There are many ways to approach the valuation of an exploration portfolio. The most appropriate methodology will depend on the scale and complexity of the portfolio as well as time/resource constraints. An expected monetary value (EMV) approach based on success case discounted cash flow analyses is generally the preferred valuation methodology. Various EMV approaches are described and when they might be applicable is discussed. For large complex portfolios the only reliable valuation methodology is a full probabilistic approach built around a Monte Carlo simulation of all possible exploration outcomes. This approach will be illustrated via a detailed case study to back up the EMV discussion. The expected value of an exploration portfolio should be the mean of all possible values of the portfolio. This is driven by all possible combinations of success and failure of the exploration drill out of the portfolio and the resource uncertainty in those wells that are successful. Short cut approaches are frequently adopted to simplify the valuation process but in many cases the final value, which may be described as the expected value, is in reality not the true expected value. These short cuts often ignore any correlation or dependency between the chances of success for prospects in the portfolio. In undeveloped or immature areas, the commercial and development synergies and dependencies can also have a very significant impact on the ultimate expected value. These short cut approaches do not capture these impacts. In most cases, therefore, the only way to derive an expected value is adopt a full probabilistic approach in a similar manner as routinely employed for resource volumes. A large worked example is presented to highlight the methodology together with some practical ways of dealing with valuation complexities that arise in many portfolios. The end product is not only a single expected value but a distribution of value of the portfolio. The ability to report a range of value similar to the range of resource volumes that are routinely reported is far more informative to decision makers or investors than the current practice of a single expected value.

Reserves estimation is crucial to the petroleum business as it involves all financial aspects of the petroleum companies. Among reserves classification, proved reserves always capture the most attention because most value is attached to it. By the SPE definition, proved reserves must be estimated by reliable methods that have a high-at least a 90% probability (P90) - that the actual quantities recovered will equal or exceed the estimate. Decline curve analysis (DCA) is one of the most commonly used methods for proved reserves estimation throughout the industry. Through the DCA, a production history is fitted with a trend line, then the line will be extrapolated to an economic limit for the reserves estimation. If linear regression is used, the line is the "best estimate" that represents the performance, which corresponds to the 50 th percentile value (P50). This practice, therefore, conflicts with the proved reserves definition. In this paper a method is derived to estimate the variation of the reserves spread (the difference between P50 and P90). Compared to Monte Carlo, the method gives good results. The analytical solution is then used to study the sensitivity analysis of the spread and a field application. The study covers all decline models (exponential, hyperbolic and harmonic models) and both cases where the decline exponent is known and unknown.

Both the Kelly Criterion and risk aversion using utility theory are methods that attempt to mitigate the risk of gambler's ruin through capital management by suggesting working interest levels for each project. Utility theory uses an assigned risk tolerance and a natural log utility function to calculate the optimum working interest for a project. The modified Kelly method does not use a utility function but rather a geometric mean-based valuation derived from the ratio of the expected value to the net present value. This ratio, called the Kelly Criterion, is a guide for capital allocation to a project as the percent of the total capital available. When the total recommended cost of all the projects exceeds the budget, the combined portfolio is optimized by further adjusting the working interest using a linear programming technique, such as Excel’s ™ Solver. This optimization process was described by MacKay (1995) at a previous HEES Symposium (SPE 30043). The objective of this paper is to compare the procedure described by MacKay (1995) to a similar procedure using the modified Kelly method. The Kelly Criterion is currently used for financial investment decisions ( Lee, 2006 ) and described in the book "Fortune’s Formula" by Poundstone (2005) .

The choice of appraisal strategy for the decision whether to develop a reservoir largely determines the amount of uncertainty that is carried forward to the development and execution phase of the project. Hence, the selection of an appraisal strategy can indirectly influence later go/no-go decisions. Reservoirs are appraised by drilling and producing wells. In unconventional reservoirs, these wells represent a small subset of possible wells that could be monetised if the decision were made to develop the reservoir, i.e. the appraisal wells sample an underlying population of possible wells. This study explores how an optimal appraisal strategy can be designed in terms of the number of appraisal phases, the number of wells to be drilled in each appraisal stage, and how long to produce the appraisal wells before deciding whether to abandon the project or to proceed to the next stage. It is demonstrated how the view on the average expected ultimate recovery of a well in an unconventional reservoir can be continually revised as new information surfaces. Production data from a large well set from the Montney Formation, which straddles British Columbia and Alberta, Canada, is used to assess how accurately initial production predicts the expected ultimate recovery of a single well.

