No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Allocation System Setup Optimization in a Cost-Benefit Perspective
Abstract
Allocation of hydrocarbons to their original production sources, also known as hydrocarbon accounting, is a key factor for the distribution of costs, revenues and taxes between interested parties in field development and production of oil and gas. When developing an allocation system, the allocation uncertainties in the system should be understood and accepted by all involved parties. Furthermore, the implemented allocation system should be cost efficient and practical to operate. One of the pivotal design questions for such an allocation system is the choice of measurement uncertainty of the individual metering stations comprising the system. In this paper, we device a framework for allocation system modelling that allows for an algorithmic solution to the problem of optimizing the allocation system setup, i.e., choosing the right meter with the right uncertainty at the right place. This includes balancing the risk associated with misallocation due to measurement uncertainty against the cost of realizing the system. The presented framework makes use of a combination of optimization and ISO Guide to the Expression of Uncertainty in Measurement (ISO GUM) compliant Monte Carlo simulations. We illustrate the usefulness of our framework by applying it to example allocation systems with different allocation principles and production rates. We review the obtained results and provide a discussion of strengths and current limitations of the proposed approach.
... Different methods have been presented in the literature or employed in the industry for performing allocation calculations [10][11][12][13][14][15]. The purpose of all of these methods is to estimate the production of a single well using the available data. ...
... A thorough comparison of their accuracy with the accuracy of the traditional allocation method, however, has not been presented. Pobitzer et al. [13] proposed an algorithm that helps choosing the right meter and its place in the allocation process. Therefore, their focus was on optimizing the allocation system setup for reducing the allocation uncertainty. ...
Although the application of multi-phase flow meters has recently increased, the production of individual wells in many fields is still monitored by occasional flow tests using test sep-arators. In the absence of flow measurement data during the time interval between two consecutive flow tests, the flow rates of wells are typically estimated using allocation techniques. As the flow rates, however, do not remain the same over the time between the tests, there is typically a large uncertainty associated with the allocated values. In this research, the effect of the frequency of flow tests on the estimated total production of wells, allocation, and hydrocarbon accounting has been investigated. Allocation calculations have been undertaken for three different cases using actual and simulated production data based on one to four flow tests per month. Allocation errors for each case have subsequently been obtained. The results show that for all the investigated cases, the average allocation error decreased when the number of flow tests per month increased. The sharpest error reduction has been observed when the frequency of the tests increased from one to two times per month. It reduced the allocation error for the three investigated cases by 0.43%, 0.45%, and 1.11% which are equivalent to $18.2M (million), $18.9M, and $46.8M reduction in the yearly cost of the allocation error for the respective cases. The reductions in the allocation error cost for the three cases were $27M, $29M, and $80M, respectively, when the flow tests have been undertaken weekly instead of monthly.
... Different methods have been presented in the literature or employed in the industry for performing allocation calculations [10][11][12][13][14][15]. The purpose of all of these methods is to estimate the production of a single well using the available data. ...
... A thorough comparison of their accuracy with the accuracy of the traditional allocation method, however, has not been presented. Pobitzer et al. [13] proposed an algorithm that helps choosing the right meter and its place in the allocation process. Therefore, their focus was on optimizing the allocation system setup for reducing the allocation uncertainty. ...
Although the application of multiphase flow meters has recently increased, the production of individual wells in many fields is still monitored by occasional flow tests using test separators. In the absence of flow measurement data during the time between two consecutive flow tests, the flow rates are typically estimated using allocation techniques. However, since the flow rates do not remain the same over time, there is typically a large uncertainty associated with the allocated values. In this research, the effect of the frequency of flow tests on the estimated total production of wells, allocation, and hydrocarbon accounting has been investigated. Allocation calculations have been undertaken for three different cases using actual and simulated production data based on one to four flow tests per month. Allocation errors for each case have subsequently been obtained. The results show that for all investigated cases, the average allocation error decreased when the number of flow tests per month increased. The sharpest error reduction has been observed when the frequency of the tests increased from one to two times per month. It reduced the allocation error for the three investigated cases by 0.43%, 0.45%, and 1.11% which are equivalent to 18.2M(Million), 18.9M, and 46.8M reduction in the yearly cost of the allocation error for the respective cases. The reductions in the allocation error cost for the three cases have been 27M, 29M, and 80M, respectively, when the flow tests have been undertaken weekly instead of monthly.
... Uncertainty and risk are calculated over a 15 years period using the framework described in [6], based on the production profiles for the three fields shown in Figure 7. ...
... In order to make a fair assessment it is important to keep in mind at this point that the CAPEX and OPEX of different metering stations will vary depending on the specifications of the metering station and the maintenance and calibration scheme chosen. This risk calculation also prepares the foundation for full cost-benefit analysis including CAPEX and OPEX as described in [6] and can be extended to include other types of risk as in [7]. ...
Current focus on cost-effective developments of new hydrocarbon fields aims at exploiting the capacity of existing production units to the maximum using a number of tie-ins to subsea developments. This results in increasingly complicated process flows, where individual multiphase streams may or may not be measured. This increased complexity makes design and analysis of allocation systems a challenging task.
