Materials Transactions, Vol. 44, No. 7 (2003) pp. 1255 to 1258 Special Issue on Growth of Ecomaterials as a Key to Eco-Society #2003 The Japan Institute of Metals

Building Materials from Waste

Karin Weimann1;*, Lutz B. Giese1,Gu¨nter Mellmann2 and Franz-Georg Simon1 1Division of Waste Treatment and Remedial Engineering, Federal Institute for Materials Research and Testing, 12200 Berlin, Germany 2Division of Building Materials, Federal Institute for Materials Research and Testing, 12200 Berlin, Germany

In Germany, about 60 million tons of demolition waste are generated annually. Approximately 70% of the demolition waste is currently recycled. Most recycling applications, for example in roadbed substructures, can be seen as a kind of downcycling. However, there are also some high-level applications for demolition waste. For instance, crushed and sieved demolition waste can be used as concrete aggregate instead of natural materials. Depending on the crushing process, about a third of the broken concrete is currently concrete . At present, concrete sand fraction (0–4 mm) is rarely reused as a concrete aggregate. This is due to the fact that some important characteristics of crushed concrete sand are different from those of natural sand. As a result, it does not reach the standards required for aggregates. Concrete containing such aggregate is of lower quality than concrete made of natural aggregate. This paper describes an experimental investigation on the treatment of concrete sand gained from demolition waste by wet processing using a jig whereby the sand is separated by grain size and, more importantly, by density. Concrete produced with concrete sand which has been treated in this manner should have the same quality as concrete prepared with natural aggregate. Recycled sand used as aggregate substitutes natural resources and avoids disposal of waste. It has all the characteristics of an Ecomaterial.

(Received January 21, 2003; Accepted April 8, 2003) Keywords: demolition waste, jig, wet treatment, ecomaterial, recycled aggregates, material properties

