Mass Balance Applied to Brazilian Conventional Broiler Houses during One Production Cycle

Brazil plays an important role in global poultry production as it is the world’s largest chicken exporter and the third largest broiler producer. Despite the development achieved by the Brazilian broiler industry as a result of the integrated production system, many obstacles still need to be overcome, in particular, housing environment. In this regard, detailed knowledge of the inputs, outputs, and primary waste generated by broiler production cycles is essential to establish a baseline for energy balance research studies and to develop environmental solutions. This article proposes a mass balance of conventional broiler houses of southern Brazil sheds in order to predict their outputs, but that may also be applicable to other regions with similar climate. Control volumes considered the heating, cooling, and rearing processes. All generated products and wastes were estimated considering litter production and the total production per cycle. Despite the variety of the microclimates, the results were very close to those reported in experiments, indicating that this model is adequate for this kind of estimation.


INTRODUCTION
Brazil is the third largest producer and the main exporter of broiler meat in the world, and its poultry industry is one of most ef cient sectors of the Brazilian agribusiness.The southern region of Brazil accounts for 63.46% of Brazilian broiler production, and for a signi cant share of chicken meat exports (ABPA, 2015).

Mass Balance Applied to Brazilian Conventional Broiler Houses during One Production Cycle
One third of the animal protein currently consumed in the world derives from chicken meat Broilers have one of the lowest production costs, and are the fastest converters of feed into animal protein (Miele & Girotto, 2004).In the 1980s, a broiler needed 70 days to reach around 2 kg body weight at a feed conversion ratio (FCR) of 3.5.In the 2000s, broilers reached 2.3 kg body weight at a feed conversion ratio of 1.8 in 42 days (Bueno & Rossi, 2006).Currently, it is possible to achieve almost 2.8 kg at a feed conversion ratio of 1.7 in the same period.However, despite of the improvement in genetics and technology, the main obstacle for higher productivity is the broiler house environment.
Many research studies have been conducted aiming at improving management and housing environment, as well as construction characteristics.In this context, mass balance studies of broiler houses are very challenging due to the wide differences in microclimates, live-weight targets, building typologies, and technologies applied.An article proposing poultry house building characteristics for the a speci c for the conditions in the Midwest Region of Brazil was published by Santos & Lucas Junior (2004).However, no other studies quantifying the input and output masses of the production process of a Brazilian broiler production cycle were found in international literature.
This article reports a mass balance applied to conventional poultry houses, according to the de nition of Brazilian broiler production standards proposed by Abreu & Abreu (2011).Predictability analyses were based on the performance of straightrun Cobb-500 ocks, according to the manual published by the genetic company for maximum genetic potential (Cobb-Vantress, 2012).The main inputs were estimated using data from literature.The solid, liquid, and gas outputs of the production cycle were calculated by the model, achieving good predictability for each element analyzed when compared with the literature.

