JDS
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Journal of Dairy Science Vol. 85 No. 12 3389-3394
© 2002 by American Dairy Science Association ®
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Monteny, G. J.
Right arrow Articles by de Boer, I.J.M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Monteny, G. J.
Right arrow Articles by de Boer, I.J.M.

Prediction of Ammonia Emission from Dairy Barns using Feed Characteristics Part II: Relation between Urinary Urea Concentration and Ammonia Emission

G. J. Monteny*, M.C.J. Smits*, G. van Duinkerken{dagger}, H. Mollenhorst{ddagger} and I.J.M. de Boer{ddagger}

* Institute of Agricultural and Environmental Engineering
{dagger} Research Institute for Animal Husbandry
{ddagger} Animal Production Systems Group, Wageningen Institute of Animal Sciences Wageningen University and Research Center, Wageningen, The Netherlands

Corresponding author:
G.J. Monteny; e-mail:
g.j.monteny{at}imag.wag-ur.nl.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Emission of NH3 from dairy barns can be reduced substantially by changing the cows’ diet. Emission of NH3 is reduced most effectively when dietary changes result in a reduction of urinary urea concentration. The objective of this research was to predict NH3 emission from dairy barns for various diets, using feed characteristics, and climate, barn, and slurry related parameters. Model results were validated using experimental data. Cows were fed one of nine diets, which was a combination of three rumen degradable protein balances and one of three roughage compositions. Each diet was repeated once. Measured parameters included herd, diet, urine, slurry, barn and climate characteristics, and emission of NH3 from the barn. For a wide range of diets and barn conditions, observed NH3 emission from a dairy barn can be predicted accurately using a combination of existing nutrition-emission models. Accuracy of prediction improved considerably, however, when observed emissions during four diet treatments were omitted due to suspected technical failure of the emission measurement equipment. Results also show that NH3 emissions in common practical situations will range from about 3.3 to 16.3 kg per cow per 190 d. To reduce NH3 emission in practice, farmers should maximize the diet’s grass content, and at the same time, minimize its rumen degradable protein balance level. Currently, however, farmers need additional information to compose such a low-emission diet, which should fulfill also the intestine digestible protein and net energy-lactation requirements of a cow.

Abbreviation key: TAN = concentration of total ammoniacal nitrogen (N, kg•m–3), UUC = urinary urea concentration (N, kg•m–3)

Key Words: ammonia emission • urinary urea concentration • dairy barns • feed characteristics


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Atmospheric NH3, mainly originating from agricultural sources (estimated around 94% in The Netherlands, Sliggers, 2001), can cause serious environmental problems related to soil acidification and eutrophication. Each country of the European Union has designated equivalent emission ceilings to limit such environmental problems. For The Netherlands, for example, maximum annual NH3 emission from 2004 is designated at 128 kton. The Dutch government, however, aims at an even lower annual emission of only 100 kton, by 2010 (Sliggers, 2001).

Monteny and Erisman (1998) reviewed possibilities to reduce NH3 emission from dairy barns. They concluded that reduction of urinary urea concentration (UUC) by nutritional measures would result in a maximum emission reduction of 39% when applied in dairy barns. Moreover, Monteny (2000) stated that a combination of a N-flow model and an emission model for dairy cows (Monteny et al., 1998) would assist animal nutritionists and producers to determine diets that reduce NH3 emission. Such an N-flow model should yield reliable UUC values based on feed characteristics, which, subsequently, can be used as input to the NH3 emission model.

In Part I, two N-flow models for dairy cows are presented and evaluated for their potential to predict UUC, using feed characteristics, i.e., a regression (van Dongen, 1999) and a mechanistic model (van Straalen, 1995). Model results were validated using experimental data. The regression model performed best in terms of prediction of observed UUC. In this paper, therefore, this regression model was used to predict UUC required as input to the NH3 emission model (Monteny et al., 1998). This NH3 emission model also uses barn, climate, and slurry related input parameters.

The objective of this research, therefore, was to predict, for various diets, NH3 emissions from dairy barns using feed characteristics and climate, barn, and slurry related parameters. The model was validated using data from an experiment at the Research Institute for Animal Husbandry in Lelystad.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Nitrogen Flow Models
Prediction of UUC using a regression model is described in detail in Part I (de Boer et al., 2002). In summary, to predict UUC, first, volume produced by a cow is predicted (Bannink et al., 1999). Second, urinary N excretion of a cow is predicted using a regression model (van Dongen, 1999) based on the Dutch protein evaluation system (Tamminga et al., 1994). Subsequently, urinary N concentration was computed as urinary N excretion divided by urine volume. Finally, the relationship between UUC and urinary N concentration was derived from experimental data.

