Conductive Concrete Melts Snow and Ice

Winter sucks. Ice sucks. Shoveling sucks. What if roads, or your driveway, could get rid of snow and ice by themselves? (…with the help of our friend, the electron.)

A few days ago we shared a project about building an epic snow-melting system right into your driveway. But for obvious reasons, it’s not that easy to do — or cheap. But [Chris Tuan], an engineering professor at the University of Nebraska thinks he can change that.

He’s created his own special formula for conductive concrete. Which means you can turn the concrete into a resistive heat load. And this isn’t just a university research project that is going nowhere; it’s actually being trialed by the FAA for use in airports . There is a patch of it in Omaha undergoing testing right now.

And it’s actually not that complex. It consists of a mixture of 20% steel shavings and carbon particles, in a regular run-of-the-mill concrete. Apparently, this is enough to cause the entire patch of concrete to become conductive, meaning if you pump enough juice through it — it’d definitely melt some ice on top.

[Thanks Michael!]

92 thoughts on “Conductive Concrete Melts Snow and Ice

  1. Not convinced it’s a great idea.

    Normal underfloor heating has insulated cables or fluid pipes. It’s inherently much safer.

    This conductive concrete will have conductive surfaces exposed, so they’d need to use low voltages to keep it safe, and high currents to get the power levels. They’d then need bulky cables and/or a power supply situated very close in order to keep losses down. Even then, if someone drops a piece of metal on it it’s going make a better path than the concrete and conduct some very large currents, possibly overheating to the point of being dangerous.

    1. I think maybe a grid of something conducting, implanted into the concrete, might be a better idea. A grid, or at least a set of parallel elements, so that if corrosion ate away some of the circuit paths, others would still function.

      I’d be as much worried about how the concrete will stand up, carbon particles aren’t sticky, might be they’ll make the concrete quite a bit weaker.

      1. Back in the 1970s, the city of Minneapolis tried something like your suggestion. They buried resistive heating elements in the sidewalks along Nicollet Mall, which was being rebuilt as a pedestrian-friendly street running the length of downtown.

        Not too many winters later, pedestrians were complaining of receiving electric shocks from the sidewalks! The city and building owners all use a lot of salt to deice the sidewalks, and continued to do so when the sidewalk heaters couldn’t keep up with the snow and cold. The salty melt-water found its way through cracks in the concrete into cracks in the insulation. The damp leather soles of the pedestrians’ shoes were conductive enough to carry the current to their skin. As far as I know nobody was injured, but the city was ridiculed in the local press.

        The insulation could have been damaged by concrete expansion in the summers, or it could simply have been poor insulation technology provided by the manufacturer. I’m totally speculating now, but I know that at that time lots of heating wire was made with asbestos insulation, which by itself isn’t waterproof; but asbestos needles are sharp enough to pierce any plastic outer coatings they might have applied if the wire was bent too sharply.

        All the modern in-floor heating installations I know of today use boiler systems and anti-freeze as a heat transfer agent. I’ve seen small electric systems designed to be installed beneath bathroom floor tiles, but nothing approaching the size or scale of a driveway. And as far as applying current directly to the concrete? How could you absolutely guarantee that only rubber tired vehicles would run on it, and that we conductive humans could never set foot on it? It just sounds like a recipe for disaster.

        1. That’s interesting! But yeah I was thinking something along those lines. Only done properly! Like, lower voltage, and don’t put salt on it. Just as an alternative to making the concrete itself conducting. Was just an idle idea, there’s my heated-concrete empire down the drain, fallen at the first hurdle.

          1. Nah, in the 70s computers could be big but they fit in cabinets. Or at least the CPU and memory did, you might have tape drives separately. By the late 70s you could buy desktop computers, running CP/M usually on an 8080. You’re thinking of the 50s.

            They had wrist computers in the early 80s, not very powerful ones, usually a little 8-bit controller with a few K of RAM, one had a dot-matrix screen, something like 12 characters by 4 lines.

    2. There are so many things wrong with your assessment.
      1) Layers. If you poured 5″ of conductive concrete and then capped it with 1″ of non-conductive all issues would go away. Concrete is an excellent thermal conductor so it would still work.
      2) Your steel bar wouldn’t matter. Heaters are typically fairly low resistance devices. If they are high resistance, you don’t get any power flow(P=v^2/r). We are also talking about some reasonably good distances(feet, not inches). Your steel bar would only increase the resistance marginally and get a little hotter than the surrounding concrete.
      3) You don’t need big wires. This thing is powering concrete. Pretty sure you can just use some steel I-beams incased in concrete. What is the worst that will happen? They get too hot?

