Cold-weather window cleaning: Upper Midwest

Sixteen years on Twin Cities routes, eight before that selling chemistry to the trade. What the standard cleaning literature gets wrong about sub-freezing work, what ice-dam meltwater actually leaves on the glass and why a routine wash will not remove it, the percarbonate-citric ladder I run on late-winter residential, and the operational reality that the Upper Midwest winter is interior-only for everyone running a real route. The complete reference for cold-weather window cleaning.

I came to this trade from the chemistry side rather than the cleaning side, and I have been a working route operator since 2010, with the eight years before that selling window-and-glass chemistry to the trade at a Minneapolis distributor. That sequence — supply-side first, route-side second — is the unusual part of my background, and it has shaped my thinking about how the chemistry actually works in cold-weather conditions in ways that pure-trade operators sometimes find surprising. This piece is the encyclopedia entry on the part of the business that the national cleaning literature treats poorly or not at all: how to work glass in a Minnesota winter, what the chemistry actually does when the glass-surface temperature is below 32°F, what ice-dam meltwater leaves behind that a four stages routine wash will not remove, and how the seasonal-revenue structure of an Upper Midwest cleaning operation has to be built around all of it.

The starting point is one I have to make explicitly because most of the national trade references skip it: the Upper Midwest winter is interior-only, and the interior-only winter is a working-season reality rather than a stylistic preference. This is the foundational fact that organizes everything else.

The interior-only winter

From roughly the third week of November through the second-to-last week of March, exterior cleaning on Twin Cities residential routes is not viable. The ambient air temperatures sit below 32°F for the great majority of the working day, with daytime highs frequently in the single digits Fahrenheit and overnight lows that routinely hit −10°F or colder during the deeper weeks of January and early February. The glass surface temperatures track the ambient temperatures with a small downward bias on shaded elevations and a small upward bias on south-facing elevations in direct sun, but the working envelope is below freezing on the great majority of exterior surfaces for the great majority of the working day, for the great majority of the season.

The implication for a working route is that the residential exterior book closes for the winter. I close mine in early-to-mid November depending on the year's specific weather and reopen it in early-to-mid April depending on the same. The closure is not soft. I do not take exterior bookings during that window; I tell new customers asking for January work that the work is interior only and that the exterior portion will be scheduled in April; and I plan the route's revenue structure around the interior-only assumption from day one of the operating year.

Operators new to the Twin Cities market often resist this and try to push exterior work into December and January on the theory that a determined operator can make any conditions work. This produces three failure modes, and I have seen all three. The first is poor finish quality — flash-icing on the glass, streak patterns the operator cannot diagnose because the streaks are not from the squeegee technique, rubber stiffening that produces visible channel drag.[^4] The second is customer disappointment — the homeowner can see that something is wrong with the result and assumes it is the cleaner's fault, which on a sense it is, because the cleaner should have rescheduled. The third is the route-economics problem — pushing exterior work in genuinely sub-freezing conditions produces a work rate of less than half the temperate-week rate, with no premium pricing to compensate. The trip is unprofitable. The customer is unhappy. The protocol does not actually work.

The right answer is to accept the seasonal reality and structure the business around it.

The seasonal-revenue structure

A working Upper Midwest residential operation needs three revenue stages stacked together to operate sustainably across the calendar:

Stage one is the temperate-season residential exterior work. April through October, with a hard peak in May and a soft secondary peak in September-October. This is the largest single revenue category and the one the trade is structured around.

Stage two is the interior-only winter book. November through March. The interior work is a smaller-margin operation than the exterior — interior glass soils more slowly, the customer demand is lower, the per-job revenue is smaller — but it provides cash flow through the worst part of the season and it preserves customer relationships across the closure of the exterior book. I run interior cleanings on roughly 40 percent of my residential book through the winter, generally on a 90-day rather than the 60-day cycle the exterior side runs in summer.

Stage three is the commercial interior contract base. This is the structural foundation. I tell every new Twin Cities operator I mentor — there are typically three or four at any given time who reach out about chemistry questions and end up asking about the business side as well — that they should plan from day one to build at least 25 percent of their annual revenue from commercial interior contracts, because the commercial interior contracts are the cash-flow buffer that gets a residential operation through February. The alternative is to face the winter with no income and inadequate financial reserves, and that alternative has put more Twin Cities cleaning shops out of business in the past sixteen years than any other single factor I have observed in the market. The pricing and contract-structure side of building a commercial route in the Upper Midwest follows the same fundamentals Jan Davenport documents for the broader Midwest, with the seasonal-buffer use case being the specific Upper Midwest twist.

