With growing environmental pressures to reduce clinker content in cement, the Canadian Standards Association introduced the Portland Limestone Cement classification. This article details the considerable lab testing done to show that PLC with up to 15% limestone is performance equivalent to portland cement in strength and durability.
By Steve Prokopy
INTRODUCTION
In response to growing pressures to reduce the clinker content in cement and hence the CO2 emissions associated with its production, the Canadian Standards Association (CSA A3001-08) introduced a new classification of cement in 2008, this being portland limestone cement (PLC) containing up to 15% limestone. PLC is now permitted in the production of all classes of concrete in Canada except for sulfate-exposure classes (CSA A23.1-09).
Considerable laboratory testing has been conducted in Canada in recent years to demonstrate that PLC with up to 15% limestone can be manufactured to produce equivalent performance to portland cement in terms of concrete strength and other properties, including durability. The equivalent performance is achieved by optimizing the PLC with regards to composition and particle-size distribution, and requires intergrinding rather than blending of the portland cement and limestone.
This article presents data from a number of recent full-scale field trials where PLC has been used in paving projects in Quebec, Alberta and Nova Scotia. In each field trial, the performance of PLC concrete has been compared with that of equivalent concrete produced using portland cement (PC), and concretes have been produced with varying levels (up to 50%) of supplementary cementing materials (SCM).
Testing of concrete produced during the field trials has included compressive strength, freeze-thaw and deicer-salt-scaling resistance, chloride permeability, and chloride diffusion. The results indicate that concrete performance is strongly influenced by the level of SCM used, but is independent of whether PLC or PC is used.
The use of such PLC has the potential to bring about a 10% reduction in the greenhouse gas emissions associated with the production of portland cement clinker. The acceptable performance of concrete containing combinations of PLC and SCM will permit further reductions in the CO2 footprint of the concrete. In one mix, the clinker content was just 41% of the total mass of cementing material.
BACKGROUND
PLC is produced by blending portland cement and limestone or, preferably, intergrinding portland cement clinker, limestone and calcium sulfate. Such cements have been allowed by the European Standard (EN197-1) since 2000, although a number of European countries allowed their use through national standards for a decade or more prior to this date. EN197-1 allows up to 20% limestone in CEM II/A cements and up to 35% in CEM II/B cements.
In Canada, PLC containing up to 15% limestone was first permitted by the cement standard (CSA A3001) in 2008 and by the concrete standard (CSA A23.1) in 2009. Prior to acceptance of PLC in the Canadian Standards, considerable laboratory testing was performed to demonstrate that PLC containing up to 15% limestone could be manufactured to provide equivalent performance to PC in terms of strength and durability in concrete mixes with and without supplementary cementing materials (SCM). Some of the laboratory data have been recently published (Thomas et al. 2010a; Hooton et al. 2010).
The equivalent performance is achieved by optimizing the PLC with regards to composition and particle-size distribution, and requires intergrinding rather than blending of the PC and limestone. It should be noted that up to 5% interground limestone is permitted in ordinary PCs in Canada, and that typically PC will contain approximately 3% to 4% limestone interground with the cement and clinker.
The acceptance of PLC as a substitute for PC potentially allows the clinker content of cements in Canada to be reduced by up to about 10%. The production of PC clinker results in significant CO2 associated with the calcination or decarbonation of limestone (CaCO3heatCaO+CO2) and the combustion of fossil fuels to achieve the clinkering temperature of 1450°C (2640°F). Approximately 1 kg of CO2 is produced for each 1 kg of clinker, although the precise amount varies depending on the fuel efficiency of the plant. It is estimated that cement production accounts for about 5% of the CO2 produced globally. Reducing the clinker content of cement by 10% will effectively reduce the CO2 emissions associated with its production by the same amount. Of course, further reductions are possible by adding SCM (such as fly ash, slag or natural pozzolans) either directly to the cement (blended cements) or by partially replacing cement with SCM at the concrete mixer.
This article presents data from the three paving trials that were conducted in Quebec, Alberta, and Nova Scotia in 2008 and 2009. In two of the trials, concrete mixtures were produced with both PC and PLC, and varying amounts of SCM. In the third trial, concrete mixtures were produced with blended cement containing PC + 15% slag or PLC + 15% slag, and varying amounts of fly ash were incorporated at the concrete mixer. The mixes were tested to determine the setting characteristics, strength development and durability including freeze-thaw, deicer-salt scaling and chloride permeability.
FIELD TRIALS
The three trial paving projects were located in Quebec, Alberta and Nova Scotia. Details of the cementing materials and concrete mixtures used at the three sites are presented in Tables 1 and 2. Note that this article uses the following nomenclature from CSA A3001 for the cementing materials:
All three projects were placed in the fall (2008 or 2009), and the pavements are exposed to frequent applications of deicing salt, cyclic freezing and thawing, and heavy truck traffic (either cement tankers or ready-mixed concrete trucks).
