flow-through chamber system Zifei Liu, Lingjuan Wang, David B. - - PowerPoint PPT Presentation

Zifei Liu, Lingjuan Wang, David B. Beasley Department of Biological and Agricultural Engineering

Modeling ammonia emissions from broiler litter with a dynamic flow-through chamber system

Ammonia emissions from broiler houses

Broiler litter NH3 NH3 Ammonia in broiler house: affect bird performance Ammonia emissions: the primary concern for regulatory reporting under CERCLA

Estimation of emission rates

• Seasonal conditions
• Regional conditions
• House design
• Management practices
• Litter properties …

Wide variations have been found!

Scientific basis of ammonia emissions from broiler litter

Chemistry of ammonia in aqueous solution Partitioning between solid and aqueous phase ammonia Ammonia generation Partitioning between aqueous and gaseous phase ammonia,

NH4

+ NH3(aq) + H+

Litter solid NH4

+ (adsorbed)

Biomass Free air stream NH3(g) Kd

Convection mass transfer

C g, 0 C g, ∞ Moisture layer

Factors that may influence ammonia emissions from broiler litter

• Air and litter temperature
• Ventilation rate
• Air velocity
• Litter pH
• Litter nitrogen content
• Litter moisture content
• …

Emission models

1. To calculate site-specific emissions, using the local design and operating parameters; 2. To quantify and evaluate the effectiveness

• f the various control strategies;

3. To simulate seasonal and geographic variations in ammonia emission factors.

Research objective

• To develop a mathematical emission model so

that ammonia emissions can be predicted under given conditions;

• To evaluate the effects of various influencing

factors on ammonia emissions from broiler litter.

The dynamic flow-through chamber system

NH3 free air Impeller ECH2O moisture sensor Vent NH3 Ambient air Flux Chamber Motor Data logger TEI NH3 analyzer Pump Litter Flow controller Carbon filter Data logger

Operating conditions

• The ventilation rates of the chamber: 10.0 to 74.0L/min
• Residence time : 40 to 300 seconds
• Air velocity at the litter surface: 0.10 to 0.99 m/s
• Room temperature: 22 oC

Litter samples and ammonia measurements

• Litter samples at various ages were taken from three

commercial broiler farms in North Carolina.

• For each test, 3000 gram litter samples with a depth of

about 5 cm.

• The ammonia analyzer + HOBO data logger record

ammonia concentrations at one-minute intervals.

• The ammonia concentration in chamber at steady-state
• nce the variation in concentration < 0.5 ppm in ten

minutes.

Properties of the tested litter samples

Variable N Mean Min. Max. TKN (μg/g) 73 41109 30844 67298 TAN (μg/g) 73 3553 636 11172 pH 73 8.11 6.20 9.08 Moisture content (w/w %) 73 32.94 13.39 101.41 Total carbon content (%) 73 37.44 27.95 46.20 Total nitrogen content (%) 73 4.04 2.56 6.85

Model structure

General mass transfer flux equation J = Km (Cg, 0 - Cg, chamber) Mass balance equation J = (Q/A) Cg, chamber J = ((Q/A)-1 + Km

• 1)-1 Cg, 0

J = Ke * Cg, 0

Model structure

J = ((Q/A)-1 + Km

• 1)-1 Cg, 0
• Km: Determination of Km is largely empirical. It

was usually calculated as a function of air velocity and temperature.

• C g, 0: Dependent on the equilibrium between

gas phase ammonia and the NH3-N content in litter.

Estimate the mass transfer coefficient Km

• The general mass transfer flux equation

J = Km (Cg, 0 - Cg, chamber) Cg, chamber = Cg, 0 - (1/Km) J

• Linear regression of Cg, chamber vs. J
• 1/Km

• Three series of tests were conducted on three

litter samples.

• The Km was estimated to have an average

value of 8.59 m/h.

