The Challenge and Promise of High-Expansion Nitrogen Foam
Portable humane killing might be an important capability, but the best use case is not yet clear.
When a flock becomes infected with highly pathogenic avian influenza (HPAI), there are no good options. Left unchecked, the virus kills its host birds in a gruesome manner and rapidly spreads through the flock with a mortality rate of up to 90-100%. The other option is “depopulation”—culling the entire flock as quickly as possible to minimize viral spread. The current USDA policy towards HPAI is to depopulate whenever there’s an infection, a strategy known as “stamping out.”
Both of these options are bad for animal welfare and also bad for business, with the current HPAI outbreak already leading to the culling of over 100 million birds. While we can't eliminate these losses (at least not without vaccination or some other major shift in USDA policy), it’s worth considering how to make the depopulation process more humane. The method matters significantly.
Currently, the most common method is ventilation shutdown (VSD), where barn ventilation is turned off until birds succumb to heat stroke. This method is considered to be particularly bad for welfare, and is supposed to be allowed only in "constrained circumstances." However, it’s also the fastest way to kill large numbers of animals, and considering the importance of limited HPAI spread for both business and welfare, many circumstances can be considered constrained.
One proposed alternative to VSD is high-expansion nitrogen foaming—a technology that encases a flock of birds in nitrogen, so birds instead die of asphyxiation. This is much more humane, but unfortunately, I’m skeptical that this will work in the US, where our farms are significantly larger and more spread out. That said, I’m still bullish on the technology, if the right use case can be identified.
The best way to go
When I worked in biotech operations, I had to handle liquid nitrogen, which can be useful in labs to keep things cold. When working with liquid nitrogen, safety protocols were paramount, because if there’s a nitrogen gas leak in an enclosed space, there is a risk the nitrogen could displace all of the oxygen. This is not only deadly, but also hard to detect. It turns out that a human’s unpleasant experience of suffocation is not actually a reaction to a lack of oxygen, but rather a reaction to too much carbon dioxide in the air. If the atmosphere is entirely nitrogen, you might lose consciousness and die of asphyxiation before you notice anything is wrong. To guard against this, we had oxygen sensors installed anywhere there was liquid nitrogen.
You don’t want a deadly gas sneaking up on you when you’re running a lab, but in the context of humane livestock depopulation, the goal is to make the process, in a sense, undetectable. Thus, the exact property of nitrogen that makes it so threatening in a lab setting actually makes it perfect for use in animal agriculture. Of course, putting this into practice can be complicated.
For one, different animals react differently to different gas mixtures, and there’s a whole science devoted to whether it’s more humane to incorporate some amount of CO2, or other inert gasses like argon. It’s also operationally challenging to create an artificial environment with specific concentrations of particular gasses, none of which are oxygen. Nitrogen in particular is a light gas that often finds ways to escape through cracks and crevices. That said, inert gas-based methods are probably the most humane way we currently have to kill large quantities of animals at once (something like a captive bolt gun, which is used for larger animals like cows and pigs, isn’t scalable for the number of chickens we farm).
One example of inert gas-based killing in practice is controlled atmosphere stunning (CAS). In the chicken meat and pork industries, animals can be put in an airtight container right before slaughter where inert gas is gradually added until animals experience a painless loss of consciousness. This is commonly considered to be the most humane way to slaughter animals for meat on a large scale, and it’s used in 5% of US poultry slaughter plants. Because slaughterhouses need to process so many animals, CAS systems must be heavily engineered and automated.
Portable humane killing
The challenge of using inert gas for depopulation is that the killing happens on farms, not at slaughter plants. It takes a sizable capital expenditure for a slaughter facility to be able to take in a large number of live animals, expose them to a highly-controlled atmosphere, then integrate the carcasses into the rest of the production line. It wouldn’t make sense for farms to invest in this kind of capability when depopulation happens so seldomly, and ideally doesn’t happen at all. Therefore, humane depopulation capability needs to be brought in from the outside when it comes to controlling HPAI.
In the US, this can sometimes involve simply transporting an airtight container to a farm so that animals can be loaded inside in order to perform CAS. However, this process is slow and requires significant manual labor. Another possibility is just to fill the entire barn with gas, but this requires the barn to be airtight.
A third option, currently in use in Europe, is high-expansion nitrogen foaming (HENF). This technology creates an artificial environment with foam consisting of bubbles filled with inert gas. A special surfactant makes these bubbles as big as possible, which serves two purposes: 1) to ensure the ratio of gas to liquid is as high as possible, minimizing the likelihood that the animal will drown rather than asphyxiate, and 2) to allow the foam to gain vertical height, so that it can completely encase the animals. With HENF, an entire barn can be filled with foam, and the movement of the animals will gradually burst the bubbles, releasing the gas and creating a controlled atmosphere. This method requires less manual labor and doesn’t require the barn to be airtight. It’s currently the best way we have to make gas-based killing portable.
