Asphalt Concrete Core of the Meijaran Dam in Brief

Dam Engineering Vol XX Issue 3 235

Asphalt concrete core of the Meijaran dam inbrief

Alireza Sharifi Soltani, Principal Geotechnical Engineer & Project Manager, BSc &Siavash Litkouhi, Chairman, PhDSoils Engineering Services Co#135, Zhoubin Cul-de-SacNorth Dastour StreetGheitariehTehran 19317-64611IranEmail: [email protected] / [email protected]

AbstractThe lack of clay borrow sources at the Meijaran dam site, along with a high annualprecipitation rate in the area, led consulting engineers to replace the clay core with anasphalt concrete core (ACC) at the rockfill dam. The steps taken for the project fromthe start, i.e. the selection of suitable aggregate sources for the asphalt concrete (AC)and the adjacent filter materials to the end, i.e. ACC placement in the dam body, arebriefly discussed. AC mix design parameters, which were finalized upon completionof laboratory tests are presented, as well as the scope of trial placements conductedfor optimizing the hand - and machine - placement of the ACC. Proper results weregained both in the final trial placement and in the dam body, in accordance with thequality control program. The air void contents measured in the AC cores were allbelow three percent. Construction of the dam took about 18 months which, whencompared with the placement of a clay core, is a very short length of time for damconstruction activities along the northern coastal region of Iran. The Yangtze ThreeGorges Technology & Economy Development Co (TGDC) supervised the wholework.

Keywords: Asphalt concrete core, clay core, rockfill dam, Marshall test, air void con-tent, trial placement, hand placement, machine placement, drilled core.

1. IntroductionFor many years clay core has been conventionally used in embankment dams, as asealing element in the dam body. However, when there is a lack of clay borrowsources at the dam site, or it is not available within an economically viable distance

Dam Engineering Vol XX Issue 3236

[5], it can be substituted with ACC. At some locations, despite the availability of clayborrow sources, due to the rainy climate, the moisture content of the clay is equal oreven above the optimum moisture content, which creates problems in the constructionof the clay core.

In addition, ACC has been chosen as an alternative for the dam body watertightnessbecause of its deformability, erosive strength and resistance to ageing [5]. It is virtual-ly impervious, offers jointless embankment core construction, and is workable andcompactable. The self-healing (self-sealing) ability of the ACC, provided due to itsviscoelastic, plastic and ductile properties (provided that the mix is properlydesigned), could compensate for the problem of cracks developing in the core wall [4].Good contact also exists between the ACC and the material of the embankment [6].Compared to a clay core, the placement and compaction of ACC is much less suscepti-ble to rainy conditions and, as opposed to earth materials, AC is man-made and itscontrolled properties can be tailored to satisfy specific design requirements [2]. Withcontinuous vertical positioning of the core, the downstream drainage layer permits fullcontrol of the seepage water by means of the drainage layer between the embankmentand core [5]. In the case of a vertical core, ACC is the smallest sealing area possiblefor a dam structure [5].

Apart from the comprehensive quality control program used in the dam construc-tion, which covers materials, AC production, placement and compaction, the dambody compaction should also be tightly controlled. The thin ACC has to follow andadjust to the movements and deformations imposed by the embankment as a whole.These deformations must be accommodated by the AC without cracking or significantshear dilation, which may lead to increased permeability. To reduce the probability ofcore cracking due to excessive static and/or dynamic embankment deformations anddistortions, the embankments have to be well compacted [2]. Dam slopes as steep as1:1.3 are reported for a well-compacted ACC embankment dam of good rockfill rest-ing on bedrock [4].

For repair purposes, the grain composition of the upstream transition zone should bedesigned to allow the sinking of boreholes for grouting purposes. Boring within the actu-al core wall, and direct sealing of the AC, is only purposeful in exceptional cases [5].However, disregarding extreme earthquake movement, where the core may be partiallysheared off (for which lowering the reservoir level and executing repairs is needed) [2],the design and construction of an ACC embankment dam should be tightly controlled asthe methods of ACC repair are costly.

With regard to all the points previously mentioned, ACC was evaluated to be a suit-able sealing element in comparison with the clay core for the Meijaran dam. In addi-tion, using this technology is very useful for dam construction technology in Iran, espe-cially where lack of clay, or a rainy climate, causes problems in the construction of thedam.