Input characteristics
The choice of the adequate absorbent poultry litter substrate is essential for broiler performance.Evaluating different litter substrates, Araujo et al. (2007) and Avila et al. (2008) did not nd any signi cant in uence on broiler performance variables, and suggested that all studied substrates could be used as broiler litter.However, Garcia et al. (2012) analyzed six different litter substrates, and indicated that wood shavings is the best choice due to its availability, low cost, absorption capacity, and reutilization possibility.Wood-shavings density ranges between 52 kg/m 3 and 160 kg/m 3 , depending on granulometry.Its moisture content is around 13% (wet basis) and it is in uenced by tree species, season, weather, and storage time (Moulin et al., 2011;Teixeira et al., 2015).Broiler litter can be used up to six 42-day grow-out cycles without any negative effects on performance.However, in shorter cycles, the same litter can be used up to eight or ten cycles.
Feed composition varies according to production system, rearing phases, and regional speci cities of the production system.Brazilian broilers feeds are typically based on corn and soybean meal.The studies of Lana et al. (2001) and Gomes et al. (2008) indicated feed intakes of 4.5 kg and 5 kg during a 42-day production cycle, respectively, as well as signi cant differences among genetic strains and between sexes.Currently, in order to achieve the same market weight, feed intake has been reduced due to genetic improvement, consequently improving feed conversion ratio.
At arrival on the farm, day-old chicks weighed about 0.046 kg, but different weight ranges are reported by other authors: 0.042-0.046kg (Gomes et al., 2008), 0.039-0.045kg (Almeida et al., 2006), and 0.042-0.056kg (Aviagen, 2012).Chicks are able increase 4-to 5-fold their body during the rst week of rearing.Productivity is in uenced by gender, genetics, and production targets.
Water is a vital input for broiler maintenance and growth to maintain homeostasis.Water intake depends on the external temperature, and its relation to feed intake is 1.5 to 2.5 L w /kg f when temperatures are within the thermal comfort zone.Water should be supplied at temperatures lower than 24 °C.Water intake is also in uenced by sodium and potassium content in the feed (Palhares & Kunz, 2011).When environmental temperature exceeds comfort limits, water consumption may increase up to 5 L w /kg f .Birds lose water through breathing and excretion.
In evaporative cooling systems of poultry houses, water is added to unsaturated air and absorbs heat by evaporation.Sprinkler systems are supplied by water lines consuming about 900 L/h and, each evaporative panel requires the same water volume (Cobb-Vantress, 2012).However, this value may change as a function of microclimates and management procedures.
Ventilation is the main method of cooling poultry houses.Ventilation systems operate differently during Mass Balance Applied to Brazilian Conventional Broiler Houses during One Production Cycle each phase of the cycle.Ventilation is responsible for providing clean air and control litter humidity, as well as for removing greenhouse gases and reducing the thermal load of the sheds.Within the thermal comfort temperature range applicable, atmospheric air is considered to be a perfect gas, composed of dry air in volume fractions of 79% N 2 and 21% O 2 (mass fractions of 76.7% N 2 and 23.3% O 2 ) and water vapor, which proportionately reduces the amount of other elements (Cengel & Boles, 2005).
Lique ed Petroleum Gas (LPG) and rewood are the main fuels used for heating, particularly for brooding.Conventional poultry farms can only use LPG, with an estimated consumption of 400-600 kg or only rewood, with consumption estimated from 2500 kg to 12000 kg per cycle.In Brazil, over 70% of the poultry houses use rewood for heating, due to its availability and low cost, but studies show that it generates a greater amount of greenhouse gases and more energy is destroyed compared with LPG (Migliavacca et al. 2016).

