<p>Out of the total 540 000 km Swedish road network, approximately 400 000 km represents gravel roads. According to data from the Swedish Road Administration (SRA), the length of the public road network administered by the government amounts to 100,000 km. About 20% of this network comprises of gravel roads. The traffic volume on these gravel roads is very small, i.e. less than 125 AADT but these roads can be of great regional importance and are often a direct prerequisite for the transportation of forest products; the largest export product of the country. These roads are also essential for the survival of sparsely populated regions.</p><p>An increased amount of particulate matter in the air could give rise to reduced traffic safety, clogging of watercourses, air pollution, damage to agriculture, and have negative health implications.</p><p>If the fines in the gravel wearing course disappear, this is likely to lead to the loss of other larger sieve fractions of the aggregate. The deficiency of gravel course material implicates more frequent maintenance, as well as increase demands of aggregate supplement, both of which will result in higher costs. Dust suppressants not only, and efficiently, reduce the amount of dust raised by traffic induced wear, but they also reduce overall gravel road maintenance costs (Kirchner, 1988). On the other hand, dust suppressants, especially chlorides, may also negatively affect the environment, and for example cause corrosion to cars or dehydration to road side vegetation.</p><p>Approximately 5%, or 370 MSEK, of the total annual SRA maintenance budget is assigned to these roads. However, there is also a significant expense associated with dust control. Dust binding is expensive, as it amounts to 30% of the total expenses for the maintenance of the gravel roads during the period when the ground is free from snow (Bergström and Grebacken, 1995). This expense is, of course, affected by the type of dust suppressant, the concentration of used dust suppressant, and the number of treatments. However, at present there is a real lack of knowledge concerning the efficiency, durability, life cycle cost, and impact on the environment from the different dust suppressants utilized.</p><p><b>Objective and Delimitations</b></p><p>The objective of this study is to evaluate dust suppressants and gravel wearing courses regarding dust control efficiency, longevity, leaching and penetration of dust suppressants, moisture retaining properties, and optimal aggregate size distribution. In addition, developing and evaluating adequate methods for analysing these properties are also within the purpose of the study. The methodology used includes literature reviews as well as laboratory and field investigations. The findings are expected to yield reduced life cycle costs for gravel roads and environmental benefits from reduced dust formation as well as reduce the exposure of chemicals to nature. The study is valid for the Nordic climate, design practices, gravel material, and maintenance tradition.</p><p>Still there are reasons to believe that the results are also valid for gravel roads in other countries when using these kinds of dust suppressants.</p><p><b>Literature study</b></p><p>Certain dust suppressants depend on absorbing moisture from the air to reduce dust. Others rely on adding cohesion to the aggregate material for the same purpose. Hygroscopic salts, such as calcium and magnesium chloride, use the former principle while lignosulphonates and bitumen emulsion primarily use the latter. Regardless of the mode of action, the result of dust suppressant treatment is temporary and therefore, at least, renewed treatment annually (U.S. Department of Transportation, 2001).</p><p>Calcium chloride is by tradition the most common used dust suppressant in Sweden, followed by, in turn, magnesium chloride, lignosulphonate, and bitumen emulsion. Treatment with either chloride or lignosulphonate has a duration of about six months, according to literature (Lohnes and Coree, 2002). Chloride is stated to be most effective with surface wearing courses with moderate content of fines whereas lignosulphonate works best with aggregates that have high content of fines (Gillies et al, 1999). Finally, bitumen emulsion (BE) is said to have a duration of approximately twelve months (Gillies et al, 1999) and to be most effective in combination with coarser grains, thus having less grain surface area to cover (Bergström and Grebacken, 1995).</p><p>A major disadvantage with chlorides is that they are water soluble and tend to migrate downwards through the roadway. They are easily washed away by rain, which may imply that more than one treatment per year is required (Foley et al, 1996). Another problem regarding chlorides is that they are corrosive to most metals.</p><p>Calcium chloride and magnesium chloride have several common qualities and work as dust suppressants on the same basis. However, up to 20 wt-% more magnesium chloride is needed to achieve the corresponding dust controlling effect as with calcium chloride (Reyier, 1972).