Allocation principles, metering system setup, use of test separator time, ownership structure, flow rates and life time profiles are all factors which affect the field and ownership allocation uncertainty. In order to find how each field or each owner is exposed to economic risk associated with measurement and allocation uncertainty, an uncertainty analysis combined with a risk-cost-benefit analysis should be carried out. Traditionally, the analysis of such allocation systems is based on analytic calculations. These calculations increase rapidly in complexity as the process flow becomes more complicated. For systems with several tie-ins and satellites, and with a fragmented ownership, powerful numerical methods are required to perform this analysis.
This paper demonstrates the calculation of field and ownership allocation uncertainty for realistic measurement setups and allocation scenarios in a multi-field setting based on industrial projects. A flexible framework for analysis of complex multi-field configurations is used in these numerical calculations, which are based on an ISO GUM (ISO/IEC, 2008) compliant Monte Carlo technique.
Further on, it is demonstrated how different field configurations, ownership structures, allocation principles, meter uncertainties and flow rates affect the total cost and risk for each owner. This investigation includes the exposure to economic risk associated with measurement uncertainty associated with the different alternatives. We also give examples of how the lifetime cost of the metering system may vary depending on choices in allocation principle, flow rate profiles, as well as placement and calibration scheme of the individual meters. A particular focus of our work is how each owner is exposed to misallocation risk. In this context it is mandatory to take into account the correlation between the uncertainties in the field-allocated streams. Failure to include these correlations may result in erroneous estimations of each owner’s economic exposure due to misallocation, and may thus potentially result in sub-optimal field developments.
Through realistic example systems based on industry projects, it is shown how an uncertainty analysis combined with a risk analysis may provide valuable insight into the exposed economic risk for each owner due to misallocation. It is demonstrated how thorough knowledge and understanding of the allocation uncertainty is essential in order to minimize each parties’ economic exposure, especially in real-life complex allocation systems.
... The principles of different methods of allocation have been explained by the Energy Institute (2012). The focus of many publications on allocation is its application in hydrocarbon accounting ( Cramer et al. 2011; Kaiser In an inverse problem, such as history matching, the characteristics of an unknown system are estimated based on its observed output data 2014;Pobitzer et al. 2016). There is a dearth of publications on the application of allocation data in reservoir analysis, reservoir management and history matching. ...
History matching is the process of modifying a numerical model (representing a reservoir) in the light of observed production data. In the oil and gas industry, production data are employed during a history matching exercise in order to reduce the uncertainty in associated reservoir models. However, production data, normally measured using commercial flowmeters that may or may not be accurate depending on factors such as maintenance schedules, or estimated using mathematical equations, inevitably has inherent errors. In other words, the data which are used to reduce the uncertainty of the model may have considerable uncertainty in itself. This problem is exacerbated for gas condensate and wet gas reservoirs as there are even greater errors associated with measuring small fractions of liquid. The influence of this uncertainty in the production data on history matching has not been addressed in the literature so far. In this paper, the effect of systematic and random flow measurement errors on history matching is investigated. Initially, 14 production data sets with different ranges of systematic and random errors, from 0 to 10%, have been employed in a history matching exercise for an oil reservoir and the results have later been evaluated based on a reference model. Subsequently, 23 data sets with errors ranging from 0 to 20% have been employed in the same process for a wet gas reservoir. The results show that for both cases systematic errors considerably affect history matching, while the effect of random errors on the considered scenarios is seen to be insignificant. Although reservoir model parameters in the wet gas reservoir were not as sensitive to the flow measurement errors as in the oil reservoir, for both cases, the future production forecast was significantly affected by the errors. Permeability was seen to be the most sensitive history matching parameter to the flow measurement errors in the oil reservoir, while for the wet gas reservoir, the most sensitive parameter was the forecast of future oil and gas production. Finally, considering the noticeable effect of systematic errors on both cases, it is suggested that flowmeter calibration and regular maintenance is prioritised, although the subsequent economic cost needs to be considered.
In upstream oil and gas production, flow measurement errors are common at multiple locations in the production system. These measurement errors lead to imbalances, which are reconciled by using allocation methods. The allocation method should redistribute the imbalance in a fair manner to all involved stakeholders. In this paper, data validation and reconciliation (DVR) is proposed as an alternative allocation method in multiphase production systems. DVR is a model-based optimization method that exploits redundancies in process data to minimize random measurement errors, and simultaneously provides a basis for the detection of gross errors. To study the applicability of DVR compared to conventional allocation methods, a model for generic multiphase production systems is developed, and the corresponding DVR problem is formulated. In order to deal with nonlinear flow effects (e.g. interphase mass transfer) the model is linearized around the operating conditions and the DVR problem is solved iteratively. From the simulation studies, it is concluded that DVR reduces measurement errors by up to 56% for gas allocation systems and 33% for multiphase systems. Gross errors in interior locations of the network can be located precisely, while allocating gross errors to exterior locations is generally not possible, due to limited network redundancy. Compared to conventional methods, DVR provides a better means for the detection of gross errors, and results in more accurate attribution of oil/gas ownership and more fair division of revenues between stakeholders.
