Design, Construction and Monitoring of an Innovative RC WaterTank Chlorination of Water Treatment Plant Totally Reinforced

with GFRP Bars

Hamdy M. Mohamed’ and Brahim Benmokrane2‘Postdoctoral Fellow, University of Sherbrooke/Pultrall Inc., Canada,

Hamdy.Mohamed@usherbrooke.ca2 Research Chair Professor, University of Sherbrooke, Canada, Brahim.Benmokrane@USherbrooke.ca

Abstract

Reinforced concrete (RC) tanks have been used for water and wastewater treatment plants for

decades. Design of these tanks requires attention flot only to strength requirements, but also to

crack control and durability. Usually, the liquid tanks are designed as crack free structures to

eliminate any leakage. This paper presents the design procedures, construction details, testing,

and monitoring resuits of the first worldwide RC water tank chlorination, totally reinforced

with glass fiber-reinforced polymers (GFRP5) bars. The project is located in Thetford Mines

City, Quebec, Canada. The tank is considered as one ofthe most important components in the

new water treatment plant in the city. The volume capacity ofthe tank is over 2500 m3, and it

has 4650 mm wall height. The foundation, vertical walls and the top slab were totally

reinforced with GFRP bars. The tank was designed for satisfying the serviceability and

strength criteria according to CAN/CSA S806-12, ACI 440.1R-06 and ACI 350/350R-06. The

tank is well instrumented at critical locations for strain data collection with fiber-optic

sensors. Also, the leakage test resuits prior to backfïlling under actual service conditions are

presented for the leaking cracks and strain behavior in the FRP bars at different location in the

tank. Site inspection showed that the FRP-RC water tank performed very well and it was able

to withstand applied loads without up normal cracks or leaking at the leakage test.

Keywords: FRP bars; Field applications; Implementation; Water tank; Design.

1. INTRODUCTIONConventionally reinforced concrete (RC) tanks have been used extensively in municipal and industrialfacilities for water and wastewater treatment plants (WWTPs) for decades. There are three kinds ofwater tanks; tanks resting on ground, underground tanks and elevated tanks. The most common typeused in WWTP is the underground tank. The walls of these tanks are subjected to water and earthlateral pressure and the base is subjected to weight of water and uplift soi! pressure. Usually, thesetanks should be covered on top to protect the water inside. Design of these tanks requires attention notonly to strength requirements, but also to crack control and durability to prevent the water leakage andcorrosion of steel reinforcements. Therefore, a conservative design for these tanks must be able towithstand applied loads without cracks. To achieve this design, more reinforcement ratio withadequate bar spacing, larger wall thickness and the use of quality concrete by proper constructionpractices are needed.

1.1 Background and Statement ofthe ProblemElectrochemical corrosion of steel is a major cause of the deterioration of the civil engineeringinfrastructure. RC tank is one of the most important structural facilities in the WWTPs, which it isusually subjected to a uniquely difficult environment where corrosion poses exceptional challenges.These concrete tanks deteriorate faster than any other structures because of direct and permanentexposure to chemical aggressive environment. However, the need for their protection is oftenidentified only after significant deterioration has occurred. For years, containment designers have triedto achieve crack-free concrete to eliminate the corrosion problem. Techniques have included specificspecial mix designs, low water/cement ratios, many different admixtures, special aggregates, andsupplementary cementitious materials, ail with limited success. Severely corrosive environments inwater and wastewater treatment facilities include those naturaily present in incoming flows and thoseresulting from the application of specific treatment methods or chemicals. The mechanisms ofcorrosion in WWTP are influenced by a variety of environmental factors. Foremost, each metal in adrinking water system is affected by contact with chlorinated water in its own characteristic manner.Secondly, each constituent that comprises the water quality characteristics such as carbonate, pH,dissolved oxygen, sulfate, and chloride, exerts its own influence on corrosion. The use of halogenssuch as chlorine in the disinfection of drinking water or treating the wastewater, as well as the ozonation, ail have a devastating effect on reinforcing steel products, no matter if they are black,galvanized, epoxy covered or even stainiess steel elements. However, the chlorine is the primaryoxidant, other than oxygen (aeration) in chemicai treatment method which it is considered being acorrosive agent in water. Evidentiy, this disinfection is an absolute necessity to treat and achieve aproper quality of the water before it is released to the surrounding homes and users. The subject ofchloride induced corrosion of steel reinforcements is very complex and depends on many things suchas concentration, temperature and pH. Only, careful selection of reinforcing materials can significantlyprevent the detrimental effects of corrosion against the aggressiveness of these products. So, thechallenge for the structural engineer and municipalities is to design these concrete tanks usingnoncorrosive materials such as fiber-reinforced polymers (FRP) reinforcing bars.

1.2 FRP Composite Reinforcing Bars: Advantages and ApplicationsStructural engineers and researchers are seeking for innovative solutions that provide longer life andrequire less maintenance than reinforcing conventional materials for RC tanks. One such innovation,which is becoming an attractive alternative for RC tank structures, is using the FRP reinforcing bars.In the last decade, there has been a widespread application in using noncorrosive FRP reinforcingcomposite bars for concrete structures due to enhanced properties and cost-effectiveness. Known to becorrosion resistant, FRP bars provide a great alternative to steel reinforcement. FRP materials ingeneral offer many advantages over the conventional steel, including one quarter to one fifth thedensity of steel, no corrosion even in harsh chemical environments, neutrality to electrical andmagnetic disturbances, and greater tensile strength than steel (Benmokrane et al. 2006; 2007; ElSalakawy et al. 2003a).

Since glass FRP (GFRP) bar is more economical than other available types (carbon and aramid) ofFRP bars, it is more attractive for infrastructure applications and to the construction industry. The

2

GFRP bars have been used extensively in different infrastructure applications such as bridges, parkinggarages, tunnels and marine structures in which the corrosion of steel reinforcement has typically ledto significant deterioration and rehabilitation needs (Mohamed and Benmokrane 201 2a; El Salakawyet al. 2003b). Many significant developments from the manufacturer, various researchers and DesignCodes along with numerous successful installations have led to a much higher comfort level andexponential use with designers and owners. After years of investigation and implementations, publicagencies and regulatory authorities in North America has now included FRP as a premium corrosionresistant reinforcing material in its corrosion protection policy. However, up to date, there are no fieldapplications and implementations have been reported in the literature to utilize the FRP bars in the RCtanks to resolve the expansive corrosion issues.

