How I Fixed Dust Infiltration

Last update Jul. 5, 2024. 

This post is slightly out of scope for this blog, being a sharing of experience and how-to piece, not a suggestion for a social experiment. Call it a case study, the conclusions of which always require confirmation.

I concluded based on copious and diverse lived evidence that I was being made sick by dust infiltrating into my apartment. The answer (symptom reduction) was to seal up all dust entry points in my apartment. This is a medium-sized, basement apartment in an older WWII-era apartment block. The entry points were many and varied. To find them all, I needed a mirror and strong light to see under every overhang. If you try this, always illuminate areas from a variety of angles. If I cannot keep my eyes open for longer than twelve seconds without blinking, I have a problem.

The applicable maxim is not “If it ain’t broke, don’t fix it,” but “When in doubt, fix it.”

I will not be imparting all my thoughts on this topic; compile a list of rules of thumb that help you discover dust entry points in your location.

This dust is extremely penetrating and may be full of static electricity; if you can  see a crack, it’s too wide. On a white surface, an important hole can be misinterpreted as a small piece of debris such as an adhering paint chip. Biological containment facility-grade workmanship may be required, which in any event will be useful against objectionable smells, hot or cold drafts, insect visitors, asbestos, mold, radon, COVID, or whatever comes down the pike next. (June, 2023: What came down the pike next was wildfire smoke.) Moreover, this dust does not seem to settle in confined spaces: a mirror set out in a closed cupboard or closet will collect nothing, even if the space contains active dust entry points.

Sealant applied near the ceiling or near any metal inexplicably develops bubbles, ridges, and holes before it can fully cure. Using the minimum amount of a fast-drying adhesive such as plumber’s glue (Gloozit) is helpful. 

Gloozit is transparent and colourless, so its presence will not be offputting to future tenants and thus the landlord, if you do a neat job. (Cleans up in Varsol.) 

Before sealing anything, vacuum the area and wipe it down with a paper towel moistened with Varsol followed by a dry wipe. Dust entry points accumulate an oily substancee, also inexplicably, which must be removed with Varsol before sealing. Another sealant, which is preferred because of its durability, is clear silicone caulking or white, paintable silicone caulking. You will need a caulking gun with the silicone caulking. Use lots of paper towels to clean excess sealant and buy a caulking tool to keep things neat. Paintable silicone is also useful for filling in pits in walls. Use the straight end of the caulking tool to create a flat surface after application.

In tile floors, the grout may crack and create dust entry points. Use a toothbrush to scrub the grout with concentrated bleach, wipe, rinse, and allow to dry. Then mask off the damaged grout with post-it notes, apply clear silicone caulking, tool, lift off the masking, and touch up if necessary. Clean smears off the tile with the straight end of the tool. Allow to cure overnight. The paper masking is set very slightly back from the edge of the tile, to allow the caulking to bridge across cracks that occur exactly at the edge.

To seal a big hole in an otherwise flat surface, apply a wall-repair patch followed by spackling applied with a 3” drywall putty knife. For straight, wide gaps, substitute sbacker rod for the wall repair patch and apply spackling as before. Contractor-grade masking tape can be useful but will not stick if you get sweat on it or on the surface it will go on. Impossibly complex areas like windows can simply be sealed off behind window film.

Because of the complexity factor, I had to isolate a closet and seal it off with weatherstripping on the frame of the door, which is now kept closed. The weatherstripping should not be too stiff, to keep down the force needed to close the door. All old paint was cleaned off the striker plate. A 1/2"-thick rectangular plywood insert was placed on the floor to hold the weatherstripping against the bottom of the door and transfer the closing force to the back wall. (A door bumper strip may also work.) I measured carefully before cutting the insert to size; this was a precision job. The weatherstripping was caulked in the corners. The crack between the insert and the floor was also caulked along the front. I ran my finger along all mating surfaces to make sure they were flat and eliminated all hollows and bumps, even small ones caused by old paint runs. Broad defects that require a straightedge to detect did not seem to be a problem. I finished with a light test on the theory that if the door doesn’t leak light, it doesn’t leak dust. Glowing sections were built up further with weatherstripping tapered at the ends and supplemented at the dicey spots with gaskets made by applying a slight excess of silicone caulking, covering it with scotch tape, closing the door firmly, and going to bed. Before starting the gasket job, put masking tape on the door to prevent silicone caulking from bonding to the paint job, which would eliminate the option of a quick undo. (A slow undo remains possible by alternately softening the caulking with Varsol and gently scraping with a fingernail.) An alternative to the light test is the stain test, based on transferring a thin layer of fresh, colored silicone caulking from the door to the weatherstripping when the door is temporarily closed. Uncolored spots on the weatherstripping indicate dust entry points.

Since your head will be close to dust entry points during this process, you must wear a respirator rated for hazardous dusts while you work and as much as possible at all other times, until you are finished. 

