Modeled daily oxidation of soil & crop residue to CO 2 for no-till continuous corn (21-21, Mead, NE) based on field measurements of initial SOC (C o ), annual biomass input (I c ), and daily temperature (T a ) Source: Pelton MP. Soil Organic Carbon Dynamics in Agriculture: Model Development & Application from Daily to Decadal Timescales. Master s thesis, Univ. Nebraska (213). CO 2 C o Ic Mostly lignin remaining

6 Daily measured vs modeled CO2 emissions, corn-corn 21-21, NE CO2 measured every few seconds, daily aggregated values shown 2 measured ERe modeled ERe measured ERe 18 2 Modeled daily respiration (g C.m-2.d-1) site1 RMSE: 1.51 gc.m-2.d-1 NAE: -9% R2:.882 ME:.84 n: : Measured daily respiration (g C.m-2.d-1) 2 We know crop residue carbon respires to CO2 Models for respiration of CO2 based on temp. & solar radiation, etc.: crop (red) crop residue (green) soil (blue) Source: Liska, in preparation

7 The same multi-pool SOC model quantifies how much carbon is left in soil & crop residue (continuous corn, 21-21): comparison of eddy covariance flux measurements of CO 2 vs. model estimates found annual avg. error of 12% (solid line); model estimated (dashed line) Removal of crop residue (6 Mg ha -1 yr -1 ) puts less carbon in the soil SOC (Mg C per hectare) no residue removal (control) residue removal carbon emitted via biofuels Residue converted to biofuels & burned to CO Time (days) 6 Source: Liska et al., Nature Climate Change 4, , 214.

8 From at Mead, NE, eddy covariance measurements of CO 2 from continuous corn (control) (top), corn residue removed in field! (~5 Mg ha -1 yr -1, ~5%) (middle), difference between the two (bottom) NEP = GPP - respiration Grain harvest NBP = GPP - respiration - C removed Cumulative sum of net ecosystem production (NEP) and net biome production (NBP) starting with harvest in 21. This field site has produced >85 peer-reviewed related publications from $8 million in funding from US DOE, USDA, and NASA. Source: Nugy-Robertson, Suyker, Zahn, Liska, Arkebauer, in preparation NBP from residue removal compared to no removal

9 US-Ne2 US-Ne1 For , comparison of eddy covariance measurements of CO 2 vs. modeled CO 2 emissions from continuous corn (control) (US-Ne1) & continuous corn with residue removal (US-Ne2): Avg Abs Error 2.6% Growing Season SOC Re Residu e Re Modeled Crop Re (hybrid-maize) Grain Re (hailstorm) Ecosystem Re Measured Ecosystem Re error (%) error (%) Avg Avg. Source: Nugy-Robertson, Suyker, Zahn, Liska, Arkebauer, in preparation

10 Extrapolation soil model applied to independent geospatial data for SOC (SSURGO), temperature (PRISM), & maize yields (USDA) in 4 simulations: R1, R2, R3, R4 (, 2, 4, 6 Mg ha -1 yr -1 removal) for each 3 x 3 m cell 4 3cm SOC -2 (Mg ha ) Km Eddy-covariance field site, center pivot irrigation in Nebraska; 48 hectares (US-Ne1) Source: Liska et al., Nature Climate Change 4, , 214. SOC (Mg C per hectare) Time (days) R1 R2 R3 R4 Geospatial modeling results were within 6% of field measurements over 9 years, avg annual error:.3 Mg ha -1 yr -1

11 Extrapolation Modeled SOC loss to CO2 in Corn Belt from residue removal to capture spatial variability Corn Belt avg. net SOC loss to CO2 Model applied at 589 million cells, at different removal rates: 3 x 3 meters, Supercomputing 2, 4, 6 Mg ha-1 yr-1 (589 M cells) (6 Terabyte experiment!) (avg. for 5 & 1 years of removal) 2 5yr.8 C Removal 5yr 1.6 C Removal 5yr 2.4 C Removal 1yr.8 C Removal 1yr 1.6 C Removal 1yr 2.4 C Removal Count (millions) R1 - R4 Mg C ha-1yr Km Net SOC loss to CO2 from removal of 6 Mg biomass per hectare per year Mg SOC per hectare per year More net SOC loss to CO2 with increased removal of residue, Source: Liska et al., Nature Climate Change 214. but less loss of SOC over time

