Embedding Automatic Risk-Level Estimation and Performance Analysis in Drilling Plans

Fig. 1—Three-dimensional representation of the admissible drilling-parameter set. For a given drill depth and confidence parameter, the safe combinations of flowrate, drillstring rotational speed, and WOB are represented by the red volume. Although quantified information is difficult to obtain from such a representation, it is clear from the comparison of the two volumes that the configuration on the left provides more safety with respect to the analyzed risks than the one on the right.

You have access to this full article to experience the outstanding content available to SPE members and JPT subscribers.

To ensure continued access to JPT's content, please Sign In, JOIN SPE, or Subscribe to JPT

It is not unusual that the difficulties encountered during a drilling operation can be tracked to choices made during planning. However, generating a robust drilling-operation plan is not easy because substantial uncertainties are often associated with the actual geological context. To address this problem, this paper presents a method that quantitatively evaluates the risk levels of a drilling-operation plan as a function of the underlying uncertainty associated with its description.


A drilling operation is always subject to operational risks (e.g., formation-fluid influx, lost circulation, borehole instability, packoff, stuck pipe, or drillstring failure). The chance of occurrence of some of those threats may be small in a particular well design. But, in other configurations, it may be known already at the planning stage that potential drilling problems are likely, because of tight operational margins for instance. The drilling challenges also may be aggravated by uncertainties in the geological context.

Making a robust drilling-operation plan is a challenge, and there are many reasons for this. One of them is connected to the difficulty in evaluating quantitatively, at the planning stage, the effect of uncertain information on the effective risk levels that may prevail during well construction. Quantitative risk assessment with management of uncertainty has been described for specific tasks, such as choosing a mud weight as a function of an uncertain geopressure window. The more risks that are simultaneously evaluated within the expected uncertainty of the coming operation, the greater the chances are that the proposed drilling-operation plan will be robust.

Drilling-Operation Risks

Numerous sources of risk exist in a drilling operation, and the quality of the evaluation made by the proposed risk-evaluation method is directly related to treating as many threat sources as possible. However, it is an overwhelming task to manage all of the conceivable potential sources of drilling problems at once. For this reason, the authors decided that, as a first step, only a subset of possible causes of drilling incidents will be considered. The causes from this initial list are associated with typical problems encountered in drilling operations in the North Sea.

A first group of drilling problems is related to geopressure margins. This includes the risk of the downhole pressure being lower than the pore pressure anywhere along the openhole section during the evaluation of the drilling operation and, therefore, risking formation-fluid influx. This risk is denoted as R1-Pore Pressure. Similarly, the downhole pressure is compared to the collapse pressure of the openhole formation rocks to estimate the risk for borehole instabilities (R2-Collapse Pressure). The risk of formation fracturing caused by downhole pressure exceeding the fracturing pressure at any depth along the openhole section is also estimated (R3-Fracturing Pressure). Finally, hydrostatic pressure being larger than the minimum horizontal stress of the exposed formation rocks leads to the evaluation of the risk of experiencing permanent mud losses in the case of a lost-circulation incident (R4-Minimum Horizontal Stress).

Another group of risks is associated with cuttings transport. High concentration of cuttings in suspension anywhere along the annulus may be the origin for a packoff incident (R5-High Cuttings Concentration). In deviated wells, large accumulations of cuttings in beds may cause packoffs, overpulls, or set-down weights (R6-High Cuttings Bed).

Risks are also associated with potential drillstring failures, such as buckling (R7-Buckling).

A last group of sources of potential problems is connected to the limitations of the drilling equipment. For instance, the pump pressure shall not exceed the rating of the rig hydraulic system (R8-Maximum Pump Pressure). The torque at the topdrive shall be lower than a maximum value, which may depend on the selected gear for topdrives with multiple gear ratios and topdrive speeds vs. torque performance curves (R9-Maximum Topdrive Torque).

In a drilling context, while using a conventional drilling method, three parameters can be controlled—the weight on bit (WOB), the drillstring rotational speed, and the flow rate.

Uncertainty Management and Propagation

When evaluating the potential risks embedded in a drilling program, accounting for the possible sources of uncertainties that may affect the estimations is important. Indeed, a combination of drilling parameters is assessed to have the potential of generating a drilling incident if an estimated parameter goes beyond a given limit. The limit, however, may not be known precisely, and the estimation of the parameter may also be uncertain.

