An Interactive 5-Day Training Course
Modelling & Managing Subsurface Uncertainty
Modelling and Managing Subsurface Uncertainty Properly
Secure your spot now!
Duration
5 Days
Price
$ 3,500
Starting From
11 Jan 2027
Home Geology and Geoscience Modelling & Managing Subsurface Uncertainty
Modelling & Managing Subsurface Uncertainty
Have Questions ?
Modelling & Managing Subsurface Uncertainty - GG-SUNCM-PEA27
| Code | Date | Time | Duration | Location | Currency | Early Bird Fee Per Person |
|---|---|---|---|---|---|---|
| GG-SUNCM-PEA27 | 11 - 15 Jan 2027 | 10 AM CST | 4 Hours Per Day |
Online |
USD |
3500 |
Need this for a group? We deliver the same course in-house — face-to-face at your location or online — tailored to your assets and team level. Contact info@peassociations.com
Boost your team's skills and your budget! Enjoy group discounts for collaborative learning. Send an inquiry to info@peassociations.com.
Modelling & Managing Subsurface Uncertainty
This training covers the modelling and management of subsurface uncertainty across the asset life cycle. It works through uncertainty types and their sources, bias and its recognition, probability and distribution construction, scenario and parameter uncertainty, dependency and correlation, uncertainty propagation to volumes and forecasts, value of information, decision framing under uncertainty and the communication of ranges to decision makers.
Description
Subsurface uncertainty is managed badly in most organisations, and the symptoms are consistent: ranges that turn out to be too narrow, outcomes falling outside the pre-drill P10 to P90 more often than one time in five, reserves revised repeatedly in the same direction, and decisions presented as though the subsurface were known. The causes are partly technical and largely behavioural, and both have to be addressed for uncertainty management to improve.
This training covers both. Uncertainty types and their sources are established first, covering measurement, interpretation, model and scenario uncertainty, and the distinction between what more data can reduce and what it cannot. Bias is then addressed directly, covering anchoring, overconfidence, availability and motivational bias, with the evidence on how much they narrow subsurface ranges and the practical techniques that counter them. Probability and distribution construction follow, covering the derivation of distributions from data, from analogues and from judgement. Scenario against parameter uncertainty is then developed. Dependency and correlation are addressed, since ignoring them is a common and consequential error. Uncertainty propagation to volumes and forecasts follows. Value of information is then developed as the tool that decides whether to acquire more data. Decision framing under uncertainty follows. The training closes with communication.
Subsurface ranges are consistently too narrow and the evidence for this is extensive. Comparisons of pre-drill prediction against outcome across hundreds of wells show that actual results fall outside the stated P10 to P90 interval far more often than ten per cent of the time, commonly thirty to forty per cent. This is overconfidence, it is systematic, and it means that a range quoted as 80 per cent confidence is actually perhaps 50 per cent. Correcting it requires deliberately widening ranges beyond what feels right, which is uncomfortable and necessary.
Correlation between parameters changes the answer substantially. If porosity and net to gross are both treated as independent uncertainties when they are in fact positively correlated, the Monte Carlo result will be too narrow, because the simulation generates combinations of high porosity with low net to gross that do not occur in nature. Conversely, treating correlated parameters as perfectly dependent overstates the range. Specifying the correlations, even approximately, is more accurate than the default assumption of independence.
Value of information is the discipline that connects uncertainty to spending. The question is not whether an uncertainty exists but whether reducing it would change a decision, and if so, whether the reduction is worth its cost. A well test costing ten million that would change the development concept has clear value; one that would refine a number nobody acts on does not, however interesting the result. Running the calculation before committing appraisal spend is routine in some organisations and absent in others, and the difference shows in their appraisal efficiency.
Finally, a range only matters if the decision changes across it. The practical test of an uncertainty assessment is whether the recommended action is the same at P90 as at P10. If it is, the uncertainty is interesting but not decision relevant and the project can proceed. If it is not, the uncertainty must either be reduced or the decision must be structured so that it holds across the range. Stating this explicitly turns an uncertainty study into a decision input rather than a technical deliverable.
By the end of this training, participants will be able to:
- Classify uncertainty by type and source
- Distinguish reducible from irreducible uncertainty
- Recognise and counter bias in subsurface estimation
- Construct probability distributions from data, analogue and judgement
- Distinguish and handle parameter and scenario uncertainty
- Specify dependencies and correlations between parameters
- Propagate uncertainty into volumes and production forecasts
- Apply value of information analysis to appraisal decisions
- Frame decisions so that they hold across the uncertainty range
- Communicate uncertainty in a form decision makers can act on
The training combines calibration exercises with asset case work. Participants make estimates with stated confidence intervals and have their calibration scored against the actual answers, which demonstrates overconfidence directly. Uncertainty assessments are built on real asset data with and without correlation and scenario treatment to compare the resulting ranges. Value of information calculations are run on appraisal options and recommendations defended to a panel.
