building rubble composed soils : contamination status and ... · building rubble composed soils:...

Building rubble composed soils:contamination status and sulfate release

vorgelegt von:M.Sc.

Abel, Stefangeb. in Prien am Chiemsee

von der Fakultat VI - Planen Bauen Umweltder Technischen Universitat Berlin

zur Erlangung des akademischen Grades

Doktor der Naturwissenschaften- Dr. rer. nat.-

genehmigte Dissertation

Promotionsausschuss:Vorsitzender: Prof. Dr. Reinhard Hinkelmann

Gutachter: Prof. Dr. Gerd WessolekGutachter: Prof. Dr. Jurgen Bottcher

Tag der wissenschaftlichen Aussprache: 08. Juli 2015

Berlin 2015

mailto:[email protected]

Acknowledgements

I would like to thank

• Prof. G. Wessolek for supervising the thesis, his confidence in my work and constructivediscussions on my findings

• Prof. J. Bottcher, the co-examiner for reviewing this thesis

• The Senatsverwaltung fur Stadtentwicklung und Umwelt for funding the project and inparticular U. Hormann and Dr. B. Fritz-Taute

• Dr. T. Nehls for constructive discussions and his contribution to Chapter 2, 3 and 4

• Dr. A. Peters for constructive discussions and his contribution to Chapter 4

• Dr. B. Kluge for constructive discussions

• all the staff members for their help with the installation of the lysimeter and the Ph.D.students of the Department of Soil Conservation for their companionship and for valu-able advices

• the laboratory staff members of the Department of Soil Science for their support withsample analysis, especially Iris Pieper

• κ . from the 10th floor for his support whenever problems with software occurred

• C. Lange (Zentraleinrichtung fr Elektronenmikroskopie, TU-Berlin) for X-ray diffrac-tion (Chapter 3)

• Last but not the least, I would like to thank my parents for supporting me throughoutmy life

AbstractBackground: Soils in urban areas frequently exhibit appreciable amounts of building rubble.Particularly in cities which have been heavily bombed during World War II, e.g., in the studyarea Berlin, rubble composed soils cover vast areas. Building rubble may possess elevatedcontents of organic and inorganic pollutants. The disordered deposition and the omission ofany protection facilities can lead thus to negative impacts on environment and human health.Particularly, high sulfate concentrations in the groundwater result from the deposition of build-ing waste. The study assesses the contamination status and the sulfate release from buildingrubble composed soils.Methods: A data set from numerous soil surveys was analyzed for heavy metals and benzo[a]-pyrene contamination patterns of rubble composed soils and substrates free of any rubble. Be-sides, the sulfate release from rubble composed soils were intensively studied. This includesanalysis of total sulfur, its binding forms, sulfate bearing minerals and readily soluble sulfatecontents. Furthermore, long-term sulfate release were measured in a lysimeter study, simu-lating various years of groundwater recharge. Based on the results, a numerical model wasdeveloped to describe the sulfate release dynamics.Results: Building rubble composed soils contain significantly higher contents of heavy metalsand benzo[a]pyrene compared to soils free of any rubble. Precautionary value of the GermanSoil Protection Ordinance are frequently exceeded. The rubble composed soils further consti-tute elevated total sulfur contents, mainly in form of inorganic sulfate minerals like gypsum.Its content is increasing with increasing sampling depth. Gypsum, existing in the fine soil andin the coarse-grained soil fraction, counts up to 12 g·kg−1.In the lysimeter study, the leachate from the rubble composed soil has shown initial concen-trations in the range of gypsum solubility (1400 mg·L−1). After the percolation of 1.8 porevolume a rapid decrease occurred and the sulfate concentrations remained over the percola-tion of several pore volumes at a moderate level in the range of the European drinking-waterthreshold value (240 mg·L−1). The sulfate release dynamics indicate a dissolution of the gyp-sum strongly constrained by small effective surface areas. The sulfate release dynamics canbe well described by a numerical model, in which two gypsum pools with different effectivesurface areas are applied.Conclusion: Rubble composed soils feature elevated contents of metals and benzo[a]pyrene.However, there is no remarkable groundwater pollution in the study area Berlin, although vastareas are covered with building rubble. This can be attributed to the alkaline characteristics,favoring the sorption of the pollutants. On the contrary, sulfate contamination of the ground-water is a severe problem. This can be associated with high contents of the readily solublegypsum in building rubble substrates. For building rubble landfills the weathering of gyp-sum will thus last several centuries. Subsequent measures to protect the groundwater shouldtherefore be considered.

AbstraktHintergrund: Aufschuttungsboden mit Trummerschutt konnen in stadtischen Gebieten einenhohen Flachenanteil einnehmen. Dies trifft insbesondere auf Stadte zu, die im 2. Weltkriegstark zerstort worden sind, wie z.B. auf das Untersuchungsgebiet Berlin. Trummerschutt kannerhohte Gehalte an organischen und anorganischen Kontaminanten aufweisen. Da der Trum-merschutt meist ungeordnet und ohne technische Sicherungsmaßnahmen abgelagert wordenist, besteht eine potentielle Gefahr fur die Umwelt und die menschliche Gesundheit. Beson-ders die Sulfatbelastung des Grundwassers ist auf Trummerschutt zuruckzufuhren. Die fol-gende Arbeit beschreibt detailliert die Kontamination und die Sulfatfreisetzung von Trum-merschuttboden.Methoden: Datensatze aus zahlreichen Bodengutachten von Trummerschutt- und naturlichenBoden wurden hinsichtlich Schwermetall- und Benzo[a]pyrengehalten ausgewertet. Wei-ter wurde das Sulfatfreisetzungspotential von Trummerschuttsubstraten intensivst untersucht.Dies umfasst Messungen von Gesamtschwefel, Schwefel-Bindungsformen, sulfathaltigen Min-eralien und leicht loslichen Sulfatgehalt. Daruber hinaus wurde anhand einer Lysimeterstudiedas langfristige Sulfatfreisetzungspotential untersucht. Auf den Ergebnissen basierend, wurdeein numerisches Modell entwickelt, welches die Sulfatfreisetzungsdynamik beschreibt.Ergebnisse: Trummerschuttboden weisen im Vergleich zu Boden frei von technogenen Kom-ponenten signifikant hohere Gehalte an Schwermetallen und Benzo[a]pyren auf, die meist dieVorsorgewerte der deutschen Bodenschutzverordnung uberschreiten. Trummerschuttbodenweisen daruber hinaus hohe Gehalte an Schwefel auf, welcher insbesondere in anorganischerForm wie Gips gebunden ist. Der Gehalt an Schwefel steigt mit zunehmender Beprobungstiefe.Der Gipsgehalt betragt bis 12 g·kg−1, wobei der Gips sowohl in der Feinkorn- als auch in derGrobkornfraktion vorliegt. In der Lysimeterstudie zeigten sich im Sickerwasser anfanglichSulfatkonzentrationen im Bereich der Sattigung mit Gips (1400 mg·L−1). Nach dem Aus-tausch von 1.8 Porenvolumen nahm die Sulfatkonzentration relativ rasch ab und blieb ubermehrere Porenvolumen im Bereich des Grenzwertes der Trinkwasserverordnung (240 mg·L−1).Der Konzentrationsverlauf deutet auf eine kinetisch limitierte Freisetzung, die stark von dereffektiven Oberflache des Gipses abhangig ist. Die Freisetzungsdynamik kann durch ein nu-merisches Modell mit Anwendung zweier Gipspools, welche unterschiedliche Oberflachenaufweisen, gut beschrieben werden.Zusammenfassung: Trummerschuttboden weisen erhohte Gehalte an Schwermetallen undBenzo[a]pyren auf. Es besteht jedoch kein bemerkenswerter Eintrag der Schadstoffe insGrundwasser, was auf die alkalische Bodenreaktion zuruckzufuhren ist. Sulfat hingegen stelltein großes Problem, besonders in Berlin, dar, was auf dem hohen Anteil an leichtloslichenGips im Trummerschutt beruht. Es ist zu erwarten, dass aus Trummerschuttdeponien nochuber mehrere Jahrhunderte Sulfat freigesetzt wird. Um die Grundwasserressourcen zu schutzen,sollten hier Gegenmaßnahmen in Betracht gezogen werden.

Contents Stefan Abel - PhD Thesis

Contents

1 General Introduction 1

1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.2 Regulatory framework and management practice . . . . . . . . . . . . . . . 11.3 Characteristics of building rubble and new C&D waste . . . . . . . . . . . . 41.4 The contaminants - impacts on environment and human health . . . . . . . . 71.5 Study area: Berlin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

1.5.1 Geomorphology . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71.5.2 Building rubble composed soils . . . . . . . . . . . . . . . . . . . . 81.5.3 Groundwater quality . . . . . . . . . . . . . . . . . . . . . . . . . . 12

1.6 Objectives of this thesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

2 Heavy Metals and Benzo(a)pyrene in Soils from Construction and Demoli-

tion Rubble 15

2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152.2 Material and Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

2.2.1 Study area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172.3 Data base . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

2.3.1 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202.4 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

2.4.1 Soil reaction and electrical conductivity . . . . . . . . . . . . . . . . 222.4.2 Cadmium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232.4.3 Lead . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242.4.4 Copper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252.4.5 Zinc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252.4.6 Other Heavy Metals . . . . . . . . . . . . . . . . . . . . . . . . . . 282.4.7 Benzo[a]pyrene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302.4.8 Correlations between the contaminants and identification of represen-

tatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302.5 Conclusion and Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

3 Pools of sulfur in urban rubble soils 33

3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

i

Contents Stefan Abel - PhD Thesis

3.2 Materials and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353.2.1 Sampling and sample preparation . . . . . . . . . . . . . . . . . . . 353.2.2 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

3.3 Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393.3.1 Composition of soils . . . . . . . . . . . . . . . . . . . . . . . . . . 393.3.2 Total sulfur and readily soluble sulfate contents . . . . . . . . . . . . 393.3.3 Arylsulfatase activity . . . . . . . . . . . . . . . . . . . . . . . . . . 443.3.4 Mineralogical characterization and speciation of sulfur . . . . . . . . 44

3.4 Conclusion and outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

4 Long-term release of sulfate from building rubble composed soil: lysimeter

study and numerical modeling 48

4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 484.2 Material and Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

4.2.1 Soil material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 504.2.2 Lysimeter preparation . . . . . . . . . . . . . . . . . . . . . . . . . 524.2.3 Numerical Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . 53

4.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 564.3.1 Lysimeter study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 564.3.2 Numerical model . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

4.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 604.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

5 Synthesis and Conclusions 64

5.1 Heavy metal and B[a]P contamination and its impacts on the environment . . 645.2 Sulfate release and its impact on the environment . . . . . . . . . . . . . . . 675.3 Transferability of the results to C&D waste . . . . . . . . . . . . . . . . . . 695.4 Outlook and future research . . . . . . . . . . . . . . . . . . . . . . . . . . . 715.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72

References 73

ii

List of Figures Stefan Abel - PhD Thesis

List of Tables1.1 Threshold values according to various directives . . . . . . . . . . . . . . . . 31.2 Pollutants in building rubble components . . . . . . . . . . . . . . . . . . . 62.1 Fractions of technogenic components and natural material . . . . . . . . . . 182.2 Analytical methods (ISO) and their detection limits (DL) for the chemical

parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212.3 Classification of contamination according to the GLIi . . . . . . . . . . . . . 222.4 General properties, statistical parameters of contaminant concentrations and

GLIi index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292.5 Spearman’s rank correlation coefficient between contaminants . . . . . . . . 313.1 Skeleton and fractions of coarse grain technogenic components. . . . . . . . 343.2 Geometric mean and ranges of contents of total carbon, total sulfur, soluble

sulfate and pH for soil horizons and main technogenic components of rubblecomposed soils. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

4.1 Stot and readily soluble SO42–-content in the fine soil ≤2 mm and content of

coarse-grained gypsum >2 mm in the topsoil and in the rubble-composed sub-strate, respectively. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

4.2 Input parameter of the numerical model with one gypsum pool and two gyp-sum pools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

5.1 Assessment scheme for the human toxicological health risk by heavy metalsand benzo(a)pyrene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

5.2 Pollutants in building rubble and C&D waste: a comparison . . . . . . . . . 70

List of Figures1.1 Composition of new C&D waste and building rubble composed soils . . . . . 41.2 Quantity of gypsum used in the building industry . . . . . . . . . . . . . . . 51.3 Berlin, 1945: street scene; rubble composed substrate from landfill . . . . . . 61.4 Geological map of Berlin . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81.5 Distance groundwater table to surface . . . . . . . . . . . . . . . . . . . . . 91.6 Map of war damage in Berlin . . . . . . . . . . . . . . . . . . . . . . . . . . 91.7 Construction of the Teufelsberg, 1951 . . . . . . . . . . . . . . . . . . . . . 101.8 Sulfate concentration in the groundwater, locations of landfills in Berlin . . . 11

iii

List of Figures Stefan Abel - PhD Thesis

1.9 Contamination status of the groundwater in Berlin . . . . . . . . . . . . . . . 132.1 Situation in Berlin, Germany 1945 . . . . . . . . . . . . . . . . . . . . . . . 172.2 Distribution of rubble containing soils in Berlin . . . . . . . . . . . . . . . . 192.3 Cumulative distribution of heavy metals and B(a)P . . . . . . . . . . . . . . 262.4 Cumulative distribution of Pb and Hg . . . . . . . . . . . . . . . . . . . . . 273.1 Typical profiles of rubble composed soils . . . . . . . . . . . . . . . . . . . 373.2 Depth profile of total sulfur and soluble sulfate . . . . . . . . . . . . . . . . 403.3 Correlation between total sulfur and soluble sulfate content for fine soil , mor-

tar and plaster components . . . . . . . . . . . . . . . . . . . . . . . . . . . 423.4 X-ray powder diffractogram . . . . . . . . . . . . . . . . . . . . . . . . . . 433.5 Fraction of different sulfur species and respective total sulfur contents in dif-

ferent soil layers as analyzed by XANES . . . . . . . . . . . . . . . . . . . . 463.6 Sulfur K-edge XANES spectra . . . . . . . . . . . . . . . . . . . . . . . . . 464.1 Typical profile of urbic technosols, Teufelsberg Berlin, Germany; gypsum par-

ticles in the coarse-grained fraction; design of the outdoor lysimeter . . . . . 514.2 Lysimter study: average water flux at the upper and lower boundary. . . . . . 574.3 SO4

2–-concentration in the leachate of the lysimeter . . . . . . . . . . . . . . 584.4 Sum of squared residuals for selected combinations of A1,t0 and A2,t0 . . . . . 594.5 Simulated SO4

2–-concentration in the range of measured data and predictedapplying two gypsum pools . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

5.1 Sorption capacity of heavy metals . . . . . . . . . . . . . . . . . . . . . . . 665.2 Predicted time period of SO4

2–-concentrations exceeding TrinkwV thresholdvalue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69

iv

1 GENERAL INTRODUCTION Stefan Abel - PhD Thesis

1 General Introduction

1.1 Background

Today, in developed countries construction and demolition (C&D) waste constitutes the majorfraction of solid waste. Its waste management and landfilling is strictly regularized preventingfrom negative impacts on the environment.However, C&D waste and building rubble1 was - and in most developing countries still is-not always landfilled in environmentally responsible manner. Particularly in Europe, whereseveral cities have been heavily destroyed by bombing raids during the World War II (WW II),rubble has been placed throughout destroyed areas and on landfills, which exhibit no technicalpollution control facilities as e.g. liners. During construction activities in the younger history,rubble continuously was deposited in urban areas and on unsecured landfills until the early70ies in BRD and until the end of the 80ies in the former GDR. Thus, urban soils are fre-quently composed of technogenic materials and cover vast areas. Due to the use of hazardoussubstances and materials in the building industry, today, such urban soils may possess elevatedcontents of organic and inorganic pollutants. Hence, rubble composed soils and in particularbuilding rubble landfills, may pose threat to the environment in view of organic and inorganicpollutants.In several German cities, e.g., Berlin, Hamburg and Dresden, elevated sulfate concentrationsin the groundwater aquifer result from weathering of sulfate bearing minerals from buildingwaste. Besides, for various sites in formerly heavily destroyed areas high metal contents inthe soil are documented, which may be attributed to the deposition of building waste. Bothsubjects, sulfate release and metal contamination will be addressed in this thesis.

1.2 Regulatory framework and management practice

C&D waste, including deconstruction waste from infrastructure and excavation material, con-stitutes the largest volume of all solid wastes in most developed countries. In Europe about461·106 Mg of C&D waste arises per year; in Germany about 53·106 Mg·a−1 of C&D wasteis reported (Cowam, 2004).C&D waste is sorted commonly by selective deconstruction works according to the system ofwaste nomenclature (European Waste Framework Directive 2001, Kreislaufwirtschafts- und

1here and in the following the term ”building rubble” and ”rubble” refers to C&D waste from World War II

1


Abfallgesetzes KrW-/AbfG, 1994, and Abfallverzeichnis-Verordnung AVV). Concrete, bricks,tiles and ceramics is classified as inert waste. Other components as wood, glass, plastics, gyp-sum, metals etc. have to or -in view of financial aspects- should be treated separately duringdeconstruction of buildings and waste management. Overall, C&D waste is recycled andreused to a large extend, e.g., for road construction. In Germany 61.6% is reused (UBA 2008),only 9.1% is deposited on special landfills.

C&D waste complying with the regulation for inert waste according to the Waste FrameworkDirective 2001 (see Tab. 1.1) and with the Landfill Directive 1999 can be deposited on speciallandfills for inert waste. According to the German Landfill Ordinance (1998c) and the LAGA(1997), this C&D waste is classified as ’Z 0’ and the respective landfill as ’Deponieklasse DK0’. C&D waste rich in gypsum or other pollutants as e.g. heavy metals do not meet theserequirements. Such C&D waste has to be dumped on landfills, which exhibit special surfaceand basal liners (at least ’Z 1.1’; ’DK I’). Slags and ashes are classified according to LAGA(1997) as well. They were formerly used as insulation material in ceiling voids and still areused to a large extent in the cement industry and in road construction,

However, in Germany these regulatories were first implemented in 1972. Thus, old landfillslike the building rubble landfills created after WW II, feature no adequate technical facilities toprotect the environment from negative impacts. Beyond this, there was no adequate separationof different kinds of building rubble. Pursuant to the Federal Soil Protection Act (BBodSchG1998b) those landfills are defined as ’Altablagerungen’ or ’Altdeponien’. Also sites whererubble have been dumped or admixed to soil may be suspected of being contaminated (’Ver-dachtsflachen’).

Contaminated sites may pose threat to to the environment and, in particular, to water resources.In order to evaluate the impacts on the environment, quantification of the contamination isobligatory. Further monitoring is necessary at moderate contamination level with respect tosite specific characteristics: In case of high contamination remediation measures to restore thefunctions of the soil and to eliminate any harmful risk to health and the environment have tobe conducted. A major goal is to ensure the quality of groundwater and water intended for hu-man consumption. The concentrations in the leachate have to pursuit the precautionary valuesaccording to the BBodSchV (Tab. 1.1). The quality standards for water bodies and ground-water, given by the European Water Framework Directive (2000), the European GroundwaterDirective (2006) and the European Drinking-Water Directive (1998, see Table 1.1) have to beconsidered thereby as well.All member states of the EU have to adopt strategies and measures to meet these requirements.

2


Table 1.1: Threshold values of mineral (C&D) waste (∗LAGA M 20 1997), in the eluate from C&D waste(⋆DepV, 1998c), in drinking-water (‡European Drinking-Water Directive Directive (1998) and Ger-man TrinkwV (2001), equivalent to threshold values of European Groundwater Directive (2006)) andin the leachate in soil (†precautionary values acc. to BBodSchV 1998a).

total amount eluate - -

Z0∗ Z1.1∗ Z1.2∗ Z2∗ DK 0⋆ DK I⋆ DK II⋆ DK III⋆ TrinkwV‡ BBodSchV(Z0∗) (Z1.1) (Z1.2) (Z2)

mg·kg−1 mg·L−1 mg·L−1 mg·L−1

As 20 30 50 150 ≤0.05 ≤0.2 ≤0.2 ≤ 2.5 0.01 0.010(0.014) (0.014) (0.020) (0.060)

Cd 0.6 1 3 10 ≤0.05 ≤0.2 ≤1 ≤5 0.003 0.005(0.0015) (0.0015) (0.003) (0.006)

Cr 50 100 200 600 ≤0.004 ≤0.05 ≤0.1 0.5 0.05 0.050(0.0125) (0.0125) (0.025) (0.06)

Cu 40 100 200 600 ≤0.2 ≤1 ≤5 ≤10 2 0.050(0.02) (0.02) (0.06) (0.10)

Hg 0.3 1 3 10 ≤0.001 ≤0.005 ≤0.02 ≤0.2 0.001 0.001(0.0005) (0.0005) (0.001) (0.002)

Pb 100 200 300 1000 0.05 0.2 0.2 2.5 0.01 0.025(0.04) (0.04) (0.08) (0.20)

Ni 40 100 200 600 ≤0.04 ≤0.2 ≤1 ≤4 0.02 0.050(0.015) (0.015) (0.020) (0.070)

Zn 120 300 500 1500 ≤0.4 ≤2 ≤5 ≤20 - 0.500(0.15) (0.15) (0.20) (0.60)

B[a]P 0 <0.5 <1 3 in terms of DOC 10−6 -SO4

2– - - - - ≤100 ≤2000 ≤2000 ≤5000 240 -

pH - - - - 5.5 - 13 5.5 - 13 5.5 - 13 4 - 13 6.5 - 9.5 -- - - - (6.5-9.5) (6.5-9.5) (6-12) (5.5-12) - -

However, member states may aim to achieve less stringent environmental objectives than thoserequired in case of disproportionate costs of remediation with regard to their environmentalbenefit or technical unfeasibility. This is e.g. the case regarding the building rubble landfills,created after the WW II, which today are covered by a well evolved vegetation.

3


Figure 1.1: Left: composition of new C&D waste acc. to COWAM (2004) with A=bricks, concrete, tiles, and,ceramics; B= wood; C= tar free bitumen mixes; D=metals and plastics; E=mixed C&D waste; F=gypsum contaminated waste. Right: composition of building rubble composed soils acc. to Wessoleket al. (2011) G= fines; H=other anthropogenic materials; I= mortar and plaster; J= slag; K= bricks;L= natural material.

