Methods and apparatus for delivering foam

i?

<img file="GB2492480B_D0001.tif" />

<img file="GB2492480B_D0002.tif" />

<img file="GB2492480B_D0003.tif" />

Methods and apparatus for delivering foam

Field of the invention

The present invention relates to methods and apparatus for generating and delivering high expansion water based foam filled with a gas. It is particularly applicable, but in no way limited, to a system and equipment, including mobile systems and equipment, for mass depopulation of livestock.

Background to the invention

Over the past 10 years there have been significant outbreaks of notifiable animal diseases that require the culling of large numbers of livestock to control and prevent the spread of the disease. These have not only had huge financial consequences to governments and the livestock industry but where they involved zoonotic diseases, such as highly pathogenic avian influenza (HPAI), there is significant risk to the health of operational personnel from exposure to the virus and the health of the wider community if the virus is not contained quickly.

There is therefore a requirement for in-situ emergency killing of poultry flocks in the event of a notifiable disease outbreak. There are many systems in use around the world for this application, the main two systems used in the UK being:

1. Whole House Gassing (WHG) where poultry sheds are sealed and liquid carbon dioxide is pumped into the sheds to reach a concentration of 45%. The advantages of this system are that it requires few personnel to be in direct contact with the birds until the carcases need to be removed, but there are limitations to the type and number of sheds that are suitable for this technique. There are also reservations on the animal welfare side, owing to the aversive nature of carbon dioxide to poultry.

2. Containerised Gassing Units (CGU) where poultry are caught and loaded into transport modules then placed in metal containers that are filled with a gas mixture of 80% Argon and 20% Carbon dioxide. This system is suitable for use on any farm with all species of poultry. However it does have the disadvantage of exposing many more people to the virus.

Whole House Gassing would be the preferred option in relation to reducing the risk to human health. However, as already mentioned, not all poultry buildings are suitable for WHG due to the fact they cannot be adequately sealed. Reports on the

06 13 outbreaks of HPAI in Holland in 2003 suggested only one third of the housing stock were suitable for this method and a report by the National Farmers Union in the UK found that 60% per cent of broiler houses over 20 years old and 92% of all housing was over 10 years old. In particular, mobile free range poultry houses would be almost impossible to seal due to the presence of pop holes and to the fact that many are mobiie and as such have no solid floor.

An alternative method of euthanasing poultry in their sheds is by direct application of low to medium density fire fighting foam using conventional fire fighting equipment or by the use of a high expansion fire fighting foam containing a gas such as carbon dioxide or an inert gas (Gerrizten 2008). The genera! concept of using fire fighting foam has been described in two patents (EP 1921921 B1 Method for the sanitary slaughter of an animal, andUS7435166 B2 Methods and devices for depopulating avian species), however no equipment has been developed to successfully create a high expansion foam containing a gas, that can be applied to animals in a container or shed. Conventionally, fire fighting foam is described by the ratio of liquid to air volumes and may be divided into three groups (see Table 1).

Table 1 Descriptive names and typical uses of foam in relation to expansion ratio

<td> Expansion ratio (Airdiquid)</td><td> Descriptive names used</td><td> Typical Uses</td>

<td> 20:1</td><td> High density Low expansion Heavy foam</td><td> Flammable and combustible hydrocarbon fires, creates foam blankets and has a long throw range. Also coverage of chemical spills.</td>

<td> 20:1 to 200:1</td><td> Medium expansion Medium density Medium foam</td><td> For 3 dimensional fires (e.g. tyres) and filling oil bunds.</td>

<td> 200:1 -1000:1</td><td> Low density High expansion Light Foam</td><td> Fire fighting in large enclosed spaces, (ships, hangers, warehouses and mines) especially where flammable vapours exist.</td>

The foam euthanasia methods developed in the United States have been based on medium expansion foam. They have been tested and conditionally approved in the United States for use in specific circumstances (Dawson et al. 2006, Benson et at. 2007). The United States Department of Agriculture (USDA) performance standards for foam euthanasia require an expansion ratio of between 40 and 135 (USDA APHIS, 2007). This method is not approved for use in Europe because of animal welfare concerns surrounding this approach, it operates as a euthanasia agent by small bubbles of foam rapidly occluding the airways of the birds causing death by mechanical hypoxia (Benson et al., 2007).

It is an object of the present invention to overcome or at least mitigate some or all of these disadvantages.

Summary of the invention

According to a first aspect of the present invention there is provided a foam generator unit, for depopulation of livestock, according to Claim 1 and its dependent claims.

co ¹ By arranging for the gas required to generate a foam to enter the unit enclosure by way of its

1” own specially designed drum, where the gas is directed towards the sides of the second end ¹ 20 of the chamber body, this enables foam to be created with an expansion ratio of 250:1 or CM above.

It is important that the gas is introduced into the chamber body downstream of the foam solution. By ‘downstream' is meant that the gas enters the chamber body at or near the second end of the chamber body and turbulent flow is established in the gas stream as the gas makes its way out of the enclosure towards the first end of the chamber body. A spray of foam solution is then introduced into the turbulent gas, upstream of the gas manifold.

By ensuring that little or no gas can escape from the vessel directly towards the end of the enclosure through which foam exits encourages the desired turbulent gas flow, and avoids a high speed jet of gas passing through the chamber.

Preferably the side(s) of the drum incorporates one or more perforations and in a particularly preferred embodiment the side(s) of the drum comprises a mesh or gauze. A mesh or gauze is a particularly effective way of breaking up the gas stream and creating turbulent gas flow.

11 13

According to a second aspect of the present invention there is provided a foam as claimed in Claim 15. When nitrogen gas is used to generate this foam it is particularly effective at depopulating livestock.

According to a third aspect of the present invention there is provided a method for the mass depopulation of livestock as claimed in Claim 16.

Brief description of the drawings

The present invention will now be described by way of example only with reference to the following drawings in which:Figure 1 illustrates two diagrammatic side elevational views and a sectional view of a foam generator unit according to a first embodiment of the present invention; and

Figure 2 illustrates a schematic view of a foam generating system;

Figure 3 illustrates a side elevational view of a foam generator unit according to a second embodiment of the present invention;

Figures 4 and 5 illustrate side and cross-sectional views respectively of the embodiment shown in Figure 3.