Through the NSERC Industrial Research Chair in Fiber-Reinforced Polymer Composite.Reinforcement for Concrete Infrastructures at University of Sherbrooke, a joint effort andcollaboration with the government organizations and private industry was established to develop andimplement FRP bars in diffident applications (Mohamed and Benmokrane 201 2b). This effort hadbeen employed through more than twenty years on developing and improving glass/carbon compositebars (Benmokrane et al. 2002a,b). After achieving satisfactory laboratory results on different concretestructural elements such as full-scale beams, slabs, columns and shear wall reinforced with these bars,the FRP bars have been used extensively through the last 12 years in different field applications suchas bridges, parking garages, marine structures; seawall and tunneling. Using and implementation ofFRP reinforcements in infrastructures have been approved that the cutting-edge technology hasemerged as one of the most cost-effective alternative solutions compared to the traditional solutions.On the other hand, these field applications and its monitoring results have been the basis to validateand improve existing design codes and guidelines, establish construction details, and evaluate theperformance of FRP bars under actual service loading and environmental conditions. Despite thebeneficial characteristics and widespread application of FRP bars in different infrastructure, these barshave flot been yet used in the corrosive wastewater and water treatment plant structures.

This paper presents the first Canadian and worldwide field application of FRP bars in concrete tank ofwater treatment plant. Design and construction details of this tank presented in this study are used toillustrate code requirements, tank analysis, design details, and construction of FRP-RC tanks. The tankwas instrumented at critical locations to monitor the cracks and strain behavior at leakage test prior tobackfilling (filling the tank with water). Also, it is intended to evaluate the in-service performance ofthe FRP-RC tank to several years after running works of the plant. The following sections providedescription ofthe tank, FRP materials, design and codes, construction details and monitoring results.

2. TANK DESCRIPTIONCities maintain safe municipal water supplies based on standards established by province, state andlocal governments. The chlorine disinfection process is used worldwide in the water treatment plantsto produce large amounts of safe drinking water as quickly as possible. This process is usuallyperformed in a big tank which is called Water Tank Chlorination. The use of chlorine in thedisinfection of drinking water has devastating effect on steel reinforcements, no matter if they areblack, galvanized, epoxy covered or even stainless steel elements. Therefore, the expansive corrosionof steel reinforcing bars stands out as a significant factor limiting the life expectancy of the water tankchlorination. The owner and location ofthe new investigated water treatment plant is Thetford Minescity, Quebec, Canada. Figure 1 presents the general overview of the plant. The Thetford Mines citydecided to use noncorrosive FRP reinforcing bars in the water tank chlorination part to extend theservice life, reduce maintenance costs and improve life-cycle cost efficiency of the new plant. One theother hand, this implementation presènts the first worldwide innovative field application of FRPcomposite reinforcing bars in concrete tank structures.

3

The structural system of the tank is rectangular under-ground rested on rock completely buried withcompacted fil-sou around the walls. The surface area of the top slab is buried 600 mm of uncompactfil-sou. The vertical walls support the top cover slab ofthe tank and it is rested on RC-rafi foundation.The volume capacity of the tank is approximately 2500 cubic meter, the wall height is 4650 mm, andthe tank has the dimensions 24.0 m wide and 23.0 m iength. The tank is designed to include twoclosed celis (C1 and C2) using continuous vertical wall (WM) at the middle of the tank, (see Figure 1).Each ceil is divided for two zones using un-continuous interior vertical wall (W1). The clear spacingbetween these walls is 5475 mm. The thickness of the exterior and middle walls (WE and WM), topfloor slab and foundation is constant and equal to 350 mm, while the thickness of the interior walls(W1) is 300 mm.

3. GFRP BARS USED IN THE TANK

Sand-coated GFRP bars were used to reinforce the three structural elements of the tank; foundation,walls and cover slab. The bars were made of continuous E-glass impregnated in a vinylester resinusing the pultrusion process, manufactured by a Canadian company [Pultrali Inc.]. The glass fibersgive the bar mechanical strength, while the resin matrix (resin, additives, and fihlers) providescorrosion resistance in harsh environments. The GFRP bars had a sand-coated surface to enhance bondperformance between bars and surrounding concrete, (see Figure 2). Two grades of these bars wereused, Grade III and Grade II as classified in the CAN/CSA S807-l0 according to the tensile young’smodulus (60 and 50 GP, respectively). GFRP bars Grade III and II were used as main and secondaryreinforcements, respectively, in the wails, cover slab and foundation of the tank. Also, two bardiameters were used in the design of the tank, No. 15 and No. 19 (Nominal cross-sectional area 199and 284 mm2, as indicated in CAN/CSA S807-10). The mechanical properties of the GFRP bars aresummarized in Table 1 as provided by the manufacture. Considerable resçarch efforts have beenconducted in the past decade to assess the suitability of FRP reinforcements in reinforced-concretestructures (Robert et al. 2009; 2010). The work of these researchers has highlighted the short and longterm performance of FRP-reinforced concrete structures or the durability of FRP reinforcing barssubjected to different conditions including immersion in aikaline solution, sustained tensile stress,elevated temperature, and freeze—thaw cycles. The resuits of these studies have reveaied that the effectof aforementioned conditions had no significant effect on ‘the tensile strength of the GFRP bars forservice life beyond 100 years.

FRP-RC Tank

1El

f

Figure 1: Over view of the water treatment plant

4

Table 1. Properties GFRP Reinforcing Bars used in the tank

Bar Modulus ofdiameter Grade

Tensile Strength Ultimate Tensile(MPa)

ElasticityStrain (%)

(mm) (GPa)

15 II 934 55.4±2.5 1.6915 III 1105 64.7 ±2.5 1.7120 III 1059 62.6±2.5 1.69

Note: ±XX standard deviation.

6. DESIGN 0F THE TANK

6.1 Codes and Design EquationsThe design was made according to the CAN/CSA S806- 12 (2012) for design and construction ofbuilding components with fibre reinforced polymers, CAN/CSA-A23.3-04 . (R20 10) - Design ofConcrete Structures, ACI 440.1 R-06 Guide for the design and construction of structural concretereinforced with FRP bars, and ACI 350/350R-01/06 (2001/2006) Code requirements forenvironmental engineering concrete structures and commentary. The loads were calculated accordingto the National Building Code of Canada (NBCC 2005). The tank was designed to determine all thepossible loading condition as result from water pressure and sou load on the walls and foundation, alsodead and live loads on the top slab were considered. According to ACI 350, the full effects ofthe soi!loads and water pressure were considered for without the benefit of resistance of the loads which couldminimize the effects of each other. The design was made using normal-weight concrete having a target28-day compressive strength of 35 MPa. The following sections present the summary of the codeprovisions that has been considered in the design.

Flexural strength:• The walls, foundation and top slab were designed as over reinforced sections as specified by the

CAN/CSA S806-02 (2012) considering the following equation:

o 7—>d 7+2000

• The ultimate strain at the extreme concrete compression fibre was assumed to be 0.0035.• FRP reinforcement has been used in ail members subjected to combined flexure load and axial

compression and tension force. However, the FRP reinforcement in the compression zone ofthe wallwas deemed to have zero compressive strength and stiffness.

