Chapter 11 Overview: The Elements of Uncertainty

Chapters 11 through 14 of Monograph 5 provide a structured framework for handling uncertainty in Type Well Profiles (TWPs). Chapter 11 introduces the key elements of uncertainty, explaining how they arise in both single-well and multi-well contexts. Chapter 12 visually illustrates these uncertainties using graphical tools such as trumpet plots and probability distributions. Chapter 13 takes a quantitative approach, providing equations and statistical methods to measure and aggregate uncertainty. Finally, Chapter 14 focuses on the practical application and documentation of TWPs, ensuring that uncertainty is properly communicated to decision-makers.

While earlier chapters of Monograph 5 have already covered bias in data selection and normalization, Chapter 11 shifts the focus to uncertainty—the inherent variability in well performance and forecasting that persists even when bias is minimized.

Chapter 11: Introduction to Elements of Uncertainty

Uncertainty is an unavoidable reality in reserves estimation. No matter how sophisticated the methodology, every TWP carries a degree of unpredictability due to the complex nature of subsurface geology, well performance, and forecasting assumptions. Chapter 11 breaks down the primary sources of uncertainty, distinguishing between those that impact single wells and those that affect multi-well analyses.

Single Well Uncertainty: The Foundation of Variability

At the core of uncertainty is the fact that no two wells are identical. Even within the same reservoir, production rates and ultimate recovery can vary due to differences in:

• Reservoir properties – Small-scale variations in porosity, permeability, and fluid saturation affect well performance.
• Completion and stimulation effectiveness – Differences in lateral length, stage spacing, and proppant volumes influence how much of the reservoir is accessed.
• Operational factors – Changes in artificial lift, workovers, and shut-ins can introduce variability in production behavior.
• Data limitations – Measurement errors, missing pressure data, and inconsistent production reporting add another layer of uncertainty.

Even if all wells were drilled and completed identically, unpredictable geologic variation alone would create uncertainty in EUR (Estimated Ultimate Recovery).

Uncertainty in Multi-Well Aggregation

When constructing a TWP, uncertainty does not simply average out across multiple wells. Several factors influence how uncertainty scales in multi-well analysis:

1. Well Count and Sample Size Effects – Larger datasets tend to reduce uncertainty, but only if the sample is representative of future wells. Small or biased datasets can lead to overconfidence in projections.
2. Correlation Among Wells – If analog wells share similar geological or operational characteristics, their uncertainties are correlated. This means that adding more wells does not necessarily reduce uncertainty as much as expected.
3. Analog Well EUR Distribution – EURs typically follow a lognormal distribution, meaning that a few high-performing wells can skew the mean. Understanding whether the dataset is dominated by outliers is critical when assessing the reliability of a TWP.
4. Forecasting Uncertainty in Undrilled Locations – When applying TWPs to new wells, uncertainty compounds because future wells may not perform like past wells due to changes in drilling techniques, depletion effects, or evolving operational constraints.

Forecasting and Uncertainty Growth Over Time

Chapter 11 also highlights how forecasting amplifies uncertainty. The longer the projection period, the greater the potential error, due to:

• Decline curve fitting choices – Different decline models (Arps, stretched exponential, numerical simulations) yield different long-term predictions.
• Economic and operational changes – Variations in oil price, well interventions, or changes in operator strategy can alter production trajectories.
• Unknown drainage effects – Early production data often reflects transient flow behavior, making long-term extrapolation less reliable.

This reinforces the importance of probabilistic TWPs that account for a range of possible outcomes, rather than relying on a single deterministic forecast.

Key Takeaways from Chapter 11

• Uncertainty is inevitable in both single-well and multi-well analyses, even when bias is minimized.
• Single-well uncertainty arises from geological variation, completion differences, operational factors, and data limitations.
• Multi-well uncertainty is influenced by well correlation, dataset representativeness, and the statistical distribution of EURs.
• Forecast uncertainty grows over time, meaning long-term projections carry more risk than short-term ones.
• Managing uncertainty requires a probabilistic approach, using tools such as P10/P50/P90 distributions rather than single-point estimates.

By laying out these foundational elements, Chapter 11 provides the necessary context for the more detailed uncertainty quantification, visualization, and application methods covered in Chapters 12–14.

Next week's topic - Graphical Displays of Uncertainty