Integrated Analysis
& Optimization Tool

CMOST employs innovative experimental design, sampling and optimization techniques to efficiently determine reservoir and operating parameters defining recovery and production of oil and gas fields.

cmostandreservoir.png
Improve efficiency & productivity with automated/optimized workflows
cmost-SAfinal.png
cmost-HMfinal.png
cmost-PAfinal.png
cmost-NOfinal.png
cmostarrow1final.png
cmost-arrow2final.png
cmost-arrow3final.png
cmost-arrow4final.png
CMOST 2016.10
Highlights include:
  • Robust Optimization
  • CMG Bayesian Engine (Probablistic Forecasting)
  • Rate and group individual experiments
  • Open and navigate text format files
  • is Now Available
    Screen Shot 2017-02-02 at 3.50.52 PM.png
    cmost_watermark_lightwhite.png
    Sensitivity Analysis: Assess Different Parameters
    Confidently identify and assess the relationship between reservoir parameters /objective functions and how they impact the history match, production forecasts and overall recovery.

    Read more...

    • Assess the sensitivity of different parameters and their relative effect on a user defined objective function (e.g. HM error, NPV etc.)
    • Investigate the effect that changes in input parameter values have on output
    • Conduct indepth analysis with: Response Surface methodology, Morris and Sobol variance-based or One parameter at a time (OPAAT)
    • Reduce simulation time by by selectively determining which parameters can be eliminated during HM
    • Analysis tools: histograms, cross-plots, time-series plots, scatter-plots, proxy model validation plots, and tornado plots

    Optimization: Improve Operational and Completion Strategies
    Modern experimental design and optimization algorithms vary dozens to hundreds of parameters simultaneously to find the optimal solution.

    Read more...

    • Easily define reservoir, well and operating variables for design and field development planning
    • Define custom user defined optimization objective function
    • Use an advanced calculation engine: CMG DECE, PSO, DE, Latin Hypercube plus proxy, or random brute force methods
    • Identify an optimum (global optima) development plan or operating conditions with the least number of simulation runs
    • Optimize well and fracture spacing to increase production, NPV and EUR
    • Histograms, cross-plots, time-series plots, scatter-plots, proxy model validation plots, and tornado plots for advanced analysis
    • Applications may include:
      • Increase recovery and reduce chemical costs by optimizing chemical injection scheme and slug sizes in cEOR process
      • Optimize well and hydraulic fracture spacing to increase production, NPV and EUR in unconventional reservoirs
      • Develop and optimize completion design and operational strategies to increase ultimate recovery and reduce steam-oil ratio (SOR) from SAGD operations

    Uncertainty Analysis: Manage Risk and Identify Opportunities
    Quantify and understand the impact reservoir and operational uncertainties will have on project economics, thereby managing risk and identifying opportunities.

    Read more...

    • Quickly run thousands of cases to assess HM variability or field optimization and the associated probabilities
    • Quantify uncertainty to manage risk and identify opportunities
    • Apply powerful statistical methods: Monte Carlo Simulation using Proxy, Monte Carlo using reservoir simulator, Probabilistic Forecast using posterior samples
    • Develop a response surface (RS) for each objective function of interest, with respect to each of the uncertain variables, using simulation results
    • Assess prediction reliability to understand the range and flexibility of potential operating constraints
    • Assess the viability of operational strategies by determining how results vary with uncertainty
    History Matching: Calibrate Models with Field Data
    History matching involves adjusting the simulation model properties to accurately reproduce past reservoir behavior and to simulate future behavior with increased confidence.

    Read more...

    • Automate the history matching process by using advanced statistical methods
    • Advanced calculation engines: CMG DECE, CMG Bayesian Engine, Particle swarm optimization (PSO), Differential Evolution (DE), Latin Hypercube plus proxy, or Random brute force methods
    • Calibrate and History match models with field production and injection data, including temperature logs and 4D seismic
    • Guide and speed-up history matching by applying Uncertainty in measured data and user defined weighting factors to parameters
    • Create and assemble the best matches, in the least amount of runs, using advanced Experimental Design (ED) techniques
    • Easily analyze the reservoir and recovery processes with automated HM

    Robust Optimization: Use Geological Uncertainty to Reduce Risk
    Account for multiple geological realizations, or other uncertain parameters, to identify the optimum results for a field development plan.

    Read more...

    • Rigorously analyze geological uncertainty, using multiple geological realizations, to create a risk-weighted optimization solution for a field development plan
    • Innovative workflow for better informed decisions, leading to a higher probability of success and profitability
    • Increase production and achieve a reservoir's full economic value by implementing the best recovery method
    • Use advanced analytics and graphical output to accurately predict probability of reservoir success with reduced risk

    Probabilistic Forecasting: Drive Informed Field Development Decisions
    Forecast using a range of possible outcomes to quantify risk and confidently make business decisions using CMG's Proxy Based Acceptance-Rejection (PAR) sampling method.

    Read more...

    • Capture uncertainties in forecasts while honouring the observed production data
    • Significantly reduce computational costs and achieve a 10x speed-up in simulation runs using CMG PAR over the MCMC method
    • CMG's PAR defines posterior probability density function by using "misfit" simulation results and measured production data
    • Improve the accuracy of the RBF proxy model in regions of good history matches
    • Use multiple history matched models to generate P10, P50 and P90 forecasts during field application