Shear strength:The shear strength was verified according to the type of axial load on the member using the followingequations CAN/CSA S806-02 (2012):

Figure 2: Sand-coated glass FRP-reinforcing straight and bent bars as shipped to the site

5

Members subjected to axial tension

=[0.052(Pckmkr(fbwdv]kska [OE3Nf

Members subjected to axial compression

=[0.052+ckmkr(f )bwdv]kska []

o.o,[i]

3.0

Where;

[0.05ickmkr(fbwdv] 0.1 2Jbd

k= 450+d

1.0, ka =--1.0, km kr =1+(EFpFW)

Vfd

Serviceability and Crack Width Limitation:Extensive laboratory work involving new developed FRP bars has confirmed that crack width atservice loads is proportional to bar stress, bond coefficient (kb) and the tensile modules of the bar.According to CAN/CSA S806-02 (2012), when the maximum strain in FRP tension reinforcementunder full service loads exceeds 0.00 15 the z value (see Equation 4) should be address to limit thecrack width and control the stress in the FRP bars. The cross-sections of maximum positive andnegative moment shail be 50 proportioned that the quantity, z, does flot exceed 45 000 N/mm forinterior exposure and 38 000 N/mm for exterior exposure. The numerical limitations for these values,respectively, correspond to limiting crack widths of 0.7 and 0.5 mm.

z= kb--fJÂ,

Where

= M +__T (M & +T), for members subjected to axial tensionnoAfd(1-k/3) floAf

M T .ff =

______________

- (M & -T), for members subjected to axial compressionnoAfd(1-k/3) flOAf

k=g2p, flf +(p nf) Pf flf, flfb,,d’

E0 = 4500,ft

For the steel liquid structures, the corresponding z values according to the ACI 350-01 are 20000 and16600 N/mm for normal environmental exposure and severe environmental exposure, respectively.The numerical limitations for these values, respectively, correspond to limiting crack widths of 0.27and 0.23 mm. These z values were established for cover equal to or less than 50 mm. Today, the ACI350-06 replaced the z factor requirements of the 2001 edition. The maximum allowable stresses arenow specified directly as a function of bar spacing as follows.

normal environmental exposure

6

20000fmax=320

36000 psi (138fsm 248MPa)pJs2+4(2+dbI2)2

severe environmental exposure260

17000fmax= 36000psI(117fmax248MPa)/3Js2+4(2+dbl2)2

Crack width is inherently subject to wide scatter even in careful laboratory work and is influenced byshrinkage and other time-dependent effects. The best crack control is obtained when the reinforcementis well distributed over the zone of maximum concrete tension. The current provisions for spacing areintended to limit surface cracks to a width that is generally acceptable in practice, but may vary widelyin a given structure, (ACI 3 50-06). The maximum allowable stress (248 MPa) in the steel bar isapproximately equal to 60% of the yield strength, which is corresponding to 1240 micro strain. Theequivalent stress in the FRP bar to this strain presents 10 to 15% of the ultimate tensile strength,according to the Grade of the bar. The first attempt in the design of the present tank was to determinethe reasonable concrete thickness to control the tensile stress in the tension side and hence eliminatethe crack. This was achieved by controlling the tensile stresses in concrete within permissible limits.Design of the tank using the CAN/CSA S806-02 for crack and stress limitation was lead to strain andstress in the FRP bars close to the aforementioned stress and strain limitation using (ACI 350-01/06).Also, using FRP bar Grade III in the design, less concrete cover with small size diameter wasoptimized the design to reinforcement ratio close and less than that used in the steel-RC tank of theplant. The outcome of this optimization with the advantage of the FRP bars as noncorrosive materialwas convinced the city to implement FRP bars in the water tank chlorination ofthe new plant.

Shrinkage ReinforcementShrinkage and temperature reinforcement is intended to limit crack width. The stiffness and strength ofreinforcing bars control this behavior. Shrinkage cracks perpendicular to the member span arerestricted by flexural reinforcement; thus, shrinkage and temperature reinforcement are only requiredin the direction perpendicular to the span (ACI 440. 1R-06). The FRP shrinkage and temperaturereinforcement perpendicular to the member span was determined for the walls, top slab and foundationby using the ACI 440.1R-06:

p =0.0018414 0.0014

ACI 440. 1R-06 limits the spacing of shrinkage and temperature FRP reinforcement to be flotexceeding three times the slab thickness 300 mm, whichever is less.

6.2 StructuraI Analysis and FRP-Reinforcements DetailsThe reinforced concrete tank is a combination of walls, foundation and top slab. The vertical walls andfoundation rigidly connected together to form a monolithic frame. The top slab was designed based onone-way loading action in the short direction, as statically indeterminate continuous four equal-spanson hinged support. Also, the loading action on the walls combined with the foundation was consideredone-way in the short direction. As, the water pressure is resisted by vertical bending moments in thewalls. Figure 3 shows the structural model of the tank for the main vertical cross section, with theloads distribution on each member for 1.0 m strip width. Each individual member must be capable ofresisting the forces acting on it, so that the determination of these forces is an essential part of thedesign process. The bending moments, shearing forces and axial forces in each member for the tankwere determined using elastic analysis computer program.

7

A haif vertical section of the outer and interior wails, foundation and top slab together with the axissymmetric reinforcement details, is shown in Figure 4. The tank was designed based on serviceabilityand ultimate limit state for FRP reinforcement requirements, while the thickness was determined usingworking stress design as recommended by ACI 350. The moment in the wall varies considerabiy fordifferent locations in the wall. The reinforcing could differ at several locations for a highiy efficientdesign. The thickness of the wall could also vary, either tapering the wall or stepping the wall.However, as common and for the sake of time, the thickness and reinforcing were kept consistent forthe entire wall. One design for the vertical maximum moments was considered to save the time of thedetailers and construction crew. Usually, crack width must be minimized in tank walls to preventleakage and corrosion of steel reinforcement, but herein there is no issue for the corrosion with FRPreinforcements. The foundation, waii and siab thicknesses were determined based on the usualprincipies ignoring tensile resistance of concrete in bending. Additionally the thickness of thesemembers was checked to be ensured that tensile stress on the water retaining face of the equivaientconcrete section did not exceed the permissibie tensile strength of concrete. The 350 mm concretethickness was chosen for exterior walls, foundation and top slab based on the permissibie allowabieworking stress for concrete. Since, the type of reinforcement, steel or FRP bars does flot govern thefirst estimation of the concrete thickness. It depends on the concrete strength and the service forcesacting on the cross section. Here, it is of interest to mention that 350 mm concrete thickness was aisoused for the tank reinforced with steel bars that has the same straining action. The benefit of usingnoncorrosive FRP reinforcing bars in the design aliowed decreasing the clear concrete cover to 50mm, as compared with that 60 mm used in the tank reinforced with steei bars.