1.3 Characteristics of building rubble and new C&D waste

Today, the composition of arising C&D waste strongly depends on the vernacular architecturalstyle and is therefore strongly country related. Concrete became the most important construc-tion material, accounting for about 70% (Clement et al., 2011). In the United States, e.g., thefraction of concretes makes up the major part of C&D (up to 50%), followed by wood (upto 30%); the fraction of gypsum drywalls is reported with a maximum of 15% , metals up to5% and asphalt roofing up to 5% (Sandler, 2003). The composition in Germany (see Fig. 1.1)differs from those in the U.S., in particular, the fraction of wood and gypsum drywall is muchsmaller than that in the U.S. However, since the use of gypsum in the building industry isincreasing (see Fig. 1.2), its fraction in the C&D waste can be expected to increase as well inthe near future.The components which are frequently deposited on landfills are commingled C&D waste,contaminated with gypsum, as well as fineries, which may exhibit an appreciable amount ofgypsum. In both cases a separation is technically infeasible in an economically acceptablemanner. Furthermore, gypsum containing material is commonly not suitable for further usesince adding the recyclate for the production of, e.g., concrete, upcoming sulfates may attackits structure. The recovery rate is the lowest compared to other components of C&D waste(76%, other components >88%, KLWB 2010). The fraction of slags in C&D waste is notdocumented. However, slags may occur in C&D waste since it is frequently used as additivein Portland cement or in the construction of roads etc.. Overall, about 9.1% of the total C&Dwaste is deposited on landfills (KLWB 2010), thereby most of it is dumped on the landfillclass DK I.

4


The composition of the building rubble deposited after WW II differs in its composition fromthe present C&D waste (see Figure 1.1). Today, the building industry is characterized by theuse of concrete lightweight and insulating materials, while before 1950 the use of brickworkswas most common. In the building rubble substrates the fine soil fraction, and in particularsand, constitutes the major fraction (see also substrates depicted in Figure 1.3, right). Thequantity of fines is high, since sand was -and still is- used as an additive in mortar and plaster.Concrete possesses a fraction less than 3%. Although intact bricks were sorted and reused toa large extend due to the scarcity of resources after WW II, they constitute the major coarse-grained fraction in rubble composed substrates. Natural products as, e.g., wood is rarelypresent, because it was reused as well. Same applies to metals as e.g. roof flashings orelectrical installations. Building rubble substrates also consists of gypsum. Even though itsuse in the building industry before WW II was not as high as nowadays (see Fig. 1.2), theoverall quantity can be appreciable and may result from stucco fragments and from the use ofgypsum as additive for cements and mortar.

Figure 1.2: Quantity of gypsum used in the building industry according to Bundesverband Baustoffe und Erdene.V. (until 1997), and Statistisches Bundesamt (since 2003), see also Vulpius, B. (2012).

5


1.4 The contaminants - impacts on environment and humanhealth

The intake of heavy metals or organic pollutants (via vegetables, drinking-water or soil par-ticles) may have severe impacts on human health. In particular, lead, cadmium, mercury andarsenic pose high risk to human health. Dependent on the doses, e.g. damage to organs mayoccur; high doses may even lead to death. Detailed information about health-related effectscan be found in e.g. Morais et al. (2012).

In contrast to heavy metals, sulfates in drinking-water do not pose serious threat to humanhealth but excessive consumption of water rich in sulfate may have laxative effect (Heizeret al., 1997). Besides, high concentrations cause negative gustatory effects on the drinking-water quality (World Health Organization (WHO), 2004) and may damage the drinking-waterdistribution system (Larson and Skold, 1958; McNeiill and Edwards, 2001). In groundwatersit provokes the attack of buried concrete facilities and building foundations (Hobbs and Taylor,2000; Freyburg and Berninger, 2003). On the natural environment, there are no direct impactsof high SO4

2–-concentrations.

Microbial reduction of sulfate to H2S may take place in eutrophic waters (Brown, 1982). Thisprocess occurs frequently in the sewer system, causing substantial damage to concrete pipes(Mori et al., 1992; OConnell et al., 2010). However, in the upper groundwater aquifer of thestudy area Berlin a reduction of sulfate can be highly excluded as the mean redox potential ofgroundwater is +143 mV (Hannappel and Reinhardt, 2002); sulfate reduction starts at about-50 mV.

1.5 Study area: Berlin

1.5.1 Geomorphology

The geomorphology of Berlin is dominated by a glacial valley in the inner city and moraineplateaus in the north and in the south, which were formed by the Weichselian glaciation (seeFig. 1.4). The natural substrates in the inner city are mostly fluvoglacial sands. Soils on theplateaus developed either from marls and loamy sands or, like in the south-western part ofBerlin, from glacial load sands or drift sands. Natural soils are commonly arenosols, luvisolsand podzols. However, in densely populated areas anthrosols and technosols predominate.In the glacial valley in the inner city the distance between groundwater table and surface isabout 3 m, whereas for areas in the north and in the south it is about 20 m (see Fig. 1.5). This

7


Figure 1.4: Geological map of Berlin from Umweltatlas Berlin 1998.

leads to different residence time of the infiltrating water in the vadose zone and different geo-chemical reaction characteristics.

1.5.2 Building rubble composed soils

In the end of the WW II Germany became heavily bombed by the Allies. Overall the totalamount of rubble occurred is approximately 1050 Mio. t (Die Welt, 2005), corresponding to620 Mio. m3. For Berlin the quantity of rubble accrued during bombing raids and in firestormsis documented between 55 and 110 Mio m3 (Fichtner, 1977) and is thus the city with the high-est war damage. For other German cities as e.g. Hamburg about 40 Mio. m3 are reported,for Cologne about 30 Mio. m3 (Pietsch and Kamieth, 1991), for smaller cities the quantity isreported with several million m3 as well.The built up area in Berlin (1944) was approximately 72 km2, and was thus much smallerthan today (374 km2). Bombing raids destroyed about 28.5 km2 (Bundesministerium fur Ver-triebene, Fluchtlinge und Kriegsgeschadigte, 1967), particularly in the inner city of Berlin(see Figure 1.6). At the end of the war 19.5% of all buildings were completely or irreparablydestroyed, in the inner city it was about 60% (Statistisches Landesamt Berlin, 1963).

8


1.5.3 Groundwater quality

Threshold values of the European and German Drinking Water Ordinance (TrinkwV 2001)and the European Groundwater Directive (2006), respectively, for heavy metals and organicpollutants in the groundwater are commonly not exceeded in Berlin. The data depicted inFigure 1.9 show that less than 1% of all groundwater samples exhibit concentrations abovethreshold value (SenstadtUm 2006, n>1561).In the upper groundwater aquifer SO4

2–-concentrations below 120 mg·L−1 are commonly de-fined as natural in Berlin. An exceeding of this concentration indicates geogenic anomaliesor anthropogenic impacts (see Hannappel and Reinhardt, 2002). Results from groundwa-ter monitoring programs, conducted by the Berliner Senat and the Berliner Wasserbetriebe,show clearly elevated SO4

2–-concentrations in the densely populated areas in the inner city(Hannappel and Reinhardt, 2002, see also Fig. 1.8). In total, 23.9% of groundwater sam-ples of numerous wells across the city exhibit concentrations above the threshold value of theTrinkwV (2001) and the European Groundwater Directive (2006), respectively, of 240 mg·L−1

(SenstadtUm 2006, see Fig. 1.9). While in the south-western part of Berlin, the high SO42–-

concentrations mainly result from degradation processes of fens, the ones in the central arearesult from the vast deposition of building rubble. Here concentrations up to 872 mg·L−1 arereported (SenstadtUm 2006). In this area, the source of SO4

2– cannot localized exactly, sincethe rubble composed substrate was spread over a vast area. Unlike, building rubble land-fills can be identified as a point-source for sulfate, causing elevated SO4

2–-concentrations inthe groundwater downstream. Overall, in the inner city, there is a slight decrease of SO4

2–-concentrations over the past ten years, pointing to advanced depletion of readily soluble sul-fates (cf. Hannappel and Reinhardt, 2002). On the contrary, close to building rubble landfillsthere is no clear tendency in SO4

2–-concentration patterns.

Drinking-water in Berlin is mainly supplied by river bank filtration at the river Spree. Hence,the SO4

2–-concentration of the river Spree directly influence the drinking-water quality. ItsSO4

2–-concentration is assumed to increase in the upcoming years due to the open pit minesin the region Lausitz, south of Berlin (Sonntag, 2013). Temporary, SO4

2–-concentration ofthe river Spree up to 285 mg·L−1 were measured in the eastern part of Berlin (RBB, 2015).Among with the high SO4

2–-concentrations in the groundwater, resulting from the buildingrubble, this circumstances exacerbate the problem of elevated sulfate concentrations in thegroundwater, a problem the water authorities have to face.

12


Figure 1.9: Contamination status of the groundwater in Berlin regarding various pollutants; redrawn after Sen-StadtUm (2006), with DL= detection limit; threshold value according to TrinkwV (2001) ; n>1561.

1.6 Objectives of this thesis

The aim of this thesis is to evaluate the (i) contamination status of rubble composed soils and(ii) to assess their potential to release sulfate. Following questions are addressed in the thesis:

• Does building rubble composed substrates contain significantly higher contents of heavymetals and B[a]P compared to substrates free of any notable rubble?

• Which are the main sulfur pools and how high is the quantity of readily soluble sulfatein rubble composed substrates?

• How long will last the sulfate release from rubble composed substrates?

To answer the question of contamination status, we reevaluated available data sets from soilsurveys, which were conducted at various sites, considering rubble composed soils and soilsfree of any rubble. Detailed information about the methods, data sets and the results are givenin Chapter 2.In order to determine the sulfur pools and quantity of readily soluble sulfate of rubble com-posed substrates we conducted various experiments. Thereby, total quantity of sulfur, sulfurspecies and source strength of sulfate of rubble composed substrates and individual compo-nents were analyzed. The results are discussed in Chapter 3.

13


In order to determine sulfate release dynamics and to assess the prospective sulfate releasefrom rubble composed substrates a lysimeter study was conducted. A numerical model wasparameterized to describe the sulfate release dynamics and to predict prospective SO4

2–-concentrations (PHREEQC) . The outcomes are described in Chapter 4.

14

2 HEAVY METALS AND B[A]P IN SOILS FROM C&D Stefan Abel - PhD Thesis

2 Heavy Metals and Benzo(a)pyrene in Soils fromConstruction and Demolition Rubble 1

Abstract

Purpose: Rubble is an important component in urban soils worldwide, especially in Europe.In Berlin, Germany, rubble composed soils cover about 17% of the total city area and 60%of the inner city. This study assesses the contamination status of rubble soil, particularly forheavy metals and benzo[a]pyrene (B[a]P).Methods: The results of 164 soil surveys from Berlin, including more than 2000 analyzedsoil samples of topsoils, rubble subsoils, and natural material have been analyzed for typicalcontamination patterns.Results: The concentrations of all contaminants range over several orders of magnitude andfollow negatively skewed log-normal distribution functions. For rubble containing subsoils aproportion of 34, 71, 67, 68, 74, and 61% of the analyzed samples exceed precautionary valuesof the German Soil Conservation Act, regarding Cd, Pb, Cu, Zn, Hg and B[a]P respectively.Similar results were found for topsoils. A minor part of the soils is contaminated with Cd,while Pb and Hg are the most typical contaminants of rubble material. In contrast to topsoilsand rubble containing subsoils, the majority of the natural subsoil material is not contaminated.Only low to moderate positive correlations were found between the contaminants.Conclusions: Compared to natural soil material, rubble containing soil materials show clearlyelevated concentrations of heavy metals and B[a]P. As the most characteristic contaminantsfor rubble are Pb and Hg, these heavy metals should first be analyzed as proxy contaminants.

2.1 Introduction

Construction and demolition rubble is one of the most common anthropogenic material inurban soils (Bridges, 2009; Meuser, 2010). Rubble composed soils, classified as Urbic Tech-nosols according to the FAO-WRB classification system (IUSS, WRB, 2006; Rossiter, 2007),prevalently contain technogenic artifacts like bricks, mortar and concrete (e.g. Blume andRunge, 1978; Short et al., 1986; Wessolek et al., 2011). Construction and demolition rubblealso contains ash, rubble from infrastructure (asphalt, ballast), slag, chars, wood and glass

1Published: S. Abel, T. Nehls, B. Mekiffer, G. Wessolek; Soils and Sediments, Special issue: SUITMA 7,Springer, 2015, The final publication is available on SpringerLink

15

http://dx.doi.org/10.1007/s11368-014-0959-4


(Blume and Runge, 1978; Hiller and Meuser, 1998a). Disposed rubble material is often pol-luted and contaminates technosols (e.g. Short et al., 1986; Burghardt, 1994; Gras et al., 2000;Shaw et al., 2010). Technogenic components frequently exhibit extreme high heavy metalconcentrations. Particularly for slags and ashes high concentrations are reported (Blume andHellriegel, 1981; Meuser and Blume, 2001; Cloquet et al., 2006). Furthermore, the concen-trations of PAHs can be elevated if e.g. tar based rubble from infrastructure or tarpaper is de-posited (Meuser, 1996b). Raised PAHs concentrations also are related to fire affected rubble(Smettan and Mekiffer, 1996). Besides the disposed technogenic material itself, the proximityto industrial facilities and traffic ways leads to elevated levels of contaminants of technosolsand natural soils in the urban areas (e.g. Manta et al., 2002; Morton-Bermea et al., 2002;El Khalil et al., 2013).Especially in Germany Urbic Technosols are widespread due to bombing raids in the end ofthe World War II and the interrelated emergence of huge amount of rubble (Burghardt, 1994).Meuser (1996b), for instance, found anthropogenic artifacts, particularly construction and de-molition rubble, in 56.3 % of analyzed soil profiles in the city Essen, Germany. At sportsground, parks and allotment gardens the ratio of rubble containing soils can even be higher(Hiller and Meuser, 1998b). Kneib et al. 1990 reported of up to 74.6% tipped soils, frequentlycontaining rubble (Wolff, 1996), in the inner city of Hamburg, Germany.The German Soil Conservation Act (Bundesministerium fur Umwelt, Naturschutz, Bau undReaktorsicherheit, 1998b) defines waste and rubble disposal areas as suspected sites of harm-ful soil changes, assuming contamination due to the deposition of techogenic material. Forrisk assessment and further rehabilitation measures of those sites, soil surveys are frequentlyconducted.Research on urban soil quality mainly concentrates on properties of the topsoil. For instance,topsoils of Berlin are already investigated and documented in detail so far (Fellmer et al.,1993; Birke and Rauch, 1997, 2000). In contrast, there is still a lack of information for sub-soils. In this study we analyze, summarize and discuss the heavy metal and benzo[a]pyrenecontamination patterns of a very large number of rubble containing soils, topsoils, and naturalsubsoils in the urban area of Berlin. We identify characteristic contaminations, calculate theirprobabilities, discuss contamination pathways, and suggest proxy indicators for a faster andcheaper soil contamination assessment, based on the statistical certainty provided by the largenumber of analyzed soil samples.

16


Figure 2.1: Left: Situation in Berlin, Germany 1945; Photo: Bundesarchiv, Bild 183-J31345/CC-BY-SA 3.0(with permission) and right: typical urbic technosols in the inner city of Berlin, Germany. Photo: B.Kluge, TU Berlin.

2.2 Material and Methods

2.2.1 Study area

The geomorphology of Berlin is dominated by a glacial valley in the inner city and moraineplateaus in the north and in the south, which were formed by the Weichselian glaciation.The natural substrates in the inner city are mostly alluvial sands, while soils on the plateausdeveloped from marl and loamy sands.Construction and demolition debris has always been dumped onsite during the city’s 775 yearslong history. However, the rubble depositions after World War II dominate in the soils interms of covered surface area. In 1945, Berlin’s built up area was approximately 72 km2, with28.5 km2 of it affected by bombing raids (Bundesministerium fur Vertriebene, Fluchtlinge undKriegsgeschadigte, 1967). At the end of the war 19.5 % of the buildings were completely orirreparably destroyed (Statistisches Landesamt Berlin, 1963, see also Fig. 2.1). The amountof material from rubble clearance is reported to range from 55 Mio. m3 to 100 Mio. m3

(Fichtner, 1977). The major part of it was tipped into landfills (Keiderling, 1999), the minorpart was deposited throughout the city at non built-up areas, into bomb craters, in parks andbackyards. Today, rubble landfills make up an area of 8.7 km2. During construction activitiesin the younger history, rubble continuously was deposited in the inner city until the early70’s. As a consequence, in the densely populated inner city undisturbed naturally layeredsoils rarely exist. So, today, the area covered with rubble soils (blackish area in Fig. 2.2) or

17


soils partly mixed with rubble (grayish area in Fig. 2.2) is about 133 km2, which is 7 % of thewhole administrative area and almost 60 % of the inner city. A typical profile of a buildingrubble composed soil is shown in Figure 2.1.

Table 2.1: Fractions of technogenic components and natural material in construction and demolition debris layersfrom soils in Berlin, Germany (sites: n=164 , soil samples: n=1771).

Component Fraction [Vol.%] n

bricks 0-70 1625slag 0-60 562mortar 0-50 714ashes 0-50 88ballast 0-50 42concrete 0-20 251charcoal 0-20 50tar 0-15 25glass 0-10 231wood 0-10 52soot 0-2 13natural material 0-86 95

2.3 Data base

Two thousand four hundred eighty-six soil data sets from 164 soil surveys from six of thetwelve district agencies of Berlin were analyzed. The surveys were carried out due to the factthat technogenic materials have been found, which according to the German Soil protectionAct, implies the suspicion of harmful soil changes. Additionally, several surveys were con-ducted as part of building and excavation operations. At the respective sites, the contaminationstatus was analyzed in order to decide on further soil material management like excavation,remediation etc. The data base comprises all types of soils and land use.The soil sampling was predominantly carried out by soil core drilling, only some soil pitswere established. On average, each survey is based on 10 soil core drilling points or soil pitswith three sampling depths. Sampling depths did not necessarily correspond to soil horizonsand samples or mixed samples were taken in steps from 0-0.3 m, 0.3-1 m, 1-2 m, 2-3 m, etc..Most of the surveyors started sampling at 0.3-1 m and ended at 4 m depth, maximum samplingdepth was 10 m. In several surveys the sampling depth of the natural material started onlyat 3 m since it was covered with construction and demolition debris. The number of topsoilsamples is smaller (n=77).From the surveys, we can distinguish between the following three substrate classes: Mate-

18


Figure 2.2: Distribution of rubble containing soils in Berlin, Germany, for the whole city and for the inner city,encircled by the city train ring. Grayish areas demark soils partly filled up with rubble, while blackishareas demark soils predominantly filled up with rubble. Modified map from Berliner Umweltatlas2008.

19


rial from the topsoil layer (TSM), material from rubble containing layers (CDM) and subsoilscomposed of natural soil material only (NSM). The TSM was mainly sampled from 0-0.3 m.The topsoils have either developed by pedogenesis or they were applied as artificial layers.Some topsoils were anthropogenically changed e.g. by addition and mixing of materials suchas compost or sand. The CDM contains appreciable amounts of rubble, with unclear origin. Itcan include old building remnants, rubble from the World War II as well as younger construc-tion and demolition debris. In approximately 90 % of the surveys the layers deeper than 0.3 mwere completely filled up with anthropogenic material. This also includes strongly anthro-pogenically influenced natural materials as e.g. crushed stones used as construction materialor sand which was used as building additive in mortars. The NSM includes all natural lay-ered subsoils or relocated natural soil materials, free of any notable rubble. The predominantnatural substrates documented in the surveys are sands. Other substrates, e.g. developed frommarl and loam, are not considered as individual substrate classes in the following, since theirfraction is less than 5% of the examined soil samples (n=21 and n=6, respectively).

2.3.1 Methods

All TSM, NSM, and CDM samples were analyzed without any spatial reference due to privacyprotection. The according data sets can belong to same or different sites, thus sample numbersof the three substrates are not equal. In the evaluation we did not distinguish between differ-ent sampling depths for CDMs and NSMs. We only included data of surveys, in which soilsubstrates and technogenic components are explicitly listed. The ratio and the compositionof the single anthropogenic components were mostly estimated by visual criteria (Tab. 2.1).Second, only datasets were included, for which chemical analyses complied with ISO stan-dards. The respective analytical methods are given in Table 2.2. The heavy metals and B[a]Pare discussed in terms of total amounts.

The statistical analyses of the data set included the distribution function of the mentionedchemical parameters. The correlations between single contaminants were determined usingthe Spearman’s rank correlation coefficient (Spearman, 1910), whereby concentrations smallerthan the detection limits have been excluded from correlation analysis. The distributions ofthe concentrations and correlations of the contaminants Cd, Pb, Cu, Hg, Zn and B[a]P wereanalyzed for CDM and NSM only, since the number of topsoil data is insufficient. The sig-nificance of the contamination differences for the three substrate classes was tested by theKruskal-Wallis-Test for p = 0.05 (Kruskal and Wallis, 1952). Same statistical test was ap-plied to evaluate whether the most common technogenic components as bricks, mortar, slag

20


Table 2.2: Analytical methods (ISO) and their detection limits (DL) for the chemical parameters as applied inthe soil surveys.

Parameter Methods DL[mg kg−1]

EC ISO 11265 -pH ISO 10390 -Pb ISO 11047 0.03Hg ISO 16772 0.006Ni ISO 11047 0.01As ISO 11047 0.02Cr ISO 11047 0.03Cd ISO 11047 0.01Cu ISO 11047 0.007Zn ISO 11047 0.01BaP ISO 13877 or 0.001

Soxhlet-Extraction

and ashes in the CDM samples differ significantly from those CDM without these respectivecomponents. For the estimation of soil contamination by heavy metals and B[a]P according tothe geochemical background, we used the geochemical load index GLIi; (Muller, 1979) whichis defined by

GLIi = log2Ci

B ·1.5(1)

with: Ci the concentration of an element in the soil [ mg kg−1] and B its background concentra-tion in the corresponding region or country. For Ci we applied the median of the correspondingcontaminant distribution, in order to characterize a distribution instead of a single value. Thebackground concentrations for subsoils in Germany are given for different texture classes byLABO 2003. Since sandy soils are predominant in the evaluated soil surveys (>95%) the con-centrations for sandy soils were applied. These background values represent the 90% quartileof subsoils from up to 608 analyzed soil profiles without any regionalization and are reportedfor the discussed contaminants except for As and B[a]P. Thus, for As and B[a]P backgroundvalues for topsoils under agricultural use were applied. The contamination classes accordingto the GLI are given in Table 2.3. GLI values smaller than 0 indicate a non-polluted or even adepleted soil.

21


Table 2.3: Classification of contamination according to the GLIi (Muller, 1979).

