Managers crave greater confidence in well construction costs from the inception of a cost estimate for investment decisions through cost estimation for budgeting to cost tracking and control during well construction and, finally, for validity of performance tracking and benchmarking.

Shortfalls in well cost estimation and control occur frequently due to three main sources: lack of defined processes, lack of discipline and reliance on outdated or poor methodologies.

Framing the issue
Well costs are a significant component of many field developments. Effective planning and control of well costs are critical to profitability. Well costs impact operators in multiple ways.

• Well feasibility costs drive economic analysis, leading to a project approval decision;
• Drilling budget and authorization for expenditure (AFE) define a significant component of the expenditure of an operating company; and
• Actual well costs define the expenditure made and provide reconciliation to the invoiced quantities and industry benchmarking metrics.

Risk management identifies, assesses and then defines mitigation or management plans. Qualitative risk analysis prioritizes risks by assessing and combining their probability and impact. Quantitative risk analysis applies profiles and numerical values to assess the effect of the risks on the project outcome in terms of schedule and cost.

The two methods are interconnected. Initially, risks are identified and assessed in a qualitative process to identify and rank the most important, and then their effect on the schedule and costs is mathematically assessed in terms of probabilistic ranges. Probabilistic estimating is a powerful tool that is very appropriate to well cost estimates; it enables the range of costs in an uncertain and risk-bound environment to be expressed and presented in a manner that can support management decision-making.

Cost tracking is an extremely important element for managing drilling and completion costs. The tracking system must be structured to capture expenditures against the control estimate as these are committed to enable reconciliation. This structure also must relate to the invoicing system such that the well costs can be reconciled against invoices and the total system closed out within a very low percentage of variation.

A standardized method of integrating scheduling and risk analysis with probabilistic estimating provides a comprehensive workflow of these highly interdependent activities for improved decision-making and cost control as shown in the drilling and completion cost estimation and control framework (Figure 1).

Poor practices
Cost estimation and control are often poorly defined and executed practices in drilling operations. Many bad practices get copied from one operation or company to another.

Uncontrolled estimate process. Casual processes for well-cost estimating create havoc. Unrecorded requests for well costs are made, and the results are used in the exploration or development project economic analysis.

Ownership. Well costs at early stages of project economic analysis are generated outside the drilling department, invalidating responsibility to execute within the prognosis of the cost range when the wells are approved. Cost estimating, budgeting and control are often split between different entities.

Single well cost value. Companies often allocate a single well cost and use it in internal processing for approvals with management before it is technically possible to generate such a defined number. As time progresses, this number becomes the “fact” that the operation must achieve, leaving the drilling department in a no-win situation.

Deterministic-based estimates. Specific-value cost estimates made for an environment that includes both uncertainties (well design changes and drilling program updates) and risks will consequently be wrong.

Contingency expenses. Adding a fixed percentage (such as 10%) has no factual foundation and is an inadequate means to capture potential cost overruns.

Probability estimating errors. Probabilistic well costs require rules to be followed, including limiting the application of the Monte Carlo simulation to rolled-up cost and time elements and incorporating correlation. Some systems disregard this and force a normal distribution where one does not exist.

Method selection. Bottom up (micro) estimating is used at an early stage where only top down (macro) estimating should be applied.

Competency. Too often oil companies treat this skill set as something a drilling engineer can undertake while managing his or her daily workload. In small operations, where the subject drilling engineer has the requisite process to follow and training in the skill set, this works. In larger operations, the workload of a properly run estimating and control team requires dedicated expert personnel.

Poor alignment with accounting. Well costs often are estimated in one format, a specific breakdown structure, while accounting uses another format and breakdown for cost allocation. Reconciliation between the tracked well costs and the invoiced well costs is very difficult.

AFE variances. Requests for increasing AFEs are often recognized and processed after the cost overruns have occurred. Earned value rules provide a means to predict cost overruns long before these occur, enabling approvals for AFE increases.

Value of probabilistic estimation
Probabilistic estimation and maturation are integrated components throughout stages of well cost estimating and control. Probabilistic estimation requires full disclosure of the uncertainties and risks in well cost estimates, quantifying their effects on the well cost estimate as easily viewed S-curve graphics.

In Figure 2, the S-curve shows the spread of results from P10 to P90, demonstrating the range of possible outcomes. In higher classes of estimate, the spread under the S-curve tends to be large due to the significance of the unknown elements in the design and execution of the wells. As the cost class progresses toward Class 2, the curved spread reduces, reflecting the improvement in understanding of the execution through increased design and programming detail.

Some typical aspects of S-curves are:
• The P90 result is usually offset further from the P50 than the P10 result due to the increased opportunity of overruns compared to underruns;
• The final S-curve P50 falls within the initial S-curve P10 to P90, reflecting reduction in uncertainty and not a change in results; and
• Development wells have a lower spread between P10 and P90 due to the more certain nature of these wells.

Adopting standardized workflow recommendations of cost estimating and control framework improves confidence in well-construction cost estimates, enabling improved management decision-making.

Acknowledgment
This article is based on SPE paper 173148-MS, presented at the SPE/IADC Drilling Conference March 17 to 19.

Comments