Available online at www.sciencedirect.com

www.fuelfirst.com

Fuel 87 (2008) 359–364

Durability of biomass fly ash concrete: Freezing and thawingand rapid chloride permeability tests

Shuangzhen Wang a, Emilio Llamazos a, Larry Baxter a,*, Fernando Fonseca b

a Department of Chemical Engineering, 350 CB, Brigham Young University, Provo, UT 84602, United Statesb Department of Civil Engineering, Brigham Young University, UT, United States

Received 1 September 2006; received in revised form 7 April 2007; accepted 3 May 2007Available online 11 June 2007

Abstract

Strict interpretation of ASTM C 618 excludes non-coal fly ashes, such as biomass fly ashes from addition in concrete. Biomass fly ashin this investigation includes (1) cofired fly ash from burning biomass with coal; (2) wood fly ash and (3) blended fly ash (wood fly ashmixing with coal fly ash). A set of experiments conducted on concrete from pure cement and cement with fly ash provide basic data toassess the effects of several biomass fly ashes on the performances of freezing and thawing (F–T) and rapid chloride permeability test(RCPT). The F–T tests indicate that all fly ash concrete has statistically equal or less weight loss than the pure cement concrete (control).The RCPT illustrate that all kinds of fly ash concrete have lower chloride permeability than the pure cement control concrete.� 2007 Elsevier Ltd. All rights reserved.

Keywords: Biomass fly ash; Permeability; Freezing and thawing

1. Introduction

Biomass combustion is assumed to be CO2 neutral pro-cess if its consumption rate is less than its growth rate,which is environmentally friendly and arouses great inter-ests of the world. However, biomass fly ash is excludedfrom addition in concrete by ASTM C 618 because of its‘‘non-coal’’ origin.

Entrainment of small air bubbles in the range of microor nanometers imbues concrete with resistance to freezingand thawing degradation [1]. The benefit of air entrainmentof concrete is shown in Table 1 and thus adequate amountof air content is crucial to the durability of concrete inrelated to F–T [2].

Coal fly ash addition has little direct effect on the con-crete performance by freezing and thawing, but it doesaffect the air content of concrete mixes through the behav-iors of air-entraining agent (AEA) [2–4] and mostly likely

0016-2361/$ - see front matter � 2007 Elsevier Ltd. All rights reserved.

doi:10.1016/j.fuel.2007.05.027

* Corresponding author. Tel.: +1 801 422 8616; fax: +1 801 422 0151.E-mail addresses: wangshuangzhen@gmail.com (S. Wang), larry_

baxter@byu.edu (L. Baxter).

brings severe air loss within 2 h of mixing [3]. Unburnedcarbon residue, the main form of LOI in the fly ash,adsorbs AEA; and the adsorption capacity not onlydepends on the amount of LOI, but also more on the car-bon forms and surface area available in LOI [5–7]. Further-more, water soluble alkalis from fly ash and cementdecrease air-entraining agent amount in concrete [8].

It is generally agreed that Class C fly ash has less carbonthan Class F; therefore, Class C fly ash need less AEAdemand than Class F in concrete mix. Furthermore, if suit-able amount of AEA is added in fly ash concrete to pro-duce desirable air void, fly ash does not affect the F–Tbehavior too much.

Biomass fly ash generally has more alkali and more LOIthan coal fly ash [9,10] and its mineralogical compositionwith coal fly ash can be also different [11,12]. High alkalicontent can cause serious alkali silica reaction (ASR)expansion concerns, and high LOI will have the same effectas that from coal fly ash, causing unstable air content lead-ing to poor durability by F–T if not dealt properly. How-ever, the research by the same authors from BYU havestudied the effect of cofired (biomass with coal) fly ash on

Table 1The effect of air content on durability factor [1]

Air content (%) Durability factor (%) by ASTM C 666

<3 <80>4 >85

360 S. Wang et al. / Fuel 87 (2008) 359–364

ASR expansion, and the results show that although bio-mass fly ash has much more available alkali than that ofClass C fly ash, it is still more effective in reducing ASRexpansion [13].

Permeability is an important property of concrete. Lowpermeability delays ionic and moisture transfer within con-crete and prevents chemical erosion or attack in chemicallysevere environments. Permeability mainly arise from largecapillary pores rather than gel pores in the cement paste;therefore, the level of concrete permeability mainlydepends on water/cement ratio, curing conditions and per-iod [1,14].