GLIi level of contamination

<0 depleted to non polluted0-1 non to slightly polluted1-2 slightly to moderately polluted2-3 moderately polluted3-4 moderately to highly polluted4-5 highly polluted>5 very highly polluted

2.4 Results and Discussion

2.4.1 Soil reaction and electrical conductivity

The pH values follow a normal distribution function. The pH values range from 6.2 to 9.3 forTSMs, from 3.5 to 12.6 for CDMs and from 4.4 to 13.0 for NSMs (Tab. 2.4). A portion of 91 %of the CDMs and 84 % of the NSMs are at least slightly alkaline. According to Gerstenberget al. 2005, the pH of natural top- and subsoils in the area of Berlin ranges from slightly tostrongly acidic. This, however, cannot be the case for subsoils of the northern moraine plateausas these developed from marl (lime and loam). Therefore, we interpret most of the pH in theslightly alkaline and alkaline range of NSM as geogenic or as the result of secondary limingdue to carbonate rich ground- and seepage waters (cf. Jim, 1998). Apart from the NSMs, theanalyzed soils exhibit elevated pH values, which indicate a strong anthropogenic impact. Thealkaline soil reaction of the TSM might be the result of deposition of alkaline urban dust (cf.Anttila, 1990) or mixing with underlying CDM. The alkaline characteristics of the CDMs canbe mainly attributed to carbonate-rich components such as mortar, slags and ashes (Meuser,1993; Burghardt, 1994; Bridges, 2009). Schleuss et al. 1998 reported of a ratio of carbonatesin Urbic Technosols of up to 14 % (w/w). The high amount of carbonates in CDMs constitutea long time buffer capacity, which persists over several hundreds of years (Hiller, 1996) andwhich will decrease the bio-availability of heavy metals (e.g. Chuan et al., 1996; Ge et al.,2000).The electrical conductivity (EC) follows a negatively skewed log-normal distribution functionfor CDMs and NSMs. The CDMs and the NSMs show extreme values of up to 9292 µScm−1

and 8020 µScm−1, respectively (Tab. 2.4). High values can be attributed to the presence ofashes in the soil (cf. Nagamori et al., 2007) or due to other anthropogenic inputs, such as

22


fertilizers or deicing salts (cf. Helmreich et al., 2010).

2.4.2 Cadmium

The Cd concentrations exhibit a negatively-skewed log-normal distribution for the NSM andCDMs (Fig. 2.3). A ratio of 51, 40, and 27 % of TSMs, CDMs, NSMs, respectively, areabove background concentration (see Tab. 2.4). The precautionary value (0.40 mg kg−1) isexceeded for 47 % of the TSMs, 34 % of CDMs, and 21 % of the NSMs. As it can be ex-pected, the examined NSMs show significantly lower Cd concentrations than CDMs (n=735;p<0.05). Between TSMs and the other two materials, no significant difference could be found,resulting from the high variability of TSMs. This finding also applies for the other contami-nants. Within the CDM, the samples containing slag and ashes show higher medians (0.28 and0.07 mg kg−1) compared to those without the respective component (0.01 and 0.04 mg kg−1,respectively). These differences are significant for slag and ash containing CDM. Neverthe-less, significant correlations between fractions of technogenic materials and Cd concentrationscould not be found. For brick and mortar, no effects on Cd contamination could be identified.The same applies for the other analyzed contaminants and thus it is not discussed in the fol-lowing.The results indicate, that higher layers are more contaminated than deeper layers, thus a con-tamination could have taken place from above after the materials have been deposited, e.g.from atmospheric deposition (Alloway and Steinnes, 1999). The median Cd concentrationis 0.34 mg kg−1 for TSMs, which is nearly identical to topsoil samples taken by Birke et al.1994 as they found a median of 0.32 mg kg−1 for the densely populated residential areas inBerlin. The elevated concentrations in CDMs might be the result of a secondary contami-nation but are partly a characteristic of the rubble itself. Cd contaminations can arise fromcadmium-based paint (cf. Blume and Hellriegel, 1981), the existence of metal works slags(cf. Meuser and Blume, 2001), or ashes. Ashes from coal combustion show concentrations of10 to 60 mg kg−1 and have been identified to contribute to the elevated Cd concentrations inthe Berlin Urbic Technosols (Blume and Hellriegel, 1981). According to the GLICd , which is<0, none of the three substrate classes show a typical contamination with Cd (Tab. 2.4). ForCDM just in few cases elevated concentrations were observed. Thus, building rubble is not acharacteristic factor enriching Cd.

23


2.4.3 Lead

The Pb concentrations exhibit a negatively-skewed log-normal distribution for the two con-sidered substrates CDM and NSM (Fig. 2.4). The concentrations range over several orders ofmagnitudes up to 14093.0 mg kg−1 for CDM. The background concentration of 12.0 mg kg−1

(LABO, 2003) is exceeded by 91 % of CDM but also of 45 % of the NSMs. The majorityof CDM (71 %) and the TSM (68 %) have Pb concentrations higher then the precautionaryvalue of 40 mg kg−1. In contrast, only 27 % of the NSMs show concentrations higher thanthe precautionary value. Again, the contamination of CDMs and NSMs is significantly dif-ferent (n=773; p<0.05). Birek et al. 2000 documented a median of 109.0 mg kg−1 Pb inBerlin topsoils, which is double of the median (55.2 mg kg−1) for TSM in our data base. Aswe have analyzed soils which have partly been disturbed and engineered and which probablyhave constructed top soils from hopefully high quality raw material a low contamination statuscan be at least expected compared to old urban soils, which have been included in Birke andRauch’s study. In contrast, in old or especially exposed urban soils, Pb contamination mightbe linked to direct industrial contamination or atmospheric deposition from industry, vehicleemissions (c.f. Galloway et al., 1982; Nriagu and Pacyna, 1988; Blume and Hellriegel, 1981)and street dust (c.f. Miguel et al., 1997). For CDMs, the mean concentration (219.5 mg kg−1)is in a comparable magnitude as the mean of 395.0 mg kg−1 for subsoils, reported by Smettanet al. 1993. The source of lead in CDMs containing soils can be linked to a variety of rubble-typical sources: lead-based paints (cf. Jacobs et al., 2002; Howard et al., 2013), leadlights andceramic glaze (Blume and Hellriegel, 1981), slag components, which can lead to concentra-tions higher than 10 000 mg kg−1 (cf. Cloquet et al., 2006) and ashes from coal combustion,with concentrations ranging from 1000 to 20 000 mg kg−1 (Blume and Hellriegel, 1981). Thespecial contribution of slag and ash to the contamination can be confirmed by our results. Asfor Cd, slag and ash containing CDM show significantly higher medians than CDM withoutthem (120.0/172.0 and 70.5/82.5 mg kg−1, respectively). Further corroded or smelted waterpipes (Smettan et al., 1993) or the deposition of lead batteries (cf. Pichtel et al., 2000) maycontribute to elevated concentrations. Bricks and mortar, which are the most common com-ponent in the documented urban soils, show rather low to moderate lead concentrations of14 to 103 and 310 mg kg−1, respectively (Meuser, 1996a; Nehls et al., 2013). The calculatedGLIPb indicates that the majority of CDMs are moderately polluted (Tab. 2.4). The majority ofNSMs feature no enrichment of Pb in comparison to the geogenic background concentrations(GLIPb<0).

24


2.4.4 Copper

The Cu concentrations follow a negatively-skewed log-normal distribution for the NSMs andCDMs (Fig. 2.3). 91 % of CDMs and 47 % of the NSMs show concentrations higher than thebackground value of 7 mg kg−1 (LABO, 2003). The CDMs and TSMs show the highest con-centration medians and more than 67 % of them exceed the precautionary value of 20 mg kg−1,but only 28 % of the NSMs. CDMs are significantly higher contaminated with Cu than theNSMs (n=631; p<0.05). Again, slags and ashes could be identified as a crucial source forcontamination. The CDMs containing these components show significantly higher medians(57.4 and 59.5 mg kg−1) than those without (24.3 and 29.0 mg kg−1, respectively). The cop-per concentrations with a median of 28.7 mg kg−1 of the topsoil samples are in a comparablemagnitude as the results of Birke et al. 2000, reporting of a median of 37.0 mg kg−1. Regard-ing CDM several sources potentially induce elevated Cu concentrations: wood preservativescan lead to high concentrations (Bergback et al., 2001), slag, a frequently found componentin rubble, possesses elevated Cu concentrations (Meuser, 1996a) as well as pipes, cables andother electric installations. For the components bricks and mortar Blume 1978 documentedmoderate Cu concentrations of 30 mg kg−1. Besides the anthropogenic artifacts in the soil,atmospheric deposition and vehicle emission might particularly raise the Cu concentration ofthe topsoils (Nriagu and Pacyna, 1988; Davis et al., 2001). On average, the CDMs show aslight to moderate pollution level (Tab. 2.4).

2.4.5 Zinc

The Zn concentrations for NSMs and CDMs exhibit a negatively skewed log-normal distri-bution (Fig. 2.3). The concentration of CDM range over several orders of magnitudes with amaximum of 14 000 mg kg−1. The median of topsoil (87.0 mg kg−1) and CDM (99.4 mg kg−1)is approximately three and four times higher than the background value (24 mg kg−1). The ma-jority of analyzed TSMs and CDMs exceed precautionary value. In contrast, the median of theNSMs is below the background and precautionary value. Meuser 1996a reported even higherlevels of Zn for soils filled up with rubble. He documented a mean of 610 mg kg−1, whichis more than two times higher than the mean of 282 mg kg−1 for the CDMs in this study. Asfor the other mentioned heavy metals the NSMs have significantly lower Zn concentrationsthan the CDMs (n=471; p<0.05). Sources of zinc in the analyzed CDMs are presumablygalvanized goods as gutters, roof parts and kitchen tools, paints (cf. Bergback et al., 2001),slags (Shaw et al., 2010), ashes and the bricks in the rubble itself. Bricks show a moderate

25


Figure 2.3: Cumulative distribution of heavy metals and B(a)P: left: parent subsoils (NPM), right: constructionand demolition rubble containing subsoil (CDM); roughly dotted line indicates background values,finely dotted line indicates precautionary value (p.v.); x-axis starts with value of detection limit.

26


Figure 2.4: Cumulative distribution of Pb and Hg; left: parent subsoils (NPM), right: construction and demoli-tion rubble containing subsoil (CDM); roughly dotted line indicates background value, finely dottedline indicates precautionary value; x-axis starts with value of detection limit.

Zn concentration above the precautionary value. With 127 and 126 mg kg−1, Meuser 1996aand Shaw et al. 2010 mentioned quite similar results of Zn in bricks. Nehls et al. 2013found concentrations ranging from 85 to 171 mg kg−1. Our results could again identify a sig-nificant influence of slags and ashes, indicated by higher medians of those CDMs containingthese components (176.0/262.5 compared to 90.0/110.0 mg kg−1 for CDMs without slags andashes, respectively). Since the data base also includes industrial and traffic sites, emissions,tire and brake pad abrasions can be additional sources for zinc in the soil, particularly in thetopsoil (cf. Davis et al., 2001; Councell et al., 2004). At least half of the topsoil and the CDMsamples show a slight to moderate pollution according to the GLIZn (Tab. 2.4). This indicatesthat, CDM is responsible for an enrichment of Zn, since the majority of NSMs show no higherconcentrations than the geogenic background.

27


2.4.6 Other Heavy Metals

As the heavy metals discussed above, concentrations of Ni, Cr, Hg and As follow a negatively-skewed log-normal distribution function for CDMs and NSMs. For both substrates Ni, Cr,and As exhibits median concentrations below precautionary value and even below the geolog-ical background value (Tab. 2.4). Thereby, the geological background of these trace metals(90 % percentile) for sandy subsoils are given by LABO 2003 without any regional or land-use specification. The findings indicate that there is no relevant contamination linked with thedeposition of rubble.

Hg concentrations show a different pattern. A portion of 37 % and 74 % exceeds precau-tionary value regarding NSMs and CDMs, respectively (Fig. 2.4). Also TSMs show elevatedconcentrations of Hg with a median of 0.24 mg kg−1. Again, the CDMs differ significantlyfrom the NSMs concerning all four heavy metals. As for the aforementioned metals, slag andashes could be identified as components responsible for elevated Hg levels. The correspond-ing substrates with slag and ashes show significantly higher medians (0.41 and 0.39 mg kg−1,respectively) than those without (0.24 mg kg−1 for both).Birke et al. 2000, mentioning a median of 0.34 mg kg−1 in the high density residential areas,ascribe the contamination to industrial land-use, but they also found high concentrations inassociation with dumped rubble. The source of Hg in those rubble composed soils might beattributed to mercury lamps or electrical inventory (cf. Rodrigues et al., 2006; Cheng and Hu,2011).

28

2H

EAV

YM

ETA

LS

AN

DB

[A]P

INSO

ILS

FRO

MC

&D

StefanA

bel-PhDT

hesis

Table 2.4: General properties, statistical parameters of contaminant concentrations and GLIi index

pH EC Cd Pb Cu Zn B[a]P Ni Cr Hg As- µS cm−1 mg kg−1 mg kg−1 mg kg−1 mg kg−1 µg kg−1 mg kg−1 mg kg−1 mg kg−1 mg kg−1

Background value - - 0.24 12 7 24 46∗ 9.4 15 0.05 4.0∗

Precautionary value - - 0.40 40 20 60 300† 15.0 30 0.10 25 ‡

TSMn 53 30 71 72 44 22 33 62 67 32 61Arithmetic mean 7.5 287 0.66 103.1 46.3 101.5 440 12.2 23 0.31 5.2Median 7.3 300 0.34 55.2 28.7 87.0 289 10.3 14 0.235 4.175 % percentile 8.2 398 0.97 81.0 40.5 117.5 613 15.9 26 0.41 7.190 % percentile 8.4 431 1.90 188.4 59.7 186.8 1036 20.9 34 0.83 11.0Max. 9.3 742 4.10 1080 527.5 449.0 2100 38.4 297 1.01 33.1GLIi (median) - - <0 1.6 1.5 1.3 2.1 <0 <0 1.6 <0

NSMn 693 350 735 773 631 471 639 617 719 708 743arithmetic mean 7.9 209 0.36 46.7 18.1 60.0 738 4.4 10 0.23 4.6Median 8.0 85 0.01 9.0 6.0 12.0 100 2.100 3 0.01 1.175 % percentile 8.4 180 0.30 46.0 22.4 0.23 230 6.7 9 0.21 2.890 % percentile 8.8 421 0.97 102.8 44.1 47.9 727 11.6 17 0.58 5.6Max. 13.0 8 020 16.2 2694 718 2284 108 000 88.2 1 690 7.00 848GLIi (median) - - <0 <0 <0 <0 <0 <0 <0 <0 <0

CDMn 1397 957 1600 1641 1328 1011 1422 1335 1544 1427 1591arithmetic mean 8.1 392 0.65 219.5 119.0 282.2 2668 14.9 15 1.56 6.7Median 8.1 189 0.04 77.2 28.4 99.4 496 8.9 10 0.29 3.675 % percentile 8.5 404 0.67 169.0 65.0 223.0 1549 14.00 17 0.65 6.290 % percentile 8.9 989 1.50 422.0 148.3 583.0 4371 30.3 30 1.60 12.0Max. 12.6 9292 25.30 14 093 25 000 14 000 460 000 428 466 1100 619GLIi (median) - - <0 2.1 1.4 1.5 2.8 <0 <0 2.0 <0

TSM: topsoils, NSM: natural soil material, CDM: construction and demolition material containing soil; ∗ background value B[a]Pand As for topsoils under agricultural use; other background values for sandy subsoils (LABO, 2003); precautionary value of the

German Soil Protection Act for sandy soils, † sandy soils with organic matter content <8 % and for ‡ playgrounds.

29


2.4.7 Benzo[a]pyrene

B[a]P is frequently used as a proxy substance for the contamination with PAHs as it is oneof the most hazardous and carcinogenic of the PAHs (e.g. Collins et al., 1991). The concen-trations of B[a]P can be described by a negatively skewed log-normal distribution (Fig. 2.3).The majority of the CDMs and the TSMs exceed background (46 µg kg−1) and precautionaryvalues (300 µg kg−1). In contrast only 21 % of the NSMs exceed the precautionary value.The concentrations of B[a]P in the NSMs are significantly lower than that of CDMs (n=639;p<0.05). Slag significantly contributes to soil pollution, as indicated by a higher median thanCDMs without (910 and 360 µg kg−1). For ashes this could not be confirmed statistically. Thehigh B[a]P contamination rate of CDM could be attributed to the presence of material fromWorld War II. These materials frequently contain high amounts of ashes, slags, and incom-pletely burned, smoldered components (black carbon) (cf. Smettan and Mekiffer, 1996; Graset al., 2000). The fire spread during the bombing raids and the related combustion of organicmaterials is the main source of B[a]P and other PAHs in rubble from the World War II. Ruoko-jarvi et al. 2000 and Wobst et al. 1999 documented high PAHs concentration in the soot and atsurfaces of building inventories, respectively, after fire. Tar based components can also serveas sources of PAHs. Meuser 1996b found total B[a]P concentrations of up to 148 000 µg kg−1

in such materials, which are frequently found in the CDMs and which can make up a ratio to15 Vol.%, according to our data base. Additionally, elevated levels of B[a]P and PAH can beattributed to recent atmospheric deposition (Baek et al., 1991). The concentration in TSMsmight further be elevated due to urban dust inputs from traffic, which can exhibit concentra-tions of up to 2200 µg kg−1 (Essumang et al., 2011). Overall, the majority of the TSMs showa moderate, the majority of CDM show a moderate to high B[a]P pollution level according tothe GLIB[a]P (Tab. 2.4). It has to be considered that we applied the background concentrationfor topsoils under agricultural use for calculating the GLIB[a]P. Subsoils can be expected tocontain B[a]P at lower levels (e.g. Krauss et al., 2000). However, data for German subsoilsare not available. This might have led to an underestimation of the GLIB[a]P and the fractionexceeding background concentration.

2.4.8 Correlations between the contaminants and identification of repre-sentatives

In general we found low to moderate correlations between Cd, Pb, Cu, Hg, Zn and B[a]P(Tab. 2.5). Between the pH and the contaminants there is no meaningful correlation (coef-

30


Table 2.5: Spearman’s rank correlation coefficient between contaminants; ∗statistically significant (p<0.05), †

highly significant (p<0.01); NSM: natural soil material, CDM: construction and demolition materialcontaining soil.

Cd Cu Zn Hg B[a]P

NSMPb 0.36† 0.55† 0.56† 0.62† 0.60†

Cd 0.40† 0.46† 0.25† 0.32∗

Cu 0.84† 0.53† 0.57†

Zn 0.63† 0.58†

Hg 0.32†

CDMPb 0.40† 0.37† 0.36† 0.19† 0.40†

Cd 0.45† 0.49† 0.26† 0.33†

Cu 0.67† 0.49† 0.36†

Zn 0.37† 0.41†

Hg 0.25†

ficient <0.16). A strong correlation was found between Cu and Zn in NSMs. Besides forCd, the correlations between the contaminants are stronger for the NSMs as for the CDMs.This might be attributed to the higher heterogeneity of the CDM, with its various technogeniccomponents in differing compositions.

Due to the low correlation between the contaminants, it is difficult to determine a single rep-resentative. Soil surveys, supporting decision making for further soil material management orremediation, should first test Pb and Hg concentrations, since these are the most characteristicones for rubble. Additionally, Pb and Hg show the highest concentrations and impose a suf-ficient health risk to imply remediation measures. Only if Pb or Hg concentrations are belowthe precautionary value, other contaminants have to be additionally analyzed.

2.5 Conclusion and Outlook

Our study shows clearly elevated concentrations of heavy metals and B[a]P in rubble soilscompared to the natural subsoil material. Thereby, the majority of these rubble soils are mod-erately to highly contaminated with Pb, Cu, Zn, Hg and B[a]P, while Cd, Ni, As and Cr isa minor concern. A high heterogeneity in composition and amounts of technogenic compo-nents leads to highly variable contaminant concentrations. Nevertheless, slag and ashes couldbe identified as crucial sources for elevated concentrations. The depth profiles of the con-taminants’ concentrations additionally suggest anthropogenic pollution after rubble materials

31


were deposited. Pb and Hg are the most pronounced contaminants for rubble soils. We suggestanalyzing them for a fast contamination assessment of the corresponding sites and further de-cision support for potentially required remediation. Only if Pb or Hg concentrations are belowthe precautionary value, analysis of further contaminants is necessary in order to assess thepollution status of the site and to decide whether remediation measures have to be conducted.

Although the area of Berlin is covered by contaminated Urbic Technosols to a large extend,groundwater monitoring does generally not show elevated heavy metal concentrations in theupper groundwater aquifer. However, recent studies have shown, that rubble depositions posea serious threat to the groundwater quality regarding sulfate.

32

3 POOLS OF SULFUR IN URBAN RUBBLE SOILS Stefan Abel - PhD Thesis

3 Pools of sulfur in urban rubble soils 2

Abstract

Purpose: Elevated concentrations of sulfate in groundwater are increasingly becoming a prob-lem in several European cities. Building rubble from the World War II is assumed to be a majorsource of sulfate. This study characterizes pools of sulfur in rubble composed technosols, andassesses their potential to release sulfate.Methods: Six urban soil profiles have been analyzed. Fractions of the main technogenic com-ponents in the skeleton fractions were determined by hand sorting approximately 100 kg ofmaterial. Total sulfur and water soluble sulfate were determined. Microplate-scale fluoromet-ric assays were applied to measure the depth-dependent enzyme activity of arylsulfatase. Themineral composition of soil samples was analyzed using powder X-ray diffractometry. Bind-ing forms of sulfur were determined using X-ray absorption near-edge structure spectroscopy.Results: The maximum total content of sulfur is 4.6 g·kg−1; that of readily soluble sulfur is2.3 g·kg−1. Both gypsum and traces of barite and ettringite were detected in some fine soiland component samples. Samples taken from deeper soil depths exhibited higher total sulfurand soluble sulfate contents. The depth profiles of sulfur and the activity of arylsulfatase sug-gest advanced leaching of inorganic sulfates from the upper horizons. Hence sulfur is mainlyorganically bound in the topsoil. In the subsoil, however, sulfates make up about 90% of totalsulfur, approximately 30% of which is readily soluble.Conclusions: The sulfur pool of rubble composed soils differs completely from natural soils.This is particularly the case for subsoils, in which high contents of sulfur are readily solu-ble. This suggests that sulfate minerals such as gypsum predominate. Urbic technosols cantherefore be assumed to be one of the main sources of sulfates in urban groundwater.

