Stabilised/Solidified Product Preparation

The main variables affecting the performance and properties of the S/S products were expected to be the APC residue/binder ratio and the water content (water/solids or w/s ratio).

Samples were prepared with CEM I or GGBS additions of 0, 10, 20 and 50 wt. % of total dry mass and w/s ratios between 0.35 and 1.0. The ranges for binder addition and w/s ratios were based on industrial applicability considerations for commercial viability (i.e. cost of binder) and workability respectively, as well as performance of the S/S matrices. CEM I or another alkali activator was not used with GGBS because the free lime and alkalis in the APC residues can activate the GGBS and initiate pozzolanic hydration reactions. It is also noted

100 that zero binder addition aims to investigate s/s of APC residues by adding only water, based on the study by Todorovic et al (2003).

The binder was thoroughly mixed with the APC residues prior to water addition using a 5L capacity mortar mixer (ELE, UK). Distilled water was then added to achieve the desired water/solids ratio. Mixing for 3 minutes produced a homogenous paste that was formed into 50mm cube samples using a vibrating table (Controls, Italy) to remove air voids from the mix.

Specimens were de-moulded after 24 hours and transferred to polyethylene re-sealable plastic bags containing a damp tissue to maintain a high humidity environment. This also minimised carbonation that could alter the properties and leaching characteristics of the S/S wastes.

Specimens were prepared and cured in a laboratory environment at room temperature (20-24

oC) and pressure.

The experimental schedule is presented in Table 4.2. No binder addition refers to the investigation of S/S of APC residues by using only water. The experimental schedule was compiled based on preliminary experiments on a laboratory scale that aimed to obtain an indication of the mixing behaviour of APC residues, coupled with a review of existing data from mixes investigated in previous studies. The literature review determined that mixes with waste additions greater than 60 wt.% are not commonly observed and therefore data are relatively thin. Formulations with low binder addition (i.e. 10 and 20 wt.%) were selected not only for evaluating the performance of low binder S/S mixes, but also with the economics of a commercial process in mind. An assessment of the economics of the S/S process for APC residues, however, was outside the scope of the study. Finally, the ranges for the w/s ratios selected were based on preliminary experiments, where water additions resulting in mixes that would not set in 24 hours or that could not be cast were not considered further.

101 4.4 Stabilised/Solidified Product Testing

4.4.1 Physical Tests

S/S products were tested for physical parameters that may determine a successful process on a commercial scale taking into account workability and the structural integrity of the S/S monolithic matrix. Physical tests conducted on S/S APC residues include:

i. Consistence (based on BS EN 4551). Consistence was determined using a flow table with a disc diameter of 255 mm and a 100 mm diameter conical mould. The test involves removing the conical mould from the mixed slurry sample, applying 15 rapid vertical displacements (jolts) to the disc and measuring the diameter of the spread of the sample.

ii. Setting time (BS EN 196-3:2005). Initial and final setting times were determined using a manual Vicat apparatus. This test involves determination of the time at which a needle of a specific weight and shape penetrates a cement paste to a given depth (initial setting time) or when it leaves no mark on the paste’s surface (final setting time).

iii. Bulk density. Bulk density was calculated using the equation:

Bulk density (BD in g/cm³) =

s s

V m

Where ms is the mass (g) of the cube specimen determined by weighing on electronic balances (Sartorius and Oertling) and Vs is the volume (cm³) calculated by measuring the specimen dimensions using Vernier callipers.

iv. Water content. Water content was determined by drying samples at 60^oC to constant weight after immersing samples in acetone to prevent further hydration, and the results were calculated on a wet mass basis.

v. Specific gravity. The specific gravity of samples was determined using a He gas volume expansion meter (Robertson Research, ASTM D5550-94).

vi. Unconfined compressive strength. S/S products were tested in triplicate at 7 and 28 days for unconfined compressive strength (UCS, Automax 5) using a loading rate of 300 N/s. Water-saturated 28-day UCS data was obtained by curing samples for 21 days as described above and then immersing them in water for 7 days before measurement of UCS.

102 vii. Porosity. Specimen porosity (p) was calculated from the bulk density, water content

and specific gravity using the equation (Stegemann et al, 1991):

Porosity (p) = 

where BD = bulk density (g/cm³), WC = water content (w/ww), SG = specific gravity (dimensionless) and d = density of water (1 g/cm³).

