Efforts to identify optimal completion technology and design parameters are complicated by the compounding impacts of broad statistical variability in operations, reservoir/fluid and completion/wellbore design. There are several analysis approaches available to identify and optimize key completion design parameters. Each approach offers limited insight on its own, but combining a set of approaches into a disciplined methodology can collectively present a unique understanding of optimal completion technology and design. Traditional parallel coordinates visualizations offer strong visual cues of correlations, but in datasets with broad statistical variability they often convey a lack of correlation and fail to distinguish statistical trends. Statistical methods are unique in their ability to provide insights into non-continuous correlations where upper and lower thresholds exist; however, they are not effective at providing a deterministic measure of an individual input's effect on an outcome. Modelling and regression analysis can provide a means to measure the effect of several input variables on an outcome, but lack transparency and are often perceived as a "black box" solution with outcomes that have limited supporting evidence, or supporting evidence that is difficult to understand.
We demonstrate a robust multivariate analysis methodology using a hybrid approach involving the principles of parallel coordinates, dimensional normalization and advanced probabilistic techniques. One of the benefits of this approach is that it can yield statistically significant insights on sample sets as small as 80 wells. The methodology involves six steps that offer transparency to the analysis and facilitate a narrative of understanding: Selection of a performance measure set Analogue well selection Selection of numerical completion design input parameters Parallel Coordinates Distributions: input parameter impact analysis Evaluation of analogue fitness and subset selection Input Optimization Distributions: input optimization process
Selection of a performance measure set
Analogue well selection
Selection of numerical completion design input parameters
Parallel Coordinates Distributions: input parameter impact analysis
Evaluation of analogue fitness and subset selection
Input Optimization Distributions: input optimization process
We found that the use of consistent dimensional normalization on both inputs and outcomes better isolates the impact of an input parameter. The shape and position of parallel coordinates distributions can illustrate nuances of impact that are lost in other multivariate approaches.
In this paper we apply and test this methodology on three major resource plays in the Western Canadian Sedimentary Basin: a gas play, a liquids-rich gas play and an oil play.
A mature reservoir is usually characterized with a number of both vertical and horizontal wells, which may both contribute to a significant recovery to-date. When considering forecasting for new wells in the geological similar areas (GSAs), it is common practice to generate a type curve by harnessing historical production data from analog wells. But, given varying well types and completion practices (e.g. different horizontal wellbore length), the analog assumption may be challenged.
When working on type curve generation, the common questions frequently asked are: Do we need to normalize analog wells for the type curve generation? How to conduct the normalization on both wellbore length and completion parameters?
Do we need to normalize analog wells for the type curve generation?
How to conduct the normalization on both wellbore length and completion parameters?
It seems that the answer to the first question is a very obvious yes if the new wells will be designed differently from the analog wells – particularly if the lateral length and completion method is not similar. As a result, Estimated Ultimate Recoveries (EURs) and the Initial Production (IPs) will need to be normalized from those analog wells to a desired new wellbore length.
After having investigated both analytically and numerically the impacts on Recovery Factors (RFs) and the IPs from those factors of horizontal wells that include horizontal wellbore length ( RF will be directly affected by the wellbore length. There is linearity between RF and Also, IP has a high positive correlation with lateral length. In fact, these two variables have high linearity. In addition, fracture spacing will have a large impact on IP rates; however, drainage area will not at all. Note that these conclusions presume that wellbore hydraulic considerations are not a constraint. Further, lateral length presumes an effectively stimulated horizontal section.
RF will be directly affected by the wellbore length. There is linearity between RF and
Also, IP has a high positive correlation with lateral length. In fact, these two variables have high linearity. In addition, fracture spacing will have a large impact on IP rates; however, drainage area will not at all. Note that these conclusions presume that wellbore hydraulic considerations are not a constraint. Further, lateral length presumes an effectively stimulated horizontal section.
The specific scope of this study is to provide a systematic normalization technique. Dry gas and wet gas case studies from the Western Canadian Sedimentary Basin (WCSB) have been adopted to demonstrate the workflow. Further, sequential accumulation statistical logic has been successfully applied to validate the premise of lateral length and fracture spacing as the key normalization variables.
It is believed that this methodology is rigorous for dry and wet gas reservoir systems. Moreover, this methodology is also applicable to richer gas-condensate and oil plays; however, broader relationships need to be established and tested before any conclusions can be drawn with wellbore hydraulic dynamics being taking into account as an effective factor.
Development projects in unconventional reservoirs can require capital expenditures of several hundred million to several billion dollars. Therefore it is essential that the assessment of these projects be as objective and unbiased as possible – that begins with the selection of analog production type curves.
Economic analysis of any development project is highly dependent on the production schedule incorporated into the cash flow model. For unconventional resource plays, project production schedules are often based on the aggregation of production type curves for an average, or representative, well within the defined project area. In general, we rely on the most representative analog wells available to define the production type curve. In established projects, these analogs typically come from the most recent wells within the project area; for new plays, the “best” analogs may be located in a different basin or play. But regardless of the data used to build type curves, the question emerges, “How do I know if my analog production type curve is representative?”
This paper presents methods for building aggregation models using Monte Carlo simulation to:
Create confidence curves to evaluate the confidence of meeting pilot objectives as a function of the type curve uncertainty and the number of wells in the pilot;
Generate Sequential Accumulation plots which project cumulative initial production rates for pilot projects onto a forecast envelope consisting of a cumulative P10-P90 envelope. By plotting the cumulative initial rates as pilot wells are brought online, the evaluator can assess whether the modeled type curve is appropriate or if it is conservative or optimistic.
Included in the paper are several examples comparing pilots that validate the selected type curve with pilots showing potential bias in the analog type curve. An important observation from these curves is the impact of sample size (e.g. the number of pilot wells) on the predicted confidence of meeting pilot objectives and the assessed validity of the analog type curve.