3.1 Introduction

Construction and demolition rubble is one of the most common anthropogenic materials inurban soils. In Europe in particular, where numerous cities were extensively destroyed inbombing raids during World War II, rubble composed soils, classified as urbic technosols,cover vast areas of urban regions (Abel et al., 2015). Enormous quantities of building rubble

2Published: S. Abel, T. Nehls, B. Mekiffer, M. Mathes, J. Thieme, G. Wessolek; Soils and Sediments, Springer,2015; The final publication is available on SpringerLink

33

http://dx.doi.org/10.1007/s11368-014-1014-1


were also deposited in specific landfills. These rubble soils are mainly composed of techno-genic components such as brick, mortar, plaster and slag (Wessolek et al., 2011). Recentstudies in Germany suggest that building rubble is one of the major sources of sulfate inthe groundwater of urban areas, e.g. Berlin (Hannappel and Reinhardt, 2002; Schonsky et al.,2013), Hamburg (Amt fur Umweltschutz, Hamburg, 2000) and Dresden (Grischek et al., 1996;Umweltamt Dresden, 2010). Sulfate concentrations in the upper groundwater aquifer gener-ally exceed the EU drinking-water standards of 240 mg·L−1 in these cities. Concentrations ofup to 872 mg·L−1 have been reported in the vicinity of rubble landfills in Berlin (Hannappelet al., 2003); background concentrations do not exceed 100 mg·L−1 (Hannappel and Jakobs,2002). One key pool of readily soluble sulfate could be gypsum.

Table 3.1: Skeleton (>2 mm) and fractions of coarse grain technogenic components.

Soil layer Skeleton Brick Mortar & Slag OtherPlaster

kg·kg−1

Fritz-Schloss-ParkAh 0.02 0.01 0.01 0.00 0.01yC1 0.42 0.13 0.09 0.01 0.07yC2 0.50 0.30 0.13 0.01 0.15yC3 0.52 0.28 0.17 0.01 0.06

TeufelsbergAh 0.05 0.02 0.01 0.00 0.02yC1 0.09 0.02 0.04 0.02 0.04yC2 0.42 0.19 0.11 0.03 0.11yC3 0.47 0.25 0.12 0.04 0.06

Martinez et al. 2012 reported gypsum contents of 1 to 10 g·kg−1 and a total sulfur (Stot) con-tent of 0.4 g·kg−1 for fresh building waste. (Jang and Townsend, 2001) determined a fractionof gypsum ranging from 15 to as much as 91 g·kg−1 in new C&D waste, which commonlycontains gypsum drywalls. In such rubble composed materials, the leachate may contain con-centrations of sulfate in the region of gypsum solubility (Jang and Townsend, 2001, 2003).Gypsum used to be employed as a binder, particularly for decorative coatings, which is why itmay occur in rubble composed soils deposited after World War II. Alaily et al. 1986 identifiedsingle gypsum particles in rubble composed soils, and assumed its fraction to be fairly high.However, no measurements were conducted. The sulfur pool in single technogenic compo-nents such as slag and brick are reported to be a maximum of 40 g·kg−1 and 2 g·kg−1 (Strayerand Davis, 1983; Nehls et al., 2013, respectively). Mineral soils were investigated in detail,

34


unlike urbic technosols for which there is still a lack of information about sulfur pools. Theirtypical sulfur contents range between 0.1 and 0.6 g·kg−1(Brown, 1982). The sulfur stock ofboth organic and mineral soils in the temperate climate zone occurs mainly in organic matter;the occurrence of gypsum is very rare. The content of organic sulfur decreases rapidly asthe soil depth increases. Freney et al. 1966 and Tabatabai et al. 1972b reported that organicsulfur accounts for up to 93% of Stot in surface soils and up to 99% of Stot in subsoils. Sulfurspeciation, particularly for organic sulfur, was thoroughly investigated for natural soils (e.g.Solomon et al., 2003; Prietzel et al., 2007, 2010). In these studies, X-ray absorption near-edgestructure spectra (XANES) was shown to be a suitable tool for identifying sulfur oxidationstates and binding forms. Further information about this method, particularly regarding thecharacterization of sulfur binding forms, can be found in Vairavamurthy 1998 and Prietzel etal. 2003. Sulfonate, ester sulfates and sulfides are the main binding forms within the organicsulfur fraction (Autry and Fitzgerald, 1990; Xia et al., 1998). Activity of the extracellularenzyme arylsulfatase is frequently used as a sign of organically bound sulfate (e.g. Tabatabaiand Bremner, 1970). High correlations between total sulfur, soil depth and activity have beendocumented accordingly by Speir et al. 1980 and Lorenz et al. 2005. Organic forms of sulfurfeature a relatively low ratio of soluble sulfate (Stanko-Golden et al., 1994), resulting in lowsulfate concentrations in the soil solution in the range of several mg·L−1 (Scherer, 2009). Thefraction of readily soluble sulfate is between 1 and 10% of Stot in mineral soils (Brown, 1982).There is still a lack of comprehensive information regarding sulfur pools in urbic technosols.The aim of our study is to characterize (i) the typical depth profile of sulfur, (ii) readily solublesulfate patterns, (iii) sulfur-bearing minerals, and (iv) the speciations of sulfur in these rub-ble composed technosols. Thus, the source function of rubble soils for sulfate in urban waterresources will be assessed.

3.2 Materials and methods

3.2.1 Sampling and sample preparation

Two typical rubble dump sites in Berlin were sampled. The first site, Teufelsberg, was createdbetween 1950 and 1972. It contains approximately 26.2 Mio. m3 rubble, originating mainlyfrom World War II damage. Soil sampling was conducted from four soil profiles along atransect downwards from a 60 m high heap. The second sampling site, Fritz-Schloss-Park,containing approximately 1 Mio. m3 rubble, also originating from buildings destroyed afterWorld War II. Waste was deposited there from 1949 to 1953. Soil material was taken from

35


two soil profiles at the top of the heap. Both sites were covered with an allochthonous Ahmaterial following the closure of the dump sites. Typical soil profiles from these sites areshown in Figure 3.1. Both sampling sites are vegetated with herbaceous plants.Soils were sampled from several depths. These included the Ah horizon (maximum depth of0.2 m); soil layers at a maximum depth of 0.3 m (yC1) with appreciable quantities of rubble andyet a moderate to high degree of penetration by roots; soil layers at a maximum depth of 0.5 m(yC2) where there is little root penetration; and layers at a maximum depth of 1.2 m (yC3)where root penetration is negligible. It was not possible to distinguish yC2 for some profiles.In order to determine the enzyme activity of arylsulfatase, composite samples of moderately tohighly root penetrated Ah and yC1 layers and individual samples of yC2 and yC3 were takenfrom Teufelsberg. These samples were immediately cooled with liquid nitrogen. Fractionsof the main technogenic components in the skeleton (>2.0 mm) were determined by handsorting approximately 100 kg air dried material. The technogenic components were clearlydistinguishable, with the exception of mortar and plaster of Paris. In this case, the componentswere differentiated using 10% HCl solution. Sub-samples were used for the fraction between2.0 mm and 6.3 mm. To calculate their fraction (kg·kg−1), water contents were analyzed usingrepresentative sub-samples of respective technogenic components.

3.2.2 Methods

Total sulfur contents (Stot) and total carbon contents (Ctot) of fine soil samples (<2.0 mm)and the main technogenic components were determined by dry combustion and coupled gas-chromatography according to ISO 15178 (Element Analyzer Vario EL III, Elementar Anal-ysensysteme GmbH). The depth profile of (Stot) for two soil profiles from Teufelsberg andFritz-Schloss-Park were analyzed in detail. All soil samples and technogenic componentswere ground prior to analysis. In order to analyze soluble sulfate of the fine soil and the maintechnogenic components, we applied batch experiments (solid:liquid ratio 1:4 and 1:10) for24 h in an overhead shaker. All samples were ground to <2.0 mm beforehand. Due to thepotential occlusion of sulfate-bearing minerals within the slag fraction, this component wasfurther ground in a ball mill. The dissolved sulfate content was analyzed by ion chromatog-raphy according to ISO 10304, using a Dionex DX-120 Ion Chromatograph. The pH solutionwas measured using Knick pH-Meter 761.In order to study the distribution of organic sulfate in a soil profile at Teufelsberg, we appliedan enzyme activity test of arylsulfatase according to ISO 22939 (Marx et al., 2001). To thisend, the cooled samples were mixed with deionized water in the laboratory (ratio 1:10). The

36


Figure 3.1: Typical profiles of rubble composed soils at the Fritz-Schloss-Park (left) and the Teufelsberg (right)Berlin, Germany.

37


fluorescence dye used was 4-methylumbelliferone-sulfate (MUF). The pH solution was ad-justed to 6.1, adding 2-(N-morpholino)ethanesulfonic acid (MES) as a buffer. Samples werehomogenized using an ultrasonic bar and placed into micro-well plates. Tests were conductedon four replicates of each sample. Fluorescence was measured after an incubation time of 3 hat 30◦ C at wavelengths of 355-365 and 450-460 nm.We characterized sulfur-bearing minerals from fine soil and technogenic components in thecoarse grain fraction by powder X-ray diffractometry (Philips diffractometer PW1050, An-ode material Cu). The diffracted rays were recorded at angular steps of 0.02◦(ranging from0◦to 80◦). All samples were ground beforehand in a ball mill. The peak patterns wereevaluated and minerals identified using the XPert Highscore software (Philips, the Nether-lands, 2001). In order to test findings of minerals, we analyzed the elemental composi-tion of the samples by X-ray fluorescence spectroscopy according to EN 15309 (Philips PW2400). In order to analyze the sulfur speciation in the soil samples from Teufelsberg, weemployed XANES at the K absorption edge of sulfur (White Line energy of different bind-ing forms ranging from approximately 2468 to 2483 eV). Measurements were carried out atbeamline 8 at the SLRI (Synchrotron Light Research Institute) in Thailand and at KMC-1 atBESSY (Berliner Elektronenspeicherring-Gesellschaft fur Synchrotronstrahlung m.b.H.) II atHelmholtz-Zentrum Berlin, Germany. Prior to analysis, coarse grain fractions and soil wereground to grain sizes <2 mm and mixed according to their original fractions. In addition,the most common technogenic components in the soil (brick, mortar and slag) were analyzedseparately. The peak areas of the spectra were analyzed using Athena software (Ravel andNewville, 2005). Standard substances with different sulfur oxidation states and different sul-fur compounds of the same oxidation state were used as references for peak positions in thespectra. Fractions for the different S oxidation states were calculated by comparing the respec-tive peak area with the sum of all peak areas within this spectrum (c.f. Manceau and Nagy,2012).Since the blend of mortar and plaster features comparable input materials, these componentswill be discussed together in this paper.

38


3.3 Results and discussion

3.3.1 Composition of soils

Both sites feature a high fraction of skeleton in the subsoils (Tab. 3.1). In addition to the yC1

layer, the magnitude of these fractions is comparable. The most common technogenic com-ponents in the coarse grain fraction are brick, mortar, plaster and slag. Most of the additionalcomponents (declared as ’others’) are gravel, glass, asphalt, concrete, char and metal. Plasterof Paris (stucco) accounts for 3.4 g·kg−1 in the skeleton fraction. There are appreciable quan-tities of slag in the skeleton fraction and ash in the fine soil fraction, which was frequentlyused as insulating material in the ceiling voids of buildings. These components also originatefrom fires following World War II bombing raids.

3.3.2 Total sulfur and readily soluble sulfate contents

In general, the Stot of fine soil increases with soil depth for all investigated profiles (Tab. 3.2).Two examples of profiles are shown in Fig. 2. The soil samples taken from Fritz-Schloss-Parkexhibit approximately five times as much sulfur as samples from the Teufelsberg site. TheStot of the two subsoils shown exceed that of natural mineral soils by far, ranging between0.1 and 0.6 g·kg−1 (Brown, 1982). This may be attributable to a high content of ash in thefine soil fraction. According to Davison et al. 1974, the Stot content of (fly) ash accounts for488 g·kg−1. The pattern of Stot content along the examined soil layers and depths is contraryto those in natural soil, exhibiting decreasing sulfur contents with increasing depth (Tabatabaiand Bremner, 1972a). The other four examined profiles (not shown in detail) feature a sulfurdistribution comparable to the profile from Teufelsberg, illustrated in Fig. 3.2. The Stot slightlydecreases in yC1, and increases again in yC3. High variances in results were found betweenprofiles, and also within one layer of the same profile. This indicates a high degree of hetero-geneity at the examined sites, which is common for urbic technosols (e.g. Burghardt, 1994).Nevertheless, the samples taken from the deepest point feature the highest Stot in all cases,with at least 0.6 g·kg−1 Stot. The prevalent components of mortar, plaster and brick exhibit ge-ometric mean sulfur contents slightly below that of fine soil (Tab. 3.2). The Stot of brick is ina comparable range to that reported by Nehls et al. 2013, namely between 0.1 and 1.2 g·kg−1.In contrast, Stot of slag is roughly twice that of fine soil. High Stot contents of slag, up to16.9 g·kg−1, are also reported by Scott el al. 1986. The Ctot contents show no clear tendency

39


Figure 3.2: Depth profile of total sulfur (Stot, top) and soluble sulfate (SSO42− , bottom). Exemplary profile from

the Teufelsberg and Fritz-Schloss-Park, Berlin, Germany.

40


along soil layers (Tab. 3.2). The high carbon content of soil at deeper layers is attributed tothe large quantity of carbonate, which is also reflected in the high pH values. However, it mayalso result from a high ratio of black carbons, common in urban soils (Schleuß et al., 1998),especially in rubble caused by bombing. Ash in the fine soil fraction may further contribute toa high fraction of carbons (Landesarbeitsgemeinschaft Abfall, 1998).The dissolved sulfate concentrations of analyzed soil and component samples range from<7 mg L−1 up to gypsum solubility, which is 1581 mg L−1 (c.f. Wisotzky, 2011). The finesoil samples exhibit higher SSO4

2−-contents in lower soil layers than in the topsoil (Fig. 3.2).The maximum content was 2.28 g·kg−1 SSO4

2− (Tab. 3.2). Concerning Stot, the SSO42− of the

other four profiles examined is comparable to that at Teufelsberg, shown in detail in Fig. 2.Although yC2 and yC3 feature a comparable composition in technogenic components, yC3

exhibits higher contents of soluble SSO42− . This suggests advanced leaching of SSO4

2− fromupper soil layers. The high sulfate release from fine soil in the subsoil could arise from ap-preciable amounts of ash, with a high fraction of readily soluble SSO4

2− , as mentioned byStrayer et al. 1983. Kutchko et al. 2006 identified gypsum and anhydrite minerals in flyash. In addition, small particles of plaster of Paris (gypsum stucco) occur within fine soil;the fraction of this component in the skeleton ranges from 0.4 to 3.5 g·kg−1. For plaster ofParis, analyzed using coarse grained particles, SO4

2–-concentrations were at an equilibriumstate with respect to gypsum. The variance of sulfate release is high across all samplingdepths, due to a high degree of heterogeneity of the composition in the horizons. Mortar,plaster and stucco samples yield the highest contents of SSO4

2− among the technogenic com-ponents analyzed (see Tab. 3.2). These high contents could be attributed to the use of gypsumas an additive in mortar and plaster blends. Readily soluble sulfate minerals such as ettrin-gite (Ca6Al2[(OH)12(SO4)3] ·26 H2O) or thaumasite (Ca3Si(CO3)(SO4)(OH)6 ·12 (H2O)) mayalso occur in these components, as stated by Crammond et al. 2002. The slag componentsexhibit the least sulfate release. This is in contrast to their high Stot, suggesting that sulfur isnot bound in the highly soluble sulfide and sulfate minerals such as hannebachite, oldhamite,gypsum and ettringite, as Roy et al 2009 and Kirsch et al. 1966 reported for slag. Grindingslag to silt particle size did not lead to higher solubility. Nevertheless, it may be possible oc-clusion of soluble sulfide and sulfate minerals within the slag matrix occur. Sulfur may alsobe bound to black carbons, which feature an appreciable quantity of Ctot in slag (Shaw et al.,2010), or present as quite insoluble minerals such as barite (BaSO4) or pyrite (FeS2).There is a strong correlation between Stot and SSO4

2− contents for fine soil and mortar/plastercomponents (see Fig. 3). This correlation indicates that 34% and 29% of Stot exist as read-ily soluble SSO4

2− for fine soil and mortar/plaster, respectively. This is considerably higher

41


Figure 3.3: Correlation between total sulfur (Stot) and soluble sulfate content (SSO42−) for fine soil (left), mortar

and plaster components (right) from the Teufelsberg and Fritz-Schloss-Park, Berlin, Germany. n>13.

than for natural soils (Brown, 1982). Nevertheless, there is still a high fraction of Stot thatis not soluble in fine soil and mortar. The reasons for this could be that (i) sulfate mineralsare coated with a calcite layer as reported by Keren et al. 1981, inhibiting the dissolutionof sulfur-bearing minerals, or (ii) it is occluded by other minerals. No significant correlationbetween Stot and the SSO4

2− was determined for brick and slag. As already discussed for slag,occlusion of potentially soluble sulfur within the brick matrix may have hampered dissolutionbecause the loamy feedstock of brick was sintered during the manufacturing process.

Table 3.2: Geometric mean and ranges of contents of total carbon (Ctot ), total sulfur (Stot), soluble sulfate as wellas the pH for soil horizons and main technogenic components of rubble composed soils. The standarddeviation of the pH is <5%. n: number of samples.

Ctot range Stot range n SSO42− range pH n

Soil horizons g·kg−1 - g·kg−1 - - g·kg−1 - -

Ah 23.2 13.7 - 32.0 0.45 0.29 - 0.63 7 0.02 <0.01 - 0.03 7.3 4yC1 15.6 8.3 - 28.9 0.41 0.21 - 0.79 4 0.01 <0.01 - 0.03 7.6 7yC2 22.9 6.6 - 82.2 0.95 0.14 - 2.91 6 0.02 <0.01 - 0.06 8.1 8yC3 17.9 8.0 - 29.2 1.21 0.12 - 4.64 8 0.13 <0.01 - 2.28 8.1 15

Component

Bricks 3.4 0.7 - 13.6 0.58 0.13 - 2.60 18 0.05 0.01 - 0.84 7.9 31Mortar & Plaster 16.7 14.1 - 20.2 0.61 0.19 - 4.92 11 0.10 0.01 - 0.73 8.1 18Slag 70.6 22.7 - 170.0 2.13 0.75 - 6.88 5 0.06 0.02 - 0.31 8.0 6

42


Figure 3.4: X-ray powder diffractogram of the components plaster of Paris, mortar and slag.

43


3.3.3 Arylsulfatase activity

In order to assess the vertical distribution of organic sulfate, we analyzed arylsulfatase ac-tivity along one soil profile at the Teufelsberg rubble landfill. The activity declines from0.58·10−9mol g−1

DM min−1 in the highly root penetrated horizons Ah and yC1 (0-0.3 m) to0.26·10−9mol g−1

DM min−1 in yC2 (0.3 to 0.5 m) and 0.05·10−9mol g−1DM min−1 in yC3 (0.5 to

1.0 m). The standard deviation of the four replicates is below 5%. The measured rate ofenzyme activity of the topsoil is in a comparable magnitude as reported for agricultural top-soils by Giacometti et al. 2014 (0.36 to 0.45·10−9mol g−1

DM min−1). Tabatabai et al. 1970 andLorenz et al. 2005 documented a strong decrease in arylsulfatase activity with soil depth innatural and urban soils, respectively, since activity is closely linked to (i) organic matter and(ii) organic sulfate content (Zucker and Zech, 1985; Autry et al., 1990). No positive corre-lation between contents of Stot and activity, as stated by Lorenz et al. 2005, was determinedbecause the sulfur pools of rubble composed soils differ from natural soils. Nevertheless, theresults point to advanced leaching of readily soluble sulfate in the upper horizons becauseactivity is generally repressed by high sulfate concentrations in the bulk solution (McGill andCole, 1981). This finding confirms the results of the batch experiments, with low to moderatesoluble sulfate contents in the upper horizons. This is verisimilar because soil substrates havebeen exposed to leaching for more than 50 years.

3.3.4 Mineralogical characterization and speciation of sulfur

The solubility of inorganic bound sulfate depends to a great extent on the respective min-eral. We conducted X-ray diffractrometry to characterize existing minerals. The fine soilfraction consists mainly of quartz, various clay minerals and calcite. Traces of the sulfatemineral barite (BaSO4) were detected in only one of the 15 samples. This is plausible becausesimultaneously elevated Ba concentrations (0.6 g·kg−1) were measured, with a backgroundconcentration of 0.2 g·kg−1 (Birke and Rauch, 2000). Traces of barite were also detected inthe slag fraction (Fig. 3.4). The corresponding Ba concentration of analyzed slag ranged from0.2 to 1.8 g·kg−1. Other authors have also reported barite as a mineral in slag (Gee et al., 1997;Fallman, 2000). No sulfur-bearing mineral was detected in the coarse brick fraction. Traces ofettringite were detected in one out of three mortar samples. This mineral is quite common inbuilding materials, creating a highly alkaline environment (e.g. Hampsoim and Bailey, 1982).The high contents of readily soluble SSO4

2− from several mortar samples also corroborate this

44


result. Gypsum (Ca[SO4] · 2 H2O) was found in plaster of Paris particles, which are usuallymade of pure gypsum (Fig. 3.4). Although we assumed that we would detect gypsum in finesoil and, particularly, in the mortar and plaster components, this was not the case. This couldbe due to the detection limit of the XRD method for gypsum, which ranges between 1 and2% (Kontoyannis et al., 1997). Nevertheless, a congruently spatial distribution of calcium andsulfur, as reported by Brettholle et al. 2011 for samples from another rubble deposit in Berlin,indicates that sulfur exists mainly in fine soil as calcium sulfate. These results were obtainedby hard x-ray fluorescence element mapping in combination with XANES. In addition, thehigh contents of SSO4

2− in subsoil layers suggest the existence of sulfate minerals such as gyp-sum, featuring high solubility.The fractions of different sulfur oxidation states of Stot are shown by the results of XANESanalysis (Fig. 3.5). The sulfide and sulfonate content, as well as other sulfur compounds withinall four soil profiles, decrease from top to bottom, while the sulfate contents increase. At adepth of 0.31-0.80 m, sulfates account for 90% of Stot. The other studied profiles (not shown)feature a comparable pattern of sulfur species along the sampling depths. The high ratio ofsulfonate in the root-penetrated soil layers (to 0.31 m) indicates organically bound sulfur. Thisis in accordance with the outcomes of Autry et al. 1990, who reported that sulfonate wasthe main form of organic sulfur, with a ratio of at least 40%. Ester sulfates and sulfides canalso account for an appreciable fraction of total organic sulfur (Xia et al., 1998), especiallyfor topsoils. The XANES spectra of the main components of brick and mortar clearly yieldsulfate as the main source of sulfur; the spectra of slag indicate several oxidation states ofsulfur (Fig. 3.6). No differentiation between organic ester sulfate and inorganic sulfates waspossible for the soil and component samples because the peak energy of 2482.5 eV overlapsfor both sources (see also Prietzel et al., 2007). Nevertheless, the previously mentioned resultsof the batch experiments and arylsulfatase activity indicate that sulfate exists mainly as readilysoluble inorganic sulfate. The results, with a sulfate fraction of 90% of Stot, further suggestthat the soluble sulfate concentration may be higher than assessed by the batch experiments,due to the aforementioned reasons.