Organisations sending participants to this training will:
- Produce ranges that contain the eventual outcome
- Reduce systematic bias in subsurface estimates
- Capture scenario uncertainty rather than parameter variation alone
- Direct appraisal spend at decision relevant uncertainty
- Frame decisions so that they hold across the range
- Improve the credibility of subsurface input to decisions
Participants will:
- Recognise bias in their own estimates
- Build distributions with defensible bases
- Handle correlation and dependency correctly
- Run value of information analysis
- Frame uncertainty for decisions
- Present ranges credibly to management
- Subsurface geoscientists and reservoir engineers
- Reservoir modellers and volumetric analysts
- Subsurface team leaders and managers
- Planning and portfolio staff
- Technical assurance and peer review staff
- Exploration and appraisal decision makers
- Commercial staff using subsurface ranges
Module 1 - Uncertainty Types and Sources
- Measurement uncertainty
- Interpretation uncertainty
- Model and conceptual uncertainty
- Scenario uncertainty
- Aleatory and epistemic distinction
- Reducible against irreducible uncertainty
- Uncertainty through the asset life cycle
- Dominant uncertainties by life cycle stage
- Uncertainty inventories
- Ranking uncertainty by impact
- Tracking uncertainty over time
Module 2 - Bias and Its Correction
- Anchoring and adjustment
- Overconfidence and narrow ranges
- Calibration evidence from the industry
- Availability and recency bias
- Confirmation bias in interpretation
- Motivational and organisational bias
- Groupthink in technical teams
- Deliberate range widening
- Reference class and outside view methods
- Red teams and devil's advocate roles
- Structured elicitation techniques
- Measuring team calibration
Module 3 - Probability and Distributions
- Probability interpretation in subsurface use
- P10, P50 and P90 conventions
- Distribution types and their selection
- Normal and lognormal distributions
- Triangular and uniform distributions
- Beta and truncated distributions
- Deriving distributions from data
- Deriving distributions from analogue
- Expert judgement elicitation
- Distribution width and its justification
- Testing distributions by stating extremes
- Documenting distribution bases
Module 4 - Scenario and Dependency
- Scenario against parameter uncertainty
- Constructing discrete scenarios
- Scenario probability assignment
- Combining scenarios and parameters
- Dependency between parameters
- Correlation specification
- Rank correlation in Monte Carlo
- Effect of correlation on range width
- Independence as a default and its errors
- Physical basis for correlations
- Common dependency structures
- Testing correlation assumptions
Module 5 - Propagation to Volumes
- Volumetric equation and its terms
- Monte Carlo simulation of volumes
- Deterministic against probabilistic volumes
- Gross rock volume uncertainty
- Net to gross and porosity uncertainty
- Saturation uncertainty
- Recovery factor uncertainty
- Area and thickness dependency
- Sensitivity identification
- Tornado and spider plots
- Aggregation across segments and fields
- Arithmetic and portfolio aggregation effects
Module 6 - Propagation to Forecasts
- Static uncertainty into dynamic models
- Multiple realisation simulation
- Assisted history matching approaches
- Forecast uncertainty bands
- Production profile ranges
- Plateau length and decline uncertainty
- Facility sizing under uncertainty
- Well count uncertainty
- Economic sensitivity to subsurface ranges
- Reserves categories and probabilistic estimates
- Updating forecasts with production data
Module 7 - Value of Information
- Value of information concept
- Decision dependence of information value
- Decision trees for appraisal options
- Expected value calculation
- Perfect against imperfect information
- Data reliability in the calculation
- Appraisal well value assessment
- Seismic acquisition value assessment
- Core and test data value
- Value of flexibility and option value
- Cases where no data is worth acquiring
- Presenting value of information results
Module 8 - Decisions and Communication
- Framing decisions under uncertainty
- Does the decision change across the range
- Development concepts that hold across the range
- Staged and phased development
- Downside protection and upside capture
- Risk appetite and its expression
- Communicating ranges to non-technical audiences
- Presenting the drivers of the range
- Avoiding false precision
- Documenting assumptions and their falsification
- Uncertainty in assurance and review
- Learning from outcomes against predictions
Upon successful completion of this training course, delegates will be awarded an official Certificate of Completion issued by the Petroleum Engineers Association (PEA), an ISO 9001:2015 certified training organization. The certificate carries 10 Credits and formally records the total learning hours completed.
Each certificate is signed by the Course Facilitator and the CEO of the Petroleum Engineers Association, and serves as verifiable proof of professional training that delegates can present to employers and professional bodies worldwide.
Frequently Asked Questions
All course bookings made through PEA are strictly non-refundable. By registering for a course, you acknowledge and accept that all fees are payable in full and are not subject to refund under any circumstances, including changes in personal or professional commitments or partial attendance.
PEA reserves the right to make reasonable adjustments to course content, trainers, or schedules where necessary, without entitling delegates to a refund. Comprehensive details of each course — including objectives, target audience, and content — are clearly outlined before enrolment, and it is the responsibility of the delegate to ensure the course's suitability prior to booking.
For any inquiries related to cancellations or bookings, please contact our support team, who will be happy to assist you.
