Stabilisation of sulphide rich soil: problems and solutions

Abstract

The research presented in this thesis can be split into two well-defined areas. Firstly, the effects of the addition of lime (either 2%, 4%, 6% or 8%Ca(OH)2 or 1.5%, 3%, 4.5% or 6%CaO) on various engineering properties of a sulphide (pyrites) rich, naturally occurring clay - Lower Oxford Clay (LOC) - were studied. Experimental tests representing, or relating to, tests outlined in the Design and Construction of Lime Stabilised Capping Specification (DoT, 1995) were utilised and consisted of Atterberg (consistency) tests, compaction tests, unconfined compressive strength tests, linear expansion measurements and durability tests. Also, sulphate (S03) analyses, pH measurements, differential thermogravinietric (DTG) analyses and X-ray diffraction (XRD) tests were employed to confirm experimental data. Environmental conditions were chosen to reflect temperatures experienced in the UK - between 5*C, 10'C and 201C. The majority of samples were mellowed (the time delay between initial mixing and compaction) for 3 days at 201C prior to testing. The results show that small additions of lime (i. e. 2%Ca(OH)2 or 1.5%CaO) modify the engineering properties of LOC by causing flocculation, and also increase S03 levels due to increased gypsum levels via pyrites oxidation. However, strength development is poor as little cementitious bonding is formed. Mellowing further modifies the engineering properties as the increased time allows greater flocculation leading to a less dense, more porous material. Also ettringite forms during mellowing allowing expansive reactions to occur prior to compaction, consuming both calcium and sulphate in the process and leading to slightly reduced pH levels compared to similar unmellowed samples. Further additions of lime give further change to the engineering properties of the LOC, up to an addition of 6%Ca(OH)2 or 4.5%CaO. Further additions result only in supplying an excess of calcium. Strength development is improved with increasing lime additions, especially after 4 weeks of curing at higher curing temperatures (20'C); however linear expansion increases and durability worsens. However, a period of mellowing prior to compaction reduces linear expansion and improves durability. Sulphate (S03) levels are also increased with increasing additions of lime due to pyrites oxidation. In samples containing >2%Ca(OH)2 or 1.5%CaO, pH levels are maintained to a level where ettringite formation is sustained rather than gypsum formation. Secondly, the effects of a secondary stabilising agent (ground granulated blast furnace slag - GGBS - incorporated at various percentage additions (2%, 4%, 6% or 8%)) on the strength development, linear expansion and durability of mellowed, lime-modified LOC is presented. For comparative purposes, Portland cement (PC) is also used. The results show that although LOC-lime-PC samples exhibit greater strengths (especially when cured for long periods at 20"C), linear expansion is lower and durability is higher in similar LOClime- GGBS samples. As the hydration process of PC produces lime, and as lime is usually consumed during GGBS hydration, then samples containing the former stabiliser will maintain pH levels well above samples containing the latter. Therefore, on soaking disruption and expansion due to the formation of ettringite (which is unstable at pH levels <10.5) is common to LOC-lime-PC samples but not in similar GGBS samples. In conclusion, the findings from this thesis indicate that soil stabilisation with lime and GGBS is particularly effective for naturally occurring sulphide rich clay soils (such as the LOC) and is an environmentally friendly alternative to PC-stabilised soils.