ty ots, tn c). Bldn
energy consumption in US, China, India and EU has been
. Although large amount of y ash has already been uti-lized in the construction industry as a partial cement
and concrete products has been reported to be 47%, while
for reclamation and restoration . However, the fused andglassy texture of coal ash makes it an ideal substitute fornatural raw materials. Using siliceous ash in concrete andcement applications requires that the loss-on-ignition(LOI) content of the ash generally be less than 6% byweight in order to obtain the required pozzolanic proper-
* Corresponding author. Tel.: +90 222 2393756x3435; fax: +90 2222393613.
E-mail address: email@example.com (H. Kurama).
Available online at www.sciencedirect.com
Construction and Building Materia
Constructionrising since the mid-1990s and this trend is expected to con-tinue. European environment agency (EEA) has recentlyreported that fossils are presently dominating the fuelsources with an 80% share. This proportion is expectedto increase slightly over the next 30 years in the EU.Despite some growth in absolute terms, renewable energyis not expected to raise its share to a signicant extent,while contribution of nuclear power is projected to decline
for CBA this rate has been reported as only 5.28% of thetotal recycling amount- the total CBA production beingabout 19.8 M tonnes for 2002. 7.6 M tonnes of this produc-tion were recycled mainly in structural lls/embankments(26.61%), road base/subbase/pavement (19.15%) and min-ing applications (10.43%). The same utilization prolecan also be given for the EU. Nearly 89% of the producedCBA was recycled, but 54% of this amount was evaluatedKeywords: Coal bottom ash; Portland cement; Unburned carbon; Treatment; Concrete
The electricity industry, particularly coal-red powerplants, has been greatly aected by the increasing publicattention being paid to the environment. Coal ash gener-ated from power plants have become an important eco-nomic and environmental objective, and thus calls forrecycling alternatives to traditional landll option due totheir high generation amount and low recycling rate. Total
replacement and/or mineral additive in cement production,the usage of coal bottom ash (CBA) is limited due to its rel-atively higher unburned carbon content and dierent struc-tural properties compared to y ash . CBA, a coarsesand to ne gravel size material collected at the bottomof the boiler, is generally used as a low cost replacementmaterial either as a base material in road construction oras a blasting grit. According to the American Coal AshAssociation (ACAA), recycling rate of y ash in concreteUsage of coal combustion bo
Eskisehir Osmangazi University, Mining Engineerin
Received 20 February 2007; received in revAvailable onlin
The present study aims to determine and evaluate the applicabiliStation-Turkey, in concrete industry. In the laboratory experimenPortland cement. In order to be able to reduce the unburned carboclassication, heavy medium separation and electrostatic separationCBA up to 10% as a replacement material for Portland cement couused in the concrete industry. The eect of operating parameters o 2007 Elsevier Ltd. All rights reserved.0950-0618/$ - see front matter 2007 Elsevier Ltd. All rights reserved.doi:10.1016/j.conbuildmat.2007.07.008tom ash in concrete mixture
*, Mine Kaya
epartment, Bat Meselik, 26480 Eskisehir, Turkey
form 14 July 2007; accepted 16 July 20077 August 2007
f an industrial bottom ash (CBA), supplied from Tuncbilek Powerhe bottom ash was used up to 25% as a partial substitute for theontent, CBA was treated by three dierent processes (particle sizeased on the obtained results, it was concluded that the addition ofimprove the mechanical properties of concrete, and thus, could betreatment processes has also been discussed in the paper.
ls 22 (2008) 19221928
ties. This requirement results from the fact that propertiesof concrete incorporating high-carbon ash are inferior tothose of concrete incorporating low-carbon ash. Theamount of water and quantity of air entraining agents usedin the mix increases signicantly as the carbon content of
loss-on-ignition analysis (LOI) and the determined LOI isaccepted as the mass of unburned carbon in the originalsample, which is a common approach in cement and con-crete applications.
Particle size distribution of the representative sampleand unburned carbon content of each size fraction aregiven in Table 2.
According to wet screen analysis, the D80 and D50 ashsize values were calculated as 2.7 and 1 mm, respectively.As seen from Table 2, the unburned carbon in CBA hasmainly accumulated at coarse fractions. However, a littleamount of carbon-rich fraction is also present at0.053 mm. The specic gravity of the sample measuredby the pycnometer method was 2.39 g/cm3.
The crystalline mineral phases in the CBA were identi-ed by using X-Ray Diraction (XRD), model S5000 dif-fractometer, with a nickel ltered Cu Ka. Regarding theXRD analysis, it was found that CBA had a relatively sim-ple mineralogy consisting of alumina, glass and varyingamount of crystalline phases of quartz, ferrite spinel andcalcite (Fig. 1).
The scanning electron micrograph of ash shows spheri-cal, rounded and irregularly shaped grains (Fig. 2).