Each element of the drilling program has an associated uncertainty that describes how precisely that information is known but also how much variation is acceptable from the planning stage to the well construction.

Unfortunately, these sources of uncertainties are not independent. For instance, the uncertainty on the wellbore position affects the depths at which formation tops will be encountered. The borehole-position uncertainty must be combined with the depth tolerances of the stratigraphy prognosis. In turn, the resulting depth variations of the occurrence of the intersection of the wellbore trajectory with a formation top have an effect on the associated geopressure gradients, which also have their own uncertainty. Thus, the uncertainty-­propagation mechanism influences the margins used to estimate risks.

Results and Analysis

The analysis process covers the configuration space, the parameter space, the drilling time, and depth space, and the risk space. Several high-dimensionality parameters need to be represented. The authors chose three different representations of the analysis results. All of the representations share the same configuration capabilities. They all depend on the chosen confidence value. This parameter, which controls the size of the admissible drilling-parameter space, can be set by the user using a slider. The representations also depend on the chosen set of drilling criteria. The nine criteria considered in the analyses can be turned on and off using check boxes. This allows assessment of individual risks or of any combination of risks.

The first visualization provides a global overview of the allowable drilling-­parameter set. For a given drill depth, a 3D display of this set is provided that allows a quick estimation of the drillability of the well at that particular location (Fig.  1 above). Even though quantified information is difficult to gain from such a visualization, a qualitative assessment can already be performed using this representation. The displayed volume will change when modifying the chosen confidence parameter; it will shrink or expand as the parameter is increased or decreased. The shape will also be modified by the chosen set of risks.

Then, 2D projections of this volume give more qualitative information. The user can fix any of the three drilling parameters and inspect the projection of the volume on the two remaining dimensions. For each of the selected criteria, the boundaries of the allowable parameters are plotted. The forbidden regions (i.e., the combinations of parameters that create drilling incidents with a probability larger than the chosen confidence value) are marked by a red area, while a color coding of some parameters of interest (rate of penetration, downhole equivalent circulating density) is visible in the allowed region.

Finally, projections for a single drilling parameter along depth are displayed. For those projections, the two other drilling parameters are fixed (the values can be modified by the user), and one has access to the allowable range, as a function of depth, for the considered parameter. Here also, the forbidden values are marked in red, while parameters of interest are displayed in the background.


  • A new methodology to evaluate the embedded risk levels in a drilling operation has been presented.
  • Risks are defined as the probability of a parameter (e.g., downhole pressure) to exceed a tolerable limit (e.g., fracturing pressure).
  • The tolerable limits are associated with a level of uncertainty that arises from the combination of several sources of uncertainty such as possible wellbore-position variations or variance in stratigraphic depths.
  • The calculation of the comparison parameters depends on the certainty of input information and the possible evolution of drilling parameters during the drilling process.
  • The modeling of possible drilling scenarios is a function of achievable rates of penetration.
  • The possible rates of penetration can be evaluated as a function of the formation-strength prognosis.
  • Various drilling advancement scenarios influence the possible variation of temperature and production of cuttings during the drilling operation.
  • The estimation of the possible temperature evolution is a function of the thermophysical properties of the formation rocks and drilling fluids, which are often known with a large level of uncertainty.
  • In addition to risk levels, obtaining pertinent assessments of performance is also possible.
This article, written by Special Publications Editor Adam Wilson, contains highlights of paper SPE 181018, “Automatic Performance Analysis and Estimation of Risk Level Embedded in Drilling-Operation Plans,” by Eric Cayeux, SPE, and Benoît Daireaux, SPE, IRIS; Mohsen Karimi Balov, Statoil; Stein Haavardstein, ConocoPhillips; Leif Magne Stokland, Wintershall; and Arild Saasen, SPE, Det Norske, prepared for the 2016 SPE Intelligent Energy International Conference and Exhibition, Aberdeen, 6–8 September. The paper has not been peer reviewed.

Embedding Automatic Risk-Level Estimation and Performance Analysis in Drilling Plans

01 September 2017

Volume: 69 | Issue: 9


Don't miss the latest content delivered to your email box weekly. Sign up for the JPT newsletter.