      1. >1) Layers. If you poured 5″ of conductive concrete and then capped it with 1″ of non-conductive
        > all issues would go away. Concrete is an excellent thermal conductor so it would still work.

        And then, one winter, there’s just too much snow or ice to melt. So someone decides to throw salt on it. The salt water seeps in to the crack beyond the layer. Someone steps onto the concrete, touches something with his bare hand, and gets electrocuted.

        If it was all so simple, someone would have already done it.

        Besides, heat cycles crack concrete. So the 1″ layer only solves all issue for a short time.

        Also, a 1″ layer on top of another layer of concrete cannot be used for a runway. The wheels of the airplane would tear that top layer straight off the lower laywers, when it brakes.

    1. Ya, likely it will be in a sandwich between a high strength structural layer on the bottom and a high wear resistance layer on the top. Otherwise the rebaring would short the resistive layer out I would guess.

  2. And all the shocked Earthworms will emerge and mix with the standing water on the concrete only to be frozen into an large Earthworm Popsicle when this thing pops a fuse. Brilliant.

  3. What impact do the steel shavings have once they change state from steel shavings into various iron oxides? There’s a lot of surface area there and it’s a corrosive and wet environment, made even more corrosive (?) by adding current through this whole thing. Unless it’s somehow protective? I am not a steel shavings inside concrete under corrosive conditions expert but this feels like a system out of a stable equilibrium.

    1. What happens to the steel shavings is largely a function of just what steel alloy is being used in this application. We should not assume that it is necessarily mild varieties.

        1. Who know/ There are an endless number of CRESS type alloys out there and a lot of steels formulated to work locked in concrete. Corrosion is an issue that would need to be considered during design, but it is not an impossible one to deal with. My guess it’s likely a high nickel variety in this case.

          1. Unlikely to be Nichrome as it is not a steel per se and the resistance would be rather high I would think for a given x-section but until we are told it’s a guess.

    2. Given that this is designed to have electricity passed through it any corrosion should be significantly slowed and easily reversed by an appropriate DC cycle. You’d want AC during normal operations to avoid this exact scenario.

  4. Regarding shock risk. . .doesn’t the electricity want to go to ground? and if that is true, it’s kinda sitting on and in the ground, so why would the electricity try to zap a person?
    also, wouldn’t a person be electrically isolated from a complete circuit standing on it?
    I’m not really an electricity guy, so I might (probably?) greatly misunderstand what I think I’m deducing.

    1. It has to be both isolated from ground for it to work, and insulated from anything on the surface to be safe. If it were not covered and, say a wet ground crewman was standing on it, then he would conceivably become part of the circuit given that he is less resistant than the material itself, and assuming his feet span towards the opposing conduction points then he would be in for all manner of unpleasantries round the crotch region. In the picture the water doesn’t appear to be boiling off so it’s probably safe to assume it’s covered with a nonconductive layer. It’s highly unlikely that it would be more conductive than the water itself as that would make for a little too short lived and spectacular snow removal solution. Resistive heating is the key word here, the energy has to put some effort into getting from A to B, creating heat along the way.
      And, seeing as the by product of it’s intended use is lots of water, and in most use scenarios you also wouldn’t want to create lots of steam which would reduce visibility, I’d say well insulated and low steady heat would be the desired effect.
      Now I’m imagining a horrible engineering fail where hundreds of barefoot passengers escaping a burning plane slide down the inflatable chute only to become human popcorn on contact with the tarmac!

        1. because you don’t want voltage potentials between the ground around it and the concrete. isolating it from ground (electrically) means that if you have one foot on wet earth and another on the concrete (lets assume there’s a fault and a non negligible amount of voltage is exposed) nothing should happen.

    1. Probably not much for radio, it’s all low-frequency. Compasses, nearby it might well throw some off, yep. I suppose you could design the circuit so the fields cancel each other out over the greater range. As long as Christopher Columbus still has his GPS, shouldn’t be a worse problem. I don’t think airliners rely on magnetic compasses as their primary equipment, especially on takeoff or landing, where there’s guide radio beams and stuff.

  5. Normal concrete also happens to be fairly conductive on it’s own, in fact conductive enough that rebar in a concrete pad is a recommended method for grounding (Ufer Ground). The real question is how much power do you need to dissipate to make reasonable progress in melting snow.

    I wonder if a passive system using a well and closed water loop could transfer enough heat to melt significant snow?

          1. Everyone bitching about the energy costs of heated pavement have to understand that in places where snow doesn’t melt in a few days after it falls, there is a significant energy cost in removing it that is largely spent in burning diesel fuel (and natural gas if it is disposed at melting stations for sewer disposal.) Before engaging in these bouts of high moral indignation please show the calculations proving that local heating uses significantly more energy and produce significantly more greenhouse gas than physical removal. And remember energy use for equipment transits must be included.