2. A brief introduction to the Meijaran damThe Meijaran dam is a 52m high rockfill dam which is located approximately 20kmsoutheast of Ramsar, a coastal city north of Iran, adjacent to the Caspian Sea. It hasbeen constructed on the Nesaroud River. Table 1 provides the main features of theMeijaran dam, and Figures 1 and 2 show typical cross-sections of the dam body andACC, respectively. Grading of the transition and rockfill shell zones are shown inFigures 3 and 4.

5

6.010.3

112.0

7 109 6

5

9 8

44

12

338.0 105.0

1.0 : 2.0

1.0 : 1.75 1.0 : 1.6

1.0 : 1.5

N.W.L. 148.08.0 153.0

1.0: 1.5

1.0 : 1.6

170.0

160.0

150.0

140.0

130.0

120.0

110.0

100.0

90.0

1- Asphalt concrete core

2- Filter

3- Transition

4- Rockfill shell

5- Protection layer

6- Toe drain

7- Upstream cofferdam

8- Concrete plinth

9- Bedrock

10- Grout curtain

Figure 1. A typical cross-section of the dam body together with compaction

specifications of the dam zones (all dimensions in metres)


4.0 4.01.5 1.5

1.0

to0.

7

1

3.0

2.0

:1.0

3.0

3.0 3.0Variable length

2.0:1.0

1.5 2.0

5.0

1.5

Flared asphalt concreteMastic

Concrete plinth

Figure 2. ACC cross-sections (all dimensions in metres)

Table 1. Main features of the Meijaran dam


3. AC aggregates and filter test resultsPrior to commencing laboratory tests, several aggregate sources (quarries) were selected,and samples collected. The decision was made to extract both the aggregates for the ACand the filter material (i.e. materials adjacent to the AC) from the same quarry. Table 2shows the test results from the selected quarry.

100

90

80

70

60

50

40

30

20

10

0

Perc

entp

assi

ng(%

)US sieve numbers/opening #2

00

#100

#50

#30

#16

#10

#4 3/8́

´

3/4́

´

1.5́

´2́

´3́

´4́

´

0.001 0.01 0.1 1 10 100 1000

Clay Silt Fine FineMedium Coarse Coarse CobblesBouldersSand Gravel

Figure 3. Grading tolerances of the transition zone

100

90

80

70

60

50

40

30

20

10

0

Perc

entp

assi

ng(%

)

US sieve numbers/opening #200

#100

#50

#30

#16

#10

#4 3/8́

´

3/4́

´

1.5́

´2́

´3́

´4́

´

0.001 0.01 0.1 1 10 100 1000

Clay SiltFine FineMedium Coarse Coarse Cobbles

BouldersSand Gravel

Figure 4. Grading tolerances of the rockfill shell zone


Grading of the ACC, together with the filter material, are shown in Figures 5 to 7,respectively.

Table 2. Some of the test results of the selected quarry materials

100

90

80

70

60

50

40

30

20

10

0

Perc

entp

assi

ng(%

)


#100

#50

#30

#16

#10

#4 3/8́

´

3/4́

´

1.5́

´2́

´3́

´4́

´

0.001 0.01 0.1 1 10 100 1000


BouldersSand Gravel

Fuller’s CurveTolerances

Figure 5. Grading tolerances of the ACC (Fuller’s Curve)


100

90

80

70

60

50

40

30

20

10

0

Perc

entp

assi

ng(%

)US sieve numbers/opening #2

00

#100

#50

#30

#16

#10

#4 3/8́

´

3/4́

´

1.5́

´2́

´3́

´4́

´

0.001 0.01 0.1 1 10 100 1000


BouldersSand Gravel

Figure 6. Grading tolerances of the filter (upstream)

100

90

80

70

60

50

40

30

20

10

0

Perc

entp

assi

ng(%

)


#100

#50

#30

#16

#10

#4 3/8́

´

3/4́

´

1.5́

´2́

´3́

´4́

´

0.001 0.01 0.1 1 10 100 1000


BouldersSand Gravel

Figure 7. Grading tolerances of the filter (downstream)


4. Bitumen test resultsB60 type bitumen was used for the mix design. Table 3 represents some of the accept-able ranges specified for this type of bitumen. The B60 sample collected for use at theMeijaran dam complied fully with the ranges specified below.

5. AC mix parametersAC test samples containing various bitumen content were prepared, ranging from 5.5-7.0percent, on which Marshall tests were performed. Table 4 presents the final bitumencontent and various other parameters of the AC laboratory samples.