Chloride creates severe corrosion concern for steel avail-able in concrete [1]. In ASTM C1202 and AASHTO T277,the chloride ions penetration is measured by the totalcharges passed exerted by an external electrical field. Thisprocedure is inappropriate and upon criticism because (1)all the ions (instead of chloride ions only) contribute tothe charges measured in this test [15]; and (2) the measuredvalues also depend on the chemistry of pore solution[16,17].

Despite what should be the appropriate method forchloride penetration measurement, the following discus-sion will focus on the fly ash’s effect on RCPT guided byASTM C 1202. One contribution to RCPT by fly ash addi-tion comes from the modification of concrete pore size dis-tribution. This process mainly depends on the particle sizedistribution of fly ash (physical filling effect) and the forma-tion of secondary C–S–H gels by pozzolanic reaction, bothof which have close relationship with fly ash/cement ratio[14,18–25]. The ions leached out from fly ash modify thepore solution, thus contributing to the rapid chloride per-meability [17,26]. Since biomass fly ash has potentiallyhigher alkali content than coal fly ash, the reduction ofRCPT by biomass fly ash addition is potentially less effi-cient than that of coal fly ash.

Although chloride ions penetration can be measured byother alternate methods such as long time chloride pound-ing test [27], results from other researchers still show theconsistency of RCPT by itself (in short time, say one monthand long term,say one year) and with other tests. At20–30% replacement ratio of cement by fly ash, fly ashconcrete is significantly less permeable than cement onlyconcrete even from one to three months curing; and thistrend goes up to one year or longer depending on the avail-ability of test data [14,28–30]. At 90-day after concrete mix-ing, results of RCPT are quite consistent with those fromchloride ponding tests [31].

Further study of pozzolanic reaction kinetics has foundthat at one month curing, coal fly ash (Class C and F) and

biomass fly ash (either cofired or wood fly ash by itself)have undergone significant pozzolanic reactions at ambienttemperatures [32], which can explain why fly ash can reduceconcrete chloride permeability at shorter period within twomonths.

2. Experimental materials and concrete mix design

2.1. Materials

There are seven types of fly ash involved in this investiga-tion, (1) two coal fly ashes, Class C and Class F; (2) twocofired (biomass and coal) fly ashes, SW1 (20% switchgrassburned with 80% Galatia coal, wt%), SW2 (10% switchgrassburned with 90% Galatia coal, wt%); (3) wood fly ash fromwood combustion; (4) two blended fly ashes, Wood C andWood F, which comes from blending wood ash (20 wt%)with either Class C or Class F (80 wt%), respectively.

All other fly ashes except wood have similar particle sizewith major portion in the range of 3–50 lm, and mostwood fly ash falls in the range of 30–130 lm. Galatia coalproduces Class F fly ash by itself; therefore, it is reasonablypostulated that SW1 and SW2 behave more like Class F flyash due to the major portion of coal in the cofiring process.

The detailed information of fly ash, such as particle sizedistribution and chemical composition, and the specifica-tion of cement and aggregates, are given in another paper[33].

The AEA used in this project is AMEX 210, which isfrom Grace Construction Products and popularly used ascommon concrete mixes.

2.2. Concrete mix design and strength build-up

The concrete mix design has the following parameters:(1) fly ash/cement = 25/75 (wt%); (2) water/(fly ash +cement) = 0.5 (wt%); (3) slump 7.5–12.5 cm (3–5 inches)and (4) air content 4–6% (volume). Detailed informationof concrete mix design is given elsewhere [33].

3. Experimental procedures

3.1. Resistance to freezing and thawing

Procedures from ASTM C666 are followed in the F–Ttest of concrete specimens. Weight loss percentage isreported for all the eight mixes while durability factorswere reported only for wood and Class F mix, becausethe freezing and thawing machine performed properly onlyduring the tests of these two mixes. Typically, three speci-mens for one concrete mix are applied for the targetedtests.

3.2. Rapid chloride permeability test

On the 55th day of moisture curing, two 5-cm thickspecimens were prepared from a 10.2-cm (diame-

S. Wang et al. / Fuel 87 (2008) 359–364 361

ter) · 20.3-cm (height) cylinder for each mix. Each speci-men was cut from the middle section of the cylinder usinga water-cooled diamond saw. A belt sander removed therugged edges and notches to produce a flat surface on bothsides of the specimen. Two specimens for each mix weretested according to ASTM C 1202-91 and the values areaveraged by mix.

