The SF6 tracer technique for methane emission estimation: The normalized mixing ratio of SF6 across days of breath sampling

GERE J. I.; WILLIAMS K. E.; GRATTON R.; JULIARENA P.; PINARES-PATIÑO C.

Banff

Conferencia; GREENHOUSE GASES AND ANIMAL AGRICULTURE CONFERENCE; 2010

The SF6 tracer technique is widely used for estimating methane (CH4) emissions, yet the mechanisms by which SF6 mixes with fermentation gases and is released from the rumen are not clear. In particular, short time-scale irregularities in SF6 release seem to occur. This issue is investigated here, where data from an experiment involving daily and extended (across 5 and 10 days) breath sample collection periods were analysed concerning the mixing ratios of SF6. Eight rumen-fistulated non-lactating cows were housed in a facility enabling both natural and artificial ventilation and fed commercial silage to achieve a fixed and common daily feed intake of 6.4 kg dry matter per cow. Each cow was deployed with a fresh SF6 permeation tube and after a 10-day diet acclimatization, breath collections for CH4 emission estimation using the SF6 technique were conducted. Eight sample lines were fitted to collection harnesses ensuring that a common mix of breath sample was available to each sampling line. Four lines collected breath samples into stainless steel cylinders using a steel bearing ball flow regulator (UNICENArgentina system, ARG), while the remaining 4 lines collected samples into PVC yokes using a capillary system as a flow regulator (AgResearch-New Zealand system, NZL). Within each system and within the 10-day timeframe, common-source breath samples were collected either daily (ARG-1 and NZL-1), across two 5-day periods (ARG 1-5, ARG 6-10, NZL 1-5, NZL 6-10) or one 10-day period (ARG 1-10, NZL 1-10). Both the ARG 1-10 and NZL 1-10 collections were conducted in duplicate. It was assumed that during extended sample collection periods, the time-averaging of micrometeorological and other conditions influencing gas mixing ratios near the animal’s snouts reduced the uncontrollable external variations. As SF6 release rates from each permeation tube are known and assumed to continue at the same rate in the rumen, a ‘normalized’ SF6 mixing ratio (NR) was found by subtracting the background mixing ratio from the observed values and dividing the difference by the release rate of each permeation tube. The NR for the ARG and NZL correlated well and the NR for the 5-day and 10-day extended sample collection periods were similar to those of average consecutive 5 and 10 daily sample collections, respectively. The between-animal variability of NR averaged over 10 days was larger than the within-animal variability of NR (35 vs. 10%). This suggests that some factors related to individual animals may affect NR. However, the effects of pathways and release rates of SF6 from the animal could not be excluded, but if present, their effects should have been largely smoothed out by the timeaveraging effect of the extended sample collection periods. mechanisms by which SF6 mixes with fermentation gases and is released from the rumen are not clear. In particular, short time-scale irregularities in SF6 release seem to occur. This issue is investigated here, where data from an experiment involving daily and extended (across 5 and 10 days) breath sample collection periods were analysed concerning the mixing ratios of SF6. Eight rumen-fistulated non-lactating cows were housed in a facility enabling both natural and artificial ventilation and fed commercial silage to achieve a fixed and common daily feed intake of 6.4 kg dry matter per cow. Each cow was deployed with a fresh SF6 permeation tube and after a 10-day diet acclimatization, breath collections for CH4 emission estimation using the SF6 technique were conducted. Eight sample lines were fitted to collection harnesses ensuring that a common mix of breath sample was available to each sampling line. Four lines collected breath samples into stainless steel cylinders using a steel bearing ball flow regulator (UNICENArgentina system, ARG), while the remaining 4 lines collected samples into PVC yokes using a capillary system as a flow regulator (AgResearch-New Zealand system, NZL). Within each system and within the 10-day timeframe, common-source breath samples were collected either daily (ARG-1 and NZL-1), across two 5-day periods (ARG 1-5, ARG 6-10, NZL 1-5, NZL 6-10) or one 10-day period (ARG 1-10, NZL 1-10). Both the ARG 1-10 and NZL 1-10 collections were conducted in duplicate. It was assumed that during extended sample collection