Description of the preferred embodiments

Anoxic gases have not thus far been used for whole house gassing because of the practical impossibility of sealing the house to the extent required to reduce the residual oxygen content within the confined space to less than 2%. The use of high expansion gas-filled foam containing an anoxic gas presents one feasible alternative delivery method of anoxic killing. As the foam containing the gas envelops the bird, the movement of the bird will break the bubbles, releasing the gas and therefore making it available for respiration by the bird. The foam situated above the bird effectively seals the space where the bird is located from access to atmospheric air. The effect is to rapidly reduce the residual oxygen content in the atmosphere around the bird to less than 2%, at which point the bird will rapidly become unconscious. When maintained in this oxygen deficient atmosphere for a period of time the bird will die from anoxia.

A system for generating a gas filled water based foam with an expansion ratio in excess of 250:1 is described. One use of this system could be to humanely kill animals in their housing, pen or other chamber where their killing is required to control the outbreak of a disease or other infection. The gas used in the system is preferably an inert gas such as nitrogen with the intention of reducing the oxygen level in the vicinity of the animal less than 2%. Alternatively the gas could be one which has a direct anaesthetic effect on the animal such as carbon dioxide and in this case a final concentration of carbon dioxide around the bird should be at 45%. The system comprises a pressurised gas supply, a pressurised water supply, a device for mixing a propriety foam concentrate solution into the water flow, a manifold system and hoses to then deliver the gas and foam solution to a series of individual foam generator units. The system is preferably required to be mobile in order that it may be rapidly set up in different places and situations.

Foam generator

Many high expansion foam generators work by spraying foam solution onto a mesh though which air is driven by a fan. These can either be fixed installations in the roof or walls of a building or more compact portable systems. The fans are often driven by hydraulic flow, and therefore do not require an external power supply beyond that required to create the water pressure. These systems have the advantage that they can be used in closed environments where the foam needs to be deployed against a back pressure, for instance in mines or inside oil storage tanks. An alternative system uses a high velocity jet nozzle to create a pressure differential across a mesh screen effectively sucking air through the unit and creating large volumes of high expansion foam, up to 1000:1 expansion ratio, without the need for a fan. These systems have been developed to fill the air space in warehouses containing dangerous chemicals of several thousand cubic metres within a few minutes. There are also foaming systems that use compressed air or gas to develop foam, but these are limited to low expansion (high density) foam.

The current embodiment of the gas-foam generator has been developed to generate high expansion water based foam filled with substantially 100% nitrogen gas with an expansion ratio of 350:1, the expansion ratio being the ratio of the volume of foam produced to the volume of liquid used to produce it. Experimental work has demonstrated that the expansion ratio of the foam must be of this order for the gas to be freely available to the animal to respire normally. Foam with an expansion ratio below 250:1 consists of much smaller bubble sizes which risk occlusion of the airway of the animal making the system unacceptable on animal welfare grounds. It is advantageous for the foam generator to produce a foam having an expansion ratio of 350:1, higher than the functional minimum requirement of 250:1, in order to increase the rate at which the space where the animals are kept is filled. If the expansion ratio is too high, then the flow properties of the foam matrix change, the foam becomes drier and less fluid and there is a risk that it may not flow around obstacles or animals within the space and leave air pockets within the foam structure reducing its effectiveness.

A key element of the invention is the foam generator units illustrated in Figure 1 and in Figures 3, 4 and 5. The foam generator unit illustrated in Figure 1 consists of an enclosure 1, which may be constructed from a metal such as stainless steel or a plastics material such as a high density plastic. The enclosure 1 preferably takes the form of a substantially tubular chamber body having a first end which is substantially open and through which foam can escape, and a second end at which a gas and a foam solution are introduced in separate supply pipes, the second end being otherwise substantially sealed in a fluid tight manner to prevent ingress of ambient air. The second end includes input connections for a gas supply and for the supply of a foam solution. It is however not essential that these connections are made through the second end of the chamber body, although this is a convenient arrangement. These connections could be made through the side wall(s) of the chamber body near to the second end of the chamber.

The tubular chamber body may be any suitable cylindrical shape, such as substantially square as shown in Figure 1C, or substantially circular as shown in Figure 5. The cylinder does not have to be of uniform cross-section along its entire length. In fact it is shown to taper towards the first end in Figures 1A and 1B.

As described above, a first end of the enclosure is open and covered with a mesh 2, and the other, second end of the enclosure is substantially closed in a fluid tight manner. In this example, a first pipe 4 carrying a foam forming solution enters the body of the foam generator through the closed end and terminates in a spray nozzle 5. The appropriate spray nozzle is selected based on its delivery characteristics such as spray pattern, spray angle and flow rate. The nozzle is required to deliver a full cone spray pattern at a flow rate to match the expansion ratio and final foam flow rate of the unit. For example to generate foam at 12 cubic metres per minute with an expansion ratio of 350:1 the discharge rate from the nozzle will need to be 34 litres per minute. A second mesh 3 is positioned part way down the enclosure in such a position that the discharge from the spray nozzle is spread uniformly across its surface. Changes to the delivery pressure of the foam solution to the nozzle will alter the shape of the spray pattern and therefore affect the distribution of foam solution across the mesh. In this example, both the first and second meshes are Vshaped in cross-sectional profile, such that neither mesh is positioned substantially parallel to the general longitudinal axis of the enclosure. Thus the plane or planes of one or both of the meshes may be out of alignment with the plane generally normal to the longitudinal axis of the enclosure.

It will be understood that the performance of the foam generator will be refined by experimental use of various spray nozzles, pressures, gas flow rates, and foam solution flow rates.

A second pipe 6 carrying a gas enters the closed end of unit and discharges into a specially designed gas inlet manifold situated within the enclosure or chamber body 1 behind or downstream of the spray nozzle 5. The relative positioning of the manifold and the spray nozzle with respect to each other is important. There is a general longitudinal axis to the foam generator extending from the substantially closed end of the foam generator body, through the gas inlet manifold, past the foam solution spray nozzle and out of the first end of the body through mesh 2, in the direction in which the foam is ejected. The end through which foam is ejected is termed the first or front end of the unit, and the closed end through which a gas enters under pressure is the second or rear end.