4. Results and discussion

4.1. Summary of fresh concrete mix and strength build-up

All concrete mixes, neat cement and all fly ash ones,have the air content and slump within the experimentaldesign range. All biomass fly ash mixes (except wood flyash, which has irregular shape) reduce water amount ascoal fly ash because of their spherical shape; all biomassfly ash mixes delay concrete setting, but quite similarly ascoal fly ash mixes perform.

Compression strength from1 day to 1 year and flexurestrength at 56 day after concrete mixes have shown thatall biomass fly ash (except wood fly ash) mixes have similarperformances as coal fly ash (Class C and Class F), whichhave lower strength than neat cement concrete from 1–7days after mixing, but statistically (at 95% confident inter-val) equal strength to neat cement concrete from onemonth to one year after mixing.

All the detailed results of fresh concrete mix andstrength build-up up to one year are available in anotherpaper [33].

4.2. Resistance to freezing and thawing

Fig. 1 illustrates that within the statistical uncertainty ofthe measurements (95% confidence interval) the weight lossof all fly ash concrete mixes except Class C fly ash is less orequal than/to the control cement mix. This result is consis-tent with previously reported results from coal fly ash alone

Fig. 1. Freezing and thawing with 95% confidence intervals for eachconcrete mix.

and it suggests that fly ash, either Class C, Class F or wood,has little impact on freezing and thawing behavior [2–4,34].The weight gain phenomena in this test might be attributedto water gain of samples in the freezing and thawingmachine, which lost too much moisture during the 26-dayexposure to air to be compensated by the ASTM standardof 2 days in fog room prior to testing. Fig. 2 showed dura-bility factors of Class F and wood mix, which correlatedroughly with the weight loss percentages in Fig. 1.

While fly ash addition does not appear to substantiallyimpact the freezing and thawing behavior of concrete witha proper amount of air entrained, fly ash addition stronglyimpacts that amount of AEA required to meet the ASTMspecification. Fig. 3 illustrates the amount of AEA requiredfor each of the concrete mixes and they can approximatelydivided into three groups: (1) SW1 fly ash requires thehighest AEA to create the same air content in concretemix; (2) SW2, Class F and Wood F have the mediumAEA requirement; (3) Wood, Wood C, Class C and purecement have the lowest AEA requirement.

A further investigation of AEA related to unburned car-bon is listed in Fig. 4, which can partially explain theresults illustrated in Fig. 3. If LOI % from cement is omit-ted and the minor difference in air content of all fly ashconcrete mixes is neglected, it implies that AEA amountincreases with the LOI % in the four kinds of fly ash: ClassC, Class F, SW1 and SW2, which are consistent with theresults from other workers [3,4]. Cement and Wood flyash lie far away from this tendency, which might resultfrom their different forms and surface areas of carbon inthe LOI [5,7,35]. The above reasoning also applies to com-parison of blended fly ash (Wood C and Wood F) withwood fly ash, which is not repeated here.

4.3. Rapid chloride permeability

Rapid chloride permeability depends strongly on thewater to cementitious material ratio, curing conditionsand testing date [14,36,37]. Since ratio of water/(cement +fly ash) is fixed, curing conditions are the same by ASTMC192 and testing date is 56th day after mixing, the effectof fly ash on RCPT is dominant and will be discussed here.

Fig. 5 summarizes the results of RCPT. (1) Wood mixeshave the similar charges passed with pure cement mix; (2)Class C has a lower permeability than Wood mixes; (3)Class F, SW1 and SW2 mixes have the lowest permeability;and (4) the blended Wood C have the permeability betweenmixes of their resource fly ash: Wood and Class C, and thatof Wood F is between Wood and Class F.

The permeability difference can be discussed in relatedto fly ash particle size, modification of pore size distribu-tion due to pozzolanic reaction and pore solution chemis-try changes. The following facts are taken intoconsideration for the evaluation of permeability of differentfly ash: (1) Wood fly ash has a much coarser particle sizethan Class C, Class F, SW1 and SW2 indicated in [33];

Relative Dynamic Modulus of Elasticity (%)

88

90

92

94

96

98

100

0 50 100 150 200 250 300

Number of Cycles

Rel

ativ

e D

ynam

ic M

odul

us o

f El

astic

ity (%

) Class FWood

Fig. 2. Durability factors of Class F and wood mix (note: all other mixes are not included here due to the improper performance of test machine).

0.00

0.50

1.00

1.50

2.00

2.50

3.00

3.00 4.00 5.00 6.00 7.00Air (%)

Surf

acta

nt (O

Z/10

0lbs

cem

entit

ious

)

Cement

Class C

Class F

SW1

SW2

Wood

Wood C

Wood F

Fig. 3. AEA requirements to create designed air content of all concrete mixes.