</p><p>Lignosulphonate is also called sulfite lye or lignin. Lignin is a major constituent of wood, i.e. 33-40% of dry weight (McDougall, 1986). It works as nature’s own adhesive and binds the wood fibres together. Lignosulphonate is obtained as a waste product during the boiling of wood for paper pulp production.</p><p>There are several different types of lignosulphonates, i.e. calcium-, sodium-, and ammonium lignosulphonate, where calcium lignosulphonate represents the most commonly used (Roads and Transportation Association of Canada, 1987). Experiences from Canada indicate that sodium and calcium lignosulphonates are equally effective (Chichak, 1991).</p><p>Lignosulphonate creates a hard surface crust on the road but the product is highly water soluble and therefore tends to leach from the road during heavy rainfall, which implies that this dust suppressant is much more effective in a dry than in a wet environment (Jones, 1999).</p><p>It seems as though the grain size distribution is especially important when dust suppressing with lignosulphonate while BE and chlorides are not as sensitive to variations in grain sizes. The adherent effect of lignosulphonate is significantly lowered if the content of fines is too low. Clay also prevents the lignosulphonate from leaching during rainfall (Svensson, 1997). In addition, if the lignosulphonates, as well as the chlorides, are mixed into the wearing course aggregate, they are noted to show greater longevity (Hoover, 1981).</p><p>The amount of emitted dust from gravel roads is considered to be a function of several different factors. These include: number of vehicles, vehicle weight, vehicle speed, number of wheels per vehicle, grain size distribution of the wearing course material, restraint of the surface fines (compaction, cohesiveness/bonding, and durability), surface moisture (humidity, amount of precipitation, and amount of evaporation), and topography (Addo and Sanders; 1995; Bolander and Yamada, 1999). Findings in literature indicate that all dust control treatments may reduce dusting of up to 80% when compared to untreated, reference roads (Hoover, 1973). Failures are usually ascribed to the loss of dust suppressant during rainfall. However climatic conditions, wearing course aggregate, traffic, and dust suppressant concentration used will all affect the effectiveness and life-time of the treatment.</p><p>Examinations of road dust may be divided into two categories: laboratory screening tests, which are used to determine if a dust suppressant actually works for preset, controlled conditions or application rate needed; and field tests, which give realistic test conditions and often include static dust collectors or mobile, measuring devices.</p><p><b>Dust formation on gravel road test sections</b></p><p>With the objective of comparing the effectiveness of different dust suppressants and of studying the influence of altered concentrations as well as new combinations of dust suppressants, a field experiment was established on four different geographic locations in Sweden: Umeå, Rättvik, Hagfors, and Halmstad. At each location a stretch of road was further divided into a number of 1000m test sections marked by signs at both ends. Different dust suppressants were then applied to these 1000m sections. Magnesium chloride solution, calcium chloride solution, magnesium chloride flakes, calcium chloride flakes, lignosulphonate, and a solution of starch were tested on all test locations. In addition, bitumen emulsion and biomass were tested in Hagfors and rape oil in Halmstad. Mesa, a material that is used for increasing the amount of fines in the aggregate, was tested in combination with calcium chloride solution, magnesium chloride solution, and lignosulphonate, respectively.</p><p>For assessing the amount of dust emitted from the different test sections, TSI DustTrak Aerosol Monitor was used in combination with visual assessments. The DustTrak apparatus uses a 90º light scattering laser diode sensor for real-time determination of aerosol mass concentrations in the range of 0.001-100 mg m-3. The amount of light scattering is proportional to the mass concentration of the aerosol for particles ranging in size from 0.1 to 10 μm (Hitchins, 2000; Wu, 2002; Veranth et al, 2003).</p><p>For the DustTrak measurements, an estate car was used. The vehicle was driven at a constant speed of 40 km/h over the entire road length and so passing all test sections. The driving pattern was to the greatest extent possible restricted to the existing wheel tracks. A PDA connected to a GPS, was programmed to register time, current test section, and vehicle velocity. These data were then combined with the obtained DustTrak data.</p><p>The results obtained through this objective method were compared to results obtained from visual inspections made in accordance with the subjective assessment directives (Vägverket, 2005). This was done with the intention of evaluating the method for recording dust emissions and hence look into the possibility of developing an unbiased method to replace the existing subjective method for assessing dust emissions on a gravel road. Compared results expose a fairly good correlation.