An estimated 365,000 tons of road salt is applied in the Twin Cities Metropolitan Area (TCMA) each year. The chloride in road salt flows into our lakes, streams, and groundwater, potentially harming our environment.
Several different types of deicing chemicals exist. Those covered in this section include chloride-based deicers, acetate-based deicers, and carbohydrates. A list of the chemicals approved for use by the MNDOT can be found here. This article summarizes environmental effects of de-icing chemicals. Other effects (e.g. on infrastructure) are discussed elsewhere in this manual.
The chloride-based deicers discussed in this section are sodium chloride (NaCl), magnesium chloride (MgCl2), and calcium chloride (CaCl2). Deicers can enter into the environment during storage, transport, and application. The distribution of the deicer is a complex process, an overview of which is provided in the figure to the right. When chloride deicers are dissolved in runoff, the anion and cation dissociate. The following section separately describes the environmental effects of anions (i.e. chloride) and cations (i.e. sodium, calcium, or magnesium).
The chloride component of chloride-based deicers does not easily precipitate, is not biodegradable, is not readily involved in biological process, and does not adsorb significantly to mineral/soil surface (Levelton Consultants Ltd., 2008). As such, chloride is highly mobile and can impact the soil, vegetation, groundwater, surface water, and air. Stefan et al. (2008) found around 30 percent of the salt applied to the roads in the TCMA makes its way to the Mississippi River. This suggests that the remaining 70 percent is either blown away, transported into the ground water, or stays within the soil, lakes, or wetlands.
Deicers reach the soil via runoff, splashing, spraying, or plowing. In general, chloride concentrations are the greatest within 2 to 3 meters from the road edge (Berthouex and Prior, 1968). Others, such as Norrstrom and Bergstedt (2001) have found salts as far out as 10 meters from the road edge, with the highest concentration within 6 meters. The distance that the salts will be transported through the soils depends on subsurface conditions. Long-term accumulation of chloride can result in reduced soil permeability and fertility, as well as increased soil alkalinity and density. As a result, there could be negative effects on the chemical properties of the soil and its ability to retain water, both of which are important to plant growth and erosion control (National Research Council, 1991). Another adverse effect of chloride in soil is its potential to release the metals sorbed to the soil particles (National Research Council, 1991; Amrhein et al., 1992; Backstrom et al., 2004).
Roadside vegetation can be negatively impacted by absorption of chloride through the plant roots, or from accumulating on the foliage and branches. The symptoms associated with salt impacts are similar to those of a drought; stunted growth, brown and falling leaves/needles, dying limbs, and premature plant depths (National Research Council, 1991). The image to the right shows the browning of needles due to elevated salt levels. The effect of chloride on plants has been seen at distances of100 to 650 feet off the road (Fischel, 2001). The level of chloride that must be reached before the plant is harmed depends on the type of vegetation. Developers and planners often use salt tolerant vegetation near the road’s edge to lessen the impact of salt.
Since chloride does not bind to soils, chlorides that enter the subsurface with infiltrating water may reach the groundwater table. Howard and Haynes (1993) found that 55 percent of salt applied to a catchment in Toronto enters temporary storage in shallow sub-surface waters. Cusack (n.d.) estimated that approximately 45 percent of chlorides applied as road salt will be carried to the groundwater. Chloride entering groundwater systems is likely to persist for a long time since there is no significant removal mechanism and groundwater moves slowly.
Chloride is naturally present in Minnesota due to weathering of geologic materials. In urban areas, most of the chloride found in shallow groundwater likely comes from the use of deicing salts. Kroening and Ferrey (2013) reported on the conditions of Minnesota’s groundwater from 2007 to 2011. Sand and gravel aquifers in the TCMA had chloride concentrations that ranged from less than 1 milligrams per liter to 8,900 milligrams per liter, with a median concentration of 86 mg/L. Approximately 27 percent of the monitoring wells located in the TMCA sand and gravel aquifers were above the Secondary Maximum Contaminant Level (SMCL) of 250 milligrams per liter. The SMCL is based on aesthetic concerns, specifically taste. Statewide the median chloride levels in sand and gravel aquifers was 17 milligrams per liter, with only 1 percent of the monitoring wells showing chloride levels above the SMCL. In bedrock aquifers, chloride concentrations ranged from less than 0.5 milligrams per liter to 680 milligrams per liter, but in general did not exceed the secondary SMCL of 250 milligrams per liter. Chloride concentrations were highest in shallow groundwater, typically at depths of 30 feet or less below the ground surface.
