How does road salt work?

Even though salt may be applied dry it does not begin its snowfighting job until it dissolves into brine. A chemist would explain the process in terms of colligative properties. The brine is a solute and the concentration of grains in the solute (in this case, salt brine) determines its freeze-point lowering potential. Any substance that dissolves in water has this effect, but each substance will have varying outcomes. While sugar or molasses can be a solute and lower water’s freezing temperature, for example, salt’s lower molecular weight gives it almost six times the effectiveness of sugar in lowering the freezing point of water – actually even more in this example since sugar isn’t an electrolyte at all. This is the same principle you use when you put antifreeze into your car’s radiator.

Salt applied as a liquid or prewet solid can begin to act immediately lowering the freezing point of water. On a pavement where the temperature is 30°F (-1° C), one pound of salt melts 46.3 pounds of ice. One inch of ice on one lane-mile of road would weigh 70 tons. To melt that much ice would take 17 tons of salt. But the objective is not to melt the snow and ice off the pavement, only to prevent or destroy the bond on the surface of the roadway between the pavement and the ice or snow. In our example lane-mile with an inch of ice, most road agencies would use 500 pounds or less, less than 2% of the amount of salt needed to melt the ice.

The objective being to prevent the bond if possible (not melt all the ice), liquids are appropriate when applied in a pre-storm anti-icing application to be in place before freezing precipitation arrives. It also explains why agencies use larger particles for application of dry salt to ice- and snowpack-covered roads since they need to have the weight and mass to bore down to the pavement where the real work is done.

The concentration of the brine and the temperature of the pavement are the key variables determining whether and how fast the salt will act. When salt dissolves in water, the resulting brine is generated at the saturation level, 25-26%, the same level as the salt crystals form in a solar saltworks. But the brine is quickly diluted by the snow or ice it contacts. As dilution proceeds, there is less salt to depress the water’s freezing point, so the freezing point will rise, assuming the temperature is unchanged. If the temperature falls, the loss of melting power accelerates. That is why intense storms may require multiple applications, to keep the brine concentration from become too dilute to do its work.

The roadway temperature is critical in choosing a deicer with temperatures below 15 degrees Fahrenheit generally requiring addition of other "hotter" deicers to salt (e.g. magnesium chloride, calcium chloride). The phase diagram illustrates the required concentration of salt to keep the brine liquid at various temperatures.

A more detailed explanation

A number of researchers have conducted experiments on particle penetration and ice disbondment characteristics of various deicers. Variables include deicer unit weight, particle size, atmospheric conditions, substrate temperature, etc.

Reaction times range from a few minutes to several tens of minutes. Data from experiments indicate the initial ice-melting reaction begins almost immediately upon application, depending on the deicer’s eutectic and the temperature of the ice, pavement, and air. Undercutting at the ice-pavement interface takes longer because the particles must penetrate the snow-ice mass to reach the pavement interface. Lower temperatures flatten the curve. For example, Nixon, et al. compared deicer penetration and ice undercutting capabilities of NaCl wetted with CaCl2 and untreated NaCl. The authors showed that NaCl penetrated ice immediately and undercut the ice in 5-6 minutes at a temperature of 10 ºF (–12.2 ºC). Their work indicates that NaCl wetted with CaCl2 accelerates penetration reaction time for the first few minutes after contact. The curves become parallel within approximately ten minutes.

McElroy, et al. (1988) compared ice undercutting rates and deicer application rates for seven deicers and mixtures of deicers. Variables include temperature, time, and application rate. The authors presented laboratory data showing that CaCl2 begins undercutting a 1/8 in layer of ice in 7-8 min at 15 ºF. Similarly, NaCl begins undercutting in 17-18 min; and KCl and urea begin undercutting in >60 min. Dickinson (1959) compared amounts of ice melted by mixtures of calcium and sodium chloride at different temperatures and time intervals. He showed that CaCl2 melted 2.5 lb ice/lb of deicer and NaCl melted 1.6 lb ice/lb in 15 min at 20 ºF. Dickinson’s Table 2 includes data for temperatures from 0 ºF to 26 ºF and time intervals of 15 min-6 hr. These data indicate that ice melting begins almost immediately. Sinke et al. (1976) Compared ice undercutting rates for NaCl and CaCl2 at various temperatures. They showed reaction times for NaCl of approximately 5-8 min at 15-25 ºF and reaction times for CaCl2 of 2-5 min under the same conditions.

Kaufmann (1960), referring to B. C. Tiney (1934), showed comparative melting capacities of calcium and sodium chlorides at various t temperatures. NaCl melted quantities of ice varying from 3.2 lb at -6.5 ºF (eutectic) to 46.3 lb at 30 ºF. 77%-80% CaCl2 melted 3.7 lb of ice at -6.5 ºF and 31.1 lb of ice at 30 ºF. Kaufmann also reported on penetration time in minutes of grains of NaCl in ice, based on grain size and temperature. A 1/8” salt grain penetrated >1” in 20 min at 25 ºF and 1” in approximately 50 min at 8 ºF. Larger particles penetrated to a greater depth. A ¼” particle penetrated 2” in 50 minutes at 25 ºF and 2” in 120 min at 8 ºF.

References

Dickinson, William E. 1959. Ice-Melting Properties and Storage Characteristics of Chemical Mixtures for Winter Maintenance. Highway Research Board Bulletin 220. 38th Annual Meeting January 5-9.

Kaufmann, Dale W. 1960. Sodium Chloride, The Production and Properties of Salt and Brine. Reinhold Publishing Corporation. Chapter 23, pp 562-565

Kersten, M.S., L.P. Peterson, and A.J. Toddie, Jr. 1959. A Laboratory Study of Ice Removal by Various Chloride Salt Mixtures. Highway Research Board Bulletin 220. 38th Annual Meeting January 5-9.

McElroy, A.D., Robert R. Blackburn, and Henry Kirchner. 1990. Comparative Study of Chemical Deicers – Undercutting and Disbondment. Transportation Research Board Annual Meeting. January.

McElroy, A.D., Robert R. Blackburn, Jules Hagymassy, and Henry Kirchner. 1988. Comparative Study of Chemical Deicers. Transportation Research Record 1157.

Nixon, J.G., D.R. Larrimore, and E.H. Mossner. 1979. A Laboratory Comparison of Prewet and Untreated Rock Salt as Ice Removal Agents. Dow Chemical Company.

Sinke, G.C. and E.H. Mossner. 1976. Laboratory Comparison of Calcium Chloride and Rock Salt as Ice Removal Agents. Transportation Research Record 598. Maintenance Management, The Federal Role…

Tiney, B.C. 1934. Highway Research Board, Proc. 13:333

Trost, Susan E., Frank J. Heng, and E.L. Cussler. 1987. Chemistry of Deicing Roads: Breaking the Bond Between Ice and Road. Jour. Transp. Engrg. Vol. 113, No. 1. January 1.