Emission Model
The NH3 emission model used in this study was developed and described in detail by Monteny et al. (1998). In summary, the model consists of three modules: urination module, urine pool module, and pit module.

Urination module.
This module simulates distribution of urinations (i.e., urine pools) over the available (slatted) floor area in the barn. First, the total number of urinations is calculated as:

Formula 1[1]
where nc is the total number of cows and uf is the urination frequency (•cow–1•day–1). Urinations are assumed to be distributed randomly over the available number of urine locations, determined as:

Formula 2[2]
where Afloor is the total area of the (slatted) floor (m2), and Apool is the floor area covered by one urination (m2).

Urine pool module.
This module describes urea conversion and NH3 emission-related processes for each urine pool. When an existing, and thus emitting, urine pool is superseded, however, by a fresh urination, the original pool is washed to the pit, and all processes start again at the conditions valid for the moment of superseding.

Urea in the pool volume (= dpool•Apool, where dpool is the depth of the urine pool in m) is converted to NH3 by the enzyme urease. This conversion is determined by urease activity. In the urine pool, NH3 (unionized) and NH4+ (ionized) are in equilibrium (dissociation). The amount of NH3 dissolved depends on pH and temperature. Henry’s equilibrium is valid for the dissociation of NH3 between the liquid and the gas phase at the pool/air boundary, with temperature as the main determining variable. Finally, volatilization of NH3 occurs at that boundary, depending on air velocity at floor level and pool temperature. In summary, the following processes and corresponding input parameters are relevant (see Figure 1Go): first, urea conversion with inputs UUC (predicted or observed), urease activity, and urine pool volume; second, NH3/NH4+ dissociation with inputs pH and temperature of urine pool; third, NH3 dissociation between the gas and liquid phase with input pool temperature; and fourth, NH3 volatilization, with inputs urination floor area (Apool), temperature of urine pool, and air velocity.


Figure 1
View larger version (14K):
[in this window]
[in a new window]

  		Figure 1. Schematic representation of the urine pool and pit module of the NH3 emission model. A solid line represents a conversion (single arrow) or an equilibrium (double arrow) process. A dashed line represents parameter influence on a specific process.

 
Pit module.
In the pit (top layer of the slurry in the pit) module, urea conversion is not modeled. Instead, the concentration of total ammoniacal N (TAN) is used as input parameter to the subsequent processes: NH3/NH4+ dissociation (inputs: TAN; pH temperature of slurry); gas/liquid NH3 equilibrium (input: slurry temperature); and NH3 volatilization (inputs: floor area of pit, slurry temperature, and air velocity in pit).

Emission prediction.
The urine pool and the pit module yield a prediction of NH3 emission. The sum of both emissions is interpreted as the predicted NH3 emission from the dairy barn.

Inputs Required for NH3 Emission Model
Urination module.
Input data for the urination module were derived from barn design characteristics (see Figure 2Go; Afloor = 207 m2), from management during the experiment (nc = 56), and from literature (Apool = 0.8 m2; Monteny, 2000). Urination frequency depended on the diet and varied from 9 to 11 urinations per cow/d (Smits, personal communications; Table 1Go). For a more detailed description of diets, see Table 1Go of Part I. Given an average frequency of 10 urinations/d, the total number of urine deposition = 560/d and is distributed over 259 (= 207/0.8) locations. This implies that each urine pool, on average, is present for 11 h (259/560•24 h) before being superseded by a fresh one.


Figure 2
View larger version (26K):
[in this window]
[in a new window]

  		Figure 2. Floor plan of the naturally ventilated cubicle dairy barn at the Waiboerhoeve, Lelystad.

 

View this table:
[in this window]
[in a new window]

  		Table 1. Variable input parameters per diet treatment.
 
