The Use of Probabilistic Analysis for Estimation of Drilling Time and Costs When Evaluating Economic Benefits of New Technologies
- Peringandoor Raman Hariharan (Hydril) | Robert Arnold Judge (Hydril) | Dat Manh Nguyen (Hydril-SMDC)
- Document ID
- Society of Petroleum Engineers
- IADC/SPE Drilling Conference, 21-23 February, Miami, Florida, USA
- Publication Date
- Document Type
- Conference Paper
- 2006. IADC/SPE Drilling Conference
- 4.1.5 Processing Equipment, 1.6 Drilling Operations, 2.1.7 Deepwater Completions Design, 1.1.2 Authority for expenditures (AFE), 1.7 Pressure Management, 4.1.2 Separation and Treating, 1.7.3 Dual Gradient Drilling, 5.7.3 Deterministic Methods, 5.7.4 Probabilistic Methods, 5.6.3 Deterministic Methods, 4.4.2 SCADA, 1.12.6 Drilling Data Management and Standards, 5.7.5 Economic Evaluations
- 0 in the last 30 days
- 672 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 8.50|
|SPE Non-Member Price:||USD 25.00|
There has been limited published work involving the applications of probabilistic methods in drilling engineering, particularly when estimating drilling times and costs. This fact is even more acute when considering the evaluation of new drilling technologies. Traditionally, drilling costs have been estimated using deterministic methods, where little scope exists to incorporate an estimation of risks and non-productive time (NPT) in a meaningful manner. In deepwater drilling where the costs of a single well can easily run into tens of millions of dollars, the value of using probabilistic methods enables the economic evaluation of risks in a quantitative and understandable manner. Within such a probabilistic framework, it is possible to more accurately evaluate the range of potential economic impacts of new technologies for work and process improvement for drilling cost reduction such that the rewards and risk can be clearly understood.
This paper presents a new approach and a tool for estimation of drilling time and costs. The tool uses a commercially available software package capable of conducting Monte Carlo simulation and analysis. The well is modeled using a user-friendly interface with the two primary variables being the timings for drilling activities and individual cost components, both of which are defined as probabilistic distributions rather than a single number. Further, non-productive-times (NPT's) are also incorporated and described as distributions where applicable. An example case is presented to illustrate the benefits of using probabilistic methods compared to deterministic methods with the main advantage clearly shown while trying to model risk. The benefits become clearly evident while trying to gauge the usefulness of novel technologies where very limited historical information is usually available.
The paper also includes an interesting survey conducted through the SPE that shows the prevalence of (or lack of) the use of probabilistic methods in the drilling industry today.
The drilling costs and time spent in the development of a field still form a major percentage of the costs involved in the production of hydrocarbons. In the case of exploratory drilling undoubtedly, these costs form the major component and thus there's always an impetus to control and keep them low. Traditionally, the drilling times and costs have been estimated on a deterministic basis which provides a single number for the time as well as the total cost for drilling of a well. As is understood, the magnitude of each component that contributes to the computation of the final number is subject to variation themselves and hence, the deterministic process has to be conducted over and over to obtain some form of a range for the final number for the estimated time and cost. Incidentally, as is well known, this repeated and random sampling from a series of values (a probabilistic distribution) for a particular variable is accomplished through the process known as Monte Carlo simulation. The value of utilizing techniques such as these prove to be extremely useful in the presence of gross and acute uncertainties that exist in these data. As we all know, drilling operations are always fraught with uncertainties.
Generally, drilling AFE's (Authorization For Expenditure) are generated based on offset data, which primarily consists of the recorded and documented experience of the drilling operations. From the analysis of these data, estimates for drilling performance and the likelihood of facing drilling complications are estimated. Added to all these is also an element of expert judgment, though subjective, that is drawn upon in the estimation of time and costs for the new well/s to be drilled. Without the use of probabilistic methods, the estimates for both of these variables suffer from the lack of a true scientific approach to predict time and costs. Also lacking is the opportunity for quickly evaluating various scenarios that would immensely aid management in evaluation of different prospects and in making decisions with regards to committing financial resources for drilling projects.
|File Size||363 KB||Number of Pages||10|