362 S. Wang et al. / Fuel 87 (2008) 359–364

(2) Class C, Class F and wood fly ash have quite similarpozzolanic reactivity indicated by the hydrated lime con-sumption at two month [32] and (3) In related to available

0

0.5

1

1.5

2

2.5

0 2 4 6 8LOI %

Air

Entr

ainm

ent A

gent

(OZ/

100

lbs

(cem

ent +

fly

ash)

Cement

Class C

Class F

SW1

SW2

Wood

Fig. 4. Relationship between air entrainment requirements and LOI % offly ash.

equivalent alkali percentage by ASTM C 33, wood fly ash1.78, Class C 1.03 and Class F 0.53 [13].

If the similar pozzolanic reactivity can bring similar poresize modification, Class C and wood fly ash will have betterpore size modification than Class F because of their higherpozzolanic reactivity [32].

Since pore solution mainly contains OH�, Na+ and K+,equivalent alkali from fly ash will play the main role ofpore solution modification. Because wood fly ash has a lar-ger particle size and higher available alkali, wood mixeshave the highest permeability among fly ash concretemixes. Class C has a higher equivalent alkali than ClassF and that is why its concrete has a higher permeability,which is consistent with other workers [14,17]. SW1 andSW2 fly ash do not have the data of equivalent alkali.Based on the small portion of biomass involved, they areassumed to have the properties of Class F fly ash due totheir coal property. In this sense, they will have a lower per-meability than Class C. The blended Wood C and Wood F,due to the influences of both particle size and availablealkali, will have the permeability between their originalfly ash sources.

Fig. 5. Rapid chloride permeability tests (averaged two samples) on 56th day.

S. Wang et al. / Fuel 87 (2008) 359–364 363

5. Conclusions

1. Special attention is taken to get appropriate amount aircontent in coal and biomass fly ash concrete mix.

2. With the similar amount of air content, the freezing andthawing tests show that all fly ashes mixes, includingcoal fly ash and all kinds of biomass fly ash, performwell and similarly to that of the pure cement concrete.

3. Biomass fly ash SW1 and SW2 have the similar perfor-mances of coal fly ash in reducing rapid chloride perme-ability of concrete.

4. The higher rapid chloride permeability of Wood and theblended Wood fly ash can be attributed to the muchcoarser particle size of wood fly ash.

References

[1] Mindness S, Young JF, Darwin D. Concrete. 2nd ed. PearsonEducation, Inc.; 2002.

[2] Woods H. Durability of concrete construction. Detroit, Mich:American Concrete Institution; 1968.

[3] Helmuth R. Fly ash in cement and concrete. 1987.[4] Wesche K. Fly ash in concrete: properties and performance. New

York: Chapman & Hall; 1991.[5] Hill R. An examination of fly ash carbon and its interaction with air

entraining agent. Cement Concrete Res 1997;27(2):193–204.[6] Kulaots I, Hurt RH, Suuberg E. Size distribution of unburned carbon

in coal fly ash and its implications. Fuel 2003;83:223–30.[7] Freeman E et al. Interactions of carbon-containing fly ash with

commercial-entraining admixtures for concrete. Fuel 1997;76(8):761–765.

[8] Pistille M. Air-void parameters developed by air-entraining admix-tures, as influenced by soluble alkalies from fly ash and Portlandcement. ACI Mater J 1983:217–22.

[9] Thy P et al. High temperature elemental losses and mineralogicalchanges in common biomass ashes. Fuel 2006;85:783–95.

[10] Turn SQ et al. Fuel characteristics of processed, high-fiber sugarcane.Fuel Process Technol 2003;81:35–55.

[11] Ollila HJ et al. A SEM–EDS and XRD study of the factors affectingthe melting behavior of biomass ash. 6th science in thermal andchemical biomass conversion. Victoria, BC, Canada: CPL Press;2004.

[12] Pietersen HS. Reactivity of fly ash and slag in cement. Geochemis-try. Delft: Delft University of Technology; 1993.

[13] Wang S et al. Mitigation of ASR expansion by biomass/coal fly ash.9th Canmet/ACI international conference on fly ash, silica fume, slagand natural pozzolans. Poland: Warsaw; 2007.

[14] Alhozaimy A, Soroushian P, Mirza F. Effects of curing conditionsand age on chloride permeability of fly ash mortar. ACI Mater J1996;93(1):87–95.