</p><p>The difference in aerosol particle mass concentration between a dust suppressed and an unbound, natural gravel road is remarkably large. Dust suppressants creating a hard surface crust, such as lignosulphonate, but also bitumen emulsion, performed acceptably well in the beginning stages but ruptured and became ineffective after a while. The lignosulphonate was one of the least effective of the products tested. However in Umeå, it was one of the most effective, if not the most effective. It was concluded, through an analysis by means of a scanning electron microscope (SEM) that the well performing lignosulphonate product used in Umeå was sodium lignosulphonate while the surface rupturing product used at all other test locations was calcium lignosulphonate. The sodium lignosulphonate had higher relative levels of sulphur, while the calcium lignosulphonate had higher relative levels of carbon and oxygen. This, together with the difference in calcium and sodium composition, could explain the differences in performance. Sections treated with starch also raised much dust as did the biomass product, probably due to fast leaching and degradation. In fact, the latter section rapidly deteriorated to such bad condition that it had to be excluded from the test and dust suppressed by conventional means. The rape oil product applied in Halmstad, formed a hard crust which was also subjected to surface rupture but which lasted almost throughout the whole summer and during this time performed very well. However the utilized concentration of the rape oil was rather high and this product is still expensive.</p><p>Solid calcium chloride is 10 to 40% or, on average from all observations, 19% more effective than magnesium chloride. This agrees well with an earlier estimation, indicating that calcium chloride is 18% more effective than magnesium chloride (Reyier, 1973). It is shown from all of the test locations that a solution of either magnesium chloride or calcium chloride is the most effective choice for dust abatement. This finding is thought to be due to a more homogenous distribution of chlorides, in combination with better penetration, for solutions than for flakes. Even though a somewhat higher chloride concentration can be obtained from purchasing a ready-to-use solution, it is probably more cost and energy effective to prepare the solution from a solid product on location, if possible out of practical reasons.</p><p>The addition of mesa, and hence fines, to the aggregate appeared to have no positive effect on the efficiency of chlorides. However the results are hard to interpret due to different initial concentrations, as the salt concentration was reduced on sections where mesa was used.</p><p><b>Laboratory methods for determination of dust suppressant residual content over time of the road test sections</b></p><p>The goal was to study how concentrations change with time. In order to do this we had to develop new methods to quantify residual chlorides and lignosulphonates.</p><p>Samples were taken of the wearing course gravel from a square measuring 150 x 150 x total thickness (usually on the order of 40 mm). The samples were taken from the same general area after different laps of time. In general the sampling site was the mid point of the section and in the right wheel track. The samples were given the same number as the section.</p><p>Chloride and lignosulphonate treated gravel road sections were analysed for decreasing residual content over time. Eight sections in total were analysed: 1.0 kg/m and 0.7 kg/m calcium chloride, 1.0 kg/m and 1.3 kg/m magnesium chloride, calcium chloride solution, magnesium chloride solution, lignosulphonate, and lignosulphonate in combination with mesa.</p><p>Deionised water measuring 50 mL, was added to 50 g of gravel wearing course material in a glass beaker. The dust suppressant residues were dissolved by submerging the beaker containing gravel into an ultrasound bath for 10 minutes. This solution was then filtered into a clean glass beaker. The chloride containing filtrate was put into Eppendorph tubes and centrifuged for further separation of fine mineral particles. Filtrate containing lignosulphonate was instead filtered once more with a PTFE-membrane, with the purpose of minimizing background disturbances from fine mineral particles.</p><p>Chloride content was analyzed calorimetrically using the Mercury (II) Thiocyanate Method. In this method, an orange-coloured complex of ferric thiocyanate is formed when chloride ions disperse the mercury ions of mercuric thiocyanate in the presence of excess ferric nitrate. The resulting colour was measured on an ion specific photometer.</p><p>Results indicate that the chloride content of gravel wearing courses treated with chloride containing dust suppressants dicreases rather fast during the summer months. It can be concluded that there is not enough chloride left in the road surface after the summer, such that yearly renewed application would be unnecessary. It is apparent that re-blading of the road impacts the chloride concentration on different spot locations within the individual test sections and re-blading was done frequently on some test sections and that will make the results more difficult to interpret and draw conclusions from.