Unconventional resource plays are not typically homogeneous; therefore the reservoirs might not be contiguous. Geostatistical based earth modeling can create hundreds of reservoir property realizations. The challenge is to select a few optimal models from these realizations for further analysis. Traditionally, candidates are selected from ranked distributions representing quantiles (e.g., P10, P50, and P90) of the reservoir volumetrics, which assumes contiguous reservoir volumes. However, such methods lack any physical geometrical information. In this paper, a two-way clustering method is used to take into account spatial property information. First, a three-dimensional (3-D) realization is transformed into a one-dimensional (1-D) array, so that each element of the array maps to one single grid cell that carries the spatial location information. Next, a matrix is created from the multiple realizations. Each column represents a single realization, and each row represents a single 3-D grid cell. Then, clustering algorithm is applied to both columns (R-mode) and rows (Q-mode). R-mode shows the similarity of different realizations, whereas Q-mode shows the clusters of different cells. Finally, connected geobodies are extracted from the clusters. This novel method was applied to a west Texas Permian basin reservoir. Domain experts can first select realization candidates from different R-mode clusters and compare the results. Then, the location and connectivity of cells from Q-mode clustering can be mapped in addition to evaluating their statistical ranking. During Q-mode cluster analysis, Euclidean distance carries the spatial property location of the cell in a distance calculation. Statistical analysis of the geobodies suggests the optimal realization, which is not necessarily the P50. Through such quantitative evaluations of all the realizations, geologists and reservoir engineers can select optimistic to pessimistic realizations to aid economic assessment of the reservoir.

The paper discusses a novel workflow and tools for stochastic optimization under uncertainty aimed at maximizing the value of exploration and early appraisal assets using a public example of an East Africa offshore gas discovery. Following an initial discovery well, the prospect enters an appraisal phase with tangible uncertainties in resource base (a P10/P90 ratio of 3.1), well productivity (P10/P90 of 2.1), and cost estimates (30%/+50% tolerance). The choice of LNG train capacity and development concept changeover point has a profound effect on asset evaluation results used for attracting new project partners. These uncertainties raise a number of critical questions: How can a facility engineer elaborate these key parameters given the three-fold uncertainty in resource base at this project stage? Will these choices change under a random walk in oil prices or given different attitudes towards risk by the management? Does maximization of project after-tax value and of production entitlement imply identical values of control parameters in these circumstances? The challenges are resolved by sampling the solution space with 54,000 full-cycle evaluation trials run in just over 13.5 hours using a combined power of an exploration decision support tool with fit-for-purpose and fit-to-data risks, resource, and value assessment that honors all risks and uncertainties; a workflow management tool for decision optimization based on a multiple-realization approach powered by various uncertainty and optimization methods; and a proprietary link between these two applications. The original operator's view on the facilities concept was found to reflect a P50 of reserves uncertainty and a P50 of worldwide LNG train capacity distribution. Optimizations were performed to address five progressively complicating problem formulations. The achieved outcomes match conclusions that could be derived with a reasoned train of thought in economy-of-scale context meaning our solution is applicable for modeling situations lacking simple analytical solutions.