The size of reinforcing FRP bars were chosen recognizing that cracking could be better controlled byusing a larger number of small diameter bars rather than fewer larger diameter bars, (ACI 350).Severai bars at moderate spacing are much more effective in controlling cracking than one or twolarger bars of equivalent area. For these reasons, bar No. 15 was used extensively in the design of thetank with spacing ranged from 90 mm to 180 mm. However, bar No. 19 was used only at one sectionin the foundation, as the maximum observed moment required a high reinforcement ratio. Usuaiiy, indesign of water RC tank is reconimended to use higher ailowabie reinforcing bar stress, so that less baris used, resulting in iess restraint shrinkage and smaller tensile stresses in the concrete. Hence, the newdeveioped GFRP bar Grade III of the higher tensile strength and modulus was used as mainreinforcement in the short direction. One the other hand, GFRP bar Grade II was preferred from a costeffectiveness point of view to be used for ail the secondary reinforcements in the long direction of thetank. GFRP bar No. 15 Grade II was used in the tank in the long direction in the exterior and middlewails, foundation and top slab with spacing equal to 250 mm. Aiso, this bar has been used in theinterior walls (W1) in the sort and long directions both sides, as the water pressure is actingsimultaneously on the two faces resulting zero moment.

Figure 3: Structurai system and loads on the main vertical section in the tank.

8

Figure 4: FRP-reinforcement details for the main vertical section in the tank.

7. TANK CONSTRUCTIONThe construction of the tank started and completed in February and July 2012, respectively. The typeof the sou underneath the foundation is rock. In February, site excavation was started in which sou,rock, and other materials were removed, typically with the use of heavy earthmoving equipment suchas excavators and bulldozers. The depth of excavation was around 5.5 m, to create a level, clean areato work, with the foundation being established in the excavated area. The FRP bars were delivered tothe site in mid-March. The placement of GFRP reinforcements for the bottom and top foundation mat,concrete casting and curing was started and finished by the end of March. Continuous plastic chairswere placed in the longitudinal direction at 0.7 m intervals under the bottom reinforcement mat tosupport the FRP bars and maintain the required clear concrete cover. In the case of the top mat, singlechairs at 0.9 m intervals in both directions were used. Figure 5 shows the FRP-raft foundation beforeandafter

Figure 5: FRP-RC raft foundation before and after casting.Through April and May, the construction was stopped in the FRP-RC tank and shifted to complete andcast the top slab of the steel-RC tank in the site. Thereafter, the construction of the walls started inMay 27, 2012, by the installation of the interior and exterior mat of the FRP reinforcing bars (verticaland horizontal bars). After, the work in the formwork, casting and curing of the walls was started andcompleted on June 5, 2012. Figures 6 and 7 show the formwork, FRP-vertical and horizontalreinforcements of the vertical walls, before and after casting and during the different stages ofconstruction. The day after casting the walls, ail the formworks of the interior walis both side wereremoved, while the exterior formwork of the outer walls was maintained and used through all theconstruction stages of the top slab. The work in the formwork of the top siab started directly after

No. 15 @ 250

r (III)

Top &/ftom (H)No. 15 @ 250

1590 No.15@180No. 15 180(III) (III)

Top & Bm (II)

/4

z

8400,,

-

JNo.iS@l4O 15@140

No.15250 No.15120— 120 Vert (II) Vert (III)Vert (1H)

No. 15 @ 300Horiz (II)

No. 15 @ 250— No. 15 @ 250

Horiz (II) Ç,/ Horiz (11)

Nj250No. 15 @ 120Vert (III) No. 15 250 Vert (1H)

Vert (II)

No. 15 250No. 15 @ 250Top & 00000m (H)Top & 00100m (11)

No. 19 130 No. 19 ( 130

. 4:::.-:No.15)4130 No.19@1130+No.15@130 [ No. 15@130

(111) (HI) (1H)300. 350 8475 .300 5475 . 350

I

9

rernoving wall formwork and was finished at rnid-June. The placement of GFRP reinforcements forthe bottom and top mat of the top slab and concrete casting was started and finished by June 22, 2012.Also, continuous plastic chairs were placed in the longitudinal direction at 0.8 m intervals under thebottom at top reinforcement mat to support the FRP bars and maintain the required clear concretecover. After casting, the slab was cured for 10 days, and after that by 4 days the forrnwork wascompletely removed. Figure 7 shows the FRP-RC top slab. Following that, cleaning works andleveling the top surface of the foundatibn using cernent mortar inside the tank were completed at thernid-July to start fihling the tank with water.

r i, Construction détails of y. bar instalh

Figure 6: Overview of vertical walls before and after casting.

10

In this tank, water stops were used at the wail-foundation connection for the interior and exteriorwalls. Also, vertical expansion joints were introduced between the walls ofthe steel and FRP tanks. Inconclusion, under these hard work conditions ail the construction stages of the tank starting fromfoundation then walis and finishing the work at the top slab, no additional precautions were consideredin concrete casting or handiing and placement of GFRP bars as compared to steel. The constructionand installation practices required when using FRP reinforcing bars were similar to those used withsteel bars in FRP and steei tank, respectively. The construction crews reacted positively, indicatingthat more FRP bars could be handled and placed in formwork in less time consuming due to their lightweight. The FRP bars did flot move or bat during concrete placement and vibration and withstood ailon-site handiing and placement with no problems.

8. INSTRUMENTATION 0F TANKThe FRP-RC tank was instrumented at critical locations to measure internai strain data using Fiberoptic sensors (FOSs). The objective of using FOS was to allow for the iong-term monitoring of thetank. The wall and top siab were instrumented with 6 and 10 FOS, respectively, at different locations.One of the exterior walls of the tank was chosen to be instrumented to coiiect the strain data atmaximum moment location for the two loading conditions, water and earth pressure. Three FOS weregiued on three vertical GFRP reinforcing bars for each side of the wali, interior and exterior mat. TheGFRP bars were instrumented at the structural laboratory of the University of Sherbrooke. Thereafter,the bars were shipped to the construction site where they were instaiied in the designated locationduring the construction stage of the wall. The interior and middie walls have flot been instrumented asthe water pressure is acting on both sides. It is of interest to mention that the FOS is capable tomeasure strain data in the range of positive and negative 2500 micro-strain. The benefit from this is tocollect the tension and compression strain in the FRP bars, as in the case of exterior wall, the momentis reversible due to the opposite effect of earth and water pressure. Figure 8(a) shows the GFRP barsinstrumented with FOS.

e 7: Overview of top slab before and during casting

- 8: Preparng and instrumentation of

11

The ten FOS in the top siab were distributed along the mid-span section between the wails for thepositive moment (top and bottom mat), also, over the waii for the negative moment (top and bottommat), as shown in Figure 8(b). These sensors were glued on the transverse GFRP reinforcing bars atlocations of expected maximum stresses. These bars were prepared with FOS in the site, because thebar length ranged from 9.0 to 16.0 m, and it was difficult to protect and ship these bars with FOS fromthe university to the site. Ail the FOS wires ofthe wall and slab were coliected and passed at one pointin the top surface of the top siab. After finishing the construction work over the tank related to the fil!soil, the wires were collected and protected inside PVC box over the top slab of the tank to haveaccess for strain reading. Eight channel data-acquisition systems were used to collect FOS readings atthe different stages of construction.