Paving Slab at Gatineau Ready-Mixed Concrete Plant, Quebec
The first field trial was conducted using PLC with 12% interground limestone produced in Lafarge’s Bath cement plant in Ontario. A total of eight concrete mixtures were produced, four with PLC and four with PC from the same plant. The total cementitious materials content of all mixtures was 355 kg/m3 (598 lb/yd3) and the water-to-cementing-materials ratio was W/CM = 0.44 to 0.45. A blended SCM (2 parts slag and 1 part fly ash) was added at the ready-mixed concrete plant at cement replacement levels of 0%, 25%, 40%, and 50%.
The concrete was used to construct a parking slab (4500 ft2, 450 m2) at the concrete plant. The concrete was placed in October 2008. Extensive laboratory testing was conducted on specimens cast during the placing of the concrete and the results were recently reported in a paper by Thomas et al. (2010). In the PLC mix with 50% SCM, the clinker only constituted approximately 41% to 42% of the total mass of cementing materials. This compares with about 91% to 92% clinker for the control mix produced with PC and no SCM (PC contains approximately 3% to 4% limestone and 5% gypsum).
Pavement at Exshaw Cement Plant, Alberta
The second field trial was conducted using PLC produced at Lafarge’s Exshaw cement plant in Alberta. The PLC was produced by intergrinding 12% limestone. This trial also incorporated 4 concrete mixes with PLC and 4 with PC, with fly ash being added at the ready mix plant at levels of 0%, 15%, 25%, and 30%. The total cementitious materials content of all mixtures was 410 kg/m3 (691 lb/yd3) and the water-to-cementing-materials ratio was W/CM = 0.37 to 0.42.
These concrete mixtures were used for paving but additional PC and PLC mixtures were also produced for two retaining walls and a section of slipformed curb. The concrete was placed in September 2009. The pavement was 0.30 to 0.45 m (12 to 18 in) thick and was reinforced with a single mat of reinforcement. The concrete was placed by pump, struck off, bull floated and tined. After finishing, the surface was treated with an evaporation retarder as it was windy. Finally, a curing membrane was applied.
Pavement at Brookfield Cement Plant, Nova Scotia
The third field trial was conducted using a blended portland limestone cement produced by intergrinding 12% limestone and 15% slag granules together with the portland cement clinker and gypsum at Lafarge’s Brookfield cement plant in Nova Scotia. The performance of this cement was compared with a similar blended PC produced with 15% slag. Six concrete mixes were produced with these two cements, with fly ash being added at the concrete plant at replacement levels of 0%, 15%, and 20% fly ash. The total cementitious materials content of the concrete mixtures was 384 to 392 kg/m3 (647 to 661 lb/yd3) and the water-to-cementing-materials ratio was W/CM = 0.42 to 0.44. The concrete was used to pave the roadway outside the main entrance of the cement plant and was placed in October 2009. The pavement was 0.30 m (12 in) thick, transverse joints were saw cut (no dowels), and the longitudinal joint between the two lanes was dowelled. Placing was carried out using a deck finishing machine and the surface was bull floated, broom textured and sprayed with a curing membrane.
TEST RESULTS
The results of tests conducted on the concrete for the paving slab in Quebec have been reported in detail elsewhere (Thomas et al. 2010b), and a summary is provided in Table 3. The results from this trial show significant improvements in the long-term strength and the resistance to chloride ion penetration as the SCM content of the concrete increases. However, at a given level of SCM there is no consistent difference between the performance of the concrete with PLC versus PC. The paving slab has now been through two winters and there are no signs of deicer salt scaling for any of the mixes even when the PC clinker represented less than 45% of the total cementing material. This confirms the satisfactory performance of these mixes in laboratory deicer salt scaling tests (Thomas et al. 2010b) and is encouraging as it demonstrates that PLC can be used with high levels of SCM (50% in this case) without compromising durability.
For all three trials, the concrete mixes with PLC generally required more air-entraining admixture than equivalent mixes with PC. No noticeable differences were observed in the water demand and placing characteristics of similar concrete mixes produced with either PC or PLC, however, mixes with PLC appeared to have a higher paste content and were easier to finish. Set times were only measured for the concretes mixtures used for the Alberta trial (see Table 2). The use of PLC resulted in a slightly accelerated set for the three fly ash mixes, but a slight retardation for the mix without fly ash. Laboratory studies generally indicate a significant reduction in the set time for concrete produced with PLC (Thomas et al. 2010a).
Figures 1 and 2 show the strength development of site-cast cylinders from the trials in Alberta and Nova Scotia, respectively. An increase in the one-day strength was observed in the PLC concretes in Alberta, but later-age tests showed no consistent differences between the strength of similar mixes produced with either PC or PLC at either location.