1. Introduction investigate the effects of wet treated fine aggregates upon the material properties of concrete and . Concrete will continue to be the dominating construction material in the future. Like in other industries, progress in 2. Ecomaterials with a Green Environmental Profile concrete technology must take into account the world wide need to conserve both resources and the environment, to Gravel and sand are natural resources with virtually reduce waste and to effectively utilize energy.1) unlimited reservoirs but also with a high exploitation rate. Using crushed concrete from demolition waste as concrete The conditions of formation are sedimentary, i.e. mainly aggregate instead of natural materials on a large scale can glacigene, marine and fluviatile.9) In 1999, 383 million tons make a significant contribution to preserving natural resour- of sand and gravel were consumed in Germany. Approxi- ces and to the reduction of the amount of demolition waste. mately 45% of the sand and gravel is used for ready mixed During the last few years, several research projects have been concrete and mortar. Other applications are in pavements or conducted on the use of recycled aggregates in concrete. base layers for roads and in additional forms of utilization in These projects have mainly investigated coarse-grained civil engineering. The exploitation damages the natural concrete obtained from demolition waste by dry recycling landscape, since large amounts of land are used up in the technologies. These studies have demonstrated that recycled process (0.02 m2 of land is used for the production of 1 ton of aggregates can be used instead of natural aggregates under rock).10) certain conditions or they can be used as supplements to Today, concrete is the most important in natural aggregates.2–4) industrialized countries. It consists of gravel, sand and Attempts on the use of recycled sand in the production of . In Germany, cement consumption was 40 million concrete have been less successful. Therefore, the possibility tons in 1999. Limestone, and other rocks and minerals of using demolition waste sand fraction in concrete is rather are used as raw material. For the calcination of the raw limited to date. In Germany, for example, recycled sand may material approximately 1000 kWh of thermal energy is not be used in concrete for external wall units.5) Considering required to produce 1 ton of cement clinker.11) Therefore, that currently 20–50% of the broken concrete is obtained as the aspect of energy consumption is at least as important for concrete sand, the potential of concrete sand fraction, cement production as the exploitation of natural resources. especially when taking into account the extensive building The substitution of fossil fuel for the calcination process by activity and concrete consumption during the last 40 years, high-calorific waste and by reducing the environmental load will inevitably result in a significant increase in concrete by reducing the calcination temperature are possibilities to demolition waste in the coming years.6) save energy resources.12) Using ashes from incineration Prior investigations on the replacement of natural sand by processes as a substitute for lime and clay is also a promising concrete sand fraction obtained by dry treatment were not route to replace natural resources by waste materials.13–15) very successful.7,8) In this study it was therefore examined The resulting product, called Ecocement, is already certified whether wet treatment can improve the material properties of in Japan (JIS R 5214:2002).16) the concrete sand fraction. A series of tests was conducted to It is currently impossible to substitute concrete as the major building material in industrialized countries by renew- *Corresponding author: E-mail address: [email protected] able resources (e.g. wood) because neither production rate 1256 K. Weimann, L. B. Giese, G. Mellmann and F.-G. Simon nor performance is sufficient. The consumption of building settleable solids and abraded cement matrix. materials is steadily increasing because the lifetime of The remaining sand fraction of 0.1–4 mm passed through buildings is limited (mostly due to technological obsoles- the air-pulsed jig. In the jig the sand fraction was separated by cence, not because of structural deterioration). Recycling density and partly by classification. The separation is based processes for waste materials should result in products which on the different densities of natural grains (2.65 t/m3 of quartz are sufficient in performance to substitute natural materials. for example), concrete (2.2–2.5 t/m3) and of hardened cement For demolition waste this means that sand and gravel can paste (1.3–1.8 t/m3). The pulsating water in the jig causes the potentially be utilized as aggregate for concrete production. particles to float within the water. Particles of different In that case, downcycling or disposal of waste could be density stratify within the material flow: heavy grains sink avoided. Materials which are not disposed of in landfills are and lighter particles rise to an upper zone, so that both regarded as Ecomaterials. Halada and Yamamoto have fractions can be discharged separately.18) Applying such a developed a useful classification of different types of wet treatment, the concrete waste sand is separated into three Ecomaterials.17) Sand from demolition waste which can be fractions: fine, light and heavy material. used as an aggregate for concrete production is, therefore, a Different test runs were carried out with various jig and good example for an Ecomaterial with a green environmental stirring unit settings. Samples of the three fractions and of the profile. raw material (the dry treated concrete sand) from every test were analyzed by chemical/physical analyses (acid solubles 3. Experimental Section and acid-insolubles, sieve analysis and microscopy) and, except for the fine fraction, building materials tests dynamic For the study, a concrete sand fraction was chosen which modulus of elasticity (Young’s modulus), compressive was produced by crushing and dry sieving concrete demoli- strength, flexural strength and workability, i.e. consistency tion waste from the selective dismantling of a combined heat- and water absorption. power plant. Laboratory tests showed that the material was Due to the grain size of the building material, tests were not contaminated. conducted on mortar test cubes, all made with the same The concrete sand fraction (0–4 mm) was treated in a pilot mortar mixture. Also a mass balance for every test run in the plant, which combines a stirring unit to remove the binder pilot plant was calculated. matrix from the grains, a hydrocyclone to remove particles below 100 mm (fine fraction) and, most importantly, a jig (see 4. Results Fig. 1) to separate the material by density (light fraction and heavy fraction). The wet treatment of the concrete sand fraction changed The pilot plant can be operated continuously. Depending the granulometric composition of the material. The mechan- on the material’s composition, the flow-rate ranges between ical load of the material in the pilot plant caused an increase 0.3 and 1.5 t/h. For the concrete sand fraction, the flow-rate of finer grains in the total mass of the treated material. This varied between 0.6–0.65 t/h. Before entering the stirring unit, rise is mainly caused by abrasion of adherent cement matrix. water was added to the material. The stirrer settings were The differences in the granulometric composition of the total either 100, 600 or 1000 revolutions per minute (rpm). Then mass and the different particle size distribution curves are the hydrocyclone removed the fine fraction consisting of shown in Figs. 2 and 3. The results of sieve analysis show - as expected - a significant reduction of settleable solids in the heavy and light fractions. The content of settleable solids in aggregates is limited by specific guidelines.5) Sieving results also show that simultaneously particles of bigger grain size accumu- lated within the heavy fraction (Fig. 3). The acid soluble fraction decreased due to the removal of adherent cement

6.0 AV 1000 4.0 AV 100 2.0 0.0 -2.0 Difference (%) -4.0 -6.0 0.10 1.00 10.00 Grain Size, D / mm

Fig. 2 Differences in the grading curves, average values (AV) of test runs Fig. 1 Jig integrated in the pilot plant for wet treatment of concrete sand. with two different stirrer settings (in revolutions per minute). Building Materials from Waste 1257

100.00 90.00 / MPa

80.00 D 20 σ 70.00 60.00 50.00 40.00 Input mat. 10 30.00 Heavy frac. Heavy fraction 20.00 Light frac. Input material

Percentage Passing Sieve 10.00 Fine frac. Light fraction 0.00 0 0.010 0.100 1.000 10.000 Compressive Strength, 0 7 14 21 28 Grain size, D / mm Time, t / days Fig. 3 Grain-size distribution curves of concrete sand before and after wet treatment. Fig. 4 Development of compressive strength. Average values of series mortar with different recycled aggregate types.