periods, the time-averaging of micrometeorological and other conditions influencing gas mixing ratios near the animal’s snouts reduced the uncontrollable external variations. As SF6 release rates from each permeation tube are known and assumed to continue at the same rate in the rumen, a ‘normalized’ SF6 mixing ratio (NR) was found by subtracting the background mixing ratio from the observed values and dividing the difference by the release rate of each permeation tube. The NR for the ARG and NZL correlated well and the NR for the 5-day and 10-day extended sample collection periods were similar to those of average consecutive 5 and 10 daily sample collections, respectively. The between-animal variability of NR averaged over 10 days was larger than the within-animal variability of NR (35 vs. 10%). This suggests that some factors related to individual animals may affect NR. However, the effects of pathways and release rates of SF6 from the animal could not be excluded, but if present, their effects should have been largely smoothed out by the timeaveraging effect of the extended sample collection periods. mechanisms by which SF6 mixes with fermentation gases and is released from the rumen are not clear. In particular, short time-scale irregularities in SF6 release seem to occur. This issue is investigated here, where data from an experiment involving daily and extended (across 5 and 10 days) breath sample collection periods were analysed concerning the mixing ratios of SF6. Eight rumen-fistulated non-lactating cows were housed in a facility enabling both natural and artificial ventilation and fed commercial silage to achieve a fixed and common daily feed intake of 6.4 kg dry matter per cow. Each cow was deployed with a fresh SF6 permeation tube and after a 10-day diet acclimatization, breath collections for CH4 emission estimation using the SF6 technique were conducted. Eight sample lines were fitted to collection harnesses ensuring that a common mix of breath sample was available to each sampling line. Four lines collected breath samples into stainless steel cylinders using a steel bearing ball flow regulator (UNICENArgentina system, ARG), while the remaining 4 lines collected samples into PVC yokes using a capillary system as a flow regulator (AgResearch-New Zealand system, NZL). Within each system and within the 10-day timeframe, common-source breath samples were collected either daily (ARG-1 and NZL-1), across two 5-day periods (ARG 1-5, ARG 6-10, NZL 1-5, NZL 6-10) or one 10-day period (ARG 1-10, NZL 1-10). Both the ARG 1-10 and NZL 1-10 collections were conducted in duplicate. It was assumed that during extended sample collection periods, the time-averaging of micrometeorological and other conditions influencing gas mixing ratios near the animal’s snouts reduced the uncontrollable external variations. As SF6 release rates from each permeation tube are known and assumed to continue at the same rate in the rumen, a ‘normalized’ SF6 mixing ratio (NR) was found by subtracting the background mixing ratio from the observed values and dividing the difference by the release rate of each permeation tube. The NR for the ARG and NZL correlated well and the NR for the 5-day and 10-day extended sample collection periods were similar to those of average consecutive 5 and 10 daily sample collections, respectively. The between-animal variability of NR averaged over 10 days was larger than the within-animal variability of NR (35 vs. 10%). This suggests that some factors related to individual animals may affect NR. However, the effects of pathways and release rates of SF6 from the animal could not be excluded, but if present, their effects should have been largely smoothed out by the timeaveraging effect of the extended sample collection periods. 6 tracer technique is widely used for estimating methane (CH4) emissions, yet the mechanisms by which SF6 mixes with fermentation gases and is released from the rumen are not clear. In particular, short time-scale irregularities in SF6 release seem to occur. This issue is investigated here, where data from an experiment involving daily and extended (across 5 and 10 days) breath sample collection periods were analysed concerning the mixing ratios of SF6. Eight rumen-fistulated non-lactating cows were housed in a facility enabling both natural and artificial ventilation and fed commercial silage to achieve a fixed and common daily feed intake of 6.4 kg dry matter per cow. Each cow was deployed with a fresh SF6 permeation tube and after a 10-day diet acclimatization, breath collections for CH4 emission