rule of thumb is that if the VOI analysis indicates no decision changes will occur, then do not waste money and time in acquiring new information!
The following key questions need to be addressed when undertaking a VOI analysis:
• Howmuch does the information cost, in terms of both acquisition, analysis and delay to development?
• Howreliable is the information?Will the measurement fail? Is there a possibility of false results (imperfect information)?
• Howuseful is the information: howsignificant is the parameter(s) to be measured and what difference will the information make?
Turning to how VOI works, Pete presented a detailed case study in which he demonstrated the use of an influence diagram – “a useful thinking tool to assemble the ‘components’ of the problem” – and a decision tree, in determining the answer to the questions ‘Should an appraisal well be drilled in the North Extension?’ and ‘Should the North Extension be developed?’. He explained that a new user of the VOI tool took less than two hours to learn the software and complete the analysis.
“Invest two hours and get a Decision Analyst to show you how to undertake a VOI calculation,” he advised, concluding with the slide opposite, which must surely be the first time Marilyn Monroe has featured in an SPE presentation!
Franck Monmont, Principal Research Scientist at Schlumberger Gould Research Centre (SGR), followed with ‘In-situ combustion: a workflow from lab experiment to detailed numerical simulation’.
One of the most economically-attractive processes currently in use to recover heavy oil, in-situ combustion (ISC) is also one of the most physically complex techniques. Until now it has not been widely implemented due to a mixed case history of successful field development and perceived operational risks – a key factor in this lack of success being that the fundamental reaction mechanisms cannot be completely elucidated with today’s analytical tools. Also, empirical reaction models developed to date have failed to capture the governing reactions and measure their apparent rates appropriately, and there has been difficulty in evaluating the applicability of in-situ combustion from the available experimental data, especially when combined with existing reservoir modelling tools.
Franck presented work that SGR has developed to resolve these issues. He defined ISC as: “A recovery process in which air is injected in the reservoir where it reacts with the fuel. The heat generated is then used to recover the unburned crude. The ‘fuel’ is not directly the crude oil in the reservoir, but is a carbon-rich residue resulting from thermal cracking and distillation of the ‘residual’ crude oil near the combustion front.”
To date, more than 45 commercial ISC projects have been implemented and more than 200 ISC pilot projects have been tested. Current ongoing projects/pilots include:
• Petrom/OMV Suplacu de Barcau Project, Romania (~8000 bbl/day) - heavy oil (2000-13000 cP), low reservoir pressure (7-10 bar)
• ONGC Santhal, Balol and Lanwa Projects, India (~15000 bbl/day) - medium oil (50-1000 cP), high reservoir pressure (~100 bar)
• Continental Resources North/South Dakota Projects, USA (~6000 bbl/day) - light oil, carbonate, high reservoir temperature & pressure.
Franck described the experimental workflow from oil sample to detailed reaction modelling. This includes phase behaviour characterisation of the oil, pseudo-reaction model characterisation, combustion tube (CT) experiments, and is outlined in the diagram below.
He then discussed the challenges involved in detailed reaction modelling of ISC. The aim of the exercise is to achieve: detailed simulation of multi-phase flow in porous media; detailed reaction model for heavy oil combustion; modelling of the ISC process at lab scale in 1D and 2D, and to extend to reservoir scale with adaptive meshing and upscaling. However, modelling ISC is difficult due to the ‘multi-scales’ (length scale and time scale) problem, which necessitates a trade-off between spatial accuracy and temporal stability.
Franck presented two solutions developed by SGR:
1. Develop an algorithm for the modelling of ISC that: • Achieves a better balance between stability and accuracy • Resolves the combustion front • Decouples pressure equation • Splits advection, reaction, diffusion operators • Uses specialised methods for sub-systems.
2. Adaptive mesh refinement (AMR) injection of hot (800K) methane gas into a cold (344K) methane-butane-nonane liquid reservoir.
In summary, these solutions have enabled Franck and his colleagues to develop a front-capturing ISC simulator with detailed reaction modelling featuring high resolution numeric and adaptive mesh refinement for mixed parabolic/hyperbolic systems, which has performed a series of detailed combustion simulations which provide a better understanding of the stability of the front.
There are, however, outstanding issues, and Franck concluded by outlining some of the “missing pieces of the puzzle”. He explained that: as yet, there is no experimental model for very high pressure combustion; the workflow has been tested on one oil only, therefore more experimental data is required, and there is no ‘systematic’ path from detailed reaction experiments to reservoir scale computations. There is also a challenge to generate flame front correlations with the detailed simulator reservoir simulator. These are elements that he and his colleagues will be focusing on going forward.
You can download both SPE London November presentations from the London Events section at: www.spe-uk.org
The meeting was kindly sponsored by Schlumberger. 09
| Page 2
| Page 3
| Page 4
| Page 5
| Page 6
| Page 7
| Page 8
| Page 9
| Page 10
| Page 11
| Page 12
| Page 13
| Page 14
| Page 15
| Page 16