[15] Andrade C. Calculation of chloride diffusion coefficients in concretefrom ionic migration measurements. Cement Concrete Res1993;23:724–42.

[16] Shi C, Stegemann J, Caldwell R. Effect of supplementary cementingmaterials on the specific conductivity of pore solution and itsimplications on the rapid chloride permeability test (AASHTOT277 and ASTM C1202) results. ACI Mater J 1998;95(4):389–94.

[17] Shi C. Effect of mixing proportions of concrete on its electricalconductivity and the rapid chloride permeability test (ASTM C1202or ASSHTO T277) results. Cement Concrete Res 2004;34:537–45.

[18] Douglas E, Bilodeau A, Malhotra VM. Properties and durability ofalkali-activated slag concrete. ACI Mater J 1992;89(5):509–16.

[19] Leng F, Feng N, Lu X. An experimental study on the properties ofresistance to diffusion of chloride ions of study and blast furnace slagconcrete. Cement Concrete Res 2000;30:989–92.

[20] Naik T et al. Long-term performance of high-volume fly ash concretepavements. Cement Concrete Res 2003;100(2):150–5.

[21] Obla KH et al. Properties of concrete containing ultra-fine fly ash.Cement Concrete Res 2003;100(5):426–33.

[22] Sengul O, Tasdemir C, Tasdemir M. Mechanical properties and rapidchloride permeability of concrete with ground fly ash. ACI Mater J2005;102(6):414–21.

[23] Li S, Roy DM. Investigation of relation between porosity, porestructure and chloride diffusion of fly ash and blended cements.Cement Concrete Res 1986;16:749–59.

[24] Shi C. Strength, pore structure and permeability of alkali-activatedslag mortars. Cement Concrete Res 1996;26(12):1789–99.

[25] Patel R et al. Development of statistical models for mixture design ofhigh-volume fly ash self-consolidating concrete. ACI Mater J2004;101(4):294–302.

[26] Wee TH, Suryavanshi AK, Tin SS. Evaluation of rapid chloridepermeability test results for concrete containing mineral admixtures.ACI Mater J 2000;97(2):221–32.

[27] Tang L, Nilsson L-O. Rapid determination of the chloride diffusivityin concrete by applying an electrical field. ACI Mater J1992;89(1):49–53.

[28] Poon C-S et al. Comparison of the strength and durability perfor-mance of normal- and high-strength pozzolanic concretes at elevatedtemperatures. Cement Concrete Res 2001;31:1291–300.

[29] Sivasundaram V, Carette GG, Malhotra VM. Mechanical properties,creep, and resistance to diffusion of chloride ions of concretes

364 S. Wang et al. / Fuel 87 (2008) 359–364

incorporating high volumes of ASTM Class F fly ash from severaldifferent sources. ACI Mater J 1991;88(4):407–16.

[30] Bouzoubaa N, Zhang MH, Malhotra VM. Mechanical propertiesand durability of concrete made with high-volume fly ash blendedcements using a coarse fly ash. Cement Concrete Res 2001;31:1393–1402.

[31] Sujjavanich S, Sida V, Suwanvitaya P. Chloride permeability andcorrosion risk of high volume fly ash concrete with mid-range watereeducer. ACI Mater J 2005;102(3):177–82.

[32] Wang S et al. Quantitative kinetics of pozzolanic reaction andstrength build-up in the mortar of biomass/coal fly ash with calciumhydroxide. 9th Canmet/ACI international conference on fly ash, silicafume, slag and natural pozzolans. Poland: Warsaw; 2007.

[33] Wang S. et al. Biomass fly ash in concrete: mixture proportioning andmechanical properties. Fuel, doi:10.1016/j.fuel.2007.05.026.

[34] Fouad FH, Copham CA, Donovan JM. Evaluation of concretecontaining fly ash with high carbon content and/or small amounts ofwood. Department of civil engineering, The University of Alabama atBirmingham. 1998.

[35] Kulaots I, Hurt RH, Suuberg E. Size distribution of unburned carbonin coal fly ash and its implications. Fuel 2004;83:223–30.

[36] Scanlon JM, Sherman MR. Fly ash concrete: an evaluation ofchloride penetration testing methods. Concrete Int 1996(7):57–62.

[37] Zhang M-H et al. Concrete incorporating supplementary cementingmaterials: effect on compressive strength and resistance to chloride-ion penetration. ACI Mater J 1999;96(2).