</p><p>It is also obvious that the chloride concentration in the gravel wearing course most often increases during some weeks after the dust control treatment. The reason for this is not yet understood or sufficiently investigated.</p><p>A theory to the increasing chloride concentration during some weeks after the point of application is that during wet weather, as initially in the season, the chlorides follows the water runoff streams downwards and hence cannot be collected when taking samples in the gravel wearing course. During periods of dry weather, the surface moisture of the road evaporates causing moisture from deeper below to travel up towards the road surface by means of capillary rise. This causes the chloride concentration to increase within the gravel wearing course since the upwards travelling water also carries the dissolved chlorides. This is supported by findings that the chloride content in the gravel wearing courses increases during a time period of less precipitation after the application activity and then decreases during periods of more heavy precipitation. However, this should be easy to investigate further by continuously collecting samples from different levels of depth within the road pavement.</p><p>The chloride residual content over time for the section treated with 1.3 kg/m solid magnesium chloride is approximately equivalent to the section treated with 1.0 kg/m solid calcium chloride and the residual content over time for the section treated with 1.0 kg/m solid magnesium chloride is approximately equivalent to the section treated with 0.7 kg/m solid calcium chloride. These observations are significant at the 5% level for at least two of the occasions of sample collection in Rättvik during 2006 and coincides with a stated theory that 30 wt-% more magnesium chloride than calcium chloride is required to achieve equal chloride concentration. The results also indicate that sections treated with chloride solutions have the lowest concentration of residuals. This might imply that a homogenous horizontal and vertical distribution of chlorides within the wearing course is equally important as a high concentration of chlorides.</p><p>A uv/vis-spectrophotometer was used to estimate the content of lignosulphonate in samples taken from wearing courses from the different field trial sections treated with the product. A calibration of the method indicated that the lignosulphonate samples may be quantified if measured at the wavelength 280 nm, corresponding to the wavelength at which lignin absorbs light. Samples containing high concentrations of lignosulphonate need to be diluted before measurement.</p><p>Also the lignosulphonate seems to rapidly decreasing with time during the summer months, probably even faster than the chlorides. Contrary to sections treated with chlorides, where residual content actually increased during a time interval after the point of dust suppressant application, the residual content of lignosulphonate started to rapidly decrease directly following the occasion of application.</p><p><b>Horizontal diffusion of particulate matter from gravel roads</b></p><p>European Council Directive 1999/30/EC states limit values for sulphur dioxide, nitrogen dioxide, lead, and particulate matter in ambient air. This directive, valid from the 1st of January 2005, states that the average daily concentration of particles smaller than 10 μm, PM10, should not exceed 50 μg/m3 for more than 30 times a calendar year and that the total average yearly concentration should not exceed 40 μg/m3. From the 1st of January 2010, the corresponding values are 50 μg/m3 per 24 hours not to be exceeded more than seven times a calendar year and a maximum of 20 μg/m3 for the average yearly concentration.</p><p>The correlation between the mass concentrations of airborne PM10 particles and distance from the road has been studied by placing DustTrak equipments at different distances from the road edge. The equipments were placed on an open field in a downwind location from a gravel road.</p><p>There seems to be a linear correlation between mean mass concentration of PM10 and dust decay with the distance from road edge. If the regression is extrapolated, it indicates that particles travel no further than 45m downwind from the road edge in prevailing wind velocities between 0 and 7 m/s. It hardly seems possible that larger particles than these would travel further.</p><p>The highest mean PM10 concentration obtained, slightly above 0.5 mg/m3, at a distance of 5 m from the road edge corresponds to a mass concentration ten times greater than the permitted mean PM10 concentration during 24 hours, according to the directive. Since this mean particle concentration was obtained during one hour, with a traffic density of, or as much as, one car per minute, it hardly seems likely that the European Council directive is infringed on locations beside the road.</p><p><b>Laboratory study of dust suppressant leaching</b></p><p>Experiments for studying dust suppressant leaching from granular material of different composition have also been performed in the laboratory environment. Plastic measuring cylinders with small holes drilled in the bottom were used for holding different material fractions. Fractions 0-0.5 mm, 0-2 mm, 0-4 mm, 0-8 mm, 0-16 mm, 2-16 mm, and 0-16 mm containing an addition of 10 wt-% clay were put into the cylinders. Unfortunately, the test material used was somewhat deficient in fines, i.e. containing about 4 wt-% passing the 0.075 mm (mesh no. 200) sieve. Compacted samples were then subjected to dust suppressant solutions and water leaching.