The effect of the main experimental variables (i.e. waste-to-binder and water-to-solids ratios) on the physical properties of the S/S matrices was further evaluated via statistical techniques such as hypothesis tests and analysis of variance (ANOVA) using commercial software STATA version 9.2.

4.4.2 Acid Neutralisation Capacity (ANC)

The acid neutralisation capacity of the S/S products was determined using a three-point ANC test at, 1 and 2 meq/g and no acid addition using 1M HNO3. Samples cured for 28 days were dried at 60^oC, crushed using a mortar and pestle and sieved to <150µm. The crushed sample was then placed in 100ml plastic bottles and distilled water and HNO3 were added to give a liquid-to-solid ratio of 10 and the desired acid addition. The bottles were sealed and rotated end over end for 48 hours. Leachate was extracted, centrifuged at 10,500 rpm for 10 minutes (Sorvall RC6 centrifuge) and the pH determined.

Part of the leachate from no acid addition (0 meq/g) samples was filtered through a 0.45µm cellulose nitrate membrane filters (Whatman International Ltd.) and analysed for chloride using the argentometric method (APHA, 2005). This involves titrations using silver nitrate (AgNO3) as the titrant and potassium chromate (K2CrO4) as indicator. It should be noted that the sample without acid addition was leached at the same liquid-to-solid ratio as is applied in the regulatory granular leaching test, BS EN12457-3. Since a smaller particle size and longer contact time were applied, the leaching results for no acid addition can be considered to be a conservative estimate of the results of BS EN12457-3.

103 4.4.3 Tank Leaching Test (NEN 7375:2004)

The monolithic (tank) leaching test (NEN 7345:2004) is a dynamic test for assessing diffusion-controlled leaching and has been used as the basis for developing the UK monolithic Waste Acceptance Criteria (monWAC) (Hall et al, 2005). The test involves leaching of monolithic specimens in sealed containers, using distilled water as the leachant to assess surface area related release. The leachant is renewed at 8 different times (fractions) over a period of 64 days and results are expressed in terms of emission of mass per unit surface area (mg/m²).

Products that had a UCS after water immersion > 1MPa were tested for diffusion-controlled leaching from the monolith. The mixes subjected to the tank test are shown in Table 4.3.

Table 4.3 Mixes subjected to the monolithic (tank) leaching test Binder Binder Addition where VL is the volume of distilled water and VS is the volume of the specimen, with renewal of the leachant at 0.25, 1, 2.25, 4, 9, 16, 36 and 64 days. The long duration of the tank test coupled with time constraints for the experimental study allowed for only replicate to be tested for each mix.

At each fraction the pH was measured using a pH meter (accuracy ±0.05 pH units), calibrated using standard buffer solutions before each measurement. An aliquot of leachate was filtered through 0.45µm cellulose nitrate membrane filters (Whatman International Ltd.) for chloride analysis by the argentometric method which involves titration using AgNO3. Liquid samples for metal analysis were preserved and matrix-matched to the ICP calibration by adding 1ml of concentrated HCl (ARISTAR Grade, VWR UK) to 9ml of sample. These samples were stored in capped, 12 ml polystyrene test tubes, prior to analysis. Metal analysis was conducted by via Inductively Couple Plasma – Optical Emission Spectroscopy (Optima 4300

104 DV - Severn Trent Laboratories, UK ICP-OES). The methods used for analysis are presented in the Appendix II.

It is noted that samples were tested by ICP-OES for sulphur (S). However, due to the nature of the mixes it is assumed that the major sulphur species in the system is sulphate (SO42-) and reference is made only to SO4^2- hereafter. S emissions were adjusted accordingly for the mass of SO42-. Speciation of sulphur in commercial GGBS for use in concrete may comprise both sulphides and sulphates with maximum permissible contents of 2.0% and 2.5% by mass respectively, according to BS EN 15167-1:2006. A study by Roy (2009) however, has shown that in activated slags it is likely that sulphides will be transformed to sulphates. Sulphur content (as SO3) in the GGBS used in this study was 0.05%.

As with the physical properties, statistical techniques were used to evaluate the effect of the main experimental variables on the leaching characteristics of the S/S matrices subjected to the tank test.