The May cabin-opening surge is the structural counterweight to the February floor. The lake-country cabin economy — the families with second homes on Minnetonka, Mille Lacs, the Brainerd Lakes, and the Boundary Waters-adjacent shores who open their cabins in late April through Memorial Day — produces a four-to-six-week revenue compression that, if you are positioned for it, can carry 35–45 percent of annual exterior revenue. The cabin work is higher-margin than routine residential (the access is typically remote, the per-stop revenue is larger, the customer expectations are clearer), and operators who have built a cabin book have a much more stable business than operators who run only Twin Cities metro residential.

The sub-freezing protocol for the edge cases

Within the interior-only winter there are edge cases where exterior work is actually necessary. The most common are commercial emergencies (a corporate building with a window-cleaning contract that includes minimum exterior visibility standards), lake-country cabin openings in late March when the cabin owners are arriving for the first weekend of the season, and isolated mid-winter callbacks where a specific exterior pane has developed an obscuring residue that the homeowner cannot ignore. For these cases the working protocol is built around the freezing-point-depression chemistry of the Cold-Weather Mix and a handful of equipment-handling disciplines that the temperate-climate cleaning literature does not cover.

The Cold-Weather Mix is one cup of 70 percent isopropyl alcohol, two drops of Dawn surfactant, and fourteen cups of distilled water. The alcohol depresses the freezing point of the solution to roughly 28°F,[^1] which extends the working envelope of the solution down to approximately 20°F glass-surface temperature. Below 20°F the working envelope closes and the right answer is to reschedule.

The mechanical protocol for sub-freezing exterior work is built around four changes from the temperate-climate baseline.

Change one: pre-warm the solution. I keep the working solution in an insulated container in the truck at roughly 65°F, which gives the applied film additional thermal mass to resist flash-icing on contact with sub-freezing glass. This is the opposite of what Jerry Davenport does in Texas, where the summer protocol pre-cools the solution to delay flash-evaporation on 130°F glass. The two protocols are mirror images of each other, organized around the same underlying chemistry — film thermal mass and the temperature differential between film and glass — but pointed in opposite directions because the working conditions are reversed.

Change two: pre-warm the equipment. Squeegee rubber stiffens substantially below 40°F and produces visible channel drag and streaking when used at sub-freezing temperatures. I keep working squeegees in a chest cooler in the truck — the cooler runs as a temperature-stabilizer rather than a refrigeration unit in winter — and rotate the working pieces through to keep the rubber within its working envelope. The replacement rate on squeegee channels in winter use is roughly double what summer use produces, and the replacement cost is a meaningful but not dominant line item in the operating budget.

Change three: shorten the working session. Sub-freezing exterior work is hard on the operator as well as on the equipment, and the protocol for working safely in genuinely cold conditions is to plan in 20-to-30-minute exterior intervals interspersed with warm-up periods in the truck. The work rate is slower than the temperate-week rate. The premium pricing for cold-weather exterior work has to account for the work-rate difference, or the trip is unprofitable.

Change four: never push the envelope. The single highest-leverage discipline in cold-weather exterior work is the willingness to reschedule when the conditions are wrong. I reschedule below 20°F ambient. I reschedule under active wind chill that brings the effective working temperature below 5°F. I reschedule when the forecast for the work day shows the temperature dropping rather than rising through the working hours, on the grounds that the protocol works best when the conditions are warming rather than cooling. The discipline of rescheduling is the part of the protocol that most matters. The work that goes well in cold weather is the work that was done under the right conditions; the work that does not go well is the work that was pushed when it should have been rescheduled.

The reschedule decision is the most important call of a cold-weather day and also the hardest one to make on instinct, because the operator who has driven thirty miles to the job site has a powerful motivated-reasoning incentive to find a reason the conditions are workable. The Cold-Weather Window Calculator takes the glass-surface temperature (not the ambient), the wind condition, the chemistry rung you are running, the trend (warming versus cooling through the working hours), and a few equipment-state inputs and returns a working-envelope window in minutes — or a do-not-proceed verdict when the envelope has closed. Using the calculator at the truck before opening the bucket is the operational discipline that converts the principle in this section into a number you can act on without arguing yourself into a bad call.

The Cold-Weather Mix referenced above is one of the named presets in the Solution Calculator — one cup of 70% isopropyl, two drops of Dawn, fourteen cups of distilled water, scaled to any working volume. The calculator also carries the higher-alcohol variant for the 10-to-20°F edge of the working envelope and the alcohol-free House Standard for the warmer-than-30°F days that occasionally show up in January. Anyone running the sub-freezing protocol should be mixing from labeled bottles; the calculator's label-print function carries the freezing-point depression number, the working temperature range, and the rinse-protocol cautions, which matters when the cleaner working in cold conditions is moving fast between trucks and panes and cannot afford a wrong-bottle mistake.

