Refining a model-based assessment strategy to estimate the ammonia emission from floors in dairy cow houses

Dennis J.W. Snoek

Research output: Thesisinternal PhD, WUAcademic

Abstract

Ammonia (NH3) emission is still high, and agriculture is still the dominant contributor. In The Netherlands, the NH3 emission from dairy cow houses is one of the most important sources. A lot of research has been conducted to understand and model NH3 emission, to measure it, and to reduce it using identified and developed reduction measures. However, our understanding of how to measure and how to reduce the NH3 emission is still limited. In addition, the set emission ceilings were lowered for 2020.

The objective of this thesis was to refine a model-based assessment strategy to estimate the ammonia emission from floors in dairy cow houses. First the most important input variables and process parameters were identified with a sensitivity analysis in currently available mechanistic NH3 emission models and theory. It was concluded that five puddle related input variables caused the largest variation in NH3 emission estimation, being the puddle pH, depth (Dp), urinary urea nitrogen concentration (UUN), surface area (Ap), and temperature (Tliq). For each input variable the available data was scarce, and it was therefore recommended to measure these five most important variables in practice. However, measurement methods were hardly available. Therefore, sensors were chosen, new measurement methods were developed, and these were combined in a protocol to measure the pH, Dp, UUN, Ap and Tliq of fresh, random and manually created urine puddles in commercial dairy cow houses.

In total 16 commercial dairy cow houses were assessed in a factorial experimental setup based on four floor-management types in two Seasons, with PREclean treatment. PREclean represented intense-floor-cleaning that was compared to on-farm manure scraping. A V-shaped asphalt floor had significantly larger values for both Ap (1.04 m2) and Dp (1.5 mm) than did the slatted and grooved floors (0.76 m2, 0.93 mm). For both Ap and Dp the variation within a farm was large, but was negligible between farms. The Dp values and variation were 3 to 6 times larger than currently assumed. With PREclean treatment the Dp resulted in about 3 times lower values compared to the on-farm scraping. In short, the potential NH3 emission reduction of good floor cleaning is large. Overall mean values were 4.27 kg m-3 for UUN, an initial pH(t=0) of 8.3, both in fresh puddles, and a pH(t=ξ) of 9.0 for random puddles at a random time. For UUN both the variation within and between farms was large, whereas the variation for pH was small. Both the mean UUN and pH showed lower values than currently assumed. In a separate 4 h time series experiment at 3 commercial farms was shown that the pH, on average, quickly increased initially, declined after 1 h and then became stable. The calculated NH3 in kg puddle-1 showed a huge range and was considerably larger than currently assumed for the reference situation.

Compared to the aforementioned sensitivity analysis outcome, the UUN range at farm level is both slightly smaller and shifts to slightly lower values, while for Dp the range and values are both larger. These two variables caused the largest variation in the estimated NH3 emissions, and not the pH. In conclusion, these two variables certainly need to be measured in individual commercial dairy cow houses to estimate the NH3 emission. For Ap, pH and Tair the measured ranges at farm level were less large. The pH turns out to be fairly stable in commercial cow houses and, related to that, it causes less variation in the estimated NH3 emission. Nevertheless, the pH still ranks as the third most important variable, and therefore needs to be measured in individual cow houses. The Ap is fairly stable between farms, but varies within farms and it still has a significant effect on the NH3 emission. The floor design clearly affects the puddle area Ap. Therefore, it is not necessary to measure Ap at each individual farm, but it is sufficient to measure the Ap in only one commercial cow house per floor design. The Tair variable is of limited importance compared to the aforementioned four variables, but it is still significant.