estimation using the SF6 technique were conducted. Eight sample lines were fitted to collection harnesses ensuring that a common mix of breath sample was available to each sampling line. Four lines collected breath samples into stainless steel cylinders using a steel bearing ball flow regulator (UNICENArgentina system, ARG), while the remaining 4 lines collected samples into PVC yokes using a capillary system as a flow regulator (AgResearch-New Zealand system, NZL). Within each system and within the 10-day timeframe, common-source breath samples were collected either daily (ARG-1 and NZL-1), across two 5-day periods (ARG 1-5, ARG 6-10, NZL 1-5, NZL 6-10) or one 10-day period (ARG 1-10, NZL 1-10). Both the ARG 1-10 and NZL 1-10 collections were conducted in duplicate. It was assumed that during extended sample collection periods, the time-averaging of micrometeorological and other conditions influencing gas mixing ratios near the animal’s snouts reduced the uncontrollable external variations. As SF6 release rates from each permeation tube are known and assumed to continue at the same rate in the rumen, a ‘normalized’ SF6 mixing ratio (NR) was found by subtracting the background mixing ratio from the observed values and dividing the difference by the release rate of each permeation tube. The NR for the ARG and NZL correlated well and the NR for the 5-day and 10-day extended sample collection periods were similar to those of average consecutive 5 and 10 daily sample collections, respectively. The between-animal variability of NR averaged over 10 days was larger than the within-animal variability of NR (35 vs. 10%). This suggests that some factors related to individual animals may affect NR. However, the effects of pathways and release rates of SF6 from the animal could not be excluded, but if present, their effects should have been largely smoothed out by the timeaveraging effect of the extended sample collection periods. 6 mixes with fermentation gases and is released from the rumen are not clear. In particular, short time-scale irregularities in SF6 release seem to occur. This issue is investigated here, where data from an experiment involving daily and extended (across 5 and 10 days) breath sample collection periods were analysed concerning the mixing ratios of SF6. Eight rumen-fistulated non-lactating cows were housed in a facility enabling both natural and artificial ventilation and fed commercial silage to achieve a fixed and common daily feed intake of 6.4 kg dry matter per cow. Each cow was deployed with a fresh SF6 permeation tube and after a 10-day diet acclimatization, breath collections for CH4 emission estimation using the SF6 technique were conducted. Eight sample lines were fitted to collection harnesses ensuring that a common mix of breath sample was available to each sampling line. Four lines collected breath samples into stainless steel cylinders using a steel bearing ball flow regulator (UNICENArgentina system, ARG), while the remaining 4 lines collected samples into PVC yokes using a capillary system as a flow regulator (AgResearch-New Zealand system, NZL). Within each system and within the 10-day timeframe, common-source breath samples were collected either daily (ARG-1 and NZL-1), across two 5-day periods (ARG 1-5, ARG 6-10, NZL 1-5, NZL 6-10) or one 10-day period (ARG 1-10, NZL 1-10). Both the ARG 1-10 and NZL 1-10 collections were conducted in duplicate. It was assumed that during extended sample collection periods, the time-averaging of micrometeorological and other conditions influencing gas mixing ratios near the animal’s snouts reduced the uncontrollable external variations. As SF6 release rates from each permeation tube are known and assumed to continue at the same rate in the rumen, a ‘normalized’ SF6 mixing ratio (NR) was found by subtracting the background mixing ratio from the observed values and dividing the difference by the release rate of each permeation tube. The NR for the ARG and NZL correlated well and the NR for the 5-day and 10-day extended sample collection periods were similar to those of average consecutive 5 and 10 daily sample collections, respectively. The between-animal variability of NR averaged over 10 days was larger than the within-animal variability of NR (35 vs. 10%). This suggests that some factors related to individual animals may affect NR. However, the effects of pathways and release rates of SF6 from the animal could not be excluded, but if present, their effects should have been largely smoothed out by the timeaveraging effect of the extended sample collection periods.