The spray nozzle needs to be located well ahead or upstream of the gas inlet manifold, such that gas mixes with spray droplets as they hit the first mesh 3. In this example the manifold comprises a drum 7 with perforated sides and with the end of the drum facing the first, open end of the enclosure being substantially sealed in a fluid tight manner. This drum is shown in more detail in Figure 1C (Section AA). The perforated drum has a solid, sealed end in the direction of flow of the gas from the inlet pipe and the gas is therefore forced to discharge sideways through the perforations and towards the sides of the enclosure/chamber body, rather than directly into the general path of the longitudinal axis of the chamber body. This conditions the gas by creating a turbulent gas flow within the enclosure/chamber body, thus reducing the velocity with which the gas travels through the meshes, and prevents a high velocity jet of gas escaping the unit which would otherwise distort the spray pattern from the spray nozzle and reduce the formation of large foam bubbles.

It will be understood that a wide variety of designs of so-called inlet manifolds can be used, associated with the enclosure, to direct and condition the stream of incoming gas in such a way as to provide turbulent gas flow. For example, directing the incoming gas flow back towards the second end of the enclosure, with or without an additional diffuser on the end of the pipe such as a mesh or other perforated region, can be used to create turbulent flow. Alternatively, a series of inwardly directed nozzles could be used, directing the flow of gas into the chamber body, with pipe work to these nozzles connected to a gas supply.

However, in experiments carried out by the inventor into suitable manifold designs, the version described above comprising a substantially circular cylindrical body, closed at one end by the second end of the chamber body and closed at the other end by a substantially circular plate, gave the best results. The cylinder wall of the manifold is optimally formed from a mesh or gauze, although other materials containing a plurality of perforations or apertures also work well. Mesh (or gauze) however is readily available in a wide variety of mesh sizes, and is also relatively inexpensive.

The foam generator unit requires a supply of pre-mixed water and surfactant (foam solution) to be delivered to at a specific concentration, pressure and flow rate. This concentration, pressure and flow rate is determined by the type of surfactant, the design properties of the spray nozzle and the required flow rate and expansion ratio of the finished gas filled foam. It will be understood that in this application, and in this design of foam generator unit, we are dealing with a water-based foam.

The foam generator also requires a supply of gas, in this case nitrogen, of sufficient pressure to maintain the flow rate requirement of the finished foam.

It should be noted that there are inevitably some losses in this type of system. We have found that not all the gas ends up inside the bubbles, i.e. some of the gas escapes. To date the best efficiency rate results in 92% of the gas supplied to the unit ending up within the foam bubbles. An advantageous way of improving the efficiency of foam generator is to have a steady gas pressure.

The second embodiment of a foam generator is illustrated in Figures 3, 4 and 5. A corresponding numbering system to that used in Figure 1 is used here. This generator is generally circular cylindrical in shape with a handle 30 and a foot 28 to aid with portability and positioning. The front or first end 22 comprises a conicalshaped mesh and a further mesh 23 extends across substantially the whole crosssectional area of the foam generator body. The rear or second end is substantially fluid tight with the exception of two couplings 24, 26 for gas and foam solution respectively.

In this example the foam solution input tube and nozzle 24, 25 are not shown. However, the foam solution inlet pipe or lance is held firmly in a lance mounting 29, supported from three points around the body of the foam generator, for improved rigidity.

It will be appreciated that by including a lance mounting arrangement, the foam solution input pipe could enter from the side wall of the unit, rather than from the end. The same is also true of the gas inlet pipe, which could enter the manifold vessel from the side, rather than from the rear of the unit.

One example of the foam generator unit has a design discharge rate of 11.5 cubic metres of foam per minute, at expansion ratio of 350:1. The flow rate of the gas into a single foam generator enclosure is required to be at least 11.5 cubic metres per minute at a pressure of 10 bar. In operation, the number of individual foam generators connected to the system will depend on the scale and nature of the operation being undertaken. For example if ten of these units were connected to the system then it would be capable of delivering a gross flow rate of 115 cubic metres per minute of finished foam. The gross gas flow rate required for the system would be at least 115 cubic metres per minute and the upstream pressure of the gas supply would have to take account of all the pressure drops in the system to maintain a steady supply of 10 bar into each foam generator unit.

Gas supply

Individual compressed gas cylinders were used for supplying the gas to a small scale prototype gas foam generator with an output of only one cubic metre per minute. However dynamic pressure losses due to the high discharge rate from the high pressure cylinders reduce the pressure and flow rate once the cylinder is half full. Manifold cylinder banks, where a number of cylinders are linked together, are an alternative solution, and were successfully used to supply gas to a single foam generator, unit developing 10 cubic metres of finished foam per minute.

The preferred gas supply for a full scale system with multiple foam generator units attached is in the form of liquid nitrogen delivered through a vaporiser 12, as shown in Figure 2. There are several different methods for vaporising liquid nitrogen and the method and equipment selected will depend on the final volume of gas required, the flow rate, the duty cycle and the mobility of the unit.

Ambient air vaporisers are often designed with an 8 hour duty before ice build up on the exterior can inhibit their heat exchange efficiency. A static vaporisation unit will have a system to switch between vaporisation units to maintain a constant supply of nitrogen gas over a 24 hour period. The gas foam system will only have a duty cycle of between 5 minutes and 1 hour for each deployment after which there will be a period of time while the system is moved to a new shed. This means that a small ambient vaporiser with an 8 hour capacity of 1000 cubic metres per hour could be used for 1 hour at 2000 cubic metres per hour then allowed to recover. An alternative way of increasing the capacity of an ambient air vaporiser that has been used with the system is to pre-heat the air passing through it. One method of doing this is to enclose the vaporiser along its length and blow hot air inside it with propane fired space heaters. An alternative energy source could be electrical heaters. One advantage of these simple vaporisers is that they can be designed to match the flow rates and pressure requirements of the foam generators without the additional use of mechanical pressure and flow control devices, beyond those required to maintain the safety of the system.