The ice-dam meltwater problem

The largest single chemistry problem of the Upper Midwest cleaning calendar is not the sub-freezing exterior work, which is an edge case. The largest problem is ice-dam meltwater, which is the protocol-defining residue of the late-winter and early-spring residential exterior work.

An ice dam forms on a heated roof when interior heat loss melts the underside of the snowpack, the resulting meltwater flows down the roof slope under the snow, and the meltwater refreezes at the colder eave and gutter line.[^2] As subsequent meltwater backs up behind the dam, some of it flows over the dam and onto the gutters; some of it flows past the dam through gaps in the roofing material; and some of it eventually flows down the exterior face of the building, including the exterior face of the windows. The water that reaches the glass has traveled across asphalt-shingle granule dust, through whatever organic debris is sitting in the gutter, past any deicer that has been applied to the roof, and over any older flashing material that may be contributing trace heavy metals to the runoff. The resulting deposit on the glass, when the meltwater dries, is a complex mineral-and-organic residue that is fundamentally different from the routine soil load that a House Standard wash addresses.

The visible signature of ice-dam meltwater residue on glass is a downward-running streak pattern that starts at the top of the window and tapers toward the bottom, with a brown-to-orange tint from the asphalt-granule fraction and a chalky texture from the mineral fraction. The streaks resist routine cleaning. An operator who treats them as routine soiling will produce a partially-cleaned pane that the homeowner will reasonably regard as a failed cleaning. The right protocol is sequential, and the sequence matters.

The percarbonate-citric ladder

The ladder protocol for ice-dam meltwater residue runs in three stages, and each stage handles a fraction the other stages do not.

Stage one: percarbonate alkaline lift. A working solution of sodium percarbonate at one-to-two tablespoons per gallon of warm water,[^3] applied to the glass with a sleeve or soft brush, allowed a short dwell, and rinsed thoroughly. The percarbonate addresses the organic fraction of the residue — the asphalt-granule dust, the decayed gutter debris, the trace heavy-metal-bound organics from older flashing — by oxidizing the organic component without compromising the underlying glass or the surrounding frame substrate. Percarbonate is the right alkaline-lift agent for this work because conventional ammoniated cleaners attack oil-based trim paint and softer wood-frame finishes, and sodium-hydroxide-based alternatives risk frame-substrate damage. The OxiClean retail formulation works for residential job-by-job work; the technical-grade sodium percarbonate available through the chemistry-supply trade works better for route operators running ice-dam meltwater protocols across multiple accounts.

Stage two: citric acid rinse. A working solution of citric acid at one tablespoon per gallon of distilled water, applied with a clean sleeve or sprayer, allowed a short dwell, and rinsed. The citric rinse addresses the mineral fraction of the residue that the alkaline lift did not remove — the calcium and magnesium components from any deicer that flowed through the meltwater path, the trace mineral content from the dissolved-asphalt fraction, the iron and manganese trace metals from older flashing that have been partially oxidized in the alkaline lift and are now suspended in the residual surface film. Citric acid is the preferred mild-acid agent for this work because it is gentle on glass surfaces, does not compromise frame substrates, and produces a complete rinse with no measurable residual.

Stage three: routine wash. The standard House Standard recipe — distilled water, three drops of Dawn surfactant per gallon — applied as the final wash to clear any residual from the alkaline and acid stages and produce the finished surface. This stage is the routine wash that the trade is built around, and it serves the same function on an ice-dam-meltwater pane that it serves on any other residential pane. The difference is that on ice-dam meltwater work, the wash is the third stage rather than the only stage.

The full three-stage ladder takes roughly twice as long as a routine wash on a comparable pane. The pricing for ice-dam-meltwater work should account for the time differential. I price the ladder work at roughly 175 percent of routine residential exterior, and the customers I have run the protocol for since 2014 consistently choose the ladder work over the alternative of leaving the residue in place until the next routine cleaning, which would be six to eight weeks out and which would let the residue set further into the glass surface.

Each of the three stages of the ladder is a sleeve-or-brush application followed by a thorough rinse and a squeegee finish, and the wet-and-dry handling that ice-dam meltwater work demands is documented across several chapters of the Technique Library. The sleeve-application motion in Chapter 04 is what carries the alkaline lift; the channel-wipe in Chapter 05 covers the squeegee-rubber handling that matters more in cold conditions than in any other season; and the detail-pass strokes in Chapter 02 are what finish the pane after the third-stage House Standard rinse. The library's value on this work is precision: each ladder stage runs differently from a routine wash, and the chapters above are calibrated for the residue chemistry rather than for the routine-soil baseline.

