We had our first major snowstorm of the season last night, and as I was shoveling the driveway I was thinking about different ways to remove snow.

Okay, I'll be honest--I was trying to figure out how to justify installing a snow-melting system when we have to replace our driveway in a few years. I still shovel the drive by hand, but I can foresee a time when I won't want to do that any more or will be traveling enough so I can't.

There are four basic ways to remove snow and ice from a driveway: shovel it by hand, clear it with a snowblower, melt it with a heated driveway, or hire a snowplow service. (You could look at a fifth possibility, melt it with chemicals, but that would require so much chemicals as to have serious environmental consequences. Chemicals are best used for stubborn patches of ice which are hard to remove mechanically.)

The Physics Perspective

The most obvious way to look at the problem of How to Remove Snow is to compare the energy required to melt snow vs. move it.  I measured our driveway and found that it is about 1,200 square feet (I'm going to use English rather than metric units because they're probably more familiar to my readers).

If we get a heavy snowfall of a foot, which translates to an inch of equivalent rainfall (Minnesota's snow tends to have one inch of rainfall equivalent for every 8-15 inches of snow), that's about 6,000 pounds of ice on the driveway which needs to be melted (which will yield about 750 gallons of water, if you're keeping track).  It takes 144 BTU to melt a pound of ice, so it will take about 850,000 BTU to melt all the snow.

In addition to melting the snow, you also have to heat the driveway itself. If there's three inches of brick over the 1,200 square foot driveway, that's about 40,000 pounds of brick. In the worst-case scenario, that brick needs to be warmed by about 100 degrees F, which will take about another 900,000 BTU. Normally a snow-melting installation includes a layer of insulation underneath the driveway, so we don't need to heat the ground underneath the driveway. In total, then, we need about 1.75 million BTU to melt a foot of snow from the driveway on a very cold day.

Calculating the energy it takes to move the snow isn't quite as straightforward since it depends on whether you push the snow (with a plow), lift the snow (with a shovel), or launch the snow (with a snowblower). Hard-to-measure factors like friction and ice adhering to the surface can matter a lot. The simplest case is the snowblower, which essentially fires the snow out a chute. If we assume that the snowblower shoots the snow out fast enough to launch it about 30 feet straight up, then it will take about 300 BTU to clear all the snow.

This is a rather lopsided result: it takes about 5,800 times as much energy to melt the snow as to clear it with a snowblower. This is not a helpful result in my quest to justify a snow melting system. It's not the end of the story, though: a snowblower turns out to be much less efficient.

It turns out to be fairly easy to convert chemical energy from natural gas into heat. Our on-demand hot water heater (which would likely be pressed into service to drive any snow-melting system) claims to be 98% efficient, and the required plumbing would have only minimal loss, so over 90% of the energy of the natural gas would be available to heat the driveway. Delivering our 1.75 million BTU to the driveway will require just a little over 1.75 million BTU of natural gas.

Small gasoline engines, like the ones used to drive snowblowers, are not very efficient. Only about 10% of the energy content of the gasoline is actually converted into mechanical energy in the driveshaft of the engine. What's more, the snowblower has a lot of internal friction, idle time, and other losses. It's probably reasonable to assume that only 10% of the output of the engine actually gets converted into flying snow. Realistically, then, it probably takes about 30,000 BTU of gasoline (or about 1/8 of a gallon) to clear the driveway.

Even accounting for the relative efficiency of melting vs. moving snow, it still takes 58 times more energy to melt the snow. This is still not a helpful result, but there's one more wrinkle: a foot of snow on a very cold day is a worst-case scenario for the snow melting system, and melting less snow on a warmer day leads to a direct reduction in the energy required. The snowblower, on the other hand, is likely to use about the same two cups of gasoline no matter how little snow fell or how warm the weather, because most of the energy is going into friction and the important factor is how long it takes to walk the machine across the entire driveway. With only an inch of snow on a warmish sunny day, the snow-melt system might require only 2-3 times as much energy as the snowblower.

The Fuel Perspective

Another way to look at the problem is to estimate the amount of fuel consumed by the different ways to remove snow. For our foot of snow, the snow-melt system will consume about 18 therms of natural gas, or about $13 of gas at recent prices from our gas company. The two cups of gasoline the snowblower consumes is about $0.30 of fuel these days.