Managing HPAI
Although HENF is commonly used for depopulation in Europe, it is not yet available in the US, and structural factors here will likely hinder its adoption. Firstly, US farms are generally much bigger than farms in Europe. The largest single depopulation ever done in the US was over 5 million birds, a scale unheard of in Europe. Secondly, farms in the US are geographically spread out, making it difficult to bring in the necessary equipment. Depopulation is managed directly by the USDA, since an infection on one farm puts all of the surrounding farms at risk. Farmers are required to report an infection, at which point the USDA comes in, depopulates, then reimburses the farmer for the cost of the birds.
Unfortunately, HENF doesn’t look practical at massive scales when compared to ventilation shutdown (VSD), the speed and efficiency of which has made it the most common depopulation method. In order to fully utilize HENF, the USDA would have to operate a large stockpile of foaming equipment that it could quickly bring to any farm across the country. VSD, on the other hand, can be done in multiple barns in parallel, and doesn’t require specialized equipment other than the occasional space heater, which is relatively easy to obtain.
To illustrate the challenge of using HENF over VSD, we can look at other depopulation methods currently in use. For example, firefighting foam is a method that the USDA considers to be more humane than VSD (although in my opinion it’s still worse than HENF). This method employs a low-expansion foam typically used for firefighting. Unlike in HENF, the bubbles are very small, so instead of asphyxiating, the animals drown. The foam also can’t gain vertical height, so is only used for animals that are reared at ground level, such as broilers, turkeys, or young chicks.
According to recent data, the average time to complete a depopulation on large farms (more than 200K head) is around 17 days using portable CO2 containers or whole-barn gassing, or 11 days using firefighting foam. With VSD, the same task is generally completed in only about 4 days. For the current HPAI outbreak, 50% of the depopulations over 60K head that could have used firefighting foam (i.e. on turkey or broiler farms) still used VSD in some capacity, and the larger a farm was, the more likely it was to use VSD. Given the operational similarities between firefighting foam and HENF, we can assume VSD would still be commonly chosen, even if HENF were available.
Therefore, if HENF became available in the US, it could be used in some niche circumstances, but it is unlikely to be a feasible alternative to VSD, at least under the USDA’s current policy framework for addressing HPAI, where speed is considered paramount to limit the spread of the disease.
Looking for a problem
This is not to say we should abandon HENF as a promising technology for humane husbandry. Fundamentally, HENF represents a fairly unique capability to bring a controlled atmosphere to anywhere it’s needed. Even if it’s not used for HPAI depopulation, there still could be a role for it to play in a future humane, scalable husbandry system. From How to Be a Techno-Optimist for Animals:
Despite the promise of these emerging technologies, I admit that we don’t yet have the ability to build a scalable system where every individual animal is free from hunger, distress and mutilation; has access to proper veterinary care; is free to express natural behaviors; and is killed in a fast, painless manner. And I can’t tell you right now what portfolio of technologies we’ll need to make this a reality… We’re in a similar position to the early days of clean technology, when we knew that we needed to transition away from fossil fuels, but we didn’t yet know how. The subsequent history of climate innovation gives hope—after decades of innovation, and volumes of research being devoted to the problem, we now have a clearer idea of how we’ll navigate the climate crisis.
It’s not clear right now how exactly HENF would fit into our future supply chain, but it seems like developing it as a capability would be worthwhile. One reason for this is because the optimally humane husbandry system probably involves a lot less livestock transportation than we currently do.
Transportation is widely recognized as a major challenge for animal welfare. During transport, animals can go for long periods without food, water, climate control, or enrichment, resulting in particularly high levels of stress and mortality, especially when weather conditions are unfavorable. For this reason, one of only two federal laws in the US explicitly targeting livestock welfare concerns limiting transportation times (also see on-farm hatching). The more we can bring capabilities to animals rather than the other way around, the less transportation we’ll need.
Another reason to invest in HENF is that it could be safer for human operators than other gas-based methods. In a theoretical future farming system, if inert gas-based killing methods were used more frequently where a human could end up in the modified atmosphere, there’s a risk they would not realize they were in danger. With foam, a human could more easily get out of the way.
Currently there are no companies that provide HENF services to the US, although there are in Europe. If we could identify a situation in which humane killing is needed, access to a suitable slaughter plant is impossible, and speed is not as paramount as it is for HPAI depopulation, HENF may work well. For example, depopulation of spent hens in the layer industry, or daily mortality management (identifying and removing sick birds that need to be euthanized) of broilers could be suitable use cases. The ways that both of these processes are currently done require lots of manual labor, and if a HENF solution could cut down on those labor costs, it could save farms money and be better for welfare.
Once a suitable market opportunity was identified in the US, companies could begin to supply these markets and build out infrastructure that would make the technology easier to use in even more contexts. In the long-term, HENF might even become a practical alternative to VSD. The challenge is finding the right problem to help us get there.