12 Count (millions) Extrapolation Contribution of CO 2 emissions from SOC & crop residue to the life cycle GHG emissions of biofuel, either cellulosic ethanol or thermochemical conversion of corn residue Net CO 2 emissions per unit energy derived from crop residue is constant for ALL biomass removal levels yr 5 yr 5yr.8 C Removal 5yr 1.6 C Removal 5yr 2.4 C Removal 1yr.8 C Removal 1yr 1.6 C Removal 1yr 2.4 C Removal g CO2e 2 per MJ ΔSOC-CO 2 = 6 = 4 = 2 = 1 Δbioenergy 6 = 4 = 2 = 1 Source: Liska et al., Nature Climate Change 214. GHG Intensity (gco 2 e MJ -1 ) CO 2 emissions from SOCresidue were not previously quantified in LCAs by DOE,EPA Gasoline 6% GHG reduction No SOC change 9 yr avg Mead, NE 5 yr avg 1 yr avg Corn Belt Corn Belt Biofuels from crop residue produce CO 2 emissions significantly above US federal standards & gasoline

13 SOC loss is mostly <1% of stock per year from top 3 cm of soil profile 1 Year 5 Year % SOC % SOC 5 Km <1% loss per year is difficult to measure by soil mass, but can be more accurately estimated by CO2 emissions measurements using eddy flux (Kutsch et al. Cambridge 29) Average initial SOC stock: 74.5 Mg C ha-1 3 cm-1 depth ~13 Mg C ha-1 6 cm-1 ~ 17 Mg C ha-1 9 cm-1 (~1 cm-1) (Schmer et al. SSSAJ 214) Source: Liska et al., Nature Climate Change Km Geospatial cells (millions) yr 5yr 5yr 1yr 1yr 1yr 589 million cells 15 1 R1-R2 R1-R3 R1-R4 R1-R2 R1-R3 R1-R Percent (%) SOC loss 12 3.

14 CO2 emissions from SOC & corn residue are highest where SOC stock is high (3 cm depth), But removal of 2, 4, or 6 Mg ha-1 yr-1 gives the same results: 7 ± 6.4 g CO2 MJ 1 (range 3 9) for 5 years 49 ± 4.3 g CO2 MJ 1 (range 33 63) for 1 years based on the premise that marginal change (not absolute change) is quantified as for indirect land use change (Searchinger et al. 28) which the EPA recognizes, but had not recognized SOC changes: RFS2 mandate would produce ~1.4-2% of US total GHG emissions for SOC High loss 1 Year 5 Year R1 - R4 gco2e MJ -1 R1 - R4 gco2e MJ Km Km Source: Liska et al., Nature Climate Change 4, ,

15 From 21 to 26, removal of 3 Mg ha 1 yr 1 of residue from continuous corn, compared to no residue removal, in three counties in Nebraska and Iowa based on IPCC projected temperature increase. Source: Pelton MP. Soil Organic Carbon Dynamics in Agriculture: Model Development & Application from Daily to Decadal Timescales. Master s thesis, Univ. Nebraska (213). First 1 years: ~.22 Mg C ha 1 yr 1 loss First 3 years, ~.11 Mg C ha 1 yr 1 loss ~52% reduction in rate of loss per year

16 Key Ratio used by US EPA CO 2 emissions from SOC per crop residue-energy, gco 2 /MJ emissions must be normalized per unit energy Net CO 2 emissions per unit energy derived from crop residue is constant for ALL biomass removal levels Count (millions) yr 5 yr 5yr.8 C Removal 5yr 1.6 C Removal 5yr 2.4 C Removal 1yr.8 C Removal 1yr 1.6 C Removal 1yr 2.4 C Removal GHG Intensity (gco 2 e MJ -1 ) CO 2 emissions from SOCresidue were not previously quantified in LCAs by DOE/EPA 6% GHG reduction 9 yr avg 5 yr avg 1 yr avg g CO2e per MJ -2 Gasoline No SOC change Mead, NE Corn Belt Corn Belt 2 Biofuels from crop residue produce CO 2 emissions significantly above US federal standards & gasoline ΔSOC-CO 2 = 6 = 4 = 2 = 1 Δbioenergy 6 = 4 = 2 = 1 Source: Liska et al., Nature Climate Change 214.