• Km: 0.36m/h to 8.28m/h - finishing pig house

(Ni, 1999)

• Km: 15.5m/h to 42.1m/h - soil following manure

spreading (Svensson and Ferm, 1993)

Estimate the mass transfer coefficient Km

Ammonia fluxes from litter vs. NH3-N content in litter

200 400 600 800 1000 1200 200 400 600 800 1000 NH3-N content in litter (μg/g) Ammonia emission fluxes (mgN m -2 h-1)

• Cg, 0 : estimated from the measured Cg, chamber, Q

and the estimated Km.

• The following nonlinear model was established.

Cg, 0 = 7.674 + 0.323* [NH3-N] - 0.0002 * [NH3-N]2 C g, 0 has unit of mgN/m3. [NH3-N] has unit of (μg/g) on a dry basis.

• The R-square of the model : 0.9272 (N=32).

The equilibrium concentration

• f gas phase ammonia at litter

surface Cg,o

The partitioning ratios of Cg,o over NH3-N content in litter

• In the range of 0.14 to 0.69 (mgN m-3) / (μg g-1).

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 200 400 600 800 1000 NH3-N content in litter (μg/g) Ratio of C g, 0 /[NH3-N] (mgN m

• 3 / μg g
• 1)

The NH3-N contents in litter

• Taking pH, moisture content, carbon content and TKN content

as independent variables, and the TAN content as dependent variable.

• [TAN] = 9133.8 - 1405.0*pH - 3.5683*107*[MC]0.5 / [TKN]

+ 2822.7*[MC] 0.5 - 104.05*[MC] - 1.1133*[C]2

• [NH3-N]/[TAN] = Kd / (10-pH+Kd)
• Log Kd = - 0.0918 - 2729.92/T (Kamin et al., 1979);

At 22oC, Kd = 10-9.3.

Summary of Model

J = Ke * Cg, 0

Ke = ((Q/A)-1 + Km

• 1)-1

Cg, 0~ Function of [NH3-N] [NH3-N] ~ Function of [TKN], pH, [MC], [C] and Kd

Sensitivity Analysis

• Relative sensitivity is defined as:

Sr = (∆J/J) / (∆x/x) In which,

– Sr is relative sensitivity, %; – ∆J is change of ammonia emission flux, mgN h-1m-2; – J is mean ammonia emission flux, mgN h-1m-2; – ∆x is change of the input variable over the range being considered; – x is mean value of the input variable.

Sensitivity Analysis

TKN content pH Moisture content Range (μg/g) Sr Range Sr Range (%) Sr 34000-36000 1.45 7.2-7.4 8.85 15-20 0.69 40000-42000 0.96 7.8-8.0 11.34 30-35 0.42 46000-48000 0.71 8.4-8.6 8.79 45-50 0.28 Total carbon content Km Q Range (%) Sr Range (m/h) Sr Range (L/min) Sr 30-32

• 0.36

4-6 0.99 10-20 0.08 36-38

• 0.58

8-10 0.97 30-40 0.03 42-44

• 0.91

12-14 0.96 50-60 0.02

Relative magnitude of Q/A and Km

Controlling factor Emissions Example condition When Q/A >> Km Ke ≈ Km C g, chamber ≈ 0 J ≈ Km*C g, 0 Open field When Q/A << Km Ke ≈ Q/A C g, chamber ≈ C g, 0 J ≈ Q/A*C g, 0 Closed chamber

J = ((Q/A)-1 + Km

• 1)-1 Cg, 0

Conclusions

• A statistical model was developed to estimate ammonia emission

flux from broiler litter based on experimental results from a dynamic flow-through chamber system.

• The model inputs: the litter TKN content, litter pH value, litter

moisture content, litter carbon content, the mass transfer coefficient Km and ventilation rate Q.

Conclusions

• Under the designed operating condition, the mass transfer

coefficient Km: an average value of 8.59 m/h.

• The model results: ammonia emission flux increased with litter

increasing TKN content, pH, litter moisture content, mass transfer coefficient and ventilation rate, and decreased with increasing litter carbon content.

• The model was most sensitive to litter pH value than to other input

variables.