Epistemic certainty: Obviously loads of uncertainty on mirror life risks and the degree to which we'd have to pressurize buildings or filter outdoor air. Moderately high certainty for the best hasty pathways for doing this in North American and a narrow subset of European buildings. Lower certainty as we move towards international buildings. Also loads of experiments and field tests are needed, possibly somewhat urgently.

Let’s discuss the threat and the parameters around it. I’m neither a scientist nor a risk analyst so we won’t spend much time here. In the future we might have a world in which mirror life is released into the world. Unmitigated, this threat could potentially extinguish all living beings. This might be via accidental leak or an omnicidal person or group. I’m not going to speak to the technical possibilities or likelihoods, but for our purposes, the outdoor air would be poisonous. To be safely inhaled we’d need highly effective filtration (99.999% fine particulate removal).

To be inhabitable buildings would have to be almost entirely resistant to air infiltration. We would achieve this by highly pressurizing buildings, which means we would blow air into them. The two main sources of building leakage are stack effect and wind, with stack being the dominant source driver of total infiltration in most climates. However, because we’re mainly concerned with peak infiltration we’ll focus on wind. We need to pressurize buildings beyond their most likely peak wind forces (not exactly but don’t worry about this yet). Pressurization can be measured, and we currently believe that most houses would be safe at a minimum pressure around 25 Pascals (Pascals are the unit of pressure we’ll use throughout this manual), although in some cases wind can exert a higher pressure so perhaps we’d need higher building pressures in especially wind-prone areas (it’s more complicated than this but we’ll cover this later).

Let’s break this problem into hardware components:

1. Separation from outdoors (building)
2. Pressure inducer (fan)
3. Particulate remover (filter)
4. Air delivery components (duct system)

Side note: isn’t it nice the way people in construction use intuitive language for hardware? We’re too dumb to remember Latin.

This manual assumes we have at least 2 weeks warning prior to the release of mirror life during which you’ll hastily prepare for the threat. We may also provide tasks that you can begin working on immediately that will help prepare you well in advance. Where we do this we’ll prioritize tasks that have broad benefits and/or are economical. We will also largely attempt to use items that are accessible and local to you. We assume no previous technical experience although following these instructions will be easier for people with backgrounds working buildings.

Buildings are Separations with Known Leakage Qualities

We inhabit buildings because they separate us from the outdoors. In general the more severe an outdoor climate, the stronger the separation. For the mirror threat, the main separation quality we’re concerned with is rate of air leakage.

You probably don’t realize this because apparently they don’t teach it in philosophy grad school, but building leakage has been quantifiable since the late 1970s with a tool called a ‘blower door’. Your author has performed many thousands of ‘blower door tests’ as well as intensive leakage reduction in ~2000 buildings in the US northeast. Many countries and US states have codified a requirement for a blower door test as part of their inspection process. Furthermore, many advanced private certifications (LEED, Passive House) require blower door testing. This means that in many countries we have a somewhat clear picture of how much air buildings leak, and the location and contribution of various leak sites.

In a blower door test we   insert a fan into a shroud in an open door in a home. This fan negatively pressurizes (unlike mirror threat which uses positive pressure) a house by sucking air out of the front door. By fixing tests at a set pressure (-50 Pa) and measuring the air across the fan in cubic feet/minute (cfm), we arrive at a number to describe leakiness in cfm@50Pa. This number allows us to make comparisons across buildings and in the same building over time. Cross-comparing buildings is complicated by the surface area (source of leaks) and volume of these buildings and the fact that several hundred years after Galileo people still struggle with the square-cube law, which we may come back to later. But for now we want to understand that there are many countries in which we have a good idea of building leakage at 50Pa, and that we can use our 50Pa number (or educated 50Pa guess) to extrapolate to a 25 Pa positive pressure which is what we need for mirror mitigation. In other words we can make informed guesses at how much air we need to pressurize houses sufficiently to overcome wind.