Subsequent to finalizing the laboratory AC mix design, additional tests were performedon the corresponding laboratory samples, the results of which are presented in Table 5.

Table 3. Some of the acceptable ranges for B60 type bitumen

Table 4. Final results of the AC laboratory test samples


6. ACC design parametersTri-axial tests were performed on the AC and filter samples, and the results were ana-lyzed for evaluating Duncan and Chang parameters [3]. The average values of severaltest groups are shown in Table 6.

7. Design of the dam body based on ACC parametersThe dam body was analyzed for both static and dynamic loading conditions. The seismicanalysis of a dam with ACC has not been extensively investigated in literature [1], anddoubts still exist regarding the behaviour of the dam body during an earthquake. In thisregard, analysis of the Meijaran dam body with ACC was an important phase of the pro-ject. Static and dynamic analysis of the dam body was performed by the project’s con-sulting engineers, using the finite element method. The hyperbolic model provided byDuncan and Chang [3] was used for the static analysis, for which the correspondingparameters of the AC and filter are presented in Table 6. For dynamic analysis, the elas-tic and plastic strains were computed using the Mohr-Coulomb model. The shear modu-lus at small strains (Gmax) was selected as 100MPa. The magnitudes for DBL and MDLwere taken as 6.0 and 6.8 on the Richter Scale, respectively.

Table 5. Results of various other tests on the AC laboratory test samples

Table 6. ACC design parameters


According to the static analysis results the ratio of the existing shear stress to the shearstrength, both at the end of construction and impounding, was less than one, whichshows the stability of the dam body.

According to the dynamic analysis for the DBL condition, core settlement was about1cm. The maximum horizontal displacement of the core was 4cm at three quarters of thedam height. In this case the maximum displacement of the dam body was 13.6cm.

In the case of the MDL loading condition, the maximum displacement of the dambody was 47.9cm at the upstream shell, and the core settlement at the crest was 8cm. Themaximum horizontal displacement of the core was 18cm at three quarters of the damheight. In addition, the sum of the vertical stress at the core and its tensile strength wasmore than the hydrostatic pressure in height. This implies that the sealing element of thedam body (i.e. ACC) will remain stable at MDL, and that the dam free hydraulic headloss is negligible. The final dam body dimensions are presented in Figure 1.

8. AC production at the siteSubsequent to completion of the laboratory tests, and finalization of the AC mixdesign, the process was repeated once more at the site in order to control the mixdesign for the AC produced by the asphalt concrete plant (ACP). This included check-ing the gradation of the AC aggregates as input and output of ACP, checking bitumenand aggregate properties, and AC production. Marshall samples were prepared and therequired tests, as indicated in Tables 4 and 5, were performed. The results showed thatthe AC produced by the ACP was suitable for commencing a trial placement.

9. Trial placementPrior to ACC placement in the dam body, a trial placement was conducted for a) practic-ing the activities needed for preparing the dam axis for ACC placement, b) ACC contin-uous placement, c) optimizing the rolling passes, heat control and investigating thesimultaneous sequence of AC and filter compaction, and d) collecting drilled cores fortesting the properties of the AC, especially the air void content.

Trial placements were also required even during construction of the dam body where,due to various reasons, the component materials of the AC (bitumen, aggregate, filler,etc), or the placement method is changed.

During this project trial placements were conducted three times. Firstly, at thebeginning of the work, and prior to construction of the dam body, for simulation ofhand placement of the ACC in the dam body. Secondly, during construction of thedam body when an ACC paver machine was imported and, lastly, when the originalquarry used for the ACC materials was changed due to the termination of extractablevolume, which was followed by a new mix of the original AC design.

However, for the main part of the work the hand placement method was applied. Atfirst the steel formworks were set up so that the centerline of the formwork rows (the


upstream and downstream rows) were located at the dam axis. The filters were thenplaced either side of the steel formworks (while the centerline was being simultaneouslycontrolled to remain at the dam axis).

After completion of the filter placement, the ACC placement began. The temperatureof the lower ACC layer was measured prior to placement, and if the layer temperaturewas below 80ºC it was preheated using infrared heaters. Upon completion of the ACCplacement the formworks were extracted, and compaction of the filters and the ACC wasconducted.

