What You Should Know About Evaluating Simulation Results-Part 2

Introduction

This is the second of a two part article on assessing simulation results. Evaluation engineers are often provided with a numerical simulation study as part of the technical support for an economic evaluation of an oil and gas property. Simulation involves many specialized techniques which many people may not be familiar with. A procedure is outlined, in this article, which will aid in assessing the applicability of simulation results to an economic evaluation.

Part 1, in the May issue of lCPT, contained the Introduction, discussed Consistency Checks and Identifying the Critical Issues. Model design was the last topic, where implementations for waterflooding, multiple (gas-oil, water-oil, or gas-water) contacts, mature production, retrograde condensation and miscible floods were discussed. In Part 2, the discussion continues with Evaluating the Simulation Technique, Report Review followed by Conclusions.

Evaluating the Simulation Technique

Check that Grid can Physically Model Reservoir Flow

Reducing the number of grid blocks can be taken too far. The following example occurred in a study proposal which was submitted to the author. The objective was to determine horizontal well potential in a tight sandstone reservoir. Production in vertical wells was uneconomic due to low rates and GOR penalties. The study was to determine if a horizontal well would increase productivity enough to enable economic production and if GORs would be low enough to avoid penalties.

Flow nets are a very useful tool to help mentally picture the flow in the reservoir. A flow pattern something like that shown in Figure 6 should exist. Since the reservoir is tight (less than 1 mD) localized gas saturations would exist. A fine grid made with small grid blocks and thin layers would be required around the well. An element of symmetry could be utilized through the centre of the well to reduce computation.

The proposal, for the situation described above, recommended three layers with the wellbore centred in blocks with 100 foot horizontal x and y dimensions. It is important to realize that the simulator would still have run. Although these large grid blocks could not correctly represent the physics in the reservoir, the model would stilI produce results. The proposal was not considered favourably by the author.

Many wells have been hydraulically fractured. Usually models are run and minor adjustments made via "pseudo well relative permeability curves." Wells that have been fracced will show slower GOR increases than unfracced wells. This will not always produce correct results if the combined propped half length of the fractures is greater than 25% of the interwell distance. Under these conditions, the deviation from a radial flow pattern is too severe to be represented radially in conjunction with pseudo well curves.

Studies by the author have shown that waterflood response in a reservoir that has large fracs is not as pronounced as predicted by radial flow models. The approach to such a simulation had to be changed-a type well study with the fracture modelled directly was more appropriate.