Progress can be judged by laying out mirrors for a fixed time and assessing the amount of dust they accumulate using a strong light such as a gooseneck desk lamp. This method can be used with multiple simultaneously exposed, identical, numbered mirrors to obtain spatial information about dust infiltration. One mirror in each corner of the room and in each corner of each window and door seems to suffice, but if you are not making progress, use more. You may have to hang some mirrors from map pins. Diagramming helps; locate the worst corner and then refine your estimate of where the entry point is by examining the diagonal greater-than relationships. The observed spatial pattern can be distorted by gravity, ambient electric fields, the updraft behind the refrigerator, your own dust-raising activities (notably running the vacuum cleaner, pouring powders, or changing clothes), and any large grounded appliances. A diverse mixture of particle sizes on the mirror indicates resuspended old dust, not newly entered dust. Close to the entry point, dust is small, numerous, and irregularly spaced; dust that has travelled far from the entry point is larger, sparser, and more regularly spaced. Be open minded and don’t get hung up on one explanation of how the dust is getting in.

A mirror suitable for elevated sampling locations 

(02-10-2025: I recently put this method on steroids by using a scientific image processing package (ImageJ, FIJI package) to  automatically count the particles on darkfield images of the exposed mirrors taken at 1.6x with a smartphone (iPhone 12 mini). The camera lens was 8” above the centre of the mirror. The processing pipeline was: convert to 8-bit TIFF, select a circular ROI covering most of the mirror, clear outside the selection, subtract background with a 10-pixel diameter rolling ball (sliding paraboloid method, no smoothing), threshold using the moments algorithm, and count particles between .05 and .002 mm2 in area. The scale was set by knowing that the mirror diameter was 114 mm. Many routine steps requiring familiarity with ImageJ have been omitted. 

Clean the mirrors with window cleaner and paper towelling, holding the mirror over your head face-side down during the final wipe so that gravity takes stray lint away from the surface. Lightly spray the back of your hand and the backs of your fingers with water before doing this, so that rising warm, humid air from your hand kills some of the static electricity on the mirror that would otherwise attract dirt. Wipe slowly. Avoid using microfibre wipes: these leave micro lint in the size range of interest that can affect the results. I applied window cleaner with a small bar of sponge and rubbed it in with my finger, avoiding the rims; the mirror rims can trap fluid, leading to smudges later. Smudges should be removed, but some large lint can be present after cleaning without affecting the results. Mirrors with many flaws in their silvering should not be used. Cleanliness should be assessed using the same lighting as the photography. Smudges can appear with a delay after cleaning if they have been hiding in liquid films. If there are no smudges, a dry wipe may suffice for cleaning. Wash your hands before handling the mirrors and handle them only by the rims.

Set the focus of the camera once at the start of each batch of images; you do not have to set it for each image. 

I processed the raw counts through Excel to get the percentage of all particles captured by all mirrors due to each mirror. These numbers were then annotated onto a diagram of the mirrors layout in the dwelling, from which I drew my conclusions. Each mirror was handled in its own transparent plastic box with an easy-to-remove lid. Technically, mirrors used this way are called “witness plates.” Mirror numbers were indicated by felt marker marks on the rims that were visible in the images. I photographed the mirrors before and after exposure and subtracted the blanks when processing. When counts are low, this can lead to negative counts due to particles disappearing during or after exposure. Possibly, they levitate under the influence of electrostatic forces. At higher counts, always subtracting the blanks absolves you of having to clean the mirrors after each observation: just update the blanks column to the previous observations.)

Cover all air inlet and outlet registers with MERV 13 furnace filters attached and sealed with masking tape or Mulco (a clear, so-called temporary sealant that comes off with a little acetone).

All masking tape seals will develop loose spots over time, which can be seen with a strong light held at a low angle and moved around, and patched up by applying Gloozit with the finger. Adhesive tapes generally will not stick very long to anything that does not feel flat to the touch. This includes the tape used to apply window film. If necessary, sand it flat and clean off the standings before applying tape. Remember that older paint may contain lead and wear your respirator while sanding.

Temporarily remove everything removable, like doors, cupboards, and range hoods, to inspect behind them for cracks and holes. You wouldn’t think a door hinge could conceal an opening in the frame, but it can. Always use a permanent repair product in such areas and avoid tapes and temporary sealants, because you don't want to have to re-access the area in a year or month. Doors are more conveniently treated one hinge at a time, after door-stopping and shimming underneath to support the weight of the door while one hinge is unfastened. Seal all empty fastener holes you find, not just in door frames, and seal around the heads of any fasteners that appear slightly canted. Ask the landlord to make any obviously needed repairs if the defect could let in dust, but don’t get known as a complainer. Compliment the landlord for the repair on Google Reviews or BBB. 

Whenever you seal anything, come back and inspect for new holes a day later. 