The process of exploration planning in the oil and gas industry has historically been time-consuming and inefficient with limited information and large uncertainties to contend with. How does an oil and gas company take a set of exploration opportunities stored in disparate systems across the organization and combine them into a single, optimized forward exploration plan? This paper will examine the current industry processes in exploration planning and compare these with the benefits of an integrated solution with direct links from the subsurface to commercial to identify and track key parameters over time for business planning decision making. This paper will first discuss the industry challenges and current practices in exploration planning. The proposed solution will then be defined as a shared repository of opportunities, consistent capture and quantification of subsurface risk and uncertainty linked with the quantification of potential hydrocarbon recoverable ranges, economic analysis across different fiscal regimes and stochastic portfolio optimization. The paper will then look at optimizing a sample set of exploration opportunities and examine the results and benefits gained after deployment at an integrated national oil company. The integration of risk and uncertainty from the reservoir properties through to stochastic portfolio optimization based on project timing and working interest is a unique solution compared to the "siloed" and disconnected approach common in the industry. During this project, a shared understanding of risk and uncertainty from the subsurface through to the economic evaluation was observed throughout the asset teams and planners. The implementation and execution of the proposed solution improved the efficiency of the exploration decision-making process and the management of the company’s exploration portfolio. Integrated exploration planning leads to significant improved overall portfolio revenue at reduced risk and costs. The proposed solution is applicable to any mid- to large-sized oil and gas company with a number of exploration opportunities to choose from for inclusion in their portfolio. The solution’s value lies in the consistent approach to directly link key risks and uncertainties in the subsurface properties of the reservoir all the way through to the commercial evaluation of a portfolio of exploration opportunities. The result of this approach is optimal decision making and increased efficiencies in exploration business planning due to more thorough understanding and mitigation of key risks and uncertainties affecting the success of exploration opportunities.

This paper presents a method for the fast evaluation of fracture-stimulated condensate reservoir economics. For the calculation of production decline in such reservoirs, an efficient numerical model with a three-phase transient analysis of pressure distribution was built and validated using the predictions from reservoir solvers and field data. This model solves for gas-, oil-, and water-flow parameters, accounting for the gas-oil phase transition, and has been realized in a numerical code and compared with predictions from commercial software and available field data, such as production-decline curves. The developed numerical model has been implemented in commercial software and used for the sensitivity analysis of reservoir productivity regarding changes of fracture size and spacing, as well as reservoir permeability in the fractured condensate reservoirs, with an account for multiphase reservoir flows and reservoir properties. A side-by-side comparison of predictions from two commercial reservoir simulators has shown that that this model accurately calculates transientpressure fields near the fractures and the productiondecline curve. The objective of the economic analysis and fracture optimization stage is reduced to finding the target function minimum in an N -dimensional parametric space using various constrained minimization techniques, including a Quasi-Monte Carlo analysis and the Active Set Method.

One of the unique aspects of the US shale boom is that firms have purchased most mineral rights from thousands of small landowners. Though several studies have focused on cross-sectional aspects of these leases, none study their dynamics. Incorporating several aspects of the unique US context, I develop and simulate a model of depletable mineral rights that matches several key aspects of data on leasing and drilling in the US Eagle Ford shale.

Reserve and resource estimates and classification are meant to provide reasoned production expectations considering uncertainties and development risks. In conventional fields several estimation methods (volumetric, material balance, analogy, and production trends) can work in tandem in order to provide these reasoned estimates which relate to static and dynamic characterization. Despite intensive efforts to utilize all available data, these estimates could be still inaccurate, but have served our industry well in making business decisions. With the introduction of the unconventional resources, this process requires re-examination and realignment to development priorities, risks, uncertainties and of course production drive mechanics. The early approach in recoverable volume estimation was based heavily on production trends and a statistical appraisal of observed variations with limited relation to production mechanisms and reservoir characteristics. Key work in many disciplines has shown that although production performance does not follow the same static and dynamic trends as in conventional reservoirs, it relates to certain static and dynamic subsurface characteristics particularly to natural fractures. Faults, fairway and background fractures become more important in liquid rich and oil plays (as fluids segregate) and invariably affect the process of hydraulic fracturing and the development of drainage patterns. Since natural fracture network characteristics may change rapidly even across the length of a horizontal well, it is difficult to expect a homogenous performance without the understanding of the natural fracture trends, reservoir characteristics and the change in the respective drive mechanisms. Most commonly used reserve and resource classification methods rely on proximity to existing production. Due to these effects, the expectation of similar performance in infill locations could be diminished. Apart from being an inaccurate prediction, this misconception may lead to the wrong development path in which wells classified as highest ranking may exhibit low performance. In the context of these observations it would be rational to draw fairway maps in relation to key geologic characteristics and dynamic effects utilizing all available information within and around the project area and to provide a basis for analogy which conforms better to the actual production drive mechanisms. In other words, using a physics based reliable technology (fairway maps) we are able to establish undeveloped areas with similar producibility at greater distances than production offset locations with reasonable certainty. Testing the validity of fairway maps with continuous development provides more confidence in predicting offset locations; classifying reserves and resources appropriately and guiding development towards better results. This process provides close alignment between development priorities and certainty levels in reserve and resource estimates which fulfills the basic requirements for rational project management. Subsequent application of fairway maps in other areas and similar plays may someday become a more general reliable technology paradigm for addressing unconventional resource assessment.