There are many important lessons from decades of blower door tests. The first is that there’s a huge amount variation across houses. In our experience in the northeast US the average existing house that hasn’t had focused air-sealing work will test-in (first test) at around 4000 cfm@50Pa. By international or modern standards this is extremely high. There are likely millions of Passive Houses in European cold climates that have leakage rates in the 300-800 cfm@50Pa range. The second is that fastidious, intensive air sealing work in leaky US homes will typically achieve a 1/4-1/3 leakage reduction. We’d generally expect the available relative reduction to be much lower in homes that are already fairly tight. The third is that layperson intuitions about leak pathways is poor. Inasmuch as reducing building leakage plays a role in threat mitigation, we want to carefully target known leakage paths vs those that people often mistakenly believe to be substantial (eg windows and doors).

In some cases we anticipate that despite our best efforts to reduce leakage, buildings may be too leaky to achieve the requisite pressure with available fan capacity. In these cases we may have to employ a strategy for cordoning off sections of the building internally, and sealing these sections both from outdoors and the adjacent space. We’ll call this strategy seal and cordon (s&c).

Inducing Pressure

We use fans to move the air that induces the building pressure necessary to mitigate the threat. These fans have a capacity rated in cfm. In countries where ducted heating and air conditioning are common (the US and Canada), the most powerful household fans will generally be furnace and air conditioning fans. To use these fans we will need to resolve the following:

• Identify the fan type (permanent split capacitor, electronically commutated motor, variable speed)
• Identify all controls for airflow rates and modify or bypass as necessary
• Determine if it’s necessary to relocate the fan from the H/AC system

International passive houses or similar: In households with dedicated ventilation via either local exhaust or ventilation we expect to use these fans to induce pressure. This should be fairly simple in Passive Houses where buildings are quite tight and mechanical ventilation rates are high. These houses almost universally have two fans that balance airflow and recover heat/enthalpy on the outgoing airstream (these are called heat or enthalpy recovery ventilators (hrv/erv)). We need field experimentation to verify, but it should be quite simple to reverse the exhaust fan on an H/ERV and thereby induce double the devices rating in CFM (from ~50-200 to 100-400) which our calculations suggest should induce a pressure >25Pa in a substantial fraction of these homes (again experimentation needed).

Other houses internationally, especially poor countries: Here things are much more ambiguous. In households it’s likely that local exhaust fans are the dominant powerful fans although these are likely slightly weak (40-80cfm) compared with ERVs and North American H/AC (800-2000 cfm). In addition these may be rarely found in a median household. It’s likely that the most powerful fans in many countries would be commercial or industrial exhaust. Presumably these could be repurposed but given how rare they are, we would likely want them to pressurize buildings with many occupants.

Moving Air Through a Filter

As of now it’s unclear which readily available materials show the greatest promise at removing fine particles. However, we understand many aspects of filtration that will sharpen our investigations.

• Filters induce pressure
• Filter pressure is an exponential function of filter surface area
• Filter pressure is a significant determinant of fan life, and we need our fans to last a long time under an extremely heavy duty cycle
• We need filters that remove particulate down to a very fine level (99.999% removal)
• We expect particulate to clog the filters over time, which we will have to address during the course of the threat

The combination of these means that we will very likely have extremely large filters. These filters will have to be serviceable, either via cleaning or replacement. In addition, it would be better for scaling if these filters could be standardized and produced quickly via eg cottage industries in the lead-up to the threat.

Delivering Air Through Ducts

Both the filter and the fan will inhabit a duct system. This system is used to draw air from outdoors and for our purposes includes sealants, fasteners and any rigid materials used to construct it. When constructing the duct system we must be cognizant of the risk of introducing contaminants via duct connections. The best way to mitigate this risk is filter placement as close as possible to the boundary separation to outdoors. Effectively sealing ducts further reduces this risk, but also maximizes the efficiency of the filter by sending as outdoor air as possible through the filter. Ducts should allow for (ideally modular) filters to be replaced, and for serviceability of the fan motor and control settings.

Hopefully this introduction is sufficient to give you basic familiarity with the components and vocabulary to loosely understand how to mitigate household risks from a mirror life catastrophe. In future writing we’ll look at specific scenarios, hardware and construction types so that we can provide specific and effective guidance.