          2. thanks for helping me “understand,” maybe before you execute any more scintillating armchair analysis you might consider some of us have backgrounds in energy infrastructures in cold climates. Stick to your runways and dog kennels where cost is not the motivating factor.

          3. Obviously an electric heater is around 100% efficient, since heat is the “waste” in most energy conversions. But generating the power, and distributing it, is something like 50%. For non-renewables, you start off with heat in the first place, then turbines, grid, etc. If you’re gonna burn fuel to power a heater, the best way is to burn it at point of delivery, rather than the other end of the country in a power station.

            Low-temperature heat is the most useless form of energy, as far as converting it into other forms to do work. Electricity is the best. If they’re gonna do it, and at least an airport makes more sense than a driveway, they’d be much better with on-site boilers. Or even better, a heat pump drawing from the ground. Using the most efficient system could make huge differences in energy consumption.

          1. Unfortunately not so or ice storms wouldn’t take down the lines. While there are indeed ohmic losses in transmission lines the very high voltages used minimise this such that the actual amperage being carried is surprisingly small. Reactive losses on the other hand, can be significant if not actively controlled for.

      1. Parts of Canada, like Quebec with lots of hydro power to waste. Not Canada overall. Not Even Close. Resistance heat still doesn’t make sense compared to cold weather heat pumps

        1. In your own links it shows electric heating as second to natural gas. In fact, electric heating is used in 39% of households as the heating source to natural gas’ 50%.

          As you can see from the above link it’s not just that hydro power is plentiful in QC giving it a boost. The reason that natural gas is the largest heating source in Canada is that it is abundant and cheap in some of the most populace provinces. In fact, in the prairies it’s not uncommon to see automobiles powered by natural gas because it’s so cheap, when you’re on the east coast you never see that because financially it’s not worth it.

    1. GSHP work on the principle that the ground temperature is constant year round. So in the winter when the snow is present the ground temperature is perhaps 8-10C. Now looking back to the article the other day with PEX loops using a gas boiler. Great way to melt snow relatively quickly, but noting that there is cracking problems doing it too quickly.
      GHSP seems logical, perhaps even without a heat exchanger/compressor cycle depending on the mass of concrete and size of ground loop installation and the speed at which it needs to melt the snow/ice. And the location & lowest temperatures that occur.
      During the summer, one can also feed the heat from the concrete back into the ground with the same system which (some studies suggest) will raise the year round ground temperature a little and over time make the system more efficient.

    2. What you mean is geothermal energy. That is, indeed, used to heat sidewalks in Reykjavik (capital of Iceland). Question is if it’s available in Omaha. There are several reasons why you can’t use it everywhere you want.

      1. Iceland use geothermal for everything, they barely burn anything there. They have outdoor heated pools. Of course being a volcanic island makes that easy.

        As far as Omaha goes, you can always dig deeper. But the best idea is a heat pump. Some residences use this system. It’s a lot like an air conditioner in reverse, you suck heat out of the ground, concentrate it, and use it to heat whatever you want heating. Even if the ground’s freezing, that’s still 270 degrees above absolute zero. It takes work, a compressor, to concentrate it, but you get many times more watts in heat than you use to run the machine. It’s something like 700% efficient, in terms of energy in to heat out.

    1. No, absolutely.

      To the melted snow refreeze, the concrete temperature should be bellow freezing. Being bellow freezing, the snow would not melt, so no water. The concrete temperature being above freezing will melt the snow AND prevent it from refreezing.

        1. Yes, but those radiation losses also affect the snow.
          So if you couldn’t prevent the ice from forming you also couldn’t melt the snow since you have to pump enough heat into it to account for the same radiation losses in order for it to melt in the first place.

    2. Where heated pavement is currently used (regardless of the way it is heated) to control snow accumulation, one runs the system while the snow is falling and consequently it both melts on contact and re-evaporates almost instantly, thus no water accumulates. They allowed some snow to accumulate in the above demo for dramatic effect.

    3. The fact that the test pad shown in the article isn’t sloped is misleading. If your concrete contractor is any good, he’ll make sure there’s enough pitch in the concrete to allow rain and snowmelt to run off the surface of the road (or driveway or whatever). So when the snow melts into water, it should behave just like fallen rain.

      Of course if your concrete contractor is an idiot he’ll pitch the slope of the driveway to point at the one spot of your house with insufficient drainage and you’ve got a flooded basement (if you’re lucky) or a weakened foundation (if you’re unlucky).