Historically, oilfield services companies seemed to be financially more sensitive to crude oil price volatility compared to companies doing exploration and production. This was evident both in times of rising and falling price, imposing serious challenges to them. The year 2014 marked, after a 40 percent drop in the price of oil, the agreement for one of the largest acquisitions in the industry between two of the biggest international oilfield service providers; that of Baker Hughes, Inc. by Halliburton Co. The implications of these price fluctuations significantly influence this sector of the industry shaping the narrative around it. This paper evaluates the extent to which crude oil price variations affect the service companies in the stock market compared to exploration and production (E&P) companies. Regression analysis is performed on two sample groups; one consisting of service companies and the other of E&Ps. Daily data from the beginning of 1986 until early 2015 is used for both stock and crude oil prices. When compared to the E&Ps group, the service companies group is more oil price sensitive throughout the whole period by a factor of almost six and a half. In particular, from 1986 to 2007 this factor is 6.3. For the post 2008 period, a reverse causality effect is observed for both groups with changes in the oil price, with the service companies group slightly less sensitive, yielding a factor of 0.9. The factors responsible for the observed phenomena are evaluated, backed with data analysis. The higher economic limits of unconventional compared to conventional energy resources, make the service companies specializing in unconventional play development struggle as these are the first to become uneconomic in times of falling price. The downstream operations most E&Ps have unlike service providers, enables them to buffer losses in their upstream sectors when price drops and vice versa. Also, the decreasing dependence of the E&Ps on service companies as they grow bigger makes the latter less adroit to price variations than the former. Finally, the rise of the national oil companies in recent decades in terms of production and reserves control, has led to additional competition against service companies by international oil companies for providing specialized technical assistance.

Decline curve analysis has dominated reserve estimation techniques for decades. Ironically, the introduction and expansion of probabilistic calculation techniques largely by-passed these kind of calculations, but not for lack of effort. Much thought and writing has been dedicated to the unique issue of defining uncertainty in future production trends. Despite two decades of slow evolution, the resulting methods have been partial and prohibitive. Methods like scenario analysis, Delphi forecasting, bootstrapping and Bayesian networks have been proposed and even sometimes utilized. Though they have demonstrated some success, they all suffer from limitations in difficulty and objectivity and aggregation. They have also all utilized smooth model fits and projections. In tandem, other kinds and ways of uncertainty analysis have been applied in the time-domain or dependent sequences using decision trees, Bayesian networks and, more recently, Markov chains. Newly developed machine-learning techniques combine these two separate streams of uncertainty analysis into a framework that promises to address both the quantification and aggregation of uncertainty in an objective way. What seems like a black-box method is instead a natural and automated extension of well-accepted techniques, and the few published studies of the technique suggest that it may be robust, particularly over the scale of years.

The translation of risk from conventional to unconventional resources moves the fulcrum from risk of finding, towards risk of commerciality. We are often certain a hydrocarbon producing system exists; but, uncertain in what manner (production profile) it will produce. This reality is incompatible with traditional methods of risking in which recoverable volumes are estimated from volumetric calculation, and risked volumes may be calculated as the product of unrisked volumes and some risk factor. Production profile is then determined after-the-fact. Unconventional resources are typically evaluated from the basis of a type well or production analog first, and then the recoverable volume is emergent from this production profile. Like the chicken-and-egg problem, uncertainty of recoverable volumes must be estimated in association with one or more production profiles. The threshold of commerciality can be calculated, and the chance of commercial success determined by the intersection of the threshold volume on the distribution of recoverable volume. This distribution is truncated and the outcomes above the threshold are rescaled to represent success outcomes… but, failure outcomes are disregarded. This paper presents a simple workflow using Frequentist and Bayesian approaches, respectively, for proper risking of uncertain recoverable volumes for an unconventional resource, taking into account the chance of false positives from appraisal well information. A subjective risk tolerance can be included to respect how aggressive or conservative a company may be in pursuit of a project. A method of re-casting production profiles is demonstrated as an improvement over the common method of scaling, or factoring, the initial production rate. The scaling approach may reach unrealistic values of initial production in an attempt to achieve the proper recoverable volumes. Productive yet uneconomic outcomes should be included in any thorough project evaluation. Failure to do so may overestimate the value of a given project. However, the chance of such outcomes occurring is often seen as a difficult-to-determine value. Quantifying the uncertainty in type well volumes allows evaluation of the imperfectness of information of early well results. Enhanced knowledge of the project risk, the chance of success, a view of the success case, and an understanding of false positive outcomes, can lead to higher quality decisions.