9. WATER LEAKAGE TEST 0F THE TANKThe leakage test was performed directly after three days of removing the formwork of the top slab, andbefore any work in the sou backfiui. The steei and FRP RC-tank were tested by the contractor andwitnessed by the consult Engineer. As mentioned before, the tank includes two ceils, so each ce!!should be considered a single tank and tested individually. One of these celis was filled up completeiywith water to check the leakage, and the other ce!! was maintained empty. After finishing the test atone side of the tank and checking ail the visible cracks, the water transferred to the other ce!!. Figure10 shows the overview of the tank through leakage test in one ce!!. The water has been kept at the testlevel for three days prior to the actuai test. The exterior wall surfaces ofthe tank were inspected duringthe period of filling the tank.

9.1 Flexural CracksVisuai inspection of the tank thorough three days indicated that the leakage test did not induce flexuralhorizontal cracks in ail the wa!!s. No water leakage was observed indicating any presence of flexuraicrack ieaking. This can be attributed to the compression zone deveioped in one side of the wai!section as a resuit of flexurai stresses could effectiveiy prevent leakage through the crack regard!ess ofthe crack width. Here, it is of interest to mention that the cracking moment resistance for the 350 mmthickness wa!i equals 75 kN.mlm. This moment is almost over the service moment in the wailresulting from the water pressure at the leakage test. The wa!!s were designed to minimize the crackwidth resulting from the one-way !oad action of the water pressure in the vertical direction. Thismeans that flexural cracks are not of concern with regard to leakage, because the liquid passagethrough the depth of the section is obstructed by the presence of the uncracked concrete in thecompression zone. However, the compression zone depth should be controiled for limiting the !iquidloss through permeability of the concrete. This result is consistence and in a good agreement with theresearch work and experimental test results conducted by (Ziari and Kianoush 2009).

Figire9: Overview ofthe completed tank through the leakage test and prior to backfil!ing

12

9.2 Shrinkage and Restraint CracksDuring the leakage test, a site inspection showed that the concrete water. tank deveioped limitedvertical shrinkage cracks which became leaks and did flot seif-seai. These cracks are common andexpected for the liquid tank at the first stage of service loading. These are minor indications and haveno real structural impact on the tank. To control leakage in water tank the occurrence of cracks cannotbe prevented, but must be minimized, and widths of cracks should be kept below a certain limit underservice loads, (Ziari and Kianoush 2009). The leakage test resuit of the FRP-tank indicated that thenumber of observed cracks in the exterior surfaces for wall WEI, WE2 and WE3 were 6, 5 and 7,respectively. The cracks were perpendicular to the direction of the maximum principle stress inducedby moment. The cracks extended from the base and propagated up to the full height of the wall. At thecorner ofthe tank, one inclined crack in each wall was observed propagating from the base toward thecorner edge with angle approximateiy equal to 45 degree. This crack stopped at level lower than themid-height of the wall. The Engineer has described the observed cracks as “minor” leaks, as it wasexpected for this structure. The crack width was measured using microscope, the measured crackwidths ranged from 0.06 to 0.18 mm, which is less than the allowable limit (ACI 350). Figure 10shows the crack patterns through one of the exterior walls (WE1). On the other side, the leakage testwas conducted for the steel-tank one month before. Despite the shrinkage reinforcement used in thesteei-tank was higher than that used in the FRP-tank (steel bar No. 15 @225 mm both sides versusGFRP bar No. 15 @ 250 mm both sides). However, the leakage test resuits of the steei-tank indicatedthat the number and crack widths of observed cracks in the exterior surfaces for similar waildimensions were insignificantly higher than that observed in the FRP tank. The measured crack widthsranged between 0.097 to 0.24 mm. Figure 11 shows the crack patterns through one of the exteriorwalls for the steel tank. The figure presents the leakage with indication to the start corrosion of internaisteel reinforcements. The following section presents a summery and explanation for the crack behaviorin the tank.

13

In general, the problem of cracking in concrete results from its low tensile strength. Once the tensilestrength of the concrete is exceeded, a crack will develop. The number and width of the cracks thatdevelop are influenced by the amount of shrinkage that occurs, the restrained volumetric deformationsand the amount and spacing of reinforcement provided (ACI 209R-92). Plastic shrinkage cracking is aproblem for large flat structures, such as top slab and the vertical walls of the present water tank, inwhich the exposed surface area is high relative to the volume of the placed concrete. Plastic-shrinkagecracks are immediately apparent, visible within O to 2 days of placement, while drying-shrinkagecracks develop over time. Restraint in the concrete wall of the tank is provided extemally from thecontinuous restraint along edge connection with foundation, and intemally by differential dryingshrinkage and FRP reinforcements. Typical cracking pattern due to continuous edge restraint of a thinsection is shown in Figure 12. This was where the foundation of the tank restrained the early thermalmovements of the wall had been cast after. Without restraint the section would contract along the lineof the base, and so with restraint a horizontal force developed, which had led to vertical full-sectioncracking. The restrained thermal strain at early age is varying from zero remote from the base up to amaximum strain adjacent to the base. When the tank is filled with water the tension in the walls causesan extension of the tank walls. This behavior is restrained at the base, so that the imposed strainsreduce to a negligible level adjacent to the base. As each crack forms, the propagation of that crack tothe full height of the wall will cause a redistribution of base restraint such that each portion of the wallwill act as an individual section between cracks. Prior to cracking, the stress in the longitudinalreinforcement of the wall subjected to shrinkage depends primarily on the differences in coefficientsof expansion between reinforcement (FRP or steel) and concrete. Where the coefficients are equal, thereinforcement becomes stressed as crack propagation reaches the reinforcement. The averagecoefficients of thermal expansion of concrete and steel are approximately 8 x 1 06 and 12 x 1 06/°C,respectively, while the GFRP bars used in the design have a coefficient of thermal expansion 7.0 x

6. So, the coefficient of thermal expansion of GFRP bars used is similar to that of concrete,eliminating large intemal stresses due to differences in thermal expansion or contraction and preventsany adverse effect.