The use of water bath or direct fired heat exchangers is also considered to be suitable for supplying gas to the system. These higher capacity vaporisers require a flow and pressure control device 18 to be included prior to the distribution manifold 16. The control device should measure the pressure and flow rate of the gas and provide feedback to the control system of the vaporiser unit. The gas flow rate and pressure should be matched to the number and capacity of the foam generators units being used.

Foam solution supply

A stored water supply should be maintained to provide an adequate volume of water to the system to complete one operation. As an additional safety feature the stored water capacity should hold at least twice the volume of water required for a single operation. The stored water is held in a tank 13 but also it could be drawn from a natural or manmade reservoir provided that the pump used to deliver the water to the system has sufficient suction lift and delivery pressure to match that required by the foam generator(s). The quality of the water used in the system should be within the limits required by the surfactant in use. However, most propriety brands of high expansion foam concentrate are suitable for use with fresh, brackish or salt water. The pump 14 should have a capacity to deliver foam solution to the foam generators at the correct pressure for the design of the nozzle within the foam generator and the flow rate demanded by the number of units connected to the system. In practice, taking into account pressure drops in the system; this is a pressure up to 10 bar and a flow rate of up to 400 litres per minute.

A proportioning system 15 to inject foam concentrate into the water is positioned downstream of the pump. In the prototype system an electro-mechanical injector was used as it is highly accurate and creates minimal pressure drop relative to conventional venturi inductors. As the system may only be operated for a few minutes in some cases it is important to achieve an accurate mixing rate of foam concentrate into the water stream immediately.

Assembly of the system

The distribution manifold 16 is designed to take the bulk flow of gas and pre-mixed foam solution and divide it equally so that each foam generator receives an individual supply, through separate hoses, of gas and foam solution at the required pressure and flow rate. The manifold is designed in such a way that there is no preferential flow to any of the generators and that pressure losses are kept to a minimum. The manifold has a series of valved quick release couplers so that different numbers of foam generator units can be connected to the system depending on the demand of the operation.

Each foam generator is connected to the manifold by two flexible hoses, one for gas and one for foam solution. The foam can either be generated within the shed or outside. If it is generated outside, then it can be delivered into the shed through large diameter flexible ducting. If the foam is generated within the shed then the foam units should be equally spaced across the shed. In sheds with floor reared or loose birds, the foam should be delivered close to the floor. For multitier caged birds then the foam should be delivered from above the cages.

Experiments to test the efficacy of the system

Monitoring the physiology and behaviour of poultry during exposure to air filled foam and to anoxic gas (nitrogen) filled foam in the laboratory. The aim of this work was to investigate whether, in principle, anoxic gas foam could be a humane method of killing for poultry. Broilers and hens were exposed to anoxic (nitrogen filled) or control (air filled) foam under standardised conditions while their behavioural and physiological responses were monitored.

Materials and Methods

Subjects and husbandry broilers (Ross 308) were obtained at day-old from a commercial supplier and reared in a single group under commercially relevant conditions. The rearing pen was furnished with deep wood shavings litter and was equipped with heat lamps. The birds had ad libitum access to food and water. All experiments were carried out under Home Office Authority.

hens (ISA Brown) were obtained at 42 weeks of age from a commercial supplier and housed in individual cages. The cages had individual ad-lib feeders and drinkers and each hen had visual and auditory contact with neighbours.

Test apparatus

The foam trials were carried out in a large Perspex box (1 m x 1 m x 1 m,). One wall of the box was removable to allow full access for bird placement and bird retrieval and foam removal after each trial. The floor of the box was covered by plastic mesh to prevent birds from slipping on the smooth surface. To record each trial and allow detailed behavioural observations a video camera was positioned to view one complete side of the box, and a further wide angle camera in a waterproof box was positioned under the box to view the base. A foam generating unit as described above was modified to run a control experiment where foam was generated with atmospheric air. To facilitate this control experiment the gas supply was removed from the unit the enclosure opened to allow air to enter from the rear of the unit.

Instrumentation to measure O₂ concentration

Measurements of gas concentration within the high expansion foam matrix was carried out using a zirconium dioxide dynamic oxygen sensor (J Dittrich Elektronic GmbH Teda MF420-O-Zr) with a signal processing and recording system developed by Solutions for Research Ltd. The sensor is mounted in a short stainless steel tube behind a sintered bronze plug to prevent the ingress of moisture; in addition the probe is heated so that any moisture coming into contact with the sensors is evaporated. Three oxygen sensors were positioned at heights of 10, 30 and 90 cm in one corner of the box. These were protected from bird movement by a wide metal mesh grid (aperture 20 mm). An example of a trial to determine the effectiveness and physiological outcome of the method is described below.

Experimental procedure

Identical experimental procedures were used with broilers and hens. Individual birds were assigned randomly to nitrogen or air filled foam treatments. Immediately prior to each trial, each bird was fitted with ECG electrodes and a Lycra harness containing a telemetry unit. The telemetry function was used to verify of the existence of high quality physiological signals on each channel, and adjustments made if necessary Signal logging was triggered and a 2 min baseline period was allowed during which the bird was placed in an open cardboard carrier in a room adjacent to the test area. After baseline recording, the bird was carried to the test area and placed in the centre of the test apparatus. A ‘clapper board’ with bird number and treatment was held in front of both cameras for identification purposes. The removable wall was replaced and a further 2 min of baseline was recorded. Foam was introduced from the top of the box and care was taken not to aim foam delivery directly at the test subject. Timings of foam start, foam touching bird, complete bird submersion and foam off (when the box was full) were noted and later confirmed with web cam recordings. Synchronisation of timings of telemetry recordings and the web cam recordings ensured behavioural changes could be related to physiological responses. In anoxic (nitrogen foam) trials, all measures continued for 3 minutes after birds became motionless. In control (air foam) trials, all measures continued for 60 seconds after submersion, after which the bird was rapidly retrieved and immediately euthanized (barbiturate overdose, administered iv).

Behavioural observations

Visual obscuration by the foam limited the extent of detailed behavioural measurements. Nevertheless a number of observations were carried out from the video recordings of for each trial. As foam was introduced, counts of headshakes, foam avoidance and escape attempts were noted. After submersion, ataxia, loss of posture, wing flapping (flapping onset, number of bouts, total flapping duration) and onset of motionless were recorded.