Understanding the oil price changes through the demand and supply dynamics is a challenging problem. Traditional models based on stochastic process and aggregated supply and demand have been successful in explaining the price trends in short periods of time where there is no big changes in either supply or demand sides of oil market. However, these models usually failed to capture the big shocks of oil price in circumstances where there are unexpected influential events, such as reports on the economic growth of the non-OECD members. In this paper, we propose a revised model of oil price based on the supply and demand dynamics. Particularly, we inspect the effects of unconventional parameters on the oil price, many of which have been ignored in the traditional models. Our model investigates the effects of these expectational parameters on the major factors that shape the supply and demand. Particularly, the oil price is still modeled to follow the supply to demand ratio, however, the supply and demand are replaced with their expected trends in the model. In other words, based on the influential events that can happen in either demand or supply sides, related variables are initiated which subsequently affect the expected trend of supply or demand, and the oil price eventually. Our model is a system dynamic model that mathematically relates the determinants of the oil price based on the underlying causal relations. The whole model numerically solves differential equations relating the marginal parameters and the main factors including the oil price, demand and supply. We separate the oil supply into US, OPEC and other non-OPEC countries supplies whose parameters differ due to conventional and unconventional resources. Moreover, we separate the oil demand into OECD, Brazil, Russia, India, China (BRIC) and non-OECD countries’ demands due to the difference in the dependence on the economic growth and oil consumption in those group of countries. We build our core model based on regression analysis on the historic data of WTI oil price, OECD and BRIC economic growth and demand, US, OPEC and other countries supply data. Then we train other parameters of the model using the historic reactions of the market to unexpected events that have occurred in the past. Finally, we simulate the model to analyze a case study of the events occurred in the 2015 including the deceleration of the growth in Chinese economy, speculations on American shale productions, speculations on the OPEC’s decisions, and conflicts in the Middle-East region. Our results show that our model can be used to reproduce the price trends compliant with the past data, and provide predicted trends of supply, demand and their influences on the price.