The amount of gasoline consumed by the snowplowing service is harder to estimate because they likely burn more gas getting to and from our driveway than they use in actually clearing the snow. Plow services tend to drive big four-wheel-drive trucks which get poor mileage (especially with a giant plow rig attached to the front), so it seems reasonable to assume they burn about 1/2 gallon (or $1.20) getting to and from each client on the route.

Finally, when I shovel the driveway by hand, it takes me about an hour and burns 720 calories according to government exercise tables. That's about three candy bars, which cost about a dollar each at the convenience store, so about $3 worth of "fuel" is required.

Here, too, there's a slight wrinkle. Our geothermal system uses waste heat to warm a storage tank for hot water, and this heat could be available for use in a snow-melt system. This could give us the first 25,000 BTU or so for free each time we run the heated driveway--not very helpful for the foot of snow on a subzero day, but a significant factor in the case where we're trying to remove a small amount of snow or ice on a warmer day. This low-use scenario could wind up costing $0.50 or less.

The Time and Money Perspective

Finally, we can look at the problem from the perspective of how much time and money it takes to clean the driveway. Right now I spend about an hour shoveling the driveway every time we have a significant snowfall, and for bigger storms this sometimes needs to be done twice or more. As already established, this costs about $3 worth of candy bars.

Clearing the driveway with a snowblower takes about a half-hour, and about $0.30 worth of fuel each time. This may seem like a no-brainer (replacing $3 of Snickers with $0.30 of unleaded and taking half the time), but the snowblower itself will cost about $500 and last perhaps five years. If I have to clear the driveway ten times a season, it's clear that buying the snowblower is the most important expense, adding about $10 to the cost of each snowfall.

Hiring a snowplow service is the most expensive option, but it takes me zero time to clear the driveway. We used to hire a service until about 10 years ago, and back then they charged a minimum of $30 every time it snowed with a surcharge for more than three inches of snow. Today it would probably cost $40-$50 for every snowfall, and our foot of snow could cost as much as $75 with surcharges.

The snow-melt system actually starts to look compelling from a time and money perspective. Like the snowplow service, it requires zero effort for snow removal, but the deep snow on a cold day will only cost about $13 in natural gas. I haven't priced the cost of installing the system, but my guess is that it would add between $2,000 and $5,000 to the cost of replacing the driveway (which will have to be done anyway in a few years). Considering that we already have a water heater capable of driving the system, we could well come in at the low end of the range.

The installation price of a snow-melt system is steep, but it should last for the life of the driveway or longer. Over 25 years, the $5,000 spent on the system will cost only $200/year, or $20 for each snowfall if we need it ten times per season. So (rounding off a little), a heavy snowfall will cost about $35 in fuel plus capital expense to melt the snow, as compared to $50-$75 for a plowing service. A light snowfall would cost only about $20 to melt (essentially just the amortized cost of installation), but $40-$50 for a service.

Concluding Thoughts

There's no question that moving snow takes much less energy than trying to melt it, and the cheapest, most efficient way to clean up after a snowstorm is to shovel by hand. I'm happy to keep doing this, but She Who Puts Up With Me has zero interest in hand-clearing our driveway.

At some point, I might not want to keep shoveling, or my business travel schedule may make it likely that I won't be in town when the snow flies. When that time comes, we can hire a service, buy a snowblower, or install a snow-melt system.

Buying a snowblower is the cheapest option, but also the least convenient--it will still require someone to spend a half-hour in the cold and blowing snow. I don't think She Who Puts Up With Me will be too excited about this, though it's still better than hand-shoveling.

That leaves hiring a service or going with the heated driveway.

If we have to choose between those options, the snow-melt system is substantially cheaper, as long as we anticipate using the service for a number of years. If we expect to need a service for only a few years (maybe we expect my travel schedule to change, or move to a different house), then the capital expense of the snow-melt system makes it more expensive.

All this is still dreaming at this point: the time to make a decision about a heated driveway is when we replace the driveway. Our current driveway is 25 years old and in poor shape, so it could be replaced at any time. On the other hand, after the geothermal system this year we're not eager to embark on another major home-improvement project for a couple years.