Post-mortem examination

After removal from the apparatus all birds were subject to a brief post-mortem examination of the mouth, oesophagus and upper airway, specifically the trachea from the glottis (the tracheal opening) to the region where the trachea bifurcates into two primary bronchi. Presence of foam and any other abnormalities were noted and photographed.

Analysis

The logged data files were uploaded into a data acquisition and analysis program (Spike 2 Version 4.2, Cambridge Electronic Design) and a combination of automated and manual analysis techniques were used to produce dedicated event channels representing heartbeats per minute (2 s bins) from the raw traces during baseline and after foam application. Where clear waveforms were present, heart rate was measured every 5 s. Visual inspection of the EEG traces allowed estimation of the timing of onset of different types of EEG activity: baseline, transitional, suppressed and isoelectric. Some simple comparisons of timings of various events and behavioural responses were carried out using T-tests.

Results

During initial trials with hens, it became apparent while reviewing physiological traces that some type of noise interference was affecting EEG recordings. This was eventually traced to moisture interacting with the temperature and respiration sensors and these were disconnected so that in later trials only ECG and EEG were recorded. To compensate for the loss of EEG data, a total of 12 hens were exposed to anoxic (nitrogen filled) foam while 8 were exposed to control (air filled) foam. Ten broilers were exposed to anoxic (nitrogen filled) foam and 10 were exposed to control (air filled) foam.

Oxygen concentration measurements

Mean measurements of oxygen concentration in nitrogen filled foam are shown in Table 3 (trials with hens) and Table 4 (trials with broilers). Mean values were taken from 1 minute after each sensor was submerged in foam and calculated over the following 5 minutes. It was apparent that very low oxygen concentrations were achieved in the foam (regularly below 1% and the majority below 2%) at 10 and 30 cm heights. However, at the 90 cm sensor the oxygen concentration was generally high and close to ambient air. This was because the top-most sensor was regularly not submerged by the foam. Oxygen concentrations in the air filled foam (data not shown) were very similar to ambient, falling in some cases to a reading of 15%, most likely due to occlusion of the sensor.

ECG responses

Figure 7 shows mean (±SE) changes in heart rate in response to exposure to air filled or nitrogen filled foam. Initial exposure to air filled foam was associated with a rise in heart rate in hens (likely associated with a fear response due to novel environmental stimuli e.g. noise of foam generation) but this response was less apparent in broilers. During submersion in air filled foam both hens and broilers exhibited a trend for a slight temporary reduction in heart rate, but after 60s rates were similar to baseline. Anoxic (nitrogen filled) foam was associated with an initial heart rate increase (again likely a fear response which was more pronounced in hens) followed by rapid and pronounced bradyarrythmia which is an expected response to anoxia. Generally, the trend of a substantial fall in heart rate was followed by varying degrees of recovery (tachycardia) and/or stabilisation before a final decline. Responses were similar in hens and broilers (see Figure 7). Throughout recording, the ECG waveform was sometimes obscured due to electromyogram activity arising from the pectoral muscles or movement artefacts. This type of data loss was more severe in broilers than in hens.

Table 3 Mean oxygen concentrations in nitrogen filled foam at 10, 30 and 90 cm during trials with hens. SD = standard deviation.

<td> Bird number</td><td> Treatment</td><td> 10 cm % O₂ (mean</td><td> SD</td><td> 30 cm % O₂ (mean)</td><td> SD</td><td> 90 cm % O₂ (mean)</td><td> SD</td>

<td> 2</td><td> N2</td><td> 0.82</td><td> (0.52)</td><td> 1.13</td><td> (1.39)</td><td> 19.80</td><td> (0.27)</td>

<td> 3</td><td> N2</td><td> 0.33</td><td> (0.40)</td><td> 1.08</td><td> (0.82)</td><td> 13.40</td><td> (7.66)</td>

<td> 4</td><td> N2</td><td> 1.82</td><td> (1.91)</td><td> 0.81</td><td> (1.36)</td><td> 19.93</td><td> (0.50)</td>

<td> 5</td><td> N2</td><td> 0.76</td><td> (0.41)</td><td> 0.84</td><td> (0.64)</td><td> 19.88</td><td> (0.19)</td>

<td> 6</td><td> N2</td><td> 3.00</td><td> (1.79)</td><td> 0.87</td><td> (0.84)</td><td> 19.82</td><td> (0.32)</td>

<td> 9</td><td> N2</td><td> 0.64</td><td> (0.53)</td><td> 4.98</td><td> (2.37)</td><td> 19.50</td><td> (0.23)</td>

<td> 11</td><td> N2</td><td> 0.96</td><td> (0.43)</td><td> 0.87</td><td> (0.37)</td><td> 19.64</td><td> (0.28)</td>

<td> 14</td><td> N2</td><td> 0.25</td><td> (0.38)</td><td> 1.56</td><td> (2.27)</td><td> 15.38</td><td> (4.86)</td>

<td> 15</td><td> N2</td><td> 0.93</td><td> (0.58)</td><td> 1.05</td><td> (0.69)</td><td> 19.57</td><td> (0.53)</td>

<td> 17</td><td> N2</td><td> 1.64</td><td> (0.70)</td><td> 0.96</td><td> (0.93)</td><td> 19.45</td><td> (0.21)</td>

<td> 18</td><td> N2</td><td> 0.58</td><td> (0.48)</td><td> 0.91</td><td> (0.88)</td><td> 17.86</td><td> (2.05)</td>

<td> 19</td><td> N2</td><td> 2.67</td><td> (0.98)</td><td> 2.28</td><td> (0.77)</td><td> 19.93</td><td> (0.80)</td>

Table 4 Mean oxygen concentrations in nitrogen filled foam at 10, 30 and 90 cm during trials with broilers. SD = standard deviation.