While a severe drop in commodity prices was expected to have an adverse valuation impact on oil & gas producers, the variability of this impact across the Exploration and Production ("E&P") company universe has surprised industry participants. What factors caused the impact to be magnified for certain companies and muted for others? In this paper, we have investigated the role of debt in the current financial crisis afflicting the E&P industry, and proposed a new approach for a financially leveraged E&P company to measure, manage, and reduce the impact of commodity cycles on its equity value. We analyzed the last three years of publicly available financial information for seventy-one publicly traded E&P companies operating in the United States. By performing statistical analyses using this information, we observed a modest correlation between a company's financial leverage and its equity value loss. It is noteworthy that the debt load by itself did not explain a majority of the performance variability, and there were obvious signs of other factors influencing the equity performance which we also investigated. Those influences (other than debt) include the geologic basin where an E&P company's reserves are located and the extent of commodity price hedging by the company. In summary, we have three key observations: (1) the degree of a company's equity value loss has a modest correlation with the amount of its financial leverage; (2) the relationship between financial leverage and equity value loss changes from basin to basin which indicates that the value loss is a function of reserve economics; and (3) a subset of relatively leveraged E&P companies, aided by their prudent management of commodity price hedges, significantly outperformed their similarly leveraged peers which highlights an important linkage between the management of debt and hedges. This paper presents an in-depth analysis of observation #3 and concludes with a discussion of hedging practices and recommendations aimed at enhancing a producer's ability to add reserves, achieve cost reductions, and preserve debt service capacity through a commodity down cycle. Since data analysis showed the valuation impact to be basin-specific (i.e., varied by basin, all else being equal), we segregated our E&P company roster into smaller subsets based on which basin holds the majority of a producer's reserves. The Appalachian Subset, which includes any SEC reporting company that has the majority of its reserves located in the Appalachian basin, constituted the largest subset of thirteen companies and was, therefore, chosen for comprehensive research (see Appendix 3). While we noticed the presence of a relationship between corporate leverage and equity value loss, there were certain exceptions within the Appalachian Subset of E&P companies, namely Company 3, Company 32, and Company 55, for whom a relatively high leverage did not cause a correspondingly high value loss. Our study of their SEC reports revealed that these outperforming companies maintained a sizeable commodity price hedge book which, we believe, immunized their equity valuations against the potential for value diminution from a relatively higher leverage profile. Among the thirteen Appalachian companies, six distinguished themselves from the rest of the group on the basis of their superior equity performance. The stocks of these six companies, referenced as the Leaders in this paper, performed largely in line with the performance of the underlying commodities (i.e., the price performance of oil and gas). However, the other seven Appalachian producers experienced an equity value loss significantly in excess of the loss experienced by the underlying commodities. Based on the use of the Devashish Method of Equity Value Immunization discussed herein, we have tried to explain the outperformance by the Leaders using detailed financial data from SEC reports, primarily 10-Q and 10-K. Through this exercise, we identified a common trend – continuous hedging to align the total hedged volume with the total financial leverage – emerging out of the disparate hedging practices employed by these Leaders. As prescribed by the Devashish Method, we have used a graphical tool to analyze whether a company's debt load could contribute to its equity value diminution following a precipitous fall in commodity prices. Under the Devashish Method, this graphical analyzer tool is called the DEVA Plot – an acronym for the Debt effect on Equity Valuation Analyzer. The DEVA Plot of one or more comparable companies is a Cartesian plot of certain debt and hedge metrics of the companies discussed herein. The Devashish Method states that, through price hedging, it should be possible for an E&P company to protect its equity value in a downturn against the value diminution resulting from a high debt load. To achieve this goal, the company's hedging program should be managed such that a graph of its debt and hedge data points falls close to or above the unit slope line on the DEVA Plot. We believe this analyzer tool could assist evaluators, such as a corporate manager or an equity research analyst, in their analysis of the adequacy of an E&P company's hedge volumes vis-à -vis its debt load. The DEVA Plot could also be used for a comparative peer group analysis at one or more points in time. Using this tool, we found the practice of partially hedging only near term production to be insufficient for the purposes of equity value immunization for a majority of the Appalachian companies we researched. So what are the best practices that come out of our research? To protect equity holders from the potentially value diminishing effect of debt, a financially leveraged oil and gas company must manage its hedge book such that the total hedged volume stays aligned with its debt load on the DEVA Plot. We urge the managers of a financially leveraged company to utilize the Devashish Method of Equity Value Immunization by (I) identifying a suitable area on the DEVA Plot, near or above the unit slope line, based on their company's asset mix, financial leverage, and business goals, and (II) adopting a disciplined hedging policy to keep their company positioned in that area of the DEVA Plot. An analysis of quarterly hedging data shows the Leaders engaged in such practice through a continuous hedging program which kept their hedged volumes aligned with their leverage profiles. A consequence of adhering to a continuous hedging program through the price cycle is the dollar cost averaging of hedged prices – which appears to be an implicit goal of these Leaders. Hence, a disciplined hedging policy must involve committing to a continuous hedging program without factoring in personal views on the future direction of commodity prices. Furthermore, a borrower would benefit from adding to its hedge position timely, in conjunction with any increase in its debt load. As evidenced by the actual historical experience presented in Appendix 3, any time gap between the incurrence of additional debt and the layering of hedges exposes a borrower to the risk of a slide in commodity prices thereby making hedges less effective in preserving the debt repayment capacity of the borrower. In the context of the Appalachian companies analyzed, the implementation of our recommendations would require a majority of these companies to hedge significantly more volumes than those they have hedged in the past.