3.4 Conclusion and outlook

Rubble composed soils exhibit higher total sulfur contents than natural soils especially in sub-soils, where the main pool is inorganically bound. At least one third is highly water solublegypsum. This leads to high sulfate concentrations in the solution up to the solubility of gyp-sum.

45


Figure 3.5: Fraction of different sulfur species (left) and respective total sulfur contents (Stot, right) in differentsoil layers (exemplary profile from the Teufelsberg, Berlin, Germany) as analyzed by XANES with:a: organic sulfide; b: sulfonate; c: sulfate; d: sum of sulfoxides and solfones.

Figure 3.6: Sulfur K-edge XANES spectra of the most common technogenic components in rubble soils. Forsulfides, the dashed lines indicate the energy ranges where white line peaks of inorganic respec-tively organic sulfur compounds occur; the dotted line is the border between both ranges. For allother sulfur species, the dashed lines indicate the averaged white line peak position of the respectivespecies.

46


We conclude that such rubble composed soils pose a threat to the quality of groundwater,particularly because the adsorption of sulfate is negligible in an alkaline environment. Thelong-term release of sulfate from old building rubble landfills can be expected to be high.Further investigations must therefore be conducted. The same problem arises for recentlydeposited rubble caused by the increased use of gypsum drywalls in the building industry.

47

4 LYSIMETER STUDY ON LONG-TERM RELEASE OF SULFATE Stefan Abel - PhD Thesis

4 Long-term release of sulfate from building rub-ble composed soil: lysimeter study and numeri-cal modeling 3

Abstract

High sulfate concentrations in the groundwater occur in several cities and particularly inBerlin, Germany. Building rubble composed soils and landfills are a major source of dis-solved sulfates. This study assesses the sulfate release dynamics of such rubble-composedsubstrates.The substrate was taken from a building rubble landfill in Berlin, which was created afterWorld War II. It was filled into two lysimeters, which were irrigated repeatedly over two yearsto simulate several years of groundwater recharge. Sulfate concentration in the leachate wasmeasured monthly. Sulfate release dynamics was effectively described with PHREEQC, as-suming either one or two sulfate pools with kinetically limited dissolution.The volume that percolated the lysimeter column was 2440 L, corresponding to 17 years oflocal groundwater recharge. At the beginning, sulfate concentrations increased from approxi-mately 10·10−3mol·L−1 to 13·10−3mol·L−1, which is close to gypsum solubility concentra-tion. After 1.8 pore volumes a decrease was observed and after 8 pore volumes concentrationswere relatively constant at levels below 3·10−3mol·L−1. The data was best described by amodel which included a kinetically limited dissolution of gypsum from two sulfate poolsdifferent in their effective surface areas. One pool can be ascribed to fine-grained gypsumparticles, while the other can be ascribed to coarse-grained ones.Overall, rubble-composed substrates can be a severe long term source of sulfates.

4.1 Introduction

In Europe and other parts of the world numerous cities were heavily bombed during the WorldWar II and enormous quantities of building rubble accrued. The rubble was commonly de-posited throughout destroyed areas and on special landfills. Thus, rubble-composed soils covervast areas in such war-battered cities (Burghardt, 1994; Wolff, 1996; Abel et al., 2015). Those

3S. Abel, A. Peters, T. Nehls, G. Wessolek; submitted to Soil Science; March 2015, Special issue on urban soil,in review; please visit Soil Science

48

http://journals.lww.com/soilsci/pages/default.aspx


soils are classified as Urbic Technosols or as Anthroportic Udorthents according to FAO-WRB(2006) and USDA (2014), respectively.The construction and demolition material frequently exhibits considerable amounts of sul-fate bearing minerals, particularly gypsum (CaSO4:2 H2O, Abel et al. 2015). The total gyp-sum content differs from country to country. In USA, e.g., Jang and Townsend (2001) reporta gypsum content up to 91 g·kg−1 in new building waste, while in Germany Muller et al.(2011) mention a content of only 9 g·kg−1. At the beginning of the 50ies, the use of gypsumstarted to be increasingly employed in the building industry, especially for the production ofdrywalls (Bundesverband der Gipsindustrie, 2009). In former times gypsum was used as abinder and, in particular, for decorative coatings. Therefore it is a common component inthe rubble-composed soils and in landfills created after World War II. Gypsum is not consid-ered as hazardous waste. However, it has a high solubility and may provoke elevated SO4

2–-concentrations in the groundwater. Several authors document building rubble to be one of themajor sources for sulfate in the groundwater in urban areas (Grischek et al., 1996; Hannappeland Reinhardt, 2002; Abel et al., 2015). In numerous cities sulfate concentrations in the up-per groundwater aquifer largely exceed the EU drinking water standards of 2.5·10−3mol·L−1

(Council Directive 98/83/EC 1998). This can be attributed to the fact that the landfills, piledup after the World War II, feature no adequate technical protection, such as e.g. liners, and canthus release sulfate. The same applies for the rubble-composed soils.Several studies document high sulfate release from sole construction and demolition (C&D)waste (e.g. Gierig et al., 2006; Musson et al., 2008; Roussat et al., 2008). Jang and Townsend(2001, 2003) report sulfate concentrations in the leachate from new C&D waste in the mag-nitude of gypsum solubility which is about 14·10−3mol·L−1. However, these studies focusedon new building waste, and did not consider weathering processes and pedogenesis of therubble-composed artificially constructed urbic technosols. This may have altered the physi-cal and chemical behavior of substrates in soils and landfills in view of sulfate release. Abelet al. (2015) point to advanced leaching of sulfate bearing minerals in the upper horizons ofrubble-composed soils, which were constructed after WW II and still lead to elevated SO4

2–-concentrations in the groundwater.Schonsky et al. (2013) describe sulfate leaching processes from small scale rubble-composedsoil columns. By means of flow interruptions and numerical modeling they found that kinet-ically limited dissolution of sulfate might be an important process. However, the long termeffectivity of rubble as a sulfate source on larger spatial scales has not been evaluated yet.The aim of this study was to (i) study sulfate mobilization and transport processes from agedrubble-composed substrates on larger spatial and temporal scales and to (ii) use a hydrochem-ical model to predict sulfate leaching for several decades.

49


Table 4.1: Stot and readily soluble SO42–-content in the fine soil ≤2 mm and content of coarse-grained gypsum

(>2 mm; CaSO4:2 H2O) in the topsoil and in the rubble-composed substrate, respectively.

Soil Layer Stot SO42– CaSO4:2 H2O

mol·kg−1 mol·kg−1 mol·kg−1

Ah (0-0.3 m) 2.3·10−3 0.8·10−3 -yC (0.3-1.8 m) 8.7·10−3 4.2·10−3 5.6·10−3

4.2 Material and Methods

4.2.1 Soil material

The rubble-composed soil material was taken from a debris landfill called ’Teufelsberg’ inBerlin. From 1950 to 1972 approximately 26.2 Mio. m3 of rubble were deposited, particu-larly originating from damage of the World War II. The landfill reaches a maximum height of60 m. The site was covered with an allochthonous Ah soil material following the closure ofthe landfill. Results from groundwater monitoring conducted in the vicinity of the site showSO4

2–-concentrations up to 8.1 mmol·L−1. This is about the septuple of the background con-centration and thus indicates a high potential of sulfate release from the deposited buildingrubble.The soil profile of the sampling site is shown in Figure 4.1. The topsoil (Ah 0 to 0.3 m) andthe subsoil (yC 0.3 to 1.8 m) were excavated separately. In sum, 3 m3 of rubble-composedsubstrate was taken. Both substrates were passed through a 6.3·10−2 m meshed sieve and ho-mogenized. Components >6.3·10−2 m were admixed after homogenization according to theiroriginal fractions. The rubble-composed substrate in yC is composed of 0.52 kg·kg−1 fine soil(≤2 mm) and of 0.48 kg·kg−1 coarse-grained particles (>2 mm), such as bricks (0.27 kg·kg−1)and mortar (0.12 kg·kg−1). Other components (0.08 kg·kg−1) in the coarse-grained fractionare gravel, slag and glass. Fractions of the main coarse-grained technogenic components weredetermined by hand sorting of an aliquot of 48 kg of air dried material. Stot-content wasdetermined according to ISO 15178; the readily soluble SO4

2–-content of fine soil was de-termined according ISO/TS 21268-2:2007 and ISO 10304 (Tab. 4.1). The SO4

2– in the finesoil is assumed to have arisen from the dissolution of gypsum as this is the most abundantsulfate-bearing mineral in rubble soils (Abel et al., 2015) and readily soluble. The total gyp-sum content in the rubble-composed soil is thus 9.8·10−3mol·kg−1.

50


Figure 4.1: Left: Typical profile of urbic technosols, Teufelsberg Berlin, Germany; bottom left: gypsum particlesin the coarse-grained fraction; right: Design of the outdoor lysimeter, with Ah: topsoil; yC: rubble-composed substrate; a: suction plates covered by a geotextile and a quartz sand layer; b: storagecontainer; c: lever-arm counterbalance.

51


4.2.2 Lysimeter preparation

Two lysimeter were installed to determine sulfate release dynamics. The soil substrate werepoured in several steps into the lysimeter and compacted after each 0.1 m. The adjusted bulkdensity was 1.5 Mg m−3. Total amount of rubble-composed soil (yC) was 2.25 Mg in eachlysimeter; 1.5 m in depth. With a total gypsum content of 9.8·10−3mol·kg−1 (see above) thetotal mass of gypsum per lysimeter is 3.79 kg. Both lysimeters were covered with 0.45 Mgtopsoil (Ah)or 0.3 m, respectively. The lysimeters were kept largely free from any vegetationduring the experiment to minimize transpiration.In order to collect the leachate from the lysimeter outflows and to establish unsaturated con-ditions, four suction plates were installed at their bottoms (see Figure 4.1). The suction plateswere covered with a geotextile and a 0.1 m thick quartz sand layer and were set to a pressurehead of -63 hPa to obtain water content at field capacity.The leachates of the suction plates were captured in 11.5 L containers and discharged at leastonce per week. The lysimeters were irrigated once to twice per week with 30 to 70 mm tapwater, starting at June 2012 (see Figure 4.2), in order to fasten leaching process and to simu-late longer time periods. During frost (December to February) and after strong rainfall eventsno additional irrigation was applied. The duration of the experiment was 24 month, namely,from November 2012 to November 2014. The total precipitation in this period was 1095 mm,the additional irrigation applied was 2530 mm. Volume of percolate was measured at sam-pling. At each sampling date, the mean concentration for each lysimeter were calculated byweighting the concentrations of each container by the relative collected volume. This en-sures that an integrated concentration at the outflow was determined. Thus, we consider onecomplete lysimeter as representative rather than the single containers in which the concentra-tions differed due to the large heterogeneity of the rubble substrate. The SO4

2–-concentrationswere analyzed by ion chromatography according to ISO 10304, using Dionex DX-120 IonChromatograph. In order to calculate the water content, the weight of the lysimeter was con-tinuously gravimetrically measured using a lever-arm counterbalance system in combinationwith a laboratory scale. The laboratory scale was installed in April 2013. Further details ofthe lysimeter setup can be found in Peters et al. (2014).

52


4.2.3 Numerical Modeling

Theory

In order to describe the sulfate release dynamics we parameterized a combined hydrochemicalmodel integrated in the program PHREEQC (Parkhurst et al., 1999).The mobile-immobile formulation of the advection-reaction-dispersion equation for constantwater contents and fluxes is used to simulate the one-dimensional transport of dissolved sul-fate:

∂cm

∂ t=−v

∂cm

∂x+D

∂ 2cm

∂x2 +R1 +R2 (2)

where cm [mol·L−1] is the SO42–-concentration of the solute in the mobile phase, t [s] is

time, x [m] is the depth coordinate, v [m·s−1] is the pore-water velocity, D [m2·s−1] is thediffusion-dispersion coefficient, R1 [mol·L−1·s−1] is the source or sink term due to geochemi-cal reactions and R2 [mol·L−1·s−1] is the source or sink term due to dual porosity and physicalnon equilibrium. In the case that diffusion is much smaller than dispersion, D can be simplyexpressed by D = DL · v, where DL [m] is the longitudinal dispersion length. The mean pore-water velocity is given by v = q/θm, where q [m3·m−2·s−1] is the flow rate (Darcy flux) andθm [m3·m−3] is the water content in the mobile fraction .The kinetically limited dissolution process can be attributed to small effective surface areasof gypsum minerals, associated particularly with gypsum in the skeleton fraction. Since thediameter of gypsum fragments ranges from <2·10−3m up to 0.1 m, n-gypsum pools can beapplied for an effective description of the dissolution process:

dcm

dt=

n

∑i=1

Ai,t0V

·(

mi

mi,t0

)η

· k · (1−SR) = R1 (3)

where Ai,t0 [m2] are the effective surfaces of minerals at t=0; V [L] volume of solution; mi

[mol] mass of minerals; mi,t0 [mol] mass of minerals at t=0; k [mol·m−2s−1] dissolution rate;SR saturation ratio; η factor to account for changes in Ai/V during dissolution; for uniformlydissolving spheres η=2/3. For further details see e.g. Lasaga (1984). For simplicity we applya maximum of two gypsum pools (n=1; 2) for an effective description of the sulfate releasedynamics.Bricks possess an intrinsic porosity, acting as stagnant zones as described by Nehls et al.(2013). Thus we applied a dual porosity model with an exchange of solutes among the mobile

53


and immobile fraction via diffusion. In PHREEQC this is defined by:

dcim

dt=

α

θim· (cm − cim) = R2 (4)

with α =De ·θim

(a · fs→1)2 (5)

where the subscripts m is mobile and im is immobile; De [m2·s−1; ] diffusion coefficient inthe sphere; α [s−1] exchange factor; a [m] radius of spheres associated with the immobilefraction; fs→1 shape factor according to van Genuchten (1985). For further details see Gerkeand Genuchten (1993) and Parkhurst et al. (1999).

Model setup and parameterization

Model setup, initial and boundary conditions and transport parameter

The total length L of the system is set to 1.5 m, equal to the depth of the rubble-composedsubstrate. The topsoil is excluded from the simulation, since the substrate is free from anyrubble and sulfate release from such natural soil substrates can be neglected (e.g. Stanko-Golden et al., 1994; Scherer, 2009). The mean Darcy flux q to calculate v corresponds to theaverage water flux at the lower boundary of the lysimeter column, measured at least once perweek. The infiltrating solution is assumed to be in equilibrium with the atmospheric CO2 andO2 at 11.2◦ C, which is the mean annual soil temperature as measured in the weather stationBerlin-Marienfelde located next to our lysimeter study site. The bulk solution is equilibratedwith atmospheric O2 as well, which is justified by the unsaturated conditions in the lysime-ter as achieved by establishing a pressure head of -63 hPa at the lower boundary. The totalair content is 0.24 m3·m−3 (total porosity minus mean water content, see below). The bulksolution is further set in equilibrium with calcite, resulting in a pH of 7.8. The initial SO4

2–-concentration cm and cim is as measured at the beginning of the experiment.The longitudal dispersion length DL is fixed to 0.15 m, representing 10% of the simulatedflow domain (L=1.5 m). θim is calculated considering the fraction of coarse-grained bricks oftotal soil (0.27). Since almost all pores of the bricks are fine pores and the experiments havebeen carried out under moist conditions, the volumetric water content of the bricks are set tofull saturation, i.e. 0.36 m3·m−3 as mentioned by Trinks et al. (2007). This yields for θim =0.097 m3·m−3. Other coarse-grained components, such as gravel, glass and slag, are consid-

54


ered as non-porous. The intrinsic pore volume of coarse-grained mortar is attributed to themobile fraction, since its hydraulic characteristics are comparable to the ones of the fine soil(see Blume and Runge, 1978). θm is calculated by subtracting θim from mean total volumetricwater content as measured by the lever-arm balance.The exchange factor α is calculated for a spherical geometry and a diameter a of 0.05 m, rep-resenting roughly the mean diameter of the bricks in used substrate. The shape factor fs→1

for spherical shape of the stagnant region is 0.21 (van Genuchten, 1985). The diffusion coef-ficient in the sphere De for SO4

2– is set to 1.43·10−10m2·s−1. This value was determined byHill (1984) in chalk samples, which feature a comparable total porosity (0.45) as the coarse-grained bricks acting as stagnant zones (0.42, Wessolek et al. (2011)). This yields an exchangefactor α of 1.13·10−7 s−1.

Solute reaction parameters

The total gypsum content of 9.8·10−3 mol·kg−1 is already dissolved to a small extend, rep-resenting the dissolved SO4

2– in the bulk solution at initial conditions (1.6·10−3 mol·kg−1).Applying one gypsum pool (n=1 in Equation (2)) the total gypsum mass mt0 is thus 8.2·10−3

mol·kg−1. If we now apply two gypsum pools (n=2 in Equation (2)) the initial mass m1,t0 andm2,t0 for each pool is required. Thereby m1,t0 is the total gypsum content (9.8·10−3 mol·kg−1)minus the dissolved fraction (1.6·10−3 mol·kg−1) minus m2,t0 . The parameter m2,t0 is set tothe gypsum pool measured in the coarse-grained fraction (5.6·10−3 mol·kg−1, see Table 4.1).This results in 2.6·10−3 mol·kg−1 for m1,t0

The dissolution rate constant k for gypsum of 2.77·10−6 mol m−2 s−1 is given by Raines andDewers (1997). They measured the low rate insitu and thus it excludes any turbulent and fastlaminar water flow as implied by other experimental setups to determine dissolution rates (e.g.such as conducted by Jeschke et al., 2001). The two remaining parameters A1,t0 and A2,t0 areestimated by parameter optimization (see below). The set of all model parameters is given inTable 4.2.

Parameter estimation

The model was adjusted to the data by minimizing the sum of squared residuals (SSQ) betweenmeasured and modeled SO4

2–-concentrations.

55


SSQ =n

∑i=1

(Oi −Pi)2 (6)

where Oi [mol/L] are the measured data, which are the weighted mean concentrations for eachlysimter at each sampling date, Pi are the corresponding modeled concentrations and n is thenumber of observations. The minimization scheme was the same as used by Schonsky et al.(2013).In order to get information of model performance we applied the root mean square error(RMSE):

RMSE =

√SSQ

n(7)

and the Nash and Sutcliffe Efficiency (NSE) defined as:

NSE = 1−

n∑

i=1(Oi −Pi)

2

n∑

i=1(Oi − O)2

(8)

where O the mean observed concentration. NSE = 1 indicates a perfect fit, while NSE. < 0.65suggest that the model is unsatisfactory. Further details of the performance evaluation aregiven by Ritter and Munoz-Carpena (2013).

4.3 Results

4.3.1 Lysimeter study

The initial soil water content was 0.17 m3·m−3. After starting irrigation in June 2013 the meanwater content (θ ) increased to a maximum of 0.24 m3·m−3 (Fig. 4.2). In the period irrigationwas applied (June 2013 to November 2014) θ was 0.20±0.04 m3·m−3. The total volume ofwater that has infiltrated into each of the two lysimeter columns during the experiment was3625 mm. The volume of water percolated through the lysimeter column was 2363 mm and2420 mm, respectively. The percolated water corresponds to 12.7 times the pore volume in themobile zone (PVm), where PVm = θm ·L, with θm = θ −θim =0.103 m3·m−3 and L = 1.8 m.Note that in this case L of the total flow domain (Ah + yC) was used to calculate the pore vol-

56


Table 4.2: Input parameter of the numerical model with one gypsum pool (n=1) and two gypsum pools (n=2); †

measured, ‡ literature, ⋆ or fitted. ct0 for cm,t0 and cim,t0 .

parameter n=1 n=2

v† m·s−1 3.66·10−7 3.66·10−7

c†t0 mol·L−1 10.3·10−3 10.3·10−3

θ †m m3·m−3 0.103 0.103

θ†im 0.097 0.097

k‡ mol·m−2·s−1 2.77·10−6 2.77·10−6

α‡ s−1 1.13·10−7 1.13·10−7

m†t0 mol·kg−1 8.2·10−3 -

m†1,t0

- 2.6·10−3

m†2,t0

- 5.6·10−3

A⋆t0 m2·mol−1 6.2·10−3 -

A⋆1,t0 - 22.0·10−3

A⋆2,t0 - 2.3·10−3

RMSE mol·L−1 2.19·10−3 1.63 ·10−3

Figure 4.2: Average water flux at the upper and lower boundary from Nov. 2012 to Nov.2014 and mean watercontent as determined by the lever-arm counterbalance system.

57


Figure 4.3: SO42–-concentration: as measured in the leachate of each suction plate of lysimeter 1 (•) and lysime-

ter 2 (◦) and weighted mean of lysimeter 1 (•) and weighted mean lysimeter 2 (◦). fitted, applyingtwo gypsum pools (n=2), fitted, applying one gypsum pool (n=1). Horizontal dashed line depictsthreshold value of EU drinking water standard; pore volume= 185.4 L.

ume percolated, whereas for chemical reaction just the length of the rubble-composed layer(yC, L=1.5 m) was considered.The total volume percolated correlates with a liquid to solid (L:S) ratio of ∼0.9 L·kg−1. TheL:S ratio represents the ratio of percolated volume (mean 2392 L) and total mass of sub-strate (2700 kg). With the local average groundwater recharge rate of 140 mm·a−1 (Gluglaet al., 1999; Umweltatlas Berlin, 2008) the percolated volume corresponds to approximately17 years under natural conditions.The SO4

2–-concentrations in the outflow of the lysimeter range between 8.5·10−3mol·L−1 and12.6·10−3mol·L−1 in the beginning of the study (Fig. 4.3). The concentrations are continu-ously increasing until 1.8 pore volumes were percolated. Afterward, the SO4

2–-concentrationsdecrease continuously to 2.06·10−3mol·L−1 to 5.81·10−3mol·L−1.