  6. I have been wondering if you embed some rebar down below the frost line if it would use the thermal difference and conduct the heat from below to the surface. I know this article is referring to electrical conduction but I’m betting the properties extend to thermodynamics as well.

  7. In the long run using heat pipes to melt the show would be cheaper for larger applications. Electric underfloor heating is fine for bathrooms but on a larger scale it becomes a nightmare.

  8. How about using the runway and other tarmac surfaces as a heat sink for the airport and other buildings in the airport complex? The excess heat generated by the building itself (since most spaces of that size are cooled year round with heat applied at strategic locations) could melt the ice without the need for an elaborate and high energy system. The voltage needed to affect a change on a scale the size of an airport tarmac would be astronomical.

    1. Well it would be done by zone heating and again the total amount of heat required to maintain the minimum temperature differential needed to prevent accumulation during a snow event (which is how these systems are used) is just not as high as some imagine.

    2. I saw a different article about this same tech which said that runways have room for plows, but airports are interested in this for areas around gates that are already crowded with baggage and fuel vehicles.

  9. Wait a sec, maybe I could get a similar effect by hooking up the output of an arc welder to the reinforcing steel in my slab….

    Hold my beer and pass me the jumper cables…..

    1. I was thinking the same thing, why add a conductor to the concrete mix when it is already reinforced with steel rebar? The rebar will be more conductive of course, but over the length of a runway I’d think the resistance should be high enough that the current wouldn’t be too obscene compared to the rest of the airport.

      1. For long stretches of highway, runways and other paved areas that can be cleared with a plow and the snow left to the side until Spring, this sort of system is not cost effective. However there are areas in high snowfall zones where physical removal is difficult and expensive, or the traffic is high and continuous, like in high density urban areas where this is the better solution and likely the more energy efficient one.

        1. Obviously it wouldn’t be practical for a highway, but for a runway the traffic is “high and continuous”. And unlike on a road, a runway can’t be used when it is being plowed and delays cause big problems.

          1. For years I worked at an aircraft maintenance facility at a major airport in a high snowfall area. Believe me they had it down to a science/performance art and the runways were kept in continuous operation in all but the most severe storms and even then, most of the delays were caused by conditions at other airports. Busy runways are also somewhat like busy highways in that the air movements created by the traffic itself helps keep the pavement clear to some extent. I can’t see heating the strips as a cost effective alternative to current practice.

  10. Cool project. I like the comments here, too.

    I had a similar idea I’ve been considering, but instead of using a conductive concrete mix (very cool, btw) I thought about experimenting by using a homemade Nickel Chromium “lath” layered within the concrete/mortar. My idea was actually based on heated tiles in my kitchen but could theoretically be applied to the sidewalk.

    For indoor tiling, I wonder if it would be feasible to use a NiCr lath on top of the concrete board/Hardiebacker before applying the thin-set base for the tile. I’d have to build a control board to regulate the current for the NiCr lath, and to rig it to a switch or two in the house…low voltage, high current (going to have to do some math base on how much NiCr wire is used.) I’ve considered rigging it to my Xbee/wireless thermostat so I can turn it on/off remotely, and of course to a manual toggle switch should Skynet go active.

    I’m a DIYer and home improvement hobbyist, not an engineer, so I appreciate your constructive feedback!

  11. Wait an organic fabric extraction method minute…..

    Didn’t the solar roadways kickstarter fix this problem? Where are they solar roadways that can be used in airport tarmacs? That idea will always work cause they said so in their video.

    (/sarcasm off)

  12. The energy requirements are huge. To run some simple back-of-the-envelope calcs:

    To melt a gram of water, that is, to turn if from ice at 0C to water at the same temperature, requires 80 Joules of energy, assuming perfect efficiency. If the ice starts out colder than 0C, or if you want the water to end up warmer than 0C, additional energy will be required, but the latent heat of fusion still dominates.

    Saying “80 Joules/gram” doesn’t give many people a good intuitive feel for how much energy we’re talking about. But since we’re probably comparing the cost of melting the ice to the cost of removing it by plowing, let’s figure out how much mechanical work we could do to that same ice, if we used the same energy to lift it up instead of to melt it.

    The energy in Joules required to lift that same gram of ice is given by mgh, where m the mass in kilograms, g is the gravitational constant (approx 9.8 m/s2), and h is the height in meters. 80 Joules will lift one gram of ice 8163 meters. If you prefer different units, that works out to about 8 km, 5mi, or 26,800 ft.

    True, snowplows don’t lift ice straight up, and they spend a lot of their energy moving themselves around. But there are also significant inefficiencies in using electricity to heat concrete to melt water. It’s not hard to think of refinements to be done to these numbers on both sides.