If there is no chance that the airspace on the far side of the surface you are sealing can become pressurized, seal it with petroleum jelly (Vaseline). This product is also useful for controlling what silicone caulking sticks to.

Standard electrical wall switches are not dust entry points if the cover plate is sealed to the switch and to the surrounding wall.

To anticipate dust entry points, I sometimes imagine my living space as completely submerged in water and ask myself, “Where could it leak in?”

You will have to do maintenance on your sealing job, so keep those mirrors.

A crack or chip in a coat of paint can be a dust entry point if the paint has separated from the underlying surface, thereby creating a narrow space in which air-suspended dust can flow. Analogously, a kitchen counter made of particleboard covered in laminated plastic can develop delaminations that allow dust entry where the board has been repeatedly wetted and/or traumatized.

A dust entry point under a faucet in the bathroom (red arrow. The entire linear feature indicated was the entry point.)
This one was really hard to find.

A massive piece of furniture moved close to a previously sealed moulding can push down the floor enough to make cracks around the moulding, either at that point or within a few feet to either side. Seal these cracks with paintable silicone caulking and move the furniture back to its intended use position while the caulking is still wet.

Do You Have All the Latest Patches?

There's about five and a half billion adult humans alive right now, and any two of them would be more powerful than you if they cooperated and your gang wasn't around, and they kind of do all cooperate, mostly subconsciously, and we call it God.

Geese flying in formation: a simple swarm intelligence phenomenon 

In a fast-moving malware landscape, commonly used programs often need quick patches (code snippets) to close newly discovered vulnerabilities and keep them useable.

Let us suppose that humans are like that. Let us suppose the existence of something real behind belief in God: a human swarm intelligence that is slow-moving and poorly accessible to consciousness. This swarm intelligence would be a product of natural selection and [the rest of this paragraph is conjecture] talks to us in a language of emotions. It tracks slow changes in the environment while the conscious mind tracks the fast ones. The two environment trackers communicate via a frequency band centered on one reciprocal week. This is also the dividing line between the bands processed by the two trackers.

 The week does not correspond to any astrophysical cycle; it is an invention (discovery?) of religion. 

What were you doing last weekend? Don’t remember, huh? You don’t need to; God will calculate.

Arguably, the slow swarm intelligence will occasionally fall behind in tracking especially rapid environmental change (e.g., global warming), show evidence of a weak spot in its frequency response, and, pardon the blasphemy, need to be patched. At least until natural selection catches up. Logically, this same natural selection should by now have come up with a solution: A proclivity of humans to self-organize into some patch-applying sub-process. This would be organized religion, of course.

The last patch applied to human nature seems to have been a package of five patches: the Pillars of Islam. Before that, we received two big patches at once: Jesus’ "Love God” and “Love thy neighbor.”

However, we now live in a world that would have been inconceivable to the 12 disciples and the early followers of Mohammed, and collectively we are facing unprecedented existential threats. Is it time for another patch?

I say yes, two in fact, and I suggest that the needed patches will be something very like this:

“Love Thy Planet,”

and

“Do Not Divide the People.”

The first of these precepts is already addressed by a recently-added fifth Mark of Mission of the Anglican church and a sixth Mark of Mission that speaks to the second patch was accepted on principle, then folded into the fourth Mark of Mission. The proposed patches are also echoed in the Anglican Baptism Covenant. Many other people have obviously been thinking along the same lines as myself. The new “commandments” will have to be presented and received as the voice of God to be effective, and I will not here tell organized religion how to do its job.

However, does organized religion itself need a patch? It seems to suffer from a worrisome lack of flexibility, leading to hand-wringing schisms. Perhaps a further needed patch is the definition of a new kind of commandment from God that could be called a “manifest commandment,” as opposed to a prophet-channeled commandment. This would be God talking to the entire human race at once.

Since there is little time, we cannot wait millennia to accumulate enough testing of new precepts in the crucible of life to confidently promote them to commandments from God. For example, the two I just mentioned. Thus, we must aggregate such human experience across space instead of time, using the Internet. Luckily, there are now about 5.5 billion of us able to participate, which should ensure a very high-powered study. Since this would essentially be an experiment on humans, the internet could also be used to collect the ethically required written, informed consents.

The above  reasoning argues against atheism but not for the status quo. Is panreligionism the only way to ensure that you have all the latest patches? Oligoreligionism may be more practical. I now have an app on my phone that notifies me of the Islamic prayer times, even the ones after dark. I don’t say the prayers but meditate instead. I wake up in the morning feeling like I have been watched over all night. 

However, the ideal human swarm intelligence will be a function of the details of the problems of existence we face, and these details will be a function of geographical variables like latitude, altitude, and distance from the ocean. Even with the same details, different populations will need different patches because of historical factors, such as the amount of Neanderthal and Denisovan DNA they have. 

Photo by Nancy Hughes on Unsplash