This paper presents a field scale reservoir evaluation and uncertainty analysis of hydrocarbon reserves estimation for the Upper Morrow B reservoir of the Farnsworth Unit (FWU), Ochiltree County, Texas. The degree of uncertainty in volumetric reserves estimation for hydrocarbon in place is controlled in larger order by the geological complexity of the reservoir and quality of available geologic data. Morrow B core and thin sections were examined to determine composition, porosity types, depositional environment and diagenetic history. Composition and porosity types determined from XRD and optical microscopy were compared with results from an ELAN analysis. This information together with additional core, well log, borehole image logs, vertical seismic profiles and 3D surface seismic data were used to characterize and subsequently create a fine scale lithofacies based geocellular model ( Ampomah et al., 2016b ). The reservoir is classified as a highly heterogeneous. Probability density functions for input uncertain variables were constructed to estimate probabilistic reserves using first order, parametric and Monte Carlo simulation methods. The relative impact of input variables from these methods were compared and analyzed based on geology, petrophyiscs and engineering knowledge from the FWU to ascertain the applicability of these approaches. The result for each method is presented with expectation curve and log-probability plot elaborating the likelihood of occurrence as P10, P50, P90 and Mean reserves. A P10-to-P90 ratio and a coefficient of variance were used to analyze the total uncertainty of the reserves estimation. Histograms were used to illustrate the sensitivity of input parameter’s contribution towards the overall uncertainty. Statistical mean reserves from the three methodologies were close to the deterministic calculations. Formation thickness and area were the most uncertain variables and have to be topmost priorities to ensure accuracy in reserves estimation for FWU. The results from this study show that an analytical procedure such as the parametric method which is easily generated within spreadsheets can be used to replace the "black box" Monte Carlo simulation of estimating hydrocarbon reserves in the oil and gas industry.

The objective of the study summarized in this paper was to establish a robust method to detect outliers in rate and pressure data used for decline analysis and forecasting, pressure and rate transient analysis, and similar workflows. Presence of outliers contributes significantly to the uncertainty and non-uniqueness associated with the rate/pressure transient analysis workflow. In time-series data usually encountered in rate/pressure transient analysis workflows, we observed that the neighboring data points of an inlying data point will be closer to the point as compared to closeness of an outlier data point. The outlier detection method we developed exploits this observation. It relies on the concept of local density around a data point, and locality is quantified by a function of distance of the data point from its k-nearest neighbors. This concept further maps into a local outlier factor, which signifies the quantitative extent of a data point being an outlier. Outliers will have a higher value of local outlier factor than will inlying data points. A threshold value of local outlier factor is used to label the data points as outliers, and this threshold is set using the histogram of local outlier factor values. Most commonly used outlier detection methods in the industry assume that rate/pressure data follow a known model (e.g., a Gaussian probability distribution model), but outliers deviate strongly from this model. In the usual parametric-model-based approach, an a priori assumption of the correct model needs to be verified, which can limit the validity of the approach. The newly proposed outlier detection method, being non-parametric, is more robust as it is independent of the assumption that data follows a known model. We present a synthetic case study demonstrating the importance of outlier detection and outlining the limitations of currently used outlier detection methods. We then validate our method through synthetic examples generated using numerical models of multi-stage hydraulically fractured wells in unconventional reservoirs. Upon validation we demonstrate application of our method using field examples from four major shale plays.

One of the biggest practical problems with the optimization of shale gas stimulation design is estimating post-fracture production rate, production decline, and ultimate recovery. Without a realistic prediction of the production decline trend resulting from a given completion and reservoir properties, it is impossible to evaluate the economic viability of producing natural gas from shale plays. Traditionally, decline curve analysis (DCA) is commonly used to predict gas production and its decline trend to determine the estimated ultimate recovery (EUR), but its analysis cannot be used to analyze what factors influence the production decline trend due to lack of underlying support of physics, which make it difficult to guide completion designs or optimize field development. In this article, we presented a unified shale gas reservoir model, which incorporates real gas transport, nano-flow mechanisms and geomechanics into fractured shale system. This model is used to predict shale gas production under different reservoir scenarios and investigate what factors control its decline trend. The results and analysis presented in the article provide us a better understanding of gas production and decline mechanisms in a shale gas well with certain conditions of the reservoir characteristics. More in-depth knowledge regarding the effects of factors controlling the behavior of the gas production can help us develop more reliable models to forecast shale gas decline trend and ultimate recovery. This article also reveals that some commonly hold beliefs may sound reasonable to infer production decline trend, but may not be true in a coupled reservoir system in reality.