    But the fact is that, for the energy required to melt ice, you could theoretically lift it to something approaching an airliner’s cruise altitude, while a snowplow merely needs to shove it a few feet to the side of the road. That should give an intuitive feel for why plowing is more cost effective and energy efficient in most cases.

  13. Many people here are whining about the energy costs and whatnot. Well listen here.

    Does anyone actually know what it costs to send out the city plow trucks? How long it takes? Does anyone here realize the cost of the salts and other melting agents that are simply dumped on the ground and washed away by the next time it snows?

    Look. It’s good to be all “energy conscious” and whatnot, but clearing snow is simply of matter of public safety on the roads. There is little excuse for the city to spend tons of money on the infrastructure they currently use already. These concrete, while drawing lots of power, would probably cost something similar to the huge costs already incurred, but with the added benefit of reduced time till clean streets. When it’s winter time, and our contracting work is slowed for the Winter, the city is happy to pay us $80 an hour to bolt on our scoops and push snow around all night. I don’t know how many other people are out there doing this, but in terms of energy use, it’s at least on par with this electrical power discharge of a few major city roads.

    I think people are just seeing this as a big waste of energy, because they’re seeing the electrical power go straight into the concrete, but they are forgetting about all the energy in other forms (often more expensive forms) that are used in the current snow clearing procedures.

    The only concern I would have is the extra load placed on a power station if an airport decides to melt a runway or two. There would need to be overhead arrangements made at the power plant, likely in the form of a contract.

    I do see this as a great way for Grandma to clear her porch for the mailman, when her grandson is still at work or class.

    What a great idea!

    1. Good post. I don’t think most of the folks here complaining about energy have the slightest idea what it costs to run a truck. I was the plow guy in a one-horse town way up north, so I have some insight here. The initial cost of the truck plus equipment is massive. Service on the truck is insane, and when things break, they aren’t cheap either. We used sand, not salt, so that wasn’t all that bad, but it adds up fast. Don’t forget you pay a person to run that truck, and you pay for a building to store it. Who knows how many trucks these airports in Canada have, dozens, maybe even hundreds of them. In places where there is a lot of snow that doesn’t melt off all year, it is a massive undertaking to keep the roads clear. I think people take it for granted that you throw fuel in a truck and off you go. In the long run, this electric concrete system may cost more financially, but in terms of environmental impact, I’m willing to be it ends up being more efficient in places that get serious snow. I believe Canada has a lot of hydro plants making power, can’t beat that for low environmental impact.

  14. What I want to know is where the water goes? When the snow melts, where is the water going? If it can’t drain, that means you have to keep the concrete heated the entire winter. What kind of cost is that going to cause? And if it drains, where to? Are we dumping water that will now freeze in storm drains? Bushes? What effect will that have?

    1. Where systems that heat pavement to control snow accumulation are used (and they are indeed used in many places) the heat is applied during snowfall and, largely because there is a HUGE amount of latent heat in snow, just keeping the surface a degree or two above freezing will cause the snow to melt on contact and evaporate. As for the storm drains many cities that get a lot of snow operate gas-fired melting stations that empty into the sewers without any re-freezing issues.

      Look folks, I live in a big (2 million plus) city that has to deal with major snow falls and accumulations that can hang around from November until April regularly. Not only that but because parts of the town have been around for over three hundred years and the streets are narrow it can’t just be shoved to one side while we wait for Spring. We use snow dumps that are miles away from where the snow fell, we use melting stations, and we use heated pavement where appropriate, and all three are standard methods, practiced in every major urban area, all over the world where snow is a seasonal issue. This is not a simple matter, with simple solutions that you can dismiss with simple reasoning.

  15. Conductive concrete is used with buried radial conductors as a ground enhancement material at cell towers and other telecom or substations to lower the earth system effective resistance. It does this by increasing soil contact over what is possible with just the bare conductor. It also keeps the copper wire in the ground by preventing it from being pulled out by copper thieves. See Erico GEM or SAE Conducrete. Keep this in mind If the heating experiments don’t pan out.

    1. Standard concrete is naturally conductive, look up “Ufer Ground”. You might need a better sort of concrete to avoid corroding the copper wire though. I have known people put in galvanized steel wire as an antenna ground plane – because of copper thieves – surprisingly, it worked well both as a ground plane and as a thief deterrent.

  16. Found a little more on this: , and . The UNL article isn’t deep but has more information about intended use. The paper’s abstract offers a little more on what’s in the mixture, and—though dated—may still serve as a basis for Tuan’s current work.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.