<td> Bird Number</td><td> Treatment</td><td> 10 cm %O₂ (mean)</td><td> SD</td><td> 30 cm % O₂ (mean)</td><td> SD</td><td> 90 cm % O₂ (mean)</td><td> SD</td>

<td> 1</td><td> N2</td><td> 0.54</td><td> 0.556</td><td> 0.30</td><td> 0.284</td><td> 11.76</td><td> 8.160</td>

<td> 3</td><td> N2</td><td> 0.30</td><td> 0.274</td><td> 0.72</td><td> 0.364</td><td> 19.17</td><td> 0.328</td>

<td> 5</td><td> N2</td><td> 0.70</td><td> 0.431</td><td> 0.75</td><td> 0.519</td><td> 0.93</td><td> 1.600</td>

<td> 6</td><td> N2</td><td> 0.64</td><td> 0.561</td><td> 0.79</td><td> 0.616</td><td> 7.29</td><td> 7.857</td>

<td> 7</td><td> N2</td><td> 0.32</td><td> 0.291</td><td> 0.39</td><td> 0.449</td><td> 19.11</td><td> 0.997</td>

<td> 13</td><td> N2</td><td> 0.51</td><td> 0.343</td><td> 1.01</td><td> 0.628</td><td> 1.73</td><td> 2.992</td>

<td> 14</td><td> N2</td><td> 5.18</td><td> 2.337</td><td> 0.93</td><td> 0.233</td><td> 0.73</td><td> 0.677</td>

<td> 17</td><td> N2</td><td> 0.76</td><td> 0.319</td><td> 0.50</td><td> 0.086</td><td> 1.07</td><td> 1.116</td>

<td> 19</td><td> N2</td><td> 0.59</td><td> 0.419</td><td> 0.54</td><td> 0.444</td><td> 19.49</td><td> 0.275</td>

<td> 20</td><td> N2</td><td> 0.32</td><td> 0.302</td><td> 0.41</td><td> 0.455</td><td> 4.85</td><td> 2.975</td>

A

<img file="GB2492480B_D0004.tif" />

B

<img file="GB2492480B_D0005.tif" />

Time (s) C

Time (s) D

<img file="GB2492480B_D0006.tif" />

Time (s)

<img file="GB2492480B_D0007.tif" />

Time (s)

Figure 7 Graphs showing changes in mean heart rate (±SE) in hens exposed to air filled foam (A), broilers exposed to air filled foam (B), hens exposed to nitrogen filled foam (C) and broilers exposed to nitrogen filled foam (D). Line markers indicate timings of foam start (solid grey line), earliest and latest submersion (dashed lines) and earliest and latest time to motionless (solid black line, C and D only).

EEG responses

As explained above, some of the EEG files were unusable, particularly towards the end of the traces (as the equipment became more affected by moisture ingress). Baseline EEG activity consisted of low amplitude, high frequency activity reflecting the birds' alert state. In some birds exposed to air filled (control) foam the EEG pattern did not deviate from the baseline state, with no noticeable EEG changes (see Figure 8 A, B, C) while in others (1 hen, 6 broilers) there was evidence of slow wave activity during submersion (see Figure 8 D, E).

During exposure to anoxic (nitrogen filled) foam, a series of consistent changes in the appearance of the EEG were apparent. Visual inspection of the traces was used to assign portions of the EEG to one of 4 phases with particular characteristics where baseline was as before foam introduction; 'transitional' was high amplitude, low frequency activity or high frequency but reduced amplitude signal; 'suppressed' was a greatly suppressed EEG but containing some slow wave activity; and 'isoelectric' was residual low-level noise indicating lack of EEG activity. Transitional EEG tended to be characterised by slow wave (high amplitude, low frequency) activity and this response was seen in all birds exposed to anoxic foam (Figure 9). Tables 5 and 6 show the timings of phase changes in hens and broilers respectively.

<img file="GB2492480B_D0008.tif" />

379 380 381 382 383 384 385 386

Figure 8 EEG trace excerpts from two birds in response to submersion in air filled foam. A - hen 20 EEG during baseline; B - hen 20 EEG after 10 s submersion in air filled foam; C - hen 20 EEG after 50 seconds submersion in air filled foam; D broiler 15 EEG during baseline; E - broiler 15 EEG after 30 s submersion in air filled foam.

Figure 9 EEG trace excerpts showing the appearance of baseline (A, B) and transitional pattern (C, D) in hen 3 and broiler 5 respectively (both exposed to nitrogen filled foam).

........I¹¹¹’-¹¹'¹¹................................

<img file="GB2492480B_D0009.tif" />

60.4 60.6 60.8 61.0 61.2 61.4 61.6 61.8 62.0 62.2 s

6.0 36.2 36.4 36.6 36.8 37.0 37.2 37.4 37.6 37.8 s

<img file="GB2492480B_D0010.tif" />

Table 5 Individual and mean timings of EEG phase changes in hens. Data loss due to EEG noise interference or artefact is indicated by a dash.

<td> Hen</td><td> Transitional</td><td> Suppressed</td><td> Isoelectric</td>

<td> 2</td><td> 10.3</td><td> -</td><td> -</td>

<td> 3</td><td> 9.2</td><td> 17.5</td><td> 53.8</td>

<td> 4</td><td> 12.1</td><td> 38.1</td><td> 74.6</td>

<td> 5</td><td> 12.9</td><td> 38.3</td><td> 64.9</td>

<td> 6</td><td> 15.0</td><td> -</td><td> -</td>

<td> 9</td><td> 8.0</td><td> 25.3</td><td> 82.4</td>

<td> 11</td><td> 5.5</td><td> 36.0</td><td> 51.7</td>

<td> 14</td><td> 8.5</td><td> -</td><td> -</td>

<td> 15</td><td> 7.6</td><td> 31.9</td><td> -</td>

<td> 17</td><td> 10.3</td><td> 27.1</td><td> 73.0</td>

<td> 18</td><td> 11.4</td><td> 26.5</td><td> -</td>

<td> 19</td><td> 6.5</td><td> 30.0</td><td> 59.3</td>

<td> Mean ± SD</td><td> 9.8 ±2.8</td><td> 30.1 ± 6.8</td><td> 65.7 ±11.5</td>

Table 6 Individual and mean timings of EEG phase changes in broilers. Data loss due to EEG noise interference or artefact is indicated by a dash.