17 Are the US results relevant for India if crop residues are used for biofuels on a large scale? India SOC average (216) = 69 Mg ha -1 1 cm -1 Most areas have SOC at ~5-1 Mg ha 1 1 cm -1 (~1/2 US) Single crop: 58.5 Mg ha 1 1 cm -1 double crop: 67.4 Mg ha 1 1 cm -1 Andhra Pradesh, India, SOC 6 cm -1 Alfisols, 52.8 Mg ha 1 (.61% OC) Inceptisols, 51.3 Mg ha 1 (.58%) Vertisols, 49.3 Mg ha 1 (.54%) Forests, 87.3 Mg ha 1 (.94%) ~5-1 Mg ha 1 1 cm -1 SOC Sources: Sreenivas et al Digital Organic and Inorganic Carbon Mapping of India, Geoderma; Venkanna et al Carbon Stocks in Major Soil Types and Land-Use Systems in Semiarid Tropical Region of Southern India. Current Science.

18 CO 2 emissions from SOC & maize residue for biofuels in India similar results as in the US India s lower SOC levels vs US (69 vs. 17 Mg ha -1 1 cm -1 ), and lower maize yields (~12 Mg ha -1 yr -1 in Andhra Pradesh vs. ~2 Mg ha -1 yr -1 maize aboveground biomass) means ~5 to 8% as much CO 2 per unit biofuel energy from residue result compared to the estimate for the US: 66% of ~7 g CO 2 MJ 1 for 5 years: ~46 g CO 2 MJ 1 = ~half the CO 2 intensity of gasoline: ~92 g CO 2 MJ 1 (time horizons in C-intensities are arbitrary [5, 1, 2, 3, 1 yrs], but should be in line with policy goals which are often near term) Location SOC g C Mg/ha MJ/ha g/mj % USA India Source: Liska, April 2, 217. More calculations can be done for more robust estimates 17

19 Can replacement of SOC loss from residue removal offset net increases in emissions? Must be of relatively equivalent amount, must be cost effective Surface manure application won t work, merely transfer: Greater concentrations of SOM in manured fields can thus be expected to be associated with declining SOM on a proportionally larger area of off-site lands. Manuring has a number of practical applications, but net carbon sequestration is not one of them. - Schlesinger, W.H., Science Cover crops can replace some, but need limited water, nitrogen, & time Residue removal, % Avg SOC, Mg ha -1 yr -1 >5% -.87 <5% -.31 Cover crops +.49 Source: Ruis SJ & Blanco-Canqui H, 217. Agronomy Journal (review article).

20 Univ. Nebraska Research Team (Funding: US Department of Energy) Modeling crop & soil CO 2 respiration CO 2 Flux Measurements Prof. Haishun Yang Matthew Pelton Agronomy MS student-213 & Horticulture Environ. Eng. Thanks to other faculty: Prof. Ming Zhan Visiting Professor Prof. Andy Suyker Huazhong Agricultural School of University, China Natural Res. Profs. Chittaranjan Ray, Shashi Verma, Ken Cassman, Tim Arkebauer Geospatial Analysis & Supercomputing LCA Dr. Tony L. Nguy-Robertson US DOD SOC Measurements Dr. Maribeth Milner Agronomy & Horticulture Prof. Steve Goddard Computer Science; Dean, Arts & Sciences Dr. Haitao Zhu PhD student-213 Computer Science Anita Fang MS student-212 Biosystems Eng. Prof. Humberto Blanco-Canqui Agronomy-Hort

