30    OPTIMUM STRATEGIES FOR TOXIC FLUID WASTE
INJECTION IN DEEP RESERVOIRS:
REACTION-TRANSPORT MODELING
Yueting Chen, Graduate Student Research Assistant
Anthony Park, Research Associate
Jun Mu, Research Associate
Peter Ortoleva, Professor
Indiana University
Department of Chemistry
Bloomington, Indiana 4740S
INTRODUCTION
Designing the optimal strategy for toxic fluid waste injection is extremely difficult in reservoirs with
complex mineralogy. Improper injection scenarios can result in permeability/porosity-destroying side
effects such as gel precipitation, formation collapse or fine migration-induced clogging. Here we
present results of a fully coupled reaction-transport simulator on the prediction of reservoir response
to variation of the composition and scenario of the injection.
As an example, we focus on the injection of acid waste into carbonate cemented sandstone. We
investigate the degree to which Fe3*-gels and excessive stress-supporting matrix dissolution can be
minimized by finding the optimal composition and injection rate of the acid.
The basis of our analysis is the reaction-transport code C1RF (chemical interaction of rock and
fluid).1,2 CIRF accounts for the following processes:
1. finite rate mineral dissolution/precipitation reactions,
2. aqueous pore fluid reaction equilibria,
3. Darcy flow with a permeability that reflects changes in grain size and shape due to
precipitation/dissolution reactions,
4. nucleation threshold for precipitating new phases,
5. conservation of pore fluid solute mass due to flow, dispersion/diffusion and aqueous
phase and mineral reactions,
6. temperature calculated via the conservation of energy equation,
7. solute species activities corrected for temperature, pressure and ionic strength,
8. evolution in one or fully two dimensions or special symmetric configurations in three
dimensions.
The program comes with a built-in database for the thermodynamic and kinetic parameters.
Graphical and other model-building and -analysis tools are included so that the reaction-transport
simulator can conveniently be used in a research or applied environment.
The input to CIRF is the spatial distribution of the grain size and number of grains per rock volume
within the reservoir for each mineral in the domain to be simulated and the time-course of the
injection fluid composition, temperature and flow rate. The output is the time-course of the spatial
distribution of porosity, permeability, mineral grain size, and volume fraction and pore fluid composition within the reservoir.
EXAMPLE SIMULATIONS: CARBONATE CEMENTED SANDSTONE RESERVOIRS
As an illustrative example, we consider a carbonate cemented sandstone reservoir. The initial state
of the reservoir is given in Table I. The network of reaction considered is given in Table II.
A typical simulation of injection is seen in Figure 1. As injection proceeds the spatial distribution of
porosity, permeability and mineralization change from their initially uniform state. Of particular
48th Purdue Industrial Waste Conference Proceedings, 1993 Lewis Publishers, Chelsea, Michigan 48118. Printed in
U.S.A.
305