New Optical Sensor System for Improved Fluid Identification and Fluid Typing During LWD Sampling Operations

Cartellieri, Ansgar; Kischkat, Tobias; Sroka, Stefan; Meister, Matthias

doi:10.2118/184717-ms

Cited by 3 publications

(1 citation statement)

References 9 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…), что может быть реализовано методами рефрактометрии [8,9]. При этом требуется учитывать физические и конструктивные ограничения [10] -в скважине флюид протекает в герметичной кювете высокого давления порядка 10 5 МПа при температуре, достигающей 150 °C (рис. 1).…”

Section: Introductionunclassified

Sensing element for the formation fluid refractometer on the basis of total internal reflection

Сергеевна¹,

Олеговна²,

Валентинович³

et al. 2021

Naučno-teh. vestn. inf. tehnol. meh. opt.

View full text Add to dashboard Cite

http://orcid.org/0000-0001-5572-926X Аннотация Предмет исследования. При разработке нефтяных месторождений стоит а ктуальная задача оперативного определения типа прокачиваемого пластового флюида. Нефтяные пластовые флюиды включают в себя газ, нефть и воду. В работе предложен сенсорный элемент нового типа для проточной рефрактометрии пластового флюида на основе эффекта полного внутреннего отражения. Сенсорный элемент представляет собой стержневой наконечник конической формы из сапфира длиной 20 мм и диаметром 20 мм. Метод. Оригинальная форма сенсорного элемента определена модифицированным методом трассировки лучей с учетом аналитических соотношений, определяющих условия обеспечения большего динамического диапазона измерений при заданных физических, технологических и конструктивных ограничениях. Основные результаты. Получена преобразовательная характеристика сенсорного элемента для длин волн 405, 1064, 3300 нм, позволяющая определять тип пластового флюида (газ/вода/нефть). Практическая значимость. Предложенная методика позволяет разрабатывать сенсорные элементы конической формы на основе полного внутреннего отражения для систем скважинного мониторинга и оптических пороговых датчиков показателя преломления. Ключевые слова конический наконечник, полное внутреннее отражение, пластовый флюид, рефрактометрия, скважинный мониторинг, пороговый сенсор показателя преломления Ссылка для цитирования:

show abstract

Section: Introductionunclassified

Sensing element for the formation fluid refractometer on the basis of total internal reflection

Сергеевна¹,

Олеговна²,

Валентинович³

et al. 2021

Naučno-teh. vestn. inf. tehnol. meh. opt.

View full text Add to dashboard Cite

show abstract

Acquiring Accurate Real-Time Formation Fluid Properties to Provide In-Situ Fluid Analysis While Drilling

Cartellieri,

Schapotschnikow,

Weinzierl

et al. 2023

SPE Offshore Europe Conference &Amp; Exhibition

View full text Add to dashboard Cite

The composition of natural gas and crude oils specifies their calorific and commercial value. To determine required processing steps for intermediate and end products of the reservoir fluid and to define the layout of production facilities, an accurate fluid analysis is vital. Currently, the most reliable way of acquiring fluid compositional data is provided by recovering samples from the downhole environment with dedicated fluid sampling services for surface laboratory analysis. Specialized PVT (Pressure-Volume-Temperature) laboratories provide a detailed fluid characterization using advanced techniques such as gas chromatography (GC). This procedure is very costly and time consuming. Therefore, improved measurement systems and alternative methods with lower cost are commonly requested. One way to hasten the analysis of reservoir fluids and to lower the cost is the introduction of sampling while drilling services. Even though, these products have been commercial for nearly a decade, most sampling jobs are still conducted with wireline services after the well has been drilled, delaying the acquisition, and incurring additional rig costs. One reason that has limited the application of logging-while-drilling (LWD) fluid sampling services is the reduced data resolution in real time. Whereas wireline services are able to transmit comprehensive fluid identification data in seconds, it takes minutes for a comparable LWD service to transmit the data even in significantly reduced resolution. The wireline data transmission rate is many thousand times higher than while drilling. A wired pipe application is available that can overcome this constraint, however, at much higher cost. For a representative fluid evaluation and the selection of the optimal fluid samples for characterization in a dedicated PVT lab, an improved prediction of the fluid properties during pump-out, especially for LWD services, is required. Considering the limited real-time bandwidth available, the evaluation must be performed downhole. Therefore, new fluid type algorithms that combine multiple sensor readings into one model have been developed and implemented in the downhole electronics to provide a more sophisticated and thus conclusive in-situ fluid analysis. This paper will discuss and present the most recent enhancements, achieved by adding machine learning algorithms and chemometrics to LWD fluid sampling services. Models are trained on data from multiple field studies and typical deepwater applications to simplify data interpretation and sample selection, resulting in comprehensive fluid type information being available at surface for real-time decisions and accurate in-situ fluid analysis.

