|
et al., 2004; Hull et al., 2005; Jani et al., 2016; Mönkäre et al., 2016;
<br />Quaghebeur et al., 2013).
<br />The importance of moisture content in the current study, how-
<br />ever, is in its influence on subsurface resistivity, since it exerts the
<br />dominant control over resistivity distribution in landfill environ-
<br />ments (Bernstone et al., 2000). Moisture content is inversely pro-
<br />portional to resistivity of subsurface materials. For instance,
<br />igneous and metamorphic rocks have higher resistivities than sed-
<br />imentary rocks given the latter’s higher porosity, and thus higher
<br />water content (Pomposiello et al., 2012). Archie’s empirical equa-
<br />tion (Archie, 1942) best describes the relationship:
<br />q ¼aqw nm sl ð1Þ
<br />where q = rock or soil resistivity,a = tortuosity or lithology con-
<br />stant,qw = pore water resistivity,n = porosity,m = cementation
<br />exponent,s = degree of saturation, and l = saturation exponent
<br />(Glover, 2016). Thus, resistivity is largely dependent on the mois-
<br />ture content, pore water resistivity, and how the pore water is dis-
<br />tributed in the mineral i.e. porosity, degree of cementation, degree
<br />of saturation, and fracture state (Clayton et al., 1995). These factors
<br />are in turn important in interpretation of the obtained electrical
<br />resistivity models.
<br />3.1.6. Trace element contents
<br />The trace element contents of the sampled waste from Alster-
<br />fors dump glass hotspots and Madesjö glass dump are shown in
<br />Table 3 in comparison with Swedish Environmental Protection
<br />Agency (SEPA) limits for hazardous waste (Avfall Sverige, 2007).
<br />The elements As, Cd, Pb and Zn were in hazardous amounts while
<br />Ba, Cu and Sb were lower than their respective SEPA limits. The
<br />hazardous amounts of some of the elements correspond to their
<br />frequent uses in glass production. As2O3 was used as a refining
<br />and decolourising agent, CdS as a colouring agent and Pb3O4 as a
<br />glass network stabiliser and modifier (Hermelin and Welander,
<br />1986). The variations in concertation of Pb especially in Alsterfors
<br />glass samples could be attributed to little use of Pb since the fac-
<br />tory did not primarily produce Pb crystal glass. Furthermore, the
<br />use of Pb as a core component of crystal glass was only enacted
<br />by an EU directive in 1969 (European Community, 1969), towards
<br />the final years of the factory. The results also imply that sample S4
<br />may have contained some Pb crystal glass.
<br />The high contents of As, Cd, Pb and Zn in hotspot glass could
<br />benefit metal extraction processes, although no economic feasibil-
<br />ity analysis was done to assess potential contribution of such sec-
<br />ondary metal sources to the high global demand. Estimations of
<br />possible economic gains in view of recycling as secondary raw
<br />materials could be useful. Furthermore, since the 2-D ERT only
<br />focuses on single depth profiles that do not give full indication of
<br />the material volumes in a whole dump, incorporation of 3-D ERT
<br />or different complementary geophysical methods is recommended
<br />in the effort to achieve economic feasibility assessments.
<br />4. Conclusions
<br />ERT was conducted at two old glass dumps (open and buried
<br />glass) to identify glass hotspots for excavation and later use as
<br />sources of secondary raw materials. Despite challenging site con-
<br />ditions, with exceptionally high contact resistances at one of the
<br />sites, good quality data was achieved, thanks to suitable survey
<br />design and careful field procedures. Identification of glass hot-
<br />spots in the buried glass dump was guided by the ERT results
<br />from the open glass dump and was based on sharp contrasts in
<br />resistivity between glass and other materials. Regions of high
<br />resistivity (>8000 Xm) were confirmed through TP excavations
<br />as glass hotspots. Physico-chemical characterisation of hotspot
<br />materials, indicating mean waste composition of 87.2% glass (up
<br />to 99% in some samples), further confirmed glass hotspots and
<br />thus the potential for ERT to identify them. Furthermore, careful
<br />excavation of TPs with ERT as the pre-excavation guide indicated
<br />the potential for obtaining ‘clean’ glass for recycling purposes,
<br />which would be challenging to obtain through random, uncoordi-
<br />nated excavations.
<br />The study, however, encountered some limitations that
<br />require caution during data acquisition and interpretation in
<br />glass waste dumps. Firstly, the similarities in resistivity between
<br />Granite bedrock and crystal glass present the risk of misinterpre-
<br />tation, especially in a site like Alsterfors where both lie close to
<br />the surface. This, however, would not be a big limitation in sites
<br />with deep-lying and different bedrock types. Secondly, the high
<br />resistivity contrasts are prone to introduce artefacts in the
<br />results, which may further increase the degree of uncertainty
<br />with depth. Furthermore, at sites with complex variation in resis-
<br />tivity the 3-D character of the variation will lead to artefacts in
<br />2-D ERT results, so called 3-D effects. This can be handled by
<br />using a 3-D ERT approach, e.g. by measuring a number of parallel
<br />2-D ERT lines and merging the data to a 3-D data set before
<br />inversion, which prevents this type of artefacts but requires more
<br />data to be collected. It is recommended, therefore, that a mod-
<br />elling study about variation of resistivity with depth and intro-
<br />duction of artefacts (their nature, magnitude and impact) in
<br />such sites be conducted, assessing 2-D as well as 3-D ERT
<br />approaches, in order to find a suitable trade-off between quality
<br />of results and survey cost.
<br />Nevertheless, given the inherent shallowness of glass waste
<br />dumps, it is concluded that ERT is applicable since uncertainty
<br />is considerably reduced near the surface. ERT could thus be a use-
<br />ful non-destructive technique towards obtaining more homoge-
<br />neous buried glass and other wastes from LFM for use as
<br />secondary raw material sources in metal extraction and other
<br />waste recycling techniques while eliminating complicated and
<br />often costly waste sorting mechanisms. These findings could con-
<br />tribute to the effort for decontamination of such old dumpsites
<br />with integration of sustainable material recovery techniques for
<br />the circular economy.
<br />Table 3
<br />Trace element concentrations of sampled waste glass compared with Swedish EPA limits for hazardous waste (Avfall Sverige, 2007).
<br />Element (mg kg
<br />1) Madesjö Glass Alsterfors Glass S1 Alsterfors Glass S2 Alsterfors Glass S3 Alsterfors Glass S4 Swedish EPA limits for hazardous waste
<br />As* 13,138 (4 5 1) 2636 (34) 3102 (36) 4468 (43) 4781 (58) 1000
<br />Ba 1221 (46) 860 (34) 522 (35) 411 (26) 376 (27) 10,000
<br />Cd* 394 (25) 443 (10) 430 (11) 522 (11) 647 (14) 100
<br />Cu 551 (44) 125 (9) 456 (13) 369 (11) 769 (18) 2500
<br />Pb* 245,822 (3494) 5929 (60) 1315 (18) 2737 (29) 11,470 (1 4 2) 2500
<br />Sb 4233 (1 4 1) 243 (14) 571 (19) 1022 (21) 1065 (23) 10,000
<br />Zn* 4032 (1 0 1) 7020 (72) 8108 (90) 14,320 (1 2 5) 10,882 (1 0 7) 2500
<br />Elements exceeding at least one limit indicated by (*); Values in brackets represent standard deviations.
<br />R.N. Mutafela et al./Waste Management 106 (2020) 213–225 223
|