Table 1 Results of laboratory tests.

Input Heavy Light Fine Natural material fraction fraction fraction sand Settleable solids < 100 mm (%) 5.30 0.34 2.23 100.00 0.55 Acid-insolubles1 (%) 77.96 83.90 73.90 59.01 99.70 Water absorption2 (%) 9.92 7.00 11.87 n.m. n.m. Apparent density3 g/cm3 1.88 1.98 1.78 n.m. 2.49 Consistency4 cm 51.40 50.40 42.00 n.m. 39.00 Compressive strength5 MPa 17.69 19.94 18.93 n.m. 40.20 Flexural strength5 MPa 3.44 4.04 3.70 n.m. 5.80 Dynamic modulus of MPa 12700 16000 10600 n.m. 27300 elasticity5 1 percentage of dry matter. 2 percentage after 10 minutes. 3 (10 Â 10 Â 10)cm3 mortar-cube after 28 days. 4 slump test, DIN 1048 - 10 minutes. 5 (4 Â 4 Â 16)cm3 mortar-prism after 28 days. n.m.: not measured. paste on the aggregate particles. The grain size distribution material, which had only been treated dryly. Further has been examined before and after processing. The important results of the laboratory tests are summarized in differences are displayed in Fig. 2 and show that there is a Table 1. mechanical disaggregation during the processing. Mainly The results of processing and separation were analyzed grain sizes bigger than 2 mm decrease whereas the fine grains using mass balance, acidic digestion and sieving of the smaller than 0.1 mm increase. The diameters in between vary fractions. An accumulation of acid-insoluble particles, which only to a minor extent. The sum of these differences are zero. is a benchmark for the loss of adherent binder matrix, can be It seems that adherent cement gets rubbed off from the cores seen in the heavy fraction, while there is an enrichment of of the grains consisting of rock material. However, results acid-soluble particles in the fine fraction. There is a from XRF and acid leaching tests show that not only cement corresponding loss of water absorption in the heavy fraction but also quartz and other rock components are involved in -usually caused by the binder matrix- and an increase in the this process and can be found in the fine fraction. light fraction. The changes in both material properties (water The mechanical disaggregation vs. frequency relationship absorption and acid-insoluble content) and the increase in the of the stirring unit is not significant. When comparing the apparent density seem to indicate that the material properties particle size distribution curves of all fractions (Fig. 2), it can are approaching those of natural aggregates due to the wet be seen that there is no significant difference in the treatment of the sand fraction. granulometric composition having treated the material at The results of strength measurements on mortar with wet 100 and 1000 rpm. In view of the great difference in energy treated heavy fraction improved in comparison to mortar that consumption of the stirring unit at these two levels, it is not contained dry treated material (Figs. 4 and 5). The measured economical to use the higher frequency. Nevertheless, not values of compressive strength were up to 27% higher and on only the stirrer but also other process units can cause this average 13% higher. Flexural strength was up to 58% above decomposition. the input material, and on average 17%. Dynamic modulus of After the wet treatment of the concrete sand fraction, the elasticity showed a rise of up to 43% above dry treated material properties were improved compared with the input material, and on average 26%. Even the wet treated light 1258 K. Weimann, L. B. Giese, G. Mellmann and F.-G. Simon

were significantly worse than those of mortar with 16000 natural sand. Even the results of former tests on mortar

/ Mpa with the sand fraction of recycled concrete demolition E 12000 waste of another input material were better than the results obtained now. 8000 Heavy fraction (5) Variances in strength properties in these tests were not Input material proportional to the variances in acid-insolubles and 4000 Light fraction settleable solids. Further tests on the sand fraction from different concrete