58


4.4 Discussion

The water content after starting irrigation of 0.20±0.04 m3·m−3 is in a comparable magnitudeas mentioned by Blume and Runge (1978) for rubble-composed substrates at field capacity(0.23 m3·m−3).The SO4

2–-concentration in the leachate in the beginning is much smaller than the theoreticalsolubility concentration of gypsum, which is 14.1·10−3mol·L−1 (Parkhurst et al., 1999). Thismoderate concentrations can be attributed to the non-equilibrium between in the concentra-tions between pore water in the mobile and the immobile fraction, where SO4

2–-dissolutionjust occurs in the mobile fraction.The sharp increase of SO4

2–-concentrations within the first pore volume points to a high quan-tity of small gypsum particles or gypsum coatings linked with a high effective surface area(first-flush). In our lysimeter study the first-flush may result from the mechanical treatment ofsoil substrate and the precipitation of gypsum from the soil solution during soil excavation andlysimeter preparation. Such a first-flush effect is not unrealistic but can also be expected wher-ever building rubble has been deposited as reported for various contaminants in the leachatefrom new C&D waste (Kalbe et al., 2007; Wehrer and Totsche, 2008). Note that in the begin-ning of the experiment (0 to 0.7 pore volumes) no irrigation was applied and thus residencetime was longer. This might additionally explain the effect of increased concentrations, al-though this is not accounted for in the effective model.The tailing of the sulfate release curve indicates an additional coarse-grained gypsum pool

with limited dissolution due to small effective surface areas as reported by several authors(e.g. Keren and O’Connor, 1982; Bolan et al., 1991; Bendz et al., 2007). The effective surfacearea of 2.3·10−3 m2mol−1 of the coarse-grained gypsum fraction applied in the simulation ismuch smaller than the ones documented for sand (>0.1 m2g−1 Scheffer and Schachtschabel(2010)). The corresponding particle diameter of the gypsum mineral is 0.18 m. In the sub-strate particles up to 0.10 m were detected, however, most of the particles were smaller. Thesurface area corresponding to the mean diameter of gypsum particles in the substrate is thuscertainly larger than the effective surface area used in the model. This indicates that otherfactors such as coatings of less soluble mineral on gypsum particles or physical heterogeneityas e.g preferential flow does occur to a certain degree. This results in a smaller contact areabetween the mineral and mobile pore-water and thus in smaller effective surface areas.Lopez Meza et al. (2010) report of equilibrium concentrations already after a 16 h residencetime of the leachate in a rubble-filled soil column. The mean residence time in the rubble-composed layer in our experiment during the application of irrigation was approximately 47

60


Figure 4.5: Simulated SO42–-concentration in the range of measured data and predicted applying two gyp-

sum pools; high flux rate as measured in the lysimeter study (3.66·10−7m·s−1) and low flux rate(4.34·10−8m·s−1), simulating natural conditions with an average annual groundwater recharge of140 mm. Horizontal dashed line depicts threshold value of EU drinking water standard.

61


days. However, the concentrations were below equilibrium concentration as the dissolution isstrongly constrained by small effective surface areas of the coarse-grained gypsum particles.After stopping irrigation (November 2014) residence time increased again and slightly in-creasing concentrations (mean of 3.5·10−3mol·L−1 and 2.5·10−3mol·L−1 for lysimeter 1 and2, respectively, not shown in Figure 4.3) could be determined in June 2015. This substantiatesthe kinetically limited dissolution of gypsum.Jang and Townsend (2001) document comparable SO4

2–-release dynamics, using new coarse-grained C&D waste. However, for leaching experiments with C&D fines they documenthigh concentrations in the leachate over several L:S ratios. Roussat et al. (2008) documentmuch higher concentrations for different types of new C&D wastes as well, ranging between5.7·10−3mol·L−1 and 14.6·10−3mol·L−1 at a L:S ratio of 1.2 L·kg−1. In their study 0.65 m3

sized lysimeters were applied. The smaller concentrations in our study can be attributed tothe overall lower content of sulfate bearing minerals in aged building waste compared to newC&D waste.Trankler et al. (1996) report readily soluble SO4

2–-contents in the range from 0.2 to 55.6·10−3

mol· kg−1 in the coarse-grained fraction and from 0.9 to 160·10−3mol·kg−1 in C&D wastefines. Lopez Meza et al. (2010) mention a comparable SO4

2–-quantity with up to 142.7·10−3

mol· kg−1. In the substrate we analyzed a readily soluble SO42–-quantity of only 9.8·10−3

mol· kg−1. Although the threshold of 5.8·10−3mol·kg−1 for landfills for inert waste (Deci-sion 2003/33/CE 2003) are exceeded, the criteria for non-hazardous waste landfills are met(208.3·10−3mol·kg−1 at L:S = 10 L·kg−1).In landfills with high percentage of soil skeleton preferential flow may occur, influencingleaching rates (e.g. Zeiss and Major, 1992; McCreanor and Reinhart, 2000). Such preferen-tial flow paths have probably formed in the lysimeter column, leading to the variability inSO4

2–-concentrations in the leachate between single suction plates in the lysimeter outflow.Furthermore, the water flow velocity employed in the simulation is related to relatively highwater contents. Considering the regional natural conditions, the precipitation and thus waterflux (q) is definitely smaller. In the simulation for predicting SO4

2–-dynamics under naturalconditions, the smaller q has been accounted for leading to higher residence times and thushigher concentrations in the beginning as well as faster depletion of readily soluble sulfates(Fig. 4.5). However, under such conditions water contents would also be lower, which wouldlead to a slight change of solute velocity (v), which has not been accounted for in the simula-tion. For a detailed long-term prediction the magnitude of the preferential flow must further betaken into account as recommended for municipal waste dump sites (Rosqvist and Destouni,2000; Rosqvist et al., 2005).

62


For a more general long term prediction of sulfate release dynamics from building rubblelandfills the initial quantity of readily soluble sulfates is the most important parameter. It de-termines the source function and total quantity of SO4

2–, which can be transported towardsthe groundwater. The surface substrate used in this study was exposed to weathering pro-cesses for more than 60 years and is thus not representative for the whole landfill. Substratefrom deeper layers in the landfill may still possess the initial amount of readily soluble sulfateminerals. The high SO4

2–-concentrations in the leachate from the overlying layers may haveimpeded their dissolution so far. Running the simulation for a 60 m high rubble column (siteBerlin Teufelsberg) using the parameters as described above but setting the initial gypsumquantity to 52.4 mmol·kg1− as reported by Muller et al. (2011) for new C&D waste results inapproximately 2470 years of SO4

2–-release with concentrations above the EU drinking waterthreshold.However, in such a period changes of the characteristics of non or weakly soluble sulfatespecies, such as e.g. co-precipitated mineral phases as mentioned by Schonsky et al. (2013),may occur. Additionally, the pore system itself will change substantially due to weatheringand soil genesis. Thus, this simplified simulation applied to describe sulfate release fromthick rubble layers feature a high uncertainty but nevertheless underlines the long term hazardof these sulfate sources for the groundwater and the need for political action.

4.5 Conclusions

Building rubble-composed substrates and soils feature a high sulfate release. Gypsum in thefine soil fraction is associated with a high effective surface area. Their dissolution and deple-tion occurs relatively fast, provoking high sulfate concentrations in the leachates after deposi-tion of building-rubble. Coarse-grained particles render more moderate levels of sulfate in theleachate, however, this fraction is responsible for the long-term release. These phenomena canbe effectively described using two gypsum pools with different effective surface areas in themodel. The dissolution of sulfate, particularly from the coarse-grained particles, is stronglylimited by the effective surfaces in contact with the leachate.Overall, the rubble-composed substrates deposited after WW II still feature an appreciablesource for sulfates. Thus, weathering of gypsum particularly from large building rubble land-fills may last several centuries and pose a risk to the groundwater quality. Subsequent measuresto protect the groundwater should be considered.

63

5 SYNTHESIS AND CONCLUSIONS Stefan Abel - PhD Thesis

5 Synthesis and Conclusions

5.1 Heavy metal and B[a]P contamination and its impacts onthe environment

Building rubble composed substrates show clearly elevated contents of all considered heavymetals and organic pollutants in comparison to natural substrates. Therefore, the hypothesisthat rubble composed soils feature significantly higher contents in heavy metals and B[a]Pthan soils without any rubble was confirmed. Precautionary values of the BBodschV (1998a)are frequently exceeded. The elements Hg and Pb are the most predominant in rubble com-posed soils. Unexpectedly, the As, Cd, Cr and Ni contamination was not as pronounced as forother metals. In the study, I could identify slags and ashes as the most contaminated compo-nents. Extremely high contents may be attributed to industrial facilities or the deposition ofrubble from industrial buildings.The soil surveys were conducted at least 30 years after the deposition of the building rubble.Thus, the initial heavy metal contents in soil can be expected to have been higher as reportedin the surveys since leaching may have reduced the overall content of the contaminants overtime. Particularly after deposition high weathering may have occurred as reported by Wehrerand Totsche (2008) for new C&D waste.

Despite elevated heavy metal contents, the majority of surveyed sites pose only low threatto the human health according to the classification system of Eikmann and Kloke (1993).The classification system defines specific threshold values for various land use and therefore,the need to protect man and groundwater resources by applying remediation measures. InTable 5.1 the most sensitive land use, namely playground, was used to classify the rubblecomposed soils and substrates. Although in most cases risk to human health is low, high con-tamination and high risk cannot be excluded a priori for all sites, since some rubble composedsoils show extremely high contents of pollutants.

Also the contents of B[a]P are significantly higher for rubble composed substrates comparedto those without any rubble. Since B[a]P is a reliable surrogate and indicator for PAHs, highcontents can be expected for other PAHs as well. This is also demonstrated by the results fromSmettan and Mekiffer (1996), reporting elevated contents of all 16 PAHs (EPA). PAHs arecommonly highly persistent in soil, nonetheless, microbial degradation via cometabolic pro-cesses, volatilization and photolysis may occur (e.g. Lu et al., 1977; Cerniglia, 1984; Juhasz

64


and Naidu, 2000). For B[a]P Doick et al. (2005) report a half-life of 2.7 years in agriculturalsoils. However, in rubble composed urban subsoils, the degradation rate may be assumed tobe not as high since the organic matter content, necessary for cometabolic processes, is lowercompared to agricultural soils, which were considered in the mentioned study. Overall, theinitial content after deposition of the rubble 60 years ago may have been higher as reported inthe soil surveys because of the degradation of PAHs.

Table 5.1: Assessment scheme for the human toxicological health risk by heavy metals and benzo(a)pyreneaccording to Eikmann and Kloke (1993); application to the data from the soil surveys, using themedian (in brackets).

very low low moderate high very high buildingmg·kg−1 rubble

As < 2 20 50 500 > low (3.6)Cd < 0.2 2 10 100 > very low (0.1)Cr < 5 50 250 2500 > low (10)Cu < 5 50 250 2500 > low (28)Hg < 0.1 0.5 10 100 > low (0.3)Ni < 4 40 200 2000 > low (9)Pb < 20 200 1000 10000 > low (77)Zn < 30 300 2000 20000 > low (99)B[a]P < 0.1 1 5 50 > low (0.5)

The groundwater contamination with regard to the considered heavy metals is negligible, eventhough vast areas are covered with building rubble substrates (SenStadtUm 2006, see Fig. 2.2on page 19). Just in few cases (<1% of all results of groundwater samples; n>1561) thethreshold values of the EU drinking-water standard and the European Groundwater Directive2006, respectively, are exceeded. The results suggest that the leaching rate of heavy metalsfrom the rubble composed soils is quite low.

As studies of the leaching behavior of heavy metals from building rubble soils do not exist,data from studies for new C&D waste and from common soil science have to be consulted todescribe this process. Roussat et al. (2008) determined concentrations in the eluate of variousnew C&D wastes, exceeding threshold value for some heavy metals. Susset and Leuchs (2008)determined heavy metal concentration exceeding threshold values only in the beginning oftheir lysimeter study, what may be partly attributed to colloidal transport. In contrast, a study,carried out on behalf of the ministry for environmental protection (LUBW, 2009) documentsconcentrations far below the threshold values in the eluate from C&D wastes. Same applies tothe results from Galvın et al. (2012).

A low leaching rate can be attributed to the C&D and the building rubble itself, fostering ahigh sorption capacity of heavy metals (Goetz and Baasch, 2002). Thereby the high pH values

65


Figure 5.1: Sorption capacity of soils of heavy metals according to the Berliner Umweltatlas 2008 and Blumeand Brummer (1991). The building rubble landfill ’Teufelsberg’ is marked with a circle.

is the primary factor, promoting the sorption to soil colloids (e.g. Alloway, 1995; Bradl, 2004).Beyond a pH value of 6.5, the extent of metal leaching becomes insignificant (Scheffer andSchachtschabel, 2010); the pH of the considered rubble soils is on average 8.1 (see Table 2.4,page 29). Furthermore, in soils in which the quantity of dissolved SO4

2– in the leachate isabundant, precipitation of sparingly soluble metal-sulfate minerals may occur, such as e.g.Anglesite PbSO4 (Illera et al., 2004). Sorption and immobilization of solved metal ions alsooccur in underlying layers, which are in several cases up to 30 m. This may prevent fromelevated concentrations in the groundwater.

Figure 5.1 depicts the sorption capacity of soils for heavy metals in the study area Berlin. Lowsorption capacity occurs only in areas where wetland or thick layers of sandy soils, developedfrom glacial load sands or drift sands, exist and pH solution is relatively low. In these areastwo rubble deposit sites exists, though one of them is the biggest in Berlin (Teufelsberg; No.1 in Fig 1.8, page 11). However, as the rubble composed soils themselves possess a highsorption capacity as abovementioned high sorption capacity is documented wherever rubblehas been dumped, coinciding with the illustration in Figure 5.1. Same applies to the inner cityarea, where soils primarily developed from sand as well and which are highly influenced bythe deposition of building rubble.

As reported for heavy metals, a notable pollution of the groundwater with regard to organicpollutants does not occur in Berlin. Concentrations exceeding the threshold value for B[a]Paccording to the TrinkwV (2001) is documented for less than 1% of all analyzed groundwater

66


samples (n=1299, SenStadtUm2006).

This can be attributed to the low solubility of PAHs in water and their affinity to partition tosoil and, in particular, to organic matter (McCarthy et al., 1989; Hattemer-Frey and Travis,1991; Maxin and Kogel-Knabner, 1995). Smettan and Mekiffer (1996) determined low tonegligible leaching rates of PAHs from building rubble composed substrates. Also Gieriget al. (2006) report a total PAH concentration below detection limit even from soil with tarbased fragments, which are commonly rich in PAHs.

Although a threat to the ground- and drinking-water can be highly excluded, contaminated soilparticles may be re-suspended by wind erosion. Hawley (1985) documents that about 50% oftotal suspended particulate matter in air is resuspended local soil. Particularly in summermonth, when soil surface is dry and soil particles are highly susceptible to erosion those sitesmay pose threat to human health in view of direct inhalation and ingestion of contaminated soil(Van Wijnen et al., 1990; Laidlaw et al., 2012). Thus, particularly during removal of buildingrubble substrates, e.g., during construction works, it has to be ensured that dust generationis minimized, as required according to the German Federal Pollution Control Act (BImSchG2002).

In the study area landfill sites are commonly used for recreation. Damages to vegetative covercaused by footsteps and vehicle ruts along trails frequently result in bare soil layers, which arevulnerable to erosion. Weather events with severe precipitation may further cause erosion ofthe top soil layer(s). Water erosion and further transport of (contaminated) soil particles intosewer and open water bodies may therefore endanger the human health.

5.2 Sulfate release and its impact on the environment

Rubble composed substrates exhibit an appreciable quantity of total sulfur (Stot) and read-ily soluble sulfate (SSO4

2−). The Stot-content is increasing with the soil depth. In the sub-soil (<2 mm) the mean Stot-content is 1.21 g·kg−1. The pattern of Stot-content is contraryto the one in natural soils, which possess the biggest sulfur pool, mainly organically bound,in the topsoil. The mean SSO4

2−-content in the fine-grained rubble composed substrate is0.13 g·kg−1. The main binding form is inorganic sulfate, particularly gypsum. The meangypsum content in the coarse-grained fraction is 1.9 g·kg−1. The total quantity of gypsumin the analyzed rubble-soils is on average 1.3 g·kg−1. The initial quantity of gypsum of thelandfilled building rubble can be expected to have been significantly higher right after deposi-tion, since the substrate used for analysis was exposed to weathering processes for more than

67


60 years.

The particle size of gypsum strongly influence the dissolution rate: fine-grained gypsum pro-vokes high concentrations in the range of gypsum solubility, however, this fraction depletesrelatively fast. On the contrary, the coarse-grained gypsum dissolves quite slowly and provokemoderate SO4

2–-concentrations in the eluate. Nevertheless, this fraction is reasonable for thelong-term leaching behavior of rubble composed substrates. The high total amount of gyp-sum results in long time periods of several centuries with high sulfate concentrations in theleachate from rubble landfills.

Today, sulfate concentrations in the upper groundwater aquifer is commonly decreasing in theinner city. This is mainly attributed to the advanced weathering of sulfate bearing minerals.However, also excavation and removal of building rubble substrate during building operationshas contributed to the decrease in SO4

2–-concentrations in the upper groundwater aquifer. TheSO4

2–-concentrations in the inner city drop, while close to the landfills this is commonly notthe case due to its vast quantity of rubble. The concentrations documented in groundwatermonitoring programs indicate constantly high, fluctuating or even increasing concentrations.But how long will last such high SO4

2–-concentrations in the leachate?

As the initial gypsum content of the landfilled building-rubble is not known and can be recalcu-lated only by condoning a high level of uncertainty, the prospective sulfate release from land-fills created after WW II cannot be predicted exactly (see also discussion in Chapter 4). How-ever, running the simulation parameterized to the results of the lysimeter study (see page 48)for various landfill depths, the results underline the long-term release of sulfate (see Fig. 5.2).Employing the initial gypsum quantity as measured in the substrate mentioned above yields inSO4

2–-concentration above TrinkwV threshold value for more than 460 years from the highestbuilding rubble landfill in Berlin, the ’Teufelsberg’. For higher initial gypsum contents, e.g.9.0 g·kg−1 as reported for new C&D waste (Muller et al., 2011), this period is longer accord-ingly. Thereby, the approximate time span until SO4

2–-concentration in the leachate dropsbelow threshold value is calculated with 41.1 year per meter depth of the building rubble com-posed substrate.

Overall, the rubble composed soils and, in particular, the landfills piled up after WW II posea relevant threat to the groundwater quality in Berlin. This is particularly the case sinceSO4

2–-adsorption to soil is negligible at high pH values (Marsh et al., 1987). The high SO42–-

concentrations measured close to rubble landfills and in parts of the inner city of Berlin cor-roborate the outcomes of the laboratory and lysimeter studies. About 23% of groundwatersamples exceed the threshold value of the TrinkwV (2001). Thus, particularly in Berlin the

68


supply of drinking-water of good quality may become difficult, if the predicted increase ofSO4

2–-concentrations in the river Spree arises.

Up to now, measures to protect from sulfate leaching from building rubble landfills are notconducted in the study area Berlin. However, this issue is taken seriously by the local gov-ernment and the public water board and is frequently discussed in local media together withprospective high sulfate load of the river Spree.

Figure 5.2: Predicted time period of SO42–-concentrations exceeding threshold value of TrinkwV (2001) at local

groundwater recharge rate of 140 mm per year; Simulation as described above applied for buildingrubble and C&D wastes layers up to 60 m thickness with a total gypsum quantity of (a) 1.7 g·kg−1 asmeasured in the lysimeter study, (b) 5.0 g·kg−1 (exemplary guess), and (c) 9.0 g·kg−1 as mentionedby Muller et al. (2011) for new C&D waste.

5.3 Transferability of the results to C&D waste

Despite of differences in the composition (see Fig. 1.1 on page 4), the considered buildingrubble substrates and most C&D waste feature heavy metal contents in a quite comparablemagnitude as shown in Table 5.2. For B[a]P this is not the case. The organic pollutants inbuilding rubble result particularly from the firestorms following the bombing raids. Thus,Smettan and Mekiffer (1996) strongly recommend to distinguish between rubble accrued inthe WW II and new C&D.

69


Regarding the sulfate release, C&D waste constitutes in most cases higher contents in readilysoluble sulfates compared to the building rubble substrates analyzed in this study (e.g. Goetzand Baasch, 2002; Roussat et al., 2008; Susset and Leuchs, 2008, see also Chapter 3). Thiscan be mainly attributed to the increasing use of gypsum in the building industry.

As the data for new C&D waste underlines its potential contamination and its high potentialto release sulfate there is urgent need for proper waste management to avoid a disordereddeposition and negative impacts on the environment and human health.

Table 5.2: Pollutants in building rubble substrates and C&D waste: a comparison. † Acc. to (Townsend et al.,2004, U.S.), ∗ acc. to (Meuser, 1996a, Germany); o acc. to (Butera et al., 2014, Denmark); acc. toABANDA (Abfallanlysedatenbank, taken from (Ramke, 2005, Germany) with ‡ uncontaminated and⋆ contaminated C&D waste (Abfallschlussel 170106 and 170107) ; • own data; AM: arithmetic meanM: median.

- As Cd Cr Cu Pb Hg Ni Zn B[a]P SSO42−

mg·kg−1 g·kg−1

C& D waste† AM 4.4 2.0 21 50 92 0.3 76 290 - -C& D waste‡ M - 0.7 26 29 82 0.2 22 140 - -C& D waste⋆ M - 17 153 249 119 0.1 187 1120 - -C& D wasteo AM 3.4 0.2 23 18 33 - 13 67 - 5.5building rubble ∗ AM 15 1.5 28 40 168 0.3 21 610 2.7 -building rubble• AM 6.7 0.7 15 119 220 1.6 15 282 2.7 0.1building rubble• M 3.6 0.04 10 28.4 77.2 0.29 8.9 99.4 0.5 -

70


5.4 Outlook and future research

Leaching of heavy metals and B[a]P from rubble composed substrates may be considered tobe negligible, because alkaline conditions prevail in such soils. Nevertheless, this may be anephemeral effect and the pH solution may decrease with advanced weathering of calcite. Thisis corroborated by the results from Goetz and Baasch (2002), reporting of increased leachingof heavy metals from recycled C&D waste with a decrease in the pH solution. Thus, rubblecomposed soils and landfills may pose threat to groundwater quality in upcoming decades. Itis necessary to keep a consistent focus on this issue and to monitor continuously such con-taminated sites. In order to evaluate the long term leaching behavior from rubble composedsubstrates further research is necessary. Therefore, long-term column or lysimeter experimentscan be an appropriate tool to investigate the weathering of calcite, patterns of pH solution andleaching of heavy metals.

Our studies have shown, that leaching of sulfates from rubble composed substrates and es-pecially the landfills poses the major threat to groundwater quality. Concentrations up to thesolubility of gypsum were analyzed. However, in the unsaturated zone the leachate fromrubble-landfills SO4

2–-concentrations never had been measured. Existing data from ground-water monitoring programs, analyzing SO4

2–-concentrations in the groundwater beneath rub-ble landfills, e.g. at the Fritz-Schloss-Park, show concentrations smaller than 1000 mg·L−1.This might result from the sampling below the groundwater table within the zone of satura-tion. In this zone dilution of the leachate from the landfill may occur. In order to analyzeconcentrations in the leachate, measurements within the capillary fringe or in the vadose zonewould be necessary to corroborate the outcomes of the study.

All experiments were conducted using surface-near substrates, which are exposed to weath-ering processes for about 60 years. Substrates in deeper zone in the landfill body may differregarding their sulfur contents and may still exhibit the initial quantity of readily soluble sul-fate, because concentrations in the percolate in the range of gypsum solubility may impedefurther leaching. In order to estimate the quantity of sulfates in deeper zone, substrates sam-pling via deep hole drilling into the landfill body is necessary.