Characteristics of products and waste
Throughout the production cycle, the high density of birds reared in a single environment causes losses.Mortality rates of 3.5% or lower are considerable acceptable (Orrico Junior et al. 2010).This results in a reduction in the desired output and an increase in waste mass.The remaining birds are delivered to the integrated company after production cycles of 30 to 49 days, with body weights between 1.0 kg and 3.0 kg, according to production goals.
The average live weight of 42-d-old broilers is estimated as 2.6 kg (Miele et al., 2010), with an estimated growth rate of 2.5 g/h (Funck & Fonseca, 2008) and feed conversion ratio (FCR) of about 1.6 to 1.85.
The unmetabolized feed is excreted in the litter substrate together with the excreta.Poultry litter consists of a mixture of litter substrate, feed residues, bird droppings, and feathers, and it is considered the main solid waste of the broiler production cycle.According to Macari & Campos (1997), the recommended litter depth for all types of sheds ranges between 5 cm and 10 cm, and should contain 20-35% moisture at the end of cycle.Lower humidity levels can generate suspended dust, whereas higher values result in unhealthy environmental conditions due to high ammonia levels (Vieira, 2011).
The amount of litter produced depends on the litter substrate, season, number of ocks, duration of cycle, and bird density.For broilers reared up to 2.3 kg live weight for 42 days, Santos (2001) estimated the generation of 1.37 kg litter per bird (on dry matter basis), including the initial substrate.This value is close to the value of 1.40 kg proposed by Lynch et al. (2013).
Litter reutilization helps to reduce the environmental impact of litter generation and disposal, and it is an alternative in areas where there is low availability of substrates (Brand, 2007).
By the end of the production cycle, litter can be also be composted and used as fertilizer, and be burned for heating the broiler house, as shown by the energetic analysis of Migliavacca et al. (2015).
An important environmental concern in intensive animal production is the emission of greenhouse gases, particularly of ammonia (NH 3 ) and carbon dioxide (CO 2 ).In a comparative study between two broiler houses, with and without minimum ventilation, Vigoderis et al. (2010) found ammonia concentrations of 29.3 ppm and 23.2 ppm, and carbon dioxide concentrations of 1527.7 and 1427.3 ppm, respectively.In a study in conventional and tunnel-ventilated broiler houses, Nääs et al. (2007) observed that air dust and NH 3 levels were above the recommended limit for human breathing, and only CO 2 emissions were within acceptable limits.Therefore, air quality of animal production environments is a critical aspect evaluated in environmental control studies.Henn et al. (2015) developed mathematical models to estimate CO 2 generation in broiler production based on the carbon balance in broilers and in poultry litter.Those authors proposed equations to calculate gas emissions from poultry litter and from breathing (g/ bird) during the rearing cycle.Miragliota et al. (2004) estimated NH 3 emissions (mg/m 2 h) in broiler houses as a function of average litter temperature and pH and bird age, and propose an equation to calculate these emissions for the environmental and operational conditions of Brazilian broiler houses.Finally, the mass of dust suspended in the air inconventional sheds was estimated by Nääs et al. (2007), who observed a maximum value of 2.5 mg/m 3 , and an increasing trend during the production cycle.However, previous studies calculated higher values, with an average of 10.58 mg/ m 3 , which are very close to the international limits of tolerance (Simpson et al., 1999).
Combustion gases generated by brooding during the starter phase of the broiler rearing cycle increase house temperature above the external environmental temperature, carrying vaporized rewood moisture

Mass Balance Applied to Brazilian Conventional Broiler Houses during One Production Cycle
and vapor produced by combustion.In addition to CO 2 , H 2 O and N 2 emissions generated by complete wood combustion, certain amounts of O 2 are present when combustion occurs with excess air.The inert material (ashes) should be extracted from the shed as solid waste of combustion.

MATERIALS AND METHODS
Based on the main inputs, products and wastes, the mass balance of the production cycle of a complete broiler rearing cycle was calculated.The owchart shown in Figure 1 illustrates the quantities and destinations of each process input.The ows are numbered to allow relating them to the equations proposed in this paper.
The ows indicated in blue represent the fraction of water contained in each component, whereas the other ows indicate the dry mass component of each input, product, and waste.The inputs, whose amounts were estimated from the literature, are listed on the left.On the right, birds and process wastes are presented.The obtained estimates were compared with the values from previous studies.All processes which altered the mass ows were estimated between the inputs and the outputs.
The process of broiler growth demands feed, water, and oxygen supply.The temperature of the shed is maintained within the thermal comfort zone (TCZ) for the birds by heating and cooling systems.In the production cycle, heating during the starter phase (days 1-14) and cooling during the nishing phase (days 29-42) are considered.Therefore, the house environment is maintained as close as possible to the ideal temperature for each production phase, considering the climate of southern Brazil.All the calculations were performed using Engineering Equation Solver (EES ® ).
Based on the speci c mass adopted for wood shavings and the average volume acquired for 1200 m 2 sheds, the absorbent mass introduced in the control volume (CV), given by the sum of Equation 1 (dry weight) and Equation 2(humidity), was determined.This mass is used in n cycle production cycles.This study considers the introduction of 10% of the initial volume for each new produced cycle.
The feed required during the cycle is a function of individual feed in take and the number of live birds one ach day of the cycle.Figure 2 shows the data extracted from the performance objective tables (Cobb, 2012), and the trend line indicates the best function for the individual feed intake over x days.This equation is speci c to each strain and sex and in accordance with performance standards.By integrating the intake function throughout the cycle, the mass of feed consumed per bird is obtained, which, multiplied by the number of birds, gives the total mass of feed required for the cycle.On the other hand, the number of live birds on each day depends on the number of chicks received and on the estimated mortality rate.Equation 3 shows feed dry matter content and Equation 4 illustrates feed moisture content, where x indicates the day of the rearing cycle.