</p><p>Analysis of the variance of means from chloride leaching shows significant differences to the 5% level among means regarding different aggregate fractions, but shows no significant difference among means between 0-16 mm material with an addition of clay and 0-16 mm material without addition of clay. Material fraction 0-2 mm is the most efficient in retaining chlorides, and chloride leaching increases with increased material fraction sizes beyond this fraction.</p><p>The 43 wt-% lignosulphonate solution, as used in Halmstad, was used to study lignosulphonate leaching. The results indicate that the lignosulphonate is at least, equally prone to leaching as are the chlorides. In similarity to chloride leaching, the coarser the granular material fraction is the higher the concentration of lignosulphonate within the leachate. It is also apparent that the addition of 10 wt-% clay into the 0-16 mm granular material fraction, initially containing about 4 wt-% fines, clearly decreases the rate of lignosulphonate leaching. However it was seen during these experiments that the water, added for rain simulation, stayed longer on the surface before draining into the material. Therefore, the addition of clay may cause excessive water retention and hence possibly a wet, smeary, and slippery road surface.</p><p><b>Influence of grain size distribution and chloride concentration on the drying process</b></p><p>Water is often referred to as nature’s own binding agent. Therefore an aggregate with good moisture retaining properties is a good investment. The mode of action regarding salts are their hygroscopic, i.e. moisture retaining, properties. In addition, the grain size distribution is also thought to influence the moisture content of a gravel wearing course, with finer material fractions retaining moisture more efficiently. Therefore the influence of grain size distribution as well as chloride type and concentration on the drying process were assessed in the laboratory.</p><p>The drying rate of the aggregate was defined as the water loss (gram) within a unit of time (hour). Since the experiment was set up to provide for relative readings, and all samples had the same surface area, the surface area was not taken into consideration when calculating the drying rate. The experiment was performed on material fractions 0-0.5 mm, 0-2 mm, 0-4 mm, 0-8 mm, 0-16 mm and 2-16 mm. Drying was first carried out without the addition of salt and subsequently with a 30 g addition of calcium chloride and magnesium chloride, respectively.</p><p>During the concentration interval which is reasonable out of economic reasons, the water retaining capacity increases with increasing chloride concentration for both calcium chloride and magnesium chloride treated aggregates.</p><p><b>Development of a laboratory method for studies of dust formation</b></p><p>Preliminary experiments to develop a fast, relatively simple, and reliable laboratory method for the examination of dust suppressant effectiveness have been carried out. This method should be able to evaluate the influence on dust control efficiency of; concentration and type of dust suppressant, grain size distribution of granular material, precipitation, temperature, and traffic intensity.</p><p>A high pressure air device was constructed with the purpose to simulate vehicle wakes which raise dust into the air. A Proctor compaction device was incorporated to achieve compaction, thus forcing the fines up towards the surface; and crushing, thus increasing the content of fines, of the aggregate material. The compaction subjected the grain particles to wear and therefore clearly accomplished a larger content of fines. Also observed was that the compaction forced the fines up towards the surface whilst the opposite was observed for the uncompacted control. The results clearly state that the compaction also caused a more well-bound fines fraction, giving rise to less dust than the uncompacted control. Further, even the smallest addition of salt, corresponding to 3.75 g CaCl2 per 2700 g granular material, eliminated dust generation. . This amount is equivalent to about 60 g/ m2, which can be compared with the typically applied amount of about 200 g/ m2 when dust suppressing with MgCl2 on the gravel road.