In our simulations, steady state conditions of the water regime were postulated. However, innatural soils and, in particular, in landfill bodies, this is not the case. Firstly, the hydraulicproperties depends on the upper boundary conditions, particularly for layers close to the sur-face. Secondly, the different kinds of the disposed waste and its differences in shape, sizeand permeability may promote preferential routing and short-circuiting of leachate flow as

71


reported for other municipal and solid waste landfills (e.g. McCreanor and Reinhart, 2000;Oman and Rosqvist, 1999). Both characteristics were not considered in the model parame-terized to the concentrations measured in the lysimeter study. Thus, in order to improve thenumerical model, further investigations of soil physical properties of rubble composed sub-strate especially in landfill bodies is necessary.

5.5 Conclusion

Building rubble composed substrates show elevated contents of heavy metals and B[a]P. Themost predominant contaminants are Pb and Hg. The metals Cd, Ni, Cr, and As are not ascharacteristic for rubble composed substrates as the other considered pollutants. Although themajority of most pollutants exceed precautionary value of the BBodSchV 1998a, there is noimmediate threat to human health and the aquatic environment. This is attributed to the alka-line characteristics of the rubble itself, which prevents from leaching. However, the preventionof deflation of contaminated soil particles by wind erosion is of importance to avoid threat tohuman health by inhalation of airborne particles.There is a high potential of rubble composed substrates to release sulfate, resulting from ap-preciable amounts of gypsum in its composition. Leaching of sulfates into the aquatic envi-ronment may cause a serious problem for drinking-water resources, especially in the studyarea Berlin. The weathering of gypsum from large building rubble landfills can be expectedto last several centuries, while the rubble composed soils in the inner city are already depletedin soluble sulfates to a large extend. Thus, subsequent measures to protect the groundwatershould be considered especially for landfills. Protection measures, such as covering the land-fills with further topsoil material provide a reduction in groundwater recharge and thus annualsulfate loads. Simultaneously such measures impede water and wind erosion of contaminatedsoil and therefore protect human health.

72

References Stefan Abel - PhD Thesis

ReferencesAbel, S., T. Nehls, B. Mekiffer, M. Mathes, J. Thieme, and G. Wessolek (2015). Pools of

sulfur in urban rubble soils. Journal of Soils and Sediments 15, 532–540.Abel, S., T. Nehls, B. Mekiffer, and G. Wessolek (2015). Heavy metals and benzo [a] pyrene

in soils from construction and demolition rubble. Journal of Soils and Sediments 15, 1771–1780.

Alaily, F., R. Grenzius, M. Renger, K. Stahr, B. Tietz, and G. Wessolek (1986). Soilscapes ofBerlin (West). In Mitteilungen der Deutschen Bodenkundlichen Gesellschaft, Volume 50,pp. 1–206.

Alloway, B. and E. Steinnes (1999). Anthropogenic additions of cadmium to soils. In Cad-mium in soils and plants, pp. 97–123. Springer.

Alloway, B. J. (1995). Heavy metals in soils. Springer Science & Business Media.Amt fur Umweltschutz, Hamburg (2000). Umsetzung der EG-Wasserrahmenrichtlinie

(WRRL). Environmental monitoring report, Amt fur Umweltschutz, Hamburg.Anttila, P. (1990). Characteristics of alkaline emissions, atmospheric aerosols and deposition.

In Acidification in Finland, pp. 111–134. Springer.Arnold, A. (1999). Bruchstucke: Trummerbahn und Trummerfrauen. OMNIS-Verl.Autry, A. and J. Fitzgerald (1990). Sulfonate S: a major form of forest soil organic sulfur.

Biology and Fertility of Soils 10(1), 50–56.Autry, A. R., J. W. Fitzgerald, and P. R. Caldwell (1990). Sulfur fractions and retention

mechanisms in forest soils. Canadian Journal of Forest Research 20(3), 337–342.Baek, S., R. Field, M. Goldstone, P. Kirk, J. Lester, and R. Perry (1991). A review of atmo-

spheric polycyclic aromatic hydrocarbons: sources, fate and behavior. Water, Air, and Soilpollution 60(3-4), 279–300.

Bendz, D., P. L. Tuchsen, and T. H. Christensen (2007). The dissolution kinetics of majorelements in municipal solid waste incineration bottom ash particles. Journal of contaminanthydrology 94(3), 178–194.

Bergback, B., K. Johansson, and U. Mohlander (2001). Urban metal flows–a case study ofstockholm. review and conclusions. Water, Air and Soil pollution: Focus 1(3-4), 3–24.

Birke, M. and U. Rauch (1994). Geochemical investigations in the urban areas of Berlin.Mineral Magazine A 58, 95–96.

Birke, M. and U. Rauch (1997). Geochemical investigations in the Berlin metropolitan area.Zeitschrift fur angewandte Geologie 43, 58–65.

Birke, M. and U. Rauch (2000). Urban geochemistry: investigations in the Berlin metropolitanarea. Environmental Geochemistry and Health 22(3), 233–248.

Blume, H.-P. and G. Brummer (1991). Prediction of heavy metal behavior in soil by means ofsimple field tests. Ecotoxicology and Environmental Safety 22(2), 164–174.

Blume, H.-P. and T. Hellriegel (1981). Blei-und Cadmium-Status Berliner Boden. Zeitschriftfur Pflanzenernahrung und Bodenkunde 144(2), 181–196.

Blume, H.-P. and M. Runge (1978). Genese und Okologie innerstadtischer Boden ausBauschutt. Zeitschrift fur Pflanzenernahrung und Bodenkunde 141(6), 727–740.

Bolan, N. S., J. K. Syers, and M. E. Sumner (1991). Dissolution of various sources of gypsumin aqueous solutions and in soil. Journal of the Science of Food and Agriculture 57(4),527–541.

Bradl, H. B. (2004). Adsorption of heavy metal ions on soils and soils constituents. Journal

73

http://link.springer.com/article/10.1007/s11368-014-1014-1





of Colloid and Interface Science 277(1), 1–18.Brettholle, M., S.-C. Gleber, B. Mekiffer, D. Legnini, I. McNulty, S. Vogt, G. Wessolek,

and J. Thieme (2011). Spatially resolved sulfur speciation in urban soils. In The 10thinternational conference on x-ray mircoscopy, Volume 1365, pp. 361–364. AIP Publishing.

Bridges, E. M. (2009). Waste Materials in Urban Soils, pp. 28–46. Blackwell Publishing Ltd.Brown, K. A. (1982). Sulphur in the environment: a review. Environmental Pollution Series

B, Chemical and Physical 3(1), 47–80.Bundesministerium fur Umwelt, Naturschutz, Bau und Reaktorsicherheit (1994). Kreis-

laufwirtschaftsgesetz, Abfallgesetz (KrW-/AbfG).Bundesministerium fur Umwelt, Naturschutz, Bau und Reaktorsicherheit (1998a). Bundes-

Bodenschutz- und Altlastenverordnung (BBodSchV).Bundesministerium fur Umwelt, Naturschutz, Bau und Reaktorsicherheit (1998b). Gesetz

zum Schutz vor schadlichen Bodenveranderungen und zur Sanierung von Altlasten, BundesBodenschutzgesetz (BBodSchG).

Bundesministerium fur Umwelt, Naturschutz, Bau und Reaktorsicherheit (1998c). Verord-nung uber Deponien und Langzeitlager (Deponieverordnung, DepV).

Bundesministerium fur Umwelt, Naturschutz, Bau und Reaktorsicherheit (2001). Verordnunguber die Qualitat von Wasser fur den menschlichen Gebrauch (Trinkwasserverordnung -TrinkwV).

Bundesministerium fur Umwelt, Naturschutz, Bau und Reaktorsicherheit (2002). Bundes-Immissionsschutzgesetz (BImSchG).

Bundesministerium fur Vertriebene, Fluchtlinge und Kriegsgeschadigte (1967). Dokumentedeutscher Kriegsschaden: Evakuierte, Kriegssachgeschadigte, Wahrungsgechadigte: diegeschichtliche und rechtliche Entwicklung, Volume 4. Bundesministerium fur Vertriebene,Fluchtlinge und Kriegsgeschadigte.

Bundesverband der Gipsindustrie (2009). IGB Handbuch Gipsputze. Technical report, Bun-desverband der Gipsindustrie e.V.

Burghardt, W. (1994). Soils in urban and industrial environments. Zeitschrift fur Pflanzen-ernahrung und Bodenkunde 157(3), 205–214.

Butera, S., T. H. Christensen, and T. F. Astrup (2014). Composition and leaching of con-struction and demolition waste: Inorganic elements and organic compounds. Journal ofhazardous materials 276, 302–311.

Cerniglia, C. E. (1984). Microbial metabolism of polycyclic aromatic hydrocarbons. Advancesin applied microbiology 30, 31–71.

Cheng, H. and Y. Hu (2011). Mercury in municipal solid waste in china and its control: areview. Environmental Science & Technology 46(2), 593–605.

Chuan, M., G. Shu, and J. Liu (1996). Solubility of heavy metals in a contaminated soil:effects of redox potential and ph. Water, Air, and Soil Pollution 90(3-4), 543–556.

Clement, D., K. Hammer, J. Schnoller, H. Daxbeck, and P. Brunner (2011). Wert-und schad-stoffe in wohngebauden. Osterreichische Wasser-und Abfallwirtschaft 63(3-4), 61–69.

Cloquet, C., J. Carignan, G. Libourel, T. Sterckeman, and E. Perdrix (2006). Tracing sourcepollution in soils using cadmium and lead isotopes. Environmental Science & Technol-ogy 40(8), 2525–2530.

Collins, J., J. Brown, S. Dawson, and M. Marty (1991). Risk assessment for benzo[a]pyrene.Regulatory Toxicology and Pharmacology 13(2), 170–184.

Councell, T., K. Duckenfield, E. Landa, and E. Callender (2004). Tire-wear particles as asource of zinc to the environment. Environmental Science & Technology 38(15), 4206–

74


4214.Cowam (2004). Construction and demolition waste management in Germany.Crammond, N. (2002). The occurrence of thaumasite in modern construction–a review. Ce-

ment and Concrete Composites 24(3), 393–402.Davis, A., M. Shokouhian, and S. Ni (2001). Loading estimates of lead, copper, cadmium,

and zinc in urban runoff from specific sources. Chemosphere 44(5), 997–1009.Davison, R. L., D. F. Natusch, J. R. Wallace, and C. A. Evans Jr (1974). Trace elements in

fly ash. Dependence of concentration on particle size. Environmental Science & Technol-ogy 8(13), 1107–1113.

Die Welt (07.05.2005). Deutschland - die Bilanz zur Stunde Null; von Katja Gerhartz.Doick, K. J., E. Klingelmann, P. Burauel, K. C. Jones, and K. T. Semple (2005). Long-term

fate of polychlorinated biphenyls and polycyclic aromatic hydrocarbons in an agriculturalsoil. Environmental science & technology 39(10), 3663–3670.

Eikmann, T. and A. Kloke (1993). Nutzungs-und schutzgutbezogene orientierungswerte fur(Schad-) Stoffe in Boden.

El Khalil, H., C. Schwartz, O. El Hamiani, J. Kubiniok, J. L. Morel, and A. Boularbah (2013).Distribution of major elements and trace metals as indicators of technosolisation of urbanand suburban soils. Journal of Soils and Sediments 13(3), 519–530.

Essumang, D., K. Kowalski, and E. Sogaard (2011). Levels, distribution and source character-ization of polycyclic aromatic hydrocarbons (pahs) in topsoils and roadside soils in esbjerg,denmark. Bulletin of Environmental Contamination and Toxicology 86(4), 438–443.

European Parliament and Council of the European Union (1998). Council Directive on thequality of water intended for human consumption 98/83/EC.

European Parliament and Council of the European Union (2000). Water Framework Directive2000/60/EC.

European Parliament and Council of the European Union (2001). Waste Framework Directive2000/532/EC.

European Parliament and Council of the European Union (2006). Groundwater Directive2006/118/EC.

Fallman, A. (2000). Leaching of chromium and barium from steel slag in laboratory and fieldtests - a solubility controlled process? Waste Management 20(2), 149–154.

Fellmer, B., T. Schneider, and R. Zweer (1993). Umweltatlas Berlin. Senatsverwaltung furStadtentwicklung und Umweltschutz, Berlin (Germany).

Fichtner, V. (1977). Die anthropogen bedingte Umwandlung des Reliefs durchTrummeraufschuttungen in Berlin (West) seit 1945. Ph. D. thesis, Selbstverlag der Ge-ographischen Instituts der Freien Universitat Berlin.

Freney, J. and F. Stevenson (1966). Organic sulfur transformations in soils. Soil Sci-ence 101(4), 307–316.

Freyburg, E. and A. Berninger (2003). Field experiences in concrete deterioration by thaum-asite formation: possibilities and problems in thaumasite analysis. Cement and ConcreteComposites 25(8), 1105–1110.

Galloway, J., J. Thornton, S. Norton, H. Volchok, and R. McLean (1982). Trace metals inatmospheric deposition: a review and assessment. Atmospheric Environment (1967) 16(7),1677–1700.

Galvın, A. P., J. Ayuso, J. R. Jimenez, and F. Agrela (2012). Comparison of batch leachingtests and influence of ph on the release of metals from construction and demolition wastes.Waste management 32(1), 88–95.

Ge, Y., P. Murray, and W. Hendershot (2000). Trace metal speciation and bioavailability in

75


urban soils. Environmental Pollution 107(1), 137–144.Gee, C., M. H. Ramsey, J. Maskall, and I. Thornton (1997). Mineralogy and weathering

processes in historical smelting slags and their effect on the mobilisation of lead. Journalof Geochemical Exploration 58(2), 249–257.

Gerke, H. and M. v. Genuchten (1993). A dual-porosity model for simulating the prefer-ential movement of water and solutes in structured porous media. Water Resources Re-search 29(2), 305–319.

Gerstenberg, J. and U. Smettan (2005). Erstellung von Karten zur Bewertung der Bodenfunk-tionen. Unveroff. Bericht im Auftrag der Senatsverwaltung fur Stadtentwicklung Berlin,Stand 15(09), 2005.

Giacometti, C., L. Cavani, G. Baldoni, C. Ciavatta, C. Marzadori, and E. Kandeler (2014).Microplate-scale fluorometric soil enzyme assays as tools to assess soil quality in a long-term agricultural field experiment. Applied Soil Ecology 75, 80–85.

Gierig, Scheithauer, Erhart, and Rexer (2006). Vergleichende Untersuchung von praxis-tauglichen in-situ-Verfahren und Laborverfahren zur Bewertung der Sickerwasserbeschaf-fenheit von Altlastenverdachtsflachen in Bayern. Abschlussbericht. Bayerisches Landesamtfur Umwelt (LfU, Referat 76, Stoff- und Chemikalienbewertung).

Gierig, M., H. Scheithauer, M. Erhard, and G. Rexter (2006). Vergleichende Untersuchungvon praxistauglichen in-situ-Verfahren und Laborverfahren zur Bewertung der Sicker-wasserbeschaffenheit von Altlastenverdachtsflachen in Bayern. Technical report, Bay-erisches Landesamt fur Umwelt.

Glugla, G., M. Goedecke, G. Wessolek, and G. Furtig (1999). Langjahrige Abflußbildung undWasserhaushalt im urbanen Gebiet Berlin. Wasserwirtschaft 89(1), 34–42.

Goetz, D. and A. Baasch (2002). Untersuchungen zur Umweltvertraglichkeit von Recycling-Baustoffen unter realen Einsatzbedingungen. Technical report, Institut fur Bodenkunde,Universitat Hamburg, Hamburg.

Gras, B., C. Jaeger, and S. Sievers (2000). Gehalte an polycyclischen aromatischen Kohlen-wasserstoffen (PAK) in Oberboden Hamburgs. Umweltwissenschaften und Schadstoff-Forschung 12(2), 75–82.

Grenzius, R. (1987). Die Boden Berlins (West): Klassifizierung, Vergesellschaftung,okologische Eigenschaften. Ph. D. thesis, Technische Universitat Berlin.

Grischek, T., W. Nestler, D. Piechniczek, and T. Fischer (1996). Urban groundwater in Dres-den, Germany. Hydrogeology Journal 4(1), 48–63.

Hampsoim, C. and J. Bailey (1982). On the structure of some precipitated calcium alumino-sulphate hydrates. Journal of Materials Science 17(11), 3341–3346.

Hannappel, S., M. Asbrand, K. Duscher, C. Kulick, and F. Jakobs (2003). Risikoabschatzungdiffuser Schadstoffquellen fur das Berliner Grundwasser. Technical report, Senatsverwal-tung fur Stadtentwicklung, Berlin.

Hannappel, S. and F. Jakobs (2002). Bericht zur Grundwasserbeschaffenheit 1995-2000 imLand Brandenburg. Technical report, Ministerium fur Landwirtschaft, Umweltschutz undRaumordnung, Brandenburg.

Hannappel, S. and S. Reinhardt (2002). Hydrochemische Charakterisierung des Grundwassersin Berlin. Technical report, Senatsverwaltung fur Stadtentwicklung, Berlin.

Hattemer-Frey, H. A. and C. C. Travis (1991). Benzo-a-pyrene: environmental partitioningand human exposure. Toxicology and Industrial Health 7(3), 141–157.

Hawley, J. K. (1985). Assessment of health risk from exposure to contaminated soil. RiskAnalysis 5(4), 289–302.

Heizer, W. D., R. S. Sandler, E. Seal, S. C. Murray, M. G. Busby, B. G. Schliebe, and S. N.

76


Pusek (1997). Intestinal effects of sulfate in drinking water on normal human subjects.Digestive diseases and sciences 42(5), 1055–1061.

Helmreich, B., R. Hilliges, A. Schriewer, and H. Horn (2010). Runoff pollutants of a highlytrafficked urban road–correlation analysis and seasonal influences. Chemosphere 80(9),991–997.

Hill, D. (1984). Diffusion coefficients of nitrate, chloride, sulphate and water in cracked anduncracked chalk. Journal of soil science 35(1), 27–33.

Hiller, D. (1996). Okologische Standorteigenschaften urban-industriell uberformter Bodendes Brucktorviertels in Oberhausen (Ruhrgebiet). Zeitschrift fur Pflanzenernahrung undBodenkunde 159(3), 241–249.

Hiller, D. and H. Meuser (1998a). Ausgangssubstrate der Bodenbildung in Stadtboden.Springer.

Hiller, D. and H. Meuser (1998b). Gefahrdungspotentiale von Stadtboden. In Urbane Boden,pp. 93–103. Springer.

Hobbs, D. and M. Taylor (2000). Nature of the thaumasite sulfate attack mechanism in fieldconcrete. Cement and concrete research 30(4), 529–533.

Howard, J., B. Dubay, and W. Daniels (2013). Artifact weathering, anthropogenic micropar-ticles and lead contamination in urban soils at former demolition sites, detroit, michigan.Environmental Pollution 179, 1–12.

Illera, V., F. Garrido, S. Serrano, and M. T. Garcıa-Gonzalez (2004). Immobilization of theheavy metals cd, cu and pb in an acid soil amended with gypsum-and lime-rich industrialby-products. European Journal of Soil Science 55(1), 135–145.

IUSS, WRB (2006). World reference base for soil resources 2006.Jacobs, D., R. Clickner, J. Zhou, S. Viet, D. Marker, J. Rogers, D. Zeldin, P. Broene, and

W. Friedman (2002). The prevalence of lead-based paint hazards in us housing. Environ-mental Health Perspectives 110(10), A599.

Jang, Y.-C. and T. Townsend (2001). Sulfate leaching from recovered construction and demo-lition debris fines. Advances in Environmental Research 5(3), 203–217.

Jang, Y.-C. and T. G. Townsend (2003). Effect of waste depth on leachate quality from labora-tory construction and demolition debris landfills. Environmental Engineering Science 20(3),183–196.

Jeschke, A. A., K. Vosbeck, and W. Dreybrodt (2001). Surface controlled dissolution ratesof gypsum in aqueous solutions exhibit nonlinear dissolution kinetics. Geochimica et Cos-mochimica Acta 65(1), 27–34.

Jim, C. (1998). Physical and chemical properties of a hong kong roadside soil in relation tourban tree growth. Urban Ecosystems 2(2-3), 171–181.

Juhasz, A. L. and R. Naidu (2000). Bioremediation of high molecular weight polycyclicaromatic hydrocarbons: a review of the microbial degradation of benzo [¡ i¿ a¡/i¿] pyrene.International Biodeterioration & Biodegradation 45(1), 57–88.

Kalbe, U., W. Berger, F.-G. Simon, J. Eckardt, and G. Christoph (2007). Results of interlab-oratory comparisons of column percolation tests. Journal of hazardous materials 148(3),714–720.

Keiderling, G. (1999). Zur Enttrummerung Berlins. Edition Luisenstadt.Keren, R. and P. Kauschansky (1981). Coating of calcium carbonate on gypsum particle

surfaces. Soil Science Society of America Journal 45(6), 1242–1244.Keren, R. and G. O’Connor (1982). Gypsum dissolution and sodic soil reclamation as affected

by water flow velocity. Soil Science Society of America Journal 46(4), 726–732.Kirsch, H. and S. Pollmann (1966). Entstehung, Phasenbestand und Reaktionen von

77


Brennstoffschlacken in Hochdruckdampfkraftwerken. Kristall und Technik 1(4), 643–651.Kneib, W. and A. Braskamp (1990). Vier Jahre Stadtbodenkartierung von Hamburg–Probleme

und Ergebnisse. Mitt. Dtsch. Bodenkundl. Gesellsch 61, 97–104.Kontoyannis, C., M. Orkoula, and P. Koutsoukos (1997). Quantitative analysis of sulfated

calcium carbonates using raman spectroscopy and x-ray powder diffraction. Analyst 122(1),33–38.

Krauss, M., W. Wilcke, and W. Zech (2000). Polycyclic aromatic hydrocarbons and polychlo-rinated biphenyls in forest soils: depth distribution as indicator of different fate. Environ-mental Pollution 110(1), 79–88.

Kreislaufwirtschaft Bau (2010). Aktueller Monitoring-Bericht Datenbasis 2010.Kruskal, W. and W. Wallis (1952). Use of ranks in one-criterion variance analysis. Journal of

the American statistical Association 47(260), 583–621.Kutchko, B. G. and A. G. Kim (2006). Fly ash characterization by sem–eds. Fuel 85(17),

2537–2544.LABO (2003). Hintergrundwerte fur anorganische und organische Stoffe in Boden. Technical

report, Landerarbeitsgemeinschaft Bodenschutz.Laidlaw, M. A., S. Zahran, H. W. Mielke, M. P. Taylor, and G. M. Filippelli (2012). Re-

suspension of lead contaminated urban soil as a dominant source of atmospheric lead inbirmingham, chicago, detroit and pittsburgh, usa. Atmospheric Environment 49, 302–310.