Figure 12: Schematic for the cracks resulting from restrained early thermal contraction andautogenous shrinkage

The aforementioned cracking pattem is independent of the amount of reinforcement used in theconcrete wall. Only, cracking behavior can be controlled by the provision of an appropriate area ofdistribution reinforcement. When sufficient reinforcement is provided to achieve the criticalreinforcement ratio the widths of these primary cracks are controlled, although secondary cracks maybe induced. The extent and size of cracking will then depend on the amount and distribution ofreinforcement provided. The role of reinforcement is to redistribute stresses after the formation of eachcrack, (ACI 224R-0 1). A higher reinforcement ratio results in the formation of a higher number ofcracks, and hence, the crack width is reduced. This could be the reason for the observed number ofcracks in the FRP-tank, which was less than that in the steel-tank. Previous research work has showedthat for unreinforced concrete wall full-section restraint cracks can be spaced in the neighborhood of1.0 to 2.0 times the height of the wall. The leakage test results showed that the cracks spacing rangedfrom 0.6 to 0.75 and from 0.5 to 0.65 times the height of the wall in the FRP and steel-tank,respectively. This attributed to the shrinkage reinforcement used to control the crack widths and hencemore cracks have been observed. On the other hand, it has been noticed that the maximum crack widthresulting at the leakage test did flot occur at the base of the wall but at a distance of about 0.25 to 0.3H

14

from the joint (where H is the height of the wall). This is because the restraint from the base preventsthe cracks from opening local to the joint. Crack width limitations are recommended by ACI 350-O 1specifically for flexural cracks in environmental engineering concrete structures. According to thiscode flexural crack widths shouid be limited to 0.23 mm and 0.27 mm for severe and normalenvironmental exposures, respectively. However, it is flot clear which type of crack is considered inthis design guideline. Also, ACI 350 provides the minimum shrinkage reinforcement ratio that shouldbe used in the longitudinal direction based on the grade of steel bars. The shrinkage reinforcement inthe FRP-wails was designed and controlied the provisions of CAN/CSA S806- 12 and ACI 440.1 R-06,in terms of limits, spacing and required reinforcement ratio. GFRP bar No. 1 5@ 250 mm both sideswith reinforcement ratio equal to 0.0045 was used (0.00225 each side). The leakage test resuitsindicated that using S806- 12 and ACI 440.1 R-06 iimited the number of the cracks and controlied thecrack widths to reasonable values.

After completing the leakage test for each ceil in the tank, the Engineer decided to repair the crackscausing the leakage using externai injection system in each wall. Crack injection has been performedfor many years. The injection procedure would permit to fill the crack in full, from front to back.Injection has shown to be effective for fiiling cracks from 0.002 to 50 mm wide. Ail the cracks wererepaired using pressure injection of polyurethane foam sealant after inserting the stainless steelinjection ports around the cracks in the vertical direction. Figure 10 shows the crack leaking, stainlesssteel injection ports and foam sealant over the crack. This material is water activated for use in wetenvironments, no need to empty the tank. The injection sealant continues to work for considerabletime after applied. So if there is future movement the sealant wiil expand and contract compensatingfor that movement. After treatment the water has stopped leaking and the wail has started to dry out.Thereafter, the construction work for the backfiil that behind the wali has been started immediateiyafter ensuring no leakage. The exterior walls have been buried with compacted fiul-soil around and thecompaction was accomplished using a vibratory-plate compactor. Finaily, the surface area of the topslab has been buried 600 mm of uncompact fili-soil. Figure 13 shows the overview of the tankcompieteiy buried with the soil.

10. STRAIN MEASUREMENTS

10.1 Wall Tension and Compression StrainFigure 14 shows the strain measurements from the FOS attached to the GFRP bars for the interior andexterior vertical reinforcements. The initial readings for the strains were recorded few hours beforecasting (zero point at x-axis). After casting, the strain readings were captured for one week, each day.Foilowing that the 6 FOS of the wall were monitored each week. Therefore, the reported strain valuesin the first 10 days presented the shrinkage of concrete. Besides, the high temperature due to thecernent hydration at early age of concrete could be observed. Figure 1 4.a showed that the maximumrecorded compression strain ranged from 40 to 70 microstrain at the leakage test. The recorded strainvalues resuited from the water pressure, the dead weight of the top slab, and the wall itself. Aiso,Figure 1 4.b showed the sudden variation in the tension strains from compression to tension strain, as aresult from the moment on the wall due to the water pressure. The maximum tension strain rangedfrom 40 to 60 microstrain, this through the leakage test and after starting the backfihling. The measuredvalues indicated that the strains in the wali were insignificant, as it presented less than 1.0 % of theultimate strain of the used GFRP bars. This attributed for two reasons. First, by considering thestraining action of these forces and determining the maximum compression and tension stresses on thewail, the results will be approximately equal to -0.8 and 0.5 MPa, respectively. Which, it isinsignificant as compared to the strength capacity of the wall cross section, 350 mm. On the otherhand, the cracking moment of the cross section of the wall is higher than that the actual momentresulted at the ieakage test. This was confirmed from the site inspection at the leakage test as noflexural cracks were observed. The work in the backfihlîng started immediately after repairing theshrinkage cracks, this was allowed to release the water pressure by the opposite action of the earthpresser. However, the strain in the FRP bars at the exterior mat was continued to increase up to addingthe uncompacted sou over the top slab. After that, the compression strain increased insignificantiy as a

15

resuit from the dead load of the sou-fil. On the other hand, the tension strain decreased and stabilizedas the resuit from the opposite action of the water and earth pressure beside the weight of the soi! overthe top slab.

80

60

40

20

CI)

j20

-40

-60

-80

a. Interior reinforcements b. Exterior reinforcements

Figure 14: Measured strains in the FRP-vertical reinforcements in the wall (WEI).