H. Kurama, M. Kaya / Construction and Building Materials 22 (2008) 19221928 1923the ash increases above 6% LOI level . Consequently,it is critical that removal/recovery technologies should bedeveloped so that high carbon CBA can be used moreextensively for industrial applications. A number of meth-ods have been developed and commercialized for removingthe unburned carbon particles from coal ash, includingcombustion of the carbon at low temperatures ,mechanical/particle size classication , gravity separa-tion , and electrostatic separation , froth o-tation [20,21] and, combinations thereof. Although, each ofthese methods may be used to remove carbon particlesfrom siliceous coal ash. However, the eciency of theemployed method and pozzolanic properties of the result-ing low-carbon siliceous fraction are highly dependent onthe physical characteristics of the original ash.
In Turkey, nearly 28% of the total primary energy issupplied from the burning of bituminous and lignite coal.The ashes produced annually from 11 power plants varyfrom 6.5 to 13 M tonnes. However, only 1% of these wastematerials are re-used, mainly in concrete industry.
The purpose of the present study is to test commonmethods (such as sink and oat test, particle size classica-tion, and electrostatic separation) as an eective and eco-nomical method for removing of unburned carbon fromTuncbilek Power Station bottom ash in order to enhanceits application as a constituent in concrete production.The paper also examines the eects of pre-treated CBAadditions on the nal concrete properties as a replacementfor Portland cement in cement mixture.
2. Materials and methods
2.1. Characterization of material
Lignite bottom ash used in this study was obtained fromTuncbilek Power Station, located in central Turkey. Thestation consumes about 2,350,000 tons of low-grade lignitecoal (caloric value varying from 2250 to 3900 kcal/kg)and generates 854,670 ton ash per year.
The chemical composition of CBA is given in Table 1.As can be followed from Table 1, the sum of SiO2 + Al2O3and Fe2O3 reach 81.06% in the composition, indicatingthat it could be classied as a type F ash as prescribed byASTM C 618. In this study, the total carbon content ofgenerated ash was determined by employing the standard
Table 1Main oxide composition of received CBA
SiO Fe O CaO Al O LOI2 2 3 2 3
54.5 11.16 4.69 15.4 8.902.2. Methods
188.8.131.52. Particle size separation. Mechanical means ofremoving carbon from siliceous ash is based on the relativeparticle size of the carbon particles and the siliceous parti-cles in the ash. In this study, the representative 500 g of
Table 2Particle size distributions of CBA and unburned carbon (C) values of eachsieve fractions
Particle size (mm) (%) R (%) C (%)
+2.36 31.49 31.49 8.972.360 + 2.000 1.96 33.45 21.682.000 + 0.600 26.98 60.43 12.960.600 + 0.212 31.14 91.57 5.220.212 + 0.149 3.59 95.16 3.010.149 + 0.075 2.62 97.78 3.790.075 + 0.053 0.48 98.26 6.990.053 1.74 100.00 13.13Fig. 1. X-ray diractogram of CBA.
eect of ash addition in cement bodies. The amounts men-tioned above were chosen so as to highlight the eects of
Sample Std BC5 BC10 BC15 BC25
Cement (g) 450 428 405 383 338Sand (g) 1350 1350 1350 1350 1350CBA (g) 22 45 67 112Water (g) 225 225 225 225 225
d BCBA was subjected to laboratory impact crusher with andwithout using 2 mm separating sieve. The crushed productwas then classied to +2.36, 2.362.00, 2.000.6, 0.60.212, 0.2120.149, 0.1490.075, 0.0750.053 and -0.053 mm size fractions by wet screening. The sink partof the screened fractions was separated by decantation inorder to remove oated particles, mainly carbon particles,before ltration (1st stage). Each collected fraction wasdried in oven at 100 C and weighed. In order to nd outthe crushing eect on the unburned carbon content of eachsize fraction, the obtained product from 1st stage crushing
Fig. 2. Morpholog
1924 H. Kurama, M. Kaya / Construction anwas then re-crushed (2nd stage).
184.108.40.206. Sink and oat tests. The sink and oat tests wereperformed on crushed samples at various densities to assessthe suitability of heavy medium separation. These experi-ments were conducted in 250 mL glass ax with a volumeof 100 mL, using an appropriate mixture of bromoformand alcohol to adjust the density of liquid between 1.02.4 g/cm3. A 50 g of representative ash sample was intro-duced in the liquid of highest density. The oating productwas removed, washed and then placed into the next lowerdensity and so on.
220.127.116.11. Electrostatic separation tests. Electrostatic separa-tion encompasses a number of dierent technologies whichare based on the electrical properties of the particles to beseparated. The electrostatic separation tests were carriedout by using a conductor/non-conductor type of separator(Boxmag Rapid Ltd-HT150).
