|
dumped there (Länsstyrelsen Kalmar, 2012). Orrefors crystal glass
<br />typically constituted SiO2 (55%), Pb3O4 (300 %), K2CO3 (13%) and
<br />other ingredients (2%) such as As2O3 (Duncan, 1995), although
<br />glass in the area generally constituted 55–85% SiO2 (Hermelin
<br />and Welander, 1986). Covering approximately 38,500 m
<br />2, Madesjö
<br />dumpsite has a small stream on its southern part flowing west-
<br />ward and another on its north-western part flowing south-west,
<br />and thus it is designated as a high-risk site (Länsstyrelsen
<br />Kalmar, 2019; SGU, 2019).
<br />2.1.2. Alsterfors glass factory dumpsite
<br />Alsterfors glass factory site, with about 4200 m
<br />2 surface area, is
<br />situated along the Alsterån River in Uppvidinge Municipality. Geo-
<br />logically, it is located in a moraine region with granite as the main
<br />underlying bedrock and an estimated soil depth of 3–10 m (SGU,
<br />2019). The factory was active between 1886 and 1980 and pro-
<br />duced different glass types (packaging glass, fibre glass and art
<br />glass). Glass factory raw materials were mainly quartz and feldspar
<br />sand (SiO2), soda (NaCO3), potash (K2O3), calcite (CaCO3), dolomite
<br />(MgCO3.CaCO3), witherite (BaCO3) and different metal oxides such
<br />as Pb3O4 and As2O3 (Hermelin and Welander, 1986; Uddh-
<br />Söderberg et al., 2019). Factory waste (crushed glass, discarded
<br />raw material batches, furnace and demolition waste, grinding
<br />and acid sludges, etc.) was mostly dumped around the factory pre-
<br />mises (Uddh-Söderberg et al., 2019), resulting in a 2600 m
<br />2 dump
<br />with approximately 5200 m
<br />3 of glass and other wastes (Höglund
<br />et al., 2007), some of which is exposed on the surface. The site is
<br />also designated as high-risk and ranked fourth out of thirty nine
<br />high-risk objects in the county (Länsstyrelsen Kronoberg, 2018).
<br />2.2. Electrical resistivity Tomography (ERT) surveys
<br />ERT was the geophysical method used in this study. The 2D
<br />resistivity method measures and maps electrical resistivity of sub-
<br />surface materials at different depths along a survey line (Loke et al.,
<br />2010; Reynolds, 2011) based on the principle of different materials
<br />having different electrical properties naturally (Wang et al., 2015).
<br />Electrodes are deployed on a site and connected to a set of multi-
<br />electrode cables arranged along a line. Electric current is transmit-
<br />ted through a sequence of different pairs of the deployed elec-
<br />trodes while measuring the resulting potential differences
<br />between one or several pairs of electrodes simultaneously. Pulses
<br />of direct current (DC) with alternating polarity or low-frequency
<br />alternating current (AC) are used (Powers et al., 1999).
<br />In this study, an ABEM Terrameter LS, which is a multi-channel
<br />resistivity-IP (Induced Polarisation) instrument, was used together
<br />with an Electrode Selector ES10-64C and a set of multi-core elec-
<br />trode cables. The surveys were made using separated electrode
<br />cable spreads for current transmission and potential measurement
<br />in order to reduce capacitive coupling in the cables, which was
<br />expected due to very high resistive surface layers known from pre-
<br />vious tests at the sites (Mutafela et al., 2018), in order to optimise
<br />the data quality (Dahlin and Leroux, 2012). Two electrode cables
<br />were connected to the Terrameter and two to the Electrode Selec-
<br />tor, with every other electrode connected to each, as shown in the
<br />sketch in Fig. 2a.
<br />Eighty-two stainless steel electrodes were inserted into the
<br />ground with an inter-electrode spacing of 1 m. The small spacing
<br />was used in order to achieve good horizontal and vertical resolu-
<br />tion of the images for a site as shallow as Alsterfors dumpsite
<br />(Dahlin and Zhou, 2004; Höglund et al., 2007). The electrode con-
<br />figuration was multiple gradient array (Dahlin and Zhou, 2006).