10.2 Top SIab Strain MeasurementsThe tension strain in the GFRP bars was captured at the positive and negative moments in the top slab.Figures 15 (a) and (b) show the variation of the strain in the bottom reinforcement at the first andsecond mid-span of the top slab. The initial readings for the strains were recorded few hours beforecasting (at day 17 on x-axis). Therefore, the reported strain values in the first 10 days after presentedthe shrinkage of concrete before demou!ding the slab. From these Figures could be noticed thatinsignificant increase in the strain of the bottom GFRP bars when the fonnwork was released. Also,the strain in these bars had not been affected with the leakage test and after wal! backfihling. However,at 60 days the work of sou-fus over the top slab started. The top s!ab was buried with 300 mm of souand comp!eted after with another 300 mm. the figures show the increase in the strain as resuit ofadding these dead loads. The maximum measured strain was 155 and 77 microstrain, respective!y, inthe first and second span. These strain values represent less than 1.0% of the strain capacity of theGFRP of 15-mm diameter used in the tank. This attributed again to the determined service moment isless than the cracking moment ofthe slab (75 kN.mlm). Whereas, the maximum demimonde servicepositive moments as resuit from the dead load of the slab at the first and second span are 21 and 10

o

Figure 13: Overview of the tank completely buried with compacted fil-sou, around the walls and 600mm over the top s!ab, FOS PVC box and capturing strain data.

Top Siab L

Leakage Test

— —— Wall Backfihlrng

16

f

(.

kN.mlm, respectively. While, the corresponding values resulting from the total dead load of the slaband the saturated sou over are 48 and 22 kN.mlm, respectively. The site inspection of the top slabthrough over 60 days confirmed that no flexural cracks were observed before starting the soil-fihlingover the slab. Similar behaviour was observed for the top reinforcing bars at the negative moment overthe walls, (see Figure 1 6.a and b). However, the maximum strain was about 104 microstrain whichrepresents again less than 1% of the strain capacity of the GFRP bars. The maximum determinedservice moment at the first interior wall is 68 kN.mlm, which it is stili under the cracking moment ofthe slab.

100

50

o

150

100

‘o

50o

o

-50

-100

150

100

‘o

50o

o

-50

-100

Figure 16: Measured strains in the top FRP- reinforcements in the top slab (negative moment).

Based on the flexural design method, the expected average strains in the GFRP bars under servicedesign load is in the range 2000 to 2200 microstrain, equivalent to 123 to 133 MPa, which is muchhigher than the measured values. The main reason for these values is the service moment on the slab isless than the cracking moment. The concrete slab did flot crack, as resuit the GFRP bars have not yetbeen activated to resist tension forces. On the other hand, the compression strain at the level of topGFRP bars was captured at two locations. Figures 17 (a) and (b) show the variation of the strain in thetop reinforcement at the first and second mid-span of the top slab. The concrete strains presented inthese figures shows the early age strain variation resulted from the hydration and shrinkage. The strainvariation showed no increase due to the dead load when the formwork was removed. The maximumrecorded concrete strain was about -100 microstrain. The strain at the first span continued increasingup to -257 microstrain after 110 days of removing the formwork. The maximum strain presents 7.0%of the ultimate concrete strain used in the design (3500 microstrain), corresponding to a compressivestress of 6.50 MPa. In conclusion, the captured strain values are insignificant as compared to the

200

150

200

150

O 0 64 80 100 120 140 1

: I — — Top SIab Demoulded

— I I Leakage Test

—

— WaII Backfilling

100

50

Q

-50

-100

oO

Days

-50

80 100 120 140 1

—— Top Slab Demoulded

Leakage Test

— — Wa]I Backfihling

-100

a. at the first mid-span b. at the second mid-spanFigure 15: Measured strains in the bottom FRP- reinforcements in the top slab (positive moment).

200 200

o

1/rZZ.il. - I Days

-vi 80 100 120 140

—•• — Top Slab Demoulded

! I Leakage Test

i — —WalI Backfihling

T8o1oo12o14o

êg i . —— Top SIab Demoulded

. I Leakage Test

I — —

WaII Backfihling

O

a. over the first interior support (W1).

O

b. over the second interior support (WM)

17

allowable stress of the service design load of the tank. The main reason for that, the tank has beendesigned limiting the tensile stress in the concrete and cracks beside strength requirements. The resuitof this design controlled the concrete dimension of the wall and slab cross section to a conservativethickness (350 mm). Thus the cracking moment capacity ofthis section is higher than that the appliedservice load on the tank. Consequently, the FRP bars in the wall and top slab are affectedinsignificantly as the stress in the tension side is less than the allowable concrete tensile strength. Theoutcome of this helped significantly to optimize the FRP needed for the tank as compared with thesteel used in the steel-RC tank.

0

-100C‘o

1150

-200 -200

-250 -250 — — Top SIab Demoulded

Leakage Test— — WalI Backfihling

-300 -300

________

I

a. at the first mid-span b. at the second mid-spanFigure 17: Measured strains in the compression side in the top slab.

11. CONCLUSIONSThis paper presents the construction details and the leakage test resuits of the water treatment planttank. The tank is Canada’s first and the first worldwide field application concrete tank totallyreinforced with GFRP bars. This tank was approved by the City of Thetford Mines, Qc, Canada to bedesigned and reinforced with GFRP bars based on corrosion resistant, high strength, durability and thesuccessful field applications of these bars in the last decade. The FRP-RC tank was well instrumentedat critical locations in the walls and top slab using fiber-optic sensors. The tank was evaluated andinspected at the leakage test and after it was monitored for service condition of backfilling and buryingthe top slab with 600 uncompacted fil! soi!. Based on the construction details, the resu!ts ofthe leakagetest and strain data captured under the service condition, the fo!lowing conclusions can be drawn:

• The GFRP bars provided an efficient solution to overcome the expansive stee! corrosion issues andrelated deterioration problems in the corrosive Water Tank Chlorination in the water treatmentplant.

• Design and construction procedure of this project are used to illustrate code requirements, tankanalysis, design details, and construction of FRP-reinforced concrete tank.

• The design provisions used in the of Water Tank Chlorination showed that the proposedreinforcement ratios adopted by the codes and guidelines: the CAN/CSA S806-1 2 for design andconstruction of building components with fibre reinforced polymers; ACI 440.1 R-06 Guide for thedesign and construction of structural concrete reinforced with FRP bars, and ACI 3501350R-06Code requirements for environmenta! engineering concrete structures and commentary, areadequate for satisfying the serviceability and strength criteria.

‘No obstac!es to construction were encountered due to the use of the GFRP bars in all theconstruction procedures of the tank. The GFRP bars withstood normal on-site handling andplacement with no prob!ems.

• The FRP-RC water tank performed very well and it was able to withstand applied !oads without upnormal cracks or leaking at the !eakage test.

• The GFRP reinforced concrete wall, foundation and slab of the investigated water tank showednorma! structura! performance in terms of strain and cracking, through 8 months in real service

o

-50

o

-50

-100C‘o

s-150o

18

condition. The maximum measured strains in the GFRP bars did not exceed 1.0% of the straincapacity of the GFRP bars employed in the project. Nevertheless, it is lower than the expectedstrains using the flexural design method. This attributed to the design concept considered to controlthe crack widths and limit the tensile concrete stresses in the tank which lead to cracking momentresistance higher than that resuit from the actual applied loads. Limited shrinkage cracks wereobserved with insignificant crack widths ranged from 0.06 to 0.18 mm, which is less than theallowable limit according to ACI 350-06. No flexural cracks have been observed in the tank.