Mass Balance Applied to Brazilian Conventional Broiler Houses during One Production Cycle
1--0.0044 0.3025 -0.1212 14.434 The rst term of the equation allows determining feed dry matter and moisture fractions.The rst part of the integral indicates the number of live birds on each day of the cycle and the last part, the function of individual feed intake.
Based on the range of values proposed in literature, bird mass was considered as the average of the values obtained indifferent previous studies.In the present study, total bird mass considered in the CV is obtained by adding Equations 5 (dry mass) and 6 (moisture).
The water used in the production system was subdivided into drinking water (used for consumption and maintenance of the homeostasis of birds), water used in sprinkling system and evaporative panel (used for the evaporative cooling of the shed).Drinking water volume was calculated relative to feed intake.The ratio of k = 2L w /1kg f is widely accepted for temperature within the TCZ, but at higher temperatures, this ratio can achieve 5L w /1kg f .Therefore, Equation 7estimates the mass of water consumed, considering its speci c mass at 20 °C, which is within the thermal comfort zone during the nishing phase, when water consumption is high.
The evaporative cooling system is used when temperatures are high (generally above 28°C), and preferably at relative humidity levels below 70%.In this study, we assumed its use in the nishing phase, during the hottest hours of the day.Equations 8 and 9 illustrate the mass required by the cooling systems, which depends on the number of sprinklers and water ow in each sprinkler, as well as the number of hours the equipment is on.The use of sprinkler systems was considered only during the nishing phase, justifying the range proposed in the sum.
Atmospheric air was quanti ed using the number of air exchanges recommended in the genetic strain manual.Ventilation types for the different phases and times were estimated according to Table 1.Minimum ventilation should be set using a timer to provide an air exchange every 8 min and to operate about 20% of the entire time.Therefore, seven complete air exchanges Considering daily air renewal rates in Table 1 and integrating these rates over time, it is possible to estimate the air volume required during each rearing stage and for the entire rearing cycle, as shown in Equation 10, where dcn indicates the number of daily air exchanges.The average air velocity inside the house for each type of ventilation can be calculated using the cross-sectional area and the air ow rate.
Considering the relative humidity and the inlet air temperature in the control volume (CV), the dry air was separated from the steam.The sequence of equations (11-13) demonstrates the calculation method, where Ø represents relative humidity, P V vapor pressure, P sat(T) air saturation pressure at the desired temperature, P 0 atmospheric pressure, P a dry air pressure, and w absolute humidity.The vapor mass contained in the atmospheric air is estimated by Equation 14, using the volume of exchanged air, V air , and the ideal gas constant for steam, R V .The O 2 and N 2 fractions are shown in Equations 15 and 16, respectively, with R a indicating the ideal gas constant for dry air.The composition of the Lique ed Petroleum Gas (LPG) considered in this study was 70% butane and 30% propane, and burnt with 5% of excess air.The use of LPG was considered only for dry cleaning with blowtorches and a 13-kg cylinder.Firewood consists of 49% carbon, 6% hydrogen, 43% oxygen, 1% nitrogen, and 1% of ash, and burnt with 20% excess air.The mass balance of the combustion process is detailed in a previous study (Migliavacca et al. 2016).The estimated use of rewood was of 1 m 3 /day during the starter phase, considering a speci c mass of 450 kg/m 3 (Funck & Fonseca, 2008).
Taking into account these initial considerations, mass balances were calculated for the global control volume (CV), birds during the rearing cycle, cooling and heating processes, and poultry litter.The law of mass conservation was applied to all these processes and, based on the owchart of Figure 1, the proposed equations are presented in Table 2.
Finally, a weight ratio of 3.29 between N 2 and O 2 was considered, because it corresponds to the volumetric ratio of 79% of N 2 and 21% of O 2 in the atmospheric air.The generation of CO 2 was estimated according to Henn et al.(2015), dust to Fernandes (2004), and NH 3 to Miragliota et al. (2004).All these equations are at the bottom of Table 2, with appropriate adjustments.
In order to adjust the amount of residual litter to the values measured in the experiments, it was assumed that part of litter mass was extracted in the particulate form from the sheds into atmospheric air.
The equations allowed estimating the mass of wastes and products generated by the production process.Equation 17gives the mass of the dead birds for the entire cycle, and the total mass of chickens is given in Equation 18.  Excreta can be estimated considering the ows m 52-

