Performance Characteristics for mortars containing MSWI ashes

Figure 6 shows the flow times of mortars M0, M5, M10, M15 and M20 measured with the LCPC mortar "workability meter". For the reference mortar M0, 3 tests were carried out to assess the dispersal of the measurements. The standard deviation for these measures was 0.75 s. Given the low value of the standard deviation, for the other mortars, one single measurement was performed. We can see that the flow time recorded increases as the ash content increases. It is, however, noteworthy that no problems were encountered with placing the mortars in molds during the preparation of the test specimens.

Figure 6: Flow times of the M0, M5, M10, M15 and M20 mortars

Figure 7 shows the initial and final setting times for the M0 to M20 mortars. Incorporating ash in the mortars considerably increases the initial and final setting times. This increase is dramatic when the ash content increases from 10 % to 15 %. The MSWI fly ash components responsible for delayed setting were not accurately identified. However, a number of studies have already shown that zinc and lead are setting retarders, or may even inhibit setting [16]. These two components are the most abundant heavy metals found in the MSWI ash used. This almost certainly explains the slow setting times observed.

Figure 7: Initial and final setting times for the M0, M5, M10, M15 and M20 mortars

Figure 8 shows compressive strengths for the M0, M5, M10, M15 and M20 mortars after (7, 28, 90 and 565) d. In the figure, the error bars represent plus or minus one standard deviation. Compressive strengths after 28 d of the M10, M'10 and M'0 mortars are shown in Table 7. From Figure 8, it would seem that, in general, the mortar compressive strength is optimized for an MSWI ash content of approximately 10 % (a lower strength was observed after 28 d, but similar tests carried out with other MSWI ashes did not exhibit this particular feature [17]). Strength increases when ash content varies from 0 % to 10 %, and falls thereafter. After 90 d, the M10 mortar strength is approximately 15 % higher than the M0 control mortar. On the other hand, the strength of the M20 mortar is approximately 5 % lower. The strength of mortars containing MSWI ash increases with time up to 90 d, but the increases in strength due to the presence of ash are obtained generally at a much earlier stage (after 7 d). Strengths obtained after 565 d are lower (M5 mortar presents a decrease in strength of about 10 %). However, it is noteworthy that no visible signs of an increase in cracking were observed in these mortars, between 90 d and 565 d.

Figure 8: Compressive strengths of the M0, M5, M10, M15 and M20 mortars after 7, 28, 90 and 565 d

Table 7 shows that incorporating washed ash (in M'10 mortar) in place of total ash (M10 mortar) results in a decrease in strength of approximately 20 %. On the other hand however, the incorporation of ash wash water (M'0 mortar) results in an increase in strength of 35 % in relation to the M10 mortar.

Incorporating MSWI fly ash in mortars (below 15 % with an optimum for 10 % ash) provides increased mechanical strength in the medium term (between 28 d and 90 d). Several assumptions may be put forward to explain these results.

Firstly, the increase in strength after 7 d due to the presence of ash is most likely not attributable to the potential pozzolanic properties of this waste material. Indeed, pozzolanic reactions are slow and their influence is only evident after longer periods (usually between 28 d and 90 d). In the cases observed, gains in strength are important after 7 d and the presence of the ash causes a decrease in mortar strength in the long term (565 d). The pozzolanic activity of the MSWI fly ash is not necessarily zero but is in any case very low and does not account for observable increases in strength.

The tests carried out with wash water (M'0 mortar) show that the soluble fraction of the ash allows a considerable increase in compressive strength. The presence of considerable quantities of chlorides in the ash wash water is certainly the cause of this increase in strength. Indeed, chlorides are well-known setting accelerators [18]. Elsewhere, the compressive strength of mortar containing 10 % washed ash is lower than that of the M10 mortar. Our research has not been conclusive in explaining categorically the influence of the non-soluble fraction of ash on reductions in compressive strength. One of the explanations we can nevertheless suggest is the following: the MSWI ash particles could be weak points in mortars. Kessler et al. [15] have shown, based on an SEM analysis of cement pastes containing MSWI the fly ash, that the cement paste / fly ash bond appears to be relatively weak.

Competition between the positive effect of chlorides and the negative effect of the non-soluble fraction could explain the optimal strength obtained for an ash content equivalent to 10 %. For ash contents of over 10 %, the negative effect of the non-soluble fraction could become predominant and cause the observed decrease in mechanical strength.

Given the reduced workability of mortars incorporating the MSWI fly ash, one could also impute the increase in strength to absorption of water by the ashes during mixing. We have shown [18] that the higher the ash content, the higher the water demand of mortars containing these MSWI ashes. According to these tests, the MSWI fly ash "absorbs" approximately 28 % of its mass in water. This water absorption then diminishes the effective W/C ratio of the cement paste, thus increasing its strength. The effective W/C ratio of the cement paste in M10 mortar (respectively M20) would in this case be 0.47 (respectively 0.44). However, although this mechanism certainly tends towards increasing strength, it does not counteract the unfavourable effect of the non-soluble fraction of the ash. Finally, the decrease in strength observed after 565 d for mortars containing MSWI fly ash could not be explained. Certain MSWI ash components are aggressive species for cement (sulphates and alkalis in particular) and may provoke degradation of the mortars. However, it is surprising that this decrease in strength should occur in the long term (after 565 d), under conditions of non-aggressive storage (20 ºC, in sealed bags).

Figures 9 and 10 present respectively the size and the mass variations of mortars M0, M10 and M20 determined over a period of approximately 5 weeks after removing the molds. One can see in Figure 9 that the shrinkages of the two mortars containing ash (M10 and M20) are practically identical. Up to 7 d, shrinkage of the two mortars containing ash is approximately twice that of the control mortar M0. Beyond that, their respective evolution is comparable. In Figure 10, one can see that mass variations of M10 and M20 mortars are lower than those of the control mortar M0.

In consideration of the small size of the test pieces used, thermal shrinkage of the mortars may be ignored. The shrinkage measured is thus due, on the one hand, to the evacuation of water from the test pieces and, on the other, to the autogenous shrinkage caused by Le Chatelier's contraction principle. The lower shrinkage values of M0 mortar cannot be explained by drying shrinkage. Indeed, mortars containing ash lose less water than the M0 mortar (Figure 10). The increase in shrinkage of these two mortars would therefore seem to be due to an increase in autogenous shrinkage.

The lower water loss of these mortars in relation to the M0 mortar can be explained by the immobilization of water due to hydration product formation. Thus, the high chloride and sulphate content of the MSWI fly ash may result in the formation of high quantities of ettringite and chloroaluminates (hydrates which mobilise a large number of water molecules). The early formation of these two hydrates could explain the high shrinkage of mortars containing fly ash observed in early stages.