Although it has been suggested that when compared to machine placement there is thepossibility of defects as a result of inadequate compaction, and a possible cooling of themix [5], using the hand placement method is inevitable especially against the foundationand abutments, and the quality of the cores in these locations can be as high as for thatplaced by a machine [4]. Plates 1a and 1b show the trial placement of ACC by a pavermachine.

Plate 1a. ACC trial placement by paver (front view)


Plate 1b. ACC trial placement by paver (rear view)

Plate 2. A completed trial layer of ACC


Upon completion of the trial placements optimized compaction passes, suitablesequence of rolling, proper quality of compacted AC, as well as air void content lessthan three percent (controlled from the bottom of the layer to the top by cutting thedrilled cores into pieces, each about 5cm in height), were achieved. Figure 1 presents thecompaction specifications of ACC and filter, along with other zones of the dam body.

The width of the trial placements (hand placed) were all 1-2m, and the compactedthickness was approximately 0.2m. Plate 2 shows a completed layer of trial ACC.

10. ACC placement in the dam bodyACC placement in the dam body commenced after termination of the first trial place-ments, using the hand placement method. The first few layers were considered a trialplacement at the dam body. A special rolling sequence was practiced during this trialplacement, as the ACC was flared in both the vertical and horizontal sections (in order toprovide more efficient water tightness at the bottom and at the abutments, thus prevent-ing leakage). Hence, a 2m wide AC had to be rolled, where the thickness at the abut-ments reached 3m. After placing the first two layers in the dam body, drilled cores werecollected at several locations, including both of the said layers, and also from upper partsof the concrete plinth (both on the foundation and abutments). Plate 3 shows some of thecollected cores from the dam body.

Plate 3. Drilled cores collected from the ACC in the dam body


The results from these drilled cores and the Marshall samples showed correct qualityof AC, suitable bitumen content and grading, and acceptable stability and flow values.The air void contents were also less than three percent. Proper contact between the twolayers (caused by preheating, using electrical infrared heaters) was also gained. Theadhesion between the first layer, the mastic layer and the concrete were suitable, both atthe foundation and abutments. The results of all visual, physical and mechanical testsconformed to the specifications. Plate 4 shows preheating of the ACC layers.

Plate 4. Preheating of the lower ACC layer


The next layers were placed upon approval of the first two layers which had been laidproperly, and conforming to the specifications. The maximum placement rate of about300ml/day was achieved, though care was taken to control the rate of heat loss in subse-quent layer placements. Research has been carried out in Norway which aimed toincrease productivity rates for ACC, either by increasing the number of asphalt layersplaced per day and/or increasing the thickness of each layer [7]. According to the resultsof this research, air void contents below three percent could also be obtained with a ACCplacement rate of four layers of 0.2m per day, or three layers of 0.3m per day. Thiswould be especially useful along deep valley sections where due to the short length ofeach layer, and despite having a high productivity rate, activities would conventionallybe delayed due to heat loss in the lower layer. Plate 5 shows the hand placement of ACCin the dam body.

To collect qualified drilled cores, the ACC needs to be left to cool for 4-5 days. Asthis method of coring and testing is still the most reliable for in situ air void contentmeasurements [7], the only way to compensate the time loss in this regard is to increasethe level of quality control for each step of the work. When there is no change in the

Plate 5. Hand placement of ACC in the dam body (loader is transporting the asphalt concrete to

be placed, while the temperature of the ACC in the bucket is being controlled)


conditions of the ACC (viz. component materials, bitumen, general conditions, com-paction, etc.), the number of coring and cooling periods could be lowered. This could becarried out after placement of enough layers where sufficient number of air void contenttests has confirmed the quality of work. Plate 6 shows the straight set of formworks inthe dam body.

Plate 6. Steel formworks set straight along the dam axis


In the dam body the transition zone as well as the rockfill shell, was placed uponcompletion of the ACC layer. To prevent water stoppage in the ACC location duringrainfall the level of the ACC, as well as the filters, was kept higher than the adjacentfills by 1-1.5m.

11. Results of the ACC tests within the dam bodyThe test results on the drilled core samples were 100 percent qualified in unit weight, airvoid content (<3 percent), and permeability (<10 E-7cm/sec), as well as for cohesion andthe angle of internal friction obtained by tri-axial tests.

The results of the Marshall tests performed on the samples collected from the ACPoutput were 99.4 percent and 95 percent qualified for stability and flow values, respec-tively. The drilled cores collected near the edge of the ACC showed proper contactbetween the AC layer, the mastic and the concrete at the abutments.