• The concrete cover has been reduced in the FRP-tank as compared with that used in the steel tank,as there is no issue for the corrosion with FRP bars.

• The cost effectiveness of using GFRP bars in the tank has been optimized by using the two Grades(II and III) in the longitudinal and transvers direction, respectively, and also by utilizing lower barsize diameter other than using larger diameter with smaller spacing.

• Finally, this successful field application shows the effective usage of the GFRP reinforcing bars inreinforced concrete tank for the water treatment plant the first time in the world. The structuralperformance of this first worldwide application of its type and scale, based on the monitoring andcontinuous observations, is as expected. No major problems or any un-expected performanceassociated troubles appeared during the construction or after being in service for six months. Thisapplication opens the door of major application of the FRP reinforcing bars in reinforced concretewater tanks in North America and World-Wide. Reinforcing the concrete water tank with GFRPbars would extend the life of the structure to 100 years or more compared to steel-reinforcedconcrete, which needs major restoration after 25 years.

ACKNOWLEDGMENTSThe authors would like to thank and express their sincere appreciation to the NSERC—CanadaResearch Chair and Industrial Research Chair group (Canada Research Chair in Advanced CompositeMaterials for Civil Structures & NSERC Research Chair in Innovative FRP Reinforcement forConcrete Infrastructure) at Department of Civil Engineering, Faculty of Engineering, University ofSherbrooke, QC, Canada, for providing technical data along with numerous testing and reports on FRPreinforcement for concrete infrastructure. Also, the authors acknowledge technical data and thecontribution to this project from Pultrall Inc., (Thetford Mines, Qc, Canada).

REFERENCESAmerican Concrete Institute (ACI). (2006). Guide for the design and construction of concrete

reinforced with FRP bars, ACI 440.1R-06, Farmington Hills, Mich.American Concrete Institute (ACI). Control of cracking in concrete structures, (ACI 224R-0 1).

Farmington Hills (MI): American Concrete Institute; 2001.American Concrete Institute (ACI). Code requirements for environmental engineering concrete

structures and commentary (ACI 350). Farmington Hills (MI): American Concrete Institute;2001/2006.

ACI Committee 209. Prediction of creep, shrinkage and temperature effects in concrete structures,(ACI 209R-92). Farmington Hills (MI): American Concrete Institute; 1992 (reapproved 1997).

ACI Committee 318. Building code requirements for structural concrete (ACI 318-08) andcommentary (ACI 31 8R-08). Farmington Hills (MI): American Concrete Institute; 2008.

Armin Ziari, M. and Reza Kianoush, “Investigation of flexural cracking and leakage in RC liquidcontaining structures”, Engineering Structures, 31, 1056-1067 (2009).

Benmokrane, B., Wang, P., Ton-That, T. M., Rahman, H., and Robert, J. F. (2002). “Durability ofglass fiber-reinforced polymer reinforcing bars in concrete environment.” J. Compos. Constr., 6(3),143—153.

Benmokrane, B., El-Salakawy, E.F., Desgagné, G., and Lackey, T. (2004). “Building a NewGeneration of Concrete Bridge Decks using FRP Bars.” Concrete International, the ACI Magazine,Vol. 26, No. 8, August, pp. 84-90.

Benmokrane, B., El-Salakawy, E., El-Ragaby, A., and Lackey, T. (2006). “Designing and Testing ofConcrete Bridge Decks Reinforced with Glass FRP Bars.” Journal of Bridge Engineering, 11(2):2 17-229.

19

Benmokrane, B., El-Salakawy, E., E1-Ragaby, A., and El-Gamal, S. (2007). Performance Evaluationof Innovative Concrete Bridge Deck Slabs Reinforced with Fibre- Reinforced Polymer Bars. CanJour of Civil Eng, 34(3): 298—310.

Benmokrane, B., Ahmed, E., Dulude, C., and Boucher, E. (2012) “Design, Construction, andMonitoring 0F the First Worldwide Two-Way Flat Slab Parking Garage Reinforced with GFRPBars.” 6th International Conference on FRP Composites in Civil Engineering.CD proceeding,CICE 2012, Rome, Italy.

CSA Technical Committee on Reinforced Concrete D sign. A23.3-04 Design of concrete structures.Rexdale (Ontario): Canadian Standards Association; 2004.

Canadian Standards Association (CSA). (2002-12). “Design and construction of building componentswith fiber reinforced polymers.” CAN/CSAS8O6-02, Rexdale, Toronto.

Canadian Standards Association (CSA). (2006-Editiôn 2010). “Canadian highway bridge designcode—Section 16, updated version for public review.” CAN/CSA-56-06, Rexdale, Toronto.

Canadian Standards Association (2010). CAN/CSA S807- 10 “Specification for fibre-reinforcedpolymers, Canadian Standards Association, Rexdale, Ontario, Canada, 44 p.

El-Salakawy, E., Benmokrane, B., and Desgagné, G. (2003a). “FRP composite bars for the concretedeck slab of Wotton Bridge.” Can. J. Civ. Eng., 30(5), 86 1—870.

El-Salakawy, E., Benmokrane, B., Masmoudi, R., Brière, F., and Eric Breaumier, E. (2003b).“Concrete Bridge Barriers Reinforced with Glass Fiber-Reinforced Polymer Composite Bars.” ACIStructural Journal, 100(6): 8 15—824.

Hamdy M. Mohamed and Brahim Benmokrane (2012) “Field Application of FRP Bars in Tunnels”Mçntreal TAC 2012, Tunnels and Underground Spaces: Sustainability and Innovation, proceedingson CD-Rom, 17-20 October, Montreal, Qc, Canada, 6p.

Hamdy M. Mohamed and Brahim Benmokrane (2012) “Recent Field Applications of FRP CompositeReinforcing Bars in Civil Engineering Infrastructures” ACUN6 —Composites and Nanocompositesin Civil, Offshore and Mining Infrastructure, 14 — 16 November 2012 Melbourne, Australia,6p.

Mathieu Robert, Cousin P, Benmokrane B. Durability of GFRP Reinforcing Bars Embedded in MoistConcrete. J of Composites for Construction. 2009; 13(2): 66-73.

Mathieu Robert, Cousin P, Wang P, Benmokrane B. Temperature as an Accelerating Factor for LongTerm Durability Testing of FRPs: Should there be any limitations? J of Composites forConstruction. 2010; 14(4): 361-367.

20