<td> Broiler</td><td> Transitional</td><td> Suppressed</td><td> Isoelectric</td>

<td> 1</td><td> 7.7</td><td> 19.2</td><td> 50.1</td>

<td> 3</td><td> 14.9</td><td> 23.3</td><td> 60.1</td>

<td> 5</td><td> 6.0</td><td> 18.9</td><td> 48.5</td>

<td> 6</td><td> 5.0</td><td> 14.4</td><td> 41.4</td>

<td> 7</td><td> 4.7</td><td> 13.1</td><td> 41.7</td>

<td> 13</td><td> -</td><td> 13.7</td><td> -</td>

<td> 14</td><td> 3.3</td><td> 21.0</td><td> 45.5</td>

<td> 17</td><td> 13.9</td><td> 18.5</td><td> 48.2</td>

<td> 19</td><td> -</td><td> -</td><td> 31.4</td>

<td> 20</td><td> 11.1</td><td> 16.4</td><td> 55.0</td>

<td> Mean ± SD</td><td> 8.3 ±4.4</td><td> 17.6 ± 3.5</td><td> 46.9 ±8.3</td>

In hens, time to transitional EEG onset ranged from 5.5 to 15.0 s (mean 9.8 ± 2.8s, n=12), while suppressed EEG onset ranged from 17.5 to 38.3 s (mean 30.1 ± 6.8s, n=10). Onset of isoelectric EEG ranged from 51.7 to 82.4 s (mean 65.7 ± 11,5s, n = 7) in hens. In broilers, time to transitional EEG onset ranged from 3.3 to 14.9 s (mean 8.3 ± 4.4s, n=8), while suppressed EEG onset ranged from 13.1 to 23.3 s (mean 17.6 ± 3.5s, n=9), and onset of isoelectric EEG ranged from 31.4 to 60.1 s (mean 46.9 ± 8.3s, n = 9). T-test comparisons revealed that broilers exhibited suppressed and isoelectric EEG significantly sooner than hens (P&lt;0.001, P=0.004 respectively). Suppressed EEG is a reliable indicator of loss of consciousness. On this basis, the maximum measured time that consciousness was a possibility during euthanasia with anoxic (nitrogen filled) foam was 60 s in hens and 24 s in broilers.

Behavioural responses

Birds exposed to air filled foam exhibited headshakes, escape attempts and wing flapping. Headshaking was seen in both hens and broilers in response to initial foam delivery (mean number 2.1 ± 3.6, range 0-10 in hens and 2.5 ± 2.5, range 0-6 in broilers). Escape attempts (vertical jumps at box wall) were seen in one hen (2 attempts) in response to control foam delivery. After submersion, some birds exhibited brief and sometimes repeated wing flapping/struggling responses (mean number of bouts 1.4 ± 1.3, range 0-4 in hens and 3.5 ± 2.1, range 0-6 in broilers). Table 7 summarises the behavioural responses exhibited by hens and broilers in response to nitrogen filled foam. As in control birds, headshakes were observed (frequency not significantly different from control foam in hens or broilers, T-test), as were escape attempts (hens only, 2 individuals). Anoxic foam also induced ataxia/loss of posture and vigorous wing flapping characteristic of anoxic death. Ataxia/loss of posture was seen between 10-23 s after submersion in hens (mean 15.5 ± 3.9s) but significantly earlier between 0-14 s in broilers (mean 9.2 ± 4.0 s, P=0.001, T-test). The onset of vigorous wing flapping occurred between 13-24 s in hens (mean 17.8 ± 3.9 s) and between 10-23 s in broilers (mean 15.3 ± 4.7 s) and was not significantly different between strains. Number of flapping bouts also did not differ between hens and broilers (mean 3.9 ± 1.4 s and 3.6 ± 1.0s for hens and broilers respectively). Flapping duration was also not significantly different, and ranged from 5-21 s in hens (mean 13.7 ± 4.2s) and 6-21 in broilers (mean 13.7 ± 5.5 s). Time to motionless was significantly shorter in broilers (P=0.004, T-test), ranging between 40-64 s (mean 51.4 ± 7.6 s) compared to 43-81 s (mean 65.2 ± 10.9 s) in hens.

Table 7 Mean ± SD and ranges of frequency, latency and duration of various behavioural responses exhibited by hens and broilers in response to nitrogen filled foam. Timings for ataxia/loss of posture, flapping onset and time to motionless are expressed in relation to submersion.

<td></td><td></td><td> Headshakes</td><td> Ataxia/loss of posture (s)</td><td> Flapping onset (s)</td><td> Flapping bouts</td><td> Flapping duration (s)</td><td> Time to motionless (s)</td>

<td> Hens</td><td> Mean ± SD</td><td> 2.0 ±1.8</td><td> 15.5 ±3.9</td><td> 17.8 ± 3.9</td><td> 3.9 ± 1.4</td><td> 13.7 ± 4.2</td><td> 65.2 ± 10.9</td>

<td></td><td> Range</td><td> 0-6</td><td> 10-23</td><td> 13-24</td><td> 2-6</td><td> 5-21</td><td> 43-81</td>

<td> Broilers</td><td> Mean ± SD</td><td> 2.4 ±2.1</td><td> 9.2 ±4.0</td><td> 15.3 ± 4.7</td><td> 3.6 ± 1.0</td><td> 13.7 ± 5.5</td><td> 51.4 ±7.6</td>

<td></td><td> Range</td><td> 0-7</td><td> 0-14</td><td> 10-23</td><td> 2-5</td><td> 6-20</td><td> 40-64</td>

<td></td><td> T test</td><td> NS</td><td> P = 0.001</td><td> NS</td><td> NS</td><td> NS</td><td> P = 0.004</td>

Post mortem observations

In response to air filled foam exposure, several birds swallowed foam so that it was visible in the mouth and oesophagus (Figure 10). Three birds (one hen, two broilers) regurgitated food during exposure to control foam. Many birds (2/8 hens and 7/10 broilers) exposed to control foam also had small amounts of foam in the tracheal opening, but the trachea was never occluded. Birds exposed to nitrogen filled foam also had foam in the mouth and oesophagus, and some regurgitated food (two hens, five broilers). Birds exposed to nitrogen filled foam regularly had foam present deeper in the trachea than controls (10/12 hens and 9/10 broilers). In one hen, foam was present as far down as the region there the bronchi, but foam was more usually observed 3-10cm from the tracheal opening.