During dam construction, a sudden flood caused a rise in the reservoir water head toel. +122m which, in turn, produced about 22m water head upstream (far below the fin-ished ACC level at the time). This water head lasted for about two weeks as a result of aproblem with the hydro mechanical equipment, during which no leakage was registeredeither at the abutments, or in the dam body.

To date the maximum reservoir free water elevation level has risen to 135m.According to the consultant engineer’s report continued monitoring of the dam hasshown no problem with leakage, either at the abutments or the foundation, as well asfrom a pore water pressure viewpoint.

12. Points to consider in ACC placementAlthough using ACC in embankment dams is useful in climates similar to that of north-ern Iran, conducting such projects needs serious attention to details, e.g.:

a) The quality of bitumen produced by the refineries (which is not necessarily under thecontractor’s control). Unqualified bitumen loads, if delivered to the site, are not allowedto be used. Hence, extra costs are imposed to the project both in budget and time.b) An elaborate and systematic quality control program should be prepared and fol-lowed. This includes component materials, tests, and specifications, as well as AC pro-duction, transportation, unloading, compaction and protection.c) As continuous and efficient placement by hand requires more manpower thanmachine placement, the quality control program should be more restricted. During con-tinuous hand placement in the dam body, sections of ACC are being dealt with simulta-neously, while various other activities are being conducted at each of the remaining sec-tions, e.g. one section is being cleaned for formwork installation, another is having theformwork installed and the filter spread, in another section the AC is being unloaded andthe lower layer preheated, while in another rolling is underway. At the same time lorry


loads of the filter are transported and unloaded around the axis, whilst the working teamcontend with the traffic caused by the machinery engaged in the construction of the dambody. To efficiently produce an ACC which meets all the specified requirements, an effi-cient and powerful management team is required to ensure that activities at the dam site,as well as the quality control, are well organized.

13. ConclusionsConsidering the merits of the asphalt concrete core in the construction of a dam body, itshould be considered an important alternative in dam construction technology in Iran.Providing the mix design is properly conducted and the AC production, placement andwork quality are properly supervised and controlled, this sealing element in the dambody will be both watertight and stable. However, when dealing with hand placement,the sequence and continuity of activities, as well as the quality control, should be moreeffectively conducted in comparison with machine placement. Where needed, thismethod of construction [7] could be an alternative for increased productivity.Enhancement of the quality control at each stage of the project, together with obtainingsufficient qualified test records of the air void content, could be an alternative fordecreasing the number of coring periods, provided that none of the details of the works(based on which the core void results are obtained) are changed.

14. AcknowledgementsThe first author wishes to acknowledge Professor Kaare Höeg for studying this paperand giving valuable comments. The authors also wish to acknowledge the MazandaranRegional Water Co, Tehran Sahab Consulting Engineers, and the Kotra Construction Cofor their cooperation with the specialist contractor both during the project and whilstpreparing this paper.

References[1] Ghanooni Mahabadi, S & Mahin Roosta, R, (2002), ‘Seismic analysis and design ofasphaltic concrete core embankment dams’, International Journal on Hydropower &Dams, Vol 9, Issue 6, pp75-78.

[2] Höeg, K, (1993), ‘Asphalt concrete cores for embankment dams”, Statkraft,Veidekke and NGI, StikkaTrykk, 1993.

[3] Duncan, J M & Chang, C Y, (1970), ‘Non-linear analysis of stress and strain insoils”, Journal of the Soil Mechanics & Foundations Division, ASCE, Vol 96, No SM5,pp1629 - 1653.


[4] Höeg, K, (1998), ‘Asphalt core embankment dams’, International Journal of DamEngineering, Vol 9, Issue 3, pp110-132.

[5] International Commission on Large Dams (ICOLD), (1992), ‘Bituminous cores forfill dams - state of the art’, Bulletin 84.

[6] International Commission on Large Dams (ICOLD), (1982), ‘Bituminous cores forearth and rockfill dams’, Bulletin 42.

[7] Saxegaard, H, (2002), ‘Asphalt core dams: increased productivity to improve speedof construction’, International Journal on Hydropower & Dams, Vol 9, Issue 6, pp72-74.

Asphalt Concrete Core of the Meijaran Dam in Brief

Documents

Transcript of Asphalt Concrete Core of the Meijaran Dam in Brief