Lander-Arbeitsgemeischaft Abfall (LAGA) (1997). Anforderungen an die stoffliche Verwer-tung von mineralischen Abfallen (LAGA M20) .

Landesanstalt fur Umwelt, Messungen und Naturschutz Baden-Wurttemberg (2009). Ana-lytische Untersuchung von Bauschuttrecyclingmaterial in Baden-Wurttemberg. Technicalreport, Landesanstalt fur Umwelt, Messungen und Naturschutz Baden-Wurttemberg.

Landesarbeitsgemeinschaft Abfall (1998). Technische Regeln Aschen und Schlacken aussteinkohlenbefeuerten Kraftwerken, Heizkraftwerken und Heizwerken). Merkblatt, Lan-desarbeitsgemeinschaft Abfall.

Larson, T. E. and R. V. Skold (1958). Laboratory studies relating mineral quality of water tocorrosion of steel and cast iron. Corrosion 14(6), 43–46.

Lasaga, A. C. (1984). Chemical kinetics of water-rock interactions. Journal of GeophysicalResearch: Solid Earth (1978–2012) 89(B6), 4009–4025.

Lopez Meza, S., U. Kalbe, W. Berger, and F.-G. Simon (2010). Effect of contact time on therelease of contaminants from granular waste materials during column leaching experiments.Waste management 30(4), 565–571.

Lorenz, K. and E. Kandeler (2005). Biochemical characterization of urban soil profiles fromstuttgart, germany. Soil Biology and Biochemistry 37(7), 1373–1385.

Lu, P.-Y., R. L. Metcalf, N. Plummer, and D. Mandel (1977). The environmental fate ofthree carcinogens: Benzo-(α)-pyrene, benzidine, and vinyl chloride evaluated in laboratorymodel ecosystems. Archives of environmental contamination and toxicology 6(1), 129–142.

Manceau, A. and K. L. Nagy (2012). Quantitative analysis of sulfur functional groups innatural organic matter by XANES spectroscopy. Geochimica et Cosmochimica Acta 99,206–223.

Manta, D., M. Angelone, R. Neri, and M. Sprovieri (2002). Heavy metals in urban soils: acase study from the city of Palermo (Sicily), Italy. Science of the Total Environment 300(1),229–243.

Marsh, K., R. Tillman, and J. Syers (1987). Charge relationships of sulfate sorption by soils.Soil Science Society of America Journal 51(2), 318–323.

Martınez-Lage, I., F. Martınez-Abella, C. Vazquez-Herrero, and J. L. Perez-Ordonez (2012).

78


Properties of plain concrete made with mixed recycled coarse aggregate. Construction andBuilding Materials 37, 171–176.

Marx, M.-C., M. Wood, and S. Jarvis (2001). A microplate fluorimetric assay for the study ofenzyme diversity in soils. Soil Biology and Biochemistry 33(12), 1633–1640.

Maxin, C. R. and I. Kogel-Knabner (1995). Partitioning of polycyclic aromatic hydrocarbons(pah) to water-soluble soil organic matter. European Journal of Soil Science 46(2), 193–204.

McCarthy, J. F., L. E. Roberson, and L. W. Burrus (1989). Association of benzo (a) pyrenewith dissolved organic matter: Prediction of k¡ sub¿ dom¡/sub¿ from structural and chemicalproperties of the organic matter. Chemosphere 19(12), 1911–1920.

McCreanor, P. T. and D. R. Reinhart (2000). Mathematical modeling of leachate routing in aleachate recirculating landfill. Water Research 34(4), 1285–1295.

McGill, W. and C. Cole (1981). Comparative aspects of cycling of organic c, n, s and p throughsoil organic matter. Geoderma 26(4), 267–286.

McNeiill, L. S. and M. Edwards (2001). Iron pipe corrosion in distribution systems. AmericanWater Works Association Journal) 93, 88–100.

Meuser, H. (1993). Technogene Substrate in Stadtboden des Ruhrgebietes. Zeitschrift furPflanzenernahrung und Bodenkunde 156(2), 137–142.

Meuser, H. (1996a). Schadstoffpotential technogener Substrate in Boden urban-industriellerVerdichtungsraume. Zeitschrift fur Pflanzenernahrung und Bodenkunde 159(6), 621–628.

Meuser, H. (1996b). Technogene Substrate als Ausgangsgestein der Boden urban-industriellerVerdichtungstraume: Dargestellt am Beispiel der Stadt Essen. Habilitationsschrift, Institutfur Pflanzenernahrung und Bodenkunde, Universitat Kiel.

Meuser, H. (2010). Contaminated urban soils, Volume 18. Springer.Meuser, H. and H.-P. Blume (2001). Characteristics and classification of anthropogenic soils

in the Osnabruck area, Germany. Journal of Plant Nutrition and Soil Science 164(4), 351–358.

Mielke, H. (1999). Lead in the inner cities. American Scientists 87, 62–73.Miguel, E., J. Llamas, E. Chacon, T. Berg, S. Larssen, O. Røyset, and M. Vadset (1997).

Origin and patterns of distribution of trace elements in street dust: unleaded petrol andurban lead. Atmospheric Environment 31(17), 2733–2740.

Morais, S., F. G. e Costa, and M. de Lourdes Pereira (2012). Heavy metals and human health.INTECH Open Access Publisher.

Mori, T., T. Nonaka, K. Tazaki, M. Koga, Y. Hikosaka, and S. Noda (1992). Interactionsof nutrients, moisture and ph on microbial corrosion of concrete sewer pipes. Water re-search 26(1), 29–37.

Morton-Bermea, O., E. H. Alvarez, I. Gaso, and N. Segovia (2002). Heavy metal concen-trations in surface soils from Mexico City. Bulletin of environmental contamination andtoxicology 68(3), 383–388.

Muller, A., E. Linß, and T. Schulz (2011). Vom Storstoff zum Rohstoff. RECYCLING Maga-zin 09, 26–30.

Muller, G. (1979). Schwermetalle in den Sedimenten des Rheins-Veranderungen seit 1971.Umschau 79(24), 778–783.

Musson, S. E., Q. Xu, and T. G. Townsend (2008). Measuring the gypsum content of c&ddebris fines. Waste management 28(11), 2091–2096.

Nagamori, M., Y. Watanabe, T. Hase, Y. Kurata, Y. Ono, and K. Kawamura (2007). A simpleand convenient empirical survey method with a soil electrical conductivity meter for incin-

79


eration residue-derived soil contamination. Journal of Material Cycles and Waste Manage-ment 9(1), 90–98.

Nehls, T., S. Rokia, B. Mekiffer, C. Schwartz, and G. Wessolek (2013). Contribution of bricksto urban soil properties. Journal of Soils and Sediments 13(3), 575–584.

Nriagu, J. and J. Pacyna (1988). Quantitative assessment of worldwide contamination of air,water and soils by trace metals. Nature 333(6169), 134–139.

OConnell, M., C. McNally, and M. G. Richardson (2010). Biochemical attack on concrete inwastewater applications: A state of the art review. Cement and Concrete Composites 32(7),479–485.

Oman, C. and H. Rosqvist (1999). Transport fate of organic compounds with water throughlandfills. Water research 33(10), 2247–2254.

Parkhurst, D. L., C. Appelo, et al. (1999). User’s guide to phreeqc (version 2): A computerprogram for speciation, batch-reaction, one-dimensional transport, and inverse geochemicalcalculations.

Parliament, E. and C. of the European Union (1999). Council Directive on the landfill ofwaste99/31/EC.

Parliament, E. and C. of the European Union (2003). Decision (2003/33/CE) establishingcriteria and procedure for the acceptance of waste at landfill pursuant to Article 16 of andAnnex II to Directive 1999/31/CE.

Peters, A., T. Nehls, H. Schonsky, and G. Wessolek (2014). Separating precipitation and evap-otranspiration from noise–a new filter routine for high-resolution lysimeter data. Hydrologyand Earth System Sciences 18(3), 1189–1198.

Pflugradt, D. (1995). Sanierung quecksilberbelasteter industriestandorte. In ProblemkreisAltlasten, pp. 315–351. Springer.

Pichtel, J., K. Kuroiwa, and H. Sawyerr (2000). Distribution of Pb, Cd and Ba in soils andplants of two contaminated sites. Environmental pollution 110(1), 171–178.

Pietsch, J. and H. Kamieth (1991). Stadtboden: Entwicklungen, Belastungen, Bewertung undPlanung. Blottner.

Prietzel, J., J. Thieme, U. Neuhausler, J. Susini, and I. Kogel-Knabner (2003). Speciation ofsulphur in soils and soil particles by x-ray spectromicroscopy. European Journal of SoilScience 54(2), 423–433.

Prietzel, J., J. Thieme, and M. Salome (2010). Assessment of sulfur and iron speciation ina soil aggregate by combined S and Fe micro-XANES: microspatial patterns and relation-ships. Journal of synchrotron radiation 17(2), 166–172.

Prietzel, J., J. Thieme, M. Salome, and H. Knicker (2007). Sulfur K-edge XANES spec-troscopy reveals differences in sulfur speciation of bulk soils, humic acid, fulvic acid, andparticle size separates. Soil Biology and Biochemistry 39(4), 877–890.

Raines, M. A. and T. A. Dewers (1997). Mixed transport/reaction control of gypsum dissolu-tion kinetics in aqueous solutions and initiation of gypsum karst. Chemical geology 140(1),29–48.

Ramke, H. (2005). Abschatzung der Konsequenzen fur die Verwertung von Massenabfallenbei verscharften Anforderungen des Boden- und Grundwasserschutzes. In Verwertung vonMassenabfallen , pp. 79–101.

Ravel, a. and M. Newville (2005). Athena, artemis, hephaestus: data analysis for x-ray ab-sorption spectroscopy using ifeffit. Journal of synchrotron radiation 12(4), 537–541.

RBB (2015). Bericht im Rundfunk Berlin-Brandenburg http://www.rbb-online.de/politik/beitrag/2015/03/berlins-trinkwasser-bedroht.html.

Ritter, A. and R. Munoz-Carpena (2013). Performance evaluation of hydrological models:

80


Statistical significance for reducing subjectivity in goodness-of-fit assessments. Journal ofHydrology 480, 33–45.

Rodrigues, S., M. Pereira, A. Duarte, F. Ajmone-Marsan, C. M. Davidson, H. Grcman, I. Hos-sack, A. S. Hursthouse, K. Ljung, C. Martini, et al. (2006). Mercury in urban soils: acomparison of local spatial variability in six european cities. Science of the Total Environ-ment 368(2), 926–936.

Rosqvist, H. and G. Destouni (2000). Solute transport through preferential pathways in mu-nicipal solid waste. Journal of contaminant hydrology 46(1), 39–60.

Rosqvist, N., L. Dollar, and A. Fourie (2005). Preferential flow in municipal solid wasteand implications for long-term leachate quality: valuation of laboratory-scale experiments.Waste management & research 23(4), 367–380.

Rossiter, D. (2007). Classification of urban and industrial soils in the world reference base forsoil resources (5 pp). Journal of Soils and Sediments 7(2), 96–100.

Roussat, N., J. Mehu, M. Abdelghafour, and P. Brula (2008). Leaching behaviour of hazardousdemolition waste. Waste management 28(11), 2032–2040.

Roy, A. (2009). Sulfur speciation in granulated blast furnace slag: An x-ray absorption spec-troscopic investigation. Cement and Concrete Research 39(8), 659–663.

Ruokojarvi, P., M. Aatamila, and J. Ruuskanen (2000). Toxic chlorinated and polyaromatichydrocarbons in simulated house fires. Chemosphere 41(6), 825–828.

Sandler, K. (2003). Analyzing what’s recyclable in C&D debris. BioCycle 44(11), 51–59.Scheffer, F. and P. Schachtschabel (2010). Lehrbuch der Bodenkunde, Volume 16. Enke

Stuttgart.Scherer, W. . H. (2009). Sulfur in soils. Journal of Plant Nutrition and Soil Science 172(3),

326–335.Schleuß, U., Q. Wu, and H.-P. Blume (1998). Variability of soils in urban and periurban areas

in northern germany. Catena 33(3), 255–270.Schonsky, H., A. Peters, F. Lang, S. Abel, B. Mekiffer, and G. Wessolek (2013). Sulfate

transport and release in technogenic soil substrates: experiments and numerical modeling.Journal of Soils and Sediments 13(3), 606–615.

Scott, P., S. Critchley, and F. Wilkinson (1986). The chemistry and mineralogy of somegranulated and pelletized blastfurnace slags. Mineralogical Magazine 50(355), 141–147.

Senatsverwaltung fur Gesundheit, Umwelt und Verbraucherschutz (2006). Grundwasser inBerlin Vorkommen, Nutzung, Schutz, Gefahrdung.

Shaw, R., M. Wilson, L. Reinhardt, J. Isleib, R. Gilkes, and N. Prakongkep (2010). Geochem-istry of artifactual coarse fragment types from selected new york city soils. In Proceedingsof the 19th World Congress of Soil Science.

Sheridan, S. K., T. G. Townsend, J. L. Price, and J. T. Connell (2000). Policy options forhazardous-building-component removal before demolition. Practice periodical of haz-ardous, toxic, and radioactive waste management 4(3), 111–117.

Short, J., D. Fanning, M. McIntosh, J. Foss, and J. Patterson (1986). Soils of the mall inwashington, dc: I. statistical summary of properties. Soil Science Society of America Jour-nal 50(3), 699–705.

Smettan, U., C. Ehrig, and J. Cersrenberg (1993). Belastung von Boden mit As, Pb undPAK in zwei Berliner Bezirken. In Mitteilgn. Dtsch. Bodenkdl. Gesellsch., Volume 72, pp.1259–1262.

Smettan, U. and B. Mekiffer (1996). Kontamination von Trummerschuttboden mit PAK.Zeitschrift fur Pflanzenernahrung und Bodenkunde 159(2), 169–175.

81


Solomon, D., J. Lehmann, and C. E. Martınez (2003). Sulfur K-edge XANES spectroscopyas a tool for understanding sulfur dynamics in soil organic matter. Soil Science Society ofAmerica Journal 67(6), 1721–1731.

Somasundaram, S., T.-W. Jeon, Y.-Y. Kang, W.-I. Kim, S.-K. Jeong, Y.-J. Kim, J.-M. Yeon,and S. K. Shin (2015). Characterization of wastes from construction and demolition sector.Environmental monitoring and assessment 187(1), 1–14.

Sonntag, H. (2013). Problem Sulfat in der Spree - Stand der Diskussion und aktuelle Trends .Spearman, C. (1910). Correlation calculated from faulty data. British Journal of Psychol-

ogy 3(3), 271–295.Speir, T., E. A. Pansier, A. Cairns, et al. (1980). A comparison of sulphatase, urease and

protease activities in planted and in fallow soils. Soil biology and biochemistry 12(3), 281–291.

Stanko-Golden, K., W. Swank, and J. Fitzgerald (1994). Factors affecting sulfate adsorption,organic sulfur formation, and mobilization in forest and grassland spodosols. Biology andfertility of soils 17(4), 289–296.

Statistisches Landesamt Berlin (1963). Geschaftszeichen stgt. 1a -0392-1.Strayer, R. F. and E. C. Davis (1983). Reduced sulfur in ashes and slags from the gasification

of coals: Availability for chemical and microbial oxidation. Applied and environmentalmicrobiology 45(3), 743–747.

Susset, B. and W. Leuchs (2008). Ableitung von Materialwerten im Eluat und Ein-baumoglichkeiten mineralischer Ersatzbaustoffe: Umsetzung der Ergebnisse des BMBF-Verbundes” Sickerwasserprognose” in konkrete Vorschlage zur Harmonisierung vonMethoden;. [Forschungsprojekt im Auftrag des Umweltbundesamtes, FuE-VorhabenForderkennzeichen (UFOPLAN) 20574251]; Abschlussbericht; Umweltbundesamt.

Tabatabai, M. and J. Bremner (1970). Factors affecting soil arylsulfatase activity. Soil ScienceSociety of America Journal 34(3), 427–429.

Tabatabai, M. and J. Bremner (1972a). Distribution of total and available sulfur in selectedsoils and soil profiles. Agronomy journal 64(1), 40–44.

Tabatabai, M. and J. Bremner (1972b). Forms of sulfur, and carbon, nitrogen and sulfurrelationships, in iowa soils. Soil science 114(5), 380–386.

Tiller, K. (1992). Urban soil contamination in Australia. Soil Research 30(6), 937–957.Tolaymat, T. M., T. G. Townsend, and H. Solo-Gabriele (2000). Chromated copper arsenate-

treated wood in recovered wood. Environmental Engineering Science 17(1), 19–28.Townsend, T., T. Tolaymat, K. Leo, and J. Jambeck (2004). Heavy metals in recovered fines

from construction and demolition debris recycling facilities in florida. Science of the totalenvironment 332(1), 1–11.

Trankler, J. O., I. Walker, and M. Dohmann (1996). Environmental impact of demolitionwastean overview on 10 years of research and experience. Waste management 16(1), 21–26.

Trinks, S., H. Stoffregen, and G. Wessolek (2007). Hydraulische Eigenschaften vonTrummerschuttboden, Tagung der Deutschen Bodenkundlichen Gesellschaft 2007 in Dres-den. Mitteilungen der Deutschen Bodenkundlichen Gesellschaft 106(1), 141–142.

Umweltamt Dresden (2010). Umweltbericht Grundwasser. Environmental monitoring report,Umweltamt Dresden.

Umweltatlas Berlin (2008). Umweltatlas Berlin, Senatsverwaltung fur Stadtentwicklung,Berlin. Technical report, Senatsverwaltung fur Stadtentwicklung, Berlin.

Umweltbundesamt (2008). Steigerung von Akzeptanz und Einsatz mineralischerSekundarrohstoffe unter Berucksichtigung schutzgutbezogener und anwendungsbezogener

82


Anforderungen .USDA, US Department of Agriculture (2014). Keys to Soil Taxonomy, 12th Edition. Techni-

cal report, Natural Resources Conservation Service.Vairavamurthy, A. (1998). Using x-ray absorption to probe sulfur oxidation states in complex

molecules. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 54(12),2009–2017.

van Genuchten, M. T. (1985). A general approach for modeling solute transport in structuredsoils. Mem. Int. Assoc. Hydrogeol 17(2), 513–526.

Van Wijnen, J., P. Clausing, and B. Brunekreef (1990). Estimated soil ingestion by children.Environmental research 51(2), 147–162.

Vulpius, B.; Unternehmerverband Mineralische Baustoffe (UVMB) e. V. (2012). Verwertungmineralischer Abflle und Recycling Stand und Perspektiven und Perspektiven.

Wahlstrom, M., J. Laine-Yliijoki, h. Jarnstrom, T. Kaartinen, M. Erlandsson, A. Palm Cousins,O. Wik, P. Suer, A. Oberender, O. Hjelmar, et al. (2014). Environmentally sustainable con-struction products and materials–assessment of release. Technical report, Nordic Innova-tion.

Wehrer, M. and K. U. Totsche (2008). Effective rates of heavy metal release from alkalinewastesquantified by column outflow experiments and inverse simulations. Journal of con-taminant hydrology 101(1), 53–66.

Wessolek, G., B. Kluge, A. Toland, T. Nehls, E. Klingelmann, Y. Rim, B. Mekiffer, andS. Trinks (2011). Urban soils in the vadose zone. In Perspectives in Urban Ecology, pp.89–133. Springer.

Wisotzky, F. (2011). Angewandte Grundwasserchemie, Hydrogeologie und hydrogeochemis-che Modellierung: Grundlagen, Anwendungen und Problemlosungen. Springer DE.

Wobst, M., H. Wichmann, and M. Bahadir (1999). Surface contamination with PASH, PAHand PCDD/F after fire accidents in private residences. Chemosphere 38(7), 1685–1691.

Wolff, R. (1996). Typische Profile Hamburger Boden. In Urbaner Bodenschutz, pp. 129–143.Springer.

World Health Organization (WHO) (2004). Sulfate in Drinking-water: Background documentfor development of WHO Guidelines for drinking water quality.

Xia, K., F. Weesner, W. Bleam, P. Helmke, P. Bloom, and U. Skyllberg (1998). XANES studiesof oxidation states of sulfur in aquatic and soil humic substances. Soil Science Society ofAmerica Journal 62(5), 1240–1246.

Zeiss, C. and W. Major (1992). Moisture flow through municipal solid waste: patterns andcharacteristics. Journal of Environmental Systems 22(3), 211–231.

Zucker, A. and W. Zech (1985). Sulfur status of four uncultivated soil profiles in northernBavaria. Geoderma 36(3), 229–240.

83


List of building rubble landfills (>25.000 m3) in Berlin; locations see Figure 1.8volume area secondary maximum

components heigth# [1000 m3] [ha] [m]

Charlottenburg-Wilmersdorf

1 Teufelsberg 26000∗ ∼100 622 Schlosspark Charlottenburg 270∗ ∼82a AVUS Nordschleife 503

2b Murellenschlucht 7403

3 Stadion Wilmersdorf 1100∗ ∼25

Friedrichshain-Kreuzberg

4 Volkspark Friedrichshain 2100∗ ∼15 34

Mitte

5 Humboldthohe 1580∗ ∼16 385a Louise-Schroeder-Platz 1203 ∼4 - -6 Fritz-Schloss-Park 995∗ ∼10

Mahrzahn-Hellersdorf

Construction and building debris, deposited in the 80ies7 Ahrensfelder Berge ca. 35 natural soil 488 Kienberg ∼20 natural soil 449 Biesdorfer Hohe ∼14 natural soil 37

Neukolln

10 Rixdorfer Hohe 750∗ ∼20 2111 Rudower Hohe 700∗ ∼7 municipal waste 2712 Dorferblick ∼8 ∼30

Pankow

13 Anton SaefkowPark ∼4 ∼514 Volkspark Prenzlauer Berg 3000† ∼23 3515 Jahn Sportpark Mauerpark 800‡ ∼5 ∼716 Andreas Hofer Platz ∼1 ∼4

Reinickendorf

17 Borsigdamm 560∗ ∼3 k.A.

Spandau

18a Wilhelm v. Siemens Park 85†

18 Fort Hahneberg 8000⋆ ∼12 municipal waste 16

Steglitz-Zehlendorf

19 Osdorfer Straße (Rodelberg) ∼3 ∼14

Tempelhof-Schoneberg

20 Marienhohe 190∗ ∼5 1021 Insulaner 1400∗ ∼13 3022 Diedersdorfer Weg 600∗ ∼8 20

⋆ planed for the the deposition of 8 Mio. m3. Finally deposited quantity is not documented.∗ according to Fichtner 1977† according to surveys‡ according to Arnold 1999

84

building rubble composed soils : contamination status and ... · building rubble composed soils:...

Documents