An earlier publication (Hall 2006 Metrologia 43 L56–61) introduced the notion of an uncertain number that can be used in data processing to represent quantity estimates with associated uncertainty. The approach can be automated, allowing data processing algorithms to be decomposed into convenient steps, so that complicated measurement procedures can be handled. This paper illustrates the uncertain-number approach using several simple measurement scenarios and two different software tools. One is an extension library for Microsoft Excel®. The other is a special-purpose calculator using the Python programming language.
An object-oriented approach to Monte Carlo calculation of uncertainties is described. Simple C++ objects are presented that can be used to perform uncertainty calculations by modelling the underlying probability distributions. The approach allows a calculation to be expressed simply, in terms of the measurement function and the probability distributions of its input quantities. Analysis of the measurement function is not required, nor is the calculation sensitive to measurement function nonlinearity. The C++ computer code is very efficient, allowing desktop computer simulations to provide detailed information on the measurand probability distribution function.
Parameter Estimation and Inverse Problems, 2e provides geoscience students and professionals with answers to common questions like how one can derive a physical model from a finite set of observations containing errors, and how one may determine the quality of such a model. This book takes on these fundamental and challenging problems, introducing students and professionals to the broad range of approaches that lie in the realm of inverse theory. The authors present both the underlying theory and practical algorithms for solving inverse problems. The authors' treatment is appropriate for geoscience graduate students and advanced undergraduates with a basic working knowledge of calculus, linear algebra, and statistics. Parameter Estimation and Inverse Problems, 2e introduces readers to both Classical and Bayesian approaches to linear and nonlinear problems with particular attention paid to computational, mathematical, and statistical issues related to their application to geophysical problems. The textbook includes Appendices covering essential linear algebra, statistics, and notation in the context of the subject. A companion website features computational examples (including all examples contained in the textbook) and useful subroutines using MATLAB. Includes appendices for review of needed concepts in linear, statistics, and vector calculus. Companion website contains comprehensive MATLAB code for all examples, which readers can reproduce, experiment with, and modify. Online instructor's guide helps professors teach, customize exercises, and select homework problems Accessible to students and professionals without a highly specialized mathematical background.
Gas lift allocation can be modeled as a nonlinear programming problem in which adjusting optimum gas injection rates and compressor pressure maximize the oil rate or other objective functions. This study describes a nonlinear programming approach to maximize daily cash flow of some gas-lifted wells in an uncertain condition for oil price. First, some solution points of each well are obtained by employing a production simulation software. Then, by use of nonlinear optimization, a model is developed for gas lift performance in each well. Then these functions are used to develop a model under capacity, pressure and other real constraints for the cash flow of production from these wells. Oil price is assumed as a triangle risk function in this model. Results show a significant increase in cash flow in comparison with old case due to appropriate gas injection parameters. Sensitivity analysis on this problem shows that oil price, compression cost and water–oil ratio variations should be considered in the long term optimization.
Oil and gas production in the United States is frequently reported on a lease or unit basis. Leases and units with more than one well pose special problems in forecasting because individual well profiles must be inferred based on well tests and the aggregate production of the lease or unit. The purpose of this paper is to develop and compare decomposition procedures for multi-well leases based on decline curve and fixed allocation techniques. Decline curve techniques use the production profile of the multi-well lease to decompose and separate the individual well profiles, while a fixed allocation method samples the lease production profile at specific points in time to allocate production between wells. Using three multi-well lease examples from Louisiana’s Haynesville shale, we compare the decline curve and fixed allocation methods and demonstrate that the techniques yield broadly similar results in severance tax exemption computations and estimates of expected recovery volumes. Leases that require alternative decomposition techniques are depicted.
In the oil industry, it is necessary to reconcile fiscally measured hydrocarbon production with estimated production from associated wells. This process is known as "allocation" and is important for a number of reasons, including accounting for field production to owners/governments, field surveillance, and volumetric input to reservoir simulators.
Traditionally, allocation is performed on a monthly basis, reconciling the "less-accurate" sum of the well tests adjusted by well uptimes with the "more-accurate" fiscal measurements. This process is subject to a number of inherent inaccuracies, including less-than-perfect well tests, lack of knowledge of precisely when the wells were interrupted, unknown well-flow changes, and uncertainty as to how to allocate effectively the difference between fiscal and well-test measurements.
Ineffective allocation can lead to financial consequences caused by inaccuracies in volumes allocated between various owners and tax regimes. Inaccuracies can also feed through to reservoir simulators, which may be used in corresponding decision-making processes (e.g., where to drill the next well). These inaccuracies are generally compounded with time because allocation is performed month after month throughout the field life cycle.
Associated with hydrocarbon accounting are key performance indicators (KPIs) (e.g., well daily production and deferment rates).
The purpose of this paper is to describe Shell's experiences in improving allocation accuracy and automatic KPI reporting through more-effective use of available production data and by using a continuous, rather than a discontinuous, hydrocarbon accounting process.