Cabin opening protocol

The May cabin-opening surge is the third major operational pattern of the Upper Midwest cleaning calendar, and it carries its own protocol implications that the routine residential framework does not address.

The cabins themselves have typically been closed for the winter, with no occupants and no interior heating for four to six months. The exterior windows have been exposed to a full Minnesota winter — snow, ice, wind, and the seasonal organic loading that the lake-country environment produces — and the interior windows have been exposed to the freeze-thaw cycle through the unheated cabin interior. The cleaning protocol for a cabin opening is therefore a hybrid: the exterior side is essentially residential ice-dam-meltwater work plus a heavy mineral-residue load from the lake-aerosol environment, and the interior side is essentially routine residential work plus a substantial accumulation of cobwebs, organic dust, and the occasional flash-ice-melt streak from a window that had condensation through a midwinter thaw cycle.

The pacing of cabin work is also different from routine residential. The cabins are typically remote — Minnetonka is the closest of the lake-country markets, the others are 90 minutes to three hours from the Twin Cities — and the work is bundled by trip rather than by stop. I plan cabin trips at one-stop-per-day or two-stops-per-day pacing rather than the four-to-five-stop-per-day pacing of routine metro residential, with the trip economics built around the longer per-stop time and the per-trip travel cost.

The water-chemistry side of cabin work also splits sharply from the Twin Cities metro work. Most of the lake-country cabin properties are on private wells rather than municipal supply, and the well water across the lake country runs the full range from soft-aquifer to genuinely hard-water karst. The cabin water-supply protocol for the cleaning side is the same protocol I would run on any hard-water account: the wash water comes from the truck rather than from the household tap, the rinse is distilled, and the household supply is treated as a household-plumbing concern rather than a glass-cleaning input. The protocol details parallel the pure water approach Jan Davenport documents for residential and small-commercial route work, with the same fundamental principle — never let unverified source water touch the cleaned pane — applied through different equipment.

The cabin book is the highest-margin part of my operation. The work is interesting, the customer relationships are stable across multiple seasons, and the work environment is the Boundary Waters in late May. The book has expanded gradually since 2012 and is now roughly 30 percent of my annual exterior revenue, with my sister Britt running the cabin trips with me from late May through July when she is on her school-librarian summer break. The cabin book is the structural counterweight to the interior-only winter, and the seasonal-revenue arithmetic of the operation only works because the cabin work is there to balance the February floor.

What the national literature gets wrong

The single largest editorial gap in the national cleaning literature on cold-weather work is the assumption that the cold-weather protocol is a modification of the temperate-climate protocol. It is not. The cold-weather protocol is a different protocol, organized around different chemistry, different equipment behavior, different seasonal structure, and a different right answer for almost every operational question the trade is built around. Operators who arrive at the Upper Midwest market with a temperate-climate framework spend their first two or three winters discovering this the hard way.

The second-largest editorial gap is the assumption that the working operator's job in winter is to find ways to make exterior cleaning work. The actual working operator's job in winter is to organize the operation around the reality that exterior cleaning does not work for most of the season, and to build the interior-only and commercial-contract revenue stages that carry the operation across the closure. The chemistry and the protocols I have described above are the protocols for the genuinely necessary exterior edge cases and the early-spring ice-dam-meltwater work. They are not the day-to-day operating reality of an Upper Midwest cleaning route. The day-to-day operating reality is interior work, commercial contracts, and the discipline of running a sustainable business through a five-month exterior closure that the trade's national framework barely acknowledges.

This is the protocol-and-business reference I would have wanted when I switched from the distributor side to the route side in 2010. The temperate-climate cleaning literature is genuinely useful for what it covers. The Upper Midwest winter is not what it covers, and the difference between the two has shaped sixteen years of working route operation in a way I have been threatening to write up since at least 2018. This is the writeup.