</p><p>Wearing course material from eight of the test sections included in the field trial in Rättvik was collected and put into laboratory containers for comparison between the equipment used in the field and the developed laboratory device. The biggest advantage with the laboratory device, in comparison to the DustTrak equipment used in the field trials, is that this apparatus, besides allowing for relatively simple laboratory tests, collects emitted dust fraction and thus renders possible complementary analyses of this fraction. The results indicate that there is no statistical evidence at the 5% level for any difference among the sample means of the different road sections. In fact, it is possible that there, are no differences among the sample means, since the samples were collected late in the season when re-blading of the whole road had been made and the concentration of dust suppressant thus most likely was low. In resemblance to the result from the field evaluations, large standard deviations were obtained for comparisons between different samples on the same section. This may indicate that dust emission varies greatly in different spot locations on the road or that the method does not offer very reproducible results. This must be the subject for further studies</p><p><b>The most important findings of this project study:</b></p><p>• Solid calcium chloride is on average 19% more effective at holding the dust than the equivalent amount of solid magnesium chloride.</p><p>• A solution of either magnesium chloride or calcium chloride is the most efficient dust suppressant.</p><p>• Sodium lignosulphonate, with a relatively larger content of sulphur and sodium, is much more effective than calcium lignosulphonate, containing a relatively larger amount of calcium, oxygen, and carbon, for dust suppressing purposes.</p><p>• An objective method for quantifying the amount of dust on gravel roads has been evaluated. The method gives the opportunity to distinguish between sections treated with different dust suppressants. The method has been compared with visual inspections and shows a relatively good correlation. It gives substantial and reliable results. However the method has only been calibrated for a constant vehicle speed of 40 km/h.</p><p>• Statistical analyses, from visual inspections of the road condition, at the 5 % level conclude that the amount of dust raised by traffic can be used as a predictor for development of both evenness and loose gravel on the road surface.</p><p>• Methods for residual content quantification of chlorides as well as lignosulphonate have been developed and seem to work satisfactory.</p><p>• Any residual content regarding both chlorides and lignosulphonate is in principle non-existent in the surface wearing courses after the summer season. Calcium lignosulphonate is especially fast leaching which probably explains why roads which are dust suppressed with this product gives of much dust already in the beginning of the summer.</p><p>• After the application occasion, the chloride residual content in the gravel wearing course increase during a period of less precipitation and then starts to decrease rapidly during periods of more heavy precipitation. This is, however, more pronounced for solid salt than for salt solutions.</p><p>• The particle emission diffuses no further than 45 m down wind from the gravel road in prevailing wind velocities of between 0 and 7 m/s.</p><p>• Even regarding locations very close to the gravel road, there is a very small risk of exceeding the maximum allowed daily average concentration of PM10 stated by the European Council directive.</p><p>• Lignosulphonate is remarkably favoured by a larger content of fines. Much less lignosulphonate is leached from the granular material if clay or mesa is added to obtain a total content of fines of about 15 wt-%. The corresponding effect, from increasing the content of fines, was not seen in neither the field nor the laboratory regarding the leaching propensity of salt.</p><p>• The moisture retaining, i.e. hygroscopic, property of magnesium- and calcium chloride increases with increasing concentrations until a peak concentration, where further increases in concentration yield no efficiency gains. However at the concentration interval which is reasonable out of economical reasons, both calcium- and magnesium chloride treated aggregates benefit from increasing the chloride concentration.</p><p>• A finer aggregate fraction does not hold water more effectively, in fact the opposite is true, but it can absorb more water and therefore is more effective from a moisture retaining perspective.</p><p>• A laboratory method, incorporating a handheld dust generating and collecting device, for pre-assessing the efficiency of dust suppressants has been developed and pre-evaluated. Preliminary results suggest that the amount of accessible dust in different local spots is highly variable. This suggests that it is difficult to obtain a good reproducibility with this method. The highly dust variable local spots are also an apparent problem on the road. The developed objective method collecting PM10 solves this problem by recording numerous sample points.</p><p><b>Future studies</b></p><p>Some future studies within this project are already planned. For example, the field study will go on during one more season. The f
| Identifer | oai:union.ndltd.org:UPSALLA/oai:DiVA.org:kth-4356 |
| Date | January 2007 |
| Creators | Oscarsson, Karin |
| Publisher | KTH, Civil and Architectural Engineering, Stockholm : Byggvetenskap |
| Source Sets | DiVA Archive at Upsalla University |
| Language | English |
| Detected Language | English |
| Type | Licentiate thesis, monograph, text |
| Relation | Trita-VT. FR, 1650-867X ; 2007:07 |
Page generated in 0.0067 seconds