Deadheading: The Hidden Cost of Snow Removal
Every mile a plow drives with its blade up is wasted fuel, wasted time, and a street still waiting for service. Municipalities spend $2–4 billion annually on winter road maintenance — and a measurable share of that spending goes to driving that accomplishes nothing.
Greg Dogum · 2026-03-01
When a snowstorm hits, municipal plow trucks fan out across the road network and start clearing. Arterials first, then collectors, then residential streets. The work is physical, urgent, and expensive. What's less visible — and rarely measured — is how much of that driving accomplishes nothing at all.
Deadheading is the industry term for driving with the blade up: repositioning between service passes without actually plowing snow. It's the plow truck equivalent of an empty taxi. The truck burns diesel, the driver logs hours, the salt spreader sits idle, and the street under the tires is already clear. Meanwhile, unplowed roads accumulate snow.
Deadheading is mathematically unavoidable. On any road network that isn't a perfect loop, some repositioning travel is necessary. But the gap between the minimum required deadheading and what most fleets actually drive is enormous — and it translates directly into wasted fuel, delayed service, and unnecessary emissions.
The scale of winter road maintenance spending
U.S. state and local agencies spend between $2.3 and $4 billion annually on winter road maintenance, according to the Federal Highway Administration and the Clear Roads research program. State highway agencies alone account for over $2 billion, with winter maintenance consuming roughly 20% of state DOT maintenance budgets. Add in municipal and county operations, and the national total approaches $4 billion.
At the city level, the numbers can be staggering. New York City budgets $98.6 million for snow removal. Boston typically budgets $18 million but spent $40 million during the record 2014–15 winter — more than double — consuming 114,000 tons of salt across 110 inches of snow. Minnesota DOT spent $133 million in a single season covering 30,426 lane miles. Even mid-sized cities routinely budget $8–15 million annually.
The economic impact extends well beyond municipal budgets. IHS Global Insight estimated that states can lose $70 million to $700 million per day in wages, retail sales, and tax revenue during a snow shutdown. When plowing is slow, everything downstream is slow.
What deadheading actually costs per truck
Municipal snow plows are heavy-duty Class 7–8 dump trucks with mounted plows and salt spreaders. Under normal conditions, they average 4.5–6.5 MPG. During active plowing — low speeds, hydraulic plow loads, four-wheel drive engaged — fuel economy drops to 3–5 MPG. A 2024 study in Transportation Research Part D found that fuel consumption rises 16–23% on heavy snowfall days compared to clear conditions, with fully loaded trucks burning 13–18% more fuel than half-loaded ones.
At current diesel prices ($3.40–$3.75 per gallon), every deadhead mile costs roughly $0.70–$1.25 in fuel alone. Add labor, vehicle wear, and opportunity cost — the streets that aren't being plowed — and the true cost per wasted mile is considerably higher.
Each gallon of diesel also produces 10,180 grams of CO₂, per the EPA. At plow-speed fuel economy, that's roughly 2 kg of carbon dioxide per mile with zero service value. Scale that across a fleet of 800 trucks running 40 shifts per season, and even a conservative deadhead rate implies over a thousand metric tons of unnecessary emissions from a single state fleet.
How bad is the problem, really?
You may have seen claims that 30–60% of snow plow driving is deadheading. That figure circulates widely in industry discussions and vendor marketing, but after reviewing the peer-reviewed literature extensively, no original research source supports that specific claim. It appears to be an unsourced rule-of-thumb that gained traction through repetition.
What the verified research does show is both more nuanced and more interesting. The most striking finding comes from a Carnegie Mellon University study of Pittsburgh's snow plow operations. Researchers developed an optimized routing model and compared it against what the city was actually running (generated by commercial software). The result: optimized routes were 3% to 156% shorter than the city's existing routes.
That upper bound is worth pausing on. A route that is 156% longer than optimal means the truck is driving 2.56 times the distance it needs to. In that worst case, the majority of driving is waste.
Route optimization studies consistently find 5–50% reductions in total route length. In at least one documented case, a city was running routes more than double the length they needed to be.
Based on Kinable, van Hoeve & Smith — IISE Transactions, 2021Other documented results include the Iowa DOT's optimization study achieving a 13.2% reduction in deadhead distance, a northern Utah study finding 14.8% fewer deadhead miles, and the Clear Roads multi-state research program reporting route length reductions of 5–10% typically, with up to 50% in one case. Denver recently reported a 20% reduction in storm clearance time after deploying an AI-assisted routing system.