Notes

The freezing-point depression of dilute isopropyl-alcohol solutions in water is approximately linear up to about 30 percent alcohol by volume, with the freezing point dropping roughly 0.5°F for every 1 percent alcohol added. A 1:14 isopropyl-to-distilled-water mix is approximately 7 percent alcohol by volume, which depresses the freezing point of the working solution to roughly 28°F — sufficient to prevent flash-icing on a glass surface at temperatures down to about 20°F under typical ambient conditions. Below that surface temperature, the depression no longer keeps pace with the heat-transfer rate from glass to film, and the solution begins to freeze on contact regardless of alcohol fraction. The pragmatic working envelope for sub-freezing exterior cleaning with the Cold-Weather Mix is therefore roughly 20°F to 32°F glass-surface temperature. Operators occasionally claim working envelopes down to 10°F or 0°F using higher alcohol fractions; this is technically possible but is not a protocol I run because the higher alcohol load degrades squeegee rubber over time and produces measurable streaking from incomplete evaporation of the alcohol fraction. The right answer below 20°F is to reschedule.
Ice-dam meltwater is the liquid water phase that forms beneath a snowpack on a heated roof when interior heat loss melts the underside of the snow, the resulting water flows down the roof slope under the snowpack, and the water refreezes at the colder eave and gutter line, forming an ice dam that backs up subsequent meltwater behind it. The meltwater that eventually flows over or through the dam and onto the exterior windows carries the chemical load of everything it has contacted on the way down: asphalt-shingle granule dust, organic gutter debris in various states of decay, dissolved mineral content from any calcium-chloride or sodium-chloride deicer applied to the roof, and trace heavy metals from older flashing. The resulting deposit on the glass is a complex mineral-and-organic residue that the simple acid-side or alkaline-side cleaning protocols typically referenced in residential literature will not address in a single pass. The protocol that works is sequential: an alkaline lift first (to address the organic fraction), an acid rinse second (to lift the mineral fraction), and a final routine wash to clear the residual. Each stage handles a fraction the other stages do not, and skipping any stage produces an incomplete result.
Sodium percarbonate (2 Na₂CO₃ · 3 H₂O₂) is a solid oxygen-based bleach that releases hydrogen peroxide and sodium carbonate when dissolved in water. The working solution is mildly alkaline (pH 10–11 at typical use concentrations of 1–2 tablespoons per gallon) and the active fraction is the released peroxide, which oxidizes organic residue without producing the chlorine-byproduct chemistry that liquid chlorine bleach generates. Percarbonate is the preferred alkaline-lift agent for ice-dam-meltwater work because it addresses the organic fraction (the asphalt-shingle granule dust and decayed gutter debris) without compromising aluminum or wood frame substrates, which conventional ammoniated cleaners or sodium-hydroxide-based alternatives would. The peroxide fraction degrades to water and oxygen on rinse, leaving no measurable chemical residue on the glass. For working purposes, percarbonate is sold under the OxiClean brand in retail and as 'sodium percarbonate' in technical-grade supply; both formulations work, with the technical-grade product offering a higher active fraction and lower filler load for route work.
Squeegee rubber stiffens substantially at sub-freezing temperatures. The natural-rubber and synthetic-rubber blends used in professional squeegee channels are formulated for working at approximate room temperature, and below roughly 40°F the rubber loses pliability in a way that produces measurable streaking and channel drag under normal squeegee pressure. The working protocol for sub-freezing exterior work is to keep the working squeegees in a chest cooler in the truck (yes, the same cooler runs in summer for the opposite reason — keeping rubber within working temperature is the actual goal, and the cooler is a temperature-stabilizer rather than a refrigeration unit), to rotate squeegees in and out of the cooler as the working pieces drop below the threshold, and to plan on a working life roughly half what the same squeegee channel would deliver in temperate-climate use. The cost-per-cleaning impact is meaningful but not dominant; it shows up in equipment-replacement frequency rather than as a per-job line item.

Sources

Minnesota Department of Health, Public Water Supply Annual Report: Twin Cities Metropolitan Area, 2024 edition.
University of Minnesota Extension, Dealing with and preventing ice dams — the extension service for this climate on why ice dams are a heat-loss problem rather than a roof problem, and why mechanical attic ventilation is not the fix
U.S. Department of Energy, Building America Solution Center, Attic Air Sealing, Insulating, and Ventilating for Ice Dam Prevention — the federal building-science guidance on the assembly that produces ice dams, and the three-part fix: air seal the ceiling plane, insulate, ventilate
American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE), Handbook of Fundamentals, Chapter 26: Climatic Design Information, 2021 edition.
Solvay Chemicals, Sodium Percarbonate: Technical Data and Application Reference, 2023.
Archer Daniels Midland (ADM), Citric Acid: Food and Industrial Grade Technical Reference, 2024.
U.S. Geological Survey, Water Hardness Map of the Contiguous United States, 2020 update — Minnesota and Upper Mississippi groundwater section
Minnesota Pollution Control Agency, Groundwater Chemistry of the Twin Cities Metropolitan Area, environmental data series, 2023.

Cleaning glass in the cold: sub-freezing exterior protocols, ice-dam meltwater chemistry, and the seasonal-revenue reality of running a cleaning route in the Upper Midwest