show abstract

A Comparative Study on Fluid Composition Determination from Near Infrared Spectra Using Deep Convolutional Neural Networks and Partial Least Squares Regression

Weinzierl,

Cartellieri,

Schapotschnikow

2024

Day 3 Wed, February 14, 2024

View full text Add to dashboard Cite

The conventional approach to fluid characterization using partial least squares (PLS) is considered a benchmark in chemometric fluid analysis. Complementary, convolutional neural networks (CNN) have been shown to provide comparable discrimination capabilities. In a comparative study, the performance for quantitative characterization of downhole fluids using near-infrared (NIR) spectra has been evaluated. Both methods are used to predict the fluid composition in fractions of water, gas, oil, and mud. PLS is a statistical technique designed to model the relationship between two sets of variables, in this case between the spectrum and the composition. It relies on the representation of the variables in a multidimensional latent space. Usually, the inference consists of three steps. First, the input (spectrum) is linearly projected into the latent space. Second, the output is calculated in the latent space. Finally, the composition is computed as a linear transformation of the latent output. Instead of using PLS for end-to-end inference, only its first step has been used for feature extraction. By using the first latent dimension for each component, features were obtained that can be conveniently associated with water, gas and oil respectively. These features are then used together with the constant baseline in a multinomial logistic regression to obtain fractional components of the present fluid types in the NIR spectra. The baseline is primarily needed for mud detection. In parallel, several CNN models were trained for fluid characterization based on NIR spectra on processed and raw data. Hyper-parameter optimization of the CNN's is performed using a tree structured Parzen estimator to obtain a best trial configuration. Scheduling of the optimization loop yielded improved inference results. Quantitative comparison of the PLS and CNN models was performed using a k-fold approach. This allows for a direct comparison of the methods performance given as input spectra of pure and mixed fluids. Both methods show high accuracy when predicting pure components. The root mean square error (RMSE) is consistently larger for PLS. The CNN models generally show larger variance in the prediction for mud, with minor fractions of water, gas and oil being inferred. A quantitative comparison of two methods in chemometric fluid analysis shows an overall improvement of predictive power for a set of deep CNN in respect to the PLS approach. Improved inference is achieved using raw NIR spectral data. This is particularly interesting as no further pre-processing of the spectra is required, thereby minimizing porting efforts in the development of embedded applications.

show abstract

New Optical Sensor System for Improved Fluid Identification and Fluid Typing During LWD Sampling Operations

Cited by 3 publications

References 9 publications

Sensing element for the formation fluid refractometer on the basis of total internal reflection

Sensing element for the formation fluid refractometer on the basis of total internal reflection

Acquiring Accurate Real-Time Formation Fluid Properties to Provide In-Situ Fluid Analysis While Drilling

A Comparative Study on Fluid Composition Determination from Near Infrared Spectra Using Deep Convolutional Neural Networks and Partial Least Squares Regression

Contact Info

Product

Resources

About