56
, with ows for dry mass and moisture.By applying a mass balance to the CV obtained from the litter, it is possible to evaluate the quantity of litter generated during the cycle, given by Equation 19.
Considering greenhouse gases and dust/particles removed by the ventilation system, as shown in Figure 1, total mass is given by Equation 20.Likewise, the amount of air moisture extracted is calculated by adding the ows presented in Equation 21

RESULTS AND DISCUSSION
The method developed to predict the output parameters of the broiler production process allowed nding the masses of the products and of the waste.The procedure was validated by comparing the results with literature data.The method took into consideration a standard scenario of a conventional broiler house with 1200-m 2 production area, 4-m height, and capacity to house a ock of 13,000 straight-run Cobb-500 broilers.Performance targets were:1.7 feed conversion ratio, 2.7 kg market weight, and 4.6 kg feed intake per broiler in 42 days (Cobb-Vantress, 2012).
An initial mass of 0.046 kg per chick and 3.5% average mortality rate were assumed.The conventional broiler house was built of masonry, with closed ends and open sides, with low walls closed with wire mesh, and equipped with curtains.Wood stoves were used for brooding, and LPG was used only for dry cleaning.The house was equipped with two fogging lines, with total of 120 sprinklers, for cooling and fans for air renewal.
Mass balances were calculated for the heating, cooling and growth processes, for the generated litter, as well as the global mass balance.All the balances produced results with excellent accuracy compared to known empirical quantities.
Table 3 presents the input masses and the corresponding output masses of broilers, solid waste and gases generated in the process.The dry mass fractions were estimated based on average values from different previous studies.For wood shavings, 55 kg/ m 3 density, 120m 3 acquired volume, and reutilization for up to 6 production cycles were considered.
Each spray nozzle has an average ow of 6.5 L/h.The relative humidity of the atmospheric air was estias 65% at P 0 = 101.3kPa and T 0 = 20 °C environmental conditions.Fuel consumption for brooding of 14 m 3 per cycle was assumed.
According to the estimates, average litter mass per bird was close to 0.51 kg in the rst cycle, but due to litter reutilization and addition of new litter substrate in the beginning of each cycle, it reached 0.76 kg in the 6 th cycle.The amount of feed offer per bird obtained in the model (4.65 kg) was close to that established in the genetic strain manual, according to the recommended feed conversion ratio.
The bird density adopted in this study, of 10.8 birds/ m 2 , is the typical density used in Brazilian conventional broiler houses.The calculated water consumption per bird is in agreement with widely accepted values reported in literature.The sprinkler system was used 4 hours/day during the nishing phase.
The procedure applied to calculate air renewal in conventional open-sided houses may be used for tunnel-ventilation houses, as the calculation does not depend on the number or on power of the exhaustion fans.When the calculation is applied to the latter, air speeds of 0.20 m/s for minimum ventilation, 0.83 m/s for transition ventilation, and 2 m/s for maximum ventilation.
Although the daily mortality is considered constant, individual bird mass increases daily during the rearing cycle, and an average dead bird mass (1.38 kg) of about 50% of the mass of live broilers at market weight is assumed.Relative to live broilers, the nal mass (2.723 kg) calculated by the proposed method was consistent with the value indicated in the performance tables of the strain manual.
An excreta moisture value of 72% was determined by m 52,54,56 in the assumed scenario.
The volume of generated litter determined by the model (1.92 kg/ nished broiler on fresh matter basis and 1.46 kg/ nished broiler on DM basis) was close to those found in literature.Garcês et al. (2013) determined a volume of 1.06 kg litter/kg live weight generated by Cobb broilers during a 35-day production cycle, which is close to the value found in our model,