<br />Three lines at Alsterfors (AL1,AL2 and AL3) and two at Madesjö
<br />(ML1 and ML2) were thus obtained.Fig. 2b and c show equipment
<br />set-up at the two sites. Since these ERT surveys in a glass dump
<br />were first to the knowledge of the authors, different instrument
<br />modes and settings were tried to reach optimum settings as shown
<br />in Table 1. At Madesjö, electrode contact resistance was particu-
<br />larly high since it’s a glass heap (exposed materials), whereas
<br />Alsterfors has soil cover (buried materials). Therefore, in order to
<br />reduce electrode contact resistance and transmit sufficient current
<br />for good signal strength, ground contact was improved using a gel
<br />based on Johnson Revert Optimum around the electrode and in
<br />some cases also by adding an extra electrode. The lines obtained
<br />at each site were visualised as 2D resistivity images after data pro-
<br />cessing and inversion (as described in Section 2.4.3), which were
<br />used in planning and setting TP excavations.
<br />2.3. Excavation and sampling procedure
<br />Test pits (TPs) were excavated along the profile lines using a 5-
<br />tonne excavator, and according to a procedure from Kaczala et al.
<br />(2017)where excavated materials were separately stockpiled from
<br />each 1 m depth interval. At Alsterfors, twelve TPs were excavated,
<br />eleven of which were 2–2.5 m deep and one was 4.5 m deep (on
<br />AL2), whereas at Madesjö eight TPs were excavated at 1.5 m max-
<br />imum depth due to unstable surface (glass waste). Identified hot-
<br />spot stockpiles at Alsterfors were later sampled according to the
<br />Nordtest Method NT ENVIR 004-1996/05, in order to assess mate-
<br />rial composition and physico-chemical properties. Thus, four hot-
<br />spot stockpiles (S1, S2, S3 and S4) were sampled (about 100 kg
<br />each) from Alsterfors for analysis. After sampling, stockpiled mate-
<br />rials in each case were returned into their respective TPs to restore
<br />the site as much as possible.
<br />2.4. Physico-chemical characterisation of hotspot materials
<br />2.4.1. Sample sieving and hand-sorting
<br />To achieve particle size distribution (PSD), the collected sam-
<br />ples were sieved using eight sieves (Tidbecks Sweden and Giuliani
<br />Technologies) with mesh sizes 63 mm, 31.5 mm, 20 mm, 16 mm,
<br />11.3 mm, 8 mm, 4 mm and 2 mm mounted on a laboratory sieve
<br />shaker (Pascall Engineering). Based on particle sizes, the sieved
<br />materials were aggregated into course (>31.5 mm), medium
<br />(11.3–31.5 mm) and fine (<11.3 mm) fractions. Furthermore, the
<br />particle sizes were assessed based on mass percentage of the par-
<br />ticles passing through each sieve size. For waste composition, the
<br />sieved materials were hand-sorted and the materials fractions cat-
<br />egorised into glass, demolition (mainly bricks), soil, organic and
<br />residual (plastics, metals, roofing material parts, etc). All fractions
<br />sized 2–8 mm were hand-sorted using magnifying lenses while
<br />the fractions <2 mm were too fine to hand-sort. All calculations
<br />were done on wet weight basis in all instances.
<br />2.4.2. Trace elements and moisture contents
<br />A portable X-Ray Fluorescence (XRF) analyser, Olympus DS-
<br />4000 (Innov-X), was used to analyse trace element contents of all
<br />samples (oven-dried) below 8 mm. The analysis was done on the
<br />hand-sorted glass fraction only and in triplicates in each case.
<br />Analysis on the glass fraction was aimed at assessing trace ele-
<br />ments potential for future recycling. Sample moisture content
<br />was analysed based on dry residue of the fine fraction (FF) of
<br />Alsterfors soil samples, Alsterfors glass hotspot samples and
<br />Madesjö glass samples according to the Swedish Standard SS-EN
<br />14346:2007.
<br />2.4.3. Data analysis and interpretation
<br />2.4.3.1. ERT profiles. Inversion:The contact resistances were excep-
<br />tionally high for the Madesjö ERT lines, ranging to over 100,000
<br />Xm with a mean of 86,000–98,000 Xm(Table 1), despite the
<br />efforts at improving electrode contact. For Alsterfors, the resis-
<br />tances ranged up to 10,000 Xm with a mean of a few thousand
<br />R.N. Mutafela et al./Waste Management 106 (2020) 213–225 215
|