Abstract

A series of simulation runs using a compositional simulator were conducted to determine the area of influence that results from injecting acid gas (H2S) for 30 years at 330 E3m3/d into a 75 meter thick high permeability sandstone aquifer. In this study, the three-phase (gas-oil-water) equilibrium calculation was used. Initially, a 5 Township by 5 Range grid model was used to determine the area of influence; subsequently, a smaller grid model 1 Township by 1 Range was used.

Boundary producers were required to properly simulate an infinite aquifer. This study found that buoyancy and relative permeability, more than other parameters such as the perforation interval and the gas-water capillary pressure, affected the area of influence. A refined 75-layer model, onemeter thick, was constructed to check the gravity effect. The results of this refined layer model agree closely with a 13-layer model after 11 years of acid gas injection. A blow down simulation was also conducted to determine the time to significant water inflow from perforations at the bottom of the zone

.

The effect of a 50 year shut-in after 30 years of injection was also simulated.

Introduction

The objective of the study was to understand the reservoir volume and the area of influence that would result from the injection of H2S into the Basal Sand aquifer of the Cambrian age Deadwood formation. This was accomplished through the use of a compositional simulator.

The Husky Lloydminster Upgrader (HLU) produces large amounts of acid gas, mainly H2S. The existing sulphur plant removes the sulphur from the H2S. Implementation of H2S disposal into the Deadwood formation would initially reduce the load on the existing HLU sulfur plant, and ultimately it might be shut-in. Any future expansion of the HLU would not require an additional sulphur plant if acid gas disposal was proven to be effective.

For safety reasons and regulatory requirements it was required to know where the injected H2S would migrate to within the aquifer over time. The resulting area of influence would determine the penetrating wells and the mineral rights owners that might ultimately be affected by H2S injection. In this study a commercial compositional simulator was used as the model can take H2S water solubility into account. In the simulation results the area of influence is defined as the area that has an H2S concentration greater than 0.00001 mole fraction. This value is based on the numerical accuracy of the simulation.

The selected simulator was developed for oil-gas-water three phase and multi-components usage. To simulate an H2SWater system a small amount of methane (0.0005 gas saturation) was used as an oil phase for initialization. To distinguish the initial methane gas from H2S in the gas phase all the gas saturation graphs used a minimum scale of 0.0005.

Deadwood Aquifer Formation and Reservoir Model

References 1 and 2 provided a detailed geological appraisal and a mineralogy study of the Cambrian Deadwood formation Basal Sand. In this study it was assumed that there were no chemical reaction between reservoir rock, reservoir brine, and H2S. It is assumed that H2S is soluble in water.

This content is only available via PDF.
You can access this article if you purchase or spend a download.