Urine pool module.
For the urine pool module, UUC was derived using the regression model as described in Part I (de Boer et al., 2002). A constant, maximum value for urease activity (i.e., urea conversion rate of 0.0027 kg•m–3•s–1; Monteny et al., 1998) was assumed, because the barn had been occupied for a long time. The depth of each urine pool (dpool) was derived from literature (approximately 0.5 mm; Monteny, 2000). The pH of a urine pool on the floor was assumed to be 1.0 pH unit higher than pHurine (Monteny, 2000), whereby pHurine was determined in a pooled urine sample (see below). Pool temperature was assumed to equal indoor air temperature, which was continuously recorded during NH3 emission measurements. Air velocity (vfloor) was set at 0.2 m/s (default in Monteny, 2000).

Pit module.
Data on TAN for the pit module were derived from samples of the top layer (upper 5 cm) of the slurry in the pit (see below). The pH of this slurry was assumed to equal 8.6 for all treatments. The corresponding slurry temperature was assumed to equal indoor air temperature, whereas air velocity in the pit was set at 0.05 m/s (default in Monteny, 2000).

Sampling and analysis of slurry and urine.
In the third week of each diet treatment (de Boer et al., 2002), the top layer of the slurry was sampled at four locations through the slats. For this purpose, a special sampling device (cup shape; 100 ml) was attached to a broomstick. Samples were collected in a jar, stored in a cooler, and transported for laboratory analysis. In the laboratory, a pooled sample was analyzed for TAN. In addition, as described in de Boer et al. (2002), a pooled sample of the urine was analyzed for pH.

Overview of variable input data.
Table 1Go shows different values for input parameters that vary in the NH3 emission model. Diet treatments, described in more detail in Table 1Go of de Boer et al. (2002), are presented in chronological order, as can be deducted from the course of temperature in Table 1Go.

Observing NH3 Emission
Emission of NH3 was observed using a concentration ratio method with SF6 (sulfur hexa fluoride) as a tracer gas. In this method, SF6 is injected near the slatted floor through injection points that were attached to the separation boards of the cubicles and feeding fences. This arrangement assures optimal distribution of SF6 near the source of NH3 emission. Air in the top of the building (assuming air exhaust occurring there) was sampled through a duct system with multiple openings and pooled. This pooled sample was analyzed for its NH3 concentration (CNH3; g•m–3) and for its SF6 concentration (CSF6 g•m–3). Assuming complete mixing of NH3 and SF6, the NH3 emission, i.e., mass flux or MFNH3 (g•h–1), was calculated using the following equation, given the known mass flux of SF6, i.e., MFSF6 (g•h–1):

Formula 3[3]
By the end of diet treatment 4, the measurement system was replaced by a newer version, because of suspected technical failure leading to overestimation of the NH3 emission. For that reason, data of NH3 emission observations were separately analyzed for periods 1 to 18 and 5 to 18.

In addition to gas ratio, indoor temperature was measured at four locations using rotronic sensors. Data were collected on an 8-min basis during wk 3 of each treatment. Results were averaged, and these weekly averages of gas ratio and temperature were used in further data analysis.

Finally, observed emissions were used to validate predicted emissions. Theoretically, observed (y) and predicted (x) emissions relate as y = x.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
As described previously, periods 1 to 18 and 5 to 18 were analyzed separately. Figure 3Go shows the relation between observed (x) and predicted (y) NH3 emission for periods 1 to 18 using observed and predicted UUC concentrations. For low emission levels, NH3 emission was predicted accurately, whereas for high levels, predictions were poor, i.e., y = 0.61x + 27 (R2 = 0.67). The intercept and the regression coefficient significantly differed from 0 and 1, respectively (P < 0.05).


Figure 3
View larger version (10K):
[in this window]
[in a new window]

  		Figure 3. Observed versus predicted NH3 emissions (g/h) of all 18 diet treatments using the observed (• solid trend line) and predicted ({blacktriangleup} dashed trend line) urinary urea concentration (UUC). Dotted line represents y = x.

 
Figure 4Go shows the relation between observed (x) and predicted (y) NH3 emissions, for periods 5 to 18 only, using observed and predicted UUC concentrations. Comparison of Figure 3Go and Figure 4Go shows that omitting data for periods 1 to 4 improved prediction of NH3 emission substantially, i.e., y = 0.91 + 5 (R2 = 0.94). Hence, results from Figures 3 and 4GoGo confirm our initial doubt about observed NH3 emissions in periods 1 to 4. It is most likely to assume that this was caused by technical failure of the measurement system, e.g., the gas chromatograph or the injection system. This could not, however, be checked.


