HEAT OF HYDRATION OF SALTSTONE MIXES-MEASUREMENT BY ISOTHERMAL CALORIMETRY

Abstract

This report provides initial results on the measurement of heat of hydration of Saltstone mixes using isothermal calorimetry. The results were obtained using a recently purchased TAM Air Model 3116 Isothermal Conduction Calorimeter. Heat of hydration is an important property of Saltstone mixes. Greater amounts of heat will increase the temperature of the curing mix in the vaults and limit the processing rate. The heat of hydration also reflects the extent of the hydraulic reactions that turn the fluid mixture into a ''stone like'' solid and consequently impacts performance properties such as permeability. Determining which factors control these reactions, as monitored by the heat of hydration, is an important goal of the variability study. Experiments with mixes of portland cement in water demonstrated that the heats measured by this technique over a seven day period match very well with the literature values of (1) seven day heats of hydration using the standard test method for heat of hydration of hydraulic cement, ASTM C 186-05 and (2) heats of hydration measured using isothermal calorimetry. The heats of hydration of portland cement or blast furnace slag in a Modular Caustic Side Solvent Extraction Unit (MCU) simulant revealed that if the cure temperature is maintained at 25 C, the amount of heat released over a seven day period is roughly 62% less than the heat released by portland cement in water. Furthermore, both the blast furnace slag and the portland cement were found to be equivalent in heat production over the seven day period in MCU. This equivalency is due to the activation of the slag by the greater than 1 Molar free hydroxide ion concentration in the simulant. Results using premix (a blend of 10% cement, 45% blast furnace slag, and 45% fly ash) in MCU, Deliquification, Dissolution and Adjustment (DDA) and Salt Waste Processing Facility (SWPF) simulants reveal that the fly ash had not significantly reacted (undergone hydration reactions) after seven days (most likely less than 5%). There were clear differences in the amount of heat released and the peak times of heat release for the three different simulants. It turns out that SWPF simulant mixes give off greater heat than does MCU and DDA simulant mixes. The temperature dependence of the heat of hydration was measured by carrying out these measurements at 25, 40 and 55 C. In general, the peak times shifted to shorter times as the isothermal temperature increased and the amount of heat released was independent of temperature for DDA and MCU but slightly higher at higher temperatures for SWPF. The goal of this study is to apply this technique to the measurement of the heat of hydration of mixes that will be made as part of the variability study. It is important to understand which variables will impact (and to what extent) the amount of heat generated and the peak times for the heat release. Those variables that can be controlled can then be tuned to adjust the heat of hydration as long as the other properties are still acceptable. The first application of heat of hydration measurements to the variability study was completed and the results presented in this report. These measurements were made using Phase VI mixes (SWPF simulants) following a statistical design that included variation in the compositional and operational variables. Variation in both the amount of heat released and the peak times for the heat release were observed. The measured ranges were 23 Joules per gram of premix for the heat release and 23 hours for the peak time of heat release at 25 C. Linear models with high R{sup 2} values and no statistical evidence for lack of fit were developed that relate the amount of heat release and the peak time for heat release for the Phase VI mixes to certain variables. The amount of heat released was a function of the aluminate and portland cement concentrations as well as the temperature of mixing. The peak time for heat release was a function of aluminate, portland cement and total nitrate plus nitrite concentrations. A comparison was made of the measured values of heat release by isothermal calorimetry to a previous study of the measurement of the heat of hydration using adiabatic calorimetry by Steimke and Fowler. After 80 hours of reaction time, the two techniques provided heat release results that were roughly in the same range. However, additional experiments at higher isothermal temperatures will be required to see how well the two measurements agree for longer times. This is due to the higher temperatures that are experienced in adiabatic calorimetry ({approx}105 C).