LanguageEnglish
QualificationDoctor of Philosophy
Awarding Institution
  • Wageningen University
Supervisors/Advisors
  • Groot Koerkamp, Peter, Promotor
  • Ogink, Nico, Co-promotor
  • Stigter, Hans, Co-promotor
Award date27 Oct 2016
Place of PublicationWageningen
Publisher
Electronic ISBNs9789462578852
DOIs
Publication statusPublished - 2016

Fingerprint

refining
ammonia
dairy cows
farms
cows
cleaning
bitumen
commercial farms
urea nitrogen
animal manures
surface temperature
sensors (equipment)
time series analysis
Netherlands
surface area
urine

Keywords

  • dairy cows
  • stalls
  • ammonia emission
  • floors
  • modeling
  • mitigation
  • sensors
  • ph
  • temperature
  • urea

Cite this

Snoek, Dennis J.W.. / Refining a model-based assessment strategy to estimate the ammonia emission from floors in dairy cow houses. Wageningen : Wageningen University, 2016. 182 p.
@phdthesis{015449935ec24f959d413dd11905c865,
title = "Refining a model-based assessment strategy to estimate the ammonia emission from floors in dairy cow houses",
abstract = "Ammonia (NH3) emission is still high, and agriculture is still the dominant contributor. In The Netherlands, the NH3 emission from dairy cow houses is one of the most important sources. A lot of research has been conducted to understand and model NH3 emission, to measure it, and to reduce it using identified and developed reduction measures. However, our understanding of how to measure and how to reduce the NH3 emission is still limited. In addition, the set emission ceilings were lowered for 2020. The objective of this thesis was to refine a model-based assessment strategy to estimate the ammonia emission from floors in dairy cow houses. First the most important input variables and process parameters were identified with a sensitivity analysis in currently available mechanistic NH3 emission models and theory. It was concluded that five puddle related input variables caused the largest variation in NH3 emission estimation, being the puddle pH, depth (Dp), urinary urea nitrogen concentration (UUN), surface area (Ap), and temperature (Tliq). For each input variable the available data was scarce, and it was therefore recommended to measure these five most important variables in practice. However, measurement methods were hardly available. Therefore, sensors were chosen, new measurement methods were developed, and these were combined in a protocol to measure the pH, Dp, UUN, Ap and Tliq of fresh, random and manually created urine puddles in commercial dairy cow houses. In total 16 commercial dairy cow houses were assessed in a factorial experimental setup based on four floor-management types in two Seasons, with PREclean treatment. PREclean represented intense-floor-cleaning that was compared to on-farm manure scraping. A V-shaped asphalt floor had significantly larger values for both Ap (1.04 m2) and Dp (1.5 mm) than did the slatted and grooved floors (0.76 m2, 0.93 mm). For both Ap and Dp the variation within a farm was large, but was negligible between farms. The Dp values and variation were 3 to 6 times larger than currently assumed. With PREclean treatment the Dp resulted in about 3 times lower values compared to the on-farm scraping. In short, the potential NH3 emission reduction of good floor cleaning is large. Overall mean values were 4.27 kg m-3 for UUN, an initial pH(t=0) of 8.3, both in fresh puddles, and a pH(t=ξ) of 9.0 for random puddles at a random time. For UUN both the variation within and between farms was large, whereas the variation for pH was small. Both the mean UUN and pH showed lower values than currently assumed. In a separate 4 h time series experiment at 3 commercial farms was shown that the pH, on average, quickly increased initially, declined after 1 h and then became stable. The calculated NH3 in kg puddle-1 showed a huge range and was considerably larger than currently assumed for the reference situation. Compared to the aforementioned sensitivity analysis outcome, the UUN range at farm level is both slightly smaller and shifts to slightly lower values, while for Dp the range and values are both larger. These two variables caused the largest variation in the estimated NH3 emissions, and not the pH. In conclusion, these two variables certainly need to be measured in individual commercial dairy cow houses to estimate the NH3 emission. For Ap, pH and Tair the measured ranges at farm level were less large. The pH turns out to be fairly stable in commercial cow houses and, related to that, it causes less variation in the estimated NH3 emission. Nevertheless, the pH still ranks as the third most important variable, and therefore needs to be measured in individual cow houses. The Ap is fairly stable between farms, but varies within farms and it still has a significant effect on the NH3 emission. The floor design clearly affects the puddle area Ap. Therefore, it is not necessary to measure Ap at each individual farm, but it is sufficient to measure the Ap in only one commercial cow house per floor design. The Tair variable is of limited importance compared to the aforementioned four variables, but it is still significant.",
keywords = "dairy cows, stalls, ammonia emission, floors, modeling, mitigation, sensors, ph, temperature, urea, melkkoeien, stallen, ammoniakemissie, vloeren, modelleren, mitigatie, sensors, ph, temperatuur, ureum",
author = "Snoek, {Dennis J.W.}",
note = "WU thesis 6483 Includes bibliographic references. - With summary in English",
year = "2016",
doi = "10.18174/387486",
language = "English",
publisher = "Wageningen University",
school = "Wageningen University",

}

Refining a model-based assessment strategy to estimate the ammonia emission from floors in dairy cow houses. / Snoek, Dennis J.W.