Figure 4
View larger version (10K):
[in this window]
[in a new window]

  		Figure 4. Observed versus predicted NH3 emissions (g/h) of treatments 5 to 18 using the observed (• solid trend line) and predicted ({blacktriangleup} dashed trend line) urinary urea concentration (UUC). Dotted line represents y = x.

 
Using observed instead of predicted UUC concentration resulted in a close to constant increase in the predicted NH3 emission for the whole range of observations. The value of the regression coefficient increased from 0.91 to 0.94 (Figure 4Go). Moreover, the intercept increased from 5 to 15 g•h–1, indicating over-prediction of NH3 emission at lower NH3 emission values.

Overall, results show that for a wide range diets and barn conditions, observed NH3 emission from a dairy barn can be predicted accurately using a combination of existing nutrition-emission models.

Figure 5Go shows the relationship between observed UUC (x) and TAN (y), computed from data in Table 1Go. Results show that percentages of TAN of UUC linearly decreases as UUC increases. This implies that the increase in UUC is larger (ranges from 2 to 12 kg N/m) than the increase in TAN concentration in the slurry top layer of the pit (ranges from only 1.2 to 2.5 kg N/m). This is due to the mixing of urine with feces, containing little or no ammoniacal N and with farm management related aspects like discharge of waste water to the pit (e.g., from cleaning the milking parlor).


Figure 5
View larger version (8K):
[in this window]
[in a new window]

  		Figure 5. Observed concentrations of total ammoniacal N in the slurry top layer (TAN) as a percentage of observed urinary urea concentrations (UUC).

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
Input Data
Except for air velocity and slurry pH (default values were used), all input data for the combined nutrition and NH3 emission model were collected at the experimental farm The Waiboerhoeve, in Lelystad. For a wide range of diets and barn conditions, observed NH3 emission from this barn could be predicted accurately using this combination of existing nutrition-emission models. At this moment, however, prediction of NH3 emission from dairy barns requires, in addition to feed, herd, and barn characteristics, measurement of urinary pH and slurry TAN. The NH3 emission model appears very sensitive to small changes in urinary pH (Cahn, 1998; Monteny, 2000). Modeling of urinary pH, given only feed characteristics, currently is not possible (Sommer and Husted, 1995; Sommer and Sherlock, 1996). Consequently, reliable prediction of NH3 emission requires measurement of urinary pH.

Like urinary pH, TAN concentration is an essential input to the NH3 emission model because it determines the maximum amount of N (at high pH values) available for volatilization. Currently, labor-intensive slurry sampling is necessary to quantify TAN concentration in the top layer of the slurry pit. The observed relationship between UUC and its percentage of TAN, however, offers a basis for reliable (see Figure 5Go) prediction of TAN given predicted values of UUC. For this purpose, this relationship between UUC and its percentage of TAN has to be determined for various management practices first.

Temperature of a urine pool and of the top layer of the slurry pit were assumed to equal indoor air temperature. The thin layer of urine and slurry, and consequently their small heat capacity, may support this assumption. Moreover, the NH3 emission model is far less sensitive to variations in temperature than in pH (Monteny, 2000). Hence, the impact of this assumption on final prediction of NH3 emission is small.

Practical Relevance
The combined nutrition-emission model appears to be a useful tool to assess the impact of dairy cow nutrition measures on NH3 emission from dairy cow barns. Diets used in the experiment are likely to represent the range in diets used on commercial dairy farms. Consequently, NH3 emission can be reduced by 80%, from 16.3 (200 g/h) to 3.3 (40 g/h) kg NH3 per cow place per 190-d housing period, by changing from a corn-based diet with high OEB level to a grass-based diet with OEB level around zero. To reduce NH3 emission in practice, therefore, farmers should maximize the diet’s grass content, and at the same time, minimize its OEB level. Current farmers need additional information to compose such a low-emission diet, which should fulfil also the DVE and NEL requirements of a cow.

Measurements show that the range in NH3 emission in practice will be significant. For comparison, all cubicle dairy barns in The Netherlands normatively emit 8.8 kg per cow place per 190-d housing period, whereas NH3 emission during the current experiment ranged from 3.3 to 16.3 kg per cow place per 190-d housing period.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
For a wide range of diets and barn conditions, observed NH3 emission from a dairy barn can be predicted accurately using a combination of a regression model for prediction of UUC and a mechanistic model for prediction of NH3 emission. Overprediction of UUC has a similar effect on prediction of NH3 emission, disregarding the level of NH3 emission. Due to lack of reliable prediction of urinary pH as a function of the diet, measurement of urinary pH will remain necessary for adequate prediction of NH3 emission. The observed linear relationship between predicted UUC and its percentage of TAN offers a basis to omit slurry sampling currently required to quantify TAN concentration in the slurry top layer of the pit.