Measuring efficiency: the deadhead ratio
PlowPath uses a metric called the deadhead ratio. It's the fraction of total distance driven that is deadhead (blade-up repositioning) versus service (blade-down plowing). Lower is better. A deadhead ratio of 0% would mean every mile driven is a plowed mile — the theoretical ideal that is physically impossible on most networks. A ratio of 50% means half of all driving is waste.
| Greedy baseline | 55.5% | 396 km |
|---|---|---|
| CARP | 17.6% | 68 km |
Blade-up driving: 83% less. Open the runs
Across the real road networks in the gallery (October 2026), a naive greedy approach (always plow the nearest unplowed street next) produces deadhead ratios from about 23% to 71% depending on the town and the fleet, and 45% at the median: close to half of all driving. Switching to a CARP-optimized routing strategy — which considers truck capacity, depot returns, and priority tiers — brings it to between under 1% and 41%, 15% at the median, on the same networks with the same trucks.
On the same town and fleet, that cuts blade-up driving by a median of 81% (between 31% and 98%, measured in kilometres), and CARP has not done worse than greedy on any of them, with no additional trucks, no additional hours, and no changes to priority policies. The same fleet, doing the same job, just sequencing streets differently. Each town's page sets the two side by side for every fleet size it was run at: Brattleboro, Vermont, above, or Canton in Baltimore, the smallest network in the gallery.
Why plow routing is harder than delivery routing
If you've used a delivery route planner — Google Maps, Route4Me, Upper — you've worked with node routing. The software picks the best order to visit a set of stops. Snow plowing is fundamentally different. It's an arc routing problem: the service happens along the road, not at a point. The plow must traverse every road segment in its zone, not just visit selected destinations.
This distinction matters because most commercial fleet routing tools are built for node routing (the Traveling Salesman Problem and its variants). They're designed for delivery trucks, not snow plows. Applying a delivery router to a plowing problem ignores the mathematical structure entirely — and the results reflect it.
The correct formulation is the Capacitated Arc Routing Problem (CARP), introduced by Golden and Wong in 1981. CARP adds real-world constraints to the theoretical Chinese Postman Problem — limited salt capacity, multiple trucks, depot returns, priority tiers — and is proven to be NP-hard. It has no efficient exact solution at municipal scale, which is why heuristic and metaheuristic solvers (like the augment-merge and NSGA-II approaches used in PlowPath) are essential.
The human cost of slow plowing
This isn't just an optimization exercise. FHWA data shows that over 1,300 people are killed and 116,800 are injured annually in crashes on snowy, slushy, or icy pavement. Average speeds on snow-covered arterials drop 30–40%. An Iowa study found that accident rates increased 1,300% during severe winter weather on a single 30-mile highway segment.
Every minute a road stays unplowed extends the window of elevated crash risk. When a plow is deadheading across already-clear streets, the streets it should be servicing are accumulating snow, degrading conditions, and increasing the probability of collisions. Route optimization is, at its core, a way to get blades on pavement faster.
What gets better when you optimize
The documented benefits of route optimization span cost, safety, and environmental impact. Fuel savings from reduced deadheading translate directly to budget relief — even a 10% reduction across a state fleet saves hundreds of thousands of gallons per season. Faster clearance times improve road safety. Lower mileage extends vehicle life and reduces maintenance costs. And because optimized routes also improve salt distribution timing, the environmental damage from both diesel emissions and road salt runoff decreases in parallel.
The math is clear. The tools exist. The question is whether municipalities measure the right things — and whether they're willing to replace routes drawn on paper maps or generated by repurposed delivery software with solutions built for the actual problem.
Explore PlowPath's city-by-city route optimization results →
References
[1] FHWA Road Weather Management Program. "Snow & Ice." ops.fhwa.dot.gov
[2] Kinable, J., van Hoeve, W.J., & Smith, S.F. (2021). "Snow Plow Route Optimization: A Constraint Programming Approach." IISE Transactions 53(6):685–703. DOI: 10.1080/24725854.2020.1831713
[3] Golden, B.L. & Wong, R.T. (1981). "Capacitated Arc Routing Problems." Networks 11(3):305–315. DOI: 10.1002/net.3230110308
[4] Iowa DOT Office of Maintenance. "Snowplow Optimization." InTrans Project 17-626. rosap.ntl.bts.gov
[5] AASHTO / NCHRP 20-07(318). "Revised Chapter 8: Winter Operations and Salt, Sand and Chemical." transportation.org
[6] "Influence of snowfall on fuel consumption of winter maintenance vehicles." Transportation Research Part D, 2024. ScienceDirect
[7] EPA Greenhouse Gas Equivalencies Calculator. epa.gov
[8] Clear Roads Pooled Fund Research Program. "Plow Route Optimization." clearroads.org
[9] Kwan, M.K. (1962). "Graphic programming using odd or even points." Chinese Mathematics 1:273–277.
[10] Corberán, Á., Eglese, R., Hasle, G., Plana, I., & Sanchis, J.M. (2021). "Arc Routing Problems: A Review of the Past, Present and Future." Networks 77(1):88–115.
[11] MnDOT 2018–19 Winter Maintenance Report. dot.state.mn.us
[12] Nguyen, M. & Tran, B. (2024). "Optimizing Snow and Ice Route Removal Operations." J. Cold Regions Engineering, ASCE 38(2). DOI: 10.1061/JCRGEI.CRENG-721