Wageningen : Wageningen University, 2016. 182 p.

Research output: Thesisinternal PhD, WUAcademic

TY - THES

T1 - Refining a model-based assessment strategy to estimate the ammonia emission from floors in dairy cow houses

AU - Snoek, Dennis J.W.

N1 - WU thesis 6483 Includes bibliographic references. - With summary in English

PY - 2016

Y1 - 2016

N2 - Ammonia (NH3) emission is still high, and agriculture is still the dominant contributor. In The Netherlands, the NH3 emission from dairy cow houses is one of the most important sources. A lot of research has been conducted to understand and model NH3 emission, to measure it, and to reduce it using identified and developed reduction measures. However, our understanding of how to measure and how to reduce the NH3 emission is still limited. In addition, the set emission ceilings were lowered for 2020. The objective of this thesis was to refine a model-based assessment strategy to estimate the ammonia emission from floors in dairy cow houses. First the most important input variables and process parameters were identified with a sensitivity analysis in currently available mechanistic NH3 emission models and theory. It was concluded that five puddle related input variables caused the largest variation in NH3 emission estimation, being the puddle pH, depth (Dp), urinary urea nitrogen concentration (UUN), surface area (Ap), and temperature (Tliq). For each input variable the available data was scarce, and it was therefore recommended to measure these five most important variables in practice. However, measurement methods were hardly available. Therefore, sensors were chosen, new measurement methods were developed, and these were combined in a protocol to measure the pH, Dp, UUN, Ap and Tliq of fresh, random and manually created urine puddles in commercial dairy cow houses. In total 16 commercial dairy cow houses were assessed in a factorial experimental setup based on four floor-management types in two Seasons, with PREclean treatment. PREclean represented intense-floor-cleaning that was compared to on-farm manure scraping. A V-shaped asphalt floor had significantly larger values for both Ap (1.04 m2) and Dp (1.5 mm) than did the slatted and grooved floors (0.76 m2, 0.93 mm). For both Ap and Dp the variation within a farm was large, but was negligible between farms. The Dp values and variation were 3 to 6 times larger than currently assumed. With PREclean treatment the Dp resulted in about 3 times lower values compared to the on-farm scraping. In short, the potential NH3 emission reduction of good floor cleaning is large. Overall mean values were 4.27 kg m-3 for UUN, an initial pH(t=0) of 8.3, both in fresh puddles, and a pH(t=ξ) of 9.0 for random puddles at a random time. For UUN both the variation within and between farms was large, whereas the variation for pH was small. Both the mean UUN and pH showed lower values than currently assumed. In a separate 4 h time series experiment at 3 commercial farms was shown that the pH, on average, quickly increased initially, declined after 1 h and then became stable. The calculated NH3 in kg puddle-1 showed a huge range and was considerably larger than currently assumed for the reference situation. Compared to the aforementioned sensitivity analysis outcome, the UUN range at farm level is both slightly smaller and shifts to slightly lower values, while for Dp the range and values are both larger. These two variables caused the largest variation in the estimated NH3 emissions, and not the pH. In conclusion, these two variables certainly need to be measured in individual commercial dairy cow houses to estimate the NH3 emission. For Ap, pH and Tair the measured ranges at farm level were less large. The pH turns out to be fairly stable in commercial cow houses and, related to that, it causes less variation in the estimated NH3 emission. Nevertheless, the pH still ranks as the third most important variable, and therefore needs to be measured in individual cow houses. The Ap is fairly stable between farms, but varies within farms and it still has a significant effect on the NH3 emission. The floor design clearly affects the puddle area Ap. Therefore, it is not necessary to measure Ap at each individual farm, but it is sufficient to measure the Ap in only one commercial cow house per floor design. The Tair variable is of limited importance compared to the aforementioned four variables, but it is still significant.