Received for publication March 15, 2002. Accepted for publication July 9, 2002.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 


Bannink, A., H. Valk, and A. M. van Vuuren. 1999. Intake and excretion of sodium, potassium and nitrogen and the effects on urine production by lactating dairy cows. J. Dairy Sci. 82:1008–1018.[Abstract]

Cahn, T. T. 1998. Ammonia emission from excreta of growing-finishing pigs as affected by dietary composition. Ph.D. Diss., Wageningen Univ., Wageningen.

de Boer, I.J.M., M.C.J. Smits, H. Mollenhorst, G. van Duinkerken, and G. J. Monteny. 2002. Prediction of ammonia emission from dairy barns using feed characteristics. Part I: Relation between feed characteristics and urinary urea concentration. J. Dairy Sci. 85:3382–3388.[Abstract/Free Full Text]

Monteny, G. J. 2000. Modeling of ammonia emissions from dairy cow houses. Ph.D. Thesis, Wageningen Univ., Report 2000-11, Institute of Agric. and Environmental Engineering, Wageningen.

Monteny, G. J., and J. W. Erisman. 1998. Ammonia emission from dairy cow buildings: a review of measurement techniques, influencing factors and possibilities for reduction. Neth. J. Agric. Sci. 46:225–247.

Monteny, G. J., D. D. Schulte, A. Elzing, and E.J.J. Lamaker. 1998. A conceptual mechanistic model for the ammonia emissions from free stall cubicle dairy cow houses. Transactions of the ASAE 41(1):193–201.

Sliggers (Ed.), J. 2001. Towards sustainable levels for health and nature: review document theme Acidification and Transboundary Air Pollution (in Dutch). Report VROM 010334/h/10-01 17529/187, Ministry of Housing, Spatial Planning and Environment (VROM), The Hague.

Smits, M.C.J. 2001. Personal communications, Institute of Agric. and Environmental Engineering, Wageningen.

Sommer, S. G., and S. Husted. 1995. A simple model of pH in slurry. J. Agric. Sci. 124:447–453.

Sommer, S. G., and R. R. Sherlock. 1996. pH and buffer component dynamics in the surface layers of animal slurries. J. Agric. Sci. 127:109–116.

Tamminga, S., W. M. van Straalen, A.P.J. Subnel, R.G.M. Meijer, A. Steg, C.J.G. Wever, and M. C. Blok. 1994. The Dutch protein evaluation system: the DVE/OEB system. Livest. Prod. Sci. 40:139-155.

van Dongen, C.F.J. 1999. Dairy cattle feeding and ammonia emission on practical farms. Pages 21–27 in Annual on Fertilizers and Fertilization, Nutrient Management Inst., Wageningen.

van Straalen, W.M. 1995. Modeling of nitrogen flow and excretion in dairy cows. Ph.D. Diss, Wageningen Univ., Wageningen.


This article has been cited by other articles:


Home page
 J. Environ. Qual.Home page
L. Li, J. Cyriac, K. F. Knowlton, L. C. Marr, S. W. Gay, M. D. Hanigan, and J. A. Ogejo
Effects of Reducing Dietary Nitrogen on Ammonia Emissions from Manure on the Floor of a Naturally Ventilated Free Stall Dairy Barn at Low (0-20{degrees}C) Temperatures
J. Environ. Qual., October 29, 2009; 38(6): 2172 - 2181.
[Abstract] [Full Text] [PDF]

Home page
 J DAIRY SCIHome page
G. van Duinkerken, G. Andre, M. C. J. Smits, G. J. Monteny, and L. B. J. Sebek
Effect of Rumen-Degradable Protein Balance and Forage Type on Bulk Milk Urea Concentration and Emission of Ammonia from Dairy Cow Houses
J Dairy Sci, March 1, 2005; 88(3): 1099 - 1112.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Monteny, G. J.
Right arrow Articles by de Boer, I.J.M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Monteny, G. J.
Right arrow Articles by de Boer, I.J.M.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS