Executive Summary

This report provides fundamental data and analyses needed to evaluate possible detonation hazards that may result from flammable gases within the waste storage tanks located at Hanford, WA. The emphasis is on the measurement and correlation of detonation cell widths. Cell-width information can be used through generally accepted correlations [Lee (1984)] to predict more direct indicators of detonation hazards, such as critical-initiation energy or the critical-tube diameter for transmission of a detonation to an unconfined space. By measuring cell widths in some representative gas mixtures, a basis for one or more correlations is made. The cell-width data can be correlated to some other length scale, for instance the reaction-zone thickness, which can be computed directly.

Experiments have been performed in the GALCIT detonation tube for the following mixtures:

• Stoichiometric hydrogen and nitrous oxide diluted by nitrogen and air.
• Stoichiometric methane and oxygen diluted by nitrogen.
• Stoichiometric methane and nitrous oxide diluted by nitrogen and air.
• Stoichiometric ammonia and oxygen diluted by nitrogen.
• Stoichiometric ammonia and nitrous oxide diluted by nitrogen and air.
• Various model tank mixtures for tanks SY-101, AW-101, AN-105, AN-104, AN-103, and A-101 diluted by air.

Detonation velocities have also been measured and found in good agreement with equilibrium thermochemical calculations. Our measured cell widths agree with data from the literature where available, but most of the mixtures we have examined have not been studied before. The oxy-acetylene driver has been studied and found to be capable of providing 10-120 kJ of initiation energy in a repeatable fashion.

Reaction zone calculations have the advantage of being generally faster and cheaper than experiments and also of being capable of a larger range of conditions and mixtures. However, a number of difficulties prevent the calculations from being straightforward. The first problem is the lack of a comprehensive reaction mechanism for the most general mixtures. In an effort to find or create such a mechanism, we have collected several mechanisms from the literature, and a large amount of experimental data for validation.

The most successful mechanism for the model tank mixtures found so far is a modified version of the the mechanism of [Miller and Bowman (1989)], although it is not as successful at methane oxidation as the GRI-Mech 2.11 [Frenklach et al. (1995)], which can not be used for ammonia oxidation. The mechanisms of [Miller et al. (1983)] and [Miller and Bowman (1989)] can be used for ammonia combustion but are not as useful for hydrocarbon combustion.

Two analysis tools are available for performing chemical kinetics calculations under constant-volume conditions or during steady, one-dimensional, compressible flow behind a shock. The constant-volume calculations are used for validation comparisons with shock tube induction time data, and the one-dimensional dynamical calculations are used to compute the reaction zone thickness in idealized planar detonation waves.

Using the experimental data mentioned above and reaction-zone thickness calculations performed with appropriate mechanisms, cell-width correlations have been created for several mixtures. For limited conditions involving fixed fuel-oxidizer stoichiometry, with variations in initial pressure or dilution, a power law correlation between cell size and reaction zone thickness appears to be very useful. A more general correlation applicable to various fuel-oxidizer systems is more elusive but currently under development.