AB - Ammonia (NH3) emission is still high, and agriculture is still the dominant contributor. In The Netherlands, the NH3 emission from dairy cow houses is one of the most important sources. A lot of research has been conducted to understand and model NH3 emission, to measure it, and to reduce it using identified and developed reduction measures. However, our understanding of how to measure and how to reduce the NH3 emission is still limited. In addition, the set emission ceilings were lowered for 2020. The objective of this thesis was to refine a model-based assessment strategy to estimate the ammonia emission from floors in dairy cow houses. First the most important input variables and process parameters were identified with a sensitivity analysis in currently available mechanistic NH3 emission models and theory. It was concluded that five puddle related input variables caused the largest variation in NH3 emission estimation, being the puddle pH, depth (Dp), urinary urea nitrogen concentration (UUN), surface area (Ap), and temperature (Tliq). For each input variable the available data was scarce, and it was therefore recommended to measure these five most important variables in practice. However, measurement methods were hardly available. Therefore, sensors were chosen, new measurement methods were developed, and these were combined in a protocol to measure the pH, Dp, UUN, Ap and Tliq of fresh, random and manually created urine puddles in commercial dairy cow houses. In total 16 commercial dairy cow houses were assessed in a factorial experimental setup based on four floor-management types in two Seasons, with PREclean treatment. PREclean represented intense-floor-cleaning that was compared to on-farm manure scraping. A V-shaped asphalt floor had significantly larger values for both Ap (1.04 m2) and Dp (1.5 mm) than did the slatted and grooved floors (0.76 m2, 0.93 mm). For both Ap and Dp the variation within a farm was large, but was negligible between farms. The Dp values and variation were 3 to 6 times larger than currently assumed. With PREclean treatment the Dp resulted in about 3 times lower values compared to the on-farm scraping. In short, the potential NH3 emission reduction of good floor cleaning is large. Overall mean values were 4.27 kg m-3 for UUN, an initial pH(t=0) of 8.3, both in fresh puddles, and a pH(t=ξ) of 9.0 for random puddles at a random time. For UUN both the variation within and between farms was large, whereas the variation for pH was small. Both the mean UUN and pH showed lower values than currently assumed. In a separate 4 h time series experiment at 3 commercial farms was shown that the pH, on average, quickly increased initially, declined after 1 h and then became stable. The calculated NH3 in kg puddle-1 showed a huge range and was considerably larger than currently assumed for the reference situation. Compared to the aforementioned sensitivity analysis outcome, the UUN range at farm level is both slightly smaller and shifts to slightly lower values, while for Dp the range and values are both larger. These two variables caused the largest variation in the estimated NH3 emissions, and not the pH. In conclusion, these two variables certainly need to be measured in individual commercial dairy cow houses to estimate the NH3 emission. For Ap, pH and Tair the measured ranges at farm level were less large. The pH turns out to be fairly stable in commercial cow houses and, related to that, it causes less variation in the estimated NH3 emission. Nevertheless, the pH still ranks as the third most important variable, and therefore needs to be measured in individual cow houses. The Ap is fairly stable between farms, but varies within farms and it still has a significant effect on the NH3 emission. The floor design clearly affects the puddle area Ap. Therefore, it is not necessary to measure Ap at each individual farm, but it is sufficient to measure the Ap in only one commercial cow house per floor design. The Tair variable is of limited importance compared to the aforementioned four variables, but it is still significant.

KW - dairy cows

KW - stalls

KW - ammonia emission

KW - floors

KW - modeling

KW - mitigation

KW - sensors

KW - ph

KW - temperature

KW - urea

KW - melkkoeien

KW - stallen

KW - ammoniakemissie

KW - vloeren

KW - modelleren

KW - mitigatie

KW - sensors

KW - ph

KW - temperatuur

KW - ureum

U2 - 10.18174/387486

DO - 10.18174/387486

M3 - internal PhD, WU

PB - Wageningen University

CY - Wageningen

ER -