The foam was visible as small bubbles clinging to the tracheal wall (Figure 10) and in no instance was the trachea even partially occluded by foam. No other abnormalities were observed in the mouth, oesophagus or upper airway. Two broilers exposed to nitrogen filled foam had a broken wing on removal from the apparatus.

Discussion

Behavioural responses to initial foam delivery were somewhat variable (hens tending to be more reactive than broilers) but were not particularly pronounced. Responses seemed to be more related to the noise from the foam generator rather than the foam itself, and birds tended to stay still while the foam enveloped them. Transient increases in heart rate (again more apparent in hens) corroborate that fear was likely to be experienced during initial foam introduction. Consistencies in responses such as headshaking suggest that the hens did not perceive any difference between the air filled and nitrogen filled foam before submersion.

Responses to submersion in air filled foam provide a control to determine responses to the foam itself (rather than any gas it contained) and there was evidence that being submerged was associated with some distress. Most birds exposed to control foam swallowed and/or inhaled small amounts of foam, some regurgitated food and all appeared distressed by the procedure on removal. However, it should be noted that control birds were submerged conscious for 60s which is more prolonged than the duration experienced by birds exposed to the nitrogen foam (on average conscious for between 18-30s after submersion). During submersion in air filled foam some birds exhibited slow-wave EEG patterns characteristic of sleep or reduced vigilance, and these are probably associated with protective eye closure in response to the foam (eye closure and the generation of slow wave EEG are closely associated in birds; (McKeegan, unpublished observations). A slight reduction in heart rate was also apparent during submersion in air filled foam, and this may be related to the dive reflex in response to submersion during which a bradycardia is evoked to conserve oxygen (Borg et al 2004).

Submersion in nitrogen filled foam led to euthanasia, characterised by a pronounced bradyarrythymia, vigorous wing flapping and altered EEG pattern. Compared to previous reports of anoxic gas killing (in nitrogen) under laboratory or commercial conditions the timing of these events was particularly rapid. For example, McKeegan et al (2007) reported that broilers undergoing anoxia exhibited wing flapping after 40 s, compared to an average 18 s here for hens and 15 s for broilers. Time to death in that study was 95 s compared to time to motionless here of 65 s for hens and 51s for broilers. Similarly, in a commercially relevant study using Argon, time to motionless in broilers was more than 180 s (Abeyesinghe et al, 2007). That nitrogen foam application results in an impressively effective anoxic death is supported by the very low oxygen concentrations measured at bird level. Foam is advantageous in this regard as it eliminates the possibility of air in the feathers of the birds increasing the oxygen concentration during introduction to the anoxic environment.

The maximum time to loss of consciousness can be defined by time to suppressed EEG, and this is a conservative approach since the slow wave transitional EEG pattern exhibited by all the birds in the transitional phase is also consistent with a reduction in vigilance state. Onset of vigorous wing flapping tended to occur in the transitional phase, after ataxia and loss of posture. The onset of isoelectric EEG and motionlessness were closely associated, making time to motionless a reasonable measure of time to death. The timings of these responses were fairly consistent between birds, with a general trend for ataxia, onset of suppressed and isoelectric EEG and motionlessness to occur sooner in broilers than hens.

During post mortem examinations small bubbles were observed in the tracheas or tracheal openings of almost all birds. Since the foam used to deliver the gas had bubble diameters of 10-20mm, the presence of these small bubbles needs to be explained. During anoxic death it was apparent that the vigorous flapping exhibited by the birds caused the foam in their immediate area to be whipped into a froth consisting of very small bubbles. It is likely that some of this froth was then inhaled in the later stages of euthanasia. However, some of these small bubbles were also seen in the tracheal openings of control birds, and it may be that the movement of birds in control foam (which sometimes included wing flaps) led to a similar froth, and the air filled foam tended to have more variability in bubble size (because of being generated with atmospheric air rather than compressed gas). Nevertheless, it is important to note that in no case was there any evidence of occlusion of airways by foam.

Conclusion

The trials with individuals show that submersion in anoxic (nitrogen filled) foam provides a highly effective and rapid method of euthanasia. Initial aversion to the foam is not extreme; although submersion of conscious birds in air filled foam for 60 s appeared to be unpleasant. The EEG pattern in birds submerged in anoxic foam began to change very rapidly (on average in less than 10 seconds), and unequivocal unconsciousness (suppressed EEG) was apparent between 17-30 s after submersion. The rapidity of the response, physiological observations and measurements of oxygen in the foam all show that the method of killing was anoxia with nitrogen gas, not occlusion of the airway. This is corroborated by post mortem findings. The results provide proof of principle that submersion in anoxic (nitrogen filled) foam having an expansion ratio in the order of 350:1 is a humane method of euthanasia in hens and broiler chickens.

References

Benson, E., Malone, G.W., Alphin, R.L., Dawson, C.R., Pope, C.R. and Van Wicken, G.L. (2007). Foam based mass emergency depopulation of floor reared meat type poultry operations Poultry Science. 86:219-224.

Dawson, M.D., Benson, E.R., Malone, G.W., Alphin, R.L., Estevez, I. and Van Wicken, G.L. (2006) Evaluation of foam based mass depopulation methodology for floor-reared meat-type poultry operations. Applied Engineering in Agriculture 22(5): 787-793.

Gerritzen, M.A.; Sparrey, J. (2008) A pilot study to assess whether high expansion CO2-enriched foam is acceptable for on-farm emergency killing of poultry.

Animal Welfare 17-3.

Gerritzen, M.A., H.G.M. Reimert, H.G.M., Hindle, V.A., McKeegan, D.E.F., Sparrey, J.M. (2010) Welfare assessment of gas-filled foam as an agent for killing poultry. Wageningen UR Livestock Research report 399.

USDA/APHIS Water-based Foam Euthanasia Performance Standards. 2006. Available at:

http://www.avma.org/issues/policv/poultry depopulation.asp#Attachment%20A