Figure 3 shows the results of salt scaling tests (ASTM C 672) for the trial mixes used in Alberta and Nova Scotia. All the results were satisfactory being well below the typical limits used by transportation agencies in Canada (maximum mass loss in the range of 800 to 1000 g/m2 or 23 to 29 oz/yd2 after 50 cycles). Slightly greater mass losses were observed in the three concretes produced with the PLC-slag blend in the Nova Scotia trial, however, the differences are small, being less than 100 g/m2 (3 oz/yd2).
Results from rapid chloride permeability tests (ASTM C 1202) are shown in Figure 4. As expected, the addition of fly ash reduced the charge passed in all cases. No consistent trend was observed with PLC for the results from Alberta, but all three concretes produced with the PLC-slag blend in Nova Scotia showed lower amounts of charge passed than comparable mixes with the PC-slag blend. The control mixes without fly ash for both trials produced slightly unexpected “permeability” results. Values in the range of 2000 Coulombs for the PC and PLC mixes without fly ash are lower than expected for concrete that does not contain any supplementary cementing material and W/CM = 0.42. Conversely, values in the range of 3500 to 4200 Coulombs are high for concrete containing blended cement with 15% slag when tested at 100 days.
CONCLUSIONS
These trials demonstrate that concrete produced with PLC with up to 12% limestone interground with the clinker provides the same level of performance as concrete produced with PC with approximately 3.5% limestone. In addition, the performance of fly ash concrete is not adversely affected by using PLC instead of PC.
The data presented from the trials in Alberta and Nova Scotia together with the previously published data from the trial in Quebec demonstrate that very substantial reductions in the amount of cement clinker used in concrete (see Table 2) can be achieved through the combined use of PLC and SCM without necessarily jeopardizing concrete performance. Using PLC with relatively high amounts of SCM (such as 40% or as 50% used in Quebec trial) can result in reductions in the clinker content of paving mixes in the range of 145 to 175 kg/m3 (240 to 290 lb/yd3). Since the manufacture of portland cement clinker results in approximately 1 kg of CO2 for every 1 kg of clinker1, the reductions in the clinker content of the concrete can be translated directly into reductions in the CO2 footprint of the concrete. This translates to more than a 1-metric ton (1.1 short ton) reduction of CO2 for every 8-m3 (10-yd3) truckload of concrete.
Of course, it is not always possible to use such high levels of SCM. The slower strength gain of many SCMs will limit the amount that can be used in many applications where the speed of construction is critical, especially in cold weather. Also, concerns over deicer-salt scaling in hand-finished flatwork such as sidewalks and scaling may also limit the amount of SCM that is used. However, even when the SCM content is limited to 15% to 25%, the combined use of PLC and SCM can realize clinker savings in the range of 75 to 120 kg/m3 (125 to 200 lb/yd3) in paving mixes.
The blended PLC-slag cement (Type GULb) used in Nova Scotia gives equivalent performance to the PC-slag blend (Type GUb) which, in turn, performs in a similar manner to a straight PC cement (Type GU) when the level of slag is limited to 15%. The clinker content of the Type GULb is just 68% as compared with 91% for a typical Type GU cement. It is feasible that most, if not all, of the Type GU cement (equivalent to ASTM C 150 Type I) used today could be replaced by Type GULb cement produced with interground limestone and a low to moderate level of SCM. This would result in very substantial reductions in the CO2 emissions associated with the concrete industry. Of course, further reductions could be achieved by the partial replacement of the Type GULb by SCM at the concrete plant on a case-by-case basis.
REFERENCES
Hooton, D., Ramezanianpour, A., and Schutz, U. 2010. “Decreasing the Clinker Component in Cementing Materials: Performance of Portland-Limestone Cements in Concrete in Combination with SCMs.” 2010 Concrete Sustainability Conference, NRMCA.
Thomas, M.D.A., Cail, K. Blair, B., Delagrave, A., and Barcelo, L. 2010a. “Equivalent Performance with Half the Clinker Content using PLC and SCM.” 2010 Concrete Sustainability Conference, NRMCA.
Thomas, M.D.A., Hooton, R.D., Cail, K., Smith, B.A. de Wal, J. and Kazanis, K.G. 2010b. “Field Trials of Concretes Produced with Portland Limestone Cement.” Concrete International, January, pp. 35-41.
1 It should be noted that the 1 kg CO2 per 1 kg clinker is an approximation. Approximately, half of this amount results from the calcination of limestone. The remainder results from the combustion of fuel. Consequently, the total amount of CO2 varies depending on the energy efficiency of the cement plant and the fuels used. Modern plants with preheater/precalciner towers will produce less than 1 kg CO2 per 1 kg of clinker.