Mass Balance Applied to Brazilian Conventional Broiler Houses during One Production Cycle
when the same production cycle duration is considered.Santos & Lucas Jr. (2004) identi ed an average of 2.19 kg of litter generated per nished broiler, as well as differences between the inputs (chicks, feed, and litter substrate) and the outputs (broilers, dead birds, and litter), including on dry matter basis.The obtained values were close to those of Fukayama (2008), who did not consider the application of law of mass conservation either.These evidences justify the hypothesis that part of the debris in the litter (feces, feed, soil and feathers) is removed from the system by the air.
Genetics, gender, market body weight, performance targets, and bird density need to be considered when estimating the generation of waste of a broiler production cycle.Santos et al. (2005) evaluated male Hubbard broilers reared at different densities and determined that the amount of waste decreases as the number of birds increases.At a same bird density, those authors determined the generation of 1.73 kg of litter (on DM basis) in the rst cycle and 1.2 kg (DM) in the second cycle; these values are lower than those estimated by the present model.
Our study was based on the performance targets set by the genetic company at time of the study, and the differences observed when comparing our results with previous literature studies may be due to the fact that those values were obtained in less developed genetic strains.
The estimation of the generation of combustion and greenhouse gases, and ashes is important as these cause environmental degradation.In addition, those estimates may aid the development of novel heating models and/or treatment of the gases generated in the process.All estimates obtained presented good  Mass Balance Applied to Brazilian Conventional Broiler Houses during One Production Cycle predictability, with percentage errors considered irrelevant due to the range of orders of magnitude adopted for the elements.Furthermore, due to the great variability in microclimates, bird management, and house structural differences, no mass balance studies were found in literature, and therefore, the proposed model is the rst to predict the output of broiler houses under different environmental and operational conditions.

CONCLUSIONS
This article proposes a method to estimate the mass balance of broiler houses during a production cycle.The proposed mass balance is based on the expected performance of broilers, according to their genetic potential, considering feed conversion ratio, weight gain and production performance.
Inputs, products, and waste of the production process of broilers housed in conventional sheds for a single production cycle were estimated.The method was shown to be effective to predict, based on input data, the amount of products and waste for different shed types, genetic strains, and sexes.The default setting used herein represents a typical conventional house in southern Brazil, and the results obtained in this particular case were very good as compared with known output data.Therefore, the present model can be applied to different microclimates, genetic strains and production targets, with good predictability.

Figure 1 :
Figure 1: Overall mass balance for a production cycle.
inputs are shown on the left, and generated products and waste on the right.The rst part of the equations, corresponding to m 44 , m 46 , m 48 and m 51 ows, indicates the dry matter body composition live and dead birds.The ows m 42 , m 43 , m 45 , m 47 , m 49 and m 50 indicate their moisture content.Based on the concept of FCR, the central part of the equations indicates the fraction of feed and water of the body of live and dead birds.Although mortality rate may be considered constant over the rearing cycle, the body weight of dead birds varies.An averaging procedure indicates that a division factor of 2 is enough to correct the average body weight of dead birds.The ows m 52 , m 53 , m 54 , m 55 and m 56 indicate the feed and water masses of the body of live and dead broilers lost by excretion or to the air by breathing.The respiratory quotient (0.85 for the broilers) estimates the ratio between inspired oxygen and expired carbon dioxide masses (m 35 /m 21 ) during breathing

Table 1 -
Estimate adopted for ventilation throughout the breeding cycle

Balance Applied to Brazilian Conventional Broiler Houses during One Production Cycle per
(Cobb-Vantress, 2009;Aviagen, 2012)lation aims at increasing air velocity inside the shed, and it is controlled by a thermostat.It provides one complete air exchange every 2 min, totaling 30 complete air exchanges per hour.Maximum ventilation should provide a complete air exchange in less than one minute to achieve75 air exchanges per hour(Cobb-Vantress, 2009;Aviagen, 2012).

Table 2 -
Equations for the mass ows in the process.

Table 3 -
Input masses and the corresponding output masses of broilers, solid wastes and gases generated in the process.