Modulus of Elasticity, sources are planned to investigate the influence of varying 0 mineralogical compositions. 0 7 14 21 28 Time, t / days Acknowledgement Fig. 5 Development of modules of elasticity. Average values of series mortar with different recycled aggregate types. The present work is being funded by the European Union. It is part of the LIFE Environment program (Demonstration project LIFE00 ENV/D/000319). fraction showed better values for compressive strength and flexural strength. The dynamic modulus of elasticity of the REFERENCES light fraction was always below the values of the input material, on average 17% lower. 1) V. S. Ramachandran: Waste and By-Products as Concrete Aggregates. Although there is a clear improvement in the material Institute for Research and Construction IRC, Canadian Building properties of the wet treated heavy fraction, these results are Digest, CBD 215 (1981). 49 below the results for mortar made with natural aggregate, and 2) V. Penttala and J. Komonen: Magazine of Concrete Research (179) (1997) 81–97. lower than previous studies with concrete sand from a 3) R. Hohberg: Materialpru¨fung 42(7–8) (2002) 297–303. 4,19) different origin. This might be caused by a different 4) H.-J. Chen, T. Yen and K.-H. Chen: Cement and Concrete Research 33 mineral composition.2,20) In the microscopic examinations (2003) 125–132. which were conducted a significant percentage of feldspar, 5) Deutscher Ausschuss fu¨r Stahlbeton: DAfStb-Richtlinie ‘‘Beton mit which is more scissile than e.g. quartz, was found in the rezykliertem Zuschlag’’ Teil 1 und Teil 2, Bonn (1998). 6) H. Go¨rg: Bauen fu¨r den Umweltschutz: Es gibt viel zu tun - nur wissen aggregate. The particles were predominantly plate-shaped. wir nicht wann! Teil 2. Altlasten Spektrum 10 (2001) 152–153. 2,8,20) Other studies have shown that the aggregate type has an 7) C. Lemmer and M. Ru¨hl: Darmstadt Concrete 12 (1997). effect on strength properties. 8) A. Ajdukiewicz and A. Kliszczewicz: Cement & Concrete Composites As a further step to investigate the applicability of concrete 24 (2002) 269–279. sand fraction for recycled concrete aggregate after a wet 9) W. Gwosdz: Origin of the most important economic deposits of industrial minerals and rocks. http://www.bgr.de/b122 (2002). treatment, additional testing will be carried out on concrete 10) R. Ii, G. Ikemoto, K. Abe, A. Nakagawa, N. Itsubo and A. Inaba: 5th demolition sand of different origin and different mineralo- International Conference on EcoBalance, Tsukuba, Japan, (The gical compositions. Society of Non-Traditional Technology, 2002) pp. S1–118. 11) F. W. Locher and J. Kro¨pp: Cement and Concrete, in: Gerhartz, W., 5. Conclusions Yamamoto, Y. S., Campbell, F. T., Pfefferkorn, R. and Rounsaville, J. F. (Editors): Ullmann’s Encyclopedia of Industrial Chemistry, Vol. A 5, (VCH-Verlag, Weinheim 1986) pp. 490–516. In this study the material properties of wet treated 12) H. Uchikawa: Journal of Materials in Civil Engineering 12(4) (2000) demolition concrete sand was investigated in comparison 320–329. with dry treated material with regard to its applicability as a 13) M. Ohsumi: International Workshop on Ecomaterials, Tokyo, (Eco- concrete aggregate. The main results can be summarized as materials Forum, The Society of Non-Traditional Technology, 2002) pp. 94–98. follows: 14) H. Maekawa, S. Aoyama and M. Kagamida: Global cement and lime (1) The following material properties of the wet treated magazine, (Environmental special issue) (2002) 23–26. material improved: content of settleable solids in heavy 15) T. Shimoda, S. Yokoyama and H. Hirao: Eco-cement: Taiheiyo and light fraction, content of acid-insolubles, water Semento Kenkyu Hokoku 138 (2000) 5–15. The 2nd Workshop on absorption and density in the heavy fraction. 16) M. Ohsumi: JIS Standardization for Ecocement, Ecomaterials, Tsukuba, Japan, (The Society of Non-Traditional (2) The following strength properties of mortar with the Technology, 2002) pp. 107–112. heavy fraction were improved in comparison to mortar 17) K. Halada and R. Yamamoto: MRS Bulletin 26(11) (2001) 871–879. with dry treated sand fraction: compressive strength, 18) K. Mesters and H. Kurkowski: Aufbereitungs-Technik 38(10) (1997) flexural strength and dynamic modulus of elasticity. 536–542. 18 (3) Although the test results of acid-insolubles, water 19) K. Behler: Baustoff Recycling + Deponietechnik (BR) (6) (2002) 25–28. absorption and density of the light fraction were worse 20) H.-J. Jansky and G. Mellmann: Baustoff Recycling + Deponietechnik than those of dry treated material, compressive strength (BR) 17(4) (2001) 19–23. and flexural strength improved. (4) Material properties of mortar with recycled aggregates