Excessive concentrations of arsenic in soil- and groundwater constitute a global issue. The spread of arsenic is due to both natural and anthropogenic effects in the environment. Historically, the anthropogenic emissions have originated from several different industrial sectors e.g. wood impregnation. In Sweden, there are approximately 85 000 contaminated sites. Most of them are contaminated due to industrial activities. The emissions of arsenic in Sweden are mainly due to the wood impregnating agent CCA (copper, chromium and arsenic). Arsenic is a toxic metalloid that predominantly occurs as the inorganic compounds arsenite As[III] and arsenate As[V] in soil and groundwater. The mobilization of the arsenic compounds in soil water is affected by the redox conditions and by other conditions in the field, e.g. pH. Arsenate adsorbs stronger to iron- and aluminum hydroxides. Arsenite is the most toxic, mobile and soluble of the two compounds.   In a previously remediated impregnation plant in Hjältevad, Småland, arsenic has started to spread in the soil and groundwater. High dissolved concentrations of arsenic have been measured in the area. A hypothesis about the recent mobilization of arsenic is that pollutants below the groundwater table which were left behind after remediation started to dissolve due to changed redox conditions. This report aims to evaluate how the mobilization and adsorption of arsenic in Hjältevad is effected by pH and redox potential. This was addressed by leaching tests and geochemical modeling in Visual MINTEQ. Soil samples were collected during autumn 2017. Soil samples from seven different sampling points, taken from different depths were collected. Leaching tests were carried out for both dry and humid soil samples. Oxalate extractable arsenic was used together with measured dissolved concentrations of cat- and anions as input in Visual MINTEQ. The input data were used to evaluate the mobilization and adsorption of arsenite, arsenate and total arsenic due to pH, redox potential and the specific surface area of ferrihydrite. The modeling showed that the mobilization and adsorption of arsenic is dependent on pH, redox potential and reactive surfaces in the soil solution. Arsenate was adsorbed more strongly between pH 4.5 and 6.5 while arsenite was adsorbed at pH values greater than 7. The conclusion is that the mobilization and adsorption of arsenic were affected to some extent by the pH value. However the redox potential and the specific surface area of ferrihydrite were more influential. To verify the trends seen in the report, additional modeling is required. The report shows that most likely, arsenic started to mobilize in Hjältevad due to the changed redox conditions.

Förorenad mark är ett problem som kan hota ekologiska värden samt skada människors hälsa. Det finns flera åtgärder för föroreningar både ex-situ och in-situ. På senare tid har användningen av nollvärt järn som in-situ åtgärd studerats, men har inte använts tidigare i Sverige. Denna studie innefattar ett pilotförök i fält där en in-situ injektering av nollvärt järn i nanostorlek (nZVI) i Hjältevad, Eksjö kommun. Där bedrev dåvarande Televerket en impregneringsanläggning för telefonstolpar där de använde koppar, krom och arsenik (CCA) som efter ett läckage i en lagringstank förorenade grundvattnet på platsen.  Innan denna studie visade skak och kolonnförsök att tillsats av nZVI till jordprover från Hjältevad minskade koncentrationen av både total arsenik och arsenit med över 90 %. Syftet med pilotförsöket var att undersöka hur långt nZVI skulle spridas i akvifären och om arsenikhalten och grundvattenkemin påverkades på samma sätt av nZVI injekteringen som i labbförsöken.  Dynamiska grundvattenprovtagningar utfördes innan behandlingen för att kunna karakterisera grundvattnet i försöksytan. I ett inledande försök uppskattades spridningen av en 5 g/L nZVI slurry med tillsatt 0,75 g/L polyakrylsyra som injekterades med direct push teknologi. Spridningen av en nZVI injektering undersöktes genom mätning av elektrisk konduktivitet samt genom att observera färgen på grundvattnet. En spridning av nZVI mellan 0,6 och 1,5 m observerades och influensradien ökade med djupet i försöksytan samt visar på större vertikal utbredning vid grundare djup. Preferentiellt flöde kunde ses och överlappande injekteringspunkter behövs för en heltäckande utbredning. Pilotförsöket omfattade fem injekteringspunkter som borrades med 2 m avstånd vid källtermen. Sex nya grundvattenrör installerades uppströms och nedströms injektionspunkterna. Vattenprover togs under 4 månader där grundvattenkemiska parametrar som pH, EH, total As, trevärd As, och Fe koncentrationer mättes. Arsenikhalten låg runt 1700 µg/L innan behandlingen och har inte minskat 4 månader efter injektering av nZVI. I grundvattenrören närmast injekteringspunkterna fanns en tendens till ökad totalhalt och andel arsenit. Möjliga orsaker diskuteras och en hypotes är att lägre temperatur i akvifären jämfört med labbförsök förklarar lägre reaktivitet för nZVI. / Soil contamination poses ecological risks and enhances risks for human exposure to chemical in many areas of the world. Various ex-situ and in-situ soil remediation strategies have therefore been developed. The in-situ use of zero-valent iron sorbents is a relatively new strategy that has not yet been used in the field in Sweden. The current study involves a pilot study of an in-situ inejction of nano-scale zerovalent iron (nZVI) in a contaminated site in Hjältevad, Eksjö municipality. The former Swedish State-owned telephone compmnay operated a plant where telegraph poles were impregnated with chromated copper arsenate (CCA) that leaked into the groundwater following a tank rupture.   Batch and column tests done prior to the current study in the lab showed that addition of nZVI to sediment sampled from the aquifer reduced the dissolved concentrations of total arsenic as well as arsenite with more than 90 %. The aim of this pilot test was therefore firstly to investigate to investigate whether the same reduction could be obtained in the field.  Dynamic groundwater sampling was carried out before the pilot test to characterize hydrology and chemistry of the aquifer. A 5 g/L nZVI slurry was injected using direct push technology initially in one injection well to investigate spread of nZVI by measuring electrical conductivity and observing the color of the groundwater. The measured electrical conductivity and color observation showed a spread of nZVI between 0.6 and 1.5 m. The radius of influence increased with depth in the test zone and shows greater vertical spread at shallower depths. Preferential flow could be seen and overlapping injection points are thus needed for a complete coverage of the source zone. Five injection well were then drilled into the source zone at a distance of 2 m. New groundwater wells were installed downstream from the source zone where samples were taken over a period of 4 months in which chemical parameters such as pH, EH, total As, trivalent As, and Fe concentrations were measured. Arsenic concentrations were 1700 µg/L before the treatment and this concentration as well as other measured groundwater paramters did not decrease significantly during the 4 months monitoring period. In the groundwater wells closest to the injection points, the concentration of total and trivalent arsenic increased significantly. Possibly reasons are discussed and it is hypothesized that the lower aquifer temperature in the field comapred to the lab experiments may explain the lower nZVI reactivity.