Gradually increasing production of natural gas from shale formations has surpassed the production from all other sources, including conventional, coal-bed, and tight formations in the U.S. (EIA, Monthly Energy Review, Jan. 2016). In fact, thanks to shale, natural gas production in the U.S. increased much faster than demand growth. Lower gas prices since 2010-11 owing to excess supply, has been encouraging more natural gas utilization in power generation and industrial sector, and exports via pipelines to Mexico and the rest of the world via liquefaction facilities. Many of these facilities are still under construction; and demand growth may not be fully felt until 2018 or later. In the meantime, shale drilling has slowed down considerably owing to low oil (and NGL) prices in addition to low natural gas prices. This situation raises concerns about the shale gas resource potential and future production capabilities. To address these concerns, the understanding of the two key resource characteristics is essential: (1) the total resource-in-place, and (2) the extractability of the resource under the given technological and economic conditions. The Marcellus play is the largest shale gas play in the U.S.; gas production from the play continued to increase despite declining prices and drilling activity. In this study, we focus on the Marcellus play to study these two characteristics. The resource base estimate for the Marcellus has been provided by geologic studies (e.g. Smye et al., in review ). The recovery remains a subject of discussion in numerous engineering studies ( Malpani et al., 2015 ; Xie and Li, 2014 ; Arogundade and Sohrabi, 2012 ). These studies, however, are mostly focused on the best drilling and completion practices, rather than looking at changes in productivity over time or comparison of marginal contribution of different inputs to productivity, which is more relevant for economic studies of the resource extraction. The present paper attempts to fill this gap, preparing the base for further economic viability analyses.

Alianza Casabe, a Technical Collaboration venture between Ecopetrol and Schlumberger in Colombia, was finding that ensuring data integrity between the planning, finance and economic models being used by different stakeholders for investment decisions required a substantial amount of interaction and data reconciliation. Economic scenarios were only appraised at a high level of granularity, reducing the ability to factor in economic metrics at the individual well or projects level, as part of the selection of the best possible portfolio. At the same time, existing systems were turning all planning milestones into protracted exercises demanding very high effort by all participants to deal with the intense manual data manipulation, reconciliation and quality control required. Alianza Casabe set out to enhance and simplify the planning processes, to improve efficiency in annual work plan preparation, facilitate performance tracking during execution and most important improve overall investment decisions quality. The deployment presented several challenges to the planning team including: alignment of expectations across the Organization, rationalization of existing processes, definition of roles and responsibilities over the new solution, natural resistance to change, and the sheer technical complexity and breadth of the multiple integrations with incumbent systems that the deployment entailed. Post deployment, Alianza Casabe experienced numerous benefits: Planning, finance, and economic models were now based on the exact same source data; the permanent availability of an "evergreen" view of the project portfolio enabled a faster reaction time to deviations from plan during execution and a faster reassessment of the portfolio in front of changing market conditions; increased execution efficiency due to better pre-requisites planning, rig scheduling and control; the unavoidable changes introduced to the plan due to rig scheduling, and their business ramifications, were visible with a much higher frequency and quality; trust in plan data and its successive updates across the stakeholder community was greatly increased; there was a substantial reduction in the effort required to handle data, due to the system integrations deployed; the generation of plan variants and what-if scenarios was significantly enhanced; quality of decisions enriched with dynamic economic information. In conclusion, over a three years cycle, Alianza Casabe has seen a substantial improvement in its planning processes: the effort involved in the generation of a single version of the plan was halved, while providing much greater confidence in the plan data and a very substantial reduction in QC time; the integrity gap they experienced between the planning, finance and economic models was eliminated; the use of economics for project high grading was utilized at a higher granularity level; robustness, traceability and reliability of planning data was greatly improved, leading to a more timely availability of reliable information; better support for the PDCA cycle (Plan, Do, Check, Act), enabled continuous improvement of the planning process. Management is now empowered to carry out faster and more frequent reappraisals of the project portfolio, with full consideration of the economic impact of business decisions.