2.2.2. Moulding of CBA paste specimens
Representative cement compositions were prepared byprogressive incorporation of pre-treated samples in placeof Portland cement (5,10,15, and 25 wt%) to observe thef the bottom ash.
Table 3Mixture proportion of bottom ash cement paste
uilding Materials 22 (2008) 19221928CBA addition. These compositions were designated asStd, BC5, BC10, BC15 and BC25, respectively. The mix-ture proportions of the ash-cement used in this study arelisted in Table 3. The pre-treated materials mixed in therequired proportions were ground in a ceramic ball millto a neness of the 25 mass% residues on a 38 lm toimprove its pozzolanic properties. The physical tests ofthe cement mixes were performed according to Turkishstandards TS EN 1971. The cementwater mixtures werestirred at a low speed for 30 s, and then with the addition ofsand, the mixture was stirred again for 5 min. Three40 40 160 mm prismatic specimens for compressiontests were prepared from each mixture. The moulded spec-imens were cured at 20 C with 95% humidity for 24 h, andthen after placed in a tap water and cured for 7, 28 and 56days.
3. Results and discussion
3.1.1. Particle size classicationIt is well known that in impact crushers, comminution is
performed by impact eect rather than compression .
Impact causes immediate fracture with no residual stresses.This stress-free condition is particularly valuable in thequarrying industry. When the crushing force is appliedinstantaneously, the impact crushers give trouble-freecrushing on ores which tend to be brittle. The impactcrushers are widely used in coal preparation, coal beingmuch more friable than the associated stones and rubbishsuch as wood, steel, etc. The results of crushing and screen-ing tests are summarized in Table 4. It is clear thatunburned carbon in CBA is mainly concentrated at coarsefractions such as 2.36 + 2.00 mm and 2.00 + 0.6 mm, validfor both employed crushing types (with and without sepa-
3.1.3. Electrostatic separation tests
The electrostatic separation test results of 2.00 + 0.6 mmand 0.6 + 0.212 mm size fractions of CBA are given inTable 6a and b, respectively. The tests were carried outunder constant feeding rate of 5.25 min/100 g and electricalpower of 30 kW. The rotor rate was steadily increased anddistribution of unburned carbon was analyzed according tothe rotor rate. The results show that electrostatic separa-tion is an ineective method for both tested samples. This
1st stagea 2nd stagea
C (%) wt% C (%) wt% C (%)
5.94 12.37 2.0 10.36. 0.73 12.4011.57 40.31 10.80 38.56 11.655.10 38.68 7.24 43.70 5.993.53 8.92 3.75 8.08 3.964.11 5.94 5.27 5.02 6.538.03 1.19 7.88 1.70 8.51
Table 5Heavy liquid test results
wt% R C(%)
(a)1.21 3.40 3.20 58.76 17.38 17.381.211.63 10.40 13.80 12.98 11.75 29.131.631.74 7.40 21.20 12.77 8.22 37.351.741.89 2.60 23.80 10.16 2.30 39.651.891.94 3.00 26.80 8.95 2.34 41.991.942.05 40.0 66.80 8.61 29.97 71.962.052.89 30.20 97.00 9.67 25.41 97.37+2.89 3.00 100.00 10.10 2.63 100.00
H. Kurama, M. Kaya / Construction and Building Materials 22 (2008) 19221928 1925rating sieve usage). Although there is a rough relationshipbetween particle size and unburned carbon as LOI, in thecase of the without separating sieve usage, the calculatedcumulative weight percent for fractions of(+2.36,0.6 + 0.212, 0.212 + 0.149 and 0.149 + 0.075)which have less than 6% LOI is 57.67%. This is more ben-ecial than fractions which have less than 6% LOI (51.78%)for separating sieve usage. This can be attributed to the fur-ther crushing eect of separation sieve on carbon particles.
In order to nd out the positive eects of other treat-ment methods, such as gravity and electrostatic separation,on the further decrease of unburned carbon content, thetreatment tests were performed for the 2.00 + 0.6 and0.6 + 0.212 mm size fractions of crushed samples. Consid-ering the relatively higher particle size of 2.36 + 2.00 mmand unsuitability of the electrostatic separation method,dierent gravity concentration method such as jiggingwas also tested for this size fraction. However, a sucientcarbon content reduction could not be achieved by employ-ing this method.
3.1.2. Sink and oat test
The analyzed unburned carbon content and weight dis-tributions of the size fractions of 2.00 + 0.6 mm and0.6 + 0.212 mm at dierent density ranges are given inTable 5a and b, respectively. It can be seen from columns3 and 6 of the Table 5a that, if separation density of 1.94is chosen, then 26.80% of ash, being lighter than 1.94, couldbe separated as a oat product. According column 6, only41.99% of the unburned carbon would be separated in thisdensity range. Conseque...