No abstract

Spring harvested corn stover was used for direct combustion in a 146 kW dual chamber boiler designed for wood logs. Stover had a very low moisture content (6.83 ± 0.17%), a gross calorific value (GCV) of 18.57 MJ/kg of dry matter (±0.32 MJ/kg DM) and an ash content of 5.88% (±1.15%). Small stover bales (8.83 ± 0.90 kg) were placed manually in the upper combustion chamber at a rate of 10.5 to 12.8 kg/h over a 24-h period, with three replications, and compared to a control wood combustion trial (12.1 kg/h during 24 h). The overall heat transfer efficiency for stover was lower than for wood (57% vs. 77%). Stover bales produced on average 7.5% ash which included about 2% of unburned residues while wood produced 1.7% ash. CO gas emissions averaged 1324 mg/m³ for stover (118 mg/m³ for wood). The corn stover showed a good calorific potential, but it would have to be densified and the boiler should be modified to improve airflow, completeness of combustion and handling of the large amount of ash formed.

The ability of process‐based soil–crop models to simulate potato (Solanum tuberosum L.) yield and N uptake for a range of N fertilization under the conditions of eastern Canada has never been tested. Our objectives were (i) to calibrate and evaluate the performance of the STICS model for the Shepody and Russet Burbank cultivars with cultivar‐specific critical N concentration dilution curves, and (ii) to quantify the gain in model performance with cultivar‐specific N concentration curves rather than a generic curve. Data sets including measurements of leaf area index (LAI), total and tuber biomass, and total and tuber N uptake for several N rates (0–280 kg N ha−1) collected at Charlottetown, PE; Fredericton, NB; and Québec, QC, Canada were used. Calibration was done with one data set from Charlottetown for Shepody and one data set from Québec for Russet Burbank, while all the other data sets were used to evaluate model performance. Following calibration, the STICS model generally performed well, with a normalized root mean square error (NRMSE) < 30% and a normalized mean error (NME) ranging from –8% to 23%, for LAI and biomass. Model performance was slightly worse for total and tuber N uptake, although cultivar‐specific N concentration curves for Shepody and Russet Burbank improved model performance compared with a generic curve with a lower NRMSE (18–50% vs. 21–63%) and NME (–9 to 23% vs. –14 to 23%). Cultivar‐specific critical N concentration curves should therefore be used for assessing the N budget of potato cropping systems.Core Ideas The STICS soil–crop model was calibrated and evaluated for two potato cultivars. Calibrated and evaluated variables were LAI, total and tuber biomass, and total and tuber N uptake. Cultivar‐specific critical N concentration curves are recommended for N budget assessment.

Combustion trials were conducted with corn stover (CS) and wheat straw (WS) round bales in a 176-kW boiler (model Farm 2000). Hot water (80 °C) stored in a 30,000-L water tank was transferred to a turkey barn through a plate exchanger. Gross calorific value measured in the laboratory was 17.0 and 18.9 MJ/kg DM (dry matter) for CS and WS, respectively. Twelve bales of CS (1974 kg DM total, moisture content of 13.6%) were burned over a 52-h period and produced 9.2% ash. Average emissions of CO, NO x and SO 2 were 2725, 9.8 and 2.1 mg/m 3 , respectively. Thermal efficiency was 40.8%. For WS, six bales (940 kg DM total, MC of 15%) were burned over a 28-h period and produced 2.6% ash. Average emissions of CO, NO x and SO 2 were 2210, 40.4 and 3.7 mg/m 3 , respectively. Thermal efficiency was 68.0%. A validation combustion trial performed a year later with 90 CS bales confirmed good heating performance and the potential to lower ash content (6.2% average).

The ability to predict short-rotation coppice (SRC) willow productivity for a given region would be very helpful for large-scale deployment of this crop. The objectives of this study were to calibrate and validate the 3PG model for two commonly used clones (SX64 and SX67) and to provide yield potential estimates for 16 sites across Canada. One dataset for each clone, including leaf area index (LAI) and stem biomass, was used for calibrating parameters controlling leaf and stem growth. All other datasets, coming from eight different willow plantations, were used for model validation. Model performance was good in predicting stem biomass for the SX64 (normalized mean error (NME) = –8%, normalized root mean square error (NRMSE) = 22%) and SX67 (NME = –3%, NRMSE = 16%) clones. Predictions were more scattered for LAI, with NRMSE close to 35% and 33% and NME of 1% and 8% for SX64 and SX67, respectively. The simulation results show that the greatest yields were obtained with the three-year rotation for the SX67 clone, whereas a two-year rotation seemed to be more appropriate for the SX64 clone. The simulation results also show that growing degree-days had a significant impact on yield potential, which varied from 10.5 to 16.5 t DM·ha−1 for SX64 and from 7.5 to 11.5 t DM·ha−1 for SX67.

No abstract

Switchgrass (Panicum virgatum L.) is a herbaceous perennial grass that can be used as bedding for livestock, planted in buffer strips, and used as biofuel, but it is still not widely grown in eastern Canada. The objectives of this study were to verify the performance of the Integrated Farm System Model (IFSM) in simulating switchgrass growth and to estimate its yield potential and production costs in eastern Canada. The performance of IFSM was assessed with dry matter (DM) yield of switchgrass (‘Cave-in-Rock’) measured over three growing seasons (2015–2017) in southern Quebec, Canada. The model performed reasonably well, with normalized root mean square errors of 19.5% for calibration and 27.9% for validation. Simulation results of potential yield and economic management over the long term (1986–2015) for five switchgrass production sites in eastern Canada indicated that average DM yields in Quebec City and Fredericton (9.6 and 9.7 t ha−1, respectively) were significantly lower than average DM yields in Saint-Hubert, Ottawa, and London (10.8, 10.4, and 11.0 t ha−1, respectively). Average annual production costs per tonne of DM for the spring harvest were higher at low-yield sites (CAD$66.67 and $64.50 for Fredericton and Quebec City, respectively) than at high-yield sites ($60.10, $62.82, and $60.08 for Saint-Hubert, Ottawa, and London, respectively). The IFSM-estimated production costs were within the range of the calculated values reported in other agro-economic analyses conducted in Ontario and Quebec.

No abstract