21 Absolute SOC change in Corn Belt (21-21) indicates loss under both continuous corn and corn-soybean rotation (less C input but more prevalent) based on methods shown above. Geospatial modeling of SOC dynamics under hypothetical continuous maize from 21 to 21 in the US Corn Belt, (A) map indicating the majority of the area is losing SOC, with the experimental site indicated (58 million cells), (B) distribution of average SOC loss per year without residue removed and estimated C input from corn-soybean rotation, R Km R Mg C ha yr Geospatial Cells (millions) R1: R3: R NetSOC R1, continuous corn R3, corn-soybean rotation Source: Liska, Zahn, Nugy-Robertson, Yang, Pelton, Suyker, in preparation 2 R1

22 Respiration (g C m -2 d -1 ) Respiration (g C m -2 d -1 ) Daily measured vs. modeled CO 2 emissions, corn-soybean RMSE 1.65 g C m -2 d -1 NAE -9% R 2.84 ME (B) Ameriflux Site 2, irrigated Models for respiration of CO 2 (based on temp. & solar radiation, etc.): crop crop residue soil Source: Liska, in preparation

23 Daily measured vs. modeled CO 2 emissions, corn-soybean (B) 2 16 RMSE 1.64 g C m -2 d -1 NAE -18% R ME (C) (A) year year Ameriflux Site 3, dryland Models for respiration of CO 2 (based on temp. & solar radiation, etc.): crop crop residue soil Source: Liska, in preparation

24 Measurements suggest that SOC levels in Andhra Pradesh, India appear to be mostly determined by carbon inputs from plant growth SOC = I C - kc oc (See slide #2 for further details) Source: Venkanna et al. 214, Current Science 16, ; these findings are further 23 supported by: Lal, 24, Science; Lal, 26, Land Degradation & Development

25 Presentation Abstract: The transformation of crop residue to soil organic carbon and CO 2 is a conserved process that occurs globally. Due to the mathematics of carbon intensity calculations found in government regulations, the amount of CO 2 emitted from crop residue per unit of energy in biofuel is largely independent of the amount of residue removed and the location of its removal, as shown by results from the US and India. Publications & Presentations, related: ( Liska AJ, Yang H, Milner M, Goddard S, Blanco-Canqui H, Pelton MP, Fang XX, Zhu H, Suyker AE. Biofuels from Crop Residue Can Reduce Soil Carbon and Increase CO 2 Emissions. Nature Climate Change 4, , 214. Liska AJ, Yang H, Pelton MP, Suyker AE. Reply to'co 2 emissions from crop residue-derived biofuels'. Nature Climate Change 4, , 214. Liska AJ. Eight Principles of Uncertainty for Life Cycle Assessment of Biofuel Systems. Chapter 11, p , IN: Bhardwaj et al. (eds.), DeGruyter, 215 Sustainable Biofuels: An Ecological Assessment of Future Energy. Nguy-Robertson AL, Suyker AE, Arkebauer TJ, Zhan M, Liska AJ. Eddy Covariance Measured CO 2 Emissions from Maize Residue Removal over Three Years. In preparation for GCB Bioenergy. Liska AJ, Zhan M, Suyker AE, Yang H, Pelton MP, Nguy-Robertson AL. Measurement and Modeling of CO 2 Flux from Maize and Soybean Indicates Soil Carbon Loss in the US Corn Belt. In preparation for Global Change Biology. Wortmann CS, Liska AJ, Ferguson RB, Klein RN, Lyon DJ, Dweikat I. Dryland Performance of Sweet Sorghum and Grain Crops for Biofuel in Nebraska. Agronomy Journal 12, , 21. Liska AJ, Perrin RK. Indirect Land Use Emissions in the Life Cycle of Biofuels: Regulations vs. Science. Biofuels, Bioproducts, and Biorefining 3, , 29. Pelton MP. Soil Organic Carbon Dynamics in Agriculture. Master s Thesis, Univ. Nebraska 213.