Chloride concentrations in surface waters tend to follow a seasonal distribution. Concentrations usually increase in the winter and decrease in the summer (Novotny et al., 2007). The chronic chloride pollution standard has been set at 230 mg/L and the acute standard at 860 mg/L by the MPCA in Minnesota Rules Chapters 7050 and 7052. These limits are based on the findings that chronic concentrations of 230 mg/L are harmful to aquatic life, while concentrations above the acute standards are lethal and sub-lethal to aquatic plants and invertebrates. Stefan et al. (2008) reported that to date there are no documented exceedances of the acute standard in the Twin Cities. However, there are 21 lakes, 22 streams, and 4 wetlands that are impaired for chloride.
Salt-containing water has a higher density than non-salt-containing water and will sink to the bottom of the water body. This can result in chemical stratification and disrupt the lake mixing patterns (New Hampshire Department of Environmental Service, N.D.; Novotny et al., 2007). Effects on surface waters may be minimized by the dilution of the deicers as they are transported to the surface water. Dilutions of 1:100 to 1:500 are estimated to mitigate negative impacts of the deicer (Fischel, 2001). Small ponds and slow streams are estimated to be most impacted by deicers because the likelihood of dilution and dispersion is lower in those environments (Fischel, 2001).
A small percentage of the total applied chloride is evidenced to be transported by air. Blomqvist and Johansson (1999) found some deicing road salts can be transported by air 40m from the application site. Kelsey and Hootman (1992) found that sodium chloride was detected at a height of 49 feet (15 meters) within 220 feet (67 meters) of the highway. Kelsey and Hootman (1992) also found evidence of a positive correlation between plume height, and the travel distance of the constituent. The Connecticut DOT found road salt powder could travel as far as 300 feet laterally under heavy traffic conditions. Chloride transported by air can affect soil and surface/groundwater, but deposition on the vegetation is the primary concern (Levelton Consultants Ltd., 2008).
The cation components of chloride-based deicers (i.e. sodium, magnesium, and calcium) can also impact the environment. Sodium ions can change the structure of soil, causing a decrease in permeability, and infiltration (Davis et al., 2012). Sodium can also reduce the amount of calcium, magnesium, and other nutrients in the soil by raising the alkalinity of the soil and reducing the ion exchange capacity (National Research Council, 1991). Magnesium and calcium can improve soil structure by causing soil particles (particularly clays) to form aggregates, resulting in improved drainage (Amrhein, and Strong, 1990). The presence of chloride, magnesium, and calcium may also result in the mobilization of heavy metals sorbed to soil particles (Amrhein, and Strong, 1990; Backstrom et al., 2003)
Sodium, magnesium, and chloride in surface and groundwater can affect the hardness of water. The hardness of water will be reduced if there are elevated levels of sodium and will be increased if there are elevated levels of calcium and magnesium (Cheng and Guthrie, 1998). An increase in water hardness has shown evidence of decreasing the toxicity of heavy metals (Lewis, 1997).
In order to reduce the corrosive effects of some of the chloride-based deicers, corrosion inhibitors can be added. Corrosion inhibitors can include heavy metals, inorganic ions, and organic substances (Levelton Consultants Ltd., 2008). The toxicity and environmental effects of corrosion inhibitors vary greatly and are dependent on the composition (Pilgrim, 2013). In general, the inhibitors that contain organic components consume oxygen during decay. The oxygen consumption can lead to anoxic conditions in the soil, groundwater, or surface water (Fischel, 2001). At colder temperatures, the rate of decomposition will decrease and there will be an increased potential for the inhibitors to reach the groundwater (Cheng and Guthrie, 1998).
Much of the information on the environmental impacts of acetate-based deicers is based on studies regarding calcium-magnesium acetate (CMA) Therefore, much of the information presented in this section is related specifically to CMA. Modeling studies have estimated that the concentrations of CMA in the runoff from highways is between 10 and 100 mg/L, with a maximum concentration of 5,000 ppm. The typical annual mass loading is estimated to be 10 tons/linear-mile (Horner, 1988). Despite high mass loading, runoff and receiving water are predicted to dilute the concentration.
The characteristics of acetate suggest it would be absorbed to the soil surface and not carried away with the runoff. Once in infiltrating water, acetate can be mobile, however Horner (1988) found that less than 10 percent of the acetate applied to test plots were found in the underlying soil and groundwater. The sodium and potassium contained in other types of acetates are less likely to adsorb to the soils and therefore have a greater potential to leach into groundwater (Cheng and Guthrie, 1998).
Horner (1988) did not note any effects of acetate on soil plasticity, moisture-density characteristics, unconfined compression strength, or shear strengths in medium texture soils. An increase in permeability was noted. In the Horner (1988) study, the test sites that received an addition of CMA were found to have an increase in permeability up to 20 times more than that of the control plots. There is uncertainty about CMA causing the release of metals from soil (Amrhein et al., 1992; Horner, 1992; Granato et al., 1995; Levelton Consultants Ltd., 2008; McFarland and O’Reilly, 1992). Another concern is that acetate-based deicers consume oxygen when degrading.
Acetate-based deicers dissociate when in water. The metal ion persists, but the acetate ion will degrade (Fortin et al, 2014). Degradation of the acetate ion consumes oxygen, which is one of the biggest environmental concerns associated with the use of acetate-based deicers. At temperatures between 10°C and 20°C, the biological oxygen demand (BOD) was fully applied within 5 to 10 days of the acetate being deposited into the water. At a water temperature of 2°C decomposition took 100 days (Horner, 1992).
Modeling studies have predicted CMA concentrations in the highway runoff range from 10 to 100 ppm, with a maximum concentration of 5,000 ppm. Evidence has shown that at a concentration of 100 ppm and a temperature of 20°C, CMA will completely deplete the oxygen in the water. At concentrations of 10 ppm the dissolved oxygen in ponds was reduced by approximately 50% (Brenner and Horner, 1992).
Dispersion and dilution are likely to mitigate the negative effects of CMA, and the environments most likely to be severely affected are slow moving streams and small ponds (Fischel, 2001). The potential for mitigation through dispersion and dilution is confirmed by two studies of CMA and BOD. McFarland and O’Reilly (1992) found that CMA did not negatively impact the dissolved oxygen (DO) levels in the surface waters in most of the scenarios that were tested. A study on Bear Creek in Clackamas County, Oregon, did not find a correlation between CMA concentrations and BOD (Tanner and Wood, 2000).
Carbohydrate-based deicers are often made from the fermentation of grains or the processing of sugars such as cane or beet sugar (Rubin et al., 2010). Small quantities of carbohydrates are sometimes used with other deicers. Alone carbohydrates do not aid in melting ice or snow; however, their use can help reduce the freezing point of ice further than salt and can help salt stick better to the road surface (Fortin et al, 2014; Rhodan and Sanburn, 2014). Carbohydrates are not corrosive to steel, and at high concentrations, carbohydrates can act as a corrosion inhibitor for salt brines.
There is evidence that the use of carbohydrates in the United States is increasing. For example, sales of a beet based product called Beet Heet were around 900,000 gallons at the end of the winter season in 2013. By February of 2014, 1.5 million gallons of Beet Heet had been sold. The Morton Arboretum in Lisle, Il uses beet juice in their deicers. The beet juice additive has minimal environmental affects, and helps the salt stick where applied. With the addition of beet juice, the arboretum is using nine times less salt, and saving an estimated $14,000 in material costs (The Morton Arboretum, 2014). Another unconventional additive that has been used is cheese brine. Wisconsin has used a cheese brine in at least six counties in the state (Rhodan and Sanburn, 2014).
Fu et al. (2012) looked at two beet molasses-based deicers in comparison with a salt brine deicer. When used as a prewetting material, there was no statistically significant difference between any of the chemicals. When used as an anti-icing material, the organic material performed 30% better.
The decay of the alternative additives in the environment will contribute to BOD (most specifically for the organic additives). Depending on the nature of the unconventional additive, nutrients could be released during the decay process which could be a potential source of pollution (Fortin et al, 2014). Brenner and Horner (2012) compared the BOD requirements of a corn based CMA and a reagent based CMA. The corn based CMA had a higher BOD than the reagent based CMA.
Sodium Ferrocyanide and Ferric Ferrocyanide have been used as anti-caking additives for deicing (CTC and Associates LLC, 2004). Cyanide is harmful to the environment if it is leached into groundwater or carried to surface water. An overview of 13 deicing products found a cyanide range from less than 0.0003 parts per million (ppm) to 0.33 ppm (Fischel, 2001). Issues related to Cyanide in groundwater are contained in the infiltration section of the MN Stormwater Manual. Other chemicals that have been found at trace levels in deicers are arsenic, lead, and mercury (Dindorf, 2008).
There are many chemicals associated with deicing that have both similar and unique properties and environmental effects. The following table summarizes the corrosion and environmental impacts of the deicing agents described in this article. Care should be taken when determining which chemicals are best for the intended application and for the environment surrounding the application area.
|Category||Type||Lowest Practical Melting Pavement Temperature 1,2||Potential for corrosion impairment3||Environmental Impact3|
|Atmospheric Corrosion to Metals||Concrete Matrix||Concrete Reinforcing||Water Quality/Aquatic Live||Air Quality||Soils||Vegetation|
|Chloride Based Deicers||Sodium Chloride||15°F||High; will initiate and accelerate corrosion||Low/moderate; Will exacerbate scaling; low risk of paste attack||High: Will initiate corrosion of rebar||Moderate: Excessive chloride loading/metals contaminants; ferrocyanide additives||Low: Leads to reduced abrasives use||Moderate/High: Sodium accumulation breaks down soil structure and decreases permeability and soil stability; potential for metals to mobilize||High: Spray causes foliage damage; osmotic stress harms roots, chloride toxicosis|
|Calcium Chloride||-20°F||High; Will initiate and accelerate corrosion; higher potential for corrosion related to hydroscopic properties||Low/moderate; Will exacerbate scaling; low risk of paste attack||High: Will initiate corrosion of rebar||Moderate: Excessive chloride loading; heavy metal contamination||Low: Leads to reduced abrasives use||Low/Moderate: Improves soil structure; increases permeability; potential for metals to mobilize||High: Spray causes foliage damage; osmotic stress harms roots, chloride toxicosis|
|Magnesium Chloride||-10°F||High; Will initiate and accelerate corrosion; higher potential for corrosion related to hydroscopic properties||Moderate/high: Will exacerbate scaling; risk of paste deterioration from magnesium||High: Will initiate corrosion of rebar, evidence suggest MgCl2 has the highest potential for corrosion of chloride produces||Moderate: Excessive chloride loading; heavy metal contamination||Low: Leads to reduced abrasives||Low/Moderate: Improves soil structure; increases permeability; potential for metals to mobilize||High: Spray causes foliage damage; osmotic stress harms roots, chloride toxicosis|
|Acetate Based Deicers||Calcium Magnesium Acetate||20°F||Low/moderate; Potential to initiate and accelerate corrosion due to elevated conductivity||Moderate/high: Will exacerbate scaling; risk of pate deterioration from magnesium reactions||Low; probably little or no effect||High: Organic content leading to oxygen demand||Low: Leads to reduced abrasives use||Low/Moderate: Improves soil structure; increases permeability; potential for metals to mobilize||Low: Little or no adverse effect; osmotic stress at high levels|
|Potassium Acetate||-15°F||Low/moderate; Potential to initiate and accelerate corrosion due to elevated conductivity||Low; probably little or no effect||High: Organic content leading to oxygen demand||Low: Leads to reduced abrasives use|
|Carbohydrates||Beet Juice||NA||Low; Potential to initiate and accelerate corrosion due to elevated conductivity clams of mitigation of corrosion require further evaluation||Low; Probably little or no effect||Low; Probably little or no effect; claims of mitigation of corrosion require further evaluation||High Organic matter leading to oxygen demand; nutrient enrichment by phosphorus and nitrogen; heavy metals||Low: Leads